Population Health in America 9780520965294

In this engaging and accessibly written book, Population Health in America weaves demographic data with social theory an

222 60 3MB

English Pages 280 [276] Year 2019

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Contents
Illustrations
Tables
Acknowledgments
1. What Is Population Health and Why Study It in the Twenty-First-Century United States?
2. Historical Trends in U. S. Population Health
3. U. S. Population Health in International Context
4. Spatial and Social Contexts of U. S. Population Health
5. Socioeconomic Status and U. S. Population Health
6. Race / Ethnicity, Nativity, and U. S. Population Health
7. Gender and U. S. Population Health
8. Policy Implications of Population Health Science
Notes
References
Index
Recommend Papers

Population Health in America
 9780520965294

  • 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

Population Health in America

Robert A. Hummer and Erin R. Hamilton

university of california press

University of California Press, one of the most distinguished university presses in the United States, enriches lives around the world by advancing scholarship in the humanities, social sciences, and natural sciences. Its activities are supported by the UC Press Foundation and by philanthropic contributions from individuals and institutions. For more information, visit www.ucpress.edu. University of California Press Oakland, California © 2019 by The Regents of the University of California Library of Congress Cataloging-in-Publication Data Names: Hummer, Robert A., author. | Hamilton, Erin R., 1979- author. Title: Population health in America / Robert A. Hummer and Erin R. Hamilton.

Description: Oakland, California : University of California Press, [2019] | Series: Sociology in the twenty-first century ; 5 | Identifiers: lccn 2018061417 (print) | lccn 2019012657 (ebook) | isbn 9780520965294 (Epub) | isbn 9780520291560 (cloth : alk. paper) | isbn 9780520291577 (pbk. : alk. paper) Subjects: LCSH: Population—Health aspects--United States. | Public health—United States. | Medical policy—United States. Classification: lcc ra418.5.p66 (ebook) | lcc ra418.5.p66 h85 2019 (print) | ddc 362.10973—dc23 LC record available at https://lccn.loc.gov/2018061417

28 10

27 26 25 24 23 22 21 9 8 7 6 5 4 3 2 1

20

19

Population Health in America

sociology in the twenty-first century Edited by John Iceland, Pennsylvania State University This series introduces students to a range of sociological issues of broad interest in the United States today and addresses topics such as race, immigration, gender, the family, education, and social inequality. Each work has a similar structure and approach as follows: •

introduction to the topic’s importance in contemporary society



overview of conceptual issues



review of empirical research including demographic data



cross-national comparisons



discussion of policy debates

These course books highlight findings from current, rigorous research and include personal narratives to illustrate major themes in an accessible manner. The similarity in approach across the series allows instructors to assign them as a featured or supplementary book in various courses. 1. A Portrait of America: The Demographic Perspective, by John Iceland 2. Race and Ethnicity in America, by John Iceland 3. Education in America, by Kimberly A. Goyette 4. Families in America, by Susan L. Brown 5. Population Health in America, by Robert A. Hummer and Erin R. Hamilton

This book is dedicated to our spouses and children: Dawn, Holly, and Chelsea, and Dave, Kai, and Annie.

Contents

List of Illustrations List of Tables Acknowledgments 1. What Is Population Health and Why Study It in the Twenty-First-Century United States? 2. Historical Trends in U. S. Population Health 3. U. S. Population Health in International Context 4. Spatial and Social Contexts of U. S. Population Health 5. Socioeconomic Status and U. S. Population Health

ix xi xiii

1 24 53 76 97

6. Race / Ethnicity, Nativity, and U. S. Population Health 7. Gender and U. S. Population Health

127

8. Policy Implications of Population Health Science

179

Notes References Index

157

195 215 255

Illustrations

figures 1. All-cause mortality, ages 45–54 for U. S. White non-Hispanics (USW), U. S. Hispanics (USH), and six comparison countries / 2 2. Estimated life expectancy at birth, by gender: Death-registration states, 1900–1928, and United States, 1929–2014 / 5 3. Sample Lexis Diagram, 1950–2000 / 13 4. Age-specific mortality rates (log scale) for Japan and the United States, 2014 / 15 5. Age-standardized death rates from four infectious diseases, U. S., 1900–1973 / 33 6. “Aunt Kate,” Kathryn Rose Green Manier, the great-great-aunt of Erin Hamilton, died during the 1918 flu epidemic / 36 7. HIV / AIDS incidence, prevalence, and deaths from 1985 to 2014 / 47 8. U. S. male and female life expectancy at birth relative to 21 other high-income countries, 1980–2006 / 55 9. Ranking of U. S. mortality rates, by age group, among 17 peer countries, 2006–2008 / 56 10. U. S. ranking in cause-specific mortality in comparison with 17 peer countries, 2012 / 56

ix

x

|

Illustrations

The Preston Curve / 58 11. Percentage of adults who are daily smokers in 26 high-income countries / 64 12. Percentage of adults who are overweight or obese in 25 highincome countries / 66 13. Rural and urban U. S. mortality rates, 1968–2008 / 84 14. Life expectancy at birth by level of urbanization across U. S. counties, 2005–2009 / 84 15. Estimated life expectancy at age 25 by educational attainment, gender, and race in the United States, 2010 / 114 16. Conceptual diagram depicting the relationship between socioeconomic status and health / 117 17. Conceptual diagram depicting the relationship between socioeconomic status and health, while taking into account intergenerational and childhood influences / 124 18. Number of health measures on which racial and ethnic groups fare better, no differently, or worse than Whites / 135 19. Infant mortality rate by race / ethnicity and place of origin / 136 20. Life expectancy at birth for Whites and Blacks, and the White– Black gap in life expectancy at birth, from 1900 to 2013 / 137 21. Conceptual model linking race / ethnicity and nativity to health and mortality / 139 Predicted K6 score, by year and DACA eligibility / 150 22. Female and male life expectancy at birth, and the female–male gap in life expectancy, 1940–2013 / 159 23. Male and female age-specific death rates, and the gender mortality rate ratio, by age, 2012 / 160

maps 1. Life expectancy at birth, men and women combined, by county, 2014 / 79 2. Infant mortality rates in the United States, by state, 2013–2015 / 81

Tables

1. Top 10 causes of death in the United States in 2014 and in 1900 / 27 2. Age-specific percentage contributions to gains in U. S. life expectancy at birth, by sex, 1900–1920 / 37 3. Trends in age-adjusted mortality rates for 10 leading causes of death, U. S. males and females, 1960–2014 / 44 4. Age-adjusted trends in adult obesity in the United States, 1960–2014 / 49 5. Measures of infant health by maternal education, United States, 2007–2010 / 110 6. Selected measures of child and adolescent health by family income level in the United States / 111 7. Educational attainment disparities in eight measures of population health for U. S. adults / 112 8. The racial and ethnic makeup of the U. S. population in 2015 / 130 9. Age-adjusted rates of mortality for five major racial / ethnic groups, and rate ratios between certain groups, for all causes and the 15 leading causes of death, 2014 / 133 10. Death rates from major causes of death for U. S. men and women and the gender rate ratio, 2014 / 162

xi

xii

|

Tables

11. Percentages of common disease and health conditions by gender / 163 12. Rates and percentages of health care usage by gender / 164 13. Age-adjusted percentages of health behaviors and risk factors for disease among adult women and men in recent years / 165

Acknowledgments

We are thankful to many people who have provided us with the assistance, encouragement, and ideas that have been essential to this book. John Iceland, Editor of the Sociology in the 21st Century Series at the University of California Press, sought our initial interest in this project and we are thankful to him for giving us the opportunity to pursue a subject about which we are so passionate. Naomi Schneider, Executive Editor at the University of California Press, has been extraordinarily supportive, enthusiastic, and patient with us; we are very thankful to have such a champion of our project. Benjy Malings at the University of California Press has also been fabulous to work with in putting the book together and we are thankful to him for his help in doing so. Michal Engelman, Rachel Kimbro, Patrick Krueger, Stefanie Mollborn, Kristi Williams, and several anonymous reviewers deserve special thanks for their thorough readings and keen insights, from which we benefited in numerous ways. This book is much better because of the time and energy each of you devoted to it along the way. At the University of North Carolina at Chapel Hill, Julianne (Jools) Gruenhagen Key and Alejandro Vazquez were fantastic research assistants and helped us with numerous tasks in putting this book together and seeing it through to publication. David Braudt, Nathan Dollar, Samuel Fishman, and Iliya Gutin also assisted at various times with reading portions of the manuscript and / or constructing tables and graphs; we are thankful to them for their expert assistance. At the University of xiii

xiv

|

Acknowledgments

California at Davis, Megan Mekelburg provided research assistance that was also helpful in the construction of several chapters. We also offer our appreciation to the departments of sociology at UNC Chapel Hill and UC Davis, the Carolina Population Center, our students, and our universities. We are incredibly fortunate to have places to work that are so intellectually stimulating and resource-rich; such contexts clearly facilitate our work in ways that go beyond even our own recognition. Finally, we offer sincere gratitude to our families. Robert Hummer offers his heartfelt love and thanks to Dawn, Holly, and Chelsea; Erin Hamilton gives her love and thanks to Dave, Kai, and Annie and to her mom Donna and dad John. Your insights, support, and patience were essential to us in completing this project, as they are fundamental to everything we do.

chapter 1

What Is Population Health and Why Study It in the TwentyFirst-Century United States?

In November of 2015, Professors Anne Case and Angus Deaton of Princeton University published a short article in the prestigious Proceedings of the National Academy of Sciences that demonstrated an increase in the mortality rate of non-Hispanic White Americans aged 45–54 between 1999 and 2013.1 Interestingly, they showed that this age-specific mortality increase was not experienced in France, Canada, the United Kingdom, Germany, Sweden, or Australia (figure 1). Nor was it experienced by either Latinos, shown in the figure, or African Americans in the United States. The article also showed that the upsurge in the mortality rate among middle-aged U. S. Whites was largely due to sharp increases in the death rate among those with a high school degree or less. The mortality rate increase among Whites was driven by rapid increases in three causes of death—drug and alcohol poisonings (i.e., overdoses), suicide, and chronic liver diseases and cirrhosis. These cause-specific mortality increases were mirrored by trends in related health problems among middle-aged White Americans over this same period of time, including increases in reports of pain, psychological distress, difficulties with routine activities of daily living, heavy alcohol use, and overall poor health. Unlike most academic articles, the Case and Deaton paper created buzz. It was covered by media outlets all over the country and world, including the New York Times, USA Today, National Public Radio, the Washington Post, Al Jazeera, and CNN. Their research methods, the study findings, and the policy implications of their study were debated 1

|

Chapter 1

450

2

deaths per 100,000 300 350 400

USW

FRA

200

250

GER USH UK CAN AUS SWE 1990

2000 year

2010

figure 1. All-cause mortality, ages 45–54 for U. S. White non-Hispanics (USW), U. S. Hispanics (USH), and six comparison countries: France (FRA), Germany (GER), the United Kingdom (UK), Canada (CAN), Australia (AUS), and Sweden (SWE). Source: Case and Deaton 2015

on dozens of social and health science blog sites across the country. On the campaign trail, presidential candidate Hillary Rodham Clinton discussed the study as part of a speech on concerns over the American middle class. And leading syndicated opinion writers, including Paul Krugman and Ross Douthat, wrote about the study in their weekly columns. Such vast attention given to an academic article focused on U. S. mortality rates was not only highly unusual; it was nearly unprecedented. What was all the attention about? While it is true that Professors Case and Deaton are prominent economists—in fact, Deaton won the Nobel Memorial Prize in Economic Science just days before the new study was published—their prominence surely did not explain the hype and debate surrounding the results of their research; something deeper seemed to be triggering the attention of the American media, scientific community, and general public. Perhaps it was the sheer numbers. Indeed, in the “Significance” section of the published study, Case and Deaton calcu-

What Is Population Health and Why Study It? |

3

lated that had the mortality rate among White Americans aged 45–54 continued at the same declining pace that it had experienced over the previous 19 years (1979–98), approximately 500,000 fewer middle-aged Whites would have died between 1999–2013—a death toll comparable to that of the 35-year-old U. S. AIDS epidemic from 1980 through 2015. By anyone’s tally, that is a lot of prematurely lost lives. Another possible reason for the extensive attention could be the fact that the troubling trends were occurring among non-Hispanic Whites, the most economically well-off racial / ethnic group in the United States. By virtue of their advantaged position in the socioeconomic and power hierarchy of American society, it was unusual that such a rise in mortality was occurring among this group and not so among less well-off and less powerful racial / ethnic minority groups. Perhaps this population health trend received so much attention because of White privilege—that is, as a society that is economically and politically dominated by Whites, media, scholarly, and political attention tends to focus on issues affecting White people more so than other groups. Indeed, far more Black Americans die prematurely in eight years than the number of White Americans who died as a result of the rising mortality rate from 1999 to 2013. And later studies show that the trend was in fact worse for American Indian and Alaskan Natives, a group that Case and Deaton left out of their analysis.2 Yet another explanation for the attention could be that in a very wealthy country like the United States, mortality rates are not supposed to increase for any group. Progress in the form of continually decreasing mortality rates and improved health is expected, while upsurges in mortality rates that reflect a decline in the health of the population are both rare and troubling to the country’s collective ego. Thus, perhaps this study attracted such rapid and widespread attention because it signaled something deeply troubling about the health of the nation as a whole. The Case-Deaton article was not the first alarm bell that recently sounded regarding the nation’s health. In 2011 and 2013, respectively, the National Research Council assembled teams of top health and social scientists to produce companion reports on the health of the United States in comparison to other high-income countries such as Canada, Sweden, Spain, Japan, Australia, and others. The 2011 volume largely focused on mortality patterns for those aged 50 and above, while the 2013 report concentrated on a broader array of health and mortality indicators for those aged 0 through 50.3 Both reports showed that the United States fared among the worst overall on nearly all indicators in comparison with the other high-income countries. The overall health

4

|

Chapter 1

and mortality indicators for American women were particularly poor in comparison with women in the other countries, but U. S. men also fared poorly on most measures. These prominent reports, along with other related research articles published around the same time, demonstrated that the United States is not only missing from the world’s best with regard to a wide range of health indicators, but that our collective health profile is close to the bottom among wealthy countries. Ironically, both volumes pointed out that the United States spends far more on health care per person than any of the other comparison countries, suggesting that the poorer overall health conditions in the U. S. are probably not due to a scarcity of health care resources.4 Other prominent studies over the past decade have documented troubling trends in the health profile of particular U. S. population subgroups. While Case and Deaton showed that the increasing mortality rate among middle-aged Whites was largely because those with a high school degree or less experienced an escalating mortality rate between 1999 and 2013, a series of studies over the past 10 years have demonstrated widening gaps in both health and mortality rates when comparing adults with a high school degree or less to those with a college degree or more.5 U. S. women with relatively low education appear to be particularly vulnerable. One study found that women with less than 12 years of schooling have an overall lower life expectancy than they did 40 years ago.6 Concern also exists with regard to racial / ethnic subgroups of the U. S. population. For example, African Americans continue to live nearly four fewer years than Whites, on average, which equates to the premature loss of approximately 83,000 African American lives each year.7 That number equates to a large airplane full of African American residents of the United States crashing without survivors every single day, day in and day out. That is an American tragedy. And while Hispanics currently have a longer life expectancy than either African Americans or Whites, the rate of Hispanic old-age disabilities is the highest in the country.8 This means that many Hispanics, while living long lives on average, face longer periods of suffering in their older years. Finally, a number of highquality studies over the past decade have documented enormous geographic differences in the health of Americans. Some neighborhoods, counties, and states appear to have health profiles much closer to those of other high-income countries, while other neighborhoods, counties, and states appear to be falling further and further behind with regard to their overall levels of health and mortality.9 For example, Christopher Murray and colleagues have shown that African American males in

What Is Population Health and Why Study It? |

5

90 80

Life expectancy in years

70 60 50 40 30 20

0

1900 1904 1908 1912 1916 1920 1924 1928 1932 1936 1940 1944 1948 1952 1956 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012

10

Overall

Male

Female

figure 2. Estimated life expectancy at birth, by gender: Death-registration states, 1900–1928, and United States, 1929–2014. Source: Murphy, Kochanek, Xu, and Arias 2015; Arias 2012

some urban areas live, on average, 20 fewer years than the average lifespan of Asian American females.10 Fortunately, not all U. S. health and mortality trends are disturbing. For example, figure 2 shows that life expectancy increased from about 47 years in 1900 to nearly 79 years in 2014, a change that in and of itself signals a massive improvement in the overall health conditions in the country in just the past 114 years. As figure 2 also shows, these incredible increases in life expectancy were experienced by both women and men, although the gender-specific increases have not always occurred in a parallel fashion. Yet another incredible improvement in health and mortality is evidenced by the long-term declining U. S. infant mortality rate. While about 10% of babies who were born in the United States died before reaching their first birthday in the year 1900, less than 1% of infants died before their first birthday in 2014.11 These increases in life expectancy and decreases in infant mortality are stunning achievements that likewise deserve attention and explanation. Thus, for reasons that are troubling and for others that are worth celebrating, this book delves into the description and explanation of health and longevity patterns and trends in the United States. We refer to the description and explanation of such patterns and trends as “population health,” a term that we formally define below. We seek to paint a clear, contemporary portrait of U. S. population health patterns by digging into

6

|

Chapter 1

patterns for the country as a whole and for different population subgroups defined by socioeconomic status, gender, sexual orientation and gender identity, race / ethnicity, and immigrant status. Throughout this portrayal, we also discuss some key trends in population health across time, depending on how far back valid data allow us to go, and how some U. S. population health indicators stack up against other wealthy countries and, internally, across geographic units. There is much to be learned through these temporal and spatial comparisons. We also aim to shed light on some of the reasons why such patterns and trends in U. S. population health exist. Indeed, there has been a tremendous amount of research published on population health over the last couple of decades by sociologists, demographers, geographers, epidemiologists, economists, social workers, nurses, biologists, and medical doctors, just to name a few of the scientific disciplines engaged in this area of research. Each of these disciplines has contributed to the scientific community’s understanding of population health patterns and trends. While scientific knowledge continues to develop at a fast and furious pace in this area, here we seek to lay out some of the key explanations for the population health patterns and trends that characterize the United States.

what is population health? Population health is an interdisciplinary topic of study that is gaining momentum across the country.12 As one indicator of such momentum, the fledging Interdisciplinary Association for Population Health Science (IAPHS), which is the only U. S. professional association fully dedicated to population health, was incorporated in 2015 and is holding just its fifth annual meeting in the fall of 2019.13 Nonetheless, the study of population health has very strong historical roots in sociology, demography, geography, public health, epidemiology, biology, social work, nursing, medicine, and public policy—and will continue to be closely aligned with those disciplines. Moreover, it is important to note that much research and policymaking focused on population health occurred well before the twenty-first century in these disciplines, although none of them focuses exclusively on this topic. Increasing scientific and policy attention on population health is in part a reaction to an overly narrow individually and medically based conceptualization of health that dominated American research and policymaking throughout the twentieth century and in many ways continues to do so today. In his recent book on this topic, James House argues that

What Is Population Health and Why Study It? |

7

this individualistic orientation to health may be a key reason why the United States spends more on health care than any other country in the world yet has the poorest population health indicators among all wealthy countries.14 While not discounting the importance of individual-level factors (e.g., genetics, behavior) in contributing to health, the modern study of population health centrally concerns itself with the ways that inequalities in social, physical, and policy contexts influence health—both for individuals and for entire geographic areas characterized by such contexts. Further, the twenty-first-century study of population health is critical of a narrow vision of health policy that is solely concerned with medical care. Indeed, there is much more to health policy, pertaining both to individuals and to entire places, than health care. Over the years, for example, U. S. society has developed many housing, environmental, civil rights, criminal justice, education, employment, and income policies that, while not often thought of as health policies, have tremendous influence on the health of the population.15 Overall, then, the study of population health is defined as the documentation of patterns and trends in health within specifically defined geographic places; the explanation of such health patterns and trends in those specific places using a multilevel set of determinants; and the translation of population health research findings into action to improve the health of those specific populations.16 This definition includes four very important components, to which we now turn. First, a core purpose of population health research is the documentation of patterns of health at one point in time and of trends in health across time in specific geographic places. Accurate description necessarily comes before explanation. And accurate description of population health patterns and trends relies on high-quality data sets that are representative of the geographic place under study. Too often, in our view, researchers do not accurately and carefully document patterns and trends before jumping toward explanations. Careful and accurate documentation is a difficult, and underappreciated, component of population health science. Second, the study of population health searches for explanations of the documented patterns and trends across a multilevel set of factors. These multilevel determinants range from the social, environmental, and policy contexts surrounding the people under study; to the social inequalities (e.g., by gender, race, and socioeconomic status) that individuals experience on a daily basis; to the behaviors, health care experiences, and biological characteristics of individuals. Notably, the study of population

8

|

Chapter 1

health pays particularly intense attention to inequalities specific to gender, race / ethnicity, and socioeconomic status (i.e., education, employment, income, and wealth) within this multilevel framework because theory and research suggest that individual behaviors, the use of health care, and even our biological systems are strongly affected by these inequalities. Importantly, Bruce Link and Jo Phelan developed fundamental cause theory over the past 25 years to highlight the critical roles that socioeconomic status (SES) and racism play in influencing overall population health and disparities in health among the U. S. population.17 We discuss fundamental cause theory at length in chapters 5 and 6; further, we provide some thoughts regarding its applicability to gender in chapter 7. In a nutshell, the key idea is that high SES embodies individuals with an array of flexible resources to use on an everyday basis that work to enhance their health and protect against the risk of death. High SES individuals “carry” these flexible resources around with them day in and day out, using them to their health advantage throughout the life course. These flexible resources include knowledge, money, power, prestige, and beneficial social connections; they are termed “flexible” because they can be used in a wide variety of ways. In contrast, racism and sexism work to limit the availability of flexible resources for health by, for example, restricting individuals in disadvantaged groups from living in certain neighborhoods, preventing them from participating in powerful social institutions, and exposing them to greater levels of stress. Third, population health researchers are collectively interested in using research findings to make a difference in improving health patterns and trends. If, for example, U. S. population health patterns and trends are affected by contextual determinants such as federal gun laws and state-level cigarette taxes, and by social inequalities structured by race and gender, why shouldn’t lawmakers and people in power within key social institutions (e.g., universities, school districts, corporations) seriously consider issues of corporate autonomy, tax policy, racial discrimination, and gender equality as population health policies? Chapter 8 examines population health policy options with an eye toward moving beyond typical discussions that focus on improving individuals’ health behavior and providing them with greater access to health care. Finally, our definition of population health relies on a geographicspecific orientation. This is an important definitional feature because researchers and policymakers must clearly understand the specific geographic area that is being studied to develop appropriate policies and programs to improve health in that place. In the case of this book, the

What Is Population Health and Why Study It? |

9

specific geographic area of focus is the United States. Such geographic specificity distinguishes this definition of population health from one that is regularly used in the medical community, where “population health” often refers to the group of patients who particular hospitals or providers are caring for.18 While perhaps useful in the health care arena, this definition of population health is overly narrow in focusing on providers and patients, with little or no applicability beyond each provider’s influence. Instead, we argue that health care is not the only determinant of population health; in fact, we contend that the influence of health care on health is dwarfed by the social and contextual factors that shape health on an everyday basis. Thus, by focusing on specific geographic areas, our definition of population health encompasses the complete set of people and the full range of factors that influence a specific population’s health.

a social demographic perspective of population health The study of population health is inherently interdisciplinary. That is, it brings together researchers from a very wide range of academic disciplines who document new patterns and trends of health, discover new explanations for such patterns and trends, and inform policies and programs to improve population health. As such, this book draws on work from a range of disciplines and on studies from interdisciplinary research teams. Nonetheless, we bring a specific social demographic perspective to the study of population health that draws upon key strengths of sociology and demography. Sociologically, we draw from the discipline’s core foci on social stratification and social context. A simple but catchy definition of social stratification is the understanding of “who gets what and why.”19 In our case, the “who” refers both to people in the country as a whole and its various population subgroups; the “what” refers to good health and long lives; and the why refers to the explanations for patterns and trends in health and longevity, both between the United States and other high-income nations and, within the United States, between population subgroups and geographic areas. More formally, social stratification refers to the systems of inequality that operate within and across societies to create differences in access to and acquisition of valued resources, including education, occupational status, income, wealth, and a healthy and safe environment.20 Thus, a social stratification perspective on racial / ethnic

10

|

Chapter 1

inequalities in health focuses on the ways that institutionalized discrimination (e.g., in schools, workplaces, financial institutions, the criminal justice system, and the health care system) works to influence health and longevity disparities by influencing racial and ethnic differences in access to critical health-related resources.21 And a social stratification perspective on gender disparities in health emphasizes the ways that gender discrimination and gendered opportunities and constraints influence the health and longevity of women and men by, again, differentiating women’s and men’s access to health-relevant resources.22 Not unrelated to issues of social stratification, the importance of social contexts in studies of population health has long been recognized by sociologists. Social contexts refer to the groups and institutions (e.g., families, friendship networks, schools, workplaces, neighborhoods, cities, counties, and states) that structure the norms, behaviors, and health of people who are exposed to such influences.23 One of Emile Durkheim’s classic contributions to the development of sociology as a scientific discipline was his work on differential suicide rates across groups and geographic areas in nineteenth-century Europe, including by religious denomination.24 He demonstrated, for example, that the religious context (e.g., primarily Catholic, Protestant, or Jewish) of different geographic areas was instrumental to the understanding of why suicide rates varied across areas. He also put forward two concepts that continue to help frame present work on understanding why social contexts matter for population health: social integration and social regulation. Social integration refers to the social ties and support that are garnered from social contexts. That is, individuals in some contexts are more likely to be involved in a network of supportive friendships and to participate in healthy social activities in comparison to those in other social contexts. Social regulation is a second important concept to consider. In this case, social contexts help shape the health of individuals living in such contexts, through mechanisms such as formal (e.g., policy) and informal regulations on behavior. In this book, we emphasize multiple levels of social context, including those of the family, friendship networks, schools and workplaces, neighborhoods, and larger geographic units such as cities, counties, and states. The centrality of social stratification and social contexts in our approach to population health does not dismiss the importance of genetic endowments, psychological traits, individual decision-making regarding health behavior, and individually tailored medical care in contributing to the health of individuals. These are all critical factors for understanding

What Is Population Health and Why Study It? |

11

individual-level health. And if, for example, medical care is outstanding in one geographic area and of poor quality in another, such differences in provider care can add up to population health disparities. In other words, health-related decision-making and the use and quality of health care are themselves constrained by social resources. Furthermore, psychological traits and genetic endowments interact with the social environment to affect health. For example, Jason Boardman has shown that people who are genetically susceptible to smoking are especially likely to smoke more in states that have lower taxes on cigarettes; in states with higher cigarette taxes, genetic susceptibility to smoking is lessened.25 Put another way, the social context is a very important modifier of “genetic effects.” Thus, the assessment of population health in the United States and disparities therein necessarily must give central priority to the level and distribution of social resources and the social contexts that influence the overall health of the country and its constituent subgroups. Our perspective on population health also draws heavily from demography, which is the scientific study of human populations. One of the fundamental strengths of demography is its obsession with population representativeness; that is, the data and methods that demographers utilize result in descriptions (e.g., rates) and relationships (e.g., correlations) that are true in the overall population and among its subgroups.26 Population representativeness is accomplished either through the collection and use of complete data for every person in an entire population or through the careful collection and use of samples from the general population who represent the population as a whole. Such population-based health data contrast with health data from hospitals, clinics, or other nonrepresentative samples, such as volunteers for a research study. Studies based on nonrepresentative data cannot make valid scientific claims about the population health of the country as a whole or among its major subgroups because the individuals included in such data sets may differ in important ways from all individuals in that population. Because the focus of this book is population health in the United States, we draw on nationally representative data or on published findings from population-based data sets that allow us to most effectively make accurate statements about patterns and trends of health for the United States as a whole and for many of its largest population subgroups and geographic locales. In short, the demographic approach to population health provides a formidable set of tools to describe population health patterns and trends in the United States and to make cross-national comparisons. The use of population representative data sets and appropriate statistical tools

12

|

Chapter 1

facilitates such accurate description. The sociological approach to population health, in turn, provides a powerful lens through which to view population health patterns and trends in the United States, albeit not to the exclusion of other potentially useful explanatory perspectives. In particular, a social stratification lens to population health focuses on how critical health-related resources are distributed and, in turn, how the distribution of such resources informs patterns and trends of population health. Moreover, a social contextual approach to population health considers the multiple levels of influence (e.g., families and households, schools, workplaces, neighborhoods, counties, states) that individuals are embedded within, thus moving well beyond an individually based approach to the understanding of health. Together, then, key features of sociology and demography combine to comprise the social demographic perspective to population health. This approach is useful not only for describing and explaining health patterns and trends but also for informing health policy at the population level because it is based on representative data and focuses on social and economic resources—like education, income, and wealth—and social contexts—like schools, workplaces, and neighborhoods—that are both health related and policy amenable.27

measuring population health This chapter has already mentioned four key measures of population health—mortality rate, life expectancy, infant mortality rate, and the rate of old age disabilities—without defining these terms or discussing how they are calculated. From our own experience watching television news and reading newspapers and websites, we know that “loose use” of such measures is common. But as scientists, it is important to understand the formal definitions of the most often-utilized measures of population health so that accurate and common understandings can be achieved. Here, we rely on the field of demography to supply us with some very useful concepts and measures to best document patterns and trends in population health. Note that we do not attempt to offer a complete overview of population health concepts and measures. Rather, we provide an introduction to the concepts and measures we use most frequently throughout the book. In each chapter, we also discuss measures of population health in a manner that is as clear as possible. Nonetheless, this short section serves to guide readers through the measures that will be seen most frequently throughout the upcoming chapters.

What Is Population Health and Why Study It? | 50

13

Period

45 Cohort 40 35

Age

30 Age 25 20 15 10 5 0 1950

1960

1970

1980

1990

2000

Time

figure 3. Sample Lexis Diagram, 1950–2000.

To begin, we refer to one of the fundamental tools of demography: The Lexis Diagram (figure 3).28 The Lexis Diagram is a useful way to frame our thinking about the measurement of population health because it specifies time in three uniquely important dimensions. First, on the y-axis of figure 3 is age, in this example ranging from 0 up to 50. Here, we have shaded age 25 across the Lexis Diagram. Using this example, we can refer to the population health of all 25-year-olds who live in a specific geographic area. We can do so for a specific year (such as 1975) or we can track trends in population health for 25-year-olds across historical time, such as between 1950 and 2000. Such age-based indicators are more specifically referred to as age-specific measures of population health. Turning back to the opening paragraphs of this chapter, Case and Deaton’s paper focused on mortality rates for the 45–54 year-old U. S. adult population as they changed between 1997 and 2011; this is a clear example of the use of an age-specific measure to track a trend in population health across time. Age-specific measures give us a sense of how measures of population health vary across stages of the life course. The x-axis in the Lexis Diagram depicts historical time, in years, which can also be referred to as period-specific time. For illustrative purposes, we have shaded the year 1975 in figure 3. Any population health indicator specific to a certain year like 1975, then, is a period-specific measure that refers to the health of persons living in a specific geographic area in that

14

|

Chapter 1

specific year. Period-specific measures are very common in the study of population health; for example, we can refer to the life expectancy of U. S. residents in 1975 or the rate of physical disabilities for the U. S. population aged 65 and over in 1975. Population health data are commonly collected and made available on a period-specific basis, thus facilitating much research using period-specific measures. Period-specific measures give us a sense of how population health varies across historical time. Third, figure 3 also shows a diagonal bar that cuts across both age and historical time. This diagonal bar refers to a birth cohort, which is defined as the complete set of individuals born in a specific year in a specific geographic place. The example provided in figure 3 highlights the birth cohort of 1950. By following this cohort from the bottom left to the top right of the Lexis Diagram, it is clear that this whole birth cohort gets older together and moves through historical time together. Measurementwise, this means that we can track the population health of the 1950 birth cohort as it moves diagonally through historical time and across different stages of the life course. Thus, for example, we can measure the probability that members of this cohort are diagnosed with diabetes by age 40; or, we can compare the population health of this cohort at age 30 (i.e., in 1980) to the population health of the birth cohort of 1970 at age 30 (i.e., in 2000) to best understand how population health differs from cohort to cohort. Cohort-specific measures give us a sense of population health for people who are a specific age within a particular generation. The recognition and understanding of these three unique dimensions of time—age, period, and cohort—are critical to the measurement of population health. Scientists must be careful when describing measures of population health so as to not confuse or mislead readers regarding patterns, which represent the distribution of health at a specific time, and trends, which represent changes in patterns of health across time. In the following three subsections, we first discuss the centrality of age for the measurement of population health, then briefly highlight the most common period- and cohort-specific measures of population health that we use throughout the book. The Fundamental Importance of Age for Measuring Population Health It is not a surprise, but nonetheless of fundamental importance to the understanding of population health, that measures of health and mortality

What Is Population Health and Why Study It? |

15

Logged probability of death

1

0.1

0.01

0.001

0.0001

Japan

95–99

105–109

85–89

75–79

65–69

55–59

45–49

25–29

35–39

15–19

0

5–9

0.00001

United States

figure 4. Age-specific mortality rates (log scale) for Japan and the United States, 2014. Source: University of California, Berkeley (USA) and Max Plank Institute for Demographic Research (Germany), n.d.

vary strikingly by age. For example, children and adolescents (thankfully) have very low rates of most chronic diseases and from most causes of death, while rates for older people tend to be much, much higher. Just this simple reminder of the very strong relationship between age and health is convincing enough that scientists simply cannot ignore age in any useful analysis of population health patterns or trends. Further, populations (e.g., the United States) and subpopulations (e.g., women and men) tend to have different age distributions. The United States, for example, has a much younger age distribution than Japan; the median age of the U. S. population in 2017 was 39.4 while Japan’s median age was 48.7.29 Given the much older age distribution in Japan compared with the United States, Japan’s crude death rate—the number of deaths per 1,000 residents of Japan during 2017—was higher than that of the United States.30 But Japan’s higher crude death rate does not mean its population health is inferior to that of the United States; it simply has an older population than the United States. In fact, Japan has lower age-specific mortality rates at all ages than the United States, as demonstrated in figure 4. Thus, when age is properly accounted for, Japan’s level of population health, at least as

16

|

Chapter 1

indicated by mortality rates, is far superior to that of the United States. Consequently, it is critically important that population health scientists take age into account, particularly when comparing country to country and subgroup to subgroup. Key Period-Specific Measures of Population Health With the central importance of age in mind, we turn for a moment back to the Lexis Diagram presented in figure 3 to discuss period-specific measures of population health. Consider the highlighted year of 1975. Within that particular year, we can consider a range of age-specific measures of population health, such as for 25-year-old individuals as shown on the figure. Most simply, population health scientists present the age-specific percentage of persons who currently are diagnosed with a particular health condition, such as heart disease, or the age-specific percentage of persons who report that they have poor or fair health. Thus, for example, chapter 7 (table 11) shows that 9.6% of U. S. boys and 8.0% of U. S. girls between the ages of 0 and 17 are reported by one of their parents to have asthma. Age-specific percentages are also referred to as prevalence in the epidemiological literature.31 More commonly, population health scientists specify period-specific measures in the form of age-specific rates. Most often, age-specific rates also pertain to a certain year, such as 1975. Age-specific rates are usually constructed in 5-year or 10-year age groupings, such as the mortality rate for 45–54 year-old White adults as discussed above with regard to the Case and Deaton study. Less commonly, but still frequent, population health scientists specify a larger age range for its rates, such as school-aged children 5–17 or older adults 65 and above. Other times, age-specific rates are specific to a single age; for example, the infant mortality rate (IMR), a very common measure of population health that reflects the overall well-being of the youngest members of a population, is specific to individuals aged 0 to 364 days and refers to the number of infants who die in a year per 1,000 infants born in that particular year. More formally, the IMR is calculated as: Number of deaths to infants aged 0–364 days in a specific year Number of live births in that year

×1,000

Thus, for example, the 2014 U. S. IMR of 5.8 means that, for every 1,000 infants born in the United States that year, approximately 6 infants died before reaching their first birthday.32

What Is Population Health and Why Study It? |

17

Age-specific rates are some of the most commonly used period measures of population health; we use them frequently throughout this book. Following Preston et al.,33 they generally take the form of: Number of occurrences (or events) in that age range Mid-year population in that age range

×1,000

This calculation yields the number of occurrences or events (e.g., number of people who report having diabetes; or the number of deaths due to suicide) in that year per 1,000 people within a certain age range. While 1,000 is most often used as the multiplier in age-specific calculations, it is not uncommon for age-specific rates to alternatively be expressed per 100,000 people within an age range. As an example, chapter 7 (table 10) shows that U. S. men aged 20–24 had a rate of 22.9 suicides per 100,000 aged 20–24 in 2014; the suicide rate for women aged 20–24 in that same year was 5.0 per 100,000. Related to age-specific mortality rates is another very commonly used and important measure of period-specific population health—life expectancy. In fact, life expectancy may be the single-most common measure of population health because it effectively summarizes the average length of life of a population. Life expectancy is calculated by using a classic demographic tool called a life table;34 life tables use the complete set of age-specific mortality rates for a population in a specific year as its inputs. Life expectancy refers to the average number of additional years that a person of an exact age can expect to live beyond that age. Thus, life expectancy at birth refers to the number of years a newborn can expect to live, based on the current (period-specific) set of age-specific death rates in that population. Because life expectancy uses age-specific death rates, as a measure it is not influenced by differences in the age structures of populations and therefore can be compared across populations with very different age structures. Figure 2 shows that in 2014 life expectancy at birth for U. S. newborn girls was 81.2 years, while life expectancy at birth for boys was 76.4. Population health scientists also often examine (remaining) life expectancy at age 25 as an indicator of adult population health and (remaining) life expectancy at age 65 as an indicator of old-age population health. Very important to note, though, is that life expectancy—whether at birth, age 25, or age 65—is a period-based measure.35 It is calculated based on the set of age-specific death rates in a particular year; it does not project to the future. That is, if cancer is miraculously cured in 2025, life

18

|

Chapter 1

expectancy at birth (and perhaps especially life expectancy at age 65) will jump up in that year and probably remain higher than it had been in the ensuing years. On the flip side, if a massive influenza epidemic strikes in 2025, life expectancy at birth will decline in that year and perhaps in the next year as well, hopefully then rebounding after that. But the current, period-based measure of life expectancy for a population does not foreshadow such future changes. Rounding out our most commonly used period-specific measures of population health are age-adjusted rates, which are also referred to in the literature as age-standardized rates. Age adjustment is utilized in cases where two populations (e.g., Japan and the United States) or two subgroups (e.g., women and men) have different age distributions, as discussed above. The statistical process of age-adjustment eliminates the influence of age so that the health of populations and subpopulations are comparable. In practice, this usually means that the age structure of one population is used as the standard and that the other population assumes the age structure of the standard population while maintaining its unique population health profile.36 In other cases, a separate age standard is applied to both (or more) populations under study. As an example, we report age-adjusted rates of cause-specific mortality by race / ethnicity in chapter 6 (table 9). Age-adjustment is necessary because the different racial / ethnic groups have such different age structures. We show, for example, that the age-adjusted rate of mortality due to Alzheimer’s disease is highest for Whites and Blacks, and lowest for Hispanics and Asians. Given that the rates are age-adjusted, we can be sure that the disparities are not due to one group having more old people than the others; the influence of age structure has been removed. Key Cohort-Specific Measures of Population Health Our book features very few examples of cohort-specific measures of population health. This is not because cohort-specific measures are unimportant. Quite to the contrary, cohort-specific measures are critical in assessing the population health trajectory of a group of people who were born in a specific year in American society. But two datarelated issues make cohort-specific measures of population health much less well utilized than period-specific measures. First, nearly all population health data collected by the U. S. government is period based rather than cohort based. The National Health Interview Survey37 and the National Health and Nutrition Examination Survey,38 both of which

What Is Population Health and Why Study It? |

19

are collected by the National Center for Health Statistics and are critical to the nation’s understanding of health (and which we use prominently in this book), are cross-sectional (one point in time), period-based surveys. In other words, they do not follow individuals as their lives unfold, within particular birth cohorts, as shown in the diagonal shading of the Lexis Diagram in figure 3. Similarly, U. S. mortality data (e.g., death records) are compiled on an annual basis, thus also facilitating the calculation of mortality patterns and trends in a period-specific manner.39 Second, cohort-specific data take a very long time to collect. Think about the research challenge of following a birth cohort across the life course, from age 0 to 100 or more; the endeavor is very difficult, timeconsuming, and expensive. The study will also inevitably outlive the researchers who begin the project. Fortunately, there are innovative U. S. nationally based data sets that follow individuals in specific birth cohorts across different portions of the life course, such as the Fragile Families and Child Wellbeing Study,40 the National Longitudinal Study of Adolescent to Adult Health,41 and the Health and Retirement Study42. These data sets are very well utilized by researchers and have contributed greatly to the scientific community’s understanding of population health. We discuss selected findings that use such data sets throughout the book. Perhaps most notably, such data sets facilitate the estimation of: (1) longitudinal patterns of health and health behavior, that is, the extent to which individuals’ health changes across time; and (2) the onset, or incidence, of new health problems, such as the age at which individuals are first diagnosed with a disease. Given, though, that cross-sectional data sets are, by far, the dominant mode of U. S.-based governmental data collection, the majority of measures provided in this book are period based.

organization of the book This chapter opened the book by discussing some of the recent troubling trends in population health characterizing the United States, along with some of the tremendous population health achievements that have occurred across a longer time horizon. We also briefly mentioned some of the disparities in population health—by gender, race / ethnicity, and educational attainment—that have characterized U. S. society and continue to do so to the present day. And we briefly discussed that U. S. population health, on the whole, is currently inferior to that of most or even all other similarly wealthy countries in the world—a deeply

20

|

Chapter 1

troubling fact that is worthy of more in-depth exploration and explanation. We began our book in this way both to incite intrigue among readers, but also to simply lay out some of the key reasons why we have become so personally passionate about the study of population health. If nothing else, the basic trends and patterns we discussed in the opening section of the book should alert readers to some critical health and social issues facing contemporary U. S. society. In this opening chapter, we also specified a definition of population health and discussed our social demographic perspective on population health. Indeed, health is still too often considered to be an individual characteristic narrowly influenced by a person’s genetics, behavior, and utilization of medical care. Instead, we actively challenge readers to think about population health as structured by critical dimensions of social stratification, including socioeconomic status, gender, sexual orientation and gender identity, race / ethnicity, and immigrant status, and by key social contexts, such as the families, friend networks, schools, workplaces, neighborhoods, counties, and states within which we live. Chapter 2 begins our more in-depth discussion of population health by laying out historical trends underlying contemporary patterns and explanations of population health. Our historical discussion of population health begins with an overview of the epidemiologic transition as it unfolded in the United States in the nineteenth and first half of the twentieth centuries, and the important roles that social and economic change and improvements in public health infrastructure played in the changing disease patterns and population health over this time. Subsequently, we provide a discussion of more recent population health trends and patterns, largely focusing from 1960 to the present. Here, we also introduce issues of population health disparities to the discussion because they are, and continue to be, important parts of the scientific discourse concerning population health over the past half-century. Such disparities are also in part responsible for the relatively poor population health position in which the United States currently finds itself. We turn to the spatial understanding of U. S. population health patterns and trends in chapter 3. Here, we focus attention on comparisons between the United States and other high-income countries. This section draws upon recent reports from the National Research Council and Institute of Medicine showing that the United States was one of the leading health achievers in the world in 1960, but since then, has fallen further and further behind our high-income peer nations. We review the evidence for this poor standing and discuss some of the key reasons

What Is Population Health and Why Study It? |

21

why the United States has fallen so far behind other similarly wealthy countries. Chapter 4 focuses on various layers of social context and population health within the United States. This chapter begins with geographic patterns of population health within the country and reviews recent research showing that population health measures exhibit tremendous variation across states and counties. Interestingly, while some states such as Virginia and Vermont exhibit population health measures that are comparable to some of the healthiest European countries, others such as West Virginia and Kentucky are characterized by population health patterns that resemble some middle-or even low-income countries around the world. These geographic disparities help to reveal the importance of place for health—how one’s local area and state impact health, and in turn help explain wide disparities in population health across locales in the United States. The latter portion of the chapter focuses on research that demonstrates the importance of more localized social contexts for the understanding of population health. Here, we explore how the neighborhoods, schools, friends, and families among which and whom individuals live, work, and play exhibit impacts on the health of those individuals and, in turn, the health of the population as a whole. The next three chapters of the book (chapters 5 through 7) focus on the demographic and social factors that are the subject of much contemporary sociological and demographic research on population health in the United States, including the relationships between health and: (1) socioeconomic status, (2) race / ethnicity and immigrant status, and (3) gender, gender identity, and sexual orientation, respectively. Our discussion of these characteristics emphasizes the social significance of each of them and their continued (albeit dynamic) importance for population health trends and patterns across the life course. Chapter 5 focuses on socioeconomic status (SES). It discusses the complex and changing ways that key dimensions of SES—education, occupation, income, and wealth—are related to population health in the United States. The relationship between SES and health has become stronger and stronger in recent decades, yielding the largest observed health inequalities by education and income in the nation’s history. We will discuss the reasons why such large socioeconomic disparities have developed, discuss the bidirectional relationship between health and socioeconomic status, and illustrate such patterns and trends with national-level data. Chapter 6 focuses on race / ethnicity and immigrant status. Our discussion concerns the ways that both rigid historically based racism and

22

|

Chapter 1

continued forms of more subtle but institutionalized discrimination influence the population health patterns of minority groups. In some important ways, the population health trends for racial / ethnic minority groups are headed in a favorable direction; for example, the gap between African American life expectancy relative to Whites shrunk from 14 years in 1900 to 3.5 years in 2016. That said, there continue to be wide Black-White and Native American–White disparities in most measures of population health that place African Americans and Native Americans at distinct disadvantages relative to Whites. This chapter reviews data on such trends and patterns and discusses the reasons for them. Moreover, we give careful attention to the rapidly increasing Hispanic and Asian American populations of the United States, drawing on their unique histories and within-group diversity to explore their population health patterns and trends. This chapter includes a discussion of three unique patterns of racial and ethnic disparities in health: the Hispanic Paradox—the fact that some Hispanic groups have health outcomes that are better than expected given their relatively low overall socioeconomic status; the Black-White mental health paradox of better mental health outcomes among African Americans as compared to Whites; and recently rising rates of mortality among middle-aged Whites. Chapter 7 reviews and explores classic debates on gender differences in health and mortality—why women tend to experience more health problems but live longer than men, and how these patterns have changed over time, particularly as a result of changing patterns of smoking. It also incorporates new research on sexual orientation and transgender health, identifying what we know—and what we do not know—about the population health of these important subgroups of the population. We close the book in chapter 8 with a discussion of policy options regarding the future of population health in the United States. Here, we review recent changes in national health care policy and ask readers to consider the options we have as a country, given our relatively poor current standing with regard to population health in comparison to other wealthy nations and the wide disparities in population health that continue to characterize our country. Borrowing from recent work in the area by leading sociologists, epidemiologists, and economists, we lay out broad options for the nation’s future population health trajectory, focusing first on health care system reform and policy, second on policies aimed at changing individual behaviors, and third on a broad package of social and economic policies that double as health policies.

What Is Population Health and Why Study It? |

23

While our future population health challenges are substantial, particularly given where we currently stand, the data that population health scientists have amassed and the discoveries that the field is making should help inform the policy solutions that will be the most promising for the nation’s collective future.

chapter 2

Historical Trends in U. S. Population Health

Chapter 1 briefly discussed the poor population health profile of the United States in the early twenty-first century in comparison with other high-income countries. Simply based on recent research findings highlighting such a comparison, it may be easy to assume that we are living in a troubled period in American history. Are we unlucky to have been born when we were born and to be living in the twenty-first century? Is our personal health at risk? Should we be worried about the health of our family members, neighbors, and coresidents of our local areas and states? As is typical with such questions, the answers are dependent upon the context(s) we are living in. Perhaps most important, it is critical that we think about population health in temporal, or historical, context. Indeed, there are dimensions of U. S. population health to worry about in the present day and, we suspect, to worry about as we plan for the coming decades. We would not be writing this book and you would most likely not be reading it if there was nothing to worry about regarding U. S. population health. And we will spend a great amount of time in this book discussing such concerns. But as we mentioned in chapter 1, there are also reasons to celebrate the population health of our country. Based on current mortality rates, for example, U. S. newborn babies have a life expectancy of 78.6 years. This compares very favorably with an estimated life expectancy at birth of less than half that (i.e., 37 years)

24

Historical Trends

|

25

in 1870.1 In addition, there are now over 53,000 centenarians (people aged 100 or older) in the United States, by far the largest number ever. And with continued decreases in mortality rates among America’s elderly population, projections suggest that the number of centenarians could increase to over 600,000 by 2060.2 Thus, while living to 100 is still very rare, it is increasingly possible to do so. And we are not just living longer than we used to; we also encounter far fewer battles with some of the most terrible diseases that used to wreak havoc on the U. S. population. For example, smallpox sickened and killed untold numbers of Americans in the seventeenth through nineteenth centuries and was one of the diseases responsible for the near genocide of the Native American population. Through the genius of early scientists like Edward Jenner and the advent and persistence of effective vaccination campaigns throughout the nineteenth and twentieth centuries, smallpox has not been responsible for any U. S. deaths since 1949 and was eradicated worldwide by 1980.3 These are all good reasons to be thankful that we are living in the twenty-first century. In this chapter, we provide an overview of the history of U. S. population health. We organize our discussion according to the three stages of the epidemiologic transition, which is defined as the long-term shift in disease and cause of death patterns once dominated by infectious disease pandemics to one now characterized by the predominance of chronic diseases and related causes of death.4 We necessarily rely on mortality data in these sections because historical death records are much more readily available and of higher quality than historical health records. But even U. S. national-level mortality records are incomplete before 1933 and, prior to 1900, completely unavailable. Thus, a majority of the discussion focuses on 1900 to the present, although we also draw on some published mortality and life expectancy figures from the 1800s that are based on either local areas or demographic estimates.5 We spend the remainder of the chapter focusing on population health trends in recent decades, particularly since 1960. It is within this last half-century or so that high-quality population health trend data have become more readily available, allowing researchers to address key questions regarding contemporary changes in population health. For example, which indicators of population health have exhibited especially favorable and unfavorable trends since 1960? What recent trends are especially concerning? And for which subgroups of the population are the trends particularly promising or particularly troubling?

26

|

Chapter 2

the epidemiologic transition Abdel Omran is credited with originating the concept of the epidemiologic transition, the processes by which the population health of societies around the world shifted from low life expectancy settings dominated by infectious diseases caused by microorganisms such as bacteria or pathogens to relatively high life expectancy contexts characterized by chronic diseases as the dominant causes of death. Thus, research on the epidemiologic transition focuses on societal changes in both longevity and disease structure, with particular emphasis placed on reductions in communicable (infectious) diseases and ensuing deaths, along with simultaneous increases in the predominance of noncommunicable (chronic) diseases and related deaths. Clearly, the major causes of death and diseases in the United States changed over the last century in fundamental ways. Table 1 shows the top 10 causes of death in the United States in 1900 and 2014.6 The cause of death structure—which reflects the deadliest diseases of each era—is vastly different in the twenty-first century compared with 1900. In 2014, heart disease and cancer dominated as the leading causes of death in American society, together accounting for 45.9% of all deaths. Moreover, chronic lower respiratory diseases, cerebrovascular diseases, Alzheimer’s disease, and diabetes mellitus—all chronic conditions—each make the top 10 cause of death list in 2014; none of these were specifically on the top 10 list in 1900. Looking back to 1900, influenza and pneumonia was the leading cause of death, and is still the eighth leading cause in 2014, but now accounts for just 2.1% of all deaths compared with an estimated 11.8% in 1900. The second and third leading causes of death in 1900 (tuberculosis; diarrhea, enteritis, and ulceration of the intestines) are no longer among the top 10 causes in the twenty-first century. But the epidemiologic transition did not simply change the causes from which people died. Indeed, we now live in an era in which individuals born in the United States can expect to live longer and healthier lives than could even be imagined just 100–150 years ago. In just this brief slice of historical time, we have aggressively and successfully fought diseases such that long life and good health, at least well into adulthood, are now often taken for granted. As a result, early deaths in the United States are now largely unexpected and, as a result, tragic when they do occur. And while we have plenty of health issues to be concerned about as we age across the life course, the data we will review in this chapter clearly show that we are living healthier lives than Americans did in the not-too-distant past.

Historical Trends table 1

|

27

top 10 causes of death in the united states in 2014 and in 1900

Panel A: 2014 Cause of Death Diseases of the Heart Malignant Neoplasms (Cancer) Chronic Lower Respiratory Diseases Accidents (unintentional injuries) Cerebrovascular Diseases Alzheimer’s Disease Diabetes Mellitus Influenza and Pneumonia Nephritis, Nephrotic Syndrome, and Nephrosis Intentional Self-Harm (suicide)

% of All Deaths 23.4 22.5 5.6 5.2 5.1 3.6 2.9 2.1 1.8 1.6

Panel B: 1900 Pneumonia (all forms) and Influenza Tuberculosis (all forms) Diarrhea, Enteritis, and Ulceration of Intestines Diseases of the Heart Intracranial Lesions of Vascular Origin Nephritis (all forms) All Accidents Cancer and Other Malignant Tumors Senility Diphtheria

11.8 11.3 8.3 8.0 6.2 5.2 4.2 3.7 2.9 2.3

sources: Heron 2016; National Center for Health Statistics 2009

In his 1971 paper that first coined the term “epidemiologic transition” and spelled out epidemiologic transition theory, Omran argued that societies progress through three stages of transition: (1) the Age of Pestilence and Famine, where life expectancy at birth is short (i.e., 20 to 40 years) and infectious diseases are dominant and exhibit unpredictable spikes, or pandemics; (2) the Age of Receding Pandemics, when life expectancy increases to 50 or so years as such unpredictable spikes in infectious diseases become much less frequent; and (3) the Age of Degenerative and Man-Made Diseases, when infectious diseases become less common causes of poor health and death and life expectancy increases and eventually stabilizes at what Omran thought at the time was a ceiling of around 70 to 75 years. The next subsections of this chapter discuss these stages as they unfolded in the United States.

28

|

Chapter 2

The Age of Pestilence and Famine The Age of Pestilence and Famine covers the majority of human history. In the case of the United States, Omran suggested that there was little sustainable improvement in mortality rates or life expectancy at birth until around 1850,7 although more refined estimates from demographic historians show that sustained U. S. mortality declines did not unfold until after the Civil War, probably around 1870.8 Data are poor prior to 1870, but some local area records (e.g., New York City) suggest that the annual crude mortality rate was typically very high (around 30 deaths per 1,000 population, compared with about 7 deaths per 1,000 population today) and exhibited substantial year-to-year fluctuation.9 Although cause of death data from the nineteenth century deserve particular skepticism, the severe upward spikes in mortality rates in certain years between 1800 and 1870 corresponded with well-known outbreaks of yellow fever, cholera, and smallpox. These are all infectious diseases, and outbreaks of them were caused by the miserable socioeconomic and environmental conditions at the time. Moreover, prevention and treatment efforts were rudimentary. As a result, U. S. life expectancy at birth in 1850 and 1860 was probably no higher than 36 years for men and 38 years for women.10 Such low life expectancy levels for men and women were not unique to that time point; while there are no precise estimates of U. S. life expectancy prior to the mid-1800s, scattered estimates from local areas suggest that life expectancy at birth probably ranged between 25 and 40 years through the country’s early history, depending on yearto-year fluctuations in infectious disease pandemics and, in the midnineteenth century, the Civil War. Indeed, the Civil War is estimated to have killed about 500,000 soldiers11 and countless others as a result of the horrific conditions of war, thus raising young men’s mortality rates (and possibly other groups) between 1861 and 1865 and seriously impeding progress toward improved population health. The Age of Pestilence and Famine is characterized by especially poor health and high mortality during infancy (age 0) and early childhood (ages 1–4); these are two of the age ranges within which humans are especially vulnerable to infectious diseases.12 In contexts where knowledge of disease processes is limited, social and environmental conditions are harsh, public health infrastructure is underdeveloped, and medical treatment is deficient, sizable fractions of people die before reaching the age of five. Indeed, the best estimates of U. S. infant and child mortality from the 1850s and 1860s suggest that up to 30% of all newborns in

Historical Trends

|

Box 1: Population Health Achievements in Low-Income Countries The progression of the United States through the epidemiologic transition was attributable to an array of societal changes all working together to advance the health of the nation: improvements in housing and availability of food; educational expansion; increasing social and economic equity for women and racial / ethnic minority groups; greater scientific understanding of the causes of disease; development of public health infrastructure; and advancements in, and widespread access to, modern medicine. Increasing national income, brought about by the Industrial Revolution, was an important factor underlying many of these changes. But money in and of itself does not create better health. To insure a healthy and long-lived society, it is essential that policymakers band together, transcend politics, and make decisions that are in the best interests of the health of the population. What factors have helped to create healthy societies in countries with far fewer economic resources than the United States? John Caldwell took up this question in a pathbreaking paper published back in 1986.a He identified 12 places—Thailand, Ghana, Costa Rica, Kenya, Tanzania, Zaire, India, Jamaica, Burma, China, Sri Lanka, and the Indian State of Kerala—that exhibited especially good population health in the early 1980s, despite their low level of national income. Focusing specifically on the cases of Sri Lanka, Kerala, Costa Rica, and China, he identified several key factors leading to their population health achievements: (1) assuring women’s autonomy, (2) providing educational opportunities for all children, and particularly girls, in places where they have been neglected, (3) providing accessibility to health services for all individuals, (4) creating efficiencies in health services, (5) assuring an egalitarian food distribution, (6) achieving universal immunization, and (7) focusing health service access on the periods of pregnancy and infancy. Caldwell’s paper emphasized that it took broad social consensus to makes these things happen and achieve outstanding population health in the context of low income. On the flip side, he argued that major social and political schisms hamper the creation of a healthy society. Recently Randall Kuhn updated Caldwell’s findings.b Using 2007 data, he identified the following 12 middle- or low-income countries as exhibiting especially favorable population health in the context of their national income: Nicaragua, Eritrea, Paraguay, Vietnam, Comoros, Cuba, Nepal, China, Peru, Costa Rica, Morocco, and Bangladesh. The list exhibits modest overlap with that of Caldwell’s from 25 years earlier (only China and Costa Rica appear on both lists), illustrating that it is possible for even relatively poor countries to develop strategies to dramatically improve their level of population health in a relatively

29

30

|

Chapter 2

short time span. At the same time, Kuhn, like Caldwell, highlighted the roles that social consensus and political will play in the achievement of exceptional population health. Countries with sharp social, ethnic, and political divides tend to not exhibit exceptional population health, perhaps because health becomes more of an afterthought on the national agenda relative to other concerns. On the other hand, places that are not sharply divided, that work together to identify population health needs, and that put their improvement plans into action are much more likely to appear on the list of achievers. a. Caldwell 1986. b. Kuhn 2010.

this period of U. S. history died before reaching age the age of five!13 Moreover, a considerable fraction of those who survived infancy and childhood did so while suffering through serious bouts of disease that potentially affected their physical growth, development, and later life health—a phenomenon that population health scholars Mark Hayward and Bridget Gorman have referred to as the “long arm of childhood.”14 In short, life in the Age of Pestilence and Famine was extraordinarily difficult and population health was far inferior in comparison to what we experience today. Up until 1870 or so, infectious disease outbreaks were common; prevention measures were largely unknown; health care was rudimentary at best; length of life was far too short for far too many people; and the health of the fortunate survivors of the perils of infancy and childhood was often compromised in serious ways throughout the remainder of their life course. Thus, it is unsurprising that chronic diseases and causes of death like heart disease, cancer, stroke, and diabetes were relatively rare until the twentieth century in the United States. Most people simply did not live long enough to develop such conditions. Age of Receding Pandemics Until 1870, the United States showed little sustained improvement in population health. While Omran identified the mid-nineteenth century as the time point in which the United States moved into the Age of Receding Pandemics, his specification of timing was based on limited evidence. The clearest evidence Omran brought to bear for the shift into the second stage of the epidemiologic transition was a graph of changing crude death

Historical Trends

|

31

rates in New York City from 1800 to 1970. This mortality data showed both fewer year-to-year peaks and valleys in the New York City crude death rate beginning in the mid-nineteenth century, along with a generally declining crude death rate from around 30 per 1,000 people in 1850– 70 to about 10 per 1,000 people in 1920.15 But this evidence was only based on one city. More recently, Michael Haines documented that overall U. S. health conditions and mortality rates did not show much improvement between 1800 and 1870, largely because the United States was rapidly urbanizing during this time period and the cities into which people were moving proved to be filthy, disease-ridden, and dangerous places to live and work.16 And the period of 1861–65, of course, was characterized by the horrific toll of the country’s most deadly war.17 Other scholars have shown the average height of U. S. men decreased between 1800 and 1870, which is an indication that population health may have been worsening, rather than improving, across this time period.18 Eventually, though, the decline in the crude death rate in New York City and in other urban and rural areas around the country created an impressive increase in U. S. life expectancy at birth, from around 37 years in 1850 and 1860 to an estimated 47 years in 1900 and 54 years in 1920.19 This very rapid 17-year increase in life expectancy at birth over just a half-century or so corresponded with fewer spikes in the crude death rate, largely due to less frequent outbreaks of infectious diseases. Such a monumental change provided Omran with solid evidence to suggest that the second stage of the epidemiologic transition had unfolded. Why did the second stage unfold? Omran argued that social and economic modernization was largely responsible for the transition into the second stage. By “social and economic modernization,” he referred to the increased economic productivity, improved transportation, better nutrition, enhanced personal hygiene, broader access to education, and more robust housing that were all largely brought about because the Industrial Revolution was unfolding and national wealth was increasing. But this theory of the epidemiologic transition was seriously critiqued in the ensuing years. For one, the explanations Omran provided were not well specified. The time ordering of how modernization factors worked together was not at all clear. For instance, the U. S. economy grew between 1800 and 1870, but a sustained increase in life expectancy occurred only after 1870. Further, Omran did not elaborate on the role that modernization played in effecting change in public health practices or medical science. Omran’s theory thus left a lot unexplained about the transition from the first to the second stage of the epidemiologic transition.

32

|

Chapter 2

Given the complex web of factors involved, it is probably impossible to fully determine the exact mechanisms by which social and economic modernization influenced the onset and progression of the U. S. epidemiologic transition. What is most clear is that modern medical practice had relatively little to do with the major declines in infectious disease outbreaks in the United States, especially prior to 1900 but also extending into the midtwentieth century. This argument was most forcefully made in a series of papers and books by the British medical doctor Thomas McKeown and his colleagues, who focused on the epidemiologic transition in the United Kingdom.20 In the United States, John and Sonja McKinlay wrote the most influential article on the same topic. Their analysis of reductions in U. S. infectious disease mortality between 1900 and 1973 demonstrated that specific medical innovations—particularly the use of immunizations to prevent disease and the use of interventions (e.g., penicillin) to treat disease—were “generally not responsible for most of the modern decline in mortality” in the United States.21 This conclusion was based on McKinlay and McKinlay’s observation that the major reductions in infectious disease mortality occurred prior to the widespread use of the medical measures that were eventually developed to treat them. Figure 5, for example, shows that the age-standardized death rate for tuberculosis exhibited a substantial decline prior to the beginning of the use of the antibiotic isoniazid to treat the disease in 1952. A major exception to this pattern was smallpox. Due to Edward Jenner’s early advances, vaccinations against it were utilized as early as 1796 and in quite widespread fashion by the 1880s.22 However, smallpox made up only a small portion of all deaths in this period.23 At the same time, it is clear that early scientific and social scientific advances fed into very important public health measures to prevent or treat infectious diseases in the mid- and late 1800s.24 Perhaps most famously, John Snow used statistical and spatial data in London to discover that a well on Broad Street was responsible for the cholera epidemic of 1849; removing the pump handle to the well was associated with a sharp decline in cholera deaths.25 The discovery of associations between contaminated water, foul-smelling air, and disease were also being recognized in the late 1800s in the United States, even though the understanding of disease transmission was only rudimentary. But increasing efforts to rid households and local areas of contaminated water, garbage, waste, and horrible smells, while not always effective, were making an impact in the reduction of infectious diseases and related deaths.26 These early efforts, mainly in cities that were character-

Historical Trends

|

33

SCARLET FEVER

MEASLES

0.10

0.10

0.05

0.05 Vaccine

Penicillin 0.00

0.00

TYPHOID

TUBERCULOSIS

2.0 1.5

0.3

1.0

0.2 Izoniazid

0.5

0.1

0.0

0.0 1900

1920

1940

1960

Chloramphenicol

1900

1920

1940

1960

figure 5. Age-standardized death rates from four infectious diseases, U. S., 1900–1973. Source: McKinlay and McKinlay 1977

ized by concentrations of filth and odor, were generally stimulated by miasmatic theories of disease. Such theories emphasized filthy environments as the causal agents of disease, although they did not recognize the actual microorganisms responsible for the transmission of disease, illness, and death.27 While early public health efforts focused on attempts to reduce environmental exposures to filth, dirty water, and foul smells, germ theory was also under development. As early as the 1850s, French scientist Louis Pasteur and other contemporaries were conducting experiments that began to demonstrate that microorganisms were the causal agents of infectious diseases. In the 1870s–90s, German microbiologist Robert Koch made substantial advances in germ theory by identifying specific microorganisms that were the causal agents for specific infectious diseases.28 Clearly one of the most important innovations in the history of human science, germ theory led to discoveries of the microorganisms that cause tuberculosis, typhoid, diphtheria, cholera, and pneumonia— all before 1900. Moreover, these ideas diffused from Europe to the

34

|

Chapter 2

United States and began to translate into medical advances before the turn of the twentieth century, such as the beginning of the use of antiseptic during surgical procedures and the use of the diphtheria antitoxin for patients diagnosed with that dreaded infectious disease.29 Although not the major force behind declining mortality, medical science and improvements in medical practice were making at least some advances between 1870 and 1900 that were contributing to improved U. S. population health during this era. What, then, were the most important reasons why U. S. population health—as gauged by reductions in infectious disease mortality and increases in life expectancy—was advancing so rapidly in the late nineteenth century and early twentieth century? For one, the expansion of public health efforts was critical.30 Such efforts were often led by local area and state governments and included important initiatives such as the building of clean water and sewage systems, the development of regulations to improve workplace safety and the cleanliness of the food supply (e.g., milk, meat), the advent of regularized garbage pickup services, improvements in the training and licensing of physicians and midwives, and the advancement of health education programs for the public.31 Obviously, the initiation and expansion of such public health ideas and programs was related to increasing scientific knowledge underlying disease causation. Scientific progress was driven by increasing national income, the expansion of public education, and the problems of filth and disease in the growing cities of the time.32 In the United States, both the urgent need for improvement and the progress already being made were especially pronounced in industrializing cities in the Northeast and Midwest. In fact, it is well documented that U. S. urban areas were characterized by significantly higher mortality rates than rural areas at the time.33 Thus, the most effective, scientifically based public health developments were most often concentrated in urban areas, where they could have the most impact and where the need was greatest.34 One especially important analysis of the impact of public health initiatives focused on the clean water technologies (i.e., filtration and chlorination) that were put into place by large U. S. cities between 1900 and 1940. Across these 40 years, U. S. life expectancy at birth increased from 47 years to 63 years, the most rapid increase throughout the nation’s history. David Cutler and Grant Miller conducted an innovative and painstaking analysis of overall- and cause-specific mortality rate reductions in 13 major U. S. cities during this time frame and demonstrated that clean water initiatives were responsible for about one-half of the overall mor-

Historical Trends

|

35

tality reduction and up to three-quarters of the reduction in infant and child mortality in the 13 cities.35 Such simple yet enormously important public health measures were the result of both an enhanced understanding of disease processes as well as the economic means to pay for the necessary infrastructure. Thus, rather than thinking about public health measures, scientific progress, urbanization, increasing education, and economic development as competitors in explaining the epidemiologic transition, it is more useful to think about them as forces of social change that, working together, helped reduce infectious disease pandemics and led to longer and healthier lives among the U. S. population. But despite the tremendous gains in life expectancy and the clear reductions in infectious disease outbreaks between 1870 and 1940, the influenza epidemic of 1918 served as a cruel reminder that the fight against infectious disease outbreaks was far from over.36 The 1918 influenza epidemic caused over 500,000 U. S. deaths in one year (and more than 20 million deaths worldwide), including Kathryn Rose Green Manier, the great-great aunt of Erin Hamilton, one of the authors of this book (see figure 6). Another 115,000 or so deaths occurred to American soldiers fighting in World War I in 1917–18,37 increasing the death rate among U. S. young men during these years and resulting in a further reduction in life expectancy. The influenza epidemic and, to a lesser extent, the casualties of World War I resulted in a remarkable short-term decline in estimated U. S. life expectancy at birth from 52 in 1916 to just 39 in 1918 (before rebounding back to 54 in 1920).38 The horrible death toll and sudden drop in life expectancy—largely caused by the influenza epidemic— served as a warning that, while chronic diseases were in the process of replacing infectious diseases as the most prominent causes of death in the United States and other high-income countries, infectious diseases were not going away and would remain forces to be reckoned with. This point was not fully appreciated or understood by many scientists, public health officials, or medical practitioners during the twentieth century, many of whom believed that infectious diseases were becoming a thing of the past. Population Subgroups and the Age of Receding Pandemics Before turning to the third stage of the epidemiologic transition, it is important to discuss the ways that the transition may have unfolded differentially across subgroups of the population. One of Omran’s key theoretical propositions was that the most rapid changes in health and disease patterns during the transition occurred among infants and children.39

36

|

Chapter 2

figure 6. “Aunt Kate,” Kathryn Rose Green Manier, the great-great-aunt of Erin Hamilton, died during the 1918 flu epidemic.

