Health Disparities in the United States PDF

ealth THIRD edition Hisparities “itUnited states WLLL iddddddddddaddddddddddddddddddddddadddddidddddddsddddddddddddddddsdddédsééide social class, race, ethnicity, and the social d...

ealth THIRD edition Hisparities “itUnited states WLLL iddddddddddaddddddddddddddddddddddadddddidddddddsddddddddddddddddsdddédsééide social class, race, ethnicity, and the social determinants of health DONALD A. BARR, MD, PhD Health Disparities in the United States  ALSO BY THE AUTHOR Introduction to US Health Policy: The Organization, Financing, and Delivery of Health Care in America, fourth edition (Johns Hopkins University Press, 2016) Introduction to Biosocial Medicine: The Social, Psychological, and Biological Determinants ofHuman Behavior and Well-Being (Johns Hopkins University Press, 2015) Questioning the Premedical Paradigm: Enhancing Diversity in the Medical Profession a Century after the Flexner Report (Johns Hopkins University Press, 2010)  Health Disparities in the United States ~ Social Class, Race, Ethnicity, and the Social Determinants of Health THIRD EDITION Donald A. Barr, MD, PhD VW JOHNS HOPKINS UNIVERSITY PRESS BALTIMORE  © 2008, 2014, 2019 Johns Hopkins University Press All rights reserved. Published 2019 Printed in the United States of America on acid-free paper 2, AS Geis S98, S58. Published in previous editions as Health Disparities in the United States: Social Class, Race, Ethnicity, and Health. Johns Hopkins University Press 2715 North Charles Street Baltimore, Maryland 21218-4363 www.press.jhu.edu Library of Congress Cataloging-in-Publication Data Names: Barr, Donald A., author. Title: Health disparities in the United States : social class, race, ethnicity, and the social determinants of health / Donald A. Barr, MD, PhD. Description: Third edition. |Baltimore : Johns Hopkins University Press, 2019. | Includes bibliographical references and index. Identifiers: LCCN 2018057306 |ISBN 9781421432571 (hardcover : alk. paper) | ISBN 1421432579 (hardcover : alk. paper) |ISBN 9781421432588 (pbk. : alk. paper) | ISBN 1421432587 (pbk. : alk. paper) |ISBN 9781421432595 (electronic) | ISBN 1421432595 (electronic) Subjects: |MESH: Health Status Disparities |Socioeconomic Factors |Healthcare Disparities—ethnology |Ethnic Groups |Prejudice—ethnology |United States Classification: LCC RA418.3.U6 |NLM WA 300 AA1 |DDC 362.1089—dc23 LC record available at https://lccn.loc.gov/2018057306 A catalog record for this book is available from the British Library. Special discounts are available for bulk purchases of this book. For more information, please contact Special Sales at 410-516-6936 or [email protected]. Johns Hopkins University Press uses environmentally friendly book materials, including recycled text paper that is composed of at least 30 percent post-consumer waste, whenever possible. For Adam  2a ; Pen eng alblSad aap " oe Rene I ; pee cieweng. thccphaadante 2 De St Se = ON teed ieee tae” of | ; : BS doe patents a “/ a _v s hited, at hoes, : Lene ae a mat [eae hater ak # yf © by? peyeares so : “Mee> hn, PaaS ahi, ; a ee bos Us Rae mote meat 'F ait ths coal Masaya oftlh ‘Diaeiatsls: Sire eae ike tiwns, hind obtaine, fea Stearns a ts Span emaiegies ones We Sit anc SAT | — ~~ ies 19 Neate tay sa: : odin |009 + da aan ei are habaeba a i eer i, Maas Dee abiat = Sara ce yu oe, scaled ai nee Ap | — en OSes (ates I eee | ader  CONTENTS i Preface xi 1 Introduction to the Social Roots of Health Disparities 1 A New Challenge: Falling Life Expectancy in the United States 3 ” Where Does HealthCome From? 4 2 What Is “Health”? How Should We Define It? How Should We Measure It? 13 Six People: Healthy or Unhealthy? 13 Defining Health at the Level of an Individual 15 Health as a Multidimensional Concept 22 The Danger of Approaching Health as a Moral Imperative 29 Measuring the Health of aCommunity or aSociety 30 Comparing Health Status Globally 33 3 The Relationship between Socioeconomic Status and Health, or, “They Call It ‘Poor Health’ foraReason” 39 Terminology: Socioeconomic Status or Socioeconomic Position? 42 What Is Socioeconomic Status? How Should We Measure It? 42 Is Race a Measure of SES Independent of Education, Income, and Occupation? 48 Developing a Model of the Various Measures of SES 51 Social Anomie and a Loss of Trustin Society 55 Understanding the Nature of a Social Hierarchy 55 The Association of Transition among SES Categories and Health Outcomes 60 The Social Roots of Health Inequality: “The Status Syndrome” 62 4 Understanding How Low Social Status Leads to Poor Health 64 The Physiological Response to Stress: The Hypothalamic-Pituitary-Adrenal Axis 65 The Physiological Effects of Chronically Elevated AllostaticLoad 68 Stress in the Workplace, Hypertension, and the Whitehall Study 71 Economic Inequality as a Magnifier of Differences in Social Status 74 Social Capital: Explaining the Association between Inequality and Health 80 Measuring Subjective Social Status: Where Do |Stand on the Ladder of Society? 88 The Other Face of Social Capital: Social Networks 90 Building a Causal Model of the Association between Low Social Status and Poor Health 94 A New Social Determinant of Health: Despair 96 Distribution of Opioid Deaths by Age, Race/Ethnicity, and Geography 98  Vit Contents 5 Race, Ethnicity,and Health 103 What Are Race and Ethnicity? 105 The Overlap between Race and Ethnicity 107 The History of the Use of “Race” as a Category in the United States 111 Is There a Genetic Basis for Race? 113 Problems in the Continued Use of Race to Measure Health Outcomes 121 Can We Use Hispanic Ethnicity to Predict Health Status? 129 What Race Is andIsn’t 129 6 Race/Ethnicity, Socioeconomic Status, and Health: Which Is More Important in Affecting Health Status? 131 Linking Race and Ethnicity to Socioeconomic Status 135 Does Race/Ethnicity Still Affect Health Status after Taking into Account Socioeconomic Status? 137 Race/Ethnicity, Socioeconomic Status, and Geography 144 Race/Ethnicity, Socioeconomic Status, and Infant Mortality 145 The “Latina Paradox”: Reduced Infant Mortality Despite Reduced Socioeconomic Status 151 Sources of the Persisting Association between Race and Health Status 153 7 Children’s Health Disparities 168 The Changing Demographics of Children in the United States 170 The Importance of Early Child Development as a Source of Health Disparities 171 Mischel’s Marshmallow Experiment and the Predictive Ability of Childhood Behaviors 177 Disparities in Childhood Asthma 181 Confronting Pollution in the Water Children Drink: The Case of Flint, Michigan 187 Disparities in Childhood Obesity 188 Addressing the Causes of Childhood Health Disparities 195 8 All Things Being Equal, Does Race/Ethnicity Affect How Physicians Treat Patients? 196 Residential Segregation and Access to Medical Care 197 Racial/Ethnic Disparities in the Treatment of Heart Disease 200 Race/Ethnicity and the Treatment of Cancer 210 Race/Ethnicity and the Treatment of Pain 215 Race/Ethnicity and the Treatment of Kidney Failure 220 Race/Ethnicity and the Treatment of Mental Illness 225 Reviewing the Patterns of Racial Disparity in Access to Care 227 9 Why Does Race/Ethnicity Affect the Way Physicians Treat Patients? 