Handbook of Physiological Research Methods in Health Psychology 9781412926058

The field of health psychology has exploded in the last decade due to progress identifying physiological mechanisms by w

135 35 3MB

English Pages 461 [473] Year 2008

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Contents
Part I: INTRODUCTIONTO THE MEASUREMENT OF PHYSIOLOGICAL PROCESSES
1 - Physiological Research Methods in Health Psychology
2 - Measuring Physiological Processes in Biopsychosocial Research
Part II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS
3 - Measurement of Cortisol
4 - Sympathetic Hormones in Health Psychology Research
5 - Assessment of Salivary α-Amylase in Biobehavioral Research
6 - The Measurement of Blood Pressure in Cardiovascular Research
7 - Cardiovascular Stress Reactivity
8 - Ambulatory Blood Pressure Monitoring
9 - Noninvasive Assessment of Autonomic Influences on the Heart
10 - Laboratory-Based Measures of Immune Parameters and Function
11 - Immunological Functioning II
Part III: BROAD MARKERS OF HEALTH AND DISEASE RISK
12 - Measuring Adiposity in Health Research
13 - The Measurement of Physical Activity, Physical Fitness, and Physical Function
14 - Metabolic Syndrome
15 - Lipid, Lipoprotein, and Inflammatory Markers of Atherosclerosis
16 - Measurement of Sleep by Polysomnography
17 - Neuroimaging
18 - Electroencephalography and Event-Related Potentials
19 - Genetic Factors in Psychophysiological Research
About the Editors
About the Contributors
Recommend Papers

Handbook of Physiological Research Methods in Health Psychology
 9781412926058

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

FM - Luecken - 45370.qxd

9/12/2007

5:43 PM

Page i

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page ii

FM - Luecken - 45370.qxd

9/12/2007

5:43 PM

Page iii

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page iv

Copyright © 2008 by Sage Publications, Inc. All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. For information: Sage Publications, Inc. 2455 Teller Road Thousand Oaks, California 91320 E-mail: [email protected]

Sage Publications India Pvt. Ltd. B 1/I 1 Mohan Cooperative Industrial Area Mathura Road, New Delhi 110 044 India

Sage Publications Ltd. 1 Oliver’s Yard 55 City Road London EC1Y 1SP United Kingdom

Sage Publications Asia-Pacific Pte. Ltd. 33 Pekin Street #02-01 Far East Square Singapore 048763

Printed in the United States of America Library of Congress Cataloging-in-Publication Data Handbook of physiological research methods in health psychology/[edited by] Linda J. Luecken, Linda C. Gallo. p. cm. Includes bibliographical references and index. ISBN 978-1-4129-2605-8 (cloth : alk. paper) 1. Clinical health psychology—Research—Methodology. I. Luecken, Linda J. II. Gallo, Linda C. [DNLM: 1. Psychophysiology—methods. 2. Biological Psychiatry. 3. Diagnostic Techniques and Procedures. WL 103 H2361 2008] R726.7H3652 2008 616.001′9—dc22

2007016667

This book is printed on acid-free paper. 08

09

10

11

12

Acquisitions Editor: Editorial Assistant: Production Editor: Copy Editor: Typesetter: Proofreader: Indexer: Cover Designer: Marketing Manager:

11

10

9

8

7

6

Cheri Dellelo Anna Marie Mesick Catherine M. Chilton Dorothy Hoffman C&M Digitals (P) Ltd. Doris Hus Sheila Bodell Bryan Fishman Amberlyn Erzinger

5

4

3

2

1

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page v

Contents Acknowledgments

ix

PART I. INTRODUCTION TO THE MEASUREMENT

OF PHYSIOLOGICAL PROCESSES

1

1. Physiological Research Methods in Health Psychology:

Applications of the Biopsychosocial Model LINDA C. GALLO

AND

LINDA J. LUECKEN

2. Measuring Physiological Processes in Biopsychosocial Research:

Basic Principles Amid Growing Complexity TIMOTHY W. SMITH

3

AND

11

BERT N. UCHINO

PART II. PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS

35

A. HORMONAL

35

3. Measurement of Cortisol

37

NANCY A. NICOLSON 4. Sympathetic Hormones in Health Psychology Research PAUL J. MILLS

AND

75

MICHAEL G. ZIEGLER

5. Assessment of Salivary α-Amylase in Biobehavioral Research

95

DOUGLAS A. GRANGER, KATIE T. KIVLIGHAN, MONA

EL-SHEIKH, ELANA B. GORDIS, AND LAURA R. STROUD

B. CARDIOVASCULAR 6. The Measurement of Blood Pressure in Cardiovascular Research

115

115

WILLIAM GERIN, TANYA M. GOYAL, ELIZABETH

MOSTOFSKY, AND DAICHI SHIMBO

7. Cardiovascular Stress Reactivity TANYA M. GOYAL, DAICHI SHIMBO, ELIZABETH

MOSTOFSKY, AND WILLIAM GERIN

133

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page vi

8. Ambulatory Blood Pressure Monitoring DENISE JANICKI-DEVERTS

AND

159

THOMAS W. KAMARCK

9. Noninvasive Assessment of Autonomic Influences on the Heart:

Impedance Cardiography and Heart Rate Variability JULIAN F. THAYER, ANITA L. HANSEN,

AND

183

BJORN HELGE JOHNSEN

C. IMMUNE

211

10. Laboratory-Based Measures of Immune Parameters and Function

211

SHAMINI JAIN, SUZI HONG, LAURA REDWINE,

AND PAUL J. MILLS

11. Immunological Functioning II: Field Measures and Viral Challenge ARIC A. PRATHER

AND

235

ANNA L. MARSLAND

PART III. BROAD MARKERS OF HEALTH AND DISEASE RISK

257

12. Measuring Adiposity in Health Research

259

MARY C. DAVIS 13. The Measurement of Physical Activity, Physical Fitness,

and Physical Function

277

JENNIFER L. ETNIER 14. Metabolic Syndrome

299

KATRI RÄIKKÖNEN, EERO KAJANTIE,

ANNA RAUTANEN, AND JOHAN G. ERIKSSON

15. Lipid, Lipoprotein, and Inflammatory Markers of Atherosclerosis

323

CATHERINE M. STONEY 16. Measurement of Sleep by Polysomnography

341

MARTICA HALL, MICHELE L. OKUN, CHARLES W. ATWOOD,

DANIEL J. BUYSSE, AND PATRICK J. STROLLO, JR.

PART IV. EMERGING TOPICS

369

17. Neuroimaging: Overview of Methods and Applications

371

LEE RYAN

AND

GENE E. ALEXANDER

18. Electroencephalography and Event-Related Potentials DAREN C. JACKSON

AND

CORY A. B. JACKSON

19. Genetic Factors in Psychophysiological Research JEANNE MCCAFFERY

395

419

FM - Luecken - 45370.qxd

9/12/2007

Index About the Editors About the Contributors

4:45 PM

Page vii

435

451

453

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page viii

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page ix

Acknowledgments

S

age Publications would like to acknowledge the following reviewers:

Professor Derek W. Johnston School of Psychology University of Aberdeen M. David Rudd, PhD, ABPP Professor and Chair Texas Tech University Psychology Department William R. Lovallo, PhD VA Medical Center and Professor of Psychiatry and Behavioral Sciences University of Oklahoma Health Sciences Center

Sandra Sgoutas-Emch, PhD Professor of Psychology, Department of Psychology Director of Gender Studies Program University of San Diego Dr. Victoria Burns School of Sport and Exercise Sciences University of Birmingham Terry Pace University of Oklahoma Diane C. Tucker University of Alabama

ix

FM - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page x

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 1

Part I

INTRODUCTION

TO THE MEASUREMENT

OF PHYSIOLOGICAL

PROCESSES

1. Physiological Research Methods in Health Psychology: Applications of the Biopsychosocial Model 2. Measuring Physiological Processes in Biopsychosocial Research: Basic Principles Amid Growing Complexity

1

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 2

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 3

CHAPTER

1

Physiological Research Methods in Health Psychology Applications of the Biopsychosocial Model LINDA C. GALLO LINDA J. LUECKEN

T

he emergence of the field of health psychology can be traced to the 1970s, when forward-thinking scientists advanced the notion that health reflects the complex interplay of physio­ logical, psychological, and social factors (Friedman & Adler, 2007). Most notably, in 1977, physician George Engel called for a radical revision of the biomedical paradigm, presenting an alternative framework, subse­ quently labeled the biopsychosocial model (Engel, 1977, 1980; Matarazzo, 1980; Schwartz & Weiss, 1978). In this view, health and illness emerge from multiple influences at the cellular, organismic, interpersonal, and environmental levels. The biopsychosocial model has been adopted as the prevailing paradigm for research, practice, and training within the field of health psychology and rela­ ted disciplines such as behavioral medicine and psychosomatic medicine (Fava & Sonino, 2000; Lipowski, 1977; Suls & Rothman,

2004). Increasingly, the model has also been influential in the realm of traditional medicine (e.g., Frankel, Quill, & McDaniel, 2003; Institute of Medicine, 2004), although the biomedical approach maintains a dominant role (Suls & Rothman, 2004; Waldstein, Neumann, Drossman, & Novack, 2001). With the biopsychosocial model as a guid­ ing framework, dramatic advances have been achieved in understanding how physiologi­ cal, psychological, and social influences interact to affect widely varied health and disease processes (Suls & Rothman, 2004). For example, the etiology and progression of cardiovascular disease is now broadly viewed as stemming from psychosocial influences, such as hostility, depression, stress, and social isolation, as well as genetic, physiological, and behavioral determinants (Everson-Rose & Lewis, 2005; Krantz & McCeney, 2002; Matthews, 2005; Smith & Ruiz, 2002). Similarly, stress and emotions are believed to 3

01 - Luecken - 45370.qxd

4

9/12/2007

4:45 PM

Page 4

PART I: INTRODUCTION TO THE MEASUREMENT OF PHYSIOLOGICAL PROCESSES

play a role in cancer etiology and progression (e.g., Andersen, 2002; Luecken & Compas, 2002), although the evidence is more mixed in this regard. Increasingly, current models of research and treatment have taken biopsy­ chosocial perspectives for many other acute and chronic diseases, including upper respi­ ratory illnesses (Cohen, 2005), diabetes (Gonder-Frederick, Cox, & Ritterband, 2002; Lloyd, Smith, & Weinger, 2005), gastroin­ testinal disorders (Blanchard & Scharff, 2002; Levenstein, 2002; Levy et al., 2006), asthma (Lehrer et al., 2002; Wright, Rodriguez, & Cohen, 1998), arthritis (Keefe et al., 2002), obesity (Chesney, Thurston, & Thomas, 2001), and chronic pain (Campbell, Clauw, & Keefe, 2003; Gatchel, 2004; Keefe, Abernethy, & Campbell, 2005). In each of these cases, illness or disease is believed to emerge and progress as a function of dynam­ ically intertwined genetic or biological pre­ dispositions and influences, psychological states and individual differences, behavior, and social-environmental processes. Given the widespread influence of the biopsychosocial model and associated theo­ retical and empirical developments, it is not surprising that major funding bodies, espe­ cially the National Institutes of Health (NIH), have issued increasingly frequent calls for integrative, transdisciplinary research efforts. For example, in their report regarding future directions for behavioral and social science researchers, a committee formed by the National Research Council advocated a central focus on “multiple pathways to diverse health outcomes” and research that integrates information from the molecular, cellular, psy­ chosocial, and community levels (Singer & Ryff, 2001). Interdisciplinary research initia­ tives are also a critical component of the NIH Roadmap (http://nihroadmap.nih.gov/initiatives .asp; downloaded September 14, 2006), which outlines pressing health issues and opportunities and how they may be optimally addressed to advance medical science. One

such initiative focuses explicitly on inter­ disciplinary research to integrate biological with behavioral and social sciences to facilitate pro­ gress in confronting the nation’s most promi­ nent and intractable health problems. Clearly, an understanding of physiological pathways and methods is critical for researchers inter­ ested in health issues who wish to remain at the forefront of their disciplines. In combination with changes and discov­ eries in the field, and associated emphases of funding organizations, technological advances in collecting and analyzing physiological indicators have contributed to heightened interest in including these variables in health psychology research efforts. Recent years have seen the development of relatively accessible, noninvasive, and low-cost methods for mea­ suring key physiological systems and param­ eters that are relevant to chronic and acute aspects of health and illness. Such technologies have helped elucidate specific biological path­ ways that link psychosocial factors with health. For example, sympathetic and parasym­ pathetic aspects of autonomic nervous system functioning (see Thayer & colleagues, this vol­ ume); hormonal pathways, including those of the hypothalamic-pituitary-adrenal axis (see Nicolson, this volume) and sympatheticadrenal medullary system (see Mills & Ziegler, this volume); and alterations in immune and inflammatory processes (see Stoney, Jain & colleagues, and Prather & Marsland, this vol­ ume) are fundamental to understanding how variables such as stress, emotions, and social relationships can contribute directly to the emergence of acute and chronic diseases. These pathways are believed to represent a common conduit connecting psychosocial factors with multiple health and disease outcomes, and they are therefore relevant to widely varied research topic areas. Further, their inclusion in research efforts has become a reality, given improvements in measurement and assay methods and advances in understanding the optimal approaches to assessment.

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 5

Physiological Research Methods in Health Psychology

In additional lines of research, scientists have embraced the inclusion of broad indica­ tors of risk to provide a holistic assessment of health, a comprehensive understanding of the roles of psychosocial factors in health, and an evaluation of the utility of psychosocial interventions for improving health. These approaches have ranged from relatively simple, low-cost strategies such as including measures of waist circumference or body mass index to address psychosocial influences on obesity and chronic disease (e.g., see Davis, this volume) to more complex approaches of examining psychological and social processes or characteristics in relation to blood markers of inflammation (for example, see chapters by Jain & colleagues and Stoney, this volume)—a process increasingly viewed as relevant in the etiology and progression of myriad chronic health problems. Other researchers have begun to explore the neural pathways involved in mind-body con­ nections (see chapters by Ryan & Alexander and Jackson & Jackson, this volume). Advances in neural imaging and examination of electrical potentials in the brain may provide a critical step in our understanding of exactly how stress, emotions, and other psychosocial characteristics and experiences can “get under the skin” to affect health and illness. Consideration of physiological variables generally requires a level of training that has not traditionally been available in many psychology training programs. Thus, for some researchers, the enthusiasm for including physiological parameters in their work may be tempered by confusion regarding what are the most appropriate indicators and how they should be conceptualized, assessed, and ana­ lyzed. The current text represents an intro­ ductory bridge for those who wish to newly incorporate physiological pathways into their existing work, or who would like a better understanding of physiological measures they encounter in research reports, whether or not they have an interest in collecting their own

physiological data. Each of the chapters is authored by recognized experts, guided by the latest in research on the optimal methods for assessment, analysis, and interpretation of the relevant physiological measure. Many excellent recent volumes have sum­ marized and critiqued the literature to date concerning interrelationships among psycho­ logical, social, and physiological factors in health (Ader et al., 2006; Baum, Revenson, & Singer, 2001; Christensen, Martin, & Smyth, 2004; Sutton, Baum, & Johnston, 2004). Additional resources address highly technical, specialized information about specific physio­ logical measures (e.g., Cacioppo, Tassinary, & Bernston, 2000). Rather than duplicate the efforts of these texts, the goal of this volume is to provide an overview for the collection, analysis, and interpretation of physiological measures that are commonly used, relatively accessible to a broad array of researchers, and do not require hospitalization of participants, extensive medical training or supervision, or invasive and expensive techniques. The text assumes little or no prior knowledge of phys­ iological systems or assessment of its readers, and therefore can be thought of as a primer, or gateway book, for researchers new to physiological measurement. In their introductory chapter, Drs. Tim Smith and Bert Uchino review basic principles of measurement and assessment, providing an important context for understanding and crit­ ically evaluating the measurement approaches described in the remainder of the volume. Their chapter also describes pressing chal­ lenges and future directions associated with the use of physiological measures in health psychology. This thorough overview is fol­ lowed by a series of chapters that describe assessment approaches across the major physiological systems relevant to the bio­ sychosocial perspective—specifically, the hor­ monal, cardiovascular, and immune systems. The first series of chapters in this section provides guidance on assessing hormonal

5

01 - Luecken - 45370.qxd

6

9/12/2007

4:45 PM

Page 6

PART I: INTRODUCTION TO THE MEASUREMENT OF PHYSIOLOGICAL PROCESSES

systems relevant to a variety of health issues and research questions, including suggestions for “best practices” in their collection, stor­ age, assay, statistical analysis, and reporting of results. First, Dr. Nancy Nicolson presents a discussion of cortisol assessment, a primary hormone associated with the hypothalamicpituitary-adrenal axis. Next, Drs. Paul Mills and Michael Ziegler discuss methods and issues in assessing the sympathetic-adrenal medullary system hormones, epinephrine and norepinephrine. Finally, Dr. Douglas Granger and his colleagues have contributed a chapter concerning a novel method for assessing sympathetic nervous system func­ tioning, via salivary alpha-amylase assay. The next set of chapters reviews various pathways and parameters intrinsic to the cardiovascular system. Drs. Bill Gerin and Tonya Goyal and their colleagues present an overview of clinic- and home-based assess­ ment of blood pressure and heart rate, and also describe assessment issues in cardiovas­ cular stress reactivity research. Drs. Denise Janicki-Deverts and Thomas Kamarck pro­ vide an excellent discussion and overview of ambulatory blood pressure monitoring. Finally, Dr. Julian Thayer and colleagues review the relatively specialized, yet extremely informative techniques of impendance cardio­ graphy and heart rate variability assessment. In the section’s final group of chapters, parameters and methods for assessing immune functioning are described. Shamini Jain and her colleagues address laboratory methods of immune assessment, including methods for evaluating immune responses to acute stress. Widely available and more spe­ cialized assessment techniques are reviewed, and their utility, advantages, and disadvan­ tages discussed. Eric Prather and Dr. Anna Marsland then describe methods for assessing immune functioning in the field, reviewing a variety of in vivo methods such as examining immune responses to viral infection, responses to immunization, and wound healing.

The third section of the book focuses on broad markers of health and disease risk that are applicable to researchers in health psychology and many other fields. First, Dr. Mary Davis discusses issues in obesity and central adiposity measurement. Cost-effective and easily accessible approaches, such as body mass index and waist circumference, as well as technologically advanced and special­ ized methods, including dual-energy x-ray absorptiometry of body fat, are described. Dr. Jennifer Etnier then provides a thorough review of various approaches to assessing physical fitness and activity, including selfreport, direct, and indirect approaches suit­ able to a broad range of research questions. Next, Dr. Katri Räikkönen and her colleagues describe issues in conceptualizing and evaluat­ ing the metabolic syndrome—a cluster of related risk factors that is becoming increas­ ingly prevalent in the United States and other industrialized nations. In the following chap­ ter, Dr. Catherine Stoney reviews issues in assessing markers of atherosclerosis, with a particular focus on lipids, lipoproteins, and inflammatory factors. Although described in the context of cardiovascular disease, the dis­ cussion of inflammatory factors is likely to be of widespread interest, given growing evidence suggesting that inflammatory processes are broadly relevant to health and disease. Finally, Dr. Martica Hall and her colleagues provide an excellent overview of polysomnographic assessment of sleep, reviewing the connec­ tion between sleep and health, and discussing both laboratory and ambulatory assessment approaches. The final chapters in the text address emerging topics in the field of health psychol­ ogy. The first two chapters review assessments of neural systems that may be of interest to researchers who wish to pursue nuanced questions regarding integration across central and peripheral physiological responses. Drs. Lee Ryan and Gene Alexander provide an overview of major neuroimaging methods,

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 7

Physiological Research Methods in Health Psychology

and Drs. Daren Jackson and Cory Jackson review assessment of electroencephalography and event-related potentials. Both of these chapters describe applications of these state­ of-the-art techniques to health psychology researchers. In the final chapter of the text, Dr. Jeanne McCafferey provides a highly interesting and informative discussion of the assessment of genetic factors in health psy­ chology research. Given the developing knowledge base regarding the roles of genetic factors in health and disease, as well as research pointing to the interactive influence of genetic predispositions and social and psy­ chological factors in determining physiological responses, this chapter provides an important perspective for health psychology researchers.

In aggregate, the book offers a compre­ hensive overview of a diverse set of phy­ siological measures that are becoming increasingly important in the field of health psychology and related research endeavors. It is our hope that the material will inform and inspire future research that incorporates multiple levels of analysis in psychological research by including theoretically meaning­ ful physiological parameters. Optimally, we hope that this information will promote the use of state-of-the-art and methodologically sound techniques for the most reliable mea­ surement, analysis, and interpretation of physiological parameters in continued efforts to understand health and disease through biopsychosocial research.

REFERENCES Ader, R., Dantzer, R., Glaser, R., Heijnen, C., Irwin, M., Padgett, D., et al. (2006). Psychoneuroimmunology (4th ed.). New York: Elsevier. Andersen, B. L. (2002). Biobehavioral outcomes following psychological inter­ ventions for cancer patients. Journal of Consulting and Clinical Psychology, 70, 590–610. Baum, A., Revenson, T. A., & Singer, J. E. (2001). Handbook of health psychology. Mahwah, NJ: Lawrence Erlbaum Associates. Blanchard, E. B., & Scharff, L. (2002). Psychosocial aspects of assessment and treat­ ment of irritable bowel syndrome in adults and recurrent abdominal pain in children. Journal of Consulting and Clinical Psychology, 70, 725–738. Cacioppo, J. T., Tassinary, L. G., & Berntson, G. G. (Eds.). (2000). Handbook of psychophysiology (2nd ed.) New York: Cambridge University Press. Campbell, L. C., Clauw, D. J., & Keefe, F. J. (2003). Persistent pain and depression: A biopsychosocial perspective. Biological Psychiatry, 54, 399–409. Chesney, M. A., Thurston, R. C., & Thomas, K. A. (2001). Creating social and public health environments to sustain behavior change: Lessons from obesity research. In N. Schneiderman, M. A. Speers, J. M. Silva, H. Tomes, & J. H. Gentry (Eds.), Integrating behavioral and social sciences with public health (pp. 31–50). Washington DC: American Psychological Association. Christensen, A. J., Martin, R., & Smyth, J. (2004). Encyclopedia of health psychol­ ogy. New York: Kluwer Academic Press. Cohen, S. (2005). Keynote Presentation at the Eighth International Congress of Behavioral Medicine: The Pittsburgh Common Cold Studies: Psychosocial predictors of susceptibility to respiratory infectious illness. International Journal of Behavioral Medicine, 12, 123–131.

7

01 - Luecken - 45370.qxd

8

9/12/2007

4:45 PM

Page 8

PART I: INTRODUCTION TO THE MEASUREMENT OF PHYSIOLOGICAL PROCESSES Engel, G. L. (1977). The need for a new medical model: A challenge for biomedicine. Science, 196, 129–136. Engel, G. L. (1980). The clinical application of the biopsychosocial model. American Journal of Psychiatry, 137, 535–544. Everson-Rose, S. A., & Lewis, T. T. (2005). Psychosocial factors and cardiovascu­ lar diseases. Annual Review of Public Health, 26, 469–500. Fava, G. A., & Sonino, N. (2000). Psychosomatic medicine: Emerging trends and perspectives. Psychotherapy & Psychosomatics, 69, 184–197. Frankel, R. M., Quill, T., & McDaniel, S. (2003). The biopsychosocial approach: Past, present, future. Rochester, NY: University of Rochester Press. Friedman, H. S., & Adler, N. E. (2007). The history and background of health psychology. In H. S. Friedman & R. C. Siler (Eds.), Foundations of health psy­ chology (pp. 1–18). New York: Oxford University Press. Gatchel, R. J. (2004). Comorbidity of chronic pain and mental health disorders: the biopsychosocial perspective. American Psychologist, 59, 795–805. Gonder-Frederick, L. A., Cox, D. J., & Ritterband, L. M. (2002). Diabetes and behavioral medicine: The second decade. Journal of Consulting and Clinical Psychology, 70, 611–625. Institute of Medicine. (2004). Improving medical education: Enhancing the behav­ ioral and social science content of medical school curricula. Washington, DC: National Academies Press. Keefe, F. J., Abernethy, A. P., & Campbell, C. (2005). Psychological approaches to understanding and treating disease-related pain. Annual Review of Psychology, 56, 601–630. Keefe, F. J., Smith, S. J., Buffington, A. L., Gibson, J., Studts, J. L., & Caldwell, D. S. (2002). Recent advances and future directions in the biopsychosocial assess­ ment and treatment of arthritis. Journal of Consulting & Clinical Psychology, 70, 640–655. Krantz, D. S., & McCeney, M. K. (2002). Effects of psychological and social factors on organic disease: A critical assessment of research on coronary heart disease. Annual Review of Psychology, 53, 341–369. Lehrer, P., Feldman, J., Giardino, N., Song, H. S., & Schmaling, K. (2002). Psychological aspects of asthma. Journal of Consulting & Clinical Psychology, 70, 691–711. Levenstein, S. (2002). Psychosocial factors in peptic ulcer and inflammatory bowel disease. Journal of Consulting and Clinical Psychology, 70, 739–750. Levy, R. L., Olden, K. W., Naliboff, B. D., Bradley, L. A., Francisconi, C., Drossman, D. A., et al. (2006). Psychosocial aspects of the functional gastroin­ testinal disorders. Gastroenterology, 130, 1447–1458. Lipowski, Z. J. (1977). Psychosomatic medicine in the seventies: An overview. American Journal of Psychiatry, 134, 233–244. Lloyd, C., Smith, J., & Weinger, K. (2005). Stress and diabetes: A review of the links. Diabetes Spectrum, 18, 121–127. Luecken, L. J., & Compas, B. E. (2002). Stress, coping, and immune function in breast cancer. Annals of Behavioral Medicine, 24, 336–344. Matarazzo, J. D. (1980). Behavioral health and behavioral medicine: Frontiers for a new health psychology. American Psychologist, 35, 807–817. Matthews, K. A. (2005). Psychological perspectives on the development of coronary heart disease. American Psychologist, 60, 783–796.

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 9

Physiological Research Methods in Health Psychology Schwartz, G. E., & Weiss, S. M. (1978). Yale Conference on Behavioral Medicine: A proposed definition and statement of goals. Journal of Behavioral Medicine, 1, 3–12. Singer, B. H., & Ryff, C. D. (2001). New horizons in health: An integrative approach. Washington DC: National Academy Press. Smith, T. W., & Ruiz, J. M. (2002). Psychosocial influences on the development and course of coronary heart disease: Current status and implications for research and practice. Journal of Consulting and Clinical Psychology, 70, 548–568. Suls, J., & Rothman, A. (2004). Evolution of the biopsychosocial model: Prospects and challenges for health psychology. Health Psychology, 23, 119–125. Sutton, S., Baum, A., & Johnston, M. (2004). The Sage handbook of health psy­ chology. London: Sage. Waldstein, S. R., Neumann, S. A., Drossman, D. A., & Novack, D. H. (2001). Teaching psychosomatic (biopsychosocial) medicine in United States medical schools: Survey findings. Psychosomatic Medicine, 63, 335–343. Wright, R. J., Rodriguez, M., & Cohen, S. (1998). Review of psychosocial stress and asthma: An integrated biopsychosocial approach. Thorax, 53, 1066–1074.

9

01 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 10

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 11

CHAPTER

2

Measuring Physiological Processes in Biopsychosocial Research Basic Principles Amid Growing Complexity TIMOTHY W. SMITH BERT N. UCHINO

T

he measurement of physiological states and processes has always been central in health psychology. After all, what would the study of associations between mind and body or behavior and health be if some sort of physiological vari­ ables were not included in many of its individ­ ual investigations? Yet, the field has also included frequent instances of imprecise con­ ceptualization and measurement of biomedi­ cal variables. Early in the history of health psychology, for example, it was not uncom­ mon for health outcomes to be measured through self-reports of physical symptoms—a strategy that assumed reporting the symp­ toms of “runny nose” or “cough” was an adequate indication of upper respiratory infection. Self-reports of symptoms and health status do contain useful information about current and future physical health (Idler & Benyamini, 1997; Orts et al., 1995), and subjective judgments about physical

symptoms or health are an important out­ come in their own right. However, such mea­ sures are ambiguous and might be misleading when used specifically to assess actual dis­ ease, because they can contain systematic variance (e.g., somatic complaining, denial or minimization of illness) that is unrelated to the construct of interest. Even when more direct and unambiguous health endpoints were used (e.g., mortality, independently diagnosed morbidity), the potential mechanistic links between psy­ chosocial inputs or risk factors (e.g., stress, personality, social relations) and subsequent disease were often not well defined in earlier studies in health psychology and related fields. “Black box” models positing—but not measuring—some often incompletely speci­ fied intervening physiological mechanism were characteristic of much early research. There are many indications of important progress in health psychology and related 11

02 - Luecken - 45370.qxd

12

9/12/2007

4:45 PM

Page 12

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

fields over the last 25 years, but few are as clear as the increased use of sophisticated physiological measurements of health end­ points and underlying mechanisms. These advances have been a key contribution to the fields’ growing impact in biomedical science and health care, and the continued use of sound physiological measures is an essential requirement for its future. Many of the methods described in this volume involve complex, multifaceted techni­ cal challenges, such as maximizing the sensi­ tivity and specificity of assays or the recording and storage of large amounts of complex physiological information in ambulatory stud­ ies. For researchers interested in the develop­ ment, evaluation, or use of these methods, these technical challenges are critical consid­ erations. The field has benefited greatly from technical advances, but beyond this pressing technical agenda is another equally important agenda having to do with traditional issues in measurement such as reliability, validity, and utility. Even when the methods are highly complex or innovative, overcoming technical challenges in physiological measurement is perhaps best seen as the “end of the begin­ ning” rather than the “beginning of the end” in the comprehensive development, evalua­ tion, and use of such methods. This volume provides valuable introduc­ tions to issues in the measurement of a wide variety of physiological variables at the core of the field’s present—and in several cases the future—status. Our purpose here is to place these overviews in context. Specifically, we attempt to place them in the context of basic principles of measurement and assess­ ment. We hope seeing the measurement of specific physiological variables in this con­ text will lead to both more effective use and critical interpretation of these measures, but also to greater interest in measurement research itself. Grand questions of the nature of links between the mind and body or between behavior and physiological function

easily capture the attention of our field, and with good reason. In contrast, smaller ques­ tions regarding the extent to which our mea­ sures accurately reflect the psychological constructs and physiological processes at the heart of these larger questions are often relatively neglected. Yet, reliable, valid, and useful measures are an obvious necessity for progress on the central issues of the field (Smith, 2007). We hope that, in addition to learning to use physiological measures effec­ tively and evaluate them critically, readers will accept our invitation to study these mea­ surement procedures. Each of these activities requires careful and skeptical thinking based in a conceptual understanding of the process of measurement. In what follows, we outline principles of measurement that serve as a general guide to evaluating any measurement procedure. We then discuss some general issues in the application of physiological measures in health psychology, and close with comments regarding future challenges in maximizing the value of physiological measurement in future research.

PRINCIPLES OF MEASUREMENT Measurement has been defined in many ways, but most approaches include Stevens’s (1951, 1959, 1968) assertion that measure­ ment is “the assignment of numbers to aspects of objects or events according to one or another rule or convention” (Stevens, 1968, p. 850). Refinements of this basic definition have suggested that systemati­ cally assigning numbers is insufficient in des­ cribing measurement. Dawes and Smith (1985) for example argued that “the assign­ ment of numbers not only must be orderly if it is to yield measurement but must also rep­ resent meaningful attributes and yield mean­ ingful predictions” (p. 511). Similarly, Judd and McClelland (1998) asserted that the

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 13

Measuring Physiological Processes in Biopsychosocial Research

assignment of numbers represents measure­ ment “only if the subsequent numbers ulti­ mately represent something of meaning, some regularity of attributes or behaviors that permits prediction” (p. 181). Most modern approaches to measure­ ment emphasize this systematic relationship between numbers assigned in a specified pro­ cess of observation and meaningful attributes that are typically only indirectly observed. Previously (Smith, 2007) we have suggested that a definition by Judd and McClelland (1998) is a useful starting point for any dis­ cussion of measurement: The compact model or description that we construct of observations through measure­ ment we will call a scale or a variable. The meaningful attribute or regularity that it is presumed to represent we will call a con­ struct. Accordingly, measurement consists of rules that assign scale or variable values to entities to represent the constructs that are thought to be theoretically meaningful. (Judd & McLelleand, 1998, p. 181)

From this perspective, it is clear that critical evaluation of the adequacy of any measurement procedure entails examining the processes and rules by which numbers are assigned to observations, as well as the corre­ spondence between the resulting values and the constructs they are intended to represent. One of the fundamental challenges for research in health psychology and the related fields of behavioral medicine or psychoso­ matic medicine stems from the fact that they are all based on the biopsychosocial model (Engel, 1977). As a result, research questions often cut across very different levels of analy­ sis within the biopsychosocial framework. For example, many of the measurement methods addressed in the present volume are often applied when testing hypotheses about the association of characteristics of environ­ ments (e.g., job strain, social isolation) or

psychological attributes of individuals (e.g., personality traits, emotional distress) with physiological processes implicated in the development or course of disease or its related symptoms. Each level of analysis in the biopsychosocial model from molecular genet­ ics to biochemistry to physiology to individual behavior to interpersonal and even sociocul­ tural processes involves its own approaches to measurement. Hence, integrative research often requires combinations of diverse assumptions, concepts, and methods in this most basic aspect of conducting science. Importantly, these research traditions often emphasize different aspects of the pro­ cess of measurement. Some are primarily con­ cerned with the technical challenges inherent in gathering the “signals” that form the basis of observations and assigning precisely scaled numbers to them. Other measurement tradi­ tions emphasize the thorny task of evaluating correspondence between more easily acquired observations and unobservable hypothetical constructs. Given these widely varying chal­ lenges and emphases, it is not surprising that measurement traditions within the biopsy­ chosocial model often involve quite different implicit assumptions regarding the correspon­ dence between measured variables and the related constructs. In the definition offered by Judd and McClelland (1998 ), this is the cor­ respondence between the “compact model or description” we construct through systematic observation and assignment of numbers, on the one hand, and the entities or constructs these numbers are intended to represent on the other. Behavioral and biomedical sciences may differ in their view of this issue of corre­ spondence. For example, seemingly a greater distance or degree of inference is involved in the assumption that scores on a self-report inventory reflect a specific personality con­ struct of interest than in the parallel assump­ tion that the number produced by a salivary assay of cortisol reflects metabolically active amounts of that specific hormone. The gap

13

02 - Luecken - 45370.qxd

14

9/12/2007

4:45 PM

Page 14

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

between the rules for assigning numbers through systematic observation and the con­ struct or entity of interest appears to be much larger and fraught with inferential complexity in the former case. Both measurement proce­ dures pose substantial challenges, but they differ perhaps in kind. The salivary cortisol assay may pose complex issues of applied bio­ chemistry and physiology, but perhaps less of the conceptual and quantitative challenges involved in evaluating the correspondence between test responses and a specific yet indi­ rectly observed psychological construct. Grappling with this diverse array of chal­ lenges is a fundamental component of the research process in health psychology and related fields. From the perspective of one level of analysis within the biopsychosocial model, the measurement procedures of another might seem strange or even less sci­ entific. Yet, for all levels of analysis, each step in the process of measurement consists of a set of assumptions, and those assump­ tions must be clearly articulated and thoughtfully—even skeptically—examined.

The Central Role of Theory The measurement traditions of the various levels of analysis of the biopsychosocial model may also differ in the extent to which they acknowledge that theory plays a role in the measurement process. For many biomed­ ical disciplines, technical challenges of mea­ surement are most salient and the gap or inferential distance between resulting scale values and the attributes they are intended to represent seems small. These disciplines are likely to also presume a small role for theory. Disciplines in which the measurement chal­ lenges inherent in the collection of observa­ tion are less technical often involve more attention to conceptual and quantitative issues in making inferences about hypotheti­ cal constructs on the basis of observations. These disciplines typically more readily

acknowledge the theory-driven nature of measurement. Psychology is generally in the latter camp, but this varies across specific fields. Further, some traditions within psy­ chology have differing stances on this issue. For example, the tradition of operationalism or operationism equates the concept to be measured with the research procedures used in a given study. This notion that the “con­ cept is synonymous with the correspond­ ing set of operations” (Bridgman, 1927, p. 5) dates to discussions of measurement in the highly inferential domain of physics. The equating of concepts and operations was embraced in some parts of psychology as an answer to the troubling ambiguities in mea­ suring complex and unobserved psycho­ logical constructs. The author originally associated with this approach later criticized its simplistic application in psychology (Bridgman, 1945), as did many others. Yet, the approach took root in many research areas, including the behavioral tradition within psychology, which has been influen­ tial in the history of health psychology and behavioral medicine. From this perspective, generalizations or inferences from research procedures (i.e., operational definitions) to unobserved entities or hypothetical constructs are typically met with considerable skepticism and may even be seen as unscientific. Hence, traces of a discredited measurement tradition that tried to close the inherent gap between operations and constructs by definitional fiat can still be found in several disciplines of the biopsychosocial model, including psychology. More modern approaches to measurement emphasize that any given concept can be “operationalized” in many different ways, and any specific operation contains both construct-relevant variance and erroneous or construct-irrelevant variance. By examining the pattern of empirical findings across stud­ ies using multiple imperfect operations, the irrelevancies tend to fall to the background as unreplicated effects and construct-relevant

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 15

Measuring Physiological Processes in Biopsychosocial Research

variance emerges in the foreground in the form of consistent findings across the vari­ ous specific operations. Hence, this general approach requires critical analysis and evalu­ ation of the correspondence between research procedures and unobserved constructs, rather than eschewing the issue. In psychology, the formal articulation of this measurement tradition dates to the mid­ 20th century (Cronbach & Meehl, 1955), and is the basis for current discussions of the multilayered theory-driven nature of mea­ surement. In an excellent recent presentation

Table 2.1

of this tradition, McFall (2005) describes eight layers or levels at which theoretical issues influence the process of measurement, a view that is applicable beyond the bound­ aries of traditional psychological science. These are presented in Table 2.1. Through this framework McFall (2005) articulates and summarizes a general guiding principle in this approach to measurement—clear and critical thinking at each of these interrelated layers or levels is essential in the develop­ ment, evaluation, refinement, and use of any measurement procedure.

McFall’s (2005) Multiple Levels of Theory in Measurement

Level

Issue or Activity

Postulates

Philosophical assumptions and untested “givens”

Formal theoretical constructions

Hypothetical constructs and nomological networks

Referents

Category labels for observable exemplars

Instrumental methods

Techniques and procedures for sampling exemplars

Measurement model

Conversion to units and assignment of numbers

Data reduction

Aggregation of numbers into statistical summaries

Data analysis

Processing of summary statistics by analytic method

Interpretation and inference

Evaluation of results relative to original question

The first level concerns basic assumptions or postulates. These theoretical propositions are often unarticulated but essential influ­ ences on the selection of measurements and related methods. For example, the assump­ tion that individuals’ display of physiological aspects of psychological stress while they are awake links aversive life circumstances or experiences with the development of disease is an example of a basic postulate common to many of the chapters of this book. Alternative or supplementary assumptions might suggest that the disruption of restorative functions (e.g., sleep and associated physiological pro­ cesses; see Hall et al., this volume) links aver­ sive life circumstances with disease, or that stress causes disease through an acceleration

of the process of genetic aging (Epel et al., 2004). These assumptions would guide the process of measurement in other directions. Based in such postulates, formal theo­ retical constructions involve conceptual definitions of hypothetical constructs and hypothesized relationships among them. The clarity and specificity of these constructions is an important influence on the quality of related measurement procedures. Imprecise conceptualizations are unlikely to lead to sound measurement procedures, as irrelevant phenomena might be included in a measure or essential features omitted. Importantly, these conceptual definitions and specifica­ tions should include descriptions of how con­ ceptually related but distinct constructs differ

15

02 - Luecken - 45370.qxd

16

9/12/2007

4:45 PM

Page 16

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

from one another, as well as the predicted relative levels of associations among distinct constructs. These conceptual specifications are essential in designing theory-driven tests of convergent and discriminant (i.e., diver­ gent) aspects of construct validity. The con­ ceptual models or nomological nets should also specify the next level of conceptualiza­ tion in measurement—referents or general categories of observable exemplars hypothe­ sized to reflect unobserved constructs. Examples of such categories include subjective experi­ ence, overt behavior, and various aspects of physiological functioning. In conducting research, clear conceptual­ ization regarding theoretically appropriate referents guides the selection of specific tasks, techniques, procedures, or instruments. This next level of the measurement process entails instrumental methods or actual research operations. For example, the concept chronic inflammatory activity could be assessed through the instrumental method of a blood assay of high sensitivity C-reactive protein (hsCRP) or interleukin 6(Il-6). Once a specific instrumental method has been selected, a measurement model must be specified. In this process, referents sampled through specific instrumental methods are converted to some sort of units and assigned numerical values on a scale. Raw data derived from this measurement model are then abstracted in the next step—data reduction, a process through which features are extracted from measurement values and summarized to reflect the construct of interest. For example, in quantifying the construct of “cardiovascu­ lar reactivity” a beat by beat measure of blood pressure scaled as millimeters of mercury (mmHg) could be reduced and quantified to capture a mean increase over a resting baseline of a specific number of readings within that sample, the peak value over baseline, or the area under the blood pressure by time curve. In the next level, data analysis, some sort of quantitative method is used to test predicted

patterns of covariation between the values on the measure under consideration and some other variable. Each method of data analysis is based on assumptions about the nature of the variables involved and the nature of the ques­ tion asked. For example, analytic techniques typically are based on assumptions regarding the level of measurement or scale of measure­ ment (i.e., nominal, ordinal, interval, or ratio). Use of an analytic technique that assumes a different level of measurement than the vari­ able(s) under consideration can produce invalid conclusions about the presence and magnitude of covariation. Hence, the specific analytic procedure selected should be consis­ tent with prior levels of conceptualization. Finally, the interpretation and inference level involves evaluating results of the analy­ ses. The knowledge gained is considered at this point in the process, and it obviously depends on the logical consistency across the prior conceptual levels or steps. McFall (2005) articulates the issue addressed at this level as “What, if anything, was learned from (7) the data analyses of (6) the summary statistics (5) generated by the measurement model (4) of the responses gathered by the instrumental methods (3) designed to sample the referents (2) for the formal theoretical constructs (1) supported by the basic assump­ tions?” (p. 318). Results consistent with expectations support theoretical models, but only to the extent that each of the steps has been consistent with the related features of the theoretical framework. Expected results can represent “false support” if decisions at any specific level are inconsistent with the theoretical framework. For example, an asso­ ciation between self-reports of chronic stress and self-reports of “swollen and inflamed joints” could be misleading support for the theory that stress causes chronic inflamma­ tion if these instrumental methods are consid­ ered to be inconsistent or uninformative with regard to the guiding theory. Similarly, unex­ pected results could reflect difficulties at any

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 17

Measuring Physiological Processes in Biopsychosocial Research

level or step, from data analysis to founda­ tional assumptions. As McFall (2005) and others (e.g., Meehl, 1978) have noted, these multiple possible slips between conclusions about statistical covariation and inferences about associations among hypothetical constructs create diffi­ culty in obtaining unambiguous evidence sup­ porting or falsifying psychological theories or models. Reducing this multilevel ambiguity is a central challenge in research, even when seemingly less ambiguous physiological mea­ surements are involved. An essential first step in managing this inherent impediment to scientific progress is clear and specific con­ ceptual reasoning across the levels McFall (2005) delineates, including the systematic

Table 2.2 • • • •

articulation and testing of assumptions and alternatives at each level. Even when the procedure under consideration is a seemingly straightforward physiological measure, this exercise in critical thinking is imperative. West and Finch (1997) have offered a sec­ ond and seemingly simpler guide to this sort of conceptual clarity about the underpinnings of measurement. Although their perspective was initially described in the context of the importance of theory-driven measurement in personality research, it is more broadly appli­ cable. Their guide takes the form of a series of questions (listed, for emphasis, in Table 2.2). When asked, thoughtfully answered, and sys­ tematically tested, these questions increase the likelihood of sound measurement.

Questions to Guide Conceptualization in Measurement

What is the expected degree of relationship among items or indicators? What is the structure of the construct? What is the stability of the construct? What is the expected pattern of relationships of measures of the construct of interest with other measures of the same construct and with measures of other constructs?

SOURCE: Adapted from West and Finch (1997).

