Biological influences on criminal behavior [Second edition] 9780429356834, 9780367360016, 9780367417284, 0367360012, 0429356838

Citation preview

Biological Influences on Criminal Behavior

Biological Influences on Criminal Behavior Second Edition

Gail S. Anderson

SFU PUBLICATIONS

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2020 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed on acid-free paper International Standard Book Number-13: 978-0-367-36001-6 (Hardback) 978-0-367-41728-4 (Paperback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

Dedication To my parents, Alec and Pamela Anderson, and my brother, David Anderson, who have always supported me in all my endeavours and to Professor Thelma Finlayson, my mentor and friend, who has lead the way for women in science.

v

Contents Preface xvii Acknowledgments xix Author xxi 1 Biology and crime

1

Introduction 1 The question of biology, crime, and the environment 2 Is writing learned environmentally or controlled genetically? 2 Is there more to biology than just genetics? 3 Resistance to biological explanations 3 The promise of biological research 4 1. Phenylketonuria 4 2.  Cystic fibrosis 5 3. Serotonin 5 Some obvious examples of biology’s effect on behavior 6 Puberty 6 Pregnancy 6 Menstruation 6 Resistance to biological criminology 7 The history of biology and crime 7 Franz Gall and phrenology, eighteenth and early nineteenth century 8 Lombroso and atavisms, nineteenth century 8 Francis Galton and the start of eugenics 9 The eugenics movement 10 Eugenics in the latter part of the twentieth century 10 Immigration and intelligence quotient testing 11 Endomorphs, mesomorphs, and ectomorphs 11 Determinism 12 Racism and sexism 12 Mental illness and medicalization of crime 13 Other cautions that must be considered 14 Predisposition 14 What is crime? 15 Cautions for all criminological research 16 1. Adolescence 17 2.  Research samples 17 3.  Underreporting of crimes 17 4. Self-reporting 18 5. Consent 18 The future of biosocial criminology 18 Conclusion 19 Questions for further study and discussion 19 References 19

vii

viii Contents

2 Evolution, natural selection, and behavior

23

Introduction 23 Natural selection 23 Types of adaptations 25 Structural modifications 25 Biochemical pathways 25 Behavioral adaptations 25 Conditions required for natural selection to act 26 1. Variation 26 2. Heritability 26 3.  Fitness consequences 26 Constraints on natural selection 28 1.  Adaptations are often compromises 28 2.  Natural selection can only work on existing traits 28 Behavior in humans and other animals 29 Innate behaviors 31 When and where do innate behaviors occur? 31 In animals with no opportunity to learn 32 Critical to get it right the first time 32 Types of signs or stimuli that trigger innate behaviors 32 Visual 32 Auditory 33 Olfactory 33 Temperature 33 Combination of signs 33 Learned behaviors 33 Learning versus maturation 34 Habituation 34 Imprinting 34 Classical conditioning 34 35 Operant conditioning Observational learning 35 Play 35 35 Insight learning Natural selection and behavior 36 Aggression 37 Natural selection and crime 38 1.  Theft and robbery 38 2.  Assault and murder 39 Infanticide 39 Siblicide 40 3.  Sexual assault and rape 40 4.  Child abuse 42 5.  Domestic violence 44 Evolutionary theories 45 Cheater theory 45 Variation within K-strategists 46 Conditional adaptation theory 46 Alternate adaptation theory 46 Evolutionary expropriative behavior 47

Contents  ix

Conclusion 47 Questions for further study and discussion 48 References 48

3 Genetic principles

51

4 Misconceptions, experimental design, and behavioral genetics

67

Introduction 51 Introduction to genetics 51 Meiosis 53 Genetics: The study of patterns of inheritance 54 Mendel’s experiments 54 Non-Mendelian types of gene interactions and inheritance patterns 57 Linked genes 57 Sex-linked traits 58 Incomplete dominance 59 Codominance 59 Pleiotropy 59 Epistasis 59 Polygenic inheritance 59 Mutations 60 Recessive alleles in the population 61 Recessive alleles and disease 61 Why aren’t we perfect? 62 Gene expression 63 Heritability 64 Conclusion 64 Questions for further study and discussion 64 References 65 Introduction 67 Some misconceptions about genetics 67 Animal cloning 67 Human cloning 68 Does all crime have the same single cause? 70 XYY man: Truth and fallacy 70 Experimental design 72 Scientific method 72 Isolating a single variable 73 Sample size 74 Crossover studies 74 Replication 74 Double-blind studies 75 Studying humans 75 Studying behavioral genetics 76 Introduction to twin studies 76 Dizygotic twins 76 Monozygotic twins 77 Explanations for twin coincidences 77 Using twins to study genetic and environmental influences on behavior 79 Using adoption to study genetic and environmental influences on behavior 80

x Contents

Conclusion 80 Questions for further study and discussion 81 References 81

5 Evidence for genetic predispositions for criminogenic behavior: Twin and adoption studies

83

Introduction 83 Twin studies 83 Early twin studies 84 Modern twin studies 85 Identical twins reared apart 87 Cautions for twin studies 88 Shared environment 88 Generalizability 88 Genetic additivity 89 Overuse of data sets 89 Adoption studies 90 Early adoption studies 90 Mednick’s Danish adoption studies 90 Bohman’s Stockholm adoption studies, 1996 92 Modern adoption studies 93 Impulsivity 94 Psychopathy 94 Substance abuse 94 Cautions for adoption studies 96 Differences between biological and adoptive parents 96 Late separation 96 Selective placement 97 Prenatal influences 97 Genetics and behavior overall 97 Protective factors 98 Family and social bonding 99 Resilience 100 Education and school experiences 102 Enrichment programs 103 104 Other protective factors Conclusion 105 Questions for further study and discussion 106 References 106

6 Candidate genes, gene × environment interactions, and epigenetics 111

Introduction 111 Candidate genes 111 Attention deficit hyperactivity disorder 112 Conduct disorder 113 Schizophrenia 113 Interactions between genes and the environment 114 Gene–environment correlations 114 1.  Passive gene–environment correlations 114 2.  Reactive gene–environment correlations 114 3.  Active gene–environment correlations 114

Contents  xi

Gene × environment interactions 115 An example of G × E interactions 115 G × E interaction models 116 1.  Diathesis stress model 116 2.  Bioecological model 116 3.  Differential susceptibility 116 4.  Social distinction model 117 5.  Social push model 117 Interventions considering G × E interactions 117 Epigenetics 118 Epigenome 119 1. DNA methylation 119 2.  Histone modification or acetylation 119 Impact of the environment on the epigenome 119 Early-life adversity 120 Post-traumatic stress disorder 122 Treatment potential 122 Can epigenetic changes be inherited? 123 Cautions with epigenetics 123 Conclusion 124 Questions for further study and discussion 124 References 125

7 The chemistry of the body: The effects of hormones on behavior

129

Introduction 129 The functions of hormones 129 Testosterone 130 Testosterone exposure 131 Prenatal testosterone and behavior 132 Prenatal testosterone and criminal behavior 133 Prenatal testosterone and risk-taking behavior 134 Postpubertal testosterone and behavior 135 Postpubertal testosterone and criminal behavior 135 1.  Natural testosterone levels in aggressive or criminal individuals 136 2.  Increasing testosterone levels 137 3.  Decreasing testosterone levels 139 Postpubertal testosterone and risk-taking behavior 144 Testosterone, competition, and dominance 145 Theoretical background 145 Competition studies 146 Serotonin and testosterone 148 Testosterone in prosocial behavior 149 Testosterone overall 149 Cortisol 149 Cortisol, stress, and abuse 150 Cortisol and antisocial behavior 151 Cortisol and testosterone 152 The dual-hormone hypothesis 152 Cortisol and psychopathy 154 Cortisol overall 155

xii Contents

Other hormones 155 Thyroid hormones 155 Menstruation 156 Infanticide 157 Conclusion 157 Questions for further study and discussion 158 References 158

8 The prenatal environment and birth complications

165

Introduction 165 Birth complications 165 Fetal development, nutrition, and pollutants 167 Fetal alcohol spectrum disorder 169 FASD and the criminal justice system 170 Gene × environment interactions in FASD 171 Intervention 171 Maternal smoking 172 Smoking’s potential effects on a fetus 172 Direct health risks 172 Poor parenting and mother–child attachment 173 Nicotine addiction 173 Maternal stress 173 Gene × environment interactions 173 Epigenetic effects 174 Maternal age 174 Maternal stress 175 Fetal maldevelopment and minor physical anomalies 176 Other birth-related difficulties 178 Twin births 178 Maternal rejection 179 Ethical issues 180 Criminalization of pregnant women in the United States 180 Some case examples 181 Conclusion 182 Questions for further study and discussion 183 References 183

9 The chemistry of the brain: The role of neurotransmitters on behavior 189

Introduction 189 Introduction to neurotransmitters 189 The mechanism of action 190 Serotonin 192 Serotonin and suicidal behavior 192 Serotonin, suicide, and stress 194 Serotonin and aggression 195 Serotonin and psychopathy 196 Serotonin and impulsivity 197 Serotonin receptor sites 198 The serotonin precursor, tryptophan 198 Selective serotonin reuptake inhibitors 199 Norepinephrine 200

Contents  xiii

Dopamine 201 The dopamine transporter, DAT-1 201 Dopamine receptor D2, DRD2 202 Dopamine receptor D4, DRD4 202 Protective factors 203 Dopamine and schizophrenia 204 Monoamine oxidase 204 Monoamine oxidase and aggression 204 MAOA and a gene × environment interaction 205 Other factors 208 Conclusion 208 Questions for further study and discussion 209 References 209

10

Traumatic brain injury and neurocognitive disorders

213

Introduction 213 Head injury 213 Frontal lobe injuries 215 Case examples of frontal lobe injury 216 Traumatic brain injuries in youth 218 Prevalence of young offenders with traumatic brain injury 220 Gender differences in youth with traumatic brain injury 221 Which came first, traumatic brain injury or risky behavior? 222 Traumatic brain injuries and schizophrenia 222 Impact of traumatic brain injury on youth during criminal justice proceedings 223 Traumatic brain injury in adults 224 Prevalence of traumatic brain injury in incarcerated populations 224 Traumatic brain injury and substance abuse 225 Was the traumatic brain injury causal? 225 Brain disorders 227 Case examples of organic brain disorders 228 Neuroimaging studies of brain abnormalities 229 Neuroimaging techniques 229 Computer tomography 229 Magnetic resonance imaging 230 Positron emission tomography 230 Single-photon emission computed tomography 231 Functional magnetic resonance imaging 231 Brain-imaging studies of offenders 231 Treatment options 233 Conclusion 236 Questions for further study and discussion 236 References 237

xiv Contents

11 The effects of pollution, toxins, and diet on behavior

241

12

283

Introduction 241 Pollution and toxins in our environment 241 Lead 244 Effects of lead on the body 245 Effects of lead on antisocial behavior in children 246 Effects of childhood lead exposure on criminal behavior in adults 248 Effects of low socioeconomic status and lead 249 Effects of banning lead in gasoline 251 Effects of intervention 252 Recommendations to reduce environmental lead 253 Manganese 253 Manganese exposure 253 Effects of manganese on adults 254 Effect of manganese on children 254 The safety threshold for manganese in drinking water 257 Other heavy metals 257 Cadmium 257 Coexposure to pollutants 258 Breast milk 259 Mercury 259 Pollution overall 259 Diet 260 Hypoglycemia (low blood sugar) 262 Diet and supplementation 265 Dietary tryptophan and serotonin 265 Vitamins and minerals 266 Fatty acids 268 Effects of fatty acid supplementation on autism spectrum disorder 270 Combination of fatty acids and other supplements 270 271 Discontinuation of fatty acid supplementation Mechanisms of fatty acid supplementation 271 Food additives and food allergies 272 274 Diet overall Conclusion 274 274 Questions for further study and discussion References 275

The role of biology in the courtroom

Introduction 283 Role of genetics 284 Judicial perceptions of genetic predispositions for criminal behavior 285 Public perceptions of genetic predispositions for criminal behavior 286 Use of genetic predispositions in court 287

Contents  xv

Role of neurotransmitters 287 Attempts to use MAOA-L during the guilt phase of a trial 288 Use of MAOA-L during the sentencing phase of a trial 288 Use of MAOA-L during the appeal phase of a trial 289 Double-edged sword? 290 Judicial perceptions of MAOA in the courts 290 Public perceptions of MAOA in the courts 291 Concerns with the use of MAOA in court 292 Use of serotonin levels in court 292 Role of traumatic or organic brain injury 293 Use of traumatic brain injury or brain trauma in court 293 Public perceptions of neuroimaging evidence 298 Concerns with the use of neuroimaging in court 299 Role of fetal alcohol spectrum disorder 301 Role of biological factors overall 303 Conclusion 304 General conclusions 305 Questions for further study and discussion 306 References 306

Epilogue 311 Index 313

Preface In modern criminology, the main schools of thought have focused on social, environmental, material, and psychological factors that can cause crime. As a biologist, I have felt for some time that most studies neglected one important aspect of the equation: the physical person. Human beings are a complex mixture of upbringing, background, environment, experiences, social structure, and biology. Over the past 30 years, valuable work on genetic background, hormone and neurotransmitter levels, diet, and physical insults such as brain trauma has been accomplished. However, with few exceptions, most introductory works in criminology seriously neglected this research. It was my hope that, in writing the first edition of this book, I might help to correct this imbalance. The first edition was written in the hope that through an integration of the biological view with mainstream social, psychological, and environmental views, we can find a new way of studying criminality, and we might gain some positive and useful explanations for criminal behavior. When the first edition of this text was published, despite the increasing evidence of biological influences on criminality, the number of such publications for students of criminology was remarkably slight. It was almost as if studying this particular influence had been forbidden in criminological texts. Wright and Miller (1998), in their excellent article entitled “Taboo Until Today,”1 measured the amount of text written about biological explanations for crime versus other sociological explanations and found that if biology was mentioned at all, it was given infinitesimal coverage. Although excellent biological studies had been conducted, especially in recent years, they were not included in mainstream criminology. In part, this is due to a fear of biology as determinism, but it is primarily due to a lack of understanding. Few scholars have both a criminological and a biological background. If we are to truly understand these valuable studies, strong backgrounds in genetics, neuroscience, and endocrinology, as well as criminology, are an asset. These backgrounds are not often found together. The first edition was written, then, to help the criminologist navigate an additional and vital new area of study: one with its own complexities and challenges. It does not attempt to suggest that biology plays the major role in criminal behavior. It was written with the belief, grounded in research, that something vital can be discovered when we assess all the factors related to the causes of crime, including biology. In the more than 13 years since the first edition of this book was published, my desire to see biology’s role in behavior considered in criminological research has come to fruition. In the last few years, there has been a tremendous turn in academic opinions of what is now termed biosocial criminology. Increased collaboration between scientists and criminologists has led to a much stronger understanding of the intricacies of biology’s role in behavior. Moreover, more criminologists have biological backgrounds. As the science involved became more complex, so too did this text. This  second edition considers the more recent and integrated research that is being conducted today to show the interaction between the environment and a person’s biology that leads to our behavior. It has even been shown that the environment acts on and actually changes the functions of some genes. In the first edition of this book, and in this present text, I have tried to make it very clear that biology is just a small piece of the puzzle and, on its own, cannot cause a person to commit a crime. Despite this, biological influences on criminal behavior have now become so mainstream that evidence of a person’s biology is now frequently introduced in court, especially as a mitigating factor, perhaps prematurely, as we still do not fully understand all the variables involved. The tremendous amount of new work that has been published over the last few years made it vital that a new edition of this text be written. The increasing use of biological factors in court makes it even more vital, and that scholars understand the role that biology may or may not play. xvii

xviii Preface

This text is written primarily for social science and law students who wish to understand this exciting area. I do not expect that you have a scientific background, but you must have an open and inquiring mind. The text begins with basic scientific principles to introduce the reader to the more in-depth discussions in each area. Hopefully, the result will be that you will gain a much greater understanding of this rapidly growing field, so that its lessons can help to inform policy, treatments, rehabilitation, and the law. Gail S. Anderson Simon Fraser University

Reference 1. Wright, R.A. and Miller, J.M. 1998. Taboo until today? The coverage of biological arguments in criminology textbooks, 1961–1970 and 1987–1996. J. Crim. Justice 26(1): 1–19.

Acknowledgments This book is the second edition of a text that began as a series of lectures in my fourth-year seminar classes in the School of Criminology at Simon Fraser University in British Columbia. It progressed from there to a study guide and, finally, after much editing and revision, to its original shape as an introduction to biological influences on crime. I must thank the many people who have helped me work on this text and the many students who have shaped its approach. This book would not have been possible without the assistance and engagement of my students, both graduate and undergraduate. Their critical comments and questions helped to shape my approach. I would also like to thank Dr. John Whatley of SFU Publications for encouraging me to write the first edition of this book and, more recently, to write a much-needed second edition, as well as for his continued enthusiastic support. Gail S. Anderson Simon Fraser University

xix

Author Gail S. Anderson earned a BSc (Honors) in zoology from Manchester University, England, and a Masters of Pest Management and PhD in medical and veterinary entomology from Simon Fraser University. Her specialty is forensic entomology, the use of insects in death investigations. Dr Anderson is one of only three board-certified forensic entomologists in Canada. She is a ­Professor in the School of Criminology at Simon Fraser University, Burnaby, British Columbia, Canada, holds a Burnaby Mountain Endowed Professorship, and is also undergraduate director and co-director of the Centre for Forensic Research and a forensic consultant to the Royal Canadian Mounted Police (RCMP) and municipal police across Canada. Her work has been featured on many television p ­ rograms, including “Journeys—Grave Testimony” and “Forbidden Places—Silent Witness” shown on Discovery Channel and Planet Education and “The Nature of Things—Postmortem,” “Dark Waters of Crime,” “Under the Sea,” and “Weird or What?” shown on Discovery Channel, History Channel, Knowledge Network, and CBC. She was a recipient of Canada’s “Top 40 under 40 Award” in 1999, a YWCA Women of Distinction Award for Science and Technology in 1999, and the Simon Fraser University Alumni Association Outstanding Alumni Award for Academic Achievement in 1995. She was listed in Time magazine as one of the top five innovators in the world, this ­century, in the field of Criminal Justice in 2001 and received the Derome Award from the Canadian Society of Forensic Sciences. In 2014, she received the Dean’s Medal for Academic Excellence, and, in 2015, she was listed as one of the six most influential ­scientists in British Columbia. In 2017, she received the American Academy of Forensic Sciences Pathology and Biology Section Award for Achievement in the Life Sciences.

xxi

1 Biology and crime

Introduction Since prehistory, we have tried to understand why people commit crime. Innumerable explanations have been put forth over time—many ludicrous, some plausible. Medieval “biological” explanations, for example, found criminal character rooted in a system of humors: large-scale influences on behavior brought about by balances and imbalances within specific organs of the body and their secretions. Moral theories of crime as an evil influence have been historically predominant. In  about 400  bce, Plato thought that criminality was caused by an obscurity of thought—the imprisoned human mind was blocked from enlightenment as if it were locked in a cave and thus acted irrationally. In  the first century ce, St Paul thought that it was caused by sinfulness, our inability to fulfill the law of God. In the nineteenth century, Cesare Lombroso put forward a more modern biological theory, a concept of facial types, or atavisms, by which one could identify the criminal; his findings were supposedly based on science. With misunderstood beliefs in genetics, the eugenics movement began in the nineteenth century and culminated in the atrocities and genocide by Nazi Germany in World War II. After the horrors of the eugenics movement were fully understood, social scientists turned 180° away from any thought of a biological role in criminal behavior, or indeed in any behavior, and focused solely on social science explanations for crime. As we will show, although the concept of eugenics is appalling, the truth is that it had nothing to do with science or genetics but was rather a racist political ideology. This violent rejection of biological explanations for criminal behavior was extremely unfortunate, as it delayed the opportunity for a much more embracing understanding of criminal behavior. However, the vast amount of legitimate research that has been conducted in this area has reached a stage where it cannot be ignored and has shown that the new field of biosocial research is not only valid but offers tremendous promise for ways in which we may be able to intervene and even prevent crime. Anyone who has ever known a newborn baby knows that babies are not born as blank slates. They already have a strong personality, which may have even made itself felt in the womb. The strong empirical body of research that has been published recently, some of which we will cover in this text, has resulted in the new field of biosocial criminology.1 Of course, biology does not  cause crime, but then neither does social status, psychology, or ­experience. It is a combination of all the factors of a person and an environment that can lead to that person behaving in a certain manner that could lead to a criminal act. It is unlikely that we will ever understand what “causes” crime, but a more informed consideration, looking at all ­factors, ­including a person’s biology, is likely to bring us closer to reliable and more humane answers. We work, therefore, on the principle that in the effort to understand a complex human phenomenon such as crime, all factors must be considered. And in this introductory chapter, we first try some critical thought experiments to clarify this complexity. With luck, these will give rise to a healthy skepticism about research in human behavior and especially about scientific research that purports 1

2  Biology and crime

to explain crime. We next explore how many of the biological bases for particular criminal behaviors are, at least in potential, treatable. This is the promise of biological research into crime. While biology may offer a limited answer to larger, more philosophical questions about crime, the approach explained in this text has some distinct advantages. From this discussion, some further cautions about studying crime in general are explored. For example, we evaluate scientific methods throughout this text, and we need to engage some of the problems that can occur in any scientific research. This chapter reviews some of the social history from which this approach to crime has arisen. Up until the very recent past, the topics discussed in this text were considered too controversial to discuss in criminology. Even when the first edition of this text was published, the subject matter was still considered contentious. One of the important areas of controversy is, of course, the use of genetics to explain criminality. When social scientists respond negatively to biological ­explanations, they often assume that they are based on genetics. Until recently, genetic explanations of criminality were much out of favor, because they were not  understood and because such misinformation has been used reductively (and dangerously) in the past. They are now often strongly associated with past mistakes, such as the eugenics movements, or with the genocidal regimes of Nazism and Stalinism, in which whole populations were denied human rights based on supposed genetic differences. There are, then, some very good reasons to be cautious of their use. In relation to the approach developed in this book, it is important for the reader to realize that any reduction of the causes of crime to genetics alone can be a serious and dangerous distortion. Through a critical look at the field’s checkered past, we hope in the following chapters to explore its much brighter future.

The question of biology, crime, and the environment What then is the general place of biology in an explanation of crime? To explore this question, we first consider a less controversial but parallel question about genetics. We could ask, for example, a good biology question: Do our genes govern how we write? One advantage we would expect in taking this “genetics alone” approach is that we should receive a clear and precise answer. This, after all, is what scientific research promises.

Is writing learned environmentally or controlled genetically? Both common sense and scientific evidence tell us that the way we write is affected by our physical ability—how dexterous we are and how well we can manipulate objects with our fingers. Dexterity could then be determined by genetic makeup, the facility with finger movements and hand-eye coordination we are born with. But without much effort, one can probably find an important ­qualification. Our ability to write is obviously also governed by non-genetic influences. Conditions in the womb, for example, have an influence on the development of hands and arms and can therefore affect later dexterity and thus later writing skills. Another qualification: we are not born able to write; we have to learn how to write. We write in a particular way because we were taught this way. We incorporate and model the scripts of the people who are significant in our early lives—teachers, older siblings, and parents. It is not too difficult to find an increasing array of other clearly environmental influences. A hand injury in childhood could also affect the way we learn to write, and equally, an injury in adulthood can change the way we write, as we learn to compensate for the injury. Considering this multiplicity of real and potential environmental influences, how much of handwriting can be considered truly genetic in origin? The  answer is that not  much of it is genetic—it is really almost all a learned behavior. But we can now  take our thought experiment a bit further and ask a question to which the reductive genetic approach might provide a better answer. The  ability to write might have

The question of biology, crime, and the environment  3

evolved with the human species, and perhaps, there is a genetic influence on this level. We might now refine our original question to “What makes us, as a species, capable of writing? What abilities do we have that other species do not? Why, for example, can a dog not write?” Part of the answer involves degree of intelligence—the relative size, organization, and ability of the brain. For this reason, most dogs probably could not figure it out (border collies may be an ­exception). In general, we would think that dogs would not have the higher-level skills in abstraction, ­sensory focus, dexterity, and memory that are required for writing and reading. What else, then, do we (and perhaps chimps) have that allows us to write while dogs cannot? Another part of the puzzle is that we have opposable thumbs and dogs do not; it is difficult to hold a pen (or type) without an opposable thumb, although some birds manage to use sticks to get at insects, and chimps can be trained to use many tools. Moreover, people who have lost hands can write and even paint by using their toes or teeth. In general, however, dogs do not have this ability. Thus, if you go back far enough into any feature or skill involving the body, it does appear to come down to genetics. Genes produce the specific attributes of the physical body that make us capable of writing—that is, a hand with fingers and a thumb, and a relatively capacious and intelligent brain that can instruct that hand to learn how to write. Is writing then “caused” by genes? The problem is that at this level, the genetic argument p ­ roduces a form of absolute certainty by what might be termed reductive generalization. Such certainties bother scientists—they are always either untrustworthy or of no real use. Yes, we can strongly and confidently assert that the brain itself is inherited and under genetic control and that opposable thumbs are likewise under genetic control. You have a brain because humans evolved one, and the trait of producing a big brain and opposable thumbs is certainly inherited. Such assertions cannot be really contested. But the problem with them is, again, that even on this more general level, they are reductive of actual behavior. You are born with a brain, but what you do with it is a complex mixture of the brain itself, the inherited part, and the social environment—the most important part. The inherited big brain, the finger dexterity we share with the great apes, and the social environment that trains you (for better or for worse) in the necessary skills of script production and interpretation are all vitally necessary for you to write well, to be understood, and to understand writing by ­others. Learning how to read and write is a highly complex social and biological matter, as any speech therapist will tell you. Genetics is certainly not the only factor in acquiring this ability. Thus, if you go back to the evolutionary level, you can relate even writing to genetics as a cause. But you may be able to now see that it is a rather one-sided explanation as far as behavior is ­concerned. Social development, education, and many other environmental influences are utterly necessary as well. Unfortunately, genetics alone cannot give us the magic explanation we wish, because so much more is involved in any aspect of our functional, social interpretive, painfully learned, and all too human behavior.

Is there more to biology than just genetics? The  argument for purely genetic explanations of behavior, and thus crime, can be qualified in another way. Although genetics is part of biology, biology, as the “science of life,” covers much more physiological territory than genetics. In this text, we certainly consider genetics, but we will also examine a host of other biological explanations, such as hormone levels; the effects of disease, diet, and pollution; neurotransmitters; brain injury; and prenatal problems. Some authors place such material under genetics, but I think such presentations can be confusing. Note again their tendency to reduction. Biology, when it is good science, cannot be reductive.

Resistance to biological explanations Over the last decades, there has been considerable resistance to even thinking through the issues involved in biological approaches to crime. A behavioral scientist recently stated that for the past

4  Biology and crime

30–40 years, “most criminologists could not even say the word ‘genetics’ without spitting.”2(p972) Even in 1992, the National Institute of Health in the United States withdrew funding for a ­conference on crime and genetics, citing that a genes-crime link “stood for racism and eugenics.”2 (p972) In many cases, social scientists were afraid even to conceive of biological causes for crime. As mentioned in the preface, Wright and Miller showed in “Taboo until Today”3 that even the highly educated boycotted the possibility of considering genetic and biological explanations. According to the article, this reaction comes from the idea’s association with the vicious prejudices of the past, including the horrors of ethnic cleansing, slavery, and the genocides of World War II. However, the fact that an idea has been abused and misused does not mean that it should be ­forgotten or that it is wrong. It is vitally important to realize that although these atrocities were horrific, they were not based on science. They were based on the prejudices and psychopathic policies of people in power who decided to misquote science to an uneducated public to fulfill their own immoral agendas. There is the danger of repeating the prejudices of the past if we do not understand the truth of the science. It was the public’s ignorance of the true facts that allowed such people to use these misrepresentations as weapons. Another reason for the resistance is that some also think that a biological explanation means there is no hope for treatment. Some fear, for instance, that such an approach may lead politicians to make laws that require incarceration of people with a biological predisposition for antisocial ­behavior—they must be locked away because they can never “get better.” This view is, again far from the truth and underscores the weakness of this kind of criticism. The biological view actually offers much more hope for those afflicted with these predispositions than social explanations. For example, as a ­society, we now seem to accept that terrible abuse during childhood could predispose a child toward ­criminality in later life. But this explanation offers no cure; we can never take away that abuse. We can try to ameliorate it, we can try to change society for the better, but we can never change the fact that it has occurred. In contrast, the outcome is actually much more hopeful if the behavior has a biological basis, because there are many biological states that we can treat or change.

The promise of biological research With these initial cautions in mind, a picture is emerging of careful testing of the areas where this research paradigm can be applied, where it cannot be applied, and where it might offer a positive benefit. The most exciting thing about looking at biological research is its promise of providing successful treatment. Here are some examples where biology certainly has a profound influence on a disease or behavior and in which successful treatment has been found. They are not all related to crime, but they set the stage for our approach in that they can show biological conditions for which, through research and treatment, predicted outcomes changed dramatically. These examples should give us some hope that a similar change in biologically based types of crime could, with adequate research, be available and that they could be arrived at in a similar way.

1. Phenylketonuria In the old days, a percentage of children were born with a genetically based disease called phenylketonuria (PKU). Children who have PKU are unable to digest or metabolize phenylalanine, an amino acid that is essential for life.4 We get some amino acids from a balanced diet and make others ourselves; phenylalanine comes from diet. In children with PKU, the amino acid builds up to toxic levels that cause severe intellectual disability by the age of 5 years. In fact, most people with intellectual disabilities in the past had this disease. The cause of the disease was discovered in the 1930s, and the genetic basis was finally understood by the 1950s,5 making treatment easy: simply keep most phenylalanine out of the diet

The promise of biological research  5

(a tiny amount is required, but no more) until children are at least 7 years old, at which point high levels can build up with much less damage. Once this was understood, children born with the inability to metabolize phenylalanine no longer showed brain impairment, and the level of intellectual disability in the population dropped dramatically. Now, all babies are tested at birth (their toes pricked for tiny blood samples), and those at risk are fed restricted diets. The dramatic result is that there are no more children with PKU. Actually, that is not true; the disease is still prevalent. In fact, it is more prevalent than before because these children have grown up normally and had children themselves, passing the disease on. But their offspring do not show the effects of the disease because we have changed the biological outcome. This was one of the first cases to prove that we are not completely ruled by our genes. With the right knowledge, we have the ability to change our genetic future.

2.  Cystic fibrosis This terrible disease has been traced to just one gene.6 People who have cystic fibrosis (CF) lack this gene and therefore develop serious lung problems that eventually result in early death. Scientists have been experimenting with ways to implant the missing gene in people. To simplify, it is possible to use a virus to place a gene somewhere; this is termed gene therapy. Scientists put the required gene into the virus, and the virus carries the gene into the person. A virus works by getting inside your cells. To kill a virus, one must kill the cell it is in. Unfortunately, in the process, the host is often killed; this is why the common cold is so difficult to cure. Scientists have experimentally placed the missing gene that causes CF into a flu virus, which is modified so that it is incapable of causing the flu but still capable of penetrating a cell.7 When a person inhales the impregnated virus, in the way one might inhale an asthma medicine, the gene rides the virus straight into the mucosal tissue and the lungs and thus directly into the target cells. The  gene only lasts in those cells for a few weeks, because cells die ­constantly. While most of us ­produce new cells that have all the right genes, people with CF do not. The gene is missing from their original DNA, so the replacement cells do not have it either—but the person can inhale it again. More recently, non-viral methods of delivering the gene are being explored.8 This treatment is still being researched, but if it continues to show promise, it might help people with CF live beyond their mid-twenties. Even more exciting, if people who are missing this gene are treated from birth, they might escape all the terrible side effects of the disease and enjoy long, normal lives. Similar gene therapy is now being used to treat numerous diseases, including blood, liver, and eye disease and cancer.9 Both these examples are unusual because they each concern just one gene, which is rare. Behavior on the other hand involves many genes and is, as we have discussed, heavily influenced by the ­environment. Here is a more crime-related example.

3. Serotonin Serotonin is a substance that the brain uses to facilitate communication among cells. Low serotonin levels have been found to result in impulsivity and violence. We will look in depth at this ­neurotransmitter in Chapter 9. Low serotonin can result from genetics, alcoholism, and various other factors, and it can be corrected by something as simple as diet. In  sum, these three examples show us that DNA  does not  necessarily mean destiny. You might now be able to see that fear of reductionism and a potential repetition of the errors and moral outrages of the past should not blind us to the potential of modern biological research to help solve some parts of the puzzle of criminogenic behavior.

6  Biology and crime

Some obvious examples of biology’s effect on behavior In case you yet have a doubt that biology has an effect on our social and emotional lives, consider the following examples of some clear biological influences on behavior.

Puberty Puberty is an example with which everyone is familiar. During the transition from childhood into a sexual world, teenagers go through some profound emotional and behavioral changes, usually suffered along with them by their parents, other relatives, and teachers. Eventually, they mature and become adults, but, in the years of puberty, the change in outlook and disturbance of ­behavior can be profound. But again, the environmental influences are crucial in how puberty, a b ­ iological change, is realized. As we all know, puberty occurs when a lot of other changes are occurring in  children’s lives. Adolescents have a growing awareness of adulthood and their role in a peer group, and they often think that they have become adults and should be treated as such, whereas adults around them continue to treat them as children or at least as teenagers. Their hormone fluctuations result in mood swings and behavior changes, and these cause very real, and often ­embarrassing, situations to occur. Eventually, at biological maturity, the hormones level off and the mood swings cease. Negotiating the new pitfalls of a social life that now includes sexuality and managing the integration of the new sexual needs and awareness (biology) with peers and adults (society) are major problems that confront teens; most make the adjustment. Can you think of any benefits of such early and biologically based behavior from an evolutionary point of view? Medically, the ideal age for childbirth is when women are young, and not very far back in the past, most women reproduced at a very young age. In addition, young boys became men early, fighting in wars, working, and marrying at an early age. The mean age of death was also lower by far than today, so if people did not reproduce when they were young, they would not live long enough to raise and protect their children. Thus, these biologically based changes in behavior may stem from earlier days when people reached maturity much earlier, competing for mates and resources and reproducing during their teenage years.

Pregnancy Pregnancy also affects hormones and therefore emotions. In many cases, a woman seems calmer than usual when pregnant, not  reacting as she normally would to the stresses of work and life. Hormones in pregnancy have a lot of functions, including effects on mood. Can you think of any evolutionary reason why women should be calmer during pregnancy? We will consider evolution and natural selection in the next chapter to explain the mechanisms, but, in most cases, for a trait to be kept through many generations, it must have some benefit (even if the benefit is rather obscure) and it must be genetically controlled. The benefit of feeling calm during pregnancy is fairly obvious. We all know that stress is bad for us. If you are stressed, you are also more likely to become sick. Stress can be even more severe during pregnancy, not only because it may harm the mother but also because it could have severe deleterious effects on the fetus. Therefore, hormones that protect against stress have major effects on behavior, which are beneficial to the fetus.

Menstruation Menstruation affects some women more than others, but many have more mood swings during, and just before, menstruation. The reason is that women’s hormones change drastically during this time, with dramatic up and down surges of estrogen and progesterone.

Resistance to biological criminology  7

We can see more clearly, then, that biology is an important influence on behavior and that scientific research can help us find out what this influence is. Our thought experiments were designed to show that biology and society, the body and behavior, learning and genetics, phases in physical development and social mores are interactive.

Resistance to biological criminology If we actually think about it, there are myriad examples of the way biology affects behavior. So why is there still such resistance to this idea? Next, we need to explore more fully where this resistance comes from. In the past, some people erred strongly in the opposite direction and pronounced biology—and especially genetics (long before genes were discovered, let alone understood, in many cases)—as the key to the understanding of all behavior, and the present resistance in the social sciences has a lot to do with this error. According to the older uses of biological models, social life did not count for much as an influence; the great “new” idea of the late nineteenth and early twentieth ­centuries was to see biology as the inevitable and sole cause of criminal behavior. In the past, there were many efforts to explain crime simply and directly, through biology. However, as we will see, the data (if they can be called that) on which this view was based were often wrong. These ideas resulted from poorly designed experiments or experiments performed in deliberate attempts to support specific ideas. Many of the individuals involved believed that behavior could be explained entirely by biology, which as we have seen with the case of genetics, is simply not  true. Today, no right-minded scientist would try to tell you that any complex behavior could be entirely biological in origin. Some very simple behaviors are under total genetic control, as we will see in Chapter  2, but no complex behaviors could be. There  is always going to be an environmental component—usually a large one. In the early nineteenth century, inaccurate science or pseudoscience was used to support many proposed theories, such as that of phrenology, a belief that the shape and characteristic bumps of the head could be used to determine personality. It was an extraordinary idea by today’s standards, but in the 1800s, it was an accepted practice. The erroneous belief that crime was entirely biological has also led to many horrible political decisions, most notably in Nazi Germany. It  is important to point out, however, that although some people believed these theories and managed to convince people in power to act on them, most people—particularly scientists—did not believe them and in fact derided the believers at the time. Unfortunately, these right-minded individuals were ignored, as most of the general public did not understand science well enough to be able to dispute the claims. The author hopes that this book will help people understand the true science of behavior and also make them ready to dispute false science or pseudoscience if it is ever misused again. Since that time, much more unbiased research has been performed, and that is what we look at throughout this book. It is not a good idea to ignore the good research that has been performed rigorously in more recent years simply because people in the past twisted the theories to support their bigotry. Honest research should not be dismissed or suppressed for the fear that politicians or others will misuse it. A tool can be used in many ways. It is one of the ethical cornerstones of this text that if scientific research into criminal behavior is used properly, it can prevent social injustice.

The history of biology and crime One of the biggest stumbling blocks facing researchers in this field for many years is the fear of determinism. Because the early attempts to understand the basis of crime were, in most cases, unscientific, biological explanations of crime and violence came to be viewed as deterministic and oppressive.

8  Biology and crime

There were major errors in many other fields as well at the time, but their failed theories are not used to judge the present work. Unfortunately, in this field, some of the old ideas resulted in horrible actions that have not been forgotten. It must be clearly understood that past atrocities were not based on true science but occurred when people with very specific agendas used public ignorance of science to further their own causes. Long before people understood almost anything about science, they were trying to understand why some people committed crimes and others did not. Because we can see and measure physical attributes, the search for the difference between those who kill and those who do not began, with attempts to correlate physical features with personality traits. Franz Gall and phrenology, eighteenth and early nineteenth century

Franz Gall (1758–1828) was a physician and anatomist who believed that there was a relationship between a person’s mental attributes and the shape and size of the head. He believed that he could determine which parts of the brain were responsible for different emotions and behaviors and that the relative size and shape of bumps in the skull over each area related to the size of the part of the brain beneath the bump and so could be used to predict a person’s personality and subsequent behaviors. This theory was termed phrenology. Phrenology now seems absurd, and there were certainly many people who considered it absurd at the time. We tend to think that it was generally accepted, but Gall had many critics at the time who pointed out the obvious holes in his theory when it was first put forward. They argued that his theories were not based on any scientific evidence or clinical data. He leapt to conclusions based on observations seen in just one patient. For example, he stated that dangerousness could be predicted based on the presence of a lump close to the ear, as he had observed such a lump both in a student who tortured animals and in an executioner.10 This was the sum total of his so-called evidence. His “research,” such as it was, was no more accepted by intelligent people then than it would be now. Nevertheless, Gall attracted an international cult following. Gall’s ideas were brought to the United States by one of his supporters, an American doctor named John Bell. Bell founded the Central Phrenology Society in the United States in 1822 and ­lectured throughout the country.10 One of the so-called studies that he was fond of quoting was written by Spurzheim, one of Gall’s students. Spurzheim had phrenologically examined 30 women who had been convicted of killing their own children. According to Spurzheim, 26 of these women had an underdeveloped brain area, which he believed was the part of the brain linked to love of ­children, and thus, he concluded that these women’s crimes were the result of their physically ­defective brains.10 These claims resemble the much more recent misunderstandings of the XYY “super male” in the 1960s, which we will discuss in Chapter 4.11(p1) In both cases, researchers looked only at offenders, a closed population, so they came to their conclusions without seeing how common the trait was in the general population. A  true control for Spurzheim’s study group could easily have included 30 women who had not killed their children, which would probably have disproved the theory immediately. When we look at any experiment, it is important to see whether it is valid, and although there are many parts of an experiment that are important, choosing an appropriate control group is vital. Lombroso and atavisms, nineteenth century

Interest in phrenology began to die out and was finally destroyed when actual experiments and data failed to support it. Public interest collapsed when the scientific community finally managed to convince the public that there was absolutely no empirical evidence to support the theory. However, the belief that it was possible to determine a person’s personality and behavioral ­patterns from assessing some aspect of their physical looks was still attractive and was taken up by the Italian criminal anthropologist Cesare Lombroso in the latter part of the nineteenth century.10 Lombroso is frequently considered to be the father of criminology.

Resistance to biological criminology  9

Lombroso performed examinations of a range of people, including convicted criminals, people in mental institutions, and cadavers, and compared his results with those from non-incarcerated ­individuals.10 He  believed that some criminals were born to be criminals and that their criminal behavior was not a product of their environment. He reported that certain features, such as sloping foreheads, jug ears, a prognathic jaw, nose size, skin wrinkles, twisted lips, and even clearly environmental features such as tattoos, were more commonly seen in criminals than in law-abiding ­citizens.10,12 Lombroso referred to these features as atavisms and declared that people possessing such atavisms had lesser or more primitive levels of development than non-criminal people.10 He was particularly interested in those who had committed violent offences, as he considered them to be inferior “morally, mentally, and physically” and likened them to Neanderthals.10(p8) Incidentally, there is no evidence that early humans were any more violent than modern humans, but people still regularly accuse uncouth and violent people of “acting like Neanderthals.” Lombroso thought that such people were born to be criminals and could not change.10 He felt that female crime rates were lower than male crime rates because women were caregivers, generally passive and weak, with low intelligence; so, when women did offend, they had to override these obstacles to crime and thus were particularly evil.13 Lombroso believed that, as biology could not be changed, such criminals should be punished owing to their perceived increased threat. He stated that approximately a third of the criminal population were born as criminals and that the rest were those who had fewer of the physical traits than the born criminal and needed an adverse environmental trigger before they offended.12 Of course, now it seems totally absurd to use the way people look to predict whether they are violent criminals. However, people do make similar personal assessments all the time. The average person puts a lot of emphasis on what can be seen. First impressions are very important; when we look at someone, we automatically size them up. We make assessments of people’s personalities and abilities by the way they look. We reassess our opinions when we get to know the individuals, but those first impressions are based entirely on looks. Women can probably relate to this fact more than men. A woman walking alone at night probably watches and assesses anyone approaching more than a man would. Most people can remember embarrassing instances of rapid assessments based on appearance, clothing, hairstyle, and so on, many of which turned out to be entirely wrong on later assessment. Of course, this is quite different than stating unequivocally that a certain look results in a dangerous person, but it does perhaps make it easier to understand why Lombroso’s theories seemed, at first, acceptable to the non-scientific and often uneducated public of the time. Francis Galton and the start of eugenics

Owing to the lack of scientific support for Gall’s and Lombroso’s theories, they eventually fell out of favor. At around this time, however, Darwin’s theory of evolution was first being discussed.14 Darwin revolutionized the way people looked at all aspects of science. His work was solid and based on years of studies. His theories had nothing to do with criminal behavior or any suggestions of attempting to control destiny, but other people began misreading his work, and dangerous and entirely erroneous relationships between evolution and crime began to be considered.10 Darwin’s theory of evolution explained how traits such as beak size and shape in birds could be selected for over time, based on environmental conditions and changes (see Chapter 2). People already knew that offspring inherited some of their parents’ features, as children resembled their parents, inheriting, for example, eye and hair color. Farmers had used such information to breed the best farm animals for centuries. Darwin’s work was sound, but evolution and natural selection can only work on heritable traits, those under genetic control (see Chapter 2). It was other people, most notably Darwin’s own half-cousin, Francis Galton, who began to suggest that just about everything was a heritable trait, including poverty and crime. Thus, not long after evolution was first understood, in some cases, it was used as a framework for developing ideas to control such things as prostitution, petty crime, poverty, promiscuousness, and ­destitution. This was a complete misunderstanding and misuse of Darwin’s work, as such characteristics are clearly not inherited. However, many people jumped on this bandwagon as a way to control the Victorian “lower classes.”

10  Biology and crime

It was then that the terrifying idea of reproductive control was first suggested. There was concern among the more powerful Victorian families that the large numbers of children born to the poor were contributing to the moral decay of society. Galton argued that the more “exemplary” members of society (meaning the rich) should have larger families, whereas the “lesser” members (meaning the poor) should be encouraged to have fewer children.10,15 Galton called his theory eugenics, after the Greek word eugenes or “good in birth.”10 Galton’s proposal was actually positive eugenics, wherein those considered most fit were to be encouraged to have more children. It is important to remember that Galton did not suggest that those considered less fit should be prevented from reproducing. Positive eugenics is what farmers have done for centuries, breeding the best to the best. Needless to say, however, it was not long before some began to advocate negative eugenics, wherein people considered unfit, for whatever reason, were actively discouraged and even prevented from reproducing.15 This frightening misunderstanding and misquoting of science drew support from around the world. At  the beginning of the twentieth century, eugenics advocates in Britain, America, and Germany began to collect information on hundreds of thousands of people and, using this supposed “hereditary data,” they began to initiate what they referred to as “genetic hygiene” measures, which ranged from segregating the supposedly unfit from the rest of society in “work colonies” to compulsory sterilization.10,15 The eugenics movement

The eugenics movement took off in the United States, as Americans were afraid that the large numbers of immigrants arriving on their shores would take their jobs and produce so many children that they would overrun the existing population (conveniently forgetting that they themselves had recently done the same to America’s indigenous peoples). Politicians pointed out that the immigrants came from countries where, traditionally, larger families were the norm, in comparison with the upper classes of New England. They declared that this would result in greatly increased rates of crime, poverty, and insanity. Some even blamed labor unrest and strikes on genetics.10 These few people managed to convince many citizens that immigration would mean the end of the average American. Things got much worse when these so-called superior people began to claim that race was a major issue and swore that their intent to control both immigration and reproduction of non-whites and eastern Europeans was not  prejudice but merely an attempt to maintain the so-called purity of the whites.10,15 This sort of terrifying propaganda resulted in the passage of immigration laws that limited the number of immigrants allowed into the Unites States from so-called undesirable countries and in miscegenation statutes that outlawed interracial marriages.10 What these proponents of negative eugenics were doing was a form of genocide, with absolutely no basis in scientific fact. In Europe, similar measures were being taken, which escalated with Hitler’s regime into genocide. Measures that had begun with public education escalated to forced abortion, sterilization, and finally death camps.10 At the end of World War II, when the enormity of the atrocities was uncovered, the general public declaimed eugenics, and the movement thankfully began to collapse, although forced sterilization continued for several decades in America and still occurs today. Eugenics in the latter part of the twentieth century

It  is often forgotten, however, that some forms of eugenics still remained in both the United States and Canada well into the 1970s, with people, frequently children, subjected to compulsory ­sterilization. It is interesting to note that while Nazi Germany at its height sterilized 75–80 ­people per 100  000  per year, Delaware sterilized 18. Although this is much less than Nazi Germany, Delaware was part of a democracy, not  a totalitarian society that shortly afterward committed genocide.16 In 1927 in Virginia, an attempt to withdraw the state’s statute on involuntary sterilization was dismissed,17,18 with the judge stating, “It is better for all the world, if instead of waiting to execute degenerate offspring for crime, or to let them starve for their imbecility, society can prevent those who are manifestly unfit from continuing their kind.”18(p946)

Resistance to biological criminology  11

Boys as young as 14  years were castrated for something as clearly non-criminal as masturbation.16 In the 1970s in America, Native American women were sterilized against their will, with estimates as high as 25% of women of childbearing age.16 Even as late as 1979, compulsory sterilization was practiced in some US states for crime prevention or punishment, and at some points, more than two-thirds of states enforced sterilization, with over 60 000 disabled people ­sterilized.16 Most were sterilized owing to mental illness or if people were of very low socioeconomic status (SES).16 There  are many people alive today who were sterilized without their knowledge, mostly at the request of parents who found them unruly or promiscuous. Today, we would like to believe that eugenics is not being practiced in our modern world—but is it? In Canada, coercive sterilizations in which indigenous women are “persuaded” to undergo tubal ligation while giving birth, in pain and under medical sedation and before being allowed to see their infants, are claimed to have occurred as recently as 2017, and cases are presently in the courts.19 There are also more subtle forms of eugenics. One could describe lengthy prison sentences as a form of eugenics. Men and women who are incarcerated for long periods of time during their reproductive years are, in a sense, being prevented from reproducing. The right to have conjugal visits under some circumstances could be said to offset this, but still, reproduction remains heavily restricted and, in most cases, curtailed. In  all these historic cases of appalling atrocities, both the theorists and average people were ignorant of the facts. They did not understand genetics, and the public did not understand that there were no valid data on which to base these theories. The media promoted the ideas of eugenics, so they were accepted by the public. Often, theories were supported by people who were already racially prejudiced. Facts that are not understood are often misused. This is why it is so important in this subject and others to prevent such public hysteria from breaking out again. There will always be people who believe this kind of propaganda, so it is essential that there also be informed people who can say with authority, “That is not true; it does not work that way.” Immigration and intelligence quotient testing

At the height of the eugenics movement, the average person did not understand genetics, but certain people found the theory useful to support their causes. In  the early twentieth century, for example, most Americans believed that intelligence was based entirely on genetics (not on a mix of genetic and environmental influences, as would be more accurate), and new immigrants were forced to take intelligence quotient (IQ) tests. The original idea behind the IQ test was that it would provide a teaching aid for children under the age of 14 years. A child’s answers to a set of questions were compared with the answers given by a large number of children of the same age. In other words, an IQ test gave an idea of where those tested stood in relation to others in their peer groups. American immigration officers used this test on adults, which made the results invalid. They misunderstood the very limited knowledge available about intelligence and used a test that had no validity in the situation and was particularly inappropriate to use on those who did not even speak English. Thousands of immigrants were sterilized before 1935 in certain states because they were not considered intelligent enough to be allowed to contribute their genes to the country. It was only with the rise of Hitler that people began to realize that what was going on in the United States was not so different, and the practice stopped. Endomorphs, mesomorphs, and ectomorphs

Meanwhile, people were still trying to use physical attributes to determine a person’s behavior. William Sheldon (1898–1977) classified people into three different groups: (1) endomorphs, (2)  ­mesomorphs, and (3) ectomorphs. Endomorphs were those with relatively soft, rounded bodies; mesomorphs were the more athletically built types; and ectomorphs were thin people.20 Sheldon thought that certain behaviors correlated with each of the three body types. Although these terms acquired support and are still referred to today, this classification, again, was not  valid. Some  studies have shown that there may be some relationship between body type and behavior, but most such studies have used small samples and have not been repeatable.

12  Biology and crime

We still judge people today by the way they look—“I wouldn’t want to meet him in a dark alley”—although we are now reluctant to admit it. It is important to remember that factors that clearly have genetic components result in major phenotypic “looks.” People’s body builds are inherited to a large extent and may affect their behavior—not in the direct manner Sheldon meant, but indirectly. For example, a large child may discover quite early that an effective way to end conflict is to use his build—in other words, to use violence—to win a conflict. A smaller person learns quickly that physical violence is not a good way to settle disputes and develops other methods to resolve disagreements, such as skillful dialogue. Early success in violence and bullying to resolve social conflicts may encourage a person to use force in adulthood. So, body build might indeed be linked to delinquency and crime in later life, but the link is most probably through social learning, not through genes. In other words, the genes give the child a large body and thus exposure to a different environment than would be experienced by a smaller child. This is yet another example of how no factor linked to crime should ever be viewed in the old terms of genetics versus the environment. Instead, we must look at both and try to understand how the two factors work together. Although the history of the misuse of biological explanations for criminal behavior is horrific and has certainly scared people away from wanting to consider biology within a criminological framework, it is not difficult to see that there was no true biology involved. Even the simplest scientific exploration exposed these theories as fraudulent, so right-minded people should not  use history as an excuse not to explore biological evidence more closely. However, there are many other considerations that contribute to this resistance.21

Determinism One of the great fears that many people have in even considering biological influences on criminal behavior is the fear of determinism—the idea that everything, including health and behavior, is predetermined, in this case by a person’s biology. Determinism can be considered the opposite of free will. This leads people to think that there could be a “gene for crime.” Most people who may have had a small amount of education in genetics at school consider only Mendelian genetics and think that there is a one-to-one relationship between genes and behavior; that is, one gene causes x type of behavior.21 This, as we will see in subsequent chapters, is patently not possible. Even before we actually look at the science, which prevents the possibility of one gene causing a complex ­behavior, there are clear criminological reasons against it. Crime is not an entity unto itself; there are vast numbers of crimes with many different nuances, from theft (shoplifting to bank robbery) to serious crime (assault to genocide). No one mechanism, be it biological or sociological, could account for all. Also, as we will see when we delve deeper into genetics, it is very rare for one gene to cause one trait, even one so simple as hair color. A large number of genes act, using a variety of mechanisms, on almost all traits and especially on highly complex and variable traits such as behavior. One overarching fear about biological determinism can be separated into two opposing concerns: people could be considered to be not responsible for their actions and therefore should not be convicted for something they could not help, or conversely, they can never be rehabilitated, and so, they need to be locked up for the rest of their lives. Even if biological factors were accepted as having a role in people’s behavior, could it be argued that this makes them less culpable?21 We will explore this concern at length in the last chapter.

Racism and sexism Another concern is that biological studies are racist, sexist,21 or many other types of “ists.” This is a very important concern, particularly in the United States, where racial tension is so high, and it has been bolstered by inaccurate and unscientific studies that purported to show that one racial

Resistance to biological criminology  13

group had higher crime rates or lower cognitive skills than another. These studies have all been shown to be completely unsound, and they also reflect the problems seen in the past. Even though such studies have been entirely discredited, there is still fear that biological studies are underhanded attempts to promote racism or sexism.21 This is clearly untrue, and an analysis of an experiment will easily determine whether it is strong science or not. However, there is an important caution that must be stated here. When we look at heritable characteristics that result in a physical attribute that in turn has social implications, it is often difficult to distinguish the biological from the social effect. For example, if possessing a particular genetic trait, such as sex or skin color, places a person at a social or educational ­disadvantage in a given society, then social failure will be attributed to the genetic characteristic,22 when, in ­reality, the genetic makeup has nothing to do with the social failure. The genes produce a particular trait, say, sex, and society decides that a person with that trait will have less opportunity than someone with the opposite trait. As a result, the person is disadvantaged, but the reason is not biological but purely environmental. In the not-so-distant past, women in Western societies were considered much less intelligent than men. Women were thought to be delicate and incapable of intelligent discourse on weighty subjects. In the Victorian age and earlier ages, it was probably true that conversing with the average woman of any social class would show that she had very little knowledge of history, geography, mathematics, science, or politics. If a researcher had used such subjects to assess the intelligence of men and women at the time, the results would surely have shown that the men were more ­intelligent. But would that have been a correct assumption? And why not? The answer is simple: men, particularly of the higher social classes, were educated, and women of any social class were not. Many women were not even taught to read, and any sort of education was considered “wasted” on a woman. After all, why did she need an education to have babies, cook, and clean? Even at social levels in which a woman was taught to read and was given an education, it was a very different education from that received by men of similar social standing. Such women were educated in music, sewing, and running a household and usually not allowed to read anything that might distress their fragile dispositions, such as newspapers. The belief then developed that women were much less intelligent than men. This was due to a simple lack of education. In such cases, it is difficult to determine what is genetic and what is purely environmental. In the past, much emphasis was placed on racial contribution to criminal behavior, and assumptions were frequently made about the intelligence of certain racial groups. However, the skin color and the racial background have, particularly in the past and even today, a tremendous effect on the level of education people receive and the type of job and life they can expect in general. In such cases, the only hereditary factor is the race or skin color, which then leads to people, environmentally, being treated very differently from those of other skin colors. Further concerns have been postulated in relation to the excellent and prolific scientific studies that have shown that men are more likely to commit crimes, particularly violent crimes, than women (as we will see throughout this text).23 Much of this relates to higher levels of impulsivity, antisocial behavior, hyperactivity, and externalizing behaviors in males than in females.21 In the next chapter, we will look at various evolutionary theories that indicate why some behavior considered to be antisocial or violent today may have been evolutionarily advantageous in our ancestral past, including gaining more resources and mate options. This has led to the fear that, if men are considered to have more biological risk factors for such behavior than women, this might be misinterpreted to support or even justify such male behavior, suggesting that men are less culpable for crimes such as violence against women.21

Mental illness and medicalization of crime There is also a concern that mental illness, which may result in a lack of self-control, and antisocial behavior, which may include many externalizing behaviors, such as aggression, stealing, lying,

14  Biology and crime

and  property damage, may be similarly confused. Such a misunderstanding could suggest that antisocial behavior is not an act of choice, indicating that it may be unavoidable in some individuals, thereby reducing its social stigma and culpability.21 There is a definite trend for the public to feel in the case of particularly brutal acts of violence that the offender must be a “crazy person” to do such things, when, in fact, the person is not mentally ill and the acts are those of choice. On the other hand, increasing the social stigma of antisocial behavior can result in labeling, particularly in youth, which can have severe overall behavioral impacts over the life course.21 There  is also a concern about medicalizing the justice system and making judges into doctors when they are asked to make decisions about whether a person should be treated pharmacologically— for example, imposing chemical or surgical castration as a punishment, treatment, or condition of release (see Chapter  7). Also, there is further concern that considering biological influences on criminal behavior could medicalize the behavior, again reducing ­culpability and indicating ­medical rather than punitive treatments.21 This leads back to the fear of determinism21 and the frequent tabloid headlines of “my genes made me do it.” This is an area that we will explore in the final chapter. There are therefore many reasons for resistance to considering biological influences on criminal behavior, most of which can be easily argued. However, we must still be cautious in our interpretation of all data to ensure accuracy and that results are not misunderstood or used in any political agenda.

Other cautions that must be considered When we start to consider studying biological influences on criminal behavior, or in fact any ­influence on criminal behavior, there are many issues that we must consider.

Predisposition It is extremely important, now and throughout this text, that you have a very clear idea of the concept of a predisposition. Any risk factor for criminal behavior is exactly that—a risk factor. It is not a certainty. No factor will definitely cause crime. Certain factors, such as low SES, child abuse, poor peer relations, and dropping out of school, may predispose a person to criminal ­behavior, but they most certainly don’t result in crime, and in fact, most people who experience one or more of these factors lead normal, non-criminal, productive lives. In the same way, any biological risk ­factor that we discuss is only a predisposition. DNA and biology in general are not destiny. They can be changed. A very simple example with which we are all familiar is whether a person with a predisposition for a heart attack is destined to have a heart attack. A person may have a genetic predisposition from both father and grandfather, but a healthy diet and regular exercise can frequently prevent an attack. Therefore, it is extremely important that we remember that any potential biological influences we may find will result only in a predisposition toward a particular behavior. There is no gene for crime, and none will ever be found; as a scientist, this author knows that genetics does not work in that way. Because even hair color is the result of multiple genes, there certainly could not be a single gene for a behavior. And even if there is a genetic predisposition, that is all it is—a predisposition. It is likely that some genes or hormones contribute to proteins that have effects on certain behaviors, which are then greatly affected by situations, environments, and social upbringing. The influences can also be indirect. For example, biological influences on intelligence might affect the chances that a person who commits a crime will be caught, and from the other direction, biological influences on impulsivity might have a bearing on whether the person will commit a crime in the first place. To explore the idea of predisposition a little further, here is another thought experiment. Let us, once again, propose a genetic basis for behavior, one like those discussed previously that have a

Other cautions that must be considered  15

known genetic risk factor. This time, however, we can find an example of criminal behavior that is well documented as having a definite genetic predisposition. The high-risk genotype for this factor is detectable at birth and often before, and the person who has it is almost 100% more likely to be convicted of a specific crime than a person with the low-risk genotype. What crime are we referring to? Rape. And the genetic predisposition? Possession of a Y ­chromosome—that is, being male. A  male is much more likely to commit rape than a female. However, maleness is only a predisposition to the crime of rape. If you think a predisposition means ­something will happen or even most likely will happen, you must believe that all men are rapists—and we know that is not true. Only a very small percentage of men are rapists. The Y chromosome is a predisposition only—nothing more, nothing less. Thus, biology and social environment always work together, and they do so in complex, subtle, nuanced ways. Almost all the research that indicates genetic or biological influences on criminal behavior also shows strong environmental components. But this is the beauty and challenge of most biological studies. By their nature, they must consider both the genes and the environment. In trying to distinguish the effects of biology from those of the environment, scientists must study both; they thus accept both. Biological studies never exist in a vacuum. Much sociological research, on the other hand, does not take biology into account at all. Biological studies of crime must look at both environment and physiochemistry in order to compare the two and determine the effect of one versus the other. All biological studies fully recognize the importance of the environment and use it as a comparable variable. In fact, biological studies have done more to prove the ­existence of an environmental influence, particularly as an ameliorating effect, than sociological studies ever have. Biology, after all, gave us the basis for the Green movement.

What is crime? We can now turn to the way in which we can use this interactionist biological explanation of behavior to approach crime. My hope here is to lay down a few of the conditions for our upcoming study. Let us begin with a fundamental question. If we are going to study the biology of crime, we will need to have a good working definition of crime. What then is crime? And a related question: What is antisocial behavior? These are actually difficult questions to answer; definitions are difficult because crime is a social construct. We cannot simply equate breaking the law with a biological cause such as disease or genetics. We might seem to do this when we refer to programs in prisons and counseling for ­juvenile offenders as treatment, which indicates that we think crim as an illness—but these programs of “treatment” are clearly not based on medicine or biology. The disease metaphor for the definition of crime is highly limited. Some forms of civil disobedience, for example, are the result of fiercely believed moral values and are designed to change policy, the law, or the prevailing practice. In the nineteenth century in Canada, Louis Riel fought for the rights of his fellow Métis and at the time was judged a criminal and hanged. Now, we recognize him as a hero. Women broke the law during the early part of the past century to win the right to vote. In the 1960s, civil rights protesters such as Martin Luther King broke the law, as they tried to eliminate racial segregation in the United States. These types of “crime” are not the result of the biological backgrounds of those who protested, any more than they are, in any determined way, the result of their environments. They arose from particular injustices and appeared at specific times—they were based on a rational moralism of choice. Legally, they broke the law, but today, they are celebrated as heroes, and no one could possibly consider them criminals. There are also many so-called crimes, such as smoking marijuana, or the ancient peyote cults that are accepted by the groups that practice them and often by society at large. Some behaviors that are labeled crimes, such as smoking and drinking, are only crimes when people below a certain age indulge in them, and this age limit varies depending on where you are in the world.24 Eminent researchers Ellis and Walsh referred to this as the “moving target perspective.”25(p5) There  are numerous examples of actions that are considered criminal in one region or country

16  Biology and crime

but are perfectly acceptable in others, such as female genital mutilation or marriage of young girls to adult men, which are widely practiced in some parts of the world and yet are considered child abuse and rape in others.25 Moreover, behaviors and actions that were once considered criminal are no longer a crime today, and vice versa. For example, in the past, homosexuality was a crime punishable by castration and even death. Conversely, in the late 1800s, in the United States, the Bayer Drug Company, now famous for producing Aspirin, legally sold heroin over the counter as a cough ­suppressant.25 The use of marijuana, which resulted in criminal records for many people for decades, is now legal in Canada and many US states. It must not be forgotten that crime in itself is a legal concept based on political processes. As such, all crime in a country could be eliminated in one sweep by simply eliminating all its criminal statutes. This would effectively eliminate crime but would have no effect whatsoever on criminal behavior.25 Although this is a bit far-fetched, countries are always changing and revamping their criminal statutes, thereby redefining what is or is not legally considered a crime. If we try to restrict the concept of antisocial behavior to the one that is socially disapproved of in practically all societies, such as killing someone, we find that, in war, even this most violent of acts is positively approved.24 In war, killing in defense of your country is usually not thought of as a crime but as an act of bravery and survival and often an act of heroism for which medals are awarded. The act is the same; the interpretation is radically different. Tendencies that are bad in some situations may thus be useful in others. Aggression may lead to violence, for instance, but in some professions, such as the military, this is useful behavior; politicians or entrepreneurs who have retiring personalities will rarely do well. Ellis and Walsh distinguished between acts that are almost always criminalized, such as intentionally killing or harming someone or stealing someone’s property, versus those behaviors and actions that are sometimes criminalized and sometimes not. The  usual practice in criminology is to refer to them as mala in se and mala prohibita, respectively. Mala in se refers to something that is inherently bad, and mala prohibita refers to something that is bad because it is prohibited. This is referred to as the stationary core perspective25(p7) in criminology and has use for this introduction. The distinction helps us understand that we are not usually concerned with biological influences when we consider socially constructed labels of criminality. When we think of crime in the context of biology, we almost automatically think of violent crime; that is, we suppose that a violent nature may be inherited or is caused by physiological determinants such as hormones. Violent crimes are defined as those that “cause or threaten to cause bodily harm to another person.”25(p27) Although in this text, as in most research into crime, we concentrate on violent crime, we must remember that violent crime is actually quite rare. We think of it more often than common crimes such as ­shoplifting because the popular media too often emphasizes and exploits violence. So, although the bulk of the research we discuss concentrates on violent crime, we must also consider non-violent crime in our study. As we will find, some research also shows a predisposition to nonviolent crime.

Cautions for all criminological research As we found with our thought experiments on genetics and its value in explaining behavior in general, any criminal research, not  just that dealing with biology, requires equal caution and an appreciation of interaction with environments. Many of the following cautions were first examined by Trasler in his chapter of The Causes of Crime: New Biological Approaches, when people were just beginning to be challenged to reconsider biology as a part of criminological study.22 These cautions are still valid today and have since been further examined by other researchers. The following basic facts about crime and crime statistics must be considered before beginning our study into biological influences.22

Other cautions that must be considered  17

1. Adolescence

Crime in adolescence is very common; in fact, it is almost the norm.26 Studies in many countries have repeatedly shown that a surprisingly large number of adolescents are involved from time to time in minor crimes, such as theft and property damage. In the past, only a few adolescents were thought to take part in such crimes, but this is clearly not so. It appears to be so common that it could almost be considered a normal part of growing up in much of Western s­ ociety. 22 In  fact, Moffitt in 1993  defined crime as either occurring in a small number of individuals throughout the life course (“life-course-persistent antisocial behavior”) or occurring in a large number of adolescents but ceasing after the adolescent period (“adolescence-limited antisocial behavior”). 27(p674) This means that much of crime is ephemeral; that is, delinquency is often a transient or passing phase. In the past, some believed that such individuals stopped because they were locked up and thereby “cured.” However, self-report studies show the opposite. People who admit that they committed crimes as juveniles are also a subset of the whole. Although the admission is voluntary and anonymous, we have to ask whether it is representative. Still, in such self-reporting studies, it appears that many young people just seem to stop committing crimes. Their behavior may be a natural part of maturation or adaptation to the different circumstances of adult life. For example, peer pressure is immensely important in the teenage years, and teenage crime is often a social activity and an adventure. As people grow up, they develop other interests and can gain satisfaction from jobs, girlfriends, or boyfriends and eventually spouses and children and have much to lose by committing a criminal act. They become less dependent on peers, and their new life patterns are inconsistent with delinquency, so they simply outgrow it.22 2.  Research samples

Many studies compare offenders with non-offenders. However, people with criminal records are not necessarily representative of all the people who have committed offenses. Studies l­ooking at the factors that affect whether a police officer will record or overlook an offense—that is, charge a person or just give a warning—indicate that such decisions can selectively exclude some people and include others.22 As a result, the people convicted of particular crimes are just a small sample of the people who commit such crimes and may be a biased sample. Other factors, such as lack of skill or intelligence, impulsivity or lack of impulsivity, planning, and plain luck, may affect whether a person is caught. Caught offenders may also not be representative of all who offend. A study in the late 1970s28 showed that there was a distinction in arrest rates between intensive offenders, who were continuously engaged in crime, were committed to criminal lifestyles, and were careful to avoid arrest, and those who committed crimes irregularly, with less care and planning. This research ­estimated that the average intensive offender committed 10 times as many crimes as the intermittent offender but was 5 times less likely to be arrested for any one crime. Once arrested, the intensive offender was also less likely to be convicted and incarcerated. So, caught offenders are not  necessarily representative of all offenders, as many offenders are not caught. This means that the converse is also true: non-offenders are not necessarily representative of people who do not commit crimes, as clearly, many people are committing crimes and have not been caught. The tremendous number of unreported sexual assaults, particularly emphasized by the #MeToo movement, shows that most rapists are not caught. 3.  Underreporting of crimes

This leads to the problem of the underreporting of a large number of crimes, from very serious crimes such as sexual assault to much more trivial crimes such as shoplifting. Many crimes, such as shoplifting and vandalism, are not reported unless the culprits are actually caught in the act. However, estimates of stock losses in retail stores show that such minor crimes take place on an enormous scale, so they must involve a large number of people.22

18  Biology and crime

4. Self-reporting

Many studies measure crime or antisocial behavior by using self-report measures, such as questionnaires and interviews. These may be inaccurate, inflated, or deflated. Also, it is often difficult for respondents to give accurate information about breaking the law. Many people may not fully understand what being arrested entails and may believe that a simple street contact with a police officer was an arrest.22 Moreover, people may underestimate the number or seriousness of the crimes that they have committed or may exaggerate them in order to boast. In some cases, incorrect reporting may be as simple as not wanting to fill in more questions. For example, people learn quickly that if they admit to an offence, they will be asked many more questions, so they may opt for “no” to avoid more questions. This is termed a testing effect.29 In one study, the self-reported arrests, jail terms, and prison terms of 700 incarcerated men were compared with their official records for the same time period and showed significant errors in self-reports of arrests, although jail and prison time reporting was more accurate. Men with very high numbers of arrests had the poorest self-report accuracy.30 5. Consent

Many studies involve incarcerated offenders; this leads to a major ethical issue. Can an incarcerated individual every truly give free consent? People have their entire lives under the control of others while incarcerated. Therefore, they may feel that participating in a study may help their chances of parole or may increase privileges in prison and that refusing may hinder their chances of parole. This means that even when consent is given, it could be considered coerced, even if the coercion is perceived rather than actual. In addition, are there differences between offenders who are willing to participate and those who are not? Do some have personality types that want to be involved, or do they wish to boast (or lie) about their criminal histories? In  many cases, it has been suggested that taking part in a study alleviates the boredom of prison life and may even result in a trip outside the prison walls. A female colleague recently stated that when an offender was asked why he was participating in the study, he said that it was because he had not seen a woman in 30 years.

The future of biosocial criminology As this text, and many others, will show, there is now no denying that biology has a role in our behavior and that includes antisocial and, yes, even criminal behavior. The  vast wealth of good empirical research, which we will barely be able to touch upon in this text, shows that our biology interacts with our environment and affects every aspect of our being. More recent understanding of gene × environment interactions and epigenetics has now shown the mechanisms by which our genotype interacts with and is molded by our environment and our life experiences to shape our behavior. In  this chapter, I have briefly discussed the misuse of biology in early criminology and its insidious impacts on subsequent research in this area. But finally, there appears to have been what many consider a paradigm shift toward a better understanding of the interrelationships between our biology, our environment, and our behavior. Others have suggested that rather than a paradigm shift, it is a natural maturing of the field,31 perhaps a coming of age. Social scientists are becoming much more open to accepting the role that biology plays in our lives and consequently our behaviors. The tremendous increase in solid empirical research published in the last few years supports this.32 A particularly telling legal decision is discussed by Gajos and colleagues in an article on genetically informed prevention, which perhaps best illustrates this paradigm shift or maturity.33 Roper v. Simmons (2005)34 was a landmark US decision in which the death penalty was abolished for juvenile offenders, and it is considered to be one of the most important and powerful legal decisions that has been made

References  19

in the last two decades.33 Although this decision was important, its most interesting aspect relates to the evidence on which the decision was based. The court did not consult the two foremost criminology societies in the United States, the American Society of Criminology (ASC) and the Academy of Justice Studies, nor did they utilize research from their highly prestigious journals; instead, the courts relied on neurological and neurocognitive scientific research, which proved that the adolescent brain was not fully developed by the age of 18 years.33 Despite the fact that the ASC had worked for years in an effort to abolish the death penalty, it was biosocial research rather than traditional social criminology that influenced this important legal decision.33 This decision is one that perhaps illustrates the fact that biosocial research can be used to impact very major and liberal justice policies.33 It now behooves us to better understand the interrelationships between and within our genotype and our environment in order to manage antisocial behavior. Biological influences on criminal behavior exist and provide us with new insight into ways to moderate, ameliorate, or even prevent poor outcomes and stimulate prosocial development.

Conclusion The study of criminology has been remiss to the extent that it has not, until quite recently, included biology as one of its necessary subdisciplines. The reasons are understandable; the atrocities of the past are powerful deterrents. But these were, as we have seen, based on bad science, inhumane political agendas, and execrable ethics. We now have many valid empirical studies and a humane ethic on which to base ­serious analysis of the biological dimension of behavior—and, as we will see, criminogenic ­behavior. The effects of biology on behavior are very clear. Inherited diseases, hormonal changes, and ­physical changes during pregnancy are some of the aspects of biology that impinge on our existence and influence our behavior in serious ways. We have also begun to see the ways in which science itself can go wrong in the study of behavior: small sample size and poor experimental design can vitiate results. This provides a good way to evaluate all the studies we will be surveying in the following pages.

Questions for further study and discussion 1. Explain the problems with attempting to use the term crime. 2. If a researcher were to compare the crime rates in people from, for example, one racial group with another and found that crime rates were much higher in one particular racial group, what major caution should be considered, and what other research would be needed? 3. There  are many reasons why there has been resistance to accepting any form of biological influence on behavior. Explain and discuss these. 4. Discuss why imprisonment could be considered a form of eugenics and explain whether it is positive eugenics or negative eugenics.

References 1. Walsh, A. and Beaver, K.M. 2009. Biosocial criminology, In: Handbook on Crime and Deviance, Korhn, M.D., Lizotte, A.J., and Hall, G.P., editors. Dordrecht, Germany: Springer. pp. 79–102. 2. Denno, D.W. 2011. Courts’ increasing consideration of behavioral genetics evidence in criminal cases: Results of a longitudinal study. Mich. St. L. Rev 967–1047. 3. Wright, R.A. and Miller, J.M. 1998. Taboo until today? The coverage of biological arguments in criminology textbooks, 1961–1970 and 1987–1996. J. Crim. Justice 26(1): 1–19.

20  Biology and crime

4. Singh, R.H., Cunningham, A.C., Mofidi, S., et  al. 2016. Updated, web-based nutrition management guideline for PKU: An evidence and consensus based approach. Mol. Genet. Metab. 118(2): 72–83. 5. Mednick, S.A. 1987. Biological factors in crime causation: The reactions of social scientists, In: The Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. New York: Cambridge University Press. 6. Rommens, J.M., Iannuzzi, M.C., Kerem, B., et  al. 1989. Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245: 1059–1065. 7. Moss, R.B., Rodman, D., Spencer, T.L., et al. 2004. Repeated adeno-associated virus serotype 2 aerosol mediated cystic fibrosis transmembrane regulator gene transfer to the lungs of patients with cystic fibrosis: A multicenter, double-blind, placebocontrolled trial. Card. Crit. Care J. 125(2): 509–521. 8. Alton, E.W.F.W., Armstrong, D.K., Ashby, D., et al. 2015. Repeated nebulisation of non-viral CFTR gene therapy in patients with cystic fibrosis: A randomised, double-blind, placebocontrolled, phase 2b trial. Lancet Resp. Med. 3(9): 684–691. 9. Naldini, L. 2015. Gene therapy returns to centre stage. Nature 526(7573): 351–360. 10. Niehoff, D. 1999. Seeds of controversy, In: Biology of Violence: How Understanding the Brain, Behavior and Environment Can Break the Vicious Circle of Aggression, Niehoff, D., editor. New York: Free Press. pp. 1–30. 11. Mednick, S.A. 1987. Introduction: Biological factors in crime causation: The  reactions of social scientists, In: The Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. Cambridge, UK: Cambridge University Press. Proceedings NATO Conference, Skiathos, Greece, September 20–24, 1982. pp. 1–6. 12. Baum, M.L. 2011. The monoamine oxidase A (MAOA) genetic predisposition to impulsive violence: Is it relevant to criminal trials? Neuroethics 6(2): 287–306. 13. Lombroso, C. and Ferrero, W. 1898. The  Female Offender. New  York: D. Appleton and Company. 14. Darwin, C. 1859. On the Origin of Species by Means of Natural Selection; or the Preservation of Favoured Races in the Struggle for Life. London: John Murray. 859 pp. 15. Kupferman, I. 1991. Hypothalamus and limbic systems: Peptidergic neurons, homeostasis and emotional behavior, In: Principles of Neural Science, Kandel, E.R., Schwarz, J.H., and Jessell, T.M., editors. East Norwalk, CT: Appleton and Lange. pp. 735–749. 16. Kaelber, L. 2012. Eugenics: Compulsory sterilization in 50  American states. Accessed November 22, 2018; http://www.uvm.edu/~lkaelber/eugenics/. 17. Buck v. Bell, 247 U.S. 200 (1927). 18. Appelbaum, P.S. 2014. The double helix takes the witness stand: Behavioral and neuropsychiatric genetics in court. Neuron 82(5): 946–949. 19. Moran, P. 2018. 13 November. Indigenous women kept from seeing their newborn babies until agreeing to sterilization, says lawyer. CBC Radio. https://www.cbc.ca/radio/­thecurrent/ the-current-for-november-13–2018-1.4902679/indigenous-women-kept-from-seeingtheir-newborn-babies-until-agreeing-to-sterilization-says-lawyer-1.4902693. 20. Sheldon, W.H., E.M., H., and McDermott, E. 1949. Varieties of Delinquent Youth: An Introduction to Constitutional Psychiatry. New York: Harper. 21. Berryessa, C.M., Martinez-Martin, N.A., and Allyse, M.A. 2013. Ethical, legal and social issues surrounding research on genetic contributions to anti-social behavior. Aggress. Violent Behav. 18(6): 605–610. 22. Trasler, G. 1987. Some cautions for the biological approach to crime causation, In: The Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. Cambridge, UK: Cambridge University Press. Proceedings NATO Conference, Skiathos, Greece, September 20–24, 1982. 23. Ellis, L. and Hoskin, A.W. 2015. Criminality and the 2D:4D ratio: Testing the prenatal androgen hypothesis. Int. J. Offender Ther. Comp. Criminol. 59(3): 295–312.

References  21

24. Rutter, M. 1996. Introduction: Concepts of antisocial behaviour, of cause and of genetic influences. Ciba Found. Symp. 194: 1–15. 25. Ellis, L. and Walsh, A. 2000. Evolutionary biosocial theories, In: Criminology: A  Global Perspective. Boston, MA: Allyn & Bacon. pp. 432–468. 26. Boyd, N. 2000. The  testosterone connection, In: The  Beast Within: Why Men Are Violent. New York: Greystone Books. pp. 115–138. 27. Moffit, T.E. 1993. Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychol. Rev. 100(4): 674–701. 28. Petersilia, J., Greenwood, P.W., and Lavin, M. 1978. Criminal Careers of Habitual Felons. Washington, DC: National Institute of Law Enforcement and Criminal Justice. 29. Krohn, M.D., Thornberry, T.P., Gibson, C.L., and Baldwin, J.M. 2010. The development and impact of self-report measures of crime and delinquency. J. Quant. Criminol. 26: 509–525. 30. Roberts, J. and Wells, W. 2010. The validity of criminal justice contacts reported by inmates: A  comparison of self-reported data with official prison records. J. Crim. Justice 38(5): 1031–1037. 31. Rocque, M. and Posick, C. 2017. Paradigm shift or normal science? The future of (biosocial) criminology. Theoret. Criminol. 21(3): 288–303. 32. Jones, O.D., Marois, R., Farah, M.J., and Greely, H.T. 2013. Law and neuroscience. J. Neurosci. 33(45): 17624–17630. 33. Gajos, J.M., Fagan, A.A., and Beaver, K.M. 2016. Use of genetically informed evidence-based prevention science to understand and prevent crime and related behavioral disorders. Crim. Public Pol. 15(3): 683–701. 34. Roper v. Simmons, 543 U.S. 551 (2005).

2 Evolution, natural selection, and behavior

Introduction This chapter introduces biological concepts such as evolution, natural selection, and behavior that should first be understood in order to follow the approach to crime that we will develop in the text. We introduce natural selection and then compare the behaviors of humans and other animals. More specifically, we look at the questions of how natural selection or evolution works, how organisms adapt to their environments, and how some traits are selected over others and passed to the next generation. These traits include not only structural, biochemical, and physiological adaptations but also, and more importantly for our study, behavioral traits, which are equally adaptive. We also look at the nature-versus-nurture argument and show that there is really not a dichotomy but an interaction between these influences; neither works in isolation. We discuss how natural selection occurs and how behavior can be either entirely genetic or a mixture of genetics and the environment, the contribution of each varying with the behavior. We will then consider criminal behaviors and crime types and discuss how, in our ancestral past, they may have been advantageous. Finally, we will consider various evolutionary theories that seek to explain why such behaviors may have been adaptive.

Natural selection Natural selection is what drives evolution, and a brief explanation of it is given. You will soon see why a basic understanding of natural selection is important in this field. All living organisms adapt to their environments, and these adaptations lead to increased survival and reproductive success for organisms that possess them. Adapting does not  mean that a single organism can adapt—it cannot. Adaptation occurs over generations, usually over a great many generations. So, when we use the term adapt, we do not mean it in the casual sense of, say, a person adapting to a new culture or a new climate after immigrating. In that case, a person is getting accustomed or acclimated to something. For example, a population of rats in New York City may adapt over many generations and become resistant to a poison designed to kill them. The individual rat does not adapt, but some of its offspring are slightly more resistant to the poison than others, so perhaps two of 100 offspring will survive and reproduce. Something in their genetic makeup makes them more likely to survive than the others; perhaps they have slightly more of a body chemical that protects them from the toxin, or maybe they are fatter and can store the toxin better, or possibly they are more shy and do not go near new potential food sources. Whatever it is, something genetic makes these rats less likely to be poisoned. So, out of that particular family of rats, only two survive and reproduce. 23

24  Evolution, natural selection, and behavior

A trait such as avoiding new foods, having extra fat, or being shy must be genetically controlled in order for evolution to act on it. So, anything that is modified by evolution has a genetic component; otherwise, it would not work. Evolution selects individuals with genes that result in traits that help them survive and reproduce. They pass on this favorable trait. So, for example, if you are short, it might be genetic. Similarly, having brown hair is genetic. These are traits that you might pass on to your offspring. However, if you were to lose an arm in an accident, would your child be missing an arm? No, of course not. A trait is selected because it is favorable, but this relates entirely to the environment and the time frame. Something may be very favorable at certain times but not at all favorable at other times. For instance, in the example of the rats, let us say that the favorable trait was avoiding new types of food. When a poison is put out, this trait is definitely a benefit, because rats with this trait will not go near it. However, the trait could cause problems if the normal food supply dwindles and the rats have to find a new food supply. These unadventurous rats will die off. Another example is seen in the finches that Darwin observed in the Galapagos Islands.1 It was these finches and many other animals in these islands that led Darwin to first understand natural selection.2 In  the natural population, the finches have considerable variation in beak or bill size (Figure  2.1). Some of the birds have very long beaks and others have very wide beaks, but most have average beaks. In 1977, there was a prolonged drought that killed 70% of the finch population. It was found that finches with the extreme beaks, whether very long or very wide, had survived, whereas the intermediate birds had died out. So, the extremes were favored by natural selection. Why? Under normal conditions, many food sources are available, such as insects (out in the open or under bark that can be peeled off), large seeds, small seeds, and everything in between, as well as cactus fruits. There is a great range of food available for birds with average beaks, so this beak size is normally favored by natural selection. That is, birds with average beaks were the most successful, so they survived longer and produced more offspring, passing on the trait of an average beak size to more offspring. The  drought brought a new selective pressure; many food sources, such as small seeds and insects, were rapidly depleted because all the population could eat them. The  small seeds were

Figure 2.1  Finches with different-sized beaks adapt better to different environmental conditions. (From SFU Publications, Burnaby, Canada.)

Natural selection  25

not replaced, and the insect population was greatly reduced. There were only a few food sources left. Only the wide-beaked birds could crack the large, hard seeds and get at the insects under the bark of trees. The long-beaked birds could open up the cactus fruits to get at the seeds, so they also survived. Meanwhile, the birds with average bills starved.1 So, something that is a good trait may become unfavorable if circumstances change. Think about this in terms of a behavior that might have been advantageous once but is no longer advantageous because society has changed. Violence may be an example.

Types of adaptations So, all living organisms, including humans, are adapted to their environment. This  adaptation takes many forms. Structural modifications

The  modifications may be structural modifications, which enhance survival and reproduction. Examples include the development of complex organs such as eyes, ears, and wings. A more specific example is that of an orchid, in which the flower has become modified so that it resembles a female wasp. Male wasps attempt to mate with the flower and will then attempt to mate with another orchid, thereby pollinating the flowers.3 The wasp does eventually learn and will find a real wasp. What would happen to wasps that failed to learn? They would never successfully mate with a conspecific, so would not pass their genes to the next generation. Many examples are seen in insects; for example, flies in the order Diptera have modified the second set of wings into a smaller club-shaped balance organ that allows them to hover and turn very accurately in flight, and fleas in the order Siphonaptera have completely eliminated wings, as they would be disadvantageous in their habitat—a mammal’s dense coat. Some bird species have evolved strategies to allow them to quickly identify and pounce on a potential food source, and many insects have evolved strategies to avoid such predation by developing structures that focus a predator’s attack on a less significant body part. Many butterflies have bull’s-eye patterns on the tips of their wings, which simulate eyes. This not only suggests that the tiny, delicate butterfly is more dangerous than it really is but also deceives the predator into striking this area, thinking it is the head. Instead, it merely strikes the outer edge of the wing, allowing the butterfly to escape. The damage to the wing is very minor in comparison with an attack to the head or body. All these modifications increase the organism’s survival and reproductive success. Biochemical pathways

The modifications may be changes in biochemical pathways, such as the development of metabolism, photosynthesis (plants’ ability to convert sunlight energy to food), and respiration, which increase survival and reproductive success. For example, the pine beetle, the forest pest, has shown metabolic adaptations that allow it to survive in colder temperatures, allowing it to spread north and east.4 Behavioral adaptations

The modifications may be behavioral adaptations that increase survival and reproductive success, such as learning to avoid predators or developing courtship dances that enable individuals to find strong, suitable mates. Some caterpillars not only develop patterns on their furthermost segments to look like a head, such as the butterflies mentioned above, but also show behavioral adaptations by making striking moves to reinforce the suggestion that this is the head end and it is dangerous. Some birds also exhibit behavioral counterstrategies, such as females that pretend to be wounded and therefore easy prey to lure predators away from their nest. A species of harmless fruit fly has developed structural modifications so that it looks like a jumping spider and behavioral adaptations so that it actually assumes the posture of a jumping spider.5

26  Evolution, natural selection, and behavior

Conditions required for natural selection to act The process that results in these adaptations is called natural selection, which is the differential ­survival and reproduction of individuals within a population. That is, some individuals in a population have a better chance of living to reproductive age and having offspring that will carry the same traits that helped their parents survive and reproduce. Thus, adaptations are behavioral, structural, physiological, or biochemical pathway changes that enhance the survival and reproductive success of an organism. Remember that natural selection acts on a population, not on an individual. There are three conditions necessary and sufficient for natural selection to occur.2 1. Variation

There must be variation among individuals. Many variations of different traits are obvious in any population. The variations can be structural, behavioral, physiological, and/or adaptations in the biochemical pathways. One can see lots of variation in our own species. Genetic variation is critical for natural selection to act. Genetic variation could be in, for example, strength, attractiveness, physical dexterity, intelligence, or longevity.6 2. Heritability

The variation must be heritable; that is, it must be under genetic control so that it can be passed on to the next generation. Natural selection only acts on the phenotype, that is, the way the genetic information is expressed—for example, the way a person behaves or looks. This will become clearer when we discuss basic genetics and phenotype and genotype. 3.  Fitness consequences

The individual must differ in its ability to survive or reproduce depending on this trait. That is, the trait must give the individual a better chance of surviving or help it produce more offspring. In the above examples, increased strength, attractiveness, physical dexterity, intelligence, and longevity would impact survival and the ability to reproduce.6 If all three conditions are met, certain individuals will leave more offspring because they are better able to survive and reproduce, and the trait will become increasingly represented in s­ ubsequent populations. Eventually, a change in the population occurs that makes it better adapted to the environment. One of the most famous examples of natural selection is the English peppered moth.7 There are two varieties of this moth. One is light colored, with splotches of dark pigment, hence the name “­peppered.” The other variety is dark all over. So here, we have variation in the population, and it is inherited. Peppered moths feed at night and rest during the day, and they like to rest on rocks and trees that are encrusted with light-colored lichens. In this situation, the light moths are well camouflaged, but the dark moths are easily spotted by their main predators, birds (Figure 2.2). Before the industrial revolution in England in the late eighteenth century, dark peppered moths were rare, presumably because they became bird food very quickly and rarely got a chance to pass on their genes to the next generation. So, one can see that this trait for color, which varies and is heritable, also has fitness consequences: because of this trait, the individual has a better or worse chance of surviving and therefore of living long enough to reproduce. In this situation, being a light-colored moth was definitely an advantage. However, the industrial revolution brought heavy pollution, and in the late 1800s, most of the lichens died off, leaving the general background of tree bark, which is much darker. The environment had changed, and the dark moth was now at an advantage, because it was well camouflaged against the dark tree bark and the light-colored moths were easier to see (Figure 2.3). After a while, there were many more dark moths around than light ones, because the light ones were more often eaten before they could reproduce. This phenomenon also occurred in many other moth species. The happy ending to this story is that now we have realized what pollution is doing to

Natural selection  27

Figure 2.2  The dark English peppered moths are easier to spot on trees covered in light-colored lichens. (From SFU Publications, Burnaby, Canada.)

Figure 2.3  The dark moths are better camouflaged and therefore selected for when the tree bark is dark colored. (From SFU Publications, Burnaby, Canada.)

28  Evolution, natural selection, and behavior

us, apart from affecting moth populations. England has made great efforts to reduce atmospheric pollution. The lichens have returned, and the moth population is once again primarily light colored. These examples demonstrate that for natural selection to act, these three characteristics must be present. There must be variation in the particular trait, the variation must be heritable, and it must affect the ability of the organism to survive and reproduce. However, there are also some constraints on the ability of natural selection to produce adaptation of organisms to their environment.

Constraints on natural selection There are two major constraints on natural selection: adaptations are often compromises, and natural selection can only act on existing traits. 1.  Adaptations are often compromises

Any organism must do many things, and an organism that develops a new structure or behavior carries a cost, because it requires energy to develop and carry out the change, and it may also reduce the ability of the organism to do something else. For example, male peacocks have very long tails to attract females. They can use their tails to give a very attractive display, as you have no doubt seen. There is a very heavy cost associated with this tail. It is “expensive” for the bird to construct and maintain, and dragging this heavy tail around greatly reduces its ability to get away from a predator. However, it also greatly increases the bird’s reproductive success, so the cost is more than made up for. Therefore, the beautiful tail is selected because peacocks with lovely tails attract more females and produce more offspring. Humans frequently enter into this process and produce unnatural selection or artificial selection. For example, we have made pets from many animals that were originally wild. Wolves were artificially selected by humans to make their descendants into dogs. Humans chose the friendliest wolf cubs and bred them to other friendly cubs to eventually produce early dogs. They then selected specific traits within the dogs to produce dogs with different functions, such as fighting, hunting, and shepherding. If pet dogs go back to the wild, natural selection takes over after a few generations and produces a wolf-/shepherd-type dog, which is obviously the best type to survive in the wild. However, because we want pet dogs and can provide things that the wild environment does not, such as protection and food, a large variety of dog breeds survive in our society just fine, usually on someone’s bed. Another classic example of humanity’s interference with natural selection is the racehorse. A wild horse is strong and stocky, with powerful legs. We have taken wild horses and bred them, originally to carry us and pack our belongings but more recently for specific characteristics such as speed. We bred the fastest individuals with the longest legs together to create today’s thoroughbreds. They are very pretty and very fast, and now, we race them and gamble on the outcome. However, they are only very fast as long as we provide flat, well-surfaced tracks for them to run on, in which case they go much faster than a wild horse ever could—over short distances. The quarter horse, which is much stockier than a thoroughbred, was bred especially to run races over a quarter mile, hence the name. These horses do not have the endurance of the wild horse, and if they are allowed to run on anything but a track, they will break their legs at the first molehill. Even on the track they frequently break their legs. Such an animal would never arise by natural selection because the cost of speed is too high: light legs that break easily. If racehorses were to be released into the wild, the ones with the strongest legs would be naturally selected for, and those with very fine legs would become food for predators early on. The population would gradually revert to the design that nature, or natural selection, intended. 2.  Natural selection can only work on existing traits

Evolution does not totally scrap the old biochemical pathways, structures, behaviors, and so forth, and start anew. It modifies existing traits as the environment or conditions change; that is, it does not rebuild from scratch.

Behavior in humans and other animals  29

Behavior in humans and other animals Everyone has heard the argument about nature versus nurture, which means “does the biological background dominate or does the environment dominate?” Although this argument has a long history, there has never been a real dichotomy. Anyone who works in biology knows that it is not a dichotomy but an interaction, nature and nurture. We are a product of what we are biologically and of the environment to which we have been exposed over the course of our lives. Our society seems to have rejected the middle ground, where biology and environment are seen as integrally linked, arguing that it has to be all one or the other. Nothing is ever likely to be all one thing or another; it is usually a mixture of many things. If everything were based on environment, we could say that someone who has had a terrible upbringing—was sexually and physically abused and grew up in poverty—would definitely end up being a criminal, but this is certainly not the case. Many contributing members of society come from just such backgrounds; conversely, many criminals come from excellent homes. Assuming that it is all environment or all biology is simply wrong, and no scientist accepts that argument. I do not think most social scientists do that either. As we understand more and more about the brain, we have learned that the environment begins to shape the nervous system before we are born. Nature and nurture (biology and environment) affect all parts of our lives, but because this text is looking at criminal behavior, we should define what behavior actually is. Behavior is simply what an animal does and how it does it; in this case, we are talking about the human animal. The  study of behavior looks at both the action itself and its motor components, that is, the brain function and how the animal actually performs the behavior. Behavior is adaptive in just the same way that a structure or a metabolic pathway can be adaptive. Studying behavior is probably the oldest form of biology. For early humans, knowledge of animal behavior was essential for survival. They had to learn about the habits of the animals around them to catch them for food and also to avoid predators. This still occurs today. Recently, a small village in Africa was exposed to severe leopard predation, with several people killed in leopard attacks. One person in the village pointed out that people were only attacked from behind—the leopards never seemed to attack a person from the front. Therefore, the villagers made masks of human faces, which they wore on the backs of their heads, so that whichever way they were facing, it appeared that they were looking at you. It worked. No one was ever attacked while wearing a mask. By ­studying leopard behavior and learning from it, the villagers learned to protect themselves and survive, probably in the same way that humans and other species have always done. The  study of animal behavior increased fitness because it increased survival and thus the person’s chance of reproduction. No doubt, the leopards had also learned from observing, and from attempted attacks, that the easiest and safest way to attack a human was from behind. Natural selection adapts an organism to its environment. It acts on everything we inherit from our parents. Mostly, we see it acting on specific things, such as the shape of a bird’s beak, but it acts equally well to produce behaviors that increase an animal’s fitness within an environment and to produce structures and biochemical pathways that also increase fitness. An animal must: ■■ ■■ ■■

Feed in a way that maximizes fitness Reduce the risk of predation, that is, of being eaten Increase its chance of mating

Feeding behavior is likely to optimize efficiency and health, and the mate chosen tends to m ­ aximize the number of healthy offspring produced and thus the contribution to the next generation. Natural selection will favor a behavior that increases fitness and reproductive success, but the behavior must have a heritable component. If there were no genetic basis, natural selection could not act on behaviors and they could not evolve. Behavior must have a genetic basis, because behaviors do evolve. Even learned behaviors typically depend on genes that create a neural system receptive to learning.

30  Evolution, natural selection, and behavior

Figure  2.4  Genetic and environmental components of behavior in lovebirds. (a) Nests made with long strips—no tucking behavior. (b) Nests made with short strips—tucking behavior. (c) Hybrid nests made with intermediate-length strips. (d) In later seasons, only head-turning behavior. (Reprinted from Campbell, N.A., and Reece, J.B., Biology, 6th ed., Pearson Education, 2002. With permission.)

There  are many different types of behavior and some are discussed here. However, first an a­ musing example that illustrates how a behavior is governed by genes.8,9 Two species of African parrot have different nesting behaviors. One is the Fischer’s lovebird, which has a red head, and the other is the peach-faced lovebird with, predictably, a peach-colored face (Figure 2.4). The females of both species make nests with strips of vegetation or, in captivity, with paper that they cut with their beaks. The female Fischer’s lovebird cuts long strips and carries them back to the nest one at a time in her beak (Figure 2.4a). The peach-faced lovebird has a different tactic. She cuts shorter strips and carries several at a time by tucking them into the feathers of the lower back (Figure 2.4b). This maneuver is quite clever because it involves tucking the strips in firmly enough to stay and then smoothing the feathers over the strips. These two species are so closely related that they can be interbred in experiments. The resultant hybrid females show an intermediate kind of nest-material gathering. The strips themselves were intermediate in length, but the most interesting part is the way that they carried them. They usually tried to tuck them in the tail feathers, but they sometimes forgot to let go when they had the strips positioned, so they kept ripping them out

Behavior in humans and other animals  31

Figure  2.5  Behaviors may have differing amounts of genetic and environmental input. (From SFU Publications.)

again; others stuffed them improperly and they fell out (Figure 2.4c). These results show that there must be an inherited component. The baby birds did not learn to handle the strips in an intermediate manner; they had an inherited ability to carry strips, but it was confused by the mixture of genes from the parents. The result was that these birds totally failed to transport strips. In the end, the birds learned to transport the strips in their beaks (Figure 2.4d). But even so, they still made token tucking attempts. After several years, these same birds still turned their heads to the rear before flying off with a strip. This experiment shows that the observed differences in behavior are based on different genetic backgrounds or genotypes and that they are a result of genes controlling these actions. However, these innate, inherited traits can be modified by experience or the environment, and so, the hybrid birds eventually learned how to transport the strips. The observable end result is an interaction between the genes and the environment. This is the point where the nature-versus-nurture debate is inevitably raised. The original argument about nature and nurture began a long time ago, when scientists in Europe and America were both studying behavior but looking at different sides of the same coin.1 In fact, genetics and the environment should be regarded as a continuum. All behaviors have some genetic contribution and some environmental contribution, and the division between the two groups of scientists was not as cut and dried as it seemed. Very few actually believed that behavior was either all genetically controlled or all environmentally controlled. The debate was really more about which input is most important. The amount of input from each varies with different behaviors, but nearly all have components of both, as Figure 2.5 shows. At the side of 1, the behavior is governed almost entirely by the genes, at the side of 2, it is mostly the environment, and then, there is a whole range in between. This  is still not  strictly true (as you will see as we proceed), as in most cases there is a third ­component, an interaction between the genes and the environment. Behaviors can be divided into two main groups: ■■ ■■

Innate behaviors, which are governed entirely by the genes. Learned behaviors, which require experience. They  still have a genetic component, but it is strongly affected by environment and experience.