Many scholars have subsequently examined the extent to which specific population subgroups shared in the longevity gains made during the epidemiologic transition. For example, table 2 shows sex-specific gains in U. S. life expectancy at birth in the 20-year time period between 1900 and 1920 and breaks those gains down into the contributions made by specific age groups.40 The overall gains in life expectancy between 1900 and 1920 were substantial, ranging from 6.7 years for females to 7.6 years for males. Table 2 also demonstrates that mortality decreases during infancy and early childhood were responsible for most of the life expectancy increases in this time frame. For example, decreasing infant (age 0) and early childhood (ages 1–4) mortality accounted for 42.4% and 25.2%, respectively,

Historical Trends

|

37

table 2 age-specific percentage contributions to gains in u. s. life expectancy at birth, by sex, 1900–1920 Age

Females

Males

0–1 1–4 5–24 25–44 45–64 65–84 85+ Total Total gain in life expectancy in years, 1900–1920

42.4 % 25.2 10.3 9.9 8.9 3.1 0.2 100.0 6.7

41.9 % 22.3 9.4 12.7 10.0 3.4 0.2 100.0 7.6

sources: Hummer et al. 2009: 528

of the 6.7-year increase in female life expectancy between 1900 and 1920. Clearly, scientific, public health, economic, and educational advances were exhibiting incredible impacts on saving the lives of infants and children during this period of U. S. history, resulting in rapid gains in life expectancy at birth. Omran’s epidemiologic transition theory says little about racial / ethnic differences in population health changes over time, other than to note that the epidemiologic transition seemed to occur earlier and at a more rapid pace for Whites compared with non-Whites.41 Data availability and quality issues have prevented much work on this topic and, in this brief section, we limit the discussion to African Americans and Whites. There are some key indications that U. S. minority populations, and particularly African Americans, may have benefited from the epidemiologic transition at least as much as, if not more than, the White population. In large part, this was because the social, economic, and health conditions were so poor for Blacks prior to the epidemiologic transition. Claude Fischer and Michael Hout’s book on the social history of the twentieth-century United States summarizes that “African-origin Americans began the twentieth century locked in rural isolation, hemmed in by legal discrimination in the South, and held back by the legacies of slavery . . . Their poverty was deeper, their lack of education and industrial skills more glaring, and the prejudice and discrimination they faced far more severe than whites.”42 In such a context, it is not surprising that Samuel Preston and Michael Haines estimated that infant and child mortality among Blacks in 1900 was 56% higher than

38

|

Chapter 2

among Whites at the time and that race was the most important variable in predicting child mortality rates in this period of history. Preston and Haines further explained that “race was a caste-like status in 1900, and the degraded social and economic circumstances of Blacks, who had virtually no chance of entering the mainstream of American life, is undoubtedly reflected in their exceptionally high mortality.”43 The earliest available estimate of U. S. life expectancy at birth for Blacks (in 1900) is only 33 years, 14.6 years shorter than for Whites in that same year.44 As the epidemiologic transition unfolded, then, any improvements in public health infrastructure, nutrition, education, and medicine—to the extent that they were shared—were reaching a Black population in desperate need. While the structural racism of the late nineteenth- and early twentieth-century United States surely favored earlier health and longevity improvements for Whites relative to Blacks, it is clear that great gains in population health and longevity were experienced by both Whites and Blacks as the transition unfolded. By 1920, for example, Black life expectancy was estimated to be 45.3 years; by 1940, Black life expectancy rose to 53.1 years (in comparison with 64.2 years for Whites in 1940).45 Thus, as the second stage of the U. S. epidemiologic transition came to a close and the country shifted into an epidemiologic era dominated by chronic diseases, Blacks still lived significantly shorter and less healthy lives than did Whites. And infectious diseases were hardly a thing of the past, especially given that Blacks still suffered from much higher rates from them than did Whites because of their lower socioeconomic status, poorer quality housing and nutrition, and much lower access to high-quality health care.46 However, the data are also clear that Blacks benefited in major ways from the epidemiologic transition. As a result of the transition, both the largest majority and minority groups in the country were increasingly avoiding the ravages of infectious disease pandemics while, at the same time, beginning to face the very different challenges that chronic diseases posed. The Age of Degenerative and Man-Made Diseases Omran argued that the third stage of the U. S. epidemiologic transition began around 1920. By 1920, U. S. life expectancy at birth was above 50 and, influenza epidemic and World War I aside, generally trending upward. And despite the scourge of the 1918 influenza epidemic, there continued to be fewer and fewer outbreaks of infectious and parasitic diseases taking a toll on the health and longevity of the population as

Historical Trends

|

39

national income increased, living conditions improved, scientific progress on the understanding of disease unfolded, public health initiatives spread, individual-level health knowledge advanced, and medical practice modernized. Moreover, death rates among infants and children continued to decrease as the battle against infectious diseases progressed. Instead, deaths were increasingly concentrated in middle and older adulthood and caused by chronic diseases (e.g., heart disease, stroke, cancer, respiratory conditions). By 1930, for example, Omran showed that heart disease had become the leading cause of death, accounting for an estimated 19% of all deaths compared with just 8% in 1900. By 1960, heart disease accounted for over one-third of all U. S. deaths. Omran also argued that “man-made” conditions (e.g., occupation hazards, suicide, drug dependency, and mental illness), which he defined as being related to stress, were becoming more prominent causes of death as well.47 The chronic conditions that emerged as prominent causes of death in the first half of the twentieth-century United States were not new diseases: heart disease, stroke, cancer, and chronic respiratory disease did not originate in the twentieth century. Their importance increased because of the country’s success in reducing the population’s exposure to infectious diseases and in improving treatment in fighting infectious disease outbreaks when they did occur. As such, the United States was joining other higher-income countries throughout the world in ushering in a new epidemiologic era during which chronic conditions such as heart disease and cancer were becoming increasingly important health conditions and causes of death—mainly due to the fact that more and more people were surviving to the ages at which those conditions tend to develop.48 These newly dominant chronic diseases develop across decades of exposure to social, environmental, behavioral, genetic, and gene-environment risks, thus resulting in deaths that most often occur in middle and older adulthood rather than in infancy, childhood, adolescence, or young adulthood. Subsequently, the scientific, public policy, and medical challenges in tackling such chronic conditions were very different than those of infectious diseases. For example, issues of long-term exposure to risk across the life course—from birth to death— would need to be considered in ways that were not nearly as critical in the battle against the microorganisms that cause infectious diseases. By 1960, female life expectancy at birth was 73.1 years, male life expectancy 66.6 years, and life expectancy for the population as a whole had reached 69.7.49 Not only were these life expectancy figures more than 20 years higher than in 1900 and more than 30 years higher

40

|

Chapter 2

than in 1870, these figures were also about average for high-income countries throughout the world.50 U. S. life expectancy, while not the highest in the world, was also not that far from the highest. Moreover, heart disease, cancer, and stroke accounted for 66% of all deaths in 1960, while the only infectious disease remaining in the top 10 list of causes was the combined category of influenza and pneumonia. There is little question that by 1960 the United States had progressed through the epidemiologic transition and that the country’s overall level of population health had improved in enormous ways since the late nineteenth century. At the same time, scientists like Omran were questioning whether or not further major increases in life expectancy were possible. Indeed, he specifically noted that the Age of Degenerative and Man-Made Diseases is “when mortality continues to decline and eventually approaches stability at a relatively low level.”51 Such a belief in the eventual flattening out of the death rate was anchored in the idea that there had to be biological limits to life; after all, how long could humans possibly live, on average, given that the newly dominant chronic diseases were such deadly forces?52

modern trends in u. s. population health Did U. S. population health continue to improve after life expectancy reached 70 around 1960? Fortunately, the answer is yes. In many ways, such progress over the last 50 or so years has been more difficult to achieve than the gains made earlier. This is because most of the recent gains have necessarily occurred among middle-aged adults and the elderly given that the vast majority of infants and children now survive into adulthood. In other words, the relatively “low hanging fruit”—the reduction of death rates and improved health among infants and children—has already been picked. On the chronic disease front, incredible scientific discoveries have fueled both public health efforts and medical innovation over the last half-century, which have in turn made particularly important inroads in the battles against heart disease, cerebrovascular disease (stroke), and cancer. U. S. life expectancy at birth escalated to a record high of 78.9 years in 2014, with women’s life expectancy hitting 81.3 years and men’s reaching 76.4. Thus, over the 54 years between 1960 and 2014, the United States added 9 years to its overall life expectancy at birth— 8.2 additional years for women and 9.8 years for men.53 Based on these increases in life expectancy between 1960 and the early 2010s, there is

Historical Trends

|

Box 2: The Global Burden of Disease Although the United States and other high-income countries have completed the epidemiologic transition, many low-income countries still suffer high rates of death from infectious and communicable diseases. Data analyzed by the World Health Organization show that the top two causes of death in low-income countries in 2015 were lower respiratory infections and diarrheal diseases; while stroke and heart disease were numbers three and four, HIV / AIDS, tuberculosis, and malaria were numbers five, six, and seven.a By contrast, only lower respiratory infections made the top 10 causes of death in high-income countries in the same year; the other causes of death were chronic and degenerative, including heart disease, Alzheimer’s disease, cancer, and diabetes. The fact that large numbers of people in low-income countries still die from preventable and treatable diseases has been called the “global burden of disease.” Are there lessons that can be applied from the epidemiologic transition in high-income countries to address the global burden of disease? To conduct such an exercise, it is critical to note first that the population health situation in low-income countries today is inherently different from that in high-income countries 100 years ago. Back then, medical technologies—vaccines and antibiotics, especially—did not exist to treat most infectious diseases. Thus, current high-income countries progressed through the epidemiologic transition largely as a result of social, economic, and public health developments. Moreover, the socioeconomic development of high-income countries rested on centuries of colonization and extraction from “less developed” parts of the world. The global burden of disease reflects this history of domination; it also reflects contemporary systems of inequality that allow for hundreds of millions of people to die around the world every year from diseases for which there are known cures. Nevertheless, it can be a useful exercise to look to the past to learn what worked in order to address contemporary problems. One such problem is the recent cholera outbreak in Haiti. Following the massive earthquake there in 2010, cholera was introduced by a UN peacekeeping camp in the Haitian countryside. Cholera had not been present in Haiti, meaning that people there had no natural immunity to the disease, and it quickly spread through the population, sickening more than 800,000 people and killing over 9,500. Cholera is a bacterial infection that causes severe diarrhea and dehydration, and it is spread through contaminated water. In the parts of Haiti hardest hit by the outbreak, most people do not have access to clean water or to systems of sewage removal. In this context, the disease spreads quickly. Cholera can be prevented with a vaccine, and it is highly treatable with rehydration therapy and antibiotics, if it is caught quickly. Although new cases

41

42

|

Chapter 2

of cholera have declined each year since 2014 in Haiti, the disease has reemerged with other crises, such as when Hurricane Matthew struck in 2016. What is the best approach to tackling the cholera outbreak in Haiti? Should an emphasis be placed on medical treatment for those who are sick, on medical prevention through vaccination, or on improving the country’s water and sanitation systems? What are the barriers, limitations, and advantages of each of these approaches? Given what we know about the causes of the epidemiologic transition in the United States, what might be the best approach? Are there reasons to think that the causes of declining deaths from infectious diseases in the United States 100 years ago may not be relevant to Haiti today? If so, what is the best solution? These are the kinds of pressing population health problems facing many low-income countries around the world today. a. See World Health Organization 2018a.

again no question from a population health perspective that it is great to be living in the twenty-first century compared to the mid-twentieth century! There were considerable doubts that such post-1960 gains would be made. In a recent review article tracing the declines in cardiovascular (heart disease and stroke) disease mortality over the past half-century, Mensah and colleagues discuss the view that many physicians shared in the 1950s and 1960s that cardiovascular disease was an inevitable feature of aging in modern societies and that relatively little could be done to change that.54 Cancer morbidity and mortality were considered even more intransigent; even into the 1960s, many forms of cancer were considered unpreventable and largely untreatable.55 But because infant, child, and young adult mortality rates had reached relatively low levels by 1960, the life expectancy gains that were made over the last halfcentury depended more and more on reductions in these major chronic diseases that sicken and kill people in middle and older adulthood. Reductions in heart disease and stroke mortality have been especially impressive over the past half-century. Table 3 summarizes changes in ageadjusted, cause-specific mortality rates in the United States for males and females between 1960 and 2014. Perhaps most impressively, the ageadjusted heart disease mortality rate declined by about 70% for both men and women over this 54-year period, while the male and female ageadjusted rates of cerebrovascular disease mortality each fell by about

Historical Trends

|

43

80%.56 Scientific advances in the understanding of the causes of heart disease and cerebrovascular disease—most importantly, high blood pressure, high cholesterol, poor diet, sedentary lifestyle, cigarette smoking, and stressful life situations—were greatly facilitated by innovative longitudinal data collection projects such as the Framingham (Massachusetts) Heart Study, which began in 1948 and continues today.57 Scientific discoveries regarding the links between these risk factors and chronic diseases fueled the development of programs, policies, and treatments that helped to either reduce the risks of these two chronic diseases or more effectively treat them once diagnosed. On the reduction of risk side, the decline in cigarette smoking from 42% of U. S. adults in 1965 to 17% in 2015 has been incredibly important in reducing the risk of heart disease and cerebrovascular disease among U. S. adults;58 that said, the fact that 15.5% of U. S. adults continue to smoke cigarettes foreshadows millions and millions of smoking-attributable cardiovascular and cerebrovascular deaths in the coming decades. On the diagnosis and treatment side, recent work by Eileen Crimmins and colleagues shows that increasing use of antihypertensive and cholesterol-reducing medications that were developed in recent decades were responsible for substantial declines in cardiovascular disease mortality during the 1990s and 2000s. Other work points to the development of surgical procedures such as angioplasty for greatly improving the survival of persons with diagnosed heart disease.59 All told, close to one-half of the very impressive reduction in heart and cerebrovascular disease mortality over the past half-century is due to the reduction of risk factors, while the other half is due to improvements in the detection and treatment of these diseases.60 Reductions in cancer mortality rates were slower to come. The trend data in table 3 show an increase in cancer mortality rates for men between 1960 and 1980, followed by decreases since then. Between 1980 and 2014, the male age-adjusted cancer mortality rate declined by 29%. For women, there was no perceptible change in the cancer mortality rate between 1960 and 2000. But since the turn of the century, the female cancer mortality rate declined by 18%. The most important reason for the decline in the male cancer mortality rate since 1980 and the decline in the female rate since 2000 is the reduction in cigarette smoking. In 1965, 51.2% of adult men smoked cigarettes in the United States; among women, 33.7% smoked.61 For decades, epidemiologic and laboratory-based scientific evidence had been accumulating on the devastating health and mortality risks of smoking.62 Nonetheless, there was substantial pushback on the part of the tobacco industry

44

|

Chapter 2

table 3 trends in age-adjusted mortality rates (per 100,000 population) for 10 leading causes of death, u. s. males and females, 1960–2014 Panel A: Males Cause of Death

1960

1980

2000

2014

Diseases of the Heart Malignant Neoplasms (Cancers) Unintentional Injuries Motor Vehicle Poisoning Respiratory Diseases Cerebrovascular Diseases (Stroke) Diabetes Mellitus Suicide Alzheimer’s Disease Influenza and Pneumonia Nephritis and Nephrosis All Causes

687.6 225.1 85.5 35.4 2.3 186.1 19.9 20.0

538.9 271.2 69.0 33.6 2.7 49.9 102.2 18.1 19.9

———

———

65.8

42.1 12.2 1348.1

320.0 248.9 49.3 21.7 6.6 55.8 62.4 27.8 17.7 15.2 28.9 16.9 1053.8

210.9 192.9 54.7 15.8 17.3 45.4 36.9 25.6 20.7 20.6 17.8 16.2 855.1

167.6 210.9 37.4 59.1 19.3 22.0 9.5 2.5 23.0 20.7 11.5 4.0 731.4

138.1 131.8 37.1 35.6 28.3 27.3 6.1 9.1 17.2 13.2 11.1 5.8 616.7

———

———

1609.0

Panel B: Females Malignant Neoplasms (Cancers) Diseases of the Heart Respiratory Diseases Cerebrovascular Diseases (Stroke) Alzheimer’s Disease Unintentional Injuries Motor Vehicle Accidents Poisonings Diabetes Mellitus Influenza and Pneumonia Nephritis and Nephrosis Suicide All Causes

168.7 447.0 170.7

166.7 320.8 14.9 91.7

———

———

40.0 11.7 1.1 24.7 43.8

26.1 11.8 1.3 18.0 25.1 7.3 5.7 817.9

———

———

5.6 1105.3

note: “———” means that data are not available for that particular year. sources: National Center for Health Statistics 2016: table 17

and its allies, including some lawmakers and prominent members of the medical community, who disputed or flat out denied the science.63 But with continued push from the parts of the scientific community and from organizations such as the American Cancer Society, momentum in the fight against cigarette smoking eventually gained steam in the U. S. government. Finally, in 1964, the U. S. Surgeon General’s Office issued its first report on the hazards of smoking for population health.64 The report

Historical Trends

|

45

drew very widespread attention. It documented a 70% higher mortality rate among smokers compared to nonsmokers in the United States and causally linked cigarette smoking to lung cancer and chronic bronchitis. It also documented a strong correlation between smoking and coronary heart disease, as well as between smoking and emphysema. Public health action soon followed. In 1965, Congress mandated warning labels on cigarette packages. In 1970, cigarette advertising was banned from television and radio. States began to raise taxes on cigarettes in the 1970s. Public smoking bans on airplanes, on school campuses, in restaurants and bars, and in other public settings began to be instituted in the 1990s. And educational campaigns directed at children and adolescents proliferated.65 As the scientific knowledge about nicotine addiction and the laws and bans regarding smoking became more widespread, norms against smoking diffused through the public, especially among the more highly educated. This diffusion of norms worked to reduce the initiation of smoking and to increase quitting, thereby substantially reducing both the incidence of cancer and mortality rates of smoking-related cancers.66 Because men’s smoking rates fell faster than women’s in the late twentieth century, and the consequences of smoking on mortality lag about 20 years, cancer mortality patterns did not begin to fall for men and women until well after the respective peak smoking years were reached. Unfortunately, an estimated 15.5% of U. S. adults (17.5% of men and 13.5% of women) continue to smoke;67 thus, the effects of smoking-related mortality will continue to be felt for decades and decades to come. At present, nearly 20% of U. S. deaths, or about 480,000 per year, are caused by smoking.68 It is thus impossible to overstate the lethal impact of smoking and its incredible toll on U. S. population health—in the past, present, and future. The future population health of the United States in part depends upon even more aggressive public health efforts to eliminate cigarette smoking; this will continue to be a difficult battle given the powerful marketing and very deep pockets of the tobacco industry. Beyond reductions in smoking, both the earlier detection and more effective treatment of many forms of cancer have been important in reducing mortality rates in recent decades. Improved detection—especially the use of mammograms for breast cancer, PSA tests for prostate cancer, and colonoscopies for colon cancer—has been especially important in identifying cancer earlier than ever before, and public health efforts have been effective in publicizing the importance of such early detection technologies. More important in reducing cancer mortality in the last couple of decades has been the development of treatments—most importantly,

46

|

Chapter 2

chemotherapy, radiation, and surgery–that just a half-century or so ago did not exist or were far less effective and more harmful. Such treatments have been the result of expensive and time-consuming scientific research, which is the key reason why the decline in the cancer mortality rate has lagged behind the heart disease mortality rate decline.69 Nonetheless, the U. S. cancer mortality decline is clearly progressing in the twenty-first century, as shown in table 3, and many forms of cancer are more treatable than ever before. Continued decreases in cancer mortality rates will most likely depend on further decreases in cigarette smoking as well as on the earlier detection and even more effective treatment methods of this dreaded disease. Some Recent Problematic Trends in U. S. Population Health As reviewed above, substantial decreases in cardiovascular disease mortality over the past 50 years and more recent decreases in cancer mortality over the past 25 or so years have been instrumental in pushing U. S. life expectancy up over 78 years. Should such progress continue, the United States could reach a life expectancy at birth figure of 80 years in the not-too-distant future, a population health accomplishment that has already been reached by over 30 countries around the world,70 but which nonetheless would be an achievement worth celebrating in our country. Unfortunately, though, not all U. S. population health trends over the past 50 years have been positive. Such problematic trends also deserve discussion, in part because current and future efforts to improve population health may be especially geared toward them. In this section, we briefly review six troubling trends that have unfolded in recent decades. Perhaps the most challenging population health problem and subsequently most encouraging trend that emerged in the recent era is that of HIV / AIDS. First discovered in the United States among five male patients in Los Angeles hospitals in 1981,71 the HIV virus and its full-blown, corresponding disease AIDS soon developed into a public health emergency because there were no known cause(s) or treatment(s) for the disease. Thousands of people—many but not all of whom were young, gay men living in San Francisco, Los Angeles, New York City, and Miami—were being sickened by and dying from the disease. The fact that HIV was eventually discovered to be an infection transmitted most often through sexual intercourse and intravenous drug use72 raised awareness that the battle against infectious diseases was far from over—even though the United States had progressed through the epidemiologic transition.73

Historical Trends

500 400

Deaths

50 300 40 30

e

nc

ale

v Pre

200

20 100

Prevalence, No. (in thousands)

Classifications and deaths, No. (in thousands)

Incidence

70 60

47

600

90 80

|

10 0

0

14 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 2099 1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 1988 1987 1986 1985 19

Year of classification

figure 7. HIV / AIDS incidence, prevalence, and deaths from 1985 to 2014. Note: Deaths of persons with HIV infection, stage 3 (AIDS) may be due to any cause. Source: Division of HIV / AIDS Prevention 2015

Progress in understanding the causes of HIV / AIDS and the development of potential treatments were especially slow throughout most of the 1980s. This was in large part because the conservative federal government of that era did not fund the scientific studies that were necessary to understand and subsequently prevent, treat, and / or cure the disease.74 Discrimination against gay people was the key reason for the slow governmental response, and tens of thousands of people died as a result. In the late 1980s, after far too many deaths had occurred, appropriate research funding and money for programmatic efforts began to be devoted to HIV / AIDS. Soon after, modes of transmission were definitively documented. Public health programs to reduce the risk of transmission were expanded. Programs to educate the public, including children, about the modes of transmission and about the disease itself were formulated. And drug treatments were developed and tested in scientific laboratories. Figure 7 shows that by 1994, the incidence rate (of newly diagnosed cases) of HIV was finally in decline. In 1996, the Centers for Disease Control and Prevention documented for the first time a decline in the rate of AIDS mortality in comparison with the previous year,75 which signaled a crucial breakthrough in the fight against this dreaded disease. Since then, as figure 7 shows, both new cases of HIV and the HIV / AIDS death rate have continued to decline, which of course are both very positive trends. At the same time, one of the most troubling continued patterns is that racial / ethnic minorities and

48

|

Chapter 2

gay / bisexual men are, by far, at highest risk for both contracting HIV and dying of HIV / AIDS. Clearly, there remains much to do in the effort to both prevent HIV and develop a cure for AIDS, the latter of which is of importance particularly for the minority and gay subpopulations who have the highest current prevalence rates of HIV. Second, as shown above in table 3, cause-specific death rates from poisonings (largely comprised of drug overdoses) and suicide increased for both men and women between 2000 and 2014.76 This troubling trend was highlighted in the opening section of chapter 1. Although table 3 shows such increases in age-adjusted mortality for all ages combined, the increases in these two cause-specific death rates have been concentrated among adults aged 30–64. Anne Case and Angus Deaton have termed such causes “deaths of despair.” These death rate increases, which have been especially large among Native American and Alaskan Natives and less-educated White women, have been accompanied by increases in reports of pain, poor self-rated health, disability, and psychological distress among U. S. adults aged 30–64 since 2000.77 Research is ongoing regarding why such despair-related health and mortality trends have unfolded in the twenty-first century. One major reason is the proliferation of opioid prescriptions (i.e., “painkillers”) written by medical professionals but brought about by the development and aggressive marketing of such drugs by the pharmaceutical industry. Other related explanations include the loss of well-paying blue-collar manufacturing jobs in many areas of the country; increasing income inequality and its effects on lowincome and low-educated individuals; the relative lack of economic supports (e.g., unemployment benefits, subsidized childcare) and healthcare options (e.g., drug treatment facilities) for those in need; and the effects of the wars in Iraq and Afghanistan on the mental health of service members and their families. The explanation(s) of these troubling trends in drug poisoning and suicide mortality will need to account for the fact that the increasing death rates from these causes were most pronounced for Native American and Alaskan Indians and relatively low-educated Whites in the United States. Third, it is most likely no surprise that the increasing prevalence of obesity is an especially troubling U. S. population health trend. For adults, obesity is defined as an individual having a body mass index (BMI) of 30 or above, with BMI calculated by dividing a person’s weight in kilograms by the square of their height in meters.78 Table 4 shows changes in the sex-specific prevalence rates of adult (ages 20–74) obesity in the United States since 1960. The trends are alarming. Among

Historical Trends

|

49

table 4 age-adjusted trends in adult (ages 20–74) obesity in the united states, 1960–2014 % Obese Years 1960–62 1971–74 1976–80 1988–94 2001–02 2013–14

Women

Men

15.8 16.6 17.0 25.9 34.1 41.0

10.7 12.1 12.7 20.5 28.3 35.5

sources: Fryar, Carroll, and Ogden 2016: table 1

adult women, obesity increased from 16% to 41% over the last halfcentury; among men, obesity prevalence more than tripled from 11% to 36%. Amazingly, cardiovascular disease mortality exhibited rapid, albeit slowing, declines over this 50-year period, as discussed above. Looking ahead, though, cardiovascular mortality rates are not likely to continue declining. One of the key reasons is because more and more children also became obese in recent decades; indeed, the most recently available national data shows that 17% of U. S. children are now considered to be obese compared with just 10% in 1990.79 As these obese children become adults, their risks of diabetes, hypertension, heart disease, stroke, and even some cancers will be especially elevated given the long length of time they are living with obesity. Samuel Preston and colleagues recently projected that obesity may have such a negative impact on future U. S. population health that life expectancy could virtually stall in the coming decades, despite continued projected declines in smoking.80 Moreover, the increased prevalence of both childhood and adult obesity is especially concerning not only because obesity exhibits major impacts on long-term chronic diseases such as diabetes, hypertension, and heart disease, but also because it has caused increasing levels of disability among middle-aged adults in recent years.81 Thus, it is a very costly condition for individuals, families, the economy, and the health care system. Without doubt, obesity is a population health problem that will need tremendous attention from researchers, teachers, clinicians, and policymakers in the coming decades. And similar to the fights against opioid abuse and cigarette smoking, reducing obesity will necessitate waging public health battles against wealthy segments of the

50

|

Chapter 2

food, beverage, restaurant, transportation, and electronics industries, each of which has played at least some part in the rapid escalation of obesity rates over the past half-century. A fourth recently emerging and concerning trend involves activity limitations among the older adult population. Activity limitations are measured on social and health surveys by asking or physically assessing whether individuals have difficulty with or cannot perform routine activities such as climbing a flight of stairs, walking across a room, feeding oneself, making a phone call, or managing money. For decades, one of the key population health concerns in the United States was whether increasing life expectancy would be accompanied by decreasing activity limitations among the elderly. In other words, are the additional years of life healthy years or unhealthy years?82 Population health data from the 1980s and 1990s clearly indicated that increasing life expectancy was being accompanied by decreases in activity limitations in older adulthood—a very positive development.83 But in more recent years, the decreasing trend in activity limitations among adults between the ages of 65–84 has stalled; moreover, although activity limitations are quite low among adults aged 55–64, recent data show increases in activity limitations among this age group.84 It is unclear why the decline in activity limitations among the elderly has stalled and why there are signs of increasing activity limitations among middle-aged adults. The increasing percentage of obesity is one possibility. Looking ahead, population health researchers will need to continue to closely monitor trends in both life expectancy and activity limitations, particularly as the cohorts of adults with high rates of obesity age into older adulthood. Turning away from specific health issues and toward worrisome trends among demographic groups, general population health trends for U. S. women are of considerable concern. We spend more time on this issue in chapter 7. In brief, a major recent report showed that U. S. women’s life expectancy increased by only 3.3 years between 1980 and 2007; in contrast, the average gain among 21 other high-income comparison countries was over five years. As a result, U. S. women’s life expectancy is now the absolute lowest among all 22 high-income countries analyzed in the report.85 Trends in life expectancy for White women with relatively low education are particularly troubling: those with a high school degree gained less than one year of life expectancy between 1990 and 2010, while those with less than a high school degree lost three years of life expectancy between 1990 and 2010.86 The troubling life expectancy trends for U. S. women are accompanied by some dis-

Historical Trends

|

51

concerting health trends. For example, in a recent comparative study of women in 11 high-income countries with high-quality data, 11.7% of U. S. women aged 50–54 were found to have high or very high cardiovascular risk, with a probability of heart attack within five years at least .20. In the 11 high-income comparison countries, just 4.5% of women aged 50–54 were deemed to have such high or very high cardiovascular risk.87 Undoubtedly, population health trends for U. S. women have not unfolded nearly as well in comparison with women in other highincome countries around the globe. A final concerning trend, and one which we will spend considerable time on in chapter 5, is that of widening socioeconomic disparities in population health. The literature on this topic is very large and the data are clear. Across a wide range of population health measures—including mortality rates and life expectancy, self-reports of physical and mental health, and health measured using biomarkers—highly educated and high-income adults are living longer and healthier lives than ever before in the nation’s history. On the other hand, those with low levels of education (e.g., a high school degree or less) or those living in households with low income (e.g., those in the bottom 25% of the income distribution) exhibit either stagnating or decreasing life expectancy and health trends that are worsening rather than improving.88 As just one indicator of current socioeconomic disparities, women with a college degree or higher are now living 10–12 years longer, on average, than women without a high school degree; for men, the gap is even larger, at 12–16 years.89 Importantly, widening socioeconomic disparities in population health is not a new trend; this trend may go as far back as 1960, a time in which socioeconomic disparities in health were manifest but not nearly as pronounced as they are now.90 It is not a stretch to say that widening socioeconomic disparities is one of the country’s most pressing population health problems of the twenty-first century. This chapter has provided an overview of long-term and recent trends in U. S. population health. While we have made enormous progress in reducing death rates and incidence rates of heart disease, cancer, and stroke in recent decades, there are also troubling signs of stagnation in these trends. And while the recent opioid epidemic has justifiably generated much public attention, obesity and smoking remain the most important risk factors for chronic disease. Reducing their impacts is going to require multifaceted approaches that focus not only on changing individual-level behavior and improving medical access and technologies,

52

|

Chapter 2

but also on reducing the harmful effects that some industries (e.g., pharmaceutical, tobacco, fast food, processed food, alcoholic and sugary beverages) have had on population health over the last century. Given the power and wealth of those industries, that will be a very tall order. Beyond the challenge of chronic diseases, the 1918 influenza epidemic, the HIV / AIDS epidemic, and numerous other infectious disease pandemics (e.g., the recent outbreaks of the Zika and Ebola viruses) have provided us with stark reminders that infectious and parasitic diseases have not only not gone away, but that they also continue to pose enormous threats to population health. Moreover, issues of climate change, environmental degradation, war, and terrorism, combined with very high levels of social and economic inequality, pose serious sociopolitical threats for infectious disease outbreaks well into the foreseeable future. Thus, infectious diseases are not going away anytime soon while, at the same time, chronic diseases are unfolding in our bodies from the time we are in utero until death. The continued fight to improve population health will undoubtedly require educational, scientific, public health, and medical advances that move beyond those that were instrumental in the unfolding of the epidemiologic transition. Moreover, the sociopolitical and industry threats mentioned above also deserve a central place on the health policy agenda to help prevent future infectious disease outbreaks and reduce the incidence and mortality rates of our key chronic diseases. In other words, the health policy agenda looking ahead cannot simply be focused on improving access to health care and developing better medical treatments for people with disease; a much broader lens is necessary.91

chapter 3

U. S. Population Health in International Context

In 2009, as President Obama led the fight to pass the U. S. Patient Protection and Affordable Care Act (the ACA), New York Times columnist Nicholas Kristof published a column titled “Unhealthy America.” “The single greatest myth in the U. S. health care debate,” he wrote, is “the selfaggrandizing delusion” that the United States has “the greatest health care system in the world.”1 In fact, he pointed out in statistic after statistic, the United States ranks below most high-income countries—and, indeed, below many middle-income countries—in numerous measures of population health, including many of the most critical ones: life expectancy at birth; infant, child, and maternal mortality rates; and rates of preventable cause mortality. In 2016, the United States ranked 34rd among nations on life expectancy and 40th on healthy life expectancy, and in 2015, it ranked 46th for maternal mortality.2 If the United States has the “greatest health care system in the world,” Kristof asked, why are we failing so dramatically in comparison to peer countries on key indicators of population health? This question animates a large body of research, which was recently summarized in two important reports by committees produced by the National Academy of Sciences. One report focused on population health after age 50, the other before age 50. These reports show that the United States does not in fact have the “greatest health care system in the world” on many measures, and the limited U. S. health care system is indeed part of the reason for the poor international health 53

54

|

Chapter 3

ranking. But the reports also make clear that health care is not the only, and certainly not the most important, determinant of a country’s population health. Rather, there is no single answer to the question of why the United States ranks so poorly in population health. Many different facets of social life, policy, and health care matter. As a result, it is difficult to summarize simply why the United States fares so poorly. However, as we review the research investigating this troubling international comparison, we invite you to think about what might tie these disparate forces together to create a uniquely inferior context for population health in the United States, as compared to other wealthy nations.

patterns in the data Why does the United States lag behind its high-income peer countries with regard to population health? Patterns in the data suggest important clues. First of all, we know that the U. S. disadvantage emerged after 1980. Figure 8, taken from the 2011 National Academy of Sciences report, shows trends in male and female life expectancy at birth in the United States each year from 1980 to 2005 in comparison to life expectancy at birth in each of 21 peer countries, with life expectancy for the U. S. marked in black.3 Whereas the United States ranked somewhere in the middle of the distribution for both men and women in 1980, it fell to the bottom of the distribution by the end of the period. Although life expectancy increased over this period in the United States as elsewhere, it grew more slowly in the U. S. than in other countries. In 2014, life expectancy declined in many high-income countries around the world, but it rebounded in nearly all places in 2015—one exception was the United States.4 A second clue involves comparisons across the life course. The U. S. health disadvantage is experienced at nearly all ages except the oldest. As figure 9 shows, the United States ranks last or near last among 17 peer countries on age-specific mortality rates from age 0 to about age 70. After that, the rank steadily improves; at the oldest ages, the United States outperforms all other peer countries. Much of the U. S. disadvantage is concentrated below age 50. Deaths before age 50 account for two-thirds of men’s disadvantage in life expectancy at birth and twofifths of women’s disadvantage in life expectancy at birth.5 Thus, it is clear that population health in the United States is especially poor among infants, children, adolescents, and young adults in comparison with other high-income countries.

International Context | 55 87 85

Life expectancy at birth

83 81 79 77 75 73 71 69 67 1980

1985

1990

1995

2000

2005

1995

2000

2005

Year

87 85

Life expectancy at birth

83 81 79 77 75 73 71 69 67 1980

1985

1990 Year

figure 8. U. S. male (top) and female (bottom) life expectancy at birth relative to 21 other high-income countries, 1980–2006. Source: National Research Council et al. 2011

A third clue is that the U. S. disadvantage is observed for a wide variety of causes of death and illnesses / conditions. Figure 10 shows the U. S. rank in 2012 relative to 17 peer countries on age-standardized mortality rates by cause of death. The United States ranks last on five rates, including the mortality rates from all causes, infectious and parasitic diseases, diseases of the genitourinary system, pregnancy and childbirth-related

1 3 5

Rank

7 9 11 13 15

0–

1 1– 4 5– 10 9 –1 15 4 –1 20 9 –2 25 4 –2 30 9 –3 35 4 –3 40 9 –4 45 4 –4 50 9 –5 55 4 –5 60 9 –6 65 4 –6 70 9 –7 75 4 –7 80 9 –8 85 4 –8 90 9 –9 95 4 –9 9

17

Age Males

Females

figure 9. Ranking of U. S. mortality rates, by age group, among 17 peer countries, 2006–2008. Source: National Research Council et al. 2013

External causes of death Congenital malformations, deformations, and abnormalities Diseases in pregnancy, childbirth, and puerperium

17 16 17

Diseases of the genitourinary system

17

Diseases of the musculoskeletal system

12.5

Diseases of the skin and subcutaneous tissue

12

Diseases of the digestive system

12

Diseases of the respiratory system

14

Diseases of the circulatory system

14

Diseases of the nervous system

16

Mental and behavioral disorders

15

Endocrine, nutritional, and metabolic diseases

15

Diseases of blood and immune mechanism Neoplasms

16 6

Infectious and parasitic diseases

17

All causes

17

figure 10. U. S. ranking in cause-specific mortality in comparison with 17 peer countries, 2012. Source: World Health Organization 2014

International Context | 57

causes, and external causes. The United States ranks among the bottom four on an additional seven measures. The United States only ranks in the top 10 for the death rate from neoplasms (cancer). Out of the same 17 peer countries, the United States also ranks last on many health-related indicators, including composite indicators for health in infancy and early childhood (ages 0–4), school ages (5–17), young adulthood (18–34), and early to mid-adulthood (ages 35–49).6 For example, among school-aged children, the United States ranks last on overweight among girls, the adolescent birth rate, and days of life lost for girls and boys, and ranks second to last on overweight and HIV among boys. Among adults in early to mid-adulthood, the United States ranks last or second to last on average body mass index, diabetes, average fasting plasma glucose, and days of life lost. Other studies also find that the United States has higher prevalence of heart disease, hypertension, high cholesterol, cerebrovascular disease, diabetes, chronic lung disease, asthma, arthritis, cancer, and activity limitations than adults in European countries and Japan.7 Adults over 50 in the United States have twice the risk of comorbidities (i.e., experiencing more than one health problem) than in Canada, Denmark, England, France, Italy, Japan, Netherlands, and Spain.8 In other words, these comparative data make it abundantly clear that people in the United States live shorter and less healthy lives than their peers in other high-income countries. The population health literature has begun to explore the reasons why the United States has fallen so far behind its high-income counterparts. Three explanations can be thought of as proximate determinants of health: the health care system, health behaviors, and the built environment. Three others are more distal: social inequality, social policy, and culture. For a variety of reasons, the research literature has focused on the health care system and health behaviors more so than on the others, at least in terms of explaining international differences in population health. This emphasis reflects less the argument that health care and behaviors are more important than more structural determinants, and more the fact that the former are easier to measure and compare across countries, as well as easier to manipulate in a health policy sense. We describe each argument, and the research literature in support of it, in turn.

the health care system The relatively poor U. S. population health ranking may reflect limitations of the health care system. Most important among these limitations

|

Chapter 3

Box 3: National Wealth and Population Health in Countries around the World This chapter has emphasized the overall poor population health of the United States in comparison to other high-income countries. This is a remarkably distressing phenomenon in part because of the generally accepted assumption that wealth is supposed to buy good health. But just how strongly is national wealth actually related to population health? In a groundbreaking study published back in 1975, Samuel Preston illustrated the very strong but complex relationship between national wealth and health.a The paper showed that nations with greater wealth were generally characterized by higher life expectancy compared to less wealthy countries, as might be expected (see the figure below, which uses much more contemporary data than Preston used in his 1975 paper). However, the wealth-health relationship is especially strong at lower levels of wealth. Among fairly-to-very wealthy countries, the relationship between wealth and health is considerably weaker. This suggests that (at the national level) extreme wealth does not necessarily buy better health—at least in comparison to countries that have relatively high levels of wealth. Clearly, as we have discussed in this chapter, the extraordinary wealth of the United States has not led to better overall population health in comparison to other coun-

The Preston Curve 90

Spain

Japan

South Korea 85

France Switzerland Ireland

Singapore

Mexico Argentina

Life expectancy at birth (years)

58

80 UK

Bangladesh 75

USA China Brazil Russia Indonesia India

70

Qatar

Saudi Arabia United Arab Emirates

65 Pakistan 60

Namibia South Africa Equatorial Guinea

55 Nigeria 50

0

20,000 40,000 60,000 80,000 100,000 120,000 Gross national income in purchasing power parity per capita, 2016

The Preston Curve. Source: Population Reference Bureau

140,000

International Context | 59

tries that are also relatively wealthy. The curvilinear relationship between national wealth and health, shown in the graph using data from 2015, has been termed the “Preston Curve” and continues to inspire debate and inform policy on population health around the world.b a. Preston 1975. b. Bloom and Canning 2007.

is the disjointed structure of the system itself: the U. S. provision and payment of health care is fractured across multiple subsystems. Indeed, the United States is unique among high-income countries for not providing a universal, publicly funded system of health care; instead, the U. S. system combines elements of public systems with a large private payment and provider system. Parts of the U. S. system are similar to the systems of other high-income countries. For example, in the United Kingdom, the National Health Service is a government-run health care system that provides health care to all British citizens, paid for through taxes. This is much like the U. S. Veterans Health Administration, a program providing health care to veterans, also paid for through taxes. In Canada, a national health insurance program pays for health care from private providers for all Canadians. This is much like the U. S. public health insurance programs Medicare, Medicaid, and the Children’s Health Insurance Program (CHIP), which provide health insurance to the elderly, the disabled, and low-income adults and children who then obtain care from private providers. While the VA and Medicare, Medicaid, and CHIP are similar to health care systems of other high-income countries, in the United States they only cover a subset of the population, whereas the comparable systems in the United Kingdom and Canada are universal. In the United States, a large portion of Americans pays for health care from private providers through employer-sponsored, private, for-profit insurance companies. Among wealthy countries, a large private, forprofit provider and payment system of health care is uniquely American. A central problem with the three-part U. S. health care system just described is that many Americans fall outside of it. They include mostly non-elderly Americans whose employers do not provide health insurance benefits, who are self-employed, or who are unemployed or out of the labor force and do not qualify for Medicaid. These folks may purchase insurance on the private market as an individual. But because insurance

60

|

Chapter 3

works by spreading the costs of health care across a pool of people, insurance plans were historically unaffordable or unavailable to individuals on the private market who do not share a group with whom to negotiate down the costs of insurance. With very high costs for individual plans, the healthy tended to opt out of buying insurance. Those who might be willing to pay the costs were typically individuals who needed care—i.e., “high risk” individuals or individuals with preexisting conditions. Insurers avoided the costs associated with high-risk individuals by refusing to provide coverage for them or by charging exorbitant premiums. The ACA attempted to address these problems by creating state-based health insurance exchanges that pool individuals into one insurance marketplace, mandating that all individuals without employer or public insurance purchase insurance through the marketplace, providing subsidies to those who could not afford the premiums, and prohibiting the denial of coverage to individuals with preexisting conditions, among other provisions. The marketplace would theoretically lower the costs of insurance by creating competition between insurance companies and by incorporating healthy individuals into the insurance pool through the mandate. The ACA has faced numerous challenges, both logistical and political. However, by one key measure the ACA is viewed as a success: the proportion of non-elderly, uninsured Americans fell from 18.2% in 2010, when the ACA was passed, to 10% in 2016.9 The “leftover” 10% of non-elderly people in the United States who lacked insurance in 2016 includes mostly poor adults who rely on emergency care, public health clinics, and hospitals. It also includes the 2.5% of the U. S. population who are undocumented immigrants, most of whom are barred from receiving federally funded health insurance, including ACA subsidies. But as we write this, the U. S. Congress is passing legislation, signed by President Trump, that eliminates the mandate that individuals purchase health insurance. The extent to which this change in the law will impact the rate of uninsured individuals is unknown, although it is very likely that there will be considerably more uninsured persons in 2018 and beyond compared with 2016– 17. Because the individuals who opt out are likely to be younger and healthier than those who continue to purchase insurance, premiums for those who opt in will most likely rise. The various systems of payment and provision in the United States mean that the U. S. health care system is highly fractured, costly, inefficient, and unequal in comparison to other high-income countries. Indeed, the United States spends more money than any other wealthy country on health care—16.9% of GDP in 2015, which equated to $9,451 per per-

International Context | 61

son in that year—yet more people in the United States lack basic access to health care than in any other Organization for Economic Co-operation and Development (OECD) country. In 2015, 10% of the U. S. population lacked public or private primary health insurance. In only eight other OECD countries—Poland, Mexico, Estonia, Chile, the Slovak Republic, Hungary, Luxembourg, and Turkey—did more than 1% of the population lack health insurance coverage in 2015. By contrast, in 14 OECD countries (Canada, Denmark, Finland, Iceland, Israel, Korea, New Zealand, Norway, Portugal, Slovenia, Sweden, Switzerland, the U. K., and the Czech Republic), 100% of the population was insured in 2015. Studies also find that Americans are less likely to receive advice about health maintenance and disease prevention than are people in other highly developed countries, that the United States has a shortage of primary care physicians, and that a large portion of households—as many as one out of three—report problems paying medical bills.10 The fact that the United States spends more on health care yet is unable to provide basic insurance coverage and primary care to everyone in the country reflects a troubling degree of inefficiency and inequality in the system. While the limitations of the U. S. health care system are well recognized and not in debate, their role in explaining the relatively poor U. S. health ranking is much less understood and much more open to debate, for several reasons. First, we might consider whether health care can explain the emerging disadvantage of U. S. population health in international comparison from 1980 to present. Although much has changed in the U. S. health care system in that period, the basic structure of the system was established prior to 1980. Several of the most important changes to the system in the past 40 years have expanded access to health care, including the creation of CHIP in the late 1990s and the passage of the ACA in 2010. It is not clear, therefore, that the health care system itself can explain why the U. S. health disadvantage has emerged since 1980. We can also ask whether variation in access to health care across age explains the fact that the U. S. health disadvantage disappears in the oldest ages. In his column mentioned at the beginning of this chapter, Kristof attributed the improvement in the U. S. international ranking in older ages to Medicare, the U. S. health insurance program for the elderly that was signed into law in 1965. Although this is a provocative assertion, universal health care is probably not the whole story behind the more favorable U. S. ranking at the oldest ages. For one, there is no evidence that the mortality rates of the lowest-income individuals in the United States change at age 65, when people become eligible for Medicare.11 A

62

|

Chapter 3

different explanation is stronger health selection on who makes it to age 65 in the United States as compared to peer countries. Those who survive the greater risks of early death in the United States may be a healthier population than those who survive to age 65 in peer countries. In other words, the age pattern of the U. S. disadvantage probably has more to do with higher mortality in the United States prior to age 65 than with lower mortality after age 65. Are those higher mortality risks prior to age 65 a result of the health care system? We can answer that question by considering whether health care can explain the U. S. disadvantage across such a large variety of causes of death and measures of health. In fact, a large portion of the poor U. S. international health ranking in mortality is explained by causes of death that cannot be attributed primarily to health care. For instance, among men under 50, deaths from homicide, motor vehicle accidents, suicide, and other injuries account for 57% of excess mortality compared with men in other high-income countries; recall as well that mortality below 50 accounts for most of the U. S. disadvantage among men. Among women, the same external causes account for 38% of excess mortality before age 50. Although health care can intervene to treat some life-threatening injuries, and in some instances health care may help prevent those injuries in the first place, it seems fair to conclude that the primary causes of deaths from external causes often fall beyond the purview of the health care system. For causes of death that are more amenable to health care, such as cancer and heart disease, the U. S. health care system actually performs relatively well in comparison to peer countries.12 For instance, one study found that more people with high blood pressure are more frequently treated, and treated more effectively, in the United States than in Canada and five European countries.13 Other studies show that the United States fares well on survival rates following a heart attack or ischemic stroke compared to peer countries.14 Samuel Preston and Jessica Ho analyzed the U. S. performance on the screening and treatment for prostate cancer and breast cancer, two diseases that are highly treatable if caught early.15 For both types of cancer, the United States screens at higher rates, uses more aggressive treatment protocols, and has higher five-year survival rates than in European countries. Preston and Ho show that death rates from both types of cancer declined more rapidly in the United States than in the average of 15 other OECD countries. Because prostate cancer does not have clear behavioral risk factors, and the behavioral risk factors for breast cancer did not change in ways that would explain the

International Context | 63

decline in breast cancer mortality after 1990, Preston and Ho concluded that the relatively good outcomes in the United States on these two indicators are due to more effective screening and treatment in the United States, i.e., the result of better health care. Their analysis is consistent with studies of U. S. survival rates from other types of cancer.16 These arguments and evidence point to the conclusion that the U. S. health care system matters in certain key ways—most clearly, in failing to provide universal access to health care—but the impact of this failure on the relatively poor U. S. population health ranking is not straightforward. In turn, the research on cross-national population health suggests that factors that fall outside of the health care system—such as health behaviors and inequality—matter as much or more than health care for population health. We turn to these explanations next.

health behaviors The evidence in support of the role of health behaviors for explaining the U. S. population health disadvantage is more conclusive than research on the health care system but, nevertheless, research suggests that health behaviors are not the whole story. Past high rates of smoking among Americans explain a substantial portion of the U. S. disadvantage in mortality statistics after age 50, but probably explain a much smaller portion of the population health disadvantage among younger adults, where most of the U. S. disadvantage is concentrated.17 And while rates of overweight and obesity are higher in the United States than in peer countries, the evidence linking body weight to mortality is not straightforward.18 Furthermore, in assessing the role of health behaviors, it is useful to refer to fundamental cause theory (introduced in chapter 1), which reminds us that if health behaviors explain part of the U. S. disadvantage, the question then becomes, why do Americans engage in worse behaviors? Even if behaviors explained the poor U. S. international health ranking, we would want to know the causes of unhealthy behaviors in the first place. Figure 11 shows that current levels of smoking in the United States are low by international standards. Data from recent years show that the United States has one of the lowest rates of smoking among 26 OECD countries. While 15.5% of U. S. adults are current smokers, just over 11% of adults are daily smokers. This latter figure compares to an average of 17.9% in the 26 countries included in the figure. However, due to the lag in the impact of smoking on health, current smokingrelated mortality reflects the impact of smoking several decades earlier.

|

Chapter 3

Iceland

Sweden

64 30

25

20

15

10

5

Greece

Hungary

Spain

Austria

France

Germany

Italy

Switzerland

Israel

United Kingdom

Ireland

Netherlands

Japan

Belgium

Korea

Finland

Denmark

Luxembourg Portugal

Canada

New Zealand

Norway

Australia

United States

0

figure 11. Percentage of adults who are daily smokers in 26 high-income countries. Source: Organization for Economic Cooperation and Development 2017

Thus, the current relatively low rates of smoking in the United States will affect the U. S. ranking in several decades, and we have to look to the past to see how the United States compared to high-income countries several decades ago to understand the contemporary consequences of (past) smoking on mortality. Indeed, the United States had higher rates of tobacco use than other high-income countries in the 1950s, but smoking began to decline after 1965 and fell earlier and faster in the United States than in other countries.19 While the mortality impact of smoking increased among men over age 50 from 1955 to 1980, it stabilized by 2003 at about 22% of deaths.20 Because women’s smoking patterns were delayed compared to men’s—starting and peaking later—the portion of deaths accounted for by smoking among women over age 50 increased from 1955 to 1980 to 2003.21 These trends suggest that we may expect smoking to account for fewer U. S. deaths in the future. How much do smoking trends account for the U. S. international disadvantage in mortality? Samuel Preston and his colleagues examined this question among adults over age 50 and found that smoking reduces life expectancy at age 50 by 2.5 years for men and 2.3 years for women, impacts that are greater than in other countries.22 Without smoking

International Context | 65

deaths, the U. S. rank in life expectancy at age 50 for men would change from 15th to 12th among 20 OECD countries, while among women it would change from 17th to 9th. These estimates imply that smoking accounted for 41% of the gap in men’s life expectancy at age 50 between the United States and nine peer countries and 78% for women in 2003. A similar analysis has not been conducted among younger adults, but the National Academy of Sciences report on health and mortality before age 50 concluded that smoking likely accounts for a smaller portion of deaths in those age groups because smoking rates have been declining rapidly among younger adults over the past several decades.23 Compared to the relatively good contemporary U. S. ranking for smoking, the United States fares poorly among high-income countries in terms of overweight and obesity. Figure 12 shows the percent of adults who are obese or overweight in 25 peer countries, using data reported by the OECD from the most recent year available after 2012 per country. Of course, body size is not a health behavior, but it is the outcome of two important health behaviors—diet and physical activity (and their interaction with physiology and genetics)—and body size is far easier to measure than either diet or physical activity. Overweight and obesity are measured using the Body Mass Index (BMI), which is the ratio of weight (in kilograms) to the square of height (in meters), with BMIs between 25 and 29 considered overweight and BMIs of 30 and over considered obese.24 As the graph makes clear, the United States outranks most other high-income countries on this measure with 63.9% of adults overweight or obese, compared to an average of 50% among all countries included in the list and lows of 23.8% and 26.1% in Japan and Korea, respectively. While the U. S. rate of overweight has remained at just over 30% since 1960, the rate of obesity has risen dramatically since then, from 14% in 1960 to over 30% in 2008.25 Other countries have also seen increases in the proportion of adults who are obese, but at lower levels.26 The timing of the growth in rates of obesity in the United States coincides with the timing of the shift toward the U. S. disadvantage in international health rankings, and the United States clearly leads the high-income world in terms of rates of overweight and obesity, both currently and historically. However, it is not straightforward how much these trends matter for the U. S. health disadvantage because the relationship between overweight / obesity and morbidity and mortality is not straightforward. Research establishes that obesity is associated with increased risk for Type II diabetes, high blood pressure, coronary heart disease, gallstones, and some cancers.27 Research also shows that obesity is associated with

|

Chapter 3

Japan

Korea

66 70 60 50 40 30 20 10

Australia

United States

Iceland

United Kingdom

Finland

Greece

Hungary

Israel

Norway

Ireland

Portugal

Spain

Canada

Belgium

Germany

Sweden

Denmark

Netherlands

Luxembourg

France

Austria

Italy

New Zealand

0

figure 12. Percentage of adults who are overweight or obese in 25 high-income countries. Source: Organization for Economic Cooperation and Development 2017

an elevated risk of mortality, especially among younger adults and at the highest levels of BMI.28 However, these estimates of the relationship between overweight / obesity and mortality are complicated by the fact that chronic disease causes weight loss prior to death, the lag time between overweight and mortality is not well understood, and BMI is an imprecise measure of overweight.29 Because of these problems, analyses of the relative role of overweight / obesity on international differences in health are uncertain in their conclusions. Nevertheless, the best existing research on this topic suggests that obesity likely explains approximately 20–35% of the U. S. disadvantage in life expectancy at age 50 relative to other high-income countries, far less than the impact of smoking.30 A similar estimate for younger people is not available, but the National Academy of Sciences report on mortality before age 50 suggests that obesity plays an important role given that the association between obesity and mortality is stronger among younger adults than it is among older adults. In terms of morbidity, studies find that even non-obese adults over 50 in the United States have twice the risk of comorbidities than non-obese adults age 50 and over in peer countries.31 That is to say, the very high level of obesity in the United States matters, but it is not the whole story.

International Context | 67

Other health behaviors may matter as well. For instance, Americans are less likely to use seatbelts and are less likely to wear helmets while riding a motorcycle than peers in other high-income countries.32 A greater portion of traffic-related deaths in the United States involve alcohol compared to other countries.33 Further, Americans are more likely to use illicit drugs34 and own more firearms than individuals in high-income peer countries.35 While each of these patterns clearly matters for the poorer population health profile of the United States relative to its peers, it is not clear the extent to which any on its own affects the poor U. S. international health ranking. However, it is safe to conclude that health behavior patterns clearly matter in these international rankings and in a very important way. The following questions inevitably arise: Why do Americans smoke more? Why do Americans have higher rates of obesity? Why do Americans drive more? Why do Americans own more guns than people in other countries? The answers likely have something to do with other characteristics of U. S. society—contexts like the built environment, social inequality, and social policy. We turn to these next.

the built environment The built environment refers to the physical spaces we inhabit: the way that cities, suburbs, and rural areas are organized; the quality, affordability, and availability of housing; access to public transportation, parks, recreational facilities, and public spaces; the location of environmental waste and toxins; and what kinds of consumer goods, including groceries, fast food, tobacco, alcohol, and firearms, are marketed and sold in different areas. In some ways, the built environment poses direct threats to health, for instance, through exposure to lead in paint36 or water37 or to airborne particulate matter.38 In other ways, the built environment places people at increased risk of poor health behaviors, for instance, by limiting their access to healthy foods or recreational spaces for exercise.39 In the United States, physical space is patterned in such a way that the built environment is considered a key contributor to health inequalities by socioeconomic status and race / ethnicity.40 The environmental justice movement arose out of these concerns to seek a remedy to environmental inequality and its disproportionate health consequences on the poor and on minority groups in the United States.41 Although the research literature establishes that the built environment matters for health and social disparities in health, we do not know

68

|

Chapter 3

very much about how the built environment affects cross-national differences in health or what role the built environment plays in the U. S.’s poor international health ranking. This is in part due to the difficulty of measuring and comparing these characteristics in a systematic way across countries. In some aspects of the built environment, the United States fares poorly internationally, particularly in terms of our heavier use of automobiles and much greater availability of firearms42—but in others, such as air pollution, the United States is not an outlier.43 The National Research Council’s report on the U. S. disadvantage in health and mortality under age 50 concluded that the built environment likely has something to do with U. S. mortality disadvantages in terms of obesity-related causes of death and deaths related to injuries (from car accidents, homicide, and suicide).44 However, it is not clear how much of a role the built environment plays in these outcomes, when and how it acts as a direct risk factor versus a factor that raises the risk of other risk factors, and whether it may also be implicated in other causes of death. Thus, much research remains to be done on this topic.

social inequality Like some aspects of the built environment, social inequality can be conceived of as a structural condition—something outside of any one individual’s control—that increases the risk of poor health as a result of multiple, complex mechanisms. One tricky issue in assessing the role of social inequality is choosing a dimension along which inequality is structured. What matters more: economic inequality, geographic inequality, or racial inequality? Indeed, all of these matter, and in the U. S. context at least, they are deeply intertwined. As we will see in the next chapter, the substantial geographic inequality within the United States can be implicated in the poor international health ranking. If all U. S. counties were like the counties with the best population health, for example, the U. S. would rank first among nations. But most research on inequality has focused on economic inequality, in part because it is the simplest to measure consistently and compare across countries. Unlike the built environment, the evidence that economic inequality matters for international rankings in health is more robust and convincing, but there is substantial debate on how and why economic inequality affects population health. What is clear in this research is that while the United States is one of the wealthiest countries in the world, it is also characterized by substantial economic inequality. In 2012, the United

International Context | 69

States had a more unequal distribution of income than any high-income peer country and ranked fourth on income inequality among all OECD countries, after Costa Rica, Mexico, and Turkey.45 A larger share of the U. S. population lived in poverty than in all other high-income countries other than Israel—17.4% after taxes and transfers.46 And the U. S. rankings on inequality and poverty have both worsened since 1980.47 These trends correspond with the U. S.’s poor international ranking on health. It is also clear that inequality is related to health outcomes. A large body of research establishes that places with greater inequality have worse population health.48 Whether this correlation reflects a causal relationship or something else, and how and why this relationship unfolds, is subject to debate. Why might economic inequality impact population health? There are three approaches to understanding the answer to this question. One, the “material deprivation” argument, makes the case that economic inequality deprives a substantial portion of the population of the material goods of society and thereby harms their health. As we will discuss in chapter 5, there is a strong socioeconomic gradient in health whereby those with fewer socioeconomic resources have worse health than those with greater socioeconomic resources, a relationship observed across countries. The socioeconomic gradient in health results from a variety of mechanisms, including health behaviors, stress, and access to health care. If a wealthy country has a high level of inequality and a substantial portion of its population lives in poverty—as is the case in the United States—its average population health will be “driven down” by the poor health of those at the bottom of the socioeconomic hierarchy. If U. S. population health is driven down by disadvantaged segments of the population, U. S. population health might actually be comparable, perhaps even better, than health in peer countries if the disadvantaged segments of the population could simply be counted out. We can examine this idea by limiting the comparisons to groups at different levels of socioeconomic status. Doing this, studies find that even high-income, highly educated, and insured people in the United States have poorer health than similarly advantaged Europeans.49 In a similar vein, the U. S. population health disadvantage is still very wide when comparisons are made between U. S. Whites and White Europeans.50 Although it is true that the U. S. international disadvantage is largest among the least well off, the international disadvantage is observed for all social groups. In other words, it is not just the health of the poor or disadvantaged in the United States that leads to the U. S. international health disadvantage.

70

|

Chapter 3

Whatever it is about the United States that undermines population health, it affects all social groups in the United States. A different way of looking at the material deprivation argument is to consider what would happen if inequality itself could be eliminated and the material deprivation of socioeconomically disadvantaged groups was lessened by the redistribution of income from the wealthy to the less well off.51 Because the health returns to a single dollar are greater for the poor than they are for the rich,52 the benefits of raising the income of the poor would arguably outweigh the harms of reducing the income of the wealthy, and population health would improve as a result. If we understand the positive correlation between inequality and population health to be causal, as some have argued it is,53 then the implication is that population health would improve with declining inequality—that is, as income is redistributed from the wealthy to the poor. However, it is not clear that material deprivation is the only reason why countries with greater inequality have worse population health. Studies find that even when accounting for a country’s level of absolute deprivation—for instance, the poverty rate or median household income— inequality has an independent effect on population health.54 Other studies find a similar pattern when considering individual socioeconomic status and health, namely that the degree of inequality in a place matters for the health of individuals who live there, even when accounting for differences across individuals in their own socioeconomic status.55 Furthermore, studies find that inequality harms the health of people at all socioeconomic levels, although it is most harmful for the disadvantaged.56 In other words, even socioeconomically well-off people have worse health if they live in a highly unequal place, which cannot logically be the result of material deprivation. These studies suggest that population health may be worse in highly unequal places not only due to material deprivation but also due to other forces. One such force is suggested by the second approach to understanding how inequality impacts population health, which argues that inequality not only deprives people of the material goods of society but also of the social goods of society.57 By this “social deprivation” argument, inequality affects the health of everyone, including the socioeconomically well off, by creating social distance and undermining trust and cohesion between people in a society. In support of this perspective, research has linked inequality to social cohesion, on the one hand, and social cohesion (and other measures of social connectedness) to health, on the other. For instance, studies have shown that people are less likely to be civically

International Context | 71

engaged and to be concerned about the welfare of others in more unequal societies.58 And studies examining the relationship between factors like trust, cohesion, social solidarity, and belonging with population health suggest that health is better when people are connected to others, feel a sense of belonging, and live in places with greater solidarity and cohesion.59 Research on the relationship between social cohesion and health supports the view that connectedness affects health both through psychosocial mechanisms—such as through reducing stress—and through behavioral responses—such as through lower rates of smoking. While these bodies of research suggest a plausible causal chain linking inequality to health through various mechanisms, more work combining the two research threads just described—that is, linking inequality to stress processes and, in turn, to health—is needed.60 A third approach to the question of whether or not the high level of socioeconomic inequality in the United States is a cause of our poor international population health ranking results in skepticism. Several prominent economists have argued that the correlation between economic inequality and population health does not reflect a causal relationship.61 That is, there could be a third, unmeasured characteristic of places that are highly unequal and have poor population health; likewise, unmeasured factors could characterize places that are highly equal and have good population health. Indeed, income inequality is related to numerous other forces that are also related to health—for instance, social policy, health care, the labor market, and racial / ethnic and socioeconomic segregation. A related issue is reverse causation: it may be the case that it is not economic inequality that undermines social solidarity, but a lack of social solidarity in the first place that leads to economic inequality. This body of literature also points out that cross-sectional studies of the relationship between economic inequality and health ignore the plausible time ordering of the causal pathway: Does inequality have an immediate impact on health, or is it an impact that is felt gradually? Is it during childhood or adulthood that absolute or relative material deprivation matters? These questions and concerns reflect the challenge of understanding such a complex relationship as the one between economic inequality and population health. With these caveats in mind, we may still wonder what role economic inequality plays in the poor U. S. international health ranking. Few studies directly examine this question. Several patterns in the data do line up, however. For instance, the socioeconomic gradient in U. S. mortality is wider among the young than among older adults,62 among whom the

72

|

Chapter 3

Box 4: A Simple Illustration of Population Data Analysis In Box 3 on page 58, we introduced the Preston Curve, which illustrates the relationship between countries’ life expectancy and gross domestic product (GDP) per capita in 2015. This graph is a good example of a common way of doing population health science with statistical data. In this box, we break this analysis down and discuss some of the problems raised in the process. The Preston Curve analysis relies on cross-sectional, cross-national data where two concepts (health and wealth) are measured by two variables (life expectancy and per capita GDP). “Cross-sectional” means that the data come from one point in time, or one period, in this case the year 2015. “Cross-national” data means that the unit of analysis is countries: each dot on the graph represents one country. The two variables, life expectancy and per capita GDP, are measured using death rates, data regarding the size of the economy, and population counts, in each country from the year 2015. The variables are then graphed such that each dot plots the life expectancy on the y (or vertical) axis of the graph and per capita GDP on the x (or horizontal) axis for a single country. The pattern across the dots reveals the relationship between life expectancy and per capita GDP across countries in 2015. As you will recall from Box 3–1, the curve shows that life expectancy is shorter in economies with smaller per capita GDP and longer in economies with larger per capita GDP. This relationship is strong from the lowest values of per capita GDP to about $5,000, and weaker at higher values of per capita GDP. Stop for a second and ask yourself whether there may be any problems with the conclusion reached in the last paragraph. Two common problems to think about are the representation of countries in the data set and the measurement of the two variables. Are all countries there? If not, which ones are omitted, why are they omitted, and might they be different from those that are included? These are issues of selection and missing data. You might also think about whether per capita GDP and life expectancy are accurately measured and whether these measures are more accurate in some countries and less accurate in others. If this were true, say, if life expectancy was overestimated in low-income countries, how would the conclusion about the relationship between life expectancy and per capita GDP revealed in this graph be erroneous? What we are really interested in here is not life expectancy and per capita GDP, but broader concepts: health and wealth. Life expectancy and per capita GDP are measures of those concepts. Thus, we might use this graph to conclude that health is better in wealthier economies, and worse in poorer economies. Assuming for the moment that there is not substantial selectivity in the countries that appear in the graph or in

International Context | 73

the measurement of life expectancy and per capita GDP, now we must ask whether life expectancy and per capita GDP are valid measures of health and wealth. Do these measures do a good job of capturing the concepts of wealth and health? To answer this question, we should start by defining wealth and health as best as we can, and then thinking about what aspects of those definitions the measures “life expectancy” and “per capita GDP” capture. For instance, we might worry if life expectancy is long but there is substantial morbidity in one country, whereas it is short but quite healthy in another. If this were true on average, life expectancy may not validly measure population health. What problems might we have with per capita GDP as a measure of wealth? Per capita GDP is a measure of central tendency—an average income measure, which distributes the total economic product in a country among all its members. In a highly skewed income distribution, does per capita GDP capture the true wealth of the average person? Finally, we might consider what kinds of inferences we are making from the data about social processes. Inferences are conclusions that extend beyond the data. One way that population health scientists—and other social scientists—use cross-sectional, cross-cultural data such as these is to infer something about processes that unfold over time. In this instance, the inference is that differences between countries at a single point in time may capture something that happens within countries over time. This allows the analyst to make predictions from the curve: for example, if Ghana’s economy grows from $1,000 per capita to $30,000 per capita, we can assume that its life expectancy will grow as well (and make a prediction as to how much). What are some problems with this inference? In answering this question, think about what it means to assume that countries at lower levels of wealth are simply less far along the same trajectory as countries with higher levels of wealth. Can we look to places like Pakistan and Bangladesh in 2015 to see what health was like in the United States 50 or 100 years ago? This kind of analysis ignores the substantial interdependency between countries, including the fact that the wealth of countries that are “farther along” the curve is related to the wealth of countries “less far along,” and that history interacts with unique local cultures to create substantial variation in social processes. These kinds of questions, critiques, and ideas are those that make what appears to be a very simple analysis indeed quite complicated.