The first asks, “What is the expected degree of relationship among individual indi­ cators or items that constitute the measure of the construct?” In many cases, this associa­ tion is expected to be large, because the mul­ tiple indicators are expected to reflect a single construct. In others, a lower level of associa­ tion might be expected, as when the construct has a broad conceptual definition and the multiple indicators are intended to reflect this diversity. In both of these cases, at least some association among items or indicators is to be expected because the individual’s standing on the construct of interest is presumed to cause the observed variation on the measured indi­ cators (Bollen & Lennox, 1991). In other cases, little or no association among indicators would be expected. As an

example, West and Finch (1997) describe the expected level of association among items within inventories assessing the individual’s exposure to major life events (e.g., Holmes & Rahe, 1967). In this case, there is little or no reason to assert that the experience of one major life event (e.g., starting a new job) should be associated with an increased prob­ ability of experiencing a second (e.g., illness of a family member). Such associations, if they occurred, might not contradict the guid­ ing conceptual model, but they are also not inherent to it. This is because responses to items of the life events scale are not an effect of the individual’s standing on the hypotheti­ cal construct of the degree of stressful life events. Rather, responses on these items or indicators—or perhaps more accurately, the

17

02 - Luecken - 45370.qxd

18

9/12/2007

4:45 PM

Page 18

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

actual events these questionnaire responses are presumed to reflect—cause the respon­ dent’s standing on the construct of interest. Bollen and Lennox (1991) argue that socioe­ conomic status (SES) is another example of this sort of measurement model, where little or no association among indicators is expected. There are many specific indicators of SES— household income, years of education, grade of employment. Importantly, variation in these indicators is not caused by the individ­ ual’s relative standing on the hypothetical construct of SES. Instead, the individual’s standing on these indicators is the cause of their position on the latent, unobserved con­ struct of SES. Significant correlations among these indicators do not challenge the underly­ ing measurement model, but neither would the absence of such associations. Obviously, when variation in measured indicators is pre­ sumed to be caused by the latent, hypotheti­ cal, or unobserved construct, the absence of such associations could constitute a strong challenge to the measurement model. Hence, the answer to West and Finch’s first question depends in part on whether one construes measured indicators as effects or causes of the unmeasured hypothetical constructs they are hypothesized to represent. The answer to this simple question might not always be obvious, even in the case of relatively simple physiological assessments. For example, the metabolic syndrome is a set of separately assessed cardiovascular risk factors, including insulin insensitivity, abdominal fat, elevated lipids, and high blood pressure (see Raikkonen et al., this volume). Metabolic syn­ drome could be construed as a simple statisti­ cal summary of otherwise unrelated risk factors. In this case it forms a “causal indica­ tor” model, which implies no association among indicators (but such associations would not necessarily contradict it, either). If these risk factors were hypothesized to form a causally interconnected physiological state, then an effect indicator model is implied and

significant associations among indicators are expected. An empirical evaluation of the ade­ quacy of measures of the metabolic syndrome would first require an answer to this concep­ tual question regarding the expected pattern of association. Interestingly, related research sup­ ports the effect indicator view of the metabolic syndrome (Shen et al., 2003). The second question in the West and Finch (1997) framework is, “What is the structure of the construct?” In many applications, multiple indicators are presumed to reflect a unidimensional construct. In others, the con­ struct is hypothesized to consist of two or more lower-order dimensions or facets. For example, the construct of cardiovascular reactivity (see Goyal et al., this volume) is often defined as a unidimensional construct reflecting individual differences in the magni­ tude of heart rate and/or blood pressure reactions to psychological stressors. Other conceptual models of this construct suggest that it can be further subdivided into individ­ ual differences in cardiac reactivity and vascu­ lar reactivity. Cardiac reactivity is presumed to reflect increases in the rate and force of myocardial contractions, whereas vascular reactivity is presumed to reflect increases in total peripheral resistance. The level and pat­ tern of intercorrelation expected for multiple indicators of cardiovascular reactivity would depend on which dimensional model was hypothesized. Importantly, these competing models can be pitted against one another empirically, and such studies have supported the two-factor model (Kamarck, Jennings, Pogue-Geile, & Manuck, 1994). The concept of cardiovascular reactivity provides another example of the importance of clear conceptualization about structure as a guide to measurement. This term actually has two uses. The first was described earlier— an individual difference construct (albeit per­ haps having distinct cardiac and vascular components) that is stable across time and sit­ uations, hypothesized to be associated with

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 19

Measuring Physiological Processes in Biopsychosocial Research

risk for future cardiovascular disease. The second use of this construct also involves risk of cardiovascular disease. However, it is not presumed to reflect a trait that is stable across time and situations; instead, it is descri­ bed conceptually as a mediating variable that links psychosocial risk factors (e.g., low social support, high levels of trait anger) with risk of cardiovascular disease. In this conceptualiza­ tion, cardiovascular reactivity is seen as a situation-specific response that is more or less pronounced as a function of particular com­ binations of personality traits and situations, as when people who are characteristically eas­ ily provoked to anger encounter harassment or frustration in their interactions with others. In that situation, trait angry persons would be expected to display high cardiovas­ cular reactivity, and less trait angry people would be expected to display smaller responses. Importantly, individual differences in trait anger would not be expected to be related to cardiovascular reactivity during situations that did not involve relevant interpersonal stres­ sors. Hence, the patterns of temporal and sit­ uation consistency in cardiovascular reactivity is somewhat more complex than in the first conceptualization of the term. Multiple occa­ sions of measurement would increase relia­ bility of assessment, but only within the situations identified as relevant by the related conceptual model. Rather than a pansitua­ tional aggregation of responses, this construct would be more accurately assessed through a situationally specific model of individual dif­ ferences that take the form of “if-then” propo­ sitions (e.g., Mischel & Shoda, 1995). If the individual encounters a specific situation, then a specific response is expected. Obviously, the optimal sampling scheme for cardiovascular reactivity would differ depending on whether one was interested in measuring the general, broadly stable trait or the situationally specific mediating response. Similarly, the empirical evaluation of the measure would vary across these two conceptual uses, as well.

The third question in the West and Finch (1997) framework is, “What is the stability of the construct?” Some physiological constructs in health psychology are hypothesized to be quite stable (e.g., individual differences in car­ diovascular reactivity, chronic inflammation). Others are quite changeable, such as any of a wide variety of acute physiological reactions. Importantly, many specific measures can be used to capture either stable or rapidly changing physiological constructs, such as sleep quality or pain. For example, resting levels of respira­ tory sinus arrhythmia (RSA; see Thayer & colleagues, this volume) are often seen as reflect­ ing a relatively stable individual difference in vagal tone—a psychobiological trait linked to self-regulatory capacity. Momentary changes in RSA, in contrast, have been linked to selfregulatory effort (Appelhans & Leuken, 2006; Butler, Wilhelm, & Gross, 2006; Ingjaldsson, Laberg, & Thayer, 2003; Segerstrom & Nes, 2007). Hence, stable and changing values of the same physiological parameter can reflect quite distinct constructs; these constructs would imply very different assumptions about the rel­ evance of temporal stability as part of the eval­ uation of the related measure. The fourth question posed by West and Finch (1997) is the most complex; “What is the pattern of relationships of a measure of the construct of interest with other measures of the same construct and with measures of other constructs?” Answering this question entails articulating the nomological net (Cronbach & Meehl, 1955) in which the con­ struct is embedded. As described previously, a nomological net consists of rules or hypothe­ ses specifying the association of (a) observ­ able properties of constructs with one another (i.e., associations among research operations), (b) constructs with observable research oper­ ations, and (c) constructs with each other. Hence, these interlocking specifications reflect not only issues of measurement but also substantive theory. In this way, the eval­ uation of measures often also involves tests of

19

02 - Luecken - 45370.qxd

20

9/12/2007

4:45 PM

Page 20

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

substantive conceptual questions, and tests of substantive questions can be a source of valuable evidence regarding the adequacy of a given measurement procedure. Before embarking on the selection, design, or evaluation of a given physiological measurement procedure (or any measure­ ment procedure for that matter), the multiple levels of McFall’s (2005) analysis should be carefully articulated, and the West and Finch questions answered. This conceptual work can then guide highly informative theorydriven evaluations of the physiological mea­ sure in question.

Reliability, Validity, and Utility The reproducibility or reliability of a mea­ sure reflects the relative levels of systematic variance and unsystematic variance. Reliability of any measurement is an essential considera­ tion, in part because it sets an upper limit on the extent to which it can reflect the construct of interest (i.e., validity). Yet, which of many forms of reliability are important in the eval­ uation of a given measure depends entirely on conceptual issues, as described previously. The relevance of internal consistency depends on whether the measure is a cause or effect indi­ cator model (Bollen & Lennox, 1991), for example, and the relevance of stability or testretest reliability of the measure depends on the conceptual description of the construct’s temporal qualities. It is safe to say that for virtually any use of a physiological measure, some aspect of reliability is relevant. Minimally, it is virtually always assumed that two mea­ surements of a given physiological parameter taken at the same time through the same pro­ cedure (e.g., two measurements of fasting cholesterol taken from a single blood draw) would correlate quite closely, but this must be demonstrated rather than assumed. When evaluating a given physiological measure, each conceptually relevant aspect of reliability should be identified and tested. It is also important to

consider the hierarchical relationships among some forms of reliability. For example, if one considers cortisol reactivity to social stress as a stable individual difference variable, the reli­ ability of a given assay of that response (i.e., the extent to which two measurements of a sample obtained at one point in time are cor­ related) sets an upper limit on the temporal stability or test-retest reliability of the mea­ sure of individual differences in cortisol reac­ tivity. To the extent that the supposedly identical measurements produce differing val­ ues, the maximum value for test-retest relia­ bility is reduced. In testing theories about the association of a physiological construct with a health out­ come, it is important to consider the reliabil­ ity of the physiological measure as a potential influence on the observed presence (i.e., statis­ tical significance) and magnitude (i.e., effect size) of the predicted covariation. For example, when testing hypotheses regarding the association of the previously described broad individual difference in cardiovascular reactivity measured in the laboratory with cardiovascular activity assessed elsewhere or with subsequent cardiovascular disease, low reliability of the measure of cardiovascular reactivity could produce inaccurate evidence of null results or underestimates of effect size (see Goyal et al., this volume). The aggrega­ tion of multiple measurements could provide more meaningful tests of this hypothesis, by increasing the reliability of the individual dif­ ference assessment (Kamarck, Debski, & Manuck, 2000; Kamarck & Lovallo, 2003). Similar to the traditional psychometric issue in which the use of a larger number of intercorrelated scale items can improve overall scale reliability, the averaging of multiple measures of reactivity can reduce the pro­ portion of unsystematic variance in the final score. The error associated with each individ­ ual score is random by definition, and therefore averaging increases the proportion of systematic variance relative to error. The

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 21

Measuring Physiological Processes in Biopsychosocial Research

beneficial effect on reliability of averaging multiple measurements is generally true, but takes on added importance when change in a physiological variable is quantified. This is because change is defined as the difference between two imperfect scores and therefore inherently includes a larger proportion of error than does a single value. It is also important to consider the poten­ tially negative effect on reliability of a physio­ logical index when it is derived through the consideration of multiple individual physio­ logical measures. For example, in measuring aspects of cardiovascular response, cardiac output (CO) is calculated as the product of stroke volume (SV) and heart rate (HR). Total peripheral resistance (TPR) is derived from CO and mean arterial pressure (MAP), as pressure is the product of output and resis­ tance (see Thayer & colleagues, this volume). Any sample of CO will be less reliable than its two components considered separately, as HR and SV are themselves not measured with perfect reliability, and hence CO multiplies two values that contain unreliability. TPR will have still less reliability, because the addi­ tional determinant of MAP will also be mea­ sured with less than perfect reliability. Hence, TPR multiplicatively combines three sources of unreliability. The likelihood of detecting statistically significant covariation is reduced as reliability of the measure(s) under consid­ eration decreases, and estimates of effect size will be similarly reduced. Hence, interpreta­ tion of patterns of significant and nonsignifi­ cant effects or the relative magnitude of effects should always take into account the reliability of the specific measures compared as a potential explanation, even when they are based on a common set of component measures. In the present example, weaker effects on CO or TPR than on HR, SV, or MAP could simply reflect the reduced reliabil­ ity resulting from combining measures as opposed to substantive differences between effects on these variables.

Systematic variance in a measure can reflect the construct of interest, a second unintended construct, or a combination of intended and unintended constructs. Loosely speaking, validity refers to the extent to which observed variation in a measure reflects the construct of interest. The limiting effect of reliability on validity is easily seen from this perspective; the lower the reliability of a measure, the smaller the proportion of its variability that is avail­ able to potentially reflect the construct of interest. Yet, high reliability (i.e., high propor­ tion of systematic variability) is no guarantee of validity, as the systematic variance could reflect something other than the intended con­ struct. The reliability of physiological mea­ surements is often easily estimated and often quite high. As is true for any type of measure, the validity of physiological measures is harder to evaluate and often less satisfactory. Most of the physiological variables described in this volume have many potential influences, and as a result, determining which constructs are involved in an association involving the physiological measure is often a difficult chal­ lenge. The seemingly more “direct” nature of physiological measures is no guarantee of unambiguous interpretations. At a simple level, the issue of validity involves the ques­ tion, “What is the evidence that this measure reflects the construct it is intended to cap­ ture?” or “What other constructs—different from the intended use—could this measure reflect?” Articulation of alternative interpreta­ tions and critical or empirical evaluation of their relative viability is the engine of research in general; it is also the central process in eval­ uating a critical component of research—the validity of measurement. For example, many of the measures dis­ cussed in this volume are used to capture health-relevant effects of stress or emotion. Yet, they can also be affected by constructirrelevant artifacts, such as physical movement. If levels of an experimental manipulation of stress or emotion also differ in the level of

21

02 - Luecken - 45370.qxd

22

9/12/2007

4:45 PM

Page 22

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

movement artifacts, this second source of systematic variance in the physiologicaldependent measure can obscure the interpreta­ tion of results. In testing the effects of anger on cardiovascular reactivity, one might compare blood pressure responses during a task involv­ ing the recall of a past anger-arousing event to one involving descriptions of a typical daily schedule. If the anger recall task also evokes more rapid or louder speaking than the com­ parison task, it would be difficult to assert that differences in blood pressure response reflect the cardiovascular effects of anger rather than the metabolic demands of more vigorous speech. Hence, experimental control over construct-irrelevant influences on physiological measures in short-term laboratory studies of stress or related physiological processes is an important concern. Yet, the precision of exper­ imental control can also undermine the realism or ecological validity of the research, creating an important but difficult trade-off or chal­ lenge in the design of physiological research. In many cases of both laboratory and ambulatory studies of physiological processes, these artifacts are frequently measured and controlled statistically rather than experimen­ tally. Hence, the reliability and validity of the assessment of the artifact (e.g., posture, muscu­ lar movement, speech) become key concerns; limitations in either reliability or validity of measured control variables can result in their undercorrection and therefore continuing via­ bility as a threat to the validity of interpreta­ tions of the physiological measure. It is also important to recall that, as is the case for all measures, validity refers to the accuracy of a specific inference based on a physiological measure in a specific context; it is not a property of the measure itself. In this sense, it is not accurate to speak of valid or invalid measures. Rather, specific infer­ ences based on a measure vary in validity. To the extent that a specific use of the measure differs from the context in which

previous evidence of validity was acquired, it is inappropriate to claim the prior evidence in support for the current use. Differences in age, gender, ethnicity, and a host of other aspects of research samples or settings could alter the extent to which a given measure reflects the intended construct. Many of the most important health outcomes of interest in the field differ as a function of race or eth­ nicity, and the underlying physiological mechanisms might vary as well. To address these important issues of diversity in health psychology (Park, Adams, & Lynch, 1998; Whitfield, Weidner, Clark, & Anderson, 2002; Yali & Revenson, 2004), one must establish empirically rather than simply assume the equivalence across ethnicity of all measures—including physiological measures. Physiological measures can be expensive, in terms of time, complexity, and demands of researcher expertise or participant burden, as well as actual costs in equipment and materi­ als. Even when selecting from various mea­ sures of a single physiological process (e.g., heart rate, blood pressure, sleep quality or duration), there are likely to be multiple options that vary widely in these aspects of expense. It is always useful to resist the default impulse to select “the best measure I can afford” and instead consider how much information will be gained or lost with vari­ ous options. This is actually two questions. First, how much of an improvement in the correspondence between the measure and the construct of interest results from selecting one measurement procedure as opposed to another? Greater information value can come from improvements in reliability or validity. And second, what is the difference in expense—in all its various forms—associated with the more informative measure? The most informative measure might not always be worthwhile, given the degree of informa­ tion gained relative to its cost.

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 23

Measuring Physiological Processes in Biopsychosocial Research

GENERAL ISSUES The principles of measurement described in the preceding sections are the basis of sound and productive use of physiological measure­ ment in all areas of health psychology. Yet, their application in this context includes some unique considerations, as do recently emerging issues in the field.

Physiological Measurement in the Three Domains of Health Psychology For the purposes of discussion, the field of health psychology/behavioral medicine can be roughly divided into three general topics— stress and disease, or psychosomatics; health behavior and prevention; and psychosocial aspects of medical illness and care (Smith, 2003, 2007). In the first, psychosocial factors are examined as potential influences on the development and course of disease, primarily through the general psychobiologic mecha­ nisms of stress and emotion. This topic includes studies of the association between risk factors (e.g., negative emotions, stressful environmental circumstances or events, social relationship factors) and health out­ comes, but also the psychobiologic mecha­ nisms underlying these associations (e.g., psychoneuroimmunology). Many of the measurements described in this volume are applied in this area of research. In the second topic, research attempts to (a) identify behavioral risk factors for impor­ tant health outcomes (e.g., smoking, inactiv­ ity, diets high in fat and calories, excessive alcohol use), (b) delineate the determinants of these health behaviors, and (c) develop and evaluate interventions to reduce risky behav­ ior and thereby prevent negative health out­ comes. The third topic includes studies of the psychosocial effects of disease (e.g., coping with acute and chronic illness), psychosocial aspects of medical care (e.g., patient physician

interaction, adherence to medical regimens, seeking medical care), and the value of psy­ chosocial interventions as adjuncts to tradi­ tional medical or surgical interventions (e.g., stress reduction therapies for chronic disease; psychological interventions for acute or chronic pain; weight loss or exercise interven­ tions for hypertension or diabetes). Clearly, these topics are not mutually exclusive, as many specific topics combine elements of one or more categories. For example, studies of lipids, adiposity, and glucose/insulin metabolism are relevant both in the psychobi­ ology of disease and in the study of health behaviors and prevention. Similarly, inflam­ matory and neuroendocrine processes are implicated both in the psychobiology of dis­ ease development and in psychological influ­ ences and effects of established disease. Physiological measurements have histori­ cally been more prominent in research in the area of psychobiology of disease, as illus­ trated by many of the chapters in this volume. Collectively, these chapters provide a valu­ able overview of this use of physiological measurement in health psychology, and demonstrate how advances in measurement are leading to progress in addressing funda­ mental questions about psychobiological influences of disease. Historically, the pri­ mary pathways in this general topic have included hormonal (see Nicholson; Mills & Ziegler; Granger et al., this volume), cardio­ vascular (see Gerin et al.; Goyal et al.; Karmarck & Janicki; Thayer & colleagues, this volume), and immune mechanisms (Jain et al.; Prather & Marsland, this volume). However, the study of these mechanisms con­ tinues to expand, more recently including abdominal adiposity, insulin and glucose activity, lipids, and inflammation (see Davis; Raikkonen et al.; Stoney, this volume). The suggestion that each of these mechanisms and their psychosocial correlates could vary as a function of genetic factors has further

23

02 - Luecken - 45370.qxd

24

9/12/2007

4:45 PM

Page 24

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

expanded the measurement agenda in this por­ tion of the field (see McCafferey, this volume). Each area involves physiological measure­ ment, however, and continuing advances in this aspect of methodology will be important to its future. In studies of health behavior and prevention, physiological measures can be used to assess the effects of behavioral inter­ ventions on predisease physiological states. For example, measures of ambulatory blood pressure levels, plasma lipids, and glucose metabolism could be measured to evaluate the effects of exercise or weight loss interven­ tions on cardiovascular risk. In other cases, biochemical methods (e.g., Glasgow et al., 1993) are used to measure smoking outcomes following behavioral interventions, because self-reports of smoking might be less than optimally accurate (Patrick et al., 1994). Similary, biomechanical measures of physical activity and related physiological measures of physical fitness (see Etnier, this volume) can be invaluable in the evaluation of interven­ tions, especially when self-reports of physical activity are seen as simply too limited in accu­ racy or definitiveness. For studies of adjunctive psychosocial treatment for medical disorders, physiological measures can be used to assess the success of treatment directly, as when measures of sleep duration and quality are used to evaluate the effects of behavioral treatments for insomnia. In some cases, evaluation of the effects of interventions on the medical outcomes of pri­ mary interest would require the use of a very large sample studied over long periods of time, given the low incidence of clinical events and their time course. In this case, physiolog­ ical measures of subclinical or intermediate medical outcomes are useful in evaluations of the potential medical benefits of psychologi­ cal interventions. Blumenthal and colleagues (2005) used this strategy in evaluating the effects of exercise and stress management interventions for coronary heart disease (CHD) patients. Measures of flow-mediated

dilation of the brachial artery and ventricular functioning were used as arterial and cardiac endpoint measures with a previously estab­ lished association with CHD risk and progno­ sis. The positive results on these intermediate physiological outcomes provide important information regarding the interventions’ potential clinical benefits. Thus, despite the fact that physiological measurements have the most obvious rele­ vance to studies of the psychobiology of dis­ ease, they are useful throughout the research agenda in the field. Despite this wide variety of contexts of use, the issues discussed previously are always relevant in physiological measure­ ment. When physiological measures are used in evaluating interventions, additional issues related to measurement are important. One benefit of most physiological measurements is that they do not employ the arbitrary metrics that are common in many instances of psychological measurement. Pounds, beats per minute, picograms per milliliter (pg/ml), mmHg, and other units of measurement in physiological research have a form of meaning that scales in psychological tests such as depression inventories simply cannot provide. The meaning of values on these physiological measurement scales have nonarbitrary exter­ nal referents, permitting inferences that are not available through arbitrary metrics com­ monly used to measure psychological con­ structs (Blanton & Jaccard, 2006). Hence, differences in treatment conditions convey information that is not as readily available when measuring other outcomes (e.g., pain, depression, functional activity). However, cal­ culating effect sizes and relating the observed magnitude of effects even on the more defini­ tive metrics to external criteria (e.g., related levels of risk associated with mean differences of a given size) are quite useful in evaluating the clinical relevance of intervention effects (Kazdin, 2006). In the era of evidence-based practice in psychology, behavioral medicine, and related fields, attention to the translation

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 25

Measuring Physiological Processes in Biopsychosocial Research

of measured outcomes to clinically relevant information is imperative. The use of nonarbi­ trary metrics available in physiological mea­ surement is an important contribution in this regard, but the additional step of linking those scale differences to independent and clinically relevant criteria is important as well.

Integration Across Systems This edited volume is necessarily divided into manageable, system-based overviews of physiological methods in health psychology. This organization also reflects views empha­ sizing links between psychosocial processes and specific physiological pathways thought to confer increased vulnerability to distinct diseases (Weiner, 1992). However, several trends emerging from the biomedical and neuroscience literatures may potentially model greater integration across physiologi­ cal systems in understanding risk for disease. Two trends relate to the role of inflammation on cardiovascular disease (see Stoney, this volume) and central nervous system (CNS) coordination of health-relevant physiological responses (see Ryan & Alexander and Jackson & Jackson, this volume). Inflammation and Disease. Traditionally, the immune system has been linked to infectious diseases and cancer (Abbas & Lichtman, 2003). Cardiovascular disease can now be added to that list of disease processes with an immunologic component (Ross, 1999). Immune processes are implicated in just about every stage of atherogenesis (Ross, 1999; Libby, 2002). At early stages of dam­ age, the endothelium begins to express adhe­ sion molecules, such as vascular adhesion molecule-1, that help in binding immune cells to the vasculature. Monocytes and Tlymphocytes are then recruited to sites of inflammation and migrate into vessel walls through various chemokines (e.g., MCP-1) released from vascular cell walls (Charo &

Taubman, 2004). Once inside vessel walls, these immune cells proliferate and release a variety of growth factors (e.g., plateletderived growth factors) and cytokines (e.g., IL-1) characteristic of the inflammatory response (Libby, 2002). The release of cytokines further up-regulates the expression of adhesion molecules while increasing the proliferation of smooth muscles cells (SMC). Macrophages expressing scavenger receptors begin to ingest lipids (e.g., oxidized low-den­ sity lipoprotein, LDL) to form the foam cell that is characteristic of the advancing lesion. Of course, inflammation can be beneficial under “normal” conditions. For instance, the inflammatory response appears to play a role in the enlargement of existing vessels and the healing of plaque following thrombosis (Charo & Taubman, 2004; Libby, 2002). However, if the source of damage is not removed, these processes can lead to a vicious cycle of inflammation as cytokines further increase LDL binding to endothelial and smooth muscle walls (Ross, 1999). The resulting advanced lesion has a fibrous cap that protrudes into the lumen and contains elements of the inflammatory response including macrophages, T-cells, foam cells, and lipids (Ross, 1999). Immune events at later stages can lead to the rupture of such plaques. For instance, macrophages are com­ mon in vulnerable plaques and can produce enzymes (e.g., metalloproteinases) that degrade the fibrous cap, while T-lymphocytes release interferon-γ, which can impede collagen for­ mation by SMC (Libby, 2002). Importantly, research is beginning to link psychosocial processes to inflammation that may impact cardiovascular disease. These links are perhaps most clearly illustrated in the context of stress-related processes. For instance, stress can directly lead to the release of inflammatory cytokines (e.g., IL-1, tumor necrosis factor alpha, or TNF-α), which are crucial mediators of the cardiovascular inflammatory response (Black & Garbutt,

25

02 - Luecken - 45370.qxd

26

9/12/2007

4:45 PM

Page 26

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

2002; Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002; Kop & Cohen, 2001). Moreover, stress-induced cytokine produc­ tion may be mediated by the release of neuroendocrine hormones (e.g., sympathetic nervous system, or SNS hormones, cate­ cholamines) that influence cytokine release and macrophage/T-cell function more gener­ ally because cells of the immune system have functional neuroendocrine receptors (Sanders, Kasprowicz, Kohm, & Swanson, 2001). In fact, stress-induced inflammatory processes are greater in individuals charac­ terized by high cardiovascular reactivity (Bosch, Berntson, Cacioppo, Dhabhar, & Marucha, 2003). It is important to note that the links between neuroendocrine and inflammatory processes are reciprocal (Uchino et al., 2007). For instance, one important control mecha­ nism is activation of the hypothalamicpituitary-adrenal (HPA) axis, which tends to suppress immunity (Munck, Guyre, & Holbrook, 1984). However, chronic stress can lead to a state of glucocorticoid resistance (Miller, Cohen, & Ritchey, 2002; Hawkley et al., in press) because proinflammatory cytokines such as TNF-α and macrophage inhibitory factor can modulate important iso­ forms of glucocorticoid recepters on cytokine­ producing cells (Hawkley et al., in press; Webster, Oakley, Jewell, & Cidlowski, 2001). All in all, these data highlight the focal role of immune-related processes in coordinating links between psychosocial factors and the leading causes of morbidity and mortality (Kiecolt-Glaser et al., 2002). This not only underscores the specific importance of immune-related processes (see Jain et al.; Prather & Marsland, this volume) in a widen­ ing set of threats to health, but also should encourage investigators to think broadly across what are often separately conceptual­ ized and discretely measured aspects of physiological response. If physiological mea­ surement is best grounded in clear conceptual

models, these models should increasingly articulate integrated, multisystem processes. CNS Coordination of Health-Relevant Biological Responses. Evidence for coordi­ nated central pathways potentially linking psychosocial factors to disease provides fur­ ther impetus for a more integrative view (Chrousos & Gold, 1992, Dunn & Berridge, 1990; Gray, 1993; Phelps, 2006). In the area of stress physiology, animal models suggest that a promising integrating mechanism may involve the activation and release of central corticotropin-releasing hormone (CRH) in emotion-based areas of the brain including the amygdala, hypothalamus, and locus coeruleus (Dunn & Berridge, 1990). Central administration of CRH mimics many of the physiological and behavioral states seen dur­ ing stress. For instance, central CRH acti­ vates both the autonomic nervous system (ANS) and HPA axis (Irwin, Hauger, Brown, & Britton, 1989) and stimulates the release of β-endorphins (Rivier, Brownstein, Speiss, Rivier, & Vale, 1982). Increased CRH in brain regions involved in emotional respon­ ses such as the amygdala, paraventricular nucleus, locus coeruleus, and cortex in turn mediate many behavioral responses to stress (Dunn & Berridge, 1990). These include increased freezing behavior (Swiegiel, Takahashi, & Kalin, 1993), decreased appet­ itive behavior (Krahn, Gosnell, Grace, & Levine, 1986), decreased sexual behavior (Sirinathsinghji, Rees, Rivier, & Vale, 1983), increased grooming behavior (Holahan, Kalin, & Kelley, 1997), and an enhanced startle response (Lee & Davis, 1997). In fact, the CRH receptor antagonist antalarmin attenuates both biological and behavioral responses during stress (Habib et al., 2000). Importantly, CRH-containing neurons and receptors are prevalent in the amygdala, hypothalamus, and locus coeruleus (Gray, 1993; Menzaghi, Heinrichs, Pichs, Weiss, & Koob, 1993; Valentino, Foote, & Page,

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 27

Measuring Physiological Processes in Biopsychosocial Research

1993). Subsequent efferent pathways provide one mechanism by which stress may be co­ ordinated to influence health. For instance, the amygdala has direct projections to the hypothalamus (Gray, 1993). Release of CRH from the hypothalamus activates the HPA axis. The hypothalamus also has efferent pro­ jections to the ANS via the sympathetic pre­ ganglionic neurons of the intermediolateral cell column, the ventral lateral medulla, and the nucleus tractus solitarus (Menzaghi et al., 1993). In combination, these pathways result in the activation of the HPA and ANS, which have been of interest to many health psycholo­ gists modeling biological pathways to disease. Although CRH activates the noradrener­ gic system, reciprocal interactions are also apparent (Chrousos & Gold, 1992; Dunn & Berridge, 1990). For instance, catecholamin­ ergic inputs are evident to hypothalamic cell regions containing CRH and appear to stim­ ulate central CRH release by means of an α1-adrenergic mechanism (Al-Damluji et al., 1987; Cunningham & Sawchenko, 1988). It is also important to emphasize that there are critical afferent (ascending) pathways to brain structures that may be activated in stressful circumstances. Vagal afference appears to play an important role in the immune response to infection, which may then have implications for stress processes (Maier & Watkins, 1998). Berntson, Sarter, and Cacioppo (2003) also reviewed evidence for the role of these afferent pathways in modulating anxiety-related physiological reactions. According to their model, these ascending pathways provide an opportunity for anxiety- or stress-induced physiological reactions to influence cortical information processing (e.g., attentional functions). For instance, activation of the locus coeruleus via afferent pathways is linked to increased elec­ troencephalographic (EEG) arousal and vigi­ lance to significant stimuli (Aston-Jones, Rajkowski, Kubiak, Valentino, & Shipley, 1996). Many of these “bottom-up” processes

appear mediated by the basal forebrain cholinergic system, which has (a) reciprocal connections with the amygdala, and (b) wide­ spread projections to the cerebral cortex (Berntson et al., 2003). As a result, it is well situated to participate in the potentiation of stress-related responses. Recent work in brain imaging (e.g., func­ tional magnetic resonance imaging [fMRI], positron emission tomography [PET]) is start­ ing to provide converging evidence to animal models for the coordinated CNS structures responsible for more peripheral physiologi­ cal activation. Of these CNS structures, the amygdala’s role in emotion-based processes is perhaps the best characterized. Phelps (2006) has argued that the distinction between cog­ nition and emotion is blurred by data on the role of the amygdala in modulating neural systems linking cognitive and social processes during emotional cues. Indeed, fMRI data across a variety of domains suggest that the amygdala may play an important role in emo­ tional learning, emotional memory, and emo­ tional influences on attention and perception (Phelps, 2006). For instance, social phobics show increased blood-oxygenation level dependent (BOLD) fMRI activity in subcorti­ cal limbic structures (e.g., amygdala, stria­ tum) in anticipation of public speaking (Lorberbaum et al., 2004). Recent research is also focusing on BOLD fMRI activation of the anterior cingulate as a potential mediator of cardiovascular reactivity during acute stress (Critchley et al., 2003; Gianaros et al., 2005). Future research, guided by neu­ roanatomic data and animal models (e.g., lesion studies), promises a more integrative view of links between central and peripheral physiological responses in health psychology research. Historically, physiological measure­ ment in health psychology has involved “everything but the brain.” Clearly, this tra­ ditional approach has rapidly become out­ dated. As a result the future of physiological measurement in health psychology should

27

02 - Luecken - 45370.qxd

28

9/12/2007

4:45 PM

Page 28

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES

incorporate the concepts and methods of brainbased measurement of cognitive, affective, and social neuroscience (see Ryan & Alexander and Jackson & Jackson, this volume).

Lifespan Perspectives:

Links to Disease Development,

Course, and Outcome

Many of the diseases of interest to health psychology researchers develop slowly over many years, and are increasingly prevalent with advancing age. Further, the physiologi­ cal mechanisms influencing their develop­ ment and course can differ across the various stages of pathophysiology in a given disease. Finally, these physiological responses them­ selves or their psychosocial determinants may also change with age. All of these issues converge to suggest that the conceptualiza­ tion and measurement of physiological pro­ cesses in health psychology research should be informed by the consideration of biologi­ cal, psychological, and social/cultural aspects of aging and development. When considering or evaluating a given physiological measure, an ancillary question at each of McFall’s (2005) levels of measurement or follow-up to each of the questions posed by West and Finch (1997) would be, “and how would this change with age?” From the selection of rel­ evant physiological constructs (and the nonselection of others) to the reliability and validity of specific measures, the interrelated issues of age, development, and time-linked aspects of disease etiology and course should be carefully considered.

CONCLUSION When learning to use and interpret physio­ logical measures, it is often the technical challenges of these methods and the biomed­ ical science in which they are embedded that are most salient and potentially daunting.

This volume presents an excellent introduc­ tion and guide to this terrain. Yet, it is impor­ tant to recognize that no matter what the focus and method, all measures are based on a set of assumptions. This multilayered set of conceptual issues must be systematically and critically evaluated. The principles we have outlined here are intended to serve as a primer of these considerations in physiologi­ cal measurement in health psychology. The scientific yield of any physiological measures will be enhanced if technical precision is matched by an equal amount of conceptual precision. Further, the field can be advanced not only through the thoughtful use of phys­ iological measures in the pursuit of major questions regarding connections between behavioral and biological processes, but also through the empirical study of the physio­ logical measures themselves. Investment in the measurement infrastructure of health psychology plays a central role in the field’s future progress (Smith, 2007). Regrettably, it is not uncommon for health psychology researchers to go to great lengths in time, expense, thoughtful deliberation, expertise, and participant burden in measur­ ing a physiological variable, and then exam­ ine its association with an inadequately conceptualized and measured psychologi­ cal construct. Sophisticated neuroendocrine assays, functional immune measures, or images of the ischemic myocardium might be correlated with a vaguely defined and hastily measured personality trait. A quickly devel­ oped set of self-report items might be bundled without adequate evaluation of the extent to which their pattern of intercorrelation corre­ sponds to the hypothesized structure, whether the resulting scores converge with indepen­ dent measures of the same construct and diverge from measures of conceptually dis­ tinct traits, or if the conceptualization of the construct is such that self-reports could be logically expected to be valid. Inferences from research operations to conceptual questions

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 29

Measuring Physiological Processes in Biopsychosocial Research

of paramount interest in the study of mind and body involve at least two measured con­ structs, and no level of sophistication and definitiveness in physiological measurement can overcome weakness on the “other side” of a pattern of observed covariation. Fortu­ nately, the principles of measurement described previously are easily applied to all measurement

activities in health psychology (Smith, 2007). Researchers need only be equally attentive to both the physiological and nonphysiological sides of their hypotheses as they attempt to ground fascinating conceptual questions of mind and body in the more mundane but equally important specifics of research operations.

REFERENCES Abbas, A.K., & Lichtman, A.H. (2003). Cellular and molecular immunology. Philadelphia: Saunders. Al-Damluji, S., Perry, L., Tomlin, S., Bouloux, P., Grossman, A., Rees, L. H., & Besser, G. M. (1987).Alpha-adrenergic stimulation of corticotrophin secretion by a specific central mechanism in man. Neuroendocrinology, 45, 68–76. Appelhans, B.M., & Leuken, L.J. (2006). Heart rate variability as an index of reg­ ulated emotional responding. Review of General Psychology, 10, 229–240. Aston-Jones, G., Rajkowski, J., Kubiak, P., Valentino, R.J., & Shipley, M.T. (1996). Role of the locus coeruleus in emotional activation. In G. Holstege, R Bandler, C.B. Saper (Eds.), Progress in brain research (pp. 379–402). New York: Elsevier. Berntson, G. G., Sarter, M., & Cacioppo, J. T. (2003). Ascending visceral regula­ tion of cortical affective information processing. European Journal of Neuroscience, 18, 2103–2109. Black, P. H., & Garbutt, L. D. (2002). Stress, inflammation and cardiovascular dis­ ease. Journal of Psychosomatic Research, 52, 1–23. Blanton, H., & Jaccard, J. (2006). Arbitrary metrics in psychology. American Psychologist, 61, 27–41. Blumenthal, J.A., Sherwood, A., Babyak, M.A., Watkins, L.L., Waugh, R., Georiades, A., et al. (2005). Effects of exercise and stress management training on markers of cardiovascular risk in patients with ischemic heart disease. JAMA, 293, 1626–1634. Bollen, K. A., & Lennox, R. (1991). Conventional wisdom on measurement: A structural equation perspective. Psychological Bulletin, 110, 305–314. Bosch, J. A., Berntson, G. G., Cacioppo, J. T., Dhabhar, F. S., & Marucha, P. T. (2003). Acute stress evokes selective mobilization of T cells that differ in chemokine receptor expression: A potential pathway linking immunologic reac­ tivity to cardiovascular disease. Brain, Behavior, and Immunity, 17, 251–259. Bridgman, P. W. (1927). The logic of modern physics. New York: Macmillan. Bridgman, P. W. (1945). Some general principles of operational analysis. Psychological Review, 52, 246–249. Butler, E. A., Wilhelm, F. H., & Gross, J. J.(2006). Respiratory sinus arrhythmia, emotion, and emotion regulation during social interaction. Psychophysiology, 43, 1–11. Charo, I. F., & Taubman, M. B. (2004). Chemokines in the pathogenesis of vascu­ lar disease. Circulation Research, 95, 858–866.

29

02 - Luecken - 45370.qxd

30

9/12/2007

4:45 PM

Page 30

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES Chrousos, G. P., & Gold, P. W. (1992). The concepts of stress and stress system disorders: Overview of physical and behavioral homeostasis. JAMA, 267, 1244–1252. Critchley, H. D., Mathias, C. J., Josephs, O., O’Doherty, J., Zanini, S., Dewar, B.-K., et al. (2003). Human cingulate cortex and autonomic control: Converging neu­ roimaging and clinical evidence. Brain, 126, 2139–2152. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Cunningham, Jr., E. T., & Sawchenko, P. E. (1988). Anatomical specificity of nora­ drenergic inputs to the paraventricular and supraoptic nuclei of the rat hypothalamus. Journal of Comparative Neurology, 274, 60–76. Dawes, R. M., & Smith, T. L. (1985). Attitude and opinion measurement. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (3rd ed., Vol. 1, pp. 509–566). New York: Random House. Dunn, A. J., & Berridge, C. W. (1990). Physiological and behavioral responses to corticotropin-releasing factor administration: Is CRF a mediator of anxiety or stress responses? Brain Research Reviews, 15, 71–100. Engel, G. L. (1977). The need for a new medical model: A challenge for biomedicine. Science, 196, 129–136. Epel, E. S., Blackburn, E. H., Lin, J., Dhabhar, F. S., Adler, N. E., Morrow, J. D., & Cawthon, R. M. (2004). Accelerated telomere shortening in response to life stress. PNAS, 101, 17312–17315. Gianaros, P. J., Derbyshire, S. W. G., May, J. C., Siegle, G. J., Gamalo, M. A., & Jennings, J. R. (2005). Anterior cingulate activity correlates with blood pres­ sure during stress. Psychophysiology, 42, 627–635. Glasgow, R. E., Mullooly, J. P., Vogt, T. M., Stevens, V. J., Lichetenstein, E., Hollis, J. F., et al. (1993). Biochemical validation of smoking status in public health setting: Pros, cons and data from four low-intensity intervention trials. Addictive Behaviors, 18, 511–527. Gray, T. S. (1993). Amygdaloid CRF pathways: Role in autonomic, neuroen­ docrine, and behavioral responses to stress. Annals of the New York Academy of Sciences, 697, 53–60. Habib, K. E., Weld, K. P., Rice, K. C., Pushkas, J., Champoux, M., Listwak, S., et al. (2000, May). Oral administration of a corticotropin-releasing hormone receptor antagonist significantly attenuates behavioral, neuroendocrine, and autonomic responses to stress in primates. PNAS, 97, 6079–6084. Hawkley, L. C., Bosch, J. A., Engeland, C. G., Marucha, P. T., & Cacioppo, J. T. (in press). Loneliness, dysphoria, stress and immunity: A role for cytokines. In N. P. Plotnikoff, R. E. Faith, & A. J. Murgo (Eds.), Cytokines: Stress and immunity (2nd ed.). Boca Raton, LA: CRC Press. Holahan, M.R., Kalin, N.H., & Kelley, A.E. (1997). Microinfusion of corticotropin-releasing factor into the nucleus accumbens shell results in increased behavioral arousal and oral motor activity. Psychopharmacology, 130, 189–196. Holmes, T. H., & Rahe, R. H. (1967). The social readjustment rating scale. Journal of Psychosomatic Research, 14, 213–218. Idler, E. L., & Benyamini, Y. (1997). Self-rated health and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38, 21–37.

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 31

Measuring Physiological Processes in Biopsychosocial Research Ingjaldsson, J. T., Laberg, J. C., & Thayer, J. F. (2003). Reduced heart rate variability in chronic alcohol abuse: Relationship with negative mood, chronic thought suppression, and compulsive drinking. Biological Psychiatry, 54, 1427–1436. Irwin, M., Hauger, R. L., Brown, M., & Britton, K. T. (1989). CRF activates auto­ nomic nervous system and reduces natural killer cytotoxicity. American Journal of Physiology, 255, R744–R747. Judd, C. M., & McClelland, G. H. (1998). Measurement. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (Vol. 2, pp. 180–232). Boston: McGraw-Hill. Kamarck, T. W., Debski, T. T., & Manuck, S. B. (2000). Enhancing the laboratory­ to-life generalizability of cardiovascular reactivity using multiple occasions of measurement. Psychophysiology, 37, 533–542. Kamarck, T. W., Jennings, J. J., Pogue-Geile, M., & Manuck, S. B. (1994). A mul­ tidimensional measurement model for cardiovascular reactivity: Stability and cross-validation in two adult samples. Health Psychology, 13, 471–478. Kamarck, T. W., & Lovallo, W. R. (2003). Cardiovascular reactivity to psycholog­ ical challenge: Conceptual and measurement considerations. Psychosomatic Medicine, 65, 9–21. Kazdin, A. E. (2006). Arbitrary metrics: Implications for identifying evidence-based treatments. American Psychologist, 61, 42–49. Kiecolt-Glaser, J. K., McGuire, L., Robles, T. F., & Glaser, R. (2002). Emotions, morbidity, and mortality: New perspectives from psychoneuroimmunology. Annual Review of Psychology, 53, 83–107. Kop, W. J., & Cohen, N. (2001). Psychological risk factors and immune system involvement in cardiovascular disease. In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneuroimmunology (Vol. 2, 3rd ed., pp. 525–544). New York: Academic Press. Krahn, D. D., Gosnell, B. A., Grace, M., & Levine, A. S. (1986). CRF antagonist partially reverses CRF- and stress-induced effects on feeding. Brain Research Bulletin, 17, 285–289. Lee, Y., & Davis, M. (1997). Role of the hippocampus, the bed nucleus of the stria terminalis, and the amygdala in the excitatory effect of corticotropin-releasing hormone on the acoustic startle reflex. Journal of Neuroscience, 17, 6434–6446. Libby, P. (2002). Inflammation in atherosclerosis. Nature, 420, 868–874. Lorberbaum, J. P., Kose, S., Johnson, M. R., Arana, G. W., Sullivan, L. K., Hamner, M. B., et al. (2004, December 22). Neural correlates of speech anticipatory anxiety in generalized social phobia. Neuroreport, 15, 2701–2705. Maier, S. F., & Watkins, L. R. (1998). Cytokines for psychologists: Implications of bidirectional immune-to-brain communication for understanding behavior, mood, and cognition. Psychological Review, 105, 83–107. McFall, R. M. (2005). Theory and utility—key themes in evidence-based assessment: Comment on the special section. Psychological Assessment, 17, 312–323. Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46, 806–834. Menzaghi, F., Heinrichs, S. C., Pich, E. M., Weiss, F., & Koob, G. F. (1993). The role of limbic and hypothalamic corticotropin-releasing factor in behavioral

31

02 - Luecken - 45370.qxd

32

9/12/2007

4:45 PM

Page 32

PART I: INTRODUCTION TO MEASUREMENT OF PHYSIOLOGICAL PROCESSES responses to stress. Annals of the New York Academy of Sciences, 697, 142–154. Miller, G. E., Cohen, S., & Ritchey, A. K. (2002). Chronic psychological stress and the regulation of pro-inflammatory cytokines: A glucocorticoid-resistance model. Health Psychology, 21, 531–541. Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personal­ ity: Reconceptualizing situations, dispositions, dynamics, and invariance in per­ sonality structure. Psychological Review, 102, 246–268. Munck, A., Guyre, P. M., & Holbrook, N. J. (1984). Physiological functions of glu­ cocorticoids in stress and their relation to pharmacological actions. Endocrine Reviews, 5, 25–44. Orts, K., Sheridan, J. F., Robinson-Whelen, S., Glaser, R., Malarkey, W. B., & Kiecolt-Glaser, J. K. (1995). The reliability and validity of a structured inter­ view for the assessment of infectious illness. Journal of Behavioral Medicine, 18, 517–530. Park, T. L., Adams, S. G., & Lynch, J. (1998). Sociodemographic factors in health psychology research: 12 years in review. Health Psychology, 17, 381–383. Patrick, D. L., Cheadle, A., Thompson, D. C., Diebr, P., Koepsell, T., & Kinne, S. (1994). The validity of self-reported smoking: A review and meta-analysis. American Journal of Public Health, 84, 1086–1093. Phelps, E. A. (2006). Emotion and cognition: Insights from studies of the human amygdala. Annual Review of Psychology, 57, 27–53. Rivier, C., Brownstein, M., Speiss, J., Rivier, J., & Vale, W. (1982). In vivo corti­ cotropin-releasing factor-induced secretion of adrenocorticotropin, β-endorphin, and corticosterone. Endocrinology, 110, 272–278. Ross, R. (1999, January). Atherosclerosis—an inflammatory disease. New England Journal of Medicine, 340, 115–126. Sanders, V. M., Kasprowicz, D. J., Kohm, A. P., & Swanson, M. A. (2001). Neurotransmitter receptors on lymphocytes and other lymphoid cells. In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneuroimmunology (Vol. 1, 3rd ed., pp. 161–196). New York: Academic Press. Segerstrom, S. C., & Nes, L. S. (2007). Heart rate variability reflects self-regulatory strength, effort, and fatigue. Psychological Science, 18, 275–281.. Shen, B., Todaro, J. F., Niaura, R., McCaffery, J., Zhang, J., Spiro, A., & Ward, K. D. (2003). Are metabolic risk factors one syndrome? Modeling the structure of the metabolic syndrome X. American Journal of Epidemiology, 157, 701–711. Sirinathsinghji, D. J. S., Rees, L. H., Rivier, J., & Vale, W. (1983). Corticotropinreleasing factor is a potent inhibitor of sexual receptivity in the female rat. Nature, 305, 232–235. Smith, T.W. (2003). Health psychology. In I.B. Weiner (Series Ed.), J.A. Schinka & W. Velicer (Vol. Eds.), Comprehensive handbook of psychology (Vol. 2): Research methods (pp. 241–270). Hoboken, NJ: John Wiley & Sons. Smith, T. W. (2007). Measurement in health psychology research. In R. Silver & H. S. Friedman (Eds.), Foundations of health psychology, (pp 19–51). New York: Oxford University Press. Stevens, S. S. (1951). Mathematics, measurement, and psychophysics. In S. S. Stevens (Ed.), Handbook of experimental psychology (pp. 1–49). New York: Wiley.