Innate behaviors All animals, including humans, carry out many behaviors without ever having seen them ­performed. As they have never seen them, they cannot have learned them, so the behaviors did not  come from the environment. Such behaviors are innate, or genetically programmed. These behaviors tend to be constant and are performed in exactly the same way by every individual of that species. The usual term for these innate behaviors is fixed action patterns (FAPs). They can be performed in a complete, stereotypical fashion by an animal without prior experience and without the animal understanding the significance of the actions. When and where do innate behaviors occur?

When would you expect to see such innate behaviors—in what animals and under what conditions?

32  Evolution, natural selection, and behavior

In animals with no opportunity to learn

Many species of animals have very short life spans, for example, the mayfly, which has just one day to mate, lay eggs, and die. If its survival and reproduction depended on learning something, it would have very low reproductive success. Moreover, many species of animal have no or very low parental care and so cannot learn from their parents. Critical to get it right the first time

In animals that do have an opportunity to learn, we still see innate behavior, usually when it is critical that the behavior is performed correctly the first time. For example: ■■ ■■

A baby bird’s first flight. Birds can fly with no previous practice. A moth avoiding a bat. If it does not carry out the proper behavior the first time, it will not get a second chance.

Types of signs or stimuli that trigger innate behaviors

These kinds of innate behavior are often related to predation, reproduction, or other very important activities. Innate behavior is often triggered by a stimulus of some sort, usually a sign. These tend to be very simple signs that will reliably occur under conditions that should lead to the appropriate response. Visual

One of the best known triggers is a visual sign of some sort. For  example, a male stickleback fish builds a nest, herds in a female to lay her eggs, fertilizes the eggs, and then shoos her out. The male cares for the young. Males will fiercely guard their nests and show aggressive behavior toward other male sticklebacks. Niko Tinbergen, a scientist, showed several models to male sticklebacks (Figure 2.6). Some looked very much like a pencil drawing of a male fish, and others looked like line drawings of circles or squares, but the bottom half of the simple diagrams was painted bright red, whereas the realistic drawing of the stickleback was in black and white.4 The one that looks like a fish obviously looks the most like another male stickleback to us, but the fish ignored it. They did, however, react to the other models that look nothing like male sticklebacks to us. They appeared to be male sticklebacks to the fish because of the red underbelly; it is this red visual sign that is the stimulus for aggressive behavior, and not the shape of the fish. A researcher first reported this behavior when he realized that his tank of sticklebacks got extremely upset whenever a

Figure 2.6  A male stickleback and several models. (From SFU Publications, Burnaby, Canada.)

Behavior in humans and other animals  33

red truck drove by. So, the fish does not have to recognize the shape and details of another fish, just the red underbelly. As the only thing in the stickleback’s natural environment that has a red lower half is a male stickleback, the simple red color is all they need to recognize an enemy. Auditory

The stimulus may be auditory. Moths have excellent hearing, to detect the ultrasonic sounds that bats use in their echolocation of prey. When the moths hear this sound, they immediately close their wings and go into a power dive—automatically—to avoid the bat. In  turkeys, the sign for parents to be protective of their young is cheeping.1 If a mother hears her babies cheeping, she will protect them, but a mother will ignore her offspring if it is placed under a glass dome. The parent can see the chick but cannot hear it, so she ignores it. A deaf mother will kill her offspring because she does not receive the “look after me” stimulus. Olfactory

The stimulus can be olfactory (a smell). Marine snails use olfaction to detect the presence of predatory sea stars and can then take immediate action. Temperature

The stimulus may be temperature. Female mosquitoes detect their hosts by heat (as well as other stimuli). Combination of signs

The stimulus can be a combination of simple signs, such as bird feeding. Adult birds feed young when they open their mouths and gape at the parent while cheeping loudly. The gaping mouth and cheeping are the signs for the parent to feed the young. The chicks’ stimulus to begin this behavior is the landing of the parent on the nest.1 Many animals behave in this automatic way. Humans respond much more to an entire situation than most other animals, integrating more information when making a choice, but we also have FAPs. Babies grasp strongly with their hands in response to touch and will smile at a drawing of a face, even if it is just a circle with two black circles inside. Smiling helps make them attractive, so that the parent will protect them. In  all these cases, natural selection has chosen the most obvious sign or stimulus to get the desired result. Experimenters often test animals with similar types of signs—like the unrealistic models for the sticklebacks—and sometimes, the animal will respond to the wrong stimulus. However, in real life, the male stickleback would do just fine attacking anything with a red underbelly, because it is almost invariably going to be a male stickleback in the real situation. If other objects with this trait occurred naturally, natural selection would select a more discriminating mechanism. For example, greylag geese assume that anything small near their nests is an egg that they have accidentally kicked out, as they are inclined to clumsiness.1 They roll anything small back into the nest, including golf balls and even toy dogs placed by experimenters. However, in real life, the only object likely to be just outside the nest is the egg, so the goose does not have to be more precise. On the other hand, other birds must be able to identify their own eggs. Cuckoos lay their eggs in other birds’ nests, and if the host bird does not realize it, the baby cuckoo will destroy the other eggs or baby birds and will be raised by its new foster parent. It is obviously very important to reproductive success for the host bird to be able to recognize intruder eggs.

Learned behaviors Learning is defined as the modification of behavior by experience. Learning can act on a number of different components of behavior, including the stimulus response (usually it narrows the range of effective stimuli), so the animal can learn which is the most important and become more precise. It also acts on motor patterns; for example, a squirrel gnawing on a nut soon learns to gnaw with the grain because that

34  Evolution, natural selection, and behavior

makes it easier to open the nut. Even learned behavior has a genetic basis. For example, rats are intelligent and can learn quickly that if they press a bar, they get food. You can teach them not to press the bar by giving their feet a tiny electric shock when they touch it. If they get the shock immediately after they touch the bar, they will learn not to touch it, but if you wait 30 minutes, they will not learn. This is similar to the time lag seen in training a puppy. Anyone who has ever toilet trained a puppy knows that if you catch it in the act of urinating on the carpet and yell at it, the puppy may connect the yelling with the act and desist, but if there is any time lag, the puppy will not understand why you are yelling and will just think you a grumpy. They will no longer associate your response with their actions as they have passed. However, this varies with the behavior. A rat that cannot learn to touch something if it gets a shock 30 minutes later can still learn about other things over a much longer period of time, for example, how to avoid poison. Rats live in colonies, and when a new food is introduced, they do not eat it all at once. They will send out a couple of food-tasters to try a little bit, and then, they will observe them for 24 hours. If the testers get sick after eating the food, the rest of the colony will learn the smell of the food and avoid it. This is the reason that rats are so difficult to poison. The difference in learning about the reason for shock and the new food shows that there is an innate, or built-in, predisposition to learn about some things but not about others. Learning versus maturation

There are many different forms of learning and a few are discussed here, but first it is important to be able to define the difference between learning and maturation. Some behaviors that are clearly innate are performed better or faster as time goes on. This is often not the result of learning but of the maturing process, for example, the development of a more coordinated muscular system for flying. A baby bird does it right the first time—it has to—but its flying improves as it matures and becomes physically stronger. Habituation

Habituation is a very simple type of learning that involves a loss of responsiveness to unimportant stimuli. There are many examples; gray squirrels react to alarm calls by members of their group, but they will stop reacting if these calls are not followed up by an actual attack—it is just like the old “cry wolf” story.1 Imprinting

Imprinting can be illustrated by a famous old study. You have, no doubt, seen a group of ducklings or goslings waddling after their mother. Natural selection would select this behavior because the mother will help to protect them, will lead them to food, and so on. Konrad Lorenz took some greylag goose eggs and raised some of them with the appropriate mother and the rest in an incubator. He was the first living thing that the goslings saw, and he spent his first few hours with them. After that, they assumed that he was their mother and followed him everywhere. When they grew up, they preferred to associate with people, not geese, and even tried to mate with people. Apparently, they had no idea of “I am a goose, you are a goose” or of “mother.” Instead, they imprinted on the first object they encountered. The most important imprinting stimulus was movement away from the goslings, preferably making a noise at the same time. In fact, they would even imprint on a box with a clock in it.4 Another bird, a whooping crane called Tex, was hand-raised in a zoo on her own, and when she was an adult, she refused to mate with birds chosen for her. She was imprinted on humans, so she had to be induced to lay eggs (which could be artificially fertilized) by “dancing” with humans (she preferred Caucasian men of average size with dark hair). Another example is salmon returning to their spawning ground. They manage to find exactly the same place because of olfactory imprinting when they were young. They can find the precise combination of smells that is “home.” Imprinting, however, can only occur during a certain critical period; with the geese, it was the first two days of life.1,4 Classical conditioning

Classical conditioning was made famous by Pavlov’s experiment with dogs in 1900. He sprayed powdered meat into their mouths, which made them salivate. Just before he sprayed them, he rang a bell. Finally, the dogs salivated whenever he rang the bell. They were conditioned to expect the meat spray.4

Behavior in humans and other animals  35

Operant conditioning

Operant conditioning is better called trial and error. Here, an animal learns to associate one of its behaviors with a reward or a punishment.4 Many well-known experiments have been done by using animals in cages in which pushing a lever meant that they received a food treat. They learned to push the lever quickly. We train dogs in much the same way. Positive reinforcement (that is, a reward for an action) works much more effectively than negative reinforcement or punishment. A rat learns more rapidly to push a lever for food than it learns not to push it if it gets an electric shock. This is the reason the best dogs are taught by encouragement rather than by abuse. One of the most famous examples of conditioning involved certain birds in England, very much like chickadees, that learned to peck through the foil tops on milk bottles to get a free breakfast. In England at the time, milk was still delivered to the door in glass bottles. By trial and error, these birds found that if you sat on the top of the milk bottle and pecked, you could get free milk. Observational learning

Observational learning occurs when an animal learns from watching another animal doing something.1 Birds watching those pecking open the foil on the milk bottles quickly learned to open milk bottles too, and the habit spread throughout the country in one summer, much to the annoyance of people who objected to sharing their milk. People do the same thing all the time; we see someone do something in a new way and try it ourselves. Play

Play is common among young animals. If you watch puppies or kittens jump and wrestle and grab at each other, you will notice that no one gets hurt. Play is costly in terms of energy, and it is also costly because it often makes young animals vulnerable to predators, as they are not watching out for them. However, it is thought to be beneficial because it provides practice for later attacks on rivals or prey, and most importantly, it is exercise that keeps the animal fit.1 Insight learning

Insight learning occurs when an animal can perform an appropriate behavior the first time it is in a situation in which it has no previous experience.4 This behavior is also called reasoning. For example, when a banana is hung high above a chimpanzee’s head and the chimp is provided with a pile of boxes, the chimp can reason that if she piles the boxes on top of one another, she can reach the banana. This behavior is most developed in mammals, with humans showing the most advanced levels of insight learning. All this demonstrates that behavior has both genetic and environmental components. Some of the behaviors are very simple and some are completely under genetic control. Remember that we are animals too; we have a slightly bigger brain, but we are still animals. We reason things out more than most animals, but we have many behaviors that are entirely genetic or automatic, such as the baby gripping a finger or smiling at a human face. But there are other entrenched human behaviors that are not learned. We can certainly improve them with learning, but the original behavior is pure biology. One of the best examples is the sex drive; it is automatic and kicks in after puberty, once all the hormones are in place. It is definitely a behavior, and it is not learned; it is automatic and therefore genetic. People can have sex with no instruction. Just think of people of certain classes in the Victorian age who never mentioned sex, so most women did not know it existed until their wedding night. People who have been raised without any instruction still manage to have sex successfully, so it is obviously mostly controlled by the genes that control our hormones. Hormones are heavily involved in sexual attraction and sexual acts and therefore the sex drive. The sex drive comes from the limbic system of the brain and arises directly from the sex hormones. Learning certainly improves the timing and manner of expression of the sex drive, but the drive itself is completely biological.

36  Evolution, natural selection, and behavior

The effect that the environment has on the behavior we observe varies dramatically. It has no effect on some behaviors and a major effect on others. In fact, most behavior is genetically predetermined to be affected by the environment; it is designed that way, to be improved and changed by learning from experience. Remember that almost everything has a component of both the environment and the genes— even things that we consider entirely genetic, for instance, eye color. But is it really completely under genetic control? Yes, mostly, but the environment can still have an effect. A rescued dog was found to be very friendly, but his eyes were the pale amber of a wolf, which made him look as if he was about to attack. We later discovered that this dog had been shut in a dark shed for the first 8 months of his life. He had not been exposed to daylight, so the color in his eyes had not developed properly. Such things also happen with people. Now, the dog’s eyes are darkening, as he lives a normal life. Identical twins are genetically identical and so should look the same. However, the environment, again, can influence this. One member of a pair of twins may be considerably shorter that the other member due to a childhood disease that stunted that twin’s growth, overriding genetic destiny. In both cases, the environment affected the original genetic plan. DNA is not always destiny; the environment also has an impact. Sometimes, the environment has little effect on the basic genetic information, such as with eye color, and sometimes, it has a tremendous impact, as with many behaviors.

Natural selection and behavior We have shown that behavior is at least partially under genetic control. It is also influenced by the environment, and the contribution of the genes ranges from complete control, as in innate behavior, to a varying level of control, as in learned behavior. As behavior does have a genetic component, it can and will evolve. However, traits such as certain behaviors will only be selected for by natural selection if they contribute to the survival and reproductive fitness of the individual who exhibits such behavior. Reproductive success is measured by an individual’s genetic contribution to future generations. It is clear how behaviors such as a baby smiling at a caregiver to increase affection and therefore care, or imprinting on a caregiver, are likely to increase survival and eventually increase the chance of that individual reaching adulthood and reproducing. However, it is more difficult to relate this to a criminal behavior. Daly and Wilson, eloquent Canadian evolutionary psychologists, are leaders in the area of violent crime and evolution. They  have said that evolutionary psychologists frequently make the error of assuming that a behavioral trait will be selected for if it results in happiness, selfactualization, or homeostasis, whereas in fact, evolutionary biology states that behaviors will be selected to maximize Darwinian fitness, in other words, to increase survival and reproduction.10 This  reminds us that a trait will survive if it results in increased offspring and if the benefits outweigh the costs, even if by today’s standards the means of procuring this reproductive benefit are abhorrent. To consider criminal behavior in terms of reproductive success, we need to consider various reproductive strategies. In mammals, female animals contribute a tremendous amount of parental investment to their young, particularly during gestation, lactation, and early child-rearing, often at considerable risk to themselves. This  helps the females ensure that their offspring, carrying their genes, will survive to reproduce, themselves. This early investment in the offspring is almost entirely provided by the female. Females have a limited reproductive potential, as they can only give birth to a finite number of offspring, whereas males can sire infinite numbers of offspring. Therefore, female commitment to the offspring is much higher than that of males. This has resulted in females who are very choosy about their mates. In  many species of mammals, the female will only mate with a male that first offers her resources such as food11 or with a male that has already secured and defended a desirable territory that offers resources.12 These are evolved reproductive strategies that follow Bateman’s principle, which states that whichever gender makes the most parental investment will be the most careful

Natural selection and behavior  37

in choosing a mate.13 The rule indicates that females who are cautious about choosing a mate will probably rear more offspring than their less picky sisters. In  so doing, these cautious females will not only contribute more of their genes to the next generation but will also help males who are good providers to contribute their genes to future generations. This would suggest that both males and females who contribute more to their offspring in terms of care and resources will be selected for, and conversely, those males who do not provide support will be selected against. If so, this would result in a reduction in the number of males who do not support females and their offspring. Evolution is more interesting than this, however, as not only does it result in the development of such a reproductive strategy, but it also results in the development of counterstrategies that evolve to circumvent the original strategy, such as rape, or pretending to provide resources for future offspring then reneging after mating.14 Although, in the past, most criminologists shied away from the concept that genetics has any influence on criminality, many scientists and social scientists have provided a great deal of evidence that proves a strong relationship, as we will see in subsequent chapters. Classic Darwinian evolution has been used to explain very specific types of crime, such as assault, murder, sexual assault and rape, child abuse, domestic violence, genocide and terrorism, and prostitution.15 Here, we will just briefly consider potential evolutionary explanations for a variety of crimes. In order for a trait to be selected, it must confer an advantage, and that advantage must inevitably result in increased reproductive fitness or, in other words, more offspring to carry on that trait. If we look away from human society, with its rules and social norms, for a minute and just think of humans as animals (which of course, we are), then we can start to consider—from an evolutionary point of view only—how some crimes might be advantageous.

Aggression Let us first consider aggression, an act that is often linked to criminal activity and can be a crime in itself. From an evolutionary perspective, aggression is an adaptive behavior; that is, it is a behavior that helps the organism survive. Remember that survival and reproduction are the two most important things in life for any organism. Aggression is a behavior in which physical or verbal force is used to counter a perceived threat. It has escalated into a maladaptive behavior, violence, which is aggression directed against the wrong target in the wrong place at the wrong time with the wrong intensity. Work on animals has frequently shown that we can genetically select for aggression. In a particularly nice series of experiments on mice, Gariépy and colleagues demonstrated not only that aggression could be selected for but also how the environment could reverse the aggression (summarized in Gariépy et al.16). The fact that we can select for aggression means that it has a genetic component, because selection only works on heritable traits. In these experiments, male mice that frequently attacked other mice were identified and mated with females with male siblings had also shown high levels of attack. The  same procedures were duplicated over many generations, with only the most aggressive being chosen for breeding. This line of mice was termed aggressive. At the same time, the least aggressive male mice were chosen and mated to females with male siblings that were non-aggressive. This line was termed low-aggression. Mice that had not been selectively bred were used as controls.16 In each generation, the male mice were tested for aggression as adults by exposing them to a control animal and observing their reaction. The mouse lines were observed for aggression over 22 generations, and the researchers found that they could rapidly select for both aggression and non-aggression over just a few generations, with clear differences seen between the two groups within one generation and a significant statistical difference observed by the fourth generation.16 Interestingly, once aggression had been selected for, the level of aggression did not increase any further, although the reduction in natural aggression in the non-aggressive line continued to decrease to close to zero, and the animals were seen to frequently “freeze” in the presence of another male mouse. The very rapid selection for both aggression and non-aggression indicates that the aggression was highly heritable.

38  Evolution, natural selection, and behavior

In  further experiments, the researchers looked at the effects of simple manipulations on this genetically selected aggression and found that the response was extremely plastic. That is, although the aggression was genetically determined, it was designed to be flexible, so that it could adapt to changing conditions.16 For example, simple socialization with other mice completely reversed the genetic selection for aggression in the aggressive mice and for freezing in the non-aggressive mice, indicating that although the original trait is genetic, it can be completely reversed by an appropriate environment. Many other tests were performed to determine the many ways in which this genetic effect could be neutralized. Therefore aggression, at least in the mice in this very robust set of experiments, was heritable. There were several replications of the experiments by people all over the world. The most interesting part was all the methods they could use to change the aggression or prevent it entirely. These are called protective factors. This was one of the studies that actually tried to identify some of the protective factors. Simply socializing the animals was a major protective factor.16 What about other animals, such as dogs? When humans first developed dog breeds, they bred deliberately for certain traits. These included physical looks, coat length and color, and temperament. Guard dogs were bred for aggression, sporting dogs for their retrieval abilities, and scent hounds for their nose. The  most successful scent hounds were bred to other successful scent hounds, to attempt to produce puppies that would grow up to be excellent hunting dogs. In the same way, aggression was also selected for. If the breeder wanted war dogs, then aggressive dogs were deliberately bred to aggressive dogs to increase the chance that the puppies would be aggressive. In more recent years, we have bred most dogs to be pets and companions, so we do not wish to develop aggressive dogs. However, temperament still differs greatly between dog breeds, based, in most cases, on what the dogs were originally bred for. Today, responsible dog breeders breed for good temperament. A responsible breeder would never breed an aggressive dog. A well-bred dog might still be aggressive and vice versa, but certainly the chances of aggression are reduced. So, we recognize that we can breed for selected traits in animals. If we can selectively breed for them, the traits are genetic, and these traits include behavior. It is difficult to suggest what might be construed as criminal behavior in an animal, but we can certainly define and see aggressive behavior. So, most of us will accept that certain behavioral traits, including aggression, are affected by breeding—in other words, by genetics—in animals. Just because the animal is genetically predisposed to be potentially aggressive does not  mean that it will be so. For  example, we have all met dogs whose breed may once have been bred for aggression but who, through good socialization and training, are extremely well tempered. Therefore, despite the fact that the animal may be predisposed to be aggressive, the environment in which it was raised has influenced its behavior. In the same way, a dog that is not predisposed to violence but that experiences abuse or a very harsh environment may become aggressive. So again, the effect of the environment has a strong influence.

Natural selection and crime Let us now  consider different crime types and how they might be evolutionarily advantageous. These particularly relate to our ancestral past rather than to the present day and must be considered in terms of evolutionary advantage rather than morals.

1.  Theft and robbery This  is fairly simple to understand. In  order to survive and reproduce, an animal requires resources. Resources improve survival and also make a mate much more attractive to another mate; the more resources a person has, the more desirable that person is likely to be and so have greater choices in a mate. The  evolutionary expropriative theory suggests that criminality is just a way of acquiring resources but is different from normal methods of acquisition because it

Natural selection and crime  39

results in victims who attempt to suppress such theft.17–19 This theory suggests that there are two ways in which to acquire resources: ■■

■■

Generative resource acquisition, which involves traditional methods of growing and raising food, building homes, and educating others in similar vocations Expropriative resource acquisition, which involves taking advantage of those who generated the resources, but also involves more risk14

2.  Assault and murder In  today’s world, violence is not  usually advantageous; that is, it does not  increase survival and reproductive fitness. In  fact, it usually results in the opposite. Violent criminal offenders are at an evolutionary disadvantage, as they are at increased risk of an early death or lengthy periods of incarceration, which curtails or limits their reproductive opportunities. They  also usually have reduced social status, do not gain resources, and often suffer from myriad health issues.6,20 However, in our ancestral past, without today’s criminal justice systems and morals, it is easy to see how assault and murder could lead to increased resources, the removal of a rival for a desired mate, or a desirable territory. A violent person would be more likely to improve social status by becoming a leader, which would also lead to increased resources and availability of desirable mates. We are inclined to think that our world today is much more violent than our bucolic vision of the past. However, there is a great deal of archaeological and historical evidence that life in the distant past was just as violent or considerably more violent than it is today.21 Therefore, alleles that increased violent behavior would have been selected for in our ancestral past, and although disadvantageous today, and therefore less common, would still exist.6 Many evolutionary psychologists have suggested that homicide occurred as almost a side effect of interpersonal conflict rather than an actual adaptation, because while extreme violence may have had inclusive fitness benefits, homicide was likely to have been disadvantageous, as it might have resulted in retribution from kin that could be fatal. 22,23 Duntley and Buss suggested that only some homicides could be considered byproducts and proposed the homicide adaptation theory, in which most homicides were a result of specific selective adaptations for homicide.24 They  argued that in very specific and recurring situations, the fitness advantages of homicide would outweigh the costs. They list a number of possible circumstances in which a lethal solution would incur fitness benefits, such as self-defense and protecting kin—in particular, biological children—from assault, which clearly involve fitness benefits. This leads to other contexts, such as killing genetically unrelated family members—for example, stepchildren—to avoid investing resources in non-kin; or killing genetically related individuals who are unlikely to survive to reproduce, again to prevent wasted resources; or even killing to remove a reproductive rival or to gain resources to attract more mates. This adaptation, they argued, has also resulted in selection for mechanisms to defend against homicide.24 Infanticide

Although men disproportionately kill more frequently than women,25,26 certain types of homicide, such as infanticide, are much more common in women. Duntley and Buss argued that infanticide, in certain circumstances, would have increased a woman’s fitness. As natural selection selects for traits that increase reproductive fitness, killing one’s own child seems counterproductive. However, Duntley and Buss suggested several contexts in which murdering her own child would actually increase a woman’s reproductive fitness, such as when a child is diseased, injured, or born with a deformity. In  such cases, the offspring’s survival and future reproduction are unlikely, meaning that parental investment in that child will not translate into reproductive success. Also, when resources are scarce, it may be unlikely that the baby will live to reproductive age, or it may be

40  Evolution, natural selection, and behavior

advantageous to invest the resources in an older, more competitive offspring rather than a weak baby. As women have a much shorter reproductive window than men, investing resources and time in an offspring that will not improve her reproductive success is not advantageous.24 Infanticide would have very different fitness benefits for men. Killing an infant that is unlikely to survive is less beneficial for men than for women, as men invest less time and resources on offspring. However, other benefits have been proposed, such as killing a mate’s infant from a previous relationship, killing an infant whose paternity is uncertain, and even bringing a woman back into estrus after giving birth to another man’s child, as, in subsistence populations, lactating can prevent a woman from ovulating.24 Siblicide

Siblicide is the murder of one’s siblings. Natural selection favors traits that increase reproductive fitness, which passes these traits to the next generation. Kin selection theory predicts that relatedness should promote altruism, in that it is important to protect close relatives, such as siblings, as they carry many of the same genes that you do, so ensuring that they survive to reproduce helps you indirectly by passing on those traits to the next generation.27 This is perhaps best illustrated in a beehive, in which only one female, the queen, is fertile and all the other related females are sterile. They act as workers to raise the offspring of the queen, their mother, and in so doing, help to pass their genes on to the next generation. Siblicide therefore seems counterintuitive. However, siblings may come into conflict when competing for resources such as parental investment. This is more commonly seen in non-human animals, particularly in bird species, in which a nestling will frequently push another nestling out of the nest in order to receive more resources from the parents. Although much rarer in mammals, siblicide does occur in some species, usually when food resources are low.27 It may also have been common in our ancestral past.

3.  Sexual assault and rape The role of evolution in rape and sexual assault is perhaps the easiest to understand. All animals want to maximize their fitness by contributing as many genes as possible to the next generation. A  single female of any species has a limited reproductive potential; therefore, a male that sires offspring with many females will have greater fitness than one who sires offspring with only one. Also, by impregnating and then leaving, the male does not need to expend resources on raising the resulting offspring. From a reproductive fitness point of view, rape has high reproductive benefits; there are cases of approximately a third of women becoming impregnated after rape during war.28 Some studies have shown that pregnancy is twice as common in rapes as in consensual sex, which has led to the suggestion that men who rape may have a different psychology or may more competitively ejaculate, as it has been shown that sperm density can be unconsciously controlled by males under certain circumstances.29 Much of evolutionary theory relating to reproductive strategies shows that females choose mates who have high status and resources, to maximize their chances of reproduction and raising offspring to reproductive age. Evidence suggests that males from many species have evolved strategies to sexually force females.29 For rape to be an evolutionary adaptation, it must have regularly resulted in increasing reproductive fitness (that is, producing more offspring) for ancestral rapists, and these reproductive benefits must have outweighed the costs.29 An animal example of such behavior can be seen in the bluegill sunfish. It is referred to as the cheater, or “cad versus dad,” theory.14 Bluegill sunfish males come in two forms, or morphs. One, the resident, or parental, male, devotes a great deal of energy and time to acquiring and protecting a territory to attract a female. Once he has built a cozy nest, he attracts a female by displaying the proposed nursery. The choosy female selects such a male, as he has clearly demonstrated the ability and desire to provide. The female enters the nest and is induced by the male to lay her eggs, whereupon the male fertilizes them. This is the way it is supposed to happen. However, the other bluegill

Natural selection and crime  41

male morph is known as a sneaker. He waits until the resident male and female are courting, and as soon as the female lays her eggs, he dives in, fertilizes them with his sperm, and swims away.14 The sneaker has evolved to be smaller and to mature earlier, and although he lives a shorter life than resident males, he accounts for a larger number of offspring.30 This is just one of many examples of such evolved behaviors. A population can only support a limited number of cheaters. If the number of cheaters rose above a certain threshold, the population of conforming males would be outnumbered, and there would not  be enough conforming males to provide resources for the offspring of cheaters. Evolutionary criminologists have suggested that such alternate reproductive strategies also exist in humans. As successful females will be choosy, males must either develop a reproductive strategy that supports this (that is, comply and provide resources, as the resident male sunfish does) or develop an alternate reproductive strategy. There are several lines of evidence that support the cheater theory. Cheater behavior is clearly genetically controlled antisocial behavior, and this theory states that it has evolved only in males. This would predict that criminal behavior is dominated by males and also most commonly seen in young males seeking mates. According to this theory, such antisocial behavior is most likely exhibited by males who cannot convince females of their ability to provide for offspring, suggesting that cheaters are more likely to be men of lower socioeconomic class. Cheater behavior is more likely to succeed in large impersonal communities than in tight-knit communities, suggesting that such behavior is more likely seen in urban rather than rural communities. The theory also suggests that criminals and psychopaths are more likely to have numerous sexual partners and that such people will exhibit high levels of deception and lying.14 Although there are obviously many parameters that contribute to criminal behavior, the above characteristics of cheaters are supported by the literature. For example, there is a large body of literature that supports the contention that most crimes, particularly serious and recidivistic crimes, are committed by young males during the early reproductive years.31 Later chapters in this text indicate that a certain amount of criminal behavior is inherited, indicating a genetic basis. Moreover, setting aside the myriad other causes of crime, criminal behavior is more frequently seen in poor, urban neighborhoods,14 and psychopaths are much more likely to move from relationship to relationship and to exhibit high levels of deception.32 As with the sunfish, women who chose mates who have resources to provide for future offspring are most likely to be successful, that is, to successfully raise the offspring to reproductive age. In the past, women’s mates were chosen by their parents, who also favored mates with high status and excellent resources. This meant that men with low status and low resources had low mate value and were unlikely to reproduce or were forced to mate with females who also had low mate value— such as older women, who were less reproductively fit—with few resources.28(p485) Theories suggest that rape evolved to allow men with low reproductive expectations to avoid female and parental choice.28 Such a rapist would fit into the category of “disadvantaged rapist” in McKibbin and colleagues’ typologies of rape, in evolutionary psychology terms29; this is commonly referred to as the mate deprivation hypothesis.33 Of course, in most situations, such a male would face high levels of retaliation from parents and family and from the woman. Moreover, the woman may be injured in the attack, which might also reduce reproductive success.28 It has therefore been suggested that rape evolved to be advantageous to men with low mate value in circumstances in which there is little cost or chance of retaliation, such as in war, or during raids on other communities, or where the woman has little chance of retaliation due to a lack of family support or inability to identify the rapist, or where the victim is unlikely to report the rape, perhaps due to youth.28 It has therefore been argued that rape is a reproductive strategy that increases fitness in low-mate-value males by evading female and parental choice in situations where there is low cost (that is, risk of retaliation). Many rape theories support the hypothesis that rape evolved to increase reproduction at times when the risk of such behavior is low.33 Other forms of rape as an evolutionary reproductive strategy have been proposed by McKibbin and colleagues, including the specialized rapist, the opportunistic rapist, the high-mating-effort

42  Evolution, natural selection, and behavior

rapist, and the partner rapist.29 The  specialized rapist is hypothesized to have evolved an ability to be more rapidly aroused and to more rapidly ejaculate than normal, which greatly minimize risk. It has been speculated that premature ejaculation may have increased reproductive success ancestrally to reduce the risk of predation or discovery by potential rivals.29 The opportunistic rapist is hypothesized to normally prefer receptive women but will also engage in rape in situations where there is very little cost,29 which is borne out by the very high numbers of rapes that occur in times of conflict (for example, the more than 1 million rapes by the Red Army after World War II).28 The  high-mating-effort rapist is hypothesized to be more sexually experienced than most, with a very high level of self-esteem and self-perceived sexual prowess. He pursues a large number of partners with little investment and resorts to rape when consensual strategies fail. This typology is linked to psychopathy.29 The final hypothesized typology is that of partner rapist, who rapes his partner at times when he believes her to be unfaithful and therefore possibly impregnated by another man. This relates to sperm competition, in which males compete to ensure that their sperm impregnates the female. Although partner rapes are very common and were not even considered illegal in many countries until relatively recently, studies have shown that partner rapes increase during breakups due to concern over infidelity.29 Although rape is much more complex than can be covered here, it is clear that, seen through a purely evolutionary lens, rape as a reproductive strategy can be very successful. It is seen frequently in many species, sometimes in an obligate form, where all copulations are forced, such as in scorpionflies,29 and sometimes in a facultative form, in which most copulations are consensual, but some are coerced under certain circumstances, such as in humans. Clearly, when we are discussing rape in this context, we are talking about evolutionary traits, which may have been successful in an ancestral setting rather than in the present day. In animals and in our ancestral past, rape could result in increased reproductive success with, in certain circumstances, little cost. However, many factors in modern society greatly impact the evolutionary impact of the act of rape. Today, rape is illegal in almost all cultures and results in high penalties, which curtail a person’s freedom for a lengthy period of time or even cost the rapist’s life, which limits or ends any chances of reproductive success. Moreover, in many (although not all) cultures, a woman has access to abortion and so can terminate the pregnancy. Even in countries where abortion is not normally legal, it is sometimes allowed after rape. So today, abortion eliminates the rapist’s reproductive success directly, and the costs of rape may greatly curtail or terminate his future reproductive success.

4.  Child abuse In  a simplistic sense, organisms have two major reproductive evolutionary strategies, which in the past were called r-selection and K-selection. More developed theories are used today, but the older categories are useful, as they give us a clear idea of parental investment. Animals who are r-­strategists produce very large quantities of offspring, with little or no parental investment in raising and ­protecting them. In contrast, K-strategists produce only a few offspring, but parental investment is high. It is a choice between quantity and quality. r-Strategists produce vast numbers of offspring, expecting most will die but assuming, as there are so many, that some will survive to pass on their genes. A common example is that of most species of insects, which produce very large numbers of offspring, with a very high mortality. A vertebrate example of an r-strategist would be that of rats, who have a very early sexual maturity and a short gestation period and produce large numbers of offspring frequently, with a very brief period of maternal care. K-strategists, on the other hand, put a great deal of investment into their offspring, with a later sexual maturity, a long gestational period, and only a few offspring produced over a lifetime, with lengthy and costly parental care. Examples would be elephants, whales, and, of course, humans. K-strategists, therefore, maximize their fitness by a very costly investment in their offspring. Therefore, abusing or neglecting one’s biological offspring is not a successful strategy. However, killing the offspring that was sired by a previous mate would be a successful strategy. We see this in

Natural selection and crime  43

nature all the time. For example, a male bear has little or no investment in his offspring. He chooses and mates a female and then leaves. The females, however, have a very costly investment, with a long gestation and several years of maternal protection and rearing. Females with offspring at foot will not come into estrus until the cubs have grown and left. A male bear, when he finds a female with cubs, will kill the cubs, so that the female will come into estrus and he can then impregnate her. Consequently, female bears will risk moving to poorer habitats with their young to protect them from males. Many species of higher animals have such strategies. Human studies on biological children and stepchildren have shown similar patterns. Many studies have shown that children are at greater risk of abuse or death by a stepparent than by a natural parent. In such a situation, just as with the bears, a non-biological parent does not gain any reproductive fitness from investing in non-related children and may even lose further fitness because the biological parent continues to invest resources in the ex-partner’s child instead of producing more children with the new partner. A study of children in Hamilton, Ontario, a mid-sized town, found that police arrests and reports of abuse were least likely when children lived with both their natural parents, but that preschoolers living with one biological parent and a stepparent were 40 times more likely to be victims of child abuse than conspecifics living with biological parents.34 Similar results were seen in studies in the United States, with risk decreasing as the child aged but still remaining at twice the level of risk of similarly aged children living with both natural parents.34 Risk was also much higher for children living with a stepparent than with a single biological parent.34 Filicide, the killing of a child by a parent, is also much higher in stepparents. In a US study, young children (aged 2 years and under) were 100 times more likely to be killed by a stepparent than by a biological parent, and in Canada, results were similar, with children 70 times more likely to die at the hands of a stepparent than by a biological parent.22,35 Even unintentional deaths, such as drowning (implying a lack of supervision), were found to be 2–15 times more likely in preschool stepchildren than comparable biological children.36 It has been argued that people in second and subsequent marriages may be more prone to aggression, which may have been the cause of their original breakup, or that they may be part of a cycle of violence if they themselves were abused as children, which might also be linked to marriage breakup. In a Swedish study, general and violent crime was found to be higher in stepfamilies.37 However, many studies have shown that stepparents discriminate between their biological children and stepchildren when both are in the home and specifically abuse non-related children.34 A large number of studies have supported this theory, with stepchildren being much more commonly killed by stepfathers than biological children,38,39 although a smaller Swedish study found that children were killed equally often by biological parents and stepparents; however, only 27 cases were considered.37 Several studies have also shown that the method of homicide and the level of brutality vary between biological parents and stepparents. Stepfathers who kill their stepchildren are more likely to kill using very violent methods, such as beating and bludgeoning, in comparison with biological fathers, who more commonly use asphyxiation and firearms, potentially causing a more merciful death.35,39 Evolutionary psychologists Daly and Wilson state that biological parents more commonly kill their children in “sorrow rather than in anger” and under a perception of need, or even “rescue,” as seen in murder–suicides.35(p208) In a large study of child homicides in Canada from 1974 to 1990 and in England and Wales from 1977 to 1990, a significant number of children killed by biological fathers were killed as part of a murder–suicide or killed along with the spouse, with almost no such cases among stepfathers.35 Daly and Wilson suggested that the greater violence seen in stepparent killings represented resentment toward stepchildren.35 These studies were replicated in a large US study looking at almost 4000 cases of the homicide of a child under the age of 5 years by a biological parent or stepparent, with very similar results.39 Children were found to be killed at a rate of 51.2 per million per annum by a stepparent and 15.6 per million per annum by a biological parent. The Daly and Wilson studies only considered stepfathers, but the US study considered both stepfathers and stepmothers and found that the annual rate for filicide by stepfathers was 60 per million, in

44  Evolution, natural selection, and behavior

contrast with 7 per million for biological fathers. The annual rate for stepmothers was 20.6 per million, in contrast with 8.6 per million for biological mothers. As with the earlier study, significantly more brutal methods of homicide were utilized by stepparents than by genetic parents, supporting the contention that there are psychological and intentional differences between the two and much greater negative feelings between stepparents and children.39 As these studies have been replicated in three different countries using three separate national databases, the results are very robust. Understanding the evolutionary psychology underlying the parent-child bond is important in developing child welfare policy. Parental investment in child welfare is greatest in biological parents, as the child will pass the parents’ genes on to the subsequent generations and thereby increase the parents’ reproductive fitness. Close kin, such as an aunt or an uncle, also share genes with the child, so it is advantageous for kin to invest resources in the child, although their genetic investment is lower than that of the parents. As the degree of relatedness decreases, the reproductive fitness benefits proportionally decrease. Therefore, close kin could be expected to invest more than distant kin, and any kin will invest more than non-kin.40 This results in two major hypotheses when considering potential foster care for a child that has been removed from the parental home. The first hypothesis is that children will be better cared for in a kinship foster placement than in a non-kin environment.40 Several studies support this hypothesis (reviewed in Herring40). For example, studies have shown that even though kinship foster homes receive less government support than non-kinship homes, the children’s success is similar in both homes. Also, kinship homes are more stable, with the resultant improved outcomes in the children. In a longitudinal, national study, children in kinship homes exhibited better mental health and wellbeing than children placed in non-kinship homes, and in another study of the same dataset, grandmothers had much better parenting scores than non-kin foster parents (reviewed in Herring40). The  second hypothesis is that kin will differentially invest in children depending on the type of kin, which relates to two concepts, paternity and sex effect.40 Maternity of a child is a certainty—a mother knows that a child is her biological child—but paternity is much less certain, so one would expect a greater level of investment from the maternal side of the family, who is sure of the biological relationship, than the paternal side. This would suggest that maternal grandmothers would invest more than paternal grandmothers.40 The sex effect relates to the fact that a woman has a much more limited reproductive potential than a male, as there is a limit to the number of children she can have, whereas the number of children a man with many mates can sire is virtually limitless. This means that women invest more in their offspring than men, while men tend to invest more in mating effort. This is also impacted by age: when a woman ages past the point of reproduction, she can still maximize her reproductive fitness by investing in close kin, whereas a male’s age does not limit his reproduction.40 Several studies also support this second hypothesis. In Japan, a study from a Japanese village from 1671 to 1871 found that a child was 35% less likely to die if the maternal grandmother was in the home, but the presence of the paternal grandmother or either grandfather increased mortality.41 In a review of 45 studies, maternal grandmothers consistently had beneficial effects on child survival; paternal grandmothers had a variable effect, sometimes positive, sometimes negative; and grandfathers had no impact on child survival.42 In a later meta-analysis of 17 studies, both maternal grandparents provided survival benefits, in contrast with paternal grandparents, whose presence had negative effects.43 Understanding the evolutionary underpinnings of parental investment may help to reduce child maltreatment and inform policy surrounding child welfare and placement.40

5.  Domestic violence Humans, unlike most mammals, are monogamous, choosing one mate and remaining with that mate for, potentially, decades. From an evolutionary perspective, monogamy has many benefits. From a woman’s point of view, a long-term partner provides physical protection for her and her children; a consistent supply of resources, which increases child survival; and assistance in raising offspring, which increases socialization and subsequent survival and reproductive success of

Evolutionary theories  45

the offspring. Men willing to be in a long-term, committed relationship improve their chances of securing a desirable woman; improve their certainty of paternity; increase the survival of their children and improve their children’s reproductive success, due to their proximity and resource provision; and improve their own status through the woman’s kin.44 Therefore, the many evolutionary benefits of a long-term relationship would suggest that maintaining a harmonious relationship would be beneficial and that domestic violence would be detrimental. However, the very high rates of domestic violence belie this concept. Wilson and Daly suggest that at least some types of domestic violence relate to controlling a woman’s sexuality to deter infidelity.45 If this theory is correct, then domestic violence should be highest when a woman is young and fertile and decrease as her fertility decreases; several studies have supported this. For example, a study of almost 4000 cases of domestic partner abuse perpetrated by a male against a female over 14 years showed that rates of domestic violence decreased as women aged, and younger women of reproductive age were almost 10 times more likely to be abused than older women past menopause.46 Many other theories have been proposed based on sexual conflict theory, which states that “sexual conflict occurs whenever there is a conflict between the evolutionary interests of individuals of two sexes.”47(p235) These include mate poachers (stealing a mate from another), sexual infidelity (which puts a man at risk of raising children who are not genetically related to him), genetic cuckoldry, resource infidelity and scarcity, mate value discrepancies, stepchildren, relationship termination, and mate reacquisition.44,48 A more in-depth analysis is beyond the scope of this book, and it must be remembered that, in all cases, many more complex issues come into play. However, such theories can help us understand why such criminal acts may have been selected for in our ancestral past. The  above examples show some ways in which criminal activity could have been advantageous in our ancestral past, increasing resources and acquiring mates by whatever means, to increase survival and maximize reproductive fitness. In our modern world, laws have made such activities illegal, with consequent punishment, which changes the playing field, as it greatly increases risk and lowers the benefits. Being caught may limit or even permanently stop reproduction (by incarceration or execution), which changes the balance of benefit and risk.

Evolutionary theories Several evolutionary theories relate to criminal behavior, and some have been mentioned above in relation to the evolution of specific crime types.

Cheater theory The cheater theory, also referred to as the cad-versus-dad theory,14 is often used to explain the evolution of psychopathic behavior. In this theory, the cheater exploits males that have developed traditional mating strategies, such as providing food, resources, and territory. As discussed previously, females have a lower reproductive rate than males, as they can only produce a limited number of offspring during a limited period of their life, whereas males can sire offspring from puberty until death and could hypothetically sire many offspring in a single day. Women somewhat compensate for this by being more choosy in their mates, selecting mates that can provide for the offspring, whereas males will increase their reproductive success by mating with a large number of women. Males can either comply with choosy females (dad) or can trick or force a female to copulate (cad). Such behavior is seen in many species, such as the bluegill sunfish mentioned previously, and also in humans. Mealey49 and Lykken50 suggested that there are two forms of this alternate reproductive strategy in human males: (1) those who are genetically predetermined to be cheaters and exhibit such symptoms from early in life and (2) those who learn the strategy and confine its use to youth, eventually growing out of it and conforming. Mealey and Lykken suggested that the first type could be classified as psychopaths and the latter as sociopaths.

46  Evolution, natural selection, and behavior

Variation within K-strategists Although humans are K-strategists overall, individuals also vary within species as to whether they are closer to r- or K-strategists. A great example of this is seen in dandelions. One of the greatest mortality factors for a dandelion is a lawnmower. Lawnmowers kill instantly, regularly, and indiscriminately.51 Frequent mowing will therefore select for an r-strategy, with early, rapid, and intense reproduction. Researchers compared reproductive rates in two populations of dandelions, one that was frequently mowed and one that was not disturbed.52 The two groups of plants, despite being from the same species, exhibited very different reproductive strategies, even if they were transplanted to the laboratory and raised under identical conditions, with no mowers in sight. Ellis53 suggested that, in the same way, although humans are K-strategists, some males are less extreme K-strategists than others. Those who are less extreme will have more children with lower birth weights and earlier onset of reproduction, as birth weight and age of onset of reproduction are considered to be major indicators of parental investment.14 Ellis believes that men who are less extreme in exhibiting a K-strategy will be more prone to criminal and antisocial behavior. This includes two major assumptions: that humans vary in K-strategy and that antisocial behavior reflects more of an r-strategy to reproduction.53 Ellis and Walsh14 explained this by pointing out that property crimes result in the rapid acquisition of resources, which might attract sexual partners, and violent sexual crimes are often motivated by maintaining or securing mating opportunities. They contend that antisocial behavior is related to an r-approach to reproduction, and, therefore, genes that support r-selected traits, such as low birth weight, large family size, and multiple births, should be related to antisocial behavior. Again, although many other variables exist that influence these traits, as will be seen in subsequent chapters, these traits are more frequently associated with criminal behavior.

Conditional adaptation theory This theory, first proposed by Belsky,54 suggests that genes have an indirect rather than a direct impact on criminal behavior. It asserts that humans have an evolved, genetic ability to subconsciously monitor their environment when very young and subsequently adapt their behavior depending on this formative environment. The most pertinent parts of the environment, in this theory, relate to availability of resources and the stability of interpersonal relationships.14 The theory indicates that if young children are exposed to interpersonal instability (such as parental arguments, abuse, and divorce) and lack of resources (such as poverty), they are more likely to become opportunistic, with an increased risk for criminal behavior in later life. They are more likely to become sexually active early and reproduce when very young, as they are now  programmed to believe that resources will be scarce and relationships will be unreliable, and thus will “speed up” their lives.14 According to this theory, such people would then be more likely to procure resources by any means, including deception. Children from stable homes with sufficient resources are more likely to acquire resources in a more orderly manner. This theory is interesting and is supported by other evidence, much of which is discussed further in later chapters, such as the beneficial effects of a stable family home. However, as will become clear, there are many other negative aspects related to poverty and lack of parental care that also influence behavior.

Alternate adaptation theory Rowe55 proposed the alternate adaptation theory, which states that people either emphasize mating effort (in other words, seeking mating partners) or parental effort (in other words, caring for offspring). He stated that the strongest predictors of these are gender and a variety of behavioral traits, such as sensation-seeking, aggression, and sex drive.14 People low in these personality traits,

Conclusion  47

particularly women, will put more effort into parenting, whereas people high in these traits, particularly men, will put more effort into mating.14 Ellis and Walsh14 also point out that intelligence would be a third predictor, as people with high intelligence would be more likely to be able to carry out long-term and complex resource accumulation and thus put more effort into parenting. Rowe’s theory suggests that criminals would be biased toward mating effort rather than parenting effort, and therefore, criminality would be biased toward males with a relatively strong sex drive. He states that criminal behavior is a very direct and immediate method of gaining resources. Rowe’s theory contrasts with the conditional adaptation theory, as it suggests that the genetics of the parents are much more important than the rearing environment.14

Evolutionary expropriative behavior In this theory, briefly discussed previously, there are two forms of acquiring resources: the usual way, in which one works to earn or produce them, and an expropriative manner, in which one takes them from those who acquired them in the normal manner. A great deal of crime actually fits into this theory, as some people acquire wealth or resources by hard work and others cheat, trick, or steal those resources. Such a system works best in large societies that create more opportunity for expropriating resources. This leads society to the necessity of developing a system that will discourage such expropriation, that is, a criminal justice system.14 This would also explain why crime is often rifer in cities than in rural areas. According to this theory, people with low intelligence and poor academic ability will be more likely to expropriate resources than those with higher education and intelligence, who will be more likely to acquire resources with fewer risks and greater longterm rewards.14

Conclusion This chapter introduced some basic biological concepts to begin our discussion of the biological approach to crime that we will be developing in subsequent chapters. We discussed natural selection and compared behaviors of humans and other animals. Natural selection works through organisms adapting to their environments; some traits are selected over others and passed to the next generation. These heritable traits include the structural, biochemical, and physiological makeup of the body but also, and more importantly for our study, behavioral traits, which are equally adaptive. In addition, we discussed a short critical narrative of the nature-versus-nurture argument. In doing this, we came easily to the idea that there is really not an opposition between these two influences but an interaction; neither works in isolation. Behavior can be either entirely genetic or a mixture of genetics and the environment, the contribution of each varying with the behavior. We then considered natural selection and crime and how many criminal behaviors may have been advantageous in our ancestral past and therefore been selected for. This means that despite the fact that most are maladaptive today due to our criminal justice system, they still persist, although with probably lower frequency than in the past. The final part of this chapter looked at evolutionary approaches to criminal behavior, including cheater theory; r/K strategies, in which amount of parental care is seen as a heritable trait; conditional adaptation theory, in which crime may be a deep structure adaptation to the amount and availability of resources in childhood; alternate adaptation theory, in which mating effort, or parental effort, and caring for offspring are inherited traits and can be predictors for criminality; and evolutionary expropriative theory, in which two major strategies of survival are heritable — generative behavior and expropriative behavior. We next discuss genetics and the interesting question of the specific means by which adaptation, genetic inheritance, and criminal behavior interact.

48  Evolution, natural selection, and behavior

Questions for further study and discussion 1. Explain why all three conditions—variation, heritability, and fitness consequences—are required for natural selection to act. 2. Explain why the sex drive is innate. Discuss how we know it is innate and explain the evolutionary advantages. How is learning involved and why is learning needed to increase success? 3. Discuss, from an evolutionary perspective only, why a man is more likely to abuse or kill his stepchildren and why such killing is likely to be particularly brutal. 4. Explain, from an evolutionary perspective only, why all men do not rape. In other words, if rape can be reproductively successful, why do some males instead remain monogamous and provide resources for a spouse and offspring? 5. Explain, from an evolutionary perspective, why it is advantageous for an uncle to support his nieces and nephews, and how an understanding of the evolutionary basis of parental investment can inform child welfare policy.

References 1. Campbell, N.A. 1996. Biology. Menlo Park, CA: Benjamin/Cummings. 1270 pp. 2. Darwin, C. 1859. On the Origin of Species by Means of Natural Selection; or the Preservation of Favoured Races in the Struggle for Life. London: John Murray. 859 pp. 3. Cain, M., Damman, H., Lue, R.A., and Yoon, C.K. 2000. Discover Biology. Sunderland, MA: Sinnauer Associates. 209 pp. 4. Reece, J.B., Urry, L.A., Cain, M.L., et al. 2018. Campbell Biology. 2nd Canadian ed. Don Mills, ON: Pearson Canada. 1346 pp. 5. Mather, M.H. and Roitberg, B.D. 1987. A sheep in wolf’s clothing: Tephritid flies mimic spider predators. Science 236: 308–310. 6. Beaver, K., Nedelec, J.L., Schwartz, J.A., and Connolly, E.J. 2014. Evolutionary behavioral genetics of violent crime, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 117–136. 7. Majerus, M.E.N. 2008. Industrial melanism in the peppered moth, Biston betularia: An excellent teaching example of Darwinian evolution in action. Evolution: Educat. Outreach 2(1): 63–74. 8. Dilger, W. 1962. The behavior of lovebirds. Sci. Am. 206(1): 89–98. 9. Purves, W.K., Orians, G.H., and Heller, H.C. 1995. Life, the Science of Biology. 4th ed. Sunderland, MA: Sinauer Associates. 1280 pp. 10. Daly, M. and Wilson, M. 2001. Risk-taking, intrasexual competition and homicide. Nebr. Symp. Motiv. 47: 1–36. 11. Ellis, L. 1989. Theories of Rape: Inquiries into the Causes of Sexual Aggression. Boca Raton, FL: Taylor & Francis Group. 175 pp. 12. Bateson, P.P.G., ed. 1983. Mate Choice. Cambridge, UK: Cambridge University Press. 480 pp. 13. Bateman, A. 1948. Intra-sexual selection in Drosophila. Heredity 2: 349–368. 14. Ellis, L. and Walsh, A. 2000. Evolutionary biosocial theories, In: Criminology: A  Global Perspective. Boston, MA: Allyn & Bacon. pp. 432–468. 15. Ellis, L. and Hoskin, A.W. 2015. The evolutionary neuroandrogenic theory of criminal behavior expanded. Aggress. Viol. Behav. 24: 61–74. 16. Gariépy, J.-L., Lewis, M.H., and Cairns, R.B. 1996. Genes, neurobiology and aggression: Time frames and functions of social behaviours in adaptation, In: Aggression and Violence, Stoff, D.M. and Cairns, R.B., editors. Mahwah, NJ: Lawrence Erlbaum. pp. 41–48. 17. Cohen, L.E. and Machalek, R. 1988. A general theory of expropriative crime: An evolutionary ecological approach. Am. J. Sociol. 84: 465–501. 18. Vila, B.J. 1994. A general paradigm for understanding criminal behavior: Extending evolutionary ecology theory. Criminology 32: 311–359.

References  49

19. Vila, B.J. and Cohen, L.E. 1993. Crime as strategy: Testing an evolutionary ecological theory of expropriative crime. Am. J. Sociol. 98: 873–912. 20. Moffit, T.E. 1993. Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychol. Rev. 100(4): 674–701. 21. Keeley, L.H. 2014. War before civilization: 15  years on, In: The  Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 23–31. 22. Daly, M. and Wilson, M. 1988. Evolutionary social psychology and family homicide. Science 242(4878): 519–524. 23. Crabb, P.B. 2000. The material culture of homicidal fantasies. Aggr. Behav. 26(3): 225–234. 24. Duntley, J.D. and Buss, D.M. 2011. Homicide adaptations. Aggress. Violent Behav. 16(5): 399–410. 25. Daly, M. and Wilson, M. 1990. Killing the competition: Female/female and male/male homicide. Hum. Nat. 1(1): 81–107. 26. Gottschalk, M. and Ellis, L. 2010. Evolutionary and genetic explanations of violent crime, In: Violent Crime: Clinical and Social Implications, Ferguson, C.J., editor. Thousand Oaks, CA: Sage Publications. pp. 57–74. 27. Salmon, C.A. and Hehman, J.A. 2014. The evolutionary psychology of sibling conflict and siblicide, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 137–158. 28. Apostolou, M. 2013. The evolution of rape: The fitness benefits and costs of a forced-sex mating strategy in an evolutionary context. Aggress. Violent Behav. 18(5): 484–490. 29. McKibbin, W.F., Shackelford, T.K., Goetz, A.T., and Starratt, V.G. 2008. Evolutionary psychological perpectives in rape, In: Darwinian Foundations of Crime and Law, Duntley, J.D. and Shackelford, T.K., editors. New York: Oxford University Press. pp. 101–120. 30. Lalumière, M.L., Harris, G.T., Quinsey, V.L., and Rice, M.E. 2005. Forced copulation in the animal kingdom, In: The Causes of Rape: Understanding Individual Differences in Male Propensity for Sexual Aggression. Washington, DC: American Pscyhological Association. 31. Boyd, N. 2000. The  testosterone connection, In: The  Beast Within: Why Men Are Violent. New York: Greystone Books. pp. 115–138. 32. Hare, R.D. 1981. Psychopathy and violence, In: Violence and the Violent Individual, Hays, J.R., Roberts, T.K., and Solway, K.S., editors. New York: SP Medical and Scientific Books. pp. 53–74. 33. Camilleri, J.A. and Stiver, K.A. 2014. Evolutionary perspectives on human sexual psychology and behavior, In: Evolutionary Perspectives on Human Sexual Psychology and Behavior, Weekes-Shackelford, V.A. and Shackelford, T.K., editors. New York: Springer. pp. 43–67. 34. Daly, M. and Wilson, M. 1985. Child abuse and other risks of not living with both parents Etholog. Sociobiology 6: 197–210. 35. Daly, M. and Wilson, M.L. 1994. Some differential attributes of lethal assaults on small children by stepfathers versus genetic fathers. Ethol. Sociobiol. 15: 207–217. 36. Tooley, G.A., Karakis, M., Stokes, M., and Ozanne-Smith, J. 2006. Generalising the Cinderella Effect to unintentional childhood fatalities. Evol. Hum. Behav. 27: 224–230. 37. Temrin, H., Nordlund, J., Rying, M., and Tullberg, B.S. 2011. Is the higher rate of parental child homicide in stepfamilies an effect of non-genetic relatedness? Current Zool. 57(3): 253–259. 38. Harris, G., Hilton, N., Rice, M., and Eke, A. 2007. Children killed by genetic parents versus stepparents. Evol. Hum. Behav. 28(2): 85–95. 39. Weekes-Shackelford, V.A. and Shackelford, T.K. 2004. Methods of filicide: Stepparents and genetic parents kill differently. Violence Victims 19(1): 75–81. 40. Herring, D.J. 2014. Evolutionary perspectives on child welfare, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 53–72. 41. Jamison, C.S., Cornell, L.L., Jamison, P.L., and Nakazato, H. 2002. Are all grandmothers equal? A review and a preliminary test of the “grandmother hypothesis” in Tokugawa Japan. Am. J. Phys. Anthropol. 119(1): 67–76. 42. Sear, R. and Mace, R. 2008. Who keeps children alive? A review of the effects of kin on child survival. Evol. Hum. Behav. 29(1): 1–18.

50  Evolution, natural selection, and behavior

43. Strassmann, B.I. and Garrard, W.M. 2011. Alternatives to the grandmother hypothesis: A meta-analysis of the association between grandparental and grandchild survival in patrilineal populations. Hum. Nat. 22(1–2): 201–222. 44. Buss, D.M. and Duntley, J.D. 2014. Intimate partner violence in evolutionary perspective, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 1–22. 45. Wilson, M. and Daly, M. 1993. An evolutionary psychological perspective on male sexual proprietariness and violence against wives. Violence Vict. 8(3): 271–294. 46. Peters, J.R., Shackelford, T.K., and Buss, D.M. 2002. Understanding domestic violence against women: Using evolutionary psychology to extend the feminist functional analysis. Violence Vict. 17(2): 255–264. 47. Parker, G.A. 2006. Sexual conflict over mating and fertilization: An overview. Phil. Trans. Royal Soc. B 361(1): 235–259. 48. Buss, D.M. and Duntley, J.D. 2011. The evolution of intimate partner violence. Aggress. Viol. Behav. 16(5): 411–419. 49. Mealey, L. 1995. The sociobiology of sociopathy: An integrated evolutionary model. Behav. Brain Sci. 18(3): 523–599. 50. Lykken, D.T. 1995. The Antisocial Personalities. Hillsdale, NJ: Laurence Erlbaum. 264 pp. 51. Daly, E.M. and Wilson, D. 1983. Life History Strategy; Sex in Evolution and Behavior. Boston, MA: Willard Grant Press. 416 pp. 52. Gadgil, M. and Solbrig, O.T. 1972. The concept of r- and K-selection: Evidence from wild flowers and some theoretical considerations. Am. Naturalist 104: 1–24. 53. Ellis, L. 1987. Criminal behavior and r/K selection: An extension of gene-based evolutionary theory. Dev. Behav. 8: 149–176. 54. Belsky, J. 1980. Child maltreatment: An ecological integration. Am. Psychol. 35: 320–335. 55. Rowe, D.C. 1996. An adaptive strategy theory of crime and delinquency, In: Delinquency and Crime: Current Theories, Hawkins, J.D., editor. Cambridge, UK: Cambridge University Press. pp. 268–314.

3 Genetic principles

Introduction This  chapter introduces some basic concepts of genetics. Without this knowledge, it would be impossible for you to decide whether genetics can or cannot influence crime, and one must understand genetics to understand how natural selection acts. This chapter provides a very basic overview of genetics and the patterns of inheritance. This will allow you to appreciate the complexity of genetics, meiosis, and patterns of inheritance and explain some of the terms used. By the end of this chapter, you should have an understanding of basic inheritance patterns and realize that although genetics influences our behavior, our genotype interacts with our environment and that DNA alone is not destiny.

Introduction to genetics Most people who are not  trained in biology tend to ignore genetics, but they still draw conclusions about biological factors and experiments. Their conclusions are usually wrong because they do not understand the basic concepts of biology. This text will not look at complex genetic issues, only at basic biology; most of you will have covered this in high school. It should, however, provide the background needed to begin to understand genetics. It will show how the body works in relation to its environment and how natural selection acts. Darwin (1809–1892) did not know about genetics, although the man who is considered the father of genetics, Gregor Mendel (1822–1884), an Austrian monk, was alive at the same time as Darwin. It is one of the curiosities of the history of science that neither of them fully understood the importance of each other’s work during their lifetimes; years later, someone else read both their works and figured out the relationship. Sadly, after his death, Darwin was found to have a copy of Mendel’s famous paper with notations in the margins of the first few pages, but he obviously had not read further. Just like every other plant and animal, humans carry their genetic material in their chromosomes, which are found in the nucleus of every cell in the body. Each chromosome consists of strings of genes. The human species has 46 chromosomes (other species have different numbers) in 23 pairs. Twenty-two of the pairs are the somatic chromosomes, and one pair is the sex chromosomes. Everyone therefore has two chromosome 1s, two chromosome 2s, and so on, and a pair of the sex chromosomes. The sex chromosomes are noted as XX if you are female and XY if you are male. Each member of a pair of chromosomes has the same genes, in the same order, as the other member of the pair, so we call them matching, or homologous, pairs. Every one of these chromosome pairs is found in nearly every single cell in the body (exceptions include the red blood cells, which do not have a nucleus). Every chromosome consists of strings of genes made of

51

52  Genetic principles

DNA (deoxyribonucleic acid), and they code for life. They are our blueprint, and they affect just about everything about us. What is a gene? It is a set of instructions that code for (that is, make the body produce) a particular protein. This protein then does something in the body. It makes something or is involved in some mechanism. When something goes wrong with a gene, it means that the protein is not produced at all, or that it is produced in lesser quantities than normal, or that too much is produced. Any error can affect the end product. Every gene has a location, or a locus (plural loci), where it is found on the chromosome, almost like an address. For example, the gene for eye color may be on chromosome 3. (This is not a good example because eye color is actually governed by at least 20 genes and possibly by hundreds on several chromosomes,1 but let us pretend, for the sake of argument, that there is only one gene involved in eye color.) Let us say that the gene for eye color is 20 genes up on chromosome 3. You will always find the gene for eye color there; that is its address or locus. The sequence of genes on each chromosome is the same in both members of the pair. This is what the human genome project is all about—mapping the addresses for all the human genes. So, in our artificial example, the gene for eye color resides on chromosome 3, but this gene can have several different forms. It can be blue or gray or brown or green or violet or hazel. Each of these different versions of the gene is called an allele. So, an allele is a gene, but it is a more specific term. A gene that has more than one allele is termed polymorphic. In this example, the gene is for eye color; the alleles would be a blue allele, a green allele, and so on. Some genes have only one allele, in which case the words gene and allele are interchangeable. Some may have two or more. The proportion of each allele in the population, however, is not equal, in that one allele may be much more common than other alleles. For example, 70% of the alleles for one trait might be A, 20% B, 8% D, and only 2% E. Figure 3.1 is a simplified picture of a pair of chromosomes, showing just one gene and two different alleles. Remember that this is a great simplification, because many genes usually govern a single trait. Also, every chromosome has many genes. There also may be more than two alleles in the population, as there are in our eye color example, but each person gets only two because they only have two chromosomes in each pair. So, although homologous chromosomes are very similar, they are not identical, because they can contain different genetic information. We inherited one chromosome of each pair from our mother and one from our father. Using the eye color example, remember that the father also has two copies of that gene and so does the mother. Your father might have a blue allele and a green allele. He passed one of them to you. Your mother might have a blue one and a brown one. She can only give you one as well, so you might have

Figure  3.1  A  simplified diagram of a pair of chromosomes, showing just one gene and two alleles. (From SFU Publications, Burnaby, Canada.)

Introduction to genetics  53

ended up with two blue alleles, a blue one and a green one, a brown one and a green one, or a blue one and a brown one; that is, you could get any combination of the above alleles. When we reproduce, we pass on our genes to our offspring. However, we cannot give a complete copy of all 46 chromosomes, because our mate would do the same and the offspring would end up with 92 chromosomes. An egg and a sperm must fuse to produce a baby. If the egg gave the baby 46  chromosomes and so did the sperm, the fetus would not  be viable. It  must have exactly the right number of chromosomes, with exactly the right number of genes on them to survive. Babies with extra sets of chromosomes do not survive. A single extra chromosome, or even just an extra little part of a chromosome, can result in death or produce an abnormality, 2 for example: ■■ ■■ ■■

Down syndrome (extra chromosome 21, physical and mental problems) Patau syndrome (extra chromosome 13, death usually after a few months) Edwards syndrome (extra chromosome 18, death in less than 6 months)

Meiosis To produce a healthy baby, the parents must be able to halve their chromosome number precisely before they pass them on. So, although each parent has 46 chromosomes, their gametes, or sex cells (the eggs and the sperm), must have only 23 chromosomes. When the egg and sperm fuse to make a single complete cell, or zygote, each of them provides 23 chromosomes, and the zygote now has the entire set. This zygote grows into a baby and then an adult by repeated cell divisions, like a photocopier making copies of that first original cell. As the organism grows, every cell contains the same genetic information. When something has only one set of the chromosomes, it is called haploid. The normal condition is diploid, which just means that there are two copies of each chromosome. For a baby to be viable, it cannot just get a random half of the genetic material from each parent. It must get one copy of every gene from each parent. This is the reason our chromosomes come in pairs. When the parent is producing sperm or eggs, one member of each pair goes to each new cell. This ensures that each gamete (egg or sperm) has a full set of genetic information from each parent. Natural selection had to develop a very precise mechanism to cut the chromosome number in half and to ensure that every egg or sperm has exactly the right number and type of chromosomes. This process is called meiosis, and it does two things: it carefully halves the chromosome number, and it also shuffles the chromosomes so that the offspring are completely different from their parents. This process occurs in the ovaries or testes, and it results in the chromosomes being first duplicated, then divided, so that each cell eventually becomes, for example, four sperm. During meiosis, the chromosomes briefly stick together, and homologous chromosomes actually exchange bits of information exactly, gene for gene, before they split. At the end of the entire process, there are four cells, each with half the chromosome number and a unique set of chromosomes. There are two major results of meiosis: ■■

■■

The  pairs of chromosomes have been separated, so the number of chromosomes has been reduced from 46 to 23, with every chromosome still represented. All new cells are different from each other and from the parent.

Thus, each sperm and each egg are unique. This is the origin of all the genetic variation in our species. Why do we want that variation? Why, if you have a successful genotype, would you want to risk changing it? If it ain’t broke, why fix it? The  obvious answer seems to be that we would not want to take the risk. Although yours might be a successful genotype now in this environment, environments and situations do change. Your genotype might not be successful in the next generation. Remember the peppered moth! You want variety in your offspring so that if things change

54  Genetic principles

drastically, then at least some of them will survive. We need variation in the population to allow natural selection to act and enable our population to adapt to changing environments. Every human being is unique. A human egg represents one of about 8 million possible chromosome combinations, and it is fertilized with one sperm, which also represents about 8 million possible chromosome combinations. Therefore, any two parents will produce a zygote with any of at least 64 trillion combinations; as a result, you might not look much like your siblings. Genes code for, or instruct the body to produce, a protein that does something; it may produce some metabolic action, it may determine that a person is blue eyed, or it may affect a behavior. In traits as complex as behavior, a very large number of genes contribute to the phenotype or expression of the behavior, each contributing a small part.

Genetics: The study of patterns of inheritance We have long known that offspring inherit qualities from their parents. Farmers, for example, have known for centuries that breeding a heavy animal with another heavy animal will likely produce heavy offspring and that the daughters of good milk-producing cows will be good producers as well. This is really just positive eugenics. If you had a particularly good sheepdog, you would breed it with another good worker, so that you always had good working dogs. If you were a sheep farmer, you would not breed a dog that was scared of sheep. So, people have understood the basic concepts of genetics since time immemorial, but the way the genetics of inheritance actually works was not understood until much more recently. In the past, people believed that each offspring inherited a blend of its parents’ features, a bit of both, because the offspring appeared to be a mixture of both parents. Even Darwin believed this premise. This idea seems logical when you think about things that are continuously graded (you might be halfway between your parents’ heights, for instance), but it does not account for most observations. If this blending theory were correct, variation would disappear. Think of different colors of paint. If you mix black and white, you get gray; if you then mix gray and gray, you get more gray. In the end, you lose the color distinction. This theory works for paint but not for organisms. If you bred, or “mixed,” a black horse with a white horse, this theory supposes that you would always get a gray horse. But that is not  true; the pair would actually produce black foals, white foals, and lots of intermediates. If inheritance worked by blending, it would eventually result in complete uniformity. Everyone would be the same height, weight, color, and so on. People tended to ignore this obvious evidence that the blending theory was incorrect because they could not explain it. In fact, the theory does not work even for continuously graded characteristics such as height. If we were always an average of our parents, we would all be the same height in just a few generations.

Mendel’s experiments Gregor Mendel was an Austrian scientist and an Augustinian friar who is widely considered to be the founder of modern genetics. He carried out the first scientific investigation of inheritance in the 1860s. He did exhaustive experiments and came up with the principles of genetics that are still valid today, yet he did not know that genes or chromosomes even existed, because the science of microscopy was not yet at a level to make such features visible. So, his work was truly remarkable. We still teach directly from his work, with some additions but no real corrections—there are not many pieces of research, in any field, about which you can say that over 150 years later. Some scientists still say that his was the best scientific paper ever written. Genetics is the study of genes and how they are passed on to the next generation. Mendel worked on pea plants, and he did very careful experiments, breeding his plants for more than 2 years before even starting to use them in experiments. He wanted to be sure that they bred true, and he used

Genetics: The study of patterns of inheritance  55

very large numbers, so that he could perform statistics on his data, which, although understood today and required in scientific studies, was a very unusual concept at the time. One of his first experiments involved crossing pea plants that always produced only purple flowers with pea plants that always produced only white flowers.3 The original plants are referred to as parentals—we call them the P generation: ×

Parentals, P generation Purple flowers

White flowers

⇩ F1 generation

All flowers purple

The  first generation, which is called the F1 (or first filial) generation, had only purple flowers. 3 So, purple had completely obscured the white. Thus, no blending was taking place at all—no pale purple or pink. This finding alone disproved the theory of blending. We call the trait that obscures the other one the dominant trait, and the concealed trait, the one that did not appear in this first cross, the recessive. So, in this case, purple is dominant to white. These were Mendel’s terms, and we still use them today. For example, brown eyes are dominant to blue eyes in people; blue eyes are recessive. Mendel then let the first set of offspring self-pollinate. Many plants do that normally because they have male and female parts and can fertilize themselves. In the next generation, he obtained purple and white flowers in a ratio of 3:1, so the recessive trait, the white, had not totally disappeared.3 It  had just been “hidden” for one generation and reappeared in the second generation. It was unchanged—still white, not mixed with purple at all—so again, no blending. It appeared, then, that the pea plants contained two hereditary factors for color. We now understand that these hereditary factors are genes. The pea plant has genes for flower color, and they come in two varieties (or alleles): purple and white. Each pea plant received one allele from the female parent and one from the male parent. Each gamete contained only one of the two possible alleles, and it could be either purple or white. When two gametes get together, the offspring have two alleles, one dominant to the other. This is the previous cross, giving letters to represent alleles. We traditionally use an uppercase letter for the dominant trait (in this case purple) and a lowercase version of the same letter for the recessive trait: Parentals, P generation

×

PP

pp

⇩ F1 generation

Pp

Both the original plants are pure breeding, meaning that both alleles of the gene are the same in each parent. One has two dominant alleles (PP), and one has two recessive alleles (pp). Now, we do a simple cross. If one parent has two Ps, it can only give a P to its offspring. The same for the white parent: it has only ps, so that is all it can give. Therefore, the offspring of the two must be Pp. Parentals, P generation Purple (PP) Gametes

×

P  P

White (pp) p  p

⇩ F1 generation

All Pp (purple)

Thus, in this case, one dominant allele is received from one parent and one recessive allele is received from the other parent. All the offspring will have one dominant and one recessive allele.

56  Genetic principles

We call this outcome heterozygous. Homozygous occurs when the offspring has two copies of the same allele for a gene, either dominant or recessive, and heterozygous occurs when the organism has two different alleles for that gene. So, in this case, both parentals were homozygous and all the offspring were heterozygous. Then, Mendel interbred this first generation, crossing Pp with Pp. Each parent donated one of two alleles to its gametes, either the dominant allele P or the recessive allele p. The following table (known as a Punnett square) shows all the possible combinations: P

p

P

PP

Pp

p

Pp

pp

The dominant gene always produces the color purple, so if an uppercase P is present, the flowers will be purple. It can only be white if no uppercase P is present. Three of the above (PP, Pp, and Pp) will be purple, and only one (pp) will be white (because it has no uppercase P). When you look at this table, you can see how the white—the recessive—reappears and that the purple and white are in a 3:1 ratio. It is useful to draw these tables when you are working out the possible outcomes of a cross. You can notice here that there are two different genetic makeups (or genotypes) that produce purple. They can be PP or Pp. Although they look identical (their phenotypes are both purple), genetically, they are different, which affects the way they pass on their genes. One has two dominant alleles (two Ps), so it can only pass on dominant alleles. The other has one dominant and one recessive allele (Pp), so it can pass on either. Dominant and recessive genes are very important when a doctor works out the risk of parents passing on a genetically controlled disease to an offspring. It may be hidden in the parents but could appear in an offspring. Genetic counselors determine the probability of a future offspring having the disease. In this cross, the heterozygous Pp looks exactly the same as the homozygous PP. They are both purple. Now we can differentiate between the genotype and the phenotype, which are not necessarily the same. The phenotype is the physical manifestation of traits, or what the organism looks like or does. In  this case, is it purple or is it white? The  genotype is the genetic makeup of the individual, that is, which alleles it possesses. So, PP and Pp have the same phenotype but different genotypes. PP has the same genotype and phenotype—purple—and pp has the same genotype and phenotype—white. Mendel published his work in the 1860s, but it was completely ignored. It was rediscovered in 1900 by three different people at about the same time, after the improvement in microscopes had resulted in the discovery of chromosomes. Then, Mendel’s work was discovered to contain a tremendous amount of data. By the time people began to recognize Mendel’s work for what it was, sadly, he had been dead for 16 years. Later, his work was applied to natural selection, which had, by then, faded somewhat in popularity. However, when Mendel’s work was put together with Darwin’s work, natural selection soared to new heights because it was now possible to understand how natural selection worked. Darwin had been right all along; he simply had not known about genetics. If he had, both he and Mendel could have understood the process of natural selection and inheritance, and Darwin, for one, would have died a happier man. Thus far, we have been looking at a single gene that guides one characteristic, flower color. Every living thing, however, has millions of genes, each guiding something. Sometimes, several genes act on one trait. In a real mating between two living things, the cross shown above would happen for every gene, so you can see how quickly it gets complicated. There are quite a few genes that behave in this straightforward Mendelian, dominant/recessive form, but most genetic traits are not that simple. We will look at a few slightly more complicated examples, so that you realize how some of the traits discussed later could be affected by genes. Mendel provided the first understanding of genetics, but he was lucky in choosing a situation in

Genetics: The study of patterns of inheritance  57

which the genetics was straightforward. Otherwise, he would not have understood it. As discussed, most genetics is much more complex, and any genetic influence on behavior will definitely be very complex, not the simple “one gene/one action” described above. Here are a few of the different types of genetic inheritance patterns.

Non-Mendelian types of gene interactions and inheritance patterns Linked genes

What happens if we look at two genes for two different things? If they are on different chromosomes, each of them will act independently, but if they are on the same chromosome, they are said to be linked. This can make it more likely that they will be inherited together. For example, if you consider two genes, A and B, which are on two separate chromosomes, and you crossed AABB with aabb, the first cross would result in offspring that are AaBb, like this: Parentals, P generation

AABB

×

aabb

⇩ F1 generation

All AaBb

If they were interbred, the next generation would look like this: AB

Ab

aB

ab

AB

AABB

AABb

AaBB

AaBb

Ab

AABb

AAbb

AaBb

Aabb

aB

AaBB

AaBb

aaBB

aaBb

ab

AaBb

Aabb

aaBb

aabb

However, if genes A and B are on the same chromosome, they will not segregate independently; that is, they will stay together. The first cross will look like this: Parentals, P generation

AB AB

×

ab ab

⇩ F1 generation

All AB ab

From a phenotypic point of view, everything seems the same at this stage, but the gametes that are produced by the F1 generation are very different. Because they are linked, most of the gametes must be either AB or ab, because the loci are physically tied together on the same chromosome. They are part of the same DNA molecule. So, the next generation would more closely resemble the single gene cross seen with purple and white pea plants, despite the fact that two genes are involved: AB

ab

AB

AABB

AaBb

ab

AaBb

aabb

They might separate when the shuffling occurs in meiosis, but the closer they are together, the more likely they are to stick together. If the genes are very close together, it is unlikely that there will be

58  Genetic principles

a shuffle occurring between them, and they will almost always be inherited together. If they are at opposite ends of the chromosome, however, they could get separated when the chromosomes get stuck together and swap portions. So, one might get a few Ab aB gametes if crossing over occurs between the two loci, but these will be rare, and the closer the two genes are on the chromosome, the rarer they will be. This is the reason that people with dark hair usually have dark-colored eyes and people with blond hair usually have light-colored eyes. Many of the genes for eye color and hair color are on the same chromosome. They may get separated at meiosis, so you may get a person with dark hair and light eyes, but they are much rarer than people with dark hair and dark eyes, because dark hair and dark eyes are genetically linked. When things are linked in this way, a gene that is very advantageous to the organism will be selected for, and along with it, it might carry some other genes that just happen to be closely linked. This is often the reason that certain genes seem to stay in the population, although they do not confer an advantage. They are closely linked to genes that do confer an advantage, and so, when those are selected for, they just sort of tag along. It also explains why some quite bad genes, such as those that cause certain diseases, remain. The desired gene outweighs the bad effects and is still selected. Sex-linked traits

Of the 46 chromosomes that a human being possesses, one pair (two chromosomes) is the sex chromosomes. The female has two X chromosomes and the male has one X and one Y chromosome. When eggs are produced, one sex chromosome goes to each egg. A female can only provide an X for the eggs. The male, however, has both an X and a Y, so a sperm might carry either one. The egg with which it meets up will be carrying an X, so the sex of the offspring is determined entirely by the male and depends on whether the sperm that he contributes carries an X or a Y. Thus, although Henry VIII beheaded and divorced wives because they did not produce sons, the sex of the child depends on the male, not the female. Certain characteristics of the female reproductive system can favor a sperm that carries a Y or an X chromosome, meaning that the female may have some influence on the sex of the offspring, but sex is primarily determined by the male. The X and Y chromosomes not only determine sex but also carry other genes. The X chromosome is quite large and carries many genes, but the Y chromosome is very small and carries only a few genes. If a gene is on a sex chromosome, it is said to be sex linked. Usually, this means that it is on the X chromosome, simply because it has more genes. We know about a few such sex-linked genes, and some cause diseases in humans. We know the most about genes that cause diseases, because they are of most interest to us. Genes that are involved in hair or eye color are not so vital and therefore do not get the same research attention. Two classic examples of sex-linked conditions are color blindness and hemophilia.2 They are both usually found in men and only very rarely in women. It is quite common for men to be color blind but not for women, and the same is true for hemophilia. The reason is that both diseases are caused by recessive genes. The dominant form is normal. If you think back about the earlier examples relating to dominant and recessive genes, you will see that people only express the recessive gene if they have two recessive alleles. If they had two dominant alleles, or one of each, then the recessive gene would not be expressed, just as the purple flower “hid” the white flower. Also, these disease-causing recessive genes are usually fairly rare, so the allele ratio may be 98% normal dominant and 2% disease-causing recessive. Except for the sex chromosomes, a person always has two of each chromosome. With the sex chromosomes, females have two X chromosomes but males only have one. This means that the female can have one recessive gene that will not  express itself because it will be protected or hidden by the dominant one. However, a male only has one X chromosome, so if he receives the recessive allele, there is not another allele to hide it, and it will be expressed. As a result, the recessive allele is expressed in the male, while the female is normally only a carrier. The female will only actually have the disease if she gets a recessive gene from each parent, which is rare. There are lots of such traits, not just disorders, that are seen more commonly in males than in females. Females who are carriers can pass recessive genes to their sons, who will express them, but their daughters will only be carriers, like their mothers.

Genetics: The study of patterns of inheritance  59

There are some genes on the Y chromosome, but not many. One causes a disorder called scaly bark skin disease. Females are not carriers because it is only found on the Y chromosome. Incomplete dominance

In the experiment with the pea flowers, the offspring always looked like one of the parents because of the complete dominance of one allele over another. But for some traits, dominance is not complete, and the offspring have an appearance somewhere between the phenotypes of the two parents.2 For  example, when a red snapdragon is crossed with a white snapdragon, all of the first generation has pink flowers. This is because the heterozygotes have lesser red pigment than the red homozygotes. However, it is important to realize that blending is still not occurring, as was the old belief, because if that were so, the original whites and reds could not be retrieved. However, if you breed the pink plants, you get red, white, and pink. Codominance

In codominance both alleles are expressed in the heterozygote. An example is the existence of three separate blood groups in humans—M, N, and MN.4 These groupings are based on two specific molecules that people have on the surface of their blood cells. People of group M have one type of molecule, people of group N have the other type, and people of group MN have both types. M individuals are homozygous for one allele, N individuals are homozygous for the other allele, and MN individuals are heterozygous. The MN is not intermediate, as both alleles are expressed. Pleiotropy

Thus far, we have looked at genes that control just one phenotypic character. However, most genes actually have multiple phenotypic effects, which means that they control several things that affect the way we look or act. This situation is called pleiotropy.2 For example, in tigers, the same allele causes both abnormal pigmentation and crossed eyes. You can see a similar effect in Siamese cats: the same allele that is responsible for their light body color and darker points (face, paws, and tail) is also responsible for crossed eyes. This is one of those unnatural selections that humans have had a hand in. We liked the color combination, so we deliberately selected for it, and in so doing we also selected for crossed eyes. These are clearly not a benefit and certainly would not help the animal survive in the wild. However, Siamese cat breeders highly value cross-eyed animals, and in their protected environment, they survive well. Epistasis

Epistasis occurs when one gene alters the way another gene is expressed. An example is coat color in Labrador retrievers. These dogs come in black, brown, and golden. Black coat color (B) is dominant to brown (b), or chocolate, as we refer to it. To be chocolate, a Labrador must have two recessive alleles (bb). The two alleles for color produce black or chocolate melanin (dark pigment). Sounds simple, but there is another gene (E) that determines whether the dog will deposit melanin in the coat hairs or not. This E gene is epistatic, controlling the other gene. If the dog has a dominant E, then it will produce the color the allele suggests—either black or chocolate, as melanin will be deposited in the coat hairs. However, if the dog has a recessive e, then this will act on the other gene and prevent it from resulting in the deposition of melanin in the coat hairs, whether black or chocolate, and the resulting dog will be golden. Note, however, that melanin is only blocked from forming in the coat, as the dog still has melanin deposited in the dark nose and eyes; the gene only affects the deposition in the coat.2 Polygenic inheritance

Mendel studied characters that were either/or—the purple and white flowers, as well as the many other flower traits he investigated—but many characters cannot be classified in this way because they vary in the population in gradations. These are called quantitative characters; examples in humans

60  Genetic principles

include skin color and height. Quantitative variation usually indicates polygenic inheritance, which is an additive effect of two or more genes on a single phenotypic character. This is the opposite of pleiotropy, where a single gene affects many phenotypic characters. Here, we have lots of genes working on just one thing. Polygenes that affect a particular quantitative trait are commonly found on many different chromosomes. Genes are very simple, and they code for simple things. Thus, complex things require the contribution of several genes. Even things that we might consider simple, such as hair color or eye color, are coded for by several genes. A good example is human skin color. Humans differ in the amount of melanin found in their skin.5 There is a great variation in the amount of melanin that different people have, but much of this variation is a result of at least three separately inherited genes. There may be more, but for simplicity’s sake, let us just look at three genes, with a dark skin allele for each (A, B, and C), each contributing one unit of darkness to the phenotype and being incompletely dominant over the other alleles (a, b, and c). Someone who has AABBCC would be very dark, and someone with aabbcc would be very light. AaBbCc would be intermediate. If you make a calculation, you will see that there will be a wide range of graded values, from very pale to very dark. You can also see how two people of intermediate skin color could still produce offspring with either very dark or very light skin. Also, skin color is not entirely determined by genotype; we can expose ourselves to the sun to produce more melanin. This is an example of the environment affecting the translation of the genotype into the phenotype. Nutrition, exercise, experience, and other events all alter the phenotype. Nutrition affects size, exercise affects body shape, and experience can affect IQ. Even identical twins are not really identical; although they are genetic equals, they still accumulate phenotypic differences as a result of their unique life experiences. In the example of the Siamese cat, I mentioned the coat color combination—a light body and dark ears, face, tail, and feet. Coat color is controlled genetically, but the environment has a strong effect on the phenotype. These darkened areas on the cat actually have a slightly lower temperature than the rest of the body. Experiments have shown that the Siamese cat has a genotype for dark fur, but it only appears at temperatures somewhat below the general body temperature. If some dark fur is shaved from the tail and the cat is then kept at a temperature higher than normal, the fur that grows back will be light. On the other hand, if the cat is shaved and made to wear an ice pack on a normally light area, the spot that is kept cool grows back with dark hair. So, genotype and environment interact to determine the phenotype of the organism. The effect that the environment has on the phenotype varies dramatically. It has no or little effect on some things, such as eye color, and a major effect on other things, such as behavior. There is certainly a genetic basis for behavior, but the environment has a strong influence. In fact, most behavior is genetically predetermined to be affected by the environment. It is designed to be improved, modified, and changed by learning and experience.

Mutations Much of genetic variation is the result of mutations, or random changes in the DNA  molecule, which can then be passed on to the next generation. From watching too many horror and sci-fi movies, we are inclined to think of mutations as terrible things resulting in monsters, but realistically, they bring about very small changes and may be beneficial, harmful, or neutral.6 An example is that of the Balinese cat. A Balinese cat looks like a long-haired Siamese cat, which is really what it is. The present-day breed is believed to originate from a mutation in a litter of Siamese cats in which some were long haired. When such differences appear in purebred litters, they are usually not what breeders want, even though the mutation may have no deleterious attributes, as they don’t fit the breed description. So, the animals are often euthanized or at the very least not allowed to breed, to prevent the mutation spreading. In this case, the breeder obviously thought that a longhaired Siamese would be attractive and so bred the long-haired kittens to eventually create a new breed. In the wild (if Siamese cats existed in the wild), it is probable that the long-haired mutated allele would not be advantageous or disadvantageous, so it would not be selected against and would therefore stick around. If female cats found it more attractive than short hair in a mate, it would

Genetics: The study of patterns of inheritance  61

be selected for, as it would be an advantage, and more long-haired cats would exist. Whereas if it proved to be a hindrance to movement, owing to severe matting, then it would be selected against and fewer long-haired cats would exist in this hypothetical situation. In wild animals, however, some similar mutations can be highly disadvantageous. On occasion, white tigers are born in the wild; this can make the cubs much easier to locate by predators and often results in their early demise.

Recessive alleles in the population Although recessive alleles are hidden by the dominant allele in a heterozygous individual, they are important in the population. In many cases, the dominant form is the normal form and the recessive is the abnormal form. For example, in a disease, the dominant allele would be normal, and the recessive allele would result in disease. However, when the dominant allele is present, the recessive allele is masked or blocked, so the person carries a gene for the disease but does not succumb to the disease. However, the person could pass on that allele to any offspring. It would only be a problem if the person who is a carrier had a partner who either had the disease or was also a carrier. In that case, they might produce a child with only recessive alleles, who would thus have the disease. Another point is that there are often unequal ratios of the different alleles. The population may include 95% dominant alleles and only 5% recessive alleles. In this case, the chance of two people with recessive genes coming together is low, so the chance of producing offspring with two recessive genes (and hence the disease) is also low. Recessive alleles explain why there are taboos against marrying close family members. Cultures throughout history and prehistory have banned close family members from marrying. These early cultures did not understand genetics but did know from experience that incest frequently resulted in unhealthy offspring. When you understand genetics, it is obvious that there is a higher chance of finding someone else with the same recessive gene in your own family. For  example, say the gene that causes some horrible disease is ‘a’, the recessive form of the gene ‘A’. A is found in 98% of the population and a in only 2%, so it is very rare. The mother is a carrier, so she is Aa; the father is normal (AA). They have eight children, who will be AA or Aa. Therefore, none of the children have the disease. In fact, the family probably does not know that the mother is a carrier or that half of the children are carriers. If their partners are strangers, the chance of meeting another person also carrying a recessive a is small, because the frequency of that gene in the population is small. It is entirely probable that the family is totally unaware of this potential time bomb in its midst. However, if siblings mate, then the chance is 50% that the mate also has that gene. So, you could easily get a cross between two carriers. Like the purple and white flowers, the result would be a 3:1 ratio—that is, a 25% chance of producing a child with the disease and a 50% chance of producing more carriers. Recessive alleles and disease

A classic example of a recessive allele for a disease is hemophilia, a clotting disorder whose sufferers can bleed to death very easily.2 Queen Victoria was a carrier, carrying the gene on her X chromosome (hemophilia is a sex-linked disease); two of her daughters were carriers and one of her sons had the disease. He transmitted it to his daughter, who did not have the disease but was a carrier. Because royal families are very inclined to marry close relatives, the disease became common in virtually every royal family in Europe, except—conveniently—Britain. Victoria’s daughter Alice was a carrier and passed the gene to her daughter, Alexandra, who was also a carrier. Alexandra did not have hemophilia, as she had a normal X chromosome from her father. She married Czar Nicholas of Russia and passed the X chromosome carrying the gene for hemophilia to their only son, Alexei, who expressed the gene, as he only had one X chromosome and so had the disease. Because of her son’s illness, Alexandra fell under the sway of the monk and mystic Rasputin, which

62  Genetic principles

was in part the reason for the fall of the Russian royal family and the massacre of all the members of the immediate family. Had Alexandra by chance passed on the normal X chromosome, her son would have been healthy, and the course of history may well have been different. Another example is sickle cell anemia.2 This disease of the blood is common in people of African descent and much rarer in others. The red blood cells of a person with sickle cell anemia cannot carry oxygen very well, so people suffering from sickle cell anemia develop a multitude of health problems, and the disease is usually fatal. It occurs in about 1 in every 400 African Americans, which is a fairly high rate. Why would something like that be more common in one ethnic group than in another? Remember that there is a higher chance of producing one of these disorders if you marry close family members, and this concept can be expanded to show that this would also apply if you marry only within a particular group of people. Most people do marry within ancestral groups, and sometimes even closer—within a village or regional area, for example. Therefore, there is a higher chance that they will meet someone else with the recessive gene. The frequency of the alleles in different ancestral groups is different. It may be, for example, 20% in the African American population but only 2% in other ancestral groups. Therefore, your chances of meeting someone with a recessive gene are higher if you marry within your own ancestral group. The same is true for Tay-Sachs disease, which is common in Jewish people, but only in those whose ancestors come from Central Europe, a group called Ashkenazic Jews. There are many other such diseases; for example, porphyria (the disease that is thought to have resulted in the belief in vampires) is more common among Caucasians. An interesting point is that sometimes natural selection has selected for a disease because it confers some sort of advantage on the individual. In the case of sickle cell anemia, you only have the disease in its severe form if you have two recessive genes, or ss. However, if you are a heterozygote for this condition (Ss), you have some normal red blood cells and some abnormal ones. Normally, this condition would still not be good, but most of the people who have this disease are from Africa, where the disease malaria is a common killer. Malaria is caused by a parasite that lives in red blood cells. People who are Ss are much more resistant to malaria than people with normal blood (SS). Malaria kills millions of children every year in Africa, so a child in Africa who is heterozygous for sickle cell anemia has a much better chance of surviving malaria than one with normal blood. Therefore, that child is more likely to survive, grow up, have children, and hence pass on those anemia genes. Thus, the disease actually confers an advantage in regions of the world where malaria is prevalent. Descendants of these people who live in North America do not need to worry about malaria, so it is no longer a benefit but is instead a detriment, because people who are heterozygous have trouble breathing at high altitudes. Therefore, the frequency of the disease in the population in North America may eventually drop. However, it is a good example of how something that might appear to be bad can actually be selected for by natural selection. Why aren’t we perfect?

If natural selection has been shaping our species and all other species forever, why are we not perfect? In other words, why has natural selection allowed us to still have genetic flaws such as diseases and antisocial behavior? There are several reasons. The  first reasons relate to “what is perfect?” What is perfect in one environment may not  be perfect in another, and our environment keeps changing. Therefore, we and all other species are forever trying to keep up, adapting to new situations. If a population is perfectly adapted to a particular environment and then that environment changes, the population will be very badly adapted to the new environment and may die out. A population needs variety within traits, so that when changes occur, some members of the population will be better adapted than others and will survive. Remember Darwin’s finches. The  second reason relates to recessive alleles and heterozygosity, which means that recessive alleles may be hidden, even if they are deleterious. We often ask, why don’t recessive alleles that sometimes cause problems, such as disease and antisocial behavior, die out? Let us look first at a very harmless example. There  are many more

Genetics: The study of patterns of inheritance  63

dark-haired people than blonds in the world. As people travel and emigrate, they mix everywhere. Despite the fact that blond hair is recessive to dominant dark hair, it does not die out. There are still blond people. This is because there is no downside to being blond; in other words, there is no selective advantage for either hair color. So, if a dark-haired person mates with a blond, they could have blond or brunette children. If two dark-haired people mate, the children could still be either, depending on the genotypes of the parents. If both are heterozygous, then the children could be blond. Blond is not lost, just hidden for a generation—remember the purple and white flowers. But what if it were a disadvantage to be blond? Say a new disease appeared that only killed blonds. What then? Would blonds survive? Well, it would certainly have an influence on the frequency of the allele, just as we saw the frequencies of the alleles for different beak sizes in the finches’ shift. But blond alleles would not die out. Why? There could be many reasons: ■■

■■

■■

The genes for blond hair may be linked to—that is, located on the same chromosomes as— genes for very advantageous characteristics that were inherited, so when the advantageous genes were selected by natural selection, the blond genes just got carried along. Mutations can occur. Even if the allele became low in frequency, it often might pop back up as the result of mutation. The gene that causes the trait can mutate to bring the trait back. When people are heterozygous for a gene, they usually do not  show the recessive trait. It  is masked by the dominant one. So, even if there was a selective disadvantage to having a trait, natural selection can only act on it if it is expressed. In other words, natural selection acts only on the phenotype. That is, if there were a disadvantage to being blond, natural selection would act on the blonds themselves, the double recessives, but it would not act on the heterozygous people, those who have an allele for dark hair and one for blond, because they are phenotypically dark haired. In this way, the allele is hidden and protected from natural selection. When two heterozygous people get together, some of the children will be blond, so the trait is never lost. Also, being heterozygous (with one of the dominant and one of the recessive alleles) might actually be more advantageous evolution-wise, as in the case of sickle cell anemia. In that case, it is better to be heterozygous.

This also explains why a trait that was perhaps very advantageous in the past—such as aggression, which might have helped to protect one’s offspring and obtain resources in the past but is maladaptive now—will stick around, for any of the reasons listed previously. The trait will still appear, although it is now no longer advantageous to possess it.

Gene expression The above text gives a very simplistic introduction to genetics, but now we need to delve a little deeper. It  was thought for decades that genes simply code for proteins, but it is actually a lot more complex than that. The actual expression of the gene includes many steps and facets, and it is greatly impacted by the environment in which it is exposed; in other words, our genes are designed to be influenced by our environment. This  is a basic tenet of genetics—that there is a certain plasticity to our gene expression. There  are many examples of this. Children who are genetically predisposed to be tall and robustly built as adults may be short and slight if they do not receive adequate nutrition in childhood. A more complex example is that of speech. We are all genetically designed to be capable of speech. It is a learned trait, but it requires the genetic backdrop of numerous features, including brain, vocal cords, tongue, lips, and so on. However, it also requires exposure to the human voice. Even when children have all the requisite physical features for speech, if they are not  exposed to human speech, they will not  be able to speak. Moreover, this is time-sensitive, in that there is a specific window of time during which the child must be exposed. Children who are not  exposed to speech until later in life will not  be able to learn to speak, despite high levels of intelligence.

64  Genetic principles

Heritability Heritability is a term that is commonly used in behavior genetics studies. It is a measure of the proportion of the variance or variability in a trait that can be attributed to genetics, as opposed to the variance attributed to the environment.7 It is usually presented as a percentage or on a scale of 0–1 (heritability coefficient). For example, a heritability of 60%, or 0.6, suggests that 60% of the variability for that trait in a population is under genetic control. This can be somewhat misleading, as usually, a large number of genes are involved, each contributing a differing amount of genetic control to the trait. Some may only have a very tiny contribution, and others much more, but the heritability statistic does not distinguish this. Moreover, in many cases, there is an interactive effect between the genes and the environment. Perhaps, most importantly, it is often forgotten that the term refers to the heritability of a trait within a population, not within an individual.8 Therefore, heritability estimates the amount of variation of the phenotype of a trait that is due to genetics, between people or organisms in a population. Mendelian genetics only works on genes that are polymorphic, that is, those that have more than one allele. So, it is possible for a trait to be completely (100%) inherited yet have a heritability of 0%.8 Burt and Simons8 give the example of the number of eyes possessed by humans. Having two eyes is of course inherited, but there is no variation in the number of eyes we can have. Therefore, having two eyes is 100% inherited but has a heritability of 0%.8 Heritability is population specific and temporally specific, so it should not be used or misused in an attempt to explain differences between populations. Moreover, a trait can have high heritability but still be greatly impacted by the environment.8

Conclusion We can now see why appalling policies such as negative eugenics could never have worked. Even ignoring the very obvious moral and ethical reasons why eugenics is wrong, scientifically, it could never work. Megalomaniacs such as Hitler are not likely to be affected by ethical or moral reasoning, but scientific reasoning can be much more persuasive, as it can show that the policies, however warped, will not work. Even if a trait chosen for eradication were purely genetic (and remember that the traits chosen by Hitler and other genocide perpetrators, such as poverty, were not genetic), killing people with that trait would not eliminate it, for the reasons listed previously (linked genes, mutations, heterozygosity). This is why it is so imperative that right-minded people understand basic science and genetics. Then, when people attempt to enact eugenicist policies, such as North America’s racist policies of the early and middle twentieth century, moral, ethical, and scientific reasons can be given to counter them. By misquoting science and claiming a scientific background for theories that were not backed by any science, politically active people brought about the implementation of laws that violated the civil rights of all peoples. As the general public did not understand science, they could not see that the rhetoric being spouted was completely flawed. Had enough members of the general public understood genetics, perhaps some of the horrors of the past could have been avoided.