230 The Study of Racial Bias, among Physicians 232 Perceptions of Racial Bias in Care among the Public and among Physicians 236  Contents ix Is Racism on the Part of Physicians a Source of Racial Disparities in Care? 238 Racial Stereotypes as a Source of Racial Bias 243 Unconscious Aversion as a Form of Racial Bias 250 Racial Bias in Police Responses to Civilians: Black Lives Do Matter 253 Differentiating among the Forms of Racial Bias 256 10 When, if Ever, Is It Appropriate to Use a Patient's Race/Ethnicity to Help Guide Medical Decisions? 261 ” Race and the Prevention of Skin Cancer 262 Race and the Treatment of Glaucoma 264 Race/Ethnicity and Breast Cancer 266 Race/Ethnicity and the Treatment of High Blood Pressure 270 Race and the Treatment of Congestive Heart Failure: The First “Blacks Only” Medication 273 College-Educated Women in the Early Stages of Pregnancy 276 11 What Should We Do to Reduce Health Disparities? 279 Eliminate Unconscious Racial Bias and Ethnic Bias as a Cause of Health Disparities 283 Monitor Patterns of Care to Identify and Eliminate Unconscious Racial/Ethnic Bias 286 Strengthen the Physician-Patient Relationship, Especially When Physician and Patient Are from Differing Backgrounds 291 v Increase Diversity of the Health Professions to Reduce Disparities 297 Access to Care as a First Step to Reducing Health Disparities 300 Addressing the Social Determinants of Health as Part of the Health Care System 303 How Far Have We Come in Reducing Health Disparities? 307 References 309 Index 347  Pars, 0a anna aia | : , vel eee -. es + sanasocbtinn iniattineeontel | a ie eRe NOE “HSrev sod kasestan ee | “ cetooy Tae balTo ald anal mate atomeeh wake 8) #2) oy Sone ok,ei tS near RAT 08mta Tag F a ; wit Latte fon Tae AS, sat eee eee a | ie it liaa in ity ae ts fliesnoe Meg hoy A relinn? 5 Rachel bor somantenk 2ta os,nee am aT at a wpa Atfacharg Agalth Steal ue GRE. ap he bt ME.. re cr ae a ee ee Le _ Pmpaiemineaareinimeiren eh iy OL. ar’ LL Sachs das the Ute Ss ba a ae sramcbeasanii rec ‘tae /E oa Mais A AR ine a = ‘oie oat a ~ TN Paige, oot wy toe gb aapPhe fared Chetty i) | 7 ; iene 7 nO uly Oe ores of gpm Neale Digroettap- We, Ts: md 5 ahaa ineg cud, Poli he Pn en ca ne tik of Sage gition anc tvs rac ig= Leet eee Peebles tehe eepatveng Bt S| eens ie re ested of Canter ay ae ee pean u aye ae ie Saat wsStabe» 2 naoer meaton ; “> h :a 1. a) i > a ie y i? “a X ee ‘ PA v. oe Poy, ieee esa " + >, n eee a - nae  PREFACE a One of the best things about being a college professor is that I have the opportu- nity to learn from my students. This book grew from things I learned working with one student in particular. She came to my office and asked if I would help her under- take an independent research project as part of the undergraduaté honors program at Stanford University. I asked her to describe the issue she would like to research. Her response was both immediate and focused: she wanted to know if African Americans in California who develop multiple myeloma, a form of bone marrow cancer, were referred by their physicians for bone marrow transplantation less often than whites with multiple myeloma. It was unusual for an undergraduate to have a project in mind that was so fully conceptualized. I asked how she had developed her interest in and knowledge of the treatment of multiple myeloma in African Americans. She proceeded to tell me the story of her mother. Her mother was in her late forties. As part of the routine care for her high blood pressure, she had undergone a series of blood tests. Her only symptom had been fa- tigue, which she had attributed to her work as a high-ranking health official in a midwestern state. An abnormally high level of protein in the blood led to additional testing and an eventual bone scan and bone marrow biopsy, establishing a diagnosis of multiple myeloma. The cancer specialist to whom she had been referred explained to her and her husband (also a health professional) that, even though the myeloma was in an early stage, little could be done. He had advised her to resume her work and return for chemotherapy when the disease became more symptomatic. Despite a sense of hopelessness, she persisted in learning more about her disease. She consulted the National Institutes of Health to obtain more information and sought a second opinion about treatment options from another cancer specialist. The second oncologist immediately recognized the importance of intervening be- fore the disease became more advanced. He acted quickly to initiate chemotherapy and referred her to a regional center for consideration for a bone marrow transplant, a treatment that had been shown to have substantial benefit in early stages of the disease. She was recommended for immediate transplant, which she underwent four months after her initial diagnosis. While she was on the transplant ward of the hospital, an African American nurse commented how pleased she was to be taking care of another black person. For the nurse, that was unusual, for, despite the higher incidence of multiple myeloma in  xii Preface the black population, nearly all the patients receiving transplants at the center were white. My student explained to me that the first cancer specialist was white, while the second specialist—the one who, in the student’s mind, had saved her mother’s life—was black. I had the opportunity to speak with the mother several years after her transplant, and I learned that at that time she remained in remission and was feeling fine. I helped the student link up with a cancer researcher at Stanford. The student was able to obtain the records of all persons in California diagnosed with multiple myeloma over a three-year period. Linking this cancer registry data with data about all patients discharged from the hospital in California during that same time pe- riod, she was able to determine that, even after taking into account socioeconomic status and the availability of health insurance, African Americans in California with multiple myeloma were less likely than whites to be treated with bone marrow transplantation. The student is well on her way to becoming a leader in the medical profession, having combined her medical training with graduate training in public health fo- cusing on the issue of health disparities. Before working with this student, I had been unaware that the incidence of multiple myeloma in African American women was 2.4 times greater than that in white women in the United States. For African American men, the rate is twice as high (data from the US National Cancer Insti- tute 2018). As well-done research so often does, the student’s project raised more questions than it answered, at least for me. Why was the rate of myeloma so much higher in blacks than in whites? There are few genetic links associated with the con- cept of race as conceptualized in the United States; I had difficulty attributing the higher incidence to genetics. Was it the environment? Knowing that the educational attainment and income level of African Americans is, on average, lower than that of whites, is it a matter of socioeconomic status? Yet the student’s mother had a graduate degree and was a highly respected health professional. Why, also, did cancer specialists in California provide bone marrow transplanta- tion, a treatment proven to be effective, to African Americans less often than to whites? I had been practicing medicine in California for more than 20 years and had never seen a case of conscious racism on the part of my physician colleagues. I had recently had the disturbing experience of treating a patient with a sore knee, whom I describe in