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 33

Measuring Physiological Processes in Biopsychosocial Research Stevens, S. S. (1959). Measurement, psychophysics, and utility. In C. W. Christensen & P. Ratoosh (Eds.), Measurement: Definitions and theories (pp. 18–63). New York: Wiley. Stevens, S. S. (1968). Measurement, statistics, and the schemapiric view. Science, 161, 849–856. Swiergiel, A. H., Takahashi, L. K., & Kalin, N. H. (1993). Attenuation of stressinduced behavior by antagonism of corticotropin-releasing factor receptors in the central amygdala in the rat. Brain Research, 623, 229–234. Uchino, B. N., Smith, T. W., Holt-Lunstad, J. L., Campo, R., & Reblin, M. (2007). Stress and illness. In J. Cacioppo, L. Tassinary, & G. Berntson (Eds.), Handbook of psychophysiology (3rd ed.; pp. 608–732). New York: Cambridge University Press. Valentino, R. J., Foote, S. L., & Page, M. E. (1993). The locus coeruleus as a site for integrating corticotropin-releasing factor and noradrenergic mediation of stress responses. Annals of the New York Academy of Sciences, 697, 173–188. Webster, J. C., Oakley, R. H., Jewell, C. M., & Cidlowski, J. A. (2001, June). Proinflammatory cytokines regulate human glucocorticoid receptor gene expression and lead to the accumulation of the dominant negative B isoform: A mechanism for the generation of glucocorticoid resistance. PNAS, 98, 6865–6870. Weiner, H. (1992). Specificity and specification: Two continuing problems in psy­ chosomatic research. Psychosomatic Medicine, 54, 567–568. West, S. G., & Finch, J. F. (1997). Personality measurement: Reliability and valid­ ity issues. In R. Hogan, J. Johnson, & S. Briggs (Eds.), Handbook of personal­ ity psychology (pp. 143–164). Dallas: Academic Press. Whitfield, K. E., Weidner, G., Clark, R., & Anderson, N. B. (2002). Sociodemographic diversity and behavioral medicine. Journal of Consulting and Clinical Psychology, 70, 463–481. Yali, A. M., & Revenson, T. A. (2004). How changes in population demographics will impact health psychology: Incorporating a broader notion of cultural com­ petence into the field. Health Psychology, 23, 147–155.

33

02 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 34

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 35

Part II

PHYSIOLOGICAL

SYSTEMS AND

ASSESSMENTS

A. Hormonal 3. Measurement of Cortisol 4. Sympathetic Hormones in Health Psychology Research 5. Salivary α-Amylase in Biobehavioral Research B. Cardiovascular 6. The Measurement of Blood Pressure in Cardiovascular Research 7. Cardiovascular Stress Reactivity 8. Ambulatory Blood Pressure Monitoring 9. Noninvasive Assessment of Autonomic Influences on the Heart: Impedance Cardiography and Heart Rate Variability C. Immune 10. Laboratory-Based Measures of Immune Parameters and Function 11. Immunological Functioning II: Field Measures and Viral Challenge

35

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 36

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 37

CHAPTER

3

Measurement of Cortisol NANCY A. NICOLSON

INTRODUCTION TO THE HYPOTHALAMIC-PITUITARY­ ADRENOCORTICAL AXIS The hypothalamic-pituitary-adrenocortical (HPA) axis and its end product, cortisol, are thought to be important mediators of the relationship between stressful life experi­ ences and health outcomes. The HPA response is a component of the organism’s adaptive system for maintaining function under changing environmental circum­ stances. Over the long term, however, chronic overactivation following repeated stressors can give rise to wear and tear or allostatic load (McEwen, 2003). Both mal­ adaptive responses to stress and distur­ bances in the functioning of the HPA axis have been implicated in a wide variety of syndromes and illnesses, including cardio­ vascular illness, insulin resistance syndrome and diabetes, cognitive decline during aging, fatigue and pain syndromes, and psychiatric disorders such as depression and posttrau­ matic stress disorder (PTSD), among others (Charmandari, Tsigos, & Chrousos, 2005). As the name indicates, the main compo­ nents of the HPA axis are the hypothalamus, the pituitary gland, and the adrenal cortex

(see Figure 3.1). The hypothalamus releases corticotropin-releasing hormone (CRH, also known as CRF) into the portal blood vessels connecting the hypothalamus to the anterior pituitary. CRH, which works synergisti­ cally with arginine vasopressin (AVP) released from the hypothalamus, then trig­ gers the pituitary to secrete adrenocorticotropic hormone (ACTH) into the bloodstream. After reaching the adrenal cortex, ACTH stimulates the release of glucocorticoids (GCs)—in humans, cortisol. This entire pro­ cess takes place within a matter of minutes. The HPA axis is regulated by a complex neg­ ative feedback system, with circulating gluco­ corticoids inhibiting activity at the level of the hippocampus, the hypothalamus, and the pituitary. In general, hippocampal structures exert inhibitory influences on the axis at the level of the hypothalamus, whereas the amygdala plays an activating role (Herman & Cullinan, 1997). Mineralocorticoid (MR) and glucocorticoid (GR) receptors in the brain are thought to play different but com­ plementary roles in regulating normal circa­ dian activity, preparing the organism to respond to external stimuli, and facilitating recovery of disturbed homeostasis after acutely stressful situations (de Kloet, 1991).

37

03 - Luecken - 45370.qxd

38

9/12/2007

4:45 PM

Page 38

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL



Hippocampus −



Hypothalamus

↑CRH −

↑AVP

Pituitary ↑ACTH Adrenal cortex

↑Cortisol

Target organs/metabolic effects

Figure 3.1

Schematic Overview of the Hypothalamic-Pituitary-Adrenocortical (HPA) Axis

NOTE: CRH = corticotropin-releasing hormone, AVP = arginine vasopressin, ACTH = adrenocorticotropic hormone. Dashed lines indicate negative feedback effects.

Activity of the HPA axis shows a pronounced circadian rhythm, controlled by the primary endogenous pacemaker, the suprachiasmatic nucleus. ACTH and cortisol are secreted in short pulsatile episodes, con­ centrated in the morning hours in humans, but occurring throughout the day, even in the absence of stressors. In a 24-hour cycle, approximately 15 to 18 ACTH pulses can be identified. In people who have a normal rou­ tine of nocturnal sleep and daytime activity, cortisol levels are lowest between 10 p.m. and 4 a.m. After a quiescent period of HPA activity lasting from 2.5 to 6 hours (Linkowski et al., 1985), cortisol levels begin to rise several hours before awakening, with an additional sharp increase in the 30 to

40 minutes following awakening. Thereafter, cortisol levels steadily decrease, except for a moderate rise following lunch. Although cor­ tisol levels decline over the rest of the after­ noon and throughout the evening until sleep onset, the slope of the diurnal curve is rela­ tively flat compared to the morning hours.

A Brief Overview of Research Approaches Because of its central role in regulating the psychobiological stress response, the HPA axis is one of the most heavily investigated physiological systems in health psychology and psychiatry. Hans Selye’s conception of the general adaptation syndrome, in particular,

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 39

Measurement of Cortisol

called attention to the importance of the HPA axis in regulating a wide range of bodily func­ tions and their disturbance by acute physio­ logical stressors, such as exposure to toxins (Selye, 1956). A deeper understanding of the effects of psychological stress on the HPA axis, however, began to emerge in the 1960s, when now-classic studies employed new methods to assess endocrine responses to stress in rodents, nonhuman primates, and humans (Levine, 2000; Mason, 1968; Rose, 1984). The widespread involvement of the HPA axis in both healthy adaptation and common disorders, combined with the increasing ease with which its activity can be measured, have led to an enormous growth over the last two decades in research on this system, in settings ranging from the laboratory to the community. Research approaches include studies of spontaneous hormone secretion throughout the day, pharmacological manipulations to determine how feedback mechanisms are functioning, and studies of reactivity to acute real-life or experimental stressors. Assessment of the HPA axis at multiple levels is not feasi­ ble in most studies, because of the invasive procedures involved. Ignoring the vast litera­ ture on animal models and clinical research, this review focuses on methods that can be more generally applied by health psycho­ logists studying human subjects in a wide variety of real-life and laboratory settings, without undue inconvenience or risk to the research participants or the need for special­ ized medical personnel. This means that measures of CRH, ACTH, GC receptor char­ acteristics, or responses of the HPA axis to challenge tests in which CRH, ACTH, or other substances are administered are not covered, despite their utility in psychoneu­ roendocrine research and clinical studies. Furthermore, this chapter does not discuss the rationale or procedures for measuring dehydroepiandrosterone (DHEA), a steroid hormone produced primarily by the adrenal

cortex, although there is evidence that DHEA may counteract some of the effects of ele­ vated glucocorticoids and play a role in stressrelated disorders such as depression and chronic fatigue (Goodyer, Park, Netherton, & Herbert, 2001; Khorram, 1996; Wolkowitz, Brizendine, & Reus, 2000). This chapter focuses specifically on corti­ sol, the end product of the HPA axis. As a cautionary note, it is important to realize that cortisol is a peripheral measure and secretory patterns can be deviant in the statistical sense without necessarily reflecting dysregulation at a higher level. In some cases, apparent abnor­ malities may be the result of an adaptive response to environmental demands. On the other hand, cortisol levels can also be per­ fectly normal when other probes indicate regulatory abnormalities; excessive CRH or ACTH secretion might, for example, be cou­ pled with decreased adrenal sensitivity. The HPA axis is a complex and dynamic system, and cortisol measures can provide only a partial window into how this system is regulated—or dysregulated.

Investigating Spontaneous Activity of the HPA Axis Basal Cortisol Levels Researchers have long been interested in obtaining overall basal measures of glucocor­ ticoid output, as overactivation of the HPA axis resulting from chronic stress or illness was expected to result in higher levels of circulating cortisol. Because of the inherent novelty of hospital settings as well as the trou­ ble and expense of bringing healthy subjects to the clinic, ambulatory procedures have distinct advantages. Numerous studies have used 24-hour urinary measures or repeated salivary sampling to examine genetic, devel­ opmental, and especially environmental influ­ ences on HPA activity in healthy adults and children. Others have investigated HPA

39

03 - Luecken - 45370.qxd

40

9/12/2007

4:45 PM

Page 40

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

abnormalities in stress-related disorders. It is now clear that not only hypercortisolism but also hypocortisolism can occur, for example in PTSD, pain, and fatigue syndromes. The processes by which stress could lead to such divergent outcomes are still poorly under­ stood, but recent reviews have summarized a number of hypotheses (Fries, Hesse, Hellhammer, & Hellhammer, 2005; Gunnar & Vazquez, 2001; Heim, Ehlert, & Hellhammer, 2000; Yehuda, 2002). For example, hypocortisolism could be the longterm effect of adverse early experiences that permanently alter the axis. Downregulation might even be seen as a protective mechanism, set in motion following longterm hyperactivation to reduce the negative effects of allostatic load. Alternatively, hypocor­ tisolism might represent a preexisting risk factor, of genetic or early developmental ori­ gin, which later undermines the individual’s ability to respond adaptively to trauma or chronic stressors. Circadian Rhythm and Diurnal Patterns of HPA Axis Activity In addition to overall cortisol levels, the diurnal patterning of hormone secretion can provide important clues to HPA axis dys­ regulation. Sophisticated chronobiological analyses of circadian rhythms (see, e.g., Posener et al., 2000; Van Cauter, Leproult, & Kupfer, 1996) require more frequent sam­ pling than is feasible in ambulatory settings, not to mention the problem of obtaining noc­ turnal measures. For this reason, simpler measures of the shape of the diurnal curve are more frequently employed, in particular the steepness of the decline in cortisol levels from morning to evening. Loss of diurnal variation, as reflected in flatter slopes, has been reported in various disorders and atrisk groups (Bower et al., 2005; Sephton, Sapolsky, Kraemer, & Spiegel, 2000).

Even if the diurnal slope is not of direct relevance to the goals of a study, collecting several samples over the course of a day is good practice; differences between groups being compared may be restricted to a certain time of day, which often cannot be predicted on theoretical grounds. For this reason, studies with only a single diurnal sampling time will inevitably raise questions about how results generalize to the rest of the day. Cortisol Response to Awakening In recent years, interest has been growing in the cortisol awakening response (CAR). Cortisol levels rise sharply (50-160% in saliva) during the first 30 to 40 minutes after wakeup, returning to the awakening baseline within 60 to 75 minutes, and declining more gradually thereafter (Clow, Thorn, Evans, & Hucklebridge, 2004; Pruessner et al., 1997; Wüst et al., 2000). The function of the CAR is not yet clear, but general agreement is that this response is a discrete aspect of cortisol’s circadian rhythm, with its own regulatory processes (Clow et al., 2004; SchmidtReinwald et al., 1999). The CAR appears to be moderately stable within persons, from day to day and over longer periods of several weeks to months, and it has a clear genetic component (Wüst et al., 2000). Nevertheless, it can vary in relation to short-term influences, such as the stressfulness of a workday compared to a weekend (Kunz-Ebrecht, Kirschbaum, Marmot, & Steptoe, 2004), or an early-shift compared to a late-shift workday (Williams, Magid, & Steptoe, 2005). In addition, the CAR may be either enhanced or blunted in chronic stress, burnout, depression, and other disorders (e.g., Bhagwagar, Hafizi, & Cowen, 2005; Grossi et al., 2005; Pruessner, Hellhammer, Pruessner, & Lupien, 2003; Stetler & Miller, 2005).

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 41

Measurement of Cortisol

Within-Person Variability One important aspect of spontaneous cortisol secretion that has received relatively little research attention, despite its potential significance as an index of HPA dysregulation, is within-person variability. Greater irregular­ ity in within-day cortisol measures has been observed in affective disorders, even when overall levels are normal (Peeters, Nicolson, & Berkhof, 2004; Posener et al., 2004; Yehuda, Teicher, Trestman, Levengood, & Siever, 1996), and may predict worse clinical out­ comes (Goodyer, Tamplin, Herbert, & Altham, 2000). There is some evidence that a subset of individuals lacks a consistent diurnal slope pattern (Smyth et al., 1997), but dayto-day variation in cortisol measures remains largely unexplored. One major obstacle is that investigating within-person variability requires many more samples per person. Summary The degree of detail with which a given study is able to characterize spontaneous cortisol secretory patterns depends on its spe­ cific goals, but also on the available budget and logistical considerations. Thus, large epidemio­ logical surveys are often restricted to obtaining only a few samples per subject and may have to choose between the response to awakening and/or a diurnal slope measure (either of which can be estimated with a minimum of two saliva samples; see, e.g., Young & Breslau, 2004), perhaps in combination with a urinary mea­ sures if nighttime or total cortisol secretion are of interest. At the other extreme, intensive daily process designs may collect 60 or more saliva samples per subject in order to estimate not only overall levels and diurnal slopes, but also the association between cortisol at a par­ ticular point in time with current mood, symp­ toms, daily hassles, and uplifts (Smyth et al., 1998; van Eck, Berkhof, Nicolson, & Sulon, 1996). As we don’t yet know which measures

of spontaneous cortisol secretion are most relevant for understanding disease processes, a conservative approach would be to obtain reli­ able measures of cortisol basal levels, diurnal slopes, and the CAR in the same protocol (see Sampling Strategy under A Framework for Designing a Study and Interpreting the Results). The availability of noninvasive sam­ pling methods (described in Measuring Activity of the HPA Axis) has greatly increased the range of research applications. These include cross-cultural field studies (Flinn, 1999; Hruschka, Kohrt, & Worthman, 2005), large-scale longitudinal studies in the community (Rosmalen et al., 2005), interven­ tion studies (Carlson, Speca, Patel, & Goodey, 2004; Gaab et al., 2003), and prediction of disease outcomes (Sephton et al., 2000).

Sensitivity of the HPA Axis to Negative Feedback Measuring the response of the HPA axis to synthetic glucocorticoids provides a measure of the strength of negative feedback inhibition. Following an oral dose of 1 mg dexametha­ sone late in the evening, cortisol levels are nor­ mally suppressed the next day; incomplete suppression or early escape from suppression indicates deficits in feedback regulatory mech­ anisms. The dexamethasone suppression test (DST) was originally developed as a diag­ nostic tool in major depression, a disorder in which hypercortisolism is often observed (Carroll et al., 1981). A low-dose (0.25–0.5 mg) version of the DST has been used to inves­ tigate more subtle deficits in feedback regula­ tion in individuals with chronic stress (Powell et al., 2002) or to determine whether sensitiv­ ity of the HPA axis to glucocorticoid negative feedback is heightened in disorders in which hypocortisolism is more frequently observed, such as PTSD (Yehuda et al., 1993) or chronic fatigue syndrome (Gaab et al., 2002).

41

03 - Luecken - 45370.qxd

42

9/12/2007

4:45 PM

Page 42

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

Response of the HPA Axis to Acute Stressors Early studies of the HPA axis focused on the hormonal response to acute stressors, and this remains one of the primary interests of health psychologists. In humans, the corti­ sol response to stress can be studied in real life or under more controlled conditions in the laboratory. Compared to the quick but short-lived response of the catecholamines, the cortisol response to acute stress is rela­ tively slow. Within minutes of the onset of a discrete stressful stimulus or event, such as public speaking, cortisol levels begin to rise, superimposed on the diurnal profile of basal HPA activity. After termination of the stres­ sor, cortisol levels gradually return to their prestress baseline; full recovery can take an hour or more, in part reflecting the approxi­ mately one-hour half-life of cortisol in blood or saliva. Basal levels of glucocorticoids act permis­ sively to prepare the individual to respond to a stressful episode. The cortisol response to stress mobilizes energy for coping with the stressor, but also shuts down the initial fight or flight responses of the sympathetic ner­ vous and immune systems to prevent them from overshooting and damaging the organ­ ism (Munck, 2000). Glucocorticoid release during stress is thus primarily a protective response. If, however, cortisol levels are delayed in their poststress recovery, or repeated stress exposures result in sensitiza­ tion instead of habituation of the HPA axis, a chronic hyperactivation of this system can be maladaptive, leading to stress-related dis­ orders (McEwen, 2003).

What Kinds of Stimuli Activate the HPA Axis? It is a common misconception, probably going back to the work of Hans Selye (1956), that the HPA axis will respond to all types of

stressful experiences and the acute cortisol response can therefore serve as the gold stan­ dard for determining whether a particular experience was stressful. Many physiological systems are involved in stress responses, and each system varies in terms of the types of stressors that activate it, its temporal dynam­ ics, and its relations to other systems (Baum & Grunberg, 1995). For example, aversive stimuli that activate the sympathetic nervous system and adrenal medulla, producing elevations in heart rate, blood pressure, and catecholamines, do not necessarily lead to measurable changes in cortisol. Certain types of psychosocial stressors do have consistent effects. Reviews of early stud­ ies in humans, rodents, and nonhuman pri­ mates concluded that situations characterized by novelty, unpredictability, or low perceived control were most likely to activate the HPA axis (Mason, 1968; Rose, 1984). A recent metaanalysis of experimental studies showed that social-evaluative threat during task per­ formance and low control over the situation were the two best predictors of acute cortisol responses in humans (Dickerson & Kemeny, 2004). Although an individual’s appraisal of the stressor, coping, and degree of distress are predicted, on the basis of transactional stress theory (Lazarus & Folkman, 1984), to mod­ erate or (in the case of distress) mediate the cortisol response, laboratory studies have shown surprisingly low correlations between individual self-reports of these variables and cortisol measures. Physical stressors such as intense exercise also activate the HPA axis. The observation that cortisol elevations are often greater dur­ ing competitive sports than during training at the same level of physical exertion (Cook, Ng, Read, Harris, & Riad-Fahmy, 1987) indicates that physical and psychosocial com­ ponents of competition have additive effects. Cortisol levels also increase following experi­ mentally induced pain (al’Absi, Petersen, & Wittmers, 2002).

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 43

Measurement of Cortisol

MEASURING ACTIVITY OF THE HPA AXIS

Salivary Cortisol Background The first assays for salivary steroids were described in 1959, but this method did not gain widespread acceptance until researchers at the Tenovus Institute in Wales developed reliable assays for steroids in small volumes of whole saliva (Riad-Fahmy, Read, Walker, & Griffiths, 1982; Walker, Riad-Fahmy, & Read, 1978). Over the past 20 years, there has been an explosive growth in the number of studies using salivary measures to assess corti­ sol levels in a wide variety of applications in psychology, psychiatry, endocrinology, and beyond. The advantages of salivary cortisol sampling, compared to traditional procedures for blood sampling, have been summarized in several reviews (Kirschbaum & Hellhammer, 1989, 1994; Vining, McGinley, Maksvytis, & Ho, 1983). In addition to the ease and non­ invasive nature of sample collection, the fact that salivary cortisol is “free,” unbound by corticosteroid-binding globulin (CBG) or other carriers, is advantageous, as free cortisol thus represents the biologically active fraction of the hormone (Mendel, 1989). As noted earlier, salivary cortisol is ideal for assessing acute responses to experimental stressors. In addition, repeated measurement by subjects in their daily environment allows a good estimate of basal levels, diurnal vari­ ation, and response to awakening; some naturalistic designs also permit individual estimates of day-to-day variability and stress reactivity. Comparison With Blood Measures Cortisol levels measured in saliva correlate highly with free cortisol in blood. However, because of partial conversion of cortisol to cortisone during passage through the salivary

glands, the absolute level of free cortisol in saliva is 10% to 35% lower than it is in blood (Vining et al., 1983). Correlations with total blood concentrations (bound and free fractions) are also high, but the slope of the regression line becomes steeper at higher cortisol concentrations, after CBG-binding sites in blood are fully occupied. CBG levels can vary both within and between individu­ als, for example during pregnancy or with oral contraceptive use. Movement of cortisol from blood to saliva occurs by passive diffusion, so that salivary levels are independent of the flow rate of saliva (Vining et al., 1983). Changes in plasma and salivary cortisol levels are closely synchro­ nized. After injections of cortisol, salivary levels increased within 1 minute (Walker, 1984), and peak concentrations in blood are seen 2 to 3 minutes later in saliva (Kirschbaum & Hellhammer, 2000). Cortisol responses to awakening and to meals appear to be more pronounced in salivary than in plasma mea­ sures, and salivary cortisol returns to baseline more slowly after psychosocial stressors (Kirschbaum & Hellhammer, 2000). Collection The popularity of salivary cortisol mea­ sures is largely due to the ease of collecting samples from participants in both labora­ tory and field settings. A number of different techniques for collecting saliva samples have been described; which is most appropriate for a given research question will depend on characteristics of the participants, the setting, and frequency with which samples will be collected, and whether other substances will be measured in the same samples. Saliva samples are usually obtained from infants and toddlers with pipettes or other devices that aspirate saliva from the floor of the mouth, cotton ropes, swabs, or sponges held by the researcher or parent (Gunnar & Talge, 2007). In older children and adults,

43

03 - Luecken - 45370.qxd

44

9/12/2007

4:45 PM

Page 44

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

cotton dental rolls—including the widely used salivettes® (Sarstedt, Nümbrecht, Germany)—have convenient features for both research participants and laboratory personnel. Because of aspecific binding to the swabs or interference by other substances such as phytoestrogens that may be pre­ sent, cotton salivettes or swabs should not be used when other steroids (e.g., DHEA, testosterone, progesterone) or salivary immunoglobulin A (IgA) are also being measured (Shirtcliff, Granger, Schwartz, & Curran, 2001). In 2007 Sarstedt began pro­ duction of a new synthetic salivette, designed to eliminate the risk of batch-to-batch varia­ tion in the performance of cotton swabs. Swabs are unnecessary if participants can collect saliva by drooling into a tube, either directly or through a straw. Drooling may be less acceptable in studies where repeated samples need to be collected as rapidly and unobtrusively as possible, for example, dur­ ing participants’ daily activities outside the home. In one comparison, subjects collected adequate amounts of saliva in 1 to 2 minutes with cotton salivettes or cellulose-cotton tip “eyespears,” whereas passive drooling took from 1 to 15 minutes to produce the same amount (Strazdins et al., 2005). For all collec­ tion methods, it is important that the plastic storage tubes and stoppers are made of mate­ rials, such as unrecycled polypropylene (IBL, 2006), that do not absorb the hormone. Stoppers also need to fit tightly, because evaporation of saliva will lead to inaccurate cortisol results. Most cortisol assays require only 20 to 50 µl of saliva per tube, and therefore twice these amounts for a duplicate assay. In practice, larger volumes of saliva need to be collected when cotton-based methods are used, because up to 450 µl of saliva can remain in the cotton after centrifuging (de Weerth, Graat, Buitelaar, & Thijssen, 2003). Specialized techniques make it possible to extract cortisol from smaller sample volumes,

which may be a great advantage in studies of infants (de Weerth et al., 2003). In subgroups with low spontaneous flow rates (e.g., babies and small children, depressed patients, the elderly), saliva flow can be stimulated with powdered drink mix crystals, candies con­ taining citric acid, or lemon juice. Salivettes prepared with citric acid are also meant to stimulate salivary flow. Extreme caution is warranted in using such procedures, how­ ever, as they can lower the pH of the resulting saliva sample. Many currently available immunoassays produce false high values when sample pH is lower than 3.5 to 4 (Kirschbaum & Hellhammer, 2000; Schwartz, Granger, Susman, Gunnar, & Laird, 1998; Talge, Donzella, Kryzer, Gierens, & Gunnar, 2005; Vialard-Miguel, Belaidi, Lembeye, & Corcuff, 2005). Chewing on an inert sub­ stance (e.g., plain salivette, sugarless chewing gum, parafilm) or just making chewing move­ ments are good alternatives for stimulating salivary flow. Instructions to Subjects Subjects should be trained how to collect saliva samples and given the opportunity to practice under supervision to ensure that they collect adequate volumes of saliva. With salivettes, subjects should be instructed to chew lightly on the swab and to keep it fully inside the mouth until it feels saturated. (This can take 1 to 2 minutes, depending on salivary flow rate.) It is standard practice to ask sub­ jects not to brush their teeth in the 30 minutes before scheduled collection of a salivary sam­ ple. Acidic drinks, milk, and use of inhaled steroids (as examples of substances that could interfere with assay performance) should be avoided shortly before taking a saliva sample. If rinsing with water is considered necessary, it should be done at least 10 minutes before saliva collection to avoid diluting the cortisol concentration. Recent food intake and smok­ ing can influence cortisol responses to acute

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 45

Measurement of Cortisol

stressors and possibly to morning awakening and should be avoided in the hour before sam­ pling. It is crucial that subjects close the tubes tightly and label them with the exact time the sample was collected! Clear instructions should be given concerning storage (i.e., whether tubes should be kept at room temper­ ature, in the refrigerator, or in the home freezer); samples should not be left exposed to heat or sunlight. Storage and Handling Saliva samples can be stored at room tem­ perature (RT) or in participants’ home refrig­ erator or freezer until they are mailed or delivered to the lab. Estimates of how long cortisol is stable at RT range from 7 days (Groschl, Wagner, Rauh, & Dorr, 2001) to at least 4 weeks (Kirschbaum & Hellhammer, 2000). Centrifuging samples before storage appears to prolong the stable period (Groschl et al., 2001); nevertheless, increasing vari­ ance as well as decreasing levels over time indicate that storage at RT for more than 2 to maximally 4 weeks should be avoided (Garde & Hansen, 2005). Salivette samples develop mold and a bad odor after about 4 days at RT; this does not affect the cortisol concentrations, but makes the work of lab technicians unpleasant. The benefits of refrigeration at 4° to 5ºC, compared to RT, are unclear. In one study (Groschl et al., 2001), cortisol levels decreased in samples refrigerated for 11 days or longer; in contrast, Garde and Hansen (2005) found no change in cortisol levels in polyester salivettes refrigerated up to 3 months. Freezing clearly prolongs the stability of sali­ vary cortisol. In samples frozen at either –20º or –80ºC, cortisol concentrations remain sta­ ble for 9 months (Aardal & Holm, 1995) to 1 year (Garde & Hansen, 2005); freezing for as long as 2 years is probably possible. In settings where there is no access to refrigerators or freezers, stability of samples

can be prolonged by adding preservatives such as sodium azide (Groschl et al., 2001), citric acid (alone or with sodium benzoate), or ethyl and propyl paraben (Nimmagudda, Ramanathan, & Putcha, 1997). Cortisol in samples treated with citric acid and sodium benzoate remained stable for 180 days at RT (Nimmagudda et al., 1997). As noted earlier, preservatives, especially those that lower pH, may invalidate certain assays. Blood spots offer an alternative to saliva when extended storage at RT is necessary (Worthman & Stallings, 1997). (See Blood Spot Measures, below.) Salivary cortisol levels are relatively insen­ sitive to repeated thawing and refreezing; in recent studies, cortisol levels remained stable in samples undergoing up to three (Groschl et al., 2001) or four (Garde & Hansen, 2005) freeze/thaw cycles prior to assay. In the labo­ ratory, samples collected by passive drool are frozen and thawed at least once before assay­ ing in order to break down mucins that can interfere with pipetting (Vining & McGinley, 1986). Centrifuging helps remove particulate matter that can interfere with immunoassay. In salivettes, clear saliva collects in the bottom of the outer tube after centrifuging. There is normally no need to transport samples to the laboratory on ice (Clements & Parker, 1998). However, when the time in transit is more than a few days, shipping on dry ice will prevent molding and may be required by some laboratories. (For infor­ mation on international shipping, see Inter­ national Air Transport Association IATA regulations; adjustments made in 2005 exempt saliva samples from regulations for hazardous biological substances.) Prior to assay, saliva samples should be checked for blood contamination, as this can artificially elevate the cortisol concentration. Deficient diet, poor oral hygiene, and overly strenuous toothbrushing can cause bleed­ ing gums. In a recent study (Kivlighan et al., 2004), subjects first brushed their teeth

45

03 - Luecken - 45370.qxd

46

9/12/2007

4:45 PM

Page 46

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

vigorously and then collected saliva by direct drool. Minor injuries to the oral cavity led to detectable blood leakage in the samples, as assessed by three different methods: trans­ ferrin immunoassay, dipsticks for detecting hemoglobin in urine, and visual inspection. A moderate degree of blood contamination (samples visibly pink) had a negligible effect on cortisol levels, but darker saliva samples were more problematic. Visual inspection and discarding of saliva samples that are dis­ colored therefore appears to be adequate to control this source of error under normal cir­ cumstances. This is good news, because assay of transferrin—the most accurate method for assessing blood contamination—is relatively expensive, and dipsticks can yield false-posi­ tive results (Worthman & Stallings, 1997). Types of Assays Free cortisol in the blood represents only 4% to 5% of total cortisol released; more­ over, during passive diffusion into the salivary glands, approximately one-third of the free cortisol is lost through conversion to cortisone. Sensitive assay procedures are therefore necessary. Several methods cur­ rently allow reliable measurement of salivary cortisol without the necessity of extraction procedures. These include radioimmuno­ assay (RIA), enzyme-linked immunosorbent assay (ELISA), fluorescence immunoassay (FIA), and chemiluminescence immunoassay (LIA). The last three are nonradioactive assays in microtiter plate format, which can be run either manually or on automated equipment. Special laboratory equipment is required. Before the late 1990s, salivary cor­ tisol assays were often adaptations of proto­ cols designed for plasma/serum measures. Currently, assay kits developed for salivary determinations have standards suspended in a saliva matrix (in contrast to a serum or buffer matrix). Regardless of the assay used,

the following procedures are recommended: (1) assay samples in duplicate and use the mean value in statistical analyses,1 (2) repeat the assay for samples with duplicate values that differ by more than 20%, and (3) mea­ sure all samples from a given subject in the same assay run. Choosing a Lab Over the last several years, commercial labs have proliferated, in some cases offering assay services for salivary cortisol as well as kits for use in the investigator’s own lab. Details are available through the laboratories’ websites, for example www.salimetrics .com, www.ibl-hamburg.com, and www.dslabs .com. In addition, many hospital and research labs have expertise in salivary assays; some use commercially available kits, while others have developed their own in-house assays. Price is an important consideration; the costs of a duplicate cortisol determination can range from roughly $6 to $30, often with a discount for large quantities. Quality should also enter into the choice, as not all assays are equally sensitive or reliable. Fortunately, assay quality and cost are likely to be inversely related, because laboratories with tailored salivary assays also tend to have more experience, higher volumes, and auto­ mated procedures. Sensitivity refers to the minimum concen­ tration of cortisol that can be distinguished from zero. Salivary cortisol assays generally have a lower detection limit of less than .01 µg/dL, which is below the concentration nor­ mally observed until late in the evening when the HPA axis becomes quiescent. The relia­ bility of an assay, which is even more crucial for most research questions, is reflected in the intra- and interassay coefficients of varia­ tion (CV). The intra-assay CV can be calcu­ lated by dividing the standard deviation by the mean and then multiplying this figure by

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 47

Measurement of Cortisol

100, over a representative subsample (high, medium, and low cortisol concentrations) of duplicate measures from the same assay run. The interassay CV is calculated over several assay runs; laboratories should be able to provide this information on request. Both CVs should be under 12% to 15% (most are lower). In general, CVs are higher for corti­ sol concentrations in the lower part of the range. Even when intra- and interassay coefficients of variation are acceptable, some laboratories may obtain higher or lower absolute cortisol values than others (Hansen, Garde, Christensen, Eller, & Netterstrom, 2003; Kraemer et al., 2006). For this reason, it is inadvisable to switch from one type of assay to another or from one laboratory to another during the same study. Other indicators of assay performance are range of calibration, range of linearity of the assay (the linear region of the standard curve should cover the range in which most salivary cortisol values are found), spike recovery, and specificity (the percentage of crossreactivity with other endogenous or exogenous substances, like cortisone or pred­ nisone, that are or may be present in saliva). Laboratories should be able to provide this information for their assays. Reviewers are likely to request more detailed information for in-house assays. Recent implementation of voluntary quality assessment programs for salivary assays (IBL, 2005) will hopefully make it easier for researchers to evaluate and compare laboratories.

Urinary Cortisol Background Urinary measures of glucocorticoid metabolites (17-hydroxycorticoids) were among the first techniques available for studying activity of the HPA axis in humans, going back to the 1950s. An impressive body of knowledge emerged from the early

psychoendocrine studies of 17-HOCs levels (Mason, 1968). New techniques soon allowed researchers to measure cortisol directly, as small amounts are excreted as free cortisol in the urine (UFC). Urine samples collected over 24 hours provide an integrated measure of total free cortisol excretion. Mean UFC values are approximately 20µg/24 h (range 3-43 µg/24 h) in healthy adult women (Murphy, 2003). To reduce participant burden, collection over shorter periods may prove adequate for a specific research question. In addition, urine collection can be scheduled in such a way that more refined analyses are possible. As an example, Jerjes and colleagues were able to investigate diurnal patterns of HPA activity by having subjects collect urine every 3 hours for 15 hours (Jerjes et al., 2006). Another recent study compared women’s uri­ nary cortisol levels when they were at home, at work, or asleep (Dettenborn et al., 2005). A distinct advantage of urinary measures is that they allow assessment of nighttime cortisol levels, which may be crucial in cer­ tain disorders in which daytime levels are often normal (anxiety: Abelson & Curtis, 1996; PTSD: Yehuda, 2002). Urinary mea­ sures also have some disadvantages, which explain why they are less popular than sali­ vary cortisol. First, integrated measures are not very informative for research questions concerning acute stress responses. Second, the burden of collecting complete urine sam­ ples should not be underestimated, as it can lead to low participation in studies or poor compliance. Finally, transporting large vol­ umes of urine from field to laboratory is cumbersome. Collection, Storage, and Handling At the beginning of the sampling period, subjects void and discard the first urine. All urine produced thereafter is collected in large

47

03 - Luecken - 45370.qxd

48

9/12/2007

4:45 PM

Page 48

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

plastic containers designed for this purpose, or several containers if the study entails separate measurements. Samples can be kept at room temperature, without preservatives, for at least 24 hours without degradation of gluco­ corticoids (Gouarne, Foury, & Duclos, 2004). Assays Urinary free cortisol represents a small fraction of total cortisol released by the adrenal cortex. Commercially available RIA kits for measuring UFC may yield falsely high values, as results can be influenced by the presence of cortisol metabolites as well as other interfering substances; UFC values obtained with these assays are potentially two to four times higher than the true values established with chro­ matography (Murphy, 2002). Immunoassays have been reported to show particularly low specificity and poor precision at low cortisol concentrations, leading to widely discrepant results in studies of adrenal suppression (Fink et al., 2002). In choosing a laboratory, it is therefore important to make sure that the assay has been validated and is monitored according to established standards for UFC; details concerning the assay (accuracy, recov­ ery, precision, antibody used, crossreactivity, extraction method) should also be reported in publications. Accurate methods, for example liquid chromatography/tandem mass spec­ trometry (McCann, Gillingwater, & Keevil, 2005; Turpeinen & Stenman, 2003), are becoming more widely available and afford­ able. Because the tiny percentage (2-3%) of UFC in relation to total urinary cortisol metabolites may vary due to changes in steroid metabolism, measuring urinary cortisone, the ratio of cortisone to cortisol, or total cortisol metabolites may provide additional insights into HPA axis function (Gouarne, Groussard, Gratas-Delamarche, Delamarche, & Duclos, 2005; Jerjes et al., 2006). UFC results are often corrected for creatinine levels.

Blood Spot Measures Background Finger-prick blood spot sampling provides an alternative to salivary measures of cortisol; this technique combines the advantages of traditional blood samples, in terms of the range of substances that can be measured, with greater ease of sample collection and more convenient storage and handling pro­ cedures (Wong, Yan, Donald, & McLean, 2004; Worthman & Stallings, 1997). Using devices designed to allow diabetics to monitor their own glucose levels, collection of fingerprick samples in capillary blood is quick and minimally invasive. Because of the tiny amount of blood required, obtaining repeated samples from an individual is feasible. Blood spot cortisol is highly correlated with serum levels. The method also has some disadvan­ tages: not all participants can be trained to collect their own samples, so that research personnel may have to be involved; fingerpricks are not entirely painless, and recruit­ ment of subjects may be more difficult for this reason; subjects may be concerned about the safety of the procedure; and so on. Collection, Handling, and Storage Capillary blood from a finger-prick is dropped, without blotting or smearing, onto specially designed filter paper of the sort widely used in neonatal screening programs. One drop (yielding a blood spot of approxi­ mately 50 µL of whole blood) is sufficient for cortisol determination. After samples on fil­ ter paper are air-dried for several hours, they can be easily stored in plastic bags for trans­ port and even be mailed directly to the lab by ordinary post. Assays Special, highly sensitive assays have been developed to determine cortisol levels in

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 49

Measurement of Cortisol

blood spots. These assays are currently performed by an increasing number of labo­ ratories, including commercial laboratories such as Salimetrics and DSL. Additional hor­ mones and other substances can be measured in blood spots, including those not measur­ able in saliva, such as prolactin and markers of immune function (McDade et al., 2000). Accurate measures of estradiol and proges­ terone can also be obtained, enabling the researcher reliably to assess the stage of the menstrual cycle (Shirtcliff, Reavis, Overman, & Granger, 2001), for example.

A FRAMEWORK FOR DESIGNING A STUDY AND INTERPRETING THE RESULTS The first key to designing an effective study with clear results is awareness of the tempo­ ral dynamics of HPA axis activity, as these will dictate the sampling strategy. For studies of stress reactivity, the choice of a stressor can have a major impact on the results and their interpretation. Finally, the design should take into consideration the range of possible moderators, mediators, and confounders that might affect the hypothesized relationship between HPA measures and biopsychosocial variables of interest. This review attempts to summarize current recommendations and practice, without claiming that evidence in all cases is so consistent and complete that researchers have reached a consensus.

Sampling Strategy As previously described, integrated (in urine, UFC) and momentary (in saliva or blood) measures of cortisol are available. For UFC, the main decision is whether to collect samples over 24 hours or shorter time peri­ ods; the choice should be based on theoreti­ cal grounds, but subject burden and logistics

often play a role. For salivary cortisol and blood spots, the optimal number and timing of samples depends on the aspects of HPA activity being investigated (e.g., basal levels, diurnal variation, response to awaken­ ing, negative feedback inhibition, or response to acute stressors) and the stability of these measures over time. Basal Cortisol Levels and Diurnal Variation Although investigators seem to agree that cortisol should be measured several times a day for a number of days to get reliable esti­ mates of mean basal levels and diurnal slope (Goodyer et al., 2001; Stewart & Seeman, 2000), clear recommendations with support­ ing data are difficult to find. Hruschka and colleagues (2005) recently presented formu­ las for determining these figures on the basis of variance estimates from multilevel regres­ sion models. Based on data from a number of studies using different sampling protocols, these calculations suggested that—depending on the spacing of the samples in time—as few as four samples taken on one day might be adequate for estimating individual mean levels, but that 14 or more days of sampling with four to five samples per day might be necessary to obtain a reliable estimate of an individual’s diurnal slope. In contrast, a study in an older population with very good protocol compliance found that five samples per day for 3 days provided a reliable esti­ mate of daytime slope; moreover, slopes based on as few as two daily time points (wake and 9:00 p.m.) correlated highly with those based on four points, and additional days did not substantially increase reliability (Kraemer et al., 2006). These findings under­ score the need for more analyses of existing “daily profile” datasets. To establish an optimal sampling protocol for a specific population, a pilot study with at least

49

03 - Luecken - 45370.qxd

50

9/12/2007

4:45 PM

Page 50

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

50 participants is desirable (Kramer et al., 2006). Until more empirical results are avail­ able, a conservative recommendation would be to collect three to five samples a day for at least 3 days if basal levels are of primary interest and for 6 to 7 days if diurnal varia­ tion is a major focus. Increasing the number of subjects can increase statistical power when reliability of the cortisol measures is not optimal. Because individual differences in sleep patterns may be associated with shifts in the circadian cycle, some researchers have cho­ sen to collect samples at fixed intervals from the habitual time of awakening, instead of at fixed times of day. Another elegant but logis­ tically simpler design is to sample at fixed times of day and then model effects of time since awakening statistically (Cohen et al., 2006). In all cases, efforts should be made to obtain accurate information concerning the actual sample collection times. The most foolproof method is some form of elec­ tronic monitoring, for example, devices that record whenever a participant opens a vial to remove a cotton swab (Broderick, Arnold, Kudielka, & Kirschbaum, 2004; Jacobs et al., 2005; Kudielka, Broderick, & Kirschbaum, 2003) or handheld computers that generate time stamps with which participants must label their tubes (Stetler, Dickerson, & Miller, 2004). Awareness that compliance is being monitored increases the probability that samples will be taken as directed (Kudielka et al., 2003). Prompting participants with an audible or vibrating sig­ nal can also help. Finally, instructions to par­ ticipants should emphasize the importance of accuracy and honesty in reporting actual col­ lection times. In older adults, self-reported collection times were close to automatically recorded times, and test-retest reliability of slope estimates was actually slightly better when based on self-reported times (Kraemer et al., 2006).

Cortisol Awakening Response (CAR) The time course of this response has been well characterized. The peak response occurs 30 to 45 minutes after awakening; by 60 min­ utes, cortisol levels are decreasing and may no longer be reliably distinguishable from the levels at awakening. At least two samples (at awakening and either 30 or 45 minutes later) are needed to characterize the response; more samples (e.g., at 0, 30, 45, and 60 minutes) may increase reliability and allow calculation of AUC measures (see Statistical Analysis). The CAR should preferably be measured on at least 2 days. Given the narrow window of response, accurate timing of samples is cru­ cial. Because participants appear to have difficulty in taking early morning samples as directed (Kudielka et al., 2003), it may be wise to reduce the sample burden to the min­ imum, emphasizing quality rather than quan­ tity. Some kind of alarm device is useful to remind the participant to collect samples at the appropriate times, and compliance should be monitored electronically (see above) if this is possible. Activity monitors can be helpful in confirming time of awakening, but this is not considered essential for all studies. Methodological issues relevant to study design have been summarized by Clow and colleagues (2004). Instructions to subjects should be standardized along the following lines: • Place all materials next to your bed before going to sleep. • Take the first sample in bed immediately after awakening, with lights on and eyes open. • Do not go back to sleep; get out of bed (within 15 minutes) before taking another sample. • The second sample should be taken [n] minutes after awakening (and so on for each sample). • Do not brush your teeth, smoke, eat, or drink anything except water until you have finished taking the [n] morning samples.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 51

Measurement of Cortisol • Remember to record the exact time each sample was taken on the tube, even if this was not the scheduled time.