Questions for further study and discussion 1. Assuming that it is hereditary, why would something unpleasant, such as morning sickness during pregnancy, actually be selected for by natural selection? 2. Why does natural selection act only upon phenotype? What are some of the implications of this? 3. Explain why people used to believe in blending of traits, and why, even before Mendel, we should have been able to understand that blending was incorrect. Then explain how Mendel proved that blending was wrong. 4. Explain, from a scientific perspective only, why negative eugenics will not work to eliminate a trait.

References  65

References 1. Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A. et al. 2010. Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet. 6(5): e1000934. 2. Griffiths, A.J.F., Wessler, S.R., Carroll, S.B., and Doebley, J. 2015. Introduction to Genetic Analysis. 11th ed. New York: W.H. Freeman & Company. 868 pp. 3. Mendel, J.G. 1866. Versuche über Plflanzenhybriden. Verhandlungen des naturforschenden Vereines in Brünn, Bd. IV für das Jahr, 1865, Abhandlungen, 3–47. Translated into English by C.T. Druery and William Bateson. 1901. Experiments in plant hybridization. J.R. Hortic. Soc. 26: 1–32. 4. Lamb, B.C. 2000. The Applied Genetics of Plants, Animals, Humans and Fungi. London: Imperial College Press. 300 pp. 5. Griffiths, A.J.F., Wessler, S.R., Lewontin, R.C., Gelbart, W.M., Suzuki, D.T., and Miller, J.H. 2005. Introduction to Genetic Analysis. 8th ed. New York: W.H. Freeman & Company. 868 pp. 6. Beaver, K., Nedelec, J.L., Schwartz, J.A., and Connolly, E.J. 2014. Evolutionary behavioral genetics of violent crime, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer. pp. 117–136. 7. Carey, G. and Gottesman, I.I. 2006. Genes and antisocial behavior: Perceived versus real threats to jurisprudence. J. Law Med. Ethics 34(2): 342–351. 8. Burt, C.H. and Simons, R.L. 2014. Pulling back the curtain on heritability studies: Biosocial criminology in the postgenomic era. Criminology 52(2): 223–262.

4 Misconceptions, experimental design, and behavioral genetics

Introduction When most criminologists consider the words biology and crime together, they assume that they refer only to genetic issues. As much of this book demonstrates, this is clearly not true. Nevertheless, much work has been done in the area of genetics that is relevant to criminology. As this is a fairly large topic, it will be covered over the next few chapters. The  first object of this introduction is to explain the important misconceptions concerning genetics, behavior, and crime (one of the earliest of which surrounds the so-called XYY man). With the brief study of genetics in the preceding chapter, we can now begin to undo some of these misconceptions. There are many myths that surround the relationship between genetics and behavior, and most have no basis in fact or at best relate more to the effects of the environment than to genetics. To understand the experimental research discussed next, it is also important to understand how to design a good experiment to test a hypothesis. Some purported genetic studies are weak, both in conception and in experimental design. As we study some examples, it will become apparent how difficult it is to conduct such studies on humans. Aside from the moral and ethical issues in behavioral research, no two humans are exactly alike, and this makes generalizations about genetics and behavior, even from simple observational research, extremely difficult. In  the final part of this chapter, we explore the implications of this stubborn fact for comparing the behaviors of identical and fraternal twins. Twins, as we will see, provide us with a natural way to separate the effects of genetics and the environment on behavior. We will then consider comparing adopted children with their biological and adoptive parents.

Some misconceptions about genetics There are many misconceptions about genetics and what it does and does not do, which have colored people’s views of its relationship to behavior, and so, it is important to clarify these before we move on.

Animal cloning The  idea that there is a genetic basis for crime is perhaps the most controversial topic in the field of biological influences on criminal behavior, but the controversy arises primarily from

67

68  Misconceptions, experimental design, and behavioral genetics

misconceptions about genetics. One of the greatest misconceptions relates to cloning, both in animals and humans, and to what an actual or hypothetical clone might be. Dolly, the cloned sheep, is a classic example of such a misconception. Dolly was a Finnish Dorset sheep who was cloned in Scotland in 1996; she died in 2003. Dolly was not conceived naturally between a male and a female sheep but was instead cloned from a single mammary cell from her mother. This  cell was then treated as a “test-tube lamb” and implanted in the mother’s womb. The experiment was a major medical breakthrough, not because the researchers were able to create a new individual from another one without benefit of egg and sperm (we have been able to do that for a long time), but rather because they used a cell that was already differentiated: it had already become a mammary cell, dedicated to mammary functions and nothing else. All the other functions of the body were turned off. When an egg and sperm of any species fuse, they create a single cell, and at that stage and for the first few divisions, each cell is capable of performing every function in the body. Each cell must be originally capable of performing all functions because even in an adult, every cell is genetically identical in a single person. But as the zygote grows into a fetus, the cells become specialized to do special jobs, such as being a nerve cell or a cell that secretes digestive enzymes. They then are said to have become determined and only perform as a specialized cell. However, as all cells are genetically identical in the body, containing the same DNA, each cell still has the blueprints for all functions. Functions other than those for which the cell has become specialized are turned off. The Dolly project proved that although cells turn off the capacity to do other things, they can still be provoked to do them again if necessary. This was a big breakthrough medically because it may mean that cells could be persuaded to take over the function of a damaged organ, such as a kidney or even the spinal column. That is, the research was related to medicine and the possibility that, in the future, people could regenerate their own cells, perhaps regenerating a part of the spinal cord or a damaged liver. However, the public and the media did not understand the research and the amazing opportunities it may lead to one day for so many people. They also believed that this was the first time an animal had been cloned, when in fact the first animals to be artificially cloned lived over 70 years ago. A furor arose about the horror-movie idea of cloning people, such as an army of white supremacists. Of course, the big question brought up regularly in the tabloids was, what if someone tried to clone Hitler?

Human cloning So, what if we did try to clone Hitler? Hypothetically speaking, could we? We will gamely ignore the basic fact that cloning requires some tissue and that Hitler was burned beyond recognition in a bunker in 1945. Even if some loyal supporter had managed to find some tissue, the preservation methods of the time were poor, so this is all a moot point. However, while we have the technical potential to genetically clone a person, we are not there yet. Moreover, what would be the point? The aim of cloning research is to work medically with individual cells in attempts to heal people with damaged organs or nerves. However, just for the sake of argument, let us consider, hypothetically only: if we were to clone Hitler, or any other evil person, today, would we have to be concerned? Think about this idea before reading further. It is fiction that Hitler could be recreated by cloning, but the fear clearly implies that many people believe that his criminal behaviors were genetic and almost entirely genetic. This is patently not true. The truth is that if Hitler were cloned, there would be very little to worry about. There is a general belief, perpetuated by television and film, that when a person is cloned, the result will inevitably be a carbon copy. We expect another identical adult, fully formed, with brain, mind, and behaviors intact. This cannot happen. Dolly was created in a test tube and was implanted in her mother, who later gave birth to Dolly in the normal manner, and she began her life as a lamb, not as a full-grown sheep. But she would not have exactly the same Dolly-sheep behaviors as the original.

Some misconceptions about genetics  69

To see why this is, consider another example. Imagine that a family lost an 18-year-old daughter in a car crash and wished to clone her. Despite today’s headlines, they would not end up with a brand-new identical 18-year-old daughter. The reasons are strongly environmental: ■■

■■

■■

■■

■■

Someone would have to carry that fetus for 9 months, give birth to her, and then raise her until the age of 18. She would thus have either a much older original mother or a completely new surrogate mother. With her older (or brand-new) mother, this girl could not be raised in an identical manner to the original; for one thing, the original mother would probably try not to make any of the mistakes she might think she had made in the past. Her immediate family would also influence her in different ways. Let us assume that in life she had a sister who was 1 year younger than her and that they grew up together during the 1960s and 1970s. The new cloned baby would, in contrast, grow up over the next two decades, and although the two might spend a lot of time together, she would be a child growing up with an adult sister, someone who was now at least 20 years old (assuming the dead sister was cloned immediately and allowing for 9 months in utero). This would be a very different sisterly relationship. The sibling rivalries would differ, and at maturity, the relationship would also differ—she would, for example, always be the “baby sister.” The more general social environment would clearly also have changed. The cloned child would be raised amid different political and social values than the original was. The 1960s had a pronounced effect on teenagers and young adults. For the Woodstock generation, social protest was high, anti-authoritarian and alternative roles were stressed, and the heroes and heroines of both political and popular culture were attractive role models. The original daughter would have known the Kennedys; the Beatles; the Marlboro Man; Vietnam; the new McDonald’s chain; Pink Floyd; Sonny Bono and Cher; Peter, Paul, and Mary; marijuana as the illicit drug of choice; and relatively cheap college tuition. These and many other cultural influences would have helped form her wider social values and expectations. Generation X of the 1980s and 1990s had different priorities: first president Bush, terrorism, the first Gulf War, Rwanda, Sting, Madonna, Starbucks, ecstasy as the probable new illicit drug of choice, health gyms, women’s cancer rates, second-wave feminism, women and agency, tattoos, body piercing, narratives of sex abuse, and much higher college tuition were probably important to this generation. Would it even be stylish to smoke or drink? The neighborhood in which the clone daughter grew up would also have a major effect on her life. Would it be a quiet town in northern British Columbia, downtown Toronto, the urban sprawl of California, or uptown New York? If it was the original daughter’s neighborhood, this would also have changed. The income level of her family would also affect her. The clone might grow up with a wealthy family, but the original’s environment could have been relatively poor, especially when her parents were younger.

With just these five social influences, the chances that she will end up exactly the same as the original sister are very low. Genetically, she has the identical genes and is probably pretty much identical in looks to the original. But as a person, will she be the same? No. Her life experiences will have been very different. Some general personality characteristics may be similar—she may still have the same type of passive personality or she may still be moody—but the environment and social situation she grew up in will have had a major effect on all these traits. She will be an entirely unique human being, even more different from the original girl than identical twins would be. Identical twins, even if raised apart, are raised in the same decades, whereas this young girl will have had completely different life experiences, making her absolutely unique. So, what about Hitler, or any other genocidal dictator? A cloned dictator would not be the same as the original. The person who was Hitler was created not only through his genes but also through upbringing, experiences, and environment (as are we all). Strict German schools, a certain family situation, and World War I had a major effect on the formation of the original personality, as did

70  Misconceptions, experimental design, and behavioral genetics

a bankrupt German economy and the Versailles Treaty. A cloned Hitler might strike fear into the hearts of everyone (even those who swear there are no biological explanations for crime), but it would be an unfounded fear—a cloned Hitler would not be the same man.

Does all crime have the same single cause? Another misconception about genetics stems from an overly reductive understanding of genetics. Some people assume that by studying the relationship between genetics and crime, we are trying to find a single cause for all criminal behavior. As we have seen, there is a clear interaction between the environment and genetic inheritance, so such an easy explanation is highly unlikely. At a minimum, we know from motivational studies that anomic stress is common in our competitive culture and that mental illness or somatic dysfunction is often involved in crime; we also know that highly dysfunctional family environments can create antisocial personality types. Person A might kill his mother “because I always hated her getting on my case”; person B might kill her mother “because I want my inheritance”; person C might do it because “she ran off with my boyfriend”; and person D might do it because he or she was mentally ill and thought the mother was a demon. It is the same crime, matricide, but the reasons for committing it are widely different. In the same way, we all might have the same reason for committing a multitude of different crimes. The interactionist and dynamic nature of gene expression tell us that we need to look for multiple causes of multiple crimes. When we look at genetic studies, it is easy to assume that something is a genetic trait when it is really an environmentally influenced trait, although it might be triggered by an underlying genetic trait. For example, a father reunited with his son after many years may be surprised to discover that his son has many of the same mannerisms he does. Traits such as posture may be similar in both men. This could not have been learned, as the two have only just met, so they assume it is genetic. As with most things, such behavior has both genetic and environmental features. The actual posture they use is not  inherited. However, their body shape is inherited, and most of us develop stances that are comfortable for our bodies. Because they both have the same body shape, the posture comes naturally to both of them. The posture is not a result of genetics, the body shape is.

XYY man: Truth and fallacy The relationship between genetics and aggression is often misconceived. The XYY chromosome episode was highly publicized, and media and even early researchers led the public to believe in a “super-male” who was highly dangerous. Let us look at the history, the facts, and the fallacies. All people have 46 chromosomes, 44 of which are somatic and 2 of which are the sex chromosomes. Women have two X chromosomes, and men have one X and one Y chromosome, so women are denoted as XX and men as XY; this is the normal configuration. The Y chromosome is considered the “male” chromosome because only men have it, but they have an X as well. You could say that they have one male and one female chromosome, but this would not be accurate. The Y chromosome does not mean “male”; XY together indicates a male (there are no people who are YY). In 1961, a man was found with an extra Y chromosome—an XYY man; this was later referred to as Jacob’s syndrome, after the British geneticist who described the condition, Patricia Jacobs.1 This  interesting genetic anomaly was published quietly in a very respected medical journal.2 The genetic event occurs once in about every 1000 male births, so it is relatively common.1 It is important to realize that this man was not  in any way aggressive, or a criminal, and the original researchers did not suggest that he was likely to become so. Later readers of the article, however, assumed that an extra Y chromosome would lead to more “maleness,” which was defined as being more aggressive and violent. Other researchers jumped on this bandwagon and began to take

Some misconceptions about genetics  71

surveys of chromosome number, but they made a simple statistical mistake: rather than using a wide sample taken from the general public, they used men in institutions for the criminally insane. These studies did show that XYY men seemed to be exceptionally violent. Then, the media learned of this and greatly exaggerated the horror of the crimes committed by these few men. This, then, suggested to the public that XYY men were “huge hulks of ‘super-maleness’ spurred on to aggressive acts by the extra Y chromosome.”3(p1) Officially, there were two groups that responded to this work and the resulting media and public hysteria in the 1970s. Kessler and Moos4 reviewed the literature and reasonably pointed out that the mental hospital findings were quite inconsistent. They were based on extremely small and carefully selected samples. A review of the same work today would result in the same response. Sarbin and Miller5 performed a second review of the original work at the same time as the Kessler and Moos’s review. However, instead of the reasoned response of Kessler and Moos, this review was highly emotional, with a strong political agenda. Sarbin and Miller referred to all genetic studies of this type as “demonism revisited,” even using this phrase in the title of their article, and wanted all such studies banned, insisting that criminological research should be limited to social, economic, and political variables and that biology should never be considered.3 This prompted more scientific study of this arguably interesting genetic phenomenon but this time using much more powerful and scientific studies. These researchers did not preselect by choosing people already incarcerated in criminal institutions but instead examined large, non-criminal populations. One study examined 4139 men (the tallest in a birth cohort, because XYY does usually produce tall men). In this group, only 12 were found to be XYY, of which five (41.7%) had committed criminal offences, which is disproportionately higher than XY men, but the offences were minor and non-violent in four of the five cases. The authors found lower IQ in the XYY men and suggested that the lower IQ may have increased their risk of being caught rather than XYY actually increasing their criminality.6 More recent studies have led many authors to agree in that the lower IQ resulted in other behavioral problems—such as lowered socioeconomic and living status and social and family problems—which they suggest was what resulted in criminality rather than the actual chromosomal abnormality. Although a slight increase in criminality was observed in one study, it was no different from controls when results were adjusted for socioeconomic status.7 Any change in the chromosome number could have major effects on a person, so having an extra Y chromosome may be problematic. Scientific, unbiased research on large samples of men found that XYY men are, in general, very tall, may have a certain level of intellectual disability,8,9 and also develop very bad acne.6 However, 75% of males with XYY are only detected accidentally, as they are normal mentally and physically, and many may never be diagnosed at all.1 What then is the truth about this anomaly? The genetic part of XYY results in men who tend to be taller than average and have bad acne and in some cases may have a slight intellectual disability. This is fact. But does it cause crime? Or at least, are XYY men more likely to commit a violent crime than XY men? Here, we are back to the original argument about whether genetic inheritance causes something directly or indirectly. We need to ask how these genetic traits are going to affect the child’s environment. Imagine this boy’s school days. Here is a boy about 12 years old, an awkward age at best for boys. He is very tall for his age. How will he be treated? His height alone might make him a bit of a hero, girls may be attracted to him, and the sports coaches will want him on their teams, so he might develop a lot of self-esteem, which boys of that age often lack, and he may do very well in later life. Now, we need to add the reduced IQ, which occurs in some cases; it is not much, not enough to prevent him from going to a regular school or getting a job in later life, but it may be definitely noticeable. He is also prone to exceptionally bad acne, which alone often has a major effect during the teenage years. This boy, therefore, is not doing well in school and is probably going to get a low-paying, boring job as an adult. It is very likely that in school this boy will be teased and even bullied simply because he is different. Therefore, what began as a purely genetic phenotype—including abnormal height, low levels of intellectual disability, and acne—is now being influenced by the environment created by the phenotype. Bullied children often turn to crime, in compensation or in retaliation. For example, if

72  Misconceptions, experimental design, and behavioral genetics

this boy is physically bullied, he might realize quickly that he cannot fight with his mind, but he can fight with his size. In many cases, children with lower IQs—and low social skills—are predisposed to antisocial and even criminal behaviors. These will not necessarily be violent crimes, but crime all the same. So, this XYY boy could end up committing criminal acts, but is it the result of his genetic makeup or because he has grown up in a bullying, non-nurturing environment? How different would it be had he grown up in a different environment: if his teachers and sports coaches had praised his height and supported him in other activities to the best of his abilities? The answer is quite clear, given that many men in our society are XYY and are not even aware of it. Most live productive and satisfying lives. The basic genetic blueprint results in a phenotype, but none of these genetic traits lead to crime. It is only when they are integrated with the environment that such predispositions could lead to crime. And one final clarification about XYY: misconceptions about this trait have been used as an argument both for and against a biological explanation for crime. XYY is caused by a random mutation at meiosis, when the chromosomes are originally replicated, so XY becomes XXYY briefly. They should then separate once and then again, until each sperm has just one of the four. In an XYY male, for some unknown reason, one of the sperm cells ends up with two Y chromosomes instead of just one. So, when it meets an egg with an X, an XYY child is produced. This is a random mutation and it is not heritable; that is, it is not passed on to the next generation. The most important fact for this study is often not mentioned at all: XYY is simply not a genetic issue in the traditional meaning of the word.

Experimental design An experiment will not generate valid, reproducible data if it is not designed correctly.

Scientific method The scientific method is a way of conducting valid research that has been used since the seventeenth century. It basically involves developing testable hypotheses, developing experiments to test them, analyzing the experiments, evaluating the data, and then refining the hypothesis. It can be broken down into several steps: 1. Make observations that lead to questions. By observation we are constantly questioning the things that happen around us, and some of these questions can be tested experimentally, such as “How did this happen?” “When did it happen?” and “Will it happen again?” 2. Formulate hypotheses. Hypotheses are possible explanations for how or why something is, which can be tested to determine which of the hypotheses is correct. Developing hypotheses is based on knowledge, training, and information from previous research in the same area. 3. Test hypotheses. Carefully controlled experiments are designed to test the various hypotheses. By controlled we mean that only one variable is tested at a time and all the rest are kept constant. 4. Evaluate and analyze the results statistically. 5. Finally, once all the results have been analyzed, reevaluate the original hypotheses to allow more testing, often by other researchers. This results in repeated focused sets of experiments, getting closer to an explanation of the original question. Most good experiments create more questions than they answer, leading to more experiments and to a deeper and more profound understanding of the phenomenon. Some research is also observational, to add to our understanding. Finally, the work is published in peer-reviewed journals.

Experimental design  73

Isolating a single variable In performing a scientific experiment, the researcher must eliminate all variables between research subjects except the one trait being studied. This is done so that the researcher can see what effect varying that one variable has on the outcome of the experiment. For example, a researcher wishes to test a new diet for chickens, to see if the chickens grow larger on the new diet. The researcher divides a group of chickens into two groups. One group, the experimental group, is fed the new diet, and the other group, the control group, is fed the old diet. The results show that the chickens fed the new, experimental diet grew to be twice as large as the controls. Does this mean conclusively that the diet increased their weight? The answer is no; there is not enough information in the above example to come to that conclusion. However, people often do leap to such conclusions. It is important to look at all experiments very carefully to see whether they were conducted correctly before we can determine whether their conclusions are valid. With just the above information, it is not possible to determine whether it was the diet that caused the change in growth. To ensure that the results were valid, the researcher would have to eliminate all other differences between the experimental and control groups. For example, what if the experimental group was kept in nice, well-lit conditions, with solid floors and lots of space in which to peck around, and the controls were kept in battery-hen conditions, with mesh floors, which chickens (and most animals) dislike? Do you still think that we could conclude that the diet made the difference? Perhaps it did, but maybe not; maybe the experimental chickens were much happier and healthier than the controls and thus naturally put on more weight. The differing conditions could be the cause of the weight gain, and it might have nothing to do with the diet. So, it is clear that we need to eliminate all other variables to clearly see the effect of just one variable. If we wish to study the effect of, for example, zinc on a plant, we would conduct an experiment with a large number of plants, divided into two identical cohorts. The plants should be genetically identical in all ways (plants are easy to propagate from a piece of plant and thus are very simple to clone). They should also be raised in exactly the same way: the same lighting, food, soil, water, and so on. The only thing changed would be the zinc level. Because zinc is the only difference between these two groups of plants, we can conclude that any differences are caused by the zinc, and we could look at thousands of plants quite easily—simple. But it is not so simple with animals. If we wish to look at the effects of vitamin B6 in rats, we could do the same sort of thing. We could test many rats (not so many as plants, of course, because it would be expensive and difficult). Rats do breed fast, but they do not have that many babies, and they won’t all be genetically identical, like the plants, because cloning is still very experimental and expensive. It is also ethically questionable, as well as very difficult. So, using the rats we have, we can go to the next step as above: we divide them into two groups and give them the same food and treatment in everything, except for the vitamin difference. We should therefore be able to find that any differences would be related to the vitamin difference, right? Not necessarily. The observed differences might be the result of other differences in the animals, for example, genetic backgrounds. Even if all rats had the same parents, they are not  clones, they are still different—like brothers and sisters. They  share some genes but are still unique. What else might make a difference? The environment is a likely suspect. Perhaps it is the way they were raised, the foods they had as babies, or the exercise they got. In a rat colony, we can control for a lot of those things, but even with a simple lab animal, it is difficult to eliminate all other variables. After we have isolated our variable, we need to see what happens if there are other factors involved, such as how the plant reacts to getting both zinc and magnesium. There might be some interaction between these two metals that prevents the uptake of one or the other. Doctors will often tell their patients that a medicine will affect the uptake of certain vitamins, so they should take supplements. Therefore, the next step would be to start adding another variable, and then another, and so on, to see what might happen in real life. If a soil high in magnesium will either prevent the uptake of zinc or, worse, cause a fatal overload of the system, there is no point in advising farmers that zinc will

74  Misconceptions, experimental design, and behavioral genetics

help their crops grow. We need to know how the factor we have isolated reacts to other parts of its environment. But to do so, we must be able to separate each factor first, see the individual results, and then start looking at all the possible combinations.

Sample size Another confounding effect in any experiment is sample size. In the above chicken diet example, a researcher has two groups of chickens kept under identical conditions, with the same-sized cages, the same lighting, and everything else the same. The only difference between the experimental and control groups is the diet. The chickens in the experimental group grow twice as large as the control chickens. Think about the reliability of the conclusion if the experiment is performed on: ■■ ■■ ■■

■■

2000 chickens: conclusive 200 chickens: still a good sample size 20 chickens: may be significant. Certainly interesting, and a feed company might like to take it further. This is not a very conclusive sample size, as 10 chickens in each group is not many. Perhaps the results would suggest that a larger experiment is warranted. 2 chickens: clearly not conclusive

In a sample size of two chickens, one chicken was fed one diet and one another, and there certainly was an effect. Unfortunately, we see people accepting this as evidence every day: “My friend tried this diet and lost 20 pounds.” But does the person check to see how many people did not lose any weight on this diet, or even gained weight? How many people make assumptions about a country based on their experience of meeting two people from that country? We should not do it, but we do. In science, a sample size of one is not an experiment. Why not? In the two-chicken sample, one chicken is definitely larger. Why would that not  be acceptable evidence? Because it could easily be the result of chance. It is similar to tossing a coin twice and coming up heads each time. If you tried that now, I am sure that many people would get this result. But what if you tossed the coin 200 times, and it came up heads every time or even 75% of the time? You would be suspicious and not want to make any bets with the owner of that coin. Therefore, random chance must be taken into account. Chance could easily explain why one chicken gained weight and the other did not. It might even explain the difference in 10 chickens. However, when the sample size increases to 100 or 1000, then chance will play only a very small role. Therefore, to eliminate the risk of chance entering into the experiment, an experiment must have a large sample size.

Crossover studies An experiment can also be strengthened by conducting a crossover study. This involves swapping the treatment and control groups halfway through the experiment. In this way, half the chickens get the new diet and half get the old diet. The researchers record weights, then feed the diets to the opposite group, and weigh them. If all the variables are balanced, the weight increase should follow the diet type. This proves that the effect is not just in those chickens.

Replication To be considered valid, an experiment must be repeatable. If a single experiment shows an effect in 2000 chickens, it sounds valid. But what if no other researcher can duplicate the result? Often, we see an experiment that has been repeated many times. This is good; it validates the results. However, it is also interesting to note whether the experiment is being replicated by the same researchers or

Experimental design  75

by a different group. Often researchers will continue the same line of thought over several years of research, so experiments get repeated in the same lab by the same people. There is nothing wrong with this practice, as it would be foolish to research one area and then suddenly change research fields. People frequently continue in the same research area, and their results are perfectly valid. However, it is always nice to see that other researchers in other labs have been able to repeat the experiment with different animals or plants and get the same general result. This does not happen always, because few people want to be part of the second group to prove something new. But when you look at a study, consider whether it is a series of studies by the same group or by different research groups. If it is a similar study performed by many different labs and researchers, then it is considered much more robust. The case gets even stronger, or more robust, if the experiments are repeated in different countries and researchers still get the same result. For example, the breeds of chickens may be quite different, but the diet still works. Or more to the point for our interests, the entire human culture may be quite different in different countries. So, a strong, or robust, experiment has a large sample size, is repeated many times by different researchers in different labs and different countries, and gets the same results.

Double-blind studies Double-blind studies are used to eliminate or reduce bias. This is an important experimental technique, especially when we measure something such as behavior, which is much harder to quantify than, say, weight gain. In double-blind studies, neither the researcher nor the experimental subject knows which group, the experimental group or the control group, the subject is in. The researchers have someone else code the subjects, so that when they assess the experiment, they can do so with as little bias as possible. For example, when we test a new drug, we frequently take 500 people with a particular disease who are desperate for a cure and treat half with the new drug and half with a placebo (a fake drug). These are generally double-blind studies, so both the experimenters and the patients are unaware of what they are taking, to eliminate the risk of either the patients wanting a cure so badly that they think they are getting better when they are not or the researchers expecting those getting the drug to improve and thus ranking them as improved when they are not. Say that for a behavior-related experiment, a number of preschoolers with aggression problems are divided into two groups. One, the experimental group, is given vitamin supplementation and the other, the control group, is given a candy that looks the same. Teachers and parents are then asked to assess their behavior to see if the vitamin has improved their behavior. If the teachers and parents know that their child has received the treatment, they may be more likely to say that the behavior has improved, as they want it to improve and believe the treatment will help. Hopefully, the researchers are less easily biased, but it is still easy when assessing such a qualitative factor as behavior to think that there has been an improvement when there has not. Double-blinding the subjects and researchers eliminates this bias, although in some cases, as we will see in future chapters, subjects correctly guess that they are receiving a treatment, which can cause bias.

Studying humans What happens when we want to look at a factor in humans? It is much more difficult than with plants or animals. Clearly, we cannot ethically treat them like lab animals. Actually, we can and do in some respects. For example, say 100 people were monitored after having heart surgery, and it was found that those who ate real butter, drank red wine, etc. had a higher survival rate. What is the problem with such an experiment? People are different in just about every way. They would have been of different ages, different exercise levels, and so on, and definitely of different genetic backgrounds. We all know that many people have a genetic predisposition for certain problems, such as heart disease or diabetes. A family doctor always asks new patients for their medical history and that of their parents, as people with parents who have had heart problems or diabetes are

76  Misconceptions, experimental design, and behavioral genetics

predisposed, or more likely, to develop these same problems. Therefore, when we do experiments with people, it is almost impossible to control for all other variables. It is difficult to know which variable is causing an effect. Is it something different in their environment? Is it one particular part of the environment? Or is it their genetics? Clearly, we cannot clone people, either ethically or even possibly, but clones do exist in nature— identical twins. Twin studies are a remarkably helpful way of studying many factors, both genetic and environmental, in humans.

Studying behavioral genetics There are several types of studies that are used to study the genetics of behavior. However, before we consider them, it is important to understand that all studies that assess the genetic basis of a behavior by necessity also assess the environmental contribution, as well as the interaction between the two. The main types of behavioral genetics studies are those of twins and adopted children.

Introduction to twin studies There are two types of twins: (1) monozygotic (MZ) twins and (2) dizygotic (DZ) twins. The terms relate to the number of zygotes required to create these two people. Mono (“one”) zygotic means that just one zygote was involved, and dizygotic means that there were two zygotes. When an egg meets a sperm, they each carry half the genetic complement needed to create a person. Moreover, the egg carries everything else needed for life. When the two meet, they create one complete zygote, that is, a single diploid cell (an organism with its chromosomes in pairs). This single cell contains everything it needs to eventually become a complete person. The single cell divides into two, then those two divide into four, those four into eight, and this division keeps on going until the entire fetus is produced. Indeed, it keeps on going until the person dies, because cells must be replaced all the time; blood cells, for example, only live for about 28 days. Normally, one zygote equals one person, and also normally, although a man releases millions of sperm, a woman usually only releases one egg at a time, once per month. Dizygotic twins

Dizygotic (DZ) twins are a result of two zygotes creating two people. Instead of releasing one egg, the female releases two. There are thus two eggs floating around and millions of sperm, so both eggs are fertilized, and the result is two babies. The only unusual thing here is that the woman released two eggs. This is relatively unusual, but it is an inherited feature. Women who release two eggs at a time tend to have daughters who also release two eggs at a time. Each individual baby is still absolutely unique, just like siblings, only they result from two eggs and two sperm that happened to fuse at the same time—within hours or up to 3 days of each other—rather than some years apart. We traditionally say that siblings share 50% of their genes with each other, although this is a little misleading, as the entire human race shares 99% of their genes with each other. This is really quite obvious if you think about it. If you were to take two random people and look at the similarities between them, you would see that we are all almost identical. We all have two arms, two legs, and two eyes; we digest in the same way, using the same enzymes, which are produced by the same metabolic processes, and so on. It is only the little things that make us different, such as facial shape, height, and hair color. We focus on these differences, but there really is not that much difference between us. Look at people around you and observe their features. Everyone is designed on the same plan. Our eyes are close to being in the same place, and our mouths are very much the same shape. There are only small differences. So, as all humans share 99% of the same genes, when we say siblings share 50% of their genes, we really mean 50% of the 1% that makes us different. In fact, 50% is a bit of an error too, because 50% is an average. By random chance, some twins will share more than 50% and some will share less.

Studying behavioral genetics  77

Full siblings inherit their genes from their parents, but they are not genetically identical to their parents. Some genes have several possible versions (or alleles), and the parents each have two copies of that gene. Their copies might both be the same allele or different ones, and they each give one to each offspring. So, siblings share the same genetic pool—they both obtained their genes from the same place—but each child can have a different allele of the gene, which is why we are all unique. Thus, DZ twins are as similar to each other genetically as regular siblings; the only difference is that they are of the same age. Monozygotic twins

Monozygotic (MZ) twins are two people that come from one zygote, a zygote that was supposed to produce only one person. During the early stages of development, each cell still has the ability to do everything. After a few more cell divisions, the cells will become determined. It is as if they decide what they are going to do with the rest of their lives, and for a cell, the choices are skin cell, nerve cell, kidney cell, blood cell, and so on. They are not changed yet—they are just becoming determined; that is, they have become committed to a particular developmental pathway. After a few more divisions, they become differentiated: they actually change into the particular type of cell to which they were committed. They then change into a skin cell, for example, and turn off all the functions that they would have needed if they had been a kidney cell or a blood cell and concentrate solely on being a skin cell. Remember that this was the important thing about Dolly, the cloned sheep: not that she was cloned (that had been done) but that a mammary cell, which was committed to being a mammary cell and had been functioning as such for years and had turned off all other functions, could be convinced to turn back on all the functions that it had turned off and that one cell could do everything again and produce another entire sheep. However, before the cells are determined and differentiated, when the zygote consists of just a very few cells, each cell can still produce an entire person. If something happens at this point (and in nature, we are not sure exactly what it is that happens), something cleaves or splits this little zygote, so that instead of being, for example, a four-celled zygote, it becomes two separate and complete two-celled zygotes, both of which go on to become separate people. Therefore, one zygote, which came from only one sperm and one egg, becomes two zygotes and, therefore, two people. The difference between monozygotic twins and dizygotic twins is that monozygotic twins are genetically identical; they have the same genetic makeup exactly, so they are natural clones. They will grow into two identical people with exactly the same versions (or alleles) of every gene. They still share 50% of that 1% of their genes with each parent, but they share 100% between each other. These are MZ twins. Unlike DZ twins, who share 50% of that 1% of their genes with each other, they share everything. Can DZ twins be a boy and a girl? Yes. The mother contributed an X to each, but the father could have contributed a Y to one and an X to the other. Can MZ twins be a boy and a girl? No. They are identical, so they share the same sex chromosomes; they can both be boys or girls, but not one of each.

Explanations for twin coincidences We have all heard of twin coincidences in the popular tabloid press, such as the many stories of two identical (monozygotic) twins who were separated at birth and then “discover each other” as adults. It is somewhat unusual for identical twins to be separated, but people who want to adopt a child, particularly a baby, often do not want two babies at once. Authorities try not to separate twins, but it is not always possible to keep them together. In this type of story, two identical twins are separated at birth and accidentally meet each other 40 years later. They look identical. We would expect that, because they are identical twins. The environment can affect the way we look, but looks are strongly influenced by our genes. Reading further, we learn all this amazing information, such as the fact that they are both firefighters, they both wear glasses and have mustaches, they both married nurses named Susan, and they both own terrier dogs called Spot! Coincidence or biology?

78  Misconceptions, experimental design, and behavioral genetics

This seems incredible. If you accept the story, then it would appear that genes not only control the way we look but also our careers and the names and careers of our spouses. This is the sort of tabloid material that gives twin studies a bad name. Let us look at it and see if we can explain all these similarities. Which of these are inherited traits? ■■

■■

■■

■■ ■■

■■

They both married women named Susan. Is there a gene for that? No. It is probably just a coincidence … or maybe not. What else might affect that? Think of the names given to children today in North America. Although some names go back 2000 years, many names had never been heard of 50 years ago or even 20 years ago. The popularity of certain names rises and falls over time. Thus, the name may have been common in the age group of people they associated with (they are the same age, remember). Also, names are often very specific to different countries. For example, extremely common names in America are not the same as common names in Japan or Germany. The twins we are discussing were probably both adopted in North America, so it is not particularly far-fetched to find that they both married women named Susan. They both have a dog named Spot, and it is a terrier. Is this an inherited trait to buy a terrier and to name him Spot? I do not think so. But maybe they both grew up at a time when a cartoon they both watched had a dog named Spot. I wonder how many people had dogs named Max after the movie remake of the well-known Dr. Seuss story How the Grinch Stole Christmas. So, the fact that they both married someone named Susan and had a dog named Spot is probably related to their mutual environment—not that they grew up together but that they were both adopted in the same country at the same time. It has nothing, whatsoever, to do with their genes. There are often sensible explanations for what seems like too much of a coincidence. They both have mustaches. Is this inherited? Men’s facial hair arrangements go in and out of fashion. Recently, goatees were popular, whereas very few men wore them a few years ago. Both these men probably succumbed to fashion, so there was a very good chance that both would have mustaches if it was fashionable at the time. This  would be an entirely environmental reason. However, there might also be a genetic component. The  ability to grow a mustache is inherited to a certain extent. Some men have much thicker facial hair than others, and the ability to grow a thick mustache may well relate to their shared genes. Did their genes make them grow a mustache? No. Well actually, yes and no. Why do we do anything like cut our hair or curl our hair or grow a beard? Usually, we do it because it suits us. So, why would both men grow a mustache? Presumably, it suited their facial shape, and that was inherited. So, there is a genetic component in there, but only a very indirect one. There is no gene that says, “Grow a mustache, young man,” but there are factors, both genetic and environmental, that might predispose a person to wear a mustache. Facial hair is also often grown to hide a defect, such as a hare lip or receding chin. The problems that are being hidden were genetic, so the facial hair growth could be said to have its roots in the genes, but only very indirectly. There is no direct genetic link that makes a person grow a mustache, but there might be an underlying genetic predisposition, a very indirect link. They both wear glasses. This could be related to a genetic predisposition for bad eyesight. They both became firemen. Could this be related to genetics? Not very likely. However, careers go through popularity surges. Some years ago, a popular television show about a country veterinarian in Yorkshire inspired many children to become veterinarians. Forensic shows and movies have dramatically increased public interest in becoming a forensic scientist. The fact that both these two men became firemen could easily have been the result of fashion or of a big drive to recruit firemen when both men were young. What else might affect their choice? Genes could have had an effect, albeit indirect once again. Certain people crave excitement more than others or are more fearless than others. Moreover, part of a person’s physical build is genetic, although it is obviously influenced by diet and exercise. They would need to be strong and physically fit to be firemen. Both men married nurses. It again sounds like an amazing coincidence, but this was an obvious career choice for women at a certain time. Both men were the same age, and they probably

Studying behavioral genetics  79

married women of similar ages, so the women had the same cultural influences. What else might influence their choice of spouse? People with professions in similar fields often meet. Both would work shift work, they might come into contact professionally, and both are in caring professions and thus might have similar outlooks. It is clear that although the story may have sounded amazing at first, most of the coincidences can be easily explained. The explanations are based on the repeatability or standardization of roles and styles in culture—not on genetics. In all the cases addressed in this book, any genetic links to behavior, and therefore potentially criminogenic behavior, will be indirect links to a predisposition for a behavior. None will be direct links. Genetics does not work that way. It is never going to be possible to say that a person’s biological makeup caused that individual to commit a crime, simply because it could not, as it is such a complex behavior. All that genetics can do is give someone a predisposition to behave in a certain way. A predisposition. This does not mean the person will behave in that way, only that this person is perhaps more likely to behave in that way than the average person. Also, the predisposition is toward a behavior, not a crime. The behavior may lead to a person committing a crime, but it may equally lead to something totally non-criminal. For example, the predisposition may be for a person to be somewhat impulsive. This may lead to shoplifting or stealing a car. It may equally well lead to shopping impulsively at garage sales. In both cases, the predisposition and the behavior it may lead to— impulsivity—are the same and may have a basis in biology; the results are quite different. Whether the behavior is criminal or not will be greatly affected by upbringing, environment, and choice. One of the reasons that people often seem fearful of considering a biological basis for behavior, and subsequently crime, is that people believe that scientists will discover something that shows “this causes that.” Behavior is not that straightforward. An example of very straightforward genetics is “this gene causes the flower to be colored red or white,” depending on the allele of the gene the plant inherits. Even this kind of genetics often gets more complicated than one gene. Even simple things such as hair and eye color are determined by several genes, so something as complex as behavior can only be influenced by many genes, and even then, they can only offer a slight predisposition for a type of behavior. Many genetic effects are additive, in that a large number of genes may influence a single behavior, so if a person has only one or two, there will only be a slight predisposition, but if they have 20, they will have a much greater predisposition. It is still a predisposition. Whatever the genes produce will be strongly affected by the environment and can be totally changed by it. When we look at the twin studies in the next chapter, we need to look at the significance of the results, just as we would look at any other data. We must not get caught up in the excitement of facts such as both men having married women named Susan and named their dogs Spot. We need to look at the more important aspects of their similarities and also at their differences; then, we can see what is significant and what is just a case of random chance. We need to compare the similarities between these two men and the general population in order to work out whether something is significant. For example, if you were to consider a random group of 1000 people in the world, is it significant that of this group, 90% are under 30 years old, 100% speak and write the English language extremely well, and 80% are wearing jeans? This would not be at all significant if your group of people were undergraduate university students in North America. It might be much more significant in a different culture or country. Therefore, before you determine whether something is significant, you must first compare the trait with the norm for that population. A behavior that might seem very significant may not be at all significant in that particular situation or population.

Using twins to study genetic and environmental influences on behavior Twins provide a totally natural, non-experimental situation, so they are a wonderful tool with which to study factors that are environmentally or genetically determined. Certain traits are often

80  Misconceptions, experimental design, and behavioral genetics

studied using concordance rates. For example, if the concordance rate for a certain trait in monozygotic twins is determined to be 70%, this means that the chance of the other twin developing the trait is 70%. We can compare the concordance rates for a specific trait between MZ and DZ twins—for example, if the trait is 70% in the MZ twins but only 5% in the DZ twins, then it would appear to have a heritable component. The twins who are genetically identical are more much more likely to share the same trait than twins who are not genetically identical; therefore, the trait relates to genetics. If the concordance rates were 70% in both MZ and DZ twins, then it would obviously be an environmental effect, as the fact that the MZ twins are genetically identical does not appear to make any difference. If the researcher went further and compared the trait in unrelated people, and if the concordance rate was 70% in both MZ and DZ twins but only 10% in unrelated people, then it would indicate either something that is environmentally specific only to twins or something that is both genetic and environmental. In summary, MZ twins raised together share 100% of their genes and 100% of their environment. However, DZ twins raised together share 50% of their genes (remember that it is just 50% of 1%) and 100% of their environment. Thus, because they share the same environment, any differences in a trait between MZ twins and DZ twins is caused by their genetics and not by the environment. If the two, MZ and DZ twins, have the same concordance rates for a trait, then the trait is more related to the environment and not to the genes. So, twins offer a perfect real-life method to study both the effect of genetics and the effect of the environment on a behavior. Contrary to the belief of many social scientists, twin studies do not just study genetic factors but also focus on environmental influences on criminal behavior, as by necessity, such studies consider the relative weights of both genetics and the environment. Twin studies expand on the environmental influences by separating the shared and non-shared environmental factors. The shared environment is the one that is the same for both twins, while the non-shared environment includes the environmental effects that are unique to each twin.10 Although it is rare, in some situations, identical twins are reared apart. This perfectly separates the genotype from the environment. A very important assumption in twin studies is that the two twins share the same environment, which in most situations is very likely, but as we will see in the next chapter, this can be a concern. If the environments are different for each twin, then the genetic contribution to a behavior will be incorrectly inflated. A second form of study, adoption studies, eliminates this concern.

Using adoption to study genetic and environmental influences on behavior Adoption studies separate the genetic and environmental influences on behavior more completely by comparing an adopted child’s behavior with the behaviors of the biological parents and siblings and those of the adoptive family. Many countries, particularly those of northern Europe, keep very thorough registries that record everything about a person—not just birth, death, and marriage but also health, therapeutic drug use, drug and alcohol abuse, adoption, criminal record, criminal arrests, and much more. Each record can be searched, as each individual has a unique identifier number and can be traced through multiple registries. Although this does raise concerns about privacy, the registries are a treasure trove for researchers, as adoption, time of adoption, length of time in biological home, length of time in adopted home, schooling, antisocial behavior, arrest, and criminal record can be determined for all adoptees, as well as their biological and adoptive families.

Conclusion We have now worked through some of the important misconceptions concerning genetics, behavior, and crime, such as cloning and the so-called XYY man. The many myths that surround the relationship between biology and behavior either have no factual basis or relate more to the effects

References  81

of the environment than to genetics. We have also seen how a good experiment can be designed to test a hypothesis. The study of twins, and later, adopted children, as we shall see, provides us with a natural way to separate the effects of genetics and the environment on behavior.

Questions for further study and discussion 1. Many people who stridently insist that behavior has no basis in biology, despite the obvious evidence, still seem concerned about the possibility of cloning a dangerous person. Explain why this, even if possible, would not be a concern. 2. Explain why we need to remove all but one of the variables in an experiment to examine that variable. If we eliminate all the other variables, how then do we work out how the variable is influenced in the real world? 3. Many people are concerned that if we look hard enough, we will find a gene for crime, which might result in such people being targeted, labeled, jailed, or even not considered responsible for their actions. Discuss the several lines of evidence that show we will never find a single gene for crime. 4. Discuss why comparing a trait in MZ twins with the trait in DZ twins is so valuable in criminological research. 5. Explain why crossover double-blind studies are so important and why replication and sample size are so important. 6. In a true-life twin example, the Jim twins were born in Ohio in 1939 and separately adopted by different families, who named them both Jim.11 Look at each of the exciting similarities below and discuss whether they are genetic or environmental or both and explain each. They met for the first time at the age of 39 and found that they both11: a. Were of the same height and weight b. Had dogs called Toy as children c. Had spent childhood vacations in St Pete’s Beach in Florida d. Had married and divorced women called Linda e. Had married a second time to women named Betty f. Had sons with identical first and middle names g. Were part-time sheriffs h. Had carpentry hobbies i. Smoked the same brand of cigarettes and drank the same brand of beer j. Suffered severe headaches k. Had identical crooked smiles l. Had identical voices m. Left love notes around the house for their wives n. Exercised everyday o. Drank orange juice

References 1. Re, L. and Birkhoff, J.M. 2015. The  47, XYY syndrome, 50  years of certainties and doubts: A systematic review. Aggress. Viol. Behav. 22: 9–17. 2. Sandberg, A.A., Koeph, G.F., Ishihara, T., and Hauschka, T.S. 1961. An XYY human male. Lancet 1: 488–489. 3. Mednick, S.A. 1987. Introduction: Biological factors in crime causation: The  reactions of social scientists. In: The Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. Cambridge, UK: Cambridge University Press. Proceedings NATO Conference, Skiathos, Greece, September 20–24, 1982. pp. 1–6.

82  Misconceptions, experimental design, and behavioral genetics

4. Kessler, S. and Moos, R.H. 1970. The XYY karyotype and criminality: A review. J. Psych. Res. 7(3): 153–170. 5. Sarbin, T.R. and Miller, J.E. 1970. Demonism revisited: The XYY chromosomal anomaly. Iss. Crim. 5: 195–207. 6. Witkin, H.A., Mednick, S.A., Schulsinger, F., et al. 1976. XYY and XXY men: Criminality and aggression. Science 193: 547–555. 7. Stochholm, K., Bojesen, A., Jensen, A.S., Juul, S., and Gravholt, C.H. 2012. Criminality in men with Klinefelter’s syndrome and XYY syndrome: A cohort study. BMJ Open 2(1): e000650. 8. Hoffman, B.F. 1977. Two new cases of XYY chromosome complement. Can. Psyc. Assoc. J. 22: 447–455. 9. Horgan, J. 1993. Eugenics revisited. Sci. Am. 254: 122–131. 10. Beaver, K.M., Rowland, M.W., Schwartz, J.A., and Nedelec, J.L. 2011. The genetic origins of psychopathic personality traits in adult males and females: Results from an adoption-based study. J. Crim. Justice 39(5): 426–432. 11. Miller, P. 2012. A thing or two about twins. Nat. Geogr. 221(1): 1–8.

5 Evidence for genetic predispositions for criminogenic behavior Twin and adoption studies

Introduction We will now  consider inheritance studies. These studies are natural experiments that involve observing differences in traits based on genetic and environmental contributions. We will start with twin studies and their results and discuss possible problems inherent in them. This will lead, in turn, to adoption studies, which more clearly separate genetics and the environment. Numerous researchers from many countries have amassed convincing data that demonstrate the relationship between genotype and behavior. Some of these studies have limitations, which will be discussed, but overall, the evidence for a relationship seems overwhelming. These genetic studies also provide some of the most important data about the role of the environment in criminogenic behavior.

Twin studies As Chapter 4 indicated, twins provide a natural way to study the effects of the environment and genetics on behavior. As we will consider now, twin studies are an extremely powerful study technique used to help us understand the genetic and environmental contributions to a human trait and have been commonly used for decades to study a wide variety of traits. Although we are primarily interested here in studies that relate to separating the genetic and environmental contributions to criminal and antisocial behavior and in understanding their interactions, it is important to realize that twin studies have a long and successful history in elucidating the factors involved in the heritability and environmental components of a vast array of human traits, such as immune system function, structure of body features, brain and heart function, respiratory function, weight maintenance, disease, social interactions, hormone function, reproduction, cardiovascular function, metabolic function, psychiatric and cognitive function, and so on. A meta-analysis considered studies on almost 18 000 different types of traits from 1958 to 2012 and involved more than 14 ­million twin pairs in over 2700 published studies.1 The results showed varying weight for genetic and environmental contributions, depending on the trait considered, but of all the thousands of various disparate traits measured, from blood pressure to weight maintenance to smoking, there was not one trait that was not under at least some genetic control. This is convincing evidence that all human traits have a heritable component.

83

84  Evidence for genetic predispositions for criminogenic behavior

It is important to remember that a very large number of genes are likely to be involved in just one trait, as most human traits are highly complex and are influenced by many genetic variants, unlike the simple Mendelian inheritance we discussed earlier. In late 2018, for example, 535 new genetic regions were discovered to be associated with high blood pressure, adding to the already known 210.2 Therefore, any trait will be influenced by myriad genes to varying extents. In  this ­meta-analysis of twin studies, the levels of heritability were clustered within certain trait types, but the average over the almost 18 000 traits was a heritability of 49%, with 69% of these due to ­additive genetic variation. The overall data did not suggest a strong influence of shared ­environment or ­non-additive genetic variation (that is, gene interaction).1

Early twin studies Twin studies have been conducted for almost 100 years in many countries, in an effort to understand the impacts of genetics and the environment on criminal behavior. The early studies varied widely in terms of the country of origin, age and sex composition, sample size, determination of zygosity—that is, how they determined whether the twins were monozygotic (MZ) or dizygotic (DZ)—and definitions of crime. Some of the early studies had very small sample sizes, although a few were adequate. However, all the studies showed greater concordance rates for criminality in MZ twins, as opposed to DZ twins. That is, every study showed a heritable or genetic component for criminogenic behavior. In a meta-analysis of 13 early studies, Raine3 averaged concordance rates and weighted them for sample sizes, giving greater weight to those studies with adequate samples. On average, he found that the studies resulted in concordances of 51.5% for MZ twins and 20.6% for DZ twins.3 Raine then went further and reanalyzed all the raw data, eliminating or weighting other potentially confounding factors, such as potential racial bias, and adjusting for zygosity determination. In the early days, twins were often determined to be MZ, based on their physical similarity, but DZ twins of the same sex look very alike, just as siblings often do, and as they are of the same age, it might be easy to confuse MZ with DZ twins. Although DNA analysis was not available at the time, some studies did use blood typing, which is more reliable than physical similarity, so Raine separately analyzed the studies that used zygosity determined by blood. He also looked at data from males and females separately. In all cases, concordance rates were much higher in MZ twins than in DZ twins, ranging from approximately 54:28 to 48:15.3 These experiments are robust, as they have been repeated by different researchers by using different samples of twins, in different countries, and even in different decades, indicating a very powerful genetic component to risk of criminal behavior. One of the largest of the early twin studies was conducted in Denmark. Denmark is a remarkable country in that it keeps official registries of many factors, including criminal activity, alcoholism, and mental illness. Denmark also permits researchers to obtain these data, including the names and addresses of the persons recorded. Although obtaining such information in many countries would not  be considered acceptable due to concerns about privacy and human rights, the Danish databases are extremely valuable in such research, as they provide comprehensive data on a very large sample of people. A criticism of studies conducted in Denmark and similar northern European countries, such as Sweden and Finland, is that the countries are very homogeneous. Extensive social systems allow most people to have a relatively similar standard of living, and most European countries have much less racial diversity than North America. The lack of diversity in the countries where much of these data originate is very valuable, because it eliminates several variables. However, it could be suggested that these data may not apply to more diverse regions such as North America. As studies on North American data sets have also been conducted and the results support the conclusions of the European researchers, it would appear that the data are applicable. In  this first major twin study, Christiansen and colleagues looked at a total twin population of 3586 twin pairs from one area in Denmark and, as with all the other studies, showed concordance rates two to three times higher for criminality in MZ twins as compared with DZ twins.

Twin studies  85

This clearly indicated a significant heritability for crime but this time with a more than adequate sample size.4,5 Higher concordance was seen for more serious crimes.4 In  the mid-1990s, Lyons performed the first of many US twin studies on subjects from the Vietnam Era Veteran Twin registry, a data set that is still being studied.6 All subjects were male, and the registry was compiled from military records of men who served in the US military during the Vietnam War. Utilizing such a database does remove certain variables, such as age, as all the men must have been within a specific age range and political and social issues must have been relatively similar for all. However, all were soldiers who served in a specific war, which might have an influence on the results. The study used blood type and photos to determine zygosity and included 3226 pairs of twins, with 55% MZ twins. Only same-sex DZ twins were included, which reduces another variable. All data were gathered using self-report questionnaires and included questions about various levels of criminal behavior, such as juvenile (aged under 15 years) and adult behaviors, juvenile and adult criminal arrests, and multiple arrests and felony convictions.6 It is useful that different levels of criminality were considered, because in many studies, it is difficult to understand the researchers’ definitions of crime, which range in some studies from very minor juvenile acts to felony convictions. Lyons found that there were significant influences from both genetic factors and the common environment on early arrests. The environment, rather than genetics, significantly influenced juvenile criminal arrests. Conversely, genetic factors, but not the environment, significantly influenced adult crime. Genetics were seen to play a role in whether respondents were arrested after the age of 15, whether they were arrested more than once after age 15, and whether they were involved in later criminal behavior.6 Twins were found to resemble one another most in early-arrest statistics, with almost three-quarters of the variance being explained by the genes, but there was also a significant and strong correlation with the shared environment. There was a significant influence of genetics, but not the environment, on later arrests and multiple arrests, as well as on later criminal behavior, whereas the environment had a very significant effect on early criminal behavior. This finding seems somewhat anomalous: Lyons found a significant effect of genetics and the environment on early arrest but only a significant effect of the environment on early criminal behavior. This difference suggests that many of these twins got away with many criminal acts as juveniles or that the crimes were too minor to be reported, such as stealing candy from a store. The difference appears to represent the committing of the crimes, which is strongly environmentally linked, and getting caught, which is both genetically and environmentally linked. It may be that genetic factors linked to getting caught, such as intelligence and impulsivity, play a greater role than the actual genetic determination of crime itself.6 This  study confirmed that there is a higher genetic influence on adult crime compared with juvenile crime and that the environment has a greater effect than genetics on juvenile delinquency. The same results have been found in many other studies. In a reanalysis of many of the older twin studies (1931–1977), the data were divided into adult and juvenile crimes, and although a strong concordance for adult crime was again seen in MZ twins as compared with DZ twins, there was little or no concordance seen for juvenile delinquency.7 In many cases, the same researchers have conducted research on both adult and juvenile crime, so the results are not subject to researcher bias.7 Lyons’ study not only claims support for a genetic role in criminal behavior, but it also included the environmental role. That is the beauty of twin studies: you cannot study one without the other. The researcher is always comparing the two, and in many cases, the results show as strong an influence of the environment as of the genes.

Modern twin studies Many twin studies have been conducted in the last two decades that, despite being larger, better controlled, and more robust, still support the results of the early studies. Sweden is also a country in which a large amount of data about the populace is recorded and easily available for research, so a number of twin studies have been conducted there, using many different registries, such as the

86  Evidence for genetic predispositions for criminogenic behavior

Swedish Twin Registry and the Swedish Crime Registry, using each person’s unique identification number to track them through the databases. In a study of 21 603 twin pairs, genetic and environmental influences were recorded for criminal behavior in general and then divided into three major types of crime: violent, white collar, and property.8 Results showed a general heritability for criminal behavior of 45% in both men and women, which was similar to that seen in a metaanalysis of 38 studies on antisocial personality disorder (56%)9 and a meta-analysis of 51 twin and adoption studies on antisocial behavior (41%),10 both from a range of countries. The Swedish study showed that the contribution of genetics and the environment were similar for all three subtypes of crime, with heritability at approximately 45% for violent crime, 54% for white-collar crime, and 54% for property crime.8 Shared environment was also important for all three subtypes (17%–24%). However, interestingly, almost 50% of the genetic influences on violent behavior were unique to that type of crime, and about 30% were unique to white-collar crime, but 0% were unique for property crime. In contrast, half the variance in shared environment was unique for property crime but only 20% for white-collar crime and 0% for violent crime.8 This supported previous research by the same researchers using a different data set from Virginia, in which specific traits and psychiatric disorders, such as externalizing behaviors, aggression, and rule-breaking, were related to a wide range of different genetic risk factors.11 This makes sense, as antisocial behavior includes myriad traits, which are liable to be under the direction of a large number of genes. In the Virginia twin sample, conduct disorder (CD) also resulted from two separate genetic elements, in which the first related to petty antisocial behaviors, such as school truancy and lying, and the second related more to physical fights and a desire to inflict pain.11 Similarly, in the Virginia twin data set, antisocial personality disorder showed two separate genetic elements, with the first including irritation, fighting, and a total lack of concern for other people’s safety and the second involving lack of responsibility and dishonesty.11 As we cover other biological influences on behavior, we will see that these and other antisocial behaviors are linked to a number of biological processes, such as neurotransmitter, neural, and hormonal deficits, all of which are under the control of many different genes and are triggered, ameliorated, or moderated by environmental triggers, so it is logical that different forms of criminal behavior will have different genetic underpinnings. A longitudinal study of criminal convictions in twins and siblings from adolescence through early adulthood in Sweden involved almost 70 000 male twin pairs and full-sibling pairs (within two years’ age difference) at ages 15–19, 20–24, and 25–29.12 Kendler and colleagues sought to determine whether there was genetic or shared environmental evidence for Moffitt’s well-known theory that adolescents who commit crimes are broken down into two separate groups, those who commit crimes during adolescence then mature out of such behavior (“adolescent-limited”) and those who continue to commit crime throughout their lives (“life-course-persistent”).13(p676) They  suggested that there could be two sets of influences on adolescents, one that acts during adolescence and one that becomes active only in early adulthood.12 Their analyses showed that there were two sets of genetic risk factors for criminal behavior: one that began in mid-adolescence and persisted, although with greatly reduced effect, into early adulthood and one that began between ages 20 and 24 and persisted through the age of 29, the end of the experiment. These results are s­ upported by other research that has shown that genetic influences are dynamic over the ­adolescent/early adulthood period, with some genetic factors being stable through adolescence and early adulthood and new genetic factors becoming active in early adulthood.12 This  also explains Lyons’ results described above.6 These results generally support Moffitt’s theory on adolescent-limited and lifecourse-persistent types of criminal offenders, although the first set of genetic influences did persist, albeit greatly reduced.12 Further work is needed to explore this issue more deeply. Although this text is concerned with criminal behavior, twin studies have been used to study multiple behaviors, from smoking to voting to food consumption and length of time watching television. Most of these have nothing whatsoever to do with criminal behavior, but some studies on non-criminal behavior can help us understand how to improve intervention methods. School performance and outcome are known predictors, or risk factors, for future success or failure. In other words, children who succeed in school academically are likely to make good peer choices and do

Twin studies  87

well academically; are unlikely to be absent, disciplined, or expelled; and will frequently exhibit very low levels of antisocial behavior, usually leading a successful future life. Conversely, children who do poorly academically often make poor peer choices and are more likely to be absent, disciplined, or expelled, and such factors often increase risk for antisocial behavior and overall poor outcome. Although this is very much a generalization, poor school outcome is a major risk factor for antisocial behavior and later criminogenic behavior. Improving school outcome has been repeatedly shown to improve outcome. One major part of school success is the ability to read. Reading disability or a reading level greatly lower than that expected based on age, intelligence, and education is found in approximately 4% of US schoolchildren.14 A number of twin studies have been conducted to determine the heritability of reading ability and have indicated that a substantial amount of variance in reading ability is under genetic control. For example, in California, a study of 605 pairs of ­10-year-old twins found that reading heritability was estimated to be 70%; in a UK longitudinal study, teacherreported heritability of reading ability at ages 7, 9, and 10 in 4000 pairs of twins was found to be consistently 60%; and a study in Colorado found heritability to be 80% or more in several d ­ ifferent measures of reading (reviewed in Kirkpatrick et al.14). A large study (n = 4886) of a mixed sample of twins and adopted children, called the Minnesota Twin Family Study, confirmed these figures with a heritability estimate of approximately 70%. There was almost no effect of shared environment for reading ability or disability, and no gene-environment interaction was noted, as the parents’ level of reading ability or education had no impact on the child’s reading ability.14 The authors do not claim that reading ability has no relationship with environment, such as teaching, learning, and practice, as clearly, reading is a learned ability. However, almost all US children are taught to read and go through many years of compulsory schooling that provides constant practice; this is universal and so fairly constant in all children in this population, showing little variance. The authors speculated that if they had studied a population in which schooling was not mandatory and children had much greater differences in level of teaching and practice, a much greater variance in environmental ­factors would have been observed. They  also suggest that there is a dynamic gene-­environment relationship, as children themselves shape their environment—in this case, by some children actively choosing to practice and learn reading skills in their spare time, while others choosing not to improve their reading skills. The authors hypothesize that this choice may also be under genetic control.14

Identical twins reared apart One of the main criticisms of twin studies is that MZ twins may be raised more similarly than DZ twins. This problem is eliminated in studies that look at identical twins raised apart. There are lots of tabloid-type stories about their amazing similarities, but there have also been some scientific studies. In  Christiansen’s5 Danish twin studies, there were eight  cases in which MZ twins were raised separately and in which one twin was registered for criminal activity. Of these eight pairs, four were concordant for crime. Although these seem to be a perfect “experimental design,” with genetically identical children raised in totally different environments, in reality, many of the earlier studies were faulty, as the children were often not separated until they were several years old, or were reunited as children, or were raised by other family members.15 A strong study was conducted by Grove and colleagues in 1990.16 They looked at 31 sets of MZ twins and one set of triplets who were separated shortly after birth and reared apart. Zygosity was assessed using several techniques, including blood groups, which they stated resulted in less than a 0.1% chance of misidentification. A score for antisocial behavior in both childhood and adulthood was derived by interviewing each subject blindly, with a standardized interview schedule. Statistically significant heritabilities were obtained for antisocial behavior in both childhood (0.41) and adulthood (0.28).16

88  Evidence for genetic predispositions for criminogenic behavior

Cautions for twin studies There have been many challenges to twin studies over the years. In their 2014 and 2015 critiques, Burt and Simons challenged twin studies on several points, frequently emphasizing challenges that have been made by earlier researchers.17,18 These papers were followed and interspersed by several rebuttal papers19–21; all the papers were published in the eminent journal Criminology. Shared environment

Burt and Simons reiterated a commonly expressed concern with twin studies—that of the assumed common environment shared by twins. It is assumed that each member of both DZ and MZ twins experiences the same environment. If MZ twins, by reason of being identical, experience more similar environments than do DZ twins, then it could accidently inflate heritability measures.22 Dizygotic twins look different and may be of different sexes, whereas MZ twins look identical and are of the same sex. It is therefore possible that the environments in which MZ twins are raised may be more similar than those in which DZ twins are raised. We often see MZ twins dressed the same way, but it seems less common in DZ twins. Also, DZ twins who are of opposite sexes are more likely to be treated differently than DZ twins of the same sex. This is why most twin studies only include same-sex twins. Studies have also shown that MZ twins are more likely to share a classroom and friends, spend time together, share bedrooms and clothes, and report more closeness than DZ twins.17 If MZ twins are treated more similarly than DZ twins, it would artificially raise the concordance rates. However, there is also evidence that some twins make a great effort to be “different,” and some twins may develop opposite (that is, submissive/dominant) roles.3 These effects are likely to be higher in MZ twins because they look alike and would want to be more different than DZ twins, who do not look the same and thus already have more of their own identity. Therefore, this factor could artificially lower heritability estimates. Generalizability

Another argument against twin studies is that of generalizability, which states that the data from twin studies may not be applicable to non-twins, as twins have certain features that are less common in singletons, such as higher rates of birth complications and lower birth weights; they also may share the same placenta, are the same age, and are usually much more emotionally bonded to their twin sibling than to a non-twin sibling.22 Such concerns have led some researchers to argue that twin studies should be abandoned in favor of other forms of study.17 In a project to address both these issues, Kendler and colleagues (who have conducted many twin studies in Sweden and the United States) examined a very large Swedish cohort of 911 009 fullsiblings raised together (FSRT), 41 872 half-siblings raised together (HSRT), and 52 590 half-siblings raised apart (HSRA) and assessed their criminal behavior by using the Swedish Crime Registry of criminal convictions.22  The ratio of shared genetic factors in MZ and DZ twins is 1.0 and 0.5, or a ratio of 2:1, which is the same ratio as seen in full-siblings compared with half-siblings (0.5 and 0.25), so the authors argued that using full- and half-siblings in the same methodological design as MZ versus DZ twins could be used to assess the value of twin studies and also show generalizability, as full- and half-siblings are much more common in familial relationships.22 The researchers found heritability for criminal behavior to range between 33% and 55% in females and 39% and 56% in males, which was very similar to the results seen in twins from the same Swedish population by the same researchers.8,22 Interestingly, shared environmental factors were not as important in full- and half-siblings as was seen in twins (1%–14% in females and 10%–23% in males), suggesting that some shared environmental factors are exclusive to twins.22 The results shown here, based on siblings and half-siblings, were very similar to those seen in the previous twin study and also in the two meta-analyses mentioned above.9,10 The results of this study are important because one of the strongest arguments against twins studies is that MZ twins have a greater similarity in environments than do DZ twins, artificially

Twin studies  89

inflating what appears to be a genetic contribution.17 This study, however, has shown that when considering siblings and half-siblings raised together or apart, instead of MZ and DZ twins, the levels of heritability still hold true at very similar levels. Also, the argument that twins are different from non-twins due to higher birth complications and other issues does not hold true, in that regular siblings and half-siblings are much more representative of the general population and do not have these increased complications, yet the results are robust. Moreover, the study provides important information about shared environment: there were consistently greater relationships between half-siblings raised together, as opposed to those raised apart, and similarly consistently greater relationships between siblings and half-siblings raised together who were close in age.22 In their rebuttal of the original Burt and Simons article, Barnes and colleagues suggested that Burt and Simons had limited their choice of studies to only those that bolstered their argument, ignoring many others that would have been identified under their search criteria, and that a greater number of studies, many of which specifically analyzed shared environment, showed that the equal environments assumption (EEA) was valid and that even in situations where the assumption was violated, it had very little, if any, influence on heritability estimates.19 Genetic additivity

Another concern relates to genetic additivity. Genetic variance for a trait is usually described as either additive or non-additive. Additive gene variance is when a number of genes contribute to a single trait, and non-additive gene variance is when the genes interact in a manner such as dominance or epistasis,17 discussed in Chapter 3. Briefly, dominance occurs when alleles at the same locus interact, with one allele being dominant over the other—that is, expressed at the cost of the other; for example, purple flower color is dominant over white in peas. Epistasis occurs when several genes at different loci interact, for example, coat color in Labrador dogs, in which a gene’s phenotypic expression will differ depending on the presence or absence of other genes. Twin studies measure additive genetic variance, but it is probable that non-additive genetic variance is also present, which if not identified, Burt and Simons argued, will artificially inflate heritability and decrease environmental effects.17 Monozygotic twins share the same interacting alleles, but DZ twins may not, so genetic non-additivity could potentially reduce the genetic relationship in DZ twins but not MZ twins, biasing the results.17 In their rebuttal, Wright and colleagues pointed out that although gene and environment interactions (G  ×  E) between additive genetic factors and shared environmental influences will increase heritability estimates, G × E interactions between additive genetic factors and non-shared environmental influences will decrease heritability estimates and instead inflate the estimation of the influence of the non-shared environment. In other words, G × E are equally likely to inflate or deflate heritability estimates and so would not bias the result overall.20 Wright and colleagues suggest that a more important question relates to determining when additive genetic factors and the shared environment interact and when additive genetic factors and non-shared environmental factors interact.20 Overuse of data sets

Another concern raised by Burt and Simons is common throughout criminology, and it is that many studies are based on the same data set. For  example, in the 20  studies considered since 2008 by Burt and Simons, 17 used the US National Longitudinal Study of Adolescent Health (Add Health) data. Burt and Simons argued that despite this being an excellent and rich data set, any deficiencies are replicated in all studies.17 For example, they point out that measurements for delinquent peers are self-reports that ask the respondents to give their impressions of their three closest friends’ substance abuse of just three substances and that questions on bullying refer only to physical violence.17 This is interesting because in their rebuttal, Barnes and colleagues show that much of Simons’ own work is based on a single data set of only 800 respondents and that Simons has frequently praised the Add Health data set.19 Moreover, data sets such as Add Health and many others contain a very rich and large amount of data, which can be mined in many ways to answer a variety of questions and may not truly involve the same data.

90  Evidence for genetic predispositions for criminogenic behavior

Adoption studies A  much more powerful way to look at the effect of genetics on behavior is to look at adoption cohorts. This  eliminates the major challenge to twin studies—that of assumed similar environment. In such studies, we look at the life and behavior of children who were adopted by non-family members and in particular at their criminal behavior in relation to the criminality of their adoptive or biological parents. Adoption better separates environmental and genetic effects. If convicted adoptees have a significantly higher number of convicted biological fathers (assuming appropriate controls), it would suggest the influence of a genetic factor in criminal behavior. This theory is supported by the fact that, in most studies, none of the adoptees knew their biological parents and that adoptees often do not even know that they are adopted.

Early adoption studies Mednick’s Danish adoption studies

One of the largest and strongest adoption studies was performed in Denmark by Mednick and colleagues.23,24 This study examined 14 427 non-familial (that is, without genetic relationship) adoptions in Denmark from 1924 to 1947. The data were based on a register of non-familial adoptions maintained in Denmark, which includes information on the adoptees and their biological and adoptive parents. Court convictions were used as an index of criminal involvement and included details of the law violated, the date, and the sanction. All details of every individual, whether they were the biological parent, the adoptive parent, or the adoptee, were recorded.24 These included sex, date of birth, address, occupation, place of birth, and size of the community into which the child was adopted. These data give an indication of socioeconomic status (SES) and whether the child lived in a rural or urban community. Most of the data pertained to males, simply because significantly more males had criminal records. Overall, the study showed that if neither the biological nor the adoptive parents had criminal records, only 13.5% of the adopted sons were convicted of a crime. If the adoptive parents were convicted but the biological parents were not convicted, this figure increased nominally to 14.7%. However, if biological parents alone were convicted, then 20% of the sons had criminal records. If both sets of parents, biological and adoptive, had criminal records, then the level of conviction in the sons increased to 24.5%.24 Altogether, the findings show a genetic basis, although this is clearly influenced by the environment. In some cases, the adoptive parents did have criminal records, although their overall levels of conviction were lower than that of the national average, and most were one-time offenders, with no criminal record for 5 years before the adoption.24 However, one could say that adoptees in these homes were exposed to a criminogenic environment, and the level is difficult to judge. One should note that the genetic influence of the biological father is completed at birth and so is easier to quantify. Although the overall data clearly indicated that the risk of adoptee conviction was greatly increased if the adoptee had a convicted biological father rather than a convicted adoptive father, the researchers were concerned that even such low levels of criminality in the adoptive environment could skew the results. So, the researchers further analyzed the data, eliminating adoptive families with criminal records, and found that whether the adoptive parent had a conviction or not had little effect on the adoptees’ rates of conviction. The rate of conviction in the biological fathers still had a major effect.24 Although violent crime is rare, the sample size in this study was large enough to allow researchers to separate violent offenses and property offenses.25 A  later reanalysis of these data showed that the genetic relationship is highly significant for property crimes but not statistically significant for violent crimes, indicating a genetic predisposition for property crime but not for violent crimes in this study.25 The researchers also looked at recidivism levels to consider whether parental recidivism resulted in higher conviction rates in the sons. In study by Mednick et al.,24 convictions

Adoption studies  91

rather than arrests were looked at, and chronic criminality was counted as more than three convictions. Just under 5% of the male adoptees were chronic offenders, but they accounted for more than two-thirds of the convictions in male adoptees. The data showed that as the number of offenses in the biological parents increased, so did the number of convictions in the chronic adoptees. The biological parents with zero, one, two, or three or more convictions had male adoptees (that is, male children who were adopted by someone else) averaging 0.30, 0.41, 0.48, and 0.70 convictions, respectively.24 In some cases, a biological parent gave up more than one child for adoption, which allowed the researchers to perform a sibling analysis. The overall research suggests that a heritable component for property crime exists, so it would be expected that there would be stronger concordance for crime between separately adopted full-siblings than between separately adopted half-siblings and that both would show more concordance than non-related adoptees.24 The data set showed that the chance of two randomly selected adoptees, with both having a criminal record, was 8.5%, so the researchers considered that as their baseline. The data showed that the concordance rate between two half-siblings was 12.9% and between full-siblings was 20%. The sample size for this subset of the data is very small, as there were only 131 male-male half-siblings and 40 full-siblings, but the results are in the predicted direction. As the degree of genetic relation increases, the level of concordance increases. The level of concordance shot up even further when considering whether the siblings had a criminal father.24 Very few convicted women were included in the data set, but the researchers did repeat the analyses by using only female data, and they found that although the numbers were small, the relationship held true.24 In fact, it was found that the relationship between the adoptee being convicted and having a biological parent convicted was actually stronger if the convicted parent was the mother. There  was also a stronger relationship between biological mother convictions and female adoptee convictions than that between biological father convictions and male adoptee convictions.24 Although this result is interesting and statistically significant, the authors felt that the numbers were too low to draw strong conclusions. The influence of SES was also considered. Using occupation as a guideline, Mednick and colleagues separated “genetic” social class and “rearing” social class. Biological parents’ social status was found to have a greater effect on risk of conviction in the sons than adoptive parents’ SES. Highest levels of convictions were seen in sons whose adoptive and biological social status were ranked as low, and lowest levels were seen when both genetic and environmental SES were ranked as high. In both cases, as SES improved, the risk of conviction decreased.24 This was one of the first studies to prove that environment does have an influence on criminogenic behavior and to show the beneficial and protective effects of improved SES.23 This was an excellent and extremely thorough study, and the researchers themselves recognized and tried to address as many caveats as possible. The adoptions covered six decades, during which major historical crises occurred, so the analyses were repeated in 5-year blocks, but the results held true. All adopted children were removed from the biological parent immediately and either adopted or placed in a group foster home, so the biological parents’ environment had no influence; but Mednick and colleagues considered it possible that length of time in a foster home might impact later behavior, so they considered age at adoption but again found no impact on the results. The researchers were also concerned that if the adoptive parents were told of the criminogenic background of the parents of their newly adopted baby, this might result in labeling the child. The adoptive parents might subconsciously raise the child differently, based on a knowledge of their biological background, which in turn might affect the likelihood that the adoptee would commit a criminal act. The researchers reanalyzed the data set based on whether the adoptive parents were aware or not aware of the criminal history of the biological parent. If the biological parent was convicted of a crime prior to the adoption, the adoption agency would inform the adoptive parents; however, if the biological parent’s conviction occurred after the adoption, they were not informed. One-third of the convicted biological parents were convicted before the adoption took place, and two-thirds were convicted after the adoption. When the groups were considered separately, the probability of

92  Evidence for genetic predispositions for criminogenic behavior

conviction in either male or female adoptees was almost exactly the same, whether the adoptive parents knew of the biological parents’ criminal records or not. Therefore, whether the adoptive parents knew of the criminal convictions of their adopted children’s parents did not influence whether the children would themselves be convicted of a crime. In summary, this study was extremely well conducted, and every caveat was thought through and reanalyzed. It was a large enough study that in most cases valid statistics could be derived. Bohman’s Stockholm adoption studies, 1996

A number of adoption studies have also been conducted in Sweden, as data on the general populace is also easily accessible there. Bohman looked not just at criminality but also at its relationship to alcoholism and antisocial personality disorder and included 862 men and 913 women who were born between 1930 and 1950. The studies began in the 1960s and continue today.26 A high correlation was found between alcohol abuse in biological fathers and adopted sons, as well as a high correlation between crime and alcohol abuse in biological fathers and adopted sons. When alcohol abuse was studied alone, there was again a high correlation, which indicates a genetic predisposition for alcohol abuse.26 In looking at individual cases, it was seen that alcoholic criminals often committed repeated violent offences, whereas non-alcoholic criminals tended to commit a small number of petty property offenses (Bohman, 1996).26 These non-alcoholic petty criminals more commonly had biological parents with histories of petty crime but rarely alcoholism. In contrast, the risk of criminality in alcohol abusers was correlated with the severity of their own alcohol abuse but not with the criminality of their biological or adoptive parents. Therefore, it appeared that criminality without alcohol abuse was characterized by petty property offenses, whereas alcohol-related criminality was more violent and usually repeated.26 When alcohol was removed from the equation, the correlation between both environment and genetics showed a very marked increase in petty criminality in the adoptees but not in violent crime. When both the genetic and environmental risks were low, the rate of criminality in male adoptees was 2.9%. When the genetic risk was low (biological parents non-criminal) but the environmental risk was high (adoptive parents criminal), the rate was 6.7%. When the genetic risk was high but the environmental risk was low, the rate was 12.1%. When both risks were high, the rate shot up to 40%.26 The  most interesting facet of this study is the interaction between the environment and the genetic background. When only the environment is taken into account, with a low genetic risk, the male adoptees had more than twice the risk of criminality than adoptees with low genetic and low environmental factors. When the genetic risks, but not the environmental ones, are taken into account, the risk was nearly fourfold, but when you add the two, the risk jumps to 40%. These experiments indicate a genetic and an environmental predisposition for both alcoholism and criminality and suggest that there are separate mechanisms at work with each. Bohman also considered the issue of serious mental illness. There  was a threefold increase in the rate of schizophrenia for both male and female offspring of biological fathers convicted of violent offenses.26 Many studies since have shown a genetic link to schizophrenia. A  metaanalysis of 12 twin studies showed a heritability of 81%.27 Bohman also looked at sex differences. Daughters with a criminal biological parent had a threefold increase in criminal behavior over daughters without a criminal biological parent. Female crime mostly involved non-violent property offenses and was usually not associated with alcohol abuse. A lack of violence is common in female crime. The effect of the environment was different for males and females. For example, prolonged institutional care increased the risks of criminality in women but not in men. In contrast, many short-term homes of very low SES increased the risk in men but not  in women. Significantly more of the women had fathers with repeated convictions than did the men.26 In all cases in this study, low SES alone was not enough to lead to petty crime, but it did increase the risk in combination with specific types of predisposition. An unstable home, repeated institutionalizations, or repeated foster homes, did contribute to the risks of both petty criminality and alcoholism.26

Adoption studies  93

Modern adoption studies Further adoption studies continue to be conducted in Sweden. In a recent study, data were obtained from over nine national registers, including the Crime Register, Hospital Discharge Register, Outpatient Care Register, Prescribed Drug Register, Health Care Register, Cause of Death Register, Total Population Register, and Suspicion Register (which includes data on people suspected of committing a crime), and individuals were tracked through all registers based on their unique identification numbers.28 A wide range of criminal convictions was considered, ranging from extremely violent crimes to petty offences. Over 18 000 adoptees were compared with 79 206 biological and 47 311 adoptive relatives. The risk for criminal behavior overall was greatly increased in an adoptee who had biological parents with at least one criminal conviction, as well as in the adoptees’ biological full- and half-siblings. The risk of criminal behavior overall was very slightly higher in adoptees with criminal biological mothers than fathers and was significantly higher in biological full- and half-siblings of adoptees with criminal convictions. The earlier twin and adoption studies showed genetic links only to petty or property crimes, but this study and others have shown links to violent crimes as well. The risk of violent crime in this data set was considerably increased in adoptees who had a parent with at least one violent criminal conviction, with no differences seen between convicted mothers or fathers. Similarly, the biological full- and half-siblings of adoptees convicted of at least one violent crime also had an increased risk of conviction for a violent crime.28 Both genetic and environmental risk factors greatly increased the chance of criminal behavior in the adoptee, with genetic risk including criminal behavior and alcohol abuse and environmental risk including adoptive parents’ health, divorce, or death. The  two risk factors compounded each other, and some specificity was seen in genetic risk for both violent and non-violent criminal behavior.28 Overall, risk for violent and non-violent criminal behavior had a similar heritability, in contrast with Mednick’s and Bohman’s studies. Importantly, this study again showed the impact of environmental factors—such as low adoptive parental education, death or illness of a parent, and divorce—on later criminogenic behavior; this had been shown before, but this study is more powerful, as it removes the genetic elements from such factors, showing only the significant effect of the environment. Moreover, adoptee criminal behavior predicted criminal behavior in the adoptive siblings and vice versa, again showing the impact of the shared environment. This study did not show an interaction between environment and genetics, in that increased environmental risks did not change the risk of criminal behavior, whether the adoptee had low or high genetic risk,28 in contrast with the results in Bohman’s study.26 In another study using Swedish data, both adopted sons and biological sons raised in the biological home were analyzed for risk of criminal conviction. The consideration of both adopted sons and sons raised in the natural home allows for greater generalizability to the general population.29 The study included 35 years of criminal data, as well as regional data showing where children were born and the area in which they were adopted, included biological and adoptive family members, and involved approximately 7.5 million individuals. These were then cross-matched to Sweden’s official crime registry, which records all crimes that are serious enough to go to court and in which a conviction was secured. Crimes were divided into violent, property, and other, and recidivism and length of sentence were also considered.29 The results showed that whether the biological parents had criminal convictions had an impact on the criminal behavior of their sons. Adopted sons with a convicted father had a 12.1% increased chance of being convicted, and those with a convicted mother had a 13.4% increased risk.29 The results were very similar for own-birth children, indicating the importance of the biological parents. The relationship held for both single and multiple convictions and in all crime categories, not just those of property crimes, again contradicting Mednick’s Danish study, which found a link only to property crimes. Also, again contrary to Mednick’s study, a significant effect of the adoptive environment was seen, particularly related to the adoptive mother, in that criminal convictions in the adoptive mother had a significant impact on risk for criminality in the adopted sons. The author suggests that this may be because the Mednick study only covered 3 years of criminal data, which might reduce the amount of data on adoptive

94  Evidence for genetic predispositions for criminogenic behavior

parents’ criminal behavior, whereas this study included 35 years of data. The authors tested this hypothesis by breaking their data into cohorts of 3 years and were rarely able to discern an impact of the adoptive parents’ criminality over such short time frames, supporting this hypothesis. Very little interaction was seen between the biological and adopted parents’ criminal histories. Notably, having more highly educated adoptive parents appeared to mitigate the risk of having criminal biological parents, indicating that higher education in adoptive parents is a protective factor.29 Impulsivity

Impulsivity is often a risk factor for criminogenic behavior, as impulsive people do not think before acting and may not consider the consequences of their actions. Impulsivity is extremely important in considerations of antisocial and aggressive behavior, as well as attention deficit hyperactivity disorder (ADHD) and other psychopathologies.30 In a meta-analysis of 41 twin, family, and adoption studies that included over 27 000 individuals from toddlers to adults, risk for impulsivity was found to be equally related to genetic (50%) and non-shared or unique environmental factors (50%).30 The genetic factors were primarily additive (38%), with 12% being non-additive. A shared environment did not have any effect, in that any similarity between twins or family members related to genetics rather than a shared environment.30 Psychopathy

Psychopathy and psychopathic personality characteristics, such as callous-unemotional traits, lack of remorse or empathy, and guiltlessness, are strongly linked to criminal behavior.31 It has been suggested that psychopathy and psychopathic personality traits are the strongest, most robust, and most reliable risk factors for predatory and violent crimes.31–33 Several studies have shown that many of these traits are under genetic control. A meta-analysis of 10 twin studies showed that 48% of the variance in psychopathy and psychopathic personality traits was under genetic influence and 52% was under unique or non-shared environmental influences, with no influence of shared environment.34 Psychopathy was further examined using an adoption cohort from the US National Longitudinal Study of Adolescent Health (Add Health) in comparison with arrest records, although conviction data were not available. No data about type of crime were available, so offences ranged from petty to serious violent crimes. The results showed that having a biological criminal father increased the risk for psychopathic personality traits in adopted-away sons but not in adopted-away daughters.31 Moreover, there was no link between criminal mothers and psychopathic personality traits in the adopted-away offspring. The relationship between criminal biological fathers and psychopathic personality traits in the offspring, however, was extremely high, with the risk of an adopted-away offspring of either sex scoring in the top 25% on the scale of psychopathic personality traits doubling with a criminal biological father, in comparison with a non-criminal father. When male and female offspring were separated, the results were significant only for males, where the risk increased by a factor of 4.3.31 When only the offspring scorings in the top 10% of the psychopathy scale were considered, the results held true, with a criminal father resulting in a slight increase in overall risk of scoring in the top 10% and with a slight reduction in risk for females, but risk in male offspring increased by a factor of 8.5.31 The authors speculated that this dramatic increase in risk suggested that the genetic influences were strongest in the most violent psychopaths.31 The authors suggested that the gender differences seen in this study may relate to the fact that female psychopaths are very rare, and it has even been speculated that the etiology for psychopathy in females is different from that in males.31 Substance abuse

Criminal behavior is highly correlated with alcohol and drug abuse, and several adoption studies have been used to look at various aspects of substance abuse. Alcohol use disorder (AUD) is an overarching description of various levels—from mild to severe—of alcohol misuse, including alcoholism, and includes any use that puts a person’s health or safety at risk. Twin studies have confirmed that there is a strong familial link to AUD, although they have not been able to fully elucidate the role of the environment.35 Early adoption studies on AUD in Denmark, Sweden, and

Adoption studies  95

the United States also showed a genetic link between adopted offspring and biological parents, but results for the effects of the environment were mixed.35 In a large Swedish adoption study of over 18 000 adoptees, using eight different data registries, AUD was found to be strongly genetically linked to biological parents and siblings, but it was also environmentally linked to adoptive parents, showing a strong linkage of both genetic and environmental components to the development of AUD. Alcohol use disorder in adoptees was also significantly predicted by drug abuse, psychiatric illness, and criminal behavior in biological parents and by drug abuse, criminal behavior, and early death in adoptive parents.35 Therefore, AUD was found to be strongly genetically linked and was also separately predicted by drug abuse and criminal behavior in the biological parents, suggesting that AUD is part of a larger range of externalizing behaviors.35 This study also found a strong environmental link to AUD, with links to AUD in the adoptive parents as well as to low SES, low parental education, and death of an adoptive parent. Even when adoptive parent AUD was removed from the analyses, adoptive parent criminal behavior and early death still increased the risk of AUD in the adopted offspring. This study expanded upon the normal adoption experimental design to include not-lived-with parents and stepparents and found that each impacted the offspring very similarly to biological and adoptive parents, respectively, only with greater significance. A history of drug abuse or criminal behavior in not-lived-with fathers increased the risk of AUD in offspring, and stepparents who had AUDs, low levels of education, and suffered an early death also strongly impacted the risk of AUD in the offspring.35 Unlike several earlier adoption studies that found an interaction between genetic and environmental influences, in this study, genetic and environmental factors were additive, not interactive. This study also identified three separate classes of AUD with different genetic influences: class 1, mostly female, with high levels of psychiatric illnesses in biological parents; class 2, mostly mild, non-recurrent cases; and class 3, mostly male, with early onset, highest link to criminal behavior, high recurrence and link to drug abuse, and biological parents with high rates of AUD and drug abuse, criminal behavior, and low educational level.35 Kendler and colleagues’ class 3 is similar to the earlier categorization by Cloninger and colleagues of Type II alcoholism, which affects mainly the male offspring of male alcoholics, is almost entirely genetic, begins in adolescence, and is often linked to criminal behavior.35,36 Class 1 is more similar to Type I alcoholics in that it is linked to both genetic and environmental factors and begins later in life.35,36 In a meta-analysis of twin and adoption studies on AUD, the results showed that, in 12 twin and 5 adoption studies, AUD was consistently found to be approximately 49% heritable, and these results were consistent in both men and women.37 As a number of different studies were considered, with different experimental designs, data sets, and in different countries, the results were fairly robust. A small (approximately 10%) shared environment effect was also seen.37 Earlier twin studies from Australia and the United States have shown that there is also a strong genetic basis for drug abuse, which has been confirmed in adoption studies. For example, the same Swedish cohort of adopted children, biological and adoptive parents, and siblings ­mentioned above was also assessed for drug abuse.38 Medical, legal, and pharmaceutical registries were searched for evidence of drug abuse in this cohort. Adopted children with biological parents who were drug abusers were at significantly greater risk of drug abuse. Risk was also greatly increased in the ­biological full- and half-siblings of adoptees who were drug abusers and in the adoptive s­ iblings of adopted children who were drug abusers. Risk for drug abuse in the adopted children was also increased not only by drug abuse in the biological family but also by alcoholism, ­psychiatric ­illnesses, and criminal history in the family.38 However, certain factors in the adoptive family ­environment also increased risk, for example, if the bond between the adoptive family and child was damaged or broken by premature death, divorce, or by adoptive parent’s alcoholism or ­adoptive sibling’s drug abuse, showing that the etiology of drug abuse is multifactorial and that genetic background has not  only a direct impact but also increases susceptibility to adverse environmental influences.38 The  increased risk from drug abuse in adoptive siblings rather than parents also suggests the ­potential influences of deviant peer relationships.38

96  Evidence for genetic predispositions for criminogenic behavior

Cautions for adoption studies Adoption studies seem to be excellent natural experiments to allow us to gain a greater understanding of the etiology of many biological factors, in particular for our interests, those of behavior. Nevertheless, they are not perfect. In twin studies, we assume that the environment is identical for twins raised together, allowing us to separate the genetic from the familial or environmental factors. We have already discussed, however, that this shared environment may well not be identical. In adoption studies, we completely separate the biological or genetic background from the adoptive environment. However, cautions must be raised when considering these studies, for several reasons. Differences between biological and adoptive parents

Adoptive parents are often quite different from biological parents for several reasons.15 Adoptive parents are carefully screened for suitability.15,17 Adopting a child is expensive, so adoptive parents are frequently quite wealthy and well established. They are usually a couple, and the strength of their marriage has been assessed. They have also demonstrated that they are free of mental illness, substance abuse, or a criminal history. They are frequently older couples who may have attempted to have biological children for some years before turning to adoption.15 People with any potentially risky behaviors or adverse environmental factors will not be selected for adoption. Biological parents are not screened in any way. They are likely to be much younger, less wealthy, and less established and have not been screened for any adverse conditions, such as substance abuse and criminal history. Also, adoptive parents have a very strong desire to parent a child and have made a strong commitment to do so,15,17 whereas biological parents may not have considered parenthood, may have accidental pregnancies, and are more likely to be single parents. This is probably particularly likely when dealing with a parent who is giving up a child for adoption. In their arguments against using heritability studies, Burt and Simons stated that adoptive families involved in studies have agreed to participate in these studies and suggest that this involves a level of self-selection, as adoptive families who provide an adverse environment would not volunteer.17 However, this is not true in most studies, as they are based on massive data sets that in some countries, such as Finland, Denmark, and Sweden, are based on nationwide public registries in which all data are recorded, and researchers extract the data of interest—for instance, adoption and criminal records—with the participants having no knowledge of the study. In other data sets, such as Add Health, data are collected on a very large number of factors, with participants having no knowledge of what particular area is being analyzed in a specific research project. Therefore, in such cases, no family has “volunteered” to be involved in an adoption study, so no self-selection has occurred. Late separation

All adoption studies assume that the baby was adopted shortly after birth and had no contact with the biological family, and although at least the former can be determined by records in some studies, it is not always known. Burt and Simons argued that, if the child is not separated from the biological parents immediately after birth, this will bias the analysis, for two reasons.17 The first is that adoption studies are based on the assumption that biological parents provide only the genetic contribution to the child and all environmental influences come from the adoptive parents. However, if the child is not removed from the biological parents immediately, then the biological parents will provide both genetic and environmental contributions to the child, confounding the environmental influence. The second reason is that if a child is not adopted immediately, the actual adoption event may have a significant disturbing influence on the child if the child has already bonded to the biological parents. Also, there is a possibility of a stigma attached to being an adopted child, which may confound results.17 However, although Burt and Simons specifically single out the Mednick Danish Adoption studies as including late-separation adoptions, they are incorrect, as Mednick and colleagues had already considered this concern almost 45 years ago. In order to further explore

Genetics and behavior overall  97

this area, they reanalyzed their data and found that none of the adopted children spent any time in the biological home, because the newborn babies were either immediately placed in the adoptive home or placed in a group foster home for later adoption. Therefore, the children studied did not spend any time in the biological home, eliminating any environmental effect of the biological home on the child.24 Mednick and colleagues then went further to examine whether, as Burt and Simons suggested, there was any effect of early institutionalization, which might have had an effect on behavior. Mednick and colleagues determined that of those who were not immediately adopted, over 50% were placed with an adoptive family within the first year, 12.8% were placed with an adoptive family in the second year, and 11.3% were placed after the age of two.24 The researchers reanalyzed all the data, separating these groups, and found that the relationships between adoptive offspring and biological and adoptive parents held for all groups, so there was no effect of age of transfer to new homes.24 Selective placement

Burt and Simons also argued that although adoption studies assume that children are not selectively placed in adoptive homes that are similar to those of the biological parents, this may not be the case, as adoptive agencies may try to fit the child into a similar family style.17 They suggest that children from the least criminal families may be placed in the “better” adoptive homes, and those from families with higher levels of criminality and mental illness may be placed in less desirable adoptive homes. This seems extremely unlikely; although adoption agencies may try to fit a child into a similar type of family, it is highly improbable that a less desirable adoptive family would be approved for adoption in the first place. Prenatal influences

The prenatal environment is extremely important to the overall health and success of the subsequent child, as we will discuss in Chapter 8. Poor diet; exposure to pollutants, toxins, drugs, or alcohol; lack of prenatal care; and poor birth conditions and care can impact the psychological and physical outcome of the child. Burt and Simons argued that negative prenatal influences are more commonly found associated with social factors linked with criminal behavior, such as low SES, neighborhood adversity, and substance abuse. They argue that these prenatal environmental factors are not being considered in adoption studies, as all influences from the biological parents are considered genetic. This could artificially conflate the genetic influence and reduce the environmental influence in calculations.17

Genetics and behavior overall An overall criticism of twin and adoption studies is that they seek to partition the variance associated with a trait as either genetic or environmental, even though we know that in most cases there is an interaction between the two, so they are not two distinct influences.17 Most recently, molecular research has shown that the human genome is not static, as previously thought, but is in fact dynamic and molded by the environment,39 as we will discuss in the next chapter. However, this is fully accepted and even embraced in heritability studies, as they clearly show the role of the environment and, importantly, remove a major problem with studies focused on environmental causations only: that of selection. Moffit and Beckley,21 in their rebuttal paper to Burt and Simons’ papers,17,18 point out that twin and adoption studies allow researchers to clearly separate the effects of selection from actual social causation.21 Selection occurs in many ways.21 Antisocial behavior provokes a reaction in other people, such as an aggressive child being severely disciplined or an antisocial adolescent being rejected by peers. Also, antisocial people look for environments that suit their behavior, such as a bully seeking out similar peers, and antisocial individuals are often placed in more criminogenic environments, such as offenders being imprisoned with career criminals. Selection therefore accounts for

98  Evidence for genetic predispositions for criminogenic behavior

much of the relationship between the environment and criminal behavior, and Moffit and Beckley argue that twin and adoption studies allow us to separate selection from the true environment.21 In a major review of behavior genetics research, Moffit concluded that genes influence approximately half of the population variance for antisocial behaviors, based on a large number of recent heritability studies.40 It is important to point out that this is an average of many studies and that it relates to the population, not to the individual. Burt and Simons argue that it is not possible to separate the effects of the genes and those of the environment, as the two interact so much,17 but Wright and colleagues correctly point out that this is not the goal of heritability studies. Heritability studies do not attempt to separate the genetic and environmental effects within an individual but within a population.20 Heritability coefficients relate to the population being studied rather than to individuals.40 Heritability studies have also shown us the important role of the environment in antisocial behavior and that interactions occur between and within genes and the environment. This leads us to the idea that many previous sociological assumptions about the role of the environment in crime causation may be wrong, as previously assumed risk factors may only be risk factors when associated with particular genotypes.40 Heritability studies have also shown us that shared environmental factors are responsible for about 20% of the variance in antisocial behavior and that non-shared or unique environmental influences are responsible for about 20% to 30%.40 Despite criticisms, a multitude of studies using different data sets, birth cohorts, and methods have shown the same overall results—that criminogenic behavior is approximately 50% heritable.40 This plethora of studies from numerous different researchers, countries, and cultures shows that these results are robust. Perhaps more important than the actual number is that antisocial behavior has a large heritable component. It is important to remember that many traits are involved in antisocial behavior, and each trait is governed by a very large number of genes, each of which contributes a very small amount of variance to the whole. Moreover, these genes interact with each other and with the environment. Understanding the role of genetics and the environment in antisocial behavior allows us to develop a better understanding of potential protective factors.