Chapter 9, so I knew that my colleagues sometimes unconsciously applied inappropriate racial stereotypes in their treatment decisions. What was going on? Being a sociologist aswell as a physician, I decided to try and find out what was going on. I immersed myself in the literature describing the nature, causes, and  Preface xtii consequences of racial and ethnic disparities in health status and access to high- quality health care. What I found seemed crucially important to medical practice and delivery of health care in the twenty-first century. I decided to organize a course for undergraduates at Stanford, many of them, like their student colleague before them, on their way to becoming future leaders in professions such as medi- cine, public health, and the law. The course has been a remarkable success, typi- cally drawing a diverse group of undergraduates. More recently, I have also been asked by the Stanford School of Medicine to broaden my focus to include issues pertaining to health disparities among children. This book is a direct outcome of the teaching I have done. To teach effectively a topic as complex as health disparities, one must have avail- able research literature from a wide range of academic disciplines. As the title of this book suggests, I will be looking at both the association of race and ethnicity with health status and the impacts of a range of other social determinants on health status. In the past decade especially, there has been growing attention to the social determinants of health (SDH). The World Health Organization (WHO) defines the social determinants of health as “the conditions in which people are born, grow, work, live, and age. These circumstances are shaped by the distribution of money, power...” The WHO then goes on to say, “The social determinants of health are mostly responsible for health inequities—the unfair and avoidable differences in health status seen within and between countries” (World Health Organization 2018). The Centers for Disease Control and Prevention of the US Department of Health and Human Services define SDH as, “the complex, integrated, and overlap- ping social structures and economic systems that are responsible for most health inequities” (CDC 2018). The types of social determinants cited by the WHO and the CDC often exert their effect on health status independent of race and ethnicity. In the United States especially, race and ethnicity have played a powerful role in affecting health status. In this book I refer to a range of research addressing both race and ethnicity and SDH as drivers of health disparities. Simply presenting research data, however, is not conducive to effective learning. The data need to be threaded together into a theoretical framework. In this book, I have attempted those tasks. Chapter | offers an introduction to the concept of health disparities, emphasiz- ing how the United States has lower levels of life expectancy and higher rates of in- fant mortality than nearly all other developed countries, despite all the money Americans spend on health care. Rather than being a result of health care expen- _ditures, disparities in health status appear instead to be a reflection of social and economic inequality. Additionally, we will see that the association between social  xiv Preface inequality and health status is a continuous one, across all levels of income and education. In Chapter 2, I consider the question, “What is health?” Before comparing health status across social, racial, or ethnic groups, we must first appreciate that the concept of health can be defined in different ways and measured with different scales, de- pending on the context. A consideration of six different individuals, each with a form of poor health, demonstrates that the way in which we define and measure health will largely determine how healthy or unhealthy is any particular individual. Chapter 3 builds on the headline of a newspaper story from 1997: “They Call It ‘Poor’ Health for a Reason.” The chapter defines the concept of socioeconomic sta- tus (SES) and traces the consistent association, over time and across places, be- tween poverty and poor health. By looking at data gathered by the federal govern- ment, we will see evidence of this association for a wide range of conditions. In Chapter 4, I extend the exploration of SES to look at the issue of inequality, the multiple forms inequality can take, and the association between inequality and poor health across the SES spectrum. I introduce the concept of allostatic load— the physiologic response to the stress of being in a position of social disadvantage, which over a period of years can result in physiologic injury and illness. Because much of the inequality I explore is inequality among racial or ethnic groups, Chapter 5 attempts to explicate the concepts of race and ethnicity as they have been used in the United States. As early as the mid-1700s, the principal racial categories used today by the US Census Bureau were described by scientists as bio- logical in nature, representing fundamental divisions of the human species. I inves- tigate the research looking at racial categorization as biologically derived versus that suggesting it is socially constructed. Additionally, I look within racial groups to see how ethnically heterogeneous they are. For example, various ethnic groups within the race defined as “black” actually have strikingly different health status and health outcomes. Chapter 6 summarizes the data showing consistently that those in minority ra- cial or ethnic groups are likely to be in a position of low SES. If the health status of those minorities is consistently lower than that of a comparable population of whites, is it because those with poor health tend to be minority, or tend to have lower SES, or both? The chapter presents consistent evidence that being in a minor- ity racial or ethnic group can be a form of social disadvantage in and of itself, even after taking into account SES. Many of the same questions are asked in Chapter 7, while focusing on health disparities among children. With children of Hispanic and other immigrant par- ents comprising a rapidly increasing share of the US population, the demographics  Preface — xv of the United States is changing. In parallel with this demographic change has been our growing understanding of and appreciation for the powerful effects social in- equality can have on early child development, both psychological and biological. The inequality of poverty and residential racial segregation experienced by many children, in particular black children in low-income families, has been shown to contribute to observed disparities in conditions such as asthma and obesity. Iden- tifying and implementing policies for early intervention in the sources of children’s health disparities will be a crucial step toward reducing future disparities when today’s children are adults with children of their own. Chapter 8 moves from disparities in health status to examining a second type of disparity: disparity in access to health care. A principal determinant of access to health care in the United States is the availability of health insurance. As is the case with health status, those from lower SES groups in the United States also have worse access to health care, based on this economic fact of life. However, a grow- ing body of research has shown that, even when people have