Dexamethasone Suppression as a Measure of Negative Feedback Sensitivity Dexamethasone (DEX) can be safely ingested by participants at home and its effects measured in salivary cortisol (Lindley, Carlson, & Benoit, 2004; Powell et al., 2002). In most studies, cortisol measures on a control day are compared with measures taken at the same times of day following intake of 0.25 to 0.5 mg (low dose) or 1.0 mg (high dose) dexamethasone late the previ­ ous evening (at 11 p.m. or an agreed-on bed­ time). The original dexamethasone suppression test (DST), developed as a diagnostic test for major depression, was scored as positive if, following administration of 1 mg DEX at 11 p.m. on day 1, cortisol at 4 p.m. on day 2 was above an established cutoff point (Carroll et al., 1981). For research purposes, analyzing the cortisol results as continuous instead of dichotomized measures yields more information. The optimal timing of the samples depends on the DEX dosage, as cor­ tisol will “escape” from suppression earlier with lower doses. Collecting a number of post-DEX saliva samples at intervals of sev­ eral hours will increase reliability of the results and gives added information about the time course of feedback inhibition.

The Cortisol Response to Acute Stressors Cortisol responses to acute stressors can be studied in the laboratory and in real life, where anticipated as well as unanticipated stressors occur. Design issues vary according to the setting. In the laboratory, important decisions include the best time of day to

schedule the experiment, how many samples are needed to characterize the stress response, the timing of these samples in relation to the stress task, the nature of the task, how to con­ trol for effects of novelty and anticipation, and habituation to repeated stressors. For a detailed overview of many of these design issues, see Dickerson and Kemeny (2004). Time of Day. Although the HPA axis is capa­ ble of responding to acute stress at any point in the diurnal cycle (Kudielka, Schommer, Hellhammer, & Kirschbaum, 2004), schedul­ ing experiments in the mid to late afternoon (roughly between 3 and 6 p.m.) has advan­ tages. First, the cortisol response to stress is more readily distinguishable in the afternoon than in the morning from background noise in the form of spontaneous pulsatile episodes and the natural decline in basal levels over the morning hours; the metaanalysis performed by Dickerson and Kemeny (2004) showed moderate effect sizes for cortisol response to stress tasks performed in the afternoon, com­ pared to small effect sizes in the morning. Second, cortisol responses are easier to pro­ voke in the afternoon than in the late evening, when the HPA axis becomes quiescent. Third, effects of potential confounders such as recent awakening and lunch (see below) are easier to exclude. Thus far, study design has been influenced mainly by such practical consider­ ations, and little attention has been paid to theoretically important issues, such as the consequences of differential activation of MR and GR systems by stressors occurring at the trough versus the peak of the diurnal cycle (Dallman, Akana, Bhatnagar, Bell, & Strack, 2000). Number and Timing of Samples. To charac­ terize the cortisol response to an acute stres­ sor, samples are taken at fixed intervals during baseline (30-40 minutes), stress expo­ sure (10-20 minutes), and recovery (40-60

51

03 - Luecken - 45370.qxd

52

9/12/2007

4:45 PM

Page 52

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

minutes) periods. Participants thus need to remain at the laboratory for a total of 1½ to 2 hours. The stress exposure includes both the preparation period, if there is one, and the actual task performance. Peak cortisol levels are usually observed 20 to 40 minutes after task onset, depending on the intensity and duration of the task, with a gradual return to baseline levels over the next hour or longer (Dickerson & Kemeny, 2004). Even with an identical task, however, there are marked individual differences in latency to peak response (Gunnar & Talge, 2007). If time and budget allow, an optimal design would include two to three baseline mea­ sures, one to two measures shortly after and possibly during the stress task, and two to three measures during the recovery period. Minimalistic designs with one prestress and one poststress measure run the risk of miss­ ing the peak response and provide no infor­ mation about speed of recovery. Controlling for Novelty and Anticipation. Prestress baseline measures are highly sensi­ tive to the novelty of the setting. Previous visits to the lab or an extended acclimation period (30 minutes or more) after arrival can reduce the probability of elevated baseline levels. Anxiety in anticipation of the task remains difficult to control. Because infor­ mation provided earlier as part of informed consent and pretask instructions can either heighten or reduce anxiety, procedures need to be fully standardized in terms of both con­ tent and timing. Obtaining a saliva sample at home, at the same time on another day, is very useful in determining whether lab base­ line levels are elevated (Nicolson, Storms, Ponds, & Sulon, 1997). This is important to know, because high baseline cortisol is often associated with a blunted response to stress (Kudielka, Schommer, et al., 2004; Young & Nolen-Hoeksema, 2001). Interestingly, infants and young children tend to show lower

cortisol levels at lab arrival than at home (Gunnar & Talge, 2007). Type of Stressor. Health psychology studies most frequently apply psychosocial stress tasks, as these are thought to have the greatest ecological validity. The HPA axis can be activated by physical stressors such as intense exercise or pain, or by pharmacological challenges; individuals’ responses to different classes of stressors, however, do not appear to be highly intercorrelated. Among the psychosocial stressors, performance tasks with elements of social-evaluative threat, uncon­ trollability, or both produce the largest and most consistent increases in cortisol. Less con­ sistent results are found for passive tasks (e.g., watching a film or other emotion induction procedures, noise exposure) and performance tasks without evaluative threat or uncontrolla­ bility (Dickerson & Kemeny, 2004). Only a few stress tasks have been described in sufficient detail that results can be compared across studies and populations. The best known of these is the Trier Social Stress Test (TSST) (Kirschbaum, Pirke, & Hellhammer, 1993). The widespread use of the TSST reflects the fact that it has been extensively studied, can be applied in sub­ jects varying in age and educational status, and induces a cortisol response in the major­ ity of participants. For the TSST and other tasks with a combination of social-evaluative threat and uncontrollability, effect sizes, on average, are large (Dickerson & Kemeny, 2004). Results are sensitive to changes in the procedure, however: when the interval between instructions to subjects about the task and performance was extended from 10 minutes to 1 hour, the cortisol response to the task was obliterated (Young & NolenHoeksema, 2001). Interpretation of Results. Although labora­ tory stress tasks are probably the single most

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 53

Measurement of Cortisol

common approach to investigating the HPA axis in health psychology, results can be difficult to interpret. Failure to detect a statistically significant cortisol response is a rather common occurrence, even when the task appears to be experienced as stressful. Weaknesses in study design are responsible for some of these negative findings. With ade­ quate sample size and the right choice of stress task and timing of measures, however, the majority of subjects are likely to show an increase in cortisol from baseline to posttask. Theoretically, one would expect the magni­ tude of this response to reflect the individ­ ual’s experience of the situation, in terms of appraised threat, coping possibilities, and the intensity of emotional distress. Unfortunately, this is rarely the case with laboratory tasks, possibly because self-report instruments are not sensitive to the relevant processes or because other aspects of the situation—for example, its novelty—are more salient. Another discouraging finding is that cortisol responses often show no correlation with personality traits linked to stress reactivity, such as neuroticism (Schommer, Kudielka, Hellhammer, & Kirschbaum, 1999). Further­ more, little is known about the generaliz­ ability of lab reactivity measures to real-life situations (for studies involving cortisol, see Houtman & Bakker, 1987; Lundberg, Melin, Fredrikson, Tuomisto, & Frankenhaeuser, 1990; van Eck, Nicolson, Berkhof, & Sulon, 1996). Giving a speech on a particular topic or performing mental arithmetic before an audience may also have different meanings for individuals or groups, depending on vari­ ables such as cognitive ability, occupation, and cultural background. The challenge for researchers is to design laboratory stressors that convincingly tap into the processes of interest in a given population and also reli­ ably activate the HPA axis. Examples include a task involving competition with an antago­ nistic peer in children (van Goozen et al.,

1998) and a standardized lecture in student teachers (Houtman & Bakker, 1987). Test-retest reliability of acute stress response measures appears to be low, proba­ bly because of both the underestimated noise introduced by spontaneous pulsatile activity (Young, Abelson, & Lightman, 2004) and the tendency of the cortisol response to habituate following repeated exposures. Low reliability of cortisol outcome measures means that laboratory stress experiments are particularly vulnerable to Type 2 error. This issue is important in all studies, but especially needs to be taken into account in inter­ vention studies, where stress reactivity is compared pre- and postintervention. More analyses are needed to determine how many cortisol measures per session and how many repeated sessions are necessary to obtain reli­ able measures of stress reactivity for different subject populations and stressors (Gunnar & Talge, 2007; Hruschka et al., 2005). Habituation and Sensitization. If the cortisol response to experimental stressors is to be considered a valid indicator of what goes on in daily life, it is important to know what happens following repetitive stressful experi­ ences. The acute response is adaptive, but is expected to habituate over repeated expo­ sures as novelty decreases and control increases. Failure to habituate or sensitization to repeated stressors, in contrast, is regarded as maladaptive, contributing in the long run to allostatic load. Habituation versus sensiti­ zation of the HPA response has been exten­ sively examined in animal models (e.g., Pitman, Ottenweller, & Natelson, 1990), but comparatively little research has been con­ ducted in humans (exceptions include al’Absi et al., 1997; Epel et al., 2000; Gerra et al., 2001; Gunnar, Hertsgaard, Larson, & Rigatuso, 1991; Kirschbaum, Prüssner et al., 1995; Wüst, Federenko, van Rossum, Koper, & Hellhammer, 2005). Findings indicate

53

03 - Luecken - 45370.qxd

54

9/12/2007

4:45 PM

Page 54

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

rapid habituation from the first to the second and later exposures, but also show marked individual differences, with a subset of indi­ viduals failing to habituate after repeated exposures. In addition, correlations between the cortisol response and trait characteristics may increase (Kirschbaum, Prüssner et al., 1995) or decrease (al’Absi & Lovallo, 1993) over successive task performances. These findings suggest that repeated testing (for example, three times in a week, or once a week for 3 weeks) is much more informative than single exposures in elucidating individ­ ual differences in stress reactivity that are rel­ evant to long-term health outcomes. Naturalistic Experiments. Responses of the HPA axis can also be investigated in response to real-life activities that entail some level of challenge or threat. Examples include exams, parachuting, sports competitions, musical performances, and occupational stressors, to name just a few. Tension-reducing activities such as yoga, meditation, and massage, on the other hand, may lower cortisol levels. Like lab experiments, these activities are usually scheduled or at least anticipated in advance, so that baseline, response, and recovery measures can be obtained. Certain activities, like parachute jumping, also lend themselves to studies of the habituation process (Deinzer, Kirschbaum, Gresele, & Hellhammer, 1997; Levine, 1978). Correlations between cortisol responses and subjective distress measures may be higher for real-life than for laboratory stressors (Nicolson, 1992). HPA axis responses to life events can also be investigated, although prestress baseline measures are rarely available because of the unpredictable nature of individual life events (e.g., rape, sudden death of a family member) and natural or manmade catastrophes (e.g., earthquakes, hurricanes, war). In this case, cortisol levels in exposed individuals are compared with those in an unexposed

comparison group and are often examined longitudinally, in relation to symptoms. Daily Hassles and Emotions. Combining repeated self-reports with salivary measures, experience sampling or ecological momentary assessment studies have investigated cortisol reactivity to daily life hassles and uplifts and accompanying emotions. Real-life stressors vary widely in duration, and participants are often unable to report exactly when a stressful situation began or when it ended. Although the timing of cortisol measures in relation to daily hassles is therefore imprecise, multilevel regression analyses can assess associations between the two. The association between daily events and cortisol is probably mediated by changes in negative affect. The finding of higher salivary cortisol in association with daily hassles or negative affects has been repli­ cated in several samples of adults (Hanson, Maas, Meijman, & Godaert, 2000; Peeters, Nicholson, & Berkhof, 2003; Smyth et al., 1998; van Eck, Berkhof et al., 1996) and children (Adam, 2006). In some but not all studies, positive affects were associated with lower cortisol (Adam, 2005; Polk, Cohen, Doyle, Skoner, & Kirschbaum, 2005).

Moderators and Confounders Interpretation of cortisol results can be facilitated by considering a number of between- and within-individual factors that can influence HPA axis activity. These include age, gender-related variables, and ethnicity; somatic variables such as illness, medications, and obesity; daily lifestyle vari­ ables such as food intake, smoking, and sleep patterns; psychosocial variables related to stress; and genetic differences. Age and Gender-Related Variables Findings concerning the effects of age and gender on the HPA axis vary from study to

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 55

Measurement of Cortisol

study in magnitude and sometimes even in direction. It remains essential to consider and if necessary to control for the independent effects of these two variables and their possi­ ble interactions. In brief summary of the most consistent findings, cortisol levels increase with age, especially in the very old, and changes also occur, sometimes depen­ dent on gender, in circadian amplitude and phase (Van Cauter et al., 1996). Age-related differences in acute stress reactivity have also been reported (Nicolson et al., 1997; Otte et al., 2005). Gender appears to have a negligible effect on basal cortisol levels or diurnal slopes, but males and females often show different responses to experimental stressors. As sum­ marized in recent reviews (Kajantie & Phillips, 2006; Kudielka & Kirschbaum, 2005), gender differences in cortisol reactiv­ ity have been attributed to the influence of female reproductive hormones as well as exogenous estrogens, but also to differ­ ences in cognitive, emotional, and behav­ ioral responses to specific stressors (e.g., Kirschbaum, Klauer, Filipp, & Hellhammer, 1995; Stroud, Salovey, & Epel, 2002). Basal levels remain stable throughout the men­ strual cycle, but both menstrual phase and oral contraceptive use may influence cortisol reactivity to stressors. Although findings are conflicting, current evidence indicates that salivary cortisol responses to psychosocial stress are blunted during the follicular as compared to the luteal phase of the men­ strual cycle, whereas responses of women in the luteal phase are similar to those of men. Oral contraceptive users show responses similar to those observed in the follicular phase (Kajantie & Phillips, 2006; Kudielka & Kirschbaum, 2005). Pubertal development in girls (Netherton, Goodyer, Tamplin, & Herbert, 2004), pregnancy (de Weerth & Buitelaar, 2005), and menopause (Kajantie & Phillips, 2006; Kudielka, BuskeKirschbaum, Hellhammer, & Kirschbaum,

2004) have all been reported to moderate either basal cortisol levels or stress reactivity. Female hormonal and reproductive status should therefore be taken into account in study design and analysis. Race/Ethnicity Studies examining ethnic differences in cortisol measures have produced mixed results (Bennett, Merritt, & Wolin, 2004; Cohen et al., 2006; Polk et al., 2005; Reynolds et al., 2006). Given race differences in other physiological measures and their rel­ evance to disparities in health outcomes, additional research is needed. Somatic Variables Illness. Participants who are acutely ill, with fever and malaise, should be excluded or rescheduled after full recovery. Chronic disorders such as Type 1 diabetes and other endocrine disorders, epilepsy, autoimmune disorders, and severe psychiatric disorders are often excluded because of their known or suspected direct effects on the HPA axis or effects of associated medications. Adrenal disorders should clearly be excluded, and it is standard practice to exclude other severe or unmanaged chronic disorders. For more prevalent and manageable disorders, exclu­ sion criteria should be considered in light of study objectives and population. In a com­ munity sample of older men and women, for example, exclusion of all with hyperten­ sion, Type 2 diabetes, asthma, fibromyalgia, osteoarthritis, or a lifetime or family history of psychiatric disorder would leave few par­ ticipants, and results of such a study would not be generalizable. Self-reports of these conditions are also unreliable, as many of these illnesses remain undiagnosed. Medication. A similar situation applies to confounding effects of medications. These

55

03 - Luecken - 45370.qxd

56

9/12/2007

4:45 PM

Page 56

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

should not be underestimated, and a conser­ vative approach (as in most clinical studies of the HPA axis) would require all subjects to be drug free. However, due to the range of med­ ications in widespread use and their small or as yet unknown effects of the HPA axis, it is not feasible in health psychology studies to exclude all subjects who take any medication. All medications should therefore be carefully recorded. Some classes of drugs, in particular systemic GCs like prednisone, predniso­ lone, and hydrocortisone, can have long-term effects on the HPA feedback system, and indi­ viduals who have used them in the past 6 months should be excluded. Anticonvulsants such as phenytoin and carbamazepine should also be excluded (Kunzel et al., 2003), as well as pure agonist opioids (Hibel et al., 2006). Use of low-dose GC inhalers, intranasal sprays, and topically applied creams can lead to mild suppression of the HPA axis in some individuals (Masharani et al., 2005), but this is unlikely to be a serious confounder (Hibel et al., 2006). Clinical dosages of zolpidem, a frequently used non-benzodiazepine hypnotic, does not alter cortisol rhythms (Copinschi et al., 1995). Use of antidepres­ sants, low-dose benzodiazepines, non­ steroidal anti-inflammatory drugs (NSAIDs), antihypertensives, and even over-the-counter drugs such as acetylsalicylic acid and acetaminophen (Hibel et al., 2006) should be evaluated in light of study goals and con­ trolled for as necessary.

in peripheral obesity. Men and women with central obesity often show elevated cortisol responses to laboratory stressors and to food intake, but there is marked heterogeneity, also in patterns of diurnal salivary cortisol secretion (Rosmond & Björntorp, 2001). Depending on the population and research questions, waist-to-hip ratio (WHR) may be a more informative measure than body mass index (BMI) (Epel et al., 2000; Ljung et al., 2000). BMI is a useful index of abnormally low body weight due to fasting or malnutri­ tion, which has also been associated with HPA axis irregularities.

Body Weight. A comprehensive review of the literature on cortisol in human obesity (Björntorp & Rosmond, 2000) indicates that cortisol secretion rate is elevated in obesity, but cortisol is removed more rapidly from the circulation; the net effect is normal or lower-than-normal basal levels. However, studies that take the type of obesity into account have found that hypercortisolemia and dysregulation of the HPA axis are more common in central, abdominal obesity than

Food Intake. Controlled experiments have shown that food intake, particularly at lunch, increases cortisol secretion. In studies where salivary cortisol was repeatedly sampled over the day under naturalistic conditions, recent food intake at any time of day was associated with higher cortisol (Peeters et al., 2003; van Eck, Berkhof, et al., 1996). Studies have shown that the magnitude of the response depends on the macronutrient composition of the meal. Protein-rich meals lead to an

Daily Activities and Lifestyle Sleep Patterns. The circadian cycle is sensi­ tive to disturbances and individual differ­ ences in the sleep-wake cycle. Studies of diurnal variation or the CAR should cer­ tainly control for effects of wake-up time, sleep duration, acute sleep loss (Leproult, Copinschi, Buxton, & Van Cauter, 1997), and disturbances in the sleep-wake pattern, including those due to jet lag and shift work. Related variables include individual differences in morningness-eveningness and seasonal changes in zeitgebers that affect HPA axis activity directly or through changes in sleep and activity patterns (Polk et al., 2005; Touitou et al., 1983). Seasonal effects may be more pronounced in certain disorders (Sher et al., 2005).

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 57

Measurement of Cortisol

increase of 50% to 100% in cortisol concen­ trations, with levels starting to rise approxi­ mately 30 minutes after meal onset, peaking around 60 minutes, and returning to base­ line within 2 hours (Gibson et al., 1999; Slag, Ahmed, Gannon, & Nuttall, 1981). Glucose intake enhances cortisol response to acute stressors (Gonzalez-Bono, Rohleder, Hellhammer, Salvador, & Kirschbaum, 2002), which implies that food intake prior to undergoing an experimental stressor should be carefully controlled. Researchers might consider offering participants a standardized snack, with uniform caloric and carbohydrate content, an hour before baseline measures or else ask participants to refrain from eating. Similarly, subjects should not eat before com­ pleting all assessments of the CAR. For assess­ ment of diurnal profiles with fixed schedules, researchers should choose sampling times long enough after usual meal times to mini­ mize the effects of food intake (e.g., 11 a.m., 4 p.m., and 9 p.m. instead of 9 a.m., 2 p.m., and 7 p.m.). In studies with more frequent measures over the entire day, participants should be asked to record whether they have eaten in the past hour. Caffeine Intake. Dietary doses of caffeine have been shown to increase cortisol secre­ tion under experimental conditions (Lovallo, Al’Absi, Blick, Whitsett, & Wilson, 1996). In a recent study, acute cortisol responses to cumulative caffeine administration during a single day were reduced but not eliminated when subjects had consumed caffeine on the preceding 5 days, compared to a placebo condition (Lovallo et al., 2005). This sug­ gests that even regular coffee drinkers may display some degree of HPA axis activation, especially in the afternoon. A single cup of coffee or tea, on the other hand, is unlikely to trigger an acute cortisol response (Quinlan, Lane, & Aspinall, 1997). Taken together, these results suggest that habitual caffeine consumption may be a relevant trait variable

to assess, especially if one suspects that groups being compared might differ in per­ centage of coffee drinkers. Caffeine intake in the past hour is not likely to be a serious con­ founder in experimental studies. Smoking. A number of studies in different age groups have reported higher cortisol levels in habitual smokers than in nonsmokers. For example, teenagers who smoked more than 10 cigarettes a day had higher basal cortisol levels than nonsmokers or light smokers; this effect was especially pronounced in girls (Canals, Colomina, Domingo, & Domenech, 1997). In college student smokers, serum cor­ tisol levels were higher than in nonsmokers (Gilbert, Stunkard, Jensen, Detwiler, & Martinko, 1996). Habitual smoking was sim­ ilarly associated with higher serum cortisol levels in postmenopausal women; this affect was not attributable to acute effects of smok­ ing cigarettes during the test day (Baron, Comi, Cryns, Brinck Johnsen, & Mercer, 1995). In middle-aged men and women, sali­ vary cortisol levels were higher throughout the day in smokers than in nonsmokers, and smokers’ cortisol responses to awakening were also greater (Steptoe & Ussher, 2006). Smokers in the process of quitting show an acute reduction in cortisol levels in the early weeks, with levels gradually returning to somewhat under the preabstinence baseline after 4 to 6 weeks (Frederick et al., 1998; Steptoe & Ussher, 2006). Because smoking status can be an important confounder and may also partially mediate effects of other variables of interest (for examples, see Cohen et al., 2006; Olff et al., 2006), it should always be assessed, including number of cigarettes or other sources of nicotine per day and recent cessation or reduction. Smoking may have greater trait than state influences on cortisol. One-day abstinence compared to ad libitum smoking had no effect on cortisol measures in habitual smok­ ers (al’Absi, Amunrud, & Wittmers, 2002).

57

03 - Luecken - 45370.qxd

58

9/12/2007

4:45 PM

Page 58

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

On the other hand, several studies have reported that recent smoking can cause tran­ sient cortisol elevations (Baron et al., 1995; Kirschbaum, Wust, & Strasburger, 1992) and attenuates the cortisol response to acute psy­ chosocial stressors (Kirschbaum, Strasburger, & Langkrar, 1993; Rohleder & Kirschbaum, 2006; Tsuda, Steptoe, West, Fieldman, & Kirschbaum, 1996). Subjects should therefore be asked to refrain from smoking for at least an hour prior to cortisol measurements in the laboratory or scheduled sampling in real life. With random sampling, participants should record whether they have smoked in the last hour so that possible effects can be controlled for in the statistical analysis. Alcohol Intake. Acute alcohol intake is asso­ ciated with an increase in cortisol in light drinkers; this response is attenuated in heavy drinkers (King, Houle, de Wit, Holdstock, & Schuster, 2002). In a study of non–alcohol­ dependent binge drinkers, moderate consump­ tion of white wine (4 glasses over roughly 2 hours) reduced the cortisol response to food intake (Kokavec & Crowe, 2001). Moderate alcohol consumption does not appear to sig­ nificantly influence basal cortisol levels, but alcohol dependence can alter both basal activ­ ity and reactivity of the HPA axis. Compared to alcohol-dependent individuals who were currently abstinent, those who were recently intoxicated displayed elevated cortisol, which increased further during withdrawal (Adinoff, Ruether, Krebaum, Iranmanesh, & Williams, 2003). Depending on study goals and popula­ tion, these findings suggest that subjects should be screened for alcohol or other sub­ stance dependence. This is often an exclusion criterion. Physical Activity. Both recent physical exer­ tion and habitual athletic training can influ­ ence HPA activity. Experiments have shown that an hour of high-intensity exercise

(at 70% VO2 peak) leads to pronounced increases in cortisol levels, whereas lower intensities or shorter durations had no sig­ nificant effects (Jacks, Sowash, Anning, McGloughlin, & Andres, 2002). Moderate increases in physical activity are thus unlikely to have a measurable effect on cortisol. Postural changes (e.g., from supine to standing) do not affect cortisol measures (Hucklebridge, Mellins, Evans, & Clow, 2002). Psychosocial Variables Past and current exposure to life events and chronic stressors are known to moderate basal cortisol levels and stress reactivity, as do individual traits like neuroticism and habitual coping styles. Depending on the research question, it may be useful to assess childhood adversity, recent life events, chronic stress (low socioeconomic status, difficulties with work, family, or other life domains), neuroticism, trait positive affect, coping styles, and current symptoms of depression, anxiety, or fatigue. Psychologi­ cal state at the time samples are collected is also important to assess, including separate measures of positive and negative affect, and recent as well as anticipated stressors. Genetic Polymorphisms There is growing evidence of individual differences in genetic vulnerability to stress, as well as gene-environment interactions (Caspi et al., 2003). Polymorphisms in genes involved in HPA axis regulation (DeRijk, Schaaf, & de Kloet, 2002; Wüst, DeRijk, et al., 2005) or in other stress-sensitive systems may predispose individuals to show different patterns of cortisol response to envi­ ronmental stressors. DNA can be obtained noninvasively from buccal swabs, salivettes, or mouth rinses (Etter, Neidhart, Bertrand, Malafosse, & Bertrand, 2005).

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 59

Measurement of Cortisol

STATISTICAL ANALYSIS OF CORTISOL DATA

Preparing the Data for Analysis Checking the Cortisol Distributions Before statistical analysis, cortisol data should be screened to eliminate outliers, in particular those values exceeding the normal physiological range. The highest unstimulated salivary cortisol levels observed in healthy subjects are approximately 45 to 50 nmol/L, most likely to occur in the early morning. Cortisol outliers can also be defined statisti­ cally, for example, as values greater than 4 standard deviations above the mean. The researcher should also define criteria for excluding individuals who have a relatively high percentage of suspect measures, even if some samples are in the normal range. Even after excluding physiologically abnormal data points, cortisol values display skewed distributions, especially in the early morning and evening hours. Data are usually logarithmically transformed prior to analy­ sis, to avoid violating assumptions of com­ mon statistical procedures. Transformation, when necessary, should be done on the mea­ sure (e.g., area under the curve, or AUC) actually being entered into the analysis.

Checking Compliance With Timing of Samples If saliva collection is unsupervised, it may be necessary to exclude samples not taken close enough to the intended collection time. This is essential for accurate measurement of the cortisol response to awakening, as the peak response occurs within a narrow time window (30-45 minutes after awakening); participants who collect saliva samples too late will thus falsely appear to have a blunted response. Later in the day, when cortisol levels are fairly stable over periods of a few hours,

deviations from planned collection times are less likely to have a serious impact on the results; here, it is appropriate to exclude sam­ ples outside preestablished windows of 30 to 60 minutes on either side of the scheduled collection time. Some researchers define even larger windows of acceptability (Cohen et al., 2006). In multilevel regression approaches, the diurnal curve can be more accurately modeled using actual collection times (self­ reported or electronically monitored) (Ranjit et al., 2005), and in this case, cortisol mea­ sures taken later than scheduled need not be excluded. Intensive, semi-random sampling schedules appear to enhance participant com­ pliance and improve the reliability of cortisol results obtained over the day, even when par­ ticipants believe that their compliance is not being monitored (Jacobs et al., 2005).

Statistical Methods Total Cortisol Concentration and Diurnal Variation Urinary cortisol provides an integrated measure of total output over the collection interval. Momentary assessments of cortisol in saliva or blood, however, need to be aggre­ gated or statistically modeled in order to test effects of other variables on daily cortisol lev­ els or slopes. Individual summary measures of basal levels or diurnal slope are also needed when cortisol is an independent variable in the analysis—for example, in longitudinal studies where cortisol measures are examined as predictors of health outcomes (Sephton et al., 2000). There are several options. For total levels, the simplest approach is to calculate the sum or average of two or more samples on a given day. With three or more daily sam­ ples, calculating an AUC for total levels takes into account that time intervals between sam­ ples may not be equal. Because the preceding measures will be heavily influenced by morn­ ing cortisol (when normative levels are high),

59

03 - Luecken - 45370.qxd

60

9/12/2007

4:45 PM

Page 60

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

other researchers have standardized cortisol values relative to the sample mean at each time of day before calculating a daily average cortisol measure (DAC; Gunnar, Morison, Chisholm, & Schuder, 2001). An underly­ ing assumption of this approach is that the biological “meaning” of a given cortisol level may vary according to the time of day, as studies in animal models suggest (Dallman et al., 2000). Advantages and disadvantages of aggregated measures have been discussed (Hruschka et al., 2005; Kraemer et al., 2006; Rosmalen et al., 2005). The most important drawback is that these measures ignore and obscure diurnal variation, which is funda­ mental to understanding the nature of HPA axis dysregulation. The current consensus that cortisol levels and slopes represent differ­ ent constructs argues against combining them in the same measure (Kraemer et al., 2006). The simplest measure of diurnal change entails calculating the difference between the morning and the evening values. When cortisol is the dependent variable, analysis of variance for repeated measures also allows differentiation between overall basal level and diurnal variation. A practical problem is that missing data at one of the sampling times results in exclusion of the entire day. When multiple days are sampled, it may be possible to reduce the percentage of missing data in the analysis by first aggregating data at each time of day, but this approach is far from elegant. In short, these models do not allow accurate modeling of diurnal variation in cortisol secretion. In comparison to these traditional approaches, multilevel regression (hierarchi­ cal linear modeling) offers more accurate and statistically powerful approaches to analyzing cortisol data. The multilevel model is a vari­ ant of multiple linear regression, appropriate for data sets with a hierarchical structure (Raudenbush & Bryk, 2002). These methods offer many advantages for analyzing cortisol data, when the hormone measures are nested

within days, and days are often nested within participants. First, they allow estimation of effects of independent variables on overall cortisol level, slope, and other characteristics of diurnal activity, as well as individual esti­ mates of these parameters. Second, because discrete or continuous explanatory variables can enter the model at any level of the hierar­ chy, trait and state influences on cortisol can be teased apart. Models can thus be extended to test associations between cortisol and such time-varying covariates as emotional state or recent food intake, as well as individual characteristics like age or gender. Third, the multilevel model makes maximum use of the data, as it can deal flexibly with missing data and does not require fixed time intervals between successive measures. This is particu­ larly useful in studies outside the laboratory, in which missing data are inevitable when participants are asleep, forget, or are unable to comply with the sampling schedule. Moreover, the problem of subjects’ failing to collect samples on schedule is reduced, as actual collection times can be accommodated in the analysis (Ranjit et al., 2005). Fourth, multilevel models explicitly take into account the dependencies among repeated cortisol measures taken within days and within indi­ viduals, thus allowing valid inferences. Finally, statistical power is increased com­ pared to analyses of aggregated data. Hruschka and colleagues (2005) and Ranjit and colleagues (2005) provide more detailed rationales for using multilevel meth­ ods in the analysis of cortisol data, as well as examples. Initially applied in experience sam­ pling (ecological momentary assessment) studies with semi-random sampling intervals (Smyth et al., 1998; van Eck, Berkhof et al., 1996), multilevel modeling is also ideal for analysis of diurnal profile data collected at fixed time points. Despite its many advan­ tages, multilevel modeling has the drawback of requiring more statistical expertise than some more familiar methods. Accurately

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 61

Measurement of Cortisol

modeling the diurnal cortisol curve, for example, is more complex than it might first appear. Moreover, although recent versions of standard statistical software packages (SAS, Stata, SPSS) offer more extensive pro­ cedures for mixed models, consultation with a statistician remains advisable for all but the simplest analyses. Cortisol Response to Awakening (CAR) The CAR is usually operationalized as the absolute change in cortisol levels from awakening to either a fixed time point (e.g., 30 minutes) or the peak value of repeated measurements over the first hour. Alternatively, an AUC can be calculated, either as the response from waking baseline or as the total area relative to zero (Pruessner, Kirschbaum, Meinlschmid, & Hellhammer, 2003). Results should be presented in such a way that it is clear whether a larger CAR is due to relatively low cortisol at awakening. Although cortisol generally shows a pronounced increase fol­ lowing awakening, declining levels (negative CARs) are also observed and are not neces­ sarily artifactual (Wüst et al., 2000). Given the current lack of consensus concerning the best way to characterize the CAR and the extent to which nonresponse is due to con­ founders, researchers are urged to provide as much information as possible (Clow et al., 2004). The CAR should be considered a dis­ crete part of the circadian cycle, and samples taken to assess the response should be excluded from analyses of basal levels and diurnal slopes. Cortisol Response to Acute Stress For assessing cortisol responses to experimental stressors, traditional statistical approaches such as repeated measures ANOVA/ANCOVA are often used. The liter­ ature varies considerably in how constructs

such as baseline, peak response, total response, and recovery are operationalized. An AUC can be computed, using a trapezoidal integra­ tion, as a measure of total response. The AUC is usually considered to be the area above the baseline level, but should also be allowed to take negative values (Grice & Jackson, 2004; Gunnar & Talge, 2007; Pruessner et al., 2003). Stress response measures should be corrected for baseline levels, as higher baseline is often associated with an attenuated response. A recent study used growth curve analysis (a form of multilevel modeling) to character­ ize changes in salivary cortisol levels during the TSST over three baseline samples, two stress response samples, and four recovery samples for each subject (Taylor et al., 2006). Estimates for the cortisol intercept and slopes of baseline, reactivity, and recovery measures were all significant and varied from person to person (random effects); to address the study’s hypotheses, effects of between-subject predictors were tested in an extension of this model (in this case, oxytocin level at baseline was a significant predictor of cortisol inter­ cept and baseline slope, whereas the presence of an audience was associated with steeper reactivity and recovery slopes). Given the gen­ eral advantages of multilevel modeling men­ tioned earlier, plus the specific advantages of being able to model separate and theoretically important aspects of cortisol variability in the laboratory, this new statistical approach is likely to become standard practice.

HOW TO REPORT FINDINGS Laboratories often report cortisol concentra­ tions in metric units (either µg/dL or ng/dL), whereas research journals may prefer molar units of measurement. For salivary measures, for example, the conversion from metric (ng/ dL) to molar units (nmol/L) simply entails dividing by 36.2. Similarly, urinary free cor­ tisol can be expressed in micrograms/24

61

03 - Luecken - 45370.qxd

62

9/12/2007

4:45 PM

Page 62

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

hours or nanomoles/24 hours (or per unit time for shorter periods); micrograms/24 hours divided by 3.62 transforms the value to nanomoles/24 hours. The methods section of papers should briefly mention the following: exclusion cri­ teria (for subjects or samples) intended to reduce confounding of HPA axis measures, frequency and timing of measures, whether compliance was electronically monitored, sample collection method, storage condi­ tions, type of assay performed (including name and location of manufacturer if a com­ mercial kit was used), and assay performance characteristics (inter- and intraassay coeffi­ cients of variation and lower detection limit; for in-house assays, reviewers will expect more details, e.g., spike recovery). For labo­ ratory studies, a detailed description of the stress task is essential, including the time of day when experiments were performed. The results section should include infor­ mation about compliance with ambulatory collection procedures and times of awaken­ ing or meals, if these could have influenced the cortisol results. Presentation of statistical results should include information concern­ ing effect sizes and observed power, so that the meaning of nonsignificant results can be

judged. Providing intraclass correlations (ICCs) for mixed models enables readers to evaluate the strength of the results (what percentage of the variance in cortisol is explained by within- or between-person variables, after controlling for time of day effects) and to estimate sample sizes needed for future studies (see Hruschka et al., 2005).

CONCLUSION The availability of reliable and noninvasive methods for measuring cortisol, in combina­ tion with an appreciation of the pivotal role of stress in psychiatric and psychosomatic disorders, has spurred many researchers to consider incorporating cortisol measures into their investigations. The goals of this chapter were to describe basic features of the HPA axis and to alert the reader to possibilities and constraints at each stage in the process from study design, sample collection, assays, and statistical analysis through interpretation and presentation of results. Attention to these details can help ensure that new studies will continue to extend our understanding of how the HPA axis contributes to human health and disease.

SPECIALIZED REVIEWS ON RELATED TOPICS Clow, A., Thorn, L., Evans, P., & Hucklebridge, F. (2004). The awakening cortisol response: Methodological issues and significance. Stress, 7(1), 29–37. de Kloet, E. R., Joëls, M., & Holsboer, F. (2005). Stress and the brain: From adap­ tation to disease. National Review of Neurosciences, 6, 463–475. Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychology Bulletin, 130(3), 355–391. Goodyer, I. M., Park, R. J., Netherton, C. M., & Herbert, J. (2001). Possible role of cortisol and dehydroepiandrosterone in human development and psycho­ pathology. British Journal of Psychiatry, 179, 243–249. Gunnar, M. R., & Talge, N. M. (2007). Neuroendocrine measures in developmen­ tal research. In L. A. Schmidt & S. J. Segalowitz (Eds.), Developmental psy­ chophysiology (pp. 343–366). Cambridge, UK: Cambridge University Press.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 63

Measurement of Cortisol Heim, C., Ehlert, U., & Hellhammer, D. H. (2000). The potential role of hypo­ cortisolism in the pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology, 25(1), 1–35. Kirschbaum, C., & Hellhammer, D. H. (2000). Salivary cortisol. In G. Fink (Ed.), Encyclopedia of stress (Vol. 3, pp. 379–383). New York: Academic Press.

NOTE 1. This standard procedure may not always be necessary. A recent study showed that the accuracy of results obtained from a single cortisol assay (LIA) was so high that using the mean value of duplicate assays did not substantially increase reliabil­ ity (Kraemer et al., 2006).

REFERENCES Aardal, E., & Holm, A. C. (1995). Cortisol in saliva—reference ranges and relation to cortisol in serum. European Journal of Clinical Chemistry and Clinical Biochemistry, 33(12), 927–932. Abelson, J. L., & Curtis, G. C. (1996). Hypothalamic-pituitary-adrenal axis activ­ ity in panic disorder. 24-hour secretion of corticotropin and cortisol. Archives of General Psychiatry, 53(4), 323–331. Adam, E. K. (2005). Momentary emotion and cortisol levels in the everyday lives of working parents. In B. Schneider & L. Waite (Eds.), Being together, working apart: Dual career families and the work-life balance (pp. 105–134). Cambridge, UK: Cambridge University Press. Adam, E. K. (2006). Transactions among adolescent trait and state emotion and diurnal and momentary cortisol activity in naturalistic settings. Psychoneuroendocrinology, 31(5), 664–679. Adinoff, B., Ruether, K., Krebaum, S., Iranmanesh, A., & Williams, M. J. (2003). Increased salivary cortisol concentrations during chronic alcohol intoxication in a naturalistic clinical sample of men. Alcoholism-Clinical and Experimental Research, 27(9), 1420–1427. al’Absi, M., Amunrud, T., & Wittmers, L. (2002). Psychophysiological effects of nicotine abstinence and behavioral challenges in habitual smokers. Pharmacology Biochemistry and Behavior, 72(3), 707. al’Absi, M., Bongard, S., Buchanan, T., Pincomb, G. A., Licinio, J., & Lovallo, W. R. (1997). Cardiovascular and neuroendocrine adjustment to public speak­ ing and mental arithmetic stressors. Psychophysiology, 34(3), 266–275.

al’Absi, M., & Lovallo, W. R. (1993). Cortisol concentrations in serum of border­ line hypertensive men exposed to a novel experimental setting. Psychoneuroendocrinology, 18(5–6), 355–363. al’Absi, M., Petersen, K. L., & Wittmers, L. E. (2002). Adrenocortical and hemo­ dynamic predictors of pain perception in men and women. Pain, 96(1–2), 197–204. Baron, J. A., Comi, R. J., Cryns, V., Brinck Johnsen, T., & Mercer, N. G. (1995). The effect of cigarette smoking on adrenal cortical hormones. Journal of Pharmacology and Experimental Therapeutics, 272(1), 151–155.

63

03 - Luecken - 45370.qxd

64

9/12/2007

4:45 PM

Page 64

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Baum, A., & Grunberg, N. (1995). Measurement of stress hormones. In S. Cohen, R. C. Kessler, & L. Underwood Gordon (Eds.), Measuring stress: A guide for health and social scientists (pp. 175–192). Oxford, UK: Oxford University Press. Bennett, G. G., Merritt, M. M., & Wolin, K. Y. (2004). Ethnicity, education, and the cortisol response to awakening: A preliminary investigation. Ethnicity & Health, 9(4), 337–347. Bhagwagar, Z., Hafizi, S., & Cowen, P. J. (2005). Increased salivary cortisol after waking in depression. Psychopharmacology (Berlin), 182(1), 54–57. Bjorntorp, P., & Rosmond, R. (2000a). Obesity and cortisol. Nutrition, 16(10), 924–936. Bjorntorp, P., & Rosmond, R. (2000b). Neuroendocrine abnormalities in visceral obesity. International Journal of Obesity, 24, S80–S85. Bower, J. E., Ganz, P. A., Dickerson, S. S., Petersen, L., Aziz, N., & Fahey, J. L. (2005). Diurnal cortisol rhythm and fatigue in breast cancer survivors. Psychoneuroendocrinology, 30(1), 92–100. Broderick, J. E., Arnold, D., Kudielka, B. M., & Kirschbaum, C. (2004). Salivary cortisol sampling compliance: Comparison of patients and healthy volunteers. Psychoneuroendocrinology, 29, 636–650. Canals, J., Colomina, M. T., Domingo, J. L., & Domenech, E. (1997). Influence of smoking and drinking habits on salivary cortisol levels. Personality and Individual Differences, 23(4), 593–599. Carlson, L. E., Speca, M., Patel, K. D., & Goodey, E. (2004). Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress and levels of cortisol, dehydroepiandrosterone sulfate (DHEAS) and melatonin in breast and prostate cancer outpatients. Psychoneuroendocrinology, 29(4), 448–474. Carroll, B. J., Feinberg, M., Greden, J. F., Tarika, J., Albala, A. A., Haskett, R. F., et al. (1981). A specific laboratory test for the diagnosis of melancholia. Standardization, validation, and clinical utility. Archives of General Psychiatry, 38(1), 15–22. Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., Harrington, H., et al. (2003). Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science, 301(5631), 386–389. Charmandari, E., Tsigos, C., & Chrousos, G. (2005). Endocrinology of the stress response. Annual Review of Physiology, 67, 259–284. Clements, A. D., & Parker, C. R. (1998). The relationship between salivary cortisol concentrations in frozen versus mailed samples. Psychoneuroendocrinology, 23(6), 613–616. Clow, A., Thorn, L., Evans, P., & Hucklebridge, F. (2004). The awakening cortisol response: Methodological issues and significance. Stress, 7(1), 29–37. Cohen, S., Schwartz, J. E., Epel, E., Kirschbaum, C., Sidney, S., & Seeman, T. (2006). Socioeconomic status, race, and diurnal cortisol decline in the Coronary Artery Risk Development in Young Adults (CARDIA) Study. Psychosomatic Medicine, 68(1), 41–50. Cook, N. J., Ng, A., Read, G. F., Harris, B., & Riad-Fahmy, D. (1987). Salivary cortisol for monitoring adrenal activity during marathon runs. Hormone Research, 25, 18–23. Copinschi, G., Akseki, E., Moreno-Reyes, R., Leproult, R., L’Hermite-Baleriaux, M., Caufriez, A., et al. (1995). Effects of bedtime administration of zolpidem on circadian and sleep-related hormonal profiles in normal women. Sleep, 18, 417–424.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 65

Measurement of Cortisol Dallman, M. F., Akana, S. F., Bhatnagar, S., Bell, M. E., & Strack, A. M. (2000). Bottomed out: Metabolic significance of the circadian trough in glucocorticoid concentrations. International Journal of Obesity, 24, S40–S46. de Kloet, E. R. (1991). Brain corticosteroid receptor balance and homeostatic con­ trol. Frontiers in Neuroendocrinology, 12(2), 95–164. de Weerth, C., & Buitelaar, J. K. (2005). Physiological stress reactivity in human pregnancy—a review. Neuroscience & Biobehavioral Reviews, 29(2), 295–312. de Weerth, C., Graat, G., Buitelaar, J. K., & Thijssen, J. H. H. (2003). Measurement of cortisol in small quantities of saliva. Clinical Chemistry, 49(4), 658–660. Deinzer, R., Kirschbaum, C., Gresele, C., & Hellhammer, D. H. (1997). Adrenocortical responses to repeated parachute jumping and subsequent h-CRH challenge in inexperienced healthy subjects. Physiology & Behavior, 61(4), 507–511. DeRijk, R. H., Schaaf, M., & de Kloet, E. R. (2002). Glucocorticoid receptor variants: Clinical implications. Journal of Steroid Biochemistry & Molecular Biology, 81(2), 103–122. Dettenborn, L., James, G. D., van Berge-Landry, H., Valdimarsdottir, H. B., Montgomery, G. H., & Bovbjerg, D. H. (2005). Heightened cortisol responses to daily stress in working women at familial risk for breast cancer. Biological Psychology, 69(2), 167–179. Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychological Bulletin, 130(3), 355–391. Epel, E. S., McEwen, B., Seeman, T., Matthews, K., Castellazzo, G., Brownell, K. D., et al. (2000). Stress and body shape: Stress-induced cortisol secretion is consis­ tently greater among women with central fat. Psychosomatic Medicine, 62(5), 623–632. Etter, J. F., Neidhart, E., Bertrand, S., Malafosse, A., & Bertrand, D. (2005). Collecting saliva by mail for genetic and cotinine analyses in participants recruited through the Internet. European Journal of Epidemiology, 20(10), 833–838. Fink, R. S., Pierre, L. N., Daley-Yates, P. T., Richards, D. H., Gibson, A., & Honour, J. W. (2002). Hypothalamic-pituitary-adrenal axis function after inhaled corticosteroids: Unreliability of urinary free cortisol estimation. Journal of Clinical Endocrinology & Metabolism, 87(10), 4541–4546. Flinn, M. V. (1999). Family environment, stress, and health during childhood. In C. Panter-Brick & C. M. Worthman (Eds.), Hormones, health, and behavior: A socio-ecological and lifespan perspective (pp. 105–138). Cambridge, UK: Cambridge University Press. Frederick, S. L., Reus, V. I., Ginsberg, D., Hall, S. M., Munoz, R. F., & Ellman, G. (1998). Cortisol and response to dexamethasone as predictors of withdrawal distress and abstinence success in smokers. Biological Psychiatry, 43(7), 525–530. Fries, E., Hesse, J., Hellhammer, J., & Hellhammer, D. H. (2005). A new view on hypocortisolism. Psychoneuroendocrinology, 30(10), 1010–1016. Gaab, J., Blattler, N., Menzi, T., Pabst, B., Stoyer, S., & Ehlert, U. (2003). Randomized controlled evaluation of the effects of cognitive-behavioral stress management on cortisol responses to acute stress in healthy subjects. Psychoneuroendocrinology, 28(6), 767–779. Gaab, J., Huster, D., Peisen, R., Engert, V., Schad, T., Schurmeyer, T. H., et al. (2002). Low-dose dexamethasone suppression test in chronic fatigue syndrome and health. Psychosomatic Medicine, 64(2), 311–318.