Protective factors Twin and adoption studies show us not only the role of genetics in behavior but, perhaps much more importantly, also the role of the environment and sometimes the interactive effects of the two. Understanding these factors can help us understand and develop appropriate intervention strategies in order to help at-risk youth. The cycle of violence is well known, in that offspring of criminal parents are more likely to exhibit high levels of antisocial and criminal behavior and are at much greater risk for arrest and conviction themselves. Many factors have been shown to play a role in this transgenerational transmission of criminal behavior, such as SES, parental stress, domestic abuse, parental education, and exposure to parental criminal behavior. Twin and adoption studies have shown that there are also a considerable number of genetic factors at play. However, environmental factors are also extremely important and give us greater insight into methods of intervention. In  an effort to elucidate the role of the rearing environment, over 1100  full-sibling and over 3000 half-sibling sets from families deemed to be at high risk were identified using several Swedish national registries. Families were deemed to be at high risk if at least one parent was registered as a drug user, had an AUD, or had a criminal conviction. In each sibling set, at least one sibling was reared by the biological parents and one had been adopted. The results showed that risk for criminal conviction was much higher in offspring who were raised by the high-risk biological parents than for siblings or half-siblings who had been adopted.41 In full-siblings, adoption reduced the risk of criminal conviction by 44% in comparison with the non-adopted siblings. If both parents of full-siblings were high risk, then the protective factor of adoption increased even further. In halfsiblings, adoption reduced the risk of conviction by 40% in comparison with the biologically raised half-sibling, and as with full-siblings, the greater the biological parental risk, the greater the ­protective value of the adoptive home. If an adoptive parent was considered to be at high risk,

Protective factors  99

the protective factor decreased. This study showed the protective effects of a good rearing environment and that these protective effects were even more valuable in offspring with a high biological family risk for offending.41 The authors speculated on the interpretation of these results. Several previous studies have shown that adopted siblings have a higher IQ and better school performance than their biologically raised siblings (reviewed in Kendler, Morris, et al.41), and it is well known that school performance is a major protective factor. Also, as with most countries, Sweden has an extremely thorough screening for potential adoptive parents, including assessing the expected durability of the marriage as well as the health and personality of the potential adopters. Known risk factors in biological parents for offspring criminal behavior include low SES, criminal history, substance abuse, divorce, single parenthood, mental illness, young maternal age, and low education. Adoptive parents are deliberately chosen to exclude such factors. In some cases in this study, some adoptive parents did have risk factors such as substance abuse or divorce, which were found to increase risk in the adopted offspring, supporting the environmental role of these factors.41 In a further examination of this same cohort, Kendler and colleagues looked at the potentially protective effects of adoption on drug abuse.42 The results showed that the risk of the offspring of high-risk parents becoming drug abusers was greatly reduced in adopted full- and half-siblings in comparison with their biologically raised full- or half-sibling, with a 45% decrease in risk in both full- and half-siblings. As seen before for criminal conviction, the protective factor was even greater when the offspring had two high-risk biological parents but was significantly decreased in the event of death, divorce, or drug abuse in the adoptive parents, strongly suggesting that the protective effect relates to the quality of the rearing environment.42 This study again shows that children who have high genetic risks for antisocial behavior can be protected by a high-quality environment.42

Family and social bonding As we will see in the following chapters, many genotypes and candidate genes have been implicated as risk factors for a variety of antisocial behaviors, and some have been linked to specific pathways, such as genes that are involved in neurotransmitter synthesis, transportation, and metabolism. In most cases, a particular allele of the gene increases risk, which indicates that other alleles for the gene are in fact protective. Moreover, a great deal of research has shown that a favorable environment can ameliorate or eliminate the expression of antisocial behavior. For  example, two genes involved in the production of the neurotransmitter dopamine (DRD2 and DAT1, which we will discuss in Chapter 9) have been linked to antisocial behavior in many studies. DRD2 codes for a dopamine receptor, and certain of its alleles have been linked to substance abuse, ADHD, antisocial behavior, and impulsivity.43 DAT1 is involved in dopamine regulation, and some alleles of this gene have been linked to ADHD, antisocial behavior, and substance abuse.43 A  large US longitudinal study, the National Youth Survey Family Study, followed participants from adolescence until they were in their mid-40s. Data were collected via questionnaires and indepth interviews annually, then every 3 years, and then at 10 years. Buccal (mouth) swabs were used to collect genetic information. Researchers used these data to compare genotypic risk, in the form of DRD2 and DAT1 high-risk alleles, with a number of environmental characteristics over the life course.43 Environmental factors included many measures, which were grouped within family closeness, school attachment, peer delinquency, and neighborhood disorganization. Other factors considered included parental unemployment, poverty, post-secondary education, household size, and ethnicity. Overall, the results showed that stable and healthy social environments could ameliorate the risks associated with adverse genetic backgrounds. Family closeness and school attachment were found to reduce the impact of genetic risk, particularly on serious and violent crime for both DRD2 and DAT1.43 For example, the high-risk allele of DAT1 was not linked to violent antisocial behavior in favorable environments but was very significantly linked to violence in those who lived in highly disorganized communities. Participants with one or two of the high-risk alleles for DAT1 were examined in comparison with family closeness or school attachment and serious delinquency, and

100  Evidence for genetic predispositions for criminogenic behavior

participants with one or two of the high-risk alleles for DRD2 were examined in comparison with family closeness or school attachment and violent delinquency. In all four scenarios, the high-risk genotype for either gene did not increase risk unless the individual had no attachment or closeness to family or school. It is important to note that this did not relate to level of social support, just to lack of support, suggesting that, in most situations in which an individual received a normal or typical level of support, the presence of the risky genotype did not result in a significant increase in violent or serious antisocial behavior.43 This is significant because it suggests that any intervention that even slightly increases school or familial bonding or provides some form of social support will greatly improve outcome. These adolescents do not have to be raised in a perfect environment; they just need some social support to overcome the genetic risks associated with these genes.

Resilience When studying biological, particularly genetic, influences on behavior, we are inclined to think of genes as risk factors for antisocial behavior and the fact that genes have been selected for their advantageous qualities by natural selection is often forgotten. Although some behaviors considered undesirable in the modern human world may actually confer an advantage (such as theft for accumulation of resources to attract a mate), most behaviors that are selected for are truly advantageous, and many actually confer protection against risky behavior.44 In other words, many genetic influences may support prosocial behavior. Although some behaviors that may once have been advantageous, such as aggression, are considered maladaptive today, one behavior that has probably always been advantageous is that of resiliency to victimization.45 Victimization is highly disadvantageous: it can result in physical and mental injury and even death, and it may result in loss of social status with consequent loss of mating opportunities. Thus, resiliency to victimization has probably always been highly advantageous, today and in our distant past. From the point of view of natural selection, people who are resilient to victimization are more likely to live longer, healthier lives and have more mate choices and therefore have greater reproductive opportunities, all of which confer an evolutionary fitness advantage.45 From a more immediate point of view, bullying causes a wide range of negative consequences, physically, psychologically, and emotionally, and can result in long-term negative sequelae. Victimization, particularly that of children and adolescents, is extremely common, so a great deal of research has been devoted to the causes and to factors that protect adolescents, such as school and family bonding, neighborhood cohesion, parental monitoring, and social connectivity, as well as personality features such as assertiveness and friendliness.45 Beaver and colleagues have suggested that resiliency is also related to genotype. For example, when two people are exposed to a certain risky environment, their reactions will be dependent on their genotypes. One may be genetically predisposed to commit a criminogenic act, while the other may be genetically resilient to a criminogenic environment and may not exhibit criminal behavior. As we will discuss in detail in later chapters, recently, certain genotypes have been found to be highly vulnerable to environmental conditions such as severe child abuse, whereas other genotypes have been found to be protective or resilient.46 Since then, other genes have been found to confer resilience to other factors such as depression47 and disease. Genes related to neurotransmitter function have been shown to increase resilience to victimization. Serotonin is an important neurotransmitter. Increased serotonin levels are associated with increased calmness, decreased antisocial behavior, and general well-being, whereas decreased serotonin levels have been linked to increased suicide ideation and risk, as well as increased impulsivity and antisocial behavior.48 As we will discuss in more detail in Chapter 9, the synthesis, function, and metabolism of any neurotransmitter are under the control of many genes, and altered function in any may impact neurotransmitter function. The serotonin transporter gene (5-HTT) has two major alleles, referred to as the shorter (S) and longer (L) alleles, so a person may be SS, LL, or SL for this gene. The S allele is less functionally efficient than the L, and studies have shown

Protective factors  101

that individuals with the S allele have increased reactivity to stress, fear conditioning, hormonal response to stress, and increased sensitivity to threat, whereas individuals with the L allele have reduced reactions to stress and selectively avoid stressful situations and seek positive environments.49 In a study of over 2200 British children in the Environmental Risk (E-Risk) study cohort, variations in this gene were found to moderate the impacts of bullying.49 Children with the LL or SL genotype were much more resilient to frequent bullying than children with the SS genotype, who were found to have much greater risk for developing emotional problems by the age of 12 years. The  results held true even when controlling for other risk factors and emotional status prior to bullying. However, no differences were found in children who were only occasionally bullied.49 The study showed that certain genotypes were more susceptible to the negative effects of bullying, while others conferred protection. Resilience may involve factors that limit or prevent being victimized, but it is also involved in reducing or ameliorating the negative effects. Many children are bullied—in fact 30%–50% of adolescents in the United States report being bullied45—and bullying can result in many emotional and behavioral issues, including a variety of internalizing and externalizing behaviors, as well as suicide ideation and self-harm.50 However, many children who are victimized do not experience such negative effects and are considered resilient. The same is also true for post-traumatic stress disorder (PTSD), as a large number—in fact, the majority—of people suffer very traumatic events at some time in their lives, yet only a small percentage actually develop PTSD; those that do not are resilient.51 Several studies have shown that adolescent victimization is strongly genetically influenced. A study of twins using the Add Health data set showed that genetic factors were responsible for 40%–45% of the variance in adolescent victimization and over 60% of repeat victimization.52 Add health is a nationally representative sample of American adolescents from 132 schools, who were in 7th–12th grade in 1994–1995. Three waves of data collection occurred over a period of 6 years, including self-report questionnaires, interviews, and a buccal swab for genotyping, with over 2500 adolescents involved in all three waves.45 Studies have consistently shown that individuals who are engaged in criminal activity are much more likely to be victimized themselves than those who do not.45 Other risk factors include low self-control, having delinquent peers, lack of a maternal bond, and lack of social support.45 In a follow-up study that continued to use the Add Health data set, four specific genes were examined, all of which have been shown to have polymorphisms that increase risk for antisocial behavior.45 The results showed that several of the polymorphisms increased resiliency, and although the studies were preliminary, they suggested possible mechanisms for understanding resilience to victimization.45 Twin studies have shown the role of the environment in emotional and behavioral resilience in children. A  study on victimization resilience in 10- to 12-year-old children, utilizing data from 1100 twin pairs and their families in the E-Risk study from England and Wales, indicated that family influences such as maternal and sibling warmth and a positive family environment reduced the negative effects of victimization on bullied children.50 Positive family relationships may help a child cope and respond to bullying. Sibling support was found to be particularly important, even more so than maternal support. The results support the need to include families in school intervention programs and indicate that stable and supportive environments can greatly reduce the impact of victimization. Low SES in childhood is a well-known predictor for antisocial behavior and other cognitive and functional problems. However, as we see frequently, not  all children who come from a deprived background have poor outcomes; in fact, many are resilient. Resilience and vulnerability to SES deprivation were studied in over 1100 twin pairs of 5-year-olds from the E-Risk study. As was seen from previous research, children from lower SES backgrounds had lower IQs and higher levels of antisocial behavior than children from less poor environments.53 Protective factors such as a more outgoing temperament, increased maternal warmth, and engagement in stimulating activities increased a child’s resilience to risk.53 Resilience was found to be under both genetic and environmental control. Overall, additive genetic effects accounted for approximately 70% of the variation in behavioral resilience and 46% of the variation in cognitive resilience.53 This was primarily related to a child’s outgoing personality. The authors suggested that the child’s temperament could

102  Evidence for genetic predispositions for criminogenic behavior

be influenced by either passive or active gene-environment correlation. A  passive gene-environment correlation would occur if both cognitive resilience and outgoing temperament are associated with each other due to both being contributed by their parents. Parents, therefore, provide both the genetic and environmental factors that would contribute to an outgoing personality, and earlier research has shown that very social parents provide a more stimulating environment for their offspring.53 Active gene-environment correlation would occur if the child’s genotype results in a personality that elicits a reaction from caregivers. For example, a more outgoing child may elicit a more positive and stimulating reaction from parents and teachers, which could improve cognition and social functioning. Environment also played a role, in particular, that of maternal interaction using stimulating activities, which increased cognitive function—although this could also be a function of genetics, as maternal interaction may relate to maternal IQ, which is partially heritable. Overall, the research shows that increasing and improving maternal interactions and providing stimulating exercises can increase resilience to SES adversity.53 This is important when developing interventions, and it suggests that relatively simple and inexpensive methods of improving a child’s resilience to SES adversity will improve cognitive function.53 The research also showed that a child’s personality can evoke increased protective factors and promote a more nurturing and prosocial environment, and resources should be provided to allow such children to reach their full potentials.53

Education and school experiences It is well known that poor academic achievement in school is a major risk factor for later criminogenic behavior. For example, early conduct problems were assessed at ages 5, 7, and 10 years and followed up at age 18 years in an English and Welsh birth cohort of over 2000 twins. Participants with a childhood history of severe conduct problems were much more likely to have higher numbers of convictions, cautions, substance abuse problems, and psychosocial problems at age 18 years than those without a childhood history of conduct problems, with concomitant poor outcomes on a number of levels and in different areas of life.54 Education can be a very strong protective factor against risk of antisocial behavior and has been shown to ameliorate or even entirely mitigate biological risk factors. Many studies have shown that receiving a good education reduces criminal behavior, and these results are robust across crime types and type of schooling.55 For example, in a study of mandated years of schooling or mandated lower school-leaving age, increased schooling resulted in lower crime rates.56 Increased education results in reduced criminal activity for a number of reasons. A good education greatly increases an individual’s chances of employability and higher remuneration. Increasing personal human capital results in higher earnings; this makes crime a greater risk, as it will result in lost wages due to time spent incarcerated and overall will lower employment expectations.55 Education increases job prospects and wages, so it reduces the need or desire to commit crime for financial benefit. Education is also believed to have direct impacts on the individual, in that more highly educated individuals are more risk-averse and patient than undereducated individuals and are therefore less likely to risk committing a crime and more likely to invest time in increasing education for higher wages.55 Success in education increases the chances of developing good peer relationships. Moreover, actual school attendance reduces the time available to commit crimes.55 Investing in education has been calculated to be much more cost-effective than incarcerating offenders, and it has many other obvious benefits.55 However, there are many types of crime, with very different motivations, and it has been suggested that education impacts different types of crime in different ways. A study using Danish twin data showed that completing upper secondary (high school) education reduced the adult conviction rates of males by 9.5% for crime overall and 8.0% and 2.5% for property and violent crime, respectively.55 Female crime was slightly reduced, but involvement in juvenile crime was also a strong predictor of later criminal behavior in both sexes. Children with attention deficit hyperactivity disorder (ADHD) invariably do poorly in school due to a range of behaviors, including inattentiveness, impulsivity, hyperactivity, antisocial behavior,

Protective factors  103

and restlessness. Such behaviors are common in most children on occasion and to a limited extent, but children with ADHD display such behaviors consistently, although severity ranges from mild to severe.57 Many studies have shown that ADHD greatly increases risk for academic failure, increasing antisocial behavior, and eventual criminal behavior,57 with approximately 7% of children in the United States afflicted.58 ADHD is also linked to conduct disorder (CD) and oppositional defiant disorder.57 In  a study of over 4600  men in a birth cohort from Finland, ADHD was linked to later criminal conviction and was accentuated by alcohol use and academic failure; this suggests that substance abuse and a lack of school bonding and engagement exacerbate the symptoms of ADHD, so early intervention and programs to increase prosocial bonding may improve academic success and reduce later risk.59 In a retrospective study utilizing Add Health data, ADHD in both males and females was found to increase antisocial behavior, which resulted in negative school experiences and relationships, creating a vicious circle, as these negative experiences then increased poor outcome. This supports the theory that ADHD impacts school experiences by impairing social bonds. However, this poor outcome was ameliorated by increased school factors such as attachment and bonding to school, teachers, and peers, and feelings of safety, fair treatment, and happiness at school, but only for females.57 This supports social bonding theory, in that children who have positive school experiences and feel socially attached are less likely to exhibit antisocial or criminal behavior; however, in this case, it was only favorable for girls. This suggests that intervention to promote and increase social bonding and engagement with school and peers may reduce antisocial behavior and increase prosocial behavior, which will improve outcome, at least in girls. More research is needed to explore the gendered result.57 Not only have education and the resulting outcomes been shown to reduce criminal behavior, but the school environment itself has been shown to be protective. In a study of at-risk maltreated adolescents, school engagement had a significant protective effect, after controlling for ADHD, poor peer relations, and attachment to carer. Increased peer deviance increased delinquency, but a strong carer bond provided protection from deviance.60

Enrichment programs Improving a child’s nutrition both prenatally (before birth) and postnatally (after birth) has been shown to not only improve health but also improve behavior in a number of ways, as we will consider in Chapter 11. Moreover, cognitive functioning can be improved with exposure to more stimulating environments, and improved cognition has many benefits, including reduced antisocial behavior and overall improved educational outcome. A number of studies have considered enriching the environment to increase protective factors. As we will consider in Chapter 8, the prenatal environment is extremely important to the physical, psychological, and behavioral welfare of the offspring, and many interventions during this period have been shown to improve subsequent behavioral outcomes. For  example, a randomized controlled trial of low-income, unmarried, first-time pregnant women involved prenatal and postnatal home visits from nurses to teach the expectant mothers about prenatal and postnatal care, nutrition, the risks of smoking and drinking on the fetus, and how to reduce these activities. The results showed a dramatic reduction in child abuse by age 2,61 and at age 15, offspring were 63% less likely to be criminally convicted in comparison with a control sample and also had fewer incidences of running away, fewer arrests, fewer sex partners, and less alcohol and cigarette use.62 Mothers were also significantly less likely to be child abusers or substance abusers.61 This was the first trial in an initiative that has grown into the nurse-family partnership, which is now found in 594 counties in 42 states in the United States63 and is also seen in other countries.61 In a study involving environmental enrichment, 100 three- to five-year-old children in Mauritius were provided with improved nutrition, increased physical exercise, and education focusing on improving cognitive skills over a 2-year period. By age 11, the children had higher levels of arousal and orienting,64 which improve information processing, and an almost 35% reduction in criminal activity at age 2365 in comparison with a matched control group.

104  Evidence for genetic predispositions for criminogenic behavior

In  a US-based preschool intellectual enrichment program, 123  children from deprived homes were divided randomly into treatment and control groups. The  children in the treatment group attended a daily preschool and had weekly home visits over a 2-year period. The preschool and home visits were designed to increase cognitive skills, such as reasoning and thinking, and improve mental stimulation.61 By age 19, the children from the treatment group had higher high school graduation rates, higher post-secondary attendance rates, higher employment rates, and lower arrest rates than children in the control group. By age 27, those from the treatment group had 50% fewer arrests, higher home ownership rates, and higher wages than those from the control group, and by age 40, participants from the treatment group had many fewer arrests for violent, property, and drug crimes; fewer overall arrests; much higher levels of education; better employment; and higher wages than those from the control group (reviewed in Rocque61). Several other such studies have also shown the benefits of such enrichment. These studies show that investment of resources can improve neurocognitive functioning in children, which will not only improve mental and physical health but also reduce risk of later offending.66

Other protective factors We will explore many other protective factors as we delve deeper into specific biological factors associated with behavior, such as hormones, brain chemistry, brain trauma, birth defects, and inadequate diet, but many different and disparate environmental factors have been associated with reducing biological risks. For example, increased exposure to green space was found to be a protective factor against developing schizophrenia. The causes of schizophrenia have been shown to be multi-faceted, with both environmental and genetic factors increasing risk. In an analysis of proximity to green space over the life course, using satellite data from the landmass of Denmark, proximity to the least amount of green space was related to a 1.52-fold increase in risk of developing schizophrenia in comparison with people living closest to green space, even after adjusting for other known predisposers. This highlights the benefits of green space, which may relate to reduced pollution, decreased stress, improved immune functioning, and improved mental health—all known to relate to exposure to a more natural environment.67 This suggests that careful city planning and town planning could improve mental and physical health and well-being—as well as aesthetics—by increasing, or better planning, the amount and proximity of green space. Fetal alcohol spectrum disorder (FASD), which is caused by prenatal exposure to alcohol, results in a range of mental and physical deficits and is a known risk factor for antisocial and later criminal behavior, which we will discuss further in Chapter 8. In a comparison of 58 incarcerated FASD young offenders with 456 incarcerated non-FASD young offenders, FASD was linked to several risk factors for criminal activity, such as early-onset alcohol use, foster care, low self-control, low selfesteem, and other behavioral disorders.68 FASD was also linked to earlier onset of criminal behavior and more frequent recidivism compared with non-FASD youth; however, this relationship did not hold true when other risk factors were included in the analysis, in particular, foster care, early use of alcohol, and low self-control. This indicates that FASD youth were only at higher risk for criminal behavior when first exposed to other risk factors, suggesting that eliminating or reducing other, earlier risk factors could protect FASD youth from later criminal behavior.68 This has important implications for the development of early intervention strategies, which could target such risk factors to ameliorate or even eliminate later risk for criminal behavior.68 Two major neuropsychological protective factors have been identified: intelligence and executive functioning.69 Intelligence, or IQ, has been consistently linked to behavior, with low IQ correlated with antisocial behavior, in particular low verbal IQ and performance IQ. Low verbal IQ is commonly found in antisocial youth, and it is believed that defects in verbal development impact self-control and ability to socialize with peers and result in antisocial behavior.69 Low IQ is also believed to interact with other risk factors such as child maltreatment, low SES, and lack of parental control to result in antisocial behavior (reviewed in Portnoy69). The converse appears to be true, in

Conclusion  105

that high IQ has been shown to be a protective factor against risk for antisocial behavior. Studies have demonstrated a direct protective impact of high verbal IQ against a variety of risk factors such as criminal parent, deprived neighborhood, and poor home environment (reviewed in Portnoy69). It has been suggested that individuals with higher IQ may be better able to cope with adverse circumstances such as stress and adverse environments.69 Executive functioning includes a range of neurological processes involved in mental control, self-conduct, and appropriate behavior designed to achieve set goals, so dysfunction can lead to antisocial behavior.69 Executive function is primarily controlled by the frontal lobe of the brain, and damage to this area can result in major deficits in executive functioning. Although results are mixed, some forms of higher executive functioning have proved be a strong buffer against certain forms of adversity such as marital conflict and low SES (reviewed in Portnoy69). It  is important to recognize that factors such as intelligence and executive functioning are not static and can be improved. The very idea of the IQ test was not that a child was tested and given a number for life, but that IQ could be used as a measure in order to determine the level of intelligence improvement. In other words, a child could be measured on a regular basis to determine the level of improvement resulting from education. Therefore, IQ can be improved with education, intervention, and improved diet, and this has been demonstrated many times. Executive functioning can also be improved with appropriate interventions,69 which is particularly important when reduced executive functioning is a result of brain trauma. As we will see in the following chapters, there are many other protective factors that reduce or eliminate biological risks, which can include hormones, neurotransmitters, and diet, much of which can be adjusted. Understanding protective as well as risk factors is very important in developing and individualizing intervention strategies, as individuals who possess certain protective factors may benefit from a different form of intervention, one that will augment their protective factors, in comparison with an individual with no or different protective factors.69 In many of these studies, we see that a stable family home is a protective factor against crime. If a child has a predisposition for crime, based on biological family, the child may be just fine in a stable home. If some of this information were better known and understood by a larger group of criminologists and sociologists, it could go a long way toward helping some of these children. Saying that a child must have a supportive, stable home is naive and not so simple in practice. Stable homes do not just appear, nor should we ever consider taking children from the natural home and putting them into an adoptive or foster home unless the natural home is dangerous. But it is possible to add structure to the life of a child, even if the home situation is not stable. A stable long-term relationship with a teacher or teacher’s aide can help; special assistants in schools and behavior assistants for CD children are now proving their worth. Teaching children to respect diversity and improving the caring atmospheres in schools can help. School attachment and bonding has been shown to be a major protective factor against genetic risk. Organizations such as Big Brothers Big Sisters can sometimes provide ongoing stable relationships exterior to the family, although these will not work with deeply antisocial children. But to work, these relationships and new structures must be stable, they must be ongoing, and the child must be reachable. It  is equally naive to think that children who have become deeply antisocial will automatically engage a new guidance structure, but long-term structural change and consistent guidance can help. If change is to come, it will come in this direction.

Conclusion Overall, there is a large body of research indicating that almost all forms of antisocial behavior are under varying levels of genetic influence, including aggression and violence. However, it must be remembered that these are, again, predispositions. Although, in most cases, these studies can indicate whether a trait does or does not have a genetic basis, they do not indicate what that genetic relationship actually is. Every trait is under the control of numerous genes, and we know that our genotype interacts with our experiences and environment.

106  Evidence for genetic predispositions for criminogenic behavior

Hopefully, we have come a long way from earlier beliefs in determinism and the entirely incorrect concept of DNA as destiny. Even characteristics that are extremely heritable, such as hair color, are easily dramatically changed.40 Moreover, those factors that are under strong genetic control do not need genetic or biological means to change. For example, a genetic predisposition for alcoholism can be completely reversed environmentally by removing alcohol from the environment.40 Even a factor such as phenylketonuria (discussed in Chapter 1), which is 100% under genetic control, can be completely controlled with appropriate interventions.40 Heritability studies have also been very important in developing our understanding of the role of the environment, not just as a risk factor but also as a protective factor. Understanding the roles of the genotype and the environment in developing antisocial behavior helps us develop intervention strategies that promote protective factors.

Questions for further study and discussion 1. Explain how twin studies separate genetic influences from environmental influences and then discuss how this would be affected if the MZ twins in the study deliberately attempted to be different from each other. 2. Explain the benefits of adoption studies over twin studies and then consider the issues related to differences between adoptive and biological parents in general and how this would affect the studies. How could this be taken into account in studies? 3. How does understanding protective factors help us develop individual intervention strategies for at-risk children? Give an example of a new intervention strategy that would help at-risk children with ADHD. 4. Explain how a more or less shared environment in twins, both MZ and DZ, would impact heritability scores. 5. You are designing an intervention program for at-risk children. Discuss and explain how you will develop such a program either for schools or in the home, based on the information from this chapter.

References 1. Polderman, T.J., Benyamin, B., de Leeuw, C.A. et al. 2015. Meta-analysis of the heritability of human traits based on fifty years of twin studies. Nature Genet. 47(7): 702–709. 2. Evangelou, E., Warren, H.R., Mosen-Ansorena, D. et al. 2018. Genetic analysis of over 1 ­million people identifies 535 new loci associated with blood pressure traits. Nature Genet. 50(10): 1412. 3. Raine, A. 1993. Genetics and crime. In: The Psychopathology of Crime: Criminal Behavior as a Clinical Disorder. San Diego: Academic Press, Elsevier Science. pp. 47–80. 4. Christianson, K.O. 1974. Seriousness of criminality and concordance among Danish twins, In: Crime, Criminology and Public Policy, Hood, R., editor. London: Heinemann. pp. 45–88. 5. Christianson, K.O. 1977. A review of studies of criminality among twins, In: Biosocial Bases of Criminal Behavior Mednick, S.A. and Christianson, K.O., editors. New York: Gardner Press. pp. 45–88. 6. Lyons, M.J. 1996. A twin study of self-reported criminal behaviour. Ciba Found. Symp. 194: 61–70. 7. Cloninger, C.R. and Gottesman, I.I. 1987. Genetic and environmental factors in antisocial behaviour disorders. In: The  Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. Cambridge, UK: Cambridge University Press. 8. Kendler, K.S., Maes, H.H., Lonn, S.L. et  al. 2015. A  Swedish national twin study of criminal behavior and its violent, white-collar and property subtypes. Psychol. Med. 45(11): 2253–2262.

References  107

9. Ferguson, C.J. 2010. Genetic contributions to antisocial personality and behavior: A metaanalytic review from an evolutionary perspective. J. Soc. Psychol. 150(2): 160–180. 10. Rhee, S.H. and Waldman, I.D. 2002. Genetic and environmental influences on antisocial behavior: A meta-analysis of twin and adoption studies. Psychol. Bull. 128(3): 490–529. 11. Kendler, K.S., Aggen, S.H., and Patrick, C.J. 2013. Familial influences on conduct disorder reflect 2 genetic factors and 1 shared environmental factor. JAMA Psychiatry 70(1): 78–86. 12. Kendler, K.S., Lonn, S.L., Maes, H.H. et al. 2015. A national Swedish longitudinal twin-sibling study of criminal convictions from adolescence through early adulthood. Twin Res. Hum. Genet. 18(3): 227–233. 13. Moffit, T.E. 1993. Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychol. Rev. 100(4): 674–701. 14. Kirkpatrick, R.M., Legrand, L.N., Iacono, W.G., and McGue, M. 2011. A twin and adoption study of reading achievement: Exploration of shared-environmental and gene-­environmentinteraction effects. Learn. Individ. Differ. 21(4): 368–375. 15. Kaplan, J.S. 2012. The effects of shared environment on adult intelligence: A critical review of adoption, twin, and MZA studies. Dev. Psychol. 48(5): 1292–1298. 16. Grove, W.M., Eckert, E.D., L., H., Bouchard, T.J.J., Segal, N.L., and Lykken, D.T. 1990. Heritability of substance abuse and antisocial behavior: A study of monozygotic twins reared apart. Biol. Psychiatry 27(12): 1293–1304. 17. Burt, C.H. and Simons, R.L. 2014. Pulling back the curtain on heritability studies: Biosocial criminology in the postgenomic era. Criminology 52(2): 223–262. 18. Burt, C.H. and Simons, R.L. 2015. Heritability studies in the postgenomic era: The fatal flaw is conceptual. Criminology 53(1): 103–112. 19. Barnes, J.C., Wright, J.P., Boutwell, B.B. et  al. 2014. Demonstrating the validity of twin research in criminology. Criminology 52(4): 588–626. 20. Wright, J.P., Barnes, J.C., Boutwell, B.B. et al. 2015. Mathematical proof is not minutiae and irreducible complexity is not a theory: A final response to Burt and Simons and a call to criminologists. Criminology 53(1): 113–120. 21. Moffitt, T.E. and Beckley, A. 2015. Abandon twin research? Embrace epigenetic research? Premature advice for criminologists. Criminology 53(1): 121–126. 22. Kendler, K.S., Lonn, S.L., Maes, H.H., Sundquist, J., and Sundquist, K. 2015. The etiologic role of genetic and environmental factors in criminal behavior as determined from full- and half-sibling pairs: An evaluation of the validity of the twin method. Psychol. Med. 45(9): 1873–1880. 23. Mednick, S.A., Gabrielli, W.F.J., and Hutchings, B. 1984. Genetic influences in criminal convictions: Evidence from an adoption cohort. Science 224(4651): 891–894. 24. Mednick, S.A., Gabrielli, W.F.J., and Hutchings, B. 1987. Genetic factors in the etiology of criminal behavior, In: The Causes of Crime: New Biological Approaches, Mednick, S.A., Moffitt, T.E., and Stack, S.A., editors. Cambridge, UK: Cambridge University Press. pp. 74–91. 25. Brennan, P.A., Mednick, S.A., and Jacobsen, B. 1996. Assessing the role of genetics in crime using adoption cohorts. Ciba Found. Symp. 194: 115–123. 26. Bohman, M. 1996. Predisposition to criminality: Swedish adoption studies in retrospect. Ciba Found. Symp. 194: 99–109. 27. Sullivan, P.F., Kendler, K.S., and Neale, M.C. 2003. Schizophrenia as a complex trait: Evidence from a meta-analysis of twin studies. Arch. Gen. Psychiatry 60(12): 1187–1192. 28. Kendler, K.S., Larsson Lonn, S., Morris, N.A., Sundquist, J., Langstrom, N., and Sundquist, K. 2014. A Swedish national adoption study of criminality. Psychol. Med. 44(9): 1913–1925. 29. Hjalmarsson, R. and Lindquist, M.J. 2013. The  origins of intergenerational associations in crime: Lessons from Swedish adoption data. Labour Econ. 20: 68–81. 30. Bezdjian, S., Baker, L.A., and Tuvblad, C. 2011. Genetic and environmental influences on impulsivity: A meta-analysis of twin, family and adoption studies. Clin. Psychol. Rev. 31(7): 1209–1223.

108  Evidence for genetic predispositions for criminogenic behavior

31. Beaver, K.M., Rowland, M.W., Schwartz, J.A., and Nedelec, J.L. 2011. The genetic origins of psychopathic personality traits in adult males and females: Results from an adoption-based study. J. Crim. Justice 39(5): 426–432. 32. DeLisi, M. 2009. Psychopathy is the unified theory of crime. Youth Violence Juv. Just. 7(3): 256–273. 33. Hare, R.D. 1996. Psychopathy: A clinical construct whose time has come. Crim. Just. Behav. 23(1): 25–54. 34. Waldman, I.D., S.H., R., LoParo, D., and Park, Y.M. 2018. Genetic and environmental influences on psychopathy and antisocial behavior, In: Handbook of Psychopathy, C.J., Patrick, editor. New York: Guilford Press. pp. 335–353. 35. Kendler, K.S., Ji, J., Edwards, A.C., Ohlsson, H., Sundquist, J., and Sundquist, K. 2015. An extended Swedish national adoption study of alcohol use disorder. JAMA Psychiatry 72(3): 211–218. 36. Cloninger, C.R., Sigvardsson, S., and Bohman, M. 1996. Type  I and Type II alcoholism: An update. Alcohol Health Res. World 20(1): 18–23. 37. Verhulst, B., Neale, M.C., and Kendler, K.S. 2015. The  heritability of alcohol use disorders: A meta-analysis of twin and adoption studies. Psychol. Med. 45(5): 1061–1072. 38. Kendler, K.S., Sundquist, K., Ohlsson, H. et al. 2012. Genetic and familial environmental influences on the risk for drug abuse: A national Swedish adoption study. Arch. Gen. Psychiatry 69(7): 690–697. 39. Charney, E. 2012. Behavior genetics and postgenomics. Behav. Brain Sci. 35(5): 331–358. 40. Moffitt, T. 2005. Genetic and environmental influences on antisocial behaviors: Evidence from behavioral–genetic research. Adv. Genet. 55: 41–104. 41. Kendler, K.S., Morris, N.A., Ohlsson, H., Lonn, S.L., Sundquist, J., and Sundquist, K. 2016. Criminal offending and the family environment: Swedish national high-risk home-reared and adopted-away co-sibling control study. Br. J. Psychiatry 209(4): 294–299. 42. Kendler, K.S., Ohlsson, H., Sundquist, K., and Sundquist, J. 2016. The rearing environment and risk for drug abuse: A Swedish national high-risk adopted and not adopted co-sibling control study. Psychol. Med. 46(7): 1359–1366. 43. Boardman, J.D., Menard, S., Roettger, M.E., Knight, K.E., Boutwell, B.B., and Smolen, A. 2014. Genes in the dopaminergic system and delinquent behaviors across the life course: The role of social controls and risks. Crim. Justice Behav. 41(6): 713–731. 44. Ferguson, C.J. and Beaver, K.M. 2009. Natural born killers: The genetic origins of extreme violence. Aggress. Viol. Behav. 14(5): 286–294. 45. Beaver, K.M., Mancini, C., DeLisi, M., and Vaughn, M.G. 2011. Resiliency to victimization: The role of genetic factors. J. Interpers. Violence 26(5): 874–898. 46. Caspi, A., McClay, J., Moffitt, T.E. et al. 2002. Role of genotype in the cycle of violence in maltreated children. Science 297(5582): 851–854. 47. Caspi, A., Sugden, K., Moffitt, T.E. et al. 2003. Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science 301(5631): 386–389. 48. Consoloni, J.L., Ibrahim, E.C., Lefebvre, M.N. et al. 2018. Serotonin transporter gene expression predicts the worsening of suicidal ideation and suicide attempts along a long-term follow-up of a Major Depressive Episode. Eur. Neuropsychopharmacol. 28(3): 401–414. 49. Sugden, K., Arseneault, L., Harrington, H., Moffitt, T.E., Williams, B., and Caspi, A. 2010. Serotonin transporter gene moderates the development of emotional problems among children following bullying victimization. J. Am. Acad. Child Adolesc. Psychiatry 49(8): 830–840. 50. Bowes, L., Maughan, B., Caspi, A., Moffitt, T.E., and Arseneault, L. 2010. Families promote emotional and behavioural resilience to bullying: Evidence of an environmental effect. J. Child Psychol. Psychiatry 51(7): 809–817. 51. Zannas, A.S., Provencal, N., and Binder, E.B. 2015. Epigenetics of posttraumatic stress disorder: Current evidence, challenges, and future directions. Biol. Psychiatry 78(5): 327–335.

References  109

52. Beaver, K.M., Boutwell, B.B., Barnes, J.C., and Cooper, J.A. 2009. The biosocial underpinnings to adolescent victimization: Results from a longitudinal sample of twins. Youth Viol. Juv. Just. 7(3): 223–238. 53. Kim-Cohen, J., Moffitt, T.E., Caspi, A., and Taylor, A. 2004. Genetic and environmental processes in young children’s resilience and vulnerability to socioeconomic deprivation. Child Deve. 75(3): 651–668. 54. Wertz, J., Agnew-Blais, J., Caspi, A. et al. 2018. From childhood conduct problems to poor functioning at age 18 years: Examining explanations in a longitudinal cohort study. J. Am. Acad. Child Adolesc. Psychiatry 57(1): 54–60 e4. 55. Bennett, P. 2018. The heterogeneous effects of education on crime: Evidence from Danish administrative twin data. Labour Econ. 52(C): 160–177. 56. Bell, B., Costa, R., and Machin, S. 2016. Crime, compulsory schooling laws and education. Econ. Education Rev. 54: 214–226. 57. Watts, S.J. 2018. ADHD symptomatology and criminal behavior during adolescence: Exploring the mediating role of school factors. Int. J. Offender Ther. Comp. Criminol. 62(1): 3–23. 58. Thomas, R., Sanders, S., Doust, J., Beller, E., and Glasziou, P. 2015. Prevalence of attention deficit/hyperactivity disorder: A systematic review and meta-analysis. Pediatrics 135: e994–e1001. 59. Savolainen, J., Mason, W.A., Bolen, J.D. et  al. 2015. The  path from childhood behavioural disorders to felony offending: Investigating the role of adolescent drinking, peer marginalisation and school failure. Crim. Behav. Ment. Health 25(5): 375–388. 60. Snyder, S.M. and Smith, R.E. 2015. The influence of school engagement on counts of delinquent behaviors among maltreated youths. Children Sch. 37(4): 199–206. 61. Rocque, M., Welsh, B.C., and Raine, A. 2012. Biosocial criminology and modern crime prevention. J. Crim. Justice 40(4): 306–312. 62. Olds, D., Henderson, C.R.J., Cole, R. et al. 1998. Long-term effects of nurse home visitation on children’s criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial. JAMA 280(14): 1238–1244. 63. Nurse-Family Partnership. 2018. Helping first time parents succeed. Accessed October 11, 2018; https://www.nursefamilypartnership.org/about/. 64. Raine, A., Venables, P.H., Dalais, C., Mellingen, K., Reynolds, C.A., and Mednick, S.A. 2001. Early educational and health enrichment at age 3–5 years is associated with increased autonomic and central nervous system arousal and orienting at age 11 years: Evidence from the Mauritius Child Health Project. Psychophysiol. 38(2): 254–266. 65. Raine, A., Mellingen, K., Liu, J., Venables, P., and Mednick, S.A. 2003. Effects of environmental enrichment at age 3–5 years on schizotypal personality and antisocial behavior at ages 17–23 years. Am. J. Psychiatry 160(9): 1–9. 66. Glenn, A.L. and Raine, A. 2014. Neurocriminology: Implications for the punishment, prediction and prevention of criminal behaviour. Nat. Rev. Neurosci. 15(1): 54–63. 67. Engemann, K., Pedersen, C.B., Arge, L., Tsirogiannis, C., Mortensen, P.B., and Svenning, J.C. 2018. Childhood exposure to green space—a novel risk-decreasing mechanism for schizophrenia? Schizophr. Res. 199: 142–148. 68. Corrado, R.R. and McCuish, E.C. 2015. The development of early onset, chronic, and versatile offending: The role of fetal alcohol spectrum disorder and mediating factors. Int. J. Child Adolesc. Health 8(2): 241–250. 69. Portnoy, J., Chen, F.R., and Raine, A. 2013. Biological protective factors for antisocial and criminal behavior. J. Crim. Justice 41(5): 292–299.

6 Candidate genes, gene × environment interactions, and epigenetics

Introduction This chapter will look more deeply into genetics and the environment. So far, we have considered twin, adoption, and family studies, which give a general idea of whether a trait has a heritable component but do not identify particular genotypes that might have higher risks for antisocial behavior. More recent studies have been conducted in an attempt to identify specific candidate genes and polymorphisms that may link to certain behaviors. Such studies use very large genomic data sets. It should be clear by now that even when an allele is found to increase a predisposition for a behavior, it does not act alone. We have always known that our genes and environment interact, but more recent studies have investigated specific gene and environment interactions, and several hypotheses have been used to model different forms of this interaction. One of the most fascinating genetic discoveries of our time is that of epigenetics, in which gene expression is changed without changing the DNA sequence. This is perfectly normal, and most epigenetic effects are under genetic control, but some epigenetic changes have been shown to be a result of the environment, and although epigenetics is in its infancy, a great deal of work is being conducted in this area. Perhaps most exciting, studies have shown that deleterious epigenetic changes can be reversed, offering promise of future intervention and treatment.

Candidate genes More and more studies are now being conducted in order to identify candidate genes, those that are believed to be related to a particular behavior. Of course, as we have seen many times, behavior is multi-faceted and highly complex, influenced to a greater or lesser extent by many genetic and environmental influences, so such studies do not  aim to identify a gene that “causes” a specific behavior but rather genes that may contribute to a behavior. As so many genes are involved in any behavior, each gene will contribute only a very small amount of the variance, and this is further complicated by the gene × gene interactions, as well as by gene × environment interactions and environment × environment interactions.1 However, a number of studies have identified certain genes that play a role in some behaviors. For example, as we will see in later chapters, a number of genes have been identified that impact neurotransmitter functions such as serotonin and dopamine, which are linked to numerous behavioral issues, and others have been linked to alcohol dependence or overall externalizing behaviors.1

111

112  Candidate genes, gene × environment interactions, and epigenetics

Candidate gene studies involve either gene-association studies or genome-wide association ­studies.2 Candidate gene association studies involve proposing a hypothetical association between a specific allele and a specific phenotype, for example, a type of behavior. Large data sets containing both genetic and behavioral data are then examined. Genome-wide association studies take an opposite approach and are not guided by a specific hypothesis. They involve looking at the entire genome of a very large number of people who exhibit the trait or behavior of interest, compared with similar numbers of people who do not possess the trait, in order to find statistical differences between certain alleles that may relate to the trait.2 Many candidate genes have been found associated with neurotransmitter systems, such as the dopamine system, the serotonin system, and the epinephrine/norepinephrine systems.3 These neurotransmitters are involved in brain function and have been linked to aggression and antisocial behavior. They will be discussed in detail in Chapter 9, but briefly, dopamine is involved in the body’s reward system and produces pleasurable feelings and arousal in reaction to everyday events such as love, sex, eating, achieving desirable goals, and mood. People with disruption of the dopamine system may have a deficiency in this reward system, meaning that they do not receive the cascade of pleasure that normal people feel but instead feel anxious and angry and require a much greater amount of a drug or alcohol to feel pleasure.4 Serotonin is strongly involved in impulsivity and impulse control, and epinephrine and norepinephrine are involved in the flight-or-fight response, fear, and emotion. All three of these systems are impacted by another neurotransmitter, that is, monoamine oxidase A. Several polymorphic genes within all four systems have been identified, certain alleles of which have been linked to a variety of antisocial behaviors, as we will discuss later. Candidate genes have been examined for myriad disorders and behaviors, and the following are just a few examples.

Attention deficit hyperactivity disorder Attention deficit hyperactivity disorder (ADHD) is one of the most commonly diagnosed childhood behavioral disorders worldwide,5 with a global prevalence of 5%.6 Children with ADHD display persistently disruptive behavior and high levels of inattention; have drastic, rapid mood swings; and exhibit uncontrollable hyperactivity and impulsivity, with a low frustration tolerance.7 Symptoms usually begin in early childhood and continue through adolescence, persisting into adulthood in some individuals.6 ADHD may result in poor academic outcome, school failure, behavioral problems in school, poor peer relationships, poor school attachment, poor grades, and increased school suspensions.8 Many studies have linked ADHD with increased alcohol and drug use.9,10 This  frequently leads to juvenile delinquency and eventually adult crime,8,11–14 with individuals diagnosed with ADHD being greatly overrepresented in adult and juvenile correctional facilities worldwide.15 In a review of studies from a number of countries, prevalence of childhood ADHD in adult prisoners ranged from 24% to 67%, and adult ADHD was found in 23% to 45%, dramatically higher than that in the general population.16 The etiology of ADHD is complex, with environmental and social elements, although many twin, adoption, and familial studies have indicated a genetic basis for the disorder, with heritability estimated at 76% over studies. 5,6 A vast number of studies, including genetic linkage studies, genome-wide association studies, and candidate-gene association studies, have been conducted to identify candidate genes for ADHD, with the understanding that the disorder is probably the result of small contributions from a large number of genes. Over 300 candidate genes have been proposed, although most have been identified in single studies that have not been replicated. In an evaluation of such studies, Li and colleagues looked at hot genes, multi-evidence supported genes, and prioritized genes.6 Hot genes were identified as those reported by at least five association studies and are considered to be the top 7% of candidate genes. Twenty-four hot genes were identified, and most were related to monoamine neurotransmitter synthesis and function; this makes sense, as a wide range of neuroimaging and pharmacological studies, as well as animal studies, suggests that dysfunction or disruption of the dopamine, serotonin,

Candidate genes  113

or epinephrine/norepinephrine systems results in or aggravates ADHD, and all of these are impacted by monoamine neurotransmitters.6 Multi-evidence genes were those identified by a number of different study types, suggesting improved robustness. Thirty-six genes were identified in this way, although no gene was supported by all study types; however, three genes were identified by three study types. Prioritized genes were determined using five different multiple-sourced gene-prioritization tools that examine genomic information. This  method identified 32 genes, most involved in nervous system development and neurotransmitter function. Overall, the review identified 70 genes that may each contribute a small increased risk for ADHD; these genes require further study.6

Conduct disorder Conduct disorder (CD) in children is characterized by many antisocial behaviors, such as stealing, lying, bullying, cruelty to animals, aggression, violence, arson, and property destruction. The disorder is considered a strong risk factor for later criminality. It has long been known that CD has a genetic component. A retrospective study of 2682 adult twin pairs from the community-based Australian Twin Registry showed that there was a substantial genetic influence on predisposition for developing CD.17 A concordance rate of 71% was found for CD, and little evidence was found for the effects of environmental influences. Although boys are much more frequently diagnosed with CD than girls, the researchers found that the level of genetic and environmental influences for developing CD did not vary between the sexes.17 This may be because boys more commonly display externalizing behaviors, which are directed outward, such as physical aggression, disobeying rules, cheating, stealing, and property destruction, whereas girls more commonly display internalizing behaviors, which are directed inward, such as fearfulness, social withdrawal, and somatic disorders; internalizing behaviors are often less noticeable, so they may make girls less likely to be diagnosed. A later study of 70 monozygotic (MZ) and 42 dizygotic (DZ) twins between the ages of 4 and 15 years found very similar results for CD, with a heritability of 68%.18 Borderline personality disorder, which is also associated with antisocial behavior, was found to have a heritability estimate of 63%, and overall, the mean heritability estimate for personality disorders in general was 58%.18 Many candidate genes have been examined for CD, and much focus has been, again, on the dopaminergic system. For example, a study of 872 males from the US National Longitudinal Study of Adolescent Health (Add Health) data set on CD and adult antisocial behavior in relation to genotype focused on the polymorphic alleles of two dopamine receptor genes, DRD2 and DRD4. Although neither gene had significant effects on CD alone, the two genes interacted with each other to predict variation in both CD in adolescence and antisocial behavior in adulthood.19 The authors pointed out, however, that CD is often comorbid with substance abuse and sensation seeking, both of which are linked to these two genes, so the link may relate more to the genetic interaction with substance abuse than to purely a gene × gene interaction.19

Schizophrenia Schizophrenia is an extremely debilitating psychiatric disease that can include delusions, hallucinations, cognitive disabilities, apathy, social withdrawal, loss of social function, and an altered perception of reality.20,21 Certain symptoms of schizophrenia are strongly linked to violence and crime, although not all individuals with schizophrenia exhibit such behaviors.22 In Sweden, a large-scale longitudinal study showed that patients with schizophrenia were more than twice as likely to commit at least one violent crime in comparison with the general population, and the risk increased in patients who were also substance abusers.23 This was also seen in an Israeli population, in which schizophrenic patients were significantly more likely to commit a violent crime than matched controls in the general public and were also more likely to be violent if they were substance abusers.24

114  Candidate genes, gene × environment interactions, and epigenetics

The etiology of schizophrenia is complex and appears to be linked to a wide range of environmental factors, such as prenatal flu, birth complications, season of birth, childhood trauma, social isolation, and prenatal and maternal starvation,21,25,26 but myriad studies have shown that it also has a very strong genetic underpinning, with heritability estimates of 85%.21 Many studies have linked genes involved in the dopamine system with increased risk for schizophrenia, in particular the D2 dopamine receptor and other genes involved in dopamine synthesis, many of which are related to the stress response.25 Some de novo mutations, which just arise and are not passed on to the next generation, have been shown to increase the risk of schizophrenia, and factors that increase the risk of occurrence of de novo mutations, such as paternal age, also increase the risk of schizophrenia.20 Most recently, a polygenic basis has been proposed.20 In  a review of large-scale genome-wide association studies, over 100 loci have been identified for alleles that each contribute to risk, with the level contributed ranging from very low to high, although moderate- and high-contributing alleles were much rarer than low-contributing ones.27 Specific genes have also been identified as risk factors for some of the cognitive deficiencies associated with schizophrenia and many other psychiatric disorders,28 and schizophrenia has been shown to have genetic links to other disorders, such as ADHD and bipolar disorder.20

Interactions between genes and the environment As we have seen and will continue to see throughout this book, genes do not act alone and neither does the environment. Genes and the environment frequently act together. They  may correlate with each other or interact. These two relationships are quite different and must be distinguished in order to understand their mechanisms.

Gene–environment correlations Gene–environment correlations can be mistaken for gene  ×  environment interactions and must be separated. It  has been hypothesized that there are three forms of gene–environment correlations.1,29,30 1.  Passive gene–environment correlations

The first is referred to as passive and occurs when a child’s genetic background and environment come from one source—that of the parents. For example, adolescents may exhibit aggressive behavior as a result of both inheriting genes that predispose them to aggression and growing up in an environment in which their parents exhibit aggression, making the two factors difficult to separate.1 2.  Reactive gene–environment correlations

The second is reactive, in which genetically influenced behavior induces a certain type of environment. For example, a genetic predisposition to shyness may make an adolescent less responsive to peer overtures, resulting in isolation that leads to further inhibition, again making it difficult to disentangle the environmental and genetic components of the behavior.1 3.  Active gene–environment correlations

Active correlation is hypothesized to occur when adolescents pick their environments based on their genetic predispositions.1 For example, an adolescent who is predisposed to antisocial behavior may seek out peers who are like-minded with whom to engage in criminal acts. In all cases, it is hard to tease apart the genetic and environmental influences.

Interactions between genes and the environment  115

Gene × environment interactions We now understand that people with different genotypes respond differently to the same environment. In other words, 2 or 20 people may be exposed to the same environment, and all may react differently. For example, we know there are multitudes of risk factors for criminal behavior (for example, severe child abuse), but many people—in fact, the majority—who are exposed to this risk factor will not commit a crime. Sociological theories do not explain why the same environment can produce such disparate results, but biosocial criminology can show that one explanation for the disparity is the closely interwoven relationships between genetic and environmental factors.31 Frequently, there is an interaction between a person’s genetic makeup and the environment experienced. This  occurs when an individual’s particular genotype is sensitive or vulnerable to environmental factors; individuals with different genotypes will react differently to different environmental triggers.1 In  other words, the effect the environment will have on an individual’s behavior depends on genetic makeup, and the effect an individual’s genetic makeup will have on behavior depends on the environment experienced. 32 This is referred to as gene × environment interactions, or G × E interactions. It is this interaction that explains why some people who experience terribly adverse environments never commit crimes and why some others who experience a nurturing environment do commit crimes. If an individual does not  have a genetic predisposition toward criminal behavior, that person may never commit a crime, even when faced with environmental adversity. If an individual does have a genetic predisposition for criminal behavior but does not experience a triggering environment, then that person too may never exhibit criminal behavior. It is only when an individual has both a genetic predisposition and experiences an adverse or triggering environment that the risk for criminal behavior is high. 32 However, although the existence of both increases risk, it still does not guarantee a criminal outcome. Moreover, both genetic and environmental risk factors act in gradients, with both increasing and ameliorating risk, and there are many individual and interacting effects within both genetic and environmental influences, all of which serve to moderate or amplify behavior. Although the focus of most biosocial studies is to determine the etiology of behavior and identifying risk factors, such studies also identify protective factors. In  particular, understanding G  ×  E interactions allows us to recognize protective environments that will prevent a risky genotype from being expressed and so can inform intervention and prevention strategies. 32 An example of G × E interactions

There are a multitude of examples of G × E interactions that are now being studied. One example is seen in the development of schizophrenia and bipolar disorder. Numerous studies have shown both environmental and genetic causes for these psychiatric disorders, and a large number of genes have been implicated, with each contributing only a small level of susceptibility.26 More recently a G  ×  E interaction has been proposed, and a large number of studies have looked at genetic interactions with a range of environmental factors. In a review of such studies, the majority considered variation in genes that code for catechol-O-methyltransferase (COMT), brainderived neurotrophic factor (BDNF), and FK506-binding protein (FKBP5).26 COMT is an enzyme that breaks down a number of substances in the body, such as some neurotransmitters and certain drugs, and is implicated in several psychiatric disorders.33 BDNF is involved in growth and survival of brain neurons.34 FKBP5 is a binding protein that regulates steroid receptors, and methylation of the gene is thought to moderate the effects of genetic and environmental risk factors for psychiatric diseases.35 Most of these studies found significant interactions between certain alleles of these genes and cannabis use, as well as early-life stress or childhood trauma, which increased risk for schizophrenia and bipolar disorder, although there were far fewer studies on bipolar disorder. A few studies also found some interactions between the alleles and infectious diseases, birth complications, and season of birth.26

116  Candidate genes, gene × environment interactions, and epigenetics

G × E interaction models

In attempting to understand the underlying mechanisms of the complex G × E interactions that underlie the expression of criminal or antisocial behavior, a number of hypotheses, or models, have been developed, based on the proposed different methods in which genetic polymorphic risk factors increase or decrease susceptibility to unfavorable environments.32 1.  Diathesis stress model

The diathesis stress model assumes that a genotype (involving many polymorphisms or different risk alleles) confers risk and will lead to an extremely adverse outcome if the individual is exposed to a negative environment. However, in a good environment the outcome will not be as negative or may not occur at all.36 This model stresses the adverse environment and its effects on a risky genotype that is vulnerable to environmental triggers, resulting in antisocial behavior.32 This model, therefore, states that the fundamental causes of antisocial behavior are environmental triggers.37 For  example, several studies have shown that genetic risk factors for antisocial behavior are potentiated in the presence of parental conflict. In a study of over 1300 twin pairs aged 17 years, genetic risk for externalizing behaviors was greatly exacerbated by increased environmental adversity and parental negativity,38 and similarly, in a study of 720 families genetic risk factors for antisocial behavior had much greater influence when parents were negative or less affectionate.39 This is perhaps the most common model used to explain G × E interactions. The social control model is very similar to the diathesis stress model but highlights the presence or absence of social resources rather than an adverse environment as the trigger.37 For example, high cigarette taxation has been found to reduce the genetic influence on developing a smoking habit, which is considered to be a form of social control.37 2.  Bioecological model

The bioecological model states that genetic effects are only expressed in a positive environment and are restricted and not expressed in a negative environment.36 In other words, this model suggests that, in some situations, genetic risk factors may reach their highest dominance in the absence of an adverse environment.40 An example of the bioecological model is parent-child conflict, in ­contrast to parent-adolescent conflict, which is believed to fit the diathesis stress model. In a US study of 500 pairs of twins, shared environment was many times more influential on antisocial behavior in ­children with high levels of parental conflict than in those with low levels of conflict, but c­ onversely, genetic factors were much more significant in the development of antisocial b ­ ehavior in children with low levels of parental conflict.40 3.  Differential susceptibility

Differential susceptibility suggests that certain genes or polymorphic genotypes do not  simply confer risk, but instead represent a type of gene plasticity or a level of malleability to the environment, which may have negative or positive outcomes.36,41(p885) In other words, a child with a risky genotype who is exposed to a negative environment will have a very high risk for a bad outcome, but if that same child with a risky genotype was placed in a more positive environment, the child may have an even more positive outcome than a child without the risky genotype.1 This  model focuses less on the adverse environment and looks more at an individual’s susceptibility to the effects of environment. For example, certain genes involved in neurotransmitter production and function (DRD4 and 5HTTLPR, which we will explore in Chapter 9) are considered to be plasticity alleles.37(p716) Studies have shown that individuals who possess certain alleles of these genes (the 7R allele in DRD4 and the shorter allele in 5HTTLPR) are not only significantly more likely to be aggressive when exposed to extremely adverse environmental conditions but also significantly less likely to be aggressive in extremely favorable environments.42 In a study of carriers of the risk allele for another dopaminergic gene, DRD2, adolescents who were homozygous for this allele were much more likely than those who were heterozygous or homozygous for a non-risk allele to exhibit very seriously antisocial behavior if they were raised in a family that was not close, but they were

Interactions between genes and the environment  117

substantially less likely to exhibit such behavior when raised in a close family, which supports differential susceptibility.37 This model is important because it suggests that we should not just focus on adverse or non-adverse environments but also consider the entire range of environments, as it indicates that normal or usual environments would not provoke any antisocial response even in those with risky genotypes.37 Belsky argued that risk alleles should instead be considered plasticity alleles and that individuals with higher numbers of plasticity alleles are more likely to be impacted, either positively or negatively, by an environment than individuals with lower numbers of plasticity alleles. Therefore G × E interactions, in this model, are the result of the same environment having a different effect on people, depending on the number of plasticity genes they possess.41,43 4.  Social distinction model

A  less referenced model is the social distinction model, which is quite different because it does not suggest that the environment triggers a genotype to act in a certain way. Genetic factors are only identifiable in the most favorable environments.37 For example, in a study on the relationship between the e4 allele of the apolipoprotein, or APOE, gene (which increases the risk of developing Alzheimer’s disease) and cognitive functioning in elderly people, the allele appeared to be risky for individuals living in organized, well-kept environments but less risky for those living in social disorder.44 The authors of this study referred to this as non-causal G × E interaction. 5.  Social push model

The social push model focuses on the differences between normal and abnormal social environments, rather than adverse environments, and suggests that genetic factors are relevant in normal environments, but in extreme environments, the social environment “pushes” the phenotype, so the genotype has less relevance.37 For example, heritability of body mass index is highest in adolescents enrolled in schools with normal body size expectations, but heritability is lower in schools in which body mass index extremes are normal, as the environment is driving the trait and the genotype cannot differentiate between individuals.37

Interventions considering G × E interactions When considering intervention strategies, it is very important to consider G × E interactions; most intervention strategies fail to recognize that individuals with particular genetic risks may respond better to changes in environmental risks.32 This can also confound studies, as individuals with certain genotypes may be more likely to participate in an experimental intervention treatment, thus biasing the results, which would not indicate the efficacy for individuals with a different genotype. Therefore, in order to truly understand the genetic and environmental influences, randomized experiments are required in which participants are not allowed to self-select. For example, a study of 440 African American families in Georgia assessed the efficacy of family-centered interventions on behaviors such as delinquency, substance abuse, and unsafe sexual practices in relation to the families’ genetic makeup.45 The genotype that the authors considered involved the transporter gene for the neurotransmitter serotonin, which we will consider in detail in Chapter 9. A particular allele for this gene, referred to as the shorter allele, has been shown to increase risky behavior, so the authors used a randomized experiment to determine whether youth who had the shorter allele would be more or less likely to exhibit risky behavior if they participated in the intervention program. The results showed that risky behavior was higher in youth with the shorter allele and that youth who had the at-risk genotype and did not participate in the intervention program were twice as likely to exhibit risky behavior in comparison with at-risk youth who did participate or youth not at risk.45 A significant interaction was seen between the genotype and treatment.45 In a follow-up study, the researchers looked at older adolescents in the same intervention program to see whether another neurotransmitter gene, the dopamine D4 receptor gene, regulated the effect of the program on substance abuse.46 The results showed that genetic risk increased substance abuse,

118  Candidate genes, gene × environment interactions, and epigenetics

but when broken down by sex, the relationship held true only in males. There was also a G × E interaction for males only,46 supporting the differential susceptibility model, as high-risk youth were found to be particularly susceptible to risk, both genetically and environmentally, but at the same time were also more susceptible to prosocial protective environments such as those provided by the intervention program.32 Understanding G  ×  E interactions will help us better identify the most important environmental factors that influence antisocial behavior and so will inform future intervention strategies. Genetically informed studies can identify personality and risk factors that are particularly amenable to intervention and result in behavioral changes.32 These studies also show the types of individuals that are most likely to benefit from intervention and at which developmental stage treatment should be aimed.32

Epigenetics Epigenetics is a very old word meaning “extra growth.” It is used today to describe a change in the expression of genes—that is, their function, or what they actually do—without changing the actual DNA sequence. The easiest and commonest explanation for the way genes work is that they code for messenger RNA, which results in protein synthesis, and these proteins then act, but in reality, it is much more complex than that. Gene activation, more commonly called gene expression, is not a single step but a very involved, complicated, and multifaceted procedure in which the environment in which the gene acts is extremely important.47 There is a difference between the alleles of a gene that are present in an individual (their genome) and the genes that are actually expressed. Just because a person has a certain gene does not mean it will be expressed—that is, the gene may be silent in the phenotype. As we saw earlier with Dolly the sheep, all cells are able to perform all functions in the early stages of zygote development, but as the cells become determined and then differentiated, many of the cell’s functions are turned off. They become differentiated, or specialized for a specific function, for example, a kidney cell. The genes in this cell will only express kidney functions, and genes related to other functions will be turned off. Perhaps a more generic example involves eye color. A person who is heterozygous for brown and blue eyes has a blue allele from one parent and a brown from the other. As the brown allele is dominant, it effectively silences the blue allele and only the brown is expressed. Recently, evidence has shown that much of the genome is capable of change. The Human Genome Project showed that up to 48% of the human genome is composed of transposable elements that can change the genetic code or sequence and that these changes, often in response to the environment, occur throughout the life span and can be inherited.2 This is movement of actual DNA sequences— genes—and is very different from epigenetics, which does not change or move DNA but impacts the expression of the genome. Epigenetics is perfectly normal; it is needed for regular body functions, such as cellular development and tissue differentiation, and is vital for normal gene function.48 Because the epigenome changes gene expression, people with identical genomes are not identical phenotypically.48 Therefore, phenotypic differences do occur in MZ twins. For example, gene expression was compared between 3-year-old and 50-year-old MZ twins, and epigenetic differences were found to be four times as common in the older twins.49 It should be noted that epigenetic changes are tissue specific, meaning that they occur in specific tissues, such as blood or brain, and not all tissues. This makes studying them very difficult, as some tissue, such as blood, is very easy to sample, but brain tissue is much harder to sample. An interesting example of epigenetics was reported in a NASA  study of identical twins Scott and Mark Kelly. Both twins are astronauts, but recently, Scott spent a year in space while his identical brother, Mark, remained on Earth. Studies before and after the space travel showed that although their DNA remained identical, Scott’s gene expression had changed. The study showed that space flight involves various stressors, such as inflammation, oxygen deprivation, and dietary changes, which can affect gene expression.50 For  example, the end caps of chromosomes, called telomeres, shorten as we age, and shortening is often also seen in many cancers and other diseases;51

Epigenetics  119

interestingly, Scott’s telomeres actually lengthened, although they returned to pre-flight lengths within 2 days of returning. Once Scott returned to Earth, he rapidly readjusted to Earth’s conditions, but some changes persisted, and 7% of his gene expression was still different from that of his brother 6 months after landing.50 Epigenetics is an area that is just beginning to be understood. The vast majority of epigenetics research relates to health, disease, and forensic science, but there is interesting new work on the potential impact epigenetics could have on criminogenic behavior, and it may help to explain G × E interactions.

Epigenome Although all our cells contain the same DNA, different genes in different cells are turned on and off, as we discussed in Chapter 4. So, although every cell carries the entire genome, only certain genes within it will act within the cell. A human being has trillions and trillions of cells, each with different genes turned on or off. The epigenome is a large array of chemicals and proteins that tells the genome how to function—which cells to turn on and off and when—so the epigenome controls the production of proteins and their subsequent functions.52 When an epigenomic chemical or protein attaches to the DNA and changes the function or expression of a gene, it does not change the actual DNA sequence but does change the manner in which the cell interprets the instructions from the genome; in other words, it changes the gene’s expression. These changes can be passed on when the cell replicates and even from generation to generation and so are heritable.52 When the epigenome modifies DNA, it is referred to as marking the genome. There are two main types of marks or markers, DNA methylation and histone modification,52 although other types of epigenetic mechanisms exist, such as non-coding RNAs, positive-effect variegation, parental imprinting, paramutation, and X-chromosome inactivation.48 1. DNA methylation

DNA methylation has a direct impact on genomic DNA. Only part of a gene, the promotor region, is responsible for determining whether it will be switched on or off. DNA  methylation involves attachment of a chemical tag called a methyl group (CH3 or carbon and three hydrogen molecules) to the nucleotides in the DNA molecule at specific sites. This “methylates” the promotor region, either silencing or reducing the activity of the gene.48 Most studies on epigenetics and crime involve methylation.48 2.  Histone modification or acetylation

Histone modification has an indirect impact on the genomic DNA. Histones are proteins involved in the packaging of DNA, which is an extremely long molecule. They are the basic core around which DNA is wrapped.48 Histones spool and package the DNA and organize it into chromosomes. Histone modification, or acetylation, involves attaching an acetyl group (CH3CO) to certain histone sites, which results in a loosening of the chromatin that packages DNA within a cell, impacting transcription, the first part of gene expression. This tells other proteins whether this part of the DNA molecule should be turned on or off and often increases gene expression.

Impact of the environment on the epigenome Although almost all changes caused by the epigenome involve specializing cells and are under genetic control, the epigenome can be affected by the environment. For example, lifestyle environmental factors such as a bad diet, smoking, and exposure to disease can impact the epigenome, putting stress or pressure on the body and producing chemical reactions that can change the epigenome. In some cases, these changes can cause damage to the body, such as cancer, but mostly, they

120  Candidate genes, gene × environment interactions, and epigenetics

allow the body to respond to a changing environment.52 Epigenetics is a new and very exciting area of disease research that is being explored to gain a greater understanding of many disease processes. From a behavioral perspective, we are more interested in the environmental effects on the epigenetic markers (DNA methylation and histone modification) in neural development, as well as other aspects of the human body that impact behavior, such as neurotransmitter function and hormones. In the past (just a few short years ago), it was accepted that once development was complete, neurons in the brain were not replaced as they died, so no new neurons were created. Research has shown that this is not true, and it is now believed that new neurons are formed throughout the life course.47 It was also accepted that the brain’s structure was completely under genetic control and could not be changed. However, research has shown for some time that the brain is plastic and even its fundamental structure is greatly impacted by a person’s environment, experiences, and activities. In other words, it is believed that certain repeated activities, actions, thoughts, and experiences actually change the wiring of the brain, including growth in areas of the brain that are used most often and reduction in areas of the brain that are not used (reviewed in Burt and Simons47). This neuroplasticity is particularly apparent during childhood and adolescence, and some studies have suggested that environmental situations considered predisposers for criminal behavior are linked to specific neurological changes (reviewed in Burt and Simons47). Early-life adversity

Evidence shows that there are specific periods during brain development in which the brain is particularly sensitive to the environment.53 Work with rodents has shown that experiences that occur during these sensitive periods induce epigenetic changes in a variety of genes, altering gene expression for the rest of the life course, and can result in specific behaviors. A seminal epigenetic study from 2004 illustrated that maternal care in rodents could result in gene expression changes in offspring in just the first week of life.54 Weaver and colleagues showed that increased maternal care in rats, such as licking and grooming, altered a hormone receptor gene in the brains of the rat pups. The  increased care resulted in changes in gene DNA  methylation that were linked to histone acetylation binding. High levels of maternal care resulted in rat offspring with a moderate stress response, resulting in less fear when exploring new situations in comparison with offspring of mothers with low maternal care. These differences related to DNA methylation of a gene in the brain, which was a direct result of maternal care level during the first week of life.54 Interestingly, an infusion of histone deacetylase inhibitor reversed all the changes, showing that epigenetic changes can be reversed.54 In a later study, abusive care for 30 minutes a day during the first week of a rat’s life resulted in significant DNA methylation in the brain-derived neurotrophic factor gene in the prefrontal cortex of the brain, which lasted through the life course.53 As with the study by Weaver et al., however, drug treatment could reverse this trajectory. This interaction was found to be specific only to the early-life period.53 Abusive parental care was found to be transgenerational, with abused rat pups growing up to abuse their litters and non-abused rats rarely exhibiting such behavior. This begs the question of whether the abuse was learned or inherited, so cross-fostering experiments were performed and showed that there was an element of transgenerational inheritance of the epigenetic changes.53 There are now multiple human studies that also show that parenting and child-parent interaction in very early life has an epigenetic effect on the child’s developing brain and can affect resilience, stress response, behavior, and cognition and that even maternal stress during pregnancy can result in changes in the offspring’s response to stress by increasing the levels of steroid hormones (reviewed in DeLisi and Vaughn48). We will learn in subsequent chapters about the many ways an adverse or positive environment can change behavior and also brain functions. Various neural imaging techniques, such as magnetic resonance imaging (MRI) and functional MRI (fMRI), have been used to look at structures in the brain. MRI gives a three-dimensional image of the brain, and fMRI shows the parts of the brain that are functioning at a given time while the person completes a given task. Several studies have shown changes in brain structure and function based on environment (reviewed in Burt and Simons47). For example, in an imaging study of

Epigenetics  121

children rescued from Romanian orphanages, where they had experienced a severely deprived and institutionalized childhood, the brains of the institutionalized children had much less total gray and white matter volume than those of control children, but the abused children showed increased volume of the amygdala (a part of the brain that is important in dealing with emotions), which may be a result of developing protective methods to deal with such severe deprivation.55 Callous unemotional aggression has been linked to reduced amygdala volume and a low fear response, meaning that such a person has little empathy for others and is less responsive or unresponsive to threat. It has been postulated that this could be a response to living in a very adverse environment, as remaining calm during danger may be a successful evolutionary response to threat.56 Neural plasticity has been shown to continue into adulthood; for example, increases in volume of areas of the brain associated with finger dexterity have been found in virtuoso violinists, and even imagining repeatedly playing notes on a piano can increase brain volume in areas related to fingers (reviewed in Burt and Simons47). From a more criminological point of view, cognitive behavior therapy has been shown in numerous studies to improve the behavior of adults with various types of psychopathology, and more importantly, these cognitive and behavioral changes have been shown by neuroimaging studies to relate to neurological changes in the brain (reviewed in Burt and Simons47). However, how much this relates to epigenetics is not yet apparent. In  a review of human epigenetic studies from 2008  to 2015  related to law, criminology, and mental health, 41 studies were identified (from an original 44,000-plus studies on epigenetics in general).48 The studies found consistent epigenetic effects in a range of genes that have been linked to suicide ideation, depression, callous-unemotional traits (as seen in psychopaths), and chronic aggression.48 Many of these polymorphic genes, those we will be discussing throughout the rest of this text, have been associated with antisocial behavior, such as those related to neurotransmitter function, brain development and function, and the hypothalamic-pituitary-adrenal (HPA) axis. The HPA axis involves complex and vital interactions within parts of the brain and neuroendocrine glands related to myriad functions, including stress response, and is linked to the limbic system, which is involved in emotion and fear. Non-genetic epigenetic influences included sexual and physical abuse, exposure to toxins, and maternal depression, all of which have been linked to violence.48 Several of the reviewed studies showed that exposure to acute or chronic stress in early childhood resulted in increased release of glucocorticoids, or stress hormones, which caused epigenetic changes to receptor genes. Researchers have suggested that this can lead to severe antisocial behavior (reviewed in DeLisi and Vaughn48). Children with reduced stress responses do not react to danger and do not exhibit suitable emotional responses to risk, which means they do not have the normal stress responses that would lead to fear or anger and may be attracted to dangerous situations. Such underarousal is common in antisocial individuals and is a core tenet of psychopathy.48 Other studies showed that epigenetic changes to the HPA  axis and the serotonergic system (which controls the neurotransmitter serotonin) resulting from adversity in childhood can increase internalizing behaviors such as fearfulness, social withdrawal, and anxiety (reviewed in DeLisi and Vaughn48). As we will see in Chapter 9, dysfunction of the serotonergic system can result in many problems, including depression, suicide ideation and actuation, antisocial behavior, aggression, and violence. Several of the studies reviewed showed that childhood maltreatment, neglect, and adversity resulted in epigenetic changes to the serotonin transporter gene, resulting in antisocial personality disorder and other disorders (reviewed in DeLisi and Vaughn48). For example, Suderman and colleagues analyzed the promoter methylation of over 20,000  genes in 40  adult (45-year-old) males and compared those who had experienced childhood abuse with those who had not. They found almost 1000 gene promoters that were differentially methylated in the men who had been abused as children.57 In a number of studies, Beach and colleagues have shown that childhood abuse produces longterm effects on methylation levels and subsequent gene regulation in adults, 58 in particular in the serotonergic system.59,60 The  SLC64A  gene, which codes for serotonin, has been very well studied in relation to serotonin function. Functional polymorphism in the promoter region, 5-HTTLPR, results in two alleles known as the short allele (S) and the long allele (L), with the

122  Candidate genes, gene × environment interactions, and epigenetics

S allele being less functional, resulting in lower levels of serotonin, which has been linked to a number of behavioral problems. The  S allele has therefore been linked to many psychiatric disorders.61 Preliminary research shows that DNA  methylation may mediate the interaction between the S allele and early-life stress to increase the vulnerability of carriers of the S allele to later psychiatric disorders.61 A study on effects of neighborhood crime on depression showed that DNA methylation of this promoter region increased or decreased gene expression, which impacted symptoms of depression.62 Post-traumatic stress disorder

Traumatizing events are relatively common, but only some people develop post-traumatic stress disorder (PTSD). Whether a person is resilient to such trauma or will develop PTSD may relate to the extent and type of the trauma, the time of life in which it is experienced, and the ­person’s genetic background, but recently, epigenetic changes have been proposed as mediators for the risk of developing PTSD, involving both a genetic and an environmental ­etiology.63 Many human and animal studies have shown that adverse life experiences can impact DNA ­methylation and result in increased susceptibility to later adversity.64 Many different candidate genes have been studied in relation to PTSD, and epigenetic changes have been seen in genes related to neurotransmitter function and stress reactions, as well as regulation of the immune system (reviewed in Zannas, Provencal, and Binder64). As with childhood trauma, DNA methylation in several genes involved in the HPA axis, particularly those related to stress hormones, has also been seen in war veterans with PTSD.64 Alterations in immune-function genes have also been linked to PTSD, via disturbance of either the HPA axis or brain function, and epigenetic changes in such genes have been shown in many studies (reviewed in Zannas, Provencal, and Binder64). It has been suggested that epigenetic changes related to trauma may be specific to certain alleles and so may only impact individuals with certain genomes.64 Timing of trauma also appears to be crucial, again suggesting certain temporal periods of vulnerability. As discussed above, early childhood is a particularly sensitive time, but other life periods may also be sensitive, such as the prenatal period or old age, and may relate to different types of PTSD.64 A  major problem with human studies on epigenetics is that, as mentioned earlier, epigenetic changes are tissue specific, so studies on live humans are restricted to blood; brain tissue can only be examined at autopsy, which adds further complications, such as problems with determining the impact of individual lifestyles or the temporal link between exposure to the triggering trauma, development of PTSD, and the epigenetic changes.65 The differences in DNA methylation between tissues in a single healthy individual have been shown to be much greater than in the same tissues between individuals.66 Treatment potential

Epigenetic changes were at first believed to be immutable, becoming a permanent part of an individual. However, rodent and human studies have shown that such changes can be reversed. This  means that epigenetic changes are a potentially excellent target for intervention and the development of prevention and treatment programs.48 Several studies have shown that drugs that prevent histone changes can reverse the epigenetic effects of childhood adversity as effectively as behavioral therapy, so treatments in this arena should be explored. ADHD has been linked to many genetic and environmental risk factors, and recent studies have also implicated a number of epigenetic effects.67 Studies indicate that there are very critical times within development at which the fetus and subsequent child are at greatest risk for environmental stressors such as prenatal smoking, violence, or alcohol use. Stress in the mother has been linked to DNA  methylation of genes in the placenta related to a change in exposure to certain steroid hormones linked to stress.67 Early-life stressors also result in epigenetic changes. It has been suggested that an understanding of epigenetic effects can inform treatment and intervention at three levels. The first would be to attempt to reduce known stressors, such as maternal smoking during

Epigenetics  123

pregnancy. The second would be to identify epigenetic markers in the newborn that might suggest later development of ADHD and to utilize intervention methods at this stage, before symptoms begin67; however, this runs the risk of labeling. The third level would be to develop personalized interventions focusing on specific parts of the genome.67

Can epigenetic changes be inherited? When cells replicate (divide to produce new cells), much of the epigenome is passed to the new cells, which makes sense, as it keeps the cells specialized, for example, as a kidney cell. This is the normal replication of cells as a person grows older; it occurs all the time, as our cells live only for a short time and must be replaced. When reproduction occurs—that is, the reproduction of a new individual—most of the epigenome is reset, but some of the chemical marks may remain on the DNA and histones of the gametes (eggs and sperm) to be inherited by the next generation.52 This can be considered a non-genetic form of inheritance, which results in passing traits from parents to children, without involving the gene sequence.48 DeLisi and Vaughn, the authors of the review of epigenetic studies mentioned earlier, somewhat facetiously suggested that epigenetics is a “vindication” of Lamarck’s ideas.48(p607) In  1801, JeanBaptiste Lamarck proposed the theory of acquired characteristics, suggesting that what an individual organism acquired through life would be passed to any offspring. This theory was proved wrong: it is fairly obvious that if a person loses a leg and a kidney in life, the person’s children are not born with just one leg and one kidney! However, despite the somewhat tongue-in-cheek title of DeLisi and Vaughn’s paper, there does appear to be a method that we now understand does allow the passing of acquired characteristics to the next generation.

Cautions with epigenetics It is important to remember that the epigenome controls perfectly normal cell behavior, turning certain genes on or off to allow specialization of cells, and is part of the normal development of the organism. It is itself under genetic control; it usually has little to do with experience or the environment.68 DNA methylation is vital for our normal function, as it ensures that a kidney cell remains a kidney cell and a skin cell remains a skin cell and does not suddenly turn into a liver cell or get replaced by a spleen cell when it dies.69 DNA methylation linked to the environment is fairly rare, with very small effect sizes, and is not yet well understood.69 Research in this area is very young, and it is hard to accomplish. Even when gene expression is affected, gene expression is a long way from enacting a behavior: a large number of changes have to occur at the cellular and organ levels before a behavior would be impacted.69 In order to be sure that methylation can affect criminal behavior, we would first need to show that stress or an adverse environment results in methylation. Then, we would have to show that the methylation results in criminal behavior, and we are nowhere near to being able to do this.69 Experiments so far relate to more obvious factors such as exposure to chemicals and adverse nutritional status but not to more subtle issues such as social stress and lack of parental care.69 Epigenetic changes are tissue specific, meaning that epigenetics, particularly in humans, is very hard to study because, if we wish to study behavior, we should really look at epigenetic changes in the brain or the nervous system in general, which cannot be sampled in a living person, so most human studies are conducted on blood samples.68 Epigenetics is being hailed as the new way forward, and several controversial papers have been published in this area, suggesting previous methods of study are obsolete.47,70 However, this is far from the truth. So far, we know very little about epigenetics as it relates to environmental

124  Candidate genes, gene × environment interactions, and epigenetics

interaction, and most of what we do know comes from limited studies on rodents. Experts in this field have stressed the need for caution when considering epigenetics and its potential role in behavior. In  fact, renowned epigeneticists Heijmans and Mill stated that “epigenetics will not be able to deliver the miracles it is sometimes claimed it will.”68(p77),71 Not only do we not yet know enough about the very small amount of epigenetics that may be influenced by the environment, but we also run the risk of serious misuse and misunderstanding of the relationship between the environment and the genome, which could lead us to repeat the problems of the past. We are well aware of the misuse of ideas about biological explanations for criminal behavior in the past—the horrible ideologies that misused these theories, resulting in many atrocities—and the subsequent stigmatization of the idea of a biological influence on criminal behavior. Hopefully, those days have passed, with a new paradigm of understanding the interactions and inter-relationships between the genome and the environment. However, Moffit and Beckley pointed out a new concern. There is a great deal of public interest in, and misunderstanding of, epigenetics, and Moffit and Beckley suggest that this may be very dangerous as the public may be using this misunderstanding to lead them back to a deterministic way of thinking. Moffit and Beckley suggest that this is even more hazardous than the incorrect beliefs of genetic determinism of the past, as the public may no longer “blame the genes” but may, in other words, blame parents or caregivers for allowing their children’s genomes to be methylated.69

Conclusion We are making great strides in identifying genetic polymorphisms that may contribute a small amount of risk, or plasticity, to antisocial behavior, and since the first edition of this text, we have come a tremendously long way in understanding G × E interactions and developing a number of models to explain potential interaction. We have also just begun to start to understand epigenetics. The concept of epigenetics is extremely appealing, and a tremendous amount of work is being conducted in this area, particularly in relation to disease and other health issues, but so far, there is only a small amount of research on behavior, or specifically, criminogenic behavior. Epigenetics clearly links the old-fashioned ideas of nature and nurture and suggests that we may eventually gain a better understanding of how the two interact. It also offers a tremendous promise of hope for a greater understanding of, and eventual intervention, treatment, and even cure for, many of life’s ills, including all levels of antisocial behavior. Maybe. However, this field is still very much in its infancy, and caution must be taken before we come to conclusions. Very little is known about this area so far, much of the research has been conducted only on rodents, and every day, new publications come out with more, sometimes contradictory, information. Many cautions have already been mentioned as we consider epigenetics, and it is important that we move forward carefully.

Questions for further study and discussion 1. Clearly explain the differences between gene–environment correlations and gene × ­environment interactions. Use examples to illustrate each. 2. Clearly explain the differences between the diathesis stress model and the differential susceptibility model. 3. Discuss the potential for using a better understanding of G × E interactions when designing interventions and treatments for antisocial youth. 4. Cautions have been expressed that misunderstanding epigenetics runs the risk of leading us back to determinism. Explain this. 5. It has been suggested that it is more correct to refer to a risk allele as a plasticity allele. Why? Discuss the potential advantages and disadvantages.

References  125

References 1. Mullineaux, P.Y. and DiLalla, L.F. 2015. Genetic influences on peer and family relationships across adolescent development: Introduction to the special issue. J. Youth Adolesc. 44(7): 1347–1359. 2. Charney, E. 2012. Behavior genetics and postgenomics. Behav. Brain Sci. 35(5): 331–358. 3. Tuvblad, C. and Beaver, K.M. 2013. Genetic and environmental influences on antisocial behavior. J. Crim. Justice 41(5): 273–276. 4. Blum, K., Sheridan, P.J., Wood, R.C., Braverman, E.R., Chen, T.J., and Comings, D.E. 1995. Dopamine D2 receptor gene variants: Association and linkage studies in impulsive-­ addictive-compulsive behaviour. Pharmocogenetics 5(3): 121–141. 5. Biederman, J. 2005. Attention-deficit/hyperactivity disorder: A  selective overview. Biol. Psychiatry 57(11): 1215–1220. 6. Li, Z., Chang, S.H., Zhang, L.Y., Gao, L., and Wang, J. 2014. Molecular genetic studies of ADHD and its candidate genes: A review. Psychiatry Res. 219(1): 10–24. 7. Dark, C., Homman-Ludiye, J., and Bryson-Richardson, R.J. 2018. The role of ADHD associated genes in neurodevelopment. Dev. Biol. 438(2): 69–83. 8. Watts, S.J. 2018. ADHD symptomatology and criminal behavior during adolescence: Exploring the mediating role of school factors. Int. J. Offender Ther. Comp. Criminol. 62(1): 3–23. 9. Johann, M., Bobbe, G., Putzhammer, A., and Wordaz, N. 2003. Comorbidity of alcohol dependence with attention-deficit hyperactivity disorder: Differences in phenotype with increased severity of the substance disorder but not  in genotype (serotonintransporter and 5-hydroxytryptamine-2c receptor) Alcoholism: Clin. Exp. Res. 27(10): 1527–1534. 10. Molina, B.S. and Pelham, W.E.J. 2003. Childhood predictors of adolescent substance use in a longitudinal study of children with ADHD. J. Abnorm. Psychol. 112(3): 497–507. 11. Satterfield, J.H. and Schell, A.M. 1997. A prospective study of hyperactive boys with conduct problems and normal boys: Adolescent and adult criminality. J. Am. Acad. Child Adolesc. Psychiatry 36(12): 1726–1735. 12. Eyestone, L. and Howell, R. 1994. An epidemiological study of attention-deficit ­hyperactivity disorder and major depression in a male prison population. Bull. Am. Acad. Psychiatr. Law 22(2): 181–193. 13. Moffitt, T. 1990. Juvenile delinquency and attention deficit disorder: Boys’ developmental trajectories from age 3 to age 15 child development. Child Develop. 61: 893–910. 14. Zagar, R., Arbit, J., Hughes, J., Buscell, R., and Busch, K. 1989. Developmental and disruptive behavior disorders among delinquents. J. Am. Acad. Child. Adolesc. Psychiatry 28(3): 437–440. 15. Schilling, C.M., Walsh, A., and Yun, I. 2011. ADHD and criminality: A primer on the genetic, neurobiological, evolutionary, and treatment literature for criminologists. J. Crim. Justice 39(1): 3–11. 16. Gudjonsson, G.H., Sigurdsson, J.F., Young, S., Newton, A.K., and Peersen, M. 2009. Attention deficit hyperactivity disorder (ADHD): How do ADHD symptoms relate to personality among prisoners? Personality Individ. Diff. 47(1): 64–68. 17. Slutske, W.S., Heath, A.C., Dinwiddie, S.H., et al. 1997. Modeling genetic and environmental influences in the etiology of conduct disorder: A study of 2,682 adult twin pairs. J. Abnorm. Psychol. 106(2): 266–279. 18. Coolidge, F.L., Thede, L.L., and Jang, K.L. 1995. Are personality disorders pscyhological manifestations of executive function deficits? Bivariate heritability evidence from a twin study. Behav. Genetics 34(1): 75–84. 19. Beaver, K.M., Wright, J.P., DeLisi, M., et al. 2007. A gene × gene interaction between DRD2 and DRD4 is associated with conduct disorder and antisocial behavior in males. Behav. Brain Funct. 3(1): 30.

126  Candidate genes, gene × environment interactions, and epigenetics

20. Rees, E., O’Donovan, M.C., and Owen, M.J. 2015. Genetics of schizophrenia. Curr. Opinion Behav. Sci. 2: 8–14. 21. St. Clair, D., Xu, M., Wang, P., et  al. 2005. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–1961. JAMA 294: 557–562. 22. Bo, S., Abu-Akel, A., Kongerslev, M., Haahr, U.H., and Simonsen, E. 2011. Risk factors for violence among patients with schizophrenia. Clin. Psychol. Rev. 31(5): 711–726. 23. Fazel, S., Langstrom, N., Hjern, A., Grann, M., and Lichtenstein, P. 2009. Schizophrenia, substance abuse, and violent crime. JAMA 301(19): 2016–2023. 24. Fleischman, A., Werbeloff, N., Yoffe, R., Davidson, M., and Weiser, M. 2014. Schizophrenia and violent crime: A population-based study. Psychol. Med. 44(14): 3051–3057. 25. Howes, O.D., McCutcheon, R., Owen, M.J., and Murray, R.M. 2017. The role of genes, stress, and dopamine in the development of schizophrenia. Biol. Psychiatry 81(1): 9–20. 26. Misiak, B., Stramecki, F., Gaweda, L., et al. 2018. Interactions between variation in candidate genes and environmental factors in the etiology of schizophrenia and bipolar disorder: A systematic review. Mol. Neurobiol. 55(6): 5075–5100. 27. Foley, C., Corvin, A., and Nakagome, S. 2017. Genetics of schizophrenia: Ready to translate? Curr. Psychiatry Rep. 19(9): 61. 28. Lencz, T., Knowles, E., Davies, G., et al. 2014. Molecular genetic evidence for overlap between general cognitive ability and risk for schizophrenia: A report from the Cognitive Genomics consorTium (COGENT). Mol. Psychiatry 19(2): 168–174. 29. Plomin, R., McClearn, G.E., Pedersen, N.L., Nesselroade, J.R., and Bergeman, C.S. 1988. Genetic influence on childhood family environment perceived retrospectively from the last half of the life span. Develop. Psychol. 24(5): 738–745. 30. Scarr, S. and McCartney, K. 1983. How people make their own environments: A theory of genotype-environment effects. Child Develop. 54: 424–435. 31. Beaver, K.M., Mancini, C., DeLisi, M., and Vaughn, M.G. 2011. Resiliency to victimization: The role of genetic factors. J. Interpers. Violence 26(5): 874–898. 32. Gajos, J.M., Fagan, A.A., and Beaver, K.M. 2016. Use of genetically informed evidence-based prevention science to understand and prevent crime and related behavioral disorders. Crim. Public Pol. 15(3): 683–701. 33. Mill, J., Dempster, E., Caspi, A., Williams, B., Moffitt, T., and Craig, I. 2006. Evidence for monozygotic twin (MZ) discordance in methylation level at two CpG sites in the promoter region of the catechol-O-methyltransferase (COMT) gene. Am. J. Med. Genet. Part B. Neuropsychiatric Genet. 141B(4): 421–425. 34. Acheson, A., Conover, J.C., Fandl, J.P., et al. 1995. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 374(6521): 450–453. 35. Weder, N., Zhang, H., Jensen, K., et al. 2014. Child abuse, depression, and methylation in genes involved with stress, neural plasticity, and brain circuitry. J. Am. Acad. Child Adolesc. Psychiatry 53(4): 417–424 e5. 36. Bersted, K.A. and DiLalla, L.F. 2016. The influence of DRD4 genotype and perinatal complications on preschoolers’ negative emotionality. J. Appl. Developmental Psychol. 42: 71–79. 37. Boardman, J.D., Menard, S., Roettger, M.E., Knight, K.E., Boutwell, B.B., and Smolen, A. 2014. Genes in the dopaminergic system and delinquent behaviors across the life course: The role of social controls and risks. Crim. Justice Behav. 41(6): 713–731. 38. Hicks, B.M., South, S.C., DiRago, A.C., Iacono, W.G., and McGue, M. 2009. Environmental adversity and increasing genetic risk for externalizing disorders. Arch. Gen. Psychiatry 66(6): 640–648. 39. Feinberg, M.E., Button, T.M.M., Neiderhiser, J.M., Reiss, D., and Hetherington, E.M. 2007. Parenting and adolescent antisocial behavior and depression. Arch. Gen. Psychiatry 4(64): 4. 40. Burt, S.A. and Klump, K.L. 2014. Parent-child conflict as an etiological moderator of childhood conduct problems: An example of a “bioecological” gene-environment interaction. Psychol. Med. 44(5): 1065–1076.

References  127

41. Belsky, J. and Pluess, M. 2009. Beyond diathesis stress: Differential susceptibility to environmental influences. Psychol. Bull. 135(6): 885–908. 42. Simons, R.L., Lei, M.K., Beach, S.R., Brody, G.H., Philibert, R.A., and Gibbons, F.X. 2011. Social environmental variation, plasticity genes, and aggression: Evidence for the differential susceptibility hypothesis. Am. Sociol. Rev. 76(6): 833–912. 43. Beaver, K., Nedelec, J.L., Schwartz, J.A., and Connolly, E.J. 2014. Evolutionary behavioral genetics of violent crime, In: The Evolution of Violence, Shackelford, T.K. and Hansen, R.D., editors. New York: Springer; pp. 117–136. 44. Boardman, J.D., Barnes, L.L., Wilson, R.S., Evans, D.A., and Mendes de Leon, C.F. 2012. Social disorder, APOE-E4 genotype, and change in cognitive function among older adults living in Chicago. Soc. Sci. Med. 74(10): 1584–1590. 45. Brody, G.H., Beach, S.R.H., Philibert, R.A., Chen, Y.-F., and Murry, V.M. 2009. Prevention effects moderate the association of 5-HTTLPR and youth risk behavior initiation: Gene environment hypotheses tested via a randomized prevention design. Child Develop. 80(3): 645–661. 46. Brody, G.H., Chen, Y.F., Beach, S.R., et al. 2014. Differential sensitivity to prevention programming: A dopaminergic polymorphism-enhanced prevention effect on protective parenting and adolescent substance use. Health Psychol. 33(2): 182–191. 47. Burt, C.H. and Simons, R.L. 2014. Pulling back the curtain on heritability studies: Biosocial criminology in the postgenomic era. Criminology 52(2): 223–262. 48. DeLisi, M. and Vaughn, M.G. 2015. The vindication of Lamarck? Epigenetics at the intersection of law and mental health. Behav. Sci. Law 33(5): 607–628. 49. Fraga, M.F., Ballesta, E., Paz, M.F., et al. 2005. Epigenetic differences arise during the lifetime of monozygotic twins. PNAS 102(30): 10604–10609. 50. Edwards, M. and Abadie, L. 2018. NASA twins study confirms preliminary findings. Accessed 2018 May 8; www.nasa.gov/feature/nasa-twins-study-confirms-preliminary-findings. 51. Eitan, E., Hutchison, E.R., and Mattson, M.P. 2014. Telomere shortening in neurological disorders: An abundance of unanswered questions. Trends Neurosci. 37(5): 256–263. 52. National Human Genome Research Institute. 2016. Epigenomics. Accessed October  13, 2018; https://www.genome.gov/27532724/epigenomics-fact-sheet/. 53. Roth, T.L. and Sweatt, J.D. 2011. Annual research review: Epigenetic mechanisms and environmental shaping of the brain during sensitive periods of development. J. Child Psychol. Psychiatry 52(4): 398–408. 54. Weaver, I.C.G., Cervoni, N., Champagne, F.A., et al. 2004. Epigenetic programming by maternal behavior. Nat. Neurosci. 7(8): 847–854. 55. Mehta, M.A., Golembo, N.I., Nosarti, C., et  al. 2009. Amygdala, hippocampal and corpus callosum size following severe early institutional deprivation: The English and Romanian Adoptees study pilot. J. Child Psychol. Psychiatry 50(8): 943–951. 56. Rutter, M. 2012. Achievements and challenges in the biology of environmental effects. Proc. Natl. Acad. Sci. USA 109(Suppl. 2): 17149–17153. 57. Suderman, M., Borghol, N., OPappas, J.J., et  al. 2014. Childhood abuse is associated with methylation of multiple loci in adult DNA. BMC Med. Genet. 7(1): 13. 58. Beach, S.R., Brody, G.H., Lei, M.K., et al. 2013. Impact of child sex abuse on adult psychopathology: A genetically and epigenetically informed investigation. J. Fam. Psychol. 27(1): 3–11. 59. Beach, S.R.H., Brody, G.H., Todorov, A.A., Gunter, T.D., and Philibert, R.A. 2010. Methylation at SLC6A4 is linked to family history of child abuse: An examination of the Iowa Adoptee sample. Am. J. Med. Genet. B Neuropsychiatr. Genet. 153B(2): 710–713. 60. Beach, S.R., Brody, G.H., Todorov, A.A., Gunter, T.D., and Philibert, R.A. 2011. Methylation at 5HTT mediates the impact of child sex abuse on women’s antisocial behavior: An examination of the Iowa adoptee sample. Psychosom. Med. 73(1): 83–87. 61. Palma-Gudiel, H. and Fananas, L. 2017. An integrative review of methylation at the serotonin transporter gene and its dialogue with environmental risk factors, psychopathology and 5-HTTLPR. Neurosci. Biobehav. Rev. 72: 190–209.

128  Candidate genes, gene × environment interactions, and epigenetics

62. Lei, M.K., Beach, S.R., Simons, R.L., and Philibert, R.A. 2015. Neighborhood crime and depressive symptoms among African American women: Genetic moderation and epigenetic mediation of effects. Soc. Sci. Med. 146: 120–128. 63. Rampp, C., Binder, E.B., and Provencal, N. 2014. Epigenetics in posttraumatic stress disorder. Prog. Mol. Biol. Transl. Sci. 128: 29–50. 64. Zannas, A.S., Provencal, N., and Binder, E.B. 2015. Epigenetics of posttraumatic stress disorder: Current evidence, challenges, and future directions. Biol. Psychiatry 78(5): 327–335. 65. Zannas, A.S. and West, A.E. 2014. Epigenetics and the regulation of stress vulnerability and resilience. Neurosci. 264: 157–170. 66. Davies, M.N., Volta, M., Pidsley, R., et  al. 2012. Functional annotation of the human brain methylome identifies tissue-specific epigenetic variation across brain and blood. Genome Biol. 13(R43): 1–14. 67. Bhat, V., Joober, R., and Sengupta, S.M. 2017. How environmental factors can get under the skin: Epigenetics in attention-deficit/hyperactivity disorder. J. Am. Acad. Child Adolesc. Psychiatry 56(4): 278–280. 68. Heijmans, B.T. and Mill, J. 2012. Commentary: The seven plagues of epigenetic epidemiology. Int. J. Epidemiol. 41(1): 74–78. 69. Moffitt, T.E. and Beckley, A. 2015. Abandon Twin Research? Embrace Epigenetic Research? Premature Advice for Criminologists. Criminology 53(1): 121–126. 70. Burt, C.H. and Simons, R.L. 2015. Heritability studies in the postgenomic era: The fatal flaw is conceptual. Criminology 53(1): 103–112. 71. Barnes, J.C., Wright, J.P., Boutwell, B.B., et  al. 2014. Demonstrating the validity of twin research in criminology. Criminology 52(4): 588–626.

7 The chemistry of the body The effects of hormones on behavior

Introduction This  chapter is one of several that considers the areas of biology that do not  directly deal with genetics, although hormones and neurotransmitters are controlled by genes. It first provides a brief overview of the normal function of hormones and how they influence behavior. There is a very fine control kept on hormone levels through negative feedback mechanisms, and this system in turn is one of the controls of behavior. One of the most hotly contested areas in biological criminology relates to the so-called male and female hormones, as well as their relationships with other hormones. Hormone imbalances are treated medically every day, which means that as greater understanding is achieved, better interventions may be developed.

The functions of hormones Hormones (from the Greek word hormon, meaning to excite or to set in motion) are chemical signals. They are released into the blood by endocrine cells (the building blocks of the endocrine system, which include the hypothalamus, pituitary, and thyroid glands) and neurosecretory cells, which are specialized nerve cells that also make hormones. These cells sometimes release the hormones directly, but they also often store them in a gland for later release (for example, the digestive glands or the pancreas). Once something gets into the bloodstream, it will reach every cell in the body, because that is what the blood is designed to do—to take oxygen and nutrients to every single cell and take away waste products, such as carbon dioxide and nitrogenous waste. However, although the hormones will get to every cell, only certain types of cells, the target cells for that particular hormone, will respond. The other cells will ignore it. Changes in hormone response could therefore be caused by the hormone itself, by the cells that produce it, or by the target cells’ ability to respond. Hormones regulate our metabolism, growth, development, and behavior. They are active in tiny amounts, and a slight change in concentration can have a dramatic effect on the body. Endocrine cells are usually clustered together in organs or glands such as the pancreas, testes, and ovaries, and they secrete a variety of hormones (for example, insulin, testosterone, progesterone, and estrogen). There are more than 50 known hormones in the human body. In many cases, a hormone is countered by an antagonistic, or opposing, hormone. For example, insulin is produced by the pancreas in response to high blood sugar. High blood sugar prevents the cells from absorbing nutrients,

129

130  The chemistry of the body

so if the sugar in the blood gets beyond a certain point, receptors in the pancreas release insulin to bring it back down, but if the blood sugar gets too low, an antagonistic hormone (that is, glucagon) is released to send the blood sugar level back up. This is a simple feedback mechanism that keeps the blood sugar at a concentration very close to a set point, whatever your diet. When this system fails, diabetes occurs. The body has many systems similar to this one that maintain the status quo. Some hormones fluctuate naturally in the body (for example, during a woman’s menstrual cycle). If hormonal balances or levels are abnormal, the effects can be severe. We can experience behavioral changes and mood swings, or at the extreme end of the range, we can become ill and even die. Hormones are vital to our normal body function, and experiments have proven that they do many things; affecting our behavior is one of their functions. When levels of hormones change, because of a change in the amount of the hormone or the antagonist, or a change in the target cells, things can go wrong. Remember that hormones are designed to influence behavior and that levels vary in different people for a variety of reasons. Several major hormones have been linked with criminal behavior.