U. S. international health disadvantage is also especially large. Furthermore, as we mentioned earlier, income inequality has been growing in the United States since the mid-1970s, roughly the same period over which the U. S. international health ranking has fallen. Finally, economic inequality is related to a diverse set of mortality and morbidity outcomes,

74

|

Chapter 3

including infant mortality, all-cause mortality, homicide, mental illness, respiratory disease, heart disease, obesity, and drug abuse.63 Each of these consistencies between the empirical patterns of inequality and health, on the one hand, and the U. S. international health ranking, on the other, suggests that economic inequality may play an important role in explaining the poor U. S. international health ranking.

social policy Each of the explanations we have just discussed—inequality, the built environment, health behaviors, and health care—are intimately tied to social policy. Inequality is related to tax, economic, education, and social welfare policies, to name a few. The built environment reflects the specific city codes, public health policies, transportation policies, and environmental policies that establish where and how roads, buildings, parks, commercial centers, neighborhoods, and public spaces are developed and maintained. Health behaviors are influenced by policies controlling the marketing, taxing, subsidizing, regulation, and sale of different goods, as well as by city codes regulating and funding public spaces, the layout of neighborhoods, roads, and cities, and so on. And much of our discussion of health care described the fundamental role of policy in structuring the disjointed, inefficient, and unequal U. S. health care system. As with inequality and some elements of the built environment, social policy may be thought of as a structural force that creates the conditions for other forces to influence health. Social policies create, perpetuate, or reduce inequality; social policies can create, aggravate, or repair a poorly functioning health care system; social policies can determine whether the built environment is healthful or harmful; social policies can influence health behaviors. That is, social policies work hand in hand with the other forces described above to influence population health. Numerous scholars and commentators have speculated that the less generous social policies in the United States, particularly in comparison to wealthy European countries, may have something to do with the poor U. S. international health ranking.64 For instance, the United States has a far less generous and less comprehensive social welfare system, including income support and public housing, compared to wealthy European countries. The United States has less generous leave policies and child care support for new parents. Its tax policy is less progressive, and labor protections are weaker. Public education starts later in the United States and we spend less on preschool than most other high-income countries.

International Context | 75

While the United States spends more than most other high-income countries on K–12 schooling per student, we perform about on average in terms of educational outcomes. And while U. S. national policy is less generous and expansive than many other high-income countries, Jennifer Karas Montez reminds us that such policy shortcomings are also evident in certain states compared with others within the United States.65 As with other structural determinants of health, it is extremely difficult to empirically assess the role of those policies for international differences in health or specifically for the U. S. disadvantage in international health rankings. Systematically measuring policies that differ quite substantially across countries, as well as disentangling policies from the political systems that create them, are extremely fraught empirical exercises. As a result, with the exception of health care policy, as described above, the role of policy in generating international differences in population health is largely speculative. We suspect social policies matter a great deal, but research supporting this perspective is almost totally lacking. This chapter summarized a growing body of research that demonstrates the overall poor population health profile of the United States relative to its high-income peers. On the one hand, it is tempting to ask whether there is one “thing” about the United States that can be held accountable for its poor international health ranking. For instance, is it the U. S. culture of individualism—the value the culture places on independence from governmental control and from other social structures like family and community—that undermines not just population health but all the social systems and forces that promote population health? It is probably impossible, if not overly simplistic and problematic, to assess whether something overarching like “cultural or political context” undermines the health of the United States relative to other high-income countries. On the other hand, this chapter also sheds light on several sets of interrelated factors that are most likely associated with poor U. S. population health, including our health care system, health behaviors, the built environment, socioeconomic inequality, and social policy. The fact that this wide array of factors all may in part be responsible for the situation we face strongly suggests that the problem does not have a simple fix and will, instead, require aggressive and sustained attention at the national, state, and local levels. We delve more deeply into such policy issues in chapter 8. For now, chapter 4 turns to other contexts—states, counties, neighborhoods, networks, and families—that are important for understanding population health patterns and trends within the United States.

chapter 4

Spatial and Social Contexts of U. S. Population Health

Similar to the ways in which population health is spatially and socially patterned across countries, population health within the United States is also patterned both spatially and socially. Indeed, there are unique patterns of health observed within the United States across states and counties, cities and neighborhoods, and schools, workplaces, and families. This patterning of health is important to study for two main reasons. One reason is practical: identifying the places and contexts within which population health is worse facilitates policy and programmatic focus where they are most needed. The second is that understanding how population health is patterned helps us understand the causes of population health disparities. Patterns help reveal mechanisms, just as mechanisms help produce patterns. Understanding the mechanisms underlying population health outcomes is also essential for developing effective policies and programs to improve population health. In this chapter, we consider a variety of contexts related to U. S. population health, some of which are spatial, such as states, counties, rural / urban areas, and neighborhoods, and others that are social contexts embedded within those spaces—schools, workplaces, and families. These are some of the key contexts that shape our daily lives and thus our health. Insofar as we share similar social contexts with others and that social contexts have influences on health behavior and health, population health will be contextually patterned. We begin with the spatial patterning of population health across U. S. states and counties. 76

Spatial and Social Contexts |

77

state- and county-level differences in population health About 1 million people live in Fairfax County, Virginia, an affluent suburb of Washington, DC.1 The population of Fairfax County is highly educated, with more than one-half of adults aged 25 and over holding at least a bachelor’s degree. A quarter of the working population is employed in professional, scientific, or management positions, and only 3.6% are unemployed. Fairfax County is home to Thomas Jefferson High School for Science and Technology, the sixth-best high school in the United States according to US News and World Report. It is also a county that boasts excellent population health. In 2014, life expectancy at birth in Fairfax County was estimated to be 83.7 years, nearly five years longer than the life expectancy for the general U. S. population, and longer than life expectancy in all but six other counties in the United States. Drive 350 miles west of Fairfax County through Shenandoah National Park into coal country, where sits Boone County, West Virginia. Boone County is one of the largest coal producers in West Virginia, the secondlargest coal producing state in the United States after Wyoming. Boone County has less than one-twentieth the population of Fairfax County, and the majority of that population lives in places of less than 2,500 people. While more than one-half of Fairfax County’s population has a college degree, in Boone County, only 38% have completed some college or more. The child poverty rate in Boone County is three times as high as it is in Fairfax County. In Boone County, population health statistics are equally troubling. Estimated 2014 life expectancy at birth in Boone County, at 73 years, is more than 10 years shorter than it is in Fairfax County. This statistic ranks Boone County number 3,119 out of 3,196 counties in the United States in terms of life expectancy. Other statistics bring the disparity in population health contrasting the two counties into even starker relief. About 3,300 years of life before age 75 are lost per 100,000 people in Fairfax County every year, a number that would ideally be far closer to zero. But in Boone County, the number is far worse: 12,700 years of life before age 75 are lost per 100,000 people every year. Less than half as many adults rank their health as “fair” or “poor” in Fairfax County as in Boone County. The rate of low birthweight among newborns is 7% in Fairfax County, but 11% in Boone County. Twenty percent of adults in Fairfax County are obese, compared with 35% in Boone County. In Fairfax County, 11% of adults smoke; in Boone County, 25% do. In perhaps the most striking difference, recent

78

|

Chapter 4

data show that are were 33 injury-related deaths per 100,000 adults in Fairfax County. That rate is about five times greater in Boone County: 167 per 100,000 adults. The highest state-level death rate from drug overdose in 2015 was in West Virginia, where it was 41.5 deaths per 100,000 people. The same rate was 83 per 100,000 people in Boone County in 2015. It was only 6 in Fairfax County. We can put within-U. S. geographic disparities in context by placing them on a global scale. If all counties in the United States were similar to Fairfax County in terms of health, the U. S. would rank first among nations in life expectancy, just ahead of Japan, whose life expectancy at birth was 83.5 years in 2014.2 By contrast, if all counties were like Boone County, the United States would rank 96th, well behind many countries with much lower wealth, including El Salvador, Jordan, Thailand, and Iran. Boone County and Fairfax County reflect the substantial county-level inequalities in population health that exists in the United States today. Trends in Boone County and Fairfax County also mirror national trends that have unfolded over the past few decades. For instance, in 1980, the difference in life expectancy between Fairfax County and Boone County was 5 years. That difference grew to 10 years by 2015. The difference doubled because, while Fairfax County’s life expectancy grew by 7 years over this period, Boone County’s increased by only 2 years. Widening disparities characterize geographic trends at the national level as well. The county with the longest life expectancy in the United States—a ranking that shifted across counties from year to year—had a life expectancy of 75 years in 1985 for men and 81 years for women. This grew to 82 years for men and 85 years for women by 2010. However, life expectancy in the lowest-ranked county in any year remained at 64 years for men and 73 years for women from 1985 to 2010.3 As a result, the gap in life expectancy at birth between the best-off and the worst-off counties grew from 11 years to 18 years for men and from 9 years to 12 years for women from 1985 to 2010. In other words, some places in the United States have reaped the benefits of health-promoting social policies, increasing education and income, and scientific and health innovation over the past 35–40 years, while others have not. The comparison between Fairfax County and Boone County also reveals the roles of geography and human settlement. If we look at all counties in the United States, we see that many of the most healthdisadvantaged places are rural and / or located in the South or the Appalachian region. Map 1 displays county life expectancy in 2014, with shorter life expectancies in white. Boone County, in southern West

Spatial and Social Contexts |

66.81–74.99

76.48–77.49

78.37–79.23

74.99–76.48

77.49–78.37

79.23–80.26

79

80.26–86.23

map 1. Life expectancy at birth, men and women combined, by county, 2014. Source: DwyerLindgren et al. 2017

Virginia, is part of a large cluster of counties with relatively short life expectancies extending west into eastern Arkansas. Another cluster of counties with short life expectancies runs along the border between Arkansas and Mississippi, and there are several counties with very low life expectancies in South Dakota. Counties with longer life expectancies, colored in gray, tend to be urban and coastal, like Fairfax County, Virginia. Map 1 also shows clusters of counties with long life expectancies in central Colorado and southern Minnesota. These regional and rural / urban gaps have also grown over time. Studies have documented a widening gap in life expectancy between rural and urban places, as well as between counties in Appalachia and the rest of the United States.4

80

|

Chapter 4

Geographic disparities in U. S. population health are large, and so are racial / ethnic and gender disparities (which we discuss in chapters 6 and 7, respectively). Drawing on these patterns, research has shown how combinations of race, gender, and geography generate even larger disparities. That is, research focusing on so-called “race-sex counties” shows disparities in life expectancy at birth as large as 33 years. For example, while Asian women in Bergen County, New Jersey, have a life expectancy at birth of 91 years, Native American men in various counties in South Dakota have a life expectancy at birth of only 58.5 Other research has demonstrated that geographic disparities in health are larger for men but have been growing more rapidly for women.6 While not as great as the variation across counties or race-sex counties, there is also significant variation in population health across U. S. states.7 Life expectancy at birth ranged from 75 years in Mississippi to 81.3 years in Hawaii in 2009, a 6.3-year gap.8 From 2013–15, the infant mortality rate was more than twice as high in Mississippi as in Massachusetts, at 9.1 deaths per 1,000 live births in Mississippi compared to 4.3 deaths per 1,000 live births in Massachusetts in that year.9 Map 2 displays the statespecific infant mortality rate between 2013–15. We see similar geographic patterns at the state level for infant mortality as we did for life expectancy at the county level in map 1; there is higher infant mortality in the South and Midwest, particularly in Mississippi, Alabama, Arkansas, Louisiana, Indiana, West Virginia, and Ohio. State disparities have also grown since 1980. For instance, in 1980, life expectancy was 72.8 years in West Virginia and 73.1 years in Virginia, a difference of less than one year.10 By 2014, that difference had grown to more than three years: West Virginia’s life expectancy increased by 3.2 years to 76, but Virginia’s life expectancy increased by nearly twice as much, 6.1 years, to 79.2 years. Why did geographic inequality in population health increase in the United States after 1980? Jennifer Karas Montez offers a compelling hypothesis.11 She argues that growing geographic inequality reflects the combined influence of policies that have deregulated industry, devolved welfare and health policy-making to the state level, and preempted local (i.e., substate) governments from making health policies. She compared New York and Mississippi, states between which the life expectancy gap grew from 1.6 years in 1980 to 5.5 years in 2014. In New York, cigarettes have become heavily taxed, policies have expanded welfare services, and local governments have not been preempted from passing health-promoting policies such as paid sick days, minimum wage laws, requiring the posting of calorie counts on certain food items, and the

Spatial and Social Contexts |

4.00–4.99

6.00–6.99

8.00–8.99

5.00–5.99

7.00–7.99

9.00–9.99

81

map 2. State infant mortality rates, 2014–2015. Source: NCHS, National Vital Statistics System; see Matthews, Ely, and Driscoll 2018

development of firearm controls. Mississippi, by contrast, has not taxed cigarettes as heavily, has been far less generous in implementing and funding welfare policies, and has preempted local governments from passing health-related laws. These macro-structural differences, in turn, shape differences in individual behaviors and risks for health and illness across states, such that individuals in New York now live substantially healthier and longer lives than individuals in Mississippi. Indeed, there is growing empirical evidence that state contexts matter for the health of individuals living in them. Studies that use statistical models to partition the amount of variation in health or mortality outcomes into the portion explained at the individual, substate (such as county or neighborhood level), and state levels find that state contextual variables exhibit important effects on the health of individuals. One such study finds that one-third of variation in disability across neighborhoods is explained at the state level, and another finds that about one-half of variation in mortality across counties is explained at the state level.12 Studies also find that state characteristics such as economic

82

|

Chapter 4

growth and inequality are strongly related to individual risk of poor health outcomes such as disability.13 Another study examining women’s mortality found that state-level characteristics such as the degree of social cohesion and the smoking policy environment accounted for more than half of state-level variation in women’s mortality, compared to only 30% of variation accounted for by the individual characteristics and behaviors of women located in different states.14 In other words, state characteristics and policies matter above and beyond the characteristics and behaviors of individuals located in different states. But it would be an oversimplification to assume that state characteristics and individual behaviors are independent of one another. Indeed, state social, economic, and policy contexts influence health by structuring individual risk of poor health outcomes. As suggested by Karas Montez, smoking might be more common in Mississippi than in New York because of differences in how the two states regulate the sale of tobacco. Research provides support for understanding individual behaviors as influenced by state context. For instance, one study of county variation in mortality jointly tested the influence of county economic context, county health context, and individual behaviors. The study found that variation in socioeconomic characteristics, which included the percent in poverty, the median household income, the educational distribution, the unemployment rate, and the racial / ethnic composition of the county together accounted for 60% of variation in county life expectancy when health behaviors were not simultaneously considered. Meanwhile, variation in behavioral and metabolic risk factors, which included the prevalence of obesity, physical activity, cigarette smoking, hypertension, and diabetes, accounted for 74% when socioeconomic characteristics were not simultaneously considered. Variation in health care factors, including the percent insured, a measure of quality, and the number of physicians per 1,000 people, only accounted for 27%. In a model including all three sets of factors, the role of socioeconomic context was significantly reduced, suggesting that behavioral and metabolic risk factors mediate the association between a county’s socioeconomic composition and mortality. In other words, this study suggested that socioeconomic factors operate to influence population health across counties largely through their impact on individual-level factors like smoking, exercise, and diet. Such findings raise key policy-related questions regarding whether population health improvement efforts should be focused on individual behaviors or the county- and state-level contexts within which such behaviors are structured.

Spatial and Social Contexts |

83

rural-urban health disparities If you were to travel back in time to the beginning of the twentieth century, as discussed in chapter 2, you would discover that U. S. cities were failing in dramatic fashion to address the problems of human congestion, pollution, waste disposal, and clean water provision. At that time, White men lived ten years longer on average in rural places than in urban places, a disparity known as the “urban mortality penalty” that seemed to be similar across all sociodemographic groups.15 Jump forward in time to 1970. By then, life expectancy was .4 years longer in metropolitan areas than in nonmetropolitan areas, although among women, there was still a slight (.1 year) advantage for those living in a nonmetropolitan area.16 The “rural mortality penalty” emerged clearly after 1980, as shown in figure 13, because mortality rates—especially from heart disease and cancer—declined more rapidly in urban places than in rural places.17 By 2005–09, life expectancy in metropolitan areas was two years longer than life expectancy in nonmetropolitan areas, an urban mortality advantage experienced by all sociodemographic groups.18 Put another way, the contemporary rural mortality penalty results in an excess of about 38,000 deaths in rural places each year.19 Figure 14 shows that the contemporary relationship between size of place and life expectancy is not just between rural and urban or metropolitan and nonmetropolitan places; there is a graded relationship across all levels of county urbanization included in the figure, such that life expectancy is lower among small urban places. Within rural areas, there are also disparities. But it is not the most isolated and sparsely populated counties with the highest excess mortality; rather, it is counties with between 2,500 and 19,999 people that are adjacent to big cities that have the highest rates of excess mortality.20 These middle-sized, rural counties on the outskirts of metropolitan areas are disproportionately located in the South. Obviously there is nothing inherent about cities and rural places that determines the relationship between urbanization / rurality and population health. It is the social contexts and populations that comprise places with large, densely settled populations (cities and metropolitan areas), versus those with small, sparsely settled populations (rural areas), that determine this relationship. The relationship has shifted over time, as the distribution of social resources shifted across places. In 1900, rural places were healthier places to live because people in rural places avoided

1,400

Mortality rate/100,000

1,300 1,200 1,100 1,000 900 800 700 1968

1973

1978

1983

1988

1993

1998

2003

2008

Year Rural US

Urban US

figure 13. Rural and urban U. S. mortality rates per 100,000 population, 1968– 2008. Source: James 2014

82

81.6 81.0

Life expectancy at birth (years)

80.4 80

79.7

79.7

79.1 78.4

77.8

78

76.9

76.5

76.7

75.8

76

75.1 74.1

74

73.8

72 Large metro county

Medium metro county

Small metro county

Urban nonmetro county

Rural nonmetro county

Level of urbanization Total

Male

Female

figure 14. Life expectancy at birth by level of urbanization across U. S. counties, 2005–2009. Source: Singh and Siahpush 2014

Spatial and Social Contexts |

85

exposure to diseases from crowding, pollution, sewage, and unclean water in cities. As public health interventions led to declining mortality from infectious and contagious diseases, urban mortality rates began to decline and population health in urban areas rapidly improved. Today, rural places are less healthy places to live than urban places, and this is largely because economic activity, wealth, and social resources, including high-quality health care and built environments that encourage everyday physical activity, are concentrated in urban places.21

neighborhoods and health While the work on county, state, and rural / urban differences in health is interested in the determinants of population health, it also serves the purpose of documentation and description—that is, of identifying population health problems located in particular places. Where is population health superior and where is it falling behind? Knowing the answers to these questions helps us understand what promotes and what threatens population health, but it also serves the basic purpose of directing attention and resources to the places that need them. In contrast, research on neighborhoods and population health has less to do with documentation and description and more to do with understanding the social mechanisms that influence health outcomes. Thus, most of the research on “neighborhood effects” is interested in whether neighborhoods matter for population health and, if so, why. Theoretically, it makes sense that a person’s neighborhood will matter for their health. When considering this relationship, students often think first about health care: does a person have access to quality health care within their neighborhood? How far do they have to travel to go, for example, to a high-quality emergency room? Is that distance a burden? Spatial location certainly matters for obtaining prompt medical treatment when it is needed, as well as for routine preventive care. Students also think about harmful exposures such as pollution, the built environment of a neighborhood, and crime. What is the housing like? How close is the neighborhood to factories, highways, or other sources of noise and air pollution? Is the neighborhood walkable and safe? Is the neighborhood full of public parks and sidewalks that encourage exercise? Can a person access public transportation from the neighborhood? Beyond the physical environment, the social characteristics of a neighborhood may matter, too. Do neighbors know and trust one another? Can children safely play outside and have neighbors keep an eye on their

86

|

Chapter 4

safety and well-being? How invested and empowered are residents to make productive changes in their residential environment? Are there nearby institutions that support kids and families, such as quality schools, child care facilities, summer youth camps, and sports clubs? These qualities of a residential neighborhood can reduce stress and create the kinds of spaces in which people can take care of themselves and others, with positive health impacts. The characteristics of the physical and social environment interact with individual characteristics to affect our health. Regardless of a person’s interests, capabilities, and resources, they will be better able to exercise in a safe neighborhood with green spaces and exercise facilities than in an unsafe neighborhood with no such spaces or facilities. Similarly, a person’s diet, substance use, and other behaviors depend on the structures and norms that surround them: Are there grocery stores nearby? How accessible are drugs, alcohol, and tobacco, and how common is it to see people using drugs, drinking, or smoking? These aspects of the residential environment shape our own health behaviors. Thus, there are many good reasons to expect that the neighborhoods in which people reside influence health above and beyond the characteristics of the people who live there. Researchers examining these propositions consider a wide variety of characteristics of neighborhoods, including access to health care, crime, and neighborhood socioeconomic conditions, such as the poverty rate, the unemployment rate, and average educational attainment. Others examine the extent to which housing units are owner occupied or female headed. Many studies focus on characteristics of the built environment, such as neighborhood walkability, land use mix, street connectivity, and access to public transportation. Much work on neighborhoods and health has focused on the socioeconomic characteristics of neighborhoods. Chapter 5 provides an overview of key measures of socioeconomic status and solid justification for why higher SES is so strongly related to better health in the United States. In a nutshell, individuals possess and use their resources of education, occupational status, income, and wealth to maintain their health, help fight off disease, and more effectively recover when they are faced with illness. Neighborhoods also have their own socioeconomic resources. Some neighborhoods have substantial resources—such as a lot of highly educated persons who work in high-status jobs, make good incomes, and own their homes. On the flip side, some neighborhoods are composed of people who have lower levels of education and who are employed in low-status jobs, make low incomes, and rent their

Spatial and Social Contexts |

87

homes. Does the socioeconomic status of a neighborhood influence the health of those individuals living in it, above and beyond the individuallevel socioeconomic status of its inhabitants? This is an important question because, if neighborhood SES does matter for the health of its inhabitants, then social policies can be crafted that focus on neighborhoods in the effort to improve population health. For example, cities can create zoning laws that mandate mixed-income housing in newly built neighborhoods to help prevent the concentration of high SES people in some neighborhoods and low SES people in others. Over the last several decades, a sizable body of research has dug into this question. One set of studies measures the SES of individuals as well as that of the communities in which they live, using neighborhood-level measures such as the percentage of adults with a college degree, the percentage of people living in poverty, and the percentage of people who own their home. Measures of individual-level health in these studies have ranged from self-rated health, to whether individuals have chronic conditions or functional limitations, to whether or not individuals died over the course of study follow-up. Catherine Ross and John Mirowsky summarize this set of studies: “Most (but not all) previous research finds that the association between high neighborhood SES and health is positive. Beyond this, the association of health with neighborhood SES is tenuous, inconsistent, and often insignificant.”22 In other words, the health of individuals living in high SES neighborhoods benefits from the resources of their neighborhoods, but it is less clear if the health of those living in low SES neighborhoods is harmed. Ross and Mirowsky’s empirical work demonstrated just a modest association between higher neighborhood socioeconomic status (measured as an index that included components of education, income, and wealth) and a lower rate of physical impairment among U. S. adults, after accounting for individual’s own socioeconomic status. At the same time, individual-level SES factors demonstrated much stronger associations with individuals’ physical impairment than did neighborhood SES. This body of work thus suggests that neighborhood SES may matter for individual-level population health, but not nearly to the extent of individual-level SES. Thus, in spite of strong theoretical reasons to expect neighborhood effects on health, the research findings on this topic have not been terribly robust—perhaps as a result of several complexities involved in estimating neighborhood effects. One such complexity is that “neighborhoods” are a difficult concept to define. One’s own conception of their neighborhood may differ from their neighbor’s, as each person has

88

|

Chapter 4

unique social connections, patterns of daily activity, and experiences of the local geography. Equally challenging are the constraints that researchers face in terms of defining neighborhoods consistently with population health data sources. In order to estimate a neighborhood effect, one must apply a geographic boundary around people, and, in order to understand the characteristics of that neighborhood, then use random sampling techniques from within that boundary to produce a representative neighborhood sample. The challenge of doing this on a national—or even city-wide—scale is formidable. As result, most researchers studying neighborhood effects define neighborhoods as census tracts, which are administrative units drawn up by the Census Bureau of roughly 5,000 people that often do not correspond in an intuitive way to a person’s onthe-ground conception of their neighborhood. One research study that improved on administrative measurement of neighborhoods was the 1995 Project on Human Development in Chicago Neighborhoods (PHDCN). The PHDCN combined geographically contiguous census tracts in Chicago to form 343 neighborhood clusters of about 8,000 people each. The definition of neighborhood clusters incorporated knowledge of geographic boundaries (such as freeways or parks) as well as local knowledge of Chicago’s neighborhoods. The neighborhood clusters were designed to be internally homogenous on key census indicators, such as socioeconomic and demographic composition. The influence of the PHDCN on modern studies of neighborhood effects has been profound; for instance, one review of the literature estimated that 8% of all studies using individual and neighborhood characteristics to understand health used data from the PHDCN.23 One such study, for example, examined neighborhood effects on birthweight.24 This study found that neighborhood poverty and residential stability are quite strongly associated with birthweight primarily through their associations with violent crime and community cohesion. A second source of complexity in the research on neighborhood effects is the problem of observing what impact a neighborhood would have if an average (or random) person moved there, as opposed to observing the health of the unique and select group of people who actually do live there. Because people sort themselves into neighborhoods nonrandomly (e.g., by income level, race / ethnicity, and other demographic characteristics), it is difficult to disentangle differences across neighborhoods that arise from the neighborhood itself—the social and built characteristics of the place—from the individual characteristics of the people who live there. Most studies of neighborhood effects try to

Spatial and Social Contexts |

89

parcel out the extent of variation in health that can be statistically attributed to individuals versus neighborhoods using statistical methods. Research using this type of approach finds some small but statistically significant neighborhood effects on health, but also many null effects.25 A better research design for estimating neighborhood effects involves experiments. In an experiment, researchers can see exactly what happens when random people are moved into particular neighborhoods. One such experiment, the Moving To Opportunity for Fair Housing study (MTO), randomly selected low-income families and gave them vouchers to move from public housing in high-poverty neighborhoods to private housing in neighborhoods with lower poverty rates. Several studies have tested the impacts of moving to a lower-poverty neighborhood through MTO on the mental and physical health of participants, comparing people selected for the program (as well as people who in fact moved) to a control group of people who were not selected. Two studies considering the impacts on adults find positive impacts on mental and physical health. One study examined the short-term mental health impacts and found that moving was associated with lower depressive symptoms and anxiety among parents.26 A second study examined longer-term mental and physical health impacts and found that moving was associated with lower psychological distress, lower risk of obesity, and lower risk of diabetes among adults.27 Studies of the impacts on children found more mixed results, with impacts varying by age at the time of the move and gender.28 The fact that the research literature on neighborhood impacts is somewhat mixed in its conclusions does not lessen the conviction that “place matters”—and that one way to improve population health would be to invest in more vibrant, more health promoting, and more equal residential spaces for all people to enjoy. Of course, more research is needed, too, particularly research that is able to both address the methodological complexity of neighborhood effects studies and to explain why, how, and for whom neighborhoods affect health.

schools, workplaces, and population health As foreshadowed in chapter 1, sociologists have long considered different levels of social context in various areas of research work, ranging from the neighborhoods just discussed to schools, workplaces, religious congregations, social clubs and sports teams, friendship networks, and families. Social contexts structure the norms and behaviors of people who are exposed to them and, as such, they may have important and

90

|

Chapter 4

possibly underappreciated influences on individual- and populationlevel health. One mechanism by which social contexts exert influence on health is through social integration and support.29 That is, social contexts create opportunities for integration and support because they bring together romantic partners, friends, acquaintances, role models, and confidants. Depending upon the characteristics of those individuals, as well as the resources that such individuals may have available, such integration and support could have important influences on the health of individuals. For example, friends, coworkers, fellow congregational members, and family members can provide loans, a place to stay in time of need, transportation, health care information, and / or simply other people to share expenses, meals, and experiences with. More formally, Debra Umberson and Jennifer Karas Montez define social integration as the overall level of involvement with informal social relationships, such as having a spouse, and with formal social relationships, such as those with religious institutions and volunteer organizations, while they define social support as the emotionally sustaining qualities of relationships.30 On the flip side, some social contexts are not conducive to either social integration or support if, for example, a school is characterized by conflict and / or a lack of resources for clubs and activities. Social isolation, defined as a relative absence of social relationships, could result from such social contexts.31 Social contexts also have the potential to regulate the behavior of individuals, thus providing a form of social control that works to influence health. Social control refers to the idea that neighbors, colleagues, friends, congregational members, and family members restrict certain behaviors through the enforcement of social norms.32 If, for example, a person who smokes takes a new job in a workplace where no one else smokes, her / his new colleagues may very well actively work to influence the smoking behavior of the new employee. Such social control can have important influences on the long-term health of individuals through, for example, helping individuals curb their smoking behavior. Moreover, if such social control is widespread, such as in the case of the contemporary strong norms against cigarette smoking among highly educated U. S. adults, then population health patterns and trends are also affected.33 Schools are a critical social context within which children, adolescents, and young adults spend a great deal of their time throughout the early portion of the life course. But despite the obvious potential impact of schools on the formation of health behavior patterns and the structuring of population health, population health science paid very little atten-

Spatial and Social Contexts |

91

tion to the influence of schools until the 1990s. In 1994–95, the National Longitudinal Study of Adolescent to Adult Health (Add Health) was funded and soon after collected its first wave of data, which featured a nationally representative sample of 20,745 adolescents between the ages of 12 and 19, who were clustered within 132 schools.34 The availability of such a large sample and a powerful study design that clustered survey respondents within schools opened up the floodgates for understanding how schools, as social contexts, influenced the friendship networks, health behaviors, and young adult health outcomes of Americans. Since then, Add Health has followed these individuals into middle adulthood and researchers continue to assess how the schooling experiences of the respondents has long-term implications for their health. One important study using Add Health data focused on school-level norms regarding teen pregnancy and the relationship between those norms and the prevalence of teenage pregnancy within schools.35 Schools that have strong norms against teen pregnancy exhibited substantially lower teen pregnancy rates compared with schools where the norm was less strong, even after the researchers controlled for an array of factors associated with the demographic and social composition of the students in those schools. Such a finding strongly suggests that teen pregnancy rates are not simply a result of individual-level behaviors that adolescents engage in, but that school contexts have important influences on population health through the establishment of norms that help to regulate behavior. An earlier study using Add Health data focused on suicide among American adolescents. In this case, Peter Bearman and James Moody uncovered two important findings related to the structure of friendships within school contexts.36 First, adolescents who were friends with someone in the same school who committed suicide were subsequently more likely to exhibit suicidal thoughts or attempt suicide compared with adolescents who did not have school friends who committed suicide. Second, adolescent girls who were socially isolated within their schools—that is, they did not report having any friends in school—were more likely to exhibit suicidal thoughts compared to girls who were more socially integrated. Moreover, adolescent girls who had friends, but whose friends were not friends with one another, were more likely to exhibit suicidal thoughts compared with girls who were embedded within a tight network of school-based friends. Such findings raise important issues about mental health within school contexts, suggesting that adolescent suicide prevention efforts cannot simply be focused on individuals, but could

92

|

Chapter 4

benefit from considering how schools facilitate friendship networks and patterns of social integration or isolation. Such pathbreaking work on school-based social network effects has helped spawn a much larger body of literature on the effects of social networks on health including, for example, the extent to which the body weight of individuals is influenced by friends, friends of friends, or even friends of friends of friends.37 Another body of work has focused on the influence of workplace contexts on health. Population health scientists have long shown that differences in workplace environments impact the health and mortality rates of workers—such as, for example, in comparing the high mortality rates exhibited by coal miners and agricultural workers in comparison with the low mortality rates exhibited by college professors and attorneys.38 While some of the health benefits of white-collar work are accounted for by higher levels of pay, greater access to insurance, and the higher status that often is associated with such positions, workplace context also matters. In other words, it is far safer and more conducive for most long-term health outcomes to work at a university compared with working in a coal mine. More recent studies, however, have begun to tap into more subtle ways that workplace contexts matter for population health. Perhaps most innovatively, the Work, Family & Health Network (WFHN) is a federally funded study that aims to better understand the extent to which workplace interventions improve the health of workers and their families. As just one example of the interesting findings based on this major study, Phyllis Moen and colleagues examined whether employees whose workplace granted them greater control over their work hours and greater supervisor support for their personal lives exhibited more favorable health one year later, compared with a control group of employees whose workplace did not offer such an intervention.39 The study found lower perceived stress and less psychological distress among employees in the intervention group, with effects particularly strong for women. Such interventions illustrate the power of social contexts for individual-level health and, potentially, on the health of the U. S. population as a whole.

family contexts and population health Families constitute yet another critical social context within which health behavior and health are influenced. Classic work on this topic examined the association between marital status and population health, thus assessing whether legally partnering with other individuals resulted in

Spatial and Social Contexts |

Box 5: Religious Involvement and Population Health in the United States Beyond neighborhoods, workplaces, schools, friends, and families, there are other social contexts that have potential impacts on population health. In an article that has now been cited over 800 times in the scientific literature, Robert Hummer and colleagues used nationally representative data from the 1980s and 1990s to show that U. S. adults who reported never attending religious services exhibited over 1.8 times the risk of death over an eight-year follow-up period compared with people who reported attending more than once a week.a The authors controlled for confounding factors that could influence both attendance at services and the risk of death, such as health at the time of the survey, age, gender, and race. The study showed that part of the reason that those who never attended had a higher mortality risk was that they were less socially involved; another was that they had less favorable health behavior than frequent attenders. Such results were consistent with the mechanisms of social integration and social regulation that help to explain why social contexts matter for population health. Soon after publication of the study, lead author Robert Hummer received calls and media inquiries regarding the study, as well as invitations to speak at conferences. Christian radio programs were interested in the study findings, with at least one host asking whether more prayer accounted for the lower mortality among the frequent attenders (there was no information on prayer in the data set). Another host even asked whether the study provided proof that “Jesus saves,” a question which Hummer did not feel like he was qualified to answer. And a prominent cardiology (i.e., heart doctor) association asked Hummer to speak at their annual conference, inviting him to give a talk that they tentatively titled, “Faith-based medicine IS evidence-based medicine” (the published study was focused on neither faith nor medicine). Other studies focusing on connections between religious involvement and population health have received similar or even greater media attention over the years, with cover stories in Time Magazine and Newsweek devoted to the topic.b Interestingly, such media reports often focus on prayer, meditation, healing, and whether or not nurses and doctors should integrate religious beliefs with medical practice, all of which is much different from the best science in the area, which focuses on connections between individuals’ participation in religious communities and the possible health benefits, or downsides, associated with such involvement. One lesson to take away from some of the reaction to the science in this area is that consumers of media must be cautious with what they are reading, seeing, or hearing regarding scientific studies.

93

94

|

Chapter 4

Reading the actual science firsthand is preferable to hearing about it second- or thirdhand, and particularly so if the report on the science is being made through a potentially biased lens. For researchers, clarity in writing and presentation is even more critical; it facilitates accurate interpretation of study findings by audiences, including those who may have the power and resources to enact policy or programs, or to treat patients. As for the science on religious involvement and health, more recent studies continue to show that religious involvement is generally associated with more favorable health and lower mortality among U. S. adults, although the strength of the association may differ across groups.c For example, lack of involvement among African Americans is strongly associated with poor health and high mortality, perhaps because of the institutional centrality of religion in the Black community.d But lack of religious involvement among highly educated American adults may not be associated with poor health or higher mortality at all, perhaps because the benefits of high educational attainment are so strong in the twenty-first-century United States.e Continued work in this fascinating area of study should thus focus on the social contexts within which the religion-health association is being studied. Indeed, religious attendance is not a dose of medicine that works to improve health for everyone in all contexts. Quite to the contrary, its health impacts seem to depend upon the contexts and groups within which it is being considered and the social integrative and regulative functions that involvement in a religious congregation may offer. a. Hummer et al. 1999. b. See http://content.time.com/time/magazine/article/0,9171,1879179–3,00. html; www.highbeam.com/doc/1G1–109670142.html. c. Ellison and Hummer 2010. d. Ellison et al. 2000; Dupre et al. 2006. e. Moulton and Sherkat 2012.

health and longevity benefits and, if so, for whom and why. In a series of studies on the topic, Linda Waite and colleagues demonstrated that married individuals have more favorable health than unmarried or divorced individuals, largely due to the pooling of socioeconomic resources within marital households, the social support that marital partners offer one another, and the social control that partners enact over each other’s health behavior.40 Later work demonstrated, however, that not all marriages were equally beneficial for health; the quality of the relationship mattered as well such that conflict-ridden relationships resulted in worse health, especially for women and for older individuals.41

Spatial and Social Contexts |

95

Outside of the partner context, family members can and do share their resources with one another and provide social integration, support, and control to one another, all of which potentially influence the health of other family members. One simple example is with the case of infant and child health. The U. S. infant mortality rate is twice as high among babies born to women with 12 or fewer years of schooling compared with babies born to women who have a college degree or more; moreover, children aged 0–17 who live in poverty are about four times as likely to be reported to be in poor or fair health by a parent compared with children who live in households with incomes more than four times the poverty line.42 Since infants and children have no say in the socioeconomic status of their parents, such disparities reflect the powerful influences that family-level resources have on the health of the youngest members of U. S. society. Given such findings, social policies such as food stamps, income taxes, and unemployment benefits have the potential to have very strong impacts on the health of all family members, given that families share resources. Chapter 8 focuses more extensively on these policy issues. Family-level resources, social integration, and social control may not only exhibit downward flows (i.e., parents to children); such influences may also be shared among adults, such as with partners / spouses, and with aging parents. For example, a small body of research examines whether adult children influence the health of their aging parents. This body of work has been more prominent in international settings where coresidence among adults and their aging parents is more common than is the case in the United States. Zachary Zimmer and colleagues, for example, showed that high educational attainment among adult children in Taiwan is strongly related to their aging parents being able to recover from the onset of disease.43 Further, such an “upward flow” of resources may not necessarily be limited to relationships within households if, for example, adult children with high socioeconomic status are able to purchase healthy living environments or high-end health care for their nonresidential parents or use their high educational attainment to effectively organize their parents’ medication regimens. Indeed, in the United States, adult children’s educational attainment is also related to the longer survival of parents, most notably because the health behavior of aging parents is positively influenced by the educational level of their children.44 This small body of research, though, needs much further development because most studies of family social context and health to date have not considered the potential upward flow of resources,

96

|

Chapter 4

support, and control from offspring to parents. Moreover, whether such associations are stronger or weaker when adult children and their parents share the same household is unknown. Further, adult children’s health—especially midlife women—has the potential to decline if they are spending time and resources on parents and other dependents, further complicating the understanding of the complex set of relationships between family contexts and health.45 This chapter makes the case that, far from being an individual-level characteristic, the health of individuals and of populations is structured by the spatial and social contexts within which individuals live, work, and play. Spatially, neighborhood, urban / rural, county, and state–level contexts all exhibit influences on health above and beyond the characteristics of the people who are living in such spaces. While data constraints and methodological challenges abound in separating out the actual contextual effects of places in comparison with the compositional effects of the people living in such places, as discussed above, research efforts that incorporate data from multiple levels and that capitalize on experimental designs are facilitating more sophisticated work than ever before in this area of study. Such research work is especially important given that it provides evidence for the effective design of policies and programs that work to improve population health that are focused on places rather than individuals. This chapter also provided insight on the multiple layers of social context in which individuals are immersed and that likewise influence health at both the individual and population levels. Schools, friendship groups, workplaces, congregations, and families are all important social contexts within which individuals live their lives and each level of context exhibits important influences on health through mechanisms such as resource sharing, social integration, and social regulation. We now turn toward more specific examinations of how another critical organizing feature of society, social stratification, works to influence U. S. population health.

chapter 5

Socioeconomic Status and U. S. Population Health

Why has the relationship between socioeconomic status (SES) and health been one of the most thoroughly researched topics in the social and health sciences over the past several decades, with literally thousands of scholarly articles devoted to the topic? While the general relationship between SES and health is not hard to grasp—i.e., higher SES is strongly associated with better health and greater longevity in the contemporary United States—the relationship is far from straightforward. For one, is education, occupational status, income, or wealth the most important dimension of SES for health and longevity? This is a very important question because it has clear policy implications: to most effectively improve population health, should we focus most intensively on increasing and / or improving education, reducing unemployment and providing more rewarding and flexible employment opportunities, raising wages, increasing wealth, or all of the above? Yet despite its importance, it is a very hard question to answer because education, occupation, income, and wealth are interrelated. Moreover, few research studies possess data of high-enough quality on all dimensions of SES, and particularly so as they change across the life course, in order to assess their specific importance for population health. Second, the SES-health relationship is far from straightforward because it is characterized by thorny issues of causality. For example, does low income cause people to fall into poor health or does poor health cause people to lose their jobs and, subsequently, their income decreases? Is 97

98

|

Chapter 5

high education related to good health because of the things students learn in school or because of the intelligence that is needed for both succeeding in school and living a healthy life? Issues of causality in the SES and health relationship are some of the most hotly debated in the population health literature today. Moreover, these issues of causality also have very important policy implications. For example, if high educational attainment does not actually cause people to have better health, then spending a lot of governmental resources to raise educational levels and / or improve the quality of schooling will most likely not improve population health. Third, the life course timing of the SES-health relationship is not well understood. To what extent is low socioeconomic status during childhood a ticket for developing poor health during adulthood and living a short life? Alternatively, is it possible that obtaining a high level of education, getting a good job, and / or earning a good salary as an adult can erase the long-term effects of growing up in a low SES household? Fourth, the SES-health relationship is not at all straightforward because it differs across time, place, and population subgroup. We provide evidence in this chapter that educational attainment has never been as important for adult health and longevity in the United States as it is today. In other words, we seem to be living in a unique period of history with regard to how important educational attainment may be for health, which is surely worthy of the attention we are devoting to the topic. At the same time, though, high levels of educational attainment may be especially beneficial for the health and long lives of White Americans and, at the same time, may not be as beneficial for the health and longevity of individuals who are members of minority groups. Such racial / ethnic differences in the relationship between education and health are important to understand, particularly if educational policy is used as a lever to improve population health and decrease population health disparities. This chapter delves into the relationship between SES and health. In doing so, we contend that the relationship between SES and health is one of the most important scientific topics in the study of population health and, arguably, in all of science today. In large part, the importance of the topic is driven by its immense social and health policy potential. As a straightforward example, if having low educational attainment and / or a poor quality of schooling causes people to have poor health and to live short lives, we can do something about that as a society by investing more money in education and raising the quantity and quality of educational attainment. The importance of the topic is further enhanced when we note that sizable fractions of the U. S. population have low SES—that

Socioeconomic Status | 99

is, they have low levels of education, bad or no jobs, little or no income, and / or little or no wealth. For example, a recent report from the Census Bureau showed that while levels of educational attainment have increased quite dramatically over the past century, it is still the case that only 32.5% of U. S. adults aged 25 and older have a bachelor’s degree or higher. Perhaps even more troubling, more than 1 in 10 adults (11.6%) has not completed a high school degree.1 Further, 12.7% of persons in the United States currently live in a household with an income below the official federal poverty line; this percentage is even higher among children, at 18%.2 Finally, the relationship between SES and health is extraordinarily important because it taps into core issues of social justice. Is it fair that people with high incomes and / or a lot of wealth tend to live much healthier and much longer lives than people with little or no income and / or wealth? Is it troubling that we live in a place and time in which SES disparities in health may be the widest ever measured? These questions tug at our collective conscience and should force us to confront core values of American society. We begin this chapter by first defining SES and discussing how it is measured in studies of population health. Second, we examine reasons why socioeconomic resources are characterized by such strong patterns of inequality in U. S. society, with a particular focus on changes in inequality over the past half-century. Third, we then describe some key contemporary patterns of SES and health and provide evidence that SES disparities in health and mortality may be wider than ever before in U. S. history. We then move to explanations of SES-health disparities, relying largely on fundamental cause theory developed by Bruce Link and Jo Phelan, but also with consideration given to the life course perspective on SES and health.3 At the same time, we discuss the importance of socio-spatial context for understanding why U. S. society seems to be currently characterized by such enormous socioeconomic disparities in population health. We close the chapter by discussing some key research needs on SES and health and policy implications of work in this critical area of study.

what is socioeconomic status and how is it measured? Formally, we define socioeconomic status (SES) as differences between individuals and groups in the possession of highly valued societal

100

|

Chapter 5

resources, most importantly educational attainment, occupational status, income, and wealth.4 This definition explicitly reflects the multidimensional character of SES. Moreover, the concept of socioeconomic status taps into the absolute level of resources that each individual or group possesses along these multiple dimensions. Thus, SES differs from the Marxist-based idea of social class, which is based upon an individual’s relationship to the means of production. According to Marx, social class is determined by whether individuals are workers (i.e., the proletariat) or owners (i.e., the bourgeoisie) in capitalist society.5 In such a conceptualization, the bourgeoisie exploit the proletariat for their labor, leading to social class conflict between those who own the economic means of production and those who provide their labor for the profit of the owners. Social class distinctions are thus economic based, relational, and structurally imposed upon individuals; individuals’ relationship to the means of production is the sole determinant of social class. Contemporary studies of U. S. population health using such a strictly social class perspective are few and far between, given that there is much greater complexity in contemporary patterns of socioeconomic status than the owner-worker dichotomy defined in the Marxist tradition. Weber not only highlighted class (economic status) as a key component of social stratification, but also added status (social standing) and party (political power) as important dimensions.6 Such a conceptualization broadened the idea of social class to include noneconomic components; as such, “socioeconomic status” is a broader concept than social class. Moreover, a socioeconomic-based conceptualization explicitly recognizes that there are more than two (e.g., owner and worker) or three (e.g., owner, manager, worker) resource groupings in contemporary society. Indeed, the multidimensional conceptualization of socioeconomic status suggests that there is a hierarchical continuum characterizing the resource distribution of society, rather than just two or three discrete categories. Further, Weber also developed the idea that individuals actively utilize their class, status, and power to improve their “life chances,” two critical components of which are health and longevity. Over the last halfcentury, then, most contemporary American sociologists and demographers have either explicitly or implicitly used Weber’s multidimensional conceptualization of socioeconomic status in studies of population health, while making use of educational attainment, occupational status, income, and / or wealth as the key dimensions of SES.7 These indicators tap into related aspects of socioeconomic status, but each dimension also contains unique attributes.

Socioeconomic Status | 101

By far, educational attainment is the most often-utilized measure of socioeconomic status in U. S. studies of population health. Educational attainment is usually determined relatively early in life (although this is trending upward) and, thus, does not typically change when individuals experience declines in their physical or mental health; this is not necessarily the case with occupational status, income, or wealth, as discussed below. Survey items on educational attainment, usually assessed by either total number of years of schooling or highest academic degree awarded, are also usually very well reported in major data collection efforts.8 Educational attainment taps into the knowledge, cognitive skills, social networks, job preparation, and learned effectiveness that individuals develop through the schooling process. In fact, sociologists John Mirowsky and Catherine Ross argue that educational attainment is the most important dimension of SES for health because it is so critical for building knowledge and developing important life skills (e.g., reading, writing, critical thinking), finding and keeping a challenging and rewarding job, earning high income, and establishing a sense of personal control over one’s life.9 Thus, educational attainment is very commonly used in U. S. studies of population health given its relative ease of measurement, nonresponsiveness to changes in adult health status, and critical importance for improving employment prospects and enhancing the prospects for income and wealth building. At the same time, while common measures of educational attainment are very well utilized in population health research, few studies tap into the type of schools (e.g., public or private), type of instruction (e.g., in-person or online), quality of education, or course content that individuals experience during the educational process. Thus, much work remains to better understand the dimensions of education that are associated with measures of population health, especially given the rapidly evolving and diversifying nature of the educational system in American society.10 Occupational status taps into the social standing, or prestige, dimension of socioeconomic status. A groundbreaking research project in the United Kingdom called the Whitehall Study, directed by Sir Michael Marmot, demonstrated that mortality rates were graded by occupational status in a stepwise fashion among male employees in the British Civil Service System, with those in the high-ranking positions exhibiting the lowest mortality rates and those in the bottom-ranked positions exhibiting the highest.11 This was a remarkable set of findings at the time because all of the employees were covered by the United Kingdom’s universal health care program; thus, differences in access to care could not account for the occupational status disparities in mortality. Moreover, it was often assumed at

102

|

Chapter 5

the time that high-ranking employees encountered the highest levels of stress on the job and, thus, probably had worse overall health than their lower-ranking counterparts. The researchers also examined income differences and behavioral and health characteristics (e.g., blood pressure, cholesterol levels, smoking, alcohol use) across the occupational status groups, and found that they explained only a relatively small portion of the mortality differences between the occupational status groups. The researchers thus moved to explanations focusing on the social connections and day-today stresses and (lack of) control among employees in the different occupational status positions, which provided a more thorough explanation of the differences. Thus, Marmot argued that occupational status is much more than an indicator of the amount of money people make or a marker for the health behavior they engage in; more holistically, occupational status provides individuals with important social networks and an enhanced sense of control over their lives, factors which are important for preserving health and living a long life. While Marmot’s work on the Whitehall Study (and its follow-up, Whitehall II, which included women) has made a very important contribution to the literature on SES and population health, there is less focus in the U. S. literature on occupational status and health in comparison to that of educational attainment and health.12 In particular, occupational status grades (or scores) are difficult or impossible to assign to those who are unemployed or not in the labor force, including retirees, the incarcerated, the disabled, students, and stay-at-home mothers and fathers. Occupational status is also subject to change due to health problems; that is, changes in health may be a determinant of unemployment or lower occupational status, rather than vice versa. That said, Sarah Burgard and colleagues have persuasively argued that more in-depth study of work experiences and conditions among American adults is necessary to better understand how SES is related to health, particularly in the rapidly changing work context of the twentyfirst century.13 Clearly, population health data sets that measure individuals’ employment changes, working conditions, work creativity, work stresses, and the policy contexts of their work (e.g., whether or not people have access to flexible scheduling, on-site childcare, and parental leave) are needed to develop a more in-depth understanding of the relationship between occupational status and health in the United States. Perhaps most straightforwardly, income reflects the current flow of financial resources that individuals can use for their own health and the health of family members. Income can be used to make mortgage or rent

Socioeconomic Status | 103

payments on a sturdy home in a safe environment, buy nutritious food, pay for a health club membership and high-quality medical and dental care coverage, take a relaxing vacation, make payments on an automobile with high-end safety protections, obtain physical therapy after an injury, and more. Even day-to-day items that might be taken for granted— such as a pair of good running shoes, a bicycle helmet, or a grab bar to prevent falls in a bathtub—require disposable income to purchase. Thus, it is clear how a bigger paycheck can be important for the health of individuals and members of a household who share income and consumption. On the flip side, individuals who have a low income or who live in a household with low income simply have far fewer monetary resources to protect their health. On a day-to-day basis, then, low income can potentially have long-term effects on health and longevity. Thus, perhaps it is no surprise that American adults who have incomes in the top 1% of the distribution were recently shown to exhibit 10 years (women) to 15 years (men) longer life expectancy than men and women in the bottom 1% of the distribution, respectively.14 Beyond the poorer material conditions that individuals may experience when living with low income, individuals who have low income while living in a wealthy society may also experience the psychological costs of relative deprivation when comparing themselves with their (much) more affluent peers.15 While there is ample evidence that higher income is strongly related to health and longevity in the contemporary United States,16 the extent to which higher income causes people to have good health and to live long lives is up for debate.17 A key issue is the bidirectional relationship between income and health. Similar to occupational status, income can plummet with the onset and / or continuation of health problems, which has been demonstrated in a number of studies on the topic.18 In another parallel to occupational status, income is difficult to measure among individuals who are out of the labor force. Finally, data on income can be difficult to collect from individuals and households because people may be unwilling to share information about their income or may not accurately know or report their income. The Health and Retirement Study (HRS), the most well-utilized population health survey of American adults aged 50 and above, has pioneered innovative ways of collecting income data, including through the use of linkages from the data provided by survey respondents to their Social Security earnings records.19 Other innovative studies have recently used IRS tax records to estimate the relationships between income, health, and mortality.20 Given this set of causality and measurement issues, it is no surprise that

104

|

Chapter 5

much work remains to better understand how and why income is related to health in the United States. Moreover, the relationship between income and health has been shown to vary across health outcomes, age groups, and gender and racial / ethnic subgroups of the population,21 further complicating the understanding of this relationship. Over the last two decades, population health researchers have also increasingly considered wealth as a fourth dimension of socioeconomic status. This conceptual expansion makes a great deal of sense given how important wealth is in American society for economic security, power, prestige, and social connections. Wealth refers to the accumulation of assets that people own, minus the debts that they owe.22 Wealth is far more unequally distributed in American society than is income. Recent estimates suggest that the median level (50th percentile) of wealth for American families is about $81,000. On the high end, average wealth for families at the 90th percentile of the distribution is $942,000; for those at the 95th percentile it is $1.9 million; and for those at the 99th percentile, families average about $7.9 million in wealth! On the low end, however, the bottom 15% of U. S. families have no wealth and fully 40% of American families have less than $40,000 in wealth (roughly the average price of a new car).23 Importantly, wealth can serve as a security blanket to protect health. If, for example, a person with multiple homes and a portfolio of stock market assets loses her or his job or is the victim of a natural disaster, they can draw on their wealth to protect themselves and their family members from such a shock. In a systematic review of studies on the relationship between wealth and health, Craig Pollack and colleagues found that higher levels of wealth generally predicted better health among American adults, even above and beyond the effects of education, occupation, and income.24 Unfortunately, however, 40% of American families have little or no assets or savings to draw upon in times of need. Moreover, on the flip side, other researchers have demonstrated that the accumulation of debt is related to a range of poor health outcomes. Katrina Walsemann and colleagues, for example, recently showed that college debt accumulation is strongly associated with worse mental health among U. S. young adults.25

the stratification of socioeconomic resources Stark inequalities are a fact of life in contemporary American society. The possession of socioeconomic resources is not only highly unequal,

Socioeconomic Status | 105

but in some cases (e.g., income and wealth) more unequal today than ever before in U. S. history.26 For example, the Gini Index of U. S. household income inequality, a measure that ranges from 0 if every household receives the same amount of income to 1 in a place where a single household receives all the income, was .481 in 2016. This compares with a Gini Index of .398 in 1976.27 After taxes and income transfer programs (e.g., Social Security, food stamps) are factored in, the Pew Research Center showed that U. S. income inequality is currently the second most unequal in the world, behind Chile.28 Moreover, U. S. wealth inequality is even more extreme, with an estimated Gini Index of .806, the highest in the world.29 U. S. levels of income and wealth inequality are now similar to those exhibited by some of the most unequal European countries prior to World War I.30 In his book outlining the American system of social stratification, Douglas Massey notes that social stratification systems, at heart, encompass two processes.31 First, people are allocated into social categories. As we will discuss in chapter 6, for example, race / ethnicity and immigrant status constitute key categories by which U. S. society has organized itself throughout its history. In this chapter, we argue that socioeconomic status is yet another. While the United States prides itself on the idea that anyone can become highly educated, earn a lot of income, and become wealthy with enough hard work, in reality there is a great deal of socioeconomic “stickiness” between generations. Indeed, Pablo Mitnik and colleagues recently documented decreasing socioeconomic mobility among U. S. young and middle-age adults since the 1970s, especially among those who ended up in relatively high or high SES positions.32 In other words, while it is not impossible for low SES children to become high SES adults, it certainly helps (and increasingly so) if you happen to be born into a high SES family. It is clear, then, that our society not only groups individuals by SES but that such groupings have intergenerational consequences. Second, Massey notes, institutional processes differentially allocate societal resources to these social categories. But how does this work? It is instructive to focus on changes in American social stratification since the 1970s that, as discussed above, have resulted in the highest-ever levels of income and wealth inequality in American history. One institutional explanation focuses on key macroeconomic changes around the world since the 1970s—rapid increases in technology, globalization of the economy, and the relative ease of labor mobility—that has led to a rising skills premium whereby there is greater and greater payoff to

106

|

Chapter 5

very highly educated, mobile, and computer-savvy workers.33 Indeed, the postindustrial economy is based on information and knowledge, automation, and the decline of manufacturing jobs. At the same time, globalization of the economy has created a fierce competition for highly educated workers with specialized skills. Those with elite, high-end skills benefit in the form of rapidly increasing wages and wealth accumulation, while those with more common, less-marketable skills lose out. While an enticing potential explanation, it is important to note that these macroeconomic changes have affected all high-income and even many middle-income countries around the globe, so they cannot explain the extraordinary case of such rapidly rising U. S. inequality. A more likely explanation for increasing SES inequality in the contemporary United States is the interaction between these macroeconomic changes and changes in government policies over the last halfcentury (specifically, labor, tax, and income transfer policies).34 As Thomas Piketty and Emmanuel Saez remarked in their analysis showing more rapid increases in U. S. income and wealth inequality versus those in Europe, “We find large changes in the levels of inequality, both over time and across countries. This reflects the fact that economic trends are not acts of God, and that country-specific institutions and historical circumstances can lead to very different inequality outcomes.”35 What specific institutional changes and circumstances have led to such a widening in the U. S. distribution of income and wealth in recent decades? One such important change is de-unionization, which has led to lower wages and fewer benefits for manufacturing and other blue-collar workers in the United States. While 29% of U. S. workers were in unions in 1975, only 12% were unionized in 2011.36 Since the end of World War II, the United States has passed extensive legislation to undermine labor unions and empower employers, which was enabled by southern Democrats uniting with pro-business Republicans in response to the organization of Black workers in the South by northern unions.37 Second, the rise of top executive compensation in U. S. corporations has increased greatly due to changes in the way corporations are governed.38 On the other hand, the federal minimum wage has exhibited a decline, due to Congressional inaction. Indeed, the real value of federal minimum wage declined from about $9.20 per hour in 1968 to $5.30 in 2005, before inching back up in recent years.39 Third, the U. S. Congress has enacted enormous changes in tax policies favoring high earners and the wealthy. Most recently, the Tax Cuts and Jobs Act (TCJA) that

Socioeconomic Status | 107

President Trump signed into law in late 2017 substantially reduces taxes on U. S. corporations, small businesses, wealthy estates, and highearning individuals. Tax rates for the highest earners, which were as high as 70% in 1980, are now 37%.40 Fourth, while aggregate federal government spending on income transfers (e.g., to the elderly, children, the disabled, the unemployed) is higher than ever, such funding has been increasingly redistributed away from the poorest people to those with higher incomes.41 Robert Moffitt argues that such policy changes indicate an increasing conceptualization of the poor in U. S. society as undeserving.42 Beyond these changes in policy that have worked to reshape the distribution of income and wealth in U. S. society, other policies have helped to increasingly concentrate those with high SES in some geographic areas and those with lower SES in other geographic areas. Lax zoning laws, the creation of municipal boundaries to maintain pockets of segregated affluence, and property tax increases in gentrifying areas are just three mechanisms that have helped to create the highest level of socioeconomic segregation in the nation’s history.43 Indeed, Sean Reardon and Kendra Bischoff showed that increasing income inequality over the last 40 years has led to increasing segregation of the well-off from everyone else in U. S. society, thus isolating those with high education, income, prestigious jobs, substantial wealth, and prominent social networks in certain areas of the country and, within urban areas, in particular neighborhoods.44 This concentration of affluence creates further inequality because funding for schools, parks, and other health-enhancing community resources can be spent on those with already high levels of SES. Most importantly, affluent areas contain very well-funded schools with college-preparatory curriculums, substantial infrastructure for sports and the arts, and well-paid teachers who help to prepare the next generation of high SES adults. On the other hand, children living in the increasingly concentrated lower SES neighborhoods are much less likely to acquire a great education or move on to an elite college or university.45 In all, then, socioeconomic resources in U. S. society, while always stratified, have become increasingly unequally distributed and geographically concentrated since the 1970s through institutional mechanisms favoring those with high levels of education, good jobs, high incomes, and substantial wealth. Such a high level of socioeconomic inequality has serious implications for population health, to which we now turn.