65

03 - Luecken - 45370.qxd

66

9/12/2007

4:45 PM

Page 66

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Garde, A. H., & Hansen, A. M. (2005). Long-term stability of salivary cortisol. Scandinavian Journal of Clinical & Laboratory Investigation, 65(5), 433–436. Gerra, G., Zaimovic, A., Mascetti, G. G., Gardini, S., Zambelli, U., Timpano, M., et al. (2001). Neuroendocrine responses to experimentally-induced psycholog­ ical stress in healthy humans. Psychoneuroendocrinology, 26(1), 91–107. Gibson, E. L., Checkley, S., Papadopoulos, A., Poon, L., Daley, S., & Wardle, J. (1999). Increased salivary cortisol reliably induced by a protein-rich midday meal. Psychosomatic Medicine, 61, 214–224. Gilbert, D. G., Stunkard, M. E., Jensen, R. A., Detwiler, F. R. J., & Martinko, J. M. (1996). Effects of exam stress on mood, cortisol, and immune functioning: Influences of neuroticism and smoker-non-smoker status. Personality and Individual Differences, 21(2), 235–246. Gonzalez-Bono, E., Rohleder, N., Hellhammer, D. H., Salvador, A., & Kirschbaum, C. (2002). Glucose but not protein or fat load amplifies the cortisol response to psychosocial stress. Hormones and Behavior, 41(3), 328–333. Goodyer, I. M., Park, R. J., Netherton, C. M., & Herbert, J. (2001). Possible role of cortisol and dehydroepiandrosterone in human development and psy­ chopathology. British Journal of Psychiatry, 179, 243–249. Goodyer, I. M., Tamplin, A., Herbert, J., & Altham, P. M. (2000). Recent life events, cortisol, dehydroepiandrosterone and the onset of major depression in high-risk adolescents. British Journal of Psychiatry, 177, 499–504. Gouarne, C., Foury, A., & Duclos, M. (2004). Critical study of common conditions of storage of glucocorticoids and catecholamines in 24-h urine collected during resting and exercising conditions. Clinica Chimica Acta, 348(1–2), 207–214. Gouarne, C., Groussard, C., Gratas-Delamarche, A., Delamarche, P., & Duclos, M. (2005). Overnight urinary cortisol and cortisone add new insights into adaptation to training. Medicine and Science in Sports and Exercise, 37(7), 1157–1167. Grice, J. E., & Jackson, R. V. (2004). Letter to the Editor: Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology, 29(4), 563–564. Groschl, M., Wagner, R., Rauh, M., & Dorr, H. G. (2001). Stability of salivary steroids: The influences of storage, food and dental care. Steroids, 66(10), 737–741. Grossi, G., Perski, A., Ekstedt, M., Johansson, T., Lindstrom, M., & Holm, K. (2005). The morning salivary cortisol response in burnout. Journal of Psycho­ somatic Research, 59(2), 103–111. Gunnar, M. R., Hertsgaard, L., Larson, M., & Rigatuso, J. (1991). Cortisol and behavioral responses to repeated stressors in the human newborn. Develpmental Psychobiology, 24(7), 487–505. Gunnar, M. R., Morison, S. J., Chisholm, K., & Schuder, M. (2001). Salivary cor­ tisol levels in children adopted from Romanian orphanages. Developmental Psychopathology, 13(3), 611–628. Gunnar, M. R., & Talge, N. M. (2007). Neuroendocrine measures in developmen­ tal research. In L. A. Schmidt & S. J. Segalowitz (Eds.), Developmental psy­ chophysiology (pp. 343–366). Cambridge, UK: Cambridge University Press. Gunnar, M. R., & Vazquez, D. M. (2001). Low cortisol and a flattening of expected daytime rhythm: Potential indices of risk in human development. Develop­ mental Psychopathology, 13(3), 515–538.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 67

Measurement of Cortisol Hansen, A. M., Garde, A. H., Christensen, J. M., Eller, N. H., & Netterstrom, B. (2003). Evaluation of a radioimmunoassay and establishment of a reference interval for salivary cortisol in healthy subjects in Denmark. Scandinavian Journal of Clinical & Laboratory Investigation, 63(4), 303–310. Hanson, E. K., Maas, C. J., Meijman, T. F., & Godaert, G. L. (2000). Cortisol secretion throughout the day, perceptions of the work environment, and nega­ tive affect. Annals of Behavioral Medicine, 22(4), 316–324. Heim, C., Ehlert, U., & Hellhammer, D. H. (2000). The potential role of hypo­ cortisolism in the pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology, 25(1), 1–35. Herman, J. P., & Cullinan, W. E. (1997). Neurocircuitry of stress: Central control of the hypothalamo-pituitary-adrenocortical axis. Trends in Neurosciences, 20, 78–84. Hibel, L. C., Granger, D. A., Kivlighan, K. T., Blair, C., & The Family Life Project Investigators. (2006). Individual differences in salivary cortisol: Associations with common over-the-counter and prescription medication status in infants and their mothers. Hormones and Behavior, 50, 293–300. Houtman, I. L., & Bakker, F. C. (1987). Stress in student teachers during real and simulated standardized lectures. Journal of Human Stress, 13(4), 180–187. Hruschka, D. J., Kohrt, B. A., & Worthman, C. M. (2005). Estimating betweenand within-individual variation in cortisol levels using multilevel models. Psychoneuroendocrinology, 30(7), 698–714. Hucklebridge, F., Mellins, J., Evans, P., & Clow, A. (2002). The awakening corti­ sol response—No evidence for an influence of body posture. Life Sciences, 71(6), 639–646. IBL. (2005). IBL quality assessment scheme (QAS) for salivary steroids. Retrieved August 10, 2007, from http://www.ibl-hamburg.com/ IBL. (2006). Saliva diagnostics. Version 2, 05/2006. Retrieved August 10, 2007, from http://www.ibl-hamburg.com/ Jacks, D. E., Sowash, J., Anning, J., McGloughlin, T., & Andres, F. (2002). Effect of exercise at three exercise intensities on salivary cortisol. Journal of Strength and Conditioning Research, 16(2), 286–289. Jacobs, N., Nicolson, N. A., Derom, C., Delespaul, P., van Os, J., & Myin-Germeys, I. (2005). Electronic monitoring of salivary cortisol sampling compliance in daily life. Life Sciences, 76(21), 2431–2443. Jerjes, W. K., Peters, T. J., Taylor, N. F., Wood, P. J., Wessely, S., & Cleare, A. J. (2006). Diurnal excretion of urinary cortisol, cortisone, and cortisol metabolites in chronic fatigue syndrome. Journal of Psychosomatic Research, 60(2), 145–153. Jessop, D. S., Dallman, M. F., Fleming, D., & Lightman, S. L. (2001). Resistance to glucocorticoid feedback in obesity. Journal of Clinical Endocrinology & Metababolism, 86(9), 4109–4114. Kajantie, E., & Phillips, D. I. (2006). The effects of sex and hormonal status on the physiological response to acute psychosocial stress. Psychoneuroendocrinology, 31(2), 151–178. Khorram, O. (1996). DHEA: A hormone with multiple effects. Current Opinion in Obstetrics and Gynecology, 8, 351–354. King, A. C., Houle, T., de Wit, H., Holdstock, L., & Schuster, A. (2002). Biphasic alcohol response differs in heavy versus light drinkers. Alcoholism: Clinical and Experimental Research, 26(6), 827–835.

67

03 - Luecken - 45370.qxd

68

9/12/2007

4:45 PM

Page 68

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Kirschbaum, C., & Hellhammer, D. H. (1989). Salivary cortisol in psychobiologi­ cal research: An overview. Neuropsychobiology, 22, 150–169. Kirschbaum, C., & Hellhammer, D. H. (1994). Salivary cortisol in psychoneuroen­ docrine research: Recent developments and applications. Psychoneuroendo­ crinology, 19(4), 313–333. Kirschbaum, C., & Hellhammer, D. H. (2000). Salivary cortisol. In G. Fink (Ed.), Encyclopedia of stress (Vol. 3, pp. 379–383). New York: Academic Press. Kirschbaum, C., Klauer, T., Filipp, S.-H., & Hellhammer, D. H. (1995). Sexspecific effects of social support on cortisol and subjective responses to acute psychological stress. Psychosomatic Medicine, 57, 23–31. Kirschbaum, C., Pirke, K.-M., & Hellhammer, D. H. (1993). The “Trier Social Stress Test”—a tool for investigating psychobiological stress responses in a lab­ oratory setting. Neuropsychobiology, 28, 76–81. Kirschbaum, C., Prüssner, J. C., Stone, A. A., Federenko, I., Gaab, J., Lintz, D., et al. (1995). Persistent high cortisol responses to repeated psychological stress in a subpopulation of healthy men. Psychosomatic Medicine, 57, 468–474. Kirschbaum, C., Strasburger, C. J., & Langkrar, J. (1993). Attenuated cortisol response to psychological stress but not to CRH or ergometry in young habit­ ual smokers. Pharmacology Biochemistry and Behavior, 44(3), 527–531. Kirschbaum, C., Wust, S., & Strasburger, C. J. (1992). “Normal” cigarette smok­ ing increases free cortisol in habitual smokers. Life Sciences, 50(6), 435–442. Kivlighan, K. T., Granger, D. A., Schwartz, E. B., Nelson, V., Curran, M., & Shirtcliff, E. A. (2004). Quantifying blood leakage into the oral mucosa and its effects on the measurement of cortisol, dehydroepiandrosterone, and testos­ terone in saliva. Hormones and Behavior, 46(1), 39–46. Kokavec, A., & Crowe, S. F. (2001). The effect of a moderate level of white wine consumption on the hypothalamic-pituitary-adrenal axis before and after a meal. Pharmacoogy Biochemistry and Behavior, 70(2-3), 243–250. Kraemer, H. C., Giese-Davis, J., Yutsis, M., O’Hara, R., Neri, E., GallagherThompson, D., et al. (2006). Design decisions to optimize reliability of daytime cortisol slopes in an older population. American Journal of Geriatric Psychiatry, 14, 325–333. Kudielka, B. M., Broderick, J. E., & Kirschbaum, C. (2003). Compliance with saliva sampling protocols: Electronic monitoring reveals invalid cortisol daytime pro­ files in noncompliant subjects. Psychosomatic Medicine, 65(2), 313–319. Kudielka, B. M., Buske-Kirschbaum, A., Hellhammer, D. H., & Kirschbaum, C. (2004). HPA axis responses to laboratory psychosocial stress in healthy elderly adults, younger adults, and children: Impact of age and gender. Psychoneuroendocrinology, 29(1), 83–98. Kudielka, B. M., & Kirschbaum, C. (2005). Sex differences in HPA axis responses to stress: A review. Biological Psychology, 69(1), 113–132. Kudielka, B. M., Schommer, N. C., Hellhammer, D. H., & Kirschbaum, C. (2004). Acute HPA axis responses, heart rate, and mood changes to psychosocial stress (TSST) in humans at different times of day. Psychoneuroendocrinology, 29(8), 983–992. Kunz-Ebrecht, S. R., Kirschbaum, C., Marmot, M., & Steptoe, A. (2004). Differences in cortisol awakening response on work days and weekends in women and men from the Whitehall II cohort. Psychoneuroendocrinology, 29(4), 516–528.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 69

Measurement of Cortisol Kunzel, H. E., Binder, E. B., Nickel, T., Ising, M., Fuchs, B., Majer, M., et al. (2003). Pharmacological and nonpharmacological factors influencing hypotha­ lamic-pituitary-adrenocortical axis reactivity in acutely depressed psychiatric in-patients, measured by the DEX-CRH test. Neuropsychopharmacology, 28(12), 2169–2178. Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Leproult, R., Copinschi, G., Buxton, O., & Van Cauter, E. (1997). Sleep loss results in an elevation of cortisol levels the next evening. Sleep, 20(10), 865–870. Levine, S. (1978). Cortisol changes following repeated experiences with parachute training. In H. Ursin & S. Levine (Eds.), Psychobiology of stress: A study of coping men (pp. 51–56). New York: Academic Press. Levine, S. (2000). Influence of psychological variables on the activity of the hypothalamic-pituitary-adrenal axis. European Journal of Pharmacology, 405(1–3), 149–160. Lindley, S. E., Carlson, E. B., & Benoit, M. (2004). Basal and dexamethasone sup­ pressed salivary cortisol concentrations in a community sample of patients with posttraumatic stress disorder. Biological Psychiatry, 55(9), 940–945. Linkowski, P., Mendlewicz, J., Leclercq, R., Brasseur, M., Hubain, P., Golstein, J., et al. (1985). The 24-hour profile of adrenocorticotropin and cortisol in major depressive illness. Journal of Clinical Endocrinology & Metabolism, 61(3), 429–438. Ljung, T., Holm, G., Friberg, P., Andersson, B., Bengtsson, B. A., Svensson, J., et al. (2000). The activity of the hypothalamic-pituitary-adrenal axis and the sympa­ thetic nervous system in relation to waist/hip circumference ratio in men. Obesity Research, 8(7), 487–495. Lovallo, W. R., Al’Absi, M., Blick, K., Whitsett, T. L., & Wilson, M. F. (1996). Stress-like adrenocorticotropin responses to caffeine in young healthy men. Pharmacology Biochemistry and Behavior, 55(3), 365–369. Lovallo, W. R., Whitsett, T. L., Al’Absi, M., Sung, B. H., Vincent, A. S., & Wilson, M. F. (2005). Caffeine stimulation of cortisol secretion across the waking hours in relation to caffeine intake levels. Psychosomatic Medicine, 67(5), 734–739. Lundberg, U., Melin, B., Fredrikson, M., Tuomisto, M., & Frankenhaeuser, M. (1990). Comparison of neuroendocrine measurements under laboratory and naturalistic conditions. Pharmacology Biochemistry and Behavior, 37(4), 697–702. Masharani, U., Shiboski, S., Eisner, M. D., Katz, P. P., Janson, S. L., Granger, D. A., et al. (2005). Impact of exogenous glucocorticoid use on salivary cortisol measurements among adults with asthma and rhinitis. Psychoneuroendocrinol­ ogy, 30(8), 744–752. Mason, J. W. (1968). A review of psychoendocrine research on the pituitary-adrenal cortical system. Psychosomatic Medicine, 30, 576–607. McCann, S. J., Gillingwater, S., & Keevil, B. G. (2005). Measurement of urinary free cortisol using liquid chromatography-tandem mass spectrometry: Comparison with the urine adapted ACS: 180 serum cortisol chemiluminescent immunoassay and development of a new reference range. Annals of Clinical Biochemistry, 42, 112–118. McDade, T. W., Stallings, J. F., Angold, A., Costello, E. J., Burleson, M., Cacioppo, J. T., et al. (2000). Epstein-Barr virus antibodies in whole blood spots: A min­ imally invasive method for assessing an aspect of cell-mediated immunity. Psychosomatic Medicine, 62(4), 560–567.

69

03 - Luecken - 45370.qxd

70

9/12/2007

4:45 PM

Page 70

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL McEwen, B. S. (2003). Interacting mediators of allostasis and allostatic load: Towards an understanding of resilience in aging. Metabolism, 52(10 Suppl 2), 10–16. Mendel, C. M. (1989). The free hormone hypothesis: A physiologically based math­ ematical model. Endocrine Reviews, 10(3), 232–274. Munck, A. (2000). Corticosteroids and stress. In G. Fink (Ed.), Encyclopedia of stress (Vol. 1, pp. 570–577): New York: Academic Press. Murphy, B. E. P. (2002). Urinary free cortisol determinations: What they measure. Endocrinologist, 12(2), 243–150. Murphy, B. E. P. (2003). Urinary free cortisol levels in PTSD offspring. Psychoneuroendocrinology, 28(4), 594–595. Netherton, C., Goodyer, I., Tamplin, A., & Herbert, J. (2004). Salivary cortisol and dehydroepiandrosterone in relation to puberty and gender. Psychoneuroen­ docrinology, 29(2), 125–140. Nicolson, N. A. (1992). Stress, coping and cortisol dynamics in daily life. In M. W. deVries (Ed.), The experience of psychopathology: Investigating mental disorders in their natural settings (pp. 219–232). Cambridge, UK: Cambridge University Press. Nicolson, N., Storms, C., Ponds, R., & Sulon, J. (1997). Salivary cortisol levels and stress reactivity in human aging. Journal of Gerontology: Medical Sciences, 52A(2), M68–M75. Nimmagudda, R. R., Ramanathan, R., & Putcha, L. (1997). A method for pre­ serving saliva samples at ambient temperatures. Biochemical Archives, 13(3), 171–178. Olff, M., Meewisse, M. L., Kleber, R. J., van der Velden, P. G., Drogendijk, A. N., van Amsterdam, J. G., et al. (2006). Tobacco usage interacts with postdisaster psychopathology on circadian salivary cortisol. International Journal of Psycho­ physiology, 59(3), 251–258. Otte, C., Hart, S., Neylan, T. C., Marmar, C. R., Yaffe, K., & Mohr, D. C. (2005). A meta-analysis of cortisol response to challenge in human aging: importance of gender. Psychoneuroendocrinology, 30(1), 80–91. Peeters, F., Nicholson, N. A., & Berkhof, J. (2003). Cortisol responses to daily events in major depressive disorder. Psychosomatic Medicine, 65, 836–841. Peeters, F., Nicolson, N. A., & Berkhof, J. (2004). Levels and variability of daily life cortisol secretion in major depression. Psychiatry Research, 126(1), 1–13. Pitman, D. L., Ottenweller, J. E., & Natelson, B. H. (1990). Effect of stressor intensity on habituation and sensitization of glucocorticoid responses in rats. Behavioral Neuroscience, 104(1), 28–36. Polk, D. E., Cohen, S., Doyle, W. J., Skoner, D. P., & Kirschbaum, C. (2005). State and trait affect as predictors of salivary cortisol in healthy adults. Psychoneuroendocrinology, 30(3), 261–272. Posener, J. A., Charles, D., Veldhuis, J. D., Province, M. A., Williams, G. H., & Schatzberg, A. F. (2004). Process irregularity of cortisol and adrenocorticotropin secretion in men with major depressive disorder. Psychoneuroendocrinology, 29(9), 1129–1137. Posener, J. A., DeBattista, C., Williams, G. H., Kraemer, H. C., Kalehzan, B. M., & Schatzberg, A. F. (2000). 24-hour monitoring of cortisol and corticotropin secretion in psychotic and nonpsychotic major depression. Archives of General Psychiatry, 57(8), 755–760.

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 71

Measurement of Cortisol Powell, L. H., Lovallo, W. R., Matthews, K. A., Meyer, P., Midgley, A. R., Baum, A., et al. (2002). Physiologic markers of chronic stress in premenopausal, middle-aged women. Psychosomatic Medicine, 64(3), 502–509. Pruessner, M., Hellhammer, D. H., Pruessner, J. C., & Lupien, S. J. (2003). Selfreported depressive symptoms and stress levels in healthy young men: Associations with the cortisol response to awakening. Psychosomatic Medicine, 65(1), 92–99. Pruessner, J. C., Kirschbaum, C., Meinlschmid, G., & Hellhammer, D. H. (2003). Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology, 28(7), 916–931. Pruessner, J. C., Wolf, O. T., Hellhammer, D. H., Buske-Kirschbaum, A., von Auer, K., Jobst, S., et al. (1997). Free cortisol levels after awakening: A reliable biological marker for the assessment of adrenocortical activity. Life Sciences, 61(26), 2539–2549. Quinlan, P., Lane, J., & Aspinall, L. (1997). Effects of hot tea, coffee and water ingestion on physiological responses and mood: the role of caffeine, water and beverage type. Psychopharmacology (Berlin), 134(2), 164–173. Ranjit, N., Young, E. A., Raghunathan, T. E., & Kaplan, G. A. (2005). Modeling cortisol rhythms in a population-based study. Psychoneuroendocrinology, 30(7), 615–624. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage. Reynolds, R. M., Fischbacher, C., Bhopal, R., Byrne, C. D., White, M., Unwin, N., et al. (2006). Differences in cortisol concentrations in South Asian and European men living in the United Kingdom. Clinical Endocrinology (Oxford), 64, 530–534. Riad-Fahmy, D., Read, G. F., Walker, R. F., & Griffiths, K. (1982). Steroids in saliva for assessing endocrine function. Endocrine Reviews, 3(4), 367–395. Rohleder, N., & Kirschbaum, C. (2006). The hypothalamic-pituitary-adrenal (HPA) axis in habitual smokers. International Journal of Psychophysiology, 59, 236–243. Rose, R. M. (1984). Overview of endocrinology of stress. In G. M. Brown, S. H. Koslow & S. Reichlin (Eds.), Neuroendocrinology and psychiatric disor­ der (pp. 95–122). New York: Raven Press. Rosmalen, J. G., Oldehinkel, A. J., Ormel, J., de Winter, A. F., Buitelaar, J. K., & Verhulst, F. C. (2005). Determinants of salivary cortisol levels in 10-12 year old children: A population-based study of individual differences. Psychoneuroen­ docrinology, 30(5), 483–495. Rosmond, R., & Björntorp, P. (2001). Alterations in the hypothalamic-pituitary­ adrenal axis in metabolic syndrome. Endocrinologist, 11(6), 491–497. Schmidt-Reinwald, A., Pruessner, J. C., Hellhammer, D. H., Federenko, I., Rohleder, N., Schürmeyer, T. H., et al. (1999). The cortisol response to awak­ ening in relation to different challenge tests and a 12-hour cortisol rhythm. Life Sciences, 64(18), 1653–1660. Schommer, N. C., Kudielka, B. M., Hellhammer, D. H., & Kirschbaum, C. (1999). No evidence for a close relationship between personality traits and circadian cortisol rhythm or a single cortisol stress response. Psychological Reports, 84(3 Pt 1), 840–842.

71

03 - Luecken - 45370.qxd

72

9/12/2007

4:45 PM

Page 72

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Schwartz, E. B., Granger, D. A., Susman, E. J., Gunnar, M. R., & Laird, B. (1998). Assessing salivary cortisol in studies of child development. Child Development, 69(6), 1503–1513. Selye, H. (1956). The stress of life (1st ed.). New York: McGraw Hill. Sephton, S. E., Sapolsky, R. M., Kraemer, H. C., & Spiegel, D. (2000). Diurnal cor­ tisol rhythm as a predictor of breast cancer survival. Journal of the National Cancer Institute, 92(12), 994–1000. Sher, L., Oquendo, M. A., Galfalvy, H. C., Zalsman, G., Cooper, T. B., & Mann, J. J. (2005). Higher cortisol levels in spring and fall in patients with major depression. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 29(4), 529–534. Shirtcliff, E. A., Granger, D. A., Schwartz, E., & Curran, M. J. (2001). Use of sali­ vary biomarkers in biobehavioral research: Cotton-based sample collection methods can interfere with salivary immunoassay results. Psychoneuroen­ docrinology, 26(2), 165–173. Shirtcliff, E. A., Reavis, R., Overman, W. H., & Granger, D. A. (2001). Measurement of gonadal hormones in dried blood spots versus serum: Verification of menstrual cycle phase. Hormones and Behavior, 39(4), 258–266. Slag, M. F., Ahmed, M., Gannon, M. C., & Nuttall, F. Q. (1981). Meal stimulation of cortisol secretion: A protein induced effect. Metabolism, 30(11), 1104–1108. Smyth, J. M., Ockenfels, M. C., Gorin, A. A., Catley, D., Porter, L. S., Kirschbaum, C., et al. (1997). Individual differences in the diurnal cycle of cortisol. Psychoneuroendocrinology, 22(2), 89–105. Smyth, J., Ockenfels, M. C., Porter, L., Kirschbaum, C., Hellhammer, D. H., & Stone, A. A. (1998). Stressors and mood measured on a momentary basis are associated with salivary cortisol secretion. Psychoneuroendocrinology, 23(4), 353–370. Steptoe, A., & Ussher, M. (2006). Smoking, cortisol and nicotine. International Journal of Psychophysiology, 59, 228–235. Stetler, C., Dickerson, S. S., & Miller, G. E. (2004). Uncoupling of social zeitgebers and diurnal cortisol secretion in clinical depression. Psychoneuroendocrinol­ ogy, 29(10), 1250–1259. Stetler, C., & Miller, G. E. (2005). Blunted cortisol response to awakening in mild to moderate depression: Regulatory influences of sleep patterns and social con­ tacts. Journal of Abnormal Psychology, 114(4), 697–705. Stewart, J., & Seeman, T. (2000). Salivary cortisol measurement. Summary of a meeting of the John D. and Catherine T. MacArthur Research Network on Socioeconomic Status and Health. Retrieved March 23, 2006, from http://www.macses.ucsf.edu/Research/Allostatic/notebook/salivarycort.html Strazdins, L., Meyerkort, S., Brent, V., D’Souza, R. M., Broom, D. H., & Kyd, J. M. (2005). Impact of saliva collection methods on sIgA and cortisol assays and acceptability to participants. Journal of Immunological Methods, 307(1-2), 167–171. Stroud, L. R., Salovey, P., & Epel, E. S. (2002). Sex differences in stress responses: Social rejection versus achievement stress. Biological Psychiatry, 52(4), 318–327. Talge, N. M., Donzella, B., Kryzer, E. M., Gierens, A., & Gunnar, M. R. (2005). It’s not that bad: Error introduced by oral stimulants in salivary cortisol research. Developmental Psychobiology, 47(4), 369–376. Taylor, S. E., Gonzaga, G. C., Klein, L. C., Hu, P., Greendale, G. A., & Seeman, T. E. (2006). Relation of oxytocin to psychological stress responses and

03 - Luecken - 45370.qxd

9/12/2007

4:45 PM

Page 73

Measurement of Cortisol hypothalamic-pituitary-adrenocortical axis activity in older women. Psycho­ somatic Medicine, 68(2), 238–245. Touitou, Y., Sulon, J., Bogdan, A., Reinberg, A., Sodoyez, J., & Demey-Ponsart, E. (1983). Adrenocortical hormones, ageing and mental condition: Seasonal and circadian rhythms of plasma 18-hydroxy-11-deoxycorticosterone, total and free cortisol and urinary corticosteroids. Journal of Endocrinology, 96, 53–64. Tsuda, A., Steptoe, A., West, R., Fieldman, G., & Kirschbaum, C. (1996). Cigarette smoking and psychophysiological stress responsiveness: Effects of recent smok­ ing and temporary abstinence. Psychopharmacology, 126(3), 226–233. Turpeinen, U., & Stenman, U. H. (2003). Determination of urinary free cortisol by liquid chromatography-tandem mass spectrometry. Scandinavian Journal of Clinical & Laboratory Investigation, 63(2), 143–150. Van Cauter, E., Leproult, R., & Kupfer, D. J. (1996). Effects of gender and age on the levels and circadian rhythmicity of plasma cortisol. Journal of Clinical Endocrinology and Metabolism, 81(7), 2468–2473. van Eck, M. M., Berkhof, H., Nicolson, N., & Sulon, J. (1996). The effects of per­ ceived stress, traits, mood states, and stressful daily events on salivary cortisol levels. Psychosomatic Medicine, 58(5), 447–458. van Eck, M. M., Nicolson, N. A., Berkhof, H., & Sulon, J. (1996). Individual dif­ ferences in cortisol responses to a laboratory speech task and their relationship to responses to stressful daily events. Biological Psychology, 43, 69–84. van Goozen, S. H., Matthys, W., Cohen-Kettenis, P. T., Gispen-de Wied, C., Wiegant, V. M., & van Engeland, H. (1998). Salivary cortisol and cardiovas­ cular activity during stress in oppositional-defiant disorder boys and normal controls. Biological Psychiatry, 43(7), 531–539. Vialard-Miguel, J., Belaidi, N., Lembeye, L., & Corcuff, J. B. (2005). Lemon juice alters cortisol assays in saliva. Clinical Endocrinology, 63(4), 478–479. Vining, R. F., & McGinley, R. A. (1986). Hormones in saliva. CRC Critical Reviews in Clinical Laboratory Sciences, 23, 95–145. Vining, R. F., McGinley, R. A., Maksvytis, J. J., & Ho, K. Y. (1983). Salivary cor­ tisol: A better measure of adrenal cortical function than serum cortisol. Annals of Clinical Biochemistry, 20, 329–335. Walker, R. F. (1984). Salivary cortisol determinations in the assessment of adrenal activity. Frontiers of Oral Physiology, 5, 33–50. Walker, R. F., Riad-Fahmy, D., & Read, G. F. (1978). Adrenal status assessed by direct radioimmunoassay of cortisol in whole saliva or parotid saliva. Clinical Chemistry, 24(9), 1460–1463. Williams, E., Magid, K., & Steptoe, A. (2005). The impact of time of waking and concurrent subjective stress on the cortisol response to awakening. Psychoneuroendocrinology, 30(2), 139–148. Wolkowitz, O. M., Brizendine, L., & Reus, V. I. (2000). The role of dehy­ droepiandrosterone (DHEA) in psychiatry. Psychiatric Annals, 30(2), 123–128. Wong, V., Yan, T., Donald, A., & McLean, M. (2004). Saliva and bloodspot corti­ sol: novel sampling methods to assess hydrocortisone replacement therapy in hypoadrenal patients. Clinical Endocrinology, 61(1), 131–137. Worthman, C. M., & Stallings, J. F. (1997). Hormone measures in finger-prick blood spot samples: New field methods for reproductive endocrinology. American Journal of Physical Anthropology, 104(1), 1–21. Wüst, S., DeRijk, R., Meijer, O., Federenko, I., Vreugdenhil, E., de Punder, K., et al. (2005). A common mineralocorticoid receptor gene variant is associated

73

03 - Luecken - 45370.qxd

74

9/12/2007

4:45 PM

Page 74

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL with cortisol and heart rate response to psychosocial stress. Journal of Psychophysiology, 19(2), 154–155. Wüst, S., Federenko, I. S., van Rossum, E. F., Koper, J. W., & Hellhammer, D. H. (2005). Habituation of cortisol responses to repeated psychosocial stressfurther characterization and impact of genetic factors. Psychoneuroendocrinology, 30(2), 199–211. Wüst, S., Wolf, J., Hellhammer, D. H., Federenko, I., Schommer, N., & Kirschbaum, C. (2000). The cortisol awakening response—normal values and confounds. Noise & Health, 7, 77–85. Yehuda, R. (2002). Current status of cortisol findings in post-traumatic stress dis­ order. Psychiatric Clinics of North America, 25(2), 341–368, vii. Yehuda, R., Southwick, S. M., Krystal, J. H., Bremner, D., Charney, D. S., & Mason, J. W. (1993). Enhanced suppression of cortisol following dexametha­ sone administration in posttraumatic stress disorder. American Journal of Psychiatry, 150(1), 83–86. Yehuda, R., Teicher, M. H., Trestman, R. L., Levengood, R. A., & Siever, L. J. (1996). Cortisol regulation in posttraumatic stress disorder and major depres­ sion: A chronobiological analysis. Biogical Psychiatry, 40(2), 79–88. Young, E. A., Abelson, J., & Lightman, S. L. (2004). Cortisol pulsatility and its role

in stress regulation and health. Front Neuroendocrinology, 25(2), 69–76.

Young, E. A., & Breslau, N. (2004). Saliva cortisol in posttraumatic stress disorder:

A community epidemiologic study. Biological Psychiatry, 56(3), 205–209.

Young, E. A., & Nolen-Hoeksema, S. (2001). Effect of ruminations on the saliva cortisol response to a social stressor. Psychoneuroendocrinology, 26(3), 319–329.

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 75

CHAPTER

4

Sympathetic Hormones in Health Psychology Research PAUL J. MILLS MICHAEL G. ZIEGLER

INTRODUCTION AND BASIC CONCEPTS There is a long tradition of assessing sympa­ thetic nervous system activity in behavioral medicine and more recently in health psy­ chology research (Frankenhaeuser et al., 1968; Weiner, 1972). The sympathetic ner­ vous system has far-ranging relevance to these health research endeavors, including areas of acute and chronic stress, mood and perception, and psychiatric disorders. The most common way of assessing sympathetic activity is to determine circulating or excreted levels of the primary sympa­ thetic neurohormones norepinephrine and epinephrine. Assessing these two catechola­ mines in blood and urine is relatively easy compared to more invasive ways of assessing sympathetic activity, such as microneurogra­ phy, which provides a direct measure of the rate of sympathetic neural firing (Esler et al., 2003). Despite certain limitations of assessing catecholamines in plasma, such as poten­ tial confounding effects of clearance (to be

discussed later) and limitations of antecubital venous sampling, circulating levels of nore­ pinephrine provide a good assessment of sym­ pathetic nervous system activity. Resting subject’s supine venous plasma norepinephrine levels, for example, exhibit high correlations with direct electrical recordings of sympa­ thetic nerve activity (r’s = 0.76 – 0.87) (Wallin, Thompson, Jennings, & Esler, 1996). In addition, levels of plasma catecholamines following challenges such as mental stressors or exercise reflect sympathetic responsiveness (Dimsdale & Ziegler, 1991; Wilkinson et al., 1998). Among the primary reasons that nore­ pinephrine and epinephrine are of interest to health psychology research are their regu­ latory effects on the cardiovascular and immune systems. When catecholamines are chronically elevated, as in disease states such as pheochromocytoma (an adrenal cate­ cholamine-secreting tumor) and heart failure and in states of chronic stress and psychiatric disorders such as PTSD, pathologic effects can be observed in the heart, vascular 75

04 - Luecken - 45370.qxd

76

9/12/2007

4:44 PM

Page 76

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

endothelium, skin, and immune cells. Catecholamines are therefore a good choice for health psychology research because, in addition to helping us understand basic hormonal mechanisms of day-to-day physio­ logical regulation, they also provide the opportunity to understand mechanisms of many diseases and disorders. This chapter focuses on the cate­ cholamines norepinephrine and epinephrine. In addition to discussing assay techniques to measure their levels in blood and urine, we discuss other relevant topics to health psy­ chology research, including sample collection and processing, catecholamine stability, the timing of obtaining samples for assaying, the rhythms and interindividual variability of catecholamines, and best practice and pub­ lishing guidelines for catecholamine data. We will also briefly discuss two other methodolo­ gies relevant to sympathetic activity that are conceptually important for those interested in pursuing catecholamine measurement: nore­ pinephrine kinetics and adrenergic receptors.

CATECHOLAMINE PHYSIOLOGY Despite the fact that catecholamine studies have always faced assay limitations, low physiological concentrations, and a short half-life, circulating catecholamine levels have routinely been taken as indices of sym­ pathetic neural and adrenomedullary activity. Historically, norepinephrine has been under­ stood as a classic neurotransmitter of sympa­ thetic nerve endings and epinephrine as a hormone coming exclusively from the adrenal medulla, but studies show that epinephrine is also produced and released from other tissue, including the kidney, atria, and red blood cells (Kennedy, Elayan, & Ziegler, 1990; Ziegler et al., 2002). In addition to the blood, norepinephrine and epinephrine can be mea­ sured in the cerebrospinal fluid and in urine. Norepinephrine is synthesized by the enzyme

dopamine β-hydroxylase (DBH) and epine­ phrine is synthesized from norepinephrine by phenylethanolamine N-methyltransferase (PNMT). Norepinephrine and epinephrine exert their physiological effects by stimu­ lating α- and β-adrenergic receptors. Catecholamines in the circulation have a half-life of only a few minutes. Blood levels of epinephrine provide a good guide to adrenomedullary stimulation.

CATECHOLAMINE ASSAY TECHNIQUES Catecholamines are present in human plasma in a concentration of less than one part per billion. Because it is difficult to assay these tiny amounts, some assay systems have had problems with reliability and reproducibility. The first fluorometric assays, for example, were plagued by interfering compounds and gave plasma norepinephrine levels 10 to 100 times too high. This section provides infor­ mation necessary to evade the common pit­ falls in catecholamine measurements. All of the currently available assay tech­ niques have idiosyncratic situations that can negatively affect their performance, so it is necessary to know something about the assay techniques just to read the catecholamine literature adequately. As indicated in the following pages, the choice of an assay system to measure catecholamines depends on indi­ vidual needs. If, for example, only nore­ pinephrine is to be measured, the PNMT radioenzymatic technique is relatively rapid and sensitive. High performance liquid chro­ matography (HPLC) provides great flexibility in measuring catecholamines and metabolites. The catechol-O-methyltransferase (COMT) radioenzymatic assay’s sensitivity permits the use of smaller blood volumes than does HPLC. Newer immunoassays are promising and also require small sample volumes. With all of these assays, it is best to obtain the

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 77

Sympathetic Hormones in Health Psychology Research

advice of a laboratory with experience in a particular technique before using it.

Fluorometric Assay and Gas Chromatography Although fluorometric and gas chro­ matography assays are now rarely used, we briefly present these methods so that the health psychology researcher can be aware of them when reviewing the literature. Many chemicals absorb light of one wave­ length and then fluoresce by emitting light of another wavelength. Fluorometric assays quantitate these chemicals by measuring the intensity of the emitted light. Catecholamines fluoresce with an excitation maximum of 285 nm and emission maximum at 325 nm. Reaction of the catecholamines with trihy­ droxyindole or ethylenediamine enhances native fluorescence. Many materials in human plasma and urine fluoresce at similar wavelengths, providing falsely high estimates of catecholamines. Most of the cate­ cholamine assays in the literature published before 1970 were based on fluorescence tech­ niques and reported values about five times as high as with current assay methods, including radioenzymatic. As a consequence, fluorometric assays are now rarely used to measure plasma catecholamines, unless com­ bined with a separation technique such as HPLC (see later section). Gas chromatography is so time-consuming and expensive it is not really suitable for more routine catecholamine analysis. The cate­ cholamines are unsuitable for direct use in gas chromatographic separation but are small enough to allow volatile derivatives to be made that can be separated by gas chro­ matography. These derivatives can then be detected by flame ionization, electron cap­ ture, or mass spectroscopy. Electron capture is sensitive enough but lacks specificity; how­ ever, mass spectroscopy has both sensitivity and specificity sufficient to reliably detect

catecholamine levels. Gas chromatography with mass spectroscopy (GCMS) is very accu­ rate because it can be standardized by the addition of deuterated catecholamines, which differ slightly in molecular weight but in no other characteristics. GCMS provides a refer­ ence standard against other less rigorous procedures and has been used to verify the accuracy of the PNMT radioenzymatic assay.

Radioenzymatic Assays In a radioenzymatic assay the compound to be measured is incubated with a radioac­ tive substrate and an enzyme that catalyzes a reaction between them. The amount of radioactive metabolite formed from cate­ cholamine levels is proportional to the level of the compound initially present. The most pop­ ular radioenzymatic assay for catecholamines uses COMT, a nonspecific enzyme that 0­ methylates most small catechols by transferring a methyl group from S-adenosylmethionine (SAM). In the assay’s simplest form, the cate­ cholamines in unextracted plasma are incubated in a buffer solution with radiola­ beled SAM. COMT converts epinephrine to metanephrine, norepinephrine to normeta­ nephrine, and dopamine to 3-methoxytyra­ mine. A more specific form of the assay separates these metabolites by thin layer chromatography to measure individual cate­ cholamines. Many drugs such as isopro­ terenol, dobutamine, and methyldopa can interfere with the assay, but when the assay is combined with appropriate separation tech­ niques, these other catechol drugs can also be measured. Plasma proteins, some compounds in urine, aluminum, and ascorbic acid inter­ fere with the enzymatic activity of COMT, giving spuriously low catecholamine mea­ surements. A technique for extracting and concentrating catecholamines prior to assay (Kennedy & Ziegler, 1990) removes interfer­ ing compounds and increases sensitivity 10­ fold. The COMT assay for catecholamines is

77

04 - Luecken - 45370.qxd

78

9/12/2007

4:44 PM

Page 78

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

quite complex, making it subject to technical error; however, it has advantages in several situations. The assay can be performed on low sample volumes of 50 µl of plasma and 5 µl of urine and is sensitive enough to mea­ sure basal levels of plasma epinephrine, and it is applicable to measuring large numbers of samples at one time. Another radioenzymatic assay for cate­ cholamines is the phenylethanolamine-N­ methyltransferase (PNMT) assay. PNMT converts norepinephrine to epinephrine in the adrenal medulla. It can be purified from cow adrenal glands and 3H-SAM can be used as a methyl donor to convert norepinephrine to 3H-epinephrine. Unlike COMT, the enzyme PNMT is relatively specific for β-hydroxy­ lated phenylethanolamines so that it does not appreciably label dopamine or further label epinephrine. It, too, can be adapted to mea­ sure large numbers of samples. Since PNMT is not inhibited by aluminum, catechola­ mines can be concentrated on alumina prior to assay. This enables the assay to be very sensitive when large volumes of plasma are concentrated onto alumina. The cate­ cholamines are then eluted into 0.1 ml of acid solution. This preconcentration step eliminates inhibiting substances that might interfere with the assay so that standardiza­ tion of the PNMT assay is easier than that of the unextracted COMT assay. However, the very specificity of the PNMT assay limits the number of compounds it can measure.

High Performance Liquid Chromatography (HPLC) Assay Assays using HPLC separate cate­ cholamines or their metabolites and an internal standard into sharp peaks. After separation, the catecholamines can be detected by native fluorescence, by the fluorescence of their chemical derivatives, or by electrochemical detection. Reverse-phase HPLC columns have been used directly to separate catecholamines

but most frequently they have been modified using “soap” chromatography with the addi­ tion of sodium heptylsulfonate or sodium octylsulfate to the mobile phase. These hydrophobic, anionic detergents are strongly absorbed to the stationary phase and transform it into a cation exchange column. This column will separate neutral and anionic substances as well as catecholamines. Microparticulate cation exchange HPLC columns are popular for separating cate­ cholamines. Because catecholamines easily oxidize, they can be detected electrochemically when they are passed by a carbon electrode with an electrical potential in the range of +600 mV. The resulting electric current passing across the electrode is proportional to the amount of catecholamine present. This process provides detection limits in the range of 25 pg/ml so that when the system is performing optimally it can detect human plasma catecholamine levels. Catecholamines can also be detected by fluorescence with HPLC. Natural fluores­ cence of the catecholamines requires several nanograms for detection, but derivatized fluorescence techniques can greatly enhance sensitivity. Catecholamines may be deriva­ tized by several methods, including trihydrox­ yindole, ethylenediamine, or fluorescamine methods to enhance their fluorescence.