the same level of health insurance and are treated for the same disease by the same physicians and hospitals, those from minority racial or ethnic groups often get worse care—either not receiv- ing care when appropriate or receiving care that is lower in quality. In Chapter 9, I present research on racial or ethnic bias on the part of physicians, distinguishing unconscious bias from the conscious racism that plagued the United States for much of the twentieth century. For a variety of medical conditions, a con- sistent stream of research has shown that, based largely on unconscious processes, physicians in a number of settings provide a different level of care to blacks or other minorities than they do to whites. Chapter 10 discusses when, if ever, a physician or other health practitioner is jus- tified in using racial or ethnic categorization in deciding the course of a patient’s treatment. The chapter looks at recent trends toward and potential problems with race-based pharmaceuticals—using one drug for white patients, and a different drug for black patients with the same condition. Chapter 11 starts with the caveat that not all disparities need to be eliminated as a matter of public policy. Irrespective of race or ethnicity, women live longer on average than do men. However, racial and ethnic health disparities are of a different type from many gender disparities. In this book I have chosen not to include an ex- tended discussion of gender disparities, nor of other disparities such as those based on age, sexual orientation, or disability. In narrowing my focus, I do not mean to suggest that these other types of disparities are unimportant. Rather, they exist in ‘a different context and stem from different causes, and they deserve their own ex- amination and policy analysis.  xvi _—Preface I end this book by suggesting a framework with which we can identify those disparities stemming from SES, race, or ethnicity that need to be reduced or elim- inated as a matter of public policy. Before proceeding into the chapters themselves, I would like to explain the no- menclature I use in this book. How to name racial or ethnic groups has been an evolving topic with both social and political overtones. As part of the 2010 census, the US Census Bureau identified the five historic racial categories, using the follow- ing nomenclature: White. Black or African American. American Indian and Alaska Native. Asian NO. Oe RS SG Native Hawaiian and Other Pacific Islander The US Census Bureau then identifies a second division, independent of race and based on ethnicity, with two possible categories: 1. Hispanic Origin 2. Not of Hispanic Origin To simplify the text, I have taken the first word or words from each of these cate- gories to describe the categories. I speak of whites, blacks, Asians, American Indi- ans, and Native Hawaiians, meaning by these terms the full description provided by the Census Bureau. In addition, when I use the term “Hispanic,” I will mean “Hispanic of any race”; when I use the terms “white” and “black,” I mean “white and not of Hispanic origin” and “black and not of Hispanic origin.” At many places throughout this book I reference data provided by the Centers for Disease Control and Prevention of the US Department of Health and Human Services. In these citations, I simply refer to the CDC, the common acronym for this essential federal agency. In the list of references at the end of this book, I will include these citations under the heading “US Department of Health and Human Services, Centers for Disease Control and Prevention” and then list them by date and by topic. In the late 1980s, the Council on Ethical and Judicial Affairs of the American Medical Association undertook a study of racial disparities in health status and health care that persisted more than 20 years after the enactment of landmark civil rights legislation. Their conclusion was not encouraging: “Persistent, and some- times substantial, differtnces continue to exist in the quality of health among Americans. Blacks have higher infant mortality rates and shorter life expectancies  Preface xvii than whites. Underlying the disparities in quality of health among Americans are differences in both need and access” (1990, p. 2344). Fifteen years later, Dr. Nicole Lurie, a former federal health official then at the RAND Corporation, looked over the previous 15 years of research and dialogue, and concluded “Less Talk, More Action.” For most of the areas studied, disparities between white patients and black pa- tients have not substantially improved during the past decade or’so. Rather than simply dismissing these findings as more documentation, how might we use them to enhance our knowledge and inform strategies to eliminate disparities? More widespread redesign of systems—particularly, outside of the traditional health care system—will be required to address the complex interplay of social determinants of health and health care outcomes, and this change will probably be longer in coming. (2005, pp. 727-29) In 2009 the Commission on Social Determinants of Health issued a report titled “Action on Health Disparities in the United States.” Looking at the continuing pat- tern of disparities in the United States and our continued low ranking relative to other developed countries, the Commission recommended increased national atten- tion to the need to “improve the circumstances in which people are born, grow, live, work, and age,” and to “tackle the inequitable distribution of power, money, and resources—the structural drivers of conditions of daily life” (p. 1170). The national effort to reduce health disparities is a continuing one. I hope this book, by contributing to the education of future—and current—health profession- als from all fields, will provide a meaningful contribution to that process.  ‘, oe slo Dalene hs |atepie ae sum abe er sie ie series vee “UA beatae OREN "erSebago gio te iredy latin meh a «sik olveirl va ietel nee oat nied tsa pie niute, bec Maeuapats goers bpd re tT ae iors Se be apie so evens aL RA =. ee ia hy ‘in Vityst omc eyties Ape aie ot A af dbassvisio Scud oe fiers ’ $3 salty rake Sprit tbs be ‘ s ar a 12 “i +45 t +e abe phe Teel], 92 , mii ia byron alahinatades or ha h*; i mie?wy hie 2Site Wwe ft Uinwiors ale St) on cpecdisn ati surlwalrreesb viqnaly 1 Maptraspei > or Shignnaers peau ris agbnlgoad rue Sonaat tag Weds aida Lie sleperbeelel ohn ney arene Ye sai norrepeadyier sah Ren , tuisee Nee He teypeakahi) pee feet sary alleteed omer:boniigaes b ob Hier tniat eee Sere dulsstl donPaatilie wsetipeds bat vcteciine ae bata by ‘ cucsit Ww) cupedenserah tive Haaren jel Oley Cpa apple £006) guicsan wt wyael od -- betas xa sow baeat io, Ms endinrereott tient rronentret peinas ih PRO RS Sede beat rteperal. ri) shoes, [tea betes: vats oj sive dah oo nomahe or witetn ani ngy Betis boanilswoo Lewy Dine dynae: Salat? sii cat 6 whrinacpiste ter. veyed inode,seerignl bystveni Una BEd iTS 2) ash heqoksrb dee. 7 owury weal eA nbyasy dilisiee ni erertanantnls wisreprengeet” ax linet ste sok” a seni alist ‘ies did verdant ae igs vey sage ue deca cath Wades hue? dead oe PTae tan wide Mermaid binary aguetllgeaeaamnaenl's- ‘nematrouttbl ite aud Vee go tier nitysi nt eA rata siting yee ligtisanbhed iene Aoi eae aE ares ts:Sesite eon corde ne ead ote riatnsho rel tenth Y 0 esata 08 citi ter yal one ae i“ 7“ flair 53 Pig cas (iis Bega Ayah ite PdSud ie We