108

|

Chapter 5

socioeconomic disparities in u. s. health and mortality Early Insights into SES Disparities in Population Health Research on SES disparities in health and mortality does not have a particularly long or rich history in the United States. Most U. S. scholars working in this area cite the classic 1960 Matched Records Study of Evelyn Kitagawa and Philip Hauser, published in 1973, as the beginning of sustained attention to SES disparities in population health at the national level.46 By linking data from deaths that occurred between May and August of 1960 with the April 1, 1960 census records of those individuals who were alive earlier that year, the matched records data facilitated the first-ever analysis of SES disparities in U. S. adult mortality at the national level. The findings demonstrated wide SES disparities in adult mortality, whether SES was measured by educational attainment, family income, or occupational category. For example, life expectancy was estimated to be 9.6 years longer for highly educated (1 or more years of college) White women compared to low-educated (0–4 years of schooling) White women. The educational disparity in life expectancy between the highestand lowest-educated White men was smaller, at 3.2 years, but still sizable. Data limitations prevented calculation of educational disparities in life expectancy for non-White populations at the time. Since the publication of the Kitagawa-Hauser book, dozens and dozens of studies have reexamined socioeconomic disparities in population health using different measures of SES and different measures of health. One set of prominent studies showed that socioeconomic disparities in mortality widened between 1960 and the mid-1980s.47 However, a follow-up analysis found that this was only the case for men.48 One possibility for the socioeconomic divergence in population health among men between 1960 and 1985 was that those with high socioeconomic status were much more likely to quit smoking during this period of time compared to men with low socioeconomic status.49 High-SES men were most likely to quit because they were the first to receive the new warnings about tobacco use, they had the economic means to purchase aids to help them quit smoking, and their high-status social networks effectively pressured them to quit as smoking became less and less fashionable.50 The most sustained scientific attention to widening socioeconomic disparities in health has unfolded since 2000, most likely because of concerns regarding the trends in rapidly increasing socioeconomic inequality in the United States in recent decades. Indeed, this large body of

Socioeconomic Status | 109

literature demonstrates that SES disparities in key measures of population health have grown wider over the past several decades for both men and women and that the disparities are now the largest that have ever been documented in the United States. For example, a prominent report by the National Academies of Sciences, Engineering, and Medicine (NASEM) projected a 12.7-year disparity in life expectancy for men born in 1960 and a 13.6-year disparity for women born in 1960 when comparing the top 20% of the lifetime income distribution compared with those in the bottom 20% of the lifetime income distribution. For persons born in 1930, just 30 years earlier, the estimated life expectancy gaps by income were much smaller: about four years for women and five years for men when comparing the top and bottom 20% of the lifetime income distribution.51 Equally as troubling, the NASEM report found that life expectancy at age 50 was projected to be lower for women born in 1960 than for women born in 1930 when looking at the bottom 40% of the income distribution. Similarly, many papers have shown widening educational disparities in U. S. population health outcomes, including mortality, over the last two decades. In fact, in a recent review of the literature on education and health, Anna Zajacova and Elizabeth Lawrence summarized that the association between education and health has become increasingly strong since the 1980s, with widening disparities in many health outcomes across this period of time.52 Zajacova and Lawrence point to growing geographic segregation between rich and poor, the mass incarceration of low SES individuals, the loss of manufacturing jobs, and increasing psychological despair among the low educated as possible explanations for the widening educational disparities in health in recent decades. At the same time, Mark Hayward and colleagues note that highly educated adults in U. S. society have greater access to technological, informational, and monetary resources than ever, resulting in more favorable population health patterns than ever before for those in the upper portion of the educational distribution.53 Clearly, the widening socioeconomic disparities in population health in recent decades reflect broader trends of increasing inequality that have unfolded in U. S. society. Contemporary SES Disparities in Population Health The above section emphasizes that the present-day United States is characterized by enormous population health disparities by socioeconomic status. This section provides a brief documentation of some of those specific

110

|

Chapter 5

table 5 measures of infant health by maternal education, united states, 2007–2010

% Low Birth Weight1 % Premature2 Infant Mortality Rate3

< High School Degree

High School Degree

Some College

College Degree

Graduate or Professional Education

9.0* 14.0* 7.9*

8.9* 13.0* 7.4*

8.0* 12.1* 5.9*

6.8* 10.3* 4.1*

7.3 10.6 3.6

note: Asterisk (*) indicates significant difference from the Graduate or Professional Education category (p < .05). sources: Linked Birth and Infant Death Certificate Files, 2007–2010 (National Vital Statistics System 2015). N = 16,493,765. 1 Defined as infants who are born weighing less than 2,500 grams (approx. 5.5 lbs) at time of birth. 2 Defined as infants who are born at less than 37 weeks of gestation. 3 Defined as the number of deaths in the first year of life per 1,000 live births.

contemporary disparities in different stages of the life course. Table 5 shows maternal educational disparities in three important measures of infant health: the percentage of low-birth-weight births (>2,500 grams), the percentage of premature births (>37 weeks of gestation), and the infant mortality rate. The pattern for both low birth weight and prematurity show that women with the lowest level of education have the highest rates of low birth weight (9.0%) and prematurity (14.0%), while women with a college degree have the lowest rate of both low birth rate (6.8%) and prematurity (10.3%). Interestingly, women with a graduate or professional degree exhibit slightly higher rates of both low birth weight and prematurity than do women with a college degree, possibly because rates of multiple births (who are more likely to be born early and / or small) have increased among women with high levels of education, due to their greater use of fertilityenhancing treatments.54 Infant mortality rates are also the highest for women with less than a high school degree (7.9 deaths per 1,000 births) and lowest for women with a college degree (4.1) and with a graduate or professional education (3.6). Thus, at the earliest ages, population health measures are strongly differentiated by (parents’) socioeconomic status. Table 6 shows disparities in basic measures of child health (ages 1–14) and adolescent health (ages 15–24) by family income level. While population health measures in these age ranges are generally quite good, there are still very wide disparities by SES. At ages 1–14, less than one percent of children (0.5%) are rated by a parent as having fair or poor health; this percentage is eight times higher (4.1%) among children who are living in poverty. Children in poverty or near poverty (i.e., 100–199% of

Socioeconomic Status | 111 table 6 selected measures of child and adolescent health by family income level in the united states Family Income in Relation to Poverty Line >100%

100–199% 200–399%

400%+

Children ages 1–14 Percent with poor or fair health1 Percent with any physical limitations2 Mean school days missed in past 12 months3

4.1* 9.5* 3.9*

2.3* 8.3* 3.5*

1.1* 7.0* 3.3*

0.5 5.8* 3.0

6.4* 8.1* 2.8*

4.4* 6.7* 2.4*

2.4* 5.4* 2.2

1.3 4.7 2.1

Adolescents ages 15–24 Percent with poor or fair health1 Percent with any physical limitations2 Mean days of bed rest in past 12 months4

note: Asterisk (*) indicates significant difference from the 400%+ income group at the p>0.05 level. sources: National Health Interview Survey, pooled 1997–2014 (Blewett et al. 2016). N = 424,438 for children and 351,099 for adolescents. However, each measure of health was not asked of all persons, and therefore each row has a different N. Estimates are weighted so that the percentages represent the U. S. noninstitutionalized population of children and adolescents, respectively. 1

Defined as parent or adolescent reporting general health status as either fair or poor on a five-point Likert scale, ranging from “excellent” to “poor.” 2 Defined as having an activity limitation due to obesity or a chronic condition. 3 Defined as an individual experiencing an illness or injury that ended up with them missing school during the preceding 12 months (including hospitalization). 4 Defined as an individual experiencing an illness or injury that kept them in bed during the preceding 12 months (including hospitalization).

the poverty line) are also much more likely to have physical limitations than children living in more affluent families; moreover, children living in or near poverty have higher mean levels of missing school days compared to their more affluent peers. Similar patterns exist among adolescents aged 15–24, with the percentage of those in fair or poor health about five times as high (6.4%) than among those living in families with an income 400% or more than the federal poverty line (1.3%). Table 7 uses nationally representative data from the National Health Interview Survey, pooled from 2000 to 2015, to show contemporary educational attainment disparities in the health of U. S. adults aged 25 to 64.55 The disparities are extremely wide. Perhaps most striking, while just 4.3% of adults who have a bachelor’s degree and 3.3% of adults who have a master’s degree or higher rate their health as fair or poor, 24.3% of adults aged 25–64 who have less than a high school degree

112

|

Chapter 5

table 7 educational attainment disparities in eight measures of population health for u. s. adults aged 25–64 Educational Attainment1

% Fair / poor health2 % Obese3 % Current smoker % 1+ Functional limitation4 % 1+ Activity limitation5 % Needs help with ADL6 % Needs help with IADL7 % In bed 8+ days last year8

> HS Degree

HS Degree

Some College

24.3* 33.6* 34.7* 36.9* 20.6* 2.4* 4.8* 9.9*

13.4* 32.8* 32.1* 33.2* 14.1* 1.3* 2.9* 8.0*

10.1* 31.9* 25.0* 31.9* 12.1* 1.1* 2.5* 8.4*

BA/BS MA/MS+ 4.3* 21.5* 12.0* 21.0 5.6* 0.5* 0.9* 4.5*

3.3* 18.3* 7.5* 20.5 5.0 0.4 0.8 4.0

note: Asterisk (*) indicates significant difference from MA / MS+ educational group at the p>0.05 level. sources: National Health Interview Survey, pooled 2000–2015 (Blewett et al. 2016). N = 766,986. However, each measure of health was not asked of all adults, and therefore has varying missing data resulting in different row Ns. Estimates are weighted so that the percentages represent the U. S. noninstitutionalized population of adults aged 25–64. 1

Defined as the highest level of schooling an individual had completed. Defined as reporting general health status as either fair or poor on a five-point Likert scale, ranging from “excellent” to “poor.” 3 Defined as having a BMI (Body Mass Index) value greater than or equal to 30.0; based on self-reported height and weight. 4 Defined as difficulty doing any of several specific activities because of a health problem (i.e., “any physical, mental, or emotional problem or illness”). For example, push or pull large objects; participate in social activities; sit for two hours. 5 Defined as needing help with or being limited in any of the following due to a health problem: personal care needs; handling routine needs; ability to work; amount of work one can do; walking without aid; difficulty remembering or confusion; any other activities. 6 Defined as needing others’ help with personal care needs such as eating, bathing, dressing, or getting around the house (i.e., Activities of Daily Living), due to a health problem. 7 Defined as needing others’ help with handling routine needs, such as everyday household chores, doing necessary business, shopping, or getting around for other purposes (i.e., Instrumental Activities of Daily Living), due to a health problem. 8 Defined as an individual experiencing an illness or injury that kept them in bed for more than 8 days during the preceding 12 months (including hospitalization). 2

report that their health is fair or poor. Moreover, it is not simply the least educated who report higher levels of poor or fair health compared with those who are very highly educated; indeed, 13.4% of those with a high school degree and 10.1% of those with some college do so as well. Table 7 further shows seven other important measures of population health for adults aged 25–64, including the percent obese, percent with a functional limitation, percent with an activity limitation, percent needing help with activities of daily living, percent needing help with instrumental activities of daily living, percent who have been sick in bed

Socioeconomic Status | 113

for eight or more days in the past month, and percent who have two or more severe mental health symptoms in the past month. Each one of them is characterized by a substantially higher prevalence among those with the least educational attainment and a substantially lower prevalence among those with either a bachelor’s degree or a master’s degree or more. Those with a high school degree or with some college do not differ much from one another and fall in-between the extremes. For example, just 4% of those with either a bachelor’s or master’s degree or higher report being sick in bed for eight or more days in the past month, compared with 8% of those with a high school degree or some college, and nearly 10% of those without a high school degree. Finally, figure 15 shows patterns of life expectancy at age 25 by educational level, based on 2010 data. In this case, the patterns are also broken down by gender-race group. In all four demographic groups shown, there is a graded relationship such that people with a higher level of education have higher life expectancy than those with lower education in every demographic group. Moreover, those with a college degree or more live especially long lives, on average, compared to those with less education. For example, the life expectancy advantage for Black women with a college degree (56.5 years) is 4.7 years higher compared with Black women with less than a high school degree (51.8 years). The gaps are even wider for Black men, White women, and White men, respectively. Among White men, for example, those with a college degree or more live 11.9 years longer, on average, compared to their White male counterparts who have less than a high school degree. Why are the life expectancy gaps by educational attainment in figure 15 wider for men than they are for women and wider for Whites compared to Blacks? The wider disparities for men compared with women is most likely because highly educated men tend to occupy particularly advantaged social and economic positions in American society—positions that are also especially good for their health.56 That is, highly educated men tend to be very well compensated in the labor market, often occupy powerful positions in society, and are well connected to other highly educated and well-compensated men. Such economic, status, and network advantages help translate high educational attainment for men into especially good health relative to their less-educated counterparts. While highly educated women also accumulate more income, power, and healthenhancing networks than their less-educated counterparts, their potential advantages are diminished by processes of gender discrimination that play out in the labor force, in rates of pay and in women’s lesser access to

114

|

Chapter 5

65 60 55 50 45 40 Black Women < High School Degree

Black Men High School Degree

White Women

White Men

Some College

College Degree or Higher

figure 15. Estimated life expectancy at age 25 by educational attainment, gender, and race in the United States, 2010. Source: Sasson 2016

the most powerful positions in American society. As a result, while measures of population health are clearly better for both men and women who have high education, the gaps between highly and low educated men are larger than those exhibited by highly and low educated women. In a similar way, educational disparities in population health are larger for Whites than they are for African Americans.57 The shallower education gradient in population health for African Americans most likely reflects both inequality in schooling opportunities and discrimination in the labor force, processes which result in lower pay and less access to key social networks for highly educated African Americans compared with Whites. As a result, there is a dampening of health benefits for highly educated Blacks relatively to highly educated Whites in U. S. society.58

fundamental cause theory: explaining ses disparities in health and mortality As the social and health science communities began to accurately document SES disparities in key population health outcomes in the 1970s and 1980s, scholars were also beginning to search for explanations of this relationship. Why were there SES disparities in health and longevity? What were the mechanisms by which education, income, occupation, and wealth worked to influence health and longevity? Leading sociologists played an important role in developing some of the most compelling early explanations. In an important review article on the topic, David Williams

Socioeconomic Status | 115

proposed that two sets of mechanisms were responsible for SES disparities in health and mortality: medical care and psychosocial factors.59 But because medical care had become so much more widely accessible with the passage of Medicare and Medicaid in 1965, Williams argued that its explanatory power for SES disparities in health and mortality was limited. On the other hand, he argued that psychosocial factors—health behaviors, social ties, perceptions of control, and experiences of stress— were highly structured by individuals’ positions within the socioeconomic hierarchy and had become the key mechanisms that transformed low socioeconomic status into poor health. Nonetheless, Williams also argued “socioeconomic position rather than psychosocial factors or medical care is the fundamental cause of SES differences in health,” while “stress, health behaviors, and other psychosocial factors are the superficial causes.”60 Thus, Williams offered one of the first and most persuasive arguments that population-based changes in socioeconomic status, rather than individual-based changes in the mechanisms that link SES to health, were necessary for reducing and / or eliminating socioeconomic disparities in health and mortality. Around the same time, James House and colleagues were studying how socioeconomic (dis)advantages, also working through psychosocial mechanisms, accumulate with age to impact later life health outcomes. In a series of papers on the topic, House et al. showed that SES differences in health tend to be relatively small in young adulthood but are much larger in middle adulthood and early old age, as the accumulation (or lack of accumulation) of socioeconomic resources unfolded across the life course.61 This body of theoretical and empirical work was very influential in helping to usher in a life course approach to research on socioeconomic status and population health—an emphasis that continues to this very day. That is, socioeconomic resources that are important for health do not usually work immediately or on a one-time basis to influence whether and when people get sick and when they die. Rather, it is the day-to-day wear and tear of low income or the day-to-day advantages of high education that work, often on a subtle basis, to influence the long-term health and ultimate longevity of individuals in U. S. society. Research on SES and health using the life course perspective is flourishing. One notable example is recent work by Jennifer Karas Montez and Mark Hayward that focuses on whether childhood socioeconomic status or adult socioeconomic status is more important for old age health in the United States.62 The data demands to conduct such life course–based work

116

|

Chapter 5

are intensive. Fortunately, longitudinal studies such as the Health and Retirement Study (HRS) and the National Longitudinal Study of Adolescent to Adult Health (Add Health) have provided the research community with opportunities to do so. Karas Montez and Hayward used data from HRS to show that both childhood SES and adult SES are very important predictors of old age functional health in the United States. That is, older adults who experienced poverty and had parents with low education, and who themselves had low educational attainment, are by far the most likely to report that they had difficulties in carrying out basic activities of daily living (e.g., dressing oneself, getting around the house, etc.). Such findings remind us that socioeconomic contexts as children have effects on our health decades later; this also means that policy efforts to reduce childhood poverty and improve early life education likely have long-lasting effects of health. At the same time, the results of Karas Montez and Hayward’s study showed that old age adults who experienced socioeconomic disadvantages in childhood, but who later were able to graduate from high school and move on to college, had much more favorable health than socioeconomically disadvantaged children who had low educational attainment as adults. Thus, high adult SES has the potential to reduce the negative health effects of childhood socioeconomic disadvantage. In all, then, socioeconomic advantages and disadvantages experienced across the life course strongly influence the health of American adults. The intensity of theoretical and empirical work on SES and health has gained even greater momentum in recent years, especially given the widening SES disparities in health and mortality discussed above. Increasingly, social scientists have turned to Bruce Link and Jo Phelan’s “fundamental cause theory,”63 which, when combined with a life course approach perspective, provides what is the most advanced thinking on the powerful role that SES plays in the health of U. S. individuals and, by extension, in the health of the population. Fundamental cause theory draws heavily on the work of David Williams, James House, and the life course perspective discussed above. It has developed into the leading current theoretical statement regarding why SES is so strongly linked with population health in the contemporary United States. Figure 16 provides a diagram to guide discussion of fundamental cause theory.64 On the left side of the diagram, and a core premise of the theory, is the idea that SES grants individuals with an array of flexible resources to use on an everyday basis that work to enhance their health and protect against the risk of death. In other words, high SES individuals “carry” these flexible resources around with them on a daily basis and, even when

Socioeconomic Status | 117 Socio-spatial and temporal contexts Flexible resources Knowledge Money Power Socioeconomic status - Education - Occupation - Income - Wealth - Neighborhood - SES

Prestige

Mechanisms - Health behaviors - Health care - Stress - Coping - Safety - Exposure to toxins

Health/ Mortality

Social connections Life course

figure 16. Conceptual diagram depicting the relationship between socioeconomic status and health.

they may not be consciously doing so, use them to their health advantage throughout the life course. These flexible resources include knowledge, money, power, prestige, and beneficial social connections. These resources are called “flexible” because they can be used in a wide variety of ways. The flexible resource of money, for example, enables individuals of high SES to purchase homes in clean and safe neighborhoods with bicycle paths, sidewalks, local grocery stores, and high-quality schools. The flexible resource of knowledge, to cite another example, is especially important in helping highly educated individuals avoid known health risks and best understand treatment options. Using such knowledge and the empowerment that accompanies it, high SES individuals can improve their own health and the health of family members much more effectively and in many ways compared to persons who do not have such knowledge. Moreover, the flexible resource of beneficial social connections means people with high SES are much more likely to be married to a hospital administrator, to have a chiropractor in their book club, or to play poker with a cancer surgeon in comparison to people with low SES. When illness strikes, people with high SES can much more quickly and easily consult with their spouse, book club member, or poker friend regarding the best course of treatment or the best place to obtain care. SES advantages or disadvantages are compounded day after day, as depicted by the life course flow at the bottom of figure 16. Thus, it is

118

|

Chapter 5

understandable how the flexible resources of money, knowledge, power, prestige, and beneficial social connections help high SES individuals reduce their risk of heart disease, cancer, infectious diseases, accidents, and many other illnesses and causes of death. Moreover, when high SES individuals do become sick, flexible resources can also be used to hasten the pace of treatment, enhance the odds of recovery, and reduce the probability of dying from the illness. Through the daily and often unrecognized use of such flexible resources, individuals with high SES end up living, on average, far healthier and longer lives than individuals with low SES, among whom the relative lack of flexible resources leads to far more daily struggles and resulting worse health as life unfolds. Multiple Mechanisms That Link Fundamental Causes to Multiple Health Outcomes The possession of a large and varied set of flexible resources does not in and of itself translate into good health and long life; instead, flexible resources are used to reduce the risk of illness and death through their influence on a wide array of mechanisms that are closely linked to disease and death. Thus, figure 16 illustrates a second core premise of fundamental cause theory: there are many mechanisms by which SES works to influence health and longevity. Figure 16 groups them into health behaviors, health care access, stressors and coping mechanisms, and exposure to dangerous environments, toxins, and infections. But each of those groupings includes many individual dimensions. For example, health behaviors range from alcohol, tobacco, illicit drug, and prescription drug use; to dietary practices and exercise patterns; to seatbelt use and driving safe driving practices; to bicycle / motorcycle helmet use and safe riding practices; to sleep; to preventative medical and dental visits; to safe sexual behavior; to gun safety practices and more—most of which are, in fact, strongly patterned by socioeconomic status.65 And the quality of the environment includes attributes such as the cleanliness of air, water, and soil; the availability of sewage systems and regular garbage removal; the presence or absence of disease-carrying insects and rodents; rates of property and violent crime; and characteristics of the physical environment such as yards, parks, sidewalks, working streetlights, and more. Moreover, these dimensions of the environment are not only important in and around the places that people live, but also in their neighborhoods, schools, workplaces, and cities. The key point here is that the flexible resources that embody SES do not simply influence

Socioeconomic Status | 119

Box 6: Diabetes Clinics and the Massive Multiplicity of Connections In one of the best empirical elaborations of fundamental cause theory, Karen Lutfey and Jeremy Freese studied two diabetes clinics, “Park Clinic,” located in a high-income neighborhood serving mostly highincome and privately insured patients, and “County Clinic,” located in a low-income neighborhood serving mostly low-income and publicly insured or uninsured patients.a Their analysis showed that a “massive multiplicity of connections” explained why patients at Park Clinic had better diabetes outcomes than patients at County Clinic. For one, the conditions of the clinic were qualitatively different. Park Clinic had greater continuity of care with patients, higher-quality in-clinic educational materials, and was serviced by doctors, while County Clinic was serviced by doctors and medical residents. The patients also differed at the two clinics. Park County patients had greater income, fewer occupational constraints, and more social support. These differences meant that Park County patients were more likely to attend appointments and more likely to adhere to treatment plans because they were more able to get paid time off and to afford child care, had access to private transportation, were better able to fill prescriptions, and perceived fewer costs of implementing lifestyle changes. Yet the health care providers at the two clinics perceived these differences in behaviors as inherent differences between groups, rather than as differences in the socioeconomic profiles of their patients, and they adjusted their care accordingly, recommending less strenuous (and less effective) management plans to the County Clinic patients. One important insight of Lutfey and Freese’s analysis is that proximate health mechanisms—things like high-quality health care—can be compensatory resources, or factors that benefit the disadvantaged more than the advantaged because the disadvantaged have more to gain. However, in the case of the diabetes clinics, and in many other cases, compensatory resources are more available to people with the least need for them. An implication of this insight is that insofar as there is structural social inequality, new developments in prevention, treatments, cures, and technology will tend to exacerbate socioeconomic health disparities rather than ameliorate them, especially in the short term. a. Lutfey and Freese 2005.

120

|

Chapter 5

health in one simple way; there are literally dozens and dozens of mechanisms at work that translate socioeconomic advantages and disadvantages into health. The issue of multiple mechanisms is one of the key reasons why a policy or programmatic focus on single mechanisms (e.g., improving people’s patterns of exercise) or single diseases (e.g., improved diagnosis of prostate cancer) is not nearly enough to reduce or eliminate SES disparities in health. As such, Link and Phelan have argued, it is far too simplistic to tell individuals that they need to eat healthier, exercise more, use bike lanes, or avoid dirty environments to facilitate good health. Such choices are not always options or, at the least, are constrained by the amount of flexible resources available to individuals. Instead, fundamental cause theory stresses that policies and programs to improve population health—and to reduce inequalities in population health—will need to much more strongly and urgently consider those that work to improve the availability of flexible resources available for individuals, especially those with low SES.66 In other words, basic inequalities in SES must be addressed if U. S. society is to reduce disparities in population health. The Reproduction of Socioeconomic Disparities in Health A third core premise of fundamental cause theory is that SES inequalities in health are constantly being reproduced, and particularly so during periods of rapid social change. Thus, when new health challenges arise, new medical breakthroughs emerge, or innovative new technologies are developed, flexible resources are especially important in helping to determine which groups of people are able to fight off the new health challenge, learn about and use the new medical science, and / or adopt the new technology.67 In times of rapid social change, such as in the United States over the past half-century, SES has become the defining, or most fundamental, social characteristic for differentiating access to the knowledge, networks, and power that is needed to take advantage of rapidly emerging health information, science, and technology. Thus, just as some health-enhancing ideas and technologies diffuse throughout the population and reduce socioeconomic disparities in health, other new health-enhancing innovations are distributed first among those with high SES and work to continually reproduce health disparities.68 A classic example of this is how SES disparities changed—or failed to change—over the course of the epidemiologic transition. There were large socioeconomic disparities in health and mortality when infectious

Socioeconomic Status | 121

Box 7: SES Health Disparities and Sudden Infant Death Syndrome An important example of the reproduction of SES health disparities occurred in the case of infant mortality in the 1990s and 2000s. In this case, infants born to highly educated parents in the United States were most likely to benefit from new research on sleeping position. Back in the late 1980s and early 1990s, breakthrough research showed that infants were much less likely to die from sudden infant death syndrome (SIDS) if they were placed to sleep on their back instead of their stomach. The reason was that infants sleeping on their backs were far less likely to suffocate in mattresses and pillows than were infants who were placed to sleep on their stomachs. News of this scientific breakthrough spread first through highly educated communities—in doctors’ offices and universities, among individuals reading scientific magazines, and so on. Thus, it was not a surprise that SIDS death rates declined most rapidly among babies born to highly educated parents, among whom knowledge of the research on infant sleep position was first diffused. Thus, by the year 2000, even though mortality from SIDS had declined across all groups, there was a wider SES disparity in the SIDS death rate than had been the case prior to this scientific discovery, most likely because knowledge of the new discovery first reached high SES individuals and only later diffused through lower SES networks.a This example, and the more general case of the reproduction of SES health disparities, suggests that for health interventions to be effective for people of all socioeconomic status groups they need to consider what preexisting resources people have to learn about them and implement them. a. Pickett, Luo, and Lauderdale 2005.

diseases were the primary cause of death. Observing this, scholars attributed the higher mortality of lower-income groups to direct risk factors for infectious disease—greater exposure to unclean water and poor housing—and argued that the effective prevention and treatment of infectious diseases would eliminate the SES disparity, as these risk factors would presumably be irrelevant to chronic diseases. However, even as chronic diseases emerged as the primary causes of morbidity and mortality in the United States, socioeconomic disparities in health persisted. The immediate risk factors may have changed from unclean water to smoking, for instance, but what did not change was that in

122

|

Chapter 5

both periods, in both disease environments, those with greater flexible resources could better prevent and treat illness and disease than those with fewer resources. The Social Context of Fundamental Causes The above sections emphasize that SES is a powerful influence on population health because of the flexible set of resources that individuals with a certain level of socioeconomic status enjoy in the contemporary United States. At the same time, it is important to note that SES has not always been so strongly associated with population health and, in fact, it may not be in other social contexts around the world. Temporally, Bruce Link has argued that social factors like socioeconomic status have become especially important in modern societies because technological advances, the rapid availability of health-enhancing information, and advanced medicine now facilitate human control over health more so than at any point in history.69 Thus, in highly technical, information-oriented, medically advanced societies, socioeconomic status and the resources embodied with it play a critical role in which individuals gain (early) access to the information, technology, and medicine that help to maintain health and lengthen life. More recently, David Baker and colleagues have built upon the importance of historical context in the SES-health relationship by proposing the “Population Education Transition Curve” as a unifying framework for understanding the changing importance of SES.70 Baker and colleagues argue and provide empirical support for the idea that high SES may initially be related to poorer health and health behavior when societies are transitioning to high-income contexts because pleasurable, health-damaging behaviors (e.g., cigarette smoking; eating fatty foods) are predominantly affordable and accessible only among high-SES individuals. But once societies fully transition to a high-income context, socioeconomic gradients in health quickly reverse so that high SES becomes strongly associated with more favorable health behavior and outcomes (and vice versa), most notably because flexible resources become so critical for tapping into health-enhancing information, science, and technology. In short, the technological-informational-medical context of modern society has helped to carve out a central role for SES as a determinant of individual- and population-level health.71 At the same time, though, it is not just the historical context that matters for the SES-health relationship; social context also makes a differ-

Socioeconomic Status | 123

ence. Consider, for example, the recent finding from Jere Behrman and colleagues that educational attainment is only weakly associated with lower rates of hospitalization and mortality in contemporary Denmark.72 Such a finding is, of course, very different than what many researchers have documented in the U. S. context. What is it about Denmark that constrains educational disparities in health? Perhaps it is the case that there are more social welfare programs in place to protect the health of the less-educated population in Denmark, at least compared with the United States? Or perhaps it is the case that the resources embodied in high levels of education in Denmark (e.g., money, prestige, etc.) are not as pronounced in comparison to the United States; in other words, income, wealth, and prestige are not nearly as unequally distributed in Denmark. Within the U. S., Jennifer Karas Montez and colleagues have similarly showed that, in some states, socioeconomic disparities in health are far wider than in other states.73 Importantly, the policy context of the states—such as minimum wage laws and the availability of Medicaid for the low-income population—seems to play an important role in influencing the magnitude of the disparities. This body of research focusing on the contextual influences of socioeconomic gradients in health is a very important topic in the sociology of health at present. Findings from this body of work have the potential to influence state and national policy in the coming decades, particularly if the research can shed clear light on how contexts can be shaped to reduce socioeconomic disparities in health.

critiques of ses as a fundamental cause of population health Despite its many strengths and broad appeal among sociologists for the contemporary understanding of SES disparities in population health, fundamental cause theory has its critics. Perhaps most important is a body of literature that suggests that strong correlations between measures of SES and measures of population health are actually due to a third set of factors that influence both SES and health. In other words, such studies suggest that SES may not be the true causal factor that produces health disparities. Instead, it may be that genetic makeup, family background, or infant and child health characteristics are the underlying factors responsible for the strong SES-health associations that are observed across the life course. These possibilities are portrayed in figure 17, which shows that at least a portion of the association

124

|

Chapter 5 Socio-spatial and temporal contexts

Birth weight

Bi rth ch Ea ild rly ho od ch Mid ild dl ho e od Ad ol es ce nc e Tr an ad t siti ul o on th oo d

Demographic characteristics Genetic predisposition Family background Child health characteristics

SES

Mechanisms

Health/ Mortality

Adulthood

figure 17. Conceptual diagram depicting the relationship between socioeconomic status and health, while taking into account intergenerational and childhood influences.

between SES and health may be due to this third set of factors. For example, one recent paper found that genes that are common to both one measure of SES (e.g., educational attainment) and two measures of adult health (e.g., depression, self-rated health) are partially responsible for the strong association between SES and these health outcomes.74 Research-wise, what more needs to be done to understand socioeconomic disparities in health in light of these challenges to fundamental cause theory? One important avenue includes efforts to build more sophisticated models of the causal effects of socioeconomic factors on health. Such work has been underway for over a decade now and has already yielded tremendous insights. For example, one innovative study found that mandatory increases in child educational attainment in the 1920s and 1930s led to lower adult mortality rates decades later.75 Another found that educational attainment and income effects on key measures of health changed very little once measures of intelligence were taken into account in the statistical model.76 Nonetheless, scientific work on the causal effect of SES on health is extremely difficult because researchers generally cannot use experimental designs in this area of study. Thus, the best efforts toward developing improved causal evidence of the SES-health association involve advanced statistical methods to try and parse out as many competing explanations

Socioeconomic Status | 125

for the SES-health relationship as possible. Other sophisticated research designs take advantage of natural experiments (e.g., changes in compulsory school laws, as mentioned above) to estimate causal effects.77 The field is also producing better data sets to address causality in the SEShealth association. For example, population health data sets such as the National Longitudinal Study of Adolescent to Adult Health and the Health and Retirement Study now include genetic data and information on family background to facilitate understanding of the extent to which SES-health associations are at least partially influenced by common genetic or family background factors.78 Moreover, such data sets are also including biomarker-based measures of health to best understand how and when socioeconomic factors “get under the skin” to influence health, such as through the immune or cardiovascular systems. Work including such life course perspectives and focusing on social-biological connections is on the cutting edge of population health science. Over the next decade, researchers employing such interdisciplinary approaches will be influential in helping the field better understand SES disparities in health.79 There is no disputing the mounting empirical evidence that socioeconomic disparities in many measures of population health in the United States are extremely wide and, for at least some key measures, have grown even wider in recent decades. This chapter argues that high SES provides individuals with access to critical health-enhancing resources that work to protect and enhance health and to do so across the complete life course. Bruce Link and Jo Phelan’s fundamental cause theory, enhanced by a life course perspective of SES and health, provides the most compelling theoretical perspective to understand this very strong association. Moreover, this chapter has suggested that the contemporary United States is a ripe context for socioeconomic disparities in health to persist, give the rapidly unfolding social and economic changes that have placed a premium on educational attainment, technological prowess, access to information, and social connections. Moreover, the distribution of socioeconomic resources themselves, particularly income and wealth, have become much more unequal since the 1970s—in part due to policy decisions such as multiple tax reductions for the wealthy and the lack of appreciable increases in the federal minimum wage. Other governmental policy decisions—including but not limited to the intentional undermining of labor unions, the weakening of the social

126

|

Chapter 5

safety net for the poor, and the lack of passage of a universal health care system—have helped to heighten the health risks for those who are in the bottom segments of the socioeconomic distribution. Thus, it is perhaps no surprise that persons with low SES live 10–15 years shorter lives than those with high SES in the contemporary United States. The conditions are unfortunately ideal for both the causes and consequences of socioeconomic inequality to flourish.

chapter 6

Race / Ethnicity, Nativity, and U. S. Population Health

Over the past century, between 60,000 and 100,000 African Americans died prematurely every year—that is, earlier than they would have if African Americans faced the same risk of death as white Americans.1 As has been pointed out in a previous chapter, this level of excess mortality is equivalent to a jumbo jet airliner full of African Americans crashing without survivors every day, day in and day out, year after year.2 In spite of these very large numbers, the premature deaths of African Americans receive little public notice. One important reason for the invisibility of these numbers is the complexity of the origins of racial and ethnic disparities in health. This chapter delves into these origins, arguing that much like socioeconomic disparities in health discussed in chapter 5, racial / ethnic disparities in health reveal and reflect basic social inequality. In this case, fundamental to that inequality is racism. To help understand racial and ethnic disparities in health, we develop a conceptual model that explains the complex interlinkages between race and ethnicity and health and mortality. Similar to the fundamental cause model laid out in chapter 5 to explain socioeconomic disparities in health, the model of racial and ethnic disparities shows how racial and ethnic stratification in social resources, which reflects the historical legacy and contemporary impacts of racism, results in different health experiences for different racial and ethnic groups. Yet, there are some instances where the relationship between race and health does not play out as expected within this context of inequality. Such population health paradoxes offer 127

128

|

Chapter 6

insights into how race / ethnicity and health are connected. Therefore, we conclude by turning our attention to three exceptions to the standard pathways linking race and ethnicity to health and mortality: the epidemiologic, or Hispanic health, paradox; the Black-White mental health paradox; and rising mortality among low-educated, middle-aged White women. We begin by defining race, ethnicity, and nativity.

defining race, ethnicity, and nativity How do you measure an individual’s race or ethnicity? Take a famous example: Barack Obama. The 44th president of the United States was born in Hawaii to a White mother from Kansas and a Black father from Kenya. His mother’s ancestry was primarily British but also included other European origins. His father was of the Luo ethnic group from western Kenya and other parts of Africa. His mother later married a man from Indonesia, where President Obama’s sister was born, and his father had seven other children in Kenya. For parts of his childhood, President Obama lived with his mother’s (White) parents in Hawaii. Later, he married Michelle Robinson, an African American woman, and they had two daughters. Given this complex mixture of origins and influences, one might measure different things depending on the criteria. Do you use an accounting of one’s ancestry, and, if so, how far back do you go? Or do you measure one’s deeply felt identity, and, if so, what does that identity reflect, if not ancestry? In 2010, President Obama’s response to the race and ethnicity question on the U. S. census form received substantial scrutiny. Although the census questionnaire allows the respondent to select multiple races (e.g., Black and White), President Obama selected only one—African American.3 Race and ethnicity are difficult to measure because these categories are socially constructed, meaning that they are created through social processes—passed down through history, embedded in institutions, and learned through interactions with others. The social construction of race and ethnicity implies that what we understand to be meaningful racial and ethnic categories changes over time and differs across contexts. The racial and ethnic categories familiar to Americans in the twenty-first century are quite different from the categories familiar to Americans in the nineteenth century or to Brazilians in the twenty-first century. Even an individual’s racial and ethnic identity can change, depending on where they are, who is asking, or what experiences they have had.4 In general, population health scientists rely on individuals’

Race/Ethnicity and Nativity | 129

self-reported race and ethnicity—the choice each person makes when responding to demographic forms and questionnaires. However, the choice set that individuals face is also socially constructed. While the racial and ethnic categories that population health scientists use reflect a studied effort to capture and simplify these complex concepts, in creating the appearance of fixed and standard racial categories, categorization can also influence how we understand race and ethnicity in the United States. Thus, the work that population health scientists do is in fact part of the social construction of race and ethnicity. Although racial and ethnic categories are not naturally occurring, the colloquial definitions of race groups in the United States implicate biology. Whereas ethnic groups are defined by common ancestral origin and cultural traits such as religion, language, and customs, race groups are often equated with biological features such as skin tone, hair color, and facial features. However, race does not capture real biological or genetic differences between groups of people; in fact, there is more genetic variation within races than there is across them.5 Rather, the putative biological origins of race have been used to legitimize the subjugation of African Americans, Native Americans, Hispanics, and Asian Americans by Whites. This is the very definition of racism: an ideology of group inferiority used to justify the unequal treatment of racial minorities (i.e., discrimination) by individuals and institutions.6 Race in the United States captures the historical and contemporary experiences of racism—economic marginalization, political exclusion, and social stigmatization—that are entwined with deeply held political, cultural, and social identities. Bringing attention to racial and ethnic disparities in population health, as we do in this chapter, is one way to track the impact of racism in the United States, as well as to understand resilience against racism. How (well) do our current systems of population data capture the concepts of race and ethnicity? Following standards set by the U. S. Office of Management and Budget, the U. S. Census Bureau groups Americans into five major racial groups: White, Black, Asian, Native Hawaiian or Pacific Islander, and American Indian or Alaska Native (AIAN). As shown in the first column of table 8, in 2015, an estimated 77% of the U. S. population identified as White, 13% as Black, 5.6% as Asian, 0.4% as Native Hawaiian or Pacific Islander, and 1.2% as American Indian or Alaska Native. However, these numbers are misleading, particularly for Whites, because they exclude Hispanics. The U. S. Census Bureau currently considers “Hispanic” as an ethnicity made up of people of different races and

130

|

table 8

Chapter 6 racial and ethnic makeup of the u. s. population in 2015

Including Hispanics

Separating Hispanics

White Black American Indian or Alaska Native Asian Native Hawaiian and Pacific Islander Two or more races

77.3 13.0 1.2

Total

100

5.6 0.4 2.4

White Black American Indian or Alaska Native Asian Native Hawaiian and Pacific Islander Two or more races Hispanic

61.8 12.2 0.78

Total

100

5.5 0.3 1.8 17.5

sources: Xu et al. 2016

defined by a common language (Spanish) and ancestral origin in either Latin America or Spain. Therefore, the census questionnaire first asks whether one is Hispanic and then about race, which confuses many people of Hispanic origin who consider “Hispanic” (or Latino / a, Chicano / a, Mexican, etc.) to be their race.7 In answering the 2010 census race question, the majority of Hispanic respondents (53%) selected “White,” but a substantial minority (37%) chose “some other race.”8 Because scholars—and most people—conceive of Hispanics as a separate “racial / ethnic group,” this particular formulation means that demographers and other social scientists are often stuck using the cumbersome terms “non-Hispanic White,” “non-Hispanic Black,” “non-Hispanic Asian,” and “Hispanics of any race.” The second column of table 8 shows that in 2015, an estimated 17.5% of the U. S. population identified as Hispanic of any race, 61.8% identified as non-Hispanic White, 12.2% as non-Hispanic Black, 5.5% as non-Hispanic Asian, 0.8% as non-Hispanic American Indian or Alaska Native, and 0.3% as non-Hispanic Native Hawaiian or Pacific Islander. Since 2000, respondents to census questionnaires have been allowed to select multiple race groups, but only a small percentage of the U. S. population does so: 2.4% in 2015 (see table 8). The low rate of multiracial identification on the census does not capture the extent of mixed racial ancestry in the population. Using genetic information from customers of the personal genetics firm 23andme, one study determined that the average self-identified African American customer had 24% White ancestry while the average self-identified White customer had less than 1% African ancestry.9 This implausible asymmetry in racial admixing may reflect

Race/Ethnicity and Nativity | 131

the social process underlying racial self-identification. Because of the legacy of the “one-drop rule,” which held that anyone with African ancestry (even just one drop) be classified as Black, most people with Black and White ancestry identify as Black. The most famous case is Barack Obama. Each race includes people from multiple ethnic groups, as well as Americans who do not identify with a particular ethnicity. For some, ethnicity, not race, is their primary social identity. Recognizing this, scholars have referred to some census race groups—in particular, Asian and Pacific Islanders (API) and AIANs—as panethnicities, which is the grouping of a variety of related ethnicities under one “panethnic” umbrella.10 Indeed, current U. S. census questionnaires lists multiple Asian and Hispanic ethnicities and ask AIAN respondents to provide their tribal affiliation, but currently census questionnaires provide no ethnic options for respondents who identify as Black or White. Missing from the census is a separate category for people of Arab, Middle Eastern, or North African descent. In 2000, the majority of respondents declaring an Arab ancestry selected White as their race.11 In 2010, Arab American activists led a movement with the slogan “Check it right; you ain’t white!” to encourage Arab Americans to select “some other race” on the 2010 census form. While there was much discussion and research into the U. S. Census Bureau potentially adding Middle Eastern or North African (MENA) as a separate racial category on the 2020 form, in the end the bureau did not do so.12 Indeed, some of the current limitations of census measurement of race and ethnicity—the arbitrary separation of Hispanic ethnicity from race, the limited ethnic options for Whites and Blacks, and the omission of a separate category for MENA respondents—was under strong consideration by the Census Bureau for inclusion on the 2020 census form. For example, the Census Bureau strongly considered a combined race / ethnicity question for 2020 with seven race / ethnic options: White, Hispanic / Latino / Spanish, Black or African American, Asian, American Indian or Alaska Native, Middle Eastern or North African, Native Hawaiian or Other Pacific Islander, or other, with either write-in or check-box style detailed ethnic options for each group.13 Despite much support for such changes from the scientific community, the 2020 race / ethnicity options will remain similar to those from 2010.14 As of February 2018, the Census Bureau had not publicly provided a reason for sticking with its 2010-based measurement of race and ethnicity. While we recognize the limits of current measurement of race and ethnicity, as population health scientists we are further constrained by what

132

|

Chapter 6

is available in population data. For the most part we believe the data we have do a fairly good job at capturing broad racial and ethnic disparities in health. Thus, in this chapter we consider five major racial and ethnic groups: (non-Hispanic) Blacks, Whites, Asian and Pacific Islanders, and American Indians and Alaskan Natives, and Hispanics of any race. Rather than draw distinctions between race and ethnicity, we consider “race / ethnicity” to refer to a single set of meaningful sociopolitical categories in the United States. When possible, we include subgroups and / or note how the results would be different under different systems of categorization. In addition to race / ethnicity, we examine differences among groups by nativity, or country of birth, distinguishing between foreign-born immigrants to the United States and those born in the United States. In 2013, 13.1% of the U. S. population was foreign born, with 28% originating from Mexico, another 24% from other countries in Latin America, 26% from Asia, and the remainder from Europe, Canada, Africa, and other origins.15 The foreign born are also referred to as first-generation immigrants, and their children and grandchildren are referred to as secondand third-generation Americans, respectively.

racial and ethnic disparities in health and mortality Contemporary U. S. disparities in health and mortality show mostly advantages for Asians and Whites; some advantages but some disadvantages for Hispanics; and disadvantages for Blacks and American Indians, with a few key exceptions. For example, Asian Americans have the longest current life expectancy, at 86.5 years, followed by Hispanics (82.8 years), Whites (78.9 years), Native Americans (76.9 years), and Blacks (74.6 years).16 A similar pattern is observed across most causes of death, as shown in table 9, which presents the age-adjusted rates of mortality for all causes and for the 15 leading causes of death. The rate of all-cause mortality for Asian and Pacific Islanders is 48% lower than that for Whites, for Hispanics it is 30% lower than that for Whites, and for Blacks it is 17% higher than that for Whites. There is some variation in this pattern across cause of death. For example, whereas Hispanics have lower or similar rates of mortality to Whites for 13 causes, they have a higher rate of mortality from diabetes and liver disease. And whereas Blacks have higher or similar rates of mortality to Whites from 10 causes of death, they have lower rates of mortality from Alzheimer’s disease, accidents, liver disease, chronic lower respiratory disease, and

sources: Kochanek et al. 2016

All causes Heart disease Cancer Chronic lower respiratory disease Accidents Cerebrovascular disease Alzheimer’s disease Diabetes Flu and pneumonia Kidney disease Suicide Septicemia Liver disease Hypertension Parkinson’s disease Pneumonitis

523.3 116.0 112.4 17.5 26.8 30.2 19.8 25.1 15.2 11.1 6.3 8.3 14.5 7.5 5.4 3.0

Hispanic 742.8 171.3 166.2 45.4 42.6 35.4 26.8 18.6 15.1 12.1 16.4 10.3 10.6 7.3 8.1 5.4

White 870.7 210.8 190.2 28.9 35.0 50.9 22.7 38.2 16.3 25.3 5.7 18.4 7.3 16.0 3.9 5.4

Black 388.3 86.1 98.9 12.5 15.1 28.3 12.1 15.0 12.9 8.2 6.0 4.8 3.5 6.7 4.3 3.2

API 0.70 0.68 0.68 0.39 0.63 0.85 0.74 1.35 1.01 0.92 0.38 0.81 1.37 1.03 0.67 0.56

Hisp:Wht 1.17 1.23 1.14 0.64 0.82 1.44 0.85 2.05 1.08 2.09 0.35 1.79 0.69 2.19 0.48 1.00

Blk:Wht 0.52 0.50 0.60 0.28 0.35 0.80 0.45 0.81 0.85 0.68 0.37 0.47 0.33 0.92 0.53 0.59

API:Wht

0.60 0.55 0.59 0.61 0.77 0.59 0.87 0.66 0.93 0.44 1.11 0.45 1.99 0.47 1.38 0.56

Hisp:Blk

table 9 age-adjusted rates of mortality for five major racial / ethnic groups, and rate ratios between certain groups, for all causes and the 15 leading causes of death, 2014

134

|

Chapter 6

especially suicide, which we will return to when we discuss the BlackWhite mental health paradox. Asian and Pacific Islanders have lower rates of mortality than Whites from all causes, and Hispanics have lower rates of mortality than Blacks from all causes except for suicide, liver disease, and Parkinson’s disease. Few studies of racial and ethnic disparities in population health include the AIAN population because it is relatively small, extremely heterogeneous, and vital statistics data for this population are considered unreliable due to misreporting of AIAN identity on death certificates. A study of racial disparities across U. S. counties was able to overcome this problem by identifying AIAN deaths on tribal reservations in the western United States.17 The life expectancy of the “Western American Indian” population was 5.9 and 4.3 years lower than men and women living in “Middle America,” which is made up of mostly Whites of slightly above-average per capita income and education. The Western American Indian population has especially high death rates from car accidents, cirrhosis of the liver, and diabetes, reflecting the severe social disadvantages of this population. Studies that use vital statistics mortality data also show elevated mortality risks for this group; for instance, a recent study of mortality among 25–64 year-olds found that the allcause mortality rate in 2012–16 was higher for the AIAN population than for any other group.18 However, there is tremendous heterogeneity within the AIAN population by tribe and ethnic identity, and very few population health studies are able to adequately disaggregate the AIAN population due to data constraints. The general pattern of racial and ethnic disparities in mortality is mostly replicated across different measures of health. In figure 18, we summarize racial and ethnic disparities along 19 measures of health, comparing five racial and ethnic groups to Whites.19 By and large, Asians fare the best of all groups. They fare better than Whites on 16 outcomes, were no different on 2, and were worse on only 1 (the rate of low birth weight). Hispanics fared better than whites on 8 outcomes, were no different on 3, and were worse on 8 (while data limitations prevented assessment on 4 outcomes). Blacks fared better on only 1 outcome (breast cancer incidence), were no different on 1, and fared worse on 17. American Indians and Alaska Natives fared better on 4 outcomes and fared worse on 13 (while data limitations prevented assessment on 2 outcomes). The mixed picture of health for Hispanics has been shown in a number of studies. On certain key population health measures—the infant mortality rate, the rate of low birth weight, and life expectancy—

Race/Ethnicity and Nativity | 135 1 1

4

17

13

8

16 3

8 2 2

1 Asian Better

Hispanic

Black

No difference

Worse

AIAN Data limitations

figure 18. Number of health measures on which racial and ethnic groups fare better, no differently, or worse than Whites. Notes: The 23 conditions include self-reported health, physical limitations, 14 or more physically unhealthy days in past 30 days, 14 or more mentally unhealthy days in past 30 days, adult obesity, child obesity, adult asthma, child asthma, diabetes diagnosis, heart disease diagnosis, HIV diagnosis, AIDS diagnosis, HIV death rate, cancer incidence rate, breast cancer rate, colorectal cancer rate, lung cancer rate, preterm birth rate, and rate of low birth weight. Sources: Kaiser Family Foundation analysis (Artiga et al. 2016b) of the 2014 National Health Interview Survey (CDC / National Center for Health Statistics 2017a), 2011–2014 National Health and Nutrition Examination Survey (CDC / National Center for Health Statistics 2015), and National Center for HIV / AIDS, Viral Hepatitis, STD, and TB Prevention (NCHHSTP) Atlas (CDC / National Center for HIV / AIDS, Viral Hepatitis, STD, and TB)

Hispanics do quite well, especially given their overall low socioeconomic status relative to Whites.20 As we discuss at greater length below, this does not appear to be the result of measurement error, such as from the misclassification of Hispanic identity on death certificates or other sources.21 Hispanics also have relatively good health on some indicators of chronic conditions, particularly asthma, lung disease, and cancer.22 On the other hand, Hispanics experience elevated rates of disability, particularly in older ages, as well as obesity, diabetes, and overall metabolic risk.23 Their high rates of disability, coupled with relatively long

136

|

Chapter 6

14 12

IMR

10 8 6 4 2

la an n A d O me Sou th ric th er a H n isp an ic

n

C ub

ca

an

Ri

ic

to

M ex

er

en

tra

Pu

C

N

at

ive

O

th

er

AP I Am er ic an

e

n

m es na

re a

et

se

Vi

Ko

o

ne

se

in lip

pa Ja

Fi

k ac

ne C hi

Bl

W

hi

te

0

US

FB

figure 19. Infant mortality rate by race / ethnicity and place of origin. Notes: Individuals not reporting a race / ethnicity are not included. State-level weights for missing infant deaths are applied to these numbers. Source: National Vital Statistics System Linked Birth and Death Certificates 2007–2008 (National Vital Statistics System 2018)

life expectancy, means that Hispanics face a relatively long period of disabled life expectancy compared to other groups.24 The across-the-board health advantage of Asians and the good health and mortality on some outcomes for Hispanics reflect to some degree the good health of large immigrant subpopulations. Across all groups, immigrants tend to have better health outcomes than their U. S.-born counterparts. Figure 19 shows the infant mortality rate (IMR)—the number of deaths to infants through their first birthday, expressed per 1,000 live births—in 2007–08 for 29 race-ethnicity-nativity groups, based on mothers’ race / ethnicity and place of birth. Differences by nativity are shown for 14 race / ethnic groups (no foreign-born group is presented for AIAN). For 13 groups, the IMR is higher among the U. S. born than among the foreign born. For example, among White mothers, U. S.-born mothers have an IMR of 5.6 deaths per 1,000 live births, which is 1.5 deaths greater than the IMR among foreign-born mothers (4.03). Among mothers of Vietnamese origin, the IMR of U. S.-born mothers (8.36) is more than 2 times greater than the IMR of foreignborn mothers (3.77). The one exception to this pattern is among Puerto Ricans, for whom there is a higher IMR among island-born Puerto Ricans compared to mainland-born Puerto Ricans.

Race/Ethnicity and Nativity | 137 85

20 18 16

Life expectancy at birth

14 65 12 55

10 8

45 6 4

White-black gap in life expectancy (years)

75

35 2 0 1900 1904 1908 1912 1916 1920 1924 1928 1932 1936 1940 1944 1948 1952 1956 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012

25

White

Black

Gap

figure 20. Life expectancy at birth for Whites and Blacks, and the White–Black gap in life expectancy at birth, from 1900 to 2013. Sources: Arias 2015; Kochanek et al. 2016

Contemporary levels and disparities of mortality reflect vast improvements that have occurred over the past century. Figure 20 presents historical trends in life expectancy at birth for Blacks and Whites, the only two groups for whom we have consistent mortality data from 1900 to 2013. The graph shows that there was exceptional growth in life expectancy for both Blacks and Whites over this period. In 1900, Blacks could expect to live 33 years, on average, given deaths rates at that time, whereas Whites could expect to live 47.6 years. By 2013, life expectancy at birth was 75.5 years for Blacks and 79.1 years for Whites, meaning that life expectancy grew by 31.5 years for Whites and by 42.5 years for Blacks over this 113-year period, a remarkable change in the longevity of these two major groups in the U. S. population. That life expectancy grew by a larger amount for Blacks than for Whites means that the gap in Black-White life expectancy shrunk over this period. In fact, it did so dramatically, especially between 1903 and 1950. In 1903, life expectancy was an estimated 17 years longer for

138

|

Chapter 6

Whites than for Blacks. By 1950 the gap had decreased by more than half, to 6.8 years. It decreased by another 47%, to 3.6 years, by 2012. The notable increase in the gap between 1984 and 1993 was largely the influence of deaths from HIV.25 Since 1993, the declining gap was driven by relatively faster improvements in mortality among Black men and women from heart disease, as well as from improvements in mortality from HIV and homicide among Black men. But it was also driven by worsening mortality from unintentional injuries—in particular, drug and alcohol-related poisoning—among Whites at least through 2012.26

why are there racial and ethnic disparities in health? Figure 21 displays a conceptual model revealing the pathways through which racial / ethnic and nativity disparities in health are created. Figure 21 is similar to the model presented in chapter 5 on socioeconomic status as a fundamental cause of health, insofar as it portrays health as the outcome of health-related, or proximate, mechanisms that vary systematically according to a person’s or group’s access to key social resources.27 The proximate mechanisms linking social resources to health and mortality capture the various ways that individuals control and affect their health by preventing and treating disease and injury. The proximate mechanisms that matter at any point in time reflect the health environment—that is, what diseases and causes of death are most prevalent. Currently, with the primary causes of death from chronic diseases such as heart disease and cancer, health behaviors such as smoking, access to high-quality health care, and the ability to control stress are key proximate mechanisms. During the nineteenth century, when the primary causes of death were infectious and parasitic, the most important proximate mechanisms included water and food quality. While the proximate mechanisms linking social resources to health may change, what does not change is the fact that people and groups with greater social resources are better able to control their health. The most obvious of these resources, and the one for which the fundamental cause model was originally developed, is socioeconomic status. Educational attainment, occupational status, income, and wealth are flexible resources that can be transferred across contexts to prevent and treat disease through the proximate mechanisms in the model. The model adds five other social resources to socioeconomic status: neighborhoods, schools, social connections, freedom, and knowledge.28 While related to

Race/Ethnicity and Nativity | 139 Geographic and temporal context Social resources

Race/ Ethnicity Racism & Nativity

- Socioeconomic - Neighborhoods - Schools - Social - connections - Freedom - Knowledge

Mechanisms - Health behavior - Health care - Stress - Safety - Coping/ resilience

Health/ Mortality

Lifecourse, gender, and immigration

figure 21. Conceptual model linking race / ethnicity and nativity to health and mortality.

each other, the model implies that each of these social resources independently affects an individual’s ability to prevent and treat disease. Importantly, each of these social resources is also stratified by race / ethnicity and by nativity. In other words, the model shows that racial / ethnic and nativity differences in health and mortality are explained by racial / ethnic and nativity differences in social resources, which in turn differentially constrain racial and ethnic and nativity groups’ ability to control their health. This is an essential point of the fundamental cause model. It implies that racial and ethnic disparities in health and mortality do not result from racial and ethnic differences in individual behaviors. Although individual behaviors matter for health, we cannot understand individual behaviors without understanding the social factors and contexts that put some people at greater or lesser risk of those behaviors. This implies an important policy implication: so long as racial / ethnic disparities in social resources persist, public health policies focused on proximate mechanisms such as smoking or exercise will not eliminate racial / ethnic disparities in health even if those policies are effective at reducing smoking or increasing exercise or changing whatever proximate mechanism they target. Key in this model is the leftmost gray arrow labeled racism. Racial and ethnic stratification in social resources is the result of contemporary and historical racism in the United States—that is, the systematic restriction of access to society’s goods, including work and fair pay, equal housing, quality schools, and so on, to members of racial and ethnic minority groups through exploitation and exclusion. The brutal history of racism in the United States—the slaughter of Native Americans and

140

|

Chapter 6

appropriation of their lands; the capture, trade, and enslavement of Africans and African Americans; and the Jim Crow era of legal segregation, among many other atrocities—created a legacy of inequality that continues to be felt today in spite of the achievements of the civil rights movement. And even 50 years after the Civil Rights Act, in an era some have called “colorblind,” racism continues to produce systematic and persistent racial stratification in social resources. One central way that racism is revealed and enacted, and racial stratification in resources is created, is through racial residential segregation, which is the separation of racial and ethnic groups in housing. Contemporary levels of segregation in the United States are high, although lower than they were in the past. In 2010, the average Black person lived in a metropolitan area where 60% of Blacks would have to move in order to achieve Black-White integration, a rate of segregation that was down from 78% in 1970.29 In 2010, the average Hispanic person lived in a metro area where 49% of Hispanics would have to move in order to achieve Hispanic-White integration, and the comparable rate for Asians was 41%.30 These levels of segregation are higher than we would expect given the socioeconomic status of minority groups, and they generally do not decline with growth in socioeconomic status of minority groups, suggesting that processes outside of socioeconomic stratification matter—that is, it is not just higher levels of poverty or lower levels of income among minority groups that prevent spatial integration.31 In spite of the 1968 Fair Housing Act and the 1974 Equal Credit Opportunity Act, which made racial discrimination in housing and lending markets illegal, segregation persists through a variety of practices, including racial steering by real estate agents, real estate pricing, low-density zoning, public housing provision, and mortgage lending practices.32 This differential treatment is documented by audit studies that send trained auditors with identical economic profiles—income, credit, and so on—to search for homes or apply for a loan. These studies invariably find that Black and other minority auditors are shown fewer and different homes and neighborhoods and receive less financial assistance and less agreeable financial terms than White, but otherwise socioeconomically identical, auditors.33 Housing segregation is also driven by individual preferences and actions. Among Whites, preferences for segregated neighborhoods are closely tied to racial prejudice.34 As a result of the forces of segregation, minority groups cannot follow a standard route of mobility: transferring income gains into mobility into better neighborhoods.35

Race/Ethnicity and Nativity | 141

Thus, segregation creates spatial and socioeconomic mobility traps for minority groups. And residents in segregated neighborhoods are nearly always subject to distinctly unequal institutions, opportunities, services, social networks, environmental risks and hazards, and levels of crime.36 When groups are segregated by both class and race, as is the case in the United States, and poverty rises, poverty becomes spatially concentrated for segregated groups, meaning that the social dislocations associated with poverty—crime, drug and alcohol addiction, and toxic physical environments, among others—also become spatially concentrated.37 As a result of segregation, prejudicial practices and actions can be concentrated on minority groups by focusing on the places they live, with little consequence for Whites who are sheltered by physical separation. For example, predatory lenders target Black neighborhoods, which resulted in high levels of foreclosures in highly segregated metropolitan areas during the Great Recession of 2008–10.38 These conditions have clear consequences for health above and beyond their impact on the socioeconomic well-being of minority residents: segregated neighborhoods have fewer healthy food options and recreational spaces, more fast food chains and convenience stores selling alcohol and tobacco products, greater exposure to toxic environmental conditions, and less access to quality medical care, a context that makes it very difficult for residents to control risks to their health.39 In a study of Chicago neighborhoods, racial disparities in cumulative biological risk—an index of eight biomarkers indicating high risk for disease—were completely explained by differences in the neighborhoods that Blacks, Whites, and Hispanics live in, whereas individual socioeconomic status did not fully account for group differences in cumulative biological risk.40 Thus, racial segregation is a structural mechanism of racial inequality, one way that racism is enacted and embedded in the structures of our society to produce unequal access to social resources, including environments that empower and enable people to prevent and treat disease. Yet even if separate spaces were materially equal, segregation imposes social distance between groups, limiting the potential for neighborly interaction and community building—the very things that might reduce prejudice and discrimination in the first place. Racism has also been documented in other social structures—particularly in employment, in schools, and in the criminal justice system. Audit studies similar to the ones testing for discrimination in housing and lending markets test for discrimination in employment and find that Black and Latino job seekers are far less likely to be given interviews than

142

|

Chapter 6

Whites, even with identical education and work experience.41 In schools, studies have found evidence of bias in the evaluations of Black students by White teachers, and in spite of the formal desegregation of schools in 1954, our education system remains highly segregated and unequal.42 Perhaps most striking is inequality in the criminal justice system. The U. S. prison incarceration rate grew fivefold between 1970 and 2005, and by the early 2000s, the United States incarcerated 1% of its adults in prisons.43 Among men without a college education, 30% of Blacks had spent time in prison by their mid-thirties, and 60% of Black men who did not complete high school had spent time in prison. The rate of incarceration is 6–8 times higher among Black men than it is among White men. These rates “establish . . . incarceration as a normal stopping point on the route to midlife” for Black men born after the 1960s.44 Growth in the incarceration of minority men is not accounted for by growth in criminal activity but has been driven instead by racial differences in policing, sentencing, and parole. Like segregation, incarceration not only reflects but also perpetuates racial inequality, as people who have been incarcerated earn lower wages, are less likely to be employed, are highly likely to reenter the criminal justice system, and have far fewer rights.45 As the model in figure 21 makes clear, this racial structuring of inequality influences proximate determinants of health, including levels of stress, access to and quality of health care, health behaviors, safety, and coping and resilience. But racism is also perpetuated through individual actions and interactions. In 2015, the U. S. Federal Bureau of Investigation reported more than 5,800 hate crime incidents, 57% of which were motivated by race, ethnicity, or ancestry bias, and the majority of those were bias against African Americans.46 Explicit racial bias is also documented in survey data. As recently as 2008, just under 30% of Whites responded that individual home sellers have the right to discriminate in sales, only 25% said they were willing to live in a neighborhood where half of residents were Black, and 25% opposed or strongly opposed a close family member marrying a Black person.47 These rates may be underestimated if respondents are reluctant to reveal discriminatory attitudes to survey takers. Indeed, 56% of Whites reported in the General Social Survey that they perceived racism against Blacks to be widespread.48 While explicit racial bias persists in the post–Civil Rights era among a substantial minority of Whites, an arguably much greater portion of individuals hold implicit racial biases. Implicit biases arise from subconscious associations based in racial stereotypes. Subconscious associations are mental shortcuts between concepts that our brains use to make quick

Race/Ethnicity and Nativity | 143

judgments and decisions without conscious deliberation. The nature of these shortcuts—what concepts are linked together—reflects social understandings of the world and include neutral links, such as between the concepts “snow” and “cold,” as well as links based in racist cultural stereotypes, such as between “Black” and “criminal.” Because we act on subconscious associations without conscious awareness, implicit racial biases have the effect of perpetuating racism even without the intent of the actor. Many observers argue that it is because of implicit racial bias that people of color are less safe in their interactions with the police. One study found that research participants instructed to “shoot” images of armed men in a video game simulation were faster and more accurate in making shooting decisions regarding armed Black men.49 Notably, the study found that the speed and accuracy of shooting decisions were not related to explicit racial prejudice. Implicit racial biases can affect all types of human interactions, from employment decisions to communication of medical information to disciplinary actions in schools. Thus, racism is embedded in the structures of our society, as revealed through segregation, schools, and the criminal justice system, as well as in individual belief systems and decision-making processes, as revealed through explicit and implicit racial biases and their impact on actions. The effect has been to continue to systematically deny equal access to social resources to members of racial and ethnic minority groups. The model in figure 21 therefore shows how racism structures racial and ethnic stratification in resources, and this stratification in turn differentially constrains the ability of members of racial and ethnic groups to control their health and prevent disease via the proximate determinants in the model. But racism also directly impacts these proximate determinants. In particular, being aware of a racist ideology—which asserts that the racial or ethnic group you belong to is inferior—and experiencing racial discrimination is stressful. Scholars of racial dynamics show how stereotype threat—the simple awareness of a stereotype about one’s group—can undermine performance and self-esteem.50 And studies find that experiences of racial violence and perceptions of discrimination are associated with a variety of health outcomes.51 Being treated differently simply because of the color of one’s skin, the texture of one’s hair, or the shape of one’s eyes is isolating, infuriating, humiliating, and degrading, all of which are bad for health. The labels on the outside bars of the model make the simple point that these processes are informed by and interact with context—when and where one lives—as well as by other key demographic contours, including

144

|

Chapter 6

life course stage, gender, and immigrant status. Experiences of racism vary across time and over place in the United States. What resources and what mechanisms matter most differ for someone in their early twenties versus someone at retirement. Race and ethnicity interact with gender and immigrant status to differentially inform the experiences of men and women, and those born in the United States versus those who migrate here. These interactions hold clues for three epidemiologic paradoxes, or empirical patterns that appear inconsistent with the model we have just described: recent immigrants tend to have better health than the native born, Blacks have better mental health outcomes than other groups, and mortality rose for lower-educated, middle-aged White women from 1999 to 2012. These paradoxes can be understood by allowing the model to vary by age, gender, and social context. We turn to these paradoxes now, beginning with the immigrant health / Hispanic health paradox.

immigrant health / hispanic health paradox In 2014, immigrants made up 13.3% of the U. S. population.52 This group of first-generation Americans is highly diverse in terms of regional origin, legal status, and social resources. Among immigrants who obtained legal permanent resident status (green cards) in 2015, 8% were from Europe, 36% from Asia, 46% from Latin America, 9% from Africa, and 1% from Oceania.53 In 2014, 47% of immigrants were naturalized citizens and another 26% were undocumented, while the remainder were legal permanent and temporary immigrants.54 In terms of education, immigrants are overrepresented at both the bottom and the top of the distribution; that is, they are more likely than U. S.-born individuals to have less than a high school degree and they are more likely than the U. S. born to have a graduate degree.55 This bimodal educational distribution is also reflected in wealth, income, and occupation: some immigrant groups have large numbers of highly paid professionals in the technology and health sectors, while others have large numbers of workers in low-paid, largely manual sectors such as agriculture and construction.56 These characteristics—regional origin, legal status, and socioeconomic resources— tend to cluster into what the sociologists Portes and Rivas have called “two streams” of immigration: one with legal status and abundant socioeconomic resources, largely from Asia (but also Europe, Oceania, and Africa), and another with greater numbers of undocumented migrants and far fewer socioeconomic resources, largely from Latin America and the Caribbean.57

Race/Ethnicity and Nativity | 145

Given the great heterogeneity among immigrants, it is remarkable that all immigrant groups have better health than their U. S.-born counterparts on key population health measures, as shown for infant mortality in figure 19. This pattern has also been documented in terms of adult mortality and some chronic conditions, including hypertension and some cancers.58 This is a remarkable pattern because regardless of their education level or other socioeconomic resources, immigrants face unique stressors. In leaving their hometowns and countries, immigrants leave behind the familiarity of culture, rules and institutions, language, friends, and family. In their new country, immigrants face the challenge of navigating a new culture and unfamiliar institutions, often in a language they did not grow up speaking. In other words, immigration is stressful. And the U. S. government does not make this process easy. Unlike in other immigrantreceiving nations, such as Canada, the U. S. government provides no support for immigrant integration and settlement, except for temporary assistance to the relatively small number of immigrants who arrive as refugees. In fact, the federal provisions regarding (nonrefugee) immigrant integration that do exist are prohibitive. Legal immigrants, those arriving with green cards to reside permanently in the United States, are barred for five years from receiving any federal benefits, such as Medicare or Medicaid, and undocumented immigrants are barred permanently from most federal and state benefits. It is not surprising then that immigrants are far less likely than the U. S. born to have health insurance coverage, although this too varies substantially by immigrant origin.59 Thus, it is all the more remarkable that on a variety of key health outcomes the expected pattern of negative health outcomes is not observed, particularly for immigrant groups with few social resources. This pattern has been referred to variously as the immigrant health advantage, the epidemiologic paradox, and the Hispanic health paradox.60 People are fascinated by the paradox in part because its answer may inform health policy interventions: if we can understand how immigrants beat the odds of socioeconomic disadvantage, perhaps we can apply the same formula to other socially disadvantaged groups. What gives rise to this paradoxical pattern? Explanations focus on data problems and processes related to migration that generate and promote a healthy immigrant population. The first explanation is that the paradox is a data artifact—the result of problems with mortality and health data specific to immigrants. This skeptical position is primarily concerned with the accurate counting of immigrant deaths in vital statistics data. If immigrants are counted in census or survey data (which

146

|

Chapter 6

provide the denominator in death rates) and later migrate back to their home countries where their death goes unrecorded in U. S. vital statistics (which provide the numerator in death rates), the death rate for that immigrant group will be underestimated, that is, lower than it actually is. The selective return migration of elderly or sickly (or less well off) individuals has been called the “salmon bias.” Although there is some evidence that the salmon bias exists, particularly among Mexican immigrants, studies find that it does not explain the paradox. For example, one study used Social Security Administration data that tracked Hispanic immigrants (and their deaths) in their home countries and found that even accounting for those deaths did not explain away the Hispanic mortality advantage relative to Whites in the United States.61 In another, the paradox was documented in terms of risk of mortality among infants as early as one day and one week after birth, periods too soon after a birth for return migration to conceivably bias the infant mortality rate.62 While some scholars argue that the salmon bias produces the paradox pattern as an artifact of data, the selection of healthy people into immigration—that is, in the opposite direction of the salmon bias—is an explanation for the paradox that suggests it is real. This explanation argues that immigrants are selected on good health, meaning that people who immigrate are healthier than those who do not.63 One reason for this is that many immigrants are motivated by higher earnings in the place of destination, and healthier people are in a better position to take advantage of those earnings than less healthy people. A second reason is that immigration is a difficult and costly process that the vast majority of people around the world never take on, even from places where there are seemingly good reasons to leave. In fact, only 3% of the world’s population are immigrants and only 13% of the U. S. population are immigrants. Given the rarity of this event and the difficulty of migrating, those who migrate are likely quite different from those who do not, and there is good evidence in support of this idea. For example, one study of immigrants to the United States from 32 of the largest immigrant-sending countries found that in 31 cases, immigrants were on average more highly educated than nonmigrants in their country of origin.64 There is also some evidence that immigrants are positively selected on health. One study found that immigrants to the United States from 18 countries had better self-rated health than nonmigrants in their home countries (and only one immigrant origin group—Mexicans—did not), and another study of U. S. legal permanent residents found that immigrants appraised their health as better than people in their home country.65 Other studies, using less sub-

Race/Ethnicity and Nativity | 147

jective measures of health, such as height, smoking, blood pressure, and blood hemoglobin (iron), find mixed support for immigrant health selection, depending on the group. For example, one study of U. S. immigrants from India, China, the Philippines, Mexico, and the Dominican Republic (DR) found that, with the exception of the DR, immigrants are taller on average than nonmigrants in their respective countries of origin, whereas another focused on Mexican immigrants found that only urban, female migrants were taller than their nonmigrant counterparts (but not rural men or women or urban men).66 Whether these selectivity patterns matter for outcomes in the country of destination is not clear, as few studies have tested the role of selectivity for generating the paradox. In one that did, country-of-origin differences in migrant selectivity did not differentiate immigrants from non-Hispanic Whites in the United States.67 That is, even immigrants from countries with “low health selection” reported better health than non-Hispanic Whites, suggesting that other factors matter for the paradox. The third and final explanation for the paradox is that immigrants are protected by or benefit from the process of immigration and / or social dynamics unique to immigration. The first of these is the co-ethnic community that immigrants join in the United States, which can provide social support and social control. Immigrants tend to cluster geographically in neighborhoods where there may be a greater degree of connection and mutual identity among immigrants, who share the common experience of immigration and ethnic origin, possibly providing a richer and denser social network than the U. S. born have access to. There is some empirical evidence in support of this idea. For example, studies find that Hispanics have lower mortality rates when they live in neighborhoods with more foreign-born or co-ethnic residents.68 And another study finds that for several large immigrant groups, immigrants experience a reduction in smoking after arrival in the United States, departing from both the expected smoking trajectory had they remained in their country of origin and from the “assimilation” trajectory—that is, had they adopted the smoking behavior of the host population.69 These “protective” effects seemed especially pronounced among groups with large co-ethnic networks. In addition to the co-ethnic community, it is possible that some immigrants benefit from the improvement in well-being they experience from immigrating. For instance, a study of newly admitted immigrants found that the average legal permanent resident incurred an annual income increase of $15,000 due to immigration, which may translate

148

|

Chapter 6

into better health.70 This change may be particularly beneficial when compared to their lives prior to migration and to their nonmigrant counterparts back home. Immigrants frequently maintain a foreign frame of reference, at least in the short term. That is, in appraising their well-being, they compare themselves not to the host society they have moved to but to the nonmigrant counterfactual: where they would have been if they had not migrated. If immigrants feel that their decision to migrate “paid off,” their health may benefit. However, there is a flip side to the paradox. Despite the fact that immigrants have better health outcomes than their U. S.-born counterparts, the immigrant health advantage may be lost over time. This suggests a disturbing process of downward health assimilation. Although some studies find short-term gains in the relative health standing of immigrants, over the long term and across generations, studies find a steady deterioration of the relative health standing of immigrants and their descendants.71 For example, studies show that child asthma and other common child health conditions and adult chronic disease, obesity, allostatic load (a measure of cumulative stress), disability rates, and mortality risk increase with duration of stay or across generations.72 This could reflect unhealthy acculturation, that is, the adoption of unhealthy behaviors as immigrants become more American in their attitudes, preferences, and behaviors. Indeed, there is substantial evidence that alongside changes in health are changes in health behaviors such as smoking, alcohol use, and unhealthy eating.73 However, as we have argued in this chapter, behavioral changes alone likely do not account for group-level changes in health over time. Rather, our conceptual model of racial and ethnic disparities in health argues that health behaviors are structured by social resources and racism: inequality is what puts some individuals at greater risk of risky behaviors. So how do we explain the paradox, and why does it unravel over time? We suggest that the relatively good, short-term health of immigrants reflects all three explanations described above. Some of the observed health advantage arises from data artifact. What remains reflects the positive health selectivity of immigrants and, in the short term, the protective benefits of the co-ethnic community and of successful immigration. However, over time, the cumulative impact of stress, difficult working conditions, few mobility routes for the low-skilled, xenophobia against immigrants and racism against people of color, limited rights and the threat of deportation for the undocumented, and a changing point of reference away from the home country and toward

Race/Ethnicity and Nativity | 149

Box 8: DACA and the Mental Health of Young Immigrants The Deferred Action for Childhood Arrivals (DACA) program was created in 2012 to grant a subset of undocumented immigrant youth who arrived to the United States as children temporary reprieve from deportation, work authorization, and other benefits. By May 2018, 819,337 initial DACA applications had been approved by U. S. Citizenship and Immigration Services, representing 74% of the 1.1 million people estimated to be eligible for the program.a In September 2017, President Trump announced plans for DACA’s termination, a decision currently challenged in the courts. Many immigration activists hailed the DACA program as a major expansion of immigrant rights and an improvement to the lives of hundreds of thousands of young people in the United States. And many research studies in the two to three years following the introduction of the program found that, indeed, DACA improved the wellbeing of participants. For instance, studies found that participants’ high school graduation rates increased, labor force participation increased, wages increased, mental health improved, and feelings of safety, hope, and belonging in this country were enhanced.b Even the children of DACA-mented parents appeared to benefit.c The research findings related to DACA are consistent with theories that view immigration policy as a key determinant of immigrant integration and immigrant legal status as a fundamental cause of health.d These theories argue that undocumented immigration status hinders the integration of immigrants and undermines their health by denying rights, placing immigrants at risk of detention and deportation, excluding immigrants from social services and institutions, provoking discrimination and stigma, and generating substantial uncertainty in the lives of immigrants. However, DACA does not provide permanent legal status. DACA status is subject to biannual renewal, and the program is subject to presidential discretion. DACA can be conceived of as “liminal legality”—a state of permanent legal “in betweenness.”e Insofar as it is an improvement over illegality, liminal legality may be empowering in the short term, but experienced over longer durations liminal legality has been shown to be dysfunctional in multiple ways because its inherent uncertainty creates stress and undermines the ability to plan and make longer-term investments in the future. Recent research extending the analysis of DACA’s impacts into the end of the Obama presidency suggests that some of the beneficial impacts of DACA may have been short-lived, particularly as they relate to mental health.f Analysis of data from the California Health Interview Survey shows that DACA had a short-term, positive impact on

150

|

Chapter 6

6

5

4

3

2

1 2007

2009

2011

All documented immigrants

2012

2013

2014

2015

2016

DACA-eligible immigrants

Predicted K6 score, by year and DACA eligibility. Source: Patler, Hamilton, and Savinar 2018

mental health, as measured by a decline in the Kessler-6 (K6) Mental Distress scale of DACA-eligible immigrants between 2012 and 2014 (see figure). In this three-year period, DACA-eligible immigrants had significantly better mental health than documented immigrants in California. However, by 2015 and especially 2016, the mean K6 score for DACA-eligible immigrants in California had returned to its pre-policy average, and the difference between the two groups disappeared. The uncertainty of the future of the DACA program, particularly with the election of a president with staunch anti-immigration views, undermined the positive mental health effects of the DACA program. What the future holds for the health of undocumented immigrants in this country—with the election of Trump, the announcement of the termination of DACA, and continued political and judicial wrangling over the future of undocumented youth in this country—is unknown. a. U. S. Citizenship and Immigration Services 2018; Krogstad and GonzalezBarrera 2014. b. Abrego 2018; Amuedo-Dorantes and Antman 2017; Kuka et al. 2018; Patler and Pirtle 2018; Pope 2016; Venkataramani et al. 2017. c. Hainmueller et al. 2017. d. Asad and Clair 2018; Bean, Brown, and Bachmeier 2015; Menjívar 2006; Portes and Zhou 1993. e. Menjívar 2006. f. Patler, Hamilton, and Savinar 2018.