Testosterone Although many people believe that only social factors are involved in criminal behavior and that biology plays no role, one of the basic tenets of criminality is that in every culture and society, irrespective of social treatment, males commit more crimes and more serious crimes than females.1 So, what is it about males? Even people who are strongly opposed to the idea that any aspect of a person’s biology can influence behavior still seem to accept that testosterone equals aggression, and the more the testosterone, the more violent the man. The idea that a violent person has too much testosterone is common knowledge. But is there any truth behind this belief? Is there a link between testosterone and aggression? And if so, is it causal? That is, even if there is a link, does the testosterone cause the aggression or does the aggression result in higher testosterone levels—or are the two mediated by something else entirely? Testosterone is a steroidal hormone, one of several androgens, or “male hormones,” that are important in developing and maintaining male’s primary and secondary sexual characteristics. These characteristics include increased muscle mass and strength, increased bone density and strength, stimulation of growth in height, growth of penis and testes, deepening of the voice, and growth of facial and body hair. Testosterone is secreted by the hypothalamic-pituitary-gonadal (HPG) axis, which involves the hypothalamus, a part of the brain that links the nervous system to the endocrine system via the pituitary gland (a small pea-sized gland at the bottom of the hypothalamus) and the gonadal glands (the testes and ovaries). Testosterone is found in both males and females and is secreted by the testes and ovaries, but it is at many times higher levels in males than in females. It is produced by the Leydig cells in the testes and, in smaller amounts, by the adrenal cortex, which is part of the adrenal gland, found on the kidneys.2 More than 120 years ago, a highly respected French doctor, Charles Sequard, shocked the scientific community, and later the world, with his discovery that testicles contained an invigorating substance that could be extracted from animals and injected into humans.3 The  elderly doctor admitted trying the substance himself and reported increased mental concentration and physical endurance, as well as many other physical and mental benefits. Although, unfortunately, Sequard was ridiculed at the time, today, he is recognized for the pioneering work he did with hormones and glands, discovering the effects of testosterone some 50 years before testosterone was identified in 1935.3 Testosterone moderates gene expression by entering the blood plasma, where it remains as testosterone or may be metabolized to dihydrotestosterone or estradiol, then binding to receptors that in turn moderate gene expression. Testosterone has a large range of effects, only one of which impacts aggression; this makes aggression difficult to isolate for study.4 In addition to its role in gene expression, testosterone may affect behavior via other mechanisms.4

Testosterone  131

It has always been understood that the testicles related in some way to sexual ability in males. Farmers knew that a castrated animal could not reproduce, and in days past, the same was seen to be true in humans. So, the testicles, and consequently testosterone, are clearly linked to sperm production and sex drive, but are they related to aggression? It was also common knowledge that castrating, or neutering, an animal reduces its intraspecies aggression and overall “spirit.” A gelded or castrated male horse is much more malleable, and a neutered dog is far less aggressive toward other male dogs. The relationship between aggression and testosterone in animals has been known for decades and has been confirmed in many studies.5 Testosterone plays a major role in sexual behavior, so in animal species it makes evolutionary sense for testosterone to be linked to aggressive behavior, as it will result in acquiring resources and, most importantly, mates. The  driving need in all animals is to reproduce and thereby pass their genes on to the next generation, so aggression would help them be “successful.” Studies have shown that levels of testosterone after puberty are highly heritable. A twin study in the Netherlands showed that 66% of the variance was due to genetics in males and 41% in females.6 In another twin and sibling study looking at testosterone and other reproductive hormones—such as inhibin B, follicle-stimulating hormone (FSH), sex ­hormone-binding protein (SHBP), and luteinizing hormone (LH)—all were found to be highly ­heritable, with heritability ranging from 56% in testosterone to 81% for SHBP and inhibin  B.7 In contrast, in a large study of twin pairs tested at 5 months of age, testosterone levels were determined to be primarily impacted by environment,8 suggesting differences in the mechanisms for prepubertal and postpubertal levels. Testosterone is known to impact several neurobiological systems that relate to behavioral modification,9,10 including: ■■

■■

■■

Reduced coupling or interaction between part of the frontal lobe (the orbitofrontal cortex) and the amygdala. As we will see in Chapter 10, these areas are heavily involved in behavior control and inhibition, so reduced coupling may result in reduced regulation of emotions. Increased activity in regions of the brain related to sensitivity to increased or decreased reward and to risk-taking. Suppressed stimulus-response of the hypothalamic-pituitary-adrenal (HPA) axis (which we will discuss later), which will result in resistance to stress.

Testosterone exposure The body is exposed to testosterone at two very different times of life in mammals, when it plays a vital role in masculinization. The first is prenatal (before birth), referred to as the organizational stage, and the second occurs years later, post-puberty, and is referred to as the activation stage.1 Almost all ­studies in the past have measured testosterone only in the latter stage, although this is changing. Although we usually think of hormones like testosterone as only having an impact after puberty, in fact, testosterone’s most important effects occur in the prenatal period. During this period, testosterone sensitizes the brain to its later effects.4 It was first shown almost 60 years ago that testosterone sensitizes, or organizes, the mammalian brain early in ontogeny (development), during sexual differentiation of the fetus, when the originally feminine brain is masculinized in a fetus that has been determined by chromosomes to be male.11 Animal studies showed that castrating a neonate (newborn) male mouse reduced adult aggression, and in contrast, injecting testosterone into a neonate mouse increased adult aggression.4 However, the relationship was not direct: the increased neonatal testosterone did not cause aggression but instead sensitized the brain to respond more strongly to testosterone in a later life period. In other words, testosterone-related aggression in later life relates to testosterone level prenatally, when the neural system became sensitized or organized.4 Because it structurally affects the brain, it is irreversible.12 In further experiments, female rodents could be induced to be as aggressive as males if sensitized to testosterone neonatally and then exposed to testosterone again in adulthood,

132  The chemistry of the body

owing to the early sensitization. Other studies showed that position in the womb of male and female rodent embryos can impact adult aggression, owing to variations in the levels of testosterone received by an embryo neighboring a male.5 Similar results have been seen in humans. Obviously, such studies cannot be conducted in humans, but several lines of evidence have suggested that the prenatal levels of testosterone sensitize the human neural system to display testosterone-related aggression when exposed to testosterone later in development.4 First, there is a disease called congenital adrenal hyperplasia, wherein levels of testosterone are extremely high in early development, resulting in increased aggression in females, closer to that normally seen in males. Second, in rare situations, pregnant women have been treated with testosterone, and both male and female children have shown heightened aggression. Third, the ratio between the length of the index finger and the ring finger (known as 2D:4D) has been shown to be an excellent marker for prenatal testosterone levels and is inversely related to aggression, suggesting that higher prenatal testosterone results in higher aggression in adulthood. Finally, girls in opposite-sex twin pairs may have levels of aggression similar to males, which may relate to the animal studies looking at position in the womb.4 Therefore, in the prenatal stage, testosterone plays a critical role in masculinization of the fetus during ontogeny and organizes the brain to prepare it for later testosterone exposure. The  next period of testosterone exposure occurs at and after puberty, when the body produces large amounts of the hormone. The next period of testosterone exposure occurs at and after puberty, at the activation stage, when the body produces large amounts of the hormone. The impact of testosterone at this stage is dependent on the testosterone sensitization which occurred during the earlier prenatal or organization stage. During puberty, testosterone is vital in the masculinization of the boy to a man, including development of the gonads and secondary sexual characteristics.

Prenatal testosterone and behavior It should be supposed that levels of prenatal and postpubertal testosterone would be correlated, but this does not always appear to be the case,13 although a correlation has been seen in some studies. High levels of salivary testosterone in aggressive soccer players correlated with low 2D:4D ratio or, in other words, with high prenatal testosterone.14 However, not all studies support this, so caution should be taken before assuming that postpubertal levels can be used to infer prenatal levels. As prenatal testosterone levels have such a critical effect in sensitizing, or organizing, the developing brain to later exposure to testosterone, it is unlikely that postpubertal blood or saliva testosterone levels will relate directly to aggressive behavior; rather, it is the levels of prenatal testosterone that are most likely to be important. This may be why there is such variation in study results attempting to find a direct relationship between aggression and testosterone levels. Although it is relatively easy to measure testosterone levels in living people, it is not possible to do so in the prenatal environment without highly invasive and damaging procedures, such as amniocentesis. Even if this were possible, a longitudinal study would have to be performed over the following two to three decades to make any connections. However, an extremely simple and non-invasive technique to measure prenatal testosterone and overall androgen levels was discovered in the late 1990s. Researchers found that the ratio of the length of the second digit (2D), or index finger, and the fourth digit (4D), or ring finger, could be used to infer prenatal testosterone levels.1 A longer ring finger in relation to the index finger—that is, a lower ratio—is linked to higher prenatal testosterone levels, and this relationship is seen not just in humans but also in rats and frogs.12 It has since been confirmed in many studies. In rare cases when amniocentesis has been medically required, direct measurement of intrauterine testosterone levels was compared with the offspring’s 2D:4D ratio at 2 years old and upheld the relationship.15 Moreover, a study of men with Klinefelter’s syndrome, in which men have an XXY chromosome complement and so lower testosterone levels, showed that affected men had 2D:4D ratios that were higher than those of their male family members or XY men, more consistent with a female hand.15

Testosterone  133

The  relationship between 2D:4D ratio and testosterone is believed to relate to bone growth. Testosterone is very important in bone growth, and the time when prenatal testosterone rises sharply coincides with a sudden growth period in the fourth digit, or ring finger. Animal studies have shown that this particular digit has much higher levels of androgen receptors than others, and when these receptors were experimentally inhibited, the digit did not grow as much relative to the second digit, indicating a more female ratio.16 Further research indicates that the ratio on the right hand is a better predictor of testosterone levels than the left, in both humans and animals.1 Prenatal testosterone and criminal behavior

Several studies have been conducted on the relationship between 2D:4D ratio, as a measure of prenatal testosterone exposure and later criminality. In an earlier study of both males and females, a lower 2D:4D ratio (that is, higher testosterone levels prenatally) was associated with higher aggression in males but not in females.17 Similarly, higher aggression and risk-taking were found in males with lower 2D:4D ratios.18 In  another study comparing criminal offenders with non-offenders, convicted males had lower ratios.19 When considering more minor offences, men in Germany with lower ratios were found to incur more traffic violations, but there was no relationship with ­sensation-seeking behavior, suggesting that several mechanisms may be involved.20 Many other studies have confirmed these results, with a lower 2D:4D ratio linked to aggression, athletic ­prowess, and dominance.21 Higher levels of prenatal testosterone have also been linked with aggression in professional sportsmen. A study looking at Italian professional soccer players using both 2D:4D and salivary tests to determine levels of testosterone both prenatally and postpubertally showed a correlation between higher levels of testosterone at both times and player aggression on the field.14 Many of the 2D:4D studies were performed on relatively small sample sizes, but a very large study that included two very different countries indicated a clear relationship between prenatal testosterone levels and criminal behavior. Over 2000 undergraduate students in Malaysia and more than 1200 in the United States were surveyed using self-report questionnaires that included 13 types of criminal or delinquent behaviors, ranging from drug use and minor vandalism to rape and serious assault.1 Participants self-determined digit length based on a 5-point system in which they scored their right second digit as much longer, slightly longer, equivalent, slightly shorter, or much shorter than the fourth digit. Statistically significant relationships were found between most crimes and reduced 2D:4D ratio (that is, higher androgen levels prenatally) in both males and females.1 Because digit length was determined by the participants in this study, it may have been incorrect, so a subsequent study repeated the experiment, this time with the 2D:4D ratio measured directly by the researchers. The results were upheld, with a strong correlation between estimated prenatal testosterone exposure and criminal behavior.13 This research was published in the journal Criminology, perhaps the flagship journal for the field of criminology, so it was considered extremely robust. Therefore, many studies have shown a link between 2D:4D ratio and antisocial behavior, in particular aggression, although some studies have shown inconsistent results and yet others have found no relationship.12 Two large meta-analyses have been conducted to date. Pratt and colleagues considered crime broadly, looking at studies that involved a variety of criminal activities and included 47 empirical studies through the end of 2015.22 A small overall effect size was shown when all 47 studies were combined, suggesting that the link is not nearly as clear as has been suggested by some studies. The  authors suggested caution when using 2D:4D ratios, as emphasis has been heavily placed on studies that show a link, with little attention to those that do not show a link.22 In the second meta-analysis, 32 empirical studies considering 2D:4D ratio and serious antisocial behavior, such as violence and aggression, were analyzed up to May 2016, extending the previous study by 6 months.12 Studies came from a variety of countries, with various ethnic backgrounds, age ranges, and study designs. The meta-analysis showed that 2D:4D ratio was not reliably linked to aggressive behavior and the overall mean effect size was quite low.12 This  meta-analysis was

134  The chemistry of the body

performed because the robust results seen in studies by Ellis1 and Hoskins13 appeared so strong that there was concern that practitioners might put too much reliance on such an easily measured criterion when considering risk factors, labeling, treatment, and even sentencing.12 The  authors of this meta-analysis caution that the effect size they found was lower than other risk factors that are only very moderately predictives of aggression, such as attention deficit hyperactivity disorder (ADHD), IQ, and self-esteem. They encouraged further research in this area but suggested caution in moving too fast with what appears to be such a useful trait.12 Aggression, of course, is just one antisocial behavior, and many studies have shown links with testosterone levels in a variety of other antisocial behaviors, as well as non-criminal behaviors. The authors do not state whether they believe that the 2D:4D ratio is simply not a good measure of prenatal testosterone levels or whether they do not believe that prenatal testosterone levels impact aggressive behavior. In a study designed to expand the use of 2D:4D ratio as a measure of prenatal androgen exposure, a number of other measures that are known to be influenced by androgens, such as height, sports prowess, strength, and muscularity, were included.23 The study included an ethnically mixed group of 445 male and female university students who answered a questionnaire on 14 different types of offence. Finger measurements were made by the researchers, and participants were asked their height and to score their muscularity, physical strength, and athletic prowess. Low 2D:4D ratio predicted some but not  all forms of criminal behavior in both males and females and the group overall.23 This makes sense, as not all crime has a single root cause. Moreover, there was no more relationship to violent crime than to property or victimless crime, suggesting, as many others have, that the risk is to antisocial behavior, which accrues resources to attract mates, rather than to violence. Using several measures of prenatal androgen levels did not increase associations, so the authors concluded that the use of 2D:4D alone was enough to draw conclusions.23 Recent research has shown more specific relationships between prenatal testosterone levels and 2D:4D ratio, linking low ratios to sensitivity to sudden increases in testosterone rather than just resting state, as well as to the ratio between testosterone and cortisol.24 Aggression, as we know, is not  caused or even impacted by just one factor, and no data have shown that men with high prenatal and/or postpubertal testosterone levels are aggressive or even criminal. Even if the link is proven, it does not mean that high testosterone levels cause criminal behavior—far from it. The  suggestion is that in some manner, it is a predisposer for antisocial behavior. Many of these studies link high prenatal testosterone levels to a variety of criminal acts, many of which may relate more to risk-taking than actual criminal offending. Prenatal testosterone and risk-taking behavior

As mentioned in the previous section, several studies have linked higher levels of p ­ ostpubertal ­testosterone with higher risk-taking in various arenas rather than with crime alone. Studies ­considering prenatal testosterone levels with later risk-taking have seen similar results. Financial risk-­taking has been likened to a form of resource-gathering, to attract mates and maximize ­reproductive success. A study of almost 100 university males using salivary samples and 2D:4D ratios to determine postpubertal and prenatal testosterone levels found that postpubertal ­circulating testosterone was related to financial risk-taking, but no link was found for prenatal levels.25 Other studies have shown that increased financial risk-taking was inversely proportional to 2D:4D ratio, but interestingly, they have also shown that there was a link to personality, in that right-hand 2D:4D was linked to emotion and left-hand 2D:4D to generally agreeable traits.26 In a large study of over 400 ethnically diverse university students, researchers looked at risktaking in a number of areas, such as financial, health, recreational, social, and ethical arenas.21 Researchers measured 2D:4D ratio and rel2—the length of the second finger relative to the length of all four fingers, which had been shown in previous work to be a better measure of sexual differences. The results showed that prenatal testosterone levels increased the adult males’ risk-taking behavior in social, financial, and recreational areas but not in ethical and health areas, when ethnically homogenous groups were considered. No effect was seen in women.21 The  authors then compared both ethnically homogenous and heterogenous students and found higher significance

Testosterone  135

within ethnic groups, with only financial risk being significant in the entire heterogenous group. The authors suggested that it is important to separate ethnic groups in such studies, and this lack may account for studies that have had negative results.21 However, in a study of offenders and non-offenders in England using 2D:4D ratio, no difference in risk-taking or 2D:4D ratio was seen between the two groups.27 Offenders had mostly committed drug and burglary crimes, but some had committed violent crimes, not including murder. All had been recently released. Although a lack of correlation between offenders and 2D:4D ratio has been seen in other studies, the lack of correlation between offending and risk-taking is surprising, as most studies indicate higher risk-taking in offenders, especially in caught offenders. The author speculated that because the risk-taking measurement involved real financial risk and gain, the fact that the offenders had only recently been released, had not earned money while incarcerated, and were mostly unemployed with very meager incomes perhaps made them more cautious.27

Postpubertal testosterone and behavior Most studies consider testosterone levels in adolescents and adults in comparison with aggression or antisocial behavior. Animal studies show a strong positive link between testosterone and aggression.5 In contrast, studies on the possible relationship between aggression and testosterone in humans show mixed results. There are several possible reasons for this. In animal studies, actual aggression can be observed in a contextual basis; in contrast, in most human studies, aggression is determined by self-report questionnaires that measure baseline (or trait) aggression rather than contextual aggression.28 As such, most studies comparing criminal with non-criminal individuals look at baseline testosterone levels, not at levels occurring just before or after an aggressive or criminal act. Baseline levels of testosterone are stable over weeks or months, whereas testosterone is known to fluctuate dramatically depending on challenge or social context, so acute levels of testosterone are much more likely to give us insight into the relationship between aggression and testosterone than baseline levels.28 Also, most human studies fail to differentiate between reactive aggression and proactive aggression, which have different evolutionary backgrounds. Reactive, or hostile, aggression is defensive and is a response to an actual or perceived threat. It  is a product of anger and impulsivity and includes high levels of physiological arousal and reduced inhibition. It is often referred to as “hotblooded.” Proactive aggression, on the other hand, is “cold-blooded”: it involves longer-term planning and low arousal and is aimed at acquiring resources.28(p936) Another reason for inconsistencies may be that it used to be very difficult to measure testosterone levels. In many of the earlier studies, only one blood test was taken.1 Testosterone is secreted episodically, so a single reading is not a reliable indicator. In addition, blood is not a good sampling medium. Blood tests are invasive, and 98% of the testosterone in the blood is bound to sex hormone-binding globulin (SHBG), which means that this 98% is not free to enter target cells and bind with receptors. So, only 2% of the testosterone in blood is physiologically active.29 More recent research uses testosterone levels in saliva. Not only is the procedure much less invasive, but the testosterone in saliva is unbound and therefore free to act, and it has been found to correlate highly with free serum testosterone.30 Postpubertal testosterone and criminal behavior

People have always recognized that most violent crime is committed by males and that a great deal of violent crime and delinquency occurs during adolescence and early adulthood, when testosterone levels are increasing dramatically.31 Innumerable studies have been conducted to determine whether testosterone is linked to aggression or criminal behavior, with no clear answer. The  first studies to be conducted were on animals, which can be much more easily manipulated than humans. The  first animal studies were conducted back in the 1930s and 1940s and showed increased aggression in chickens and rodents.32 Several animal studies have shown that acute changes in testosterone levels occur due to competitive interactions and that levels rise after

136  The chemistry of the body

victory.33 Moreover, a rise in testosterone levels after winning appeared to moderate future aggression. Research in fish showed that the increase in testosterone after winning a fight increased the chance of winning future fights, but this could be blocked by experimentally reducing testosterone after a win.33 Similar results have been seen in humans.31 Increased aggression, as measured by the point subtraction aggression paradigm (PSAP), an accepted measure of human reactive aggression, was found after competition in males but not in females in several studies.32 Interestingly, an intervention treatment for kindergarten children considered at risk for antisocial behavior was assessed 20 years later and was found to be successful, as it had reduced testosterone reactivity to provocation in the young adults.32 Most human studies can be divided into three general types3: 1. Natural testosterone levels in aggressive or criminal men 2. Increasing testosterone levels in normal men 3. Decreasing testosterone levels, castration 1.  Natural testosterone levels in aggressive or criminal individuals

This is one of the simplest studies, in which levels of natural testosterone are measured in an offender or criminal group and compared with levels in a matching non-offender group. The normal warning applies of course—that is, the control group may well include successful (that is, uncaught) offenders. Many early studies compared baseline testosterone levels in violent offenders with non-violent offenders and found higher levels of testosterone in violent offenders. Several of these studies were conducted by Dabbs and colleagues on incarcerated offenders. For example, in a very large prison study including almost 4500 inmates, Dabbs and colleagues found high levels of testosterone were positively correlated with a variety of problems, including delinquency, substance abuse, conflicts with authority figures, and promiscuity34; however, many of these problems have myriad causes, so it is impossible to blame just one source. In an earlier study on incarcerated young males, the same research group found that high testosterone levels were positively correlated with more violent crimes, rule violations in prison, and parole board decisions against release.35 Although testosterone is always considered a male hormone, it is also found in women, produced by cells in the ovaries and placenta. In a prison study of incarcerated women, a correlation was also found between high testosterone levels and violence.36 High testosterone was found to be specific to the type of violence. It was highest in female prisoners who had committed unprovoked assaults, and lowest in inmates who had only reacted violently when they were physically assaulted.36 In a later study of 87 female inmates in a maximum-security prison, Dabbs and colleagues compared both the women’s ages and testosterone levels with the level of violence of the crime for which they had been convicted, as well as their aggression and dominance while in prison.37 The researchers found a direct link between testosterone, criminal behavior, and aggressively dominant behavior in prison and also found that criminal violence and aggressive dominance decreased with age. 37 This may have been due to reduced testosterone levels in older women, but it could also be linked to many other factors linked to increased age, as well as to length of time spent in prison and the total numbers of years spent incarcerated, which were not considered. Moreover, the level of violence of the crime committed before incarceration may well indicate the level of violence once imprisoned. Interestingly, the five women who had the lowest testosterone levels were reported by prison staff to be very treacherous and sly. The researchers suggested that because there is a strong link between dominance and testosterone, those women at the bottom of the dominance scale due to low testosterone levels may have had to resort to less confrontational methods of dealing with others.37 Dabbs et al. also linked testosterone levels to specific types of crimes, such as those involving sexual assault or violence. They measured salivary testosterone levels in 692 adult incarcerated men and compared these with their criminal records and their prison records of behavior while incarcerated.38 Offenders convicted of crimes that involved interpersonal violence, such as sexual assault had higher levels of testosterone than offenders convicted of property crimes or drug offenses.

Testosterone  137

During incarceration, offenders with higher levels of testosterone were more likely to break prison rules and be overly confrontational.38 This  study was extended in 2001, examining data from a subset of 230 male prisoners from the previous sample in order to investigate not only the type of crime committed but also the way in which the men had actually committed their crime(s), such as level of planning, whether the consequences had been intended, the level of violence of the crime, whether the action was especially callous, whether the victim was a stranger, and whether sexual assault occurred.39 The results varied depending on the crime. In homicide cases, inmates with higher testosterone levels were significantly more likely to kill people they knew and also more likely to plan the action ahead of time. They tended to be more callous than those with low testosterone levels but not significantly so.39 Interestingly, testosterone level was not related to these parameters in inmates who committed crimes of robbery, assault, and sexual assault, without homicide. The  authors interpreted the results to mean that high-testosterone killers were more ruthless than low-testosterone killers. However, the sample size, although originally good, becomes much smaller when sub-sampled in this manner. More recently a meta-analysis of seven studies on testosterone levels in sex offenders in comparison with non-sex offenders looked at baseline testosterone levels and found no significant relationship, although some differences between rapists and child molesters were noted.40 In another study of sex offenders, several androgens were assessed, including free and total testosterone, as well as gonadotrophic hormones, FSH, and LH, which activate the Leydig cells to produce testosterone. Indicators of sex drive and hostility were also measured. Over 770 adult male sex offenders were tested and then retested up to 20 years after the offenders had been free in the community. The gonadotrophic hormones were better predictors of sexual violence recidivism than testosterone and correlated positively with self-reports.41 All these studies, however, look at baseline testosterone levels, and a large number of studies since have shown that baseline testosterone levels are at best only very weakly correlated with aggression. In actuality, testosterone fluctuates diurnally and also acutely in response to certain conditions, and many studies have shown that acute changes in testosterone levels are much more strongly and reliably correlated with aggression.31 Moreover, these studies compare testosterone levels while in prison with a past offence, which assumes that the testosterone levels were the same at both times. Baseline testosterone levels are relatively stable, but this assumes that the testosterone at the time of the offence is the same as that measured months or years later, which is highly unlikely. Testosterone increases rapidly in certain social conditions, and it is extremely likely that testosterone levels at the time of the offence were different from baseline. Also, just because certain offenders have higher baseline levels than others, it does not  mean this influenced their earlier aggressive behavior. Perhaps they have increased levels of testosterone due to incarceration and became more aggressive while in prison.28 2.  Increasing testosterone levels

Many studies have shown that there are higher testosterone levels in aggressive individuals, but none have shown that the testosterone causes the violence, although it is frequently intimated. There can be, and no doubt are, many other factors at work to create aggressive behavior, and the link to testosterone levels may be very indirect, if it exists at all. Clearly not everyone with a high testosterone level is violent; in fact, the great majority are not, and they frequently live very successful lives. This would support the idea that the true relationship may not be between testosterone and aggression but via a mediator such as dominance or competition. However, a simple experiment to determine the effects of increased testosterone would be to increase testosterone artificially in normal people and observe the effects. A limit in human studies is that levels of administered testosterone cannot be too high, as they could pose a health risk, so test ranges are small. Animal studies have shown that administering testosterone increases an animal’s dependency on reward and decreases its sensitivity to punishment, which is a situation often seen in psychopaths.42 In a human study using the Iowa gambling task (IGT), which involves rewards and punishments, 12 women were given a dose of testosterone or a placebo sublingually (under the tongue).

138  The chemistry of the body

Women who received testosterone made riskier and poorer decisions than women given the placebo, showing a decreased response to punishment and an increased desire for reward, which may suggest that testosterone levels may be part of the mechanism involved in psychopathy.42 There have been an increasing number of studies that have shown that administering a single dose of exogenous testosterone affects several behavioral and physiological parameters that relate to aggression.32 For example, in studies of men, a single pharmacological challenge with testosterone rapidly increased activity in the amygdala, hypothalamus, and periaqueductal gray regions of the brain,43 which are regions that play a large role in developing aggression in response to threat or provocation.32 Several studies using a single dose of testosterone have shown increases in brain activity related to perceived threats in both healthy men43 and women and an increased belief by the men in their own dominance and masculinity, in particular in men with lower baseline testosterone levels.44 Dominance also plays a role in moderating testosterone in women, as a single dose of testosterone increased competitive motivation after a win only in women with high trait dominance.32 A double-blind placebo study of healthy young males involved a sequence of four blood assays, the first drawn before administration of injectable testosterone or placebo and the others following certain tests. The study indicated that testosterone alone did not result in aggressive behavior directly but the expression of aggression depended on the person’s own dominance and self-­ control.33 In other words, men with low natural dominance or high self-control were unaffected by the administration of testosterone, but aggression was increased in men who already had traits of high dominance or low self-control.33 This experiment builds on many other recent studies that suggest the link between testosterone and aggression is not clear cut and is moderated by individual traits. The  authors suggest that testosterone may activate aggressive behavior in dominant men through increasing amygdala reactivity in the brain to social messages of threats. Moreover, low self-control is a known potentiating agent for risky behavior, as it involves less frontal lobe inhibition in the orbitofrontal region of the brain, which is frequently linked to impulsive and aggressive behavior (see Chapter 10). Low self-control is frequently seen in patients with high risk for antisocial behavior such as conduct disorder, intermittent explosive disorder, borderline personality disorder, and antisocial personality disorder.33 Many studies show that the impact of testosterone is moderated by personality and socio-­ cultural differences between individuals. The  above study shows the relationship between trait levels of dominance and increased aggression in relation to acute testosterone levels, and other studies show that it impacts mating strategies as well. In women, trait dominance impacts later decision-making after competitive success.45 Introvert women with low sensation-seeking behaviors exhibited lower testosterone levels when in a committed relationship, but this was not seen in extrovert or ­sensation-seeking women.45 In  another study, men with high levels of baseline grandiose narcissism (who believe they are superior to others) were compared with those with ­v ulnerable narcissism (who are introvert and more modest). Both forms of narcissists self-reported higher rates of reactive and proactive aggression, but only grandiose narcissists exhibited increased ­testosterone levels and aggression in tests.46 Finally, in competition studies, personal differences in trait anxiety levels moderated the effect of increased testosterone during competitions. Increased testosterone levels increased subsequent aggression only in individuals who already had low levels of baseline anxiety.47 All of this suggests that effects of testosterone levels are moderated by a person’s individual makeup, as trait levels of dominance, aggression, and anxiety, and many other factors, vary between people. This leads us to the notion of self-construal, which is a “culturally-relevant difference in how individuals define the self in relation to others,” or how they think of themselves as “independent from others or interdependent with them.”45(p117–118) It has been suggested that individuals with higher levels of interdependence with others will avoid aggression in order to avoid damaging social relationships, and some research has supported this. This would suggest that basal levels of testosterone are low in humans because interdependents suppress aggression when testosterone is increased. Many studies have suggested that acute changes in testosterone are a more

Testosterone  139

robust predictor of reactive aggression than basal levels, in which case, again, more-interdependent individuals with increased testosterone may avoid aggression to maintain social connections.45 Therefore, two experiments were designed to test whether the association between basal testosterone and aggression was different between interdependents and independents—in other words, whether self-construal moderated the relationship between acute testosterone fluctuations and aggression.45 Salivary testosterone levels were taken at a number of times during which men played boxing games against a computer opponent, that were manipulated to determine which participants won or lost. Participants were given questionnaires in between bouts to gauge difficulty of the game, and they also completed a variety of other tasks. The results showed that basal testosterone was not associated with aggression in independents, but there was a negative interaction with interdependents. This  was not  statistically significant, but the pattern suggested that there was a positive trend in relationship between testosterone and aggression in independents but not in interdependents. The  second experiment was much larger and included both men and women. Although no direct effect was seen, an integrated analysis of both studies showed that acute testosterone increases were positively associated with increased aggression and risk-taking in men with more independent self-construals. In contrast, basal levels of testosterone were negatively associated with men with more interdependent self-construals.45 The authors suggest that the way men self-identify in relation to others moderates the influence of testosterone on aggression and might explain some of the variance seen in testosterone studies.45 This supports Carré and colleagues’ study, mentioned earlier, showing that the link between testosterone and aggression appears to be moderated by individual traits.33 This is very interesting and paints a much bigger picture, one in which various personality factors moderate the effects of testosterone, and these factors are themselves under both biological and environmental control, showing an overall interaction. 3.  Decreasing testosterone levels

Removal of parts of the body that relate to the production of testosterone has been used to curb behavior in animals for centuries. The most obvious method of removing testosterone is to remove the testes, and castration, of men or animals, has a long history. However, although the testes actually produce testosterone, it is the brain that controls its production. The part of the brain that is responsible for controlling the production of testosterone is the amygdala, and past attempts to reduce aggression involved trying to remove or destroy that part of the brain. As discussed in a later chapter, brain damage of any sort is liable to influence behavior, and no part of the brain is responsible for only one action; thus, such surgery was highly speculative and carried high risks. Amygdalectomy surgeries did achieve a certain popularity in the mid-1960s and 1970s, with predictably inconsistent results.3 Not surprisingly, this has fallen out of favor but what may be surprising to you is that castration has not. Surgical—or more commonly, chemical—castration is still alive and well and is used in many countries, including the United States. In Canada, castration cannot be enforced on offenders as punishment but can be included as a condition of parole. In Europe, it is practiced in Germany and the Czech Republic and is available in Denmark, Finland, Norway, Sweden, and Poland. Korea also recently enacted castration laws. There has been interest in enacting the use of castration in Australia, New Zealand, and Russia, for child molesters in particular.48 The aim of either chemical or surgical castration is to reduce the offender’s testosterone to prepubescent levels in order to prevent sexually deviant fantasies, desires, and sexual urges, with an overall intent of preventing recidivism.49 Surgical castration involves the permanent removal of the testes and was the most common method in the past, usually as punishment. It  is still used today in the traditional neutering of pets to prevent overpopulation. In some countries, such as the Czech Republic and Germany, a testicular pulpectomy is performed, in which part of the testes are removed, without obvious disfigurement, and testosterone production is greatly reduced but not eliminated.50 Such castration is, clearly, irreversible.

140  The chemistry of the body

Chemical castration is much more commonly used today, as it is much less drastic, reversible, and does not sterilize the man except while he is undergoing treatment. Chemical castration involves ongoing and regular administration of anti-androgen drugs that reduce sex drive, ability to be sexually aroused, and sexual fantasies. Several types of chemicals can be used to chemically castrate a man51: ■■

■■

■■

Cyproterone acetate (CPA or Androcur)—administered as a daily or weekly oral pill or as an intramuscular injection. It is used in about 40% of forensic psychiatric facilities in Germany and about 38% of treatment facilities in Canada. It is not licensed in the United States because animal studies suggest that it can cause liver cancer.49 Medroxyprogesterone acetate (MPA  or Depo-Provera)—administered as a daily oral pill or a weekly intramuscular injection. It is used primarily in the United States, where it is used in 17% of treatment programs. It was originally developed as a form of female birth control. Interestingly, its use in Europe was stopped due to its severe side effects.49 Analogues of gonadotropin-releasing hormone (GnRH)—administered as a monthly or trimonthly intramuscular injection. It  is used to treat sex offenders in about 53% of forensic psychiatric facilities in Germany, about 14% in the United States, and about 59% in Canada. The GnRH analogues are considered the most potent; they reduce testosterone to surgical castration levels.51 They are more expensive than the previous options. They are medically used to treat some forms of prostate cancer and have fewer side effects than CPA or MPA, although reductions in bone density may still occur.49

These drugs are not without side effects. All reduce testosterone levels, and because testosterone is a critical hormone in many functions, its loss can cause many health issues. Anti-androgen drugs are frequently used medically in conditions such as prostate cancer, so there are many studies available on the side effects. Anti-androgen drugs reduce bone density and lean body mass, both of which increase the risk of bone fractures. Other side effects include femininization, loss of body hair, and breast development.50 Anti-androgen drugs have also been associated with the development of diabetes and increased risk for metabolic syndrome and cardiovascular disease.51 They cause andropause, the male equivalent of menopause, which results in hot flashes and other unpleasant effects; it can cause depression and, perhaps most concerning, severe mood instability.49 Aside from chemical castration’s side effects, an obvious concern is that the offender, once released, may stop taking the medication, in which case normal libido will return in a few weeks. In fact, a study in South Korea showed that cessation of anti-androgenic drugs resulted in a very dramatic upsurge of testosterone levels, with similarly dramatic increases in sex drive, which offenders said included “uncontrollable” sexual intensities and fantasies.52(p566) This was a totally unexpected result for the researchers, but it is critically important because it means that if a person is non-compliant, he may be at an even higher risk for recidivism than without any treatment at all. Also, many jurisdictions include chemical castration as a condition of parole, so it is vital to take the risk of this upsurge into consideration when the individual completes his parole period and is no longer required to take the drugs, as even an individual who has been perfectly behaved during his parole period may suffer a major setback after drug cessation. Another concern is that testosterone is a valid medical treatment in prostate cancer, types of hair loss, and sexual problems such as erectile dysfunction, or is given to older men with reduced testosterone levels. Hence, it is widely available, and it might be possible for a chemically castrated individual to obtain testosterone to reverse the effects. Possibly, mandating frequent tests of salivary testosterone during parole may be of value.53 Surgical castration, on the other hand, is permanent, and no further action is required. Hence, several countries continue to use surgical castration, including Germany, the Czech Republic, and the United States, although most have replaced surgical castration with chemical castration. In some countries and states, castration is mandatory and part of punishment, usually for the sexual assault of a minor or, in some states, for second sexual offences; or, it may be offered in exchange

Testosterone  141

for a reduced prison sentence. In seven US states, the laws allow the government to keep sexual offenders incarcerated indefinitely after their sentence expires. In order to avoid lifetime incarceration, 15 repeat offenders in California requested surgical castration in the first part of 2018 alone.53 California is one of several states that require mandatory chemical or surgical castration for repeat offenders before they can ever be released. Ethics of castration  In many countries, sex offenders are given the option of chemical castration versus incarceration for many years, indefinite incarceration, or even the death sentence. As chemical castration is a medical treatment, there has been a great deal of concern about medicalizing justice and allowing judges to make medical decisions. There are medical and psychological side effects to chemical castration. Moreover, there are other long-standing ethical concerns. Even when castration is offered as an option rather than as a mandatory requirement, is consent truly willingly given? First of all, can an incarcerated individual actually give free consent? Anything he agrees to or disagrees with could conceivably be used for or against him in parole or other decisions. This is a long-standing concern with any experiments involving incarcerated individuals. Second, when the alternative is many years in prison or even capital punishment, can it truly be said that the offender had a choice? In  2009 and 2012, the European Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (CPT) visited the Czech Republic and Germany. They submitted reports stating that surgical castration was degrading and demanded that both countries end its use.50 Both countries rebutted this demand, citing the efficacy of castration and arguing that such surgery was only ever provided when requested by the offender with informed consent. However, as the alternative to the surgery was prolonged incarceration, the CPT argued that the informed consent could only be coercive. The CPT stated that they had many concerns about consent. The first was that there were few options for treatment other than surgical castration in the Czech Republic, although there did seem to be more in Germany; cost of chemical castration was considered prohibitive in the Czech Republic. Second, they said men told them in interviews that they had not been given adequate information on the side effects, such as osteoporosis. Third, the consent was not true consent because their only other option was long-term incarceration. In a review of these reports, and a discussion on the issue of consent, McMillan of the Bioethics Centre in Dunedin, New Zealand, considered many aspects of the issue but felt that, overall, the decision to allow a prisoner to consent to such surgery was valid as long as threats and coercion were not involved and that such surgery may be actually beneficial in helping a person begin his life again.50 He provided an example of a true case, Mr. K, who had a very long history of child sexual abuse, including abusing his own children. He  had spent his life in and out of prison for such offences, and he had decided that he wanted to be physically castrated before his release, so that he could start to live normally. He was offered chemical castration, but he turned it down in favor of surgical castration, as he wanted a permanent solution. In such a case, McMillan argued, this may be a major step forward in allowing a person to win back his life.50 In some situations, especially those in which the offender has sexually assaulted his own children, castration may be the only option for him to have any form of access to his children. In some cases, there may be less of an ethical dilemma, as some men have asked for castration because they are genuinely concerned that they will recidivate, especially if it may be against their own children. When offenders are given the option of either castration or further—possibly lifetime—incarceration, it seems clear that any sort of consent would be coerced. It is not considered acceptable to enforce medical interventions on competent adults, because it threatens their autonomy or control over their own lives,49 although incarceration alone obviously takes away an offender’s autonomy. Therefore, it has been argued that even if a truly valid consent can never be given, enforcing castration may actually increase an offender’s autonomy by increasing his own control of his life.49 A hypothetical example of a pedophile, Jeremy, was presented in a paper in the journal Bioethical Inquiry. Jeremy is presented as a 55-year-old repeat sexual predator of children, presently serving one of many periods of incarceration. He knows that he will reoffend once released, as he always

142  The chemistry of the body

does, but despises this and is remorseful for his crimes. He has constant and vivid sexual fantasies and cannot stop thinking of deviant sexual acts for more than a few minutes at a time. He desperately wants to be rid of these fantasies so that he can stop reoffending and begin to develop normal, non-sexual, interests. He requests chemical castration; his deviant sexual fantasies are greatly reduced, and he finds that he can now repress them easily.49 Although this case is hypothetical, it was closely based on a true case.49 The authors argued that allowing Jeremy to undergo castration actually increased his autonomy rather than decreasing it; therefore, they considered that despite the impossibility of obtaining true voluntary consent, castration is ethically valid.49 Public attitude  Public outrage at sexual offenders, particularly aimed at those who rape children, frequently calls for castration as punishment and to prevent recidivism. It is certainly punishment, not just in the clear manner intended but also in an evolutionary context, in that a castrated man can neither produce offspring nor contribute to the next generation, so it can also be considered a form of eugenics. In fact, the use of castration in many countries in the first half or more of the twentieth century was to prevent “undesirables” from reproducing and had little or nothing to do with criminal behavior, never mind sexual offences. Throughout the history of castration in the Netherlands (which continued until 1968, with no formal legal framework to support it), there was never a clear statement on whether the surgical castration was to curb sexual assaults or to sterilize offenders because of the belief that sexual crime was hereditary.54 In general, the public seem to support the concept. A study in France on the acceptability of chemical castration presented to physicians and members of the public a variety of sexual assault scenarios, changing the age of victim, age of offender, psychiatric status of offender, and offender’s family’s attitude to castration. In all scenarios, a high percentage of participants considered that castration was fully justified.55 Moreover, chemical castration seems to most people to be much less drastic than surgical castration and, if effective, to result in releasing a person to hopefully live a normal and productive life without the damaging impacts of deviant sexual fantasies and desires. Efficacy of preventing sexual crime recidivism  Perhaps the most important question that

needs to be asked is, does it work? Does castration prevent or reduce recidivism? A number of studies have been conducted on the efficacy of castration to reduce sexual offending recidivism. Robust studies are hard, as you cannot give a placebo to dangerous offenders and release them, so obtaining objective controls is difficult. Nevertheless, several studies have been conducted. Older studies from the 1960s and 1970s, primarily from Germany, found a dramatic decrease in sexual recidivism in castrated individuals (reviewed in Khan and Mashru51). In  one study of over 1000 castrated males 20  years after release, recidivism was only 2.6%, in comparison with  39.1% in a matched sample of non-castrated sex offenders. 51 Recidivism was highest in males castrated in their 20s. In another study, recidivism rates after 10 years were 7.4% and 52% in ­castrated and non-castrated sexual offenders, respectively.51 However, these studies were conducted before randomized control trials (RCT) were introduced, and some even used unethical data from the Nazi era.51 In a review of RCT studies using chemical castration methods, there were some promising results but no strong evidence to support such treatments. 51 Most studies had major methodological flaws. Six studies looked at testosterone-suppressing drugs but only two considered recidivism. One compared MPA  and imaginal desensitization with imaginal desensitization alone. Imaginal desensitization is a relaxation technique that uses images to help people with a number of disorders, such as explosive aggression, sexual paraphilia, kleptomania, pathological gambling, and compulsive eating. 56 The sample size was small, and no ­recidivism was observed in the treatment group, with only one recidivist in the desensitization-alone group.51 In the second study, offenders treated with MPA were compared with offenders treated psychologically as well as with MPA and offenders treated only psychologically. Recidivism was 20% in the combined treatment and 50% in the psychological treatment alone, but dropout rate was high in those treated with MPA alone. 51 Frequency of deviant sexual fantasies appeared to be reduced but not the deviancy type.51

Testosterone  143

In a search of more than 3000 published and unpublished reports on sex-offender treatments, 29 studies were found to be comparable for meta-analysis. These included almost 5000 treated and almost 5500 untreated sex offenders.57 The results showed a large effect size for surgical castration, a fairly strong effect size for chemical castration, and a moderate effect size for cognitive-behavioral treatment.57 Several studies indicated that chemical castration reduced frequency and intensity of sexual thoughts and fantasies, which in turn improved the efficacy of psychotherapy,51 an important point. A study from Germany compared behavior in men before and after voluntary surgical castration,58 although again, the concept of voluntary is debatable. The men were surgically castrated and then followed up after release to determine recidivism. In this study, 99 castrated sex offenders and 35 non-castrated sex offenders were followed for, on average, 11 years after release. They accounted for about 25% of all castrations in the period from 1970 to 1980, so they are reasonably representative of this population. Obviously, you cannot randomly assign the men to the experimental or control groups, but the controls were sound: all 35 had originally requested castration and then changed their minds, so they were about as close to a control group as could be achieved ethically in such a situation. The control group was matched for previous criminal record, age, intelligence, social background, and marital status. The recidivism rate over an average of 11 years post-release was 3% in castrated offenders, compared with 46% in non-castrated offenders.58 It has been argued that the control group may have represented people unwilling to stop their sexual offending, and hence they backed out of castration, whereas the castration group may have felt more strongly about changing their behavior.50 Moreover, men in castration studies may represent a sub-population of sex offenders, as they are willing to sacrifice libido for freedom, so they may not be representative of the majority of sex offenders. Nonetheless, it is a powerful result. In a follow-up, 70% of those castrated were satisfied with their choice, 20% were not, and 10% were not sure.58 In a metaanalysis, treated offenders showed 37% less recidivism than controls, although treatment included both chemical and surgical castration, so the relative merits of each cannot be determined. Similar reductions were also seen in violent and general crime. Castration was much more effective than psychosocial interventions alone.59 In  a recent study in South Korea, after 3 months of chemical castration offenders reported a 71%–76% reduction in the intensity and frequency of sexual thoughts and masturbation.52 As the drugs had little effect on roughly a quarter of the sample after 3 months, it was recommended that offenders remain in custody for this period of time to be assessed for efficacy before being deemed safe to release.52 A common goal of sex-offender treatment programs is to change the offender’s focus from criminal sexual acts to normal, healthy sexual relations with an appropriate consenting partner. One of the criticisms of castration is that any such goal is no longer attainable, as the individual is no longer capable of any sexual act, deviant or otherwise.48 However, in a review of studies, in some cases, surgically castrated men were still able to function sexually.60 In one study, 35% of castrated men maintained sexual potency and drive, 50% exhibited full erections while visually stimulated, another 18% retained some sexual functioning 11 years after castration, and 21% maintained sexual relations with their partner.60 Although this questions the efficacy of the treatment and raises concerns about recidivism, if castration can reduce sexual fantasies and deviant urges enough to allow offenders to participate successfully in psychotherapy treatment programs and yet can still allow some sexual function, then the goal has been achieved. So now, we go back to the most important question in all this: does it work? In other words, does castrating a man prevent further sexual criminal recidivism? The results show that it appears to do so in some cases. We all know that sexual assault is not just about sex but also involves power, and impotent men can still sexually assault a victim by using inanimate objects such as guns or knives to penetrate. In the infamous 2012 rape case in Delhi, India, in which a young female paramedical student was gang raped on a bus and tortured and later died from her injuries, iron rods similar to tire irons were involved.61 When a sex offender is motivated solely by sex, castration may be effective, but many sexual assaults are related to other crimes, such as psychopathy or substance abuse, in which

144  The chemistry of the body

case castration may have little effect. There are several types of sex offenders, and some studies have shown that castration is much more effective against some types over others. For example, paraphiliacs are a specific group of sexual offenders who feel compelled to commit deviant sexual crimes in order to act out specific fantasies.62 Paraphilia includes (but is not limited to) pedophilia, fetishism, voyeurism, frotteurism, exhibitionism, sadomasochism, and sexual sadism, but not all paraphiliacs commit crimes.60 Studies have shown that castration is most effective against paraphiliacs over other types of sex offenders, but most countries do not distinguish between them when considering castration.61 Castration has been shown to be ineffective in sex offenders who do not suffer from paraphilias, such as antisocial or psychopathic sex offenders.49 After the Delhi gang-rape case mentioned above, there was tremendous public outrage and worldwide denunciation from the United Nations and UNICEF, with a demand for the castration of rapists. During this time, Dr. Aniruddha Malpani, the medical director of the Malpani Infertility Clinic and the Health Education Library for People in Mumbai, said that chemical castration may work for repeat sexual offenders, serial rapists, and pedophiles but that “most rapes are one-off assaults, done by men who have the opportunity to take advantage of a woman perceived to be helpless and defenseless by the offender.”63(pf64) This was exactly the situation in this particular case, as the woman had accidently gotten on the wrong bus: instead of a regular bus, it was a party bus, illegally taken out by a group of drunk young men. Malpani went on to say that while chemical castration made “great sound bites … by diverting energies to explore these, one might squander the chance to find more effective solutions.”63(pf64) He finished eloquently, saying, “the solutions are not likely to be medical, because the problem is largely non-medical.”63(pf64) In a discussion of the case, it was stated that castration in India would be considered a form of torture under the Geneva Declaration and the Universal Declaration of Human Rights. Ironically, instead, all men were sentenced to death, with the exception of the sole youth among them.63 Overall, castration does appear to greatly reduce sexual violence recidivism in certain types of sexual offenders, and GnRH has been stated to “probably constitute the most promising treatment for sex offenders at high risk of sexual violence, such as pedophiles and serial rapists.”60(p643) However, the studies show that there is a great range in efficacy in recidivism reduction, with only some sex offenders improving. Is castration truly a good method to ensure that a person does not reoffend? Remember that, in many cases, castration is offered in lieu of lengthy prison sentences, meaning that sex offenders are being released back into society in the belief that they are no longer a risk, due to castration, with no guarantee that they will not immediately recidivate. However, studies do show that castration can increase the success of psychotherapy treatment, as the offender is better able to concentrate, so a combination approach is usually recommended.60 Overall effect of castration on criminal behavior  So overall, what do studies on castration, or decreasing testosterone, tell us about the role of testosterone in crime? They appear to show, as we knew already, that testosterone is deeply embedded in sexuality rather than necessarily in sexual offending. Removing or greatly lowering testosterone reduces some forms of sexual offending but has little effect on other forms. Clearly, the role of testosterone in offending is not direct. Postpubertal testosterone and risk-taking behavior

Perhaps testosterone’s putative link with crime relates more to risk-taking than actual crime. Several studies have linked higher levels of circulating testosterone with risky behavior rather than directly with delinquent or criminal behavior. In a large study of men’s health, men with higher levels of testosterone engaged in more health-related risks, such as multiple sex partners, resulting in higher risk of sexually transmitted diseases, smoking, and alcohol and drug use.64 Testosterone does have health benefits also, but this benefit was lost in men with very high levels of testosterone.64 Financial risk-taking has also been shown to be higher in men with high testosterone levels. In a study of London men actively engaged in trading on a trading floor, levels of salivary testosterone taken in the morning predicted the traders’ success that day.2

Testosterone  145

In a study using the IGT, both men and women with higher baseline salivary testosterone ­levels took greater economic risks than those with lower levels.65 In addition, men and women with low baseline testosterone levels learned to avoid financial losses over the course of the game, but high-­testosterone individuals did not. Overall, the study showed that high-testosterone individuals were more likely to take monetary risks and did not learn from their mistakes. The authors suggested that such studies may not only provide information on future investment risks but also potential for p ­ roblem ­gambling.65 A follow-up study using the same IGT test found that the relationship between endogenous testo­ sterone and risk-taking was not linear but rather U-shaped. Thus, those with low and high testosterone levels were more willing to take financial risks than those with intermediate levels.66

Testosterone, competition, and dominance As we have seen, there has been a tremendous amount of discourse on the relationship, putative or otherwise, between testosterone levels and criminal behavior, in particular violence. Most studies were originally designed to determine whether testosterone regulated aggression, and although there is clear evidence that this is the case in animals,5 the data are not  quite so consistent in humans. Aggression in animals and early humans would have been evolutionarily beneficial, especially in male animals, as it allows them to increase and maintain social status and resources, attracting mates and thereby maximizing fitness. In the modern human world, aggression is less desirable. Therefore, it has been suggested that testosterone may relate more to dominance than aggression in modern humans, as dominance relates to gaining assets or status rather than directly injuring another person.40 Theoretical background

Ellis and colleagues proposed an evolutionary neuroandrogenic (ENA) theory, which argues that males commit higher levels of crime than females because females have evolved a preference for mates who appear to be strong resource providers, and this in turn has made males evolve genetic mechanisms to increase their levels of testosterone and other hormones to increase their competitiveness, often at the expense of hurting others.67 They state that this evolved because women who chose stable, long-term resource providers had more successful offspring, which increased their success. This female bias impacted male evolution by selecting for certain traits that increased the males’ ability to accrue resources and hence gain mates and so produce more offspring. Thus, high levels of competitiveness in males were favored in the past by natural selection, but in the modern day, the highest levels are controlled by criminal justice systems, which curtail competition when it includes victimization.67 Ellis and colleagues suggested that, in adolescence, the forms of competition are fairly rudimentary, but with learning, this changes to more sophisticated forms in adulthood. The rudimentary types of competition are those more associated with antisocial behavior.67 Maturity and learning allow males to develop more legal and accepted forms of competitive behavior. The ENA theory argues that testosterone and other androgens have evolved to organize the brain so that the male is highly motivated for competition in order to acquire resources, even at the risk of harming others.67 The  hormonal impact on the brain occurs first prenatally, when the brain is configured by hormone levels and sensitized to react to later hormonal exposure, and then at puberty, when the brain is activated by the hormones.67 Males with excellent cognitive skills and learning abilities will mature their competitive skills, but those with poor learning skills or reduced cognition may be less able to refine their ability to compete in a legal manner or even to differentiate between legal and non-legal forms of competition. The ENA theory also explains why males, who are exposed prenatally to much higher levels of testosterone than females, are much more criminal and violent.1 The ENA theory states that levels of testosterone and other androgens need to be very high in both exposure stages—prenatal and postpuberty—to maximize risk for highly competitive and victimizing behaviors.1

146  The chemistry of the body

In 1985, Mazur first proposed a biosocial model of status (BMS), originally in primates.68 Mazur predicted that testosterone levels during a competition will vary depending on performance, with higher levels of testosterone in winners. Mazur predicted that from an evolutionary perspective, variations in testosterone level in competitions will moderate future behaviors involving competition and dominance, as winners are likely to face future challenges to their social status. Therefore, higher levels of testosterone would increase aggression and therefore ability to achieve and maintain higher status.68 Losers, on the other hand, would have decreased testosterone when losing, which may increase submissive—and consequently, safer—behavior. Many studies have looked at this relationship over the last 25 years. Competition studies

In  a review, Carré and Archer discussed the increasing number of studies showing that acute changes in testosterone level related more closely to competition or social provocation than directly to aggression.32 It is now thought that testosterone is actually a “competition hormone” that moderates competitive behavior, owing to its plasticity during winning and losing.69(p924) It is believed that testosterone responds to social environments adaptively to coordinate a behavioral response. This approach suggests that testosterone acts as a competition hormone by activating a range of behavioral and cognitive adaptations—such as competitiveness, aggression, risk-taking, perseverance, and displays of status—to maximize fitness.69 In a review of competition studies, 25 showed higher testosterone levels in winners over losers, although a further 31 showed no difference between the two and 5 showed that losers had higher levels than winners.31 Overall, studies show that acute changes in testosterone levels related to competition do impact present and future social behavior and interactions, but this is not limited to winners.31 This is somewhat different from Mazur’s BMS model, which predicted that only winners would have an increase in testosterone levels, to promote dominance behavior, and losers would have lower testosterone levels, to promote submissive behavior.68 Although many studies do show an increase in testosterone levels in winners, consistent with BMS theory, increased testosterone in both winners and losers is positively associated with increased competitiveness and aggression, and decreased testosterone is positively associated with reduced competitiveness and aggression, again in both winners and losers, which does not fit BMS and suggests that although competition does impact testosterone levels, the results may be similar in both winners and losers.31 In one study based on competitive tasks, testosterone levels decreased significantly in winners and increased very slightly in losers, which contradicts the BMS theory.70 However, this was moderated by personalized power. Studies have shown that power motivation also regulates testosterone levels and behavior. Implicit power motivation is related to the emotional gratification that an individual feels in relation to positive behaviors, such as doing a good deed; negative behaviors, such as hurting someone; or behaviors that maintain or increase status.70 A more specific power is that of personalized power (p Power), which is the need to persuade others of your own ideas, for your own purposes, as opposed to socialized power (s Power), which involves prosocial or altruistic acts to influence others.70 As p Power involves dominating others, it is considered egotistical and more related to winning in competition. It is proposed that men with high p Power experience a much greater feeling of dominance when they win and an even greater response to losing because of the value they place on the potential effect of such a loss on others. In this study, although, overall, testosterone levels were lower in winners, contrary to BMS, p Power moderated testosterone, with the decreases in testosterone only seen in male winners with low p Power and not in men with high or moderate p Power. Type of aggression was also analyzed. Both winners and losers exhibited more proactive aggression in relation to increasing p Power, although winners had a higher increase. In contrast, although both winners and losers showed an increase in reactive aggression, with losers showing a sharper increase, testosterone levels were not linked to either form of aggression.70 In the second part of this study, the impact of testosterone levels after competition on emotion recognition was considered. Earlier studies have shown that increasing people’s power, through such things as role playing, increased their ability to recognize other people’s emotions. In this study,

Testosterone  147

as p Power increased, winning improved an individual’s ability to correctly gauge other people’s emotions. Emotion recognition is extremely important in social interactions, as it is crucial in all aspects of relationships, such as negotiations, leadership, marriages, and social adjustment.70 In  a 2017 meta-analysis of studies on human competition and testosterone, Geniole and colleagues analyzed 62 studies over 35 years, considering sample size, type of testing (physical exertion or mental test), sex, and location of test (laboratory or field).71 The combined studies involved over 2500 participants and confirmed that winners of a competition exhibited a greater increase in testosterone levels in comparison with losers, with an effect size (D) of 0.20 that was similar in both men (D  =  0.23) and women (D  =  0.22).71 However, the results were quite varied, being much more robust in field studies, such as those conducted in actual sports events (D = 0.43), in comparison with those conducted in laboratory settings, where the effect sizes were weak or trivial (D = 0.8 overall, D = 0.15 in men and D = 0.04 in women). It was suggested that this may be due to heightened investment in the competition in true sports events, which were more likely to involve physical rather than cerebral competition. Also, they commonly involved an audience, and some studies have shown that the number of male or female spectators impacted competitors’ testosterone levels, with men exhibiting higher testosterone levels in front of largely female spectators and females exhibiting higher levels in front of largely male spectators.71 Effect sizes were also higher when baseline testosterone levels were tested more than 10 minutes before competition, suggesting that testosterone begins to increase in anticipation of the event, so some studies may not be obtaining a true baseline with which to compare.71 Overall, the meta-analysis supported the competitive role of testosterone, but the large heterogeneity between studies and relatively small effect sizes in many indicate that there are many other factors involved. Several studies in humans have shown that increased testosterone is not only positively linked to competitive aggression but also to a desire to compete again, and athletes receiving encouragement from coaches or watching motivational and aggressive video clips before competing had increased testosterone levels, improved athletic performance, and increased physical strength.31 In animal studies, increased testosterone levels after a win continued to increase aggression more than 24 hours later, and it was suggested that the earlier acute testosterone levels may enforce behaviors associated with winning, such as aggression.31 In simple tests in humans, increased testosterone during a puzzle forecast improved performance in the same task 24 hours later, also showing that the brief, acute increase in testosterone could still modulate behavior long after the acute increase occurred. Interestingly, the relationship was seen in winners and losers.31 A number of studies have shown that increases in testosterone during a variety of competitive interactions often forecast later heightened aggression, and effect sizes have been moderate to large, indicating that the impact of the brief, acute change in testosterone levels had a continued effect.31 Similar results are found in many social species, in that competitive behavior is moderated by previous experience. Those that win continue to compete and win, whereas those that lose no longer compete and defer to winners. This supports the BMS theory. The studies above mostly examined dominance-type competitions, in which status is achieved by domination, fear, and intimidation, but a recent study looked at a different type of victory, that of prestige, or victory based on respect, due to skills, knowledge, and success rather than fear.69 A large-scale community of semi-professional musicians in a university marching band was examined. The group was highly cohesive, meeting and interacting regularly, and was highly respected, regularly winning numerous awards and accolades. Participants were asked to nominate who they most respected musically and who they would go to for advice, together with submitting saliva for testosterone assay, before and after band rehearsal. The results mirrored those in dominance studies, in that men who had achieved prestige in early stages of group cohesion showed a rise in testosterone levels over the following 8 weeks, whereas men with low prestige within the group exhibited a decrease in testosterone or no change. No changes were seen in women.69 Although studies in both men and women have shown that testosterone levels change acutely during competition, the overall response to this acute fluctuation varies between the sexes, with several studies showing that although increased testosterone in men during competition increased aggression, the same was not seen in women,31 although changes in behavior were still noted.

148  The chemistry of the body

In a study of recreational (as opposed to elite) female athletes tested before, during, and after competition, significant increases in testosterone and another hormone, cortisol, were found throughout the competition, and levels in both winners and losers related to their own perceptions of their success rather than their actual success. This may have been an accurate measure of performance or may have been due to an increased personal belief in themselves due to elevated testosterone.72 Interestingly, in one study of female competitors, changes in testosterone levels predicted increased prosocial behaviors rather than competitive aggression. Salivary testosterone was measured before, during, and after competitions and was found to increase during competition and then decrease within 30  minutes. However, whether they won or lost, women with testosterone levels that stayed higher for longer were more motivated to reconcile with their opponents than those whose levels dropped more rapidly.73 The authors suggested that testosterone may moderate behavior after a competition to promote and increase status, using social bonding in women rather than aggression or dominance measures.73

Serotonin and testosterone Almost nothing in the body acts in isolation. Hormones, genes, and neurotransmitters (brain messengers) all interact and influence each other. It has been suggested that the purported relationship between testosterone and aggression may not be direct and may instead be more closely related to other chemicals in the body that are modulated by testosterone. One of the most likely candidates is serotonin. Serotonin is a neurotransmitter in the brain that has been repeatedly linked to aggression and impulsivity (see Chapter 9 for a discussion of serotonin). High testosterone levels are often correlated with low levels of serotonin, which have in turn been strongly linked to impulsivity. In a study of rapists, plasma levels of testosterone, as well as other androgens, were significantly higher than in controls, but levels of 5-hydroxyindoleacetic acid (5-HIAA) (the major metabolite of serotonin) were significantly lower.74 Much of the recent work we have just discussed suggests that testosterone is more directly related to dominance than to aggression per se, and this is supported by the many studies that indicate increased testosterone levels during sporting competitions and games. Although dominance could lead to aggression, it is more related to improving status and is usually non-aggressive. It has been suggested that aggressive men may have a combination of high levels of testosterone and low levels of serotonin, and that aggression may ensue when such a man is frustrated in his attempts to achieve status or dominance. People with low serotonin levels are impulsive and inclined to overrespond to aversive stimuli. It is thus speculated that high testosterone levels would encourage dominance-­seeking ­behavior, which if it led to failure, could result in violence associated with low serotonin levels.75 Results from animal studies have indicated that testosterone may relate to aggressive drive and serotonin to regulating the intensity, threshold, and frequency of the expression of aggression. Therefore, males with high testosterone levels but normal serotonin levels might exhibit aggression in certain situations and be dominant but would be unlikely to be violent. Conversely, males with high testosterone levels but low serotonin levels would have greater impulsivity and a lower threshold for aggression, resulting in more frequent episodes of violence. Moreover, such males would be less able to prevent aggression from escalating into violence and injury.76 In a study of adolescent male rhesus monkeys, testosterone levels alone positively correlated with competitive aggression but not impulsivity; serotonin levels negatively correlated with impulsive behavior and severe unrestrained bouts of violence but not general aggression overall; and males with low serotonin and high testosterone levels exhibited an increased number of incidents and an increased intensity of the aggression.76 This animal study does support an interaction between serotonin and testosterone. Such an interaction is logical and supports the idea that high testosterone alone does not result in violence but merely in competitive aggression. This competitive aggression could relate to attaining resources or a mate. The findings also point to a certain definition of aggression in the animal kingdom: unrestrained, forceful, and highly violent behavior, including biting and beating, but a

Cortisol  149

pattern that is goal directed. Does the same definition hold for humans? One could argue that most successful people are competitively aggressive to a certain extent or they would not be successful. This is not to suggest, however, that such people are aggressive per se, or violent, but merely that competition for status is fairly normal. However, if such a drive is combined with other behavioral conditions such as low neurotransmitter function, then behavioral disorders could result.

Testosterone in prosocial behavior Testosterone has a bad reputation in the public eye, and the dominance of literature looking at criminal behavior, due to the assumed link between testosterone levels and aggression or other forms of criminal behavior, reinforces this view. However, testosterone has myriad important functions in the body, and some of these are highly prosocial. Research has shown that testosterone increases altruistic and extremely prosocial behaviors in people who have socially protective positions in society, such as firefighters and police officers.77 It has been suggested that instead of just moderating aggression, testosterone is an adaptive social hormone that is critical in supporting status-improving behaviors.9 Several studies have looked at the prosocial effects of testosterone. In a study of lying, increasing testosterone in healthy men resulted in a reduction in self-serving lies.78 In  women, administering exogenous testosterone sublingually resulted in a significant increase in fair bargaining behavior, which is highly prosocial, as it reduces conflicts and increases successful social interactions.79 Interestingly, women who believed that they had received testosterone (whether they had or not) behaved much more unfairly than those who believed that they had received the placebo, which the authors suggested shows that the general public (from which the women were chosen) expect a negative outcome when given testosterone.79 Increased social cooperation was seen in individuals given exogenous testosterone but only in those who had low prenatal testosterone levels (as determined by 2D:4D ratio). It was suggested that those with low prenatal testosterone levels would metabolize testosterone more slowly.80

Testosterone overall Overall, testosterone clearly plays a role in our behavior, as well as in many other aspects of our health. The relationship between testosterone and aggression in animals is fairly clear and makes evolutionary sense, whereas the relationship in humans, perhaps once the same as in other animals, appears to have evolved and changed. In modern humans, the relationship between testosterone and behavior appears to be more related to dominance, competition, and social status; these make more sense than overt aggression, as the latter is constrained by our criminal justice system, which effectively minimizes reproductive fitness by incarcerating offenders. However, this does not mean there is no relationship between testosterone and violence, as dominance, competition, and desire for social status can easily motivate violence and crime. Testosterone also has a role in risk-taking, and while this may be harmless, it can also result in harm of the self or others.

Cortisol Cortisol is also a steroid hormone. It is produced by the adrenal gland and is the end-product of the HPA axis. It is often considered a “stress hormone,” being released under conditions of physical or psychological stress or low blood sugar.81 It has many roles in metabolism and the immune system. High cortisol levels are linked to psychological and physical stress, anxiety, and social avoidance, whereas low levels are linked to more prosocial behaviors, social approach, relaxation, feelings of submission, and low levels of aggression.81–83 The HPA axis is involved in adaptation, both physiological and behavioral, to changing situations and environments.84 It is a complex interaction between, as it sounds, the hypothalamus (an important region of the brain which links the neural system to the endocrine or

150  The chemistry of the body

hormonal system), the pituitary gland (found in the brain and involved in the production and regulation of many different hormones), and the adrenal glands (found on the kidneys and also involved in the production and regulation of a number of hormones, including adrenalin as well as steroids).

Cortisol, stress, and abuse The HPA axis is considered to be the body’s stress–response system, so disruption of this system, via neurotransmitters, hormones, or brain pathology, can result in abnormally low cortisol levels. This hormone is also involved in autonomic arousal, which means that those who are low in cortisol do not fear punishment unless it is immediate. Studies in animals have shown that experiencing psychological stress early in life, such as maternal stress or separation, can permanently affect the HPA axis.85 Reduced cortisol levels were found in animal offspring following maternal stress, as well as after repeated maternal separation.85 In humans, there have been a few studies on traumatized children in clinical settings which have confirmed these results. Also, studies on abused children and adolescents and on adult psychiatric patients with a history of childhood abuse have shown long-term changes in the functioning of the HPA axis.85 A number of studies on Romanian children rescued from orphanages with conditions of horrific social isolation and deprivation after Ceauşescu’s rule showed that the children had blunted cortisol diurnal rhythms. Normally, cortisol rises to a peak in the morning and drops to a low in the evening, but these deprived children, who had been kept in small cots well into their teens and had little to no contact with people, exhibited a relatively flat cortisol profile. This  is similar to experimental studies on social deprivation in non-human primates and suggests that a lack of socialization and stimulation in early development can result in permanent damage to the HPA axis and normal functionality (reviewed in van Goozen et al.85). A study of children who were physically or sexually abused in the first 5 years of life compared with children who were not abused showed that the abuse resulted in a diurnal decrease in cortisol and increased depression and internalizing behaviors, suggesting neuroendocrine disruption. This result was specific to early-onset physical or sexual abuse, as opposed to later physical and sexual abuse or to early neglect or emotional abuse.86 Childhood abuse and bullying were studied in relation to cortisol levels in a cohort of 190 ­children aged 12 years, and researchers found that bullied or abused children had significantly lower cortisol responses (that is, blunted responses) to stress than children not bullied or abused, who had a normal rise in cortisol in response to stress. Children with a blunted response also exhibited higher social and behavioral problems than children with normal cortisol responses.84 Some studies have shown that preventative interventions, both family-based and attachmentbased, can regularize HPA  axis function and thus reduce later aggression, so the authors suggested that cortisol measurements be included in intervention strategies.84 This blunted cortisol response to bullying was shown, in a study of monozygotic (MZ) twins in which only one was bullied, to be unrelated to genetics or familial environments, which suggested that the bullying had a direct effect on HPA axis function.87 However, in another study using both dizygotic (DZ) and MZ twins, both genetics and environment were involved. Bullying was associated with depression via both genetic and environmental etiologies.88 Bullying, or peer victimization, often results in increased depressive symptoms and internalizing behaviors but only in some victims. Similarly, it is estimated that 40%–90% of the general population experience a traumatic event, yet only 7%–12% develop post-traumatic stress disorder.89 It is therefore thought that biological factors may increase either resilience or sensitivity to stressors, making some people more likely to be adversely affected by stress than others.88 As cortisol is known to play a critical role in physiological responses to environmental stressors, it has been suggested that baseline cortisol levels may be an indicator of an individual’s sensitivity or resilience to a stressor.88 Studies have linked low cortisol levels with externalizing behaviors in adolescents and high cortisol levels with internalizing behaviors. They have also shown that higher morning cortisol levels, as well as cortisol reaction to stress, can predict increased depression (reviewed in Brendgen et al.88). In order

Cortisol  151

to test whether genetic or environmental factors explained the relationship between cortisol, peer victimization, and depression, cortisol levels were measured when children first woke up (cortisol awakening response, or CAR) over several days in 159 MZ twin pairs and 120 DZ twin pairs, aged 14 years. Peer victimization and depression were assessed by self-reports. The results showed that as CAR increased, the environmental association between bullying and depression consistently increased, and the genetic association also varied, although less strongly.88 Children with a high genetic risk for depression together with increased physiological reactivity were at greater risk of peer victimization. Moreover, children with increased CAR were more likely to suffer from depression and other internalizing behaviors when bullied. The authors suggest that the results support the contention that peer victimization increases internalizing behaviors in youth who have an increased biological sensitivity to adversity and implicates the HPA axis in this reactivity.88 They suggest that treatment interventions that modulate the HPA axis by improving stress regulation would be beneficial in children who are physiologically vulnerable to stress.88 The  effect of childhood abuse and maltreatment in the etiology of criminal behavior is well known in criminology and is usually attributed to the direct social impact of the abuse. Child abuse and neglect are universally recognized as risk factors for later violence, criminal behavior, and delinquency, as well as mental and physical health issues.90 However, a recurrent theme within this text is the link between childhood abuse and the damage that it does to the biology of the child, which may later have behavioral impacts. In a study of 67 adults given endogenous cortisol and assessed for a history of abuse, psychopathy, impulsivity, base-level (trait) and active (state) aggression, abuse, and neglect predicted reduced HPA activity and response and indicated a nonsignificant trend toward base-level and active aggression.90 Including all factors together explained 58% of the variance in trait aggression and 26% of the variance in state aggression.90 This suggests that all these factors—HPA axis activity, child abuse, and psychopathy—interact to increase risk for aggression.90 In a meta-analysis of 27 studies on childhood abuse and cortisol dysregulation, an overall small effect size was seen between reduced CAR and abuse, although this became stronger when only children reported by an official agency as abused were considered, as opposed to those who self-reported abuse.91 Clearly, abuse or other stressors can biologically change the HPA axis, which impacts the stress response. Reducing cortisol activity reduces autonomic arousal, which in turn reduces sensitivity to punishment and ability to think ahead to potential consequences of an act. This in itself can increase risky behavior.

Cortisol and antisocial behavior As lower cortisol levels blunt response to social environment and result in lower autonomic arousal, we might expect to see reduced levels of cortisol in antisocial people, and many studies have demonstrated such a link. For example, one study found that low cortisol levels were associated with both early onset of aggression and persistent aggression. Boys whose levels were low at both times they were measured (2 years apart) had triple the number of aggressive symptoms as boys whose levels were low at only one reading.92 Persistent low levels of cortisol were therefore considered to be a good indicator of aggression, rather than a low level at a single time.92 In another study, over 300 preadolescents were tested for resting salivary cortisol levels and then tested 5 years later for aggression and personality traits.93 Low preadolescent cortisol levels were associated with low self-control, low harm avoidance, and heightened aggressive behavior 5 years later, with low self-control as the main personality moderator. The authors believed that low cortisol levels more directly predicted low self-control, which impacted aggression, rather than the cortisol impacting aggression directly.93 In general, rates of antisocial behavior increase dramatically during adolescence, but for most adolescents, this is just a passing phase. However, a very small fraction of individuals continue to show persistent (life-course) antisocial behavior beyond adolescence.94 It is this small fraction of individuals who are considered to have neurobiological deficits that increase the risk of antisocial behavior.95

152  The chemistry of the body

In normal children, activity increases in the HPA axis during adolescence, usually between the ages of 15 and 17, and this can be assessed by measuring CAR. This increase is related to certain developmental stages in adolescents96 and is part of normal development, but it is not seen in clinical (­hospitalized) samples of people with depression or anxiety. Also, many studies in children and some in adults have shown that antisocial behavior is associated with reduced HPA axis activity, as measured by CAR.96 Results in adolescents, however, have been mixed. As adolescence is normally associated with increased antisocial behavior, and for the majority, this is just a passing phase, it is unlikely that a clear correlation between CAR and antisocial behavior would be seen in adolescents, which would explain why the results are mixed. Two very different forms of antisocial behavior, aggression and rule-breaking, have been shown by research to have very different origins and trajectories. Rule-breaking peaks during adolescence, whereas aggression usually begins in childhood and usually decreases in adolescence, although the most aggressive children frequently remain aggressive throughout life.96 It has been suggested that aggression is more related to neurobiological issues than rule-breaking.96 A study of 390 adolescents in a Dutch cohort were tested three times, at ages 15, 16, and 17 years, for CAR, aggression, and rule-breaking. Adolescents with persistently high aggression exhibited a low CAR consistently over the sampling period in comparison with adolescents with low aggression; however, no differences in CAR were seen between adolescents who persistently broke rules and those who did not.96 This supports the earlier theory that aggression and rule-breaking have different etiologies, with aggression being more probably related to neurobiological deficits.96 It also indicates that further studies, especially on adolescents, need to factor the type of antisocial behavior into the experiment. It is suggested that aggression is more grounded in genetics and biology, whereas rule-breaking is more closely linked to peer relationships, socioeconomic status, and social environment. It also supports the use of CAR as a risk factor for aggression.96

Cortisol and testosterone There is a great deal of inconsistency in the results of studies attempting to find a clear relationship between testosterone and antisocial behavior, risk-taking, and competition, although most show there is some sort of relationship. No part of the body works in isolation, and many hormones and neurotransmitters are highly linked. A  possible explanation for the inconsistency in results has linked testosterone to cortisol. It is believed that testosterone and cortisol act both independently and together to influence antisocial behaviors.97 Baseline levels of endogenous testosterone and cortisol are relatively stable at a given time of day and are believed to relate to an individual’s personal levels of aggression, submission, and dominance. However, as we have discussed above, testosterone is designed to respond acutely to long-term and immediate social conditions. Cortisol is similar in that its levels fluctuate based on social context.81 The dual-hormone hypothesis

Testosterone and cortisol act both independently and with each other. Cortisol can inhibit the activity of the HPG axis, which secretes testosterone, inhibits testosterone activity, and blocks testosterone receptors in the prefrontal cortex region of the brain.83 In turn, testosterone can inhibit the activity of the HPA axis, which secretes cortisol.81 High levels of cortisol have been shown to reduce the levels of circulating testosterone and also block its cellular effects.98 Hence, the dualhormone hypothesis suggests that the impact of testosterone on human aggression is moderated by cortisol, so that testosterone is linked to aggression only when cortisol levels are low.99 This was first discovered in two notable studies. First, a study of late adolescent male offenders (aged 17–18 years) showed that those with high testosterone levels had committed more violent offences, were less likely to receive parole, and had higher levels of within-prison infractions than those with low testosterone, a result similar to many studies on offenders and testosterone. No direct effects of baseline cortisol were seen.100 More interestingly, a significant interaction was seen between testosterone and cortisol levels, with the link between high testosterone levels and

Cortisol  153

violence seen only in participants with low cortisol.100 In a subsequent study of delinquent boys (average age 13.7 years), cortisol was also found to moderate testosterone. A significant positive relationship was found between overt aggression and high levels of testosterone, but only in individuals with low cortisol levels. In participants with high cortisol levels, no relationship was seen.99 Mehta and Josephs further examined these findings in dual studies on leadership and competition.83 Leadership is extremely important in status and dominance but has been little studied. In their first study, both males and females were videotaped in a position of leadership, and then, the videos were judged for levels of dominance and compared with salivary assays for testosterone and cortisol. Testosterone levels were correlated with dominance only in men and women with low cortisol levels. In individuals with high cortisol levels, there was no relationship between testosterone and dominance, and the authors concluded the link had been blocked by the high cortisol.83 In the second study, only men were studied in a competitive task. Salivary samples were taken, and then, pairs of men competed in a cognitive task that was manipulated to control victory and defeat. After the task, dominance was measured by determining whether the competitors wished to re-challenge their opponents. Similar results were seen: individuals with high testosterone and low cortisol showed increased dominance, but in individuals with high cortisol, the relationship was actually reversed. The second study also indicated that the hormonal relationship to dominance particularly occurred after social threat or provocation.83 Based on these and the earlier studies, Mehta and Josephs proposed the dual-hormone hypothesis and stated that the axes of hormonal stress (HPA) and reproduction (HPG) interact to moderate dominance. The authors argued that as dominance relates to achieving and maintaining social status, testosterone will only have a positive relationship with dominance and hence status, when cortisol is low, but when cortisol is high, high testosterone will not result in increased dominance and may even result in increased submissiveness and promote the seeking of lower status.83 In an expansion of the above studies on aggression and dominance, a further set of studies was performed on the relationship between the two hormones and risk-taking.82 In two studies, one with a mixed sex sample and one with males only, high testosterone levels predicted increased risk, but only in individuals with low cortisol, further supporting the dual-hormone hypothesis.82 Risktaking is considered important in achieving and maintaining social status, and hence acquiring resources and mates. Males in particular take risks in order to attract females. Studies have shown that risk-taking increases in adolescents when witnessed by peers, so it has been suggested that risk-taking is a behavioral evolutionary strategy to heighten personal status in the eyes of peers.101 In  a study of almost 1000 adolescents, hair (although not  salivary) testosterone significantly predicted heightened aggression only in individuals with low hair cortisol. In  individuals with high hair cortisol, testosterone showed no relationship with aggression.10 Similar results have been found in other studies of adolescents.102 In a study of women performing a manipulated competitive test against males, winning or losing resulted in increases in testosterone but not in cortisol.98 Other studies have either supported this result or found higher testosterone only in winners or have found no relationship.98 The authors suggest that the differences relate to experimental design and frequency of sampling, as sampling in this study was done often and immediately after competition. Also, higher testosterone levels were associated with much greater competition accuracy but only when cortisol levels were low, and this relationship was blocked in individuals with high cortisol. More interestingly, this relationship was only seen in individuals who were about to lose the game and were being given negative comments, in other words, competitors who were threatened with losing their status.98 This study supports the dual-hormone hypothesis in a competition setting, using the more objective measurement of accuracy, and also reinforces previous studies that showed that hormone levels related to desire to compete again.98 The authors suggested that the mechanism of action may be, as previously mentioned, that high levels of cortisol block the effects of testosterone. But they also suggest that there may be a more neural explanation. In this study, testosterone and cortisol interacted only when an individual was threatened with loss of status. Neuroimaging studies have shown increased activation in several areas of the brain—including the amygdala, hypothalamus,

154  The chemistry of the body

and periaqueductal gray—in response to threat, as we discussed earlier, and these are the same areas of the brain that are activated when exogenous testosterone is administered.43 Other imaging studies in women showed heightened activation in the amygdala and hypothalamus in response to social threat, and this was particularly strong in individuals with high baseline testosterone and low baseline cortisol.103 In light of this, Henry and colleagues suggested that levels of testosterone and cortisol may affect neurotransmission between these regions of the brain in response to a threat to status, which then impacts behavior. They suggest, therefore, that high testosterone and low cortisol may heighten threat processing but high levels of both may disturb threat processing. Heightened threat processing may increase incentive to respond to the threat to status.98 Many studies support the dual-hormone hypothesis, but some have found a reversal81; it has been suggested that the reversal occurred because individuals with high cortisol have increased reactive aggression in the face of provocation. In a study of female university students in which the participants were personally provoked and then allowed to retaliate with loud blasts of noise, a reversal was found in which baseline testosterone levels predicted reactive aggression as well as state dominance (personal belief that they are better than the competition) only in women with high cortisol baselines.81 The authors suggested that high cortisol levels indicated greatly increased reaction to socially threatening situations. Other studies have shown that exogenous administration of cortisol increased aggression in women but not in men, and neural imaging showed higher amygdala activity in individuals with high cortisol rather than low.81 The authors suggested that the combination of high levels of testosterone, together with the increased emotional reactivity associated with high cortisol, may increase the risk of aggressive reaction or retribution to provocation.81 However, other neural imaging studies have shown that the relationship is much more complex: high testosterone levels predicted increased activity in areas of the brain linked to arousal, emotion, and cognitive control when participants were asked to control or restrain their anger to provocation.104 It has been suggested that men and women may react differently to the testosterone-to-cortisol ratio. Because men are expected to be more dominant and aggressive, high cortisol may reduce dominance in men, but in women, high cortisol may increase stress reactivity and negative emotions, which may result in reactive aggression.81

Cortisol and psychopathy Cortisol has been positively linked to fear, sensitivity to punishment, and withdrawal behavior, all of which have been noted to be deficient in psychopaths.97 Studies have shown differences in cortisol secretion between psychopathic criminals and non-psychopathic criminals, with psychopaths having lower diurnal levels, and within psychopaths, cortisol levels were negatively correlated with violence. Also, men with higher psychopathy scores have exhibited lower cortisol reactivity to stress than those with low psychopathy scores.97 In a study of 237 male and female university students, participants completed the Self-Report Psychopathy, Short Form, and provided saliva samples for testosterone and cortisol assay.97 A weakly positive relationship was seen between psychopathy scores and testosterone in both sexes, but within men only, cortisol was positively linked to psychopathy and moderated the link between testosterone and psychopathy. High testosterone levels were linked to higher levels of ­psychopathy when cortisol was high but were negatively linked when cortisol was low. This is c­ onsistent with the theory that testosterone and cortisol mutually regulate antisocial behaviors and dominance behaviors.97 Many other studies have shown that testosterone levels predict dominance levels but only in people with low cortisol.97 However, this study showed a positive relationship between testosterone and psychopathy when cortisol was high, which is a reversal of the dual-hormone hypothesis. Similar results were seen in adolescent males, which indicated that psychopaths have a higher coupling of both testosterone and cortisol.105 The authors suggest that the normally mutually inhibiting effects of testosterone and cortisol may not occur in males with high psychopathy scores.97 However, in other studies, low cortisol activity has been associated with psychopathy. A ­non-clinical sample of men and women was tested for psychopathic traits, and then, high- and low-­scoring individuals were assigned to one of two types of stressor, based on either performance

Other hormones  155

or social rejection. Salivary cortisol tests were taken just before the stressor and then immediately after the stressor and at 20, 40, and 60 minutes post-stressor. Both males and females with high psychopathy scores had lower cortisol responses to performance-based stress, but only highscoring males had lower cortisol responses to social rejection, although when menstrual phase was removed from the equation, the results were more significant.106 Some studies have suggested that low cortisol levels could be used as a biological marker for specific psychopathy traits. However, a study on adolescents in a Dutch youth detention center did not find a lower basal cortisol level associated with psychopathy, although a relationship with impulsivity was noted in the comparison youth, which has been seen in other studies.107 As in all things, the testosterone-to-cortisol ratio and cortisol levels form only part of the equation and cannot explain all variance within violence, criminal behavior, or psychopathy.

Cortisol overall Cortisol is well known as a stress hormone, and clearly, the HPA axis, which secretes it, can be disrupted or damaged, particularly in early development. The HPA axis relates specifically to stress, which explains why stress events, such as childhood abuse, can disrupt its function and blunt its response. Many studies have also shown that cortisol levels play a major role in the effects of posttraumatic stress disorder and in the resilience to it. The HPA axis does not act in isolation, and its function closely interacts with that of the HPG axis, which secretes testosterone, so it makes sense that the two interact. The above studies show that the relationship between testosterone and aggression or antisocial behavior is not direct and is moderated by cortisol, as well as, no doubt, many other hormones and neurotransmitters, to increase or decrease the risk of antisocial behavior. The complexities of these systems are only just beginning to be understood.

Other hormones Thyroid hormones The thyroid is a gland in the throat that produces two hormones: thyroxin, or tetra-iodothymine (T4), and tri-iodothyronine (T3). These in turn are regulated by thyroid-stimulating hormone (TSH), which is produced by the pituitary gland. Thyroid levels in dogs are very important in regulating aggression. When people buy purebred dogs, they usually ask whether the parents’ thyroid levels are normal. Because thyroid levels are known to be genetic, dogs with abnormal levels (too low or too high) should not be bred. Any dog with a high thyroid level may be aggressive, but many aggressive dogs are unnecessarily put down because their owners do not understand that diet and medication can lower the thyroid level. It is possible that the same sort of relationship exists in people. In humans, the thyroid plays a role in mood regulation, body weight, physical activity, and attention. Many studies have linked thyroid dysfunction with aggression, psychiatric illness, and learning disabilities, although others have also failed to find such a link. Again, this is not  particularly surprising, because aggression, no doubt, has myriad causes. Thyroid insufficiency may be implicated in some. Thyroid hormones also interact with serotonin and norepinephrine (which we will discuss in Chapter 9), neurotransmitters that are known to have links to pathways for a number of behaviors.108 In a large Danish sample of over 2500 individuals diagnosed with hyperthyroidism and over 10 000 controls, patients with hyperthyroidism were significantly more likely to be hospitalized for psychiatric disorders and treated with anti-psychotics than controls.108 In a study of juvenile delinquency, levels of T3 and TSH were measured in adult males between the ages of 38 and 42 years.109 The men were divided into three groups: (1) those who had been considered juvenile delinquents at age 15 but no longer exhibited antisocial behavior, (2) those

156  The chemistry of the body

who were delinquent at age 15 and continued this delinquency into adulthood, and (3) controls. The researchers found that the life-course criminals (group 2) had much higher levels of T3 than either the men who had been juvenile delinquents but had desisted in adulthood or the controls.109 In another study of 61 non-psychotic men undergoing psychiatric evaluations and 66 controls, elevated T3 levels were found to be highly correlated with type II alcoholism (which is highly genetically correlated), psychopathy, criminality, and several personality disorders, including borderline personality disorder, although free thyroxine in the blood was found to be negatively correlated.110 One case study of a merchant seaman is suggestive, as his thyroxin levels were linked to his later violent behavior.111 At age 31, this sailor developed hyperthyroidism and so had most of his thyroid gland removed. Once the thyroid is removed, thyroxin is no longer produced, so it must be supplemented to keep a normal level in the body. The man took the supplements for some time, but then, while at sea, he was unable to obtain his medication. He became depressed, paranoid, and delusionary, convinced that his shipmates planned to attack and rape him. During one such hallucination, he murdered one of his shipmates. Once he was put back on his medication, all the delusions ceased, and he was found not guilty by reason of insanity (that is, not criminally responsible by reason of mental disorder).111 Although only a single case history, the results are interesting and were certainly considered mitigating by the US courts.

Menstruation This chapter would not be complete without a brief comment about menstruation and hormone fluctuations. Most studies looking at hormones consider testosterone the main hormone that could be involved in antisocial behavior. However, there are many other hormones in a person’s body that affect behavior, among other things. Hormone levels in women in general are well studied, not for criminogenic reasons but simply because they fluctuate so greatly during an adult woman’s life. Many hormones in a woman’s body fluctuate dramatically over a monthly cycle, from puberty to menopause, as a woman’s body prepares every month to be ready to accept a fertilized egg (zygote). If the woman is not impregnated, then the body goes through all those hormonal fluctuations again a month later. If she is impregnated, different fluctuations occur. During ovulation, when a woman releases an egg (which occurs roughly midway between menstrual periods), estrogen and progesterone levels are high. Progesterone levels drop prior to menstruation and the other hormones increase. In some women, these changes result in the clinical condition of premenstrual syndrome (PMS), more properly called late luteal phase dysphoric disorder. In some women, it can result in severely disabling symptoms, including feeling overwhelmed, increased irritability, concentration problems, and emotional changes such as depression and aggression. Some people believe that PMS can also escalate into an increased propensity for violent and antisocial behavior. Since the 1800s, anti-feminist reports have tried to relate menstruation to many criminal acts, suggesting that women are “unstable” at such times, and this belief carried well into the 1900s. More serious studies were conducted by Katherina Dalton, a pioneer in research into serious PMS in the 1950s and 1960s. In a study of 150 newly convicted incarcerated women (who all committed non-violent crimes), she found that almost half of the crimes for which the women were convicted had occurred either 4 days before or 4 days after menstruation (chance alone would have been closer to 30%).112 Following this, the results of other studies have suggested that a small group of women may be more susceptible to cyclical fluctuations in hormone levels that make them more prone to anger during menstruation.113 Although most women show few or no effects of PMS during menstruation, some do seem to suffer severely. Whether those women who suffer more are more likely to be involved in crime is an entirely different matter. There is a problem with any study that looks at incarcerated women in that women who spend time together frequently cycle together. No intimate contact is needed, just simply being around other women. Also, if a male is present only at certain times, then a group of women will cycle so

Conclusion  157

that they are fertile when the male is present, again with no intimacy suggested, merely presence. This cycling may be a hangover from prehistoric days, when men went out and hunted and were only present rarely. All the women being fertile at the same time would have increased reproductive success and therefore survival. Critics of the idea that PMS can result in aggression point out that stress can also affect the menstrual cycle, suggesting that stress caused by committing the crime and the subsequent arrest and police interrogation could potentially trigger menstruation, which would mean that the stress brought on menstruation rather than vice versa.29 Moreover, there are many other changes in body chemistry taking place during menstruation, including a reduction in both norepinephrine and blood glucose levels.113 Both of these factors have separately been linked to violence (discussed in later chapters), so any of these factors, or an interaction between them, could explain what may appear to be a possible link between PMS and violence.29 It has been suggested that because PMS occurs during the most “expensive” stage of the menstrual cycle (from a metabolic point of view), energy depletion at this time may result in reduced self-control, as self-control is known to consume a lot of glucose; thus, PMS does not cause aggression or behavioral changes per se, but decreased self-control reduces the inhibition of normal impulses.114 The idea that PMS is linked to crime has been blown out of proportion by tabloid newspapers, and much more research is necessary before we can determine whether any such link actually exists. There are obvious very serious legal and ethical implications. PMS distress has been used as a defense, in some cases successfully. Little work has been done in this area for some time, but if further work does suggest a link, we have to consider how it might affect women’s rights.

Infanticide To finish our discussion of hormones, it is important to mention the fascinating exception the legal system has made in the case of infanticide. It is only very recently—in fact since the first edition of this text—that biological influences on criminal behavior have even been considered as mitigating circumstances in a crime, and they are, and should be, extremely controversial (see Chapter 12 for further discussion). However, it has been accepted for a very long time that infanticide—the deliberate killing of a baby under the age of 12 months by the biological mother—is caused by biological influences. This acceptance is premised on the belief that the mother had not recovered from the effects of the birth, so her mind was disturbed. Despite the fact that the killing was premeditated, so should have been judged first-degree murder, it is usually considered manslaughter, or specifically infanticide, in most countries today and is either not prosecuted at all or treated much more lightly than if the killing had occurred after 12 months or if the baby had been killed by anyone other than the biological mother. This is all premised on the idea that the birth affected the mother’s mens rea.

Conclusion The results of these studies are leading us away from the idea that testosterone equals violence, to a much better understanding of the complexities and interactions that occur between all our body systems, including but not limited to our body chemistry, or hormones. It is interesting to note that even people who believe that biology and genetics have absolutely no role in a person’s behavior still accept that increased aggression and criminal behavior in youth are due to testosterone.115 This seems to be deeply embedded in our psyche and is frequently not even considered as a biological influence. The many studies discussed above show that testosterone’s role in antisocial behavior is not nearly as straightforward as thought in the past, and its effect is moderated by a number of other systems, in particular the HPA axis and cortisol. The results are intriguing, but caution should be exercised in interpreting them; although many of these studies show exciting links, meta-analyses often show a fairly small overall effect size, indicating that although a relationship exists, there are a large number of other variables involved. This result is in part due to attempting to compare studies with different

158  The chemistry of the body

methodological designs, tests, and participants, but it still shows a need for caution. This is to be expected in our exploration of biological predispositions and risk factors for antisocial or even criminal behavior. We have said from the beginning that our biology does not make us commit criminal acts, only that it may contribute risk factors to increase antisocial behavior. Many of these risk factors may be very small and, together with the myriad other factors, both social and environmental, that impact our behavior, may be entirely insignificant. Even if the risk factors are a little higher, if those social and environmental factors are protective or prosocial, then the risk for antisocial behavior is greatly reduced and may never amount to anything. However, when those hormonal risk factors interact with negative social and environmental factors, the risk increases. In some cases, this may be a simple summation of a number of risk factors, biological, social, and environmental, but in others, it may be a direct interaction—risk factors such as child abuse actually change a person’s biology and make it more sensitive to certain other risk factors. Understanding these interactions is critical, not only for our understanding but also for developing tools and intervention strategies to help individuals. If we could use a simple non-invasive test such as saliva or hair to assess hormone levels and possible future risk, it would be much easier to prioritize individuals, particularly children, for intervention and could help us choose the correct intervention strategy. For example, in the rather drastic use of castration to reduce sexual recidivism, a reduction in testosterone production allowed offenders to successfully participate in treatment programs in which they had previously been unable to engage. At the same time, using such a simple test when so many other factors are at play runs the risk of labeling.

Questions for further study and discussion 1. Discuss the possible evolutionary benefits of risk-taking. 2. Is it ethical to offer a prisoner the option of either castration or indefinite incarceration? Why or why not? 3. Violence is abhorred in our society, so why does it still exist, from an evolutionary perspective? 4. From an evolutionary perspective only, why might rape be an evolutionary strategy? Think of this in the context of early humans rather than our modern society. 5. Discuss the relationship between negative eugenics and lengthy incarceration. 6. What are some of the ethical issues raised by the arguments about premenstrual tension and crime? 7. Why is baseline testosterone so much less useful than acute testosterone levels to explain behavior?

References 1. Ellis, L. and Hoskin, A.W. 2015. Criminality and the 2D:4D ratio: Testing the prenatal androgen hypothesis. Int. J. Offender Ther. Comp. Criminol. 59(3): 295–312. 2. Coates, J.M. and Herbert, J. 2008. Endogenous steroids and financial risk taking on a London trading floor. Proc. Nat. Acad. Sci. USA 105(16): 6167–6172. 3. Boyd, N. 2000. The  testosterone connection, In: The  Beast Within: Why Men Are Violent. New York: Greystone Books. pp. 115–138. 4. Haller, J. 2014. Neurobiological Bases of Abnormal Aggression and Violent Behaviour. London: Springer. 220 p. 5. Brain, P.B. and Haug, M. 1992. Hormonal and neurochemical correlates of various forms of animal “aggression.” Psychoneuroendocrinology 17(6): 537–551. 6. Harris, J.A., Vernon, P.A., and Boomsma, D.I. 1998. The heritability of testosterone. Behav. Genet. 28(3): 165–171.

References  159

7. Kuijper, E.A., Lambalk, C.B., Boomsma, D.I. et al. 2007. Heritability of reproductive hormones in adult male twins. Hum. Reprod. 22(8): 2153–2159. 8. Caramaschi, D., Booij, L., Petitclerc, A., Boivin, M., and Tremblay, R.E. 2012. Genetic and environmental contributions to saliva testosterone levels in male and female infant twins. Psychoneuroendocrinology 37(12): 1954–1959. 9. Eisenegger, C., Haushofer, J., and Fehr, E. 2011. The role of testosterone in social interaction. Trends Cogn. Sci. 15(6): 263–271. 10. Grotzinger, A.D., Mann, F.D., Patterson, M.W., Tackett, J.L., Tucker-Drob, E.M., and Harden, K.P. 2018. Hair and salivary testosterone, hair cortisol, and externalizing behaviors in adolescents. Psychol. Sci. 29(5): 688–699. 11. Phoenix, C.H., Goy, R.W., Gerall, A.A., and Young, W.C. 1959. Organizing action of prenatally administered testosterone propionate on tissues mediating mating behavior in female guinea pig. Endocrinology. 65(3): 369–382. 12. Turanovic, J.J., Pratt, T.C., and Piquero, A.R. 2017. Exposure to fetal testosterone, aggression, and violent behavior: A meta-analysis of the 2D:4D digit ratio. Aggress. Viol. Behav. 33: 51–61. 13. Hoskin, A.W. and Ellis, L. 2015. Fetal testosterone and criminality: Test of evolutionary neuroandrogenic theory. Criminology 53(1): 54–73. 14. Perciavalle, V., Di Corrado, D., Petralia, M.C., Gurrisi, L., Massimino, S., and Coco, M. 2013. The second-to-fourth digit ratio correlates with aggressive behavior in professional soccer players. Mol. Med. Rep. 7(6): 1733–1738. 15. Lutchmaya, S., Baron-Cohen, S., Raggatt, P., Knickmeyer, R., and Manning, J.T. 2004. 2nd to 4th digit ratios, fetal testosterone and estradiol. Early Hum. Dev. 77(1–2): 23–28. 16. Zheng, Z. and Cohn, M.J. 2011. Developmental basis of sexually dimorphic digit ratios. Proc. Natl. Acad. Sci. USA 108(39): 16289–16294. 17. Bailey, A.A. and Hurd, P.L. 2005. Finger length ratio (2D:4D) correlates with physical aggression in men but not in women. Biol. Psychol. 68(3): 215–222. 18. Hönekopp, J. 2011. Relationships between digit ratio 2D:4D and self-reported aggression and risk taking in an online study. Person. Individ. Diff. 51(1): 77–80. 19. Hanoch, Y., Gummerum, M., and Rolison, J. 2012. Second-to-fourth digit ratio and impulsivity: A comparison between offenders and nonoffenders. PLoS One 7(10): e47140. 20. Schwerdtfeger, A., Heims, R., and Heer, J. 2010. Digit ratio (2D:4D) is associated with traffic violations for male frequent car drivers. Accid. Anal. Prev. 42(1): 269–274. 21. Stenstrom, E., Saad, G., Nepomuceno, M.V., and Mendenhall, Z. 2011. Testosterone and domain-specific risk: Digit ratios (2D:4D and rel2) as predictors of recreational, financial, and social risk-taking behaviors. Personal. Individ. Diff. 51(4): 412–416. 22. Pratt, T.C., Turanovic, J.J., and Cullen, F.T. 2016. Revisiting the criminological consequences of exposure to fetal testosterone: A meta-analysis of the 2D:4D digit ratio. Criminology 54(4): 587–620. 23. Hoskin, A.W. 2017. Male sex hormones and criminal behavior: The  predictive power of a two-factor model of organizational androgen exposure. Personal. Individ. Diff. 108: 86–90. 24. Crewther, B., Cook, C., Kilduff, L., and Manning, J. 2015. Digit ratio (2D:4D) and salivary testosterone, oestradiol and cortisol levels under challenge: Evidence for prenatal effects on adult endocrine responses. Early Hum. Dev. 91(8): 451–456. 25. Apicella, C., Dreber, A., Campbell, B., Gray, P., Hoffman, M., and Little, A. 2008. Testosterone and financial risk preferences. Evol. Hum. Behav. 29(6): 384–390. 26. Kim, Y., Kim, K., and Kim, T.H. 2014. Domain specific relationships of 2D:4D digit ratio in risk perception and risk behavior. J. Gen. Psychol. 141(4): 373–392. 27. Anderson, T. 2012. Comparing risk-taking and digit ratio (2D:4D) in offenders and nonoffenders. Plymouth Stud. Sci. 5(2): 105–120. 28. Carré, J.M., McCormick, C.M., and Hariri, A.R. 2011. The social neuroendocrinology of human aggression. Psychoneuroendocrinology 36(7): 935–944.