me : w “Aires Cees! «i Paredes oo chy thy Bitiss atv oe hale shined Veet ia chete pitetiond, takale whee 4éictae cere re PReeo eororqem pyaar? tect” orc. be ihe tie he atraltya y oh eae aly Mac share ac etn lag the hac Cl cal mien ol int; errr dint: Onipoade Koma pelt onBh ‘ id thant: ik: : (edhe ars RB, the Vanna hp tli has yeppenciede eee em rt he wae ent tt ae are ae Mace ae thes BP sr narchnne oir Tapp actor. Tgate-cerbainasekss i ere. MPuiitios Hence oe a ee be gubbuomecis. —< Health Disparities in the United States  b e t i n ’ o e ? a l 2oltt' aid ztade x ne = P { ee Ait 7 = x hives a ee ae I ; Ly 4dr) teen whan ren) A , ‘  CHAPTER ONE SSS Introduction to the Social Roots of Health Disparities SS Why are some people healthy and others not? This is a question that has gained increas- ing attention in the United States over the last several years. It is also the title of a book published in 1994. The contributors to that book come to two basic conclusions. First, in trying to improve health in developed societies, we have tradition- ally focused most of our attention on improving the quality and availability of health care. Modern societies devote a very large proportion of their economic resources to the production of health care... Such massive efforts reflect a widespread belief that the availability and use of health care is central to the health of both indi- viduals and populations. (Evans and Stoddart 1994, p. 27) Second, most of the variability in health status we find in the United States and other developed countries has little to do with health care and everything to do with one’s position in the social hierarchy. Some of the best-kept secrets of longevity and good health are to be found in one’s social, economic, and cultural circumstances.... The largest gap lies between the richest and the poorest. But the middle classes are also affected. The lower one is situated on the social hierarchy... the lower one’s probability of staying in good health and the lower one’s life expectancy. (Renaud 1994, p. 322) Nowhere has the emphasis on improving health through advancing the technology of health care been more evident than in the United States. As shown in Figure 1.1, the proportion of the US economy that goes for providing health care rose from just under 11 percent in 1988 to 17.9 percent in 2016. Some of the most recent increases in health care costs are a result of the expan- sion in access to health insurance under the Affordable Care Act (ACA). Most of the cost increases are due to our heavy investment in new medications, new facili- ties, and the new technology of advanced procedures, reflecting the faith of the  2 Health Disparities in the United States 19% 18% 5 17% 16% - 15% 14% = 13% 12% 1% 0% 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Figure 1.1. National health expenditures as percentage of GDP, 1988-2016. Source: US Centers for Medicare and Medicaid Services, National Health Expenditures. American public in the power of health care to improve health status. Yet, in compar- ing ourselves to other developed countries that have made different policy decisions about investing national resources in health care, we find a striking inconsistency. We spend much more of our national economy on health care than any other developed country in the world (Table 1.1). However, using two common measures of popula- tion health, our population has worse health status than the population in any of these other countries. Life expectancy estimates how many years, on average, a baby born today can expect to live. (Given consistent difference between males and females, this figure is typically broken down by gender.) Babies born in the United States in 2015 can expect to live between two and six years less than babies born in other developed countries. Infant mortality estimates the following statistic: of a thousand babies born alive, how many will die before their first birthday? We see that the United States has the highest infant mortality rate of any of the countries listed. In fact, in 2015 the United States had a higher infant mortality rate than 31 of the 35 countries in the Organ- isation for Economic Co-operation and Development, an association of the most developed countries in the world. The US infant mortality rate was only better than the rates found in Chile, Turkey, and Mexico. i  Introduction to the Social Roots ofHealth Disparities 3 Table 1.1. Health Indexes, Selected OECD Countries, 2016-17 % GDP Spent on Infant Mortality Male Life Expectancy Female Life Expectancy Country Health Care (2017) Rate (2016) at Birth (2017) at Birth (2017) Se ee ee Japan 10.7 2.0 81.0 87.1 Sweden 10.9 2.5 80.6 84.1 France HS 37 79.2 85.5 Germany 11.3 3.4 78.6 83.5 Switzerland 12.3 3.6 817 85.6 Greece - 8.4 42 78.9 s 84.0 Canada 10.4 47 79.8 83.9 United Kingdom 97 3.8 79.4 83.0 United States 17.1 59 76.1 81.1 Source: Data from OECD, http://www.oecd.org/statistics/, accessed 7/2/18, Infant mortality data for Canada are from the World Bank, https://data.worldbank.org/. A New Challenge: Falling Life Expectancy in the United States In December 2017, the National Center for Health Statistics, the principal govern- ment data source that tracks population health statistics, reported some startling news. After rising consistently every year since 1950, life expectancy in the United States fell for two consecutive years (Kochanek et al. 2016). There had been some warning earlier that the rate of decline in death due to cardiovascular disease was decreasing (Sidney et al. 2016). While death rates due to heart disease and cancer— the top two causes of death—continue to decrease, albeit at a somewhat slower rate, the death rate for unintentional injuries, the third leading cause of death, in- creased 17 percent between 2014 and 2016. Unintentional injuries include acci- dents and poisonings. The principal increase in this category has been due to what the US Department of Health and Human Services (2018) has described as the “Opioid Epidemic,” involving deaths from prescription opioids as well as illicit opi- oids. Often death from an overdose of opioids is intentional rather than acciden- tal, suggesting a substantial overlap between accidental opioid deaths and suicide by opioid overdose. Oquendo and Volkow (2018) describe suicide as “A Silent Con- tributor to Opioid-Overdose Deaths.” They conclude that “The significant in- creases in both opioid-overdose deaths and suicide rates in our country have con- tributed to reduced life expectancy for Americans” (p. 1569). Case and Deaton (2017) refer to these deaths as “deaths of despair.” Stein and colleagues (2017) have analyzed the patterns of these deaths over time and conclude that they occurred principally among non-Hispanic whites living outside large urban areas, and “were ‘ primarily caused by self-destructive health behaviors likely related to underlying so- cial and economic factors in these communities” (p. 1541).  