Race/Ethnicity and Nativity | 151

the host country—these various disadvantages begin to take their toll. That is, group health over time reflects how immigrants are incorporated into U. S. society. For relatively advantaged groups—the stream of high SES immigrants largely from Asia that Portes and Rivas referred to—this means sustained good health. For relatively disadvantaged groups— the second stream largely from Latin America and the Caribbean—this means the loss of a health advantage and the emergence of relative disadvantage over time. Changes in the health of these groups thus reflect how immigrants are incorporated into the U. S. system of racial stratification.

black / white mental health paradox A second paradoxical set of research findings can be seen in table 9: Black Americans have the lowest rate of mortality from suicide of all major racial / ethnic groups. The rate of suicide mortality for Blacks is 65% lower than the rate of mortality from suicide for Whites, 9.5% lower than the rate for Hispanics, and 5% lower than the rate for Asians and Pacific Islanders. Black Americans also have lower lifetime risk for depression and other psychiatric disorders, including anxiety, social phobia, substance use disorders, and panic disorders, than other groups.74 For example, in three national data sets collected between 2001 and 2003, it was found that 6.8% of African Americans experienced a past-year major depression and 12.3% experienced a lifetime major depression, compared to 8.3% and 20.4% of Whites and 8% and 14.5% of Mexican Americans.75 Because social disadvantage, discrimination, and stress are associated with mental health problems just as they are with physical health problems, scholars have referred to this set of findings as the “Black-White mental health paradox.”76 Explanations for the Black-White mental health paradox have focused on data artifact and protective mechanisms. One concern regarding data is measurement error. Depression and other psychiatric disorders are measured in surveys by scales that combine responses to questions regarding the frequency, duration, and / or severity of psychiatric symptoms such as sadness, worry, loss of appetite, and disruptions to sleep. If there are systematic differences across racial / ethnic groups in how individuals respond to psychiatric scale questions, then the scales may not be valid; that is, they may incorrectly measure true levels of depression or other psychiatric conditions. While there is some

152

|

Chapter 6

evidence that some scale items are responded to differently by African Americans and Whites, research finds that these differences do not account for racial disparities in the overall scale.77 Indeed, the scales used in most national surveys have been extensively tested and validated, including for different racial and ethnic groups.78 A second data concern is the underrepresentation of African Americans in national surveys of the noninstitutionalized population as a result of the high levels of incarceration of Black men. Given the association between crime and mental health issues, survey data that exclude incarcerated populations may underestimate the extent of mental health problems among African Americans. However, the mental health paradox is revealed not just in survey data, but also in suicide data, as described above, which are not affected by this source of bias. The third data concern is that mental health stigma and racism may prevent African Americans from identifying and seeking care for mental health problems, resulting in underreporting and underdiagnosis. In one focus group study with African Americans in Baltimore, participants described fears of being labeled as weak, dangerous, hopeless, and sinful, of being written off by family and friends as worthless, and of losing status in their community if they were diagnosed with a mental illness.79 A distinct fear the participants expressed was that mental health providers would not treat them with respect, would not ensure their confidentiality, and would force them into a damaging label or treatment program against their will. Combining these two concerns, participants worried that seeking care for a mental health problem would confirm preexisting biases that African Americans are weak or dangerous. It is not clear the extent to which these concerns affect actual rates of mental health problems in the African American community. Two explanations for the Black-White mental health paradox emphasize potential protective mechanisms within the African American community. The first considers the mental health benefits of religious involvement. African Americans are more likely than Whites to belong to a church, attend church services frequently, engage in other church activities, read religious materials, watch or listen to religious programs on television or radio, and to pray or request prayer from others.80 And multiple studies document that religious participation, particularly service attendance and a sense of purpose derived from religious involvement, is associated with lower levels of depression and anxiety.81 The second explanation emphasizes a potentially salubrious interaction among stress, mental health, and unhealthy behaviors such as eating

Race/Ethnicity and Nativity | 153

tasty but fatty, low-fiber foods (“comfort food”), smoking, and drinking, an idea referred to as the Environmental Affordances Model.82 This model argues that the uniquely disadvantaged environments in which many low-income Blacks live offer certain “affordances,” specifically the ability to engage in unhealthy but stress-coping behaviors with greater ease given the material and cultural setting. Eating comfort food, smoking, and drinking each change the body’s physiological response to stress: eating comfort foods inhibits the release of anxiety-causing hormones, consuming alcohol increases the release of pleasure-causing hormones, and consuming tobacco produces a calming reaction to certain stress hormones.83 The Environmental Affordances Model suggests that these unhealthy behaviors have opposite effects on mental and physical health in the context of severe deprivation: while eating high-fat, high-salt foods, smoking, and drinking increase the risk of physical health problems, they may serve as effective short-term coping techniques for dealing with stress caused by severe adversity, thereby minimizing the impact of stress on the development of psychiatric disorders such as major depression. This implies that the mental health advantage and the physical health disadvantage that Blacks experience are related. Insofar as individuals facing extreme adversity use unhealthy behaviors to effectively cope with stress, reducing the risk of serious mental health problems, the long-term consequence is borne out silently in the deterioration of physical health. The Environmental Affordances Model makes two central research predictions. One is that there is an interaction between the severe deprivation experienced by some low-income Blacks and unhealthy behaviors. The second prediction is that the interaction between disadvantage and unhealthy behaviors should account for racial disparities in mental health. Research testing these two hypotheses is limited, reflecting the relatively recent development of the Environmental Affordances Model. However, there are a number of empirical patterns that are consistent with the model. Research suggests that the Black mental health advantage relative to Whites is greater among adults with low levels of education, consistent with the idea that the mental health protections of coping behaviors may be especially pronounced among disadvantaged Blacks.84 Evidence regarding the extent and type of mental health problems that Blacks face also provides some clues. The Black mental health advantage appears to be driven by a related set of “internalizing” conditions, which may share a similar etiology, possibly related to the kinds of protections described above. And research finds that while the prevalence of depression and

154

|

Chapter 6

anxiety disorders is lower among African Americans than other groups, African Americans do not experience less distress overall, and they experience a higher prevalence of dysthymic disorder, or low-grade, perpetual depression.85 On the other hand, among adults with major depressive disorder, African Americans report more severe, chronic, and disabling forms.86 The fact that Blacks experience greater low-grade distress and worse depression, but less prevalence of depression, is consistent with the idea that unhealthy coping behaviors provide a minimal form of protection from mental health problems.

rising rates of mortality among middle-aged americans The third topic has not been termed a paradox, but the fact that White midlife mortality rose in recent decades has been treated as such. As you will recall from chapter 1, the Case-Deaton article reporting this finding received far more attention than is common among published works of population health science. The central finding—that all-cause mortality rose among Whites between ages 45 and 54 after 1998—was remarkable because increases in mortality are virtually unprecedented in the post-epidemiologic transition period. This is particularly the case for U. S. Whites, who, because of their relative advantage and power, tend to be buffered from changes to the disease environment (such as the emergence of HIV) or the social environment (such as the Great Recession) that might increase mortality among more vulnerable groups. The increase was driven entirely by increases in mortality among Whites with a high school degree or less—that is, a similar increase was not observed among better educated whites—and the trend was most pronounced for women.87 A substantial part of the increase in all-cause mortality was due to increases in deaths from external causes—primarily suicides and drug and alcohol poisoning.88 But declines in mortality from nonexternal causes also stalled during this period; indeed, all-cause mortality for middle-aged Whites would have declined had it not been for stalling declines in mortality from nonexternal causes such as heart disease.89 Following the publication of the Case and Deaton study, other researchers have clarified the trend. Recently, Steven Woolf and colleagues analyzed midlife (ages 25–64) mortality trends from 1999–2016 for Whites in comparison to the four other major racial / ethnic groups in the United States by cause of death.90 The Woolf study showed that American Indians and Alaska Natives experienced rising mortality over

Race/Ethnicity and Nativity | 155

the period that frequently outpaced trends for Whites, including for allcause mortality and mortality from alcohol liver disease. The study also showed that although midlife mortality had been declining or stable for Blacks and Hispanics prior to 2012, after 2012 mortality began to rise for a variety of causes of death, including drug overdoses, alcoholrelated deaths, and mental and behavioral disorders. Several explanations for this trend have been suggested, but to date there is no direct evidence in support of one or the other. However, a distinct component of the problem is the opioid epidemic. One study found that the clearest cause of rising rates of mortality among middleaged Whites was drug-related mortality.91 Indeed, the rate of deaths from drug overdoses increased by 188.6% among White women between 1999 and 2010 but only by 55.3% among Black women, 50% among Asian women, and 68% among Hispanic women over the same period, resulting in the flipping of the racial disparity for this cause of death between 1999 and 2013.92 By 2010, the death rate from drug overdoses was higher among White women than all other groups except American Indian / Alaska Native women.93 Opioid painkillers are more frequently prescribed to White patients, particularly for conditions associated with drug-seeking behavior, such as migraine and back pain, possibly reflecting an implicit bias associating Black patients with drug seeking.94 In other words, an unintended consequence of racial discrimination in opioid prescribing may have been to raise the relative risk of drug overdose mortality among Whites. However, prescribing differences are not the whole story, as Whites are also more likely than other race / ethnic groups to abuse painkillers.95 Other causes of death matter, as well, and different proximate mechanisms will matter for different causes of death.96 However, the clustering of poisoning deaths with suicides and liver disease, alongside various changes in the morbidity of middle-aged Whites over the same period, led Case and Deaton to conclude that the rising rate of mortality reflects the despair of decades of economic stagnation—of a “lost generation” becoming “the first to find, in midlife, that they will not be better off than were their parents.”97 Studies show that socioeconomic mobility in the United States has declined dramatically over the course of the twentieth century and beginning of the twenty-first. Among Americans born in 1940, 92% eventually earned more than their parents did, but this declined dramatically across cohorts—to 60% among those born in 1961 and 50% among those born in 1984.98 As lesseducated Baby Boomers faced the economic realities of middle age at

156

|

Chapter 6

the end of the twentieth and beginning of the twenty-first centuries, this despair has been played out in physical form through diverse proximate determinants: drug abuse, alcohol abuse, and mental health problems. Racial / ethnic and nativity disparities remain one of the most important topics in the study of U. / S. population health. This chapter has outlined a fundamental cause of health model through which current disparities may be understood. This model reflects the central importance of racism, both historically and in the contemporary United States, in structuring racial / ethnic health disparities. While African Americans and Native Americans have exhibited much improved population health relative to Whites in recent decades, major disparities remain. Moreover, given continued group disparities in access to high-quality education, jobs, income, wealth, safe and high-quality housing, encounters with the criminal justice system, and more, we unfortunately expect to find continued disparities between these groups for decades into the future unless very aggressive changes in U. S. social and health policies are made.99 Although they are both heterogeneous groups, Hispanics and Asian Americans on the whole exhibit quite favorable patterns of population health, and particularly so among the immigrant generations of these groups. However, U. S.-born Hispanics in particular also exhibit less favorable indicators of population health than the immigrant generation, most likely reflecting the process of incorporation into the U. S. systems of racial and socioeconomic stratification. One of the most important roles for population health scientists in the coming decades will be to monitor the health patterns of the rapidly growing Hispanic population in the United States, among whom the pattern of highly favorable health at the time of immigration seems to wear away across time and generations. Given the stresses of the immigration and incorporation processes that some subgroups of Hispanic are experiencing, the long-term positive population health patterns of this group are not at all assured.100

chapter 7

Gender and U. S. Population Health

As chapter 3 made clear, the population health of American women has been particularly worrisome over the past three decades. Although both men’s and women’s relative international health ranking declined, women’s ranking declined more rapidly and to a greater extent than did men’s, such that in recent years, American women’s life expectancy is the lowest among women in 22 high-income countries.1 Trends are especially concerning for women of low socioeconomic status. From 1986 to 2006, mortality rates declined for women with a college education, but they actually increased for women with less than a high school degree.2 A similar trend was observed for older men between the ages of 65–84, but in fact the steepest increases in mortality were observed for women between the ages of 45–54. Indeed, it was this age group that was identified by the now well-cited Case and Deaton paper showing rising mortality for middle-aged Whites. Moreover, in subsequent analyses, it became clear that it was women among this age group for whom the problematic midlife health and mortality trends were most pronounced.3 Why would recent population health trends be worse for women than for men? Why are there gender differences in health to begin with? These questions animate a large body of population health research that attempts to document and explain health inequalities between men and women. In the case of gender, the research is especially complex because of the close conceptual link between the biological underpinnings of 157

158

|

Chapter 7

gender—what we call “sex”—and the social and identity construct that we define as “gender.” Therefore, to begin, we define these terms.

what is gender? Sociologists use the term “sex” to refer to biological differences between men and women and “gender” to refer to the meanings and social arrangements that create different experiences for men and women. Sex thus refers to nature, and gender to nurture. Sex to chromosomes, genes, hormones, and reproductive anatomy, and gender to socialization, expectations, identities, and structures of power. Sex to differences that would be observed regardless of social configuration, gender to the impact of the social configuration itself. In reality, these differences are not so neat or distinct. As we will see in this chapter, gender structures the body, translating the cultural and the social into the physical. As with socioeconomic and racial / ethnic disparities in population health, disparities in health between men and women, and girls and boys, reveal social inequalities. However, in this case, the fact that gender is so closely linked to sex—the complex biological systems that are used to define male and female bodies—gives the appearance that health and mortality differences between men and women are the result of biology—of sex—rather than of gender. In reality, as the biologist Anne Fausto-Sterling observed, “there are very few absolute sex differences, and without complete social equality we cannot know for sure what they are.”4 While we will return to this conceptual discussion throughout the chapter to understand what it means for gender differences in health and mortality, for now we simply clarify that we use the term “gender” to refer to observed differences in the health and mortality patterns of boys and girls and men and women, as none of the differences we observe exist outside of social understandings of gender. An individual’s sex and gender commonly correspond, but in cases where they do not, we give priority to gender—that is, to the social identity (woman or man or something different) that one claims for oneself. Given that the population of transgender people is relatively small, and most population data generally do not (yet) identify transgender respondents or people who identify outside of the gender binary, our discussion of gender differences in population health necessarily refers to and relies on the gender binary (male / female; man / woman). We return to the issue of transgender identities and transgender health again toward the end of the chapter.

Gender | 159

9 8

80 Life expectancy at birth

7 75 6 70

5

65

4 3

60 2 55

1

50

Female-male gap in life expectancy

85

Female

2012

2009

2006

2003

2000

1994

Male

1997

1991

1985

1988

1979

1982

1976

1970

1960

1950

1940

0

Gap

figure 22. Female and male life expectancy at birth, and the female–male gap in life expectancy, 1940–2013. Note: Decadal data presented (in dots) prior to 1975; yearly data (in lines) presented after 1975. Source: Xu et al. 2016

patterns and trends in women’s and men’s mortality and morbidity Currently, women live longer than men do in most places. In 2014, women around the world outlived men by an average of 4 years.5 But the gender gap in life expectancy varies substantially across regions and countries, with women in Syria and several former Soviet states, including Russia and Belarus, outliving men by 10 or more years, and women in many sub-Saharan African countries outliving men by far less, 1 to 3 years. Swaziland is the only country in the world where men’s life expectancy is currently longer than women’s.6 If we were to rank countries in terms of the gender gap in mortality, the United States would fall squarely in the middle of the list. Given current death rates, American women can expect to live on average 81.2 years, whereas American men can expect to live 76.4 years, a 4.8year gap (figure 22). Although this gap implies a substantial mortality advantage for women, it is the smallest gender mortality gap observed

160

|

Chapter 7

3 2.5 2 500 1.5

Rate ratio

Death rate (logged)

5000

1

50

0.5 5

Male

≥85

75 to 84

65 to 74

55 to 64

45 to 54

35 to 44

25 to 34

15 to 19

Female

20 to 24

10 to 14

1 to 4

5 to 9

0

Rate ratio

figure 23. Male and female age-specific death rates, and the gender mortality rate ratio, by age, 2012. Note: Death rates are graphed on the logarithmic scale; the dotted line marks a rate ratio of 1, or an equal age-specific death rate for men and women. Source: Heron 2015

in the United States in about 75 years. As shown in figure 22, life expectancy in the United States grew by 16 years for both women and men from 1940 to 2012. However, life expectancy grew faster for women in the first half of the period, resulting in a peak gender gap of 7.8 years in the late 1970s, and then faster for men in the second half, reducing the gap to 4.8 years by 2012. Preston and Wang’s detailed analysis of gender-specific death rates in the second half of the twentieth century shows that these changes were structured on a cohort basis rather than a period basis.7 Specifically, the gender mortality gap grew among cohorts born in the late 1800s up until the birth cohort of 1904–08, and they narrowed among cohorts born after 1908. Contemporary gender differences in the risk of mortality across the life course are graphed in figure 23, using age-specific death rates for U. S. men and women from 2012.8 The ratio between men’s and women’s age-specific death rates is also graphed in order to reveal how the gender difference in mortality changes across the life course; ratios greater than 1 (i.e., above the dashed line) indicate that men exhibit a higher rate of death than women at that age. The figure shows that both men’s and women’s age-specific mortality follows the standard J-curve,

Gender | 161

with the rate of death declining from infancy to a low around age 10 and rising thereafter to its highest level in the oldest ages. At all ages men have a higher rate of death than women, but this difference is widest in early adulthood. Between the ages of 20–24, men are 2.8 times more likely to die than women, whereas at the youngest and oldest age groups, gender differences decline to near equality, although there is a female advantage even among infants. In 2013, the infant mortality rate was 5.38 deaths per 1,000 live births among female infants and 6.52 deaths per 1,000 live births among male infants.9 Men’s higher mortality reflects an increased rate of death from nearly all causes of death. In table 10, age-adjusted, cause-specific death rates (per 100,000 population) are listed for men and women in all age groups; in the second column, we limit to the cause-specific comparisons to 20–24 year-olds because this is the age group for which the largest gender mortality rate ratio is observed. We report death rates for the top 10 leading causes of death for men and women of all ages, as well as causes that rank in the top 10 for men and / or women from ages 20–24, which adds homicide, congenital malformations, legal intervention, HIV, pregnancy and childbirth, and septicemia. For all ages and for the age 20–24 column, the gender cause-of-death rate ratio is listed as well, again with rate ratios above 1 indicating a higher rate of death for men. Focusing first on the entire population (i.e., men and women of all ages), men exhibit an elevated rate of death from all major causes except for Alzheimer’s disease, for which the male death rate is about 30% lower than the female death rate, and cerebrovascular disease (stroke), for which the death rates are equal. The male death rate is at least two times higher than the female death rate for accidents, suicide, homicide, and HIV, and between 10 and 80% higher than the female death rate for most other causes, including the most common two causes of death, heart disease and cancer. Relative gender differences in mortality are larger for 20–24 year-olds for nearly all causes. For example, while the gender mortality rate ratio is 1.6 for heart disease among all age groups, implying a 60% higher rate for men, it is 1.8 (80% higher for men) among 20–24 year-old young adults. The gender mortality rate ratio among these young adults is especially large for suicide and homicide. Men between the ages of 20–24 are 4.6 times more likely to die from suicide and 6.6 times more likely to die from homicide than similarly aged women. Legal intervention—death by law enforcement actions, including legal executions—is the eighthleading cause of death for 20–24 year-old men, killing .5 men per every

162

|

Chapter 7

table 10 death rates (per 100,000) from major causes of death for u. s. men and women and the gender rate ratio, all ages and ages 20–24, 2014 All Ages*

Diseases of heart Malignant neoplasms Chronic lower respiratory diseases Cerebrovascular diseases Accidents Alzheimer’s disease Diabetes mellitus Influenza and pneumonia Kidney disease Suicide Homicide Congenital malformations and chromosomal abnormalities Legal intervention HIV Pregnancy and childbirth Septicemia All causes

20–24

Women

Men

Rate ratio

Women

Men

Rate ratio

131.8 138.1 37.1

210.9 192.9 45.4

1.6 1.4 1.2

2.0 3.2 0.4

3.6 5.1 0.7

1.8 1.6 1.8

35.6 27.3 28.3 17.2 13.2 11.1 5.8 2.1 3.0

36.9 54.7 20.6 25.6 17.8 16.2 20.7 8.0 3.2

1.0 2.0 0.7 1.5 1.3 1.5 3.6 3.8 1.1

0.4 17.8 — 0.5 0.4 0.2 5.0 3.1 0.7

0.6 52.0 — 0.8 0.7 0.3 22.9 20.4 1.0

1.5 2.9 — 1.6 1.8 1.5 4.6 6.6 1.4

0.3 1.1 0.8 9.9 616.7

0.0 3.0 — 11.8 855.1

— 2.7 — 1.2 1.4

— 0.3 1.2 0.4 44.0

0.5 0.6 — 0.4 121.7

— 2.0 — 1.0 2.8

notes: “—” indicates there were 20 or fewer deaths by that cause for that age / gender group. * Rates for all ages are age-adjusted. sources: Centers for Disease Control and Prevention (CDC) 2018a

100,000 among those in the age 20–24 range (for a total of 58 deaths of 20–24 year-old men in 2014). There were fewer than 20 deaths from this cause to women in this age group, resulting in the CDC reporting no rate. Just over one woman for every 100,000 women ages 20–24 dies from deaths related to pregnancy and childbirth (for a total of 132 deaths of 20–24 year-old women in 2014), whereas of course no men died from these causes.10 The fact that U. S. women’s mortality is lower than men’s in the 20–24 age group and throughout the life course does not mean that women necessarily experience less illness and disease. Using data from the 2011– 12 National Survey of Children’s Health and the 2012–14 National Health Interview Survey, table 11 reports rates of common health measures among U. S. children and adults by gender. Among children, girls

Gender | 163 table 11 percentages of common disease and health conditions by gender Girls / Women

Boys / Men

Among children One or more chronic condition, ages >18 Current ADD or ADHD, ages 2–17 Current asthma, ages >18 Overweight or obese, ages 10–17

19.7 4.9 8.0 27.8

27.3 10.7 9.6 34.7

Among adults ages 18+ Heart disease Cancer Arthritis Diabetes Lung disease Asthma Psychological distress Fair or poor health

9.7 8.4 23.6 8.1 5.2 9.3 3.8 28.3

12.1 7.6 17.9 9.0 3.8 5.4 2.7 11.5

Among adults ages 65+ Any physical functioning limitation Any personal care needs Any routine care needs

71.9 7.6 14.1

58.6 5.4 9.1

note: Adult percentages are age-adjusted; all are weighted to be representative of the U. S. population. sources: National Survey of Children’s Health 2011–2012 (Child and Adolescent Health Measurement Initiative [n.d.]); National Health Interview Survey 2012–2014 (CDC / National Center for Health Statistics 2017a)

have lower rates of chronic conditions, attention deficit disorder and attention deficit hyperactivity disorder (ADD / ADHD), asthma, and overweight / obesity. The gender difference in ADD / ADHD is particularly striking, with boys two times more likely than girls to be diagnosed. Among adults, women have higher rates of most health conditions and age-related limitations. Out of the 11 health concerns listed, women are more likely to report nine, including cancer, arthritis, lung disease, asthma, psychological distress, fair or poor self-rated health, physical functioning limitations, need for personal care assistance, and need for routine care assistance. Some of these differences are rather pronounced; for example, women are nearly three times more likely than men to rate their health as fair or poor. Men, by contrast, have slightly elevated rates of heart disease and diabetes compared to women. The combination of women’s generally higher morbidity and lower mortality means that women live far more years with health problems than men do, including

164

|

table 12

Chapter 7 rates and percentages of health care usage, by gender Girls / Women

Among children ages >18 (%)a Has a usual source of care Saw a doctor for preventive care at least once in past year Among adults ages > = 18 Physician visits (per 100,000)b, c ER visits (per 100,000)c Flu vaccine (%)d Pneumococcal vaccine (%)d Appendectomy (per 100,000)e Heart surgery (per 100,000)e Total hip replacement (per 100,000)e Endoscopy of small intestine (per 100,000)e

Boys / Men

91.4 85.2

91.3 83.7

46129.5 4860.7 42.4 20.7 10.2 2396.3 13.4 46.4

31555.5 3960.3 35.4 19.1 10.3 1781.4 13.3 44.4

sources a. National Survey of Children’s Health 2011–2012 (Child and Adolescent Health Measurement Initiative [n.d.]) b. National Ambulatory Medical Care Survey (NAMCS) (CDC / National Center for Health Statistics 2017b) c. National Hospital Ambulatory Medical Care Survey (NHAMCS) 2008–2010 (CDC / National Center for Health Statistics 2017c) d. National Health Interview Survey 2012–2014 (CDC / National Center for Health Statistics 2017a) e. National Hospital Discharge Survey (NHDS) 2010 (CDC / National Center for Health Statistics 2011)

with conditions such as cardiovascular disease that have historically been considered “men’s diseases.”11 The phenomenon of lower death rates but less healthy lives for women compared with men has been referred to by some researchers as the “gender paradox.”12 Although not shown in these data, some of the most substantial gender differences in health are observed in the expression of psychological distress. Whereas U. S. women are more than two times as likely to be depressed than men, men are more than two times as likely to abuse drugs and alcohol.13 But there are no gender differences in mental health when depression, anxiety, addiction, and antisocial behavior disorders are combined, suggesting that gender differences may arise in how men and women express distress rather than in overall mental health itself.14 Men and women also vary in their use of health care, and this could reflect both gendered differences in need for care and in practices related to the use of care. In general, women use more health care than men, as revealed in table 12. Whereas boys and girls have similar access to

Gender | 165 table 13 age-adjusted percentages of health behaviors and risk factors for disease among adult women and men in recent years

High cholesterol (%a, ages 20+) High cholesterol (%, ages 65+) Hypertension (%a, ages 20+) Hypertension (%a, ages 65+) Overweight or obese (%, ages 20+)a Obese (%, ages 20+)a Physical activity (%, ages 18+)b Current smoker (%, ages 18+)b Former smoker (%, ages 18+)b

Women

Men

27.9 55.6 13.0 37.3 64.5 35.8 17.3 15.5 17.9

27.8 50.9 15.6 29.2 72.3 34.2 25.0 19.9 24.9

sources a. National Health and Nutrition Examination Survey (NHANES) 2009–2012 (CDC / National Center for Health Statistics 2015) b. National Health Interview Survey 2012–2014 (CDC / National Center for Health Statistics 2017a)

medical care, as measured by whether the child has a usual source of care, girls are slightly more likely to have seen a doctor in the past year for preventive care (85.2% vs. 83.7%). These differences are much more pronounced in adulthood. Women’s level of physician visits is about 50% higher than men’s, and their use of the emergency room is 22% higher. They are significantly more likely to get the flu vaccine and to have heart surgery. However, rates of pneumonia vaccine, appendectomies, hip replacement, and endoscopies of the small intestine do not vary between men and women. With the important exceptions of smoking and the abuse of drugs and alcohol, most health risks and behaviors tend to disadvantage women, as shown in table 13. At older ages, women are more likely than men to have high cholesterol and hypertension. They are also slightly more likely than men to be obese and less likely to engage in physical activity. Men, however, are more likely than women to be current or former smokers. Given smoking’s death toll, this gender difference is responsible for a large share of the gender difference in mortality, and has been since the early 1900s, although long-term changes in men’s and women’s smoking patterns have reduced the relative contribution of smoking to the gender gap in mortality in recent decades.15 We return to this issue in more detail below.

166

|

Chapter 7

explanations for gender differences What causes gender differences in population health? Scholars identify four general categories of determinants of gender differences in health and mortality: biological, social-structural, psychosocial, and behavioral.16 Biological differences (i.e., sex differences) refer to differences between men and women in their genetic makeup, physiologies, anatomies, and hormones, which result in different health risk profiles, as measured by inflammation, immune function, fat distribution, glucose control, bone density, blood pressure, and cholesterol, among other things. Social-structural differences refer to differences between men and women in their social status and position in society and more specifically to men’s greater social status and power, occupational prestige and rewards, and control of women. Psychosocial differences include men’s and women’s stress and coping mechanisms, including differences in social support. Behavioral differences refer to men’s and women’s health-related actions, including exercise and diet, use of substances, and use of medical care. We discuss the evidence for and against each of these explanations by focusing on two central topics of research on gender disparities in health and mortality— first, the gender paradox of women’s longer life but worse health, and, second, changes in the U. S. gender gap in mortality over the past century. The “Paradox” of Women’s Longer Life But Worse Health The fact that women live longer lives but experience greater health problems than men has been called a paradox, as generally we expect different measures of population health (e.g., morbidity, disability, and mortality) to coincide with one another. Part of the story involves “external” causes of death—accidents, homicide, and suicide—and the fact that men are more likely to die from these causes of death than women. But that is not the whole story, as men’s mortality disadvantage is observed for almost all causes of death. In fact, the paradox is entirely accounted for by the different prevalence and severity of the health conditions that men and women face.17 Table 11 shows, for example, that men have higher rates of heart disease, the leading cause of death in the United States. Whereas women suffer to a greater extent from depression and anxiety, men are more likely to abuse substances or commit violence against themselves or others, with greater mortality consequences.18 Thus, women suffer disproportionately from short-term and chronic but nonfatal diseases that

Gender | 167

cause significant disability and use of health care but less commonly result in mortality, whereas men suffer to a greater extent from life-threatening conditions.19 At the same time, among men and women with the same conditions, men tend to experience more severe forms, particularly of lung and heart disease.20 In other words, the fact that women have lower mortality but higher morbidity than men is not a paradox at all but an outcome of their different disease profiles. What is not well understood is why women and men experience different disease profiles. A common explanation is that women benefit from biological advantages that protect against some key fatal illnesses. At least two lines of evidence support this conclusion. First, the social determinants of health—social status, psychosocial supports, and behaviors—appear to counterbalance each other when it comes to mortality.21 Men are more likely to smoke and abuse alcohol and are less likely to seek preventive health care, but women have lower socioeconomic status and are less likely to exercise. As a result, gender differences in socioeconomic status and health behavior do not explain the gender gap in mortality22 or in life-threatening conditions.23 A second line of evidence involves biological differences between men and women. For instance, the weaker link formed in the male XY chromosomal pair than in the female XX pair is implicated in the higher miscarriage rate for male fetuses.24 Hormonal differences may also play a role, as testosterone is a risk factor for heart disease25 and may be an immunosuppressant, thus providing men with a biological disadvantage.26 The systems that allow a woman to carry fetuses to term and breastfeed infants give certain biological advantages, including a flexible cardiovascular system, which allows a woman’s blood volume to increase during pregnancy and may also protect against high blood pressure. Moreover, women may also have a stronger and different immune system than men, which provides passive immunity to infants during pregnancy and through breastmilk and may also protect women from certain chronic conditions.27 Although these biological differences suggest an advantage for women, there are few studies that test for the role of biology in gender differences in morbidity or mortality using population data,28 and those that have done so have not found strong support for the biological argument.29 Richard Rogers and colleagues found that U. S. women have higher average levels of inflammation and cardiovascular risk than men, and that accounting for these differences actually widens the gender mortality

168

|

Chapter 7

gap.30 In other words, their study found that if men and women had similar levels of inflammation and cardiovascular risk, women would have even lower mortality rates than men. A problem with testing the biological explanation for gender differences in mortality is that research methods are only now beginning to disentangle interactions between biology and social forces.31 For example, testosterone increases as a result of engaging in competitive sports,32 meaning that differences in testosterone levels between men and women partially reflect social forces. While premenopausal estrogen protects women from heart disease, smoking may lead to early menopause, reducing this biological protection.33 While there is some evidence to suggest that biology may play a role in explaining men’s higher rate of mortality, there is less evidence that biology explains the greater extent of nonfatal morbidity among women.34 Rather, women’s unfavorable disease profile likely emerges as a result of their social disadvantages relative to men.35 In this sense, gender differences in health may be another reflection of how fundamental social causes—money, knowledge, prestige, power, and beneficial social connections—work to systematically disadvantage a social group (women) across a wide variety of population health measures.36 While women do indeed outlive men in most countries of the world, women’s lower socioeconomic standing with regard to power, prestige, income, and wealth, in particular, may serve to harm their population health relative to men across a wide variety of outcomes. A number of research studies support this conclusion. For instance, one study found that women have higher morbidity and mortality in U. S. states where they have less economic and political autonomy,37 and another found that increases in women’s education and labor force participation have narrowed the gender gap in self-rated health.38 Studies also find that women’s social structural disadvantage—including their relatively lower levels of education (at least historically), occupational prestige, and political power—has psychosocial and behavioral consequences, which in turn affect health.39 Burdens that inordinately fall on women, such as caretaking and housework, also contribute to women’s nonfatal morbidity.40 Caretaking, particularly of the elderly and / or ill, is associated with psychological distress among women,41 and inequity in housework contributes to the gender gap in depression.42 These research studies suggest that socioeconomic equality between men and women would improve women’s health relative to men. Thus, we can draw several conclusions about what underlies gender differences in health from research on the so-called “paradox” of women’s

Gender | 169

Box 9: Women’s Reproductive Health Population health scientists have contributed substantial knowledge to an area of intense political debate—women’s reproductive health and specifically abortion. One finding emerging from this research is that contraceptive use and abortion tend to be substitutes. That is to say, when women and men have access to safe, effective, and affordable contraception, they are able to more effectively avoid abortion. For example, researchers have studied the case of Russia, where abortion was legal after 1955 but access to safe, effective, and affordable contraception was substantially limited.a When contraception became more accessible after the fall of the Soviet Union, the abortion rate declined rapidly and the rate of contraceptive use rose rapidly. As women gained access to contraception, the need for and reliance on abortion to control their fertility declined. A second clear finding from this area of research is that restricting access to safe contraception and legal abortion increases maternal mortality. Population health scientists have documented this in the case of Romania.b There, too, access to safe, effective, and affordable contraception was severely limited during the Communist era. But in Romania, unlike the Soviet Union, abortion was illegal. Nicolae Ceausescu, the leader of Romania from 1965 to 1989, implemented a pronatalist policy that penalized women if they did not have routine gynecological exams, rewarded physicians for achieving a monthly birth quota, taxed unmarried and childless adults, forbade the sale of contraception, and severely penalized women and physicians for receiving or conducting abortions. Although these policies led to a short-term rise in the birth rate, the birth rate declined slowly back to its pre-Ceausescu level by 1989. Maternal mortality, on the other hand, increased significantly. By 1989, when Ceausescu was assassinated, the maternal mortality ratio was 170 deaths per 1,000 births, meaning that 1 in 210 women died as a result of pregnancy or childbirth. Maternal deaths from abortion made up 87% of all maternal deaths. With the end of the regime, and the legalization of abortion and contraception, the maternal mortality ratio declined by half within a year and continued to decline thereafter. Since Roe v. Wade was decided by the Supreme Court in 1973, women in the United States have the legal right to an abortion. Most women in America also have access to safe and effective contraception, although cost and access to health care impose barriers to contraception. Although abortion substituted for contraception in the Soviet Union and Romania, where access to contraception was limited, in the United States women rely on both to control the number and timing of births. Guttmacher Institute researchers have determined that between one in three and one in four women in the United States will have an

170

|

Chapter 7

abortion by the end of their childbearing years.c If contraception is available, why do women still have abortions? There are a number of reasons, but most important among them are contraceptive failure, unplanned pregnancies, and health concerns. Abortion is an important component of women’s reproductive health even in a context where contraception is available. That up to one in three women will have an abortion is an important statistic to keep in mind in the ongoing efforts to curtail access to abortion in the United States. Recent state laws and regulatory decisions to restrict access to abortion include cutting funding for family planning from abortion providers; provider restrictions such as licensing regulation and hospital admission requirements; gestational age restrictions; payment restrictions, such as preventing private insurance from covering abortions; and service restrictions, including waiting times, required counseling, and parent involvement in the case of minors. Population health scientists in Texas led by Joseph Potter have studied the impacts of state laws and budgetary decisions restricting access to abortion in that state and found that while abortion has declined, so too has the use of highly effective contraceptives as a result of limited access to family planning; as a result, pregnancies by women covered by Medicaid have risen.d a. b. c. d.

Deschner and Cohen 2003. Haddad and Nour 2009; Johnson, Horga, and Fajans 2004. Jones and Jerman 2017. See https://liberalarts.utexas.edu/txpep/.

lower mortality and greater morbidity. First, mortality differences between men and women reflect gendered differences in morbidity—men die at higher rates than women because they face more severe disease, whereas women suffer disproportionately from nonfatal disease. Second, mortality and morbidity differences between men and women reflect a combination of biological and social forces and their interaction. The weight of evidence argues for the role of social forces, particularly as they relate to women’s greater degree of nonfatal morbidity. However, population health research has a long way to go toward a better accounting for biological differences between men and women, as well as for interactions between the biological and the social as they unfold in complex ways across the life course. We turn next to an area of research in which biology is unlikely to explain gendered differences in health and mortality.

Gender | 171

Historical Changes in the U. S. Gender Mortality Gap One would be hard-pressed to argue that biological differences between populations located in different places, social groups, or points in time explain gender differences in mortality across those contexts. For instance, we would not likely conclude that biological differences between populations explain why the gender gap in life expectancy exceeds 10 years in some former Soviet states but is only 1–3 years in some sub-Saharan African countries. Rather, men’s mortality is elevated by substance abuse and violence in the former Soviet Union, whereas the undue burdens of poverty, political turmoil, and the AIDS epidemic drive up women’s mortality in sub-Saharan Africa.43 These and other forms of contextual variation in the gender mortality gap reveal how gender itself varies and is transformed across time, space, and social context. One topic that has been the focus of substantial research has been the changing gender gap in mortality in the United States. As we saw in figure 22, the gender mortality gap grew prior to the 1970s. This reflects the end of a longer historical trend of women’s growing mortality advantage, driven initially by women’s declining maternal mortality rate in the late 1800s.44 During the first half of the twentieth century, the gender gap in life expectancy grew almost entirely due to increases in men’s smoking; as a result, women’s life expectancy increased more rapidly than did men’s in this period of time.45 As late as 1972, men’s higher rate of smoking in the United States accounted for nearly half of the gender gap in mortality.46 The declining gender gap after 1970 has been the subject of somewhat greater debate. It is tempting to argue that the declining gap reflects an unfortunate consequence of women’s growing equality. The timing of second-wave feminism, women’s mass entrance into the workforce, declining wage gaps, and the growing representation of women in high-status positions coincides with the onset of the declining mortality gap, given a two-decade lag through which social change impacts mortality.47 But we know from a rich body of research that higher socioeconomic status is associated with lower mortality.48 Indeed, the relationship between socioeconomic status and mortality is similar for men and women, and studies find that if women had the same socioeconomic status as men, the gender mortality gap would be wider than it is.49 Thus, the idea that women’s greater social equality with men would lead to a narrowing of the gender gap in life expectancy flies in the face of a large and powerful literature on the beneficial health effects of high socioeconomic status.

172

|

Chapter 7

Others have argued that growing social equality between women and men, while not conferring a direct mortality disadvantage for women, may result in women adopting traditionally male-oriented unhealthy behaviors such as smoking, thereby undermining women’s health.50 Indeed, the United Nations referred to the convergence in men’s and women’s smoking (and other unhealthy behaviors) as the “dark side of equality.”51 While it is true that men’s and women’s rates of smoking have converged (although in the United States men are still more likely to smoke than women; see table 13), convergence results both from increases in women’s smoking and from declines in men’s smoking, a process driven by changes in the smoking behavior among U. S. birth cohorts.52 The prevalence of smoking among men peaked earlier, and at a much higher level, than it did among women. Smoking increased across cohorts of men born up through 1914 (i.e., men who were in their forties in the 1950s), who spent an average of 18 years smoking prior to age 40, and declined in cohorts born thereafter, whereas smoking increased across cohorts of women born up to 1944 (i.e., who were in their forties in the 1980s), who spent an average of about 10.5 years smoking prior to age 40, and declined thereafter.53 As a result of these different trends, after 1960, declines in men’s smoking drove convergence to a greater extent than increases in women’s smoking,54 and smoking contributes to less and less of the gender gap in mortality, from explaining 47% in 196255 to 22% in 1998–2000.56 Fred Pampel explains that differences in the smoking patterns of men and women reflect the uneven process of the social diffusion of “technologies” such as cigarette smoking.57 Social diffusion tends to follow a pattern wherein those with higher social status adopt new technologies first. In this case, mass-marketed cigarettes and, later, information about the deleterious effects of cigarette smoking are both considered technologies: men began smoking earlier and at higher rates than women, but they also abated sooner and more quickly once information about the deleterious effects of cigarette smoking became available, thereby closing the smoking gap with women. Knowing that gender differences in the timing of smoking initiation and cessation, and in the prevalence of smoking, explain a great deal of the gender mortality gap begs the question of why gender differences in smoking exist in the first place. Pampel’s diffusion argument makes the case that with new technologies such as smoking it is those with the highest status (i.e., men, and, among men, those of higher socioeconomic status) who are first to adopt and then also to abate.58 A different way of looking at this issue is offered by an emerging body of work on men’s health, which

Gender | 173

argues that health behaviors are gendered behaviors more generally—that is, health is one avenue of life through which gender is enacted.59 Men’s higher rates of smoking, alongside their heavier drinking, greater likelihood to drink and drive, more frequent violent behavior, less frequent use of health care, and so on, are expressions of masculinity through which men enact the social script of manhood as the more powerful, “stronger sex.”60 Endorsement of this type of masculine ideal is associated with a variety of health risks and unhealthy behaviors among men.61 Masculinity varies among men by other forms of status—by socioeconomic status, race, and sexuality—such that men with lower status enact masculinities in different (often more deleterious) ways.62 Because gender is defined and created through relationships, these deleterious behaviors are rewarded with social acceptance and recognition, as well as reciprocated by important others, such as health care providers.63 For example, studies have suggested that physicians are far less likely to diagnose depression in men than in women who have similar depressive symptoms.64 This argument as applied to men echoes a longer-existing argument that women’s health and health behaviors are reflections of femininity. In the case of women, the argument has frequently sought to minimize or question the extent of their suffering—for instance, to make the claim that women exaggerate health problems and overuse health care. Research does not support either claim. For example, when comparing men and women with the same diagnosed health conditions—say, cancer or arthritis—they report the same average health,65 and accounting for disability and disease explains women’s worse self-rated health.66 Given a similar distribution of health conditions to men, women would be even more likely to use the hospital than they currently do.67 These findings suggest that women’s reporting of health and use of health care is explained by their different health conditions, which, as detailed above, most likely arise from their social-structural disadvantages. Research on gender disparities in health and mortality runs the risk of essentializing gender, that is, of giving the impression that gender is an immutable trait, the outcome of biological forces with biological consequences, rather than a social product.68 The mere appearance of demographic differences can have the effect of confirming prior-held beliefs that men and women are essentially and naturally different. Often the interpretation offered by scholars examining gender differences stops short of acknowledging that both the social and the biological matter without saying how they matter in tandem.69 However, research on gender and health, including work by population health scientists, reaches

174

|

Chapter 7

more nuanced conclusions emphasizing the interplay between the social and the biological.70 First, the social meanings of gender, rather than the biological, explain the extent of and change in gender disparities in health and mortality over time and across space. Second, biological differences between men and women are attenuated or exaggerated by social forces; that is, there is an interaction between the biological and the social. Finally, biological differences between men and women are ultimately understood and made meaningful through language, interactions, and social structures—that is, the biological does not exist outside of its social interpretation. Turning back to the issue of the declining gender gap in U. S. life expectancy, the evidence is clear that the later adoption of smoking by women and their later declines in smoking have been important in explaining the smaller gender gap in life expectancy in the 2010s relative to the 1970s. This narrowing gender gap due to the cohort-based patterns of men’s and women’s smoking could continue to unfold for the next couple of decades.71 At the same time, population health patterns among women with low socioeconomic status have suffered in recent decades, as they continue to absorb the consequences of their lower wealth, incomes, power, prestige, and other social resources relative to men in U. S. society.72 Until women obtain full social and economic equality with men, we expect that their data on most measures of population health will continue to reflect higher levels of morbidity and disability than their male counterparts.

transgender health One area of rapid growth in population health research is research on transgender health. Transgender individuals, who have a gender identity or expression that is different from the gender corresponding to their sex assigned at birth, challenge gender essentialism and the gender binary.73 Research on the health of transgender populations holds the potential to transform our understanding of gender differences in health, as well as of gender more generally. Population data on the transgender population in the United States are currently rare, as most available national surveys, censuses, and population registers ask about sex / gender using the male / female binary.74 One exception is the Behavioral Risk Factor Surveillance System (BRFSS), which collects population data in U. S. states and in 2014 included an optional survey module on transgender identity, which 19

Gender | 175

states adopted. Based on those data, Flores et al. estimated that 0.6% of the U. S. population, or 1.4 million adults, identify as transgender, with the number highest among 18 to 24 year-old individuals.75 This estimate is two times higher than the previous national estimate, which relied on two smaller state-level data sets.76 The vast majority of research on the health of transgender individuals uses convenience samples and finds that transgender respondents have higher rates of HIV infection and mental health problems, and more limited access to health care, compared to the general population.77 However, one study using population data from Massachusetts found few health differences between transgender and nontransgender respondents.78 Transgender respondents were less likely to be overweight and more likely to currently smoke than cisgender respondents but were no different in terms of self-rated health assessments, activity limitations, recent bouts with poor physical health, exercise, HIV screening, diabetes, heart disease, asthma, binge drinking, mental health, and access to needed care for mental health. It is not clear whether the difference in findings between the research using convenience samples and the Massachusetts study reflects who participates in convenience samples versus those who respond to a household-based telephone survey, or something else. More data sets with detailed measures of gender identity and more research on this topic are sorely needed. Fortunately, better data—and presumably more research—on transgender health are in the works. The U. S. Office of Disease Prevention and Health Promotion has established “increasing the number of population-based data systems . . . that include in their core a standardized set of questions that identify lesbian, gay, bisexual, and transgender populations” as one of its Healthy People 2020 objectives. This includes expanding the number of states that adopt the module on transgender identity in the BRFSS. Furthermore, three data collection efforts dedicated to understanding the experiences of transgender individuals were in the field as of 2016: TransPop, the first national probability sample of transgender individuals;79 the U. S. Trans Survey, the largest national, nonprobability survey of transgender individuals;80 and Project Affirm, the first longitudinal study of transgender individuals.81 Various national data sets were also in the process of incorporating questions regarding gender identity into their designs, including Wave V of the National Longitudinal Study of Adolescent to Adult Health, the Medicare Current Beneficiary Survey, the Population Assessment of Tobacco Health Study, and the Recognize Intervene Support Empower Study. The NIH

176

|

Chapter 7

Sexual and Gender Minority Research Coordinating Committee has recommended that the National Health Interview Survey (NHIS), the National Health and Nutrition Examination Survey (NHANES), and the National Survey of Family Growth (NSFG) all incorporate questions regarding gender identity into their survey instruments. As these data become available within the next decade or so, much more research on the health of transgender individuals will be conducted.

sexual orientation and health Although distinct from gender identity, sexual minorities are often grouped with gender minorities under the label “LGBTQIA,” which stands for Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, and Asexual.82 Research on the health of sexual minorities among this group—the lesbian, gay, and bisexual (LGB) population—is farther along than research on transgender health.83 The AIDS epidemic brought attention to the particular health risks of the LGB population by highlighting both the importance of sexual behavior in the risk of contracting HIV as well as the substantial social barriers to recognizing, funding, and responding to a disease that at first predominantly affected the gay community. The specific health needs of sexual minorities are also important because of the physical and mental health strains imposed by social stigmatization, homophobia, and limits on the civil liberties of LGBTQIA people.84 In addition to the BRFSS described above, government health surveys with questions about LGB identity include NHANES, NSFG, the National Alcohol Survey, the National Comorbidity Survey, and, starting in 2013, the NHIS. Studies using these and other population data document some health differences between the LGB and the straight populations, but recent analyses of national data do not reveal a systematic disadvantage for sexual minorities, particularly gay men and lesbians.85 For instance, based on NHIS data, gay and lesbian respondents were more likely than straight respondents to be current smokers, to have consumed five or more alcoholic drinks in one day at least once in the past year, and to have experienced serious psychological distress in the past year, but they were also more likely than straight respondents to have met the federal guidelines for physical activity, to have received a flu vaccine, and to have been tested for HIV, and they were no different from straight respondents in terms of their self-reported health or level of obesity.86 The study documented a more concerning profile for bisexual respondents, whose rate of psychological distress was double that of the other groups, and who also

Gender | 177

had a higher rate of obesity and a lower rate of good self-reported health compared with the gay, lesbian, or straight populations. Other studies also find especially elevated health risks among bisexual respondents, including high rates of suicidal ideation, victimization from intimate partner violence, and risky alcohol and drug use.87 Prior to the recent availability of data that has allowed individuals to express their sexual orientation, some researchers have creatively attempted to assess the health of the LGB population by using the sex composition of marital and cohabiting unions in household-based surveys—or “relationship orientation.” This research is limited of course to partnered people and has tended to focus on health differences between married and cohabiting couples. The research finds that same-sex cohabiters have better health than different-sex cohabiters, an advantage driven by the higher socioeconomic status of same-sex cohabiters.88 However, same-sex cohabiters are disadvantaged compared with different-sex married adults, as are, by extension, different-sex cohabiters, consistent with a large body of research on the health benefits of marriage.89 Other work finds that children in married same-sex households have similar health to children in married different-sex households and better health than children in cohabiting households, whether same- or different-sexed.90 This work suggests that that the recent federal legalization of gay marriage may improve the health of gay and lesbian couples and their children.91 Overall, as with research on the transgender population, the increased availability of data on LGB individuals should lead to a much greater understanding of the population health of the LGB community in the coming decades. This is a long overdue and very welcome change. Better understanding of LGB health, in an era of both increased acceptance of LGB individuals and continued discrimination against them, should help lead to programs and policies that are important for the health and well-being of an important subgroup of Americans. This chapter suggests that gender expressions of femininity and masculinity may have a lot to do with gender disparities in life expectancy; most important, stereotypically masculine behaviors such as driving too fast, using violence, abusing alcohol and drugs, smoking cigarettes, and eating fatty foods have seriously cut into men’s longevity relative to women’s over the course of U. S. history and continue to do so until the present day. Looking down the road from a population health perspective, it will be important that our society work toward de-coupling the idea that being male means behaving in ways that seriously cut into longevity. Such

178

|

Chapter 7

efforts may be especially important in childhood and adolescence—when ideas of gender are most intensely socialized. At the same time, our chapter has shown that U. S. women’s health and longevity is harmed by lower socioeconomic status relative to men. In this way, we have suggested that fundamental cause theory is an important lens through which women’s generally poorer population health relative to men’s may be understood and combatted. That is, improvements in women’s population health relative to men’s will take continued efforts to achieve equity in women’s career options and trajectories, pay, wealth, and power in both social relationships and institutions. Unfortunately, U. S. society is still far away from achieving such equity. Both the de-coupling of men’s damaging health behavior from the meaning of being a strong male and the achievement of true gender equality in social status between women and men will necessitate that the meaning of gender needs to change in important ways for U. S. population health to improve for both women and men. This chapter also dug into the relatively sparse research on population health among the transgender and LGB populations. Just 10 or so years ago, a book like this probably would not have included such sections at all. Fortunately, our society has finally moved toward greater acceptance of LGBTQIA individuals, and national data sets are collecting information that will allow the population health community to arrive at a much greater understanding of health disparities across these historically marginalized groups. We expect that some of the greatest gains to population health knowledge over the next decade will occur in our understanding of LGBTQIA health, which will be a very welcome addition to this area of study.

chapter 8

Policy Implications of Population Health Science

The Universal Declaration of Human Rights, adopted by the United Nations General Assembly in 1948, proclaims that “everyone has the right to a standard of living adequate for the health and well-being” of her- or himself.1 Population health science begins from this premise: health is a basic human right. Recognizing this, population health scientists are fundamentally interested in societal well-being—how well society is doing at guaranteeing this human right to everyone. Population health scientists study how long individuals within a population live and with what degree of health versus illness as a way of understanding societal well-being as a whole. There are of course many other human rights, such as the right to education, to work, and to security, and other ways of measuring societal well-being, such as average levels of education attainment. But population health is a fundamental measure of societal well-being because health is a platform on which all other social activities take place: good health enables learning, work, and caretaking. Health ties our social experiences to our bodies and reminds us of our basic humanity—the fact that we will all ultimately face illness and death, in spite of our differences. Thus, as measures of societal well-being, population health statistics reflect how well society is doing at taking care of its people. This is easy to appreciate when we consider infant mortality, an important agespecific death rate because it so intuitively captures this concept. Infants are completely dependent on their caretakers and environment, so the 179

180

|

Chapter 8

rate of infant survival through the first year of life captures how well society is doing to support its most vulnerable members. But, of course, this same understanding applies across the life course when we consider that our health is embedded in and dependent on the physical and social environments in which we live at all ages. Thus, when we observe that life expectancy in the United States increased from 47 years in 1900 to over 78 years in 2016, we appreciate the remarkable social progress that unfolded over that era in order to allow a far greater number of individuals to live for much longer than they ever did before. Indeed, the fact that life expectancy reflects societal well-being is why it is a key component of the United Nations’ Human Development Index and why we rely on it so heavily in this book. As we introduced in chapter 1, population health as a subject matter of study is interested in documenting the patterns and trends in health within specific places and periods in history, explaining those patterns and trends, and translating research findings into action to improve population health. A great deal of population health research involves the first critical step: the documentation of patterns and trends using the best available population data. This work demands the use of transparent, intuitive, and comprehensive measures of population health and high-quality data that are representative of the populations they come from. Only once prevailing patterns and trends of population health are accurately documented using valid measures of population health can the next two jobs be done—the work of explanation and translation. In our experience, it is not hard to convince students that population health is a worthy area of study. Students in our classes do not typically question the idea that health is a human right. Students are also easily assured that carefully constructed population health statistics that use high-quality nationally representative data accurately reflect societal levels of health. The power of demographic data and techniques to describe patterns and trends in population health is not lost on them. What is harder to convince students of is why such patterns and trends emerge and what to do about them—the explanation and translation pieces. Our experience suggests that the most challenging tasks for population health scientists is the work of explaining why population health patterns and trends unfold as they do and what can be done about them. Put simply, this is incredibly hard work—both given the scientific challenges in explaining patterns and trends in human behavior, as well as because such explanations and translation ideas often challenge common understandings of health and social / health policy.

Policy Implications | 181

Indeed, population health scientists face something of an uphill battle in their work on explanation and translation because common ideas about health and health policy are frequently at odds with science. We have observed two common fallacies related to population health among our students. The first is that health is the outcome of individual characteristics and behaviors, an idea influenced by the predominant U. S. culture of individualism. As this idea goes, population health reflects the accumulation of individual characteristics and behaviors, and differences in population health across groups reflect differences in the characteristics and behaviors of individuals in those groups. The policy implication of such an individualist understanding of population health is to encourage, manipulate, or mandate certain behaviors—or, more cynically, to not intervene at all. This approach to understanding population health is appealing because it provides the false reassurance that the quality of one’s life and the timing of one’s death are either out of individual control because they were determined at birth or completely within individual control, the result of personal health-related choices and actions. But as this book has made clear, while health behaviors are important proximate determinants of health, they are shaped by other, more fundamental social forces. If population health measures reveal how well a society takes care of its members, then health disparities reveal social inequality. An individual’s ability to take action on behalf of her or his own health, whether that be through exercise, a healthy diet, or avoiding risky behaviors, depends on the resources, power, and knowledge available to them and the freedom they have to act. The implication of this insight is that the extent to which health is within our individual control depends on forces well outside our control: when and where we are born, into what body, and with what preexisting privilege in our families, and the social systems in those periods and places that structure our opportunities and give meaning to the shape, anatomy, color, and identities we embody. A hidden problem of the individualist argument is that it rhetorically rewards those who are already most able to take action in their lives and demeans those who are least powerful to take action. The second common fallacy we have discovered among our students regarding population health is that health care is the most important determinant of health and health disparities. This fallacy is also at least in part influenced by the predominant U. S. culture of individualism, but in this case as exemplified in the miraculous capacity of physicians and medical practice. Students are aware that the United States has a health care problem, has a large number of uninsured people, and spends a lot

182

|

Chapter 8

of money on health care. While students may not be deeply familiar with the nuances of policy debates surrounding the Affordable Care Act, they have heard of Obamacare, know it is controversial, and believe there is a serious problem that the policy attempts to address (even if they disagree with the potential solutions). If students do not fully buy that health outcomes are the result of individual characteristics and behaviors, the determinant they most often identify is access to medical care. Simply put, though, that idea gives far too much credit (and blame) to medicine / physicians and far too little blame (and credit) to other social institutions. Much scientific work has shown that access to and utilization of health care is responsible for only a modest amount of variation in population health across societies and subgroups within societies.2 Yet, inevitably, any serious health policy discussion at the national level in the United States focuses exclusively on access to medical care. While having access to health care certainly matters for preventing some diseases and effectively treating others, a lack of health care is not the cause of disease or injury (though it may be a stressor). Even in cases where health care can clearly affect the progression of illness or injury, it is insufficient to say that inequality in access to health care is itself the cause of population health problems. Moreover, inequality in access to care is not inevitable, for nearly every high-income country other than the United States has achieved universal access to health care. What causes inequality in access to health care in the United States? What are the political, economic, cultural, and social systems that create such inequality in the first place? What the study of population health reveals is that individual control over one’s health and length of life—whether that be through behaviors or access to health care—is profoundly influenced and constrained by social forces. So, what have we learned? The population health science covered in this book has shown that in certain respects, the United States is doing a pretty good job at taking care of its people. Life expectancy and other general measures of population health improved considerably and without historical precedent over the past 150 years as a result of socioeconomic development, major scientific and medical advances, and the investments that cities, states, and the federal government made to distribute that development and those advances to the population at large through public health projects, social institutions like schools and hospitals, and the social welfare system. The Black-White gap in life expectancy declined dramatically over the same period, reflecting improving material conditions and growing human rights, particularly as African Americans and

Policy Implications | 183

their allies fought for opportunities and equality through the Jim Crow Era, the Civil Rights Movement, and the Black Lives Matter Movement, among many other private and social efforts. The gender gap in population health also declined over this period, as women gained greater rights and socioeconomic power, gender differences in smoking narrowed, and slow cultural change began to unravel how masculine and feminine ideals inform the behaviors, social psychology, and identities of men and women. We are hopeful that these improvements will continue into the future and that, within our lifetimes, race, gender, immigrant status, sexual minority status, and other identities will no longer differentiate how long Americans live and the degree of suffering subgroups face while alive. Given, however, the ongoing struggles for social and economic equality among U. S. women, transgender individuals, racial and ethnic minority groups, sexual minorities, and immigrants, we know that much work remains to be done. This is especially the case in the context of the current Trump administration, which has not established an environment conducive for social and economic equality across groups. In other ways, the story of U. S. population health is even less hopeful. Chapter 3 emphasized that the United States lags behind other highincome countries in terms of population health, an international health disadvantage observed across a variety of measures of health and for nearly all subgroups. The problem is multifaceted and complex, but it captures something uniquely unhealthy about the United States. Based on the patterns and trends discussed in chapters 4 and 5, we suspect that this problem is closely tied to another U. S. population health problem—growing geographic and socioeconomic inequalities in health. Since 1980, the same period over which the U. S. international disadvantage emerged, inequalities in health between the best-performing U. S. counties / states and the worst-performing counties / states grew, and inequality in health between those with college degrees and high income and those with less than a high school degree and low income grew. Some groups of people and places in the United States are reaping the health benefits of rapid social, economic, and technological change over the last 40 years or so, whereas others are not. Of particular and immediate concern is the opioid crisis, which captures the linked processes of growing geographic and socioeconomic inequalities. The opioid crisis is uniquely American—no other highincome country faces its ravages. It is felt most acutely among the rural poor, among people with less than a high school degree, and among those who live in places like southern West Virginia, where Boone

184

|

Chapter 8

County is located—places that have been left behind by the postindustrial economic transition that has unfolded over the past 40 years. It is in these places where the increasing availability of prescription opioids and, more recently, of heroin and the synthetic opioid fentanyl have intersected with increasing levels of social and economic despair to produce an addiction and overdose epidemic of a scale never seen before.3 What can be done about this and other population health problems? Admittedly, population health scientists working in academic environments do more work documenting and explaining than they do translating their research into action to improve population health. Designing and implementing population health policy requires much more than an understanding of the how’s and why’s of population health. Although the how’s and why’s are critically important, so too is understanding the political process, legislative and bureaucratic rules, stakeholder priorities, financial and budgetary analyses, and existing policies—all areas of expertise that most often fall outside the expertise of population health scientists. Nonetheless, we can draw some clear implications of the population health approach for policy direction, which of course is what it means to do the work of translation. We begin by directly discussing policy approaches that stem from the two fallacies we just described: health care policy and health policies narrowly focused on modifying individual behaviors. We then turn to a discussion of policies that have the potential to more directly address the fundamental causes of population health and of population health disparities.

health care policy As we described in chapter 3, the U. S. health care system is a patchwork of different types of providers and payers, operating in similar ways to other high-income countries for subsets of the U. S. population. Medicare, for instance, is similar to the National Health Insurance (NHI) system in Taiwan, but the NHI covers everyone in Taiwan, whereas Medicare only covers those 65 and over in the United States. Many nonelderly adults are covered by private insurers, with premiums shared between themselves and their employers, similar to Germany. Thus, the payment of medical services in the United States is fragmented across many different subsystems, which leads to several major problems. Perhaps the most obvious is that a large portion of the population is left out. Even after various steps were taken through the Affordable Care Act (ACA) to expand insurance coverage, about one out of 10 nonelderly people in the

Policy Implications | 185

United States still lacked health insurance coverage in 2016, well after the passage of the ACA.4 The second problem is that the system is costly. A system of multiple payers increases the administrative work of fee negotiation and billing. In the United States, more than a quarter of all medical expenditures go toward administrative costs.5 Furthermore, the fact that prices are negotiated across multiple payers means that the negotiating power of any single payer is reduced, and prices increase as a result. In places like Japan and Germany, where private insurers pay for medical care, prices are set by the government. Price control achieves both lower costs and equity, as with price controls, providers cannot charge more for superior care. The third problem is that quality is compromised in a fragmented system, as a large number of uninsured people may forgo needed care due to cost, the pricing structure incentivizes the provision of highcost services over prioritizing healthy outcomes, and communication between providers is limited. The NHI in Taiwan offers an exemplary model of health care provision and payment.6 Everyone in Taiwan is mandated to obtain coverage through the NHI, which is paid for by a combination of premiums, copays, and government funds through general taxation. Individual premiums are equal to a percent of monthly salary (up to a ceiling; in 2018, the rate was 4.69%), and premiums are shared among individuals, employers, and the government, depending on the individual’s occupation, employment status, and income (e.g., the government subsidizes premiums for individuals from low-income households).7 Coverage in Taiwan is comprehensive, including in- and out-patient services, preventive care, prescription drugs, emergency care, surgery, in-home nursing care, renal dialysis, Chinese medicine, and mental health care. Copays are waived for many basic services. Providers in Taiwan are private employees and they compete for patients by providing high-quality care and attractive services, such as after-hours care or in-home visits. Fees for services are set by the NHI Administration with input from a multiple-stakeholder committee. This structure addresses several of the major problems of the U. S. health care system. First of all, no one in Taiwan is uninsured, and no one goes into bankruptcy paying medical bills. Because there is a single payer—and also because Taiwan invested in a sophisticated and comprehensive information technology system— administrative costs are very low, only 1% of all medical costs. People are highly satisfied with their system, and population health outcomes are very good. In 2014, life expectancy at birth in Taiwan was 77 years for men and 83 years for women, one and two years longer, respectively,

186

|

Chapter 8

than in the United States, even though Taiwan’s gross domestic product (GDP) per capita is $10,000 less than in the United States.8 What is the state of health care reform in the United States? The ACA took several measures to expand insurance coverage, including prohibiting the denial of insurance coverage on the basis of preexisting conditions, expanding the upper age limit (to 26) for children to be covered by their parents’ plans, mandating that everyone be insured, creating staterun insurance exchanges for individuals to purchase coverage on the private market, subsidizing the cost of insurance for low-income households, and expanding eligibility for Medicaid. Several of these efforts have been politically curtailed, and others have not worked as well as hoped. As of 2018, 18 states had refused to expand Medicaid coverage, leaving a swath of low-income households uninsured.9 Although the Supreme Court ruled that the federal government could mandate that individuals purchase insurance through taxation, the tax reform legislation passed at the end of 2017 eliminated the penalty for not complying with the mandate. It is feared that those who need insurance the least—the young and the healthy—will opt out, causing insurance premiums to rise to levels that will be unaffordable for individuals who are in real need on the private market. Finally, owing to a variety of causes, premiums have not stabilized and some state exchanges have not been as competitive as intended, reducing options for those on the individual market and presenting the concern that the system is unsustainable without further measures to control costs. Thus, there is quite a lot of work left to be done to overcome the problems of quality, access, and cost that plague the U. S. health care system. As we reflected earlier, health care is not the primary determinant of population health, and this is not just because health care cannot prevent all illness and early death. It is also because the structure of the health care system reflects the political, economic, and social systems of the country, and those systems affect health through a variety of mechanisms. It is without doubt that there are other reasons why Taiwanese population health statistics are superior to the U. S.’s in addition to its health care system. But it is also possible that there is something that differs between Taiwan (and other high-income countries with comprehensive health insurance coverage) and the United States that explains both the structure of the health care system and population health outcomes. For instance, one could hypothesize that an ideological commitment to caring for others, to seeing one’s fate as tied to the fate of one’s neighbors, and to believing in the importance of equality of opportunity and equity in outcomes,

Policy Implications | 187

results in health care policies that prioritize universal access to health care over the financial interests of private insurance companies, health care providers, the pharmaceutical industry, and their investors. This kind of ideological context would presumably produce other policies that invest in people, provide security, ensure equality, and maximize well-being, and these kinds of policies would arguably benefit population health in myriad ways. But as we concluded in chapter 3, there is not good research evidence assessing this type of hypothesis, and it would be difficult to construct a study that did so while also meeting the standards of population health science. Perhaps the closest we might come is to better understand how political, economic, and social context influences population health patterns, trends, and disparities in the 50 U. S. states, which could provide important insights on what contexts might also be most favorable for population health at the national level. Important work in this area is ongoing.10

policies that influence individual health behaviors Health care policy is a “supply side” policy—it attempts to affect population health by changing the supply of medical care. But, as James House argues, supply-side policy will do little to affect demand: the extent to which a population needs health care in the first place.11 Demand-side policies, House argues, are far more effective population health policies. Following from the idea that health is the outcome of our individual behaviors, probably the most common way of thinking about demand-side health policy relates to efforts to change individual health behaviors, which indeed can be extremely effective. Tobacco regulation policies offer a great example. Through a multipronged public health approach that began in the mid-1960s and which has included information campaigns, sales and advertising regulations, product warning labels, prohibitions on smoking in public spaces, workplace initiatives to cut smoking, and increases in tobacco taxes, rates of smoking in the United States were reduced by more than 50 percent between 1964 and 2014.12 This multipronged strategy and resulting reduction in smoking behavior has arguably been the country’s most important public health achievement in the past half-century, saving literally millions of lives and helping to improve the health of many millions more. Nonetheless, much work remains: smoking still kills roughly 480,000 Americans every year, prematurely sickens hundreds of thousands of others, and costs U. S.