Immunoassays The enzyme-linked immunosorbent assay (ELISA) and the radioimmunoassay (RIA) work on the same principles. They use enzyme-linked antibodies (ELISA) or radioiso­ tope-labeled antibodies (RIA) that are spe­ cific for the analyte of interest, in this case, catecholamines. Immunoassays are easily the most widely used techniques in research lab­ oratories. Their methodologies are highly standardized, easy to learn, provide a high degree of specificity and sensitivity, and are relatively inexpensive. The specifics of these

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 79

Sympathetic Hormones in Health Psychology Research

types of assays are discussed in more detail in chapter 10 by Jain, Hong, Redwine, and Mills. We only briefly review the assay prin­ ciples and then review findings using these methods to assay catecholamines. ELISAs for measuring plasma and urine catecholamines have been developed rela­ tively recently. RIAs, on the other hand, have been available for several decades. Early RIAs showed poor performance in reliability, reporting very high basal levels and mini­ mal physiologic variation because of crossre­ activity by catecholamine sulfates. However, advances in the ability to develop mono­ clonal antibodies of high specificity and acetylation of catecholamines have improved the specificity, sensitivity, and reliability of these methods, and reduced crossreactivity with other compounds to less than 1%. Several studies have evaluated the perfor­ mance of these newer ELISA and RIA kits for determining catecholamines by comparing them to the reversed phase HPLC with elec­ trochemical detection method (Westermann, Hubl, Kaiser, & Salewski, 2002; Wassell, Reed, Kane, & Weinkove, 1999). The essence of the ELISA assay is that sam­ ples containing catecholamines are incubated with enzymes that are chemically conjugated to antibodies specific to the catecholamines of interest. The assay procedure includes the steps of sample extraction, chemical and enzymatic derivatization, and immunological reactions. The volume requirement for urine is 10 µL and 300 µL for plasma. The assay typically involves a series of two or three sep­ arate incubations, which can range from 20 minutes to 2 hours in duration depending on the sensitivity of the assay. The unbound antibody or antigen is removed by washing. Following the antigen-antibody reaction and washing, a chromogenic substrate of the enzyme is added and the enzyme changes the color or fluorescence of the substrate, demon­ strating that the antigen-antibody reaction has taken place. The intensity of the color of

the unknown sample is compared with a set of standards of known concentration and associated color intensity (i.e., a calibration or standard curve). One disadvantage of ELISA is that the antigen must be quite free of extran­ eous contaminating materials that may quench the antigen-antibody interaction. Comparison of values obtained from ELISA versus HPLC yields regression coefficients of 0.97 for epinephrine and norepinephrine in urine and regression coefficients of 0.93 for epinephrine and 0.95 for norepinephrine in plasma (Westermann et al., 2002). The epinephrine and norepinephrine values in plasma tend to be somewhat lower than obtained by HPLC. Such data indicate that ELISA can provide accurate values for epinephrine and norepinephrine concentrations compared to more gold standard methods. The principles of the RIA are identical with the ELISA except that radioactive com­ pounds are conjugated to antigens or anti­ bodies instead of enzymes. The concentration of the catecholamine in the unknown sample is obtained by incubation with a limited amount of anti-catecholamine antibody and a 125I-radiolabeled catecholamine. The unknown catecholamine competes for and inhibits binding of the radiolabeled tracer for the limited and constant number of binding sites on the antibody. For catecholamines, the incubation time of the RIA is much longer than the ELISA, up to 18 hours. After separating the antibody-bound from the free tracer and counting the bound fraction, the amount of radioactivity of each unknown sample is compared to a standard curve with increasing known amounts of antigen. Performance characteristics of RIA for cate­ cholamines are comparable to those of the ELISA reviewed earlier. Compared to other techniques, advantages of immunoassays such as ELISA and RIA include being less time-consuming, much eas­ ier to use, more economical to set up (rela­ tively inexpensive equipment to purchase and

79

04 - Luecken - 45370.qxd

80

9/12/2007

4:44 PM

Page 80

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

maintain), and requiring fairly small sample volumes. ASSAY RECOMMENDATIONS AND COSTS All catecholamine assays have shortcomings, so it is important to choose an assay that is appropriate for the sample being tested. Some antibody-based assays crossreact with catecholamine sulfates, which will show up as part of the total catecholamines measured. Those assays are of little use. Antibody-based assays that first derivatize the catecholamines can minimize this shortcoming. The major difficulty with current assays for plasma catecholamines is inadequate sen­ sitivity. This is particularly a problem with epinephrine levels, which are about onetenth of norepinephrine levels. Urine cate­ cholamines are 100 times as concentrated as basal plasma levels, so most assays are satis­ factory for urine. Despite the advertised sen­ sitivity of many commercial assays, most have difficulty reliably measuring basal plasma epinephrine levels of less than 25 pg/ml. Insensitive assays will also have diffi­ culty detecting catecholamine responses to mild stressors such as mental arithmetic. Most HPLC assays are reliable except those with electrochemical detection, which may have low sensitivity. The claimed sensitivity of ELISAs and RIAs is best verified by the

Table 4.1

user. An examination of the representative standard curves provided by commercially available assays provides better insight into assay sensitivity than advertised claims. The most sensitive assays are the HPLC methods that make postcolumn fluorescent deriva­ tives and radioenzymatic techniques that concentrate catecholamines prior to assay. These assays are typically carried out in spe­ cialized laboratories. If performance charac­ teristics of ELISA and RIA are shown to be equal, preference should be given to the ELISA because it needs no radioactive sub­ stances (which necessitate more strict envi­ ronmental restrictions and regulations). The cost of catecholamine assays vary widely. Most academic laboratories charge anywhere from $25 to $50 per sample. Commercial laboratories charge approxi­ mately $125 per sample. VALUES FOR PLASMA CATECHOLAMINES Table 4.1 shows typical normal plasma nore­ pinephrine and epinephrine values in a number of settings that are averaged from sev­ eral hundred healthy subjects. Catecholamines in the blood are typically reported in the U.S. literature in units of picograms per milliliter (pg/ml) and in the urine as micrograms per hour (µg/h) and elsewhere in International System of Units (SI).

Plasma Catecholamine Levels in Healthy Individuals

Condition

Norepinephrine (pg/ml)

Epinephrine (pg/ml)

Supine

282 (±56)

24 (±3)

Sitting

403 (±34)

31 (±13)

Standing

513 (±67)

46 (±10)

Giving a public speech

348 (±110)

70 (±81)

Cold pressor test

528 (±76)

39 (±61)

Exercise @ 65% V02 max

747 (±296)

98 (±83)

NOTE: Numbers indicate mean ± SD.

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 81

Sympathetic Hormones in Health Psychology Research

Dynamic changes in catecholamine concentrations have physiological effects. A small increase in a resting subject’s blood epinephrine levels from a low normal of 20 pg/ml to a high normal of 80 pg/ml is suffi­ cient to alter glucose metabolism. Norepine­ phrine is also present in the adrenal gland, but most blood norepinephrine comes from sympathetic nerves. Blood levels of nore­ pinephrine in the normal range for a resting, recumbent subject of 150 to 500 pg/ml have little physiologic effect, but blood levels of 1,000 pg/ml cause hemodynamic changes. Norepinephrine’s major effect follows its release from sympathetic nerves across a synapse onto adjacent adrenergic receptors. A small fraction of this norepinephrine finds its way into the bloodstream. Blood levels of catecholamines vary with circadian and ultradian rhythms, but mean levels are generally very consistent from week to week and month to month, thus provid­ ing a stable basis for longitudinal studies (Burleson et al., 2003; Mills, Berry, Dimsdale, Nelesen, & Ziegler, 1993). Naturally, cate­ cholamine levels will vary depending on the type of assay employed to assay them.

STRESSORS AND THE TIMING OF BLOOD SAMPLE COLLECTION There are numerous reasons why examining the catecholamine response to stressors is of interest to health psychology researchers. In addition to reflecting potential group and individual differences in sympathetic ner­ vous system responses to a certain type of stressor, examining catecholamine responses can also provide information on mecha­ nisms responsible for cardiovascular and immune responses to stressors. For example, the typical lymphocytosis (increase in number of lymphocytes in the peripheral cir­ culation) seen in response to acute stress is to a large extent driven by the increase in

catecholamines in response to the acute stress (Mills, Berry, et al., 1995). In stress studies, blood samples are usually obtained before the stressor is imposed (i.e., the resting or baseline sample), immediately after the stressor is imposed (i.e., the stress response sample), and for those interested in possible differences in sympathetic nervous system recovery from stress activation, some minutes following cessation of the stressor (i.e., the recovery sample). Regarding the recovery sample, catecholamines are cleared from the circulation in a matter of minutes following the cessation of a stressor; to be able to observe potential group or individual differences in recovery, therefore, samples should be gathered within 3 to 5 minutes or so. Regarding the resting sample, it is impor­ tant that it be drawn under conditions as close as possible to a true rest, or baseline condition. The fact that catecholamines are so exquisitely responsive to stressors means that even the stress of venipuncture can raise catecholamine levels in most people. Thus, an intravenous catheter is needed when sam­ pling blood for catecholamines, to enable blood to be sampled at a fixed time follow­ ing the stress of venipuncture and also noninvasive repeated sampling. Once the catheter is in place, it is common to wait 15 to 20 minutes for the person to recover from any stress response to the venipuncture before drawing blood for the resting sam­ ple. Policies on who can place an intravenous catheter—that is, a phlebotomist, a nurse, or a physician—vary from institution to institution. There are numerous types of stressors that investigators have used to examine catecho­ lamine (and other) physiological responses, including mathematics tasks, public speech stressors, video and role-playing games, aca­ demic examinations, exercise challenges, heat stress, marital interaction paradigms, and even parachute jumping. The type of task an investigator chooses really depends

81

04 - Luecken - 45370.qxd

82

9/12/2007

4:44 PM

Page 82

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

on the research question at hand. For example, studies have demonstrated nuances in norepinephrine versus epinephrine res­ ponses to different types of stressors. Dimsdale (1984) compared plasma catecholamines obtained after a laboratory mathematics task and a naturalistic public speaking task. The math task led to increases of 80 pg/ml and 42 pg/ml in norepinephrine and epinephrine, respectively, while public speaking led to increases of 300 pg/ml and 75 pg/ml in nore­ pinephrine and epinephrine, respectively. Goldstein and colleagues (1982) showed an average 24% (85 pg/ml) increase in nore­ pinephrine and 308% (194 pg/ml) increase in epinephrine following a real-life stressor (dental surgery). In general, any drug or stressor that alters heart rate or blood pres­ sure is likely to alter blood and urine cate­ cholamine levels as well. The timing of sample withdrawal also influences catecholamine measures in these settings. During the course of a 15­ minute public speaking task, epinephrine increases rapidly in the initial 3 minutes, but returns to baseline levels by the time the speaker finishes speaking (Dimsdale & Moss, 1980). Other studies show that in contrast to epinephrine, which reaches its peak in the first few minutes of a stressor, norepinephrine reaches its peak 6 to 8 minutes later, sometimes after the stressor is over. Thus, both the setting of a stressor (field versus laboratory) and the timing of sample acquisition can markedly affect the nature of the derived catecholamine data. A final and related consideration is the relation­ ship, or possible interaction, between the setting and the nature of the task. That is, a naturalistic stressor such as public speaking may be minimally evocative with friends, more evocative in a laboratory setting, and very stressful before an unsympathetic group.

CIRCADIAN AND ULTRADIAN RHYTHMS Plasma norepinephrine has a circadian rhythm with lowest levels at 3 a.m. Peak lev­ els double shortly after awakening and dou­ ble again upon standing (Nishihara, Mori, Endo, Ohta, & Ohara, 1985; Prinz, Halter, Benedetti, & Raskind, 1979). This sharp rise in norepinephrine in the early morning hours has been implicated in the peak of cardiovas­ cular mortality. Basal plasma epinephrine levels are quite low, making it more difficult to demonstrate the circadian rhythm for epinephrine. Epinephrine levels increase about 20% with upright posture. Studies of urinary secretion of catecholamines suggest that the epinephrine rhythm is present even in the absence of sleep. Although there is a significant correlation between the diurnal changes in epinephrine and norepinephrine, there are also separate factors controlling circulating levels of each compound. The ultradian fluctuation of plasma nore­ pinephrine in humans is very large, with the highest values typically twice as great as the lowest values obtained over a period of sev­ eral hours. The rhythm cycles over about 90 minutes, and the plasma norepinephrine and epinephrine levels are generally correlated. Generally, however, they have not been shown to correlate with blood pressure or even with heart rate. This may be explained by the observation that levels of plasma nore­ pinephrine and epinephrine alternate from one side of the body to the other and that norepinephrine levels drawn from one arm is frequently twice as high as norepinephrine levels from the contralateral arm (Kennedy, Ziegler, & Shannahoff-Khalsa, 1986). It is not possible at this time to make a clear recommendation whether the same arm should be used over time, since no systematic research has addressed that question. However, sampling at the same time of day

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 83

Sympathetic Hormones in Health Psychology Research

is optimal because plasma catecholamines have a circadian rhythm. Since this rhythm has only been shown for day versus night, a wide range of a given part of the day should be acceptable for catecholamine sampling.

CATECHOLAMINES IN URINE In addition to assessing circulating cate­ cholamine levels, some research paradigms can benefit from assessing them in urine. For healthy individuals, norepinephrine ranges in urine are 10 to 75 µg/24 hours and epinephrine ranges are 10 to 25 µg/24 hours. Urinary excretion of catecholamines is con­ sidered to provide a more integrated measure of sympathetic activation than circulating lev­ els. Catecholamines are much more concen­ trated in urine than plasma and are derived from filtered plasma, renal nerves, and renal catecholamine synthesis. Other potential sources for urinary catecholamines are circu­ lating catecholamine sulfates, which might be deconjugated by a renal sulfatase. The exact origin of urinary catecholamines is unknown, but it is apparent that urinary catecholamines do not simply reflect filtered plasma cate­ cholamines. Urinary levels of norepinephrine and epinephrine are low during nighttime, at rest, and increase gradually during morning hours, reaching a peak between 12 noon and 2 p.m. (Hansen, Garde, Skovgaard, & Christensen, 2001; Lakatua et al., 1987). Urine epinephrine has a somewhat more pro­ nounced diurnal pattern than does nore­ pinephrine and is relatively independent of sleep-wakefulness patterns. Urine nore­ pinephrine levels reflect variations in both posture and activity. The total amount of urinary catecholamines tends to increase with age. As with plasma levels, urinary norepinephrine levels increase with age, although the effect may be due to the loss of

physical fitness (Lehmann & Keul, 1986; Silverman & Mazzeo, 1996). During rest and relaxation there are generally no differences in urine catecholamines between men and women after adjusting for body surface area. Ideally, to obtain an integrated measure of catecholamines, urine would be collected over a 24-hour period. In our research studies, we typically have the participants collect their 24-hour urine in two separate aliquots—one from bedtime to awakening and one from awakening to bedtime. These samples pro­ vide information on sympathetic activity dur­ ing the night and during the day, respectively, or can be combined to provide a measure of 24-hour excretion. If 24-hour sampling is not feasible, one could gather either daytime or nighttime excretion only, depending on the research question at hand.

CATECHOLAMINE STABILITY, PROCESSING, AND STORAGE Unlike corticosteroids, which can sit at room temperature for hours if not days without appreciable decay, blood and urine samples obtained for catecholamine assaying require immediate processing. As far as major equip­ ment for blood sample processing, researchers need a refrigerated centrifuge and an ultra-low freezer capable of reaching temperatures of –70° to –80° C. An ultra-low freezer is needed because catecholamines are unstable at room temperature and appreciably deteriorate at 2° C in 2 hours and in a regular –20° C freezer in 2 weeks. Samples are stable for at least a year at –70° to –80° C. Storage in an antioxi­ dant or acid retards apparent degradation. Blood is typically preserved with ethylenedi­ aminetetraacetic acid (EDTA) or heparin and immediately put on ice. Ideally, then, samples should be processed within a few hours in a refrigerated centrifuge and stored at –70° to –80° C until assay.

83

04 - Luecken - 45370.qxd

84

9/12/2007

4:44 PM

Page 84

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

Catecholamine conjugates are excreted into urine, where they are present in rela­ tively high levels. Preservation of urine with acid, particularly if the sample is not kept cold, can lead to cleavage of these conjugates and erroneously high catecholamine mea­ surements. Antioxidants such as sodium metabisulfite or acid are adequate to prevent destruction of catecholamines in urine. Ideally, samples should be refrigerated imme­ diately, but studies suggest that samples are stable at room temperature for up to 24 hours, provided a preservative is used (Chan, Wee, & Ho, 2000). In contrast to blood, where nore­ pinephrine is stable at room temperature for at least 30 minutes, norepinephrine is extremely labile in cerebral spinal fluid (CSF). Ascorbic acid, which is both an antioxidant and an acidifying agent, helps preserve nore­ pinephrine in CSF, as does freezing. Samples collected on ice without preservative can lose three-fourths of their norepinephrine content in 1 hour. The best technique for collecting CSF is to have it drip directly from the lum­ bar puncture needle into a tube containing an antioxidant, with the collection tube placed in a container of dry ice. These exacting conditions have only rarely been followed in studies of CSF catecholamines; therefore, degradation of the catecholamines is an important factor to consider in evaluating the literature on CSF catecholamine levels.

BEST MEASUREMENT PRACTICES The vast majority of health psychology research uses urine and/or plasma samples for determining catecholamine levels. Procedures for obtaining CSF are extremely invasive compared to those for obtaining blood or urine. As previously discussed, urine samples are relatively easy to obtain and provide an integrated measure of sympathetic activation

over longer periods of time. Studies of chronic stress, for example, would be a good fit for collecting urine for catecholamine analysis. Blood catecholamine levels are especially useful for assessing the responsiveness of the sympathetic system to short-term perturba­ tions such as stressors. Urine samples would be inappropriate in such circumstances. Regarding relevant controls and/or poten­ tial confounds that should be considered when collecting catecholamine samples, Table 4.2 lists a number of factors that influ­ ence human catecholamine levels. Some of these factors will affect more basal levels but not responses to stressors and vice versa. For example, while sodium, hypertension, and race have been shown to have significant effects on resting levels of plasma nore­ pinephrine, these factors appear to have no effect on catecholamine reactivity (Dimsdale, Ziegler, Mills, Delehanty, & Berry, 1990). Suppressed anger may be associated with increased plasma norepinephrine responses to mental stressors, while increased expres­ sion of anger has been associated with decreased plasma norepinephrine responses to a mental stressor (Mills, Schneider, & Dimsdale, 1989; Suarez, Kuhn, Schanberg, Williams, & Zimmermann, 1998). Age has important effects on both the release of and response to norepinephrine. The resting levels of plasma norepinephrine in a 60-year-old man are approximately twice that of a 10-year-old child. In response to a mental stressor, older subjects will exhibit greater norepinephrine respon­ ses (approximately 40% higher) (Barnes, Raskind, Gumbrecht, & Halter, 1982; Jansen, Lenders, Thien, & Hoefnagels, 1989). Data suggest that age influences sym­ pathetic neural but not adrenomedullary responses since epinephrine responses show no significant age effects. Thus, age should be considered a potentially important source of individual variation in norepinephrine

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 85

Sympathetic Hormones in Health Psychology Research Table 4.2

Factors That Influence Human Catecholamine Levels

Age

Physical fitness

Caffeine consumption

Sleep disorders (e.g., sleep apnea)

Congestive heart failure

Spaceflight

Depression

Sodium intake

Environment of subject (external stimuli)

Stressors (acute and chronic)

Hypertension

Temperature

Medications (including diuretics, vasodilators,

β-antagonists, β-agonists, antidepressants,

amphetamines, neuroleptics)

Thyroid disease

Meditation and relaxation

Time of day

Posture and exercise

Venipuncture

reactivity studies. Another important factor to control for is caffeine intake (Lane, Adcock, Williams, & Kuhn, 1990; Papadelis et al., 2003). An equivalent caffeine con­ sumption of two to three cups of coffee per day can more than double epinephrine and cortisol responses to psychological stressors (Lane et al, 1990). Two cups of coffee (200 mg caffeine) doubles basal catecholamine levels in caffeine-naïve subjects but has little effect in regular coffee drinkers. Since caf­ feine consumption is so widespread, careful attention should be given to its control in reactivity studies. Regarding best assaying practices, it is important to use the same assay technique when assaying catecholamine samples from any one study; that is, don’t switch from one assay method to another during the course of a study. In the case of multiple samples/ subject, it is ideal if all of the samples from an individual are run within the same assay batch. Thus, if a single commercial ELISA assay kit can determine catecholamine levels in 42 unknown samples in duplicate, then when multiple samples are being obtained on each subject, all of the samples from a single

Tobacco smoking

subject should be run in that same assay. In addition, for the best results, samples should be run in duplicate or triplicate, if feasible. Regarding the selection of an assay method, the assay should have good repeatability coefficients. The interassay coefficient of variability provides information on the repro­ ducibility of the assay from one assay to the next. The intra-assay coefficient of variability provides information on the reproducibility of the assay within a single assay. Ideally, the interassay coefficient of variability should be less than 10% to 12% and the intra-assay coefficient of variability les than 5% to 7%.

GUIDELINES FOR PUBLICATION OF DATA Reviewers for scientific journals will want the following information presented in a manuscript: specifics on the setting and timing of blood and/or urine sampling, the preservative used, how quickly the samples were processed and at what temperature, the temperature at which the samples were stored prior to assay, the assay methodology

85

04 - Luecken - 45370.qxd

86

9/12/2007

4:44 PM

Page 86

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

used to determine the catecholamine levels, and the assay’s sensitivity and inter- and intra-assay coefficients of variation.

OTHER RELEVANT ASSAY METHODOLOGIES In addition to determining levels of cate­ cholamines in blood and urine, two other relevant methodologies are presented in the following sections. These techniques are not as readily accessible to health psychology investigators as those reviewed previously, but they are important to know about for their conceptual relevance to studying cate­ cholamines and their potential importance to clinical research. Depending on your research questions, it might be worth explor­ ing the potential availability of these meth­ ods at your local institutions.

Catecholamine Kinetics Once catecholamine concentrations are successfully measured in blood, the interpre­ tation of their value is not straightforward. Venous and arterial norepinephrine levels reflect not only release, but metabolic degra­ dation, reuptake, diffusion, and regional and local circulation. Catecholamines are cleared by reuptake into nerves (referred to as uptake 1) and by uptake into non-neuronal tissue (referred to as uptake 2). Catecholamines can stimulate their own clearance, and epinephrine is even more potent than norepinephrine in stimulating metabolic clearance of cate­ cholamines. Thus, catecholamine levels may only approximate the complex web of sym­ pathetic activity, a web whose many threads are themselves subject to regulation. There are methods to measure the release rate and the clearance rate of norepinephrine that pro­ vide insight into the factors that contribute to circulating levels. Before discussing these techniques in detail, we provide an example

of how useful the methodologies can be by presenting work we have done in patients with the sleep disorder obstructive sleep apnea. Example of the Usefulness of This Technique Our group has spent several years study­ ing the effects of sleep apnea on the sym­ pathetic nervous system. Our findings and those of others show consistent effects of sleep apnea on sympathetic arousal, as deter­ mined by elevated circulating norepinephrine levels, increased norepinephrine excretion, and desensitized and/or down-regulated adrenergic receptors (Dimsdale et al., 1997; Mills, Dimsdale, et al., 1995; Nelesen et al., 1996). These effects are closely related to blood pressure and poorer clinical outcomes in apneics (Bao, Nelesen, Loredo, Dimsdale, & Zeigler, 2002). Fortunately, a common treatment for sleep apnea, continuous posi­ tive airway pressure (CPAP), reverses these effects and restores sympathetic activation, including catecholamine levels, to normal levels (Bao et al., 2002; Ziegler, Mills, Loredo, Ancoli-Israel, & Dimsdale, 2001). We wondered whether CPAP normalized norepinephrine levels by reducing the release and/or increasing the clearance of nore­ pinephrine. To address this question, we used the norepinephrine kinetics assay. We studied 50 CPAP-naive sleep apnea patients before and after 14 days of CPAP or placeboCPAP (CPAP administered at ineffective pressure) (Mills, Kennedy, Loredo, Dimsdale, & Ziegler, 2006). Our endpoints were norepinephrine clearance and release rates, circulating norepinephrine levels, urinary norepinephrine excretion, and blood pressure. As expected from prior studies, we found that CPAP led to significant decreases in plasma norepinephrine levels (p ≤ 0.018) and daytime (p < 0.001) and nighttime (p < 0.05) norepinephrine excretion. Regarding nore­ pinephrine kinetics, we found that CPAP led

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 87

Sympathetic Hormones in Health Psychology Research

to an increase in the clearance of nore­ pinephrine (p ≤ 0.01) but not a significant change in the release rate. In addition, clinically, the drop in posttreatment blood pressure was best predicted by the patient’s pretreatment blood pressure and the post­ treatment norepinephrine clearance and release rate (p < 0.01). Thus, this nore­ pinephrine kinetics technique was useful in determining that CPAP treatment reduced norepinephrine levels through the mecha­ nism of increasing norepinephrine clearance from the circulation. Catecholamine Kinetics Methodology If the rate at which norepinephrine is cleared from the plasma is known and the plasma level of norepinephrine is known, then we can calculate the rate at which nore­ pinephrine appears in the plasma. This is known as the norepinephrine spillover or release rate. Clearance of norepinephrine has been measured by two infusion techniques. In one technique, norepinephrine is infused at a rate that increases plasma levels but has no marked effect on hemodynamics. Blood is sampled to measure plateau levels of nore­ pinephrine and norepinephrine decay rates to determine the rate of norepinephrine clear­ ance. A technically simpler method for evalu­ ating norepinephrine clearance rate is to infuse radiolabeled norepinephrine (3H-NE) and then measure the disappearance rate of radioactivity in sequential blood samples (Ziegler, Kennedy, Morrissey, & O’Connor, 1990). Results of these two techniques have not been entirely identical since 3H-NE infu­ sions take longer to reach plateau level than does unlabeled norepinephrine. Both tech­ niques indicate that the initial half-life of norepinephrine after infusion has stopped is in the range of 2 to 3 minutes in humans. During radiolabeled norepinephrine infusion, it is possible to sample arterial 3H-NE blood levels and the venous drainage of 3H-NE from

individual organs. Arterial and venous cate­ cholamine samplings do not represent the same things. Arterial blood catecholamines are representative of the mixed venous drainage from the lungs, an organ that is very active in catecholamine uptake and release. Venous catecholamines represent arterial catecholamine levels further modified by the venous organ drained, usually the fore­ arm and hand. Most published values are for venous blood obtained from the forearm. When these data are combined with measure­ ments of endogenous plasma norepinephrine, the contribution of individual vascular beds to release and clearance of norepinephrine can be calculated by the following formula: clearance (L/min) = 3H-NE infused/ min ÷ 3H-NE/L plasma. At steady state, plasma norepinephrine level times clearance of norepinephrine equals the apparent release rate or spillover into plasma, that is, release rate (ng/min) = clearance X plasma norepinephrine. This rate is one step closer to a real measure of sympathetic activity than plasma nore­ pinephrine levels alone. It does not measure the amount of norepinephrine released into the synapse, but no technique currently available does that. If the rate of nore­ pinephrine clearance is normal, then nore­ pinephrine levels may be adequate for determining sympathetic activity. Norepine­ phrine clearance varies widely between nor­ mal individuals, however, and norepinephrine clearance is abnormal in many illnesses, including depression, sleep disorders, hyper­ tension, and heart failure, to name a few (Esler & Kaye, 2000; FitzGerald et al., 1979; Jacobs, Lenders, Willemsen, & Thien, 1997; Ziegler et al., 1997). Drugs, too, including β-adrenergic receptor antagonists (β blockers)

87

04 - Luecken - 45370.qxd

88

9/12/2007

4:44 PM

Page 88

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

and antidepressants such as desipramine, can diminish clearance and reuptake. Of addi­ tional relevance to health psychology, nore­ pinephrine release rates have been shown to change in response to mental stressors (Goldstein et al., 1987).

Adrenergic Receptors Inasmuch as assessing norepinephrine kinetics provides important information that catecholamine levels alone cannot provide, assessing adrenergic receptors provides other important information on sympathetic regu­ lation and the actions of catecholamines. Catecholamines initiate biochemical and physiological events by binding to the three subclasses of β-adrenergic and the six sub­ classes of α-adrenergic receptors (Hall, 2004; Piascik & Perez, 2001; Taylor & Bristow, 2004). These adrenergic receptors provide the functional link between catecholamines and the numerous end organ responses they generate. In addition to modulating cate­ cholamine release and reuptake, as discussed earlier, they mediate end organ responses such as blood pressure, heart rate, myocar­ dial contractility, vascular constriction and relaxation, and renin release and inhibition, as well as a host of immune functions such as immune cell trafficking, adhesion, and cytokine responses, all of relevance to health psychology (Brodde, 1990; Sanders, 1995). Since the sensitivity (e.g., desensitization) and density (e.g., down-regulation) of agonist receptors are dynamically regulated in response to changing concentrations of adrenergic agonists, it is important in certain research models to be able to measure them directly. There are both in vivo and in vitro techniques to determine the functionality of adrenergic receptors. In vivo techniques involve infusing adrenergic agonists and assessing a specific end organ response. As with the nore­ pinephrine kinetics protocol, these methods

are typically carried out in a clinical research setting and require medical oversight. In vitro techniques typically involve isolating periph­ eral cells or specific organ tissue and quanti­ fying either the number or sensitivity of the adrenergic receptors expressed in that tissue. We will briefly review the methodologies of these techniques. In Vivo Techniques to Assess Adrenergic Receptors An in vivo technique for assessing β-adrenergic receptor sensitivity involves infusing the β adrenergic agonist isopro­ terenol and then measuring the heart rate response. An in vivo technique for assessing α-adrenergic receptor sensitivity involves infusing the α-adrenergic agonist phenyle­ phrine and then measuring the blood pres­ sure response. The general approach to both methods is similar. We will present details of the method to assess β-adrenergic receptor sensitivity. This technique is called the “chronotropic 25 dose”, or CD25 for short. We have used this technique to assess β­ adrenergic receptor sensitivity in a number of clinical studies (Dimsdale & Mills, 2002; Mills, Dimsdale, Ancoli-Israel, Clausen, & Loredo, 1998). The method involves intra­ venously infusing a series of bolus doses of isoproterenol and then measuring the heart rate response. Doses are typically 0, 0.10, 0.25, 0.50, 1.0, 2.0, and 4.0 µg. Heart rate is charted continuously by electrocardiography (ECG), and the maximum heart rate response to each dose is recorded. CD25 is calculated using the following formula: CD25 (µg iso­ proterenol = [(basal heart rate + 25)— intercept]/slope. The slope and intercept for each individual’s heart rate response to iso­ proterenol is calculated by linear regression. The CD25 value can then be tested for differ­ ences between groups by t-test or analysis of variance (ANOVA).

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 89

Sympathetic Hormones in Health Psychology Research

In Vitro Techniques to Assess Adrenergic Receptors Whereas the in vivo techniques can pro­ vide information only on the functional sen­ sitivity of adrenergic receptors, in vitro techniques can provide information on both sensitivity and density of adrenergic recep­ tors. In addition to conducting receptor assays on specific tissues of interest, such as cardiac or lung tissue for the β-adrenergic receptor or adipose tissue for the β-adrenergic receptor, there are less invasive ways to access these receptors by using peripheral blood cells. Lymphocytes, for example, express β2-adrenergic receptors, which can serve as a model for β-adrenergic receptors on the heart and lung. Platelets contain α2­ adrenergic receptors, which have been used in psychiatry and behavioral medicine research as a model of human α-adrenergic receptors and drug responsiveness. For these in vitro experiments using peripheral cells, lymphocytes and platelets are isolated from whole blood using a variety of techniques and then washed in prepara­ tion for the sensitivity and/or density assays. We have used these techniques widely in our research, including studies on stress, hyper­ tension, antihypertensive drug therapy, sleep apnea, and spaceflight (Bao et al., 2005; Meck et al., 2004; Mills & Dimsdale, 1988; Mills, Perez, Adler, & Ziegler, 2002). There are limitations to the use of peripheral blood cells as models of adrenergic receptors that researchers should be aware of (Mills & Dimsdale, 1993). As with the in vivo techniques, the in vitro techniques for assessing adrenergic receptor sensitivity involve stimulating the receptor of interest with an agonist and then measuring the response. Since many of the adrenergic receptors act through the activation of the membrane-bound enzyme adenylate cyclase, which catalyzes the conversion of adenosine

triphosphate (ATP) to cyclic adenosine monophosphate (cAMP), the amount of cAMP following receptor stimulation can be used as an index of receptor sensitivity. Typically, the greater the sensitivity and den­ sity of β-adrenergic receptors, the greater the amount of cAMP that is generated in the cell in response to stimulation. In the case of the β-adrenergic receptor, stimulation with acute isoproterenol stimulation results in a threeto fivefold increase in lymphocyte intracellu­ lar cAMP levels. The in vitro assay for determining adren­ ergic receptor density is called radioligand binding, which can be performed on intact whole cells or membranes from fractionated cells. Radioligand binding involves incubat­ ing a radioligand with the cell of interest under highly controlled conditions. When peripheral cells are used, typical radioligands for β-adrenergic receptors are 125I-iodocyanopin­ dolol and 125I-iodopindolol. 3H-prazosin, 3H­ rauwolscine, and 3H-yohimbine are ligands for α-adrenergic receptors. Upon termination of the incubation, usually by dilution, the unbound radioligand is removed by filtra­ tion. The remaining radioligand that is bound to the cell surface receptors is then measured and used to calculate the density of the receptor. A number of important issues need to be addressed to ensure that optimal experimen­ tal conditions have been met and the binding experiments measure the specific receptors of interest. Although radioligands are designed to bind specifically to the receptor of interest, there is always some nonspecific binding to other membrane proteins. This amount of nonspecific binding is determined by incubat­ ing the radioligand and tissue in the presence of a nonradioactive competing ligand that will bind to nearly all of the specific recep­ tors of interest and leave the radioligand binding to only the nonspecific sites. The non­ radioactive ligand is usually propranolol for

89

04 - Luecken - 45370.qxd

90

9/12/2007

4:44 PM

Page 90

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

β-adrenergic receptors and phentolamine for α-adrenergic receptors. By subtracting the radioactivity observed in the presence of the unlabeled drug (nonspecific binding) from that obtained in the absence of the unlabeled drug (total binding), the amount of specific binding is obtained. Specific binding repre­ sents the binding of interest. To determine the receptor density and binding affinity, radioli­ gand binding isotherms are used. Bmax, or the maximum amount of radioligand bound to the receptors, is the number of receptors expressed on the whole cell or the density expressed on cellular membranes. Kd, or the dissociation constant or binding affinity of the radioligand for the receptor, is the con­ centration of radioligand that binds to half of the specific receptors. Radioligand-binding isotherms involve incubating six to eight con­ centrations of the radioligand with a constant number of cells or membranes. The data derived are then mathematically transformed and analyzed by nonlinear regression to yield Bmax and Kd (Motulsky, 2001). Depending on factors such as age, fitness hypertension, use of adrenergic receptor antagonists, and so on, receptor binding typically yields a Bmax of 600 to 2,000 β2-adrenerigc receptors per lympho­ cyte and a Bmax of 240 to 600 α2-adrenerigc receptors per platelet.

SUMMARY Determining levels of catecholamines offers a useful guide to sympathetic activity pro­ viding that proper controls and guidelines are adhered to. Although catecholamines are unstable compounds, by following proper collection and storage procedures the researcher can help ensure that assessed lev­ els reflect, as near as possible, the in vivo lev­ els. Their value to health psychology is vast, including that their levels are related to mood, trauma, and behavior and are res­ ponsive to antidepressant therapy and acute stressors. Catecholamine levels in plasma depend on rate of release and rate of clear­ ance, which are rapid and vary in response to drugs and disease. Blood sampling sites need to be uniform, but there is not one best site for all circumstances. Interpretation of uri­ nary catecholamines is less clear than for plasma, since excreted catecholamines may come from plasma, renal nerves, or blood metabolic substrates. Even though the origin of urinary catecholamines is not resolved, their levels parallel changes in plasma cate­ cholamine levels. The advent of more reliable and available assay techniques has enabled more widespread use of catecholamine assessment in health psychology research.

SUGGESTED FURTHER READING Cohen, S., Doyle, W. J., & Baum, A. (2006). Socioeconomic status is associated with stress hormones. Psychosomatic Medicine 68(3), 414–420. Dimsdale, J. E, & Ziegler, M. G. (1991). What do plasma and urinary measures of catecholamines tell us about human response to stressors? Circulation 83(4 Suppl), II36–II42. Grant, I., McKibbin, C. L., Taylor, M. J., Mills, P. J., Dimsdale, J., Ziegler, M. G., & Patterson, T. L. (2003). In-home respite intervention reduces plasma epinephrine in stressed Alzheimer caregivers. American Journal of Geriatric Psychiatry 11, 62–72. Mills, P. J., Kennedy, B. P., Loredo, J., Dimsdale, J. E., & Ziegler, M. G. (2006). Effects of nasal continuous positive airway pressure and oxygen on

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 91

Sympathetic Hormones in Health Psychology Research norepinephrine kinetics and cardiovascular responses in obstructive sleep

apnea. Journal of Applied Physiology 100(1), 343–348.

Oberbeck, R. (2006). Catecholamines: Physiological immunomodulators during health and illness. Current Medical Chemistry 13(17), 1979–1989.

REFERENCES Bao, X., Mills, P. J., Rana, B. K, Dimsdale, J. E., Schork, N. J., Smith, D. W., et al. (2005). Interactive effects of common beta2-adrenoceptor haplotypes and age on susceptibility to hypertension and receptor function. Hypertension 46(2), 301–307. Bao, X., Nelesen, R. A., Loredo, J. S., Dimsdale, J. E., & Ziegler, M. G. (2002). Blood pressure variability in obstructive sleep apnea: Role of sympathetic ner­ vous activity and effect of continuous positive airway pressure. Blood Pressure Monitoring 7(6):301–307. Barnes, R. F., Raskind, M., Gumbrecht, G., & Halter, J. B. (1982). The effects of age on the plasma catecholamine response to mental stress in man. Journal of Clinical Endocrinology & Metabolism 54(1), 64–69. Brodde, O. E. (1990). Physiology and pharmacology of cardiovascular cate­ cholamine receptors: Implications for treatment of chronic heart failure. American Heart Journal, 120(6 Pt 2), 1565–1572. Burleson, M. H., Poehlmann, K. M., Hawkley, L. C., Ernst, J. M., Berntson, G. G., Malarkey, W. B., et al. (2003). Neuroendocrine and cardiovascular reactivity to stress in mid-aged and older women: Long-term temporal consistency of individual differences. Psychophysiology 40(3), 358–369. Chan, E. C., Wee, P. Y., & Ho, P. C. (2000). Evaluation of degradation of urinary catecholamines and metanephrines and deconjugation of their sulfoconjugates using stability-indicating reversed-phase ion-pair HPLC with electrochemical detection. Journal of Pharmaceutical and Biomedical Analysis 22(3), 515–526. Dimsdale, J. E. (1984). Generalizing from laboratory studies to field studies of human stress physiology. Psychosomatic Medicine 46(5), 463–469. Dimsdale, J. E., Coy, T., Ancoli-Israel, S., Mills, P., Clausen, J., & Ziegler, M. G. (1997). Sympathetic nervous system alterations in sleep apnea. The relative importance of respiratory disturbance, hypoxia, and sleep quality. Chest, 111(3), 639–642. Dimsdale, J. E., & Mills, P. J. (2002). An unanticipated effect of meditation on car­ diovascular pharmacology and physiology. American Journal of Cardiology, 90(8), 908–909. Dimsdale, J. E., & Moss, J. (1980). Short-term catecholamine response to psycho­ logical stress. Psychosomatic Medicine 42(5), 493–497. Dimsdale, J. E., & Ziegler, M. G. (1991). What do plasma and urinary measures of catecholamines tell us about human response to stressors? Circulation, 83(4 Suppl), II36–II42. Dimsdale, J. E., Ziegler, M., Mills, P., Delehanty, S. G., & Berry, C. (1990). Effects of salt, race, and hypertension on reactivity to stressors. Hypertension, 16(5), 573–580. Elfering, A., Grebner, S., Semmer, N. K., Byland, C., & Gerber, H. (2003). Two uri­ nary catecholamine measurement indices for applied stress research: effects of time and temperature until freezing. Human Factors, 45(4), 563–574.

91

04 - Luecken - 45370.qxd

92

9/12/2007

4:44 PM

Page 92

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Esler, M., & Kaye, D. (2000). Measurement of sympathetic nervous system activity in heart failure: The role of norepinephrine kinetics. Heart Failure Reviews, 5(1), 17–25. Esler, M., Lambert, G., Brunner-La Rocca, H. P., Vaddadi, G., & Kaye, D. (2003). Sympathetic nerve activity and neurotransmitter release in humans: Translation from pathophysiology into clinical practice. Acta Physiologica Scandinavica, 177(3), 275–284. FitzGerald, G. A., Hossmann, V., Hamilton, C. A., Reid, J. L., Davies, D. S., & Dollery, C. T. (1979). Interindividual variation in kinetics of infused epinephrine. Clinical Pharmacology & Therapeutics, 26(6), 669–675. Frankenhaeuser, M., Mellis, I., Rissler, A., Bjorkvall, C., & Patkai, P. (1968). Catecholamine excretion as related to cognitive and emotional reaction pat­ terns. Psychosomatic Medicine, 30(1), 109–124. Goldstein, D. S., Dionne, R., Sweet, J., Gracely, R., Brewer, H. B., Jr., Gregg, R., & Keiser, H. R. (1982). Circulatory, plasma catecholamine, cortisol, lipid, and psychological responses to a real-life stress (third molar extractions): Effects of diazepam sedation and of inclusion of epinephrine with the local anesthetic. Psychosomatic Medicine, 44(3), 259–272. Goldstein, D. S., Eisenhofer, G., Sax, F. L., Keiser, H. R., & Kopin, I. J. (1987). Plasma norepinephrine pharmacokinetics during mental challenge. Psychosomatic Medicine, 49(6), 591-605. Hall, R. A. (2004). Beta-adrenergic receptors and their interacting proteins. Seminars in Cell & Developmental Biology, 15(3), 281–288. Hansen, A. M., Garde, A. H., Skovgaard, L. T., & Christensen, J. M. (2001). Seasonal and biological variation of urinary epinephrine, norepinephrine, and cortisol in healthy women. Clinica Chimica Acta, 309(1), 25–35. Jacobs, M. C., Lenders, J. W., Willemsen, J. J., & Thien, T. (1997). Adrenomedullary secretion of epinephrine is increased in mild essential hypertension. Hypertension, 29(6):1303–1308. Jansen, R. W., Lenders, J. W., Thien, T., & Hoefnagels, W. H. (1989). The influ­ ence of age and blood pressure on the hemodynamic and humoral response to head-up tilt. Journal of the Amerocan Geriatric Society, 37(6), 528–532. Kennedy, B., Elayan, H., & Ziegler, M. G. (1990). Lung epinephrine synthesis. American Journal of Physiology, 258(4 Pt 1), L227–L231. Kennedy, B., & Ziegler, M. G. (1990). A more sensitive and specific radioenzymatic assay for catecholamines. Life Sciences, 47, 2143–2153. Kennedy, B., Ziegler, M. G., & Shannahoff-Khalsa, D. S. (1986). Alternating later­ alization of plasma catecholamines and nasal patency in humans. Life Sciences, 38(13), 1203–1214. Lakatua, D. J., Nicolau, G. Y., Bogdan, C., Plinga, L., Jachimowicz, A., SackettLundeen, L., et al. (1987). Chronobiology of catecholamine excretion in differ­ ent age groups. Progress in Clinical and Biological Research, 227B, 31–50. Lane, J. D., Adcock, R. A., Williams, R. B., & Kuhn, C. M. (1990). Caffeine effects on cardiovascular and neuroendocrine responses to acute psychosocial stress and their relationship to level of habitual caffeine consumption. Psychosomatic Medicine, 52(3), 320–336. Lehmann, M., & Keul, J. (1986). Age-associated changes of exercise-induced plasma catecholamine responses. European Journal of Applied Physiology and Occupational Physiology, 55(3), 302–306.

04 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 93

Sympathetic Hormones in Health Psychology Research Meck, J. V., Waters, W. W., Ziegler, M. G., deBlock, H. F., Mills, P. J., Robertson, D., & Huang, P. L. (2004). Mechanisms of postspaceflight orthostatic hypoten­ sion: low alpha1-adrenergic receptor responses before flight and central auto­ nomic dysregulation postflight. American Journal of Physiology, Heart and Circatory Physiology, 286(4), H1486–H1495. Mills, P. J., Berry, C. C., Dimsdale, J. E., Nelesen, R. A., & Ziegler, M. G. (1993). Temporal stability of task-induced cardiovascular, adrenergic, and psychological responses: The effects of race and hypertension. Psychophysiology, 30(2), 197–204. Mills, P. J., Berry, C. C., Dimsdale, J. E., Ziegler, M. G., Nelesen, R. A., & Kennedy, B. P. (1995). Lymphocyte subset redistribution in response to acute experimen­ tal stress: Effects of gender, ethnicity, hypertension, and the sympathetic ner­ vous system. Brain, Behavior, and Immunity, 9, 61–69. Mills, P. J., & Dimsdale, J. E. (1988). The promise of receptor studies in psy­ chophysiologic research. Psychosomatic Medicine, 50(6), 555–566. Mills, P. J., & Dimsdale, J. E. (1993). The promise of adrenergic receptor studies in psychophysiologic research II: Applications, limitations, and progress. Psychosomatic Medicine, 55(5), 448–457. Mills, P. J., Dimsdale, J. E., Ancoli-Israel, S., Clausen, J., & Loredo, J. S. (1998). The effects of hypoxia and sleep apnea on isoproterenol sensitivity. Sleep, 21(7), 731–735. Mills, P. J., Dimsdale, J. E., Coy, T. V., Ancoli-Israel, S., Clausen, J. L., & Nelesen, R. A. (1995). Beta 2-adrenergic receptor characteristics in sleep apnea patients. Sleep, 18(1), 39–42. Mills, P. J., Kennedy, B. P., Loredo, J. S., Dimsdale, J. E., & Ziegler, M. G. (2006). Effects of nasal continuous positive airway pressure and oxygen supplementa­ tion on norepinephrine kinetics and cardiovascular responses in obstructive sleep apnea. Journal of Applied Physiology, 100(1), 343–348. Mills, P. J., Perez, C. J., Adler, K. A., & Ziegler, M. G. (2002). The effects of space­ flight on adrenergic receptors and agonists and cell adhesion molecule expres­ sion. Journal of Neuroimmunology, 132(1–2):173–179. Mills, P. J., Schneider, R. H., & Dimsdale, J. E. (1989). Anger assessment and reactivity to stress. Journal of Psychosomatic Research, 33(3), 379–382. Motulsky, H. (2001). The GraphPad Guide to Analyzing Radioligand Binding Data, GraphPad, San Diego, CA. Available at http://www.graphpad.com/www/ radiolig/radiolig.htm Nelesen, R. A., Dimsdale, J. E., Mills, P. J., Clausen, J. L., Ziegler, M. G., & AncoliIsrael, S. (1996). Altered cardiac contractility in sleep apnea. Sleep, 19(2), 139–144. Nishihara, K., Mori, K., Endo, S., Ohta, T., & Ohara, K. (1985). Relationship between sleep efficiency and urinary excretion of catecholamines in bed-rested humans. Sleep, 8(2), 110–117. Papadelis, C., Kourtidou-Papadeli, C., Vlachogiannis, E., Skepastianos, P., Bamidis, P., Maglaveras, N., & Pappas, K. (2003). Effects of mental workload and caf­ feine on catecholamines and blood pressure compared to performance varia­ tions. Brain and Cognition, 51(1), 143–154. Piascik, M. T., & Perez, D. M. (2001). Alpha1-adrenergic receptors: New insights and directions. Journal of Pharmacology and Experimental Therapeutics, 298(2), 403–410.