4 Health Disparities in the United States I discuss these issues in greater depth in Chapter 4. There is extensive data con- firming the impression that deaths of despair and the opioid epidemic are strongly associated with social, racial and ethnic, economic, as well as geographic differences among population groups. Where Does Health Come From? How is it that we, as a country, invest so much of our economy in providing health care, but we get so little in return in the form of health—at least health as measured by these common statistics? Perhaps our basic assumption—that more health care will lead, necessarily, to better health—is flawed. Perhaps something other than health care drives the health of a community or a society. Economist Victor Fuchs addressed this issue in 1983 in his seminal book Who Shall Live? Fuchs tells us “A Tale of Two States” (pp. 52-54), comparing demo- graphic characteristics and health status in two adjacent states in the United States—Nevada and Utah. Using data from the 1960s and 1970s, he found that the two states had similar populations using measures such as income, education, age distribution, and access to health cave. However, the health of those who live in Nevada was substantially worse than that of those who live in Utah. Infant mor- tality was 40 percent higher in Nevada than in Utah. Death rates were consistently higher in Nevada for all age groups—about 40 percent higher for young adults, 50 to 70 percent higher for those ages 40 to-49. When one looks at two specific illnesses with known causes, the differences in death rates are even more striking. The com- bined death rate in Nevada from lung cancer (associated with cigarette smoking) and cirrhosis of the liver (associated with alcohol abuse) ranged from 100 percent higher to nearly 600 percent higher than the comparable rate in Utah, depending on the age range and the gender studied. You will no doubt recognize why Nevada and Utah had such strikingly differ- ent health statistics, despite having populations that were generally similar in de- mographic characteristics and their access to health care. The influence of the Mor- mon Church in Utah has led to much lower rates of smoking and alcohol abuse in that state. In addition, those in Utah experienced lower rates of divorce and migra- tion. Nevada, on the other hand, has an economy and a culture that includes higher rates of smoking and alcohol consumption, higher divorce rates, and higher rates of geographic migration. Life in Nevada is fundamentally different from life in Utah. Those lifestyle factors, and not the availability of health care, drove the dif- ferences between the twb states in terms of health status. As described by Victor Fuchs, “The basic finding is the following: when the state of medical science and  Introduction to the Social Roots ofHealth Disparities 5 SSS o Chemotherapy — 500 Ss = 400 BCG a Vaccination 2 300 e = 5 200 a 100 0 —+—+__+——_+——_+ —+ Sas — 1930 1940 1950 1960 1970 Year Figure 1.2. Mean annual death rates (standardized to 1901 population) from respiratory tuberculosis, England and Wales, 1935-70. Sources: Evans et al. (1994); McKeown 1979, other health-determining variables are held constant, the marginal contribution of medical care to health is very small in modern nations.... For most of man’s his- tory, [per capita] income has been the primary determinant of health and life expectancy—the major explanation for differences in health among nations and among groups within a nation” (1986, pp. 274-76). To illustrate the relationship between the availability of health care, the stan- dard of living, and mortality rates, let us look at death rates from tuberculosis (TB) in England and Wales. There was a decline in the death rate from TB fol- lowing the discovery in the late 1940s of the first drugs that were effective against TB and the development in the 1950s of the first vaccine used to prevent TB (Figure 1.2). It appears that the advent of antituberculosis drugs and a tuberculosis vaccine— two key additions to the health care regimen available to treat TB—were effective in reducing the death rate. However, when one takes a longer time frame, the pic- ture that emerges is quite different. Figure 1.3 shows the decline in the death rate from TB in England and Wales from 1840 to 1970. From the longer-range perspective, it appears that the health care developments of the mid-twentieth century had relatively little effect on the overall death rate from TB. In fact, half of the decline in the rate of death from TB took place before the tubercle bacillus responsible for TB was discovered. While the number of deaths declined by 51 percent between 1948 and 1971 following the introduction of medi- cations and a vaccine, most of the overall decline had taken place before any effec- tive medical treatment was available (McKeown 1979).  6 Health Disparities in the United States 4,000 3,500 3,000 2,500 2,000 BCG 1,500 million) (per Rate Death 1,000 500 1840 1860 1880 1900 1920 1940 1960 Year Figure 1.3. Mean annual death rates (standardized to 1901 population) from respiratory tuberculosis, England and Wales, 1840-1970. Sources: Evans et al. (1994); McKeown 1979. If health care was not responsible for the decline in death rates from TB, then what was? Consistent with Fuchs’s comments above, the rising standard of living seen in England and Wales in the nineteenth and twentieth centuries can explain the falling death rates. Better nutrition, better sanitation, better housing, and less crowding, combined with public health measures to prevent the spread of TB, ac- counted for most of the decline in death rates over the last 200 years. To illustrate this point, and to underscore the role income and standard of liv- ing play in reducing the chance of death from TB, let us consider two fictional persons, one from the world of music and one from the field of literature. The opera La Bohéme, first performed in 1896, tells the story of Mimi, a poor seam- stress struggling to survive in the Latin quarter of Paris. She meets and falls in love with Rodolfo, an equally poor poet who has to burn the pages of a play he was writing to stay warm. The opera tells us of the love between Rodolfo and Mimi and of Rodolfo’s hesitance when he learns Mimi has TB. The lovers struggle to stay together (and to stay warm), only to have Mimi.die tragically, succumbing to her disease. Thomas Mann first began writing his novel Magic Mountain in 1912. It tells the story of Hans Castorp, the son of a well-to-do German family who goes to visit his cousin in a TB sanatorium high in the Swiss Alps. What was originally intended as a stay of only a few weeks extends to seven years of intensive therapy, as it turns out that Hans himself has developed TB. Seven years of sumptuous meals, fresh air, exercise, and companionship help Hans to fight off his disease, only to be drafted  Introduction to the Social Roots ofHealth Disparities vA into the German army at the beginning of World War I (a conflict in which he is likely to be killed). What if their roles had been reversed? If Mimi had been from a wealthy French family and Hans from a poor German family, which one would have survived TB? For more than a century, where you are on the social hierarchy has been a strong predictor of whether you live or die from a disease such as TB. The standard of living available to those lower on.the social hierarchy has been a powerful predictor of how many of them would die from the many infectious dis- eases that ravaged Europe and North America during much of the twentieth century. From 1900 through 1970, the United States experienced a steady decline in the overall death rate. As was the case in England and Wales, the rising standard of living seen in the United States during this period was associated with a steady fall in the rate of death from infectious diseases