188

|

Chapter 8

individuals and taxpayers nearly $300 billion per year.13 And disparities in cigarette smoking behavior, especially between the less- and highly educated segments of the population, are extremely wide.14 Thus, even this incredibly successful, multidimensional health policy effort targeted at changing individual behavior has not solved what continues to be an American (and increasingly worldwide) tragedy. Other health policies that target individual behaviors include minimum drinking ages, school-based physical education programs, seatbelt laws, speed limits, soda taxes, mandatory vaccination for public school enrollment, health screening recommendations, media campaigns about drinking and driving, gun regulations, and much more. There is a very large research literature testing the effectiveness of these policies on behaviors, health, and mortality, which shows that many of them have effectively changed health behaviors, prevented illness and injury, and lengthened life.15 But while many behaviorally focused population health policies may be effective overall, population health science and fundamental cause theory raise concerns about the potential of such policies to actually widen, rather than reduce, health disparities. One problem is that focusing on a particular health risk factor or disease outcome will not eliminate social gradients in health because individuals with more resources, power, knowledge, and freedom will always be better able to act on behalf of their own health, even as the behavioral or disease environment changes. That is, even if individually focused population health policies effectively eliminate the risk of a certain disease—say, of smallpox through universal vaccination—the social gradient in that disease (smallpox) will be eliminated, but the gradient will still exist for any other disease that one can direct knowledge, money, and other efforts toward avoiding and treating. Another important issue regarding specific behavior- or diseasefocused policies is that, given inequalities in resources, power, knowledge, and freedom, not all people are equally able to access or follow public health advice. For instance, the U. S. Department of Agriculture releases healthy eating guidelines (e.g., the Food Pyramid), and the Food and Drug Administration requires chain restaurants to post nutritional information on their menus. The idea is to encourage nutritious food choices in order to improve population health. But who is best able to access and implement this advice and take into consideration nutritional information when making food choices? Most likely, it is people who read English; have resources to eat at restaurants; who already have substantial knowledge of nutritional science and recommended best prac-

Policy Implications | 189

tices for diet and health; who have sufficient resources with which to make healthy choices, which includes the ability to search and travel for different food options; who are able to satisfy cravings and seek pleasure from healthy, nonfood outlets and tasty, nutritious food sources; and who are surrounded by other people who share, reflect, and affirm these dispositions, tastes, and habits. In other words, it is often those who are already substantially socioeconomically advantaged who are in the best position to take advantage of individually oriented public health policies. If this is the case, then individually based public health policies can reproduce or even widen social disparities, as perhaps most clearly shown in the tremendous increase in socioeconomic disparities in cigarette smoking within the context of the overall decline in smoking since the mid-1960s.16 Is it possible to focus on population health policy that improves individual behaviors or health outcomes while not aggravating inequality? As Link and Phelan wrote in their seminal article that first described fundamental cause theory, one approach to the problem is to “contextualize risk.”17 What they meant by this is instead of trying to modify individual risk factors or behaviors through policy, policy should instead focus on changing the environment so that individual risk factors and behaviors become irrelevant. For example, instead of urging individuals to filter, boil, or otherwise decontaminate water before consuming it, health policy should seek to ensure that all water is clean before anyone drinks, washes, or cooks with it. Instead of urging individuals to cook meat to a certain temperature, food inspectors could ensure that meat is free of disease-causing bacteria before it is sold to consumers. Policies that contextualize risk by changing the environment can have the effect of improving health while reducing socioeconomic gradients in health, insofar as those who have the fewest resources—those who are least able to access the information and enact change on behalf of their own health—will benefit most from the environmental change. Because not all individual health behaviors substitute so easily with environmental change, though, contextualizing risk also means considering how resources, knowledge, and practice may prevent some individuals from acting on public health policies. Policies should then incorporate means for overcoming those barriers. For instance, if vaccinations are required for enrollment in public schools, the cost of vaccinations should be subsidized for low-income households. If public health campaigns seek to motivate exercise by encouraging outside activity, then city-planning and land-use policymakers should invest in equal access

190

|

Chapter 8

to safe, green spaces across all neighborhoods and not just those in newer, wealthier housing developments. When public health interventions and policies are designed, an important element should be efforts to make them available to all and minimize the degree to which their effectiveness depends on preexisting resources, power, knowledge, and freedom.

“non-health” social policies Taking the insight of contextualizing risk one step further, we arrive at a much broader set of demand-side policies: policies that affect the level and distribution of fundamental causes of health, including resources, power, knowledge, and freedom. Social and economic policies are not typically thought of as health policy, but James House makes the case that they should be.18 Indeed, he argues, all public policy should be assessed according to its health impact, just as policies are assessed according to their environmental impact. He argues that social and economic policies should include a Health Impact Assessment (HIA) to address whether there is evidence that the policy will impact health and how it does so and whether the policy is feasible, cost-effective, politically palatable, and likely to have other positive effects and few negative externalities. Recognizing the health impacts of putative non-health policies may indeed shift the cost-benefit analysis of a variety of public programs toward clear support for their enactment. As a simple example, increasing the federal minimum wage may not only put a bit more money into the pockets of low-wage workers; it may also have (considerable) health spillover effects. A variety of programs seek to increase the level and equality in resources, power, knowledge, and freedom within U. S. society. These include education policies, such as increased funding for early childhood, secondary, and postsecondary schools; social assistance policies such as Temporary Assistance for Needy Families (TANF), Supplemental Security Income (SSI), food assistance (food stamps), the Earned Income Tax Credit (EITC), housing aid, and the Women, Infants, and Children (WIC) program; social insurance programs such as Social Security, Medicare / Medicaid, unemployment insurance, and workers’ compensation; and a large class of policies that seek to guarantee equality, including civil rights, equal rights, LGBT rights, and voting rights policies. Economic policy—including tax policy, minimum-wage law, and corporate regulation—also have large social impacts on the level and distribution of social resources. Over the next decade or more, for example, it

Policy Implications | 191

will be important for researchers to study the social, economic, and health impacts of the 2017 Tax Cuts and Jobs Act to best understand how substantial tax cuts, aimed especially at high earners and businesses, influence population health and health disparities. Indeed, this new tax law was debated and passed without any discussion regarding its potential impacts on population health or health disparities. A fourth class of policies that clearly impact health include those that regulate land use, environmental impacts, and development. Although we know quite well that these policies impact the level and distribution of social resources within a population, we echo James House’s call for consistent, systematic research on the impacts of these policies on health. As an example of considering non-health social policy as health policy, we turn back to one of the most important U. S. population health problems that we identified above—very large and growing geographic and socioeconomic inequalities in health. Policy-wise, what should or could be done to reduce socioeconomic and geographic disparities in health while, at the same time, improving overall population health? First, we argue that policymakers cannot wait for publication of the ideal causal evidence in the scientific literature; there will never be a randomized control trial that proves that a certain residential configuration of people or a certain amount of additional education, income, or wealth will have a particular impact on a specific health outcome(s). Population health science is not conducted in a controlled, laboratory setting. And even if there was evidence from such a hypothetical randomized control trial, such a causal effect would most likely be contextual; that is, it would be specific to a certain time and place.19 Instead, as emphasized throughout this book, it is critical for policymakers to recognize that the population health research community has been building a large and increasingly sophisticated body of research over the last few decades. This research clearly demonstrates that, at least for the contemporary United States, socioeconomic and geographic disparities in health are wider than ever, that multiple health outcomes are affected by such inequalities through a large set of mechanisms, and that the fundamental causes of the socioeconomic and geographic disparities are in need of serious policy attention. Most simply, people with low education, those who are unemployed, those who have low income, those who have little or no wealth, and those who live in poor and rural areas are sicker in many ways and die from a range of causes much earlier than those who are better off. This is true for infants and children, as well as for adults. In a wealthy country such as the United States, we

192

|

Chapter 8

argue that this is evidence enough for aggressive action in treating the social and economic inequalities that are fundamentally related to health and longevity. Second, of the four dimensions of socioeconomic status, educational attainment seems to exhibit the clearest causal impacts on a range of population health outcomes.20 Moreover, it also plays a major role in influencing the other dimensions of SES, including occupational status, income, and wealth. It seems prudent that an incredibly wealthy country like the United States should make sure that all preschoolers, children, and adolescents have access to the high-quality education that would facilitate lifelong good health for all. Such an effort could very well take major property tax redistribution efforts to make sure that children in all parts of the country—rural, suburban, and urban; south, west, north, and east—and those in both poor and wealthy areas and neighborhoods have access to schools that start young, have high-quality instruction, and achieve uniformly high outcomes. Our nation’s children, far too many of whom experience less-than-stellar educational opportunities, deserve better than what they have been offered in recent decades. Our nation’s future well-being—both in terms of population health and in other key arenas—depends upon such aggressive policy change. Third, we also documented that increasing socioeconomic disparities in health over the last 40–50 years in health are consistent with broader U. S. trends in increasing socioeconomic inequality. The income and income distributions have widened and high SES individuals increasingly live in areas geographically separated from low SES individuals. Population health trends are mirroring these social and economic trends; moreover, the population health science literature we reviewed in chapters 4 and 5 strongly suggests that it is likely that the broader trends have also played a key role in the widening of health inequalities. Having learned such a lesson, it seems clear at this point in the nation’s history that aggressive policy actions must now be taken to reduce social and economic inequalities and, thus in turn, also reduce health inequalities. Although racial and gender disparities in health have decreased over the same period, there is also more to be done in these domains as well, and policy should continue to create equal opportunities and access to resources for people of all backgrounds and identities. But focusing on social and economic disparities in the form of education, employment, income, wealth, and geography may not be enough. Consider three of the major epidemics we discussed throughout the

Policy Implications | 193

book that have wreaked incredible havoc on U. S. population health over the past century: tobacco, obesity, and opioids. While all are influenced by individual health behavior, inequalities in access to medical care, and socioeconomic disadvantage, they are also all the product of larger corporate influences and government failures. Most recently, of course, the opioid epidemic was created by a profit-driven pharmaceutical industry whose addictive product was approved by governmental regulators and far too loosely prescribed by medical doctors. The fact that the opioid epidemic has thus far been a uniquely American phenomenon strongly suggests that our unique corporate culture and lax governmental regulation may be an important contributor to our unique pattern of overall poor population health compared with our highincome counterparts. As such, we argue that it is time that American social and economic policy takes cues from other countries in reining in the power of corporations and increasing the role of the federal government in more effectively regulating harmful (and very profitable) products that have taken such a toll on our nation’s health. We are far from the first population health scholars to issue this call for aggressive social and economic policy change. A decade ago, for example, Robert Schoeni and colleagues documented an array of educational, income, civil rights, employment, welfare, and housing policies that have made important impacts on improving U. S. population health over the last half-century or so.21 They urged policymakers to build on such a base and take seriously the idea that social and economic policies double as health policies. But as our nation’s population health data reflect, the social and economic policy agenda emphasized to date has not been nearly enough. Mauricio Avendano and Ichiro Kawachi, in analyzing the overall worse population health of the United States compared with our high-income counterparts, summarized that “there is ample evidence that social policies and programmes potentially affecting health across the life course are less comprehensive in the US than in other OECD countries.”22 We echo one of Avendano and Kawachi’s conclusions that more research is needed to better understand which social and economic policies are most effective for improving population health and reducing disparities. At the same time, though, we cannot wait on such research to take policy action. We are losing far too many lives too early and the collective health of our nation is nowhere near where it should be, especially given our bounty of resources. Corporate greed, lax governmental regulation, low education, unemployment, homelessness, residential segregation, poverty,

194

|

Chapter 8

and entrenched discrimination against racial / ethnic, gender, sexual minority, transgender, and immigrant groups all remain vexing issues that are affecting national health and well-being. The time to shift momentum and move to a more comprehensive, equitably oriented social and economic policy agenda is now.

Notes

chapter 1 1. Case and Deaton 2015. 2. Woolf et al. 2018. 3. National Research Council et al. 2011, 2013. 4. James House (2015: 1) refers to the poor U. S. international health standing in combination with the highest level of health care spending in the world as America’s “paradoxical crisis of health care and health.” 5. See, e.g., Liu and Hummer 2008; Lynch 2003; Masters, Hummer, and Powers 2012; Karas Montez et al. 2011. 6. Olshansky et al. 2012. 7. Kochanek, Anderson, and Arias 2015; Satcher et al. 2005. 8. Hayward et al. 2014; Sheftel and Heiland 2018. 9. Fenelon 2013a. 10. Murray et al. 2006. 11. Murphy, Kochanek, Xu, and Arias 2015. 12. Adler et al. 2013. 13. See the Interdisciplinary Association for Population Health Science website at https://iaphs.org (Bigman et al. 2018). 14. House 2015. 15. House 2015. See also Schoeni et al. (2008) for an excellent volume on how social and economic policies double as health policies. 16. Adler et al. 2013; Kindig and Stoddart 2003. 17. See, e.g., Link and Phelan 1995; Phelan and Link 2015; Phelan, Link, and Tehranifar 2010. 18. Adler et al. 2013. 19. See, e.g., Kerbo 2011. 20. See Grusky and Weisshaar 2014; Kerbo 2011. 195

196

|

Notes to Chapter 1

21. Phelan and Link 2015; Williams and Sternthal 2010. 22. Bird and Rieker 2008; Read and Gorman 2010a. 23. Entwisle 2007; Harris and McDade 2018. 24. Durkheim [1897] 1951. 25. Boardman 2009. 26. Duncan 2008. 27. Again here, see House 2015 and Schoeni et al. 2008 for excellent examples. 28. See Preston, Heuveline, and Guillot (2001) for a more in-depth and technical description of the Lexis Diagram. Preston and colleagues also provide a thorough and technical, yet accessible, description of the most important demographic concepts and measures used in that field. We draw on Preston and colleagues here in our attempt to most efficiently explain how population health is measured. We refer interested readers to the Preston et al. book for a much deeper understanding of the core concepts and measures of demography. 29. Central Intelligence Agency 2018. 30. World Bank Group 2018. 31. See, e.g., New York State Department of Health 2018. 32. Murphy, Kochanek, Xu, and Arias 2015. 33. Preston, Heuveline, and Guillot 2001. 34. Again, Preston, Heuveline, and Guillot (2001) is a very clear and accessible text to understand the mechanics behind life tables. Common statistical programs can be used to calculate life expectancy using life tables. 35. As with most things, there is an exception. Life expectancy can be calculated for a birth cohort of individuals if everyone in that cohort has died. In that case, life expectancy simply equates to the mean age of death for that cohort of individuals. But this cohort-specific measure of life expectancy can only be calculated after 100 years or so, that is, after the cohort is extinct. Thus, cohortbased life expectancy is not an often-used measure of population health. 36. See Preston, Heuveline, and Guillot (2001) for an accessible, yet technically precise, discussion of age-adjustment (i.e., age-standardization). 37. For details on the National Health Interview Survey, see www.cdc.gov /nchs/nhis/index.htm (CDC / National Center for Health Statistics 2017a). 38. For details on the National Health and Nutrition Examination Survey, see www.cdc.gov/nchs/nhanes/index.htm (CDC / National Center for Health Statistics 2015). 39. For details on U. S. National Vital Statistics System—Mortality Data, see www.cdc.gov/nchs/nvss/deaths.htm (CDC / National Center for Health Statistics 2018). 40. For details on the Fragile Families and Child Wellbeing Study, see https:// fragilefamilies.princeton.edu/ (Princeton University 2018). 41. For details on the National Longitudinal Study of Adolescent to Adult Health, see www.cpc.unc.edu/projects/addhealth (UNC Carolina Population Center 2018). 42. For details on the Health and Retirement Study, see http://hrsonline.isr .umich.edu/ (Health and Retirement Study 2017).

Notes to Chapter 2 |

197

chapter 2 1. Murphy, Kochanek, Xu, and Heron 2015; Haines 2001. 2. Mather, Jacobsen, and Pollard 2015. 3. Barquet and Domingo 1997. 4. Omran 1971, 1977. 5. In their classic book Fatal Years: Child Mortality in Late NineteenthCentury America, Samuel Preston and Michael Haines (1991) provided a pioneering overview of U. S. population health history in the late nineteenth and early twentieth centuries, with a focus on child health and mortality. We draw heavily on their historical work in this portion of the chapter. 6. These data are available from the invaluable website of the National Center for Health Statistics (2018). Caution is warranted in the interpretation of cause of death data from 1900 for at least two reasons: (1) the U. S. death registration system was not considered to be complete until the 1930s; and (2) knowledge and documentation of causes of death at the time were far inferior to what is the case today. Nonetheless, these comparative data from 1900 and 2014 provide a useful summary of the U. S. epidemiologic transition. 7. See Omran 1977: 33. 8. Haines 2001. 9. Omran 1977. 10. Haines 1979. 11. See the fact sheet produced by the U. S. Department of Veterans Affairs 2017. 12. The elderly population is, of course, the other especially vulnerable age range. But keep in mind that, prior to the epidemiologic transition, the U. S. population did not have many elderly people; most people were not fortunate enough to live into old age. Moreover, birth rates were very high in historical times. Thus, the Age of Pestilence and Famine, characterized by extremely high levels of infectious diseases, was especially harmful for infants and children. 13. Preston and Haines 1991: table 2–2. 14. Hayward and Gorman 2004. 15. Omran 1977. 16. Haines 2001. 17. U. S. Department of Veterans Affairs 2017. 18. Haines 2004. Unfortunately, there is no historical data available on changes in women’s height during this time. 19. The National Center for Health Statistics provides very useful yearly updates on U. S. life expectancy. In this case, the report being drawn upon is Arias et al. 2017. These reports typically include historical data in addition to the yearly updates. 20. See McKeown (1976) for the most comprehensive outline of his argument. 21. McKinlay and McKinlay 1977: 406. 22. Barquet and Domingo 1997. 23. McKeown 1976. 24. Deaton 2013.

198

|

Notes to Chapter 2

25. Riley 2001. 26. Duffy 1971. 27. Deaton 2013; Preston and Haines 1991. 28. Riley 2001. 29. Preston and Haines 1991. 30. Deaton 2013. 31. Preston and Haines 1991. 32. Deaton 2013. 33. For example, U. S. life expectancy for Whites in rural areas was estimated to be nine years longer in 1900 compared with estimated life expectancy for Whites in urban areas (Preston and Haines 1991: 58; see also Haines 2001). Later in the book, we will discuss the reversal of this pattern in the current era. 34. Duffy 1971. 35. Cutler and Miller 2005. 36. Riley 2001. 37. U. S. Department of Veterans Affairs 2017. 38. Arias et al. 2017. 39. Omran 1977: 43. 40. This evidence is drawn from Hummer et al. 2009: 528. 41. Omran 1971. 42. Fischer and Hout 2006. 43. Preston and Haines 1991: 210. 44. Arias et al. 2017: 45–47. 45. Arias et al. 2017: 45–47. 46. Masters et al. 2014. 47. Omran 1971. 48. Deaton 2013. 49. U. S. Department of Health, Education, and Welfare 1964. 50. Avendano and Kawachi 2014. 51. Omran 1977: 35. 52. See also Olshansky and Ault (1986) regarding this point. They proposed a fourth stage of the epidemiologic transition, The Age of Delayed Degenerative Diseases, largely because continued decreases in mortality rates and continued increases in life expectancy in the late 1960s and 1970s were so unexpected (and welcome) after what had seemed to be the achievement of peak life expectancy of about 70 in 1960. 53. Murphy, Kochanek, Xu, and Heron 2015. 54. Mensah et al. 2017. 55. See, for example, the historical section of the webpage of the American Cancer Society (2017). While the ACS was founded in 1913, its research program did not begin until the late 1940s. Moreover, its reach was limited until the 1960s, when it was important in developing the Surgeon General’s 1964 report on smoking and cancer, and the 1970s, when it contributed to the passage of the National Cancer Act in 1971, which expanded funding for the National Cancer Institute to conduct innovative research on the causes, detection, and treatments of cancer. 56. See also Mensah et al. 2017.

Notes to Chapter 2 |

199

57. For an overview of the Framingham Heart Study, see Framingham Heart Study 2018. 58. See National Center for Health Statistics 2016: table 47. 59. Mensah et al. 2017. 60. Mensah et al. 2017. 61. National Center for Health Statistics 2016: table 47. 62. For a very insightful early examination of smoking behavior and mortality rates using family history record data collected in his laboratory, see Pearl 1938. 63. For fascinating historical overviews of the incredible death toll of tobacco on health and mortality, both in the United States and around the globe, see Ravenholt 1990; Brandt 2007. 64. U. S. Department of Health, Education, and Welfare 1964. 65. U. S. Department of Health and Human Services 2014. 66. Pampel 2003. 67. Jamal et al. 2018. 68. For such estimates, see Fenelon and Preston 2012; Carter et al. 2015; Lariscy et al. 2018. Far from a threat of the past, very aggressive policy and programmatic efforts will be needed to reduce and eventually eliminate the health problems and death tolls caused by smoking. 69. Deaton 2013. 70. See the 2014 data from the United Nations at United Nations Statistics Division 2018. All told, 35 unique countries had a life expectancy at birth of 80 or above as of 2014. 71. Gottlieb et al. 1981. 72. Other means of spread included blood transfusions and mother-to-fetus transmission. 73. See, for example, Olshansky et al. (1998), who proposed yet another stage of the epidemiologic transition, the “Reemergence of Infectious and Parasitic Diseases.” 74. Randy Schilts (1987) most thoroughly documents the apathetic U. S. governmental response to HIV / AIDS in his controversial and gripping book And the Band Played On, which was later developed into an HBO film. 75. Ventura et al. 1997. 76. A related increase in deaths due to liver disease (including alcoholinduced cirrhosis) has also been documented. See Case and Deaton 2015. 77. See particularly Case and Deaton 2015, 2017; Woolf et al. 2018. 78. See the Centers for Disease Control and Prevention website (Centers for Disease Control and Prevention [CDC] 2016). Note that childhood obesity is calculated differently than adults. 79. Ogden et al. 2016. 80. Preston et al. 2014. 81. Bhattacharya, Choudhry, and Lakdawalla 2008. 82. In an important 1980 paper, James Fries argued that U. S. life expectancy would likely peak at 85 and that poor health (morbidity) would be compressed into a smaller portion of old age prior to death. This “compression of morbidity” thesis would undergo rigorous debate in the ensuing decades. As a result, a rich body of literature on both old age mortality and health trends unfolded.

200

|

Notes to Chapter 3

83. Freedman et al. 2004. 84. Freedman et al. 2013. 85. See National Research Council et al. 2013. The other 21 countries in the comparison included Australia, Austria, Belgium, Canada, Denmark, Finland, France, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and West Germany. 86. Sasson 2016. 87. National Research Council et al. 2013. 88. For some reviews of this massive literature, see Braveman et al. 2010; Elo 2009; Pampel, Krueger, and Denney 2010. 89. Rostron, Boies, and Arias 2010. 90. For the most thorough documentation of socioeconomic differences in U. S. population health as of 1960, see Kitagawa and Hauser 1973. 91. See, for example, the health policy ideas put forward by James House (2015).

chapter 3 1. www.nytimes.com/2009/11/05/opinion/05kristof.html. 2. World Health Organization 2018b. 3. The 16 peer countries are Australia, Austria, Canada, Denmark, Finland, France, Germany, Italy, Japan, Norway, Portugal, Spain, Sweden, Switzerland, the Netherlands, and the United Kingdom. These countries were also selected because reliable population health data could be obtained. 2012 is the most recent year for which mortality rates by cause of death were available for all 16 countries. 4. Ho and Hendi 2018. 5. Ho 2013. 6. National Research Council et al. 2013. 7. Avendano et al. 2009; Banks et al. 2006; Reynolds et al. 2008; Thorpe et al. 2007. 8. National Research Council et al. 2011. 9. Foutz et al. 2017. 10. Macinko, Starfield, and Shi 2003; Institute of Medicine 2014. 11. Chetty et al. 2016. 12. As Preston and Ho (2009) point out, research has ranked the United States last among high-income countries on mortality amenable to health care, but the categorization of causes of death “amenable” to health care has been seriously challenged. 13. Wolf-Maier et al. 2004. 14. Moise 2003; Kaul et al. 2004; Moon et al. 2003. 15. Preston and Ho 2009. 16. Gatta et al. 2000; Verdecchia et al. 2007 17. Preston 2010; National Research Council et al. 2013. 18. Alley et al. 2010. 19. Forey et al. 2002; Garrett et al. 2011; Pampel 2010.

Notes to Chapter 3 |

201

20. Preston et al. 2010. 21. Preston et al. 2010. 22. Preston et al. 2010. 23. National Research Council et al. 2013. 24. To calculate your Body Mass Index using weight in pounds and height in inches, divide your weight in pounds by your height in inches squared, and multiply by 703. 25. Ogden and Carroll 2010. 26. National Research Council et al. 2011: Figure 3–2. 27. Hu 2008; National Research Council et al. 2011. 28. National Research Council et al. 2011. 29. Alley et al. 2010. 30. National Research Council et al. 2011. 31. National Research Council et al. 2011. 32. World Health Organization 2017. 33. National Research Council et al. 2013. 34. United Nations Office on Drugs and Crime 2010. 35. National Research Council et al. 2013. 36. Needleman 2004. 37. Hanna-Attisha et al. 2016. 38. Brunekreef and Holgate 2002. 39. French, Story, and Jeffrey 2001. 40. Cutter et al. 2003; Gordon-Larsen et al. 2006; LaVeist et al. 2011. 41. Brulle and Pellow 2006. 42. World Bank Group 2017; Hepburn and Hemenway 2004; Richardson and Bae 2004. 43. Baldasano et al. 2003; Parker et al. 2011. 44. National Research Council et al. 2013. 45. Organization for Economic Co-operation and Development 2017. 46. Organization for Economic Co-operation and Development 2017. 47. National Research Council et al. 2011. 48. For reviews, see Pickett and Wilkinson 2015; Subramanian and Kawachi 2004. 49. Banks et al. 2006; Avedano et al. 2009, 2011. 50. Avendano et al. 2011; Banks et al. 2006. 51. Preston 1975. 52. Chetty et al. 2016; Subramanian and Kawachi 2004. 53. Pickett and Wilkinson 2015. 54. Chetty et al. 2016; Subramanian and Kawachi 2004. 55. Kondo et al. 2009; Lochner et al. 2001; Subramanian and Kawachi 2004; Wolfson et al. 2000. 56. Subramanian and Kawachi 2006. 57. Pickett and Wilkinson 2015. 58. Lancee and Van de Worfhorst 2012; Paskov and Dewilde 2012. 59. Kawachi and Berkman 2000. 60. Pickett and Wilkinson 2015. 61. Deaton and Paxson 2001; Currie and Schwandt 2016.

202

|

Notes to Chapter 4

62. Kitagawa and Hauser 1973; Crimmins et al. 2009. 63. Pickett and Wilkinson 2015. 64. Avendano and Kawachi 2014; National Research Council et al. 2011, 2013. 65. Karas Montez 2017.

chapter 4 1. For the Fairfax–Boone County comparison, we draw on statistics from http://countyhealthrankings.org (University of Wisconsin Population Health Institute 2017); https://vizhub.healthdata.org/subnational/usa (University of Washington Institute for Health Metrics and Evaluation 2018c); http://maps .gcir.org/ (Grantmakers Concerned With Immigrants 2009); http://politico.com (“POLITICO” 2018); www.cdc.gov/drugoverdose/data/statedeaths.html (Centers for Disease Control and Prevention [CDC] 2018); www.cdc.gov/Community Health/profile/currentprofile/VA/Fairfax/. 2. World Health Organization 2016. 3. Wang et al. 2013. 4. Singh and Siahpush 2014; Singh, Kogan, and Slifkin 2017. 5. Murray et al. 2006. 6. Wilmoth, Boe, and Barbieri 2010. 7. Wilmoth, Boe, and Barbieri 2010. It is a well-known tendency in geographic analyses that there is more variability and lower correlation between variables when using smaller units of analysis (such as counties versus states). 8. Henry J. Kaiser Family Foundation 2018. 9. Mathews, Ely, and Driscoll 2018. 10. University of Washington Institute for Health Metrics and Evaluation 2018b. 11. Karas Montez 2017. 12. Arcaya et al. 2012; Karas Montez, Hayward, and Wolf 2017. 13. Karas Montez, Hayward, and Wolf 2017. 14. Karas Montez, Zajacova, and Hayward 2016. 15. Cutler and Miller 2005. 16. Singh and Siahpush 2014. 17. Cossman et al. 2010; James 2014. 18. Singh and Siahpush 2014; Singh et al. 2017. 19. James 2014. 20. James 2014. 21. Anderson et al. 2015; Galea and Vlahov 2005; Lawrence, Hummer, and Harris 2017. 22. See Ross and Mirowsky 2008: 169. There are many other papers on this topic, using similar “multilevel methodology” (i.e., individuals clustered within neighborhoods). Stephanie Robert (1999) provided a very thorough review of such studies. This area of study has since moved on to more advanced theoretical and methodological work, as discussed below. 23. Arcaya et al. 2016.

Notes to Chapter 5 |

203

24. Morenoff 2003. 25. Diez Roux 2001; Oakes et al. 2015. 26. Leventhal and Brooks-Gunn 2003. 27. Ludwig et al. 2013. 28. Kessler et al. 2014; Leventhal and Brooks-Gunn 2003; Ludwig et al. 2013. 29. Entwisle 2007; Harris and McDade 2018. 30. Umberson and Karas Montez 2010. 31. Umberson and Karas Montez 2010. 32. Umberson and Karas Montez 2010. 33. Pampel 2006. 34. National Longitudinal Study of Adolescent to Adult Health; see www .cpc.unc.edu/projects/addhealth (UNC Carolina Population Center 2018). 35. Mollborn et al. 2014. 36. Bearman and Moody 2004. 37. See, for example, the work of Christakis and Fowler 2009. 38. Krueger and Burgard 2011. 39. Moen et al. 2016. 40. Lillard and Waite 1995; Waite 1995. 41. Umberson et al. 2006; Liu and Waite 2014. 42. Braveman et al. 2010. 43. Zimmer et al. 2007. 44. Friedman and Mare 2014. 45. Christakis and Allison 2006.

chapter 5 1. Ryan and Bauman 2016. 2. These figures are taken from the U. S. Census Bureau’s annual report on income and poverty; see Semega, Fontenot, and Kollar 2017. The poverty line is a complex measure that is updated by the Census Bureau each year, based on the cost of specific goods, inflation, the number of people living in households, and more. The poverty line was $12,228 for a person living in a household by herself or himself in 2016; for a household of four persons, the poverty line was $24,563. These very low cut-points make it easy to grasp that living in poverty in the United States is not easy, to say the least. 3. Fundamental cause theory was originally developed by Link and Phelan (1995). In doing so, they drew heavily on the work of House, Kessler, and Herzog 1990; House et al. 1994; Lieberson 1985; Williams 1990. 4. See Glymour, Avendano, and Kawachi (2014) for an insightful discussion of socioeconomic status and health. This includes remarks on the definition of SES and its distinction from the epidemiological concept of socioeconomic position. We draw on their conceptual ideas in this section of our chapter. 5. See the chapter by Marx entitled “Classes in Capitalism and Pre-Capitalism” in the edited volume by Grusky and Weisshaar (2014). 6. See the chapter by Weber entitled “Class, Status, Party” in the edited volume by Grusky and Weisshaar (2014).

204

|

Notes to Chapter 5

7. See the very informative review of SES and health in the Annual Review of Sociology by Elo (2009). Galobardes, Lynch, and Smith (2007) and Chandola and Marmot (2010) have also provided very thorough reviews of this literature. 8. Elo 2009. 9. Mirowsky and Ross 2003a. 10. See Hummer and Hernandez (2013) for a short discussion of what we do not know regarding educational attainment and health in the United States. Even though the literature in this area of study is very large, we may know less than we think we do given that our measures of education are often onedimensional (e.g., years of schooling) and thus do not tap into the complex educational careers of most individuals. 11. Marmot and colleagues have written an extensive set of papers using the Whitehall Study data. The key findings and arguments from this body of work are summarized in Marmot’s 2005 book The Status Syndrome. 12. But see Burgard and Lin (2013) for an excellent overview on studies in this area. 13. Burgard and Lin 2013. 14. Raj Chetty and colleagues have built an amazing data set that links individual income data from over 1 billion U. S. tax records to follow-up mortality data for the people who are included in the tax records. Their estimates of life expectancy by income, published in Chetty et al. (2016), not only demonstrate huge differences in life expectancy by income among American adults, but also that such disparities vary by geographic location and have widened between 1999 and 2014. Thus, for example, it is far worse to have a low income in Detroit than in San Francisco, possibly due to the relative lack of governmental programs for poor people in Detroit relative to San Francisco. The growing income gap in life expectancy during the twenty-first century is also troubling; it suggests that while people living in high-income households are living longer (and most likely healthier) lives than ever before, those with low incomes are living somewhat shorter lives than they were living a couple of decades ago. 15. Galobardes, Lynch, and Smith 2007. 16. Braveman et al. 2010. 17. Kawachi, Adler, and Dow (2010) provide an excellent summary of these thorny issues. 18. See, e.g., Smith 2004. 19. See http://hrs.isr.umich.edu for detailed information about the HRS (Health and Retirement Study 2017). 20. Chetty et al. 2016; see also note 17. 21. Braveman et al. 2010. 22. Keister 2000. 23. Urban Institute 2017. 24. Pollack et al. 2007. 25. Walsemann, Gee, and Gentile 2015. 26. Piketty and Saez 2014. 27. Semega, Fontenot, and Kollar 2017. 28. Desilver 2013. 29. Sherman 2015.

Notes to Chapter 5 |

205

30. Piketty and Saez 2014. 31. Massey 2007. 32. Mitnik, Cumberworth, and Grusky 2015. 33. Piketty and Saez 2014. 34. D. Brady 2009; Piketty and Saez 2014. 35. Piketty and Saez 2014: 838. 36. Iceland 2013. 37. Massey 2007. 38. Iceland 2014. 39. Massey 2007. 40. Iceland 2014; for details on the TCJA, see K. Brady 2017. 41. Moffitt 2015. 42. Moffitt 2015. 43. Massey 2007. 44. Reardon and Bischoff 2011. 45. Massey 2007. 46. Kitagawa and Hauser 1973. 47. Feldman et al. 1989; Pappas et al. 1993. 48. Preston and Elo 1995. 49. Feldman et al. 1989. 50. Pampel 2005. 51. National Academies of Sciences, Engineering, and Medicine et al. 2015. 52. Zajacova and Lawrence 2018. 53. Hayward, Hummer, and Sasson 2015. 54. Hamilton et al. 2015. 55. We also estimated disparities in these same measures of health by level of family income. The income-based disparities were similar to those of the education-based disparities; thus, we show and discuss only the education disparities here. 56. Hayward, Hummer, and Sasson 2015. 57. Farmer and Ferraro 2005. 58. Montez, Hummer, and Hayward 2012. 59. Williams 1990. 60. Williams 1990: 91. 61. House and Robbins 1983; House, Kessler, and Herzog 1990; House et al. 1994. 62. Montez and Hayward 2014. 63. Link and Phelan’s 1995 paper in the Journal of Health and Social Behavior provided their initial statement of fundamental cause theory. Subsequently, they updated the theory in many other papers, including Link and Phelan 1996; Link and Phelan 2002; Link 2008; Phelan, Link, and Tehranifar 2010; Phelan and Link 2015. Lutfey and Freese (2005) also provide important insights into fundamental cause theory. We draw from all of these papers in this discussion. 64. This diagram is based on our interpretation of fundamental cause theory. 65. See, e.g., Pampel, Krueger, and Denney 2010; Skalamera and Hummer 2016. 66. This point is made especially clear in Phelan, Link, and Tehranifar 2010.

206

|

Notes to Chapter 6

67. Phelan, Link, and Tehranifar 2010. 68. See, e.g., Hayward, Hummer, and Sasson (2015) for a discussion of educational disparities in U. S. health within a society that is undergoing very rapid social and technological change. 69. Link 2008. 70. Baker et al. 2017. 71. See also Hayward, Hummer, and Sasson 2015. 72. Behrman et al. 2011. 73. Karas Montez, Zajacova, and Hayward 2016; Karas Montez, Hayward, and Wolf 2017; Karas Montez, Zajacova, and Hayward 2017. 74. Boardman, Domingue, and Daw 2015. 75. Lleras-Muney 2005. 76. Link 2008. 77. For an example of a propensity score modeling approach to educational disparities in health that helps approximate a causal model, see Lawrence 2017. For an example of a natural experiment research design that utilizes changes in compulsory education laws to understand educational effects on adult mortality, see Lleras-Muney 2005. 78. See, for example, Harris et al. 2013. 79. For an example of SES working through biological mechanisms to influence adult physical functioning across the life course, see Yang et al. 2017.

chapter 6 1. Levine et al. 2001. 2. Tavernise 2016. 3. Chang 2010; Harris-Lacewell 2010; P. J. Williams 2010. 4. For example, a study found that a sizeable proportion of adolescents respond differently to the same question of racial identity at home and school (Harris and Sim 2002). 5. Lewontin 1974. 6. D. R. Williams and Mohammed 2013. 7. Rodríguez 2000; Campbell and Rogalin 2006. 8. Humes, Jones, and Ramirez 2011. 9. Bryc et al. 2015. 10. Okamoto and Mora 2014. 11. De la Cruz and Brittingham 2003. 12. U. S. Census Bureau 2018. 13. Mathews et al. 2017. 14. U. S. Census Bureau 2018. 15. Pew Research Center 2015. 16. Henry J. Kaiser Family Foundation 2014. 17. Murray et al. 2006. 18. Woolf et al. 2018. 19. The 19 conditions include self-reported health, physical limitations, 14 or more physically unhealthy days in past 30 days, 14 or more mentally

Notes to Chapter 6 |

207

unhealthy days in past 30 days, adult obesity, child obesity, adult asthma, child asthma, diabetes diagnosis, heart disease diagnosis, HIV diagnosis, AIDS diagnosis, HIV death rate, cancer incidence rate, breast cancer rate, colorectal cancer rate, lung cancer rate, preterm birth rate, and the rate of low birth weight. Data limitations prevented the separate analysis of Native Hawaiians and Other Pacific Islanders; on most outcomes they are grouped with Asians. 20. Fenelon 2013b. 21. Arias 2010; Elo et al. 2004. 22. Artiga et al. 2016b; Zhang, Hayward, and Lu 2012. 23. Artiga et al. 2016b; Hayward et al. 2014. 24. Hayward et al. 2014. 25. Harper et al. 2007. 26. Harper et al. 2007; Harper, Rushani, and Kaufman 2012; Woolf et al. 2018. 27. In addition to drawing heavily on work on fundamental causality (Link and Phelan 1995; Phelan and Link 2015), we build on earlier models by Hummer (1996) and Williams (D. R. Williams and Mohammed 2013). 28. Hummer 1996; Phelan and Link 2015. 29. Rugh and Massey 2014. 30. Rugh and Massey 2014. 31. Massey and Denton 1993. 32. Studies find evidence of racial housing discrimination in lending, credit score determinations, real estate advertising (Yinger 1995), in access to real estate agents and landlords (Massey and Lundy 2001), in providing information about and showing homes to renters and buyers by real estate agents (i.e., racial steering). 33. Turner et al. 2002; Ross and Yinger 2002. 34. Charles 2003. 35. Alba, Logan, and Stults 2000; Logan et al. 1996. 36. Cutler and Glaeser 1997. 37. Massey 1990; Quillian 2012. 38. Rugh and Massey 2010. 39. LaVeist and Wallace 2000; Hofrichter 2000; Moore et al. 2008; MorelloFrosch and Jesdale 2006; Morello-Frosch and Lopez 2006; Institute of Medicine 2003. 40. King, Morenoff, and House 2011. 41. Pager 2003; Pedulla 2014. 42. Downey and Pribesh 2004. 43. Wakefield and Uggen 2010. 44. Pettit and Western 2004: 164. 45. From Pettit and Western 2004: Western, Kling, and Weiman 2001; Hagan and Dinovitzer 1999; Sampson and Laub 1993; Uggen and Manza 2002; Hirsch et al. 2002. 46. Federal Bureau of Investigation n.d. 47. Bobo et al. 2012. 48. Jones 2016.

208

|

Notes to Chapter 6

49. Correll et al. 2002. 50. Steele and Aronson 1995; Steele 1997. 51. Karlsen and Nazroo 2002; Paradies 2006; D. R. Williams, Neighbors, and Jackson 2003. 52. Zong, Batalova, and Hallock 2018. 53. Johnson, Bersin, and Rosenblum 2016. 54. The Pew Research Center estimates the undocumented population in 2014 at 11.1 million; the total immigrant population was estimated to be 42.4 million (Zong, Batalova, and Hallock 2018; Passel and Cohn 2016). Zong, Batalova, and Hallock (2018) also provided naturalization statistics. 55. National Academies of Sciences, Engineering, and Medicine et al. 2016. 56. National Academies of Sciences, Engineering, and Medicine et al. 2016. 57. Portes and Rivas 2011. 58. Eschbach, Mahnken, and Goodwin 2005; Hummer et al. 1999, 2000; Singh and Siahpush 2002. 59. Carrasquillo, Carrasquillo, and Shea 2000; Ku 2009. 60. Markides and Coreil 1986. 61. Turra and Elo 2008. 62. Hummer et al. 2007. 63. Jasso et al. 2004. 64. Puerto Rico was the exception, probably because of its status as a U. S. Commonwealth and Puerto Ricans’ facility migrating as U. S. citizens (Feliciano 2005). 65. Ro, Fleischer, and Blebu 2016; Akresh and Frank 2008. 66. Riosmena, Kuhn, and Jochem 2017. 67. Ro, Fleischer, and Blebu 2016. 68. Bond Huie, Hummer, and Rogers 2002; Eschbach, Mahnken, and Goodwin 2005; Kimbro 2009. 69. Riosmena, Kuhn, and Jochem 2017. 70. Jasso et al. 2004. 71. Regarding short-term improvements, see Teitler, Hutto, and Reichman 2012. 72. Angel et al. 2010; Cho et al. 2004; Choi 2012; Finch, Do, and Frank 2009; Hamilton et al. 2011; Singh and Siahpush 2002. 73. Akresh 2007; Lopez-Gonzalez, Aravena, and Hummer 2005. 74. Breslau et al. 2006; González et al. 2010; Kessler et al. 2005; D. R. Williams et al. 2007. 75. Rates were lower (than African Americans’) for Chinese, Filipino, and Vietnamese respondents, and higher for Cuban, Puerto Rican, and Black Caribbean respondents (González et al. 2010). 76. Lorant et al. 2003; Mezuk et al. 2010. 77. Breslau et al. 2005. 78. Mezuk et al. 2013; D. R. Williams et al. 2007. 79. Mishra et al. 2009. 80. Chatters et al. 2009.

Notes to Chapter 7 |

209

81. Ellison 1995; Koenig and Larson 2001; D. R. Williams and Sternthal 2010; Strawbridge et al. 2001. 82. Jackson, Knight, and Rafferty 2010. 83. Jackson, Knight, and Rafferty 2010. 84. Breslau et al. 2006. 85. Riolo et al. 2005. 86. Breslau et al. 2005; González et al. 2010; D. R. Williams et al. 2007. 87. Gelman and Auerbach 2016; Montez and Zajacova 2013. 88. Case and Deaton 2015. 89. Meara and Skinner 2015. 90. Woolf et al. 2018 91. Masters, Tilstra, and Simon 2017. 92. Case and Deaton 2015; Centers for Disease Control and Prevention (CDC) 2013. 93. Centers for Disease Control and Prevention (CDC) 2013. 94. Pletcher et al. 2008; Singhal, Tien, and Hsia 2016. 95. Han et al. 2015. 96. Yet others have questioned that the causes of death necessarily indicate a common cause, as changes in mortality rates from suicide, drug-related deaths, and other external causes from 1980 to 2013 show distinct historical patterns, with suicides peaking after the 2007 Great Recession, and drug-related deaths rising steadily across the period (Masters, Tilstra, and Simon 2017). 97. Case and Deaton 2015: 15081. 98. Chetty et al. 2017. 99. Hummer and Gutin 2018. 100. Hummer and Hayward 2015.

chapter 7 1. National Research Council et al. 2013. 2. Karas Montez et al. 2011. 3. Case and Deaton 2015; Gelman and Auerbach 2016. 4. Fausto-Sterling 1992: 269. 5. Population Reference Bureau 2018. 6. Population Reference Bureau 2018. 7. Preston and Wang 2006. 8. Note that the death rates are graphed on the logarithmic scale in order to make differences visible at the lowest levels of mortality. 9. Xu et al. 2016. 10. Some transmen (i.e., individuals assigned the female sex at birth but who identify as men) may die from causes related to pregnancy and childbirth. However, it is unclear how the gender of transmen and transwomen is coded on death certificates. If transmen who die from pregnancy- and childbirth-related causes are coded as men on death certificates, the CDC data indicate that there were fewer than 20 of them in this age group in 2014. 11. Crimmins, Kim, and Hagedorn 2002. 12. Liu 2014.

210

|

Notes to Chapter 7

13. Kessler 2004; Mirowsky and Ross 2003b; Regier et al. 1993. 14. Kessler and Zhao 1999. 15. Preston and Wang 2006. 16. Read and Gorman 2010a, 2010b. 17. Case and Paxson 2005. 18. Rosenfeld and Mouzon 2013. 19. Crimmins, Kim, and Solé-Auró 2011; Gorman and Read 2006. 20. Case and Paxson 2005. 21. Verbrugge 1989; Rogers, Hummer, and Nam 2000; Rogers et al. 2010. 22. Rogers, Hummer, and Nam 2000; Rogers et al. 2010. 23. Gorman and Read 2006. 24. Harrison 1978. 25. Newman and Brach 2001. 26. Owens 2002. 27. Bird and Rieker 1999. 28. Bird and Rieker 1999; Rieker, Bird, and Lang 2010. 29. For example, see Rogers et al. 2010. 30. Rogers et al. 2010. 31. Harris 2010; Johnson, Greaves, and Repta 2009. 32. Edwards, Wetzel, and Wyner 2006. 33. Grundtvig et al. 2009. 34. Bird and Rieker 1999. One exception may be women’s elevated risk of autoimmune disorders. 35. Read and Gorman 2010a, 2010b. 36. Link and Phelan 1995; Phelan, Link, and Tehranifar 2010. 37. Kawachi et al. 1999. 38. Schnittker 2007. 39. Chen et al. 2005; Denton, Prus, and Walters 2004; Mirowsky 1996; Ross and Mirowsky 2002, 2006. 40. Bird 1999; Pavalko and Woodbury 2000. 41. Pavalko and Woodbury 2000. 42. Bird 1999. 43. Andoh et al. 2006; Leon 2011. 44. Berin, Stolnitz, and Tenebein 1989; Omran 1971; Noymer and Garenne 2000. The trend of a growing female mortality advantage across the epidemiologic transition was interrupted in the United States by the Spanish flu epidemic in 1918, which narrowed the gender mortality gap as a result of the flu’s disproportionate impact on men with tuberculosis; men’s mortality during the epidemic precluded their later deaths and narrowed the gender mortality gap through about 1930. 45. Miller and Gerstein 1983; Preston 1970; Retherford 1975. 46. Retherford 1975. 47. Pampel 2002. 48. See chapter 4. 49. Rogers, Hummer, and Nam 2000; Rogers et al. 2010. 50. Trovato and Lalu 1996; Waldron 1993.

Notes to Chapter 7 |

211

51. United Nations Statistical Office 1991: 56. 52. Preston and Wang 2006; see also Pampel 2001, 2002. Cohort-driven change in smoking reflects how historical (i.e., period) changes in information about the impacts of smoking, smoking-related policy, cost of tobacco, and smoking culture (i.e., the “social context of smoking”) impact age groups differently. In particular, the social context of smoking has a greater impact on age groups making decisions about whether to take up smoking (adolescents and young adults). 53. Preston and Wang 2006: fig. 1. 54. Pampel 2002. 55. Retherford 1972. 56. Rogers et al. 2010; see also Preston and Wang 2006. 57. Pampel 2001, 2002. 58. Pampel 2001, 2002. 59. Connell 2012. 60. Courtenay 2000. 61. Neff, Prihoda, and Hoppe 1991; Mahalik, Burns, and Syzdek 2007. 62. Courtenay 2000. 63. Courtenay 2000. 64. Potts, Burnam, and Wells 1991. 65. Case and Paxson 2005. 66. Crimmins, Kim, and Vasunilashorn 2010. 67. Case and Paxson 2005. 68. Connell 2012; Courtenay 2000. 69. Bird and Rieker 1999. 70. Fausto-Sterling 2005; Verbrugge 1989. 71. Preston and Wang 2006. 72. Karas Montez et al. 2011. 73. Stroumsa 2014. In 2012, the board of the American Psychiatric Association (APA) voted to eliminate Gender Identity Disorder (GID), which pathologized gender nonconformity as a mental disorder, from the Diagnostic and Statistical Manual of Mental Disorders (DSM) and introduced gender dysphoria, or the distress resulting from the incongruence between one’s gender and one’s sex assignment, in its place. Although a diagnosis of gender dysphoria facilitates access to and coverage for treatment (such as hormone therapy and sex reassignment surgery) for transgender individuals, there is controversy over its categorization as a mental illness, which has the effect of locating the problem within the individual as opposed to within a discriminatory society. 74. Stroumsa 2014. 75. Flores, Brown, and Herman 2016. 76. Gates 2011. 77. Clements-Nolle et al. 2001; Garofalo et al. 2006; Herbst et al. 2008. 78. Conron et al. 2012. 79. See http://transpop.org. 80. See http://ustranssurvey.org. 81. See http://projectaffirm.org.

212

|

Notes to Chapter 8

82. Mayer et al. 2008. 83. Due to real limits on how gender identity and sexual orientation are measured, queer, intersex, and asexual people are rarely identified in population data. 84. LGBT-QIA stands for Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, and Asexual. 85. Conron, Mimiaga, and Landers 2010; Cochran and Mays 2000; Cochran, Mays, and Sullivan 2003; Drabble, Midanik, and Trocki 2005; Gilman et al. 2001; Ward et al. 2014. 86. Ward et al. 2014. 87. Conron, Mimiaga, and Landers 2010. 88. Denney, Gorman, and Barrera 2013; Liu, Reczek, and Brown 2013. 89. Brown 2000; Waite and Gallagher 2000. 90. Reczek et al. 2016, 2017. 91. Denney, Gorman, and Barrera 2013; Liu, Reczek, and Brown 2013.

chapter 8 1. United Nations 1948. 2. House 2015; Schroeder 2007. 3. Case and Deaton 2015, 2017. 4. See www.kff.org/uninsured/fact-sheet/key-facts-about-the-uninsuredpopulation/ (Henry J. Kaiser Family Foundation 2017). 5. Billing and insurance-related administrative costs make up 18%; other administrative activities, such as maintaining medical records and scheduling, make up 9.4% (Jiwani et al. 2014). 6. Cheng 2015. 7. See www.nhi.gov.tw/english/ (National Health Insurance Administration 2015). 8. Life expectancy data taken from the Population Reference Bureau Data Finder at www.prb.org/DataFinder (Population Reference Bureau 2018); GDP per capita data taken from the CIA World Factbook at www.cia.gov/library /publications/the-world-factbook/rankorder/2004rank.html (Central Intelligence Agency 2018). 9. Henry J. Kaiser Family Foundation 2018. 10. See, for example, the project on this topic led by Jennifer Karas Montez at https://jennkarasmontez.com/research/ (Karas Montez 2016). 11. House 2015. 12. U. S. Department of Health and Human Services 2014. 13. U. S. Department of Health and Human Services 2014. 14. U. S. Department of Health and Human Services 2014. 15. See, for example, the review article by Brownson, Haire-Joshu, and Luke 2006. 16. Pampel, Krueger, and Denney 2010. 17. Link and Phelan 1995. 18. House 2015; see also House’s co-edited volume (Schoeni et al. 2008), entitled Making Americans Healthier: Social and Economic Policy as Health Policy. This edited volume provides an overview of educational, income, civil

Notes to Chapter 8 |

213

rights, employment, welfare, and housing policies that have been shown to have important effects on population health in the United States. Nonetheless, such areas of policy are rarely considered to be associated with health. 19. Hayward, Hummer, and Sasson 2015. 20. Hummer and Hernandez 2013; Kawachi, Adler, and Dow 2010; Smith 2007. 21. Schoeni et al. 2008. 22. Avendano and Kawachi 2014.

References

Abrego, Leisy J. 2018. “Renewed Optimism and Spatial Mobility: Legal Consciousness of Latino Deferred Action for Childhood Arrivals Recipients and Their Families in Los Angeles.” Ethnicities 18 (2): 192–207. Adler, Nancy E., Christine Bachrach, Dorothy Daley, and Michelle Frisco. 2013. Building the Science for a Population Health Movement. Washington, DC: Institute of Medicine of the National Academies. www.nam.edu/wpcontent/uploads/2015/06/BPH-BuildingTheScience.pdf. Akresh, Ilana Redstone. 2007. “Dietary Assimilation and Health among Hispanic Immigrants to the United States.” Journal of Health and Social Behavior 48 (4): 404–17. https://doi.org/10.1177/002214650704800405. and Reanne Frank. 2008. “Health Selection among New Immigrants.” American Journal of Public Health 98 (11): 2058–64. https://doi.org/10 .2105/AJPH.2006.100974. Alba, Richard D., John R. Logan, and Brian J. Stults. 2000. “The Changing Neighborhood Contexts of the Immigrant Metropolis.” Social Forces 79 (2): 587–621. https://doi.org/10.1093/sf/79.2.363. Alley, Dawn E., Jennifer Lloyd, and Michelle Shardell. 2010. “Can Obesity Account for Cross-National Differences in Life-Expectancy Trends?” In International Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 164–92. Washington, DC: National Academies Press. https://doi.org/10.17226/12945. American Cancer Society. 2017. “Our History.” July 13. www.cancer.org /about-us/who-we-are/our-history.html. Amuedo-Dorantes, Catalina, and Francisca Antman. 2017. “Schooling and Labor Market Effects of Temporary Authorization: Evidence from DACA.” Journal of Population Economics 30 (1): 339–73.

215

216

|

References

Anderson, Timothy J., Daniel M. Saman, Martin S. Lipsky, and M. Nawal Lutfiyya. 2015. “A Cross-Sectional Study on Health Differences between Rural and Non-Rural U. S. Counties Using the County Health Rankings.” BMC Health Services Research 15: 441–49. https://doi.org/10.1186/s12913–015– 1053–3. Andoh, Samuel Y., Masahiro Umezaki, Keiko Nakamura, M. Kizuki, and Takehito Takano. 2006. “Correlation between National Income, HIV / AIDS and Political Status and Mortalities in African Countries.” Public Health 120 (7): 624–33. https://doi.org/10.1016/j.puhe.2006.04.008. Angel, Ronald J., Jacqueline L. Angel, Carlos Díaz Venegas, and Claude Bonazzo. 2010. “Shorter Stay, Longer Life: Age at Migration and Mortality among the Older Mexican-Origin Population.” Journal of Aging and Health 22 (7): 914–31. https://doi.org/10.1177/0898264310376540. Arcaya, Mariana C., Mark Brewster, Corwin M. Zigler, and S. V. Subramanian. 2012. “Area Variations in Health: A Spatial Multilevel Modeling Approach.” Health and Place 18 (4): 824–31. https://doi.org/10.1016/j.healthplace.2012 .03.010. Arcaya, Mariana C., Reginald D. Tucker-Seeley, Rockli Kim, Alina SchnakeMahl, Marvin So, and S. V. Subramanian. 2016. “Research on Neighborhood Effects on Health in the United States: A Systematic Review of Study Characteristics.” Social Science and Medicine 168: 16–29. https://doi .org/10.1016/j.socscimed.2016.08.047. Arias, Elizabeth. 2010. “United States Life Tables by Hispanic Origin.” Vital and Health Statistics: Series 2, Data Evaluation and Methods Research 2 (152): 1–33. Hyattsville, MD: National Center for Health Statistics. www .ncbi.nlm.nih.gov/pubmed/21313788. . 2012. “United States Life Tables, 2008.” National Vital Statistics Reports 61 (3): 1–63. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/24974590. . 2015. “United States Life Tables, 2011.” National Vital Statistics Reports 64 (11): 1–63. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/26460931. , Melonie Heron, and Jiaquan Xu. 2017. “United States Life Tables, 2013.” National Vital Statistics Reports 66 (3): 1–64. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/28437241. Artiga, Samantha, Julia Foutz, Elizabeth Cornachione, and Rachel Garfield. 2016a. “Key Facts on Health and Health Care by Race and Ethnicity.” Henry J. Kaiser Family Foundation, June 7. www.kff.org/disparities-policy /report/key-facts-on-health-and-health-care-by-race-and-ethnicity/. . 2016b. “Key Facts on Health and Health Care by Race and Ethnicity—Tables.” Henry J. Kaiser Family Foundation, June 7. www.kff.org /report-section/key-facts-on-health-and-health-care-by-race-and-ethnicityappendix/. Asad, A. L., and M. Clair. 2018. “Racialized Legal Status as a Social Determinant of Health.” Social Science and Medicine 199: 19–28. Avendano, Mauricio, M. Maria Glymour, James Banks, and Johan P. Mackenbach. 2009. “Health Disadvantage in US Adults Aged 50 to 74 Years: A

References | 217

Comparison of the Health of Rich and Poor Americans with That of Europeans.” American Journal of Public Health 99 (3): 540–48. https://doi .org/10.2105/AJPH.2008.139469. Avendano, Mauricio, and Ichiro Kawachi. 2014. “Why Do Americans Have Shorter Life Expectancy and Worse Health Than Do People in Other HighIncome Countries?” Annual Review of Public Health 35 (January): 307–25. https://doi.org/10.1146/annurev-publhealth-032013–182411. Avendano, Mauricio, Renske Kok, Maria Glymour, Lisa Berkman, Ichiro Kawachi, Anton Kunst, Johan Mackenbach, and with support from members of the Eurothine Consortium. 2011. “Do Americans Have Higher Mortality Than Europeans at All Levels of the Education Distribution?: A Comparison of the United States and 14 European Countries.” In International Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 313–32. Washington, DC: National Academies Press. https://doi.org/10.17226/12945. Baker, David P., William C. Smith, Ismael G. Muñoz, Haram Jeon, Tian Fu, Juan Leon, Daniel Salinas, and Renata Horvatek. 2017. “The Population Education Transition Curve: Education Gradients across Population Exposure to New Health Risks.” Demography 54 (5): 1873–95. https://doi .org/10.1007/s13524–017–0608–9. Baldasano, Jose M., E. Valera, and Pedro Jiménez. 2003. “Air Quality Data from Large Cities.” Science of the Total Environment 307 (1–3): 141–65. https://doi.org/10.1016/S0048–9697(02)00537–5. Banks, James, Michael Marmot, Zoe Oldfield, and James P. Smith. 2006. “Disease and Disadvantage in the United States and in England.” JAMA: Journal of the American Medical Association 295 (17): 2037–45. https://doi .org/10.1001/jama.295.17.2037. Barquet, Nicolau, and Pere Domingo. 1997. “Smallpox: The Triumph over the Most Terrible of the Ministers of Death.” Annals of Internal Medicine 127 (8 Pt 1): 635–42. www.ncbi.nlm.nih.gov/pubmed/9341063. Bean, Frank D., Susan K. Brown and James D. Bachmeier. 2015. Parents without Papers: The Progress and Pitfalls of Mexican American Integration. New York: Russell Sage Foundation. Bearman, Peter S., and James Moody. 2004. “Suicide and Friendships among American Adolescents.” American Journal of Public Health 94 (1): 89–95. https://doi.org/10.2105/AJPH.94.1.89. Behrman, Jere R., Hans-Peter Kohler, Vibeke Myrup Jensen, Dorthe Pedersen, Inge Petersen, Paul Bingley, and Kaare Christensen. 2011. “Does More Schooling Reduce Hospitalization and Delay Mortality? New Evidence Based on Danish Twins.” Demography 48 (4): 1347–75. https://doi.org/10.1007 /s13524–011–0052–1. Berin, Barnet N., George J. Stolnitz, and A. Tenebein. 1989. “Mortality Trends of Males and Females over the Ages.” Transactions of Society of Actuaries 41: 9–32. Bhattacharya, Jay, Kavita Choudhry, and Darius Lakdawalla. 2008. “Chronic Disease and Severe Disability among Working-Age Populations.” Medical Care 46 (1): 92–100. https://doi.org/10.1097/MLR.0b013e3181484335.

218

|

References

Bigman, Cabral, Iaphs Staff, Mia DeFino, and Jackie White Hughto. 2018. “Interdisciplinary Association for Population Health Science.” IAPHS— Interdisciplinary Association for Population Health Science. https://iaphs .org/. Bird, Chloe E. 1999. “Gender, Household Labor, and Psychological Distress: The Impact of the Amount and Division of Housework.” Journal of Health and Social Behavior 40 (1): 32–45. https://doi.org/10.2307/2676377. Bird, Chloe E., and Patricia P. Rieker. 1999. “Gender Matters: An Integrated Model for Understanding Men’s and Women’s Health.” Social Science and Medicine 48 (6): 745–55. https://doi.org/10.1016/S0277–9536(98)00402-X. . 2008. Gender and Health: The Effects of Constrained Choices and Social Policies. Cambridge: Cambridge University Press. Blewett, Lynne, Julia A. Rivera, Drew R. Griffin, M. L. King, and K. Williams. 2016. “IPUMS Health Surveys: National Health Interview Survey, Version 6.2.” Minneapolis: University of Minnesota. Bloom, David E., and David Canning. 2008. “Population Health and Economic Growth.” Commission on Growth and Development Working Paper No. 24. Washington, DC: World Bank. https://openknowledge.worldbank.org /handle/10986/28036. Boardman, Jason D. 2009. “State-Level Moderation of Genetic Tendencies to Smoke.” American Journal of Public Health 99 (3): 480–86. https://doi .org/10.2105/AJPH.2008.134932. , Benjamin W. Domingue, and Jonathan Daw. 2015. “What Can Genes Tell Us about the Relationship between Education and Health?” Social Science and Medicine 127 (February): 171–80. https://doi.org/10.1016/j.socscimed .2014.08.001. Bobo, Lawrence D., Camille Z. Charles, Maria Krysan, and Alicia D. Simmons. 2012. “The Real Record on Racial Attitudes.” In Social Trends in American Life: Findings from the General Social Survey since 1972, edited by Peter V. Marsden, 38–83. Princeton, NJ: Princeton University Press. Bond Huie, Stephanie A., Robert A. Hummer, and Richard G. Rogers. 2002. “Individual and Contextual Risks of Death among Race and Ethnic Groups in the United States.” Journal of Health and Social Behavior 43 (3): 359–81. www.ncbi.nlm.nih.gov/pubmed/12467258. Brady, David. 2009. Rich Democracies, Poor People: How Politics Explain Poverty. New York: Oxford University Press. Brady, Kevin. 2017. An Act to Provide for Reconciliation pursuant to Titles II and V of the Concurrent Resolution on the Budget for Fiscal Year 2018. www.congress.gov/bill/115th-congress/house-bill/1/text. Brandt, Allan M. 2007. The Cigarette Century. New York: Basic Books. Braveman, Paula A., Catherine Cubbin, Susan Egerter, David R. Williams, and Elsie Pamuk. 2010. “Socioeconomic Disparities in Health in the United States: What the Patterns Tell Us.” American Journal of Public Health 100 (Supplement 1): S186–96. https://doi.org/10.2105/AJPH.2009.166082. Breslau, Joshua, Sergio Aguilar-Gaxiola, Kenneth S. Kendler, Maxwell Su, David Williams, and Ronald C. Kessler. 2006. “Specifying RaceEthnic Differences in Risk for Psychiatric Disorder in a USA National

References | 219

Sample.” Psychological Medicine 36 (1): 57–68. https://doi.org/10.1017 /S0033291705006161. Breslau, Joshua, Kenneth S. Kendler, Maxwell Su, Sergio Gaxiola-Aguilar, and Ronald C. Kessler. 2005. “Lifetime Risk and Persistence of Psychiatric Disorders across Ethnic Groups in the United States.” Psychological Medicine 35 (3): 317–27. www.ncbi.nlm.nih.gov/pubmed/15841868. Brown, Susan L. 2000. “The Effect of Union Type on Psychological Well-Being: Depression among Cohabitors versus Marrieds.” Journal of Health and Social Behavior 41 (3): 241–55. www.ncbi.nlm.nih.gov/pubmed/11011503. Brownson, Ross C., Debra Haire-Joshu, and Douglas A. Luke. 2006. “Shaping the Context of Health: A Review of Environmental and Policy Approaches in the Prevention of Chronic Diseases.” Annual Review of Public Health 27: 341–70. https://doi.org/10.1146/annurev.publhealth.27.021405.102137. Brulle, Robert J., and David N. Pellow. 2006. “Environmental Justice: Human Health and Environmental Inequalities.” Annual Review of Public Health 27: 103–24. https://doi.org/10.1146/annurev.publhealth.27.021405.102124. Brunekreef, Bert, and Stephen T. Holgate. 2002. “Air Pollution and Health.” Lancet 360 (9341): 1233–42. https://doi.org/10.1016/S0140–6736(02)11274–8. Bryc, Katarzyna, Eric Y. Durand, J. Michael Macpherson, David Reich, and Joanna L. Mountain. 2015. “The Genetic Ancestry of African Americans, Latinos, and European Americans across the United States.” American Journal of Human Genetics 96 (1): 37–53. https://doi.org/10.1016/j.ajhg.2014.11.010. Burgard, Sarah A., and Katherine Y. Lin. 2013. “Bad Jobs, Bad Health? How Work and Working Conditions Contribute to Health Disparities.” American Behavioral Scientist 57 (8): 1105–27. https://doi.org/10.1177 /0002764213487347. Caldwell, John C. 1986. “Routes to Low Mortality in Poor Countries.” Population and Development Review 12 (2): 171–220. Campbell, Mary E., and Christabel L. Rogalin. 2006. “Categorical Imperatives: The Interaction of Latino and Racial Identification.” Social Science Quarterly 87 (5): 1030–52. www.jstor.org/stable/42956593. Carrasquillo, Olveen, Angeles I. Carrasquillo, and Steven Shea. 2000. “Health Insurance Coverage of Immigrants Living in the United States: Differences by Citizenship Status and Country of Origin.” American Journal of Public Health 90 (6): 917–23. www.ncbi.nlm.nih.gov/pubmed/10846509. Carter, Brian D., Christian C. Abnet, Diane Feskanich, Neal D. Freedman, Patricia Hartge, Cora E. Lewis, Judith K. Ockene, et al. 2015. “Smoking and Mortality—Beyond Established Causes.” New England Journal of Medicine 372 (7): 631–40. https://doi.org/10.1056/NEJMsa1407211. Case, Anne, and Angus Deaton. 2015. “Rising Morbidity and Mortality in Midlife among White Non-Hispanic Americans in the 21st Century.” Proceedings of the National Academy of Sciences of the United States of America 112 (49): 15078–83. https://doi.org/10.1073/pnas.1518393112. . 2017. “Mortality and Morbidity in the 21st Century.” Brookings Papers on Economic Activity, Spring (March): 397–476. Washington, DC: Brookings Institution. www.brookings.edu/bpea-articles/mortality-andmorbidity-in-the-21st-century/.