160  The chemistry of the body

29. Raine, A. 1993. Other biological factors: Head injury, pregnancy and birth complications, physical appearance, hormones, diet and lead, In: The Psychopathology of Crime: Criminal Behavior as a Clinical Disorder. San Diego: Academic Press, Elsevier Science. pp. 191–214. 30. Rilling, J.K., Worthman, C.M., Campbell, B.C., Stallings, J.F., and Mbizva, M. 1996. Ratios of plasma and salivary testosterone throughout puberty: Production versus bioavailability. Steroids 61: 374–378. 31. Carré, J.M. and Olmstead, N.A. 2015. Social neuroendocrinology of human aggression: Examining the role of competition-induced testosterone dynamics. Neuroscience 286: 171–186. 32. Carré, J.M. and Archer, J. 2018. Testosterone and human behavior: The role of individual and contextual variables. Curr. Opin. Psychol. 19: 149–153. 33. Carré, J.M., Geniole, S.N., Ortiz, T.L., Bird, B.M., Videto, A., and Bonin, P.L. 2017. Exogenous testosterone rapidly increases aggressive behavior in dominant and impulsive men. Biol. Psychiatry 82(4): 249–256. 34. Dabbs, J.M. and Morris, R. 1990. Testosterone, social class and antisocial behavior in a sample of 4,462 men. Psychol. Sci. 1: 209–211. 35. Dabbs, J.M., Frady, R.L., Carr, T.S., and Besch, N.F. 1987. Saliva testosterone and criminal violence in young adult prison inmates. Psychosomatic Med. 49: 174–182. 36. Dabbs, J.M., Ruback, G.J., Frady, R.L., Hopper, C.H., and Sgoutas, D.S. 1988. Saliva testosterone and criminal violence among women. Person. Individ. Diff. 9: 269–275. 37. Dabbs, J.M. and Hargrove, M.F. 1997. Age, testosterone and behavior among female prison inmates. Psychosomatic Med. 59(5): 477–480. 38. Dabbs, J.M., Carr, T.S., Frady, R.L., and Riad, J.K. 1995. Testosterone, crime and misbehavior among 692 male prison inmates. Person. Individ. Diff. 18(5): 627–633. 39. Dabbs, J.M., Riad, J.K., and Chance, S.E. 2001. Testosterone and ruthless homicide. Person. Individ. Diff. 31: 599–603. 40. Wong, J.S. and Gravel, J. 2018. Do sex offenders have higher levels of testosterone? Results from a meta-analysis. Sex Abuse 30(2): 147–168. 41. Kingston, D.A., Seto, M.C., Ahmed, A.G., Fedoroff, P., Firestone, P., and Bradford, J.M. 2012. The role of central and peripheral hormones in sexual and violent recidivism in sex offenders. J. Am. Acad. Psychiatry Law 40: 476–485. 42. van Honk, J., Schutter, D.J., Hermans, E.J., Putman, P., Tuiten, A., and Koppeschaar, H. 2004. Testosterone shifts the balance between sensitivity for punishment and reward in healthy young women. Psychoneuroendocrinology 29(7): 937–943. 43. Goetz, S.M., Tang, L., Thomason, M.E., Diamond, M.P., Hariri, A.R., and Carré, J.M. 2014. Testosterone rapidly increases neural reactivity to threat in healthy men: A novel two-step pharmacological challenge paradigm. Biol. Psychiatry 76(4): 324–331. 44. Welling, L.L., Moreau, B.J., Bird, B.M., Hansen, S., and Carré, J.M. 2016. Exogenous testosterone increases men’s perceptions of their own physical dominance. Psychoneuroendocrinology 64: 136–142. 45. Welker, K.M., Norman, R.E., Goetz, S., Moreau, B.J.P., Kitayama, S., and Carré, J.M. 2017. Preliminary evidence that testosterone’s association with aggression depends on self-­ construal. Horm. Behav. 92: 117–127. 46. Lobbestael, J., Baumeister, R.F., Fiebig, T., and Eckel, L.A. 2014. The role of grandiose and vulnerable narcissism in self-reported and laboratory aggression and testosterone reactivity. Person. Individ. Diff. 69: 22–27. 47. Norman, R.E., Moreau, B.J.P., Welker, K.M., and Carré, J.M. 2014. Trait anxiety moderates the relationship between testosterone responses to competition and aggressive behavior. Adapt. Hum. Behav. Physiol. 1(3): 312–324. 48. Oswald, Z.E. 2013. “Off with his—”: Analyzing the sex disparity in chemical castration sentences. Mich. J. Gender and Law 19: 471–503.

References  161

49. Douglas, T., Bonte, P., Focquaert, F., Devolder, K., and Sterckx, S. 2013. Coercion, incarceration, and chemical castration: An argument from autonomy. J. Bioeth. Inq. 10(3): 393–405. 50. McMillan, J. 2014. The kindest cut? Surgical castration, sex offenders and coercive offers. J. Med. Ethics 40(9): 583–590. 51. Khan, O. and Mashru, A. 2016. The efficacy, safety and ethics of the use of testosteronesuppressing agents in the management of sex offending. Curr. Opin. Endocrinol. Diabetes Obes. 23(3): 271–278. 52. Koo, K.C., Shim, G.S., Park, H.H., Rha, K.H. et al. 2013. Treatment outcomes of chemical castration on Korean sex offenders. J. Forensic Leg. Med. 20(6): 563–566. 53. Sealey, G. 2 March 2017. Some sex offenders opt for castration. abcNEWS. 54. van der Meer, T. 2014. Voluntary and therapeutic castration of sex offenders in the Netherlands (1938–1968). Int. J. Law Psychiatry 37(1): 50–62. 55. Sedkaoui, H. and Mullet, E. 2016. Mapping French people’s views on chemical castration of child and adolescent sex offenders. Univ Psychol 15(3): 1–7. 56. Blaszczynski, A. and Nower, L. 2003. Imaginal desensitisation: A  relaxation technique for impulse control disorders. J. Clin. Activit. Assign. Handout Psychotherapy Pract. 2(4): 1–14. 57. Schmucker, M. and Lösel, F. 2015. The effects of sexual offender treatment on recidivism: An international meta-analysis of sound quality evaluations. J. Exp. Criminol. 11(4): 597–630. 58. Wille, R. and Beier, K.M. 1989. Castration in Germany. Ann. Sex. Res. 2: 103–134. 59. Lösel, F. and Schmucker, M. 2005. The  effectiveness of treatment for sexual offenders: A comprehensive meta-analysis. J. Exp. Criminol. 1(1): 117–146. 60. Thibaut, F., De La Barra, F., Gordon, H., Cosyns, P., Bradford, J.M., and Disorders, W.T.F.o. S. 2010. The World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for the biological treatment of paraphilias. World J. Biol. Psychiatry. 11(4): 604–655. 61. Goswami, G.K. 2014. Castration laws: Marching towards imperfect justice. Medico-Legal J. 82: 70–79. 62. Chism, L.S. 2013. The case for castration: A “shot” towards rehabilitation of sexual offenders. Law Psychol. Rev. 37: 193–209. 63. Bhaumik, S. 2013. Indian medical experts shun chemical castration for rapists. BMJ 346: f64. 64. Booth, A., Johnson, D.R., and Granger, D.A. 1999. Testosterone and men’s health. J. Behav. Med. 22(1): 1–19. 65. Stanton, S.J., Mullette-Gillman, O.A., McLaurin, R.E. et al. 2011. Low- and high-testosterone individuals exhibit decreased aversion to economic risk. Psychol. Sci. 22(4): 447–453. 66. Stanton, S.J., Liening, S.H., and Schultheiss, O.C. 2011. Testosterone is positively associated with risk taking in the Iowa Gambling Task. Horm. Behav. 59(2): 252–256. 67. Ellis, L. and Hoskin, A.W. 2015. The evolutionary neuroandrogenic theory of criminal behavior expanded. Aggress. Viol. Behav. 24: 61–74. 68. Mazur, A. 1985. A biosocial model of status in face-to-face primate groups. Soc. Forces 64: 377–402. 69. Cheng, J.T., Korniemnko, O., and Granger, D.A. 2018. Prestige in a large-scale social group predicts longitudinal changes in testosterone. J. Personality Social. Psychol. 114(6): 924–944. 70. Vongas, J.G. and Al Hajj, R. 2017. The effects of competition and implicit power motive on men’s testosterone, emotion recognition, and aggression. Horm. Behav. 92: 57–71. 71. Geniole, S.N., Bird, B.M., Ruddick, E.L., and Carré, J.M. 2017. Effects of competition outcome on testosterone concentrations in humans: An updated meta-analysis. Horm. Behav. 92: 37–50. 72. Casto, K.V., Rivell, A., and Edwards, D.A. 2017. Competition-related testosterone, cortisol, and perceived personal success in recreational women athletes. Horm. Behav. 92: 29–36. 73. Casto, K.V. and Edwards, D.A. 2015. Testosterone and reconciliation among women: Aftercompetition testosterone predicts prosocial attitudes towards opponents. Adapt. Hum. Behav. Physiol. 2(3): 220–233.

162  The chemistry of the body

74. Giotakos, O., Markianos, M., Vaidakis, N., and Christodoulou, G.N. 2003. Aggression, impulsivity, plasma sex hormones, and biogenic amine turnover in a forensic population of rapists. J. Sex Marital Ther. 29(3): 215–225. 75. Bernhardt, P.C. 1997. Influences of serotonin and testosterone aggression and dominance: Convergence with social psychology. Curr. Dir. Psychol. Sci. 6(2): 44–48. 76. Higley, J.D., Mehlman, P.T., Poland, R.E. et al. 1996. CSF Testosterone and 5-HIAA correlate with different types of aggressive behaviors. Biol. Psychiatry 40: 1067–1082. 77. van Honk, J., Terburg, D., and Bos, P.A. 2011. Further notes on testosterone as a social hormone. Trends Cogn. Sci. 15(7): 291–292. 78. Wibral, M., Dohmen, T., Klingmuller, D., Weber, B., and Falk, A. 2012. Testosterone administration reduces lying in men. PLoS One 7(10): e46774. 79. Eisenegger, C., Naef, M., Snozzi, R., Heinrichs, M., and Fehr, E. 2010. Prejudice and truth about the effect of testosterone on human bargaining behaviour. Nature 463(7279): 356–359. 80. van Honk, J., Montoya, E.R., Bos, P.A., van Vugt, M., and Terburg, D. 2012. New evidence on testosterone and cooperation. Nature 485(7399): E4–E6. 81. Denson, T.F., Mehta, P.H., and Ho Tan, D. 2013. Endogenous testosterone and cortisol jointly influence reactive aggression in women. Psychoneuroendocrinology 38(3): 416–424. 82. Mehta, P.H., Welker, K.M., Zilioli, S., and Carré, J.M. 2015. Testosterone and cortisol jointly modulate risk-taking. Psychoneuroendocrinology 56: 88–99. 83. Mehta, P.H. and Josephs, R.A. 2010. Testosterone and cortisol jointly regulate dominance: Evidence for a dual-hormone hypothesis. Horm. Behav. 58(5): 898–906. 84. Ouellet-Morin, I., Odgers, C.L., Danese, A., Bowes, L. et al. 2011. Blunted cortisol responses to stress signal social and behavioral problems among maltreated/bullied 12-yearold children. Biol. Psychiatry 70(11): 1016–1023. 85. van Goozen, S.H., Fairchild, G., Snoek, H., and Harold, G.T. 2007. The evidence for a neurobiological model of childhood antisocial behavior. Psychol. Bull. 133(1): 149–182. 86. Cicchetti, D., Gunnar, M.R., Rogosch, F.A., and Toth, S.L. 2010. The  differential impacts of early physical and sexual abuse and internalizing problems on daytime cortisol thythm in school-aged children. Child Develop. 81(1): 252–269. 87. Ouellet-Morin, I., Danese, A., Bowes, L. et al. 2011. A discordant monozygotic twin design shows blunted cortisol reactivity among bullied children. J. Am. Acad. Child Adolesc. Psychiatry 50(6): 574–582. 88. Brendgen, M., Ouellet-Morin, I., Lupien, S., Vitaro, F., Dionne, G., and Boivin, M. 2017. Does cortisol moderate the environmental association between peer victimization and depression symptoms? A genetically informed twin study. Psychoneuroendocrinology 84: 42–50. 89. Rampp, C., Binder, E.B., and Provencal, N. 2014. Epigenetics in posttraumatic stress disorder. Prog. Mol. Biol. Transl. Sci. 128: 29–50. 90. Gowin, J.L., Green, C.E., Alcorn, J.L., 3rd, Swann, A.C., Moeller, F.G., and Lane, S.D. 2013. The role of cortisol and psychopathy in the cycle of violence. Psychopharmacology (Berl) 227(4): 661–672. 91. Bernard, K., Frost, A., Bennett, C.B., and Lindhiem, O. 2017. Maltreatment and diurnal cortisol regulation: A meta-analysis. Psychoneuroendocrinology 78: 57–67. 92. McBurnett, K., Lahey, B.B., Rathouz, P.J., and Loeber, R. 2000. Low salivary cortisol and persistent aggression in boys referred for disruptive behaviour. Arch. Gen. Psychol. 57(1): 38–43. 93. Shoal, G.D., Giancola, P.R., and Kirillova, G.P. 2003. Salivary cortisol, personality, and aggressive behavior in adolescent boys: A  5-year longitudinal study. J. Am. Acad. Child Adolesc. Psychiatry 42(9): 1101–1107. 94. Moffit, T.E. 1993. Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychol. Rev. 100(4): 674–701. 95. Raine, A., Moffitt, T.E., Caspi, A., Loeber, R., Stouthamer-Loeber, M., and Lynam, D. 2005. Neurocognitive impairments in boys on the life-course persistent antisocial path. J. Abnorm. Psychol. 114(1): 38–49.

References  163

96. Platje, E., Jansen, L.M., Raine, A. et  al. 2013. Longitudinal associations in adolescence between cortisol and persistent aggressive or rule-breaking behavior. Biol. Psychol. 93(1): 132–137. 97. Welker, K.M., Lozoya, E., Campbell, J.A., Neumann, C.S., and Carré, J.M. 2014. Testosterone, cortisol, and psychopathic traits in men and women. Physiol. Behav. 129: 230–236. 98. Henry, A., Sattizahn, J.R., Norman, G.J., Beilock, S.L., and Maestripieri, D. 2017. Performance during competition and competition outcome in relation to testosterone and cortisol among women. Horm. Behav. 92: 82–92. 99. Popma, A., Vermeiren, R., Geluk, C.A. et  al. 2007. Cortisol moderates the relationship between testosterone and aggression in delinquent male adolescents. Biol. Psychiatry 61(3): 405–411. 100. Dabbs, J.M., Jurkovic, G.J., and Frady, R.L. 1991. Saliva testosterone and cortisol among late adolescent male offenders. J. Abn. Child Psychol. 19(4): 469–478. 101. Ellis, B.J., Del Giudice, M., Dishion, T.J. et al. 2012. The evolutionary basis of risky adolescent behavior: Implications for science, policy, and practice. Dev. Psychol. 48(3): 598–623. 102. Platje, E., Popma, A., Vermeiren, R.R. et  al. 2015. Testosterone and cortisol in relation to aggression in a non-clinical sample of boys and girls. Aggress. Behav. 41(5): 478–487. 103. Hermans, E.J., Ramsey, N.F., and van Honk, J. 2008. Exogenous testosterone enhances responsiveness to social threat in the neural circuitry of social aggression in humans. Biol. Psychiatry 63(3): 263–270. 104. Denson, T.F., Ronay, R., von Hippel, W., and Schira, M.M. 2013. Endogenous testosterone and cortisol modulate neural responses during induced anger control. Soc. Neurosci. 8(2): 165–177. 105. Johnson, M.M., Dismukes, A.R., Vitacco, M.J., Breiman, C., Fleury, D., and Shirtcliff, E.A. 2014. Psychopathy’s influence on the coupling between hypothalamic-pituitary-adrenal and -gonadal axes among incarcerated adolescents. Dev. Psychobiol. 56(3): 448–458. 106. O’Leary, M.M., Taylor, J., and Eckel, L. 2010. Psychopathic personality traits and cortisol response to stress: The role of sex, type of stressor, and menstrual phase. Horm. Behav. 58(2): 250–256. 107. Feilhauer, J., Cima, M., Korebrits, A., and Nicolson, N.A. 2013. Salivary cortisol and psychopathy dimensions in detained antisocial adolescents. Psychoneuroendocrinology 38(9): 1586–1595. 108. Brandt, F., Thvilum, M., Almind, D. et al. 2014. Hyperthyroidism and psychiatric morbidity: Evidence from a Danish nationwide register study. Eur. J. Endocrinol. 170(2): 341–348. 109. Alm, B., af Klintenberg, B., Humble, K., Leppert, J., Sorenson, S., Tegelman, R. et al. Lidberg, L. 1996. Criminality and psychopathy as related to thyroid activity in former juvenile ­delinquents. Acta. Psych. Scand. 94(2): 112–117. 110. Stalenheim, E.G., Von Knorring, L., and Wide, L. 1998. Serum levels of thyroid hormones as biological markers in a Swedish forensic psychiatric population. Biol. Psychol. 43(10): 755–761. 111. Tardiff, K. 1998. Unusual diagnoses among violent patients. Psych. Clinics N. Am. 21(3): 567–576. 112. Dalton, K. 1961. Menstruation and crime. Br. Med. J. 2: 1752–1753. 113. Fishbein, D.H. 1992. The psychobiology of female aggression. Crim. Just. Behav. 19: 99–1126. 114. Gailliot, M.T., Hildebrandt, B., Eckel, L.A., and Baumeister, R.F. 2010. A theory of limited metabolic energy and premenstrual syndrome symptoms: Increased metabolic demands during the luteal phase divert metabolic resources from and impair self-control. Rev. Gen. Psychol. 14(3): 269–282. 115. Pieri, E. and Levitt, M. 2008. Risky individuals and the politics of genetic research into aggressiveness and violence. Bioethics 22(9): 509–518.

8 The prenatal environment and birth complications

Introduction This  chapter complements the examination of hormones in the previous chapter. During fetal development, all of a baby’s systems are formed. Any damage that occurs during the mother’s pregnancy or at the time of birth can have critical implications in the later life of the child. Many complications can occur during pregnancy or at birth, and they may damage the brain or other systems in the body. Even maternal stress can have a major impact on the developing fetus. Damage in the early stages of development can be dramatically more serious than similar levels of damage in adulthood.

Birth complications Complications at birth can cause myriad injuries, from physical defects to direct brain damage. Because the brain is the seat of all behaviors, brain damage can obviously affect behavior and learning. Therefore, it is probable that birth, or perinatal (peri meaning “at the time of,” natal meaning “birth”), complications can cause problems in later life, including criminal and violent behavior. Since the 1930s, many studies have shown a link between birth complications and later behavioral problems in children. The first studies looked at the link between birth complications and schizophrenia. A meta-analysis of studies found that three types of complications were linked to ­schizophrenia: (1) complications of the pregnancy, such as hemorrhage, diabetes, and p ­ re-­eclampsia; (2)  abnormal fetal growth, such as low birth weight, small head circumference, and congenital deformities; and (3) delivery complications, such as emergency cesarean section and asphyxia. The most important seemed to be fetal asphyxia.1 This is not surprising, as asphyxia, or hypoxia, causes a lack of oxygen to the brain, which is known to cause brain damage. Delays or ­difficulties in birth can easily result in hypoxia if not corrected rapidly. Such damage can result in later problems, such as learning disabilities, impaired cognitive abilities, attention deficit hyperactivity disorder (ADHD), depression, and eating disorders, as well as schizophrenia.2 Many animal studies have shown that hypoxia can damage the prefrontal cortex, hippocampus, and dopamine systems and results in lowered dopamine transmission.1 The prefrontal cortex is involved in inhibitory control, so damage will impact behavior control.3 Hypoxia can also result in physical disabilities due to a lack of nervous control of muscles. As many studies have shown links between birth complications and a number of psychotic disorders, studies were conducted to determine whether this link held true with personality disorder.2 Individuals with personality disorder in Sweden were identified through the national registry of 165

166  The prenatal environment and birth complications

psychiatric evaluations and compared with two control groups: a sample of offenders who had been forensically evaluated but did not have any psychiatric disorder and a sample of the general public who had no psychiatric history. A large range of birth complications were considered, including delivery complications, bleeding, prematurity, cesarean section, and infant-related factors such as low birth weight, being small for gestational age, and anoxia or asphyxia at birth. The main issues that were clearly linked with being diagnosed with personality disorder in adulthood were low birth weight and being born prematurely.2 Both of these relationships were statistically significant when compared with either control group; however, the link between primarily asphyxia at birth, as well as the majority of other birth complications and infant factors considered, was only significant when compared with the general population and not with the offender group.2 Of course, it is still difficult to determine whether personality disorder is directly related to birth complications, as it is quite possible that the mothers also suffered from personality disorder, which may have affected their personal prenatal care and potential substance abuse,2 as well as their ability to parent. Extremely preterm births (below 28  weeks) have shown higher risks of neurodevelopmental problems, which cause sequelae throughout the life course—attention and regulatory problems, seen in early years; emotional and peer relationship issues; and inattention at school, with such issues persisting into adulthood.4 Approximately a quarter of such individuals have psychiatric disorders, but it is suggested that closer to half may have functional difficulties.4 A vast number of studies have looked at the effects of birth complications on future behavior, and many have found links, often quite strong links, between antisocial behavior and birth complications; however, another major factor that must be included is the environment in which the child grows up. Environment in child development obviously has a major influence. Early brain and central nervous system damage may have no effect on a child raised in a good and stable environment. Longitudinal studies have shown that prenatal and perinatal complications had negligible or no long-term effects in advantaged families, whereas in disadvantaged families, they did predict problems.5 Alternatively, a child with behavioral problems associated with birth complications may be more difficult to parent. It has been demonstrated that children with neurological problems have an increased risk for parental child abuse and often receive harsher disciplinary measures than normal children, so it is possible that the birth complications are only indirectly responsible for the delinquent behavior, with the negative and unsupportive child-rearing environment having a greater effect.6 Brain imaging has allowed a deeper analysis of the effects of certain birth complications on the developing brain. Magnetic resonance imaging (MRI) (see Chapter 10) was used to compare the brains of 65 eight-year-old children born prematurely with a sample of 70 children who were born at full term.7 The  preterm children showed an overall reduction in brain volume, although this seemed to have a greater influence on boys over girls. Girls showed normal volumes of white matter but reduced gray matter, whereas boys had significant reductions in both gray and white matter.7 Most of the deficit in boys was seen in the temporal lobe and deep cerebral region of the brain, which are important in emotion, learning, reading, and attention and are often impaired in children born prematurely. As our ability to safely image brains has improved over the last few years, so has the use of imaging techniques in understanding the effects of a variety of birth complications. For example, it has been shown that premature babies are at greatly increased risk for cerebellar hemorrhage, with almost 50% of such babies affected.8 A variety of MRI techniques are now commonly used during pregnancy to assess the development of the brain and to identify potential problems. Brain imaging studies are now  commonly used, and many studies have shown that people with psychotic illnesses have abnormal brain morphology, with the most common being reduced gray matter volume in the cortex.9 The three distinct parts of the cortex mature at different times, ­ranging from prior to 24 months to shortly before puberty, and changes continue throughout life.9 MRI scans of 36 adult male patients with early-stage psychosis were compared with their extensive birth histories; these scans revealed that birth complications were associated with cortical ­congenital malformation.9

Fetal development, nutrition, and pollutants  167

Another obvious risk occurs when the mother is subjected to trauma during the pregnancy itself, rather than just birth. The mother and the fetus could be injured in an accident; however, maternal and fetal damage might also result from abuse. In the latter case, the child has a double risk: the actual physical damage caused during gestation and also the potential risks from physical, ­emotional, and psychological damage when born into a dysfunctional and abusive environment. Of course, all this research looks only at the time of birth and the complications that occur over a matter of hours. There are many things that could affect a fetus much earlier.

Fetal development, nutrition, and pollutants Growing bodies of evidence show us that prenatal conditions play a major role in all aspects of fetal,  childhood, adolescent, and adult life, impacting many aspects of future health as well as behavioral outcomes. When a fetus is developing, it is entirely dependent on its mother for all resources, including nutrition. A very common birth issue related to a multitude of health, behavioral, and cognitive problems is that of low birth weight.10 Low birth weight and prematurity have frequently been linked to high mortality and morbidity, as well as behavioral, neurological, and developmental problems.11 In a large twin study in both the United Kingdom and New Zealand, low birth weight was linked to lower IQ and was a risk factor for reduced neurological health.12 Many studies have linked low birth weight with a number of negative neurodevelopmental, cognitive, and behavioral outcomes, although some have suggested that its link to higher crime rates is due to the interaction between low birth weight and other adversities.10,11 Low birth weight and other prenatal issues may result from the mother simply not  receiving enough nutrition. Several studies have shown that major nutritional deficits in the mother can have a large impact on fetal development. The Nazi blockade in the Netherlands from October 1944 to May 1945 caused severe famine, and studies on the offspring of women who were pregnant ­during this time have shown that severe nutritional deprivation during pregnancy can result in many sequelae, including increased birth mortality, obesity, a reduction in glucose tolerance, and even altered birth weights into the next generation.13 At age 18, male children were called to m ­ ilitary service and received extensive physical and psychological testing. Extensive records allowed researchers to compare men who were prenatally exposed to severe nutritional deprivation with those who were not. These indicated that those men whose mothers were starved during the first and second trimesters had a significantly higher risk of developing antisocial personality d ­ isorder, suggesting that the nutritional deficits impacted the developing brain.13 A much larger study on prenatal exposure to famine during the Chinese famine of 1959–1961 showed that the prevalence of schizophrenia almost doubled in people born during and just after the famine.14 Animal studies suggest that prenatal maternal nutritional deprivation affects both metabolic development and neurodevelopment and ultimately affects the offspring’s executive functioning and its reward processing system, resulting in cognitive problems and problems with normal conditioning with reward and pleasure.15 As understanding of brain development in utero increases, it has become clear that nutrition plays a major role in neural development. Hormones play a crucial role in fetal growth, mediating the use of the available nutrients, that is, handling the nutrients and determining what goes where. One of those hormones is called insulin-like growth factor 1 (IGF1), which is involved in the regulation of  fetal growth; levels of IGF1 in blood correlate with birth weight. Some researchers have shown that polyunsaturated fatty acids (PUFAs) may be very important in brain development and birth weight, as well as in myelination (covering) of the nerve cells in the brain, which is vital for nerve function. These essential fatty acids are required for the formation of the myelin sheath.16 They are available only from the placenta or diet. So, a maternal diet that is high in these fatty acids is a protective factor in nerve maturation. On the other hand, if the mother’s diet is low in these fatty acids, then myelination can be delayed or damaged. The result is reduced connections between the two sides, or hemispheres, of the brain, and brain problems involving cognitive, perceptual, and sensorimotor functions may occur.

168  The prenatal environment and birth complications

For  example,  for a person to be able to perform coordination tasks, communication must occur between the two hemispheres of the brain. Such developmental problems could eventually result in behavior disorders, developmental dyslexia, and other learning disorders.17 In an experiment to test the role of PUFAs, pregnant rats were treated with an immune activator that is known to instigate a number of neurodevelopmental problems in offspring, including behavioral and physiological issues. Subsequent supplementation of the young offspring with PUFAs reversed the abnormal behavior and produced protective effects against neuronal ­damage.16 In a human study in Mexico involving over 1000 pregnant women, half the women were supplemented with docosahexanoic acid (DHA), a major PUFA involved in the development of the central nervous system, from second trimester to delivery, and their offspring were assessed at age 5.18 Children were assessed using several tests of cognitive development, executive functioning, and behavior. Although most scales were similar in both treated and untreated groups, children born of supplemented mothers exhibited improved sustained attention. Interestingly, a healthy, positive, stable home environment increased general cognitive abilities in the treated group but not in the untreated group, showing the interaction of good home environment with other protective factors.18 Maternal stress during pregnancy can have several deleterious effects on the offspring, as will be discussed shortly, but studies have shown that supplementing PUFAs during pregnancy can ameliorate stress and result in improved positive mood, although the causal relationship is not fully understood. It  is thought to relate to the major role that PUFAs play in brain development and function, which may be deficient in pregnancy due to either poor diet or the extra demands that the fetus makes on the maternal system.19 Most studies have assessed the effect of supplementing the diet with PUFAs to determine effects, but others have measured the actual levels of PUFAs in the fetus. The level of PUFAs can be measured at birth from the umbilicus, showing the level to which the fetus has been exposed. A study compared levels of two PUFAs, DHA and arachidonic acid (AA), at birth with neurodevelopment at age 9, when children have completed a major stage of mental development.20 Higher levels of DHA only were associated in boys, but not girls, with higher neurodevelopment at age 9.20 Thus, an optimum amount of fatty acids in the mother’s diet is critical to brain and neural development, but studies have also shown that if the amount is too high, the result may be poor development and aggression. Animal studies have shown that high fat consumption during pregnancy and lactation increases the risk of a number of health issues, as well as behavioral problems and disruption of brain development in offspring.21 Maternal prenatal and perinatal nutrition has been demonstrated to have profound effects on a number of genes governing neurological development. Epigenetic DNA  methylation is believed to be a major mechanism by which gene expression is impacted by high fat consumption. Experiments in mice have shown that these risks can be ameliorated by the addition of folate to the diet. 21 In other studies, increasing maternal levels of antioxidants, in particular vitamins A and C, as well as selenium and zinc, reduced the negative effects of stress during pregnancy and the consequent impacts on child behavior at 30 months.22 Many chemicals in our environment can have a major impact on our own health and are likely to have a much greater impact on a developing fetus. For example, prenatal exposure to lead has been linked with later criminal behavior.23 Bisphenol A (BPA) is a very common chemical found in a vast number of ordinary things we use every day, such as plastics, food packaging, thermal paper receipts, and can linings.24 Both human and animal studies have shown that BPA can disrupt hormonal and other physiological systems, alter steroid receptors, and disturb both the catecholaminergic and serotonergic signaling systems, which are important in the expression of attention deficit hyperactivity syndrome (ADHD).24 A  meta-analysis showed that exposure to BPA prenatally was a risk factor for developing hyperactivity, with a strong relationship seen in animals and a more moderate relationship in humans, 24 probably due to the fewer human studies available. No doubt, many other environmental pollutants known to cause health issues in humans also have a detrimental effect on the developing fetus.

Fetal alcohol spectrum disorder  169

Not only is the fetus at risk if the mother misses vital nutrients, but the risks increase if the mother has consumed damaging or undesirable substances during pregnancy, in particular, drugs, alcohol, and cigarettes. All of these can affect the health of the fetus; in the case of drugs, the child may be born addicted. Again, the child suffers the double burden of potential fetal damage and an addicted mother, with the consequent dysfunctional environment.

Fetal alcohol spectrum disorder Fetal alcohol spectrum disorder (FASD) has been known for centuries but was first officially described in 1968. The  term covers several diagnoses related to prenatal alcohol exposure, but all include structural and chemical brain damage, leading to a wide variety of health and behavioral problems that continue throughout the life span. Prenatal alcohol exposure is the number one cause of cognitive disability in the Western world, and yet, it is entirely preventable. FASD is caused by prenatal exposure to alcohol. This does not mean that the baby is born addicted to alcohol, as is seen in drug-addicted babies born to drug users. Instead, prenatal alcohol results in classic teratogenic effects. A teratogen is something that causes birth defects. German measles (rubella) is a well-known example of a teratogen, and pregnant women are now tested for the antibodies to rubella. The disease of rubella itself is not generally severe, so it does not normally have much effect on a person. However, if a woman contracts it when she is pregnant, the fetus may be miscarried, or if it survives, it has a very high chance of being disabled physically and mentally. This  has been recognized for decades, and in the past, if a child contracted the disease, all their friends would be invited to a German measles party to expose all the children to the mild virus and therefore give them immunity for life. In this way, the chances of a young woman developing the disease during pregnancy were greatly reduced, as she would be protected by the antibodies she had developed when she harmlessly contracted it as a child. Boys were also deliberately exposed, so they could not infect their sisters and lovers. Now, most of us are vaccinated against rubella as children and the days of German measles parties are long gone. However, protection against other, more insidious teratogens is not so simple. In the past, much attention was focused on the physically observable changes in the facial region of FASD-affected individuals, but recent work has shown that most people with FASD do not exhibit any sort of physical differences.25 As this was believed to be a marker for FASD, facial malformation was used to identify children with FASD in the past, but as this is now shown to occur only in a small percentage of affected individuals, it means that prevalence estimates in the past were vastly underestimated and the majority of individuals were not identified. In most cases, affected individuals are not identified until early or middle school, when externalizing behavior and problems become more evident.25 Recent estimates indicate that the prevalence of FASD ranges from 1.1% to 5% in the United States26 and from 1% to 4% in Canada.25 In  addition to the damage that alcohol does to the developing brain, FASD has specific effects on other biological systems that have been shown to be predisposers for crime. Prenatal exposure to alcohol affects the development of the neurotransmitter systems. It  seriously impairs the development of the system that produces and regulates serotonin, a very important neurotransmitter. Serotonin is a major factor in criminogenic acts, as further discussed in Chapter 9. Low serotonin has been consistently associated with aggression and impulsivity, and alcohol exposure before birth results in low serotonin levels, so this mechanism alone could be related to criminal acts. Prenatal exposure to alcohol also affects the other neurotransmitter systems: it impairs the dopamine producing system, the norepinephrine producing system, and the acetylcholine producing system, all of which are vital in life. There is evidence that impairment of any of these systems may result in antisocial behavior. Alcohol also affects the hormone systems as they develop, and it can affect the immune system and other functions such as hearing.

170  The prenatal environment and birth complications

FASD and the criminal justice system The organic brain damage caused by prenatal exposure to alcohol causes a wide variety of behavioral problems, including cognitive disabilities, hyperactivity, attention and perceptual deficits, peer and communication problems, motor and sensory abnormalities, and problems with learning from mistakes, and the manifestation of these problems varies widely in people affected.25 In a study of adolescents and adults, IQ ranged from 20 (severe intellectual disability) to 105 (average range), with academic level ranging from second to fourth grade.27 Many studies have shown a high incidence of FASD-affected individuals involved in the criminal justice system, suggesting a strong link between FASD and criminal behavior.28 Youths in Canada with FASD have been estimated to be 19 times more likely to be imprisoned than healthy youths,29 and some studies show rates as high as 40 times.30 Costs associated with FASD in the Canadian criminal justice system have been estimated at Can$356.2 million for adults and Can$17.5 ­million for juveniles. To put this in perspective, this is 8.6% of the total cost of Canadian corrections and 50% of the cost of correctional services for offences committed while under the influence of alcohol.28 There is also increased risk of sexual offending in individuals with FASD, as reduction in ­cognitive abilities means that affected individuals do not understand social boundaries and are frequently impulsive, with no thought or understanding of consequences. This is a particular concern in adolescence, when sexual urges begin in a person with the body of an adult and the cognitive awareness of a child and there is a lack of ability to understand social mores.31 In a US study of 415 FASD patients, 60% had problems with the criminal justice system, 50% had been imprisoned or placed in a psychiatric institution, 49% had repeatedly exhibited inappropriate sexual behaviors, and 35% had drug or alcohol problems.32 When people commit crimes or break rules or social boundaries, it is assumed they are doing this deliberately, with malice aforethought, but individuals with FASD are, in many cases, unable to understand the consequences of their actions or comprehend causality, and a lack of diagnosis together with a lack of a full understanding of FASD results in conflict with the criminal justice system and high incarceration rates.31 The effects of incarceration compound the underlying effects of the disorder, as FASD individuals are frequently victimized and are highly suggestible, which may lead to association with delinquent peer groups, all of which may lead to increased recidivism.31 It has been estimated that FASD is 95% underreported,30 and this is supported by the fact that in a survey of over 3 million prison inmates in the United States, only one had been actually diagnosed with FASD,33 despite the very large percentages of inmates identified in targeted research.31 There are several factors at work with FASD children that may predispose them to criminogenic behavior. The  children are prenatally exposed, so that is biological. However, they are probably also born into dysfunctional families. Although we as a society are aware of the risks of drinking when pregnant, the prevalence of maternal consumption of alcohol during pregnancy ranges from 14% in Canada to 89% in Ireland, with binge drinking during pregnancy—which causes the most damage—ranging from 7.4% to 9.5%.28 As we know the risks, and they are posted in every bar and liquor store, women who continue to drink are either too addicted to stop drinking or are not well invested in the health of their fetus. In either of these cases, the resulting parenting situation is likely to be dysfunctional, adding an environmental insult to compound the offspring’s biological problems. Many parents of FASD children also have FASD themselves, which means they probably have low academic outcomes, dysfunctional behavior, and addictions, among other deficits, which contribute to intergenerational poor parenting outcomes. Moreover, a child who has cognitive and behavioral difficulties and problems with social interaction is more likely to have a difficult time at school with peers and bullies, which results in additional environmental and social pressures. FASD youth are also more likely to be exposed to other risk factors, such as placement in foster homes, having low self-control and poor self-identity, and having an earlier onset of first alcohol use, than non-FASD youth.34

Fetal alcohol spectrum disorder  171

Gene × environment interactions in FASD Although the drinking of alcohol while pregnant is an environmental insult, it causes severe biological impairments. However, there is growing evidence of various genetic mechanisms at work, as well as gene × environment interactions, which also impact the development of FASD. Not all children exposed prenatally to alcohol will develop FASD, or they may only develop one of the lesser diagnoses within the syndrome, suggesting that factors other than alcohol exposure are involved. In a twin study, monozygotic (MZ) twins were 100% concordant for FASD, whereas dizygotic (DZ) twins were 64% concordant, suggesting a genetic effect. Moreover, severity varied between twins, suggesting the possibility of genetic protective factors, 35 although chorionicity (whether they share a placenta or not), differential blood supply, and position in the womb could also have contributed. Animal studies have shown that there are both fetal and maternal genes that impact the potential development of FASD. 36 Alcohol dehydrogenase is the main enzyme that breaks down alcohol into its metabolites, and it is well known that different ancestral groups, sexes, and individuals have varying levels of this enzyme. Not surprisingly, the genes governing the production of alcohol dehydrogenase have been implicated as both risk and protective factors for FASD, depending on the variants, and other candidate genes, such as those involved in neurotransmitter systems, have also been suggested. 36 A number of genes are involved in alcohol metabolism, and polymorphisms (different alleles) in these genes could impact the level of alcohol exposure to the fetus. Some metabolites of alcohol, such as aldehyde, have also been shown to be teratogenic.37 It is suggested that many genes may interact with alcohol, and certain alleles are more sensitive to the teratogenic effects of alcohol, making some fetuses more susceptible. The number of genes and their alleles that could potentially be involved would explain the wide variety of manifestations of FASD. 37

Intervention The solution to FASD seems obvious: women who drink should stop drinking when they are pregnant, and those who cannot stop drinking should not get pregnant. But of course, it is not that easy. If it were, FASD would not exist. FASD remains one of the major causes of intellectual disability today.28 In some cases, a woman may not even be aware that she is pregnant, and thus not realize that she should not consume alcohol, until the damage has been done. It is estimated that almost 50% of pregnancies are unplanned, and as a large percentage (over 75%) of women report drinking alcohol, it is likely that normal drinking patterns would continue until the pregnancy is detected.25 Resources need to focus not only on the pregnant mother but also on the already affected developing child. FASD children are often not identified until the school years, and many children are not diagnosed and can easily slip through the cracks, receiving no help. Of course, it is a doubleedged sword, because children identified as FASD and allocated extra resources and assistance are then labeled as such, potentially creating further problems. A fine line must be drawn that allows children the help they need, without stigmatization. Individuals with FASD are easily led and bow to peer pressure, and if they make poor peer choices, they can easily be led into crime.31 However, the reverse is also true, and good peer relationships and mentorship may assist such individuals. Many children with FASD are raised in foster or government homes,32 but evidence has shown that children who are raised in healthy, stable environments, or who are identified early and receive support, are two to four times less likely to develop the myriad adverse outcomes normally seen in FASD.32 Despite this, many people with severe FASD need much more than a stable home. Children placed in foster care are often moved from home to home, often despite having good foster parents as they do not have the skill sets to help such children. In Oregon, a 3-year study on FASD children in foster care involved developing accommodations for the children in school, home, health and social services and involved

172  The prenatal environment and birth complications

a large number of professionals and carers who were trained in rethinking the meaning of the child’s behaviors.30 Although the study was short term and only involved a small sample, remarkable changes were seen in children, with one child going from a state where he had to be put in four-point physical restraints several times a day to being in a public school on an individualized education plan.30 Recommendations were also made for adults with FASD to help them succeed, such as using experiential techniques in learning, teaching time management and organization through the use of lists, use of tablets, supporters who act as external brains, and bank programs that assist in financial management.30 Such programs show that support and peer mentorship could greatly reduce the risk of criminal involvement, not only helping the individuals but also reducing the heavy financial and emotional burden placed on society.

Maternal smoking Although it is very well recognized by most people today that pregnant women should not drink, the risks associated with smoking are less well understood. We know that smoking is bad for us and causes many health risks, but these health risks also extend to the unborn child; not only can smoking cause health problems in the child, but it has also been associated with a predisposition for hyperactivity, ADHD, conduct disorder (CD), aggression, and criminogenic behavior.38 It affects the development of the neurotransmitter systems, including cholinergic, dopaminergic, and serotonergic systems, all of which impact behavior and also relate to anxiety and depression.39 Smoking has been shown to cause structural changes in the fetus, such as reduced growth of the head. In animal studies, maternal exposure to nicotine has been shown to affect learning ability in adult mice.40 Many studies have shown a link between maternal smoking and later antisocial behavior. A  ­meta-analysis of studies showed a clear link between prenatal smoking and offspring health, educational outcomes, and risk of later risky and criminal activity.41 Clearly smoking during ­pregnancy has negative effects on the offspring, including health issues such as low birth weight and ­prematurity, which on their own have been shown to be risk factors for later antisocial ­behavior, and many studies indicate increased risk for antisocial and criminogenic behaviors. But is the actual maternal smoking causal, and if so, how?

Smoking’s potential effects on a fetus There are several ways in which smoking might impact offspring behavior, biologically, environmentally, and in an interaction between the two. Direct health risks

First, innumerable direct health risks with smoking are well known for the smoker, but nicotine and other products of smoking pass through the placenta and cross the blood–brain barrier and are found at levels 15% higher in the fetus than in the mother.42 Many of the chemicals in cigarette smoke are ­teratogenic and bioaccumulate in the fetus, and animal studies have shown that they have direct effects on brain structure and function. Thus, cigarette smoke may directly impact cognitive ­development, resulting in later behavioral problems.42 In a large study of almost 2000 women in France, maternal smoking was found to have a direct effect on increased risk of hyperactivity and inattention in ­offspring.43 To account for problems seen in previous studies, the authors used negative ­controls (paternal smoking); looked at smoking before, during, and after pregnancy; fully assessed the child’s social and family environment; used inverse probability weights to account for covariates; and ­studied several different aspects of the child’s behavior up to age 5.43 The relationship was still strong even after accounting for all these variables. In a study of neural development in adolescents,

Maternal smoking  173

MRI showed that several areas of the cerebral cortex were thinner in adolescents prenatally exposed to cigarette smoke than in matched controls, and this was more evident in females.44 The cerebral ­cortex is the surface layer of gray matter of the cerebrum (which includes the frontal, parietal, t­ emporal, and occipital lobes) and is important in coordinating motor and sensory information. Poor parenting and mother–child attachment

Second, as the effects of smoking on health are well known, it can be assumed that most women would attempt to abstain from smoking during pregnancy, and if they do not, it suggests a potential lack of concern for their own health and that of their fetus.43 If so, it is possible that such women would not engage in good prenatal health and may not bond well to their child, resulting in the risk of indifferent parenting. A meta-analysis showed that women who quit smoking during pregnancy had a more secure attachment to the child and exhibited better parenting skills.38 Heavy smoking during pregnancy has been associated with reduced bonding with the fetus and the infant,43 and in its turn, insecure attachment has been shown to have a clear association with development of ADHD.45 Prenatal smoking has also been implicated as a risk factor for developing ADHD.46 This is not surprising, as ADHD has been linked to disruptions in the dopaminergic system, and nicotinic receptors modulate dopaminergic activity.46 This can be part of a vicious circle, as ADHD has a genetic component and mothers with ADHD may have difficulty bonding with their infant, meaning there are genetic, biological (damage prenatally due to smoking), and environmental insults. Nicotine addiction

Third, it may be that women who smoke attempt to quit but are unable to, as nicotine is strongly addictive. If so, such addictive behavior may be partially genetic and passed to the offspring, together with behavioral issues. A failure to stop smoking has been shown to be related to m ­ aternal CD,43 and smoking during pregnancy may be an indicator of psychological or biological risk ­factors.43 Moreover, women who are unable to quit smoking during pregnancy are unlikely to quit during lactation and subsequent child rearing, exposing the child to secondhand smoke. A study on Inuit children found an association with maternal smoking and increased externalizing behaviors and inattention, as well as symptoms of ADHD, CD, and oppositional defiant order, but the authors pointed out that it was hard to separate the effects of prenatal and postnatal smoking.47 Maternal stress

Fourth, continuing to smoke prenatally may be an indicator of maternal stress or psychological problems, which may also impact the fetus and the growing child.43 Gene × environment interactions

Fifth, smoking might be a risk factor that regulates other risk factors, either biological or environmental, which would mean that prenatal smoking is only a problem when it coexists with other risk factors.42 This fits with the hypotheses of Boardman and colleagues, who suggested that biological risks only predict criminal or violent behavior when there is an additional environmental trigger.48 This speaks to the work by Caspi and colleagues on the interactions between certain genotypes, such as a variant of the MAOA gene, and the environment, in that case, violent physical abuse in the child.49 More and more, we are seeing that biological predisposers are only a concern in the presence of environmental adversity, which is not only important for understanding the systems involved but extremely ­important in developing programs to reduce environmental triggers. In a twin study, genetic risk factors were shown to have a significant impact on externalizing behaviors, but only in female offspring, and once genetic factors were excluded, maternal smoking did not have a significant effect on offspring behavior; however, the genetic risk was greatest in offspring exposed to prenatal ­smoking, indicating once again a gene × environment interaction.42 It is interesting that this result was only positive in females, but other research has suggested that some gene × environment interactions may be gendered.42

174  The prenatal environment and birth complications

Epigenetic effects

Finally, a relatively large percentage of fetuses is exposed prenatally to smoke, with up to 20% of women continuing to smoke during pregnancy in North America and up to 50% ­continuing to do so in some countries.44 However, only a small proportion of the offspring suffer from health and behavioral issues in later life. It  has been suggested that this can be explained by ­epigenetic effects on the fetus’ changing gene expression. 50 Studies have shown that not  only are the expression of over 600 genes and the DNA methylation of over 1000 CpG dinucleotides (regions of DNA) changed significantly in smokers, but smoking in the mother also affects ­placenta methylation, which obviously has a major impact on the health, development, and birth weight of the fetus. 50

Maternal age Many studies have shown a link between maternal age and expression of externalizing behaviors in offspring, with a strong linear relationship decreasing as maternal age increases. 51 However, no causal relationship has been fully identified. Moreover, there are many studies that have linked low birth weight to later antisocial and criminal activity, but again, no causal relationship has been found. It has been suggested that the purported link between low birth weight and negative outcomes is an interaction between low birth weight and some form of adversity and that low birth weight is disadvantageous only when it is triggered by adversity. In a study in Taiwan of over 700 000 males, a comparison of criminal convictions with extensive birth data, including birth weight, parents’ age and education, mother’s marital status, and number of siblings and birth order, indicated that low birth weight and criminality were correlated only when the maternal age at birth was below 18 or over 40, which the authors interpreted as indicating a disadvantage, which, combined with low birth weight, increased the risk of committing a violent crime. They suggested that children born to very young or older mothers may be unwanted and unanticipated, which could result in greater maternal rejection, which has been shown to increase risk.11 However, there is no actual evidence of this, and the children may have been deeply wanted and valued. In  a large study of almost 4000  pregnancies in Rhode Island, extensive data were collected from the perinatal time through age 7  and compared with later arrest records up to age 40.10 Babies born with low birth weight had an increased risk of adult arrest but only if they were born to ­adolescent mothers. The  authors looked at all possible confounding parameters, such as ­prematurity, race, maternal smoking, marital status, and socioeconomic status (SES), and the relationship stood. Babies born to mothers over 18 years old did not show this link, despite the low birth weight, suggesting a protective effect of maternal maturity. This study again h ­ ighlights the relationship between a biological effect (low birth weight) and the environment (maternal age), with the deleterious effects of low birth weight being counteracted or even completely ameliorated by maternal age. Many studies have, therefore, indicated that offspring of teenage mothers are at greater risk for exhibiting risky behavior, but no causal relationship has been shown. Two hypotheses have been suggested to explain the relationship between teenage mothers and risk of crime in adult offspring: the social influence hypothesis and the social selection hypothesis. 52(p18) The  social influence hypothesis suggests that having a baby while so young affects the normal development of the mother, resulting in social and economic stress, such as lack of education and poverty, which is likely to lead to poor parenting. 52 It is also probable that a teenage mother may be raising her child without a partner. It has also been shown that teenage mothers have higher levels of antisocial behavior, which could mean that there is a genetic link to negative behavior in the offspring. It also suggests that a young woman who engages in risky sexual behavior, such as unprotected sex, is more likely to engage in other forms of risky behavior. 53 Alternatively, the social selection hypothesis suggests that there are biological, psychological, and social factors

Maternal stress  175

that may make some women more at risk for engaging in risky sexual behavior and becoming pregnant. Teenage mothers often come from an underprivileged home environment, with poor nutrition, poverty, dangerous communities, and low educational achievement, so they are at more risk than those from advantaged families. 52 In a large Swedish sibling and twin study, the authors found that offspring of teenage mothers were more likely to have criminal convictions than those born to slightly older mothers; they concluded that their data supported the social influence hypothesis, as the relationship between maternal age and offspring criminal convictions were independently associated. 52 They compared the antisocial behavior of the offspring of siblings and twins in situations where one sibling, MZ twin, or DZ twin had given birth in adolescence and the other had delayed reproduction, and the relationship between maternal age and criminal convictions held true. This relationship was similar in both MZ and DZ twins, suggesting a non-genetic basis. Understanding the impact of maternal age on later risk of criminality in the offspring makes it clear that additional health care support, prenatal services, and postnatal nurse visitations may help support a young mother and therefore support her offspring, reducing the risk of later antisocial behavior. Such studies help health care providers prioritize resources.

Maternal stress It is well known that stress has a very negative impact on our bodies, mood, and welfare, so it makes sense that maternal stress during pregnancy could also impact the fetus. Many animal studies have been conducted to determine the effects of stress during pregnancy, in which pregnant animals are randomly assigned to stressful or non-stressful conditions. Rodent studies have shown that in stressful situations, maternal glucocorticoids (a type of steroidal hormone) can pass through the placenta and alter the hypothalamic-pituitary-adrenal axis, 54 a major ­neuroendocrine system that regulates many body processes, including the immune system, mood, emotions, and, perhaps most importantly in this case, how a body processes stress. Glucocorticoids have also been shown to alter the brain development of the fetus, and even mild stress in non-human primates has been shown to impact many aspects of the immune system, 55 as well as the behavioral and metabolic systems.54 Studies in mice showed that prenatal stress, such as restraint, cold, immersion in water, and electric shock, resulted in changes in the serotonin and glutamate ­systems in adult offspring but not in adolescent or prepubertal offspring and resulted in behaviors similar to those seen in schizophrenia.56 Although animal studies are very useful in projecting the probable impact of stress on humans, determining the actual impact of stress on pregnant women is difficult, as it would obviously be unethical to conduct experiments subjecting women to stress. However, it is possible to look ­retrospectively at people who were born during and just after known stressors have occurred. The 1998 ice storm in Quebec was one of Canada’s worst natural disasters, with heavy ice destroying power lines, resulting in power failures that lasted up to 6 weeks, leaving people without heat or services during an unusually frigid winter. Women who were pregnant during the ice storm or conceived shortly afterwards were recruited, and they and their offspring were examined over the subsequent 13 years. Level of stress experienced by the mother during the storm was correlated with epigenetic effects caused by DNA methylation in almost 1000 genes in the offspring, meaning that the normal expression of these genes was changed.54 The ice storm research looked at a single, albeit prolonged, stress event. Other work has considered effects of one-time major stressors, such as bombings, explosions, and even 9/11, and has shown that stress, particularly in the first trimester, can result in low birth weight and preterm babies.57 Although this finding is interesting, it has been suggested that such catastrophic single events might impact more than just the mother’s psychological well-being: they might impact access to health care or even increased pollution, making it hard to isolate stress as a causal agent.57 Therefore, other studies have looked at the effects that sustained violence may have on

176  The prenatal environment and birth complications

the prenatal environment. A modest relationship was seen between low birth weight in infants born during or just after the Second Intifada in 2000, which resulted in over 4000 Palestinian deaths, although the authors admit that the effect may not have been limited to psychological stress, as the conflict also resulted in poor nutrition during this time and in other adverse conditions, such as limited access to prenatal care and lack of transit.58 In Brazil, exposure to violence, as measured by homicide rates in different communities, confirmed that exposure to violence during the first trimester of pregnancy led to a small increase in the risk of preterm and low-birth-weight babies. The effects were consistent in small rural communities, where homicides were rarer and therefore perhaps more impactful, and in large urban communities, where homicide was commonplace. The effects were greatest in children of lower SES with mothers with low levels of education, suggesting that increased stress interacts with preexisting risk factors.57 The  authors comment, however, that increased stress may impact other risky activities, such as alcohol consumption and smoking, which were not measured.57 Interestingly, in Mexico, a similar study of effect of homicide rates on prenatal environment in the first trimester indicated a reduced risk of low birth weight and an increase in birth weight with increased stress.59 This surprising result appeared to be due to an increase in healthy behavior by the mother, resulting in greater prenatal care, although the results were significant only in women of low SES living in urban areas. The authors suggested that the increased anxiety due to proximal violence affected women of lower SES more than those of higher social standing and promoted a desire to protect their fetus. However, only women in urban centers had easy access to prenatal health care, and hence, the improvement was not seen in rural areas.59 Overall, maternal depression and anxiety have both pre- and postnatal negative effects on child misbehavior, with depression having a stronger impact on externalizing behaviors and verbal IQ and anxiety having impact on more internalizing behaviors.60 A study of 56 women during ­pregnancy, 1 week, and 6 weeks after birth compared stress levels with level of birth complications and showed that, as expected, mothers who had undergone greater birth complications had higher stress, but these were negated by good family coping skills, including accepting social assistance.61 Although it is ­difficult to eliminate stress from our lives, it is clear that proactive prenatal care, together with support postnatally, can greatly improve fetal health and reduce risks caused by stress-related factors.

Fetal maldevelopment and minor physical anomalies Some children are born with what are termed minor physical anomalies, or MPAs. These are very small defects that can occur during pregnancy, and several studies have shown that they can relate to later criminal behavior.5,62 MPAs are thought to result from some form of fetal maldevelopment that occurs in about the third month of pregnancy, a period when the brain and many ­neurological structures are being formed. MPAs are minor physical defects, such as low-seated ears, ear lobes that attach to the skull, a slightly curved fifth finger, a single crease in the palm instead of the two that most people have (incidentally, that is also seen in children with Down syndrome), gaps between the first and second toes, and so on.63 Early research identified only 18, but today, over 50 are studied.64 These defects are physically minor and do not affect the attractiveness of the child. These defects result from some disturbance in the development of the fetus. For example, when the ears develop, they begin low down on the neck and slowly drift into the usual position. If they stop moving and remain low seated, it is suggestive of some disturbance in fetal development during this time.5 MPAs certainly do not cause crime, but they have been found to be very useful in pinpointing timing and acting as indices of some trauma or damage during the pregnancy; therefore, they may also be indicative of other damage that is not visible but is perhaps more serious, such as neuro­ developmental impairment that could affect behavior. Something that can change the development of the ears could certainly also damage such vital things as the central nervous system.

Fetal maldevelopment and minor physical anomalies  177

Higher numbers of minor physical anomalies have been recorded in people with schizophrenia,64 autism,65 and hyperactivity,66 as well as general aggressive behavior,62 violent offending,5 and violent recidivism.67 When birth complications are also considered, the risk of violence increases, as this would add a further insult to the fetus.68 Interestingly, in this study, MPAs only predicted for violent offending when they were coupled with an unstable home life. If a child with MPAs had a strong, supportive, stable home, then the effects of the neurodevelopmental damage were negated, and the child grew up non-criminal. We find this result again and again when we study the biology of crime—not so much the causes of crime but, more importantly, the protective factors. In a longitudinal Canadian study, 170 adolescent males from a lower-income, inner-city region were examined for the presence or absence of minor physical anomalies, family adversity, and rates of delinquency.62 Delinquency was measured by self-reports, teachers’ reports, and official crime records. Family adversity included parents’ occupational socioeconomic level, parents’ age at birth of first child, parents’ educational levels, and whether the child lived with both parents. The  researchers found that the total number of MPAs showed a significant positive correlation with violent delinquency in adolescence. In particular, MPAs in the mouth showed the highest correlation with violence. When MPAs of the mouth were removed from the analysis, MPAs in other regions of the body no longer predicted for violence, although family adversity still did.62 There was no correlation between MPAs and non-violent delinquency. The authors suggested two reasons for the high correlation between number of MPAs in the mouth and violence. First, as has been frequently suggested before, MPAs are probably a marker that indicates when deeper and more profound damage occurred in utero. A fetus develops from a single-celled zygote to a fully functional baby in just 9 short months. The  central nervous ­system and organs in the fetus develop in a very preset, hierarchical way. Therefore, any insult that interrupts this process will have the greatest effect on the system that is at its critical period of ­development at that time. The critical period for the development of MPAs in the palate of the mouth starts at the ninth week of gestation.62 So, the authors suggest that any trauma, whether physical or chemical, that occurs at this time results in neurological damage that could predispose to violence and that also causes MPAs in the mouth region. Second, the authors suggest that even small irregularities can affect the child in other ways, such as affecting the sucking reflex required to feed as a young baby. This in turn not only may have a nutritional effect but may also affect the mother–child bond. Infants who have feeding problems have been reported to have other ­behavioral problems.62 In a longitudinal study in Taiwan, 184 preschool children were assessed for MPAs and then tested with a number of psychiatric assessment scales 4 years later. An association was found between MPAs and several mental health issues, such as anxiety, depression, and paranoia, and the authors suggest that they should be used as easily identifiable risk factors that can be used to identify persons at risk, so that early interventions can be used to ameliorate the risks.69 MPAs are commonly noted in persons with schizophrenia and are considered to be risk markers for the disease.70 As schizophrenia affects a range of brain functions in a variety of regions, MPAs are useful markers to help understand the neurodevelopmental impairment underlying the disease, including timing and type of insult. Although known to be found much more commonly in persons with schizophrenia than healthy persons, they are also found to be more common in first-degree relatives of persons with schizophrenia than in healthy controls, particularly in the head and facial region.64 This was confirmed in a recent meta-analysis of a number of studies from 1980 to 2017, but effect sizes varied, which suggested that the risk of schizophrenia was not necessarily neurodevelopmental in all affected persons and that protective or resilience factors might be shielding close relatives with neurodevelopmental damage from developing schizophrenia.71 Although there are many etiologies for schizophrenia, aberrant neurodevelopment is considered to be a major causal agent. The manifestations of schizophrenia vary dramatically, with a wide variety of symptoms, which can, but do not always, include aggression and violence. Neuroimaging studies of schizophrenic persons consistently show brain abnormalities in both the temporal and frontal lobes in those who were aggressive, but not in those who did not exhibit aggression.72 In a study of

178  The prenatal environment and birth complications

schizophrenic patients who had committed or attempted to commit homicide, MPAs were much more common in those in the homicide group than in those who had never attempted homicide or in healthy people.73 This supports the idea that neurodevelopment abnormalities, probably in the frontal and temporal lobes, are linked to the etiology of the more severe forms of schizophrenia and that underlying serious neurodevelopmental damage can be indicated by MPAs, particularly in the face and mouth region.73 Earlier studies have shown MPAs have been linked to aggression and impulsivity in persons without any form of psychoses,74 suggesting that the brain damage is responsible for the behavior and is not specifically linked to schizophrenia. There is a great deal of interest in studying MPAs, as they are a very easy and non-invasive way to assess and provide information about the timing and type of neurodevelopmental insult a fetus has received in the womb. For example, growing evidence suggests that there may be a neurodevelopmental basis for some forms of pedophilia, and increased levels of MPAs in pedophiles have been used to support this hypothesis and to suggest that hypoxia, due to a range of factors, may be partially responsible.75

Other birth-related difficulties Twin births It is now clear that birth trauma can result in damage to infants. What effect might this fact have on some of the broader studies in this field? Remember that many of the studies on genetic effects on behavior look at twins, where the likelihood of birth trauma is greater because the birth process is much more complex. Mortality during birth is four times higher in twins than in singletons, and twin pregnancies have higher rates of miscarriage and higher risk of hypertension, anemia, hemorrhage, premature birth, low birth weight, structural abnormalities, and growth restriction in utero.76 Mothers of twins have been shown to suffer from greater parenting stress than mothers of singletons, and having two babies at once can reduce the mother’s individual attention to each, which can cause a decreased verbal environment.77 Moreover, twins are often born preterm and spend time in incubators, so that in the first few days or weeks after birth, they have much less human contact than a normally birthed baby. This means that they also have less contact with their mother at a time when the main bonding takes place. In a study of very preterm (VP) (less than 32 weeks gestational age) twins compared with matched mothers of VP singletons, mother–infant interaction was studied while in the hospital and at home at age 3 months, using a variety of measurement scales. The  results suggested that there was some evidence that VP twins had lower-­ quality interactions with their mother than similar VP singletons. Mothers of twins scored less on the responsivity scale, which measures sensitivity and responsivity to offspring, and exhibited higher levels of parenting stress than mothers of VP singletons. Also, VP twins did not provide as clear signals and were less responsive to their mothers one to one.77 This suggests the need for higher intervention and support in situations of VP twins, such as instruction on providing better one-to-one interactions when dealing with two infants who are less responsive, as well as other forms of assistance to allow parents to spend more time with their babies.77 Some twin studies have included birth complications as a factor, but can we measure the effect? Could the birth complications have affected one twin more than the other? Also, are birth complications more likely in monozygotic twins than dizygotic twins? Chorionicity refers to the number of placentas. DZ twins each have their own placenta, whereas 70% of MZ twins share a single placenta. Risks are higher for monochorionic twins.76 Studies have shown a significant difference between twin and non-twin behavior problems. A study showed that twins showed small but ­consistently higher levels of problem behaviors than non-twins and suggested that these results may be the result of perinatal injury, which has a higher incidence in the twin population.78 It could also be the result of the environment in which twins grow up, and the researchers have admitted that they did not take regional differences or method of study into account.

Other birth-related difficulties  179

Another point to consider in multiple births is position in the womb in relation to the co-twin and hormone exposure. In animals that have litters, it is known that hormones in the uterus and position with respect to male and female littermates influence sexually related adult behavior. There may be a similar effect in twins. Sensation-seeking was studied in 422 twin pairs, including 51 oppositesex twins. Previous work has shown that sensation-seeking is higher in males. This study found that girls who had a male twin had higher levels of sensation-seeking, although the opposite was not true of male twins, which does suggest some sort of in utero hormonal influence on later behavioral development.79

Maternal rejection Early maternal rejection and separation has been shown in animal studies to cause morphological changes in brain structure that are often not identifiable until adulthood.80 To determine whether the normal development and changes that occur in the brain of a young animal as it matures are affected by maternal separation, baby rats were separated from their mothers for four hours a day between the ages of 2 and 20 days.80 In comparison with baby rats that were not separated from their mothers, the separated baby rats showed delayed development of the hippocampus, a region of the temporal lobe and limbic system that is involved in learning, memory, and emotion.80 Birth defects themselves may be implicated in early maternal rejection and in the way the child is subsequently raised. For example, are birth complications likely to increase maternal rejection, especially when the child may not have been with the mother from birth because of medical needs? Research has shown that birth complications interact with early maternal rejection in predisposing individuals to violence at age 18.81 The study considered more than 4000 male births in Denmark and assessed birth complications and maternal rejection before the age of 1 year. Criminal status was determined at age 17–19 years. A significant interaction was seen between birth complications and maternal rejection.81 As observed in previous studies, it is the interaction between both birth complications and early maternal rejection that was the predisposing factor, rather than either one on its own. However, this does not tell us whether the predisposition for violence relates to the fact that the mother rejected the child as an infant, or whether she rejected the child because of her own trauma or because the baby was not in her arms for the first days after birth. This study is very robust, as it has been replicated in Canada, the United States, Finland, and Sweden.82 A follow-up study looked at the history of criminal offenses by age 34 in the same group. When the age was extended, the number of subjects with violent offenses increased threefold. In the original study, 145 were classified as violent, 540 as non-violent, and 3584 as non-criminals. In the new study, 466 were violent, 844 were non-violent, and 2959 were non-criminals. The study showed a significant interaction between birth complications and early rejection in predicting violence at 34 years of age. The components of maternal rejection that seemed most important were attempts to abort the fetus and institutionalization of the infant; not wanting the pregnancy did not appear to be an issue. Maternal psychiatric illness had no effect on the interaction. Offspring’s perception of maternal rejection has been linked to poor academic performance in school83,84 and poor psychosocial adjustment and results in higher levels of bullying mediated by depression, but interestingly, these effects were ameliorated by paternal acceptance.85 There are obviously many reasons why a mother might reject her child, but there is a great deal of research that could be utilized to assist maternal bonding. The mother might be allowed to get closer to the baby, even if the baby is in an incubator. She could hold and feed her child rather than having a nurse do that. Not so long ago, babies were swept away from their mother, and both were kept in the hospital for up to 2 weeks, with the baby brought out only for feedings. Now, we recognize that those first days are vital for bonding, and babies stay with their mothers from the beginning. Fathers are allowed in the delivery room. It is all important to the bonding process on all sides. The baby needs to bond, but for the baby’s sake, the parents must also bond. However, when babies are sick, premature (as twins often are), or have birth trauma, they are kept under special

180  The prenatal environment and birth complications

care, and their mothers may be squeezed out of the bonding process. Perhaps changes that would allow them to be closer to their babies more often would help with problems caused both by maternal rejection and by birth trauma.

Ethical issues There  are major ethical issues to be considered in this area. It  is clear from many of the factors ­covered here that maternal health and environment can have a major impact on the developing fetus and the subsequent life of the child and adult. Does a woman have complete rights over her body and her actions despite the fact that these actions may damage her unborn child? Could a woman be forced to stop using drugs or alcohol to protect her fetus? Perhaps even incarcerated “for her baby’s own good”? If so, why stop there? Could we not then force her to stop smoking? If we accept that maternal age impacts the fetus, could we mandate that women only give birth between certain ages? If this sounds far-fetched and ridiculous, think again. Women in the United States have been arrested, incarcerated, and had interventions forced on them for decades. A case that stands out for me was one presented at a conference I attended years ago. A very pregnant woman was abused by her partner, who beat her, kicked her in the abdomen, and pushed her down the stairs. She was taken by ambulance to the hospital, where she lost the baby. Blood tests showed she had had a small amount of alcohol to drink. She was charged and convicted in the death of her unborn child for the alcohol use, despite it clearly having had no role in her losing the baby. The partner was not even arrested.

Criminalization of pregnant women in the United States During America’s “war on drugs,” concerns arose about a potential crack cocaine “epidemic,” which would, so people believed, result in a vast number of damaged and addicted babies.86 The  issue became a major target for politicians, lawmakers, and certain health care providers, who told the public that there was a crack epidemic and that something had to be done to protect the helpless babies. This created a moral panic.86 In the late 1980s, South Carolina, a major drug corridor with a high poverty rate, introduced policies to charge pregnant women with dealing drugs to their unborn children. 86 The Medical University of South Carolina developed a testing program. Initial screening of pregnant women was done without their knowledge of the consequences of the test and, interestingly, almost entirely included women who did not  have private medical coverage. An interagency policy was developed that involved testing women and babies, which could be considered a normal medical practice, and then providing these private medical details to police, which is against a person’s rights. Frequently, women were arrested, shackled, and jailed immediately after giving birth, while they were weak and still bleeding from birth and often vomiting. All were African American, except one white woman who had an African American partner.86 Some were imprisoned, but most were sent for treatment and counseling, which might have been somewhat helpful, even though enforced, except that there were very few treatment facilities available and women were left with no support and yet were still prosecuted for not taking treatment, a vicious circle. The majority of women were of color, poor, using crack, and had suffered from physical and sexual abuse.86 Moreover, the facilities had no accepted protocols to treat women with crack addiction, and even women who were referred to a clinic were unable to get transport or child care to attend.86 In 1992, Cornelia Whitner was convicted of giving cocaine metabolites to her fetus. She was sentenced to 8 years but appealed. On appeal, it was decided that a viable fetus had the same rights as a child, establishing this precedent in South Carolina.86,87 Of course, such rulings, rather than deterring women from abusing substances, resulted in deterring them from seeking help for their addictions, or even much-needed

Ethical issues  181

normal prenatal help. It also created a mistrust between women and their medical care p ­ roviders.88 In many cases, women avoided prenatal care after being threatened, avoided hospital births and prenatal care, or, in one case, literally fled while in labor to another hospital to avoid enforced cesarean section.89 A big concern when looking at these cases is that it was not just the legal community that ­literally took the law into their own hands, deliberately misinterpreting existing laws that were never designed to restrict a woman’s liberty, but also health care professionals. Medical professionals broke their own medical rules and ethics not only by not protecting the private medical i­ nformation of their patients but also by blatantly collaborating with police in obtaining medical information for the single purpose of incriminating their patients, breaking all laws of ­confidentiality and patient–doctor privilege, as well as state laws. In 112 cases, information was given to the police via members of the medical profession.89 Following South Carolina, many states began to develop similar drug policies, but these were put on hold when 10 of the women who had been previously arrested filed a civil suit90 for “discrimination and violation of their constitutional rights under the Fourth and Fourteenth Amendments.”86(p695) In this precedent-setting case, the judge ruled that if the tests had been done for medical ­purposes and the results kept confidential, they would have been constitutional, but as they were done purely in order to provide the police with evidence to incriminate their own patients, they violated the Fourth Amendment of the US Constitution.86 This  major decision strengthened the rights of women against unreasonable searches. Later, women in 24 of 25 states successfully challenged their charges based on this ruling. However, in South Carolina, even now, women can still be charged with child abuse if they cause any harm to a fetus in the last trimester of pregnancy.86 Similar cases pertain to the use of alcohol during pregnancy, with state policies ranging from providing warnings in liquor stores and restaurant washrooms of the risk of using alcohol when pregnant, which we are all familiar with, to priority treatment and mandatory reporting, all the way to civil commitment, which includes mandatory involuntary incarceration.88 A US study looked at cases between 1973 to 2005 in which women were criminalized and their liberty either threatened or actually taken away due to their being pregnant.89 Over 400 cases were identified across 44 states and District of Columbia (DC), but due to difficulty in obtaining some records, this is considered to be a gross underestimation.89 Almost all the women were of low SES, and 59% were women of color, including 52% African Americans, with racial differences being much higher in some states, particularly in the South. In most cases, there was no actual evidence of harm to the fetus and no or very little scientific evidence ever presented. In some cases, people with no scientific background were allowed to testify to scientific evidence. Some case examples

The first case of a woman being arrested and prosecuted for prenatal drug use in the United States was that of Pamela Rae Stewart, who, in 1986, tested positive for methamphetamines while ­giving birth. The  child was born with brain damage and subsequently died at 6  weeks. Stewart was charged with providing drugs to a minor—her unborn child.86 She  was acquitted, but the case was one of many used in attempts to reduce women’s reproductive rights and increase fetal rights. Interestingly, she was a white married woman and was not using what was considered a poor person’s drug of choice, crack cocaine. Most subsequent cases involved African American women using crack cocaine.86 Three cases followed, in which the women were incarcerated while pregnant. In one case, the judge, when sentencing the pregnant woman for check forgery, explicitly stated that he was keeping her in jail to protect the fetus, as he believed she was addictive. Notably, of the three, the only white woman served minimal time.86 A number of rather enlightening case histories are included in Paltrow’s 2013 review.89 In South Carolina, Regina McKnight, a 21-year-old African American, suffered a stillbirth due to ­infection, but she was found to have used cocaine and so was convicted of homicide due to child abuse. She was sentenced to 12 years’ imprisonment. In 2008, her conviction was overturned, with the court saying that cocaine was no more harmful than other factors seen in extreme poverty, such as poor nutrition

182  The prenatal environment and birth complications

and maternal care.89 To eliminate the risk of a retrial and potentially a greater sentence, McKnight “confessed” to manslaughter and was released due to time served, 8 years.89 In another case, a married 20-year-old woman went to a Wisconsin hospital to get help with her addiction to therapeutic opiates. State law allowed staff to take her into custody, and she was incarcerated in a psychiatric ward against her will, a distance away from her husband and first child. She was given no prenatal care and was prescribed drugs that have high risks of causing birth defects. Eventually, a doctor stated at her hearing that her addiction was not putting the fetus at risk, although she was still kept at the hospital for several days subsequently and was under close supervision for the duration of her pregnancy. As a result of the incarceration, the woman lost her job and her husband had to take a leave from work.89 A  homeless Native American woman in North Dakota was arrested in her first trimester for putting her unborn child at risk due to paint sniffing. She spent two weeks in jail and then was allowed to attend a medical appointment, where she requested and obtained an abortion. The state dismissed the charges, as the fetus was no longer at risk,89 which poses very interesting ethical questions relating to the rights of the fetus. Not all cases relate to substance abuse. In Florida, in the late 1990s, Laura Pemberton, a white woman, decided to have a vaginal birth after having had a previous child by cesarean section. While she was actually in labor, doctors, attempting to force her to have another cesarean section, requested a court order. Unbelievably, sheriffs attended her home, arrested her, “strapped her legs together,” and took her to the hospital while an emergency hearing was conducted.89(p306) The  judge compelled Pemberton to have a cesarean section despite the view of Pemberton and her husband that such a surgery was unnecessary. She  later sued based on the violation of her rights, but she lost, as the court ruled that the rights of the fetus outweighed her First, Fourth, and Fourteenth Amendment rights. Incidentally, she gave birth to three subsequent children, all vaginally, ­indicating that it had not been medically necessary to perform the cesarean section.89 Although, at first blush, it seems sensible to attempt to assist women who may have addiction problems, which may impact the subsequent health of their babies, incarceration is clearly not the answer. Providing a trustworthy source of prenatal support to assist women with such addictions is paramount and cannot possibly be achieved if a woman knows or even considers that she might be arrested and incarcerated if she comes forward. In that way lies disaster, as the women who need the most support will avoid seeking it. Also, if we accept that the rights of the unborn fetus outweigh the rights of the mother, then where will it end? I speculated earlier, somewhat facetiously, that if we accept that a fetus has more rights than the mother, then women who smoke, or women of certain ages, could be arrested to “protect the fetus.” This is, unfortunately, not far-fetched. In the United States, unbelievably, women have been arrested for not getting enough doctor-prescribed rest, being late to the hospital when in labor, having prenatal diabetes, not requesting prenatal care, deciding to have a baby at home, and suffering a miscarriage.89 In one case, with strong shades of eugenics, the state used the fact that the woman had refused involuntary sterilization as a point against her in court.89 It behooves us to be very careful when considering issues related to pregnancy outcomes and depriving women of their liberty. It is interesting to note that all the cases discussed above resulted from the interpretation of criminal laws in ways that were never intended and occurred despite no ­formal laws with which to uphold the charges. No state (except South Carolina) has ever passed a law that could deprive a woman of her liberty for using potentially dangerous substances, make her liable for the results of her pregnancy, change drug laws to make in utero drug use considered ­dealing, or consider child abuse to occur in the womb.89 Despite no such laws existing, at least 413 cases of severe mis­ carriages of justice have occurred, with some women serving many years in prison.

Conclusion Understanding good pre- and postnatal care is vital for the protection of both mother and child. This holds not only for their physical health at birth but also for the mental well-being of the developing child. Care during the prenatal and perinatal periods clearly should be targeted for improvement.91,92

References  183

Nurses could instruct pregnant women on caring for themselves during pregnancy and in postnatal care after the baby is born. This is an area that is often neglected, as a new young mother frequently finds herself suddenly alone with a baby, with very little training on how to care for the infant. It often appears to be thought that once a woman becomes a mother, she just “naturally knows”; however, this is not necessarily true, especially if the mother herself was a victim of poor parenting skills. Instruction on maternal health during and after the birth is very important for the health of the child and also the continuing health of the mother. This could include advice on appropriate diet and supplements such as folic acid and discussions on the risks of smoking and drinking alcohol. It could also extend to advice on how to protect a baby or young child’s head from head injury, such as the selection of appropriate car seats, toys, and safety helmets. Nursing programs in hospitals and in the community could greatly increase parental knowledge about pregnancy and child care.91,92 Unfortunately, many young women do not seek medical advice when pregnant and only come to the attention of medical services at childbirth. Others give birth without medical assistance. In such cases, hospital nursing plans are unlikely to be of any assistance. Community services and outreach programs may be able to assist in getting the information to those who need it most.

Questions for further study and discussion 1. It is clear that many actions of the mother can impact the fetus and eventually the grown child throughout life. Discuss the many ethical issues this raises in relation to the rights of women, as well as the obvious issues that this relates to in long-term personal and public costs of the birth of a child damaged by prenatal environment. 2. Compare and contrast the various potential mechanisms for FASD. At what points might it be possible to step in to support FASD children and their families? 3. Weigh the evidence relating to maternal smoking and offspring antisocial behavior. Do you think it is causal? Why or why not? 4. Twin births involve much more risk than singletons. How might this affect the many twin studies that have been conducted?

References 1. Cannon, M., Jones, P.B., and Murray, R.M. 2002. Obstetric complications and schizophrenia: Historical and meta-analytic review. Am. J. Psychiatry 159: 1080–1092. 2. Fazel, S., Bakiyeva, L., Cnattingius, S. et al. 2012. Perinatal risk factors in offenders with severe personality disorder: A population-based investigation. J. Personality Disord. 26(5): 737–750. 3. Liu, J., Raine, A., Wuerker, A., Venables, P.H., and Mednick, S. 2009. The association of birth ­complications and externalizing behavior in early adolescents: Direct and mediating effects. J. Res. Adolesc. 19(1): 93–111. 4. Johnson, S. and Marlow, N. 2014. Growing up after extremely preterm birth: Lifespan ­mental health outcomes. Semin. Fetal Neonatal Med. 19(2): 97–104. 5. Mednick, S.A. and Kandel, E. 1988. Genetic and perinatal factors in violence, In: Biological Contributions to Crime Causation, Mednick, S.A. and Moffitt, T., editors. Dordrecht, the Netherlands: Martinus Nijhof. 6. Quinsey, V.L., Skilling, T.A., LaLumiere, M.L., and Craig, W.M. 2004. Juvenile Delinquency: Understanding the Origins of Individual Differences. Washington, DC: American Psychological Association. 7. Reiss, A., Kesler, S., Vohr, B., et al. 2004. Sex differences in cerebral volumes of 8-year-olds born preterm. J. Pediatrics 145(2): 242–249. 8. Ferriero, D.M. 2016. The vulnerable newborn brain: Imaging patterns of acquired perinatal injury. Neonatology 109(4): 345–351.

184  The prenatal environment and birth complications

9. Smith, G.N., Thornton, A.E., Lang, D.J., et al. 2015. Cortical morphology and early adverse birth events in men with first-episode psychosis. Psychol. Med. 45(9): 1825–1837. 10. Vaske, J.C., Newsome, J., Boisvert, D.L., Piquero, A.R., Paradis, A.D., and Buka, S.L. 2015. The impact of low birth weight and maternal age on adulthood offending. J. Crim. Justice 43(1): 49–56. 11. Chen, W.-C., Lin, M.-J., and Liu, J.-T. 2010. Maternal age as a crucial factor between low birth weight and crime: Evidence from Taiwan’s National Data—A  research note. Soc. Sci. Res. 39(6): 1047–1058. 12. Newcombe, R., Milne, B.J., Caspi, A., Poulton, R., and Moffitt, T.E. 2007. Birthweight predicts IQ: Fact or artefact? Twin Res. Hum. Genet. 10(4): 581–586. 13. Neugebauer, R., Hoek, H.W., and E., S. 1999. Prenatal exposure to wartime famine and development of antisocial personality disorder in early adulthood. JAMA 282: 455–462. 14. St. Clair, D., Xu, M., Wang, P., et  al. 2005. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–1961. JAMA 294: 557–562. 15. Smith, B.L. and Reyes, T.M. 2017. Offspring neuroimmune consequences of maternal ­malnutrition: Potential mechanism for behavioral impairments that underlie metabolic and neuro­developmental disorders. Front. Neuroendocri. 47: 109–122. 16. Fortunato, J.J., da Rosa, N., Martins Laurentino, A.O., Goulart, M., et  al. 2017. Effects of omega-3 fatty acids on stereotypical behavior and social interactions in Wistar rats prenatally exposed to lipopolysaccarides. Nutrition 35: 119–127. 17. Saugstad, L.F. 1997. Optimal foetal growth in the reduction of learning and behavior ­disorder and prevention of sudden infant death after the first month. Int. J. Psychophys. 27: 101–121. 18. Ramakrishnan, U., Gonzalez-Casanova, I., Schnaas, L., et al. 2016. Prenatal supplementation with DHA improves attention at 5 y of age: A randomized controlled trial. Am. J. Clin. Nutr. 104(4): 1075–1082. 19. Lindsay, K.L., Buss, C., Wadhwa, P.D., and Entringer, S. 2017. The interplay between maternal nutrition and stress during pregnancy: Issues and considerations. Ann. Nutr. Metab. 70(3): 191–200. 20. de Jong, C., Kikkert, H.K., Seggers, J., Boehm, G., Decsi, T., and Hadders-Algra, M. 2015. Neonatal fatty acid status and neurodevelopmental outcome at 9  years. Early Hum. Dev. 91(10): 587–591. 21. Yan, Z., Jiao, F., Yan, X., and Ou, H. 2017. Maternal chronic folate supplementation ameliorates behavior disorders induced by prenatal high-fat diet through methylation alteration of BDNF and Grin2b in offspring hippocampus. Mol. Nutr. Food Res. 61(12): 1700461. 22. Lipton, L.R., Brunst, K.J., Kannan, S., Ni, Y.M., Ganguri, H.B., Wright, R.J., and Bosquet Enlow, M. 2017. Associations among prenatal stress, maternal antioxidant intakes in pregnancy, and child temperament at age 30 months. J. Dev. Orig. Health Dis. 8(6): 638–648. 23. Wright, J.P., Dietrich, K.N., Ris, M.D., et al. 2008. Association of prenatal and childhood blood lead concentrations with criminal arrests in early adulthood. PLoS Med. 5(5): 732–740. 24. Rochester, J.R., Bolden, A.L., and Kwiatkowski, C.F. 2018. Prenatal exposure to bisphenol A and hyperactivity in children: A systematic review and meta-analysis. Environ. Int. 114: 343–356. 25. CanFASD. 2018. FASD Fact Sheet. Accessed June 27, 2018; https://canfasd.ca/topics/basicinformation/. It clearly states date accessed – June 27, 2018. 26. May, P.A., Chambers, C.D., Kalberg, W.O., et al. 2018. Prevalence of fetal alcohol spectrum disorders in 4 US communities. JAMA 319(5): 474–482. 27. Streissguth, A.P., Aase, J.M., Clarren, S.K., Randels, S.P., LaDue, R.A., and Smith, D.F. 1991. Fetal alcohol syndrome in adolescents and adults. JAMA 265(15): 1961–1967. 28. Popova, S., Lange, S., Burd, L., and Rehm, J. 2015. Cost attributable to fetal alcohol spectrum disorder in the Canadian correctional system. Int. J. Law Psychiatry 41: 76–81.

References  185

29. Popova, S., Lange, S., Bekmuradov, D., Mihic, A., and Rehm, J. 2011. Fetal alcohol spectrum disorder prevalence estimates in correctional systems: A systematic literature review. Can. J. Public Health 102(5): 336–340. 30. Malbin, D.V. 2004. Fetal alcohol spectrum disorder (FASD) and the role of family court judges in improving outcomes for children and families. Juvenile Family Court J. 55(2): 53–63. 31. Brown, J.M., Long-McGie, J., Wartnik, A., et al. 2014. Fetal alcohol spectrum disorders in the criminal justice system: A review. J. Law Enforce. 3(6): 1–10. 32. Streissguth, A.P., Bookstein, F.L., Barr, H.M., Sampson, P.D., O’Malley, K., and Young, J.K. 2004. Risk factors for adverse life outcomes in fetal alcohol syndrome and fetal alcohol effects. J. Develop. Behav. Pediatrics 25(4): 228–238. 33. Burd, L., Selfridge, R., Klug, M., and Bakko, S. 2004. Fetal alcohol syndrome in the United States corrections system. Addict. Biol. 9(2): 169–176. 34. Corrado, R.R. and McCuish, E.C. 2015. The development of early onset, chronic, and versatile offending: The role of fetal alcohol spectrum disorder and mediating factors. Int. J. Child Adolesc. Health 8(2): 241–250. 35. Streissguth, A.P. and Dehaene, P. 1993. Fetal alcohol syndrome in twins of alcoholic m ­ others. Am. J. Med. Genetics (Neurpsych. Genet.) 47: 857–861. 36. Warren, K.R. and Li, T.K. 2005. Genetic polymorphisms: Impact on the risk of fetal alcohol spectrum disorders. Birth Defects Res. Pt. A 73(4): 195–203. 37. Eberhart, J.K. and Parnell, S.E. 2016. The genetics of fetal alcohol spectrum disorders. Alcohol Clin. Exp. Res. 40(6): 1154–1165. 38. Massey, S.H. and Compton, M.T. 2013. Psychological differences between smokers who spontaneously quit during pregnancy and those who do not: A  review of observational studies and directions for future research. Nicotine Tob. Res. 15(2): 307–319. 39. Ashford, J., van Lier, P.A., Timmermans, M., Cuijpers, P., and Koot, H.M. 2008. Prenatal smoking and internalizing and externalizing problems in children studied from childhood to late adolescence. J. Am. Acad. Child Adolesc: Psychiatry 47(7): 779–787. 40. Ma, X.D., Li, B.P., Han, Y., Tian, Y.P., Wu, L., and Wang, H. 2018. Influence of exposure to ­nicotine during pregnancy on the learning and memory for adult offspring. Exp. Ther. Med. 15(3): 2404–2410. 41. Bell, K., Corbacho, B., Ronaldson, S., Richardson, G., Torgerson, D., Robling, M., and Building Blocks Trial Group 2018. The impact of pre and perinatal lifestyle factors on child long term health and social outcomes: A systematic review. Health Econ. Rev. 8(1): 2. 42. Petkovsek, M.A., Boutwell, B.B., Beaver, K.M., and Barnes, J.C. 2014. Prenatal smoking and genetic risk: Examining the childhood origins of externalizing behavioral problems. Soc. Sci. Med. 111: 17–24. 43. Melchior, M., Hersi, R., van der Waerden, J., et al. 2015. Maternal tobacco smoking in pregnancy and children’s socio-emotional development at age 5: The EDEN mother-child birth cohort study. Eur. Psychiatry 30(5): 562–568. 44. Toro, R., Leonard, G., Lerner, J.V., et al. 2008. Prenatal exposure to maternal cigarette ­smoking and the adolescent cerebral cortex. Neuropsychopharmacology 33(5): 1019–1027. 45. Storebo, O.J., Rasmussen, P.D., and Simonsen, E. 2016. Association between insecure ­attachment and ADHD: Environmental mediating factors. J. Atten. Disord. 20(2): 187–196. 46. Biederman, J. 2005. Attention-deficit/hyperactivity disorder: A  selective overview. Biol. Psychiatry 57(11): 1215–1220. 47. Desrosiers, C., Boucher, O., Forget-Dubois, N., et  al. 2013. Associations between prenatal cigarette smoke exposure and externalized behaviors at school age among Inuit children exposed to environmental contaminants. Neurotoxicol. Teratol. 39: 84–90. 48. Boardman, J.D., Menard, S., Roettger, M.E., Knight, K.E., Boutwell, B.B., and Smolen, A. 2014. Genes in the dopaminergic system and delinquent behaviors across the life course: The role of social controls and risks. Crim. Justice Behav. 41(6): 713–731.

186  The prenatal environment and birth complications

49. Caspi, A., McClay, J., Moffitt, T.E., et al. 2002. Role of genotype in the cycle of violence in maltreated children. Science 297(5582): 851–854. 50. Suter, M., Ma, J., Harris, A., et  al. 2011. Maternal tobacco use modestly alters correlated epigenome-wide placental DNA  methylation and gene expression. Epigenetics 6(11): 1284–1294. 51. Orlebeke, J.F. 2001. Recent decreasing trend in U.S. juvenile delinquency attribute to changes in maternal age. Psychol. Rep. 88: 399–402. 52. Coyne, C.A., Langstrom, N., Rickert, M.E., Lichtenstein, P., and D’Onofrio, B.M. 2013. Maternal age at first birth and offspring criminality: Using the children of twins design to test causal hypotheses. Dev. Psychopathol. 25(1): 17–35. 53. Coyne, C.A., Fontaine, N.M.G., Långström, N., Lichtenstein, P., and D’Onofrio, B.M. 2013. Teenage childbirth and young adult criminal convictions: A  quasi-experimental study of criminal outcomes for teenage mothers. J. Criminal Jus. 41(5): 318–323. 54. Cao-Lei, L., Massart, R., Suderman, M.J., et al. 2014. DNA methylation signatures triggered by prenatal maternal stress exposure to a natural disaster: Project Ice Storm. PLoS One 9(9): e107653. 55. Veru, F., Laplante, D.P., Luheshi, G., and King, S. 2014. Prenatal maternal stress exposure and immune function in the offspring. Stress 17(2): 133–148. 56. Holloway, T., Moreno, J.L., Umali, A., et al. 2013. Prenatal stress induces schizophrenia like alterations of serotonin 2A and metabotropic glutamate 2 receptors in the adult offspring: Role of maternal immune system. J. Neurosci. 33(3): 1088–1098. 57. Koppensteiner, M.F. and Manacorda, M. 2016. Violence and birth outcomes: Evidence from homicides in Brazil. J. Develop. Econ. 119: 16–33. 58. Mansour, H. and Rees, D.I. 2012. Armed conflict and birth weight: Evidence from the al-Aqsa Intifada. J. Develop. Econ. 99(1): 190–199. 59. Torche, F. and Villarreal, A. 2014. Prenatal exposure to violence and birth weight in Mexico: Selectivity, exposure, and behavioral responses. Am. Sociol. Rev. 79(5): 966–992. 60. Barker, E.D., Jaffee, S.R., Uher, R., and Maughan, B. 2011. The contribution of prenatal and postnatal maternal anxiety and depression to child maladjustment. Depress. Anxiety 28(8): 696–702. 61. Janis, B.M., Callahan, J.L., Shelton, A.J., and Aubuchon-Endsley, N.L. 2016. Birth complications and parental stress reactions: Moderated by family coping. Pract. Innov. 1(4): 243–252. 62. Arseneault, L., Tremblay, R.E., Boulerice, B., Seguin, J.R., and Saucier, J.-F. 2000. Minor physical anomalies and family adversity as risk factors for violent delinquency in adolescence. Am. J. Psychiatry 157: 917–923. 63. Raine, A. 1993. The  Psychopathology of Crime: Criminal Behavior as a Clinical Disorder. San Diego, CA: Academic Press, Elsevier Science, 377 p. 64. Hajnal, A., Csabi, G., Herold, R., et  al. 2016. Minor physical anomalies are more common among the first-degree unaffected relatives of schizophrenia patients—Results with the Méhes Scale. Psychiatry Res. 237: 224–228. 65. Myers, L., Anderlid, B.M., Nordgren, A., et al. 2017. Minor physical anomalies in neurodevelopmental disorders: A twin study. Child Adolesc. Psychiatry Ment. Health 11: 57. 66. Fogel, C.A., Mednick, S.A., and Michelson, N. 1985. Hyperactive behaviour and minor ­physical anomalies. Acta Psychiatr. Scand. 72: 551–556. 67. Kandel, E. 1989. Genetic and perinatal factors in antisocial personality in a birth cohort. J. Crim. Justice 12(2): 61–77. 68. Brennan, P.A. and Mednick, S.A. 1993. Genetic perspectives on crime. Acta Psychiatr. Scand. Suppl. 370: 19–26. 69. Cheng, H., Chang, C.C., Chang, Y.C., Lee, W.K., and Tzang, R.F. 2014. A pilot study: Association between minor physical anomalies in childhood and future mental problems. Psychiatry Investig. 11(3): 228–231.

References  187

70. Compton, M.T., Chan, R.C., Walker, E.F., and Buckley, P.F. 2011. Minor physical anomalies: Potentially informative vestiges of fetal developmental disruptions in schizophrenia. Int. J. Dev. Neurosci. 29(3): 245–250. 71. Akgul, O., Bora, E., Akdele, B.B., and Alptekin, K. 2018. A  meta-analysis of minor physical anomalies in first-degree unaffected relatives of patients with schizophrenia. In: 6th Biennial Conference of the Schizophrenia International Research Society (SIRS). Florence, Italy. 72. Soyka, M. 2011. Neurobiology of aggression and violence in schizophrenia. Schizophr. Bull. 37(5): 913–920. 73. Tenyi, T., Halmai, T., Antal, A., et al. 2015. Minor physical anomalies are more common in schizophrenia patients with the history of homicide. Psychiatry Res. 225(3): 702–705. 74. Raine, A. 2002. Annotation: The role of prefrontal deficits, low autonomic arousal and early health factors in the development of antisocial and aggressive behavior in children. J. Child Psychol. Psychiatr. Allied Disc. 43(4): 417–434. 75. Dyshniku, F., Murray, M.E., Fazio, R.L., Lykins, A.D., and Cantor, J.M. 2015. Minor physical anomalies as a window into the prenatal origins of pedophilia. Arch. Sex. Behav. 44(8): 2151–2159. 76. Aziz, S. and Nargis, S. 2012. Twin births and their complications in women of low socioeconomic profile. J. Pakistan Med. Assoc. 62(11): 1204–1208. 77. Beer, C., Israel, C., Johnson, S., Marlow, N., Whitelaw, A., and Glazebrook, C. 2013. Twin birth: An additional risk factor for poorer quality maternal interactions with very preterm infants? Early Hum. Dev. 89(8): 555–559. 78. Gau, J.S., Silberg, J.L., Erickson, M.T., and Hewitt, J.K. 1992. Childhood behavior problems: A comparison of twin and non-twin samples. Acta Genet. Med. Gemellol. Roma 41(1): 53–63. 79. Resnick, S.M., Gottesman, II, and McGue, M. 1993. Sensation seeking in opposite-sex twins: An effect of prenatal hormones? Behav. Genet. 23(4): 323–329. 80. Andersen, S.L. and Teicher, M.H. 2004. Delayed effects of early childhood stress on hippocampal development. Neuropsychopharmacology 29(11): 1988–1993. 81. Raine, A., Brennan, P., and Mednick, S.A. 1994. Birth complications combined with early maternal rejection at age 1  year predispose to violent crime at age 18  years. Arch. Gen. Psychiatry 51(12): 984–988. 82. Liu, J. 2011. Early health risk factors for violence: Conceptualization, review of the evidence, and implications. Aggress. Violent Behav. 16(1): 63–73. 83. Vézina-Gagnon, P., Daigneault, I., Daignault, I.V., and Dupré, M.-P. 2016. La relation entre le rejet maternel et le succès scolaire: Une étude de médiation. Can. J. Behav. Sci./Rev. can. sci. comport. 48(2): 132–141. 84. Khan, S., Haynes, L., Armstrong, A., and Rohner, R.P. 2010. Perceived Teacher acceptance, parental acceptance, academic achievement, and school conduct of middle school s­ tudents in the Mississippi Delta Region of the United States. Cross-Cult. Res. 44(3): 283–294. 85. Papadaki, E. and Giovazolias, T. 2013. The protective role of father acceptance in the relationship between maternal rejection and bullying: A moderated-mediation model. J. Child Family Stud. 24(2): 330–340. 86. Wolff, K.B. 2011. Panic in the ER: Maternal drug use, the right to bodily integrity, privacy, and informed consent. Politics & Policy 39(5): 679–714. 87. Whitner v. State, 492 S.E.2d 777 (S.C. 1997) see https://www.courtlistener.com/ opinion/1388496/whitner-v-state/. 88. Drabble, L., Thomas, S., O’Connor, L., and Roberts, S.C. 2014. State responses to alcohol use and pregnancy: Findings from the Alcohol Policy Information System (APIS). J. Soc. Work Pract. Addict. 14(2): 191–206. 89. Paltrow, L.M. and Flavin, J. 2013. Arrests of and forced interventions on pregnant women in the United States, 1973–2005: Implications for women’s legal status and public health. J. Health Politics Policy Law 38(2): 299–343.

188  The prenatal environment and birth complications

90. Ferguson et  al. v. City of Charleston, 532  U.S. 67 (2001). https://supreme.justia.com/cases/ federal/us/532/67/. 91. Liu, J. and Wuerker, A. 2005. Biosocial bases of aggressive and violent behavior—Implications for nursing studies. Int. J. Nurs. Stud. 42(2): 229–241. 92. Liu, J. and Raine, A. 2000. Prevention. In: Encyclopedia of Violence in the United States, Gottesman, R. and Mazon M., editors. New York: Scribner.

9 The chemistry of the brain The role of neurotransmitters on behavior

Introduction The chemistry of the brain is not often covered in criminology texts, probably because it requires a fairly deep understanding of neuroscience. However, this is an extremely important area in the discussion of behavior, as much of a person’s behavior is governed by neurochemistry. A brief introduction to the brain and to neurotransmitters can help the reader navigate the general research in this area. The aim of this chapter is to present a general overview of how neurotransmitters work and the roles that the major neurotransmitters—in particular serotonin, dopamine, and monoamine oxidase A (MAOA)—play in behavior. A fine balance exists between neurotransmitters and behavior, so slight changes can alter behavior in significant ways, most notably through affecting impulsivity and inhibition. As with so many biological issues, neurotransmitter imbalances that cause behavioral changes can sometimes be treated, in some cases with medication or something as simple as a change in diet.