such as measles, tuberculosis, pneu- monia, diphtheria, typhoid, and polio. It was also the case that the antibiotics and vaccines to treat these diseases were among the important medical discoveries of the twentieth century. However, as with the case of tuberculosis in England and Wales, most of the decline in the death rate for each of nine most feared infectious diseases occurred before the medical treatment for that disease was discovered (McKinlay and McKinlay 1997). Better health, in this case measured by reduced death rates from infectious diseases, was associated principally with rising levels of income and improvements in the standard of living. Advances in medical care played a smaller role in reducing the death rate from these illnesses. By the end of the twentieth century, deaths from infectious diseases had declined substantially and were largely replaced by deaths from three major chronic diseases: heart disease, cancer, and stroke. With the tremendous advances in medical care seen during the last part of the twentieth century, we might expect to see death rates from these and other chronic diseases falling substantially. Yet, when we look at the actual death rates in the United States from the six leading causes of death, adjusted for changes in the age distribution of the population over this time, we see a mixed record of success (Jemal et al. 2005). The death rate from heart disease and acci- dents showed a sharp decline from 1970 through 2002, while the death rate from stroke declined at a slower rate. The death rate from cancer and diabetes changed relatively little during this time, and the death rate from chronic lung disease actu- ally rose. By 2011 heart disease, cancer, chronic lung disease, and stroke accounted for 59 percent of all deaths in the United States (National Center for Health Statistics, 2013). If measles, tuberculosis, and pneumonia have now been replaced by heart disease, cancer, and stroke as principal causes of death, will Fuchs’s point—that  8 Health Disparities in the United States differences in income are the principal determinant of differences in health status— still hold? An answer to this question has been provided by the Whitehall study, conducted over a period of several decades in England. The study has been follow- ing employees in the British Civil Service, recording many aspects of their health status over time. It compares the health of four principal groups of employees: 1. the administrators who work at the highest ranks of the Civil Service 2. the professional and executive employees who carry out much of the work of the Civil Service, under the supervision of the administrators 3. the clerical workers who provide the staff support for their professional and executive supervisors 4. the other workers, not in any of these employment categories, who clean the floors, take out the trash, and serve the food in the cafeteria As one might expect, the level of education and training required to work in each of these categories is quite different. The administrators tend to have more educa- tion than the executive employees, who in turn have more education than the cleri- cal workers, and so on down the line. Also not surprisingly, the income of each category tends to be higher than the income of the category immediately below it. Each category of worker is fully employed, and each tends to work in similar physical surroundings. Each has access to the full level of health care provided by the British National Health Service. Will we see the same hierarchical difference in death rates in British Civil Service workers that we saw between Mimi and Hans when faced with tuberculosis? Figure 1.4 provides an answer to this question. Figure 1.4 graphs the cumulative death rate from all causes over 10 years of ob- servation for each of the four categories of employee described above. It seems clear that administrative employees have a substantially lower rate of death than the jani- tors, cafeteria workers, and “others” in this category. If Hans had been an admin- istrator in the British Civil Service and Mimi a maid who cleans the bathrooms for the Civil Service, she would be more likely to die during this time period than he. However, the comparison between the highest class of worker and the lowest is only part of the story. Each category of worker has a lower death rate than the category immediately below it. With each step down the employment hierarchy, health deteriorates and the chance of death goes up. Repeating Renaud’s com- ment from the beginning of this chapter, “the lower one is situated on the social hierarchy... the lower one’s probability of staying in good health and the lower one’s life expectancy” (Renaud 1994, p. 322). The Whitehall study is not a study of how poverty is related to the chance of death or illness. It is a study of people who are employed and who have regular ac-  Introduction to the Social Roots ofHealth Disparities 9 18 =Zz 6 Other — So = et 2 Ss g& 2 Clerical 5 = 10 =Ss 8 Professional/ rawv wu A Executive 8 evs i= & 5 A aS eur Administrative uv = 2 2 £ Ss 4 iL | Year of Follow-Up Figure 1.4. Cumulative probability of death based on occupational category within the British Civil Service—the Whitehall study. Source: Marmot et al. 1984, used by permission of Elsevier Limited. cess to health care. It is important to appreciate that the relationship between in- come and health is not dichotomous, with only those who fall below some thresh- old of poverty suffering the health consequences of being at the bottom of the social hierarchy. Even though the majority of research published between 1975 and 2000 on the relationship between social class and health focused on the effects of pov- erty, it is clear from the Whitehall study and numerous others that the relationship between social class and health—between one’s position on the social hierarchy and one’s likelihood of either illness or death—is a continuous relationship that spans all levels of the social hierarchy, from the very lowest to the very highest (Adler and Ostrove 1999). Figure 1.5 illustrates the two conceptual models of the relationship between in- come and health. The horizontal axis measures income—typically family income. The vertical axis represents the likelihood that an individual at a given income level will enjoy good health, whether measured by the rate of illness or the chance of death. In the threshold model, those who live in poverty have lower health status than those who live above the poverty line (the poverty line measured by the federal government and used to qualify those who fall below it for a variety of benefits). Once a family receives a level of income sufficient to meet its basic needs for food, clothing, and housing, there is no further health benefit to a rising standard of living. By contrast, the continuous model in Figure 1.5 suggests that there is, indeed, a health benefit to escaping poverty. However, the farther a family gets from  10 Health Disparities in the United States -- mie =-- Continuous - - ma - - Health Level Oe Status oon Threshold Income Level Figure 1.5. Two models of the association between income level and health status. poverty—the higher its income and associated standard of living—the better its health becomes. Even though neither lives in poverty, an executive employee brings in more income than a clerical employee and accordingly enjoys better health. That better health status extends to the members of the executive’s family as well. An administrative employee, in turn, enjoys a higher standard of living and better health than the executive employee who works under their direction. This is the message of the Whitehall study. Health psychologist Nancy Adler and colleagues have looked thoroughly at the association between socioeconomic status (SES) and health status. (SES measures status within the social hierarchy according to a number of measures in addition to income. Chapter 3 details how income and SES are associated.) In a series of pub- lications (Adler et al. 1994; Adler and Ostrove 1999), they summarized the re- search in this area that firmly supports the continuous model of this association. They analyzed the results of eight separate studies, each measuring the association between SES and the mortality rate—either the overall adult mortality rate for those at a given level of SES or the infant mortality rate.