220

|

References

Case, Anne, and Christina Paxson. 2005. “Sex Differences in Morbidity and Mortality.” Demography 42 (2): 189–214. www.ncbi.nlm.nih.gov/pubmed /15986983. CDC / National Center for Health Statistics. 2009. “National Vital Statistics System—Natality (Births).” www.cdc.gov/nchs/nvss/birth_methods.htm. . 2011. “National Hospital Discharge Survey (NHDS).” www.cdc.gov /nchs/nhds/nhds_questionnaires.htm. . 2015. “National Health and Nutrition Examination Survey (NHANES).” www.cdc.gov/nchs/nhanes/nhanes_questionnaires.htm. . 2017a. “National Health Interview Survey (NHIS).” www.cdc.gov /nchs/nhis/data-questionnaires-documentation.htm. . 2017b. “National Ambulatory Medical Care Survey (NAMCS).” ftp:// ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NAMCS. . 2017c. “National Hospital Ambulatory Medical Care Survey (NHAMCS).” ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets /NHAMCS. . 2018. “NVSS—Mortality Data.” http://cdc.gov/nchs/nvss/deaths.htm. CDC / National Center for HIV / AIDS, Viral Hepatitis, STD, and TB Prevention. 2017. “AtlasPlus | NCHHSTP | CDC.” August 30, 2017. www.cdc .gov/nchhstp/atlas/. Centers for Disease Control and Prevention (CDC). 2013. “Vital Signs: Overdoses of Prescription Opioid Pain Relievers and Other Drugs among Women— United States, 1999–2010.” Morbidity and Mortality Weekly Report 62 (26): 537–42. www.ncbi.nlm.nih.gov/pubmed/23820967. . 2016. “Defining Adult Overweight and Obesity.” Centers for Disease Control and Prevention, June 16. www.cdc.gov/obesity/adult/defining .html. . 2018a. “CDC WONDER.” Centers for Disease Control and Prevention, February 9. https://wonder.cdc.gov. . 2018b. “Drug Overdose Death Data.” CDC Injury Center, January 5. www.cdc.gov/drugoverdose/data/statedeaths.html. Central Intelligence Agency. 2018. “The World Factbook.” www.cia.gov /library/publications/the-world-factbook/fields/2177.html. Chandola, Tarani, and Michael G. Marmot. 2010. “Socio-Economic Position and Health.” In Handbook of Behavioral Medicine: Methods and Applications, edited by Andrew Steptoe, 307–20. New York: Springer. https://doi. org/10.1007/978–0–387–09488–5_22. Chang, Elizabeth. 2010. “Why Obama Should Not Have Checked ‘Black’ on His Census Form.” Washington Post, April 28. www.washingtonpost.com /wp-dyn/content/article/2010/04/28/AR2010042804156.html. Charles, Camille Zubrinsky. 2003. “The Dynamics of Racial Residential Segregation.” Annual Review of Sociology 29 (1): 167–207. https://doi.org /10.1146/annurev.soc.29.010202.100002. Chatters, Linda M., Robert Joseph Taylor, Kai Mckeever Bullard, and James S. Jackson. 2009. “Race and Ethnic Differences in Religious Involvement: African Americans, Caribbean Blacks and Non-Hispanic Whites.” Ethnic and Racial Studies 32 (7): 1143–63. https://doi.org/10.1080/01419870802334531.

References | 221

Chen, Ying-Yeh, S. V. Subramanian, Doloros Acevedo-Garcia, and Ichiro Kawachi. 2005. “Women’s Status and Depressive Symptoms: A Multilevel Analysis.” Social Science and Medicine 60 (1): 49–60. https://doi.org/10 .1016/j.socscimed.2004.04.030. Cheng, Tsung-Mei. 2015. “Reflections on the 20th Anniversary of Taiwan’s Single-Payer National Health Insurance System.” Health Affairs 34 (3): 502– 10. https://doi.org/10.1377/hlthaff.2014.1332. Chetty, Raj, David Grusky, Maximilian Hell, Nathaniel Hendren, Robert Manduca, and Jimmy Narang. 2017. “The Fading American Dream: Trends in Absolute Income Mobility since 1940.” Science 356 (6336): 398–406. https://doi.org/10.1126/science.aal4617. Chetty, Raj, Michael Stepner, Sarah Abraham, Shelby Lin, Benjamin Scuderi, Nicholas Turner, Augustin Bergeron, and David Cutler. 2016. “The Association between Income and Life Expectancy in the United States, 2001–2014.” JAMA: Journal of the American Medical Association 315 (16): 1750–66. https://doi.org/10.1001/jama.2016.4226. Child and Adolescent Health Measurement Initiative. n.d. “National Survey of Children’s Health (NSCH).” Data Resource Center for Child and Adolescent Health: A Project of the Child and Adolescent Health Measurement Initiative. http://childhealthdata.org/browse/data-snapshots/nsch-profiles /customizable-profile. Cho, Youngtae, W. Parker Frisbie, Robert A. Hummer, and Richard G. Rogers. 2004. “Nativity, Duration of Residence, and the Health of Hispanic Adults in the United States.” International Migration Review 38 (1): 184–211. Choi, Sunha H. 2012. “Testing Healthy Immigrant Effects among Late Life Immigrants in the United States: Using Multiple Indicators.” Journal of Aging and Health 24 (3): 475–506. https://doi.org/10.1177/0898264311425596. Christakis, Nicholas A., and Paul D. Allison. 2006. “Mortality after the Hospitalization of a Spouse.” New England Journal of Medicine 354: 719–30. https://doi.org/10.1056/NEJMsa050196. Christakis, Nicholas A., and James H. Fowler. 2009. Connected: How Your Friends’ Friends’ Friends Affect Everything You Feel, Think, and Do. New York: Back Bay Books. Clements-Nolle, Kristen, Rani Marx, Robert Guzman, and Mitchell Katz. 2001. “HIV Prevalence, Risk Behaviors, Health Care Use, and Mental Health Status of Transgender Persons: Implications for Public Health Intervention.” American Journal of Public Health 91 (6): 915–21. www.ncbi .nlm.nih.gov/pubmed/11392934. Cochran, Susan D., and Vickie M. Mays. 2000. “Relation between Psychiatric Syndromes and Behaviorally Defined Sexual Orientation in a Sample of the US Population.” American Journal of Epidemiology 151 (5): 516–23. www .ncbi.nlm.nih.gov/pubmed/10707921. Cochran, Susan D., Vickie M. Mays, and J. Greer Sullivan. 2003. “Prevalence of Mental Disorders, Psychological Distress, and Mental Health Services Use among Lesbian, Gay, and Bisexual Adults in the United States.” Journal of Consulting and Clinical Psychology 71 (1): 53–61. www.ncbi.nlm.nih.gov /pubmed/12602425.

222

|

References

Connell, Raewyn. 2012. “Gender, Health and Theory: Conceptualizing the Issue, in Local and World Perspective.” Social Science and Medicine 74 (11): 1675–83. https://doi.org/10.1016/j.socscimed.2011.06.006. Conron, Kerith J., Matthew J. Mimiaga, and Stewart J. Landers. 2010. “A Population-Based Study of Sexual Orientation Identity and Gender Differences in Adult Health.” American Journal of Public Health 100 (10): 1953– 60. https://doi.org/10.2105/AJPH.2009.174169. Conron, Kerith J., Gunner Scott, Grace Sterling Stowell, and Stewart J. Landers. 2012. “Transgender Health in Massachusetts: Results from a Household Probability Sample of Adults.” American Journal of Public Health 102 (1): 118–22. https://doi.org/10.2105/AJPH.2011.300315. Correll, Joshua, Bernadette Park, Charles M. Judd, and Bernd Wittenbrink. 2002. “The Police Officer’s Dilemma: Using Ethnicity to Disambiguate Potentially Threatening Individuals.” Journal of Personality and Social Psychology 83 (6): 1314–29. www.ncbi.nlm.nih.gov/pubmed/12500813. Cossman, J. S., W. L. James, A. G. Cosby, and R. E. Cossman. 2010. “Underlying Causes of the Emerging Nonmetropolitan Mortality Penalty.” American Journal of Public Health 100 (8): 1417–19. https://doi.org/10.2105/AJPH .2009.174185. Courtenay, Will H. 2000. “Constructions of Masculinity and Their Influence on Men’s Well-Being: A Theory of Gender and Health.” Social Science and Medicine 50 (10): 1385–1401. www.ncbi.nlm.nih.gov/pubmed/10741575. Crimmins, Eileen M., Jung Ki Kim, and Aaron Hagedorn. 2002. “Life with and without Disease: Women Experience More of Both.” Journal of Women and Aging 14 (1–2): 47–59. https://doi.org/10.1300/J074v14n01_04. Crimmins, Eileen M., Jung K. Kim, and Teresa E. Seeman. 2009. “Poverty and Biological Risk: The Earlier ‘Aging’ of the Poor.” Journals of Gerontology: Series A, Biological Sciences and Medical Sciences 64 (2): 286–92. https:// doi.org/10.1093/gerona/gln010. Crimmins, Eileen M., Jung Ki Kim, and Aïda Solé-Auró. 2011. “Gender Differences in Health: Results from SHARE, ELSA and HRS.” European Journal of Public Health 21 (1): 81–91. https://doi.org/10.1093/eurpub /ckq022. Crimmins, Eileen M., Jung Ki Kim, and Sarinnapha Vasunilashorn. 2010. “Biodemography: New Approaches to Understanding Trends and Differences in Population Health and Mortality.” Demography 47 Suppl: S41–64. www .ncbi.nlm.nih.gov/pubmed/21302421. Currie, Janet, and Hannes Schwandt. 2016. “Mortality Inequality: The Good News from a County-Level Approach.” Journal of Economic Perspectives: A Journal of the American Economic Association 30 (2): 29–52. https://doi .org/10.1257/jep.30.2.29. Cutler, David M., and Edward L. Glaeser. 1997. “Are Ghettos Good or Bad?” Quarterly Journal of Economics 112 (3): 827–72. www.jstor.org/stable /2951257. Cutler, David, and Grant Miller. 2005. “The Role of Public Health Improvements in Health Advances: The Twentieth-Century United States.” Demography 42 (1): 1–22. https://doi.org/10.1353/dem.2005.0002.

References | 223

Cutter, Susan L., Bryan J. Boruff, and W. Lynn Shirley. 2003. “Social Vulnerability to Environmental Hazards.” Social Science Quarterly 84 (2): 242–61. https://doi.org/10.1111/1540–6237.8402002. De la Cruz, G. Patricia, and Angela Brittingham. 2003. The Arab Population: 2000. U. S. Census Bureau. www.census.gov/prod/2003pubs/c2kbr-23.pdf. Deaton, Angus. 2013. The Great Escape: Health, Wealth, and the Origins of Inequality. Princeton, NJ: Princeton University Press. and Christina Paxson. 2001. “Mortality, Education, Income, and Inequality among American Cohorts.” In Themes in the Economics of Aging, edited by David A. Wise, 129–70. Chicago: University of Chicago Press. www.nber.org/chapters/c10324.pdf. Denney, Justin T., Bridget K. Gorman, and Cristina B. Barrera. 2013. “Families, Resources, and Adult Health: Where Do Sexual Minorities Fit?” Journal of Health and Social Behavior 54 (1): 46–63. https://doi.org/10.1177 /0022146512469629. Denton, Margaret, Steven Prus, and Vivienne Walters. 2004. “Gender Differences in Health: A Canadian Study of the Psychosocial, Structural and Behavioural Determinants of Health.” Social Science and Medicine 58 (12): 2585–600. https://doi.org/10.1016/j.socscimed.2003.09.008. Deschner, A., and S. A. Cohen. 2003. “Contraceptive Use Is Key to Reducing Abortion Worldwide.” Guttmacher Report on Public Policy 6 (4): 7–10. Desilver, Drew. 2013. “Global Inequality: How the U. S. Compares.” Fact Tank: News in the Numbers. Pew Research Center. www.pewresearch.org/facttank/2013/12/19/global-inequality-how-the-u-s-compares/. Diez Roux, Ana V. 2001. “Investigating Neighborhood and Area Effects on Health.” American Journal of Public Health 91 (11): 1783–89. https://doi.org /10.2105/AJPH.91.11.1783. Division of HIV / AIDS Prevention. 2015. Trends in HIV Infection, Stage 3 (AIDS). National Center for HIV / AIDS, Viral Hepatitis, STD, and TB Prevention. www.cdc.gov/hiv/pdf/library/slidesets/cdc-hiv-infection-trends-stage3– 2015.pdf. Downey, Douglas B., and Shana Pribesh. 2004. “When Race Matters: Teachers’ Evaluations of Students’ Classroom Behavior.” Sociology of Education 77 (4): 267–82. www.jstor.org/stable/3649390. Drabble, Laurie, Lorraine T. Midanik, and Karen Trocki. 2005. “Reports of Alcohol Consumption and Alcohol-Related Problems among Homosexual, Bisexual and Heterosexual Respondents: Results from the 2000 National Alcohol Survey.” Journal of Studies on Alcohol 66 (1): 111–20. https://doi .org/10.15288/jsa.2005.66.111. Duffy, John. 1971. “Social Impact of Disease in the Late Nineteenth Century.” Bulletin of the New York Academy of Medicine 47 (7): 797–810. www.ncbi .nlm.nih.gov/pubmed/5283134. Duncan, Greg J. 2008. “When to Promote, and When to Avoid, a Population Perspective.” Demography 45 (4): 763–84. www.ncbi.nlm.nih.gov/pubmed /19110896. Dupre, Matthew E., Alexis T. Franzese, and Emilio A. Parrado. 2006. “Religious Attendance and Mortality: Implications for the Black-White Mortality

224

|

References

Crossover.” Demography 43 (1): 141–64. https://doi.org/10.1353/dem .2006.0004. Durkheim, Emile. [1897] 1951. Suicide: A Study in Sociology. New York: Free Press. Dwyer-Lindgren, Laura, Amelia Bertozzi-Villa, Rebecca W. Stubbs, Chloe Morozoff, Johan P. Mackenbach, Frank J. van Lenthe, Ali H. Mokdad, and Christopher J. L. Murray. 2017. “Inequalities in Life Expectancy among US Counties, 1980 to 2014: Temporal Trends and Key Drivers.” JAMA Internal Medicine 177 (7): 1003–11. https://doi.org/10.1001/jamainternmed .2017.0918. Edwards, David A., Karen Wetzel, and Dana R. Wyner. 2006. “Intercollegiate Soccer: Saliva Cortisol and Testosterone Are Elevated during Competition, and Testosterone Is Related to Status and Social Connectedness with Team Mates.” Physiology and Behavior 87 (1): 135–43. https://doi.org/10.1016/j .physbeh.2005.09.007. Ellison, Christopher G. 1995. “Race, Religious Involvement and Depressive Symptomatology in a Southeastern U. S. Community.” Social Science and Medicine 40 (11): 1561–72. https://doi.org/10.1016/0277–9536(94)00273-V. Ellison, Christopher G., Robert A. Hummer, Shannon Cormier, and Richard G. Rogers. 2000. “Religious Involvement and Mortality Risk among African American Adults.” Research on Aging 22: 630–37. https://doi.org/10.1177 /0164027500226003. Ellison, Christopher G., and Robert A. Hummer, eds. 2010. Religion, Families, and Health: Population-Based Research in the United States. New Brunswick, NJ: Rutgers University Press. Elo, Irma T. 2009. “Social Class Differentials in Health and Mortality: Patterns and Explanations in Comparative Perspective.” Annual Review of Sociology 35 (1): 553–72. https://doi.org/10.1146/annurev-soc-070308–115929. , Cassio M. Turra, Bert Kestenbaum, and B. Reneé Ferguson. 2004. “Mortality among Elderly Hispanics in the United States: Past Evidence and New Results.” Demography 41 (1): 109–28. www.ncbi.nlm.nih.gov/pubmed /15074127. Entwisle, Barbara. 2007. “Putting People into Place.” Demography 44 (4): 687–703. https://doi.org/10.1353/dem.2007.0045. Eschbach, Karl, Jonathan D. Mahnken, and James S. Goodwin. 2005. “Neighborhood Composition and Incidence of Cancer among Hispanics in the United States.” Cancer 103 (5): 1036–44. https://doi.org/10.1002/cncr.20885. Farmer, Melissa M., and Kenneth F. Ferraro. 2005. “Are Racial Disparities in Health Conditional on Socioeconomic Status?” Social Science and Medicine 60 (1): 191–204. https://doi.org/10.1016/j.socscimed.2004.04.026. Fausto-Sterling, Anne. 2005. “The Bare Bones of Sex: Part 1—Sex and Gender.” Signs: Journal of Women in Culture and Society 30 (2): 1491–527. https://doi.org/10.1086/424932. Federal Bureau of Investigation. n.d. “Uniform Crime Report: Incidents and Offenses.” https://ucr.fbi.gov/hate-crime/2015/topic-pages/incidentsand offenses_final.

References | 225

Feldman, Jacob J., Diane M. Makuc, Joel C. Kleinman, and Joan CornoniHuntley. 1989. “National Trends in Educational Differentials in Mortality.” American Journal of Epidemiology 129 (5): 919–33. www.ncbi.nlm.nih.gov /pubmed/2705434. Feliciano, Cynthia. 2005. “Educational Selectivity in U. S. Immigration: How Do Immigrants Compare to Those Left Behind?” Demography 42 (1): 131– 52. https://doi.org/10.1353/dem.2005.0001. Fenelon, Andrew. 2013a. “Geographic Divergence in Mortality in the United States.” Population and Development Review 39 (4): 611–34. https://doi .org/10.1111/j.1728–4457.2013.00630.x. . 2013b. “Revisiting the Hispanic Mortality Advantage in the United States: The Role of Smoking.” Social Science and Medicine 82 (April): 1–9. https://doi.org/10.1016/j.socscimed.2012.12.028. and Samuel H. Preston. 2012. “Estimating Smoking-Attributable Mortality in the United States.” Demography 49 (3): 797–818. https://doi .org/10.1007/s13524–012–0108-x. Finch, B. K., D. P. Do, and R. Frank. 2009. “Could ‘Acculturation’ Effects Be Explained by Latent Health Disadvantages among Mexican Immigrants?” International Migration 43 (3): 471–95. https://doi.org/10.1111/j.1747–7379 .2009.00774.x. Fischer, Claude S., and Michael Hout. 2006. Century of Difference: How America Changed in the Last One Hundred Years. New York: Russell Sage Foundation. Flores, Andrew R., Taylor N. T. Brown, and Jody Herman. 2016. Race and Ethnicity of Adults Who Identify as Transgender in the United States. Williams Institute, UCLA School of Law. https://williamsinstitute.law.ucla.edu /wp-content/uploads/Race-and-Ethnicity-of-Transgender-Identified-Adultsin-the-US.pdf. Forey, Barbara, Jan Hamling, Peter Lee, and Nicholas Wald, eds. 2002. International Smoking Statistics: A Collection of Historical Data from 30 Economically Developed Countries. 2nd ed. Oxford: Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198508564.001.0001. Foutz, Julia, Anthony Damico, Ellen Squires, and Rachel Garfield. 2017. “The Uninsured: A Primer—Key Facts about Health Insurance and the Uninsured under the Affordable Care Act.” Henry J. Kaiser Family Foundation, December 14. www.kff.org/uninsured/report/the-uninsured-a-primer-key-factsabout-health-insurance-and-the-uninsured-under-the-affordable-care-act/. Framingham Heart Study. 2018. “Framingham Heart Study.” www.framingham heartstudy.org/. Freedman, Vicki A., Eileen Crimmins, Robert F. Schoeni, Brenda C. Spillman, Hakan Aykan, Ellen Kramarow, Kenneth Land, et al. 2004. “Resolving Inconsistencies in Trends in Old-Age Disability: Report from a Technical Working Group.” Demography 41 (3): 417–41. www.ncbi.nlm.nih.gov/pubmed /15461008. Freedman, Vicki A., Brenda C. Spillman, Patti M. Andreski, Jennifer C. Cornman, Eileen M. Crimmins, Ellen Kramarow, James Lubitz, et al. 2013.

226

|

References

“Trends in Late-Life Activity Limitations in the United States: An Update from Five National Surveys.” Demography 50 (2): 661–71. https://doi .org/10.1007/s13524–012–0167-z. French, Simone A., Mary Story, and Robert W. Jeffery. 2001. “Environmental Influences on Eating and Physical Activity.” Annual Review of Public Health 22: 309–35. https://doi.org/10.1146/annurev.publhealth.22.1.309. Friedman, Esther M., and Robert D. Mare. 2014. “The Schooling of Offspring and the Survival of Parents.” Demography 51 (4): 1271–93. https://doi .org/10.1007/s13524–014–0303-z. Fries, James F. 1980. “Aging, Natural Death, and the Compression of Morbidity.” New England Journal of Medicine 303 (3): 130–35. https://doi.org /10.1056/NEJM198007173030304. Fryar, Cheryl D., Margaret D. Carroll, and Cynthia L. Ogden. 2016. “Prevalence of Overweight, Obesity, and Extreme Obesity among Adults Aged 20 and Over: United States, 1960–1962 through 2013–2014.” Health E-Stats, July, 1–6. Hyattsville, MD: National Center for Health Statistics. Galea, Sandro, and David Vlahov. 2005. “Urban Health: Evidence, Challenges, and Directions.” Annual Review of Public Health 26: 341–65. https://doi .org/10.1146/annurev.publhealth.26.021304.144708. Galobardes, Bruna, John Lynch, and George Davey Smith. 2007. “Measuring Socioeconomic Position in Health Research.” British Medical Bulletin 81–82 (February): 21–37. https://doi.org/10.1093/bmb/ldm001. Garofalo, Robert, Joanne Deleon, Elizabeth Osmer, Mary Doll, and Gary W. Harper. 2006. “Overlooked, Misunderstood and At-Risk: Exploring the Lives and HIV Risk of Ethnic Minority Male-to-Female Transgender Youth.” Journal of Adolescent Health 38 (3): 230–36. https://doi.org/10.1016/j.jadohealth.2005.03.023. Garrett, Bridgette E., Shanta R. Dube, M. S. Angela Trosclair, Ralph S. Caraballo, and Terry F. Pechacek. 2011. “Cigarette Smoking—United States, 1965– 2008.” Morbidity and Mortality Weekly Report. Surveillance Summaries 60 (01): 109–13. www.cdc.gov/mmwr/preview/mmwrhtml/su6001a24.htm. Gatta, Gemma, Riccardo Capocaccia, Michel P. Coleman, Lynn A. Gloeckler Ries, Timo Hakulinen, Andrea Micheli, Milena Sant, Arduino Verdecchia, and Franco Berrino. 2000. “Toward a Comparison of Survival in American and European Cancer Patients.” Cancer 89 (4): 893–900. https://doi .org/10.1002/1097–0142(20000815)89:43.0.CO. Gates, Gary J. 2011. “How Many People Are Lesbian, Gay, Bisexual and Transgender?” Williams Institute, UCLA School of Law. https://escholarship.org /uc/item/09h684X2. Gelman, Andrew, and Jonathan Auerbach. 2016. “Age-Aggregation Bias in Mortality Trends.” Proceedings of the National Academy of Sciences of the United States of America 113 (7): E816–17. https://doi.org/10.1073/pnas .1523465113. Gilman, Stephen E., Susan D. Cochran, Vickie M. Mays, Michael Hughes, David Ostrow, and Ronald C. Kessler. 2001. “Risk of Psychiatric Disorders among Individuals Reporting Same-Sex Sexual Partners in the National

References | 227

Comorbidity Survey.” American Journal of Public Health 91 (6): 933–39. www.ncbi.nlm.nih.gov/pubmed/11392937. Glymour, M. Maria, Mauricio Avendano, and Ichiro Kawachi. 2014. “Socioeconomic Status and Health.” In Social Epidemiology, edited by Lisa F. Berkman, Ichiro Kawachi, and M. Maria Glymour, 17–62. New York: Oxford University Press. https://doi.org/10.1093/med/9780195377903.003 .0002. González, Hector M., Wassim Tarraf, Keith E. Whitfield, and William A. Vega. 2010. “The Epidemiology of Major Depression and Ethnicity in the United States.” Journal of Psychiatric Research 44 (15): 1043–51. https://doi .org/10.1016/j.jpsychires.2010.03.017. Gordon-Larsen, Penny, Melissa C. Nelson, Phil Page, and Barry M. Popkin. 2006. “Inequality in the Built Environment Underlies Key Health Disparities in Physical Activity and Obesity.” Pediatrics 117 (2): 417–24. https://doi .org/10.1542/peds.2005–0058. Gorman, Bridget K., and Jen’nan Ghazal Read. 2006. “Gender Disparities in Adult Health: An Examination of Three Measures of Morbidity.” Journal of Health and Social Behavior 47 (2): 95–110. https://doi.org/10.1177 /002214650604700201. Gottlieb, M. S., H. M. Schanker, P. T. Fan, A. Saxon, J. D. Weisman, and I. Pozalski. 1981. “Pneumocystis Pneumonia—Los Angeles.” Morbidity and Mortality Weekly Report. Surveillance Summaries 30 (21): 1–3. Atlanta: Centers for Disease Control and Prevention. www.cdc.gov/mmwr/preview /mmwrhtml/june_5.htm. Grantmakers Concerned with Immigrants. 2009. “Immigrants in the United States: County-Level Data on the Foreign-Born.” http://maps.gcir.org/. Grundtvig, Morten, Terje P. Hagen, Mikael German, and Asmund Reikvam. 2009. “Sex-Based Differences in Premature First Myocardial Infarction Caused by Smoking: Twice as Many Years Lost by Women as by Men.” European Journal of Cardiovascular Prevention and Rehabilitation 16 (2): 174–79. https://doi.org/10.1097/HJR.0b013e328325d7f0. Grusky, David B., and Katherine R. Weisshaar. 2014. “Introduction.” In Social Stratification: Class, Race, and Gender in Sociological Perspective, edited by David B. Grusky and Katherine R. Weisshaar, 4th ed., 1–17. Boulder, CO: Westview Press. Haddad, L. B., and N. M. Nour. 2009. “Unsafe Abortion: Unnecessary Maternal Mortality.” Reviews in Obstetrics and Gynecology 2 (2): 122. Hagan, John, and Ronit Dinovitzer. 1999. “Collateral Consequences of Imprisonment for Children, Communities, and Prisoners.” Crime and Justice 26: 121–62. https://doi.org/10.1086/449296. Haines, Michael R. 1979. “The Use of Model Life Tables to Estimate Mortality for the United States in the Late Nineteenth Century.” Demography 16 (2): 289–312. www.ncbi.nlm.nih.gov/pubmed/378722. . 2001. “The Urban Mortality Transition in the United States, 1800– 1940.” Annales de Démographie Historique 101 (1): 33–64. www.cairn .info/revue-annales-de-demographie-historique-2001–1-page-33.htm.

228

|

References

. 2004. “Growing Incomes, Shrinking People—Can Economic Development Be Hazardous to Your Health?: Historical Evidence for the United States, England, and the Netherlands in the Nineteenth Century.” Social Science History 28 (2): 249–70. https://doi.org/10.1017/S0145553200013158. Hainmueller, Jens, Duncan Lawrence, Linna Martén, Bernard Black, Lucila Figueroa, Michael Hotard, Tomás R. Jiménez, Fernando Mendoza, Maria I. Rodriguez, Jonas J. Swartz, and David D. Laitin. 2017. “Protecting Unauthorized Immigrant Mothers Improves Their Children’s Mental Health.” Science 357 (6355): 1041–44. Hamilton, Brady E., Joyce A. Martin, Michelle J. K. Osterman, Sally C. Curtin, and T. J. Matthews. 2015. “Births: Final Data for 2014.” National Vital Statistics Reports 64 (12): 1–64. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/26727629. Hamilton, Erin R., Jodi Berger Cardoso, Robert A. Hummer, and Yolanda C. Padilla. 2011. “Assimilation and Emerging Health Disparities among New Generations of U. S. Children.” Demographic Research 25 (25): 783–818. Han, Beth, Wilson M. Compton, Christopher M. Jones, and Rong Cai. 2015. “Nonmedical Prescription Opioid Use and Use Disorders among Adults Aged 18 through 64 Years in the United States, 2003–2013.” JAMA: Journal of the American Medical Association 314 (14): 1468–78. https://doi .org/10.1001/jama.2015.11859. Hanna-Attisha, Mona, Jenny LaChance, Richard Casey Sadler, and Allison Champney Schnepp. 2016. “Elevated Blood Lead Levels in Children Associated with the Flint Drinking Water Crisis: A Spatial Analysis of Risk and Public Health Response.” American Journal of Public Health 106 (2): 283– 90. https://doi.org/10.2105/AJPH.2015.303003. Harper, Sam, John Lynch, Scott Burris, and George Davey Smith. 2007. “Trends in the Black–White Life Expectancy Gap in the United States, 1983–2003.” JAMA: Journal of the American Medical Association 297 (11): 1224–32. https://doi.org/10.1001/jama.297.11.1224. Harper, Sam, Dinela Rushani, and Jay S. Kaufman. 2012. “Trends in the Black– White Life Expectancy Gap, 2003–2008.” JAMA: Journal of the American Medical Association 307 (21): 2257–59. https://doi.org/10.1001/jama.2012 .5059. Harris, Kathleen Mullan. 2010. “An Integrative Approach to Health.” Demography 47 (1): 1–22. www.ncbi.nlm.nih.gov/pubmed/20355681. , Carolyn Tucker Halpern, Jon Hussey, Eric A. Whitsel, Ley KilleyaJones, Joyce Tabor, Glen Elder, et al. 2013. “Social, Behavioral, and Genetic Linkages from Adolescence into Adulthood.” American Journal of Public Health 103 (Suppl 1): S25–32. https://doi.org/10.2105/AJPH.2012.301181. Harris, Kathleen Mullan, and Thomas W. McDade. 2018. “The Biosocial Approach to Human Development, Behavior, and Health Across the Life Course.” RSF: Russell Sage Foundation Journal of the Social Sciences 4 (4): 2–26. https://doi.org/10.7758/RSF.2018.4.4.01. Harris, David R., and Jeremiah Joseph Sim. 2002. “Who Is Multiracial? Assessing the Complexity of Lived Race.” American Sociological Review 67 (4): 614–27. https://doi.org/10.2307/3088948.

References | 229

Harris-Lacewell, Melissa Victoria. 2010. “Black by Choice.” Nation 290 (17): 10. Harrison, James. 1978. “Warning: The Male Sex Role May Be Dangerous to Your Health.” Journal of Social Issues 34 (1): 65–86. https://doi.org/10.1111 /j.1540–4560.1978.tb02541.x. Hayward, Mark D., and Bridget K. Gorman. 2004. “The Long Arm of Childhood: The Influence of Early-Life Social Conditions on Men’s Mortality.” Demography 41 (1): 87–107. www.ncbi.nlm.nih.gov/pubmed/15074126. Hayward, Mark D., Robert A. Hummer, Chi-Tsun Chiu, César GonzálezGonzález, and Rebeca Wong. 2014. “Does the Hispanic Paradox in U. S. Adult Mortality Extend to Disability?” Population Research and Policy Review 33 (1): 81–96. https://doi.org/10.1007/s11113–013–9312–7. Hayward, Mark D., Robert A. Hummer, and Isaac Sasson. 2015. “Trends and Group Differences in the Association between Educational Attainment and U. S. Adult Mortality: Implications for Understanding Education’s Causal Influence.” Social Science and Medicine 127 (February): 8–18. https://doi .org/10.1016/j.socscimed.2014.11.024. Health and Retirement Study (HRS). 2017. “Health and Retirement Study (HRS).” https://hrs.isr.umich.edu/. Henry J. Kaiser Family Foundation. 2014. “Life Expectancy at Birth (in Years), by Race/Ethnicity.” August 27. www.kff.org/other/state-indicator/lifeexpectancy-by-re/?currentTimeframe=0. . 2017. “Key Facts about the Uninsured Population.” September 19. www .kff.org/uninsured/fact-sheet/key-facts-about-the-uninsured-population/. . 2018. “Current Status of State Medicaid Expansion Decisions.” January 16. www.kff.org/health-reform/slide/current-status-of-the-medicaidexpansion-decision/. Hepburn, Lisa M., and David Hemenway. 2004. “Firearm Availability and Homicide: A Review of the Literature.” Aggression and Violent Behavior 9 (4): 417–40. https://doi.org/10.1016/S1359–1789(03)00044–2. Herbst, Jeffrey H., Elizabeth D. Jacobs, Teresa J. Finlayson, Vel S. McKleroy, Mary Spink Neumann, Nicole Crepaz, and HIV / AIDS Prevention Research Synthesis Team. 2008. “Estimating HIV Prevalence and Risk Behaviors of Transgender Persons in the United States: A Systematic Review.” AIDS and Behavior 12 (1): 1–17. https://doi.org/10.1007/s10461–007– 9299–3. Heron, Melonie. 2015. “Deaths: Leading Causes for 2012.” National Vital Statistics Reports 64 (10). Hyattsville, MD: National Center for Health Statistics. www.cdc.gov/nchs//data/nvsr/nvsr64/nvsr64_10.pdf. . 2016. “Deaths: Leading Causes for 2014.” National Vital Statistics Reports 65 (5). Hyattsville, MD: National Center for Health Statistics. www.cdc.gov/nchs/data/nvsr/nvsr65/nvsr65_05.pdf. Hirsch, Amy E., Sharon M. Dietrich, Rue Landau, Peter D. Schneider, Irv Ackelsberg, Judith Bernstein-Baker, and Joseph Hohenstein. 2002. Every Door Closed: Barriers Facing Parents with Criminal Records. Washington DC: Center for Law and Social Policy. www.clasp.org/sites/default/files/publications /2018/01/every_door_closed.pdf.

230

|

References

Ho, J. Y. (2013). “Mortality under Age 50 Accounts for Much of the Fact That US Life Expectancy Lags That of Other High-Income Countries.” Health Affairs 32 (3): 459–67. and A. S. Hendi (2018). “Recent Trends in Life Expectancy across High Income Countries: Retrospective Observational Study.” British Medical Journal 362: k2562. https://doi.org/10.1136/bmj.k2562. Hofrichter, Richard. 2000. Reclaiming the Environmental Debate: The Politics of Health in a Toxic Culture. Cambridge, MA: MIT Press. House, James S. 2015. Beyond Obamacare: Life, Death, and Social Policy. New York: Russell Sage Foundation. , Ronald C. Kessler, and A. Regula Herzog. 1990. “Age, Socioeconomic Status, and Health.” Milbank Quarterly 68 (3): 383–411. www.ncbi.nlm .nih.gov/pubmed/2266924. House, James S., James M. Lepkowski, Ann M. Kinney, Richard P. Mero, Ronald C. Kessler, and A. Regula Herzog. 1994. “The Social Stratification of Aging and Health.” Journal of Health and Social Behavior 35 (3): 213–34. https://doi.org/10.2307/2137277. House, James S., and Cynthia Robbins. 1983. “Age, Psychosocial Stress, and Health.” In Aging in Society: Selected Reviews of Recent Research, edited by M. W. Riley, B. Hess, and K. Bond, 175–97. Hillsdale, NJ: Erlbaum. Hu, Frank, ed. 2008. Obesity Epidemiology. New York: Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195312911.001.0001. Humes, Karen R., Nicholas A. Jones, and Roberto R. Ramirez. 2011. Overview of Race and Hispanic Origin: 2010. U. S. Department of Commerce, Economics and Statistics Administration, U. S. Census Bureau. www.census.gov /prod/cen2010/briefs/c2010br-02.pdf. Hummer, Robert A. 1996. “Black–White Differences in Health and Mortality: A Review and Conceptual Model.” Sociological Quarterly 37 (1): 105–25. https://doi.org/10.1111/j.1533–8525.1996.tb02333.x. , Monique Biegler, Peter B. De Turk, Douglas Forbes, W. Parker Frisbie, Ying Hong, and Starling G. Pullum. 1999. “Race / Ethnicity, Nativity, and Infant Mortality in the United States.” Social Forces 77 (3): 1083–117. https:// doi.org/10.1093/sf/77.3.1083. Hummer, Robert A., and Iliya Gutin. 2018. “Racial / Ethnic and Nativity Disparities in the Health of Older US Men and Women.” Chapter 3 (pp. 31–66) in Future Directions for the Demography of Aging: Proceedings of a Workshop, edited by M. D. Hayward and M. K. Majmundar. Committee on Population, Division of Behavioral and Social Sciences and Education, National Academy of Sciences, Engineering, and Medicine. Washington, DC: National Academies Press. Hummer, Robert A., and Mark D. Hayward. 2015. “Hispanic Older Adult Health and Longevity in the United States: Current Patterns and Concerns for the Future.” Daedalus 144 (2): 20–30. Hummer, Robert A., and Elaine M. Hernandez. 2013. “The Effect of Educational Attainment on Adult Mortality in the United States.” Population Bulletin 68 (1): 1–16. www.ncbi.nlm.nih.gov/pubmed/25995521.

References | 231

Hummer, Robert A., Daniel A. Powers, Starling G. Pullum, Ginger L. Gossman, and W. Parker Frisbie. 2007. “Paradox Found (Again): Infant Mortality among the Mexican-Origin Population in the United States.” Demography 44 (3): 441–57. www.ncbi.nlm.nih.gov/pubmed/17913005. Hummer, Robert A., Richard G. Rogers, Sarit H. Amir, Douglas Forbes, and W. Parker Frisbie. 2000. “Adult Mortality Differentials among Hispanic Subgroups and Non-Hispanic Whites.” Social Science Quarterly 81 (1): 459–76. www.ncbi.nlm.nih.gov/pubmed/17879490. Hummer, Robert A., Richard G. Rogers, Ryan K. Masters, and Jarron M. Saint Onge. 2009. “Mortality Patterns in Late Life.” In International Handbook of Population Aging, edited by P. Uhlenberg, 521–42. International Handbooks of Population. Dordrecht, NL: Springer. https://doi.org/10.1007/978– 1–4020–8356–3_23. Hummer, Robert A., Richard G. Rogers, Charles B. Nam, and Christopher G. Ellison. 1999. “Religious Involvement and U. S. Adult Mortality.” Demography 36: 273–85. https://doi.org/10.2307/2648114. Iceland, John. 2013. Poverty in America: A Handbook. Berkeley: University of California Press. . 2014. A Portrait of America: The Demographic Perspective. Berkeley: University of California Press. Institute of Medicine (U. S.) Committee on Quality of Health Care in America. 2014. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academies Press. https://doi.org/10.17226 /10027. Institute of Medicine (U. S.). 2003. “The Healthcare Environment and Its Relation to Disparities.” In Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care, edited by Brian D. Smedley, Adrienne Y. Stith, and Alan R. Nelson, 80–124. Washington, DC: National Academies Press. https://doi.org/10.17226/10260. Jackson, James S., Katherine M. Knight, and Jane A. Rafferty. 2010. “Race and Unhealthy Behaviors: Chronic Stress, the HPA Axis, and Physical and Mental Health Disparities over the Life Course.” American Journal of Public Health 100 (5): 933–39. https://doi.org/10.2105/AJPH.2008.143446. Jamal, Ahmed, Elyse Phillips, Andrea S. Gentzke, David M. Homa, Stephen D. Babb, Brian A. King, and Linda J. Neff. 2018. “Current Cigarette Smoking among Adults—United States, 2016.” Morbidity and Mortality Weekly Report 67 (2): 53–59. Atlanta: Centers for Disease Control and Prevention. http://doi.org/10.15585/mmwr.mm6702a1. James, Wesley L. 2014. “All Rural Places Are Not Created Equal: Revisiting the Rural Mortality Penalty in the United States.” American Journal of Public Health 104 (11): 2122–29. https://doi.org/ 10.2105/AJPH.2014.301989. Jasso, Guillermina, Douglas S. Massey, Mark R. Rosenzweig, and James P. Smith. 2004. “Immigrant Health: Selectivity and Acculturation.” In Critical Perspectives on Racial and Ethnic Differences in Health in Late Life, edited by Norman B. Anderson, Rodolfo A. Bulatao, and Barney Cohen, 227–66. Washington, DC: National Academies Press.

232

|

References

Jiwani, Aliya, David Himmelstein, Steffie Woolhandler, and James G. Kahn. 2014. “Billing and Insurance-Related Administrative Costs in United States’ Health Care: Synthesis of Micro-Costing Evidence.” BMC Health Services Research 14 (November): 556. https://doi.org/10.1186/s12913–014–0556–7. Johnson, B. R., M. Horga, and P. Fajans. 2004. “A Strategic Assessment of Abortion and Contraception in Romania.” Reproductive Health Matters 12 (Suppl 24): 184–94. Johnson, Jeh, Alan Bersin, and Marc R. Rosenblum. 2016. “Yearbook of Immigration Statistics: 2015.” Washington, DC: U. S. Department of Homeland Security, Office of Immigration Statistics. www.dhs.gov/immigrationstatistics/yearbook/2015. Johnson, Joy L., Lorraine Greaves, and Robin Repta. 2009. “Better Science with Sex and Gender: Facilitating the Use of a Sex and Gender-Based Analysis in Health Research.” International Journal for Equity in Health 8 (1): 14. https://doi.org/10.1186/1475–9276–8–14. Jones, Jeffrey M. 2016. “Six in 10 Americans Say Racism against Blacks Is Widespread.” Gallup News, August 17. http://news.gallup.com/poll/194657 /six-americans-say-racism-against-blacks-widespread.aspx. Jones, R. K., and J. Jerman. 2017. “Population Group Abortion Rates and Lifetime Incidence of Abortion: United States, 2008–2014.” American Journal of Public Health 107 (12): 1904–9. Karas Montez, Jennifer. 2016. “Research.” September 12. https://jennkarasmontez.com/research/. . 2017. “Deregulation, Devolution, and State Preemption Laws’ Impact on US Mortality Trends.” American Journal of Public Health 107 (11): 1749–50. https://doi.org/10.2105/AJPH.2017.304080. and Mark D. Hayward. 2014. “Cumulative Childhood Adversity, Educational Attainment, and Active Life Expectancy among U. S. Adults.” Demography 51 (2): 413–35. https://doi.org/10.1007/s13524–013–0261-x. Karas Montez, Jennifer, Mark D. Hayward, and Douglas A. Wolf. 2017. “Do US States’ Socioeconomic and Policy Contexts Shape Adult Disability?” Social Science and Medicine 178 (April): 115–26. https://doi.org/10.1016/j .socscimed.2017.02.012. Karas Montez, Jennifer, Robert A. Hummer, and Mark D. Hayward. 2012. “Educational Attainment and Adult Mortality in the United States: A Systematic Analysis of Functional Form.” Demography 49 (1): 315–36. https:// doi.org/10.1007/s13524–011–0082–8. Karas Montez, Jennifer, Robert A. Hummer, Mark D. Hayward, Hyeyoung Woo, and Richard G. Rogers. 2011. “Trends in the Educational Gradient of US Adult Mortality from 1986 to 2006 by Race, Gender, and Age Group.” Research on Aging 33 (2): 145–71. https://doi.org/10.1177/0164027510392388. Karas Montez, Jennifer, and Anna Zajacova. 2013. “Explaining the Widening Education Gap in Mortality among US White Women.” Journal of Health and Social Behavior 54 (2): 166–82. http://journals.sagepub.com/doi /abs/10.1177/0022146513481230. Karas Montez, Jennifer, Anna Zajacova, and Mark D. Hayward. 2016. “Explaining Inequalities in Women’s Mortality between US States.” SSM—

References | 233

Population Health 2 (December): 561–71. https://doi.org/10.1016/j.ssmph .2016.07.004. . 2017. “Disparities in Disability by Educational Attainment across US States.” American Journal of Public Health 107 (7): 1101–8. https://doi .org/10.2105/AJPH.2017.303768. Karlsen, Saffron, and James Y. Nazroo. 2002. “Relation between Racial Discrimination, Social Class, and Health among Ethnic Minority Groups.” American Journal of Public Health 92 (4): 624–31. https://doi.org/10.2105 /AJPH.92.4.624. Kawachi, Ichiro, Nancy E. Adler, and William H. Dow. 2010. “Money, Schooling, and Health: Mechanisms and Causal Evidence.” Annals of the New York Academy of Sciences 1186 (February): 56–68. https://doi.org/10.1111/j .1749–6632.2009.05340.x. Kawachi, Ichiro, and Lisa F. Berkman. 2000. “Social Capital, Social Cohesion, and Health.” In Social Epidemiology, edited by Lisa F. Berkman and Ichiro Kawachi, 174–90. New York: Oxford University Press. Kawachi, Ichiro, Bruce P. Kennedy, Vanita Gupta, and Deborah ProthrowStith. 1999. “Women’s Status and the Health of Women and Men: A View from the States.” Social Science and Medicine 48 (1): 21–32. https://doi .org/10.1016/S0277–9536(98)00286-X. Kaul, Padma, Paul W. Armstrong, Wei-Ching Chang, C. David Naylor, Christopher B. Granger, Kerry L. Lee, Eric D. Peterson, Robert M. Califf, Eric J. Topol, and Daniel B. Mark. 2004. “Long-Term Mortality of Patients with Acute Myocardial Infarction in the United States and Canada.” Circulation 110 (13): 1754–60. https://doi.org/10.1161/01.CIR.0000142671.06167.91. Keister, Lisa A. 2000. Wealth in America: Trends in Wealth Inequality. Cambridge: Cambridge University Press. Kerbo, Harold R. 2011. Social Stratification and Inequality: Class Conflict in Historical, Comparative, and Global Perspective. 8th ed. Boston: McGrawHill Education. Kessler, Ronald C. 2004. “Gender Differences in Mental Disorders in the U. S. National Comorbidity Survey.” In The Economics of Gender and Mental Illness, edited by David E. Marcotte, 15: 7–25. Emerald Group. https://doi .org/10.1016/S0194–3960(04)15002–6. , Patricia Berglund, Olga Demler, Robert Jin, Kathleen R. Merikangas, and Ellen E. Walters. 2005. “Lifetime Prevalence and Age-of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication.” Archives of General Psychiatry 62 (6): 593–602. https://doi .org/10.1001/archpsyc.62.6.593. Kessler, Ronald C., Greg J. Duncan, Lisa A. Gennetian, Lawrence F. Katz, Jeffrey R. Kling, Nancy A. Sampson, . . . and Jens Ludwig. 2014. “Associations of Housing Mobility Interventions for Children in High-Poverty Neighborhoods with Subsequent Mental Disorders during Adolescence.” JAMA: Journal of the American Medical Association 311 (9): 937–47. https://doi .org/10.1001/jama.2014.607. Kessler, Ronald C., and Shanyang Zhao. 1999. “Overview of Descriptive Epidemiology of Mental Disorders.” In Handbook of the Sociology of Mental

234

|

References

Health, edited by Carol S. Aneshensel and Jo C. Phelan, 127–50. Handbooks of Sociology and Social Research. Boston: Springer. https://doi.org /10.1007/0–387–36223–1_7. Kimbro, Rachel Tolbert. 2009. “Acculturation in Context: Gender, Age at Migration, Neighborhood Ethnicity, and Health Behaviors.” Social Science Quarterly 90 (5): 1145–66. http://onlinelibrary.wiley.com/doi/10.1111 /j.1540–6237.2009.00651.x/full. Kindig, David, and Greg Stoddart. 2003. “What Is Population Health?” American Journal of Public Health 93 (3): 380–83. http://ajph.aphapublications. org/doi/full/10.2105/AJPH.93.3.380. King, Katherine E., Jeffrey D. Morenoff, and James S. House. 2011. “Neighborhood Context and Social Disparities in Cumulative Biological Risk Factors.” Psychosomatic Medicine 73 (7): 572–79. https://doi.org/10.1097/PSY .0b013e318227b062. Kitagawa, Evelyn M., and Philip M. Hauser. 1973. Differential Mortality in the United States: A Study in Socioeconomic Epidemiology. Cambridge, MA: Harvard University Press. https://repository.library.georgetown.edu/handle /10822/764543. Kochanek, Kenneth D., Robert N. Anderson, and Elizabeth Arias. 2015. “Leading Causes of Death Contributing to Decrease in Life Expectancy Gap between Black and White Populations: United States, 1999–2013.” NCHS Data Brief, no. 218 (November): 1–8. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/26556175. Kochanek, Kenneth D., Sherry L. Murphy, Jiaquan Xu, and Betzaida TejadaVera. 2016. “Deaths: Final Data for 2014.” National Vital Statistics Reports 65 (4): 1–122. https://www.cdc.gov/nchs/data/nvsr/nvsr65/nvsr65_04 .pdf. Koenig, Harold G., and David B. Larson. 2001. “Religion and Mental Health: Evidence for an Association.” International Review of Psychiatry 13 (2): 67–78. https://doi.org/10.1080/09540260124661. Kondo, Naoki, Grace Sembajwe, Ichiro Kawachi, Rob M. van Dam, S. V. Subramanian, and Zentaro Yamagata. 2009. “Income Inequality, Mortality, and Self Rated Health: Meta-Analysis of Multilevel Studies.” British Medical Journal 339 (7731): 1178–81. https://doi.org/10.1136/bmj.b4471. Krogstad, Jens Manuel and Ana Gonzalez-Barrera. 2014. “If Original DACA Program Is a Guide, How Many Eligible Immigrants Will Apply for Deportation Relief?” Pew Research Center. www.pewresearch.org/fact-tank/2014 /12/05/if-original-daca-program-is-a-guide-many-eligible-immigrants-willapply-for-deportation-relief/. Krueger, Patrick M., and Sarah A. Burgard. 2011. “Work, Occupation, Income, and Mortality.” Chapter 13 (pp. 263–89) in International Handbook of Adult Mortality, edited by R. G. Rogers and E. M. Crimmins. Dordrecht, Neth.: Springer. Ku, Leighton. 2009. “Health Insurance Coverage and Medical Expenditures of Immigrants and Native-Born Citizens in the United States.” American Journal of Public Health 99 (7): 1322–28. https://doi.org/10.2105/AJPH .2008.144733.

References | 235

Kuhn, Randall. 2010. “Routes to Low Mortality in Poor Countries Revisited.” Population and Development Review 36 (4): 655–92. https://doi.org/10 .1111/j.1728–4457.2010.00353.x. Kuka, Elira, Na’ama Shenhav, and Kevin Shih. 2018. “Do Human Capital Decisions Respond to the Returns to Education? Evidence from DACA (Working Paper 24315).” Cambridge, MA: National Bureau of Economic Research Working Paper Series. www.nber.org/papers/w24315. Lancee, Bram, and Herman G. Van de Werfhorst. 2012. “Income Inequality and Participation: A Comparison of 24 European Countries.” Social Science Research 41 (5): 1166–78. https://doi.org/10.1016/j.ssresearch.2012.04.005. Lariscy, Joseph T., Robert A. Hummer, and Richard G. Rogers. 2018. “Cigarette Smoking and All-Cause and Cause-Specific Mortality in the United States.” Demography 55 (5). LaVeist, Thomas, Keshia Pollack, Roland Thorpe Jr., Ruth Fesahazion, and Darrell Gaskin. 2011. “Place, Not Race: Disparities Dissipate in Southwest Baltimore When Blacks and Whites Live under Similar Conditions.” Health Affairs 30 (10): 1880–87. https://doi.org/10.1377/hlthaff.2011.0640. LaVeist, T. A., and J. M. Wallace Jr. 2000. “Health Risk and Inequitable Distribution of Liquor Stores in African American Neighborhood.” Social Science and Medicine 51 (4): 613–17. www.ncbi.nlm.nih.gov/pubmed/10868674. Lawrence, Elizabeth M. 2017. “Why Do College Graduates Behave More Healthfully Than Those Who Are Less Educated?” Journal of Health and Social Behavior 58 (3): 291–306. https://doi.org/10.1177/0022146517715671. , Robert A. Hummer, and Kathleen Mullan Harris. 2017. “The Cardiovascular Health of Young Adults: Disparities along the Urban–Rural Continuum.” Annals of the American Academy of Political and Social Science 672: 257–81. https://doi.org/10.1177/0002716217711426. Leon, David A. 2011. “Trends in European Life Expectancy: A Salutary View.” International Journal of Epidemiology 40 (2): 271–77. https://doi.org /10.1093/ije/dyr061. Leventhal, Tama, and Jeanne Brooks-Gunn. 2003. “Moving to Opportunity: An Experimental Study of Neighborhood Effects on Mental Health.” American Journal of Public Health 93 (9): 1576–82. https://doi.org/ 10.2105/AJPH .93.9.1576. Levine, R. S., J. E. Foster, R. E. Fullilove, M. T. Fullilove, N. C. Briggs, P. C. Hull, B. A. Husaini, and C. H. Hennekens. 2001. “Black-White Inequalities in Mortality and Life Expectancy, 1933–1999: Implications for Healthy People 2010.” Public Health Reports 116 (5): 474–83. https://doi.org/10.1093/phr/116.5.474. Lewontin, Richard C. 1974. The Genetic Basis of Evolutionary Change. New York: Columbia University Press. https://authors.library.caltech.edu/5456 /01/hrst.mit.edu/hrs/evolution/public/papers/lewontin1974/lewontin1974_ chap1andfront.pdf. Lieberson, Stanley. 1985. Making It Count: The Improvement of Social Research and Theory. Berkeley: University of California Press. Lillard, Lee A., and Linda J. Waite. 1995. “Till Death Do Us Part: Marital Disruption and Mortality.” American Journal of Sociology 100 (5): 1131–56. https://doi.org/10.1086/230634.

236

|

References

Link, Bruce G. 2008. “Epidemiological Sociology and the Social Shaping of Population Health.” Journal of Health and Social Behavior 49 (4): 367–84. https://doi.org/10.1177/002214650804900401. Link, Bruce G., and Jo C. Phelan. 1995. “Social Conditions as Fundamental Causes of Disease.” Journal of Health and Social Behavior, Extra Issue: 80–94. www.ncbi.nlm.nih.gov/pubmed/7560851. . 1996. “Understanding Sociodemographic Differences in Health: The Role of Fundamental Social Causes.” American Journal of Public Health 86 (4): 471–73. www.ncbi.nlm.nih.gov/pubmed/8604773. . 2002. “McKeown and the Idea That Social Conditions Are Fundamental Causes of Disease.” American Journal of Public Health 92 (5): 730–32. www.ncbi.nlm.nih.gov/pubmed/11988436. Liu, Hui. 2014. “Gender Paradox (and the Health Myth).” In The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society, edited by William C. Cockerham, Robert Dingwall, and Stella R. Quah, 629–33. New York: John Wiley and Sons. https://doi.org/10.1002/9781118410868.wbehibs110. and Robert A. Hummer. 2008. “Are Educational Differences in U. S. Self-Rated Health Increasing? An Examination by Gender and Race.” Social Science and Medicine 67 (11): 1898–906. https://doi.org/10.1016/j.socscimed.2008.09.021. Liu, Hui, Corinne Reczek, and Dustin Brown. 2013. “Same-Sex Cohabitors and Health: The Role of Race-Ethnicity, Gender, and Socioeconomic Status.” Journal of Health and Social Behavior 54 (1): 25–45. https://doi.org /10.1177/0022146512468280. Liu, Hui, and Linda Waite. 2014. “Bad Marriage, Broken Heart? Age and Gender Differences in the Link between Marital Quality and Cardiovascular Risks among Older Adults.” Journal of Health and Social Behavior 55: 403–23. https://doi.org/10.1177/0022146514556893. Lleras-Muney, Adriana. 2005. “The Relationship between Education and Adult Mortality in the United States.” Review of Economic Studies 72 (1): 189– 221. https://doi.org/10.1111/0034–6527.00329. Lochner, Kim, Elsie Pamuk, Diane Makuc, Bruce P. Kennedy, and Ichiro Kawachi. 2001. “State-Level Income Inequality and Individual Mortality Risk: A Prospective, Multilevel Study.” American Journal of Public Health 91 (3): 385–91. www.ncbi.nlm.nih.gov/pubmed/11236402. Logan, John R., Richard D. Alba, Tom McNulty, and Brian Fisher. 1996. “Making a Place in the Metropolis: Locational Attainment in Cities and Suburbs.” Demography 33 (4): 443–53. www.ncbi.nlm.nih.gov/pubmed /8939417. Lopez-Gonzalez, Lorena, Veronica C. Aravena, and Robert A. Hummer. 2005. “Immigrant Acculturation, Gender and Health Behavior: A Research Note.” Social Forces 84 (1): 581–93. https://doi.org/10.1353/sof.2005.0112. Lorant, V., D. Deliège, W. Eaton, A. Robert, P. Philippot, and M. Ansseau. 2003. “Socioeconomic Inequalities in Depression: A Meta-Analysis.” American Journal of Epidemiology 157 (2): 98–112. www.ncbi.nlm.nih.gov /pubmed/12522017.

References | 237

Ludwig, Jens, Greg J. Duncan, Lisa A. Gennetian, Lawrence F. Katz, Ronald C. Kessler, Jeffrey R. Kling, and Lisa Sanbonmatsu. 2013. “Long-Term Neighborhood Effects on Low-Income Families: Evidence from Moving to Opportunity.” American Economic Review 103 (3): 226–31. https://doi.org/10 .1257/aer.103.3.226. Lutfey, Karen, and Jeremy Freese. 2005. “Toward Some Fundamentals of Fundamental Causality: Socioeconomic Status and Health in the Routine Clinic Visit for Diabetes.” American Journal of Sociology 110 (5): 1326–72. https:// doi.org/10.1086/428914. Lynch, Scott M. 2003. “Cohort and Life-Course Patterns in the Relationship between Education and Health: A Hierarchical Approach.” Demography 40 (2): 309–31. www.ncbi.nlm.nih.gov/pubmed/12846134. Macinko, James, Barbara Starfield, and Leiyu Shi. 2003. “The Contribution of Primary Care Systems to Health Outcomes within Organization for Economic Cooperation and Development (OECD) Countries, 1970–1998.” Health Services Research 38 (3): 831–65. www.ncbi.nlm.nih.gov/pubmed /12822915. Mahalik, James R., Shaun M. Burns, and Matthew Syzdek. 2007. “Masculinity and Perceived Normative Health Behaviors as Predictors of Men’s Health Behaviors.” Social Science and Medicine 64 (11): 2201–9. https://doi.org /10.1016/j.socscimed.2007.02.035. Markides, Kyriakos S., and Jeannine Coreil. 1986. “The Health of Hispanics in the Southwestern United States: An Epidemiologic Paradox.” Public Health Reports 101 (3): 253. www.ncbi.nlm.nih.gov/pmc/articles/PMC1477704/. Marmot, Michael. 2005. The Status Syndrome: How Social Standing Affects Our Health and Longevity. New York: Henry Holt. Marx, Karl. 2014. “Classes in Capitalism and Pre-Capitalism.” In Social Stratification: Class, Race, and Gender in Sociological Perspective, edited by David B. Grusky and Katherine R. Weisshaar, 4th ed., 131–41. Boulder, CO: Westview Press. Massey, Douglas S. 1990. “Social Structure, Household Strategies, and the Cumulative Causation of Migration.” Population Index 56 (1): 3–26. www .ncbi.nlm.nih.gov/pubmed/12316385. . 2007. Categorically Unequal: The American Stratification System. New York: Russell Sage Foundation. and Nancy A. Denton. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge MA: Harvard University Press. Massey, Douglas S., and Garvey Lundy. 2001. “Use of Black English and Racial Discrimination in Urban Housing Markets: New Methods and Findings.” Urban Affairs Review 36 (4): 452–69. https://doi.org/10.1177 /10780870122184957. Masters, Ryan K., Robert A. Hummer, and Daniel A. Powers. 2012. “Educational Differences in US Adult Mortality: A Cohort Perspective.” American Sociological Review 77 (4): 548–72. https://doi.org/10.1177/0003122412451019. Masters, Ryan K., Robert A. Hummer, Daniel A. Powers, Audrey Beck, ShihFan Lin, and Brian Karl Finch. 2014. “Long-Term Trends in Adult Mortality for US Blacks and Whites: An Examination of Period- and Cohort-Based

238

|

References

Changes.” Demography 51 (6): 2047–73. https://doi.org/10.1007/s13524– 014–0343–4. Masters, Ryan K., Andrea M. Tilstra, and Daniel H. Simon. 2017. “Mortality from Suicide, Chronic Liver Disease, and Drug Poisonings among MiddleAged US White Men and Women, 1980–2013.” Biodemography and Social Biology 63 (1): 31–37. www.tandfonline.com/doi/abs/10.1080/19485565.2016 .1248892. Mather, Mark, Linda Alane Jacobsen, and Kelvin M. Pollard. 2015. “Aging in the United States.” Population Bulletin 70 (2): 1–20. Population Reference Bureau. Mathews, Kelly, Jessica Phelan, Nicholas A. Jones, Sarah Konya, Rachel Marks, Beverly M. Pratt, Julia Coombs, and Michael Bentley. 2017. “2015 National Content Test: Race and Ethnicity Analysis Report.” Suitland, MD: U. S. Census Bureau. www2.census.gov/programs-surveys/decennial/2020/programmanagement/final-analysis-reports/2015nct-race-ethnicity-analysis.pdf. Mathews, T. J., Danielle M. Ely, and Anne K. Driscoll. 2018. “State Variations in Infant Mortality by ace and Hispanic Origin of Mother, 2013–2015.” NCHS Data Brief, no. 295 (January): 1–8. Hyattsville, MD: National Center for Health Statistics. www.cdc.gov/nchs/products/databriefs/db295.htm. Mayer, Kenneth H., Judith B. Bradford, Harvey J. Makadon, Ron Stall, Hilary Goldhammer, and Stewart Landers. 2008. “Sexual and Gender Minority Health: What We Know and What Needs to Be Done.” American Journal of Public Health 98 (6): 989–95. https://doi.org/10.2105/AJPH.2007.127811. McKeown, Thomas. 1976. The Modern Rise of Population. London: Edward Arnold. McKinlay, John B., and Sonja M. McKinlay. 1977. “The Questionable Contribution of Medical Measures to the Decline of Mortality in the United States in the Twentieth Century.” Milbank Memorial Fund Quarterly: Health and Society 55 (3): 405–28. www.ncbi.nlm.nih.gov/pubmed/413067. Meara, Ellen, and Jonathan Skinner. 2015. “Losing Ground at Midlife in America.” Proceedings of the National Academy of Sciences of the United States of America 112 (49): 15006–7. https://doi.org/10.1073/pnas.1519763112. Menjívar, Cecilia. 2006. “Liminal Legality: Salvadoran and Guatemalan Immigrants’ Lives in the United States.” American Journal of Sociology 111 (4): 999–1037. Mensah, George A., Gina S. Wei, Paul D. Sorlie, Lawrence J. Fine, Yves Rosenberg, Peter G. Kaufmann, Michael E. Mussolino, et al. 2017. “Decline in Cardiovascular Mortality: Possible Causes and Implications.” Circulation Research 120 (2): 366–80. https://doi.org/10.1161/CIRCRESAHA.116.309115. Mezuk, Briana, Jane A. Rafferty, Kiarri N. Kershaw, Darrell Hudson, Cleopatra M. Abdou, Hedwig Lee, William W. Eaton, and James S. Jackson. 2010. “Reconsidering the Role of Social Disadvantage in Physical and Mental Health: Stressful Life Events, Health Behaviors, Race, and Depression.” American Journal of Epidemiology 172 (11): 1238–49. https://doi.org /10.1093/aje/kwq283. Mezuk, Briana, Cleopatra M. Abdou, Darrell Hudson, Kiarri N. Kershaw, Jane A. Rafferty, Hedwig Lee, and James S. Jackson. 2013. “ ‘White Box’ Epide-

References | 239

miology and the Social Neuroscience of Health Behaviors: The Environmental Affordances Model.” Society and Mental Health 3 (2). https://doi .org/10.1177/2156869313480892. Miller, G. H., and Dean R. Gerstein. 1983. “The Life Expectancy of Nonsmoking Men and Women.” Public Health Reports 98 (4): 343–49. www.ncbi .nlm.nih.gov/pubmed/6604290. Mirowsky, John. 1996. “Age and the Gender Gap in Depression.” Journal of Health and Social Behavior 37 (4): 362–80. www.ncbi.nlm.nih.gov/pubmed /8997891. Mirowsky, John, and Catherine E. Ross. 2003a. Education, Social Status, and Health. New York: Aldine De Gruyter. . 2003b. Social Causes of Psychological Distress. Hawthorne, New York: Routledge. https://doi.org/10.4324/9781315129464. Mishra, Shiraz I., Alicia Lucksted, Deborah Gioia, Beth Barnet, and Claudia R. Baquet. 2009. “Needs and Preferences for Receiving Mental Health Information in an African American Focus Group Sample.” Community Mental Health Journal 45 (2): 117–26. Mitnik, Pablo A., Erin Cumberworth, and David B. Grusky. 2015. “Social Mobility in a High-Inequality Regime.” Annals of the American Academy of Political and Social Science 663 (1): 140–84. https://doi.org/10.1177 /0002716215596971. Moen, Phyllis, Erin L. Kelly, Wen Fan, Shi-Rong Lee, David Almeida, Ellen Ernst Kossek, and Orfeu M. Buxton. 2016. “Does a Flexibility / Support Organizational Initiative Improve High-Tech Employees’ Well-Being? Evidence from the Work, Family, and Health Network.” American Sociological Review 81 (1): 134–64. https://doi.org/10.1177/0003122415622391. Moffitt, Robert A. 2015. “The Deserving Poor, the Family, and the U. S. Welfare System.” Demography 52 (3): 729–49. https://doi.org/10.1007/s13524– 015–0395–0. Moïse, Pierre. 2003. “The Heart of the Health Care System: Summary of the Ischaemic Heart Disease Part of the OECD Ageing-Related Diseases Study.” In A Disease-Based Comparison of Health Systems: What Is Best and at What Cost?, 27–52. Paris: Organisation for Economic Co-operation and Development. https://doi.org/10.1787/9789264100053-en. Moon, Lynelle, Pierre Moïse, and Stéphane Jacobzone. 2003. “Stroke Care in OECD Countries: A Comparison of Treatment, Costs and Outcomes in 17 Countries.” Paris: Organisation for Economic Co-operation and Development. https://doi.org/10.1787/380362605045. Moore, Heather, Rhonda Jones-Webb, Traci Toomey, and Kathleen Lenk. 2008. “Alcohol Advertising on Billboards, Transit Shelters, and Bus Benches in Inner-City Neighborhoods.” Contemporary Drug Problems 35 (2–3): 509–32. https://doi.org/10.1177/009145090803500214. Morello-Frosch, Rachel, and Bill M. Jesdale. 2006. “Separate and Unequal: Residential Segregation and Estimated Cancer Risks Associated with Ambient Air Toxics in US Metropolitan Areas.” Environmental Health Perspectives 114 (3): 386. National Institute of Environmental Health Science. www.ncbi.nlm.nih.gov/pmc/articles/pmc1392233/.