Introduction to neurotransmitters The brain is the basis for all behavior, both innate and learned. Today, we have a good understanding of how brain chemistry affects problems such as depression, which has, for some time, been accepted as a medical disorder and not simply a behavioral problem. We certainly do not understand everything yet, but current research reveals a good deal about how the brain works and which chemicals cause what reaction. This research has made it possible to help many people both medically and psychologically. However, it is only recently that there has been an exponential increase in research on brain biochemistry and its possible relationship to criminal behavior. Many criminology texts now include critical discussions of the link between genetics and crime, and most people immediately think of genetics when they think of biology and crime. However, very few criminology texts mention neurotransmitters, because neurochemistry is difficult and specialized. One does not need to be a scientist to understand basic genetics or to get a grasp of the basic genetic research in the field of biology and crime. Neurochemistry, on the other hand, is much more difficult to understand and uses many technical terms, and most people do not have the background to understand it. The field of neurochemistry includes some of the most exciting recent discoveries in the field, together with clear evidence of not just the effect of neurochemistry on behavior but also its interaction with the environment to cumulatively impact behavior. Moreover, as our understanding of neurochemistry increases, so does our ability to modify and even treat some chemical 189

190  The chemistry of the brain

imbalances. We have known for years that dysfunction in the dopamine system is involved in ­schizophrenia1 and that low levels of serotonin, norepinephrine, and dopamine relate to depression.2 It is not ­suggested that these chemicals are the cause of the disorder, but they form a part of the total picture. More importantly, it has also been proven that certain chemical imbalances can be treated, providing patients with a hopeful prognosis. It  is therefore possible that imbalances linked to ­antisocial behavior can also be modified.

The mechanism of action One can think of the body as having two communication systems: one system uses hormones and the other uses the neural system. The hormones are biological chemicals secreted into the bloodstream by endocrine organs, or glands, found throughout the body, for example, the testes, which produce testosterone, or the pancreas, which produces insulin. Most hormones communicate fairly slowly because they must travel through the bloodstream to get to where they are going to take effect (relatively speaking, that is; an exception may be adrenaline). Neural signals, on the other hand, are electric signals produced, received, and translated by nerve cells, and they act extremely quickly. Neurons (or nerve cells) are specialized cells in the nervous system that transmit signals from one part of the body to another and instruct the body on how to act. For a greatly simplified example, imagine nerve cells sending signals through the neural systems to the hand to pick up a fork, use it to pick up a piece of broccoli, and then move the hand up to the mouth to deposit the food therein. Other signals tell the mouth to close the lips, so the food will not fall out, and signals are sent to the mouth to chew, to the glands in the mouth to start secreting digestive enzymes, and so on. This  is greatly simplified but gives an idea of the messages going through your body in nanoseconds. There are many other kinds of nerve cells that we do not need to go into here, but it is essential to learn a little more about how nerve cells communicate. The communications take place between junctions called synapses. Figure 9.1 shows two nerve cells communicating. The cell sending the message uses the axon’s synaptic terminals, and the dendrites of the receiving cell

Figure 9.1  The presynaptic cell and the postsynaptic cell. (From SFU Publications, British Columbia, UK.)

The mechanism of action  191

receive the message. The transmitting cell is called the presynaptic cell (Figure 9.1), and the cell receiving the signal is called the postsynaptic cell (inset). There are two main types of synapses: (1) ­electrical synapses, which conduct electrical signals, and (2) chemical synapses, which conduct chemical ­signals. It is the chemical ones in which we are interested. At a chemical synapse, there is a very narrow gap called a synaptic cleft that separates the transmitting, or presynaptic, cell from the receiving, or postsynaptic, cell. An electrical signal cannot span the gap, but a chemical signal can. So, the electrical signal in the first cell is converted into a chemical messenger (or signal), so it can travel over the cleft. It is then converted back into an e­ lectrical signal in the receiving cell, so the receiving cell knows what the message is. Of course, this only describes communication between two cells. For a message to get from the brain, say to move one’s hand, millions of cells must communicate (involving countless ­neurotransmitters) in a split second. The chemicals that do the transmitting are called neurotransmitters, and they are stored in sacs called synaptic vesicles. Many neurotransmitters are known, including amino acids and biogenic amines. We are most interested here in the amines—which include such basic neurotransmitters as dopamine, norepinephrine, and serotonin—because they are the basis for information processing and communication in the human brain. As a result, they are the basis of almost all behavior, from sensation (the ability to touch and feel) to perception, learning, memory, eating and drinking, and, more controversially, antisocial behavior. As they are produced and discharged into the cleft, these neurotransmitters transmit information throughout the brain. All three of these n ­ eurotransmitters are thought to be very important in the context of brain behavior. Both medical diagnoses and research into criminogenic behavior need to measure the l­evels of these neurotransmitters in the central nervous system. There  are several methods used. Neurotransmitters are made of precursors that are the basic proteins of life, amino acids. Each neurotransmitter has a particular amino acid precursor, which is synthesized to make the ­neurotransmitter. Moreover, there is a particular enzyme that catalyzes that reaction to convert the amino acid into a neurotransmitter. It then must bind to receptors. Once the neurotransmitter has done its job, it breaks down into its metabolites. There are sometimes other chemicals that fluctuate with the neurotransmitter, so they can tell us the levels of the neurotransmitters indirectly. You can measure any one of these elements in the chemical process (the precursor, the enzyme, the metabolites, or an indirect indicator) to get the levels of the neurotransmitter. The best method seems to be measuring the metabolites.2 It is rather like measuring the level of cocaine or heroin metabolites to determine the amount of drug a person has taken. In many cases, the original drug breaks down very rapidly; for example, heroin breaks down in a matter of minutes, so the heroin levels themselves are not measured but rather the levels of the metabolites, the chemicals that the body breaks heroin into, such as morphine. Because we know what levels of morphine reflect ­specific levels of heroin, we can determine the original level. Three different body fluids can be tested for these chemicals in a living person: (1) cerebrospinal fluid (CSF), (2) blood, and (3) urine. Many of the studies we discuss next use one or more of these fluids. Neurotransmitters are synthesized in the brain and the metabolites go directly into the CSF. Thus, CSF is clearly the most accurate fluid to study. However, obtaining CSF involves performing a lumbar puncture, which is an extremely painful and fairly major medical procedure. It is an invasive method to measure neurotransmitters in a living person, and most people will not consent to it. The procedure brings up an ethical issue: many studies do use CSF, and many of them use incarcerated offenders. Such studies balance on a thin line of the supposed voluntarism of an incarcerated population. Researchers in this area use such experimental designs with caution and presumably with a greater value in mind. From the CSF, the metabolites enter the bloodstream, so blood samples are the next most accurate measure of neurotransmitter levels.2 Blood sampling is much less problematic than CSF sampling, although still invasive. Finally, the eventual metabolites are excreted via the urine. By this time, they have been broken down significantly, they are contaminated with other breakdown products, and considerable time has passed since they were utilized, making urine sampling not only the least

192  The chemistry of the brain

effective but also the least invasive method of measuring neurotransmitter levels. Although these body fluids are examined in many studies, neuroimaging is now also used frequently to ­measure the levels and activity of neurotransmitters in the brain, as well as receptor density. We know that neurotransmitters affect behavior; that is what they are biologically designed to do. Research is thus guided by the following goal: if it can be shown that neurotransmitters can be related to antisocial behavior, then altering the levels of the neurotransmitters, or their precursors, or the enzymes that synthesize them, should reduce the undesirable behavior. There are many neurotransmitters that have been studied, the most common being serotonin, dopamine, norepinephrine, and MAOA.

Serotonin Serotonin (also known as 5-hydroxytryptamine, or 5-HT) is manufactured in the brain by the raphe nuclei found in the brainstem. It is synthesized from tryptophan, an essential amino acid that we cannot synthesize ourselves but instead acquire from our diet. Through various chemical processes, the tryptophan is converted into serotonin and stored in the presynaptic vesicles until used. After it has been released into the synapses, it is taken up again by the serotonin transporter protein (5-HTT) and metabolized by monoamine oxidase (MAO) into hydroxyindole acetaldehyde and then into 5-hydroxyindolacetic acid (5-HIAA) and excreted in urine.3 There is a great deal of information that suggests that serotonin has a role in impulsive-­aggressive behavior. It is an important neurotransmitter, as it regulates the hypothalamic-­pituitary-adrenal axis (HPA) stress response. As discussed in Chapter  7, the HPA  axis is an interaction between the hypothalamus (a region of the brain that links the neural system to the endocrine system), the  ­pituitary gland (found in the brain and involved in the production and regulation of many ­different hormones), and the adrenal glands (found on the kidneys and also involved in the production and regulation of a number of hormones, including adrenalin, as well as steroids). The HPA axis is crucial in the regulation of many of our internal systems, such as the reproduction system, cardiovascular system, immune system, and central nervous system. Most importantly, it is involved in mood regulation and its dysfunction is related to many disorders such as attention deficit hyperactivity disorder (ADHD), bipolar disorder, borderline personality disorder, major depressive disorder, and alcoholism. It plays a major role in the effects of stressors on an individual. Serotonin is thought to act as a behavioral inhibitor, and so, it normally displays an inverse relationship with impulsive-aggressive behavior. That is, the lower the serotonin levels, the more impulsive-aggressive the person is; the higher the levels are, the calmer the person is.4 A  link between low levels of serotonin and aggression was first shown in a study in 1959, and the findings have been confirmed many times since then. Lower levels of serotonin (or its ­metabolites or precursors) have been associated with increased aggression, irritability, hostility, and ­impulsivity.4 The increase in aggression may occur because such behaviors are normally inhibited by a balance of neurotransmitters such as serotonin and dopamine. People with lower levels of these neurotransmitters lose that balance, so their behavior is no longer inhibited. Serotonin has many other functions, such as mood regulation, which could affect people more indirectly and push them into aggression. However, it is rare for a biological system to be influenced by just one thing. Usually, there are inhibitory and excitatory influences that interact and lead to some form of balance between the factors we can observe behaviorally.4

Serotonin and suicidal behavior The first serotonin behavior studies, in the late 1960s, looked at its role in suicidal behavior. Suicide can be considered an act of violence although it is directed at oneself. The early research looked at the levels of serotonin and its metabolite 5-HIAA in the brains of individuals who had committed

Serotonin  193

suicide compared with those who had died in accidents, and researchers found reduced levels of serotonin in the suicide victims compared with the accident victims.5 In further experiments, on living people, depressed patients who had a history of suicide attempts were found to have lower levels of 5-HIAA  than patients who were clinically depressed but had never attempted suicide. This  relationship was particularly clear in patients who had used violent methods in their suicide attempts.6 Obviously, not all suicides have low serotonin levels, but this relationship has been repeated in so many studies, some sort of link is clear. Dysfunction within the serotonergic system is presently the most commonly studied predisposition for suicide.7 Many aspects of the entire serotonergic system have been studied, from precursors to metabolites, with similar results. For example, the serotonin transporter protein (5-HTT) regulates the entire serotonergic system by reuptaking the neurotransmitter after it has been released, thereby controlling serotonin levels. 5-HTT is controlled by a gene called SLC6A4  on chromosome 17.7 Expression of the SLC6A4 gene was found to be stable in healthy control subjects, but the baseline level and changes in the expression of the gene in subjects who had suffered a major depressive episode predicted increasing suicidal ideation within the subsequent 2 months, and further changes during those 2 months predicted a suicide attempt within 7 months.7 Such expression levels can be determined with a blood test, and it is suggested that this could be a helpful indicator of future self-harm for physicians, to allow for intervention.7 Serotonin levels in a number of areas of the body have been studied. Blood is obviously easier to study than the brain, and platelet cells store serotonin, so they can be used as a measure of s­ erotonin levels. In Romania, a study of psychiatric patients showed that platelet serotonin levels were l­owest in individuals at highest risk for suicide and in those diagnosed with bipolar disorder.8 With today’s imaging advances, levels of serotonin in the brain of a living person can be studied using positron emission tomography (PET) after injecting a radio-tracing element. In  a study to look at serotonin transporter binding, 51 individuals with major depressive disorder (MDD), 15 of whom had attempted suicide, and 32 healthy controls were scanned.9 Serotonin transporter binding was significantly lower in suicide attempters with depression than it was in depressed non-attempters or normal controls, but there was no difference between depressed and healthy individuals ­overall, suggesting that serotonin transporter binding is associated with suicide directly and not  with depression per se.9 A  number of genes throughout the entire serotonergic system have been studied to determine their relationship with suicidal behavior, including tryptophan hydrolase genes (TPH1 and TPH2), which are involved in the metabolism of tryptophan to serotonin; serotonin transporter genes (5-HTTLPR in SLC6A4); serotonin receptor genes (HTR1A, HTR2A, HTR1B, and HTR2C); and  MAOA.10 In  a meta-analysis of studies worldwide (published in English) on all the above genes, Antypa and colleagues found relationships between certain polymorphisms (several alleles or ­variants of the gene) in 5-HTTLPR and in the TPH1 gene with violent suicide. 5-HTTLPR is a ­f unctional polymorphic region within the SLC6A4 gene, which governs the 5HTT gene, which in turn controls serotonin levels, as it uptakes serotonin from the synaptic cleft after it has been deployed. Humans have been found to have two alleles, or variants, of 5-HTTLPR: a short v­ ersion (SS) and a long version (LL). A  person can be homozygous for the short version (SS), meaning that both their alleles for this gene are the short version; homozygous for the long version (LL); or ­heterozygous (SL), carrying both versions. This  meta-analysis considered 14  studies from around the world, looking at a large range of ­ethnicities, and found a strong relationship between the S allele and suicidal behavior.10 Moreover, a greater relationship was seen between the SS genotype and higher success in completion of ­suicide, as well as the violence of the methods chosen.10 In  contrast, two studies noted that the L allele was associated with higher risk for attempting suicide in a Chinese population, and some studies found no association. Nevertheless, the large number of studies and other meta-analyses do show a ­significant relationship between the S allele of 5-HTTLPR and suicidal behavior, and it is suggested that differences in studies may relate to variation in measures to assess violence and lethality of suicide attempts, as well as ethnic differences.10

194  The chemistry of the brain

TPH1 is one of the genes involved in the conversion of dietary tryptophan to serotonin. This metaanalysis involved 25 studies on polymorphisms of this gene. Most studies showed a link between some variants of this gene and violent suicide attempts, although different alleles were involved, and some studies showed no links. The strongest links were among Caucasians.10 There was no consistent relationship between the other candidate genes and suicidal behavior unless very s­ pecific types of suicide were considered separately, such as level of violence, anger, and impulsivity, in which cases positive links were seen between TPH1, HTR2A, and MAO. Gene × environment interactions were also seen in some polymorphisms of 5-HTTLPR, HTR2A, and MAO with adverse life experiences.10 Serotonin, suicide, and stress

So, why would low serotonin levels increase the risk for suicide? Many studies have shown that there is a genetic link between low serotonin levels and the ability to handle stress. Everyone experiences stress, and most people are able to cope with such stress very well. However, some people appear to be more vulnerable to stress, and succumb to it, resulting in depression and in some cases suicide attempts. Evidence from animal and human studies suggests that some people are genetically more vulnerable to chronic stress. In seminal work, Caspi and colleagues11 showed that the S allele of 5-HTTLPR makes an individual more vulnerable to stress, whereas the L allele makes them more resistant. In a prospective long-term study in New Zealand, 847 individuals were divided into those who were homozygous for the short, or stress-sensitive, allele of the 5-HTT gene (17%), those who were homozygous for the long, or stress-protective, version (31%), and those who were heterozygous (51%). They then looked at individuals within each group who had suffered multiple stressful life events between the ages of 21 and 26 years. Among those who had suffered multiple stresses, 43% of those who were homozygous for the short version of the allele developed depression, in comparison with only 17% who were homozygous for the long version. Heterozygous individuals were i­ntermediate. Moreover, 11% of those who were homozygous for the short version of the gene attempted suicide, in comparison with only 4% in those with the long.11 This is a gene × environment (G × E) interaction and is evidence of the ways that the environment can modify genotypes. In many ­chapters, we have seen that a good, stable upbringing and life can be a major protective factor against genetic predisposition for criminogenic behavior. Perhaps in many of these cases, the mode of action is similar to the one described above, where a genotype is deleterious only under certain life ­circumstances. This takes us back to the diathesis stress model and other G × E interactions ­discussed in Chapter 6. We will see many more examples of this shortly. This groundbreaking study has been replicated many times. A meta-analysis of 54 studies that investigated the interaction between the 5-HTTLPR S allele and stress and its relationship to depression strongly indicated that 5-HTTLPR regulates the relationship between stress and depression, with levels of depression increasing with levels of stress.12 More in-depth analyses of the studies showed that there was a much greater association between the genotype and increased sensitivity to stress in individuals who had experienced early-life stress—in particular, childhood abuse—over those with stressful life events.12 Both early stress and later stress have been shown to have differential effects on epigenetic DNA methylation of 5-HTTLPR.13 A more recent meta-analysis considered epigenetic impacts on the stress-related gene SLC6A4. Research has indicated that epigenetics may moderate serotonin by changing the transcriptional function of the gene, without actually changing the DNA  of the gene itself.14 Nineteen studies from 2010  to 2016  that looked at the relationship between early adversity and DNA  methylation of SLC6A4 in individuals from 0 to 18 years old were considered. Adversity ranged from prenatal to childhood abuse to other early-life stressors. Early-life trauma appeared to trigger hyper­methylation at a number of sites within the gene. The  actual effect of this is still under investigation.14

Serotonin  195

Serotonin and aggression Brown and colleagues15 were the first to report a clear relationship between levels of the serotonin metabolite 5-HIAA in CSF and aggression, rather than suicide, in humans. Their work began by looking at young male navy recruits who had personality disorders but were neither depressed nor substance abusers. They used medical and military records to assess aggression. They found an inverse relationship between aggression and 5-HIAA levels; that is, the lower the serotonin metabolite, the higher the levels of aggression. However, although the subjects ranged in levels of aggression, the researchers did not look at any normal controls. However, many similar studies followed, confirming the link between serotonin levels and aggression. Some researchers have not found a link, or found a much weaker link, although this is hardly surprising, as no one suggests that all aggression has only one cause. Early work suggested a simplistic inverse relationship between serotonin levels and aggression, but more recent work suggests that the relationship is much more complex and that any dysfunction within the serotonergic system that impacts the fragile balance between neurotransmitters, their precursors, receptor sites, and metabolites, as well as with other neurotransmitters, can impact antisocial behavior. It is probable that such alterations in serotonin homeostasis relate to the interaction between genes, and the environment, with early adversity having the greatest effect during vital developmental stages.16 In a treatment study, aggressive individuals (male and female, from a range of ethnicities) with low serotonin levels were unable to inhibit violent reactions to experimental provocations, but when given paroxetine (Paxil), which rapidly increases central serotonergic activity, their reaction to provocation was inhibited to the same level as non-aggressive matched individuals.17 This study was augmented by a later study looking at aggressive psychopaths and found that primary psychopathy was related to aggressive responses to provocation, but this affect could be ameliorated with serotonin augmentation due to administration of paroxetine,18 supporting the concept that some forms of aggression in psychopaths may be modulated by a deficit in the serotonergic system. In a contrasting study, experimental reduction of serotonin by depleting the participant’s dietary intake of tryptophan, the precursor for serotonin, resulted in changes in normal individuals’ conventional behavior. Using games to assess behavior, the study found that participants with reduced serotonin were much more likely to punish other players who were unfair to them but not those who were unfair to other players.19 The authors suggested that serotonin plays a role in regulating normal social behavior—retaliation and fairness—affecting an individual’s social responses.19 Several genes have been implicated in the link between aggression and violence. One gene that has been studied well is the one previously discussed as linked to the ability to handle stress,11 SLC6A4, which governs 5-HTT. The  shorter allele, previously linked to the inability to handle stress appropriately, is also a predisposer for increased impulsivity, aggression, and hostility.3 The fact that the SS allele of SLC6A4 has already been shown to impact an individual’s ability to handle stress explains why it might also result in antisocial behavior, especially under stressful circumstances. Many studies have suggested that the SS genotype does not  have a direct effect on behavior but instead involves a gene × environment interaction, being moderated by stressful environmental triggers.3 We will see that this is a common theme. We know that genes rarely act in isolation, but more recent research is showing again and again that many possible genetic predispositions for crime only become risk factors in the presence of certain environmental experiences or stressors. An Australian study of 815 male and female adolescent offspring of mothers with depression looked at the potential interaction between various stress levels and the SLC6A4 genotypes.20 The participants were assessed at ages 15 and 20 for aggression and stress levels, using multiple measures. Youth with the SS or SL genotype indicated significantly more aggression on checklists in comparison with youth with the LL genotype, when exposed to chronic stress in the preceding 12 months. The response was the same for both sexes, and presence of acute stress or genotype alone had no effect.20

196  The chemistry of the brain

Epigenetic mechanisms may also alter the expression of the 5-HTTLPR alleles. PET scans measuring serotonin synthesis in the brain have shown that adult men who were highly aggressive as children but not as adults had lower synthesis of serotonin in areas of the brain. It was postulated that this was caused by differential epigenetic changes to certain genes within the serotonergic system.21 DNA methylation of SLC6A4 promoter in blood cells was compared with synthesis of serotonin in the brains of healthy adult men with or without childhood-limited aggression, using PET. DNA methylation was correlated with lower serotonin synthesis in the brain and high levels of childhood aggression. In fact, men who had been aggressive as children but not as adults had higher levels of methylation than those without.21

Serotonin and psychopathy Most work on 5-HTTLPR has shown that the S allele is the “risk allele” and is linked to a variety of antisocial behaviors and to increased sensitivity to stress, resulting in depression, anxiety, and substance abuse, but work has shown that the L allele may also predispose to risk factors, in particular those for psychopathy.22 The L allele is considered to be protective against stress and so results in reduced reactivity or emotional response. In a review of studies considering the L allele, Glenn argued that these traits are similar to those seen in psychopaths, who exhibit reduced emotion and callous, shallow, impulsive, and aggressive behavior. Glenn reviewed 19  studies comparing traits found in psychopaths (in comparison with normal individuals) in relation to 18 studies comparing the same traits in LL individuals compared with SL or SS. She noted striking similarities between both groups. At  the brain level, both psychopaths and LL individuals exhibited reduced activity in the amygdala in response to negative stimuli, reduced connectivity between the amygdala and ventromedial prefrontal cortex, and reduced concern when making cognitive errors. At the psychophysiology level, both exhibited reduced resting heart rate, reduced fear, reduced potential startle, reduced skin conductance under fear-inducing conditions, and increased heart rate variability. At  the hormonal level, both showed reduced cortisol response and elevated serotonin functioning. At  the neuropsychology level, both showed reduced attention to negative stimuli, poor passive avoidance learning, poor decisions based on punishment and reward, and increased risk-taking.22 Of course, there are also many studies that do not find similarities between the two groups. An apparent inconsistency is that aggression has long been associated with the S allele of 5-HTTLPR, but Glenn pointed out that there are two major types of aggression: reactive aggression, which is impulsive, and instrumental aggression, which is more controlled, proactive, and predatory. Most disorders involve the former, but psychopathy alone is associated with both reactive and instrumental aggressions. In fact, it is elevated instrumental aggression that is often used to distinguish psychopaths.22 Most studies do not distinguish between the two forms of aggression and those that do focus on reactive aggression. Reactive aggression is impulsive and unplanned, often involves a very low threshold for reactivity to a stimulus, and frequently results in overreaction, fear, and anxiety, leading to aggression.22 This fits with all we have discussed with regard to low serotonin levels and the S allele. The S allele appears to confer a risk for hyperreactivity to stressors or adverse triggers, which can result in violence. However, it is suggested that instrumental aggression has a very different etiology. Instrumental violence in psychopaths is believed to evolve from the very opposite condition to the hyperreactivity seen above, as psychopaths are characterized by a lack of sensitivity and empathy and are insensitive to punishment or another’s distress.22 Glenn therefore suggested that instrumental aggression, as opposed to reactive aggression, is more likely to be related to the L allele.22 In Caucasians, 36% of the population is LL, 48% LS, and 16% SS. However, the prevalence of psychopathy in the same population is 1%, so the vast majority of individuals who are LL are not psychopaths, in the same way that all SS individuals do not have clinical depression.22 Glenn therefore suggested that LL may be one of the many risk factors for psychopathy.

Serotonin  197

A more recent study looking at resting heart rate together with the LL 5-HTTLPR genotype tends to support Glenn’s hypothesis. An incarcerated population of 90 males, convicted of either violent offences or property crimes, was tested for genotype and resting heart rate.23 An interaction was found, with higher arrest rates for violent crimes for individuals with both a low resting heart rate and the LL allele.23 Low resting heart rate has long been shown to have a close correlation with antisocial behavior and aggressive behavior in children, adolescents, and adults. A recent meta-analysis of 114 reports showed that there was a strong, highly replicable relationship between low resting heart rate and numerous manifestations of antisocial behavior, psychopathy, and aggression.24 The work on serotonin as a whole suggests that dysfunction in any part of the serotonergic system appears to have negative effects on behavior. As always, it must be remembered that all this work refers only to predispositions and risk factors and that in most studies, the effect sizes are small, meaning that the risk factor may only be a very small amount of the variance.

Serotonin and impulsivity Linnoila and colleagues25 were the first to suggest that there was also a relationship between serotonin and impulsivity. They studied violent offenders in a Finnish forensic facility, looking at the number of violent crimes committed and whether these were impulsive or premeditated crimes. CSF levels of the serotonin metabolite 5HIAA were found to be significantly lower in repeat violent offenders compared with those who had committed only one violent crime.4,25 However, perhaps more interestingly, the levels of the serotonin metabolite were also ­significantly lower in those who had committed impulsive violent offenses compared with those who had ­committed premeditated violent offences.25 In addition, violent offenders with histories of suicide attempts had lower levels than those who had never tried to commit suicide. This was the first ­evidence that related low serotonin levels to impulsivity. This was extremely important because it indicated that serotonin’s link to criminal behavior is probably mediated by impulsivity. These ­findings may also explain why some researchers have not found a link with serotonin levels and aggression. If the link is serotonin and impulsivity, the impulsivity might result in violence, and serotonin would only show a relationship in violent offenders whose offenses related to impulsivity. Reduced levels of serotonin have also been shown to impact an individual’s ability to learn from punishment, which is a critical factor in emotional learning of boundaries. A lack of under­standing of negative consequences can increase impulsivity. Functional magnetic resonance imaging (fMRI) was used to assess healthy young women after either serotonin depletion or supplementation, using games (go, no ­punishment; no-go, punishment).26 Depleting serotonin in areas of the brain involved in ­emotion regulation resulted in reduced neural sensitivity to negative consequences and increased ­impulsivity, and the magnitude of the effect increased with depletion levels of serotonin.26 There are several theories on the mechanism of serotonin regulation of impulsive and aggressive behaviors. The first, based on Linnoila and colleagues’ early research, is that, as serotonin regulates behavior, low serotonin levels result in a lack of restraint or inhibition and result in increased impulsivity.27 Another theory is that of the “irritable aggression model.”27(p297) This suggests that individuals with overall low serotonin levels will have higher levels of irritability and therefore lower thresholds for reacting to adverse stimuli. This is supported by other studies that have shown that provocation is required to stimulate aggressive behavior in individuals with low serotonin,17 which provides further support for a gene × environment interaction. Many researchers have now demonstrated a link between certain types of aggression and low serotonin levels. However, it is still not  fully understood whether serotonin directly influences aggression or whether it regulates impulsivity, so low levels mean that a person has less control. However, lack of control, or impulsivity, is frequently linked to crime, and serotonin normally acts as a behavioral inhibitor. Low serotonin would therefore be logically expected to result in less inhibition and increased impulsivity.

198  The chemistry of the brain

It  is important to note that there are a range of types of impulsivity and that it is not  just a single modality. This  means that serotonin may regulate some forms of impulsivity but not  all. For example, in human and animal studies, severely depleting serotonin levels affects certain types of impulsivity but not others.28

Serotonin receptor sites Neurotransmitters such as serotonin need to bind to a site in the brain or on platelets in the blood in order to act. If there is no binding or receptor site, there is no effect. Therefore, levels of serotonin themselves may be normal in some cases, but reduced or changed receptor sites will impact efficacy. Early studies on the brains of suicide and accident victims showed that suicide victims had fewer binding sites than did accident victims.29 This finding means that normal levels of serotonin could be present but unable to act. This  theory is still controversial because the relationship between receptor sites and serotonin activity is not fully understood, although more recent work is helping to elucidate the relationship. One of the serotonin receptors, 1A (or serotonin1A), has been extensively studied. Higher numbers of serotonin1A receptors, or increased binding potential, relate to lower levels of serotonin release in the CSF. Early studies of cadavers showed higher brainstem serotonin1A in successful suicides than non-suicides.30 More recently, a study using PET scans over a 10-year sample of individuals with a major depressive disorder showed no difference in levels of serotonin1A in the brainstems of suicide attempters and non-attempters but did show a 45% increase in attempters with higher levels of lethality in previous suicide attempts.30 Another serotonin receptor, 1B (or 5-HT1BR), has been shown to be involved in aggression and impulsivity in animal studies, and recent work in human subjects has supported this. 5-HT1BRs are found both before and after the synapse. A number of studies have shown that drugs that activate these receptors blocked aggressive behavior in rodents, and rodents with the receptors knocked out were highly aggressive.31 In a human study, PET scans of 19 violent, incarcerated offenders and 24 healthy non-offenders were used to measure 5-HT1BR binding in three regions of the brain: the anterior cingulate cortex (involved in attention allocation, reward expectancy, and ethical reasoning), the orbitofrontal cortex (found in the frontal lobes and involved in decision-making), and the corpus striatum (found in the forebrain and involved in cognition, reward expectancy, and decision making).31 Only high 5-HT1BR binding in the corpus striatum was linked to high levels of anger and psychopathy in the violent offenders but not in controls, but no overall group differences were observed. This was not the expected result, and the authors suggested several hypotheses to explain the results.31 The first is that high 5-HT1BR binding might be indicative of low serotonin levels in the synapse; the second is that violent offenders may have responded differently due to being out of the prison environment for the tests, although the authors believe that this is unlikely; and finally, that 5-HT1BR binding may be a result of a genetic propensity for higher than normal density of receptors, which is supported by animal research, but further work is needed to better understand the role of the 5-HT1BR receptor. The authors suggested that 5-HT1BR might be a good potential target for future drug development.31 Platelet cells in the blood also have receptor sites for serotonin, and several studies have shown that reduced numbers of binding sites were inversely proportional to levels of aggression,4 as well as poor impulse control and violent recidivism.3

The serotonin precursor, tryptophan Neurotransmitters are synthesized from basic amino acids. Each neurotransmitter has a precursor amino acid. The precursor for serotonin is tryptophan, which cannot be synthesized in the body but rather must be ingested in the diet. If the diet is low in tryptophan, even people

Serotonin  199

with normal serotonergic activity will have low levels of serotonin, because they are unable to synthesize appropriate amounts of serotonin. Many serotonin studies have either experimentally depleted or supplemented tryptophan to reduce or elevate available serotonin. Medically, it is simple to acutely deplete tryptophan in individuals for a short period. A review of such studies has shown that although the results vary, general healthy individuals do not show any change in mood, but healthy individuals who have a familial history of depression do show mood depression, although still within the normal range.32 In depressed patients who are recovered and are on antidepressant treatment that modulates the serotonergic system, at least half experience a regression to full depression, but in those no longer on medications, the response is much smaller.32 It appears clear that blocking serotonin production can depress mood in some people and seems to have a greater effect on people vulnerable to depression. Many studies have also used acute tryptophan depletion to study aggression, and most found increased irritability or aggression with depleted tryptophan, but unlike mood, it did not relate to whether the individual was normally aggressive.32 In other studies, tryptophan was supplemented to increase serotonin synthesis. Many human and animal studies have shown the prosocial benefits of increased serotonin levels, with supplementation reducing irritability and increasing friendly behavior.32 Although many studies indicate that serotonin and aggression are related, many also show that low serotonin levels seem to relate more directly to impulsivity than to aggression. Thus, it may be that high impulsivity is the direct effect and aggression is a consequence. The reason that this work on lack of serotonin is exciting is that it is so easily affected by diet. In many people, increasing dietary levels of tryptophan can solve a problem caused by a lack of the precursor in the diet. High levels of dietary tryptophan can increase the levels of serotonin, thereby reducing impulsivity and often creating a general feeling of well-being and agreeable behavior, although the review above suggests that the mechanisms for each are different, as more agreeable behavior can occur without mood change.32 Traditional Christmas and Thanksgiving meals involve ingesting large quantities of roast turkey, and frequently, this is followed by guests relaxing, snoozing, and generally feeling mellow and peaceful. This is often thought to be related to simply eating too much, but actually, turkey is very high in tryptophan and the mellowing effects are most probably due to the resultant high levels of serotonin. Of course, there may be many other reasons why serotonin levels are low, and these may not be rectified by simple dietary correction. These reasons include an innate inability to convert tryptophan to serotonin, or a lack of or dysfunction in receptor sites. Still, when something is the result of a chemical imbalance, it should be possible to eventually readjust that balance.

Selective serotonin reuptake inhibitors Selective serotonin reuptake inhibitors (SSRIs) are a class of antidepressant drugs that are used to treat depression and other similar disorders.33 In general, they increase the levels of serotonin to  improve mood and feelings of well-being. They  have also been successfully used to reduce ­impulsive aggression in highly impulsive-aggressive individuals, and although it only completely prevents aggression in 30% of patients, almost half benefit from treatment.27 Work has shown that for SSRIs to work, the serotonin synaptic function must be working at least partially, as SSRIs attach to serotonin transporters, increasing serotonin levels within the synapse.27 Evidence suggests that low serotonin transporter numbers predispose to high levels of aggression, so fewer transporters would mean less to which the SSRIs could bind, resulting in blocking serotonin uptake and so ­reducing serotonin at the synapse.27 This  explains why individuals carrying the S  allele of 5-HTTLPR do not  respond well to SSRI treatment, as they have lower levels of synthesis of the transporter ­protein.27 Neuroimaging studies have shown that SSRIs and other aggression ­treatments normalize the brain’s overreaction to threatening stimuli (from 122% to less than 0.1%).27 Selective serotonin reuptake inhibitors are one of the most commonly prescribed ­psychiatric medications, with a study in a large Swedish population showing that over 14% of all women and 7%

200  The chemistry of the brain

of all men were prescribed SSRIs.34 Several studies have shown decreases in lethal and violent crime in individuals prescribed SSRIs, but controversially, others have shown an increase in violence in both adults and youth. A large Swedish study that included over 850 000 ­individuals prescribed SSRIs, in comparison with the entire population of the country (almost 8 million), showed a significant increase in violent crime arrests for young people (aged 15–24  years) using SSRIs, but not for other age cohorts, and found a similar association between this age cohort and non-violent ­criminal arrests, non-fatal accidents, and emergency treatment for alcohol misuse.34 Other studies have shown that youth and young adults react differently to SSRIs than other age groups, not just in propensity for violence when using such drugs but also in risk for suicide. Several meta-analyses have shown an increased risk of suicide attempts in people under 25 when prescribed SSRIs.33 Of course, SSRIs are prescribed for depression and anxiety, which have higher overall risks for suicide in general. In Denmark, each citizen has a unique civil registration number that allows researchers to follow all aspects of an individual’s life, including drugs prescribed and redeemed and suicide attempts. A large study was performed on all Danish people born between 1983 and 1989 (almost 400 000), with follow-ups until 2011, to determine the effect of SSRIs on suicide attempts. The results indicated a significant association with SSRIs and suicide risk. The ­hazard ratio was 5.23, meaning youth prescribed SSRIs were 5.23 times more likely to attempt suicide than youth not taking such drugs, with risk dramatically increasing within the first 3 months of SSRI use.33 This is obviously a major concern, as SSRIs are very commonly prescribed for people with depression, yet there are clearly major contrasting risks when prescribed to youth. The simple relationship between serotonin levels and behavior that was first considered is clearly not as simple as first hoped. Overall, it seems that dysfunction in the serotonergic system leads to impulsivity and a lack of inhibition to provocation or adverse stimuli. Individuals with normal serotonergic systems are able to respond appropriately to such stimuli, but those hypo-serotonin individuals will overreact, often in a violent or aggressive manner. For example, a provocation that could easily be dealt with verbally may escalate to physical violence in a hypo-serotonergic individual. Further research into the genetic control of the serotonergic system and the polymorphisms involved, together with the impact of epigenetic factors on gene expression, will help to elucidate the complexities of the system. Of course, the serotonergic system does not act in a vacuum, and levels of all neurotransmitters are interrelated.

Norepinephrine Norepinephrine is a common neurotransmitter that also functions as a hormone, but it is not as well studied as serotonin. It is designed to be produced to prepare the body for the classic fightor-flight reaction. In other words, it prepares the body for rapid action, such as fighting or running away to preserve life. Its levels in the body rise and fall during the day, being lowest during sleep and higher during the day, but can reach very high levels during periods of stress. During a stress event, it impacts most body functions to prepare for action—including increasing heart rate and blood pressure, increasing release of sugar for energy, and increasing blood flow to the muscles—and shuts down functions not involved in fight or flight, such as digestion. Its function alone therefore lends itself to potential aggressive behavior. Animal studies have shown that high levels of norepinephrine are correlated with aggression, and some studies in humans have suggested a similar relationship.4 Norepinephrine has been shown to aggravate aggressive behavior and to be higher in dominating, aggressive individuals. It is designed to be produced and released in highest quantities during provocation or adverse stimuli.35 Higher levels of norepinephrine were found in aggressive over less aggressive incarcerated prisoners in a small human study in the country of Georgia.35 Very few recent human studies have been conducted, but in many cases, we can infer information from studies that had objectives other than behavioral, such as drug tests. Certain drugs are known to reduce norepinephrine levels, so we can observe whether they also have an effect on aggressive behavior.

Dopamine  201

For  example, reserpine reduces the levels of norepinephrine, and it has been shown to reduce aggressive behavior. On the other hand, drugs that are known to increase norepinephrine levels, for example, tricyclic antidepressants, usually increase aggressive behavior in agitated, depressed patients.2 So, in general, the studies show that increased levels of norepinephrine result in increased aggression, whereas a decrease in norepinephrine results in a decrease in aggression. However, there has still not been enough research in this area to draw strong conclusions.

Dopamine Dopamine, a part of another significant neurotransmitter system, is highly important to the body’s reward system. Dopamine produces strong positive feelings that occur regularly in the course of everyday events. These pathways provide the pleasure drives for sex, love, food, and other such sought-after goals. The body thus produces its own natural rewards through a cascade of pleasurable feelings, and these feelings depend on the release of dopamine.36 Unnatural or artificial substances such as illicit drugs, alcohol, and nicotine can also target this region and bring on this cascade of pleasure, so the dopamine system is the major target of most drugs of abuse, due to the pleasure that the stimulation of this system produces.37 Such drugs rapidly induce elevations of dopamine in the brain, which result in intense euphoria, the typical “highs” of illicit drugs. Although different drugs have different mechanisms of action, all elevate dopamine. For example, cocaine blocks the dopamine transporter (DAT-1), preventing the uptake of dopamine, resulting in an increase in available dopamine levels. This heightened dopamine level then triggers adaptive plasticity in the neurons in the corpus striatum in the brain, which can lead to addictive behavior.38 Both illegal and therapeutic drugs, such as amphetamines and cocaine, greatly increase dopamine production and sometimes also increase human aggression.4 Although many people smoke, drink alcohol, and even take drugs, only some of these people become addicted. It has been suggested that these people suffer from a reward deficiency syndrome. 39 Such people have deficiencies in this normal reward cascade and thus do not receive the normal pleasures from life and hence crave greater stimulation. They are at a much greater risk for addiction than people with normal systems. Because dopamine is the main neurotransmitter in the reward pathway, variants of the genes involved in this pathway are most probably responsible, and some have been identified. Type II alcoholism, which is very strongly influenced by genetics, has also been linked to disruption of several of the genes related to dopamine receptors.39 Obesity, smoking, and pathological gambling have also been linked to variants in the dopamine receptors, in particular dopamine receptor D2 (DRD2). 39 However, serotonin, norepinephrine, and other neurotransmitters also regulate dopamine, so variant genes in any of these pathways might contribute to a predisposition for addiction.

The dopamine transporter, DAT-1 As with serotonin, the dopaminergic system is controlled by a number of genes, some of which have polymorphisms that have been suggested as candidate genes for antisocial behavior. For ­example, DAT-1 variable number tandem repeat (VNTR) functional polymorphism is responsible for expressing the dopamine transporter protein, which controls dopamine levels in the synapses by reuptaking dopamine and releasing it into the presynaptic terminal.40 Different polymorphisms or alleles result in differing expression of the dopamine transporter, meaning that different alleles result in greater or lesser efficiency of the protein, hence affecting dopamine levels. An early study looked at 790  children from the Colorado Longitudinal Twin Study and the Colorado Adoption Project and evaluated their behavior, based on parents’ ratings, at ages 4, 7, and 9 years.41 The researchers found that a particular allele of the DAT-1 gene involved in dopamine

202  The chemistry of the brain

transport was a significant risk factor for behavioral problems at ages 4 and 7 years. This allele, the 9-repeat variant (9R) of the DAT-1 gene, has also been suggested as a risk factor in alcoholism, and a different variant of the same gene, the 10-repeat variant (10R), has been linked to ADHD.41 Although 9R was not directly linked to externalizing behavior at age 9 years, the researchers did find a significant association across the three ages. This work supports the hypothesis that dopamine and changes in the dopamine system can result in undesirable behavior.41 In a study in Pakistan, 370 men convicted of murder and 359 non-incarcerated, non-violent men were genotyped for DAT-1, and a significantly higher prevalence of DAT-1 9R allele was found in the convicted murderers.40 In contrast, a Russian study of violent and non-violent incarcerated offenders found higher levels of 10R in non-violent offenders, but no association with 9R was observed in violent offenders.42 Polymorphisms of this gene have also been associated with neuropsychiatric disorders, schizophrenia, ADHD, substance abuse, and a number of other disorders.40

Dopamine receptor D2, DRD2 Another well-studied allele within the dopaminergic system is the TaqI A polymorphism of the DRD2 gene. There are two alleles, A1 and A2. The A2 allele is considered the normal allele, with 70% of the US population carrying this allele. The  A1  allele results in a 30%–40% reduction in number of D2 dopamine receptors and so results in lower levels of dopamine. 39 It has been linked to substance abuse and violence in youth and adults40 and to antisocial personality and pathological gambling.43

Dopamine receptor D4, DRD4 A further gene within the dopaminergic system that has been greatly studied is DRD4, another receptor that has 10 polymorphisms, or alleles, involving 2–11 repeats. Repeats from 7 to 11 have been considered risk alleles linked to a number of antisocial behaviors and disorders, such as ADHD, conduct disorder (CD), and binge drinking,43 although others also consider the 5R variant to be a risk factor for ADHD.44 The receptors from the 7R variant are less efficacious in binding dopamine than the more common 2R and 4R variants, and 7R is also very strongly correlated with a risk for developing ADHD. Studies also show that it increases risk for aggression in all life stages.45 It has been suggested that the risky alleles that relate to ADHD provide the link to a next step of antisocial and aggressive behavior. However, most people with ADHD do not exhibit criminal behavior, and a Russian study found that the prevalence of 5R and 7R was higher in violent offenders than in the general population, yet these offenders had no history of ADHD, suggesting that the alleles may predispose to violence alone, without being mediated by ADHD.45 A great deal of research has looked at all three genes (DAT1, DRD2 and DRD2) and has shown that polymorphisms of these genes are linked to many antisocial behaviors, such as impulsivity, violence, aggression, adolescent delinquency, lack of self-control, and substance abuse (see Beaver, Vaughn, et al.46 for review). Moreover, using the US National Longitudinal Study of Adolescent Health (Add Health) database of over 90 000 students, variants of both the DRD2 and DRD4 genes have been shown to be risk factors for psychopathy.47 Interactions between some polymorphisms of these genes have also been shown to impact behavioral disorders. Using the same database, Beaver and colleagues looked at the relationship between CD and polymorphisms of DRD2 and DRD4 and found no independent link between either gene and CD or gene and antisocial behavior but did see an interaction between certain alleles of the genes that were a predictor of risk for CD in youth and antisocial behavior in adults.48 So, in many cases, risk relates to an interaction between genes as well as with the environment.

Dopamine  203

Protective factors In many of the cases we have discussed, although various factors such as hormones, neurotransmitters, brain injury, and genes can provide risk factors for antisocial behavior, there are also many protective factors that reduce or eliminate the risk. One well-accepted protective factor is that of academic success in school. Failing or dropping out of high school has been linked to a plethora of unfavorable outcomes, such as substance abuse, antisocial behavior, and delinquency.46 Students who succeed academically have much more favorable outcomes. Although most studies have linked environmental factors such as family status and socioeconomic status (SES) with success in school, recent research is also considering the genetic underpinnings of academic success. A  variety of studies looking at different samples of students have reliably shown that around 50% of the variance in academic achievement is under genetic control.46 Of course, as in all behavioral genetics, behaviors are under the control of many genes, and each one makes only a small contribution to such complex traits as behaviors. Moreover, academic success is not a behavior or function on its own, so in itself, it is not under genetic or other types of control but is dependent on other factors that do impact success, such as memory, information retention, motivation, reward response, speed of information processing, and attention.46 As the dopaminergic system is heavily involved in all these factors, it has been suggested that it plays a role in educational success.46 Beaver and colleagues reviewed many studies that overall show three major lines of evidence that support this hypothesis. The first is the large number of studies that link intelligence to school success, and intelligence has been shown to be highly heritable. Many studies have shown a link between the dopaminergic system and higher or lower intelligence and cognition, as well as motivation, although not all studies support this.46 Second, research has shown that the DRD2 A1 allele plays a role in learning and that the DRD4 gene is involved in retaining information and in the formation and retention of memory.46 Finally, the more studied role of the dopaminergic system in antisocial behavior suggests that, as certain alleles have been implicated in dysfunctional behavior, it is likely to also play a role in academic failure, in that antisocial behavior is strongly linked to a high risk for dropping out or being expelled from school.46 A study using the Add Health database assessed genotype and academic success in adolescents at several points in their school career. The  first wave of surveys and interviews (approximately 21 000 students, aged around 16 years, and their primary caregivers) assessed antisocial behavior, peer relationships, family dynamics, delinquency, intimate relations, school grades, and other issues. The second wave was conducted approximately 1–2 years later and involved reinterviewing approximately 15 000 students with similar questions. The third wave occurred when the participants were young adults (aged around 22 years), with questions tailored more to employment, marital status, and financial status. At this third wave of questioning, a subsample of participants with a sibling in the study was genotyped for polymorphisms at DAT-1, DRD2, and DRD4. The results indicated that DRD2 was associated with grades in English, mathematics, history, and science at wave 1; DRD4 was associated with all subjects at wave 1 but only science and history at wave 2; and DAT-1 only had a small effect on English alone at wave 1 and not at wave 2.46 The authors suggested that these mixed results may relate to the different skill sets required for different subjects that depend on different regions of the brain. For example, success in English relies primarily on verbal skills, while mathematics and science require more visuospatial skills,46 and others have found variance in heritability scores between English and mathematics/science.49 Looking at the three genes cumulatively, there was a modest effect on overall GPA, with an 11% reduction in GPA in males with three of the risk alleles and a 14% reduction in GPA in females with three of the risk alleles, compared with gendermatched students with none. This is easily enough to affect university acceptance.46 Bullying and victimization in school-age children are major problems, with studies suggesting that up to 50% of children are bullied, with a rate of 20% by as young as age 6.50 Bullying can increase behavioral and mental health problems, impacts peer relationships, increases both externalizing and internalizing behaviors, increases suicide ideation, and can result in later psychotic problems.50

204  The chemistry of the brain

However, some children seem more resilient to the effects of such victimization than others. A study designed to test the relationship between childhood victimization, DRD4 ­polymorphisms, and antisocial behavior was conducted on 174  children between ages 6 and 10, using child and caregiver questionnaires to assess bullying and behavior, followed by genotyping. Results showed a significant interaction between certain variants of the DRD4 gene and the environment, in this case bullying. Some polymorphisms on the gene appeared to confer protection from the negative effects of bullying. Children with the DRD4-7R had significantly lower antisocial behavior if they were not bullied but had the highest levels of antisocial behavior if they were victimized, indicating a gene × environment interaction. More importantly, children with alleles other than DRD4-7R did not exhibit differential externalizing behavior whether they experienced victimization or not, suggesting that alleles other than 7R provided protection against the potential effects of bullying. No evidence was seen of a gene × environment interaction with internalizing behaviors.50 Studies on gene × environment interactions look not just at the risk factors in both the genotype and the environment but also at protective factors. A longitudinal study of over 1700 ­adolescents and young adults using interviews together with genotyping for dopaminergic system genes showed that a stable school and home environment had a greater protective influence than the level of risk provided by a bad environment of poor peers. The authors suggest that there is strong evidence, therefore, that a good social environment has a greater effect on reducing the genetic risk than a bad environment has on triggering the risk.51

Dopamine and schizophrenia The  impact of the dopaminergic system on the etiology of schizophrenia has been studied for over 40 years. Clinical studies have shown that increased levels of dopamine are associated with schizophrenia and that dopaminergic stimulants worsen the symptoms of schizophrenics and can ­actually induce psychosis in healthy individuals.1 Effective antipsychotic drugs affect the dopaminergic system by binding dopamine receptors and so reducing dopamine levels. Schizophrenia has many potential etiologies, genetic, environmental, and an interaction between the two. Some environmental causes, such as trauma in childhood and social isolation, have also been shown to affect the dopaminergic system, which might suggest one mechanism of schizophrenia.1 It has been suggested that, as the dopamine D2 receptor and genes involved in dopamine synthesis are involved in stress reaction, they may interact with environmental stressors or triggers to sensitize the dopaminergic system to stress, making it highly vulnerable and potentially resulting in ­psychosis.1 It is hoped that a better understanding of the relationships will lead to better treatment and perhaps even prevention methods.1

Monoamine oxidase Monoamine oxidase and aggression As we continue looking at neurotransmitters, we find that the mechanism of their influence becomes more and more complex. MAO is an enzyme in the mitochondria (the powerhouses found in large numbers in every cell, which provide the cell with energy). MAO is the enzyme responsible for breaking down several neurotransmitters, including serotonin, dopamine, and ­norepinephrine.52 Therefore, it also affects the levels of these neurotransmitters in the body. Low  MAO activity results in increased levels of these neurotransmitters in the brain. Animal studies have shown that when MAO is experimentally removed, the animals show increased aggression.10 There  are two MAO enzymes, MAOA and MAOB, and they are produced by two different genes that have loci (or ­positions) close to each other on the short end of the X chromosome.53 This means that these genes are sex-linked.

Monoamine oxidase  205

It is well known medically that inhibiting the production of MAO, particularly MAOA, using drugs, is a very effective way of treating depression. So, by inference, it would seem that altering MAO enzymic activity can affect the brain.53 Brunner and colleagues54 identified a large Dutch family with an abnormality in MAO production, some male members of which had a complex behavioral syndrome. These males were borderline intellectually disabled and had numerous behavioral problems, including aggressive behavior, which was mostly verbal but could become violent. They had also committed many crimes, including rape, assault, and arson.53 The ­problem was only ever seen in males, but it was transmitted through the females who were normal. This showed that the condition was sex-linked and carried on the X chromosome, in a similar manner to hemophilia and color blindness (discussed in Chapter 3). The  family had known about the problem for years, and one family member had even documented all the occurrences over the previous 35 years. All family members clearly knew who had the problem and who did not. It was found to be a result of a variant, or allele, of the gene that produced MAOA.53 Affected males could not produce MAOA, which again highlights that even a minute change in a single gene can be devastating. Proper gene function can be likened to opening a safe: if you have the combination, you can open it, but if you get just one number wrong, it will not open. Whether you have one wrong number in 6000 or in 6 million does not matter—one wrong number and the safe will not open. It is the same with genes. Everything must be perfect; a minute mutation can prevent proper functioning. In this case, the affected males could not produce any MAOA, which affected their ability to break down serotonin and dopamine. This was a single point mutation at C to T, position 936.54 These males produced serotonin normally, but it was not broken down, and as a result, their levels of serotonin were higher than normal. This  finding seems to go against everything we have just discussed. Although a lot of the information we have is not so clear, the relationship between ­lowered serotonin levels and impulsiveness, and therefore aggression, is fairly obvious. These men are impulsive and aggressive, and thus, one would expect them to have lowered serotonin levels. However, we also know that any sort of problem with the normal levels of body chemistry could cause problems. Brunner, who first discovered this family, speculated that if brain serotonin levels were high throughout life, from conception onward, as they would be when the MAOA gene is defective, it might have had an effect on the number and density of the receptor cells for serotonin. If so, it could mean that although there is too much serotonin, it has a lesser effect because there are not enough receptor cells; that is, the overall effect would be less serotonin activity. The evidence shows that MAOA inhibition does lead to significantly raised levels of serotonin in the brain, which causes changes in the brain.53 It is believed that any type of dysregulation of neurotransmitters can cause behavior changes. This early work on MAOA identified a very rare point mutation that resulted in the complete absence of the neurotransmitter, but this is unusual, and later work has shown that there are several gene variants that control MAOA production.

MAOA and a gene × environment interaction Many people who were abused as children grow up to be criminals or abusers themselves, and this is often considered a result of their abusive background. However, many equally abused children grow up to be highly productive, non-criminal adults. Although it is unlikely that there is a single explanation for this discrepancy, exciting research into the gene that controls production of MAOA suggests that one of the differences between the two groups may be a gene × environment interaction (Caspi et al., 2002).11 The Dutch kindred was very unusual, with an allele that resulted in no MAOA being produced, and is the only one known, but later work indicated that the MAOA gene has a VNTR functional polymorphism (meaning that the gene includes several variants, or alleles) producing either normal levels of MAOA, referred to as high efficiency alleles (MAOA-H), or producing low levels,

206  The chemistry of the brain

referred to as low-activity variants (MAOA-L). In  seminal work, Caspi and coworkers analyzed data from 442 male adult New Zealanders and identified 154 who had been abused as children, including 33 who had been severely abused.55 Each participant was evaluated to determine whether he had a fully functioning variant of the MAOA gene or whether he had the low-activity variant, MAOA-L and then was assessed for antisocial behavior based on convictions for violent crime, CD, antisocial personality disorder (ASPD), and a diagnosis of violence. Although only 12% of the cohort possessed the MAOA-L genotype, they were responsible for 44% of the violent criminal convictions.55 Moreover, 85% of the severely abused participants with MAOA-L developed some form of antisocial behavior, whereas severe childhood abuse alone increased the risk slightly. In contrast, participants who had the fully functioning variant of the gene rarely exhibited aggressive or criminal behavior or CD as adults, even if they had been severely abused as children.55 Importantly, no significant correlation was seen between antisocial behavior and genotype alone. The authors contended that the combination of childhood abuse with MAOA-L could predict antisocial behavior in the same way that high cholesterol predicts heart disease.55 Simple possession of MAOA-L did not  predispose for aggression, and child abuse only slightly. Only those who had MAOA-L and who also experienced abuse in childhood were predisposed to aggression. This indicates that the presence of a fully functioning MAOA  gene provides protection against the effects of early childhood abuse and could explain why only some victims of early abuse grow up to be abusers.55 The MAOA-L genotype appeared to affect a boy’s resilience or sensitivity to abuse. As the gene is on the X chromosome, women can be carriers but would only be affected if they inherited the aberrant alleles from both parents, making the phenotype much more common in males, and the authors suggested that this could be one of the reasons that antisocial behavior is so much more prevalent among men. This was the very first study to show a gene × environment interaction resulting in a behavioral change and the first to indicate a predisposition for violence. Since this exciting breakthrough, many other studies have replicated this work globally, showing that the result is very robust and clearly indicates a gene × environment interaction, although some studies have not.56 In a meta-analysis of 27 English-language studies, the low activity genotype, in combination with early abuse, was a significant predisposer in males for later antisocial behavior and indicated a possibility of poor control over antisocial behavior and aggression due to poor executive functioning.57 This relationship stood up to meta-analysis, despite the differences in methods, study groups, and definitions of adversity, and some data have suggested that the level of adversity is proportional to the risk for and level of violence in later life.58 Moreover, carriers of MAOA-L who have experienced abuse also have higher risks for ADHD, impulsivity, and ASPD.59 Other types of adversity, such as parental neglect or violence between parents, have also been studied, and links have been found to other disorders such as CD, and again, it was found that the level of adversity predicted severity.60 Interestingly, in this study, MAOA-L children who were not abused were less likely to be aggressive than MAOA-H children.60 Although most of the work shows that the link only results in antisocial behavior in boys, some work has shown a different effect in girls. MAOA-L in tandem with child sexual abuse, rather than physical abuse, was shown to increase the risk of substance abuse and antisocial personality in girls, in a sample of Native American women.61 As the MAOA gene is on the X chromosome, girls can be homozygous or heterozygous for MAOA-L. Women homozygous for MAOA-L had the highest rates of alcoholism and antisocial behavior, and heterozygous women had intermediate rates, when compared with women who were homozygous for MAOA-H. There was no relationship between the genotype in women who had not been sexually abused. This work suggests that the MAOA-L genotype increases a girl’s sensitivity to childhood trauma in the same way that physical abuse works in boys.61 In contrast, however, a later study showed that the gene × environment interaction and later initiation into substance abuse held true for boys but not for girls.62 More research is needed to more fully understand the effect of gender. A large study of high school students in Sweden, using genotyping and self-assessed information on delinquency and childhood physical and sexual abuse, not only confirmed the relationship between abuse and MAOA-L in boys but also showed a relationship with MAOA-H in girls, which has also been seen in other studies.63

Monoamine oxidase  207

Although the relationship between maltreatment and antisocial behavior was present, the relationship was weaker than that seen in other studies but similar to that seen in earlier Swedish studies. The authors suggest that the gene × environment relationship may relate more to impulsivity that results in violence than to violence alone and that this is moderated by culture. Sweden has a very non-violent culture, with few violent crimes compared with many other countries, and they suggest that culture may influence the expression of impulsivity.63 Other studies have also suggested that the link relates to impulsivity and a greater than normal reaction to a sudden trigger. Most work has looked at childhood adversity, but it has been suggested that there may also be a link to adult adversity. A study of US university students looked at distal stress (related to childhood maltreatment) and proximal stress (related to life stressors during the previous 12 months, such as health, financial, and family issues) and their effect on criminal behavior.64 In MAOA-L males, both distal and proximal stress increased the relationship with criminal behavior, which the authors suggested meant that MAOA-L was not a risk allele as such but a marker for a greater sensitivity to the environment, compounding the effects of the stressors.64 Most studies have found no effect of MAOA-L alone. In  fact, it is estimated that MAOA-L is found in 30%–40% of Caucasians and 60% of African Americans and Asians.65 Despite this, the relationship between the genotype and childhood abuse only holds true strongly in Caucasians so far, with very mixed results in other ethnic groups, suggesting that the relationship may be specific to Caucasian males65 or simply that further research on other ethnic groups is required. In one study, MAOA-L was found to predict ASPD in Caucasians but not in African Americans, but without physical abuse, which is in contrast to almost all other studies. It also indicated the protective factor of the MAOA-H genotype.66 The MAOA gene is functionally polymorphic, meaning that different alleles relate to different levels of enzymatic activity, and there are several alleles, or variants, including 2-, 3-, 3.5-, 4-, and 5-repeat alleles, of which 3- and 4-repeat alleles represent 95% of the variation in Caucasians.57 The 2-, 3-, and 5-repeat alleles are considered low activity, and the 3.5- and 4-repeat alleles are considered normal or high activity. Most researchers clump the alleles into either high or low, which means it is hard to distinguish the effects of the rarer alleles. However, some work has been done on the much rarer 2-repeat allele, which is only found in 0.1% of Caucasian males and 5.2% of African American males, and has shown that there is a significant link between possessing the 2-repeat allele and the commission of shootings and stabbings in African American males.67 Of particular interest is that the presence of the 2-repeat allele was the predisposer alone, independent of any environmental factors, meaning that simple possession of the allele was the risk factor, with no childhood abuse involved as a trigger.67 However, a relatively small sample size was involved, and so few Caucasians were found with the allele that the study focused only on African Americans; so, further work is needed in this area on larger sample sizes. These studies are groundbreaking. They illustrate beautifully the relationship between the environment and genetics. The genes provide the basic framework, but how the person will grow up depends on the environment working with the genes. We can see this clearly when we look at a physical phenotype, but it is much more difficult to envision when we talk of behavior. If children are born to tall, large parents, it is probable that they will also grow up to be large. However, if these children experience extreme illness or malnutrition, they will not reach their full genetic potential. The genes were the blueprint for size, but the environment influenced the final outcome and overrode the genetics. The MAOA-L studies illustrate a similar situation with behavior. Children with the low-activity variant of the MAOA gene are genetically programmed to grow up normally, with no adverse consequences of a lower level of MAOA—unless the environment is harsh, in which case children with this genotype will be more prone to behavioral problems, including impulsivity and aggression. The genes create the framework, but the environment will be what influences the final outcome. Again, the protective value of a stable, non-violent upbringing overcomes any genetic predisposition for antisocial behavior. The value of providing children with stability and care cannot be overstated. Funding for families, child-care support, mentoring programs, and school support would be repaid a thousand-fold.

208  The chemistry of the brain

This  gene  ×  environment interaction is very exciting and greatly expands our understanding of interactions involved in behavior. Once again, though, we must remember that the interaction results in a predisposition for violent behavior, a risk factor, but the level of risk varies with many factors, including level and timing of adversity, and even within such factors, it varies between studies. Therefore, it may or may not contribute to an individual’s behavior. Moreover, none of the research suggests that such individuals are incapable of controlling their impulsivity, nor that they are unable to understand right from wrong. However, MAOA-L and other biological risk factors are now being introduced into the criminal justice system, which can be worrisome, as they are not yet fully understood (see Chapter 13). Frighteningly, it has even been suggested that newborn testing for the genotype should be mandated, so that parents can receive support to prevent abuse.68 The author of this statement justified this by pointing out that there is mandatory testing for phenylketonuria (PKU), which has a frequency of 0.007% in the population, dramatically lower than the frequency of MAOA-L.68 However, the circumstances are quite different: PKU, if you remember from Chapter 1, is 100% likely to result in very severe intellectual disability if not treated and is linked 100% to genotype. MAOA-L is quite different, with a much smaller risk factor that is only present in the presence of other risk factors. Moreover, the support suggested is aimed primarily at the child, whereas it is the parental environment that would be the more appropriate target. There is even a risk that a child might be taken from parents who were considered inappropriate. If the home environment is abusive, then child services might remove the child anyway, but it has been suggested by some that the bar should be set lower for an MAOA-L child, although the author did admit that removal would also be an adversity for the child and might cause more damage.68 The risk of labeling in this situation is extremely high, and it behooves us all to be very careful in interpreting these studies into real-life outcomes until more is understood.

Other factors Sociological studies often show a relationship between crime and poverty, or low SES. However, the phrase “low socioeconomic status” simply means that many things could be influencing behavior. In particular, low SES often involves a bad diet or poor nutrition.2 We will look at diet again, but here, we need to note that for the body to produce neurotransmitters, the diet must include the precursors, such as tryptophan in the case of serotonin and tyrosine and phenylalanine in the case of dopamine. Certain things in the diet can also block the body’s ability to take up these substances. So, if the diet is low in these precursors or includes other elements that block their uptake, then the person will have low levels of the neurotransmitters. In some cases, even if the dietary problems are rectified, the brain levels never fully recover, which indicates permanent damage. It stands to reason that those of low SES will probably have poor diets, and the same is true of drug users and alcoholics. Therefore, poor nutrition in people of low SES, including during pregnancy, could very well affect neurotransmitter levels throughout life. This is just one more example of how biological factors do not act in a vacuum. Social, environmental, and biological factors interact to influence behavior, which itself may alter the original factors. In this case, the ­environment—that is, the low SES—could result in a poor diet that causes a biological problem.

Conclusion Research in this exciting area is growing exponentially, not  just from a criminological standpoint but also from a medical standpoint, as there is a need to understand the underpinnings of many antisocial behaviors and disorders to increase the chance of treatment and even prevention. The body functions as a result of interactions between numerous systems and chemicals. Therefore, none of these systems works alone. They all indirectly and directly affect each other.

References  209

However, it is important for us to remember that although many polymorphisms of these genes can be risk factors for antisocial behavior, each contributes only a small risk, and it is the totality of genetic risk factors, together with a variety of environmental factors, that will ultimately result in a particular behavior. It is easy to look at these studies and conclude that risk factor X doubles the risk for Y behavior and so is of tremendous concern. We hear such things on the news all the time. For example, we often hear that X—for example, drinking pop, coffee or alcohol—doubles a person’s risk for Y (insert horrible disease of choice here). It sounds terrifying: it doubles the risk! But the risk is really not so concerning, as what they don’t tell you is that your chance of getting disease Y anyway was only 0.001%, so doubling it only gets you to a 0.002% chance. Therefore, “risk factors” must be kept in perspective.

Questions for further study and discussion 1. Why does it make sense that, as the SS allele for the serotonin transporter increases risk for impulsivity and hypersensitivity to stress and the LL allele inhibits these effects, the LL allele is more common in psychopaths? 2. Why, of the major neurotransmitters we have discussed here, is it only MAOA deficits that seem to impact mainly males and very rarely females? 3. Why, of all the neurotransmitters we discussed, is dopamine most related to substance abuse? 4. Name at least two protective factors that can ameliorate risk from neurotransmitter dysfunction and explain their mechanism.

References 1. Howes, O.D., McCutcheon, R., Owen, M.J., and Murray, R.M. 2017. The role of genes, stress, and dopamine in the development of schizophrenia. Biol. Psychiatry 81(1): 9–20. 2. Raine, A. 1993. Neurochemistry, In: The  Psychopathology of Crime: Criminal Behavior as a Clinical Disorder. San Diego, CA: Academic Press, Elsevier Science. pp. 81–102. 3. Glick, A.R. 2015. The role of serotonin in impulsive aggression, suicide and homicide in adolescents and adults: A literature review. Int. J. Adolesc. Med. Health 27(2): 143–150. 4. Coccaro, E.F. and Kavoussi, R.J. 1996. Neurotransmitter correlates of impluse aggression, In:  Aggression and Violence: Genetic, Neurobiological and Biosocial Perspectives, Stoff D. M. and Cairns, R.B., editors. Mahwah, NJ: Erlbaum. pp. 66–85. 5. Shaw, D.M., Camps, F.E., and Eccleston, E.G. 1967. 5-hydroxytryptamine in the hindbrain of depressive suicides. Brit. J. Psychiatry 113(505): 1407–1411. 6. Asberg, M., Traksman, L., and Thoren, P. 1976. 5-HIAA in the cerebrospinal fluid: A biochemical suicide predictor? Arch. Gen. Psychiatry 33: 1193–1197. 7. Consoloni, J.L., Ibrahim, E.C., Lefebvre, M.N., et al. 2018. Serotonin transporter gene expression predicts the worsening of suicidal ideation and suicide attempts along a long-term follow-up of a Major Depressive Episode. Eur. Neuropsychopharmacol. 28(3): 401–414. 8. Giurgiuca, A., Nemes, B., Schipor, S., Caragheorgheopol, A., Cozman, D., and Tudose, C. 2017. Suicide risk is associated with low levels of platelet serotonin in bipolar I disorder. Roman. J. Legal Med. 25(2): 205–210. 9. Miller, J.M., Hesselgrave, N., Ogden, R.T., et al. 2013. Positron emission tomography quantification of serotonin transporter in suicide attempters with major depressive disorder. Biol. Psychiatry 74(4): 287–295. 10. Antypa, N., Serretti, A., and Rujescu, D. 2013. Serotonergic genes and suicide: A systematic review. Eur. Neuropsychopharmacol. 23(10): 1125–1142. 11. Caspi, A., Sugden, K., Moffitt, T.E., et  al. 2003. Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science 301(5631): 386–389.

210  The chemistry of the brain

12. Karg, K., Burmeister, M., Shedden, K., and Sen, S. 2011. The serotonin transporter promoter variant (5-HTTLPR), stress, and depression meta-analysis revisited: Evidence of genetic moderation. Arch. Gen. Psychiatry 68(5): 444–454. 13. Duman, E.A. and Canli, T. 2015. Influence of life stress, 5-HTTLPR genotype, and SLC6A4 methylation on gene expression and stress response in healthy Caucasian males. Biol. Mood Anxiety Disord. 5: 2. 14. Provenzi, L., Giorda, R., Beri, S., and Montirosso, R. 2016. SLC6A4 methylation as an ­epigenetic marker of life adversity exposures in humans: A systematic review of literature. Neurosci. Biobehav. Rev. 71: 7–20. 15. Brown, G.L., Goodwin, F.K., Ballenger, J.C., Goyer, P.F., and Major, L.F. 1979. Aggression in humans correlates with decerbrospinal fluid amine metabolites. Psychiatry Res. 1(2): 131–139. 16. Bortolato, M., Pivac, N., Muck Seler, D., Nikolac Perkovic, M., Pessia, M., and Di Giovanni, G. 2013. The  role of the serotonergic system at the interface of aggression and suicide. Neuroscience 236: 160–185. 17. Berman, M.E., McCloskey, M.S., Fanning, J.R., Schumacher, J.A., and Coccaro, E.F. 2009. Serotonin augmentation reduces response to attack in aggressive individuals. Psych. Sci. 20(5): 714–720. 18. Fanning, J.R., Berman, M.E., Guillot, C.R., Marsic, A., and McCloskey, M.S. 2014. Serotonin (5-HT) augmentation reduces provoked aggression associated with primary psychopathy traits. J. Personal. Disorder 28(3): 449–461. 19. Crockett, M.J., Apergis-Schoute, A., Herrmann, B., et al. 2013. Serotonin modulates striatal responses to fairness and retaliation in humans. J. Neurosci. 33(8): 3505–3513. 20. Conway, C.C., Keenan-Miller, D., Hammen, C., Lind, P.A., Najman, J.M., and Brennan, P.A. 2012. Coaction of stress and serotonin transporter genotype in predicting aggression at the transition to adulthood. J. Clin. Child Adolesc. Psychol. 41(1): 53–63. 21. Wang, D., Szyf, M., Benkelfat, C., et al. 2012. Peripheral SLC6A4 DNA methylation is associated with in vivo measures of human brain serotonin synthesis and childhood physical aggression. PLoS One 7(6): e39501. 22. Glenn, A.L. 2011. The other allele: Exploring the long allele of the serotonin transporter gene as a potential risk factor for psychopathy: A  review of the parallels in findings. Neurosci. Biobehav. Rev. 35(3): 612–620. 23. Armstrong, T.A., Boisvert, D., Flores, S., Symonds, M., and Gangitano, D. 2017. Heart rate, serotonin transporter linked polymorphic region (5-HTTLPR) genotype, and violence in an incarcerated sample. J. Crim. Justice 51: 1–8. 24. Portnoy, J. and Farrington, D.P. 2015. Resting heart rate and antisocial behavior: An updated systematic review and meta-analysis. Aggress. Viol. Behav. 22: 33–45. 25. Linnoila, M., Virkkunen, M., Scheinin, M., Nuutila, A., Rimon, R., and Goodwin, F.K. 1983. Low  cerebrospinal fluid 5-hydroxyindolacetic acid concentration differentiates impulsive from non-impulsive violent behavior. Life Sci. 33: 2609–2614. 26. Helmbold, K., Zvyagintsev, M., Dahmen, B. et  al. 2015. Effects of serotonin depletion on ­punishment processing in the orbitofrontal and anterior cingulate cortices of healthy women. Eur. Neuropsychopharmacol. 25(6): 846–856. 27. Coccaro, E.F., Fanning, J.R., Phan, K.L., and Lee, R. 2015. Serotonin and impulsive aggression. CNS Spectr. 20(3): 295–302. 28. Dalley, J.W. and Roiser, J.P. 2012. Dopamine, serotonin and impulsivity. Neuroscience 215: 42–58. 29. Coccaro, E.F. and Astill, J.L. 1990. Central serotonergic system function in parasuicide. Progressive Neuro-Psychopharm. Biol. Psych. 14: 663–674. 30. Sullivan, G.M., Oquendo, M.A., Milak, M., et al. 2015. Positron emission tomography quantification of serotonin (1A) receptor binding in suicide attempters with major depressive disorder. JAMA Psychiatry 72(2): 169–178.

References  211

31. da Cunha-Bang, S., Hjordt, L.V., Perfalk, E., et  al. 2017. Serotonin 1B receptor binding is ­associated with trait anger and level of psychopathy in violent offenders. Biol. Psychiatry 82(4): 267–274. 32. Young, S.N. 2013. The effect of raising and lowering tryptophan levels on human mood and social behaviour. Phil. Trans. R. Soc. Lond. B. Biol. Sci. 368(1615): 20110375. 33. Christiansen, E., Agerbo, E., Bilenberg, N., and Stenager, E. 2016. SSRIs and risk of suicide attempts in young people—A Danish observational register-based historical cohort study, using propensity score. Nord. J. Psychiatry 70(3): 167–175. 34. Molero, Y., Lichtenstein, P., Zetterqvist, J., Gumpert, C.H., and Fazel, S. 2015. Selective serotonin reuptake inhibitors and violent crime: A cohort study. PLoS Med. 12(9): e1001875. 35. Chichinadze, K.N., Domianidze, T.R., Matitaishvili, T.T., Chichinadze, N.K., and Lazarashvili, A.G. 2010. Possible relation of plasma testosterone level to aggressive behavior of male prisoners. Bull. Exp. Biol. Med. 149(1): 7–9. 36. Comings, D.E. and Blum, K. 2000. Reward deficiency syndrome: Genetic aspects of ­behavioral disorders. Prog. Brain. Res. 126: 325–341. 37. Volkow, N.D., Fowler, J.S., and Wang, G.J. 2003. Positron emission tomography and single photon emission computed tomography in substance abuse research. Sem. Nuclear Med. 33(2): 114–128. 38. Dos Santos, M., Cahill, E.N., Bo, G.D., et al. 2018. Cocaine increases dopaminergic c­ onnectivity in the nucleus accumbens. Brain Struct. Funct. 223(2): 913–923. 39. Blum, K., Febo, M., Thanos, P.K., Baron, D., Fratantonio, J., and Gold, M. 2015. Clinically combating Reward Deficiency Syndrome (RDS) with dopamine agonist therapy as a paradigm shift: Dopamine for dinner? Mol. Neurobiol. 52(3): 1862–1869. 40. Qadeer, M.I., Amar, A., Mann, J.J., and Hasnain, S. 2017. Polymorphisms in dopaminergic system genes: Association with criminal behavior and self-reported aggression in violent prison inmates from Pakistan. PLoS One 12(6): e0173571. 41. Young, S.E., Smolan, A., Corley, R.P., et al. 2002. Dopamine transporter polymorphism associated with externalizing behavior problems in children. Am. J. Med. Genet. 114(2): 144–149. 42. Cherepkova, E.V., Aftanas, L.I., Maksimov, N., and Menshanov, P.N. 2016. Frequency of 3” VNTR polymorphism in the dopamine transporter gene SLC6A3 in humans predisposed to antisocial behavior. Bull. Exp. Biol. Med. 162(1): 82–85. 43. Yun, I., Lee, J., and Kim, S.G. 2015. Dopaminergic polymorphisms, academic achievement, and violent delinquency. Int. J. Offender Ther. Comp. Criminol. 59(13): 1409–1428. 44. Takeuchi, H., Tomita, H., Taki, Y., et al. 2015. Cognitive and neural correlates of the 5-repeat allele of the dopamine D4  receptor gene in a population lacking the 7-repeat allele. Neuroimage 110: 124–135. 45. Cherepkova, E.V., Maksimov, V.N., Aftanas, L.I., and Menshanov, P.N. 2015. Genotype and haplotype frequencies of the DRD4  VNTR polymorphism in the men with no history of ADHD, convicted of violent crimes. J. Crim. Justice 43(6): 464–469. 46. Beaver, K.M., Vaughn, M.G., Wright, J.P., DeLisi, M., and Howard, M.O. 2010. Three dopaminergic polymorphisms are associated with academic achievement in middle and high school. Intelligence 38(6): 596–604. 47. Wu, T. and Barnes, J.C. 2013. Two dopamine receptor genes (DRD2 and DRD4) predict psychopathic personality traits in a sample of American adults. J. Crim. Justice 41(3): 188–195. 48. Beaver, K.M., Wright, J.P., DeLisi, M., et al. 2007. A gene × gene interaction between DRD2 and DRD4 is associated with conduct disorder and antisocial behavior in males. Behav. Brain Funct. 3: 30. 49. Haworth, C.M.A., Kovas, Y., Dale, P.S., and Plomin, R. 2008. Science in elementary school: Generalist genes and school environments. Intelligence 36(6): 694–701. 50. DiLalla, L.F., Bersted, K., and John, S.G. 2015. Peer victimization and DRD4 genotype influence problem behaviors in young children. J. Youth Adolesc. 44(8): 1478–1493.

212  The chemistry of the brain

51. Boardman, J.D., Menard, S., Roettger, M.E., Knight, K.E., Boutwell, B.B., and Smolen, A. 2014. Genes in the dopaminergic system and delinquent behaviors across the life course: The role of social controls and risks. Crim. Justice Behav. 41(6): 713–731. 52. Beitchman, J.H., Mik, H.M., Ehtesham, S., Douglas, L., and Kennedy, J.L. 2004. MAOA  and persistent, pervasive childhood aggression. Mol. Psychiatry 9(6): 546–547. 53. Brunner, H.G. 1996. MAOA  deficiency and abnormal behaviour: Perspectives on an­association. Ciba Found. Symp. 194: 155–164. 54. Brunner, H.G., Nelen, M., Breakefield, X.O., Ropers, H.H., and van Oost, B.A. 1993. Abnormal behavior associated with a point mutation in the structural gene for monoamine oxidase A. Science 262(5133): 578–580. 55. Caspi, A., McClay, J., Moffitt, T.E., et al. 2002. Role of genotype in the cycle of violence in maltreated children. Science 297(5582): 851–854. 56. Haberstick, B.C., Lessem, J.M., Hewitt, J.K., et al. 2014. MAOA genotype, childhood maltreatment, and their interaction in the etiology of adult antisocial behaviors. Biol. Psychiatry 75(1): 25–30. 57. Byrd, A.L. and Manuck, S.B. 2014. MAOA, childhood maltreatment, and antisocial behavior: Meta-analysis of a gene-environment interaction. Biol. Psychiatry 75(1): 9–17. 58. Weder, N., Yang, B.Z., Douglas-Palumberi, H., et al. 2009. MAOA genotype, maltreatment, and aggressive behavior: The  changing impact of genotype at varying levels of trauma. Biol. Psychiatry 65(5): 417–424. 59. Beach, S.R., Brody, G.H., Gunter, T.D., Packer, H., Wernett, P., and Philibert, R.A. 2010. Child maltreatment moderates the association of MAOA  with symptoms of depression and ­antisocial personality disorder. J. Fam. Psychol. 24(1): 12–20. 60. Foley, D.L., Eaves, L.J., Wormley, B., et al. 2004. Childhood adversity, monoamine oxidase A genotype and risk for conduct disorder. Arch. Gen. Psychiatry 61: 738–744. 61. Ducci, F., Enoch, M.A., Hodgkinson, C., et  al. 2008. Interaction between a functional MAOA  locus and childhood sexual abuse predicts alcoholism and antisocial personality disorder in adult women. Mol. Psychiatry 13(3): 334–347. 62. Stogner, J.M. and Gibson, C.L. 2013. Stressful life events and adolescent drug use: Moderating influences of the MAOA gene. J. Crim. Justice 41(5): 357–363. 63. Aslund, C., Nordquist, N., Comasco, E., Leppert, J., Oreland, L., and Nilsson, K.W. 2011. Maltreatment, MAOA, and delinquency: Sex differences in gene-environment interaction in a large population-based cohort of adolescents. Behav. Genet. 41(2): 262–272. 64. Wells, J., Armstrong, T., Boisvert, D., Lewis, R., Gangitano, D., and Hughes-Stamm, S. 2017. Stress, genes, and generalizability across gender: Effects of MAOA and stress sensitivity on crime and delinquency. Criminology 55(3): 548–574. 65. McSwiggan, S., Elger, B., and Appelbaum, P.S. 2017. The forensic use of behavioral genetics in criminal proceedings: Case of the MAOA-L genotype. Int. J. Law Psychiatry 50: 17–23. 66. Reti, I.M., Xu, J.Z., Yanofski, J., et al. 2011. Monoamine oxidase A regulates antisocial personality in whites with no history of physical abuse. Compr. Psychiatry 52(2): 188–194. 67. Beaver, K.M., Barnes, J.C., and Boutwell, B.B. 2014. The 2-repeat allele of the MAOA gene confers an increased risk for shooting and stabbing behaviors. Psychiatr. Q. 85(3): 257–265. 68. Brooks-Crozier, J. 2011. The nature and nurture of violence: Early intervention services for the families of MAOA-low children as a means to reduce violent crime and the costs of ­violent crime. Conn. L. Rev. 531(2): 531–573.

10 Traumatic brain injury and neurocognitive disorders

Introduction This chapter examines brain damage and its many causes and implications for our study of biology and criminality. Because the brain is the seat of all behavior, it is logical that any damage to the brain is likely to affect our actions. Although brain damage can be severe at any age, damage to the developing brain is most severe. Brain damage is surprisingly common in society, and modern medical techniques have made it possible for us to understand much more about the brain and brain damage than in the past. In many cases, even minor injuries to brain function can have a major influence on future behavior. Of course, this behavior is not usually criminogenic on its own, but it can lead to criminal behavior, such as increased impulsivity, irritability, and reduced behavioral inhibition. Brain damage can be caused by a direct injury such as a head injury in a vehicular accident, or from child abuse, but it can also be caused by illness and disease. This chapter looks at a number of important studies focusing on the effect of this type of damage on antisocial and criminal activity. A correct understanding of the types of brain injury and their results can also help in finding ways to ameliorate their effects and treat offenders to reduce recidivism.

Head injury In the previous chapter, we discussed the ways that a chemical imbalance in the brain can affect behavior. If impeding or changing the tiny chemical messengers in the brain can impact behavior to such an extent, imagine the results of direct mechanical damage to the brain. It  is important to remember when discussing the impact of biological factors on behavior, and potential criminogenic behavior in particular, that the great majority are not biological factors that a person was born with. Many are directly related to environmental factors that change a person’s biology during their lifetime. Our biology changes depending on the way we live our lives. We are well aware that many things that we do or do not do will impact our health over time. For example, a nutritionally balanced diet and regular exercise is more likely to lead to general good health and longevity than a nutritionally poor or high fat and sugar diet, or the use of drugs, cigarettes, or alcohol. Ill health is a biological change, often caused by just such environmental factors. Such environmental factors can impact all aspects of life, including behavior. This is most obvious when we look at trauma or injuries. Social scientists often equate biological factors with genetics alone and do not realize that many biological factors that may predispose for crime are directly caused by environmental influences. This includes such events as head injury, exposure to neurotoxins, diet, and birth complications. Biology is not destiny; it is just life. It can be affected by many things that change it over a lifetime. 213

214  Traumatic brain injury and neurocognitive disorders

The brain is protected by the bony protective case of the skull. Inside the skull there are three layers of tissue called the meninges, which surround and cushion the brain. It is these membranes that are affected if a person develops meningitis, or swelling of the meninges, due to trauma or infection. The swelling puts pressure on the brain and can cause damage and even death. The brain is divided into two matching halves (or hemispheres), the right and the left, which are mirror images of each other. The two halves are separate, but they are connected by a bundle of nerve fibers called the corpus callosum. The brain can also be divided into lobes, or areas. The frontal lobe is the largest and takes up about one-third of the brain. The frontal lobe begins around the middle of the top of the skull to the most anterior part of the brain, so it includes the part of the brain above the eyes and the forehead. The front part of the frontal lobe, the part above the eyes, receives all of the body’s sensory information. The frontal lobe is involved in inhibiting aggression, and injury often results in impulsivity and lack of inhibition,1 so it is obviously very important from the point of view of antisocial behavior. Behind the frontal lobe lies the parietal lobe and beneath this is the occipital lobe. The temporal lobes are symmetrically placed on either side of the head. The central nervous system has both gray and white matter. The gray matter contains nerve cell bodies, dendrites, and axon terminals, as well as all the synapses, and is found in the outer layers of the lobes of the brain, in front of the brain stem, and along the central areas of the spinal column. White matter contains the axons, or sending parts of the neuron, which connect the different regions of gray matter together (see Figure 9.1). The brain is the basis for all behavior. Whether learned or genetic, the knowledge required to remember, retain, retrieve, and perform behaviors is contained in the brain. The  behavior may be simple (for example, eating, lifting a hand, walking) or highly complex (for example, thinking about and analyzing a situation and then reacting to it). Some are innate, or genetically controlled, behaviors and others are learned behaviors, such as language. Learning is a process that is both psychological and biological. As a person learns, the brain actually changes physically. It does not simply learn the information; rather, the acquisition of knowledge as environmental information is taken in, processed, and understood causes changes to take place in the biochemistry and cell structure of the brain. Most people are inclined to assume that a baby is born with a complete brain, which is a blank slate that only needs to be filled with information. This is not true: the brain grows and changes in response to the environment. Even genetically governed behavior is predisposed and designed to be changed by the environment. One of the brain’s normal functions is to be changed physically by experience; that is, the brain does not fully develop without experience. The brain is therefore a complex mixture of genetics and the environment, just like behavior. Our behaviors depend on the brain for storage, retrieval, and performance, so any damage to the brain can have a devastating effect. Brain development requires not just the genetics to develop the neural circuits but also input from the psychological environment and the cognitive environment. It is a bit like a computer that comes with all the hardware to do important things but is useless until you add software programs to it. For example, biology provides the neural base for prefrontal lobe development, but learning and experience are also necessary for complete development to occur. Traumatic brain injuries (TBIs) are injuries that result in a disturbance of normal brain function and can range from very mild, with sequelae that last just a few days, to devastating lifetime consequences. TBI can be the result of a penetrating injury, a direct strike to the head, or coup or contrecoup injuries, in which the brain moves in the skull, either at the site of the blow or at the opposite side of the skull, causing bruising and damage.2 The Centers for Disease Control and Prevention in the US states that TBI can result in diminished cognitive skills, diminished memory retrieval and retention, reduced sensations, and emotional disorders such as personality changes, anxiety, and depression.3 The most common causes of TBI are falls, vehicular accidents, and being struck by or striking an object (79%), with 11% due to assault. Of course, not all head injuries result in a TBI, as many minor bumps may leave nothing more than a bruise. Assessing the severity of a TBI often involves determining whether the person suffered a loss of consciousness (LOC) and if so, for how long. The Glasgow Coma Scale is often used to assess severity.2 It is used to asses a number of responses, including verbal, motor, and eye-opening, with

Head injury  215

scores ranging from 15 (best) down to below 8 (indicating coma) and 3  or less (unresponsive). Alternatively, the length of time that a person was unable to process new memories may also be used to rank severity.2 A concern when looking at TBI and offending is the old chicken-and-egg conundrum. Did the TBI cause a previously non-criminal person to develop risky behaviors, or did the intrinsic risky behavior result in a TBI, which exacerbated the problem? Some studies indicate whether or not the TBI occurred prior to offending, but not all. A further concern is that people who have suffered TBIs and who then find themselves in conflict with the criminal justice system may be more vulnerable than other offenders, as they may have a reduced ability to discourse with police or legal counsel or understand instructions, further increasing their risk.2

Frontal lobe injuries Behavior is affected by various parts of the brain. However, much of it is controlled by the frontal lobe. The frontal lobe is the part of the brain that extends over the eyes into the forehead. Think about what its position means. As this area is at the front of the face and sticks out, it is one of the most likely parts to be injured in an accident. In a car accident, it is the first part to hit the windshield. Because it is also the area involved in inhibiting inappropriate aggression or violence, an injury here is most likely to have an influence on behavior. Although damage to any area of the brain can cause behavioral changes, the most studied and most commonly injured area is the frontal lobe, and it is the area most commonly related to crime. However, it interconnects with many regions of the brain, so damage in other areas may link to the frontal lobe. Another part of the brain known to be heavily involved in emotion and behavior is the limbic system. This interconnected area is found within several of the lobes and includes the amygdala, the hippocampus, and the hypothalamus.4 The limbic area is primarily associated with emotion and the expression of emotion in relation to survival, which includes a faster heartbeat, faster respiration, trembling, sweating, facial expressions, and basic primary drives such as sex, hunger, thirst, flight (as in running away from a conflict), and fight (such as defense and attack). It is also involved in memory and retrieval of memories that induce intense emotion.4 The  amygdala is critical in communication between the prefrontal cortex (the front of the frontal lobe) and the hypothalamus, and damage to this region can result in many signals from the amygdala for behaviors that are no longer inhibited.1 Much evidence shows that damage to, or malformation (such as a birth defect) of, the frontal lobe can have significant implications for a person’s day-to-day functioning, emotions, and inhibitions. Frontal lobe damage can occur in a number of different ways: mechanical damage (for example, a blow to the head or a car accident), a developmental malformation (for example, premature suture closure or Korsakoff ’s syndrome), or disease (for example, a brain tumor). Damage to the frontal lobe has been linked to increased anger, violence, and irritability, decreased inhibition of inappropriate behavior, and an inability to understand the consequences of actions, which can increase impulsivity.5 People who have suffered frontal lobe trauma may also lose their social graces, including self-control and patience, so they demand immediate gratification rather than understanding the need to work to obtain a long-term goal. They may also exhibit changes in personality, anxiety, and depression.1 Such individuals often possess very few planning skills. A classic example of poor planning skills is given in Lewis’s book Guilty by Reason of Insanity.6 She cites the case of a death-row inmate who saved the pecan pie from his last meal so that he could have a midnight snack after his execution. These potential outcomes derive from case histories, so obviously they do not  represent controlled experiments. Some symptoms might appear in only a few subjects, and others might occur in most subjects. The  results of damage to the frontal lobe vary greatly from person to person. The extent of the damage itself varies, but so does the pre-trauma profile of the injured person, which will have an impact on how much damage is sustained and how the person will respond both to the damage and to any subsequent treatment.