-In each of these studies, with every step down the SES hierarchy, the chance of death, either adult death or in- fant death, goes up. Adler and colleagues also summarized studies investigating the association be- tween SES and health measured as the rateof certain chronic illness within a given SES group. Again, the relationship is clearly a continuous one. With every step down the SES hierarchy, the chances go up of having arthritis, high blood pressure, and other chronic diseases that reduce one’s quality of life. Of course, this does not mean  Introduction to the Social Roots ofHealth Disparities il that every person in a lower SES category will have a higher rate of illness and an earlier death than every person in a higher SES category. The mortality rates and illness rates are statistical averages that accurately state the likelihood that an indi- vidual within the indicated category will have the outcome that is being measured. In 1900 a poor seamstress was more likely to die from tuberculosis than was the son of an affluent family. In 2015 a poor seamstress was more likely to contract high blood pressure and arthritis, to have her infant die before its first birthday, and her- self to die earlier than the son of an affluent family. While the citcumsiances in which Mimi and Hans Castorp lived have changed dramatically over the decades following their fictional lives, with countless medical advances, their health status relative to each other has changed little. This association between social status and health status holds both for the United Kingdom and for the United States. In the United Kingdom, all residents are pro- vided health care through the British National Health Service. The United States has taken a more fragmented and incremental approach to providing health care to its residents. Rather than a right of citizenship, health care in the United States has historically been treated largely as a market commodity, available to those who are willing (and able) to pay for it. As described in 1986 by health economist Uwe Rein- hardt: “Americans have... decided to treat health care as essentially a private con- sumer good of which the poor might be guaranteed a basic package, but which is otherwise to be distributed more and more on the basis of ability to pay” (Reinhardt and Relman 1986, p. 23). Since the 1960s the federal and state governments have provided coverage for cer- tain of the most vulnerable segments of our society, including the elderly and the very poor. However, in 2012, two years after the Affordable Care Act (ACA) was signed into law, 48 million people, most of them in low- to moderate-income working fami- lies, had no health insurance and as a result had little access to health care (US Census Bureau 2013). Beginning in 2014, the changes enacted as part of ACA reduced the number of uninsured Americans by nearly 18 million people, while still leaving 27 million people without health insurance (American Community Survey 2016). In understanding the association between social position and health in the United States, we cannot ignore the very real economic barriers to health care that still exist. It seems clear that income and other measures of social position affect health directly, as well as affecting it indirectly through impaired access to health care. This relationship is illustrated in Figure 1.6, in which the width of the arrows indicates the strength of the relationship. Before we can understand these relationships and the causal factors behind them more fully, we first need to look at the SES concept in more depth. What are the  2 Health Disparities in the United States Socioeconomic Health Status Status Access to Health Care Figure 1.6. The association between social position, access to health care, and health status. measures of SES other than income? Does it matter how long one is in a position of disadvantaged SES? Chapter 3 explores these and other aspects of the relation- ship between SES and health in more depth. Before looking at the relationship between SES and health, we must first under- stand what we mean by “health.” Is there more than one way to measure health? Which is more important, health as perceived by the individual or health as mea- sured by objective scales or the impressions of health professionals? These are the subjects of Chapter 2.  CHAPTER TWO What Is “Health”? How Should We Define It? How Should We Measure It? In Chapter 1 we saw that the United States fares relatively poorly when compared to other developed countries using measures of a population’s health status such as life expectancy and infant mortality. In order to compare the health of individuals rather than of countries, however, we need different types of measures. How “healthy” one person appears, or feels, relative to another person will depend to a large extent on how health is defined and measured in this context. Consider, for example, the following descriptions of six different people. Each is drawn from an actual person I have known in my years as a practicing physi- cian and professor. As you read the descriptions, ask yourself the following two questions: 1. Who is the healthiest? 2. Who is the least healthy? Six People: Healthy or Unhealthy? 1. The CEO of a large company who has between 6 and 10 alcoholic drinks every 24 hours 2. A student at Stanford University in the middle of studying for final exams 3. A teenager born weighing two pounds who exhibits problems with coordi- nation and cognition 4. A paraplegic person (that is, someone with a spinal cord injury who is para- lyzed from the waist down) who bikes to work each morning as an engineer at a major high-tech company (using a hand-crank bike) 5. A teenager who is 25 pounds overweight and is sad and discouraged because she has few friends * 6. A 94-year-old who has lived in the same house for 20 years and who has the beginnings of kidney failure and is in the early stages of Alzheimer disease  14 Health Disparities in the United States When I present this list to my students at Stanford, there typically is little consen- sus on the answer to either question. The CEO is obviously successful and prob- ably has a substantial income. He would be at the top of most social hierarchies. However, he is an alcoholic, exhibiting the characteristics of chronic alcohol abuse identified by the National Institute on Alcohol Abuse and Alcoholism. There are few outward