240

|

References

Morello-Frosch, Rachel, and Russ Lopez. 2006. “The Riskscape and the Color Line: Examining the Role of Segregation in Environmental Health Disparities.” Environmental Research 102 (2): 181–96. https://doi.org/10.1016/j .envres.2006.05.007. Morenoff, Jeffrey D. 2003. “Neighborhood Mechanisms and the Spatial Dynamics of Birth Weight.” American Journal of Sociology 108 (5): 976–1017. Moulton, Benjamin, and Darren E. Sherkat. 2012. “Specifying the Effects of Religion and Educational Attainment on Mortality Risk for U. S. Adults.” Sociological Spectrum 32 (1): 1–19. Murphy, Sherry L., Kenneth D. Kochanek, Jiaquan Xu, and Elizabeth Arias. 2015. “Mortality in the United States, 2014.” NCHS Data Brief, no. 229 (December): 1–8. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/26727391. Murphy, Sherry L., Kenneth D. Kochanek, Jiaquan Xu, and Melonie Heron. 2015. “Deaths: Final Data for 2012.” National Vital Statistics Reports 63 (9). Hyattsville, MD: National Center for Health Statistics. Murray, Christopher J. L., Sandeep C. Kulkarni, Catherine Michaud, Niels Tomijima, Maria T. Bulzacchelli, Terrell J. Iandiorio, and Majid Ezzati. 2006. “Eight Americas: Investigating Mortality Disparities across Races, Counties, and Race-Counties in the United States.” PLoS Medicine 3 (9): e260. https://doi.org/10.1371/journal.pmed.0030260. National Academies of Sciences, Engineering, and Medicine, Division of Behavioral and Social Sciences and Education, Committee on Population, and Panel on the Integration of Immigrants into American Society. 2016. The Integration of Immigrants into American Society. Edited by M. C. Waters and M. G. Pineau. Washington, DC: National Academies Press. National Academies of Sciences, Engineering, and Medicine; Division on Engineering and Physical Sciences; Board on Mathematical Sciences and Their Applications; Division of Behavioral and Social Sciences and Education; Committee on Population; and Committee on the Long-Run Macroeconomic Effects of the Aging U. S. Population—Phase II. 2015. The Growing Gap in Life Expectancy by Income: Implications for Federal Programs and Policy Responses. Washington, DC: National Academies Press. National Center for Health Statistics. 2009. Leading Causes of Death, 1900– 1998. Hyattsville, MD: National Center for Health Statistics. www.cdc.gov /nchs/data/dvs/lead1900_98.pdf. . 2016. Health, United States, 2015: With Special Feature on Racial and Ethnic Health Disparities. Hyattsville, MD: National Center for Health Statistics. www.cdc.gov/nchs/data/hus/hus15.pdf. . 2018. “National Center for Health Statistics.” January 31. www.cdc .gov/nchs/. National Health Insurance Administration. 2015. “National Health Insurance Administration Ministry of Health and Welfare.” www.nhi.gov.tw/english/. National Research Council, Division of Behavioral and Social Sciences and Education, Committee on Population, and Panel on Understanding Divergent Trends in Longevity in High-Income Countries. 2011. Explaining Divergent Levels of Longevity in High-Income Countries. Edited by Eileen

References | 241

M. Crimmins, Samuel H. Preston, and B. Cohen. Washington, DC: National Academies Press. https://doi.org/10.17226/13089. National Research Council, Institute of Medicine, Board on Population Health and Public Health Practice, Division of Behavioral and Social Sciences and Education, Committee on Population, and Panel on Understanding CrossNational Health Differences Among High-Income Countries. 2013. US Health in International Perspective: Shorter Lives, Poorer Health. Washington, DC: National Academies Press. https://doi.org/10.17226/13497. National Vital Statistics System. 2015. “NVSS—Linked Birth and Infant Death Data.” CDC / National Vital Statistics System, August 6. www.cdc.gov/nchs /nvss/linked-birth.htm. Needleman, Herbert. 2004. “Lead Poisoning.” Annual Review of Medicine 55: 209–22. https://doi.org/10.1146/annurev.med.55.091902.103653. Neff, Jeff A., Thoma J. Prihoda, and Sue K. Hoppe. 1991. “ ‘Machismo,’ SelfEsteem, Education and High Maximum Drinking among Anglo, Black, and Mexican-American Male Drinkers.” Journal of Studies on Alcohol 52 (5): 458–63. www.jsad.com/doi/abs/10.15288/jsa.1991.52.458. New York State Department of Health. 2018. “Basic Statistics: About Incidence, Prevalence, Morbidity, and Mortality—Statistics Teaching Tools.” www.health.ny.gov/diseases/chronic/basicstat.htm. Newman, Anne B., and Jennifer S. Brach. 2001. “Gender Gap in Longevity and Disability in Older Persons.” Epidemiologic Reviews 23 (2): 343–50. www .ncbi.nlm.nih.gov/pubmed/12192741. Noymer, Andrew, and Michel Garenne. 2000. “The 1918 Influenza Epidemic’s Effects on Sex Differentials in Mortality in the United States.” Population and Development Review 26 (3): 565–81. https://doi.org/10.1111 /j.1728–4457.2000.00565.x. Oakes, J. Michael, Kate E. Andrade, Ifrah M. Biyoow, and Logan T. Cowan. 2015. “Twenty Years of Neighborhood Effect Research: An Assessment.” Current Epidemiology Reports 2 (1): 80–87. https://doi.org/10.1007/s40471–015– 0035–7. Ogden, Cynthia L., and Margaret D. Carroll. 2010. “Prevalence of Overweight, Obesity, and Extreme Obesity among Adults: United States, Trends 1960–1962 through 2007–2008.” Hyattsville, MD: National Center for Health Statistics. www.cdc.gov/nchs/data/hestat/obesity_adult_07_08/obesity_adult_07_08 .htm. Ogden, Cynthia L., Margaret D. Carroll, Hannah G. Lawman, Cheryl D. Fryar, Deanna Kruszon-Moran, Brian K. Kit, and Katherine M. Flegal. 2016. “Trends in Obesity Prevalence among Children and Adolescents in the United States, 1988–1994 through 2013–2014.” JAMA: The Journal of the American Medical Association 315 (21): 2292–99. https://doi.org/10.1001/jama.2016.6361. Okamoto, Dina, and G. Cristina Mora. 2014. “Panethnicity.” Annual Review of Sociology 40 (1): 219–39. https://doi.org/10.1146/annurev-soc-071913– 043201. Olshansky, S. Jay, Toni Antonucci, Lisa Berkman, Robert H. Binstock, Axel Boersch-Supan, John T. Cacioppo, Bruce A. Carnes, et al. 2012. “Differences in Life Expectancy Due to Race and Educational Differences Are Widening,

242

|

References

and Many May Not Catch up.” Health Affairs 31 (8): 1803–13. https://doi .org/10.1377/hlthaff.2011.0746. Olshansky, S. Jay, and A. Brian Ault. 1986. “The Fourth Stage of the Epidemiologic Transition: The Age of Delayed Degenerative Diseases.” Milbank Quarterly 64 (3): 355–91. www.ncbi.nlm.nih.gov/pubmed/3762504. Olshansky, S. Jay, Bruce A. Carnes, Richard G. Rogers, and Len Smith. 1998. “Emerging Infectious Diseases: The Fifth Stage of the Epidemiologic Transition?” World Health Statistics Quarterly: Rapport Trimestriel de Statistiques Sanitaires Mondiales 51 (2–4): 207–17. www.popline.org/node/527054. Omran, A. R. 1971. “The Epidemiologic Transition: A Theory of the Epidemiology of Population Change.” Milbank Memorial Fund Quarterly 49 (4): 509–38. www.ncbi.nlm.nih.gov/pubmed/5155251. . 1977. “A Century of Epidemiologic Transition in the United States.” Preventive Medicine 6 (1): 30–51. www.ncbi.nlm.nih.gov/pubmed/322119. Organization for Economic Co-operation and Development. 2017. “Income Distribution and Poverty.” OECD.Stat., December 15. http://stats.oecd.org /Index.aspx?DataSetCode=IDD. Owens, Ian P. F. 2002. “Sex Differences in Mortality Rate.” Science 297 (5589): 2008–9. https://doi.org/10.1126/science.1076813. Pager, Devah. 2003. “The Mark of a Criminal Record.” American Journal of Sociology 108 (5): 937–75. https://doi.org/10.1086/374403. Pampel, Fred C. 2001. “Cigarette Diffusion and Sex Differences in Smoking.” Journal of Health and Social Behavior 42 (4): 388–404. www.ncbi.nlm.nih .gov/pubmed/11831139. . 2002. “Inequality, Diffusion, and the Status Gradient in Smoking.” Social Problems 49 (1): 35–57. https://doi.org/10.1525/sp.2002.49.1.35. . 2003. “Declining Sex Differences in Mortality from Lung Cancer in High-Income Nations.” Demography 40 (1): 45–65. www.ncbi.nlm.nih.gov /pubmed/12647513. . 2005. “Diffusion, Cohort Change, and Social Patterns of Smoking.” Social Science Research 34 (1): 117–39. www.ncbi.nlm.nih.gov/pubmed/22485056. . 2006. “Socioeconomic Distinction, Cultural Tastes, and Cigarette Smoking.” Social Science Quarterly 87 (1): 19–35. https://doi.org/10.1111 /j.0038–4941.2006.00366.x. . 2010. “Divergent Patterns of Smoking across High-Income Nations.” In International Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 132–63. Washington, DC: National Academies Press. https://doi .org/10.17226/12945. , Patrick M. Krueger, and Justin T. Denney. 2010. “Socioeconomic Disparities in Health Behaviors.” Annual Review of Sociology 36 (August): 349–70. https://doi.org/10.1146/annurev.soc.012809.102529. Pappas, Gregory, Susan Queen, Wilbur Hadden, and Gail Fisher. 1993. “The Increasing Disparity in Mortality between Socioeconomic Groups in the United States, 1960 and 1986.” New England Journal of Medicine 329 (2): 103–9. https://doi.org/10.1056/NEJM199307083290207.

References | 243

Paradies, Yin. 2006. “A Systematic Review of Empirical Research on SelfReported Racism and Health.” International Journal of Epidemiology 35 (4): 888–901. https://doi.org/10.1093/ije/dyl056. Parker, Jennifer D., David Q. Rich, Svetlana V. Glinianaia, Jong Han Leem, Daniel Wartenberg, Michelle L. Bell, Matteo Bonzini, et al. 2011. “The International Collaboration on Air Pollution and Pregnancy Outcomes: Initial Results.” Environmental Health Perspectives 119 (7): 1023–28. https:// doi.org/10.1289/ehp.1002725. Paskov, Marii, and Caroline Dewilde. 2012. “Income Inequality and Solidarity in Europe.” Research in Social Stratification and Mobility 30 (4): 415–32. https://doi.org/10.1016/j.rssm.2012.06.002. Passel, Jeffrey S., and D’Vera Cohn. 2016. “Unauthorized Immigrant Population Stable for Half a Decade.” Pew Research Center: Fact Tank, September 21. www.pewresearch.org/fact-tank/2016/09/21/unauthorized-immigrantpopulation-stable-for-half-a-decade/. Patler, Caitlin, Erin Hamilton, and Robin Savinar. 2018. “DACA Uncertainty May Undermine Its Positive Impact on Wellbeing.” Center for Poverty Research Policy Brief 6 (3). https://poverty.ucdavis.edu/policy-brief/dacauncertainty-may-undermine-its-positive-impact-wellbeing. Patler, Caitlin, and Whitney N. Laster Pirtle. 2018. “From Undocumented to Lawfully Present: Do Changes to Legal Status Impact Psychological Wellbeing among Latino Immigrant Young Adults?” Social Science and Medicine 199 (1): 39–48. Pavalko, Eliza K., and Suzanne L. Woodbury. 2000. “Social Roles as Process: Caregiving Careers and Women’s Health.” Journal of Health and Social Behavior 41 (1): 91–105. www.ncbi.nlm.nih.gov/pubmed/10750324. Pearl, Raymond. 1938. “Tobacco Smoking and Longevity.” Science 87 (2253): 216–17. https://doi.org/10.1126/science.87.2253.216. Pedulla, David S. 2014. “The Positive Consequences of Negative Stereotypes: Race, Sexual Orientation, and the Job Application Process.” Social Psychology Quarterly 77 (1): 75–94. https://doi.org/10.1177/0190272513506229. Pettit, Becky, and Bruce Western. 2004. “Mass Imprisonment and the Life Course: Race and Class Inequality in U. S. Incarceration.” American Sociological Review 69 (2): 151–69. https://doi.org/10.1177/000312240406900201. Pew Research Center. 2015. “U. S. Foreign-Born Population Trends.” In Modern Immigration Wave Brings 59 Million to U. S., Driving Population Growth and Change through 2065: Views of Immigration’s Impact on U. S. Society Mixed, 65–76. Washington, DC: Pew Research Center. www .pewhispanic.org/2015/09/28/chapter-5-u-s-foreign-born-population-trends/. Phelan, Jo C., and Bruce G. Link. 2015. “Is Racism a Fundamental Cause of Inequalities in Health?” Annual Review of Sociology 41 (1): 311–30. https:// doi.org/10.1146/annurev-soc-073014–112305. Phelan, Jo C., Bruce G. Link, and Parisa Tehranifar. 2010. “Social Conditions as Fundamental Causes of Health Inequalities: Theory, Evidence, and Policy Implications.” Journal of Health and Social Behavior 51 Suppl: S28–40. https://doi.org/10.1177/0022146510383498.

244

|

References

Pickett, Kate E., Ye Luo, and Diane S. Lauderdale. 2005. “Widening Social Inequalities in Risk for Sudden Infant Death Syndrome.” American Journal of Public Health 95 (11): 1976–81. https://doi.org/10.2105/AJPH.2004 .059063. Pickett, Kate E., and Richard G. Wilkinson. 2015. “Income Inequality and Health: A Causal Review.” Social Science and Medicine 128 (March): 316– 26. https://doi.org/10.1016/j.socscimed.2014.12.031. Piketty, Thomas, and Emmanuel Saez. 2014. “Inequality in the Long Run.” Science 344 (6186): 838–43. https://doi.org/10.1126/science.1251936. Pletcher, Mark J., Stefan G. Kertesz, Michael A. Kohn, and Ralph Gonzales. 2008. “Trends in Opioid Prescribing by Race/ethnicity for Patients Seeking Care in US Emergency Departments.” JAMA: Journal of the American Medical Association 299 (1): 70–78. https://doi.org/10.1001/jama.2007.64. Pollack, Craig Evan, Sekai Chideya, Catherine Cubbin, Brie Williams, Mercedes Dekker, and Paula Braveman. 2007. “Should Health Studies Measure Wealth? A Systematic Review.” American Journal of Preventive Medicine 33 (3): 250–64. https://doi.org/10.1016/j.amepre.2007.04.033. “POLITICO.” 2018. POLITICO. http://politi.co/. Pope, Nolan G. 2016. “The Effects of Dacamentation: The Impact of Deferred Action for Childhood Arrivals on Unauthorized Immigrants.” Journal of Public Economics 143: 98–114. Population Reference Bureau. 2017. 2017 World Population Data Sheet: With a Special Focus on Youth. www.prb.org/wp-content/uploads/2017/08 /WPDS-2017.pdf. . 2018. “DataFinder | U. S. and World Data by the Population Reference Bureau.” Washington, DC. www.prb.org/DataFinder. Portes, Alejandro, and Alejandro Rivas. 2011. “The Adaptation of Migrant Children.” Future of Children 21 (1): 219–46. www.ncbi.nlm.nih.gov /pubmed/21465862. Portes, Alejandro and Min Zhou. 1993. “The New Second Generation: Segmented Assimilation and Its Variants.” ANNALS of the American Academy of Political and Social Science 530 (November): 74–96. Potts, Marilyn K., M. Audrey Burnam, and Kenneth B. Wells. 1991. “Gender Differences in Depression Detection: A Comparison of Clinician Diagnosis and Standardized Assessment.” Psychological Assessment 3 (4): 609. http:// psycnet.apa.org/journals/pas/3/4/609/. Preston, Samuel H. 1970. “The Age-Incidence of Death from Smoking.” Journal of the American Statistical Association 65 (331): 1125–30. https://doi .org/10.1080/01621459.1970.10481148. . 1975. “The Changing Relation between Mortality and Level of Economic Development.” Population Studies 29 (2): 231–48. https://doi .org/10.2307/2173509. and Irma T. Elo. 1995. “Are Educational Differentials in Adult Mortality Increasing in the United States?” Journal of Aging and Health 7 (4): 476–96. https://doi.org/10.1177/089826439500700402. Preston, Samuel H., Dana A. Glei, and John R. Wilmoth. 2010. “Contribution of Smoking to International Differences in Life Expectancy.” In Interna-

References | 245

tional Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 105– 31. Washington, DC: National Academies Press. https://doi.org/10.17226 /12945. Preston, Samuel H., and Michael R. Haines. 1991. Fatal Years: Child Mortality in Late Nineteenth-Century America. Princeton, NJ: Princeton University Press. www.jstor.org/stable/j.ctt7ztqqf. Preston, Samuel H., Patrick Heuveline, and Michel Guillot. 2001. Demography: Measuring and Modeling Population Processes. Malden, MA: Blackwell. Preston, Samuel H., and Jessica Y. Ho. 2009. “Low Life Expectancy in the United States: Is the Health Care System at Fault?” In International Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 259–98. Washington, DC: National Academies Press. https://doi.org/10.17226/12945. Preston, Samuel H., Andrew Stokes, Neil K. Mehta, and Bochen Cao. 2014. “Projecting the Effect of Changes in Smoking and Obesity on Future Life Expectancy in the United States.” Demography 51 (1): 27–49. https://doi .org/10.1007/s13524–013–0246–9. Preston, Samuel H., and Haidong Wang. 2006. “Sex Mortality Differences in the United States: The Role of Cohort Smoking Patterns.” Demography 43 (4): 631–46. www.ncbi.nlm.nih.gov/pubmed/17236538. Princeton University. 2018. “Fragile Families and Child Wellbeing Study.” https://fragilefamilies.princeton.edu/. Quillian, Lincoln. 2012. “Segregation and Poverty Concentration: The Role of Three Segregations.” American Sociological Review 77 (3): 354–79. https:// doi.org/10.1177/0003122412447793. Ravenholt, R. T. 1990. “Tobacco’s Global Death March.” Population and Development Review 16 (2): 213–40. https://doi.org/10.2307/1971589. Read, Jen’nan Ghazal, and Bridget K. Gorman. 2010a. “Gender and Health Inequality.” Annual Review of Sociology 36 (1): 371–86. https://doi .org/10.1146/annurev.soc.012809.102535. . 2010b. “Gender and Health Revisited.” In Handbook of the Sociology of Health, Illness, and Healing: Blueprint for the 21st Century, edited by Bernice A. Pescosolido, Jack K. Martin, Jane D. McLeod, and Anne Rogers, 411–30. New York: Springer Science and Business Media. Reardon, Sean F., and Kendra Bischoff. 2011. “Income Inequality and Income Segregation.” American Journal of Sociology 116 (4): 1092–153. www.ncbi .nlm.nih.gov/pubmed/21648248. Reczek, Corinne, Russell Spiker, Hui Liu, and Robert Crosnoe. 2016. “Family Structure and Child Health: Does the Sex Composition of Parents Matter?” Demography 53 (5): 1605–30. https://doi.org/10.1007/s13524–016–0501-y. . 2017. “The Promise and Perils of Population Research on Same-Sex Families.” Demography 54 (6): 2385–97. https://doi.org/10.1007/s13524– 017–0630-y. Regier, Darrel A., M. E. Farmer, D. S. Rae, J. K. Myers, M. Kramer, L. N. Robins, L. K. George, M. Karno, and B. Z. Locke. 1993. “One-Month Prevalence of Mental Disorders in the United States and Sociodemographic

246

|

References

Characteristics: The Epidemiologic Catchment Area Study.” Acta Psychiatrica Scandinavica 88 (1): 35–47. https://doi.org/10.1111/j.1600–0447.1993. tb03411.x. Retherford, Robert D. 1972. “Tobacco Smoking and the Sex Mortality Differential.” Demography 9 (2): 203–16. www.ncbi.nlm.nih.gov/pubmed/4670319. . 1975. The Changing Sex Differential in Mortality. Westport, CT: Greenwood Press. Reynolds, Sandra L., Aaron Hagedorn, Jihye Yeom, Yasuhiko Saito, Eise Yokoyama, and Eileen M. Crimmins. 2008. “A Tale of Two Countries—the United States and Japan: Are Differences in Health Due to Differences in Overweight?” Journal of Epidemiology / Japan Epidemiological Association 18 (6): 280–90. www.ncbi.nlm.nih.gov/pubmed/19057112. Richardson, Harry Ward, and Chang-Hee Christine Bae. 2004. Urban Sprawl in Western Europe and the United States. New York: Routledge. Rieker, Patricia P., Chloe E. Bird, and Martha E. Lang. 2010. “Understanding Gender and Health: Old Patterns, New Trends, and Future Directions.” In Handbook of Medical Sociology, edited by Chloe E. Bird, Peter Conrad, Allen M. Fremont, and Stefan Timmermans, 52–74. Nashville, TN: Vanderbilt University Press. Riley, James C. 2001. Rising Life Expectancy: A Global History. Cambridge: Cambridge University Press. Riolo, Stephanie A., Tuan Anh Nguyen, John F. Greden, and Cheryl A. King. 2005. “Prevalence of Depression by Race / Ethnicity: Findings from the National Health and Nutrition Examination Survey III.” American Journal of Public Health 95 (6): 998–1000. https://doi.org/10.2105/AJPH.2004 .047225. Riosmena, Fernando, Randall Kuhn, and Warren C. Jochem. 2017. “Explaining the Immigrant Health Advantage: Self-Selection and Protection in Health-Related Factors among Five Major National-Origin Immigrant Groups in the United States.” Demography 54 (1): 175–200. https://doi .org/10.1007/s13524–016–0542–2. Ro, Annie, Nancy L. Fleischer, and Bridgette Blebu. 2016. “An Examination of Health Selection among U. S. Immigrants Using Multi-National Data.” Social Science and Medicine 158 (June): 114–21. https://doi.org/10.1016/j .socscimed.2016.04.023. Robert, Stephanie. 1999. “Socioeconomic Position and Health: The Independent Contribution of Community Socioeconomic Context.” Annual Review of Sociology 25: 489–516. Rodríguez, Clara E. 2000. Changing Race: Latinos, the Census, and the History of Ethnicity in the United States. New York: New York University Press. Rogers, Richard G., Bethany G. Everett, Jarron M. Saint Onge, and Patrick M. Krueger. 2010. “Social, Behavioral, and Biological Factors, and Sex Differences in Mortality.” Demography 47 (3): 555–78. https://doi.org/10.1353 /dem.0.0119. Rogers, Richard G., Robert A. Hummer, and Charles B. Nam. 2000. Living and Dying in the USA: Behavioral, Health, and Social Differentials of Adult Mortality. San Diego, CA: Academic Press.

References | 247

Rosenfield, Sarah, and Dawne Mouzon. 2013. “Gender and Mental Health.” In Handbook of the Sociology of Mental Health, edited by Carol S. Aneshensel, Jo C. Phelan, and Alex Bierman, 277–96. Handbooks of Sociology and Social Research. Dordrecht, Neth.: Springer. https://doi.org/10.1007/978– 94–007–4276–5_14. Ross, Catherine E., and John Mirowsky. 2002. “Age and the Gender Gap in the Sense of Personal Control.” Social Psychology Quarterly 65 (2): 125–45. https://doi.org/10.2307/3090097. . 2006. “Sex Differences in the Effect of Education on Depression: Resource Multiplication or Resource Substitution?” Social Science and Medicine 63 (5): 1400–13. https://doi.org/10.1016/j.socscimed.2006.03.013. . 2008. “Neighborhood Socioeconomic Status and Health: Context or Composition?” City and Community 7 (2): 163–79. https://doi.org/10.1111 /j.1540–6040.2008.00251.x. Ross, Stephen L., and John Yinger. 2002. The Color of Credit: Mortgage Discrimination, Research Methodology, and Fair-Lending Enforcement. Cambridge, MA: MIT Press. https://ideas.repec.org/b/mtp/titles/0262182289 .html. Rostron, Brian L., John L. Boies, and Elizabeth Arias. 2010. “Education Reporting and Classification on Death Certificates in the United States.” Vital and Health Statistics, Series 2: Data Evaluation and Methods Research. 151 (May): 1–21. www.ncbi.nlm.nih.gov/pubmed/25093685. Rugh, Jacob S., and Douglas S. Massey. 2010. “Racial Segregation and the American Foreclosure Crisis.” American Sociological Review 75 (5): 629– 51. https://doi.org/10.1177/0003122410380868. . 2014. “Segregation in Post–Civil Rights America: Stalled Integration or End of the Segregated Century?” Du Bois Review: Social Science Research on Race 11 (2): 205–32. https://doi.org/10.1017/S1742058X13000180. Ryan, Camille L., and Kurt Bauman. 2016. “Educational Attainment in the United States: 2015.” Current Population Reports P60–259 (March). Washington, DC: US Government Printing Office. www.census.gov/content/dam /Census/library/publications/2017/demo/P60–259.pdf. Sampson, Robert J., and John H. Laub. 1993. Crime in the Making: Pathways and Turning Points through Life. Cambridge, MA: Harvard University Press. Sasson, Isaac. 2016. “Trends in Life Expectancy and Lifespan Variation by Educational Attainment: United States, 1990–2010.” Demography 53 (2): 269– 93. https://doi.org/10.1007/s13524–015–0453–7. Satcher, David, George E. Fryer Jr, Jessica McCann, Adewale Troutman, Steven H. Woolf, and George Rust. 2005. “What If We Were Equal? A Comparison of the Black-White Mortality Gap in 1960 and 2000.” Health Affairs 24 (2): 459–64. https://doi.org/10.1377/hlthaff.24.2.459. Schnittker, Jason. 2007. “Working More and Feeling Better: Women’s Health, Employment, and Family Life, 1974–2004.” American Sociological Review 72 (2): 221–38. https://doi.org/10.1177/000312240707200205. Schoeni, Robert F., James S. House, George A. Kaplan, and Harold Pollack, eds. 2008. Social and Economic Policy as Health Policy. New York: Russell Sage Foundation. www.jstor.org/stable/10.7758/9781610444873.

248

|

References

Schroeder, Steven A. 2007. “We Can Do Better—Improving the Health of the American People.” New England Journal of Medicine 357 (12): 1221–28. https://doi.org/10.1056/NEJMsa073350. Schilts, Randy. 1987. And the Band Played On. New York: St. Martin’s Press. Semega, Jessica L., Kayla R. Fontenot, and Melissa A. Kollar. 2017. “Income and Poverty in the United States: 2016.” Current Population Reports P20–578 (September). Washington, DC: US Government Printing Office. https://census .gov/content/dam/Census/library/publications/2017/demo/P60–259.pdf. Sheftel, Mara Getz, and Frank W. Heiland. 2018. “Disability Crossover: Is There a Hispanic Immigrant Health Advantage That Reverses from Working to Old Age?” Demographic Research 39 (7): 209–250. www.demographicresearch.org/Volumes/Vol39/7/. Sherman, Erik. 2015. “America Is the Richest, and Most Unequal, Country in the World.” Fortune, September 30. http://fortune.com/2015/09/30 /america-wealth-inequality/. Singh, Gopal K., Michael D. Kogan, and Rebecca T. Slifkin. 2017. “Widening Disparities in Infant Mortality and Life Expectancy between Appalachia and the Rest of the United States, 1990–2013.” Health Affairs 36 (8): 1423–32. https://doi.org/10.1377/hlthaff.2016.1571. Singh, Gopal K., and Mohammad Siahpush. 2002. “Ethnic–Immigrant Differentials in Health Behaviors, Morbidity, and Cause-Specific Mortality in the United States: An Analysis of Two National Data Bases.” Human Biology 74 (1): 83–109. www.jstor.org/stable/41466033. . 2014. “Widening Rural–Urban Disparities in Life Expectancy, US, 1969–2009.” American Journal of Preventive Medicine 46 (2): e19–29. https://doi.org/10.1016/j.amepre.2013.10.017. Singhal, Astha, Yu-Yu Tien, and Renee Y. Hsia. 2016. “Racial-Ethnic Disparities in Opioid Prescriptions at Emergency Department Visits for Conditions Commonly Associated with Prescription Drug Abuse.” PloS One 11 (8): e0159224. https://doi.org/10.1371/journal.pone.0159224. Skalamera, Julie, and Robert A. Hummer. 2016. “Educational Attainment and the Clustering of Health-Related Behavior among U. S. Young Adults.” Preventive Medicine 84 (March): 83–89. https://doi.org/10.1016/j.ypmed.2015 .12.011. Smith, James P. 2004. “Unraveling the SES–Health Connection.” Population and Development Review 30: 108–32. www.jstor.org/stable/3401465. . 2007. “The Impact of Socioeconomic Status on Health over the LifeCourse.” Journal of Human Resources XLII (4): 739–64. https://doi .org/10.3368/jhr.XLII.4.739. Steele, Claude M. 1997. “A Threat in the Air: How Stereotypes Shape Intellectual Identity and Performance.” American Psychologist 52 (6): 613–29. www.ncbi.nlm.nih.gov/pubmed/9174398. and Joshua Aronson. 1995. “Stereotype Threat and the Intellectual Test Performance of African Americans.” Journal of Personality and Social Psychology 69 (5): 797. http://doi.apa.org/journals/psp/69/5/797.html. Strawbridge, William J., Sarah J. Shema, Richard D. Cohen, and George A. Kaplan. 2001. “Religious Attendance Increases Survival by Improving and

References | 249

Maintaining Good Health Behaviors, Mental Health, and Social Relationships.” Annals of Behavioral Medicine 23 (1): 68–74. https://doi.org/10.1207 /S15324796ABM2301_1. Stroumsa, Daphna. 2014. “The State of Transgender Health Care: Policy, Law, and Medical Frameworks.” American Journal of Public Health 104 (3): e31–38. https://doi.org/10.2105/AJPH.2013.301789. Subramanian, S. V., and Ichiro Kawachi. 2004. “Income Inequality and Health: What Have We Learned So Far?” Epidemiologic Reviews 26: 78–91. https:// doi.org/10.1093/epirev/mxh003. . 2006. “Whose Health Is Affected by Income Inequality? A Multilevel Interaction Analysis of Contemporaneous and Lagged Effects of State Income Inequality on Individual Self-Rated Health in the United States.” Health and Place 12 (2): 141–56. https://doi.org/10.1016/j.healthplace.2004.11.001. Tavernise, Sabrina. 2016. “Black Americans See Gains in Life Expectancy.” New York Times, May 8. www.nytimes.com/2016/05/09/health/blacks-seegains-in-life-expectancy.html. Teitler, Julien O., Nathan Hutto, and Nancy E. Reichman. 2012. “Birthweight of Children of Immigrants by Maternal Duration of Residence in the United States.” Social Science and Medicine 75 (3): 459–68. https://doi.org/10.1016/j .socscimed.2012.03.038. Thorpe, Kenneth E., David H. Howard, and Katya Galactionova. 2007. “Differences in Disease Prevalence as a Source of the US–European Health Care Spending Gap.” Health Affairs 26 (6): w678–86. https://doi.org/10.1377 /hlthaff.26.6.w678. Trovato, Frank, and N. M. Lalu. 1996. “Narrowing Sex Differentials in Life Expectancy in the Industrialized World: Early 1970’s to Early 1990’s.” Social Biology 43 (1–2): 20–37. www.ncbi.nlm.nih.gov/pubmed/8909108. Turner, Margery Austin, Stephen L. Ross, George C. Galster, and John Yinger. 2002. Discrimination in Metropolitan Housing Markets: National Results from Phase I HDS 2000. Washington, DC: Urban Institute. www.urban.org /sites/default/files/publication/60776/410821-Discrimination-in-MetropolitanHousing-Markets.PDF. Turra, Cassio M., and Irma T. Elo. 2008. “The Impact of Salmon Bias on the Hispanic Mortality Advantage: New Evidence from Social Security Data.” Population Research and Policy Review 27 (5): 515–30. https://doi .org/10.1007/s11113–008–9087–4. Uggen, Christopher, and Jeff Manza. 2002. “Democratic Contraction? Political Consequences of Felon Disenfranchisement in the United States.” American Sociological Review 67 (6): 777–803. https://doi.org/10.2307 /3088970. Umberson, Debra, and Jennifer Karas Montez. 2010. “Social Relationships and Health: A Flashpoint for Health Policy.” Journal of Health and Social Behavior 51 (Suppl): S54–66. https://doi.org/10.1177/0022146510383501. UNC Carolina Population Center. 2018. “Add Health.” http://cpc.unc.edu /projects/addhealth/. United Nations. 1948. “Universal Declaration of Human Rights.” www.un.org /en/universal-declaration-human-rights/index.html.

250

|

References

United Nations Office on Drugs and Crime. 2010. “World Drug Report 2010.” United Nations Publications. United Nations Statistical Office. 1991. “The World’s Women, 1970–1990: Trends and Statistics.” New York: United Nations. United Nations Statistics Division. 2018. “Record View | Life Expectancy at Birth, Total (Years).” UNdata. http://data.un.org/Data.aspx?d=WDI&f= Indicator_Code%3aSP.DYN.LE00.IN. University of California, Berkeley, and Max Planck Institute for Demographic Research (Germany). N.d. “Human Mortality Database.” www.mortality.org / or http://www.humanmortality.de/. University of Washington Institute for Health Metrics and Evaluation. 2018a. “United States Life Expectancy and Age-Specific Mortality Risk by County 1980–2014.” Global Health Data Exchange. http://ghdx.healthdata.org /record/united-states-life-expectancy-and-age-specific-mortality-risk-county1980–2014. . 2018b. “US Data.” Global Health Data Exchange. http://ghdx.healthdata.org/us-data. . 2018c. “US Health Map.” Viz Hub. https://vizhub.healthdata.org/subnational/usa. University of Wisconsin Population Health Institute. 2017. “County Health Rankings.” County Health Rankings & Roadmaps. http://countyhealthrankings.org/. Urban Institute. 2017. “Nine Charts about Wealth Inequality in America.” http://apps.urban.org/features/wealth-inequality-charts/. U. S. Cancer Statistics Working Group. 2017. “United States Cancer Statistics: 1999–2014 Incidence and Mortality Web-Based Report.” Atlanta: U. S. Department of Health and Human Services, Centers for Disease Control and Prevention (CDC), and National Cancer Institute. https://nccd.cdc.gov /uscs/. U. S. Census Bureau. 2018. “Census Bureau Statement on 2020 Census Race and Ethnicity Questions.” Release Number: CB18-RTQ.02, January 26. Suitland, MD: U. S. Census Bureau. https://census.gov/newsroom/pressreleases/2018/2020-race-questions.html. U. S. Citizenship and Immigration Services. 2018. Number of Form I-821D, Consideration of Deferred Action for Childhood Arrivals, by Fiscal Year, Quarter, Intake, and Case Status, Fiscal Year 2012–2018. www.uscis.gov /sites/default/files/USCIS/Resources/Reports%20and%20Studies/Immigration%20Forms%20Data/All%20Form%20Types/DACA/daca_performancedata_fy2018_qtr2_plus_may.pdf. U. S. Department of Health and Human Services. 2014. “The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General.” Atlanta: Centers for Disease Control and Prevention. www.ncbi.nlm .nih.gov/pubmed/24455788. U. S. Department of Health, Education, and Welfare. 1964. “Smoking and Health: Report of the Advisory Committee to the Surgeon General.” 1103. Washington, DC.

References | 251

U. S. Department of Veterans Affairs. 2017. America’s Wars. Washington, DC: Office of Public Affairs. www.va.gov/opa/publications/factsheets/fs_americas_ wars.pdf. Venkataramani, Atheendar S., Sachin J. Shah, Rourke O’Brien, Ichiro Kawachi, and Alexander C. Tsai. 2017. “Health Consequences of the U. S. Deferred Action for Childhood Arrivals (DACA) Immigration Programme: A QuasiExperimental Study.” Lancet 2 (4): e175–81. https://doi.org/10.1016 /S2468–2667(17)30047–6. Ventura, Stephanie J., Kimberley D. Peters, Joyce A. Martin, and Jeffrey D. Maurer. 1997. “Births and Deaths: United States, 1996.” Monthly Vital Statistics Reports 46 (1 Suppl 2): 1–40. Hyattsville, MD: National Center for Health Statistics. www.ncbi.nlm.nih.gov/pubmed/9404390. Verbrugge, Lois M. 1989. “The Twain Meet: Empirical Explanations of Sex Differences in Health and Mortality.” Journal of Health and Social Behavior 30 (3): 282–304. www.ncbi.nlm.nih.gov/pubmed/2778300. Verdecchia, Arduino, Silvia Francisci, Hermann Brenner, Gemma Gatta, Andrea Micheli, Lucia Mangone, Ian Kunkler, and EUROCARE-4 Working Group. 2007. “Recent Cancer Survival in Europe: A 2000–02 Period Analysis of EUROCARE-4 Data.” Lancet Oncology 8 (9): 784–96. https://doi .org/10.1016/S1470–2045(07)70246–2. Wang, Haidong, Austin E. Schumacher, Carly E. Levitz, Ali H. Mokdad, and Christopher J. L. Murray. 2013. “Left Behind: Widening Disparities for Males and Females in US County Life Expectancy, 1985–2010.” Population Health Metrics 11 (1): 8. https://doi.org/10.1186/1478–7954–11–8. Waite, Linda J. 1995. “Does Marriage Matter?” Demography 32 (4): 487–503. https://doi.org/10.2307/2061670. and Maggie Gallagher. 2000. The Case for Marriage: Why Married People Are Healthier, Happier, and Better-off Financially. Westminster, MD: Broadway Books. Wakefield, Sara, and Christopher Uggen. 2010. “Incarceration and Stratification.” Annual Review of Sociology 36 (1): 387–406. https://doi.org/10.1146 /annurev.soc.012809.102551. Waldron, Ingrid. 1993. “Recent Trends in Sex Mortality Ratios for Adults in Developed Countries.” Social Science and Medicine 36 (4): 451–62. www .ncbi.nlm.nih.gov/pubmed/8434270. Walsemann, Katrina M., Gilbert C. Gee, and Danielle Gentile. 2015. “Sick of Our Loans: Student Borrowing and the Mental Health of Young Adults in the United States.” Social Science and Medicine 124 (January): 85–93. https://doi.org/10.1016/j.socscimed.2014.11.027. Ward, Brian W., James M. Dahlhamer, Adena M. Galinsky, and Sarah S. Joestl. 2014. “Sexual Orientation and Health among US Adults: National Health Interview Survey, 2013.” National Health Statistics Reports, no. 77 (July): 1–10. www.ncbi.nlm.nih.gov/pubmed/25025690. Weber, Max. 2014. “Class, Status, Party.” In Social Stratification: Class, Race, and Gender in Sociological Perspective, edited by David B. Grusky and Katherine R. Weisshaar, 4th ed., 165–74. Boulder, CO: Westview Press.

252

|

References

Western, Bruce, Jeffrey R. Kling, and David F. Weiman. 2001. “The Labor Market Consequences of Incarceration.” Crime and Delinquency 47 (3): 410– 27. https://doi.org/10.1177/0011128701047003007. Williams, David R. 1990. “Socioeconomic Differentials in Health: A Review and Redirection.” Social Psychology Quarterly 53 (2): 81–99. https://doi .org/10.2307/2786672. , Hector M. González, Harold Neighbors, Randolph Nesse, Jamie M. Abelson, Julie Sweetman, and James S. Jackson. 2007. “Prevalence and Distribution of Major Depressive Disorder in African Americans, Caribbean Blacks, and Non-Hispanic Whites: Results from the National Survey of American Life.” Archives of General Psychiatry 64 (3): 305–15. https://doi .org/10.1001/archpsyc.64.3.305. Williams, David R., and Selina A. Mohammed. 2013. “Racism and Health I: Pathways and Scientific Evidence.” American Behavioral Scientist 57 (8): 1152–73. https://doi.org/10.1177/0002764213487340. Williams, David R., Harold W. Neighbors, and James S. Jackson. 2003. “Racial / Ethnic Discrimination and Health: Findings from Community Studies.” American Journal of Public Health 93 (2): 200–208. www.ncbi.nlm .nih.gov/pubmed/12554570. Williams, David R., and Michelle Sternthal. 2010. “Understanding Racial-Ethnic Disparities in Health: Sociological Contributions.” Journal of Health and Social Behavior 51 Suppl: S15–27. https://doi.org/10.1177/0022146510383838. Williams, Patricia J. 2010. “Not-Black by Default.” Nation, April 22. www .thenation.com/article/not-black-default/. Wilmoth, John R., Carl Boe, and Magali Barbieri. 2010. “Geographic Differences in Life Expectancy at Age 50 in the United States Compared with Other HighIncome Countries.” In International Differences in Mortality at Older Ages: Dimensions and Sources, edited by Eileen M. Crimmins, Samuel H. Preston, and Barney Cohen, 19: 333–66. Washington, DC: National Academies Press. https://doi.org/10.17226/12945. Wolf-Maier, Katharina, Richard S. Cooper, Holly Kramer, José R. Banegas, Simona Giampaoli, Michel R. Joffres, Neil Poulter, Paola Primatesta, Birgitta Stegmayr, and Michael Thamm. 2004. “Hypertension Treatment and Control in Five European Countries, Canada, and the United States.” Hypertension 43 (1): 10–17. https://doi.org/10.1161/01.HYP.0000103630.72812.10. Wolfson, Michael C., George Kaplan, John Lynch, Nancy Ross, and Eric Backlund. 2000. “Relation between Income Inequality and Mortality: Empirical Demonstration.” Western Journal of Medicine 172 (1): 22–24. www.ncbi.nlm .nih.gov/pubmed/18751209. Woolf, S. H., D. A. Chapman, J. M. Buchanich, K. J. Bobby, E. B. Zimmerman, and S. M. Blackburn. 2018. “Changes in Midlife Death Rates Across Racial and Ethnic Groups in the United States: Systematic Analysis of Vital Statistics.” British Medical Journal 362: k3096. https://doi.org/10.1136/bmj.k3096. World Bank Group. 2017. “Indicators | Data.” World Bank: Data. https://data .worldbank.org/indicator. . 2018. “Death Rate, Crude (per 1,000 People).” https://data.worldbank .org/indicator/SP.DYN.CDRT.IN.

References | 253

World Health Organization. 2014. “Mortality Database,” November. www .who.int/healthinfo/mortality_data/en/. . 2016. “Life Expectancy—Data by Country.” Global Health Observatory. Geneva, Switz.: World Health Organization. June 6, 2016. http://apps .who.int/gho/data/node.main.688. . 2017. “WHO Statistical Information System.” Geneva, Switz.: World Health Organization. http://apps.who.int/whosis/data/. . 2018a. “The Top 10 Causes of Death.” Geneva, Switz.: World Health Organization. May 24. www.who.int/mediacentre/factsheets/fs310/en/index1 .html. . 2018b. World Health Statistics 2018: Annex B. Geneva, Switz.: World Health Organization. www.who.int/gho/publications/world_health_statistics /2018/EN_WHS2018_AnnexB.pdf?ua=1. Xu, Jiaquan, Sherry L. Murphy, Kenneth D. Kochanek, and Brigham A. Bastian. 2016. “Deaths: Final Data for 2013.” National Vital Statistics Reports 64 (2): 1–119. Hyattsville, MD: National Center for Health Statistics. www .ncbi.nlm.nih.gov/pubmed/26905861. Yang, Yang Claire, Karen Gerken, Kristen Schorpp, Courtney Boen, and Kathleen Mullan Harris. 2017. “Early-Life Socioeconomic Status and Adult Physiological Functioning: A Life Course Examination of Biosocial Mechanisms.” Biodemography and Social Biology 63 (2): 87–103. https://doi.org /10.1080/19485565.2017.1279536. Yinger, John. 1995. Closed Doors, Opportunities Lost: The Continuing Costs of Housing Discrimination. New York: Russell Sage Foundation. Zajacova, Anna, and Elizabeth M. Lawrence. 2018. “The Relationship between Education and Health: Reducing Disparities through a Contextual Approach.” Annual Review of Public Health 39: 273–89. https://doi.org/10 .1146/annurev-publhealth-031816–044628. Zhang, Zhenmei, Mark D. Hayward, and Chuntian Lu. 2012. “Is There a Hispanic Epidemiologic Paradox in Later Life? A Closer Look at Chronic Morbidity.” Research on Aging 34 (5): 548–71. https://doi.org/10.1177 /0164027511429807. Zimmer, Zachary, Linda G. Martin, Mary Beth Ofstedal, and Yi-Li Chuang. 2007. “Education of Adult Children and Mortality of Their Elderly Parents in Taiwan.” Demography 44 (2): 289–305. www.ncbi.nlm.nih.gov/pubmed /17583306. Zong, Jie, Jeanne Batalova, and Jeffrey Hallock. 2018. “Frequently Requested Statistics on Immigrants and Immigration in the United States.” Migration Policy Institute, February 8. www.migrationpolicy.org/article/frequentlyrequested-statistics-immigrants-and-immigration-united-states.

Index

abortion, 169–170 accidents, 27t., 44t., 62, 68, 118, 132, 133t., 134, 161, 162t., 166 Affordable Care Act (ACA), 53, 60–61, 182, 184–186 Age of Degenerative and Man-made Diseases, 38–40 Age of Pestilence and Famine, 28, 30 Age of Receding Pandemics, 30–35 age-specific measures, 16–18, 54, 56fig. AIDS, 3, 41, 46–48, 52, 135fig., 171, 176, See also HIV Alzheimer’s disease, 18, 26, 27t., 41, 44t., 132, 133t., 161, 162t. arthritis, 57, 163, 173 asthma, 16, 57, 135, 148, 163, 175, 207n19 attention deficit disorder (ADD), 163 attention deficit hyperactivity disorder (ADHD), 163 Avendano, Mauricio, 193 Baker, David, 122 Bearman, Peter, 91 Behavioral Risk Factor Surveillance System (BRFSS), 174, 175, 176 Behrman, Jere, 123 bisexuals, 48, 175, 176–177 Black Lives Matter Movement, 183 blood pressure, 43, 62, 65, 102, 147, 166, 167 body mass index (BMI), 48, 57, 65–66, 112t.

Boone County, 77–78 Burgard, Sarah, 102 Caldwell, John, 29–30 cancer: cause of death, 26, 27t., 30, 39–40, 41, 62, 138, 161; cure for, 17; elevated risk for obese, 49, 65; incidence in Hispanics, 135; incidence in women, 163, 173; mortality rate, 42–43, 44t., 45–46, 51, 57, 62–63, 83, 133t., 145; reduced risk for high SES individuals, 118; smoking and, 198n55 car accidents, 68, 134 cardiovascular diseases, 42–43, 46, 49, 51, 164 Case, Anne, 1–4, 13, 16, 48, 154–155, 157 Ceausescu, Nicolae, 169 census, 88, 108, 128, 130–131, 145, 174 Census Bureau, 88, 99, 129, 131, 203n2 centenarians, 25 cerebrovascular diseases, 26, 27t., 40, 42–43, 44t., 57, 133t., 161, 162t. chemotherapy, 46 Children’s Health Insurance Program (CHIP), 59, 61 cholera, 28, 32, 33, 41–42 cholesterol, 43, 57, 102, 165, 166 chronic diseases: cause of death, 35, 39, 41, 121, 138, 145; in children, 15; dominant pattern of, 25–26, 35, 40; in girls, 163; in Hispanics, 135; in

255

256

|

Index

chronic diseases (continued) immigrants, 148; liver, 1; lower respiratory, 27t.; and measures of health, 87; obesity impacting, 49, 51, 66; in older adults, 15; and persistence of infectious diseases, 52; rarity prior to twentieth century, 30; reductions in, 42–43, 52; replacing infectious diseases, 25, 26, 35, 38, 39; smoking and, 51; in women, 166–167 chronic lower respiratory diseases, 26, 27t., 132–133, 162t. cirrhosis, 1, 134 cities, 10, 31, 32, 34–35, 67, 74, 76, 83, 85, 87, 118, 182 civil rights, 7, 140, 142, 183, 190, 193 Civil Rights Act, 140 Civil War, 28 Clinton, Hillary Rodham, 2 cohort-specific measures, 14, 18–19, 196n35 colonoscopies, 45 compensatory resources, 119 contraception, 169–170 counties, 4, 10, 12, 20–21, 68, 75, 76, 77–82, 86, 84fig., 134, 183 criminal justice system, 10, 141–143, 156 Crimmins, Eileen, 43 Cutler, David, 34 Deaton, Angus, 1–4, 13, 16, 48, 154–155, 157 Deferred Action for Childhood Arrivals (DACA), 149–150 Democrats, 106 demography, 6, 9, 11–13 diabetes clinics, 119 diabetes mellitus: cause of death, 26, 27t., 30, 41, 44t., 162t.; clinics, 119; elevated rates in Hispanics, 135; elevated rates in men, 163; elevated risk for obese, 49, 65; lower risk of, 89; mortality rates, 132, 133t., 134; in transgenders, 175; and U.S. in international context, 57 diarrhea, 26, 27t., 41 diet, 43, 65, 82, 86, 118, 166, 181, 189 diphtheria, 27t., 33–34 disabilities, 4, 12, 14, 48–49, 81–82, 135, 148, 166–167, 173–174 documentation, 7, 85, 109, 180, 197n6 Douthat, Ross, 2 drug abuse, 1, 39, 46, 48, 67, 74, 78, 86, 118, 138, 141, 154–156, 164–165, 177 Durkheim, Emile, 10

Earned Income Tax Credit (EITC), 190 Ebola virus, 52 education: above-average, 134; access to, 29, 31; advances, 37–38, 52; dimensions of, 101; disparities, 108–109, 110, 114, 123, 192; distribution, 82, 109, 144; early life, 116; expansion of, 29, 34, 35, 78; health programs, 34; high, 98, 107, 108, 110, 113, 114, 115, 123, 156, 192; as a human right, 179; immigrant profile, 144–145; and inequalities in social policies, 74; and infant health, 110t.; lack of, 37; low, 4, 50–51, 86, 99, 108, 110, 113, 116, 153, 168, 191, 193; policies, 7, 190–191, 193; segregation in system, 142; and SES, 8, 21, 87, 97, 157; social resources, 12, 86; and systems of inequalities, 9; U.S. in international context, 75; warning of smoking dangers, 45; wealth predicts better health, 104; women and, 157, 168 educational attainment: average, 86, 179; disparities, 19, 111, 112t., 113, 123, 125; as a flexible resource, 138; high, 94, 95, 98, 113; increasing levels, 99; and life expectancy, 113, 114fig.; low, 98, 113, 116; measures of, 101; quality of, 98; and SES, 100–101, 102, 108, 124, 125, 192; social resources, 100; in studies of population health, 101 emphysema, 45 enteritis, 26, 27t. Environmental Affordances Model, 153 epidemiologic transition, 20, 25–27, 29–32, 35–38, 40–42, 46, 52, 120, 154, 198n52 Equal Credit Opportunity Act, 140 equality, 8, 158, 161, 168, 171–172, 174, 178, 183, 186–187, 190 ethnic stratification, 127, 139, 143 exercise, 67, 82, 85–86, 118, 120, 139, 166–167, 175, 181, 189 explanations: for Black-White mental health paradox, 151–152; for deaths of despair, 48; for differences in occupational statuses, 102; for educational disparities, 109; factors outside of health care system, 63; for gender differences, 166–168; for health of those over, 65, 62; Hispanic health paradox, 145–148; for macroeconomic changes, 105–106; Omran’s, 31; proximate determinants of health, 57; of SES-health disparities, 99, 106, 114, 124; and social policy, 74,

Index 180–181; of trends, 7, 9, 155, 180; of U.S. population health inferiority, 20 Fairfax County, 77–79 Fair Housing Act, 140 families: affluent, 111; cost of obesity, 49; level of wealth, 104; low-income, 89; patterns of health, 76; privilege in, 181; of service members, 48; sharing resources, 95; social contexts, 10, 12, 20–21, 75, 76, 89, 92–93, 95–96; and supportive institutions, 86; of workers, 92 feminism, 171 firearms, 67–68 Fischer, Claude, 37 flexible resources, 8, 116–118, 120, 122, 138 Flores, Andrew R., 175 food assistance (food stamps), 190 food supply, 34 Framingham Heart Study, 43 Freese, Jeremy, 119 fundamental cause theory: contextualizing risk, 189–190; critiques of SES as fundamental cause, 123–125; gender and, 178; health behaviors, 63; immigration policy and, 149; multiple health outcomes, 118, 120, 191; policies and, 184, 188; racial and ethnic disparities, 139, 156; SES disparities, 8, 99, 114–118, 120, 125, 127, 138, 191; social context of, 122 gallstones, 65 garbage removal, 34, 118, See also waste disposal gay population, 46–48, 175, 176–177 gender: definition, 158; disparities, 177; equality, 178; femininity, 177, 183; LGBTQIA (Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, and Asexual), 176–178; masculinity, 177, 183; mortality gender gap, 159–165, 171–174; sexual orientation, 6, 20, 21, 22, 176–177; transgender health, 174–176; women’s health paradox, 166–168, 170, See also sex genetics, 7, 10–11, 20, 39, 65, 123, 125, 129–130, 166 genitourinary, 55, 56fig. geographic inequality, 68–69 geographic-specific orientation, 8–9 geography, 6, 78, 80, 88, 192 Gini index, 105 Gorman, Bridget, 30

|

257

gross domestic product (GDP), 60, 72–73, 186 guns, See firearms Haines, Michael, 31, 37–38 Hauser, Philip, 108 Hayward, Mark, 30, 109, 115–116 Health and Retirement Study (HRS), 103, 116, 125 health care systems, 10, 22, 49, 53, 57, 59–63, 74–75, 126, 184–186 Health Impact Assessment (HIA), 190 health insurance, 59–61, 145, 184–186 heart disease: cause of death, 26, 30, 39–41, 62, 118, 138, 161, 166; correlation between smoking and, 45; diagnosis of, 16, 43, 135fig.; elevated rates in men, 163, 166; mortality, 42, 46, 74, 83, 133t., 138, 154, 161; reducing rates of, 51; reducing risk of, 43; risk elevated for obese, 49, 65; severity in genders, 167–168; in transgenders, 175; understanding of the causes of, 43; U.S. in international context, 57 height decrease, 31 high school degree, 1, 4, 50–51, 99, 110–113, 114fig., 144, 154, 157, 183 Hispanic health paradox, 22, 144–148, 151 historical trends, 20, 24–25 HIV, 41, 46–48, 52, 57, 135fig., 138, 154, 161, 162t., 175–176 Ho, Jessica, 62–63 homicide, 62, 68, 74, 138, 161, 162t., 166 House, James, 6, 115, 116, 187, 190–191 Hout, Michael, 37 human rights, 179–180, 182 hygiene, 31 hypertension, 49, 57, 82, 133t., 145, 165 immigrant health paradox, See Hispanic health paradox individualism, 75, 181 industrialization, 34, 106, 184 Industrial Revolution, 29, 31 infant mortality rates (IMR), 5, 42, 53, 80, 95, 110, 134, 136, 146, 161 infectious diseases: cause of death, 118, 120–121, 138; HIV, 46, 52; in low-income countries, 41; medical advances against, 34; microorganisms as causal agents, 33, 39; mortality, 32, 34, 55, 56fig., 85; outbreaks, 30–32; reduction of, 32, 34–35, 39; shift to

258

|

Index

infectious diseases (continued) chronic diseases, 25, 26, 27, 35, 38–40; socioeconomic and environmental causes, 28 influenza, 18, 26, 27t., 35, 38, 40, 44t., 52, 162t., 164t. Interdisciplinary Association for Population Health Science (IAPHS), 6 intestinal ulceration, 26, 27t. IRS, 103 Jenner, Edward, 25, 32 Jim Crow Era, 140, 183 Kawachi, Ichiro, 193 Kitagawa, Evelyn, 108 Koch, Robert, 33 Kristof, Nicholas, 53, 61 Krugman, Paul, 2 Kuhn, Randall, 29–30 labor unions, 106 Lawrence, Elizabeth, 109 lesbians, 175, 176–177 Lexis Diagram, 13–14, 16, 19 LGBTQIA population (Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, and Asexual), 176–178 life expectancy, 4–5, 12, 14, 17–18, 22, 24, 25, 26–27, 28, 31, 34–35 Link, Bruce, 8, 99, 116, 120, 122, 125, 189 longevity: disparities, 10, 114; education beneficial for, 98, 115; epidemiologic transition, 26, 36, 38; life expectancy and, 137; low income effects on, 103; marital status and, 94; men’s, 177; patterns, 5, 9; SES-health relationship, 97–98, 118; social and economic inequalities related to, 192; women’s, 177–178 lower respiratory diseases, 26, 27fig., 41, 132, 133t., 162t. Lutfey, Karen, 119 McKeown, Thomas, 32 McKinlay, John, 32 McKinlay, Sonja, 32 malaria, 41 mammograms, 45 Manier, Kathryn Rose Green, 35 Marmot, Sir Michael, 101–102 Marxism, 100 Massey, Douglas, 105 Matched Records Study, 108

material deprivation, 69–70, 71 maternal educational disparities, 110 Medicaid, 59, 115, 123, 145, 170, 186, 190 Medicare, 59, 61, 115, 145, 184, 190 Medicare Current Beneficiary Survey, 175 mental illness, 39, 74, 152, 211n73 microorganisms, 26, 33, 39 Miller, Grant, 34 Mirowsky, John, 87, 101 Moen, Phyllis, 92 Montez, Jennifer Karas, 75, 80, 82, 90, 115–116, 123 Moody, James, 91 mortality rates: Age of Degenerative and Man-made Diseases, 40; Age of Pestilence and Famine, 28; Age of Receding Pandemics, 31–32, 34–38; age-specific, 1, 54, 56fig., 132, 133t., 160fig.; of American Indian Alaskan Natives (AIAN), 134; Case-Deaton article, 1–5; cause-specific, 1, 42–43, 44t., 45–49, 55, 56fig., 132, 133t., 134; of chronic diseases, 52; decreases in, 5, 24–25, 157; drug-related, 155; excess, 83; gender-specific, 160fig., 161, 168; of Hispanics, 147; increases in, 22, 48, 155, 157; infant (IMR), 5, 42, 53, 80, 95, 110, 134, 136, 146, 161; of low-income individuals, 61; maternal, 53; measures of, 12–19, 51; of preventable causes, 53; race-specific, 132, 133t., 134, 151; rural and urban, 84fig., 85; U.S. in international context, 63–64; of workers, 92, 101 Moving to Opportunity for Fair Housing (MTO), 89 National Academies of Sciences, Engineering, and Medicine (NASEM), 109 National Academy of Sciences, 53, 54, 65–66 National Alcohol Survey, 176 National Comorbidity Survey, 176 National Health and Nutrition Examination Survey (NHANES), 176 National Health Insurance (NHI), 184, 185 National Health Interview Survey (NHIS), 111, 162, 176 National Longitudinal Study of Adolescent to Adult Health (Add Health), 91, 116, 125, 175 National Survey of Children’s Health, 162 National Survey of Family Growth (NSFG), 176

Index neighborhoods: built environment, 74; environmental dimensions, 118; equal access to green spaces, 190; equal access to schools, 192; health profiles, 4; high SES, 117, 119; immigrants clustering in, 147; lower SES, 107, 119; patterns of health, 76; population health and, 85–89; racism and sexism restrictions, 8, 140–142; segregation, 107, 140–141; social contexts, 10, 12, 20–21, 75; social resources, 138, 139fig.; spatial contexts, 76, 96; and state contextual variables, 81 New York Times, 53 NIH Sexual and Gender Minority Research Coordinating Committee, 176 Nobel Prize, 2 Obama, Barack, 53, 128, 131, 149 Obamacare, 182 obesity: in bisexuals, 177; definition, 48, 65; elevated rates in Hispanics, 135; in gays and lesbians, 176; in immigrants, 148; lower rates in girls, 163; mortality, 63, 68, 74, 82; population health and, 193; risk factor for chronic diseases, 51; U.S. health trend, 48–50; U.S. in international context, 65–67 occupational status, 9, 86, 97, 100–103, 138, 192 Office of Disease Prevention and Health Promotion, 175 Office of Management and Budget, 129 Omran, Abdel, 26–28, 30–31, 35, 37, 38–40 one-drop rule, 131 opioid abuse, 48–49, 51, 155, 183–184, 193 Organization for Economic Co-operation and Development (OECD), 61–63, 65, 69, 193 Pampel, Fred, 172 panethnicities, 131 parasitic diseases, 38, 52, 55, 56fig., 138 Pasteur, Louis, 33 period-specific measures, 16–18 personal hygiene, 31 Phelan, Jo, 8, 99, 116, 120, 125, 189 Piketty, Thomas, 106 pneumonia, 26, 27t., 33, 40, 44t., 133t., 162t., 164t., 165 poisonings, 1, 44t., 48, 138, 154–155 policies: economic, 190–191; and explanations, 74, 180–181; and fundamental cause theory, 184, 188; health as a basic human right, 179; health care, 181–182,

|

259

184–187; individualism, 180–181; measures of, 179–180; non-health social, 190–194; population health, 8, 12, 22–23, 96; public health, 74, 139, 187, 189, 190; socioeconomic, 120, 192; supply-side, 187 Pollack, Craig, 104 Population Assessment of Tobacco Health Study, 175 Population Education Transition Curve, 122 population health: built environment, 67–68, 75; changes in, 25; chronic disease reduction, 51–52; cohort-specific measures of, 18–19; data patterns of U.S. inferiority, 54–57; definition, 7–9, 20; demographic perspective, 11–12, 21; disparities in, 8, 11, 19–22; educational attainment and, 98; epidemiologic transition, 25, 26–27, 35–37, 38; family contexts, 92, 94–96; fundamental cause theory, 8; geographic patterns, 21; health behaviors, 63–67, 74–75; health care system, 57, 59–63, 75; history of, 20, 24–25; importance of age, 14–16; improvement, 97; indicators of, 53; infectious and parasitic diseases, 52; international context, 20–21, 53–54, 57–59; literature, 98; measures of, 12–19, 21, 53; modern trends, 20, 40, 42–46; neighborhoods and, 85–89; policies, 8, 12, 22–23, 96; problematic trends, 46–51; public health, 6; race and ethnicity, 21–22; and religious involvement, 93–94; research on, 21; rural-urban differences in, 83–85; in schools, workplaces, and social networks, 89–92, 96; social contexts, 21, 96; social policy, 74–75; socioeconomic inequality, 68–71, 73–74, 75; sociological approach to, 12; spatial context, 20, 76, 96; state- and county-level differences in, 77–82; study of, 6–7, 9–12, 20; subgroups, 35–38; trends, 3, 5–8, 11–12, 19–22, 25; U.S. in international context, 24, 53–54, 57, 58, 61–62, 73, 75 population representativeness, 11 poverty: in African Americans, 37; association with birthweight, 88; children in, 95, 99, 110–111, 116; driving women’s mortality in Africa, 171; high, 89, 140; low, 89; older adults in, 116; rate, 70, 77, 82, 86–87; segregation and, 140–141; social policy, 193–194; U.S. in international context, 69

260

|

Index

pregnancy, 29, 55, 56fig., 91, 161–162, 167, 169–170, 209n10 Preston Curve, 58–59, 72 Preston, Samuel, 17, 37–38, 49, 62–63, 64, 160 Project Affirm, 175 Project on Human Development in Chicago Neighborhoods (PHDCN), 88 PSA tests, 45 psychological traits, 10–11 public health: achievements, 187; action, 45; advances, 37, 52; advice, 188; battles, 49; clinics, 60; developments, 34, 41; efforts, 33, 34, 40, 45; emergency, 46; infrastructure, 20, 28, 29, 38; initiatives, 34, 39; interventions, 85, 190; measures, 32, 35; officials, 35; policies, 74, 139, 187, 189, 190; population health, 6; practices, 31; programs, 34, 47, 59; projects, 182 race: Black-White mental health paradox, 151–154; definition, 128–132; disparities in health and mortality, 127, 132–138, 139–144, 156; Hispanic health paradox, 22, 144–148, 151, 156; immigrant health paradox, 144–148, 151; measures of, 128–132; rising midlife mortality rates, 154–156; social stratification, 20 racial segregation, 140–141 racial stratification, 127, 139–140, 143, 151, 156 racism, 8, 21, 38, 127, 129, 139–144, 148, 152, 156 radiation, 46 Recognize Intervene Support Empower Study, 175 Republicans, 106 research findings, 8 respiratory diseases, 26, 27fig., 39, 41, 44t., 56fig., 74, 132, 133t., 162t. Robinson, Michelle, 128 Rogers, Richard, 167 Ross, Catherine, 87, 101 rural areas, 31, 34, 67, 76, 78–79, 83–85, 96, 191–192 rural mortality penalty, 83 Saez, Emmanuel, 106 salmon bias, 146 Schoeni, Robert, 193 schools: access to, 192; funding for, 107; high education related to good health,

98; high-quality, 117–118; institutionalized discrimination, 10; patterns of health in, 76; preschools, 74; public, 101, 188; quality, 86; racism in, 141–143; segregation in, 142, 143; smoking on campuses, 45; social contexts, 10, 12, 20–21, 89–92, 96; social resources, 138–139, 182; suicide in, 91; teen pregnancy in, 91; type of, 101, 190; U.S. in international context, 57, 74–75 sex, 158, 166, 173, 174, 177, 209n10, 211n73, See also gender sex assignment, 211n73 sexism, 8 sexual behavior, 118, 176 sexual intercourse, 46 sexuality, 173 sexual orientation, 6, 20, 21, 22, 176–178 smallpox, 25, 28, 32, 188 smoking: behavioral characteristic, 102, 122, 138, 147–148, 153, 165, 172, 177, 188; behavior norms, 86, 90; cause of death, 45; cause of heart and cerebrovascular diseases, 43; declines in, 43, 46, 49, 189; fight against, 44–45, 187; gender gap, 167, 171–174, 183; leading to early menopause, 168; lower rates of, 71; measures of health, 147; patterns of, 22, 165, 172; policies, 82, 139, 189, 211n52; quitting, 108; risk factor for chronic diseases, 51, 121; socioeconomic disparities in, 189; susceptibility to, 11; U.S. in international context, 63–67 Snow, John, 32 social deprivation, 70–71 Social Security, 103, 105, 146, 190 social stratification, 9–10, 12, 20, 96, 100, 105 socioeconomic status (SES): causal effects of SES on health, 123–125; contemporary disparities, 109–114, 125; critiques of SES as a fundamental cause, 123–125; definition, 99–100; disparities in health and mortality, 108–109; fundamental cause theory, 8, 99, 114–118, 120, 125, 127, 138, 191; of individuals, 7–8; measures of, 86, 101–104, 125; of neighborhoods, 86–87; policies, 120; relationship between health and, 21, 97–99; reproduction of socioeconomic disparities in health, 120–122; social context of fundamental causes,

Index 122–123; social stratification, 20; stratification, 104–107, 140, 156; subgroups defined by, 6 sociology, 6, 9–10, 12, 123 spatial patterns, 76, 96 stroke, 30, 39–40, 41, 42, 44t., 49, 51, 62, 161 sudden infant death syndrome (SIDS), 121 suicide: among adolescents, 91; cause of death, 1, 27t., 39, 44t., 48, 68, 133t., 134, 154, 162t., 166; data, 152; mortality, 44t., 48, 62, 68, 133t., 134, 151, 154–155; rates, 10, 17, 48, 134, 151, 155, 161, 162t. Supplemental Security Income (SSI), 190 supply-side policy, 187 surgery, 46, 164t., 165, 185, 211n73 Tax Cuts and Jobs Act (TCJA), 106–107, 191 taxes: cigarette, 8, 11, 45, 80–81, 187; health care paid by, 59, 185–186; income, 95; income inequality and, 105; increases in, 107; policies, 106–107, 169, 190–191; poverty and, 69; property, 107, 192; records, 103, 204n14; reductions, 107, 125; soda, 188 Temporary Assistance for Needy Families (TANF), 190 tobacco, 64, 67, 82, 86, 108, 118, 141, 153, 187, 193 tobacco industry, 43, 45, 52 transgendered population, 22, 158, 174–178, 183, 194, 209n10, 211n73 TransPop, 175 transportation, 31, 50, 67, 74, 85–86, 90, 119 Trans Survey, 175 tuberculosis, 26, 27t., 32, 33, 41, 210n44 typhoid, 33

|

261

Umberson, Debra, 90 “Unhealthy America,” 53 United Nations, 172, 179–180 Universal Declaration of Human Rights, 179 urban areas, 5, 31, 34, 76, 79, 83–85, 96, 107, 192 urban mortality penalty, 83 US News and World Report, 77 Waite, Linda, 94 Wang, Haidong, 160 waste disposal, 83, See also garbage removal wealth: 8–9, 12, 21, 58–59, 72–73, 85–87, 97, 99–101, 104, 105, 106, 107, 114, 117fig., 123, 125, 138, 144, 156, 168, 174, 178, 191, 192; concentrated in urban places, 85; economic resources, 12, 86–87; educational distribution of immigrants and, 144; flexible resources, 138; health-wealth relationship, 114, 117fig.; health-wealth relationship in international context, 58–59; inequalities in, 8, 9, 104, 105–107, 123, 125, 156; key dimension of SES, 21; policies, 107; Preston Curve, 72–73; SES and, 87, 97, 99–101, 104 Whitehall Study, 101–102 Williams, David, 114–115, 116 Women, Infants, and Children (WIC) program, 190 Work, Family & Health Network (WFHN), 92 workplaces, 10, 12, 20, 34, 76, 89–90, 92, 93, 96, 118, 187 yellow fever, 28 Zajacova, Anna, 109 Zika virus, 52 Zimmer, Zachary, 95