93

04 - Luecken - 45370.qxd

94

9/12/2007

4:44 PM

Page 94

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Prinz, P. N., Halter, J., Benedetti, C., & Raskind, M. (1979). Circadian variation of plasma catecholamines in young and old men: Relation to rapid eye movement and slow wave sleep. Journal of Clinical Endocrinology & Metababolism, 49(2), 300–304. Sanders, V. M. (1995). The role of adrenoceptor-mediated signals in the modulation of lymphocyte function. Advances in Neuroimmunology, 5(3), 283–298. Silverman, H. G., & Mazzeo, R. S. (1996). Hormonal responses to maximal and submaximal exercise in trained and untrained men of various ages. Journals of Gerontology. Series A, Biological Science and Medical Science, 51(1), B30–B37. Suarez, E. C., Kuhn, C. M., Schanberg, S. M., Williams, R. B., Jr., & Zimmermann, E. A. (1998). Neuroendocrine, cardiovascular, and emotional responses of hostile men: The role of interpersonal challenge. Psychosomatic Medicine, 60(1), 78–88. Taylor, M. R., & Bristow, M. R. (2004). The emerging pharmacogenomics of the beta-adrenergic receptors. Congestive Heart Failure, 10(6), 281–288. Wallin, B. G., Thompson, J. M., Jennings, G. L., & Esler, M. D. (1996). Renal nora­ drenaline spillover correlates with muscle sympathetic activity in humans. Journal of Physiology, 491(Pt 3), 881–887. Wassell, J., Reed, P., Kane, J., & Weinkove, C. (1999). Freedom from drug interfer­ ence in new immunoassay for urinary catecholamines and metanephrines. Clinical Chemistry, 45, 2216–2223. Weiner, H. (1972). Presidential address: Some comments on the transduction of experience by the brain: Implications for our understanding of the relationship of mind to body. Psychosomatic Medicine, 34(4), 355 –380. Westermann, J., Hubl, W., Kaiser, N., & Salewski, L. (2002). Simple, rapid and sen­ sitive determination of epinephrine and norepinephrine in urine and plasma by non-competitive enzyme immunoassay, compared with HPLC method. Clinical Laboratory, 48(1–2), 61–71. Wilkinson, D. J., Thompson, J. M., Lambert, G. W., Jennings, G. L., Schwarz, R. G., Jefferys, D., et al. (1998). Sympathetic activity in patients with panic disorder at rest, under laboratory mental stress, and during panic attacks. Archives of General Psychiatry, 55(6), 511–520. Ziegler, M. G., Bao, X., Kennedy, B. P., Joyner, A., & Enns, R. (2002). Location, development, control, and function of extraadrenal phenylethanolamine N-methyltransferase. Annals of the New York Academy of Sciences, 971, 76–82. Ziegler, M. G., Kennedy, B., Morrissey, E., & O’Connor, D. T. (1990). Norepinephrine clearance, chromogranin A and dopamine beta hydroxylase in renal failure. Kidney International, 37(5), 1357–1362. Ziegler, M. G., Mills, P. J., Loredo, J. S., Ancoli-Israel, S., & Dimsdale, J. E. (2001). Effect of continuous positive airway pressure and placebo treatment on sympa­ thetic nervous activity in patients with obstructive sleep apnea. Chest, 120(3),887–893. Ziegler, M. G., Nelesen, R., Mills, P., Ancoli-Israel, S., Kennedy, B., & Dimsdale, J. E. (1997). Sleep apnea, norepinephrine-release rate, and daytime hypertension. Sleep, 20(3), 224–231.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 95

CHAPTER

5

Assessment of Salivary α-Amylase in Biobehavioral Research DOUGLAS A. GRANGER KATIE T. KIVLIGHAN MONA EL-SHEIKH ELANA B. GORDIS LAURA R. STROUD

S

alivary α-amylase (sAA) emerged as a surrogate marker of the sympathetic nervous system component of the stress response in the late 1980s. Surprisingly, it has (with a few noted exceptions) only rarely been employed in mainstream biobehavioral research. Recent technical advances have increased accessibility, and several new studies have generated renewed interest in sAA as

a measure of individual differences in stress vulnerability (Granger et al., 2006, 2007). In this chapter, we provide an introduction to sAA for behavioral scientists. We review its properties and basic functions and present illustrative findings. We provide practical information to enable investigators to collect, handle, prepare, and store samples for assay. We also describe the measurement of sAA by

AUTHORS’ NOTE: Components of the research were supported in part by the Behavioral Endocrinology Laboratory and the Child Youth and Families Consortium at The Pennsylvania State University, as well as the National Institute of Child Health and Development (PO1HD39667-01A1, K23HD041428), National Science Foundation (0126584), Alabama Agricultural Experiment Station (ALA010-008), and a Lindsey Foundation Grant. Thanks are due to our colleagues who made many of the preliminary studies possible: Joe Buckhalt, Kathryn Handwerger, Leah Hibel, Laura Klein, Jared Lisonbee, Jackie Mize, Elisabeth Susman, and Sheila West as well as to Mary Curran, Becky Hamilton, Vincent Nelson, and Eve Schwartz for biotechnical support. 95

05 - Luecken - 45370.qxd

96

9/12/2007

4:44 PM

Page 96

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

kinetic reaction assay, make recommenda­ tions regarding data transformation and analytical strategies, and outline a conceptual framework to guide the interpretation of results. Our intent is to accelerate the learn­ ing curve to help investigators avoid poten­ tial pitfalls associated with integrating this unique salivary analyte into the next genera­ tion of biobehavioral research. Technical advances that have made possi­ ble the assessment of biomarkers in saliva have enabled researchers to study how biolog­ ical and social processes related to stress vulnerability interact to influence health and behavior. The noninvasive nature of salivary measures is especially valuable because it allows biosocial models to be studied in quasinaturalistic social contexts and in response to the trials and tribulations of everyday life. Until recently, empirical attention has nar­ rowly focused mostly on the activity of the hypothalamic-pituitary-adrenal (HPA) axis as indexed by individual differences and intraindividual change in salivary cortisol (e.g., Kirschbaum, Read & Hellhammer, 1992). Experts on stress physiology, however, have emphasized the need to include multi­ ple measures of stress across multiple stress systems, and have suggested that the psy­ chobiology of the stress response is much more complex than can be modeled by just measuring salivary cortisol alone (e.g., Lovallo & Thomas, 2000). Two main systems com­ prise the neuroendocrine response to stress. The HPA axis is activated through the secre­ tion of corticotropin-releasing hormone, even­ tuating the release of glucocorticoids (e.g., cortisol) from the adrenal cortex into periph­ eral circulation. A second, and faster acting system, involves activation of the locus ceruleus/autonomic (sympathetic) nervous system and the release of catecholamines (e.g., Chrousos & Gold, 1992). To advance our understanding of how biological, social, cog­ nitive, and behavioral processes interact to determine risk or resilience, it has been argued

that multiple measurements of stress-related biological processes should be included in our conceptual and analytical models (e.g., Bauer, Quas, & Boyce, 2002; Donzella, Gunnar, Krueger & Alwin, 2000; Granger & Kivlighan, 2003). Chapters in this edited volume (see chapters by Kamark and by Thayer et al.) review and discuss the psychophysiological methods avail­ able to assess individual differences in activa­ tion of the autonomic (sympathetic) nervous system (SNS). As noted, these measurements typically involve electrodes, a computerized recording apparatus, and sophisticated data reduction algorithms. Whereas these methods are undeniably valuable, they are typically designed for use in controlled experimental (lab) conditions. Thus, they are difficult to employ to study biosocial models in the con­ text of the everyday social worlds of infants, children, and adults. Monitoring the SNS subcomponent of the stress response in saliva would be highly prac­ tical for studying individual differences in the stress response to ecologically valid stressors that cannot be reproduced in the controlled setting of the laboratory. In addition, this strategy would allow modeling of crosssituational consistency and habituation of the stress response to the dynamics of social land­ scapes. Further, because dissociations among various measures of physiological reactivity, even those activated by a common system such as the SNS, are frequently found (Quas, Hong, Alkon, & Boyce, 2000), and such responses are subject to the individual speci­ ficity phenomenon (Engel, 1972), in which participants may consistently react in one spe­ cific physiological modality, the examination of sAA is likely to provide unique important information in addition to knowledge based on other physiological measures influenced by the SNS. Thus, technical advances that would enable us to monitor the SNS stress response in saliva set the stage for biobehavioral science to fill several critical gaps in knowledge.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 97

Assessment of Salivary α -Amylase in Biobehavioral Research

The most common serological measures of SNS activation are catecholamines such as epinephrine (EPI) and norepinephrine (NE). Typically, catecholamines are mea­ sured using high pressure liquid chromatog­ raphy (HPLC) in plasma or urine specimens. These high-complexity measurements involve elaborate sample collection, extrac­ tion/separation, and preparation protocols, expensive laboratory equipment, and highly skilled technical operators (e.g., Mills & Ziegler in this volume). More recently, less complex and expensive antibody-based immunoassay protocols have become avail­ able for these compounds. Unfortunately, however, measuring catecholamines in saliva by either HPLC or immunoassay has proven challenging. Direct measurements of EPI or NE in saliva seem not to reflect SNS activity (e.g., Schwab, Heubel & Bartels, 1992). Consequently, researchers have been actively searching for an indirect or surrogate marker of SNS activity in saliva. EPI and NE are released by sympathetic nerve endings in tissues and glands through the body (e.g., lymph nodes, salivary glands) as well as in response to HPA activation via the adrenal medulla. The salivary gland, glandular duct cells, and vascular bed of the salivary glands are rich with β-adrenoreceptors (e.g., Nederfors & Dahlof, 1992). SNS activation affects the release of catecholamines from nerve endings, and the action of these com­ pounds on adrenergic receptors influences the activity of the salivary glands. A small, but rapidly growing literature, suggests that sAA might serve as a noninvasive and easily obtainable surrogate marker of SNS activity. Salivary α-amylase is an enzyme produced by the salivary gland. Salivary α-amylase levels are influenced by SNS activation via adrener­ gic receptors, as noted earlier. A computer lit­ erature search reveals some publications in the late 1980s, relatively little new information or publications on the topic until 2004, and then a number of presentations and papers under

review or in press. Renewed interest in sAA clearly exists among a subset of investigators, but we suspect most biobehavioral researchers remain unfamiliar with this analyte.

SALIVARY α-AMYLASE BASIC PROPERTIES AND FUNCTION Common salivary biomarkers employed in biobehavioral research include andro­ gens (e.g., dehydroepiandrosterone, and­ rostenedione), reproductive hormones (e.g., testosterone, estradiol, progesterone), gluco­ corticoids (e.g., cortisol), immunoglobulins (e.g., secretory IgA), and drugs and meta­ bolites of substance use (e.g., cotinine). In contrast, sAA is an enzyme. For most health psychologists there are several unique and important properties that distinguish sAA from the common hormones and drugs assessed in saliva in biobehavioral research. We briefly review these properties and refer to seminal works for interested readers.

Production by the Salivary Gland In contrast to most salivary analytes (the exception being sIgA, surrogate markers of inflammation, or disease-related antigens specific to oral fluids), sAA is not actively transported, nor does it diffuse passively, into saliva from the general circulation. Salivary α-amylase is produced locally in the mouth by the salivary gland. Under conditions of normal oral health, α-amylase is present in saliva in relatively high concentrations.

Digestive Enzyme Glycosyl hydrolases are a group of enzymes that hydrolyze the glycosidic bond between two or more carbohydrates, or between a carbohydrate and a noncarbohy­ drate. A primary biological function of sAA is the digestion of macromolecules such as

97

05 - Luecken - 45370.qxd

98

9/12/2007

4:44 PM

Page 98

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

carbohydrates and starch. α-Amylase is also produced within the gastrointestinal system by the pancreas. Levels of α-amylase in these two compartments (oral and gastrointesti­ nal) are produced by independent sources and are not correlated. Salivary α-amylase digests a portion of ingested starch in the stomach before it enters the intestine. Higher α-amylase activity is associated with increased caloric intake from the digestion of carbohydrates (i.e., breads, potatoes, rice, and pasta) into sugars that can be absorbed through the intestinal wall. α-Amylase inhibitors (AAIs) are molecules that interfere with the action of the α­ amylase enzyme (e.g., Doleckova-Maresova, Pavlik, Horn & Mares, 2005). Clinical stud­ ies have explored AAIs as dietary interven­ tions. The use of so called starch blockers has gained in popularity with the success and growth of carbohydrate-restricted diets. Kataoka and Dimagno (1999) reported the effects of an AAI included slowing of the rate of digestion and gastric emptying, thereby prolonging the feeling of fullness and delay­ ing the urge to consume more food. Interestingly, beyond links between sAA and response to stress, the apparent link between sAA and eating behavior may be of more than passing interest. For instance, Susman and colleagues (2006) report a positive association between sAA activity and body mass index in adolescents. It is tempting to speculate that stress-related increases in sAA, when combined with high carbohydrate/starch diet and sedentary lifestyle could in part be attributed to the weight-related disease epi­ demics facing many westernized societies.

Role in Oral Health A secondary role of sAA is bacterial clear­ ance from the mouth and prevention of bac­ terial attachment to oral surfaces (see Marcotte & Lavoie, 1998). Indeed, much of what is known about the biobehavioral

implications of individual differences in sAA levels is documented in the literature on oral biology and disease. Higher sAA activity is associated with reduced risk for a variety of processes related to oral health (bacteria load, caries, and periodontal disease). Atypically low salivary α-amylase activity is associated with oral disease. Although oral health has been a relatively minor concern for mainstream health psychology, the dental literature presents volumes of work that links psychosocial stress and social factors to com­ promised oral biology (e.g., Quinonez, Keels, Vann, McIver, Heller, & Whitt, 2001). Developmental Issues At birth, sAA is not present in the oral or gastrointestinal compartments (O’Donnell & Miller, 1980). Correspondingly, newborns do not have the same capacity as do children and adults to digest complex macromolecules. This transitional physiological state is con­ sidered adaptive because, for example, macromolecules (such as carbohydrates) are generally highly immunogenic and represent a significant source of threat to the neonatal immune system. Moreover, during this period of immune immaturity, newborns are primar­ ily protected against foreign antigens by pas­ sive immunity received from their mothers (i.e., maternal antibodies, IgA) in colostrum or breast milk. One possibility is that in the absence of sAA, maternal antibodies can be ingested without being destroyed (essentially digested) by the infant. Salivary α-amylase lev­ els shows a sharp rise in the 2- to 24-month period, reaching maximum levels by 2 to 6 years of age (Davis, Kannan, Marucut, Granger, & Sandman, 2007; O’Donnell & Miller, 1980). The age of onset in the rise of sAA levels parallels the timing of the introduc­ tion of solid foods in the diet and the emer­ gence of dentition needed to chew those solids. Individual differences in sAA activity have been reported to be associated with age

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 99

Assessment of Salivary α -Amylase in Biobehavioral Research

and pubertal development in later childhood (El-Sheikh, Mize, & Granger, 2005; Stroud et al., 2006b; Susman et al., 2006)

SALIVARY α-AMYLASE: RELATION TO PHYSICAL AND PSYCHOLOGICAL STRESS Of most interest to health psychologists is the pattern across studies showing that salivary α-amylase levels rise in response to stress. Early studies with adults revealed that levels of sAA increase in response to exercise, heat and cold stress, and written examinations (Chatterton, Vogelsong, Lu, Ellman, & Hudgens, 1996; Chatterton, Vogelsong, Lu, & Hudgens, 1997; Skosnik, Chatterton, Swisher, & Park, 2000). More recent studies show sAA increases in response to the Trier Social Stress Test (Nater et al., 2005; Nater, La Marca, et al., 2006), watching highly neg­ ative emotional pictures (van Stegeren, Rohleder, Everaerd, & Wolf, 2006), partici­ pating in athletic competition (Kivlighan & Granger, 2006), and confronting achieve­ ment (i.e., speech, arithmetic, tracing) or interpersonal (social rejection) stressors (see Stroud, Handwerger, Kivlighan, Granger, & Niaura, 2005; Gordis, Granger, Susman, & Tricket, 2006a). In contrast to the growing body of literature on older children and adults, relatively less research has been con­ ducted with salivary α-amylase in early child­ hood (but see Granger et al., 2006; Davis et al., 2007; Hill et al, 2007; Spinrad, Granger, & Eisenberg, 2007). The typical sAA response profile is consistent with our knowledge of the rapid activation and recov­ ery that characterizes the response of the SNS to stress. Findings from recent studies are illustrative. Gordis and colleagues (2006a) report that, in response to the Trier Social Stress Test (TSST), adolescents’ sAA levels increased 145%, on average over pretask lev­ els, with 65% of the sample showing greater

than 10% increases in sAA levels from pretask baseline to peak. Salivary α-amylase levels returned quickly to baseline within 10 minutes post stressor (see also Nater et al., 2005; Nater, La Marca, et al., 2006).

SALIVARY α-AMYLASE AND THE ADRENERGIC COMPONENT OF THE STRESS RESPONSE In the late 1990s, a series of elegant studies by Chatterton and colleagues suggested a strong positive association between levels of sAA and the SNS component of the stress response (e.g., Chatterton et al., 1996, 1997; Skosnik et al., 2000). Salivary α-amylase concentra­ tions were associated with baseline plasma catecholamine levels, particularly NE, and were also highly correlated with NE change in response to stress (Chatterton et al., 1996; see also Rohleder, Nater, Wolf, Ehlert, & Kirschbaum, 2004). The strong positive asso­ ciations between SNS activation and salivary α-amylase were corroborated by a recent placebo-controlled study showing that stressrelated increases in salivary α-amylase can be inhibited by administration of the adrenergic blocker propranolol (van Stegeren et al., 2006). In addition, β-adrenergic agonists are capable of stimulating salivary α-amylase release without increasing salivary flow (Gallacher & Petersen, 1983). This evidence suggests the same stimuli that result in the release of catecholamines in peripheral tissues activate sympathetic input to the salivary glands. The early work suggested that circulat­ ing levels of NE, associated with the stress response of the locus ceruleus/autonomic (sympathetic) nervous system, could be esti­ mated by the concentrations of α-amylase in whole saliva specimens, and that salivary α-amylase measurements may be employed as a noninvasive measure of plasma NE concentrations in human participants. More

99

05 - Luecken - 45370.qxd

100

9/12/2007

4:44 PM

Page 100

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

recent studies corroborate that sAA responds to physical and psychological stress, and relates to sympathetic tone or adrenergic activation generally. However, these studies also raise hard to answer questions about the specific association between sAA and stress-related change in catecholamines (Nater et al., 2005; Nater, La Marca, et al., 2006). Thus, whereas considering salivary α-amylase as a correlate to the adrenergic component of the stress response seems reasonable, based on the available data, using sAA as a specific or direct marker of NE in our conceptual and measurement models seems less appropriate at this time.

SALIVARY α-AMYLASE AND THE HPA AXIS RESPONSE TO STRESS The profile of stress-related change in sAA lev­ els is distinct from the response profile mea­ sured by salivary cortisol (see Dickerson & Kemeny, 2004, for review). Salivary α-amylase reaches its peak response faster and recovers to pretask baseline levels faster than cortisol (Gordis et al., 2006a; Nater et al., 2005; Nater, La Marca, et al., 2006). Generally, this kinetic response pattern fits our expectations of the differences in the timing of the SNS (quicker) and HPA (slower) stress responses (Chrousos & Gold, 1992). Our preliminary studies also show sAA may increase over pretask levels in response to stress in a larger percentage of cases, and the magnitude of the rise in sAA may be larger, on average, than that of salivary cortisol (Gordis et al., 2006a; Kivlighan et al., 2005). These differences may be due to the more sensitive threshold of reac­ tivity in the SNS (i.e., sAA) than in the HPA axis (e.g., Lovallo & Thomas, 2000). Given these differences in kinetic profiles and, potentially, sensitivity to stress, the lack of correlation between levels of salivary corti­ sol and sAA at baseline, in response to stress, or during recovery is not surprising (Granger et al., 2006; Nater, La Marca, et al, 2006;

with noted exceptions, Gordis et al, 2006b; Kivlighan & Granger, 2006). The statistical independence of these measures suggests that individual differences in sAA reactivity are not redundant with change in HPA measures, and likely are indexing a different stress response system (e.g., Chatterton et al., 1996). Whereas the HPA and SNS systems overall work in coordination to generate the physiologic changes associated with the stress response, the exact nature of the coordination (e.g., additive or interactive; opposing or comple­ mentary) is a subject of debate (e.g., Fries, Hesse, Hellhammer & Hellhammer, 2005; Golczynka, Lenders, & Goldetsein, 1995; Kvetanasky, Fukunara, Pacak, Cissa, Goldstein & Kopin, 1993). We believe that health psychologists should be interested in understanding how social forces and psychological states influence the coordination of these stress-response systems (see Bauer et al., 2002; Fries et al., 2005; Henry, 1992). Of course, this idea is not nec­ essarily novel (see Frankenhaeuser, 1982), but the noninvasive means of assessing both the HPA and SNS via saliva should enable the field to do so not just in the laboratory but in the context of everyday life and/or within the same individuals cross-situationally or on repeated occasions. We anticipate that the findings generated should provide insight into how disruption of physiological processes contributes to social, behavioral, health, and cognitive problems. Studies aimed at deter­ mining both internal, individual factors and social environmental events that predict corre­ lated versus dissociated rates of change between salivary cortisol and sAA would seem worthwhile (see following section). SALIVARY α-AMYLASE AND CARDIOVASCULAR PSYCHOPHYSIOLOGY Several studies have reported relationships between sAA and cardiovascular psychophy­ siology. Chatterton and colleagues (1996)

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 101

Assessment of Salivary α -Amylase in Biobehavioral Research

reported a positive relationship between sAA and heart rate (HR) that strengthened with the intensity of physical stress. Bosch and colleagues (2003) observed increases in sympathetic activity (shortened pre-ejection period, PEP) and sAA secretion and decreases in parasympathetic activity (decrease in heart rate variability) in response to a laboratory stressor. Nater, LaMarca, and colleagues (2006) report a positive relationship between sAA and sympathetic tone during stress. West, Granger, Kivlighan, and Hurston (2006) showed that, relative to resting base­ line values, individuals with the largest increases in sAA after a cold pressor task showed larger concurrent reductions in PEP, increased systolic BP, increased HR, and increased cardiac output. The association between sAA and PEP is particularly note­ worthy as PEP is among the few psycho­ physiological measures (the other is skin conductance) thought to exclusively assess SNS activity (e.g., Newlin & Levenson, 1979). Klein and colleagues (2006) showed that, in response to caffeine administration, BP, HR, and sAA increased, and caffeineinduced change in sAA was associated with increased HR. Similarly, Van Stegeren and colleagues (2006) showed that beta-blockade was successful in reducing HR, sAA, and sys­ tolic (but not diastolic) BP. Our studies have explored similar relation­ ships in youth. In a sample of participants ages 7 to 17 years, Stroud and colleagues (2005) found high-magnitude positive associations between sAA and SBP reactivity (defined as the difference between maximum poststress and baseline values) to laboratory stressors, but no associations between sAA and DBP or HR reactivity. In an elementary school-aged sample (ages 8–9 years), El-Sheikh and col­ leagues (2005, 2007) studied the relationship between respiratory sinus arrhythmia (RSA) and sAA reactivity to a stressful laboratory procedure. Salivary α-amylase was associated with deficits in vagal suppression (lower levels of RSA during challenge tasks in comparison

to baseline conditions). Whereas vagal sup­ pression is the typical response to envir­ onmental challenges, and is associated with positive child outcomes, children with higher levels of sAA exhibited vagal augmentation. El-Sheikh and colleagues (2005, 2007) also reported that children’s skin conductance levels (SCL) were positively and moderately associated with sAA levels during baseline conditions. Finally, Kivlighan, WeWerka, Gunnar, and Granger (2006) reported that during preparation for the TSST speech task, vagal tone (VT) and sAA were positively asso­ ciated. Peak sAA levels were also associated with VT levels early post task, and sAA and VT were related during both the early and late postspeech phase of the TSST.

Summary The accumulating evidence suggests a pat­ tern of associations between sAA and mea­ sures of cardiovascular psychophysiology. Higher sAA is associated with higher HR, SBP, and VT and reduced PEP and lower levels of vagal suppression to challenges. The small number of studies, variation in study design, and characteristics of participants pre­ vents us from making meaningful interpreta­ tion of the cross-study differences. However, in general, the direction of effects is consistent with the notion that sAA is associated with sympathetic/parasympathetic nervous system activation. Missing from the literature are studies that enable us to speculate about the meaning of these associations for cardiovas­ cular disease-related symptoms, outcomes, or treatment responsiveness.

CONCOMITANTS OF INDIVIDUAL DIFFERENCES IN SALIVARY α-AMYLASE The previous sections have documented that individual differences in sAA may be related to age and pubertal development, and sensitive to

101

05 - Luecken - 45370.qxd

102

9/12/2007

4:44 PM

Page 102

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

specific experiences that involve physical, social, and cognitive demands. Studies by our informal collaborative network also reveal that individual differences in sAA also may be, inde­ pendently or in combination with salivary cor­ tisol, associated with a wide range of behavior. We provide examples of these findings to give the reader an idea of potential patterns and themes that may be relevant to research pro­ grams in health psychology.

Infancy Four studies have measured sAA in motherinfant dyads as the infants participated in a task designed to illicit emotional distress (Davis et al., 2007; Hill et al., 2007; Kivlighan et al., 2005; Shea et al., 2006). Each study showed that maternal and infant sAA levels are positively correlated. Shea and colleagues (2006) report this association to be mediated by pre- and postnatal maternal depression. Davis and colleagues (2007) report that while sAA levels were detected as early as 2 months of age, sAA reactivity to inoculation stress was not evi­ dent until 6 months. Hill and colleagues (2007) measured sAA in infants (12 months old) as they participated with their mothers in the strange situation procedure involving brief sep­ arations and reunions. Children classified as “avoidant” had higher mean levels of sAA than “securely” attached children. These prelimi­ nary findings suggest that individual differences in sAA activity are associated (or attuned) between mothers and infants, and that infants’ sAA may be influenced by the nature of their social relationships with caregivers.

Preschool Period Spinrad, Granger, and Eisenberg (2007) examined preschoolers’ sAA reactivity to a “not sharing task.” Over 57% of the children exhibited increases of 10% or greater in sAA between the pretest and 10 minutes after the stressor. Increases in sAA reactivity were

related to lower positive affect and higher expressions of anger during the task in girls, but not in boys. Mize and colleagues (Mize, Lisonbee & Granger, 2005) examined sAA and cortisol as a function of stress, teacherchild relationship quality, and health in preschoolers. Children participated in a series of five developmentally appropriate challenge tasks or games intended to provoke mild frus­ tration or disappointment (e.g., disappoint­ ment experience, impossible puzzle task, delay of gratification task). Children with greater sAA increases from prechallenge to follow-up had more illness and less close relationships with teachers. The association between sAA and teacher-child relationships may support and extend the observations noted previously that suggest individual dif­ ferences in sAA in infants are related to characteristics of their relationships with caregivers. The relationship between sAA and health is novel, and if replicated, would obvi­ ously be of interest to health psychologists.

Middle Childhood El-Sheikh, Buckhalt, Granger, and Mize (2006), examined relations between cortisol, sAA, and problem behavior in children (8–9 years old). Saliva samples were obtained dur­ ing baseline and following two laboratory stressors. Baseline levels of AA were positively associated with boys’ aggression. Poststress sAA was associated with girls’ internalizing behavior problems, and higher sAA reactivity from pre- to poststress was related to higher levels of girls’ internalizing symptoms. In sup­ port of Bauer’s propositions regarding the importance of interactions between the HPA and SNS systems in the prediction of psy­ chopathology (Bauer et al., 2002), interaction effects between SNS activity (indexed by sAA and skin conductance level) and cortisol explained moderate amounts of unique vari­ ance in children’s externalizing and internaliz­ ing problems.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 103

Assessment of Salivary α -Amylase in Biobehavioral Research

More specifically, the highest levels of externalizing and internalizing problems were found among children with symmetrical HPA and SNS activity—particularly among children with high activity in both domains. Conversely, children with the lowest levels of adjustment problems had asymmetrical pat­ terns of HPA and SNS activity (i.e., high activity in one system and low activity in the other). These findings are supportive of the “additive” model put forth by Bauer et al. (2002), which suggests that redundant actions of the SNS and HPA could result in hyperarousal when both systems are high in activity or hypoarousal when both systems are low in activity. Supportive of the robust nature of effects, an identical pattern of effects was observed in interactions between cortisol and either of the two SNS measures, namely, skin conductance level or sAA. Buckhalt, El-Sheikh, and Granger (under review), examined relations between sAA, cortisol, and children’s cognitive and aca­ demic functioning. Higher baseline cortisol and sAA were broadly associated with poorer cognitive/academic performance. Poststress cortisol and sAA as well as cortisol reactivity (pre- to poststressors) and sAA reactivity seemed to have different relations with cognitive functioning for boys and girls. Specifically, higher levels on these variables were related to poorer performance for girls and better performance for boys. Further, and consistent with Bauer and colleagues’ (2002) speculations, regression analyses demonstrated significant interactions between cortisol and sAA in the prediction of child functioning. Specifically, worse academic/ cognitive functioning was found for children with higher levels of resting sAA in conjunc­ tion with higher levels of cortisol. Additional analyses with the same sample of healthy (no chronic or acute illnesses) 8- to 9-year-olds yielded interesting relations between sAA levels and children’s, especially girls’, physical health problems as reported

by parents and teachers (El-Sheikh, Mize, & Granger, 2005). Higher levels of poststress sAA and reactivity from pre- to postchal­ lenge conditions, were associated with increased health problems including respira­ tory problems, fatigue, and frequency of ill­ ness. Further, sAA levels were significantly and positively associated with sIgA during both poststress and reactivity from pre- to poststress conditions. In summary, in studies with elementary school-aged children, individual differences in sAA appear to relate to problem social behav­ ior. Given the well-established association between attachment classification early in life and problem behavior later in middle child­ hood (e.g., Lyons-Ruth, Easterbrooks & Cibelli, 1997), it is tempting to speculate about a pattern linking sAA and social behavior/relationships across childhood. The link between sAA and cortisol and children’s academic (and task-related) performance is also noteworthy and is generally consistent with Frankenhaeuser’s seminal works (Lundberg & Frankenhaeuser, 1980; Rauste-von Wright, von Wright, & Frankenhaeuser, 1981) on SNS arousal, attention, and achievement.

Adolescence Three studies with adolescents reveal asso­ ciations between individual differences in sAA and problem behavior. Stroud and colleagues (2006a) examined associations between salivary cortisol and sAA reactivity to the TSST. Greater sAA reactivity was asso­ ciated with decreases in feelings of relax­ ation (“relax”) and increases in feeling “upset” between baseline and stress periods. Participants with high cortisol and low sAA responses (based on percent change from baseline to maximum level achieved during stress) showed higher scores on the total and externalizing scales (Child Behavior Check­ list, CBCL) as well as the social problems, thought problems, and aggression/delinquency

103

05 - Luecken - 45370.qxd

104

9/12/2007

4:44 PM

Page 104

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

behavior CBCL subscales than did partici­ pants with low cortisol and high sAA reactiv­ ity. Similarly, Gordis and colleagues (2006a) report that asymmetrical cortisol and sAA reactivity to the TSST accounted for signifi­ cant variance in parent-reported adolescent aggression. At lower levels of sAA reactivity, lower cortisol reactivity corresponded to higher aggression ratings, but at high sAA reactivity levels, cortisol reactivity was unrelated to aggression. Finally, Susman and colleagues (2006) observed that sAA reactiv­ ity, based on changed from pre- to postTSST, was negatively related to adolescent reported symptoms of oppositional defiant disorder (ODD) and conduct disorder (CD). That is, low change from pre- to postSST corresponded to more ODD and CD symptoms according to adolescents’ self-reports. In summary, studies with adolescents con­ sistently find that sAA increases in response to the TSST. Attenuated sAA stress reactivity may be a potential risk for aggression and symptoms of disruptive behavior disorders. Gender differences seem to exist in the behavioral correlates of sAA in studies of elementary school-aged children and adoles­ cents. Findings show support for Bauer’s propositions regarding the importance of interactions between the HPA and SNS systems in the prediction of psychopathology (Bauer et al., 2002).

Adults In adults, few studies have explored corre­ lates of individual differences in sAA in rela­ tion to behavior. Two studies are illustrative. Kivlighan and Granger (2006) examined indi­ vidual differences in sAA response to compe­ tition. Salivary α-amylase was higher across the competition for varsity than for novice athletes, and was positively associated with performance and interest in team bonding. Salivary α-amylase reactivity to competition

explained individual differences in domi­ nance, and symmetry in sAA and cortisol reactivity to competition (low-low) was asso­ ciated with high perceived dominance. Nater and colleagues (Nater, Rohleder, Schlotz, Nakkas, Kirschbaum & Ehlert, 2006) explored correlates of the diurnal pattern of sAA activity. In two independent studies, saliva samples were collected directly after waking up, 30 and 60 minutes later, and each full hour until 9 p.m. Results indicate that sAA activity shows a distinct diurnal profile pattern with a trough in the morning and a steady increase of activity during the day. Multilevel-modeling failed to show a withinsubject association of sAA with acute stress, but significant associations emerged with chronic stress, mood, and alertness.

Summary This series of studies reveals distinct associ­ ations between sAA, health behavior, and cog­ nition, and supports the tentative conclusion that sAA may index physiological processes that affect, and are influenced by, psychologi­ cal, behavioral, and social processes.

SALIVARY α-AMYLASE MEASUREMENT ISSUES In contrast to the majority of salivary biomarkers measured using antibody-based assays (radio-, chemoluminescent, enzyme immunoassays), sAA is measured by a kinetic reaction. The assay system we have employed uses a chromagenic substrate linked to mal­ totriose. The enzymatic action of α-amylase on this substrate yields a yellow color, which can be measured spectrophotometrically (i.e., optical density at 405 nm). Basically, more α­ amylase in the sample will lead to the degra­ dation of more maltotriose. Correspondingly, more yellow color will develop in the test well. The amount of α-amylase activity

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 105

Assessment of Salivary α -Amylase in Biobehavioral Research

present in the sample is directly proportional to the increase (over a 2-minute period) in absorbance (optical density). This assay pro­ tocol recently became commercially available (Salimetrics, State College, PA), and thus we present only a brief overview here.

Assay Protocol Saliva samples (10 µl) are diluted 1:200 in assay diluent and well mixed; 8 µl of diluted sample or control is then pipetted into indi­ vidual wells of a 96-well microtiter plate. 320 µl of preheated (37 °C) α-amylase substrate solution is added to each well and mixed for 3 minutes. Results are computed in U/ml of α­ amylase. Intrassay variation (CV) computed for the mean of 10 replicate tests of low (17.7 U/ml), medium (108.8 U/mL), and high (474.6 U/ml) concentration samples were 7.2%, 6.7%, and 2.5%, respectively. Interas­ say variation computed for the mean of aver­ age duplicates for eight separate runs for lower (10.6 U/ml) and higher (166.0 U/ml) concentration samples were 5.8% and 3.6%.

Data Analysis Salivary α-amylase distributions are typi­ cally leptokurtic and positively skewed. In our collective experience, square root trans­ formation is sufficient to correct sAA distri­ butions. Even after transformation, there are wide-ranging differences between individuals in sAA levels (range 0–400 U/ml). As noted earlier, sAA levels are sensitive to change as a result of environmental and psychological challenge. However, within-individual differ­ ences in sAA levels are moderately correlated over time (Granger et al., 2007).

Screening and Exclusionary Criterion Use of any prescription or over-the­ counter medication (OTC) with potential to influence the parasympathetic or sympathetic

nervous system (adrenergic agonists or antagonists) should be avoided. In particular, use of prescription medications for angina or high blood pressure that have beta-blocking properties (e.g., Inderal, Tenormin, Coreg, Lopressor) or consumables that stimulate the SNS (e.g., caffeine) should be controlled in studies employing sAA measurements (e.g., Klein et al., 2006). Similarly, OTC use of supplements with AAI properties should be cause for exclusion. Nicotine use is associ­ ated with activation of the SNS (e.g., Grassi et al., 1994), raising the possibility that tobacco smoke exposure would be posi­ tively related to sAA. Studies show, however, that the highly acidic aldehydes in tobacco smoke inactivate the α-amylase enzyme (e.g., Nagler, Lischinsky, Diamond, Drigues, Klein & Reznick, 2000). Thus, contrary to the activation effects of nicotine on the SNS, exposure to tobacco smoke is likely to be associated with lower rather than higher sAA activity (Granger et al., in press). Nicotine use (e.g., gum, patch, water) and tobacco smoke exposure should be avoided or con­ trolled in studies involving α-amylase. The findings of Klein and colleagues (2006) sug­ gest that caffeine doses equivalent to one or two 8 oz cups of coffee (~200 mg) should also be of concern. Caffeine is an ingredient in many teas, sodas, waters, juices, and OTC medications (e.g., Nodose, Excedrin, Anacin) as well as most chocolate- and coffee-fla­ vored foods. However, dosages vary widely, and, for most food items, caffeine levels greater than 100 mg are rare in single serv­ ings of foods. In general, we recommend monitoring use of caffeine in studies measur­ ing sAA. Theoretically, because of the role of sAA in the digestion of carbohydrates and starches, increased salivary α-amylase activ­ ity may be found in saliva samples collected after consumption of high carbohydrate/ starch meals (e.g., Fortunato, Kivlighan, Davis, Granger, & The Family Life Project Investigators, 2007). At least one study

105

05 - Luecken - 45370.qxd

106

9/12/2007

4:44 PM

Page 106

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

reveals that pregnancy is associated with lower sAA activity and attenuated responses to psychosocial stress (Nierop, Bratsikas, Klinkenberg, Nater, Zimmerman & Ehlert, 2006). Finally, given the well-established links between oral disease and sAA activity, individual differences in oral health may also be an important covariate.

Sample Collection, Handling, and Storage Welcome news is that the stress-induced increase in sAA is independent of saliva flow rate (Rohleder, Wolf, Maldonado & Kirschbaum, 2006) and the assay of sAA requires only 10 µl of sample (less than 1/16th of an eye dropper drop of saliva). The most common saliva collection methods typically involve absorbing the sample with cottonbased products. While this method seems appropriate for some salivary markers (i.e., cotinine) cotton-based products cause sub­ stantial interference in the assay of many others including sIgA, dehydroepiandros­ terone, testosterone, estradiol, and proges­ terone (Shirtcliff, Granger, Schwartz & Curran, 2001). Granger and colleagues (2006) reveal that saliva samples to be assayed for sAA can be collected by passive drool, cotton swabs, or microsponge without compromis­ ing assay validity. Moreover, multiple freezethaw cycles do not have significant effects on the assay of sAA. While our preliminary study suggested otherwise (see Granger et al., 2006), more comprehensive evaluation suggests that salivary α-amylase activity may decline when saliva samples are stored at room temperature (RT) for periods as long as or longer than 24 hours. Ancedotal reports also suggest that sAA levels may vary depending on where in the mouth samples are collected. Our pilot data show that sAA levels were higher when collection swabs were placed between the upper cheek and gum than when they were put under the tongue.

Timing of Sample Collection Sample collection clearly must be on a dif­ ferent time course to capture the response pro­ files of salivary cortisol and sAA accurately in the same study. For sAA measures, appropri­ ate sample collection times might include pretask, 5 minutes posttask (peak), and 10, 20, or 30 minutes posttask (recovery). However, it is important to note that this “recommended” time course has been “established” most often using standard lab challenges (mostly varia­ tions of the TSST), and investigators should expect the time course to differ depending on the nature of the stressor (e.g., duration, inten­ sity, pain, social evaluation, novelty, threat) in combination with characteristics of the indi­ vidual (previous experience, coping resources, social support).

Summary The quantitative measurement of sAA in saliva can be efficiently accomplished with a kinetic assay approach. Reagents are available so investigators can construct their own in­ house assay (Sigma Chemical Corporation, St. Louis, MO), and they are commercially available in kit form (Salimetrics, State College, PA). Most recently, Yamaguchi, Deguchi, and Wakasugi (2005) have reported the validity of a wearable microelectro­ mechanical system (flat chip sensor) to moni­ tor sAA. Saliva samples to be assayed for α-amylase may be conveniently collected using the most commonly employed methods— passive drool, cotton, or hydrocellulose absorbent materials. As with all salivary biomarkers, special care in screening, sample collection, handling, and storage are necessary to ensure results are reliable and valid. CONCLUDING COMMENTS A small, but growing, literature reveals age-, gender-, and stress-related differences in sAA

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 107

Assessment of Salivary α -Amylase in Biobehavioral Research

levels and patterns of intra-individual sAA change in response to challenges that dis­ tinctly differ from those measured by salivary cortisol. In addition, findings suggest associa­ tions between sAA levels and social behavior and relationships, health, negative affectivity, cognitive/academic problems, and cardiovas­ cular activity. These findings underscore that integration of noninvasive measurements of the adrenergic component of the locus ceruleus/ autonomic (sympathetic) nervous system, as indexed by sAA, may extend our understanding of health-related biobehav­ ioral phenomena to new limits. Across studies, a robust pattern emerges suggesting dissociation between measures of salivary cortisol and sAA reactivity to challenge. The overwhelming pattern extends Chatterton’s conclusion that sAA is not responding to a stress signal related to the HPA axis. That these measures are not redun­ dant implies that the inclusion of both salivary cortisol and α-amylase in biosocial studies may potentially improve our explanation of individual differences in stress-related vulnera­ bility and resilience. Future studies that exam­ ine additive or interactive effects of salivary cortisol and sAA in biosocial models seem well justified. Further, studies aimed to delin­ eate features (e.g., frequency, intensity, dura­ tion) of social stressors that activate salivary cortisol only, sAA only, or both, would seem particularly worthwhile. Contemporary behav­ ioral endocrinologists have downplayed the significance of biological determinism in their models, and rarely ever endorse the notion of direct hormone-behavior relation­ ships. Instead, they champion reciprocal effects between biological and behavioral pro­ cesses, and emphasize that these reciprocal effects are highly dependent on social context. The current findings suggest that individual differences in sAA levels and reactivity are also largely determined by social forces. These forces include at least the nature of age-appropriate social stressors (e.g., arm restraint by a stranger,

frustration tasks and interactions with teachers, social evaluation, and watching your infant in distress), but also the influence of children’s social relationships (with mothers and teachers). Clearly, these observations sug­ gest we need to know more about the features of social landscapes and social relationships that moderate individual differences and intraindividual change in sAA levels. Our findings suggest age-related differences in sAA reactiv­ ity to social stressors. Although individual dif­ ferences existed, on average, neither infants (but see Davis et al., 2007), preschoolers, nor elementary school-aged children as a group showed stress-related sAA increases. By con­ trast, our studies of adolescents, young adult mothers, and collegiate athletes revealed sig­ nificant intra-individual change (increases) in sAA reactivity. Given that these are the only developmental studies of this phenomenon, we cannot be sure whether this is a true devel­ opmental difference or the result of many method differences between studies. A number of scattered findings here sug­ gest links between individual differences in sAA and health. With respect to implications for the field of health psychology, several findings are particularly noteworthy. For example, associations emerged between sAA and illness susceptibility. Generally, the find­ ing is consistent with volumes of research on the linkages among the brain, behavior, and immunity (see Ader, Cohen, & Felten, 1995). More specifically, stress-related increases in SNS activation have been associated with immune suppression. Deductive logic sug­ gests that compromised immune function subsequently leads to increased susceptibility to negative health outcomes. The invasive nature of the measurements needed to study relationships (venipuncture) among the brain, behavior, and immunity has significantly slowed the study of these relationships in youth and special populations, and in the context of everyday social life. In addition, the earlier literature suggests a link between

107

05 - Luecken - 45370.qxd

108

9/12/2007

4:44 PM

Page 108

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL

salivary α-amylase and eating behavior, and recent preliminary findings revealed an asso­ ciation with body mass index in adolescents (Susman et al., 2006). Stress-related increases in sAA consistently emerged across studies among adolescents in response to social challenge (TSST). Stress-related individual differences in sAA, combined with high­ carbohydrate/starch diet and sedentary lifestyle could potentially contribute to the weight-related disease epidemics facing many youth in westernized societies. Finally, the numerous associations between individual differences and intra-individual change in sAA and standard measures of cardiovascular psychophysiology beg questions regarding the utility of this noninvasive measure in research aimed at preventing, monitoring, or remedy­ ing cardiovascular disease-related symptoms. Our review reveals that behaviorally trained scientists with appropriate training can accom­ plish the collection, handling, and storage requirements for samples to be assayed for sAA. As with other salivary biomarkers, special issues (e.g., medication use, smoking,

caffeine) need to be addressed to ensure the integrity of assay results. Assay materials and reagents are accessible, and relatively inexpen­ sive. The assay protocol for processing samples requires relatively basic skills, and the equip­ ment needed to do so is standard in laborato­ ries set up to conduct immunoassays. In conclusion, research on sAA is burgeon­ ing, but in-depth study is in the beginning stages. Many of the “representative findings” noted are from work that has been recently, or will soon be, presented at scientific meetings, or is in the process of peer review. Thus, while we remain optimistic, we are cautious about drawing any conclusions from available empir­ ical evidence. In particular, a very important direction for future research is to add to our knowledge regarding what individual differ­ ences in sAA reactivity to stress are actually measuring. The involvement of researchers who study social health psychology in this endeavor is critical because, now more than ever before, research considers social forces to be key elements in how biobehavioral pro­ cesses set the stage for risk or resilience.

SUGGESTED READINGS Chatterton, R. T., Vogelsong, K. M., Lu, Y, Ellman, A. B., & Hudgens, G. A. (1996). Salivary alpha-amylase as a measure of endogenous adrenergic activity. Clinical Physiology, 16, 433–448. Gordis, E. B., Granger, D. A., Susman, E. J., & Trickett, P. K. (2006). Asymmetry between salivary cortisol and alpha-amylase reactivity to stress: Relation to aggressive behavior in adolescents. Psychoneuroendocrinology, 31, 976–987. Kivlighan, K. T., & Granger, D. A (2006). Stress responsivity to competition: Gender and experiential differences in salivary alpha-amylase and cortisol activity. Psychoneuroendocrinology, 31, 703–714. Nater, U. M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller, M. M., & Ehlert, U. (2006). Stress-induced changes in human salivary alpha-amylase activity— associations with adrenergic activity. Psychoneuroendocrinology, 31, 49–58. Nater, U. M., Rohleder, N., Gaab, J., Berger, S., Jud, A., Kirschbaum, C. & Ehlert, U. (2005). Human salivary alpha-amylase reactivity in a psychosocial stress paradigm. International Journal of Psychophysiology, 55, 333–342. Rohleder, N., Nater, U. M., Wolf, J. M., Ehlert, U., & Kirschbaum, C. (2004). Psychoso­ cial stress-induced activation of salivary alpha-amylase: An indicator of sympathetic activity? Annals of the New York Academy of Sciences, 1032, 258–263.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 109

Assessment of Salivary α -Amylase in Biobehavioral Research

REFERENCES Ader, R., Cohen, N., & Felten, D. (1995). Psychoneuroimmunology. San Diego: Academic Press. Bauer, A. M., Quas, J. A., & Boyce, W. T. (2002). Associations between physiolog­ ical reactivity and children’s behavior: Advantages of a multisystem approach. Journal of Developmental Behavioral Pediatrics, 23, 102–113. Bosch, J.A., de Geus, E.J., Veerman, E.C., Hoogstraten, J., & Nieuw Amerongen, A.V. (2003). Innate secretory immunity in response to laboratory stressors that

evoke distinct patterns of cardiac autonomic activity. Psychosomatic Medicine,

65, 245–258.