216  Traumatic brain injury and neurocognitive disorders

The  frontal lobe is very sensitive to problems because it has an extraordinarily high demand for oxygen and nutrients. The frontal lobe needs very high levels of arousal to function. Even in “normal” individuals, frontal lobe functioning can deteriorate when they are tired, and this effect is much worse for people with brain damage. Most of us have noticed that when we have been working hard all day doing things that involve the brain (for example, thinking or writing a paper), as opposed to working physically (for example, gardening), our brain begins to shut down and our cognitive ability to analyze the work gets increasingly more difficult, until eventually we stop understanding what we are reading. The next day we come back to it fresh, and it is easy. For proper frontal lobe function, a person must be wide awake and fresh; even tiredness affects this functioning, so any brain damage makes mental work much more difficult. Head trauma is one of the most obvious ways that the frontal lobe can be damaged. Head trauma can occur due to accidents, fights, sports, or falls. Difficult births can result in brain damage, as already discussed. Another, and unfortunately very common, method of head injury is child abuse. Remember that children’s brains are still developing, so a brain injury in a young child is likely to have a much worse prognosis than the same level of damage in an adult. Brain damage in childhood may also be due to maternal use of alcohol during pregnancy resulting in fetal alcohol spectrum disorder (FASD), maternal use of drugs, lead poisoning, malnutrition, or many other things. Case examples of frontal lobe injury

One of the most terrifying examples of frontal lobe injury is also the first to be documented and the most famous: the case of Phineas Gage. Because the frontal lobe of the brain is the main site of our personality, damage to this area can dramatically change a person’s personality but leave the memory and intellect unchanged. This is a terrifying idea when you think about it. In  1848, Phineas Gage was a 25-year-old supervisor of a railroad construction crew. He  was packing gunpowder into a rock with a tamping iron when it prematurely ignited. The tamping iron shot up through his left cheek and went right through the top of his skull. The iron was 3 cm wide and 109 cm long.7 The injury was terrible, but Gage was able to speak almost immediately afterwards and able to sit in the cart taking him home, where he got out and walked with only slight aid. When the attending physician, Dr Harlow, saw him, he was conscious but suffering from a loss of blood. The doctor could clearly see the brain through the hole in the skull. He observed the rod and saw it was covered in brain matter, and rather graphically, he stated that shortly after he arrived to treat him, Gage vomited, and the action “pressed out about half a teacupful of brain, which fell upon the floor.”8(p16) Despite this unbelievably horrific accident, and the limitations of medicine in the nineteenth century, Gage survived, but his frontal lobe was destroyed. He recovered from this terrible injury and eventually went back to work. His memory, intellect, speech, and ability to work were not damaged, but according to his family and friends, he was simply “no longer Gage.”7(p1102) His personality had dramatically altered. He had previously been a kind, soft-spoken, polite, and considerate young man. After the accident, he became profane, irritable, irresponsible, and violent—in short, a different person. Although he had retained his intelligence and the knowledge and technical skills to perform his job, he had lost his social graces and his ability to work with others, so he lost his job and wandered for years. He died 11 years after the accident, of a series of epileptic seizures. This case was the start of understanding that certain parts of the brain may have specific responsibility for social behaviors and reason.7 Gage’s death passed unnoticed, and it was several years afterwards that Dr Harlow, his original attending physician, appealed to his family to allow him to exhume and study his skull. Gage had been buried with the tamping iron, which he had carried always, so the two could be examined. These were preserved and a reexamination of Gage’s skull more than 140 years after the accident indicated that the metal rod passed through the ventromedial portion of his prefrontal cortex, and his described personality and behavioral changes correlated well with those seen in modern patients with similar trauma.7 In 28 patients with frontal lobe damage, 12 had similar damage. These 12 patients, like Gage, were perfectly capable of performing mathematical analyses, logically solving abstract problems, and remembering information, but

Head injury  217

they were unable to make appropriate social and personal decisions.7 The ventromedial portion of the frontal lobe is involved in social decision making, emotion, and behavior, and animal studies have shown a high number of serotonin receptors in this area.7 What is so terrifying in this case is that Gage was able to function without a large part of his brain, yet his personality changed so dramatically. We can accept that someone might be damaged after an accident, have a limp or a scar, but for the person to still look just like the person you knew and loved but be no longer there anymore—“no longer Gage”? This situation is difficult to accept or adjust to. A more modern example concerns a self-employed 51-year-old man who, in 1994, was happily married with nine children. He was injured in a severe car accident and received a depressed bone fracture to his frontal lobe (a closed injury). Before the accident, he had no history of mental illness (personally or in his family) and was a hard-working, energetic, optimistic man with a successful and happy family life.9 After the accident, his intelligence, cognition, and memory were unaffected; however, his personality changed dramatically, just as Phineas Gage’s did. He had sudden unprovoked outbursts of anger, triggered by trivial things. His behavior became very uninhibited: he would make obscene sexual remarks and would sexually touch any females around (not just his wife but also his children, friends, and family).9 He became hypersexual and placed abnormal sexual demands on his wife; he became suspicious, verbally and physically aggressive, indifferent to people’s reactions to his behavior, and irritable and anxious, and he exhibited poor concentration and judgment.9 Of course, as before in case studies, the findings from one case cannot be generalized to all people with frontal lobe injuries, but this case is a classic example of the kinds of changes that can take place and is almost a modern repetition of the symptoms seen in the case of Phineas Gage, although with a closed injury. “Alfred” was a 20-year-old pizza delivery person who had a normal childhood, was never aggressive, was appropriately behaved and respectful to women, and was generally a nice, kind young man.10 In  a traffic accident, he received a TBI causing damage to the orbitofrontal cortex and ­parietal regions of the brain. His behavior changed dramatically: he would catch and kill stray cats, would steal from stores and become violent if confronted for the theft, and became physically and verbally abusive with women and inappropriately sexual in their presence. He masturbated in front of houseguests and attempted to rape several women, including his mother, repeatedly. Eventually his father could not leave him alone with his mother, and he became physically violent toward his father. He could not be reasoned with, and his parents committed him to a rehabilitation center focused on brain injuries. Within the facility he continued with impulsive and uninhibited ­v iolent and sexual behavior. After 4 months of multidisciplinary and intensive treatment, his ­sexual behavior toward his mother ceased, and after 6 months his aggression was under control. After 20 months of treatment he was able to return to his family home, obtain supervised employment, and live in an appropriate manner.10 Korsakoff syndrome is a form of brain damage brought on by extreme and long-term alcoholism. “Philip” was a successful electrical engineer with a normal childhood and successful school and college career, married with no children.10 He drank throughout his adult life, and his increased drinking led to frequent fights with his wife in which he would become violent with objects, such as punching walls. He  developed paranoid behavior, convinced his wife was unfaithful. In  one fight he punched his wife in the face, and she filed for divorce. He  began stalking his ex-wife, stealing her mail and watching her with her new boyfriend. One day he broke into her house and assaulted her, breaking her arm, but she was successful in locking herself in a room and calling police. Philip was arrested at the scene and immediately confessed. Philip’s lawyer noticed that Philip had memory problems, as he kept reintroducing himself to the lawyer, even shortly after they had been in lengthy meetings. The lawyer requested a competency exam. Neuroimaging showed that large areas of Philip’s brain, including the frontal lobes, cerebellum, and hippocampus, were severely atrophied, and testing revealed deficits in memory, psychomotor control, and executive functioning.10 He was diagnosed with Korsakoff syndrome and found to be incompetent to stand trial. He was incarcerated in a mental health facility and is unlikely to be ever considered fit to leave, due to his permanent brain damage.10

218  Traumatic brain injury and neurocognitive disorders

Traumatic brain injuries in youth The  brain is not  fully developed at birth: it continues to develop and change biochemically for years. Most development is completed in the early twenties, but the brain is plastic and continues to change and adapt throughout life. Some brain structures in youth develop faster than others. For example, the limbic area of the brain is not well developed until after adolescence, meaning that adolescents may misread a protagonist’s emotion and are more likely to rely on emotion than sound decision-making, resulting in impulsivity.4 The prefrontal cortex is one of the last to mature, completing development in the mid-twenties, and is extremely important in social behavior and the inhibition of inappropriate behavior.4 It is the part of the frontal lobe found just behind the forehead, so is a frequently injured area, in a vehicular accident or in child abuse. It is involved in cognition, abstract thought, and behavioral responses. During adolescence, the amount of white matter in the brain increases and creates more and more neurocircuitry between gray matter regions of the brain, resulting in better communication between brain regions. It greatly increases in the corpus callosum, the part of the brain that connects the left and right hemispheres.4 The influence of the environment and experience is vital in healthy development of the brain, as the amount of gray matter peaks at around age 11 (girls) and 12 (boys) and then is reduced and refined by synaptic pruning, moderated by experience and the environment. This eliminates connections that are not commonly used and structurally refines the brain based on experience, increasing efficiency, and specialization in different parts of the brain.11 Risk-taking, which tests people, results in new experiences, and allows a better understanding of the world around them, can be both beneficial and dangerous and appears to be a normal adolescent activity, allowing them to develop and maximize their potential; this development of the mind continues late into adolescence.4 In the past, it was believed that the brain completed development by puberty, but neuroimaging has shown that it is not complete until age 25. This has an impact on legal policy, as we consider a person to be an adult at 18 for most criminal issues, yet we know now that the adolescent brain is not yet fully developed, and many studies have shown that the adolescent brain is quite different from the adult brain. Therefore, should we consider a person under 25 to have the same culpability as an adult or as a child?11 In the United States and in many other countries, we argue that people over 18 are adults for the purposes of criminal offending, being conscripted into the armed forces (and potentially dying for their country), marrying (and having children), and voting for their government, yet not mature enough to make decisions about alcohol consumption. Thus, there is a conflict here that frequently plays out in court when deciding culpability or in determining sentence. The developmental phase of the brain during which injury occurs will thus have an effect on the severity of damage to the brain. For  example, as the frontal lobe is involved in control and inhibition, there is often a lesser impact from frontal lobe trauma on adults (who normally have already developed good internal control systems) than on children. Children, however, have not yet developed all their internal control mechanisms, nor learned socially acceptable behaviors. Brain damage need not be the result of direct trauma but can result from disruption of the normal environment required for healthy brain development during the early years. Many animal studies have shown that just small stressors can impact brain development. For example, a study on baby mice showed that maternal separation for just 3 hours a day resulted in deficits in locomotor ability, memory, and brain chemistry.12 In a case study of two young people who suffered TBIs as babies, one from a brain tumor and one from an accidental head injury, both recovered physically but developed serious antisocial behavioral problems as they grew up.13 As young adults, they were intelligent but exhibited both antisocial and amoral behaviors. The young woman, although academically intelligent, stole from family and friends, lied, was physically and verbally abusive, was sexually promiscuous, and was entirely lacking in empathy toward anyone, including her own baby. The  young man was reckless, slovenly, and unmotivated and committed both property theft and violent assaults. He had also sired a baby and showed no interest in the child. Neither person showed any remorse or guilt

Traumatic brain injuries in youth  219

for their actions, and both resembled psychopaths in much of their behavior, although in general they were more impulsive than directed and more childlike in their behavior.13 Both subjects came from loving and stable family homes with well-adjusted siblings. Importantly, although both young people exhibited the same sort of deficits seen in adults with prefrontal cortex damage, such as a failure to consider the consequences of their actions, a desire for immediate gratification, and a lack of social graces, they also exhibited defective social and moral reasoning.13 The authors suggested that people who are injured as adults possess the knowledge of social and moral rules, although their injury may prevent them from following them, whereas those injured as infants or children may never have been able to acquire the normal understanding of social graces and moral reasoning before their impairment. Therefore, such children would never be able to grasp basic social rules, which could result in much greater levels of antisocial and amoral behavior than those seen in similarly injured adults. This may mean that such children would be much more difficult to treat than similarly injured adults whose brains had developed normally prior to the injury. TBIs in children and youth can result in many neurocognitive deficits, including impulsivity; reduced cognitive ability, social skills, empathy, and communication skills; lack of inhibition of overly aggressive responses to perceived threats; and inability to respond to other people’s emotions suitably. Many of these problems lead to poor academic success in school and risk for developing unfavorable peer associations.2 These deficiencies and their corollaries are frequently considered to be predispositions for risky behavior. Neurocognitive deficiencies are highly correlated in the literature with both early onset and lifelong criminal activities.2 Child abuse is a major cause of brain injury. It can occur from just one slap to the head, because a child’s head is very soft. Shaking a baby roughly can cause whiplash injuries that shear the fiber tracts. Even a single blow to the head can cause multiple lesions in the brain. Such injuries are probably very common; they leave no external mark, and thus they are not discovered and reported. The majority of serious head injuries in infants are the result of child abuse. Many studies showing high rates of head injury in violent offenders also report high rates of severe child abuse, which makes it likely that such people have had at least two TBIs. Thus, child abuse can add to a person’s risk of predisposition for criminal behavior, by first biologically damaging the brain and then by predetermining a criminogenic social environment. It is estimated that more than 3 million children globally receive a TBI, of which approximately 80% are ranked as mild.14 Mild TBI includes LOC up to 30  minutes, confusion, disorientation, and memory loss. Even mild TBI can have significant functional effects. TBI in children, whose brains are developing, can cause disruptions in any part of the brain and often impact regions concerned with emotional and social performance.15 Social and emotional maturation in the brain is referred to as the “social brain,” and this has a slow and complex maturation involving many brain regions, including the temporal lobes, prefrontal cortex, and the limbic system.15(p284) Youthful play is ­considered to be one of the main instruments of developing a healthy social brain, as it has many evolutionary benefits, including developing social competence, appropriate emotional communication, and the ability to work, interact with, and assess peers, all of which increases behavioral adaptability.15 Although many issues, such as reduced cognitive functioning, are implicated in poor peer relationships and long-term risk, studies have shown that TBI may greatly impact social development, leading to poor social interactions and peer relationships. Studies on children between 8 and 13 years old with a severe TBI, a mild to moderate TBI, or an orthopedic injury (OI) as a control showed that despite normal intelligence, children who had a TBI were less likely to have mutual friends (that is, to be named as a friend by several children), were more likely to be victimized and rejected, and were rated lower in popularity and acceptance.16 Unfavorable peer relations in children with severe TBIs were linked to a decrease in the white matter of the brain in areas related to the processing of social information.16 Animal studies with induced TBIs indicated that even mild TBIs altered play in young rats, and non-TBI rats avoided playing with TBI rats and particularly excluded and rejected female TBI rats. Housing TBI male rats with non-TBI conspecifics allowed the male TBI rats to learn normal play tactics, but not the females, whose responses worsened.15 Similarly, studies have shown that children with TBI may

220  Traumatic brain injury and neurocognitive disorders

have social impairments, particularly in the first 6 months after the trauma. Social participation with peers was a major problem and increased with increased severity of injury and reduced communication skills.17 Such studies show not only how the juvenile brain develops and adapts and the impacts of TBI on the developing brain but also how the TBI affects the development of such a normally accepted action as play. If normal play strategies are disrupted, the social brain does not develop properly, leading to poor peer relationships. Although, on its own, this has nothing to do with criminality, poor peer relations are linked to poor academic outcomes, and poor peer relations will lead to making unfavorable peer choices and by this route can lead to antisocial behavior. More importantly, such research indicates methods that could be used to assist youth with developing the social brain, such as the use of play therapy.15

Prevalence of young offenders with traumatic brain injury In a large longitudinal birth cohort study in the United Kingdom, over 11 500 youth under 16 years of age were assessed for mild TBI, OI, and no trauma.14 At age 17, they were re-assessed for substance use, criminal behavior, and psychiatric symptoms. Youth with TBIs exhibited increased dangerous alcohol use, over both OI youth and non-traumatized youth, and a higher risk of criminal behavior and psychiatric symptoms over non-traumatized youth.14 Timing of injury impacted outcome in that children who received a mild TBI prior to age 11 were at higher risk of psychiatric symptoms by age 17, whereas youth who received a TBI between 12 and 16 years of age had higher risks for criminal behavior and substance abuse.14 OI was used as a negative control, and in this study, both types of trauma were associated with increased offending, although only the TBI group had been in trouble with law enforcement. The authors suggested that other forms of trauma, such as breaking bones, may be evidence of different risk factors, such as sensation-seeking.14 The most important findings in this study were that even mild TBIs can increase risk factors, as well as indicating that there is a clear link between TBIs and alcohol abuse. This has also been seen in other studies.18 Alcohol abuse is well known to increase recidivism and crime, in particular violent crime, so could be an exacerbating factor for TBIs and recidivism. A Swedish study looked at hospital records over a 35-year period and found a 5.8% increase in violent criminal activity after a TBI in childhood when compared with non-injured siblings and the general population.19 Similarly, in Finland an unselected birth cohort of over 12 000 individuals was followed until age 31 to determine the impact of TBIs in childhood. TBI was found to double the risk of developing mental disorders and quadruple the risk of criminal offending in males.20 In the United Kingdom, neurocognitive and mental health assessments of 93 male youth incarcerated in a secure facility revealed that 82% had experienced at least one TBI, and almost half reported that they had ongoing neurological and psychological sequelae.21 Many studies have shown that incarcerated youth have much higher levels of prior traumatic brain injuries than non-incarcerated and non-delinquent youth. A review of 10 such studies from the United States, the United Kingdom, and Australia showed that TBI prevalence rates within incarcerated youth varied between 16.5% and 72.1%.2 There  were many differences between the studies, in population studied, method of assessing TBI (such as self-report, interview, medical screen, neuropsychiatric evaluation, or medical records), and level of injury, probably accounting for at least some of the variation seen between studies.2 In the UK study, prevalence of TBI in young incarcerated males between 16 and 18 years old was 72.1%, with over 40% having suffered an LOC.22 Alcohol use also increased with severity of the TBI, as is seen in other studies.22 Only a few of the reviewed studies included a control group with which to compare. In one study, when a very broad definition of head injury, including scrapes and cuts, was used, there was virtually no difference between incarcerated youth and matched controls. However, when more rigorous definitions of head injury were used, such as resulting in a concussion, 16.5% of offenders had received a TBI, versus 11.7% of matched controls, and when the severity of the concussion increased, the ratio was

Traumatic brain injuries in youth  221

3.5:1.5.2 In another of the studies, 55% of the offenders had received a head injury requiring medical attention, in comparison with 24% of a matched control group.2 Another of the studies looked at TBI only in the preceding 12 months and found that 12.5% of incarcerated youth had received a head injury, in comparison with 5.8% in a control group of school children.2 Rates of head injury in the general population in a number of studies from Canada, the United States, and New Zealand ranged from 20% to 35%, with severe TBI in approximately 15%. Only one study indicated a much lower range.2 The studies showed that the more serious the TBI, the higher the prevalence in the incarcerated population in comparison with controls. LOC for any time period resulted in prevalence rates between 32% and 50%, compared with surveys of college students, which indicated rates between 5% and 24%. When LOC exceeded 20 minutes, TBIs were almost quadrupled in the ­incarcerated population.2 Some of the studies considered multiple TBIs. In the UK study, almost 46% had received more than one TBI and almost 23% indicated they had received at least four TBIs.22 One Australia study indicated that 13% of incarcerated individuals in their study had lost consciousness at least twice.2 Data from New Zealand and the United States indicate that the number of people who have experienced more than one TBI in the general population is 9%–12%.2 One of the most disturbing reports considered in this review was an earlier study from 1988 in which 14 juvenile death-row inmates had suffered serious head injuries prior to their offense, so their violent lifestyle was not the cause of the TBI.23 Of these 14 inmates, nine had major brain impairment, seven suffered from severe psychotic problems that began prior to incarceration, seven had severe organic dysfunction, and only two had IQ scores above 90, which means that nearly all of them could be considered intellectually disabled. Almost all had also suffered severe physical abuse. Perhaps more disturbing yet is that these children were on death row. In 2004, 71 individuals were on death row for crimes committed as juveniles. In 2005, the US Supreme Court banned the death penalty for juveniles, calling it “cruel and unusual punishment.”24 However, the United States still has the highest number of incarcerated youth in the world, with 10 times the incarnation rate of other developed countries.25 Studies have also shown an increase in recidivism in juveniles with TBIs. In a study of 186 young male offenders in the United Kingdom, 46% had received a TBI with LOC, with 29.6% reporting LOC resulting in mild TBI and 16.6% reporting moderate to severe TBI. Almost a third reported repeated TBIs.26 The number of convictions increased with number of TBIs, and increased violence was noted in offenders with more than three TBIs.26 Those with repeated TBIs also had higher ­levels of mental health disorders,26 although timing of TBI was not presented. In a study of over 4000 juvenile offenders in Texas, using rigorous testing, 21.9% to 41.3% were found to have had at least one TBI, and more importantly, the majority received their TBI prior to their offence, from assault, sports, abuse, falls, and vehicular accidents, indicating their TBI was not a result of a criminal lifestyle.27 These studies show not only that is there a very high proportion of incarcerated youth who have suffered at least one, and sometimes many, TBIs, and that increased numbers and severity of TBIs increase recidivism and violence, but also that there is a great need for intervention. Despite the high rates of TBI in youth, most juvenile correctional facilities do not routinely screen inmates for TBI.27 As TBIs can result in a range of neurocognitive deficits, it is probable that normal treatment programs are unlikely to be beneficial to such youth, and specialized and individualized intervention and treatment plans should be developed.

Gender differences in youth with traumatic brain injury There  has been debate about the number of female youth in the corrections system who have received TBIs. Most of the studies on TBIs only considered males, but some also included females, with conflicting results. A  study in the United States of 720  incarcerated youth, 80% of whom were males, reported that almost 20% of incarcerated males and only 9.6% of females had suffered

222  Traumatic brain injury and neurocognitive disorders

at least one TBI, although sample size for females was small. Timing of TBI was not examined.28 An Australian study also reported that 37.7% of incarcerated male youth had suffered a TBI, in comparison with only 5.3% of females.2 In contrast, two studies, also from the United States and Australia, reported almost identical rates, with 50% of incarcerated males and 49% of females reporting TBIs in New York (22% of the sample being female) and TBIs primarily resulting from assault and falls,29 and 32.1% of males and 33.3% of females in New South Wales, although only 12% of the sample were female. Causes of TBI included assault, particularly in females, and sports, particularly in males.18 The differing results here are interesting, as the literature consistently reports that TBIs are much more common in young males,2 but this may result from small sample sizes of female offenders. A study in France of adult female offenders reported an incidence of 21% TBIs, with 8% in juvenile offenders (although sample size for juveniles was very small). TBIs were caused by vehicular accidents, sports, and falls, with a third related to violence. Three-quarters of the TBIs occurred prior to imprisonment.30 Males and females have also been shown to have differing responses to childhood TBI. A New Zealand study compared 65 non-incarcerated adults who had experienced a mild to severe TBI during childhood with a matched group who had experienced an OI. It  looked at substance abuse, ­criminal offending, anxiety, and depression, and found major differences between groups and sexes.31 Adults who had suffered any form of TBI had much higher levels of problem behaviors than those who had OI. Interestingly, females had a much higher history of internalizing behaviors, such as fearfulness, social withdrawal, and anxiety, than males, whereas males reported much higher levels of externalizing behaviors, such as physical aggression, disobeying rules, cheating, stealing, and property destruction.31 This is an important issue, as internalizing behaviors are often missed by medical professionals because they are less obvious and involve less antisocial behavior than externalizing behaviors, but they are clearly an important factor, so must be considered in treatment plans.31

Which came first, traumatic brain injury or risky behavior? The question is always whether the TBI was implicated in the later criminal behavior due to any of the myriad potential effects of a TBI or whether the TBI was a consequence of risky or even violent behavior. In the studies that reported timing of the TBI, in most cases, the TBI preceded the first offence. Even in the studies that do not report timing, they often include causes of TBIs, which are mostly non-violent, such as vehicular accidents, falls, and sports injuries. In  some of these studies, the cause is simply presented as assault, but as these are youth, it is quite possible that the assault may have been an assault by an adult, such as child abuse, and not necessarily the result of a criminal lifestyle. In a large US study of young offenders in Pennsylvania and Arizona, TBI was higher in males and was associated not only with impulsivity and antisocial behavior but also with higher levels of victimization and abuse.32 In a meta-analysis of a number of studies on TBI in youth, causes of TBI in delinquent youth included falls, vehicular accidents, sports, and fights, whereas non-delinquent youth primarily received TBIs from sports, which the author suggested shows that the delinquent behavior may result in the TBI rather than the other way around33; although other researchers have found an increase in child aggression after TBI, in comparison with prior, and increased numbers of TBI have been shown in several studies to increase violence and recidivism.26,33,34

Traumatic brain injuries and schizophrenia Childhood head injuries have also been linked to the development of schizophrenia and to later violent behavior. Many factors have been implicated in the etiology of schizophrenia, including genetics,35 prenatal starvation,36 and birth complications, and evidence suggests that mild childhood

Traumatic brain injuries in youth  223

head injuries may influence the expression of schizophrenia in families with a genetic predisposition for the disorder.37 In  a study of children from 23  families with a family history of multiple cases of schizophrenia, children with schizophrenia were compared with their unaffected siblings. The schizophrenic subjects were much more likely to have suffered from childhood head injuries than their unaffected siblings.37 Also, when considering only the schizophrenic individuals, those who had suffered head injuries developed schizophrenia approximately 5 years earlier than those without head injuries. All the head injuries had been judged to be mild, but within this category, the severity of the head injury correlated with the age of onset of schizophrenia.37 Differences have been noted in the brains of schizophrenics in comparison with healthy individuals, in particular with abnormalities in the frontal and temporal lobes in aggressive schizophrenics, with reduction in the orbitofrontal and temporal regions.38

Impact of traumatic brain injury on youth during criminal justice proceedings Clearly, there is a large percentage of youth in conflict with the law who have suffered one or more TBIs. Many of the sequelae of a TBI can result in cognitive impairment and deficits in language, including reduced comprehension and ability to express oneself clearly. Interactions with the criminal justice system are highly complex to navigate and require excellent cognitive and communication skills. Youth who have suffered a TBI may have deficits in language comprehension and expression.25 This is particularly so when the children received their TBIs early in life, when their language skills were first being learned and honed. Evidence suggests that the earlier the damage, the greater the language deficits later.25 Lack of recognition of emotion in others will also impact comprehension, as much of what we say is interpreted through body and facial language together with the spoken word, and a lack of understanding of such subtle signals may result in unsuitable responses and overreaction. Think of how many emails are misread when the reader does not realize that the sender is being ironic or flippant, with the resultant escalation of misunderstanding, which would not have occurred had there been the addition of body language to show that the writer was not serious. In a study of 16 adolescents with TBI and a matched group of 16 normally developing adolescents, TBI students performed equally on tasks involving recognizing sarcasm, lying, and emotion when provided with prompts but were less able to differentiate between sarcastic or sincere conversations when provided with no prompts except for a person’s demeanor.39 This again highlights the lack of the ability to interpret meaning in some situations. Youth with TBIs also suffer from deficits in the expression of language. They  may have a reduced vocabulary and slower processing speeds, but more importantly, they may suffer from deficiencies in “pragmatic language.”25(p87) Pragmatic language deficits include such things as not providing the same amount of information that a peer would provide, badly organized wording and speech, and language deficiencies that reduce the ability to explain what they have done or negotiate a reduced charge.25 If TBIs can increase a person’s risk of offending, then the deficits that result will also greatly impact their journey through the criminal justice system, making them much more vulnerable than non-TBI offenders. It has been suggested that youth with TBI face difficulties with each stage of the criminal justice system. 25 In  their first encounters with the police, they may not  answer questions well and may misinterpret the question or answer inappropriately, leading to potential further investigation, which may not have been warranted if they had normal verbal skills. When interrogated and read their rights, they may not  fully comprehend their entitlements, and at the trial stage, they may not be able to successfully communicate and discuss strategy with their lawyer.25 These issues may be similar to other youth with impairments such as FASD. Youth with both TBIs and FASD are not routinely screened for such deficits, and there is a call to increase screening, so that lawyers and the courts are aware of potential vulnerabilities.25

224  Traumatic brain injury and neurocognitive disorders

Traumatic brain injury in adults Numerous studies have linked TBI in adults to violence and crime, in both incarcerated and non-incarcerated populations. Some of the first studies showed a link between TBI and interpersonal or domestic abuse. In a seminal study of non-incarcerated individuals, the majority of men (61.3%) with problems of aggression against their families had suffered severe head injury.40 In  a follow-up study, the same research group compared the incidence of head injuries among men who had beaten their spouses; non-violent, happily married men; and non-violent, unhappily married men (n = 53, 45, and 32, respectively).41 Head injuries were medically assessed by a doctor who was unaware of their domestic and criminal history. More than half of the men who had been convicted of spousal abuse had suffered an earlier significant head injury (51%), as compared with only a quarter of the unhappily married, non-violent men and 16% of the happily married, non-violent men.41 Spousal abuse is often thought to stem from marital discord and general unhappiness with the relationship, but the dramatic difference between men who abuse and those who were unhappily married but non-violent shows that the abuse was more likely to be related to the head injury than to general marital unhappiness. Alcohol is also often presented as a predisposer for marital violence, but the researchers controlled for alcohol abuse and no differences were found between batterers and non-batterers.41 A meta-analysis of studies, which included 222 men who committed intimate partner violence, found that over half had received a TBI, which is significantly higher than the incidence of TBI in the general public, which studies suggest ranges from 10% to 35%.42 In a review of TBI studies, individuals who had suffered a TBI reported increases in loss of temper, loss of control, difficulty in communication, verbal abuse, violence, aggression, agitation, and frustration in comparison with prior behavior, and levels appeared to increase with severity of TBI.5 Overall, approximately a third of TBI patients exhibited changed temperaments, with increased antisocial behavior. Many caregivers reported concern over increased aggression, which could partially relate to the increased frustration and agitation often seen in TBI patients.5 The increased aggression is usually seen very shortly after the TBI but can remain for years, and a study showed that 25% of patients with moderate to severe TBI were still aggressive more than 5 years after injury, and in other studies, 20% still reported increased violence after 15 years post injury.5

Prevalence of traumatic brain injury in incarcerated populations In a study of incarcerated men and women from 6 US federal prisons, 88% had suffered at least one TBI,43 and in a meta-analysis of such studies, the average prevalence was 60% of offenders.44 In a more recent meta-analysis of studies of incarcerated offenders from the United States, Australia, and the United Kingdom, prevalence ranged from 9.6% to 100%, with an average of 46%.45 In Canada, prevalence of TBI in incarcerated men and women was 50.4% and 38%, respectively, and overall, 34% reported two TBIs and 22% reported three or more.46 In a review of TBI studies, both violent (approximately 61%) and non-violent (approximately 46%) offenders had suffered TBIs. Moreover, when considering released offenders in halfway houses, 50% of non-violent offenders had TBIs, 83% of whom reported the TBI occurred prior to their first criminal offense.5 Therefore, TBIs correlate with non-violent offending as well as violence. Several of the studies reviewed suggested that increased violence and antisocial behavior may be a result of loss of executive functioning and decreased IQ and verbal communication skills caused by the TBI, especially with frontal lobe damage.5 Moreover, those most severely aggressive had the lowest scores on executive functioning tasks.5 Not only are offenders more likely to have much higher prevalence rates of TBI than the general population, but they are also more likely to recidivate than those without TBI. A study in Ohio ­followed over 150 inmates after release and compared recidivism rates in those with TBIs and those

Traumatic brain injury in adults  225

without TBIs. Those with TBIs were more likely to be re-arrested post release than offenders with no TBIs, more likely to be arrested sooner, and more likely to have psychiatric diagnoses, as well as a higher number of arrests overall and a higher number of arrests for interpersonal crimes.47 Other studies have also reported higher recidivism rates in offenders with TBIs, as well as earlier age for first offense, increased prevalence of violent crimes, longer sentences, and higher rates of ­infractions while incarcerated.48

Traumatic brain injury and substance abuse Comorbidity of substance abuse and TBI is frequently reported in the literature. In a large study using the US Traumatic Brain Injury Model System National Database (TBIMSND), factors preand post-TBI were investigated. Rates of substance abuse pre-TBI were found to be higher than in the general population. Substance abuse is a well-known predisposer for antisocial behavior and offending, so it is possible that premorbid drug and alcohol use may lead an individual to risky behavior that could lead to TBI.49 Studies have shown a very high alcohol and substance abuse both pre- and post-TBI and this comorbidity greatly complicates treatment.46 Female offenders with TBI have been shown to have higher alcohol use than non-TBI offenders, 30 and dangerous use of alcohol with TBI was also seen in a non-offender birth cohort.14 Alcohol use post-TBI has been shown to consistently increase with severity of TBI.22 Alcohol depresses inhibitions, so when someone with frontal lobe dysfunction consumes alcohol, it may have a compounding effect. The inhibition for violence is significantly reduced by alcohol and compounded by brain dysfunction. So, substance abuse may precede the TBI as well as be exacerbated by it afterwards. On the other hand, impairment of frontal lobe activity has been found to be a risk factor for alcoholism and destructive drinking, as well as other forms of substance abuse, and may be a form of self-­ medicating. So, substance abuse may also be a direct risk factor of the TBI. Drug abuse and dependency in general is dramatically higher in incarcerated offenders than in the general population. In the United States, 50% of inmates are drug abusers, in comparison with 0.4% in the general population and 1.3% for alcohol abuse.50 Both drug abuse and TBI damage the brain and so may exacerbate the desire for further substance abuse, due to loss of inhibition, reduced cognition, and need for immediate reward. In a study of inmates in South Carolina, those with TBI before age 13 had greater drug use before age 13, with over 50% beginning drug use prior to the TBI. This relationship held for both males and females, although severity of drug abuse was higher in females.50 Offenders with earlier TBIs had higher drug use, higher aggression, and earlier age of initial drug use. Earlier age at drug-use onset predicted higher levels of aggression irrespective of TBI.50 It is therefore hard to say whether TBI is causal in increased substance abuse or merely aggravates an already existing predisposition for substance abuse. If the individual already had risk factors for substance abuse, then these factors, or the substance abuse itself, may have increased the risk of receiving a TBI. It has been suggested that the relationship between TBI and substance abuse is bidirectional, in that the two contribute to each other, with substance abuse increasing risk of receiving a TBI and the TBI being a catalyst for subsequent abuse.51

Was the traumatic brain injury causal? Again, as with youth, the question is whether the TBI caused the violent or antisocial behavior or whether the person already displayed such behavior prior to the TBI, which may then have exacerbated it. Cases such as that of Phineas Gage show that there can be a complete change in personality and behavior, and other cases suggest that some antisocial behavior may have been present prior to

226  Traumatic brain injury and neurocognitive disorders

the TBI but the TBI increased the levels of antisocial behavior. Children may be more vulnerable to TBI, as much of their social and neural development is not complete, but injury in adults can also have a very severe effect on personality and antisocial behavior. Some studies have considered behavior prior to TBI compared with after and found increased aggression after TBI.1 In four case studies, severe TBI resulted in greatly increased violence after the TBI, although all had previous risk factors.1 The first case was that of a 37-year-old male who suffered a severe TBI in a vehicular accident. Prior to the TBI, he had multiple substance abuses, with substance abuse arrests; however, after the TBI, he became extremely violent, committed a number of assaults, and exhibited inappropriate sexual behavior, resulting in three periods of incarceration.1 In the second case, a 20-year-old male suffered a severe TBI due to vehicular accident. Prior to the TBI, he was diagnosed with attention deficit hyperactivity disorder (ADHD), and he had also had a difficult childhood due to parental conflict. Post TBI, however, he became extremely verbally and physically aggressive to his caregivers, once kicking a pregnant staff member in the stomach, and his records indicated that he was considered to be extremely dangerous.1 In the third case, a 22-yearold male had two severe TBIs, the first from a vehicular accident while he was intoxicated and the second from a fall while in rehabilitation. His parents said that he was aggressive prior to the injuries but that these magnified after trauma. After the TBIs, he became extremely physically and verbally aggressive with hospital staff and caregivers, as well as to objects, punching windows and throwing items. He also exhibited inappropriate sexual behavior. Police were involved repeatedly, and the individual showed no remorse for his actions.1 The final case was that of a male who suffered a cerebral hemorrhage and two severe TBIs in his 20s. The first TBI was from an assault and the second from being dragged from his bed by an inmate while he was incarcerated for robbery. Prior to the injuries, he was a multiple substance abuser, including PCP (phencyclidine or angel dust) and ecstasy (3,4-methylenedioxymethamphetamine or MDMA), and did exhibit some antisocial behavior, but after the TBIs, he became extremely verbally and physically abusive and was incarcerated twice, once for robbery after the first TBI and once for assaulting a staff member and a fellow patient in residential care.1 In these four case studies, although each individual had risk factors prior to the TBIs, verbal and physical aggression and violent assaults only began after the injuries.1 Factors that are known predisposers for criminal behavior and incarceration, such as poor academic success, substance abuse, low SES, mental illness, and male gender, are also predisposers for TBI.1 These risk factors may result in aggression, or, alternatively, they may lead to TBI, which results in aggression.1 In a Canadian study, 54.3% of incarcerated females and 31.7% of incarcerated males received their first TBI before or in the same year as their first criminal offence. Women with TBI also had higher rates of substance and alcohol abuse than men with TBI, and men reported consuming alcohol earlier and criminally offending at an earlier age than non-TBI incarcerated males.46 Interestingly, women with TBI had suffered a significantly higher rate of physical and sexual abuse than non-TBI women or men, which may explain why so many more females suffered their TBI prior to offending, suggesting that their TBIs related to an abusive life, which is also a risk factor for later antisocial behavior.46 It has been suggested that childhood emotional stress and exposure to violence can have a greater effect on normal brain development in children than TBI.52 Therefore, when childhood violence and TBI co-occur, the children are exposed to both forms of neural damage. The fact that so many incarcerated females suffered from abuse prior to their TBI suggests that screening for TBI and abusive history of children in care or those assessed for psychological disorders should be performed to identify individuals at risk in order to develop appropriate treatment and interventions to reduce conflict with the criminal justice system.46 Most studies seem to indicate that criminal offending postdated the TBI, suggesting that damage caused by the TBI was a risk factor for later criminal offending. However, in contrast with this, a large study comparing pre-TBI factors with post-TBI offending, using longitudinal data from the TBIMSND, considered several thousand people who had suffered a TBI and found that pre-TBI arrests were strongly positively associated with post-TBI arrests.49 The authors suggested that the individuals may have been at higher risk for continued rearrests whether or not they had a TBI and certainly were at higher risk for arrests without the TBI than the general public.49 Moreover,

Brain disorders  227

several risk factors for criminal offending after TBI, such as being young, single, uneducated, and male, seemed to be exactly the same in both the TBI group and the general population, suggesting that these are overarching risk factors.49 Perhaps there is a greater risk for TBI within a population that already suffers from a number of other adversities, such as low socioeconomic status (SES), less social and family support, a violent environment, increased dysfunction, and risky behavior, which would predispose them to both a TBI and potential conflict with the law.50 So, is TBI a causal factor in violent and non-violent offending? So far, there is not enough data to prove cause, but there certainly appears to be a correlation. Perhaps more important to consider is that the over-representation of incarcerated youth and adults with TBI is fact, and more needs to be done to help such individuals reintegrate successfully back into society to reduce recidivism and decrease public risk.

Brain disorders The brain can also be damaged due to a number of organic brain disorders. These are referred to as acquired, as opposed to traumatic, brain injuries and usually are caused by pressure being applied to part of the brain in some manner. They can be caused by stroke, cancer, birth trauma (congenital brain injury), toxins, substance abuse, aneurysms, and diseases such as syphilis, meningitis, alcoholism, premature suture closure, and Wernicke-Korsakoff syndrome, which may result from alcoholism and brain hemorrhage. Brain damage can also be caused by treatments for other diseases, such as chemotherapy. As the adolescent brain is still developing, many factors can damage or alter its healthy development, including abuse of substances such as alcohol, cannabis, and nicotine, which are frequently used by youth.4 Animal studies have shown that exposure to alcohol causes long-term modification of neural and physiological systems.4 In rat studies, alcohol exposure via vapors caused long-term changes in sleep patterns, which were similar to those seen in aging.4 Cannabis, or marijuana, is very commonly used by adolescents and may adversely affect the endocannabinoid system, which plays a vital role in neural maturation during adolescence. The endocannabinoid system includes endocannabinoids or types of neurotransmitters that bind to cannabinoid receptors and cannabinoid receptor proteins. The  system is important in many processes, including memory, learning, and cognition, but in adolescence is particularly important in neural development and maturation.4 Cannabis use during this time can disrupt this system. Animal studies have shown that cannabis use can impact emotional reaction, anxiety, and depression and can have long-term effects on learning and memory.4 A review of human studies indicated long-term effects of both chronic and acute cannabis use, including psychotic disorders, cognitive malfunction, mood disorders, and increased substance abuse.4 Cannabis use has been shown to impact many parts of the brain, including prefrontal, medial, temporal, and striatal regions, and chronic use can double the risk for developing schizophrenia.53 Lesions (any form of abnormal tissue) may occur in all regions of the brain, and some seem to link temporally to later criminal behavior, in that the criminal behavior occurred only after the lesion. However, many lesions may result in other neuropsychological disorders but not in crime. It has been suggested that this is because there may not be a direct causal relationship between the lesion site and behavior, but instead the behavior relates to other regions of the brain with which the lesion site is interlinked.54 Studying the location and connectivity of brain lesions and their effect on behavior can allow scientists and clinicians to map the brain networks that may link to crime.54 Researchers studied literature reports of cases in which patients exhibited a range of criminogenic behavior, from non-violent crimes, such as fraud, to violent crimes, such as rape and homicide, after developing a brain lesion. Seventeen cases were identified in which criminogenic behavior only began after the lesion.54 The lesions were all located in different areas of the brain, but all were interconnected with a common functional network. This connection was found to be quite different from the neural networks seen in lesions causing other neurological syndromes that

228  Traumatic brain injury and neurocognitive disorders

did not result in criminal behavior. The authors considered this to be a network uniquely resulting in criminal behavior, so despite the lesions being located in different regions, all were connected to other regions that uniquely linked to criminal behavior in a way in which lesions in other regions of the brain did not. The lesions related to criminal behavior were all linked to regions of the brain that are activated by moral decision making, reward-based decision making, and theory of mind (which involves understanding how another person thinks, acts, and responds to assist in dealing with social encounters).54 This network map, developed on patients who were known to have exhibited criminal behavior only after the lesion, was successfully tested on a group of 23 cases in which the timing of the lesion in relation to crime was unknown, supporting the hypothesis that lesions in a disparate range of brain regions may increase risk for criminal behavior when located in a specific, unique interconnected network.54

Case examples of organic brain disorders Perhaps one of the most cited examples of a brain pathology possibly resulting in an extreme adverse personality change with resultant violence is that of Charles Whitman. Whitman was a highly intelligent, polite, quiet child, a boy scout and one of the youngest boys to become an eagle scout.55 Whitman was highly proficient with guns and as a boy, an avid hunter with his father. He joined the US Marines at 18, in 1959, and was fully medically examined at that time.55 He became an expert marksman, before returning to the United States as an engineering student at the University of Texas and marrying a fellow student. He returned to active duty for a period of time, where he was involved in a jeep accident in 1964 and again received a full medical examination, which revealed nothing unusual. He served his tour of duty, was honorably discharged, and returned to his ­studies, but he began to complain of headaches and irrational thought processes. In his suicide note, he asked for a full autopsy, as he feared he had some sort of biological anomaly driving his thoughts. He asked for his money to be used for research to prevent further tragedies such as his, suggesting he was well aware that there was something mentally wrong with him.56 At midnight on August 1, 1966, he went to his mother’s home and stabbed her to death and then returned to his own home, where he stabbed his wife to death. He then phoned both their workplaces to tell them that both his mother and wife were too sick to come in to work and proceeded to leave careful instructions about his dog and possessions,56 so he was clearly in control of his faculties and was carefully planning his crimes. Whitman then went to three separate shops and purchased guns and ammunition, drove to the University, climbed the bell tower, and, with no obvious motivation or previous violent behavior, opened fire on people, killing a further 14 and injuring 31 over a period of an hour and a half, before being killed by police.55 Investigations revealed that Whitman had visited numerous doctors prior to the shooting, exhibiting extreme hostility, and even confided in one that he dreamed of shooting people from the tower.55 At autopsy, a large tumor the size of a walnut was found in his temporal lobe. The pathologist did not consider that the tumor had any effect on Whitman’s actions, but the governor of Texas immediately established the Connally Commission to investigate further. The task force included a large number of neuropathologists, psychiatrists, neuroanatomists, neurophysiologists, neurosurgeons, and other neural specialists.55 Although the fatal gunshot wound to the head made it difficult to fully assess his brain, the tumor was identified as a glioblastoma multiforme, an extremely aggressive brain tumor.55 In their report, some members of the committee felt that the tumor could have resulted in increased aggressive behavior and an inability to control his violence, and others stated that research was increasingly showing that the temporal lobe was involved in abnormal and normal emotion-related behaviors, but overall, they felt that science at the time was not at a point to be able to explain Whitman’s actions based on his neurology.55 To this day, we do not know whether the tumor was causal or instrumental in any way in Whitman’s crimes, although it has been suggested that the tumor was pressing on the amygdala, part of the limbic system involved in emotion and aggression control.57

Neuroimaging studies of brain abnormalities  229

In another case study, a middle-aged, married schoolteacher with no history of abnormal sexual impulses suddenly changed his normal behavior.58 Originally a happily married man with a normal sexual appetite, he suddenly developed an obsession for pornography, began purchasing sexual favors, and made sexual advances toward young children, including his own stepdaughter. This  escalated into pedophilia, and he was convicted of child molestation. He  continued to display inappropriate sexual behavior and even propositioned the female instructor of a treatment program, resulting in his expulsion from the program. He later went into the hospital complaining of severe headaches and confided his fears that he was about to rape a woman. He was observed to have balance problems and showed no concern when he soiled himself. A magnetic resonance imaging (MRI) scan indicated that the man suffered from a large tumor in the frontal lobe, specifically the orbitofrontal region, or the area just above the eyes. This area is responsible for self-control, judgment, and correct performance of acceptable social behavior.58 The  tumor was successfully removed, and the man’s behavior returned to normal. He  was able to successfully complete the rehabilitation program and return home. Sometime later, his abnormal behavior returned, and a second MRI scan revealed that the tumor had returned. Again, its removal reversed the condition.58 The position of the tumor was critical, as it affected an area vital in judgment and control. However, the impact on his behavior may also have been hormonal, as tumors can result in hormonal changes, which can also impact behavior.58 The legal implications of this case will be discussed in the final chapter. In an adolescent case, a teenage boy developed severe personality changes, becoming violent, withdrawn, and very tense, and experienced temporal lobe seizures. Doctors believed that his antisocial behavior related to a disturbed home life and did not  investigate further. However, at age 19,  he slipped into a coma and later died. At  autopsy, a tumor was found in his right hippocampus.59

Neuroimaging studies of brain abnormalities Neuroimaging studies have opened up the field of brain trauma and dysfunction greatly, ­a llowing us to see the structure and functioning of the brain in life. In many cases, a person can be imaged while performing a task or looking at images, to see which part of the brain is ­responding. Early techniques showed brain structure alone, but more recent techniques combine structure with f­unction, allowing researchers to see which parts of the brain are active under a variety of ­conditions. There are several different imaging techniques available, and there are advantages and disadvantages to each.

Neuroimaging techniques Computer tomography

Computer tomography (CT), also sometimes referred to as computed axial tomography (CAT), is a medical diagnostic test that combines X-rays with computer technology. Extremely sophisticated X-ray equipment is used to obtain image information of the brain from every angle. The information is processed by computer to produce a three-dimensional (3D) image that can be examined at every plane, as if a cross-section were to be cut through the brain at any desired site. The technique can also be used on any other tissue, such as heart, lungs, and blood vessels. It is frequently used to diagnose cancers and spinal injuries. The brain is X-rayed first in one plane, making numerous scans at different positions and then it is moved 1°, and the scans are repeated until the brain has been scanned at all 360°. The images are analyzed by the computer to create a 3D image of the brain or area scanned. Of all the imaging techniques, CT is probably the cheapest to use; however, an obvious drawback is that the subject is exposed to X-rays, which can be damaging. Moreover, smaller brain structures may be difficult to individualize.

230  Traumatic brain injury and neurocognitive disorders

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is an advanced medical diagnostic technique that uses a very powerful magnet, radiofrequency waves, and computer imaging techniques to provide an image of the body’s internal organs or structures. It can be used to image the brain, as well as any other structure. In the news, we frequently hear of sports figures awaiting results of the MRI of a recent injury to determine when and if they can return to their profession. The development of the MRI was a tremendous step forward in the medical world. It allows a physician to see inside the body, from any angle, at any site, without invasive procedures, making the term exploratory surgery frequently a thing of the past. The  MRI machine creates a magnetic field, then sends radio waves into the body and subsequently measures the cellular response. Different tissues, as well as diseased and normal tissue, produce different resonance and thus can be distinguished at a very fine level. In particular, MRI gives clear images of soft tissue near and around bones, so not only is it the technique of choice for examining spinal and joint injuries, but it also means that there is no bone artifact, a further advantage of MRI over CT, particularly when examining joints or the brain. The technique is also safer because radio waves, rather than X-rays, are used. MRI does have some limitations. As a powerful magnet is used, people with any metallic objects within their body cannot be imaged. These include people with any implants, as well as pacemakers and heart clips. Also, the equipment required is extremely expensive and requires specialized technical support. Both CT and MRI, although extremely valuable techniques, only provide structural information about the brain. Other techniques can measure metabolic activity in the brain and thus can provide information about brain function. Because such tests are performed on live, conscious patients, these functional tests allow doctors and researchers to study brain function while the subject performs tasks, such as viewing images. Positron emission tomography

Positron emission tomography (PET) was the first brain scanning technique developed that allowed doctors and researchers to study brain function rather than structure. PET measures the emission of positrons (anti-electrons) from the brain after a small amount of radioactive isotope is injected into the subject’s bloodstream. The isotope used is a short-lived radioactive tracer that decays by emitting a positron. It  is combined with a metabolically active molecule, such as a sugar. Once injected, there is a short waiting period while the sugar becomes concentrated in the tissues being studied. The subject is placed in an imaging scanner, and the presence of the isotope in the tissues is detected as it decays, emitting positrons. This, therefore, shows the site of brain metabolism or function. PET has been used extensively by doctors in efforts to understand which parts of the brain are involved in many different neurologic illnesses, including seizures, schizophrenia, and Parkinson’s disease. PET has also been used to answer other interesting questions, such as which specific part or parts of the brain are involved in certain activities, such as solving mathematic problems, word puzzles, or cognitive tests. PET is an excellent and direct measure of brain activity and can be used to determine brain function during various tasks or during cognitive tests. The  resolution is excellent and allows a physician or researcher to focus on a very small region of the brain. If a dysfunction in a particular region is suspected, that area can be examined minutely while the subject is challenged with a task that is thought to be controlled in that area. The glucose metabolism can then be measured and compared with other areas. There  are, of course, several rather obvious drawbacks. The  first is that the test is somewhat invasive, requiring the injection of a substance, followed by a series of blood tests. Moreover, the substance itself is radioactive, involving some risk to the subject, and repeated tests on the same subject are not recommended. Perhaps the biggest drawback is the high cost of the cyclotron that is required to produce the isotope.

Neuroimaging studies of brain abnormalities  231

Single-photon emission computed tomography

Single-photon emission computed tomography (SPECT) is used to study the regional cerebral blood flow (rCBF) as a measure of the metabolic rate of glucose in certain areas of the brain. Glucose is the only energy source utilized by the brain, so measuring blood flow in different regions of the brain can be used to indicate levels of brain function. SPECT can be used to produce a 3D image of the rCBF, using gamma rays. The subject is injected with a gamma-emitting radiopharmaceutical. Then a series of images are recorded by a gamma camera. The gamma camera revolves around the patient, recording images over a 360° rotation. The images are processed by a computer to produce the final 3D image. The image can then be sliced in any plane. This technique is not as precise as PET, but it does not require a cyclotron, so it is easier and cheaper. There is also less exposure to radiation, although some exposure is still required. Functional magnetic resonance imaging

Functional magnetic resonance imaging (fMRI) couples MRI with measurement of blood flow, as increased blood flow in a region of the brain indicates increased activity, showing that the person is using that part of the brain. It  can be used to measure brain activity during certain tasks to determine which parts of the brain are active, for example, when recalling a memory, learning new information, or subject to disturbing images or verbal aggression; hence, it is extremely useful in criminological studies. It does not require that the individual be injected with any substance or required to ingest material or be irradiated. To provide an image in court, the operator assigns different colors to different areas of activity, but this can be misleading, as the colors chosen may suggest very dramatic differences in activity when they are actually showing very slight differences. Moreover, the produced image looks a bit like a photograph, whereas it is really a composite of activities. This can be misleading in court, as we will discuss in the final chapter.

Brain-imaging studies of offenders A large number of studies using a range or combination of imaging techniques have been used to see whether there are differences between violent or non-violent offenders and healthy controls and which brain regions are affected. They are also used to assess level of TBI and efficacy of treatment measures. In many studies, brain differences have been seen between violent offenders, non-violent offenders, and normal healthy controls. A caveat in such studies that must be remembered from the beginning of this book is that normal healthy controls may also include both violent and nonviolent offenders who have not been caught. Psychopathy is a construct of many personality disorders, including extreme callousness, lack of concern for the effects of actions on others, lack of empathy, lack of fear, lack of ability to form emotional bonds, disinhibited impulsivity, and antisocial behavior.60 Psychopaths also exhibit a lack of remorse, shallowness, and manipulation and demonstrate superficial charm.61 Although the prevalence of psychopathy in the general population is low, psychopaths commit a disproportionate number of violent crimes. Their level of gratuitous violence is also extremely high.62 Moreover, they have very high levels of recidivism, with psychopaths consistently being three times more likely to recidivate within 1 year of prison release than non-psychopaths and four times more likely to specifically commit violent crimes repeatedly.63 They adjust poorly to prison settings and ­treatment programs.63 Psychopathy is considered to be one of the most robust predictors of crime and recidivism. Many studies have considered the differences in the brains of psychopaths in comparison with non-psychopaths. Several MRI studies have shown significantly lower volumes of gray matter in the prefrontal cortex in patients with high levels of psychopathy in comparison with normal controls,64 and in another study violent offenders with both antisocial personality disorder (ASPD) and psychopathy had significantly reduced gray matter volumes in the prefrontal cortex than offenders with

232  Traumatic brain injury and neurocognitive disorders

just ASPD or normal controls.65 Volume of gray matter was similar in offenders with just ASPD and normal controls, suggesting that there are neural differences between psychopaths and individuals with ASPD alone.65 A Canadian study using fMRI to compare neural activity of psychopathic criminals with non-psychopathic criminals and non-criminal controls, using a memory task, showed that the psychopaths had reduced activity in the limbic system.61 This  is the area, remember, that is primarily involved in emotion and is a vital part of communication between the prefrontal lobe and the hypothalamus, which are heavily involved in behavioral control and inhibition. So, changes in this area will result in reduced behavioral control. The psychopaths in this study also exhibited overactivity in the frontal and temporal regions of the brain, which are involved in cognitive decision-making and impulse control. The authors of this study noted that the results were consistent with the belief that psychopaths process information differently from non-psychopaths, whether criminal or otherwise. No obvious structural deficits were noted.61 More recent studies have shown that persistent violent offending and psychopathy are linked to reduced gray matter volume in the limbic and paralimbic areas, particularly in the amygdala, the hippocampus, and the orbitofrontal cortex.64 Neuroimaging of high-risk male violent offenders and matched normal controls found that psychopathic traits and risk of violent crime recidivism was linked with volume of gray matter in regions of the brain linked to criminal behavior.64 Although reduced gray matter volume was found in the prefrontal cortex, an increase in gray matter volume was found in other areas of the brain, such as the basal ganglia, the cerebellar cortex, and the supplementary motor area (SMA) in psychopaths, all of which are regions of the brain associated with high levels of motor function and cognition.64 Moreover, within offenders, increased gray matter volume in the basal ganglia, cerebellar cortex, and SMA correlated significantly with a number of psychopathic traits, including antisocial behavior, anger management, and recidivism.64 Cerebellar gray matter volume was also correlated with interpersonal problems, anger, and other risk factors, and the authors considered that it may relate to higher levels of impulsivity and lower levels of behavioral inhibition. Moreover, the cerebellum is involved in moral behavior and aggression.64 Reduced gray matter volume in the prefrontal cortex was seen in this and many other studies in antisocial and psychopathic individuals, using a range of subjects and imaging methods. This makes sense, as we know the prefrontal cortex region of the frontal lobe is heavily involved in high levels of executive functioning, such as emotional responses, decision-making, and antisocial behavioral control, so deficits in this region are likely to result in dysfunction in these areas.64 The  study also found reduced gray matter volume in the amygdala, which correlated with criminal recidivism, as had previous studies. In another study, fMRI showed that psychopaths had reduced fear conditioning and reduced amygdala activity in comparison with controls.66 Several studies using fMRI have shown reduced amygdala activity in psychopaths in comparison with normal controls when shown images of amoral behavior and fear, and during aversive stimuli.63 Similarly, youth with conduct disorder (CD) or callous or unemotional traits showed reduced amygdala activity.63 The amygdala is important in fear and emotion conditioning and relearning, in which the brain learns from past mistakes. Fear conditioning occurs when one learns from an aversive experience; for example, touching a pot on the stove hurts, so the brain learns to avoid repeating the action. Psychopaths are known to have very little fear and do not learn from or consider the consequences of their actions. Therefore, a reduction in gray matter volume in the amygdala would explain a lack of fear awareness and ability to learn from past experiences and an inability to relate emotionally to important events.64 It has been suggested that reduced amygdala gray matter volume might be a risk factor for violence, although whether it is causal or treatable is unknown.64 Neuroimaging studies could be useful in assessing treatment options.64 In  a follow-up study, violent offenders were compared with non-violent controls using fMRI, and it was found that violent offenders had greater cerebellar–amygdala connectivity and that the controls had greater functional connectivity between the cerebellum and orbitofrontal cortex.67 The  authors considered that this resulted from the normal relationship between the cerebellum and other brain regions in regulating moral behavior and aggression, and as such, the heightened

Treatment options  233

connectivity in the cerebellar–amygdala region in violent offenders might relate to dysfunction in handling moral emotions, and the increased connectivity between the frontal cortex and cerebellum in controls might be a result of better emotional control.67 They suggest that violent offending relates to impairment in the brain regions that involve the cerebellum.67 Studies have suggested for decades that the cerebellum is important in emotion and cognition, and more recent studies have indicated that it is important in moral judgment, empathy, and deception.67 It also interconnects with other regions of the brain involved in aggression, violence, moral judgment, and emotional regulation, such as the limbic system and the prefrontal cortex.67 Psychopaths are known to be impulsive and aggressive and to be unable to consider long-term goals or consequences of their actions. In a recent fMRI study of almost 50 incarcerated offenders, the results indicated that much of the psychopath’s impulsivity and lack of long-term planning related to a need for instant gratification, and this in turn correlated with increased activity in the ventral striatum, which is a major part of the reward system in the brain.68 The participants were offered choices of a potential small amount of money immediately or a larger sum later; in other words, this measured their need for instant remuneration versus considering a long-term s­ trategy. The higher the offender scored for psychopathy, the greater the activity in the striatum, indicating a dysfunction in the reward system in which it overvalued immediate gratification. Scans not only showed much higher activity in the ventral striatum, but the authors were also able to map the connections between the ventral striatum and other regions of the brain that are involved in decision-making, in particular the prefrontal cortex, which regulates the ventral ­striatum.68 Network mapping showed much weaker connections between the ventral striatum and the prefrontal cortex in psychopaths than normal. The striatum considers reward, but it is the prefrontal cortex that considers long-term consequences, so loss of connectivity in this area can result in poor decision-making. The connection was strong enough for the researchers to be able to predict the number of convictions for each participant.68 Interestingly, neuropathological differences have been found between psychopaths who have been caught for their crimes and those who have not. Two groups of psychopaths were compared with each other and with normal controls. Both groups of psychopaths had high psychopathy checklist–revised (PCL-R)69 scores, and both groups reported identical criminal histories, but psychopaths in one group had been caught and incarcerated for their crimes and so were considered to be “unsuccessful,” and the psychopaths in the other group had evaded detection and so were considered to be “successful” (although they were only successful in avoiding capture, not under other measures).70(p547) Unsuccessful psychopaths showed structural abnormalities and reduced thickness and gray matter volume in the prefrontal cortex and reduced volume of the amygdala compared with both successful psychopaths and normal controls.70 The  authors suggested that unsuccessful psychopaths were a subgroup and that the structural abnormalities may relate to not  just psychopathy alone but also to the interaction between psychopathy and risky behavior, which was instrumental in their being caught.70 This is extremely intriguing, as almost all studies on psychopaths are conducted on incarcerated offenders, in other words, unsuccessful psychopaths. As we have discussed earlier, in all such studies comparing offenders with non-offenders, the division may not be so clear, as many so-called non-offender groups may include offenders who have simply not been caught. This study indicates that there are clear neural differences between offenders who have been caught and those who have managed to avoid detection, despite similar PCL-R scores, which raises many questions and warrants further study.

Treatment options Different brain traumas result in different cognitive and behavioral changes, and it is important to realize that treatment for one type of brain injury may not be effective for another type of injury. Moreover, it is clear that although many recognized disorders, such as CD and impulsive explosive aggression, may not be the result of diagnosed brain trauma, the resulting deficits are

234  Traumatic brain injury and neurocognitive disorders

very similar to certain types of brain injury. This may indicate that they are the result of a brain pathology unlinked to direct trauma but perhaps linked to birth complications, genetic deficits, ­environmental trauma, neurotoxins, or myriad other causes. Despite the etiology, what is perhaps more important is that the deficit is identified, as this would mean that treatment options that are effective in people with brain trauma may be valid in people with other mental disorders that may mimic the symptoms of brain injury. On the other side of the coin, it is equally important that offenders who are not necessarily considered to have brain trauma be assessed medically and neurologically, as mild brain injury may never have been reported, and many people do not consider that they have ever received such an injury. Nevertheless, a number of the above studies have indicated that many offenders exhibit the results of brain trauma, whether from direct injury or other causes. In  such cases, not  only should such ­information be made available to the courts, but it should also most certainly be taken into account during sentencing and subsequent treatment. Cognitive and insight-oriented therapy, often used in many treatment programs, are unlikely to be effective with people who have sustained frontal lobe damage. Incarcerated individuals with TBIs are more likely to receive lengthier sentences than those without TBIs, which may relate to the earlier discussion on youth with TBIs in that affected individuals are less likely to be able to assist in their own defense.51 Also, individuals with TBIs may commit more violent offences than those without TBIs and often have higher levels of recidivism, resulting in the courts imposing higher sentences. However, due to the TBI, many such offenders do not have the capacity to understand the consequences of their crimes, so incarceration is not a deterrent, which is clear from the high level of recidivism.51 It has also been shown that once incarcerated, offenders with TBI spend longer in prison than non-TBI offenders as a result of in-prison infractions due to their TBIs, with males being 86% and females being 144% more likely to break prison rules than their non-TBI peers.51 It is therefore important that offenders with TBIs be identified and appropriate treatment plans developed. It is important to assess the risk factors for offenders in order to understand the risk of recidivism and particularly, violent recidivism. When assessing offenders, certain risk factors are considered static—that is, they cannot be changed—such as poor school adjustment and age of first offense, but others are considered dynamic—that is, they can be potentially treated and changed—such as impulsivity.64(p194) This applies to brain damage. In the United States, in 1996, a program was developed to provide grants to criminal justice systems to help people with TBI; however, by 2016, only one state, Minnesota, had taken advantage of the funding.51 However, Minnesota has developed a successful program to identify offenders with TBIs, train staff, and successfully rehabilitate offenders and return them to society, which provides an excellent template for other programs.51 From this program, a number of steps were recommended to assist offenders reintegrate successfully.51 The first step is assessment of offenders to determine whether they have suffered a TBI. This can be done by self-assessment and also by staff trained to watch for behaviors indicative of TBI, followed by interviews with the offenders. The second step is psychoeducation and treatment, as many TBI sufferers are not aware of the effects of their TBI, and education can assist them to recognize their problems and limitations and assist them to move on toward helping themselves. This should be followed by cognitive behavioral therapy, which has been very effective in individuals with mild and moderate TBIs but less effective in severe cases. Anger management can also be helpful. The third step is staff training in order to educate people working with this population about symptoms of TBI and methods that work best for them, including developing very strict and consistent structure in their routines. Individuals with TBIs may overreact to stimuli such as noise and the presence of peers and may not heed instructions in such situations, so staff need to be aware that such individuals may need to move to a quiet area in order to understand or may need to be given a break. Staff also need training in communication with this population, as the offenders may not understand instructions the first time and staff may need to repeat themselves or may need to ask the individual to repeat the instructions back to ensure that they are understood, rather than assume that the individual is deliberately being difficult.51

Treatment options  235

In  Britain, an organization called the Linkworker Service was piloted in an attempt to assist offenders in the criminal justice system who had suffered a TBI.48 The  Linkworkers themselves were psychology graduates using the opportunity to gain experience before pursuing a profession in clinical psychology or a similar field. The Linkworkers were assessed and appointed by the Disabilities Trust Fund, whose mandate includes brain injury rehabilitation. The Linkworkers were specially trained in causes and range of symptoms for people with TBIs, as well as expectations for behavior and functionality, and were provided with techniques and strategies to aid individuals with TBI. The Linkworkers were assigned specific prisoners and would aid them in understanding their injury, help them develop plans to manage and cope with their disabilities, develop their future plans, and link with other staff to explain the affected individual’s behavior and discuss ways to manage it and provide interventions for support. They also linked with external ­personnel and helped to develop a plan and network of support once the offender was released, including working with family and friends in the home setting, and guided support people in referring the offender to medical specialists, housing advocates, social assistance, and so on.48 They then continued to provide support to the offenders once released, helping them apply strategies developed in prison to real-world situations and liaising with probation officers, explaining behavioral issues and interventions. In a discussion of several case histories of offenders with very severe behavioral challenges due to severe TBIs, a mixture of general and very specific and personalized assistance allowed the offenders to improve their conditions and, in some cases, succeed well outside prison. The system is low cost, owing to the desire to gain valuable experience by young graduates, and is argued to be sustainable and successful.48 It involves very intensive assistance and support but has been developed in a way to be defensible. Similar projects on larger scales will be needed to show overall improvement, but it does indicate that offenders with TBI can succeed outside the prison environment if given enough support. In  a study on a non-offender group (military service members with mild TBI), a number of different interventions were helpful. Four regimes were randomly tested. The  first was psychoeducation with concussion and medical management of symptoms, considered to be the “normal standard of care” to which the others were compared.71(p195) The  other three were computerized cognitive rehabilitation (CR), traditional CR led by a therapist, and traditional therapist-led CR combined with cognitive behavioral psychotherapy.71 Psychoeducation involved participants being provided with materials related to TBIs and symptom management, together with medical care. Computerized CR included psychoeducation as well as a total of 10 supervised in-clinic treatments, which included computer programs on specific skill training, such as increasing attention, as well as general cognitive improvement programs to improve brain health. Traditional CR included basic psychoeducation and 10 hours of group and individual therapy with speech and occupational therapists. The final regime included both psychoeducation and therapist-led CR as well as self-directed mindfulness-based stress reduction therapy and therapist-led psychotherapy.72 Results indicated that a team approach was most helpful and that patients with comorbid disorders, such as depression, showed the greatest positive responses. Importantly, however, the researchers discovered that self-use computerized CR without therapist intervention was actually harmful.71 In a study of more direct treatment, 86 normal adults were studied in a double-blind, ­placebo, stratified, randomized trial in which half of the group received electrical stimulation of the prefrontal cortex and the others received a placebo stimulation (one of much shorter duration).73 Participants then answered a questionnaire the following day involving a number of different vignettes designed to test intent to commit a violent act and personal perception of moral wrongness and performed a computer exercise in which they were asked to consider a computer-­ generated voodoo doll as a close friend or spouse and then to insert as many computer pins into it as they liked. Individuals who received the stimulation reported that they were much less likely to commit a physical or sexual assault and also considered that such acts were much more morally wrong than the placebo-treated individuals.73 This work supported much previous work showing that the prefrontal cortex is important in judging right from wrong and in aggression and shows the role of moral decision-making in aggression.73 Interestingly, the treatment group did show

236  Traumatic brain injury and neurocognitive disorders

a non-significant increase in aggression, which has sometimes been seen in other patients receiving similar types of stimulation. This was a very preliminary experiment, but the authors argue that it warrants further study, as it is non-invasive and has very minor side effects.73 Treatment options are available for a variety of behavioral and cognitive outcomes in individuals with TBI or other brain disorders, but in order for them to be of any value, affected i­ ndividuals must be identified before they can be assisted. Studies have shown that intense interventions can assist individuals in rehabilitation and in reducing antisocial behavior and recidivisms, but very few ­correctional facilities even screen for TBI, let alone have specific programming available. The few pilot studies discussed above indicate that interventions can help in both prognosis and recidivism, but in order for them to be of any value, correctional institutions and governments need to start instigating them. Considering the financial burden of violent crime and recidivism, it should be fairly obvious that intervention is overdue.

Conclusion This chapter examined brain damage and its influence on criminality. We found that because the brain is the seat of all behavior, any damage to the brain is likely to affect behavior. Thus, accidental trauma, birth trauma, abuse as a child, FASD, or biochemical imbalances can affect action and can, often in concert with environmental or other influences, result in criminal behavior. We found that brain damage can be severe at any age, but damage to the developing brain is most severe. Modern medical techniques, including MRI, PET, and CT scans, have made it possible for us to understand much more about the brain and brain damage than in the past and have allowed us to see the links to criminal behavior. In addition to mechanical damage, chemicals and other toxins can have ­permanent and traumatic effects on the brain. Large-scale brain trauma, as shown in cases such as that of Phineas Gage, can result in major personality changes and disorders, including increased violence, spousal abuse, and hypersexuality. Damage to different areas of the brain, such as the temporal lobe, corpus callosum, amygdala, and hippocampus, cause different types of behavioral changes. In many cases, even minor injuries to brain function can have a major influence on future behavior. TBIs can result in deficits in cognitive skills, ability to think and reason, social skills, self-esteem, and problem-solving and in academic and work-related problems, all of which in their own ways predispose to crime and violence. Of course, head injuries do not result in violence and crime in all cases; in fact, in most cases, such damage does not result in violence. It may be that head injuries precipitate violence in those people who are already predisposed to violence through other ­factors, whether they are biological, environmental, or social. Head injuries usually reduce coping skills, judgment, and restraint, and thus, they might result in violence in normally non-violent (pre-injury) individuals who are already deficient in those skills.

Questions for further study and discussion 1. We consider children and youth under 18 to be minors and therefore less culpable in criminal offending. However, the adolescent brain is still developing until age 25. Should we therefore consider anyone under the age of 25 to be a minor? Is this scientifically valid? Discuss the legal and ethical pros and cons of such an approach. 2. Evaluate the evidence available to us today to determine whether TBI is causal for criminal behavior or whether TBIs just occur in individuals who indulge in risky behavior. 3. Explain why youthful TBIs are more likely to result in antisocial behavior than in adults, from both a biological and an environmental point of view. 4. Why would you expect that there would be gender differences in the prevalence of TBI and expression of behaviors?

References  237

5. We consider that committing a crime is a matter of free will. But can we use the free will argument when brain damage has caused major behavioral changes (increased aggression and hypersexuality) that have led to criminal actions? Is the person still culpable? Are they able to control their impulsive or aggressive behavior, or do they just have lower levels of control? Discuss.

References 1. Lane, K.S., St Pierre, M.E., Lauterbach, M.D., and Koliatsos, V.E. 2017. Patient profiles of ­criminal behavior in the context of traumatic brain injury. J. Forensic Sci. 62(2): 545–548. 2. Hughes, N., Williams, W.H., Chitsabesan, P., Walesby, R.C., Mounce, L.T., and Clasby, B. 2015. The prevalence of traumatic brain injury among young offenders in custody: A systematic review. J. Head Trauma Rehabil. 30(2): 94–105. 3. Center for Disease Control and Prevention. 2017. Traumatic Brain Injury. Accessed 25 July, 2018; http:/www.cdc.gov/TraumaticBrainInjury/. 4. Arain, M., Haque, M., Johal, L. et al. 2013. Maturation of the adolescent brain. Neuropsychiatr. Dis. Treat. 9: 449–461. 5. Bannon, S.M., Salis, K.L., and Daniel O’Leary, K. 2015. Structural brain abnormalities in aggression and violent behavior. Aggress. Viol. Beh. 25: 323–331. 6. Lewis, D.O. 1998. Guilty by Reason of Insanity: A  Psychiatrist Explores the Minds of Killers. New York: Fawcett Columbine. 7. Damasio, H., Grabowski, T., Frank, R., Galaburda, A., and Damasio, A. 1994. The  return of Phineas Gage: Clues about the brain from the skull of a famous patient. Science 264(5162): 1102–1105. 8. Bigelow, H.J. 1850. Dr. Harlow’s case of recovery from the passage of an iron bar through the head. Am. J. Med. Sci. 39: 13–22. 9. Sebit, M.B., Acuda, W., and Chibanda, D. 1996. A case of the frontal lobe syndrome following head injury in Harare, Zimbabwe. Central Afr. J. Med. 42(4): 51–53. 10. León-Carrión, J. and Chacartegui-Ramos, F. 2010. Brain injuries and violent crime, In: Violent Crime: Clinical and Social Implications, Ferguson, C.J., editor. Thousand Oaks, CA: Sage Publications. pp. 99–118. 11. Johnson, S.B., Blum, R.W., and Giedd, J.N. 2009. Adolescent maturity and the brain: The promise and pitfalls of neuroscience research in adolescent health policy. J. Adolesc. Health 45(3): 216–221. 12. Reshetnikov, V.V., Studenikina, A.A., Ryabushkina, J.A., Merkulova, T.I., and Bondar, N.P. 2018. The  impact of early-life stress on the expression of HPA-associated genes in the adult murine brain. Behaviour 155(2–3): 181–203. 13. Anderson, S.W., Bechara, A., Damasio, H., Tranel, D., and Damasio, A.R. 1999. Impairment of social and moral behavior related to early damage in human prefrontal cortex. Nature Neurosci. 2(11): 1032–1037. 14. Kennedy, E., Heron, J., and Munafo, M. 2017. Substance use, criminal behaviour and psychiatric symptoms following childhood traumatic brain injury: Findings from the ­ ALSPAC cohort. Eur. Child Adolesc. Psychiatry 26(10): 1197–1206. 15. Mychasiuk, R., Hehar, H., Farran, A., and Esser, M.J. 2014. Mean girls: Sex differences in the effects of mild traumatic brain injury on the social dynamics of juvenile rat play behaviour. Behav. Brain Res. 259: 284–291. 16. Yeates, K.O., Gerhardt, C.A., Bigler, E.D. et  al. 2013. Peer relationships of children with ­traumatic brain injury. J. Int. Neuropsychol. Soc. 19(5): 518–527. 17. Anderson, V., Beauchamp, M.H., Yeates, K.O., Crossley, L., Hearps, S.J., and Catroppa, C. 2013. Social competence at 6  months following childhood traumatic brain injury. J. Int. Neuropsychol. Soc. 19(5): 539–550.

238  Traumatic brain injury and neurocognitive disorders

18. Moore, E., Indig, D., and Haysom, L. 2014. Traumatic brain injury, mental health, substance use, and offending among incarcerated young people. J. Head Trauma Rehabil. 29(3): 239–247. 19. Fazel, S., Lichtenstein, P., Grann, M., and Langstrom, N. 2011. Risk of violent crime in individuals with epilepsy and traumatic brain injury: A 35-year Swedish population study. PLoS Med 8(12): e1001150. 20. Timonen, M., Miettunen, J., Hakko, H. et al. 2002. The association of preceding traumatic brain injury with mental disorders, alcoholism and criminality: The Northern Finland 1966 Birth Cohort Study. Psychiatr. Res. 113: 217–226. 21. Chitsabesan, P., Lennox, C., Williams, H., Tariq, O., and Shaw, J. 2015. Traumatic brain injury in juvenile offenders: Findings from the comprehensive health assessment tool study and the development of a specialist linkworker service. J. Head Trauma Rehabil. 30(2): 106–115. 22. Davies, R.C., Williams, W.H., Hinder, D., Burgess, C.N., and Mounce, L.T. 2012. Self-reported traumatic brain injury and postconcussion symptoms in incarcerated youth. J. Head Trauma Rehabil. 27(3): E21–E27. 23. Lewis, D.O., Pincus, J.H., Bard, B. et al. 1988. Neuropsychiatric, psychoeducational, and f­ amily characteristics of 14 juveniles condemned to death in the United States. Am. J. Psychiatry 145(5): 584–589. 24. Death Penalty Information Center. 2018. Facts about the death penalty. Accessed 30 November, 2018; https://deathpenaltyinfo.org. 25. Wszalek, J.A. and Turkstra, L.S. 2015. Language impairments in youths with traumatic brain injury: Implications for participation in criminal proceedings. J. Head Trauma Rehabil. 30(2): 86–93. 26. Williams, W.H., Cordan, G., Mewse, A.J., Tonks, J., and Burgess, C.N. 2010. Self-reported traumatic brain injury in male young offenders: A risk factor for re-offending, poor mental health and violence? Neuropsychol. Rehabil. 20(6): 801–812. 27. Gordon, W.A., Spielman, L.A., Hahn-Ketter, A.E., and Sy, K.T.L. 2017. The relationship between traumatic brain injury and criminality in juvenile offenders. J. Head Trauma Rehabil. 32(6): 393–403. 28. Perron, B.E. and Howard, M.O. 2008. Prevalence and correlates of traumatic brain injury among delinquent youths. Crim. Behav. Ment. Health 18(4): 243–255. 29. Kaba, F., Diamond, P., Haque, A., MacDonald, R., and Venters, H. 2014. Traumatic brain injury among newly admitted adolescents in the New York city jail system. J. Adolesc. Health 54(5): 615–617. 30. Durand, E., Watier, L., Lecu, A. et al. 2017. Traumatic brain injury among female offenders in a prison population: Results of the Fleury TBI study. Brain. Behav. 7(1): e00535. 31. Scott, C., McKinlay, A., McLellan, T., Britt, E., Grace, R., and MacFarlane, M. 2015. A comparison of adult outcomes for males compared to females following pediatric traumatic brain injury. Neuropsychol. 29(4): 501–508. 32. Vaughn, M.G., Salas-Wright, C.P., DeLisi, M., and Perron, B. 2014. Correlates of traumatic brain injury among juvenile offenders: A multi-site study. Crim. Behav. Ment. Health 24(3): 188–203. 33. Farrer, T.J., Frost, R.B., and Hedges, D.W. 2013. Prevalence of traumatic brain injury in juvenile offenders: A meta-analysis. Child Neuropsychol. 19(3): 225–234. 34. Cole, W.R., Gerring, J.P., Gray, R.M. et  al. 2008. Prevalence of aggressive behaviour after severe paediatric traumatic brain injury. Brain Injury. 22(12): 932–939. 35. Rees, E., O’Donovan, M.C., and Owen, M.J. 2015. Genetics of schizophrenia. Curr. Opinion Behav. Sci. 2: 8–14. 36. St. Clair, D., Xu, M., Wang, P. et  al. 2005. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–1961. JAMA 294: 557–562.

References  239

37. AbdelMalik, P., Husted, J., Chow, E.W.C., and Bassett, A.S. 2003. Childhood head injuries and expression of schizophrenia in multiply affected families. Arch. Gen. Psychiatry 60(3): 231–236. 38. Soyka, M. 2011. Neurobiology of aggression and violence in schizophrenia. Schizophr. Bull. 37(5): 913–920. 39. McDonald, S., English, T., Randall, R., Longman, T., Togher, L., and Tate, R.L. 2013. Assessing  social cognition and pragmatic language in adolescents with traumatic brain injuries. J. Int. Neuropsychol. Soc. 19(5): 528–538. 40. Rosenbaum, A. and Hoge, S.K. 1989. Head injury and marital aggression. Am. J. Psych. 146: 1048–1051. 41. Rosenbaum, A., Hoge, S.K., Adelman, S.A., Warnken, W.J., Fletcher, K.E., and Kane, R.L. 1994. Head injury in partner-abusive men. J. Consulting Clin. Psychol. 62(6): 1187–1193. 42. Farrer, T.J., Frost, R.B., and Hedges, D.W. 2012. Prevalence of traumatic brain injury in i­ ntimate partner violence offenders compared to the general population: A meta-analysis. Trauma Violence Abuse 13(2): 77–82. 43. Diamond, P.M., Harske, A.J., Magaletta, P.R., Cummins, A.G., and Frankowski, R. 2007. Screening for traumatic brain injury in an offender sample: A first look at the reliability and validity of the traumatic brain injury questionnaire. J. Head Trauma Rehab. 22(6): 330–338. 44. Shiroma, E.J., Ferguson, P.L., and Pickelsimer, E.E. 2010. Prevalence of traumatic brain injury in an offender population: A meta-analysis. J. Correct. Health Care 16(2): 147–159. 45. Durand, E., Chevignard, M., Ruet, A., Dereix, A., Jourdan, C., and Pradat-Diehl, P. 2017. History of traumatic brain injury in prison populations: A systematic review. Ann. Phys. Rehabil. Med. 60(2): 95–101. 46. Colantonio, A., Kim, H., Allen, S., Asbridge, M., Petgrave, J., and Brochu, S. 2014. Traumatic brain injury and early life experiences among men and women in a prison population. J. Correct. Health Care 20(4): 271–279. 47. Ray, B. and Richardson, N.J. 2017. Traumatic brain injury and recidivism among returning inmates. Criminal Just. Behav. 44(3): 472–486. 48. Ramos, S.D.S., Oddy, M., Liddement, J., and Fortescue, D. 2018. Brain injury and offending: The development and field testing of a Linkworker Intervention. Int. J. Offender Ther. Comp. Criminol. 62(7): 1854–1868. 49. Elbogen, E.B., Wolfe, J.R., Cueva, M., Sullivan, C., and Johnson, J. 2015. Longitudinal p ­ redictors of criminal arrest after traumatic brain injury: Results from the Traumatic Brain Injury Model System National Database. J. Head Trauma Rehabil. 30(5): E3–E13. 50. Fishbein, D., Dariotis, J.K., Ferguson, P.L., and Pickelsimer, E.E. 2016. Relationships between traumatic brain injury and illicit drug use and their association with aggression in inmates. Int. J. Offender Ther. Comp. Criminol. 60(5): 575–597. 51. Horn, M.L. and Lutz, D.J. 2016. Traumatic brain injury in the criminal justice system: Identification and response to neurological trauma. Appl. Psych. Crim. Justice 12(2): 71–85. 52. Brewer-Smyth, K., Cornelius, M.E., and Pickelsimer, E.E. 2015. Childhood adversity, mental health, and violent crime. J. Forensic Nurs. 11(1): 4–14. 53. Sami, M.B., Rabiner, E.A., and Bhattacharyya, S. 2015. Does cannabis affect dopaminergic signaling in the human brain? A  systematic review of evidence to date. Eur. Neuropsychopharmacol. 25(8): 1201–1224. 54. Darby, R.R., Horn, A., Cushman, F., and Fox, M.D. 2018. Lesion network localization of ­criminal behavior. Proc. Natl. Acad. Sci. USA 115(3): 601–606. 55. Connally Commission. 1966. Report to the Governor. Medical Aspects, Charles J. Whitman Catastrophe, Austin, TX: Austin History Center, 21 p. 56. Whitman, C. 1966. The Whitman Letter. The Whitman Archives. Austin American-Statesman, Austin,TX. 57. Freberg, L.A. 2009. Discovering Biological Psychology. Belmont, CA: Cengage Learning, 608 p.

240  Traumatic brain injury and neurocognitive disorders

58. Burns, J.M. and Swerdlow, R.H. 2003. Right orbitofrontal tumor with pedophilia symptom and constructional apraxia sign. Arch. Neurol. 60(3): 437–440. 59. Tardiff, K. 1998. Unusual diagnoses among violent patients. Psych. Clinics N. Am. 21(3): 567–576. 60. Hare, R.D. and Neumann, C.S. 2008. Psychopathy as a clinical and empirical construct. Annu. Rev. Clin. Psychol. 4: 217–246. 61. Kiehl, K.A., Smith, A.M., and Hare, R.D. 2001. Limbic abnormalities in affective ­processing by criminal psychopaths as revealed by functional magnetic resonance imaging. Biol. Psychiatry 50(9): 677–684. 62. Anderson, N.E. and Kiehl, K.A. 2014. Psychopathy and aggression: When paralimbic dysfunction leads to violence. Curr. Top. Behav. Neurosci. 17: 369–393. 63. Anderson, N.E. and Kiehl, K.A. 2012. The psychopath magnetized: Insights from brain imaging. Trends Cogn. Sci. 16(1): 52–60. 64. Leutgeb, V., Leitner, M., Wabnegger, A. et  al. 2015. Brain abnormalities in high-risk ­violent offenders and their association with psychopathic traits and criminal recidivism. Neuroscience 308: 194–201. 65. Gregory, S., Ffytche, D., Simmons, A., Kumari, V., Howard, M., Hodgins, S., and Blackwood, N. 2012. The antisocial brain: Psychopathy matters. Arch. Gen. Psychiatry 69(9): 962–972. 66. Birbaumer, N., Veit, R., Lotze, M. et  al. 2005. Deficient fear conditioning in psychopathy: A functional magnetic resonance imaging study. Arch. Gen. Psychiatry 62(7): 799–805. 67. Leutgeb, V., Wabnegger, A., Leitner, M. et al. 2016. Altered cerebellar-amygdala connectivity in violent offenders: A resting-state fMRI study. Neurosci. Lett. 610: 160–164. 68. Hosking, J.G., Kastman, E.K., Dorfman, H.M. et al. 2017. Disrupted prefrontal regulation of striatal subjective value signals in psychopathy. Neuron 95(1): 221–231.e4. 69. Hare, R.D. 2003. Manual for the Revised Psychopathy Checklist 2nd ed. Toronto: Multi-Health Systems. 70. Yang, Y., Raine, A., Colletti, P., Toga, A.W., and Narr, K.L. 2010. Morphological alterations in the prefrontal cortex and the amygdala in unsuccessful psychopaths. J. Abnorm. Psychol. 119(3): 546–554. 71. Vanderploeg, R.D., Cooper, D.B., Curtiss, G., Kennedy, J.E., Tate, D.F., and Bowles, A.O. 2018. Predicting treatment response to cognitive rehabilitation in military service members with mild traumatic brain injury. Rehabil. Psychol. 63(2): 194–204. 72. Cooper, D.B., Bowles, A.O., Kennedy, J.E. et al. 2017. Cognitive rehabilitation for military service members with mild traumatic brain injury: A randomized clinical trial. J. Head Trauma Rehabil. 32(3): E1–E15. 73. Choy, O, Raine, A. and Hamilton, R.H. 2018. Stimulation of the prefrontal cortex reduces intentions to commit aggression: A randomized, double-blind, placebo-controlled, stratified, parallel-group trial. J. Neurosci. 38(29):6505-6512.

11 The effects of pollution, toxins, and diet on behavior

Introduction We are constantly exposed to many elements in our environment that can have biological effects. These include pollutants and toxins in the air we breathe, the food we eat, and the water we drink. These pollutants are insidious but can have major effects on antisocial behavior and often compound other existing inequalities, such as socioeconomic status (SES). Moreover, there are many essential nutrients that we require from our diet to develop and maintain a healthy neural system. A lack of any of these nutrients can impact our health and behavior. Chemical imbalances in the body have already been shown to influence behavior. Such imbalances are frequently the result of bad diet and can sometimes be rectified by correcting that diet. This chapter, therefore, will consider the effects of pollution and diet on criminal behavior. Removing or reducing pollution can result in major health benefits, including mental and behavioral health. Moreover, types of food, the body’s ways of processing it, and sensitivities to food additives vary widely from person to person. Many people are allergic or intolerant to some dietary components, and this alone can affect behavior. This variation between people is often biologically based, although the trigger, the food, is environmental. The objectives of this chapter are to consider various toxins and dietary components and to examine their effect on the body and subsequent behavior, with a view to considering potential intervention strategies.

Pollution and toxins in our environment We all know that pollution is bad for us, and in some cities, air quality advisories are constant. People with health issues, the very young, and the very old are regularly advised to stay inside during poor air-quality advisories, and in some parts of the world, the air quality is constantly poor. For example, in India, air pollution is the number one cause of death, killing more than 1.6 ­million people a year, and in Beijing, simply breathing the air has been considered to be equivalent to smoking 40 cigarettes a day.1 Even in the United States, it is estimated that 142 million people live in areas with dangerously high levels of pollution.1 Also, pollution in our world is insidious; even when air quality is considered safe, it can still have major effects on our health and behavior. Although, as you will see in the next section, some specific pollutants have been closely examined, overall, pollution has been generally neglected until now. Many studies have been conducted on the effects of pollution on health, but it is only recently that studies have considered behavioral effects, specifically criminal behavior. 241

242  The effects of pollution, toxins, and diet on behavior

Air pollution is considered to be the largest source of environmentally caused neurological inflammation, and nanoparticles have been shown to enter the brain, causing neurological damage and degeneration.2 A study of air pollution and wind direction in greater London, England, over a 1-year period showed that heightened air pollution was statistically significantly related to an increase in crime in general, with a greater effect size seen with less severe types of crime.3 London has a high number of air-monitoring stations and has safety and pollution levels similar to many cities in North America, allowing the data to be generalized. Over the year studied, 1.8 ­million criminal offences took place in London. An increase of 10 points on the air quality index (AQI: range 0–100) increased the crime rate by 0.9%, and when AQI was above 35, crime increased by 2.8%. The authors state that this is equivalent to a 9% decrease in police activity.3 Robustness and placebo exercises were also conducted to further validate the data. Although almost all crime types were affected, much larger effects were seen with relatively petty crimes, such as pickpocketing, rather than violent crimes and others risking lengthy prison sentences. The study also indicated that pollution levels varied across the city and were related to SES. The  authors separated their sample based on average house prices as a representation of income. A U-shaped relationship was seen in that the effect of pollution on crime was higher in the more affluent neighborhoods and also in the least affluent neighborhoods. The authors suggested that this was because areas with higher exposure were more sensitive to increases in pollution and those with lower exposure may have been more vulnerable for other reasons.3 The authors considered the results in the light of rational choice theory, which argues that an offender weighs the costs, benefits, and risks before making a rational choice to commit a crime.4 The authors looked at lottery sales as a measure of risk-taking and discovered that a 10-point AQI increase was associated with a 1.5% decrease in lottery sales, suggesting that pollution reduces rather than increases risk-taking. They suggested that instead, within the rational choice framework, the changes do not relate to perceived risk but rather that the rise in crime rate may be caused by changes in an offender’s perceived gain versus perceived punishment, and the change could be either a rise in benefit perception or a drop in cost perception. Research has shown that exposure to pollution can increase the secretion of stress hormones, including cortisol,5 and other experiments have shown that increases in cortisol increased an individual’s desire for small, immediate rewards rather than larger, postponed rewards.6 The authors proposed, therefore, that increased pollution may increase crime by causing an individual to discount the cost of prospective punishment, reaching for immediate reward and not  considering the consequences, on days of high pollution.3 In all cases, the levels of pollution were well below the current levels considered to be of enough concern to trigger an air quality advisory in England or the United States, suggesting that such levels are perhaps too high and should be reconsidered.3 The authors suggested that tightening policy on environmental pollution would be a cost-effective way of reducing crime and gaining major health benefits. A  similar study was conducted on crime and pollution in Chicago and Los Angeles, also using wind direction to consider levels of ambient pollution. In Chicago, over an 11-year period, relative criminal activity on either side of a major freeway was considered when the wind blew in the direction of the freeway, and in LA, over an 8-year period, crime rates in the foothills of the San Gabriel Mountains were compared with other areas when the wind blew inland from the ocean.7 This  allowed researchers to compare high- and low-pollution areas, based on the wind ­spreading particulate matter, on the same days. Violent crime rates in Chicago increased by 2.2% per day on the downwind side of the freeway, but no effect was seen on property crime, and violent assaults increased by 6.14% in LA.7 In further analysis, a 10-partsper-billion increase in daily fine particulate pollution resulted in a 17% increase in assaults.8 Interestingly, this study showed an effect on violent crime only, whereas in the London study, although impacting all types of crime, pollution mainly affected minor crimes, suggesting very different etiologies. The authors of the US study suggested several reasons why pollution might increase violent crime. First, they cited studies that showed that increased pollution may increase aggression via increased physical distress. Second, they considered a number of studies that showed that pollution can decrease serotonin levels, which as we know from Chapter 9

Pollution and toxins in our environment  243

increases impulsive aggression9 and also decreases risk aversion10; this could mean that lowered serotonin due to pollution may make some individuals more likely to be aggressive and others more likely to be victimized due to reduced risk aversion.7 Third, air pollutants often result in inflammation in the nervous system, which has been linked to aggression in animal studies. Finally, the authors suggested that increased pollution may affect other body systems, such as the hormonal s­ ystem, as animal studies have shown pollution can cause an increase in testosterone, or oxygen ­deficiency. A very unusual experiment on university students in Israel showed that carbon monoxide can impact cognition, learning, abstract thinking, and memory, and even very low levels affected the s­ tudents’ ability to think clearly.11 In  a comprehensive 9-year study of over 9000  US cities, increase in air pollution resulted in ­statistically significant increases in crime rates in six categories of crime, including murder, rape, assault, robbery, burglary, and motor vehicle theft, but not in larceny, which the authors suspect is due to the fact that it is usually unreported.1 The results remained highly significant after robustness checks and controlling for many variables, such as unemployment rate, inflation, per capita income, poverty rate, city-level demographics, city population, and law enforcement demographics.1  The  researchers followed up these results with studies on the direct effects of air pollution. They first looked at the psychological effects of increased air pollution on unethical behavior and showed that simply psychologically experiencing pollution increased a person’s unethical behavior. The next study investigated the role of anxiety in moderating unethical behavior, as previous s­ tudies have shown that pollution increases anxiety and anxiety can increase unethical behavior (cited in Lu et  al.1). Populations in both the United States and India, a country with severe air pollution, were recruited. Participants were provided with a series of photographs of the same ­cities on either a very polluted or a very clean day and were asked to describe living in such a city. They were then tested for unethical behavior in a variety of tests for cheating and other forms of unethical behavior. Increased anxiety due to observing and imagining pollution was associated with increased cheating and lying. Although the latter tests were lab simulations, the authors p ­ ostulated that anxiety caused by increased pollution is one of the moderators of crime rate increases related to pollution.1 Most studies have been performed at the community level, looking at overall pollution levels in relation to aggregate crime rates, but a study in California looked at the effects of pollution on adolescents at the individual level. Researchers examined almost 700 children at age 9 and ­followed them up annually until age 18, analyzing parent-reported delinquent behavior in relation to ­neighborhood exposure to particulate air pollutants. Baseline and cumulative exposures to ­particulate matter were statistically significantly linked to increased delinquent behavior, and the effect sizes were considered to be comparable to the delinquency differences expected between children 3 and 4 years apart in age, even when controlled for sex, ethnicity, SES, and neighborhood quality.12 Higher pollution was found near freeways and in areas with less green space, areas in which lower-income housing is usually situated, hence increasing exposure and risk in an alreadycompromised community. Aggravating factors included poor parental relations and poor ­parent– child relations, causing increased family stress and making the child even more vulnerable to environmental toxins.12 It  is suggested that air pollution can cause direct and indirect neurological deficits, which in turn affect cognitive abilities and reasoning and increase vulnerability to stress. One hypothesized mechanism is that increased exposure to pollution causes an inflammatory response in cytokines, which are proteins important in signaling between cells. They are produced in central and peripheral immune cells and affect memory and learning by changing maturation and plasticity in the nerve synapses.13 Inflammatory cytokines have been linked to a number of diseases, such as depression and neural degeneration, and some cytokines—for example, interleukin-6, which is released with immune reactivity related to trauma—are elevated in some people with antisocial behavior and aggressive behavior. It is suggested that increased exposure to environmental pollution can lead to a pre-inflammatory condition, which could make a person react adversely to future stressors.13 Other kinds of pollution in our environment can also impact our behavior, such as mold growth in our homes. Mold is common in many homes due to leaks in windows, pipes, showers, and

244  The effects of pollution, toxins, and diet on behavior

bathtubs. Fungi and mold release a number of substances that can cause health and behavioral problems, such as mycotoxins and fragments of fungus.14 Studies have shown that mold exposure can impact various neurological functions, such as balance and reaction time, as well as mood and depression.14 In a study of the effects of mold exposure on young children, 277 babies were followed for 6 years as part of a major health cohort in Poland. The presence of visible mold in the homes of these children was monitored from birth to age 6, together with many other potentially confounding factors, such as lead levels, breastfeeding, sex, maternal education, exposure to tobacco smoke, and number of siblings. The children were assessed using the Wechsler Intelligence Scale for Children–Revised (WISC-R) at age 6. Once all potentially confounding factors were considered, mold exposure was found to significantly reduce IQ. Children who were exposed to mold for long periods (greater than 2 years) had IQs 10 points lower than children with no exposure, although prenatal exposure had no impact.14 Overall, the direct and indirect effects of air pollution have clear negative effects on health, and these and other studies have shown links between air pollution levels and criminal behavior. Actual mechanisms of action are unknown and, so far, have simply been speculated on, from both biological and criminological theory perspectives. Large ecological studies are certainly exciting, and all have controlled for as many variables as possible, but it is impossible to eliminate all confounding variables, so a direct causal relationship is hard to prove. More detailed studies on ­specific toxins such as lead at both the individual and community levels give us more direct insight into cause and effect.

Lead Lead is a dangerous toxin that accumulates in the body tissues over time and affects many body systems, accumulating in the liver, brain, kidneys, and bones. During pregnancy, lead is released from the mother’s bones, impacting the fetus, and can increase the risk of premature birth or stillbirth, miscarriage, and low birth weight.15 It is also extremely damaging to young children and can result in permanent brain and nervous system damage. We obtain lead from the air we breathe, the water we drink, and the food we eat; it is everywhere. Mining, smelting, and recycling result in major environmental lead contamination, as well as the continued use of leaded paint, gasoline, and aviation fuels in some countries. Over 75% of lead use today is in lead-acid batteries for motor vehicles, although it is also found in jewelry, cosmetics, glassware, ammunition, and ceramic glazes. In Mexico, lead in ceramics is a significant contributor of lead transfer to food.16 Lead is also found in unregulated cosmetics and traditional medicines, as some practitioners consider heavy metals to be therapeutic.15,17 Lead is also still found in fishing weights, curtain weights, bullets, and toy jewelry, all of which can be swallowed by small children, as well as in some Mexican candies, kohl (eye makeup), pool cue chalk, and older vinyl miniblinds.17 Lead is also found in some water pipes and solder, so drinking water is frequently contaminated with lead, and recent tests in the United States have shown that drinking water in many US cities contains high lead levels.18 It has long been known that lead is bad for us, so we as a society have tried to reduce lead levels by banning lead in gasoline and paint, but old paint, water pipes, and other sources of lead are still around today. Today’s paint, in the developed world, may no longer contain lead, but much of the world is painted with older paint, meaning that the lead can still leach into the environment. In  the United States, it is estimated that 38  million housing units constructed before 1978 still have lead-based paint (40% of housing units),19,20 and only 30% of countries have banned the use of leaded paint.15 Using New Orleans as an example, researchers compared the level of lead that could be released into the atmosphere from paint with levels of lead from cars. It was estimated that vehicles in New Orleans between 1950 and 1985 emitted 10 179 metric tons of lead. Assuming that the paint in homes was 25.7% lead and that the average house has a painted surface area of 370 m2 (or 4000 sq ft) and that all old homes were sanded down at the same time, it has been estimated that

Pollution and toxins in our environment  245

a further 1811 metric tons of lead would be released.21 This is a substantial amount, and this particulate lead is preserved in soil, which acts as a reservoir and releases substantial amounts of lead in some seasons.21 The release of lead from paint is a major source of lead exposure for children in the developed world, but in the developing world, lead is much more prevalent, from emissions, food and drink containers, water pipes, and traditional medicine.20 Recycling lead-containing material, in particular vehicle batteries, has led to mass lead poisonings and many deaths of children in Nigeria, Senegal, and other countries.15 Effects of lead on the body

Lead accumulates, so the more a person is exposed to, the higher the lead levels in the body. It affects the neurological, gastrointestinal, cardiovascular, and renal systems.15 It also alters hormonal and neurotransmitter systems,21 which, as we have already explored, can increase antisocial behavior when disrupted. It has also been associated with defects in the brain’s hippocampus and cerebral cortex, with changes in nerve development, myelination, and processing. It also interferes with neurotransmitter function and synaptic pruning, both important in behavior and learning.17 Some studies have shown that gray and white matter and parietal lobe and frontal lobe matter, as well as total brain volume, may be reduced after chronic occupational exposure.17 Neuroimaging studies have shown that lead exposure in children can result in neural deficits in adulthood, with structural defects in gray and white matter in regions of the brain related to decision-making.22 Studies on childhood lead exposure, brain deficits, and psychopathy scores indicated that early lead exposure may increase psychopathy scores via damage to regions of the brain.22 Children are particularly at risk from lead exposure, as they absorb four to five times the amount of lead, from any source, that an adult does, and their habit of putting things in their mouths and curiously picking at paint or dirt increases their overall exposure. To compound this, children with poor nutrition are at even greater risk if their diet is lacking in calcium and iron, as this results in their absorption of even higher levels of lead.15 This means that children from lower SES homes are at greatest risk, not only due to increased exposure from older buildings but also because they absorb more lead, compounding the other risk factors associated with lower SES. Children are also at higher risk because their higher brain functions and nervous systems are only just developing, and human and animal studies have shown that lead affects the formation of neural networks and neurotransmitter function, which increase the risk for antisocial behavior.19 Studies have shown that lead causes decreased IQ and verbal abilities, reduced academic success, increased impulsivity, and increased risk of learning disabilities.23 At very high levels, lead exposure can result in coma, seizures, or death, and children who s­ urvive are frequently brain-damaged. However, of more interest to us here from the point of view of behavior is the much more insidious lower levels of exposure that do not result in any obvious outward health problems, yet are known to cause myriad forms of damage to multiple organ ­systems, in particular the brain.15 In the past, blood lead levels (BLLs) below 10 µg/dL were c­ onsidered safe, but the US Centers for Disease Control and Prevention (CDC) reduced this to 5 µg/dL or less in 2012.19 However, research continued to show that such BLLs were associated with reduced cognition, antisocial behavior, and learning disabilities, so today, no BLL is considered safe.15 Even at levels previously considered to be safe, we now know that lead affects healthy brain development, resulting in lowered IQ, reduced attention, poor academic performance, and increased antisocial behavior, as well as other health problems. Recently, early childhood exposure to lead has been shown to cause DNA methylation (change in gene expression) in several genes that can affect disease expression and result in neurological disorders.24 In  2016, the Institute for Health Metrics and Evaluation estimated that lead exposure caused  540  000  deaths, the loss of almost 14  million years of healthy life, and almost 64% of ­developmental intellectual disability globally.15 The  CDC stated that, in 2018, at least 4 million ­children in the United States were exposed to high lead levels at home, and 500  000 children between 1 and 5 years old have BLLs above 5 µg/dL.25 In 2000, the World Health Organization estimated that 40% of children under 5 years of age have BLLs linked to brain damage, and 97% of

246  The effects of pollution, toxins, and diet on behavior

these children live in developing countries.20 However, the risk in the developed world is still high. The scare in Flint, Michigan, in which a major lead influx into the water system led to high lead levels in people’s drinking water, was highly publicized. Yet, a 2016 US report stated that this was by no means a singular event: over 3000 US communities had a greater percentage of children with BLLs higher than 5 µg/dL than did Flint.26 Lead therefore clearly has major health and behavioral risks and is still very pervasive in our society. Effects of lead on antisocial behavior in children

As lead can cause brain damage, it is likely to impact behavior, school performance, IQ, and attention span, all of which increase the risk for antisocial behavior. Lead has been shown to cause substantial cognitive and behavioral deficits.20 We have already shown in previous chapters that negative impacts on a child’s social and learning experiences can have sequelae decades later, so it is not surprising that a toxin that causes such myriad deficits can have long-term effects on behavior and health. Numerous studies have shown that high lead levels in a child’s environment are related to c­ ognitive, learning, and attention deficits and that even relatively low lead levels can cause ­measurable ­deficits. As early as the 1940s, scientists began to note the side effects of lead exposure on antisocial behavior in children. In fact, Randolph Byers, pediatrician and pediatric neurologist, and Elizabeth Lord, ­psychologist, realized in 1943 that children they had treated for lead p ­ oisoning years e­ arlier were being referred to them as adults for assessment of their violent and antisocial behavior.27 A  few ­studies were conducted, but it was not  until Needleman and colleagues’ seminal studies on the effects of lead on children that people began to take notice. Their first study ­examined children with high and low lead levels in the dentine of their teeth. Dentine is considered a better medium than blood, as lead is stored in bones and teeth. Shed deciduous teeth were tested for lead levels, and ­children were presented with a number of academic tests. Children with higher lead levels scored worse on the WISC-R and on verbal, attention, and speech-processing tests. Teacher evaluations of classroom behavior also indicated increased antisocial behavior in children with higher lead levels.28 At the time, these were highly unpopular results which were attacked by the massive lead industry and incited comments that the work provoked “labeling, guilt, anger, recrimination, lawsuits, and witch-hunts.”29(p163) A  New England Journal of Medicine paper on attempts by special ­interest groups to silence scientists stated, “The lead industry hobbled the work of Needleman and ­colleagues on the health risks of low-level lead exposure and intimidated others through coordinated attacks at scientific meetings and skillful manipulation of the procedures for investigating scientific misconduct.”30(p1176) The response by the lead industry to Needleman’s work is quite frightening, but unfortunately not uncommon when big industries try to silence honest science. The history of science is replete with such examples, going back centuries. For example, Dr Pierre Louis, a French physician, was pilloried over 200 years ago for daring to suggest that bloodletting was not therapeutic.30 Unbelievably, the first doctor to suggest washing hands in between surgeries, autopsies, and patient care was vilified by the medical profession, fired and eventually died young in an asylum. Unfortunately, similar vilification of science and scientists still occurs today, especially when the results go against the needs of large industry or political aspirations. Needleman’s seminal and continued work was viciously attacked, with public relations firms and scientific consultants hired to discredit both his work and his integrity. This constant pressure resulted in his investigation for scientific misconduct by his university, the University of Pittsburgh, and the National Institutes of Health, Office of Scientific Integrity. His integrity and his work, of course, were proven to be sound.31 Although this shows the danger of a major industry attempting to gag science (and history, of course, forgets the number of times this has been successful, as such science disappears—think of the number of very successful attempts to produce electric cars over the last century or more that have simply disappeared until public demands exceeded opposition), it also stimulated a new generation of researchers who were electrified by Needleman’s results.31 But Needleman himself stated in an interview

Pollution and toxins in our environment  247

much later that although he eventually was vindicated, he was concerned about what might have happened and what can still happen to scientists who have not yet achieved tenure.31 Nevertheless, Needleman persevered and more studies followed. Needleman and colleagues followed 212 boys in the Pittsburgh public schools from ages 7 through 11  years.32 None of the children had any obvious observable signs of lead toxicity. The  researchers calculated the boys’ bone lead concentrations using a technique called K X-ray fluorescence, which measures cumulative exposure to lead. The study was conducted over 4 years, and during this time teachers and parents periodically filled out questionnaires evaluating the children for aggression, delinquency, and other behavioral problems. In addition, the boys themselves were asked to report whether or not they had engaged in any antisocial behavior, so the study included both external evaluations and self-reports of aggression. Only a slight association between lead levels and behavior was seen at age 7, but by age 11 the children with elevated lead levels were judged by both parents and teachers to be more aggressive, have higher delinquent scores, and have more medical complaints than their low-lead counterparts, and the children themselves reported lead-related increases in antisocial acts.32 The researchers also controlled for the effects of maternal intelligence, SES, and quality of child-rearing. Other problems associated with high lead levels included anxiety, depression, social problems, attention deficits, and somatic complaints.32 The researchers concluded that lead exposure is associated with increased risk for antisocial and delinquent behavior. They felt that if the study was representative of youth in the United States, then lead exposure could be making a major contribution to delinquency. In a later study of offenders, Needleman and colleagues compared 194 convicted juvenile offenders with 146 control adolescents from high schools, and bone lead levels were found to be considerably higher in the offender group in both Caucasians and African Americans.33 The offenders were four times more likely to have high lead levels. Of course, this study does not prove that the lead is causal, because there may be many other biological and environmental differences between the two groups that could result in both higher lead levels and higher offending rates. However, considering that lead has such a range of damaging effects on the brain and behavior, the study does strongly suggest a relationship. These findings in children also appear to hold true in adult violent offenders. Since these earlier, seminal studies, many more recent studies have explored the impacts of lead exposure on children. In a large study of almost 3800 urban fourth-grade children, those with moderate BLLs (10–19 µg/dL, measured when under 3 years old) were suspended from school almost three times more frequently than children with low BLLs (