Buckhalt, J. A., El-Sheikh, M., & Granger, D. A. (under review). Children’s cogni­ tive functioning and academic performance: The role of cortisol and alphaamylase. Manuscript submitted for publication. Chatterton, R. T., Vogelsong, K. M., Lu, Y, Ellman, A. B., & Hudgens, G. A. (1996). Salivary alpha-amylase as a measure of endogenous adrenergic activity. Clinical Physiology, 16, 433–448. Chatterton, R. T., Vogelson, K. M., Lu, Y, & Hudgens, G. A. (1997). Hormonal responses to psychological stress in men preparing for skydiving. Journal of Clinical Endocrinology and Metabolism, 82, 2503–2509. Chrousos, G. P., & Gold, P. W. (1992). The concepts of stress and stress system dis­ orders. Journal of the American Medical Association, 267, 1244–1252. Davis, E. P., Kannan, I., Marucut, J., Granger, D. A., & Sandman, C. A. (2007, March). Salivary alpha-amylase response to an inoculation stressor. Presented at the Biennial Meeting of the Society for Research on Child Development, Boston, MA. Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychological Bulletin, 130, 355–391. Doleckova-Maresova, L., Pavlik, M., Horn, M., & Mares, M. (2005). De novo design of alpha-amylase inhibitor: A small linear mimetic of macromolecular proteinaceous ligands. Chemical Biology, 12, 1257–1258. Donzella, B., Gunnar, M. R., Krueger, W. K., & Alwin, J. (2000). Cortisol and vagal tone responses to competitive challenge in preschoolers: Associations with tem­ perament. Developmental Psychobiology, 37, 209–220. El-Sheikh, M., Buckhalt, J. A., Erath, S. A. Granger, D. A., & Mize, J. (2007). Cortisol and children’s adjustment: The moderating role of sympathetic ner­ vous system activity. Manuscript submitted for publication. El-Sheikh, M., Buckhalt, J., Granger, D. A., & Mize, J. (2006, March). Salivary alpha amylase and cortisol: Their association with children’s adjustment, health, sleep, and cognitive functioning. Paper presented at the biennial meet­ ing of the Society for Research on Adolescence, San Francisco, CA. El-Sheikh, M., Mize, J., & Granger, D. A. (2005, March). Endocrine and parasym­ pathetic responses to stress predict child adjustment, physical health, and cog­ nitive functioning. Paper presented at the biennial meeting of Society for Research in Child Development, Altanta, GA. Engel, B. T. (1972). Response specificity. In N. S. Greenfield & R. A. Sternbach (Eds.), Handbook of psychophysiology (pp. 571–576). New York: Holt, Rinehart, & Winston.

109

05 - Luecken - 45370.qxd

110

9/12/2007

4:44 PM

Page 110

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Fortunato, C. K., Kivlighan, K. T., Davis, L., Granger, D. A., & The Family Life Project Investigators. (2007, March). Trials and tribulations of collecting saliva samples from infants: Prevalence and impact on salivary alpha-amylase and cortisol. Presented at the Biennial Meeting of the Society for Research on Child Development, Boston, MA. Frankenhaeuser, M. (1982). Challenge-control interaction as reflected in sympa­ thetic-adrenal and pituitary-adrenal activity: Comparison between the sexes. Scandinavian Journal of Psychology, Supplement 1, 158–164. Fries, E., Hesse, J., Hellhammer, J., & Hellhammer, D. (2005). A new view on hypocortisolism. Psychoneuroendocrinology, 30, 1010–1016. Gallacher, D. V., & Petersen, O. H. (1983). Stimulus-secretion coupling in mam­ malian salivary glands. International Review of Physiology, 28, 1–52. Golczynka, A., Lenders, J. W., & Goldstein, D. S.(1995). Glucocorticoid-induced sympathoinhibition in humans. Clinical Pharmacology Therapy, 58, 90–98. Gordis, E. B., Granger, D. A., Susman, E. J., & Trickett, P. K. (2006a). Asymmetry between salivary cortisol and alpha-amylase reactivity to stress: Relation to aggressive behavior in adolescents. Psychoneuroendocrinology, 31, 976–987 Gordis, E., B., Granger, D. A., Susman, E. J., & Trickett, P. K. (2006b, March). Salivary alpha amylase and cortisol responses to social stress among maltreated and comparison youth. Paper presented at the biennial meeting of the Society for Research on Adolescence, San Francisco, CA. Granger, D. A., Blair, C., Willoughby, M., Kivlighan, K. T., Hibel., L. C., Weigand, L. E., & Family Life Project Investigators (in press). Individual differences in salivary cortisol and alpha-amylase in mothers and their infants: Relation to tobacco smoke exposure. Developmental Psychobiology. Granger, D. A., & Kivlighan, K. T. (2003). Integrating biological, behavioral, and social levels of analysis in early child development research: Progress, problems, and prospects. Child Development, 74, 1058–1063. Granger, D. A., Kivlighan, K. T., Blair, C. El-Sheikh, M., Mize, J., Lisonbee, J.A., et al. (2006). Integrating the measurement of salivary alpha-amylase into stud­ ies of child health, development, and social relationships. Journal of Personal and Social Relationships. Special Issue: Physiology and Human Relationships, 23, 267–290. Granger, D. A. Kivlighan, K. T., El-Sheikh, M., Gordis, E., & Stroud, L. R. (2007). Salivary alpha-amylase in biobehavioral research: Recent developments and applications. Annals of the New York Academy of Sciences, 1098, 122–144. Grassi, G., Servalle, G., Calhoun, D. A., Bolla, G. B., Giannattasio, C., Marabini, M., et al. (1994). Mechanisms responsible for sympathetic activation by cigarette smoking in humans. Circulation, 90, 248–253. Henry, J. P. (1992). Biological basis of the stress response. Integrative Physiology and Behavioral Science, 27, 66–83. Hill, A., Mills-Kounce, R., Propper, C., Calkins, S., Granger, D. A., Gariepy, J-L., & Cox, M. (2007, March). Physiological responses as a function of attachment status in infants and mothers. Presented at the Biennial Meeting of the Society for Research on Child Development, Boston, MA. Kataoka, K., & Dimagno, E. P. (1999). Effect of prolonged intraluminal alphaamylase inhibition on eating, weight, and the small intestine of rats. Nutrition, 15, 123–129. Kirschbaum, C., Read, G. F., & Hellhammer, D. H. (1992). A ssessment of hormones and drugs in saliva in biobehavioral research. Gottingen, Germany: Hogrefe & Huber, Gottingen.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 111

Assessment of Salivary α -Amylase in Biobehavioral Research Kivlighan, K. T., & Granger, D. A (2006). Stress responsivity to competition: Gender and experiential differences in salivary alpha-amylase and cortisol activity. Psychoneuroendocrinology, 31, 703–714. Kivlighan, K. T., Granger, D. A., Blair, C., & The Family Life Project Investigators. (2005, March). Salivary alpha amylase and cortisol: levels and stress reactivity in 6-month old infants and their mothers. Paper presented at the biennial meet­ ing of Society for Research in Child Development, Atlanta, GA. Kivlighan, K .T., WeWerka, S. M., Gunnar, M. R., & Granger, D. A. (2006, March). Salivary alpha amylase reactivity to the Trier Social Stress Test: Relation to cortisol and autonomic responses in normally developing adoles­ cents. Paper presented at the biennial meeting of the Society for Research on Adolescence, San Francisco, CA. Klein, L. C., Whetzel, C. A., Bennett, J. M., Ritter, F. E., & Granger, D. A. (2006, March). Effects of caffeine and stress on salivary alpha-amylase in young men: A salivary biomarker of sympathetic activity. Paper presented at the annual meeting of the American Psychosomatic Society, Denver, CO. Kvetnansky, R., Fukunara, K., Pacak, K., Cissa, G., Goldstein, D. S., & Kopin, I. J. (1993). Endogenous glucocorticoids restrain catecholamine synthesis and release at rest and during immobilization stress in rats. Endocrinology, 133, 1411–1419. Lovallo, W. R., & Thomas, T. L. (2000). Stress hormones in psychophysiological research: Emotional, behavioral, and cognitive implications. In J. T. Cacioppo, L. G. Tassinary, & G. G. Bernston. (Eds.), Handbook of psychophysiology (2nd ed.; pp. 342–367). New York:Cambridge University Press. Lundberg, U., & Frankenhaeuser, M. (1980). Pituitary-adrenal and sympatheticadrenal correlates of distress and effort. Journal of Psychosomatic Research, 24, 125–130. Lyons-Ruth, K., Easterbrooks, M. A., & Cibelli, C. D. (1997). Infant attachment strategies, infant mental lag, and maternal depressive symptoms: Predictors of internalizing and externalizing problems at age 7. Developmental Psychology, 33, 681–692. Marcotte, H., & Lavoie, M. C. (1998). Oral microbial ecology and the role of salivary immunoglobulin A. Microbiology and Molecular Biology Reviews, 62, 71–109. Mize, J., Lisonbee, J., & Granger, D. A. (2005, March). Stress in child care: Cortisol and alpha-amylase may reflect different components of the stress response. Paper presented at the biennial meeting of Society for Research in Child Development, Altanta, GA. Nagler, R., Lischinsky, S., Diamond, E., Drigues, N., Klein, I., & Resnick, A. Z. (2000). Effect of cigarette smoke on salivary proteins and enzyme activities. Archives of Biochemistry and Biophysics, 379, 229–236. Nater, U. M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller, M. M., & Ehlert, U. (2006). Stress-induced changes in human salivary alpha-amylase activity—associations with adrenergic activity. Psychoneuroendocrinology, 31, 49–58. Nater, U. M., Rohleder, N., Gaab, J., Berger, S., Jud, A., Kirschbaum, C., & Ehlert, U. (2005). Human salivary alpha-amylase reactivity in a psychosocial stress paradigm. International Journal of Psychophysiology, 55, 333–342. Nater, Y. M., Rohleder, N., Scholotz, W., Nakkas, C., Kirschbaum, C., & Ehlert, U. (2006, March). Determinants of diurnal course of salivary alpha-amylase activity. Paper presented at the annual meeting of the American Psychosomatic Society, Denver, CO.

111

05 - Luecken - 45370.qxd

112

9/12/2007

4:44 PM

Page 112

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: HORMONAL Nederfors, T., & Dahlof, C. (1992). Effects of the beta-adrenoceptor antagonists atenolol and propranolol on human whole saliva flow rate and composition. Archives of Oral Biology, 37, 579–584. Newlin, D. B., & Levenson, R. W. (1979). Pre-ejection period: Measuring betaadrenergic influences upon the heart. Psychophysiology, 16, 546–553. Nierop, A., Bratsikas, A., Klinkenberg, A., Nater, U. M., Zimmerman, R., & Ehlert, U. (2006). Prolonged salivary cortisol recovery in second trimester pregnant women and attenuated salivary alpha-amylase responses to psychosocial stress in human pregnancy. Journal of Clinical Endocrinology and Metabolism, 91, 1329–1335. O’Donnell, M. D., & Miller, N. J. (1980). Plasma pancreatic and salivary-type amy­ lase and immunoreactive trypsin concentrations: Variations with age and refer­ ence ranges for children. Clinical Chemistry Acta, 104, 265–273. Quas, J. A., Hong, M., Alkon, A., & Boyce, W. T. (2000). Dissociations between psychobi­ ologic reactivity and emotional expression in children. Developmental Psychobiology, 37, 153–175. Quinonez, R. B., Keels, M. A., Vann, W. F., Jr., McIver, F. T., Heller, K., & Whitt, J. K. (2001). Early childhood caries: Analysis of psychosocial and biological factors in a high-risk population. Caries Research, 35, 376–383. Rauste-von Wright, M., von Wright, J., & Frankenhaeuser, M. (1981). Relationships between sex-related psychological characteristics during adolescence and catecholamine excretion during achievement stress. Psychophysiology, 18, 362–370. Rohleder, N., Nater, U. M., Wolf, J. M., Ehlert, U., & Kirschbaum, C. (2004). Psychosocial stress-induced activation of salivary alpha-amylase: An indicator of sympathetic activity? Annals of the New York Academy of Sciences, 1032, 258–263. Rohleder, N., Wolf, J. M., Maldonado, E. F., & Kirschbaum, C. (2006).The psy­ chosocial stress-induced increase in salivary alpha-amylase is independent of saliva flow rate. Psychophysiology, 43, 645–652. Schwab, K. O., Heubel, G., & Bartels, H. (1992). Free epinephrine, norepinephrine, and dopamine in saliva and plasma of healthy subjects. In C. Kirschbaum, G. F. Read, & D. H. Hellhammer, (Eds). Assessment of hormones and drugs in saliva in biobehavioral research (pp. 331–336) Gottingen, Germany: Hogrefe & Huber. Shea, A. K., Steiner, M., Brennan, P., Walker, E., Newport, D. J., Knight, B., et al. (2006, March). Maternal depression and salivary alpha amylase response to stress in their infants. Presented at the annual meeting of the American Psychosomatic Society, Denver, CO. Shirtcliff, E. A., Granger, D.A., Schwartz, E, & Curran, M.J. (2001). Use of salivary biomarkers in biobehavioral research: Cotton based sample collection methods can interfere with salivary immunoassay results. Psychoneuroendocrinology, 26, 165–173. Skosnik, P. D., Chatterton, R. T., Swisher, T., & Park, S. (2000). Modulation of attentional inhibition by norepinephrine and cortisol after psychological stress. International Journal of Psychophysiology, 36, 59–68. Spinrad, T. L., Granger, D. A., & Eisenberg, N. (2007, March). Individual differ­ ences in preschoolers’ salivary cortisol and alpha-amylase reactivity. Presented at the Biennial Meeting of the Society for Research on Child Development. Boston, MA.

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 113

Assessment of Salivary α -Amylase in Biobehavioral Research Stroud, L. R., Handwerger, K., Granger, D. A., Kivlighan, K. T., & Solomon, C. (2006a, March). Saliva alpha-amylase stress reactivity in children and adoles­ cents: Validity, associations with cortisol, and links behavior. Paper presented at the annual meeting of the American Psychosomatic Society, Denver, CO. Stroud, L. R., Handwerger, K., Granger, D. A., Solomon, C., Kivlighan, K. T., & Niaura, R. (2006b, March). Alpha amylase responses to achievement and inter­ personal stressors over adolescence: Developmental differences and associations with cortisol and cardiovascular responses. Paper presented at the biennial meet­ ing of the Society for Research on Adolescence, San Francisco, CA. Stroud, L. R., Handwerger, K., Kivlighan, K. T., Granger, D. A., & Niaura, R. (2005, March). Alpha amylase stress-reactivity in youth: Age differences and relation to cortisol, cardiovascular, and affective responses. Paper presented at the biennial meeting of Society for Research in Child Development, Altanta, GA. Susman, E. J., Granger, D. A., Dockray, S., Heaton, J., & Dorn, L. D. (2006, March). Alpha amylase, timing of puberty and disruptive behavior in young adolescents: A test of the attenuation hypothesis. Paper presented at the bien­ nial meeting of the Society for Research on Adolescence, San Francisco, CA. Van Stegeren, A., Rohleder, N., Everaerd, W., & Wolf, O. T. (2006). Salivary alpha amylase as marker for adrenergic activity during stress: Effect of beta blockade. Psychoneuroendocrinology, 31, 137–141. West, S. G., Granger, D. A., Kivlighan, K. T., & Hurston, K. L. (2006, March). Salivary alpha-amylase response to the cold pressor is correlated with cardiac markers of sympathetic activation. Paper presented at the annual meeting of the American Psychosomatic Society, Denver, CO. Yamaguchi, M., Deguchi, M., & Wakasugi, J. (2005). Flat-chip microanalytical enzyme sensor for salivary amylase activity. Biomedical Microdevices, 7, 295–300.

113

05 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 114

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 115

CHAPTER

6

The Measurement of Blood Pressure in Cardiovascular Research WILLIAM GERIN TANYA M. GOYAL ELIZABETH MOSTOFSKY DAICHI SHIMBO

I

n spite of its long history, the measure­ ment of blood pressure remains contro­ versial. This is largely due to the variety of factors that influence the measurement, as well as its interpretation. A blood pressure measurement changes depending on who takes it, in what locale it is taken, where it is measured in the body, the subject’s state of mind, caffeine and tobacco use in the hours prior to the measurement, the digit-preference bias of the assessor, the type of measuring device used, and the white coat effect; and this list is not exhaustive. In addition, the interpretation of the measure is also contro­ versial. Little disagreement arises over blood pressure as a measure of cardiovascular risk;

however, the extent to which the blood pressure measurement may legitimately be used as an index of a cognitive or emotional activation remains a topic of debate.

MEASURING BLOOD PRESSURE FOR RESEARCH PURPOSES Most of what is written on the subject of blood pressure measurement assumes a clini­ cal orientation. We are concerned, however, with the assessment of blood pressure for research purposes, which often has different considerations. Blood pressure might be mea­ sured in observational studies, experimental

AUTHORS’ NOTE: Preparation of this manuscript was supported by National Institutes of Health, Bethesda, MD, USA, Grants HL47540 and HL76857. 115

06 - Luecken - 45370.qxd

116

9/12/2007

4:44 PM

Page 116

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: CARDIOVASCULAR

studies, and intervention trials. In an observa­ tional study, a researcher may be interested in measuring blood pressure in different work­ ing populations, or by socioeconomic strata, cross-sectionally and/or over time. This might entail, as one example, going to different work sites and taking measurements of every­ one in a particular job description, such as clerical workers or forklift drivers; or possibly taking measurements at church-sponsored health fairs. In an interventional study, a typ­ ical strategy would be to bring patients into the laboratory, take baseline measurements, and then randomly assign them to a usual care or intervention condition. At a specified follow-up period, the patients would return to the laboratory, where their blood pres­ sures would be measured once again, and the change from baseline to follow-up computed to determine the effect of the intervention. Alternatively, the measurements may be abstracted from clinic records and patient charts. These design strategies all appear to be straightforward, and in many ways they are. However, there are important considerations that must be addressed to obtain reliable and valid measurements. When it is time to explain the measurement strategies used to those who review grant proposals and manuscripts, reviewers will want an explicit statement concerning the rationale for choos­ ing one method of blood pressure assessment over others.

AN OFTEN UNRELIABLE MEASURE Blood pressure is one of the most ubiquitous of clinical measurements; however, a blood pressure measurement often contains a great deal of both systematic and unsystematic error (Gerin, Pieper, & Pickering, 1993), which means that great care must be taken in its measurement if it is to be useful—to the

clinician for diagnosis; and to the researcher for prediction. The blood pressure changes with each beat of the heart; it is inherently quite variable, even between beats (Pickering, 1991). It is also highly responsive to any number of environmental influences, such as position (seated, standing), isometric tension (measurements should be taken with the back supported, feet on the floor), caffeine and nicotine use, stress and emotionality, and being asleep versus being awake, which is the source of the greatest variability (Pickering, 1991). When these sources of systematic variability are of interest, and are measured, their influ­ ence on blood pressure—if it is unwanted influence—can be reduced. Similarly, using careful measurement techniques and multiple measurements, the random error also can be reduced. This chapter focuses on reducing these systematic and unsystematic errors that occur in measuring blood pressure to provide measures that are appropriate for use in research.

WHAT IS BLOOD PRESSURE? Blood is carried from the heart to the rest of the body through the arteries; blood pressure is a measure of the force of the blood on the walls of blood vessels. Unless otherwise indi­ cated, blood pressure refers to the mean arte­ rial pressure, that is, the pressure in the large arteries, such as the brachial artery, which is higher than the blood pressure in other ves­ sels. The peak pressure in the arteries during the cardiac cycle is referred to as the systolic pressure, and the lowest pressure (at the rest­ ing phase of the cardiac cycle) is the diastolic pressure. Elevated blood pressure may cause damage to the target organs, including the heart, brain, kidney, and blood vessels; and has been associated with excess risk for car­ diovascular morbidity and mortality.

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 117

Measurement of Blood Pressure in Cardiovascular Research

A BRIEF HISTORY OF BLOOD PRESSURE MEASUREMENT One should feel the pulse at the place of the “cubit” and at the place of the “inch” and one should observe whether the pulse is superficial or whether it is deep, whether it is regular or uneven; and then it becomes evident where the disease originates. When the pulse is quick and contains six beats to one cycle of respiration, then it indicates heart trouble; and when the pulse is large the disease becomes grave. —Ch’i Po (Nei Ching: The Yellow Emperor’s Classic of Internal Medicine, 2698–2598 BCE)

In Western thought, the Greek physician Galen first proposed the existence of blood in the human body around 2,100 years ago. As the above quotations suggest, however, other cultures, including the ancient Chinese, had developed theories about the role of the pul­ sation of the heart in disease centuries earlier. In Western thought, based on the ideas of Hippocrates (460–361 BCE), Galen proposed that the body comprised three systems: the brain and nerves (responsible for sensation and thought), the liver and veins (which pro­ vided the body with nourishment and pro­ moted growth), and the blood and arteries. Galen proposed that the heart constantly produced blood, filling the body with lifegiving energy; however, in 1616, William Harvey concluded that this assertion by Galen was incorrect. Instead, he suggested, a finite amount of blood existed at any given moment, and circulated in the body. The first recorded measurement of blood pressure was done in 1733 by the Reverend Stephen Hales, a British veterinarian. Hales inserted a brass pipe into an artery of various

animals, and the height to which the animal’s blood spurted up into the tube gave a mea­ sure of the force propelling the blood. One of Hales’ most dramatic experiments using this simple manometer involved a mare, tied flat on the ground to a stable door. The glass tube in this instance was 12 ft, 9 in. long (3.8 m), and the horse’s blood rose in it to a height of 9 ft, 6 in. (2.9 m). It was not until more than 100 years later, in 1847, that human blood pressure was recorded. In 1828, Jean Leonard Marie Poiseuille replaced the long glass tube with a U-shaped tube filled with mercury, calibrated to record pressure levels in millimeters of mercury (mmHg). Carl Ludwig modified Poiseuille’s manometer in 1847, adding a revolving cylinder and float with a revolving drum on which the blood pressure was recorded by a quill. This device was called a kymograph (“wave-writer,” in Greek). As with Hales’ method, however, blood pres­ sure could still be measured only by invasive means. A few years later, in 1855, Karl van Vierordt found that an inflatable cuff could be placed around the arm, constricting the brachial artery, to obliterate the arterial pulse. This technique was further developed in 1860 by Etienne Jules Mary; his sphyg­ mograph could accurately measure the pulse rate, but was unreliable in determining the blood pressure. Yet this design was the first that could be used clinically with some mea­ sure of success. In 1876, Samuel Siegfried von Basch devel­ oped the first noninvasive blood pressure measuring device, which he called a sphyg­ momanometer. This was replaced in 1896 by the more accurate sphygmomanometer developed by Scipione Riva-Rocci, and Riva­ Rocci’s device has served as the prototype of today’s standard instrument, which uses an arm cuff inflated until the blood flow through

117

06 - Luecken - 45370.qxd

118

9/12/2007

4:44 PM

Page 118

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: CARDIOVASCULAR

the arteries is completely occluded. The air in the cuff is then released, and the pressure is measured on a mercury manometer at the instant the pulse reappears. Riva-Rocci’s instrument was accurate, but it measured only the pressure within the artery when the heart is contracting (i.e., the systolic pres­ sure). In 1905, however, Nikolai Korotkoff, a Russian physician, suggested that a stetho­ scope be used to listen to the blood flow in the brachial artery. Korotkoff’s method, which is employed by physicians today, uses the tapping sounds heard in the artery. The mercury level in the manometer when the tap­ ping begins indicates the systolic pressure; the mercury level at the instant the sounds disap­ pear indicates the diastolic blood pressure (the pressure that occurs between heart contractions). The technique of placing a stethoscope over the brachial artery at the antecubital fossa, distal to the arm cuff, and listening for the Korotkoff sounds is referred to as the auscultatory method. Mercury-column sphygmomanometers that use the Riva-Rocci/Korotkoff method remain the standard in many physicians’ offices. However, several physicians and hos­ pitals have gone to methods that avoid the use of mercury, and this trend will continue. In experimental use, for the most part, listen­ ing for Korotkoff sounds to assess blood pressure is awkward, and other methods, including those using automatic electronic devices, are preferable. These are discussed in the following sections.

BLOOD PRESSURE MEASUREMENT TECHNIQUES Several methods for measuring blood pres­ sure have been developed. First, we present a description of the different methods, as well as information concerning their use and accu­ racy. Following these sections, we present a

more general section on the use of the devices for research purposes, addressing such issues as the position and posture that subjects should assume prior to the measurement, how long the subject should rest before the measurement is taken, and so forth.

Intraarterial Monitoring Intraarterial blood pressure monitoring has been a theoretical concept since the 18th century but has come into common clinical practice only since the electromechanical rev­ olution of the 1960s. In 1969, Bevan and his colleagues (Bevan, Honour, & Stott, 1969) reported the use of direct arterial pressure recording in ambulatory subjects, showing that surprising blood pressure fluctuations occurred during normal activities; and noc­ turnal blood pressure tended to be cons­ iderably lower than awake pressure. These reports set the stage for a great deal of the research undertaken today. Intraarterial measurement is regarded as the gold standard for blood pressure assess­ ment. It provides a direct measure of the intraar­ terial pressure, as well as a continuous, beatto-beat profile of the blood pressure. It is a valuable monitoring tool with inherent risks, and therefore requires competent, knowl­ edgeable nursing care for its safe and effective use. The intrusive nature of the measurement represents a great limitation, and because of the subsequent development of accurate non­ invasive techniques, intraarterial monitoring is rarely used now in psychophysiological studies. However, this technique led directly to the refinement of noninvasive techniques using devices that utilize auscultatory or oscillometric (described in a later section) recording. Training. This is an in-patient procedure, and must be conducted by highly trained medical personnel only.

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 119

Measurement of Blood Pressure in Cardiovascular Research

Usefulness for Research Purposes. Because of the invasive nature of this method, and the availability of reliable noninvasive moni­ tors, intraarterial monitoring has become obsolete for the purposes of most research undertaken.

The Auscultatory Method Traditionally, this method involves the use of a mercury column sphygmomano­ meter or an aneroid device, both of which require a human operator. There are, how­ ever, automated devices, including ambula­ tory monitors (see also ch. 8, in this volume) and home monitors, that use a microphone positioned over the brachial artery to listen for sounds. This is addressed in a later sec­ tion; here we focus on manual techniques. Both mercury column sphygmomanome­ ters and aneroid devices use an arm cuff and require a human listener to hear Korotkoff sounds through a stethoscope. Pressures are measured in millimeters of mercury, much like a barometer measures atmospheric pres­ sure. An arm cuff is wrapped around the subject’s upper arm, just above the elbow. The subject’s arm should be at rest, supported at heart level; the arm cuff bladder should encir­ cle at least 80% of the upper arm, with the length-to-width ratio of the bladder being

Phase 1 A tapping sound

135

Phase 2 A soft, swishing sound

130

125

approximately 2:1. It is important to use an appropriately sized cuff (refer to Table 6.2, later in this chapter). The listener then pal­ pates the subject’s arm to locate the brachial artery, and, once it is located, a stethoscope is placed on the hollow of the elbow, over the artery. (It can be difficult to locate the artery in some persons.) To “read” the blood pressure, one must listen for the onset and offset of the Korotk­ off sounds, using a stethoscope. It is gener­ ally accepted that there are five phases of Korotkoff sounds. Each phase is characterized by the volume and quality of sound heard. Figure 6.1 illustrates the phases, using an example with a systolic blood pressure of 135 mmHg and a diastolic pressure of 90 mmHg. The listener, however, is required to identify only the Phases I (systole) and V (diastole). The cuff is inflated 20 to 30 mmHg above the palpated systolic pressure, compressing the brachial artery, and causing the artery to collapse once the systolic pressure (the maxi­ mum pressure exerted by the blood against the wall of the brachial artery when the heart beats) has been exceeded. The valve on the pump is then slowly opened, to allow the pressure in the cuff to decrease; the mercury is lowered slowly, by 2 to 3 mm per second, or per heart beat. Once the systolic pressure is reached, the brachial artery opens a bit, which

Phase 3 A crisp sound

Phase 4 A blowing sound

120

Phase 5 Silence

115

Cuff pressure (in mm Hg)

Systolic pressure

Figure 6.1

Illustration of the Five Phases of Korotkoff Sounds

Diastolic pressure

0

119

06 - Luecken - 45370.qxd

120

9/12/2007

4:44 PM

Page 120

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: CARDIOVASCULAR

causes volatile blood flow through the artery. This produces audible vibrations against the artery walls (the Korotkoff sounds), which can be heard through a stethoscope. The mer­ cury column sphygmomanometer and the aneroid monitor allow the blood pressure to be read on either a glass tube (mercury col­ umn) or a gradated dial (aneroid). The read­ ing at the instant at which the sounds are first heard is the systolic pressure. As the cuff con­ tinues to deflate, it allows more and more blood to flow through the artery. At the instant the Phase V sounds are no longer heard, the diastolic pressure is read. The blood pressure is read to the nearest 2 mmHg. Training. The following procedures are recommended by the British Hypertension Society (BHS) for training research personnel in the use of the auscultatory method (Beevers, Lip, & O’Brien, 2001): • Level of accuracy 90% of systolic and dias­ tolic blood pressure within 5 mmHg; 100% within 10 mmHg of an expert observer • Audiogram to check auditory acuity • Instruction in the theory of hypertension and blood pressure measurement • Reading materials on blood pressure mea­ surement1 • Tutorial sessions with demonstrations using a binaural or multiaural stethoscope • Interactive demonstration1 • Interactive assessment1 • Repeat interactive assessment until level of accuracy achieved1 • Training and assessment repeated every 3 months1

Mercury Column Sphygmomanometer For many physicians and researchers, this method has become the gold standard for blood pressure measurement, even more than intraarterial monitoring. Because intraarterial monitoring remains impractical for clinical or, mostly, research purposes, and because it

cannot be used to collect epidemiological data, the mercury column sphygmomanome­ ter represents the point of reference for many clinicians and researchers. The mercury sphygmomanometer has a reservoir that is filled with mercury; air pres­ sure applied to the cuff displaces the mercury in the reservoir and forces it into a graduated tube. The systolic and diastolic pressures are read from gradations marked on the tube. Reliability and Validity. Because they are gravity based, these models tend to be reli­ able. Once they are calibrated, the absence of moving parts practically eliminates the need for recalibration. Usefulness for Research Purposes. Mercury column sphygmomanometers have several disadvantages for research purposes. The main disadvantage is the need for a human lis­ tener, with its drawbacks concerning environ­ mental factors and the human biases that may affect the listener’s perception of, or recording of, the blood pressure values. These have been well-documented, and include the effects of a noisy environment on the listener’s ability to hear accurately, as well as possible hearing problems on the part of the listener. Aneroid Manometer Aneroid (“without liquid”) devices use metal that acts like a spring to measure blood pressure. Rather than transferring pressure to a mercury column, these units employ a bellows, or diaphragm, that expands when the cuff is inflated. As the pressure in the cuff rises, a pin resting on the expanding bellows is lifted, engaging a mechanism that moves a gauge needle around a graduated scale plate to point at the corresponding pressure (the gauge shows gradations in mmHg, to make measurements comparable to the mercury column sphygmomanome­ ter). Aneroid monitors, like mercury column

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 121

Measurement of Blood Pressure in Cardiovascular Research

sphygmomanometers, require a human lis­ tener to use a stethoscope to detect Korotkoff sounds. Usefulness for Research Purposes. Aneroid devices have several advantages compared to mercury column sphygmomanometers. Aneroid units are mercury-free, and tend to be lighter weight and more portable. The cal­ ibration is easy to check and can be used on most patients. However, the metal in the aneroid devices can fatigue, making readings inaccurate. In addition, aneroid units are eas­ ily damaged during use, and inaccuracy can result from the device being dropped or knocked out of calibration. Because of this, aneroid devices must be calibrated relatively frequently (i.e., around every 6 months), and immediately after being dropped or badly bumped. In addition, aneroid devices suffer from many of the same problems as mercury column sphygmomanometers, as they require a human listener to record Korotkoff sounds.

Oscillometric Measuring Devices Oscillometric instruments do not use a stethoscope or a microphone, although they do require an inflatable arm cuff. The oscil­ lometric method uses the pulsation of the heart as the basis for measurement. When the cuff is inflated past the systolic pressure, and the artery is completely occluded, there is no pulsation below the cuff. As the pressure in the cuff decreases (with the gradual opening of the valve), a sensor located in the monitor detects air pressure fluctuations in the cuff, resulting from the arterial volume changes that occur with pulsatile flow of blood. The pressure at which the oscillations peak corre­ sponds to the mean arterial pressure. Based on the fluctuation in magnitude of these oscillations, the device uses algorithms to cal­ culate the systolic and diastolic pressures. Modern oscillometric devices are either semiautomated (a rubber bulb is used to inflate

the arm cuff, but the systolic and diastolic blood pressures are calculated automatically and displayed on an LCD or other type of readout) or fully automated (same as semiau­ tomatic devices, but cuff inflation is accom­ plished by means of an electric pump). Several types of devices use an oscillomet­ ric algorithm to measure blood pressure. These include ambulatory monitors (see ch. 8, in this volume), home monitors, and office or clinic monitors, which are heavier-duty devices (than home monitors) that in some cases can be programmed to take several measurements, at a specified interval (e.g., 2 minutes). As automated devices are relatively inexpensive and easily available, semiauto­ matic devices are not discussed here. To use an automated oscillometric monitor, an arm cuff is wrapped around the subject’s upper arm, just above the elbow, and an “on” switch is pressed. Training. Very little training is required for the oscillometric method. Usefulness for Research Purposes. One great advantage to the use of oscillometric moni­ tors is that no human listener is required to determine the systolic and diastolic blood pressure. In addition, it is easy to position the cuff. This is important whether it is a research subject (or a patient) taking his or her own pressure at home or a research assis­ tant taking blood pressure measurements in the laboratory or clinic, as the brachial artery can be difficult to locate in some persons. One just has to wrap the cuff around the arm (or wrist or finger in some units). Another advantage is that some monitors can be pro­ grammed to automatically take several mea­ surements serially, and may be accompanied by software that records the individual mea­ surements and computes summary statistics. Some units have a user-adjustable inflation pressure, or they will automatically inflate to the appropriate level, usually 20 to 30 mmHg

121

06 - Luecken - 45370.qxd

122

9/12/2007

4:44 PM

Page 122

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: CARDIOVASCULAR

above the predicted systolic reading, and several models have a memory facility that stores the readings, which can then be down­ loaded later. Automated oscillometric devices are inexpensive and readily available, and, if (noninvasive) beat-to-beat blood pressure (discussed in the next section) is not going to be measured, this method is preferred for research purposes. Cautionary note: these devices may be less accurate in older people, because stiffer vessels interfere with the oscil­ lometric method of signal detection.

Noninvasive Continuous (Beat-to-Beat) Measurement Devices For measurements taken in the labora­ tory, it is highly desirable to be able to inspect the continuous blood pressure series over a period of time, rather than the inter­ mittent measurements provided by ausculta­ tory or oscillometric devices. This is not a concern for epidemiological studies or inter­ vention trials, in which discrete blood pres­ sure measures, usually based on the average of several measurements, are required. In the laboratory, however, when the aim is to study blood pressure variability over time, often measured in different phases of the study (for example, resting baseline, task, baseline recovery from the task; see ch. 7, in this volume, for details), beat-to-beat mea­ sures provide an invaluable tool. Because there are so few options, specific devices and manufacturers are discussed in this section. The Penáz Method (FinapresTM, PortapresTM, and FinometerTM) Until late in the 20th century, the only method that provided a means to inspect the beat-to-beat pattern was intraarterial monitoring, which, because of its invasive nature, has significant limitations for use in most laboratory studies. This changed in the

1980s, with the development of the Finapres monitor (Finapres, Ohmeda Inc., Ohmeda, CO). The operation of this device is based on a model proposed by Penáz in 1973 (Penáz, 1973). Briefly, the Finapres uses an inflatable finger cuff equipped with an infrared photoplethysmograph that measures the blood vol­ ume of the artery under the cuff. The finger cuff contains a valve that is connected to a source of compressed air, an electromagnetic transducer, and the electronics for the plethysmograph. Arterial blood volume is clamped at a predetermined setpoint value at which the arterial wall is considered as unloaded—that is, at a point where the exter­ nal pressure equals the internal pressure. This is why this method is also referred to as the vascular unloading technique. This setpoint is thereafter maintained by continu­ ous adjustments of the cuff pressure, which are triggered by the device in response to intraarterial pressure changes through a servo system. The workings of this device are described in detail by Parati and col­ leagues (Parati, Casadei, Groppelli, DiRienzo, & Mancia, 1989). Unfortunately, the Finapres is no longer manufactured or distributed. Many are still in use in laboratories, although it is almost impossible to find replacement parts or to have them repaired. However, both station­ ary and portable versions have been devel­ oped that use the same algorithms as the original machine (Portapres and Finometer, TNO, Amsterdam, Netherlands). The portable version, the Portapres, is equipped with a hydrostatic height correction system to account for dynamic changes in the position of the instrumented finger as referred to the heart level. The Portapres can, in theory, be used for 24-hour monitoring; however, the cost of the unit (approximately $40,000 US at this time; the stationary Finometer is the same price) may make that impractical. Either machine can be used in the laboratory

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 123

Measurement of Blood Pressure in Cardiovascular Research

for noninvasive beat-to-beat blood pressure monitoring. The Finometer also allows elec­ trocardiographic recording. Both units come with sophisticated software, Modelflow, which uses a three-element Windkessel model to mimic the three most important proper­ ties of aortic input impedance: characteristic impedance, Windkessel compliance, and peripheral resistance (Parati, Omboni, & Mancia, 1995). This allows the noninvasive computation of a number of hemodynamic parameters, including systemic vascular resis­ tance and cardiac output. Reliability and Validity. The Portapres and Finometer employ the identical measuring algorithms as the Finapres; thus, the reliabil­ ity and validity data collected on the Finapres apply to the Finometer as well (Guelen et al., 2003). The Finapres has been demonstrated as a useful alternative to intraarterial blood pressure (BP) measurement in laboratory test­ ing (Blum et al., 1997), as well as in clinical practice (Wesseling, ten Harkel, & van Lieshout, 1991). It has been shown to track intraarterial readings extremely well, even during sudden blood pressure changes (Blum et al., 1997), making it a good candidate for use during laboratory testing. In addition, the measurements taken using the Portapres showed a high correspondence with intraarte­ rial recording between subjects as well as dif­ ferent experimental conditions. This was also true for the beat-to-beat variability, assessed either as the standard deviation or as fre­ quency components, estimated using spectral analysis (Castiglioni et al., 1999; Omboni et al., 1998). The measures are extremely reli­ able because of the large number of measure­ ments that can be collected (Gerin et al., 1993). The TNO Finometer has been vali­ dated, and has met the criteria set out by the American Association of Medical instru­ mentation (Zorn, Wilson, Angel, Zanella, & Alpert, 1997).

Arterial Tonometry:

The Colin BP-508

Tonometry uses a sensor that is placed over the radial artery (wrist) to detect pulse amplitude; the entire sensor moves about (“creeps”) to the site where the widest pulse pressure is detected, using an array of piezo­ elements. Measurement of pulse amplitude provides only the shape of the pulse wave; thus, a calibration using oscillometric brachial artery blood pressure is automati­ cally performed at the beginning of a session, allowing the calibrated continuous measure­ ment of blood pressure to be obtained. To minimize motion artifacts, the sensor is fixed to the wrist by means of a strap wrapped around a forearm support. Reliability and Validity. Some validation data are available (e.g., Imholz, Wieling, vanMontfrans, & Wesseling, 1998); however, such data for different patient populations remain scant at this time. This device has been accepted by the Food and Drug Administra­ tion for blood pressure monitoring. Training Required. Researchers and labora­ tory assistants will need to be specifically trained in the use of these techniques to mea­ sure beat-to-beat blood pressure, as they are quite different from arm cuff-based methods. The manufacturers offer training courses, and use and interpretation may be a topic for conference-based seminars, for example, at the annual meeting of the Society for Psy­ chophysiological Research. Usefulness for Research Purposes. Devices that allow the measurement of (noninvasive) continuous measurements are neither useful nor, in most instances, necessary for epidemi­ ological studies or clinical trials. However, they are extremely useful for laboratory stud­ ies, which focus on processes and mechanisms.

123

06 - Luecken - 45370.qxd

124

9/12/2007

4:44 PM

Page 124

PART II: PHYSIOLOGICAL SYSTEMS AND ASSESSMENTS: CARDIOVASCULAR

The Portapres can, as previously described, be used outside the laboratory, for ambula­ tory monitoring; however, that may not be practical because of its cost, and the risk of its being damaged or not returned. All these devices are useful for tracking changes across laboratory conditions, and assessing the nat­ ural blood pressure variability under differ­ ent controlled conditions; however, they are less useful for determining the absolute blood pressure (Imholz et al., 1998; Zorn et al., 1997). For that purpose, an auscultatory or oscillometric device is more appropriate. It is worth noting that, regarding the Finometer and the Portapres devices, valida­ tion data are still relatively limited at this time (Parati et al, 1995). These devices work on the same principles as the original Finapres, for which a great deal of data is available, but they are not identical. One additional concern with these machines is the automatic servo adjustments that occur when recalibration is initiated during the measurement process, which can cause loss of data during such dynamic changes as occur during mental or physical stress.

BLOOD PRESSURE MEASUREMENT METHODOLOGY The British Hypertension Society provides a great deal of useful information on its website (http://www.bhsoc.org/Blood_pressure_ Publications.stm), including basic rules regarding blood pressure assessment. These guidelines pertain to all monitors that use an arm cuff; many are not relevant for devices such as the Finapres or the Colin monitors. We have included rules and guidelines from other sources as well. Guidelines specific to auscultatory devices were provided in an ear­ lier section. • Ask subjects in advance to wear loosefitting clothing; remove or loosen tight clothing subject might be wearing.

• Instruct subjects to avoid caffeine, exercise, and smoking for 4 hours prior to the measurement. • Subjects should rest quietly in a chair with back supported (not on an exam table) for at least 5 minutes in a quiet room before measurements are taken. • The subject’s arm should be relaxed, and supported at heart level; legs should be uncrossed and feet on the floor. • Subjects should be instructed to relax as much as possible, and not to speak during the measurements. • Use an appropriately-sized arm cuff (Table 6.1 is taken from the BHS guidelines); using too large a cuff can result in underestima­ tion of BP levels; similarly, too small a cuff will lead to overestimation. When looking to purchase any BP monitor, ensure that appropriate-size cuffs are available. • Take a minimum of three measurements, at least 1 minute apart.

The procedures and standards concerning the measurement of blood pressure differ somewhat when blood pressure is measured for research rather than clinical purposes. Shapiro and colleagues published a report that presents guidelines for conducting blood pressure studies and publishing blood pres­ sure findings (Shapiro et al., 1996). Their gen­ eral recommendations are to (a) clearly and completely present details of blood pressure methodology and relevant procedures; (b) use equipment and methods that are supported by established standards, published research, or other experimental evidence; and (c) con­ trol for factors that can influence blood pres­ sure or the interpretation of blood pressure findings. The reader is referred to this report for further detail regarding these recommen­ dations. In the following sections, we review several important issues related to the design and conduct of blood pressure studies.

Methodological Controls Many factors systematically affect blood pressure, including sex, pregnancy,

06 - Luecken - 45370.qxd

9/12/2007

4:44 PM

Page 125

Measurement of Blood Pressure in Cardiovascular Research Table 6.1

Blood Pressure Cuff Sizes for Mercury Sphygmomanometers and Semiautomatic and Ambulatory Monitors (in cm) Widtha,b

Lengtha,b

BHS Guidelines: Bladder Width and Length

Arm Circumference

Small adult/child

10–12

18–24

12 × 18

75th percentile

IGT or IFG

None

None

None

Glucose

(See insulin resistance, above)

IGT or IFG (but not diabetes)

IGT or IFG (but not diabetes)

≥100 mg/dl (includes diabetes)

≥110 mg/dl (includes diabetes)

100 mg/dl or on drug treatment for elevated glucose

Obesity

Men: waist-to-hip ratio >0.90; women: waist-tohip ratio >0.85 and/or Body Mass Index >30 kg/m2

Waist circumference ≥94 cm in men or ≥80 cm in women

Body mass index ≥25 kg/m2

Increased waist Waist circumference circumference >102 (population specific) cm (>40 in.) in men or >88 cm (>35 in.) in women

ATP III (2001)

Page 301

EGIR (1999)

4:44 PM

WHO (1998)

9/12/2007

Assigns Pathogenesis of Metabolic Syndrome to One Core Underlying Feature

Does Not Assign Pathogenesis of Metabolic Syndrome to One Core Underlying Feature

14 - Luecken - 45370.qxd

Table 14.1

Waist circumference >102 cm (>40 in.) in men, >88 cm (>35 in.) in women (population specific)

(Continued)

301

Assigns Pathogenesis of Metabolic Syndrome to One Core Underlying Feature

Does Not Assign Pathogenesis of Metabolic Syndrome to One Core Underlying Feature ATP III (2004-2005)

IDF (2005)

ATP III (2001)

Lipid

Triglycerides ≥150 mg/dl and/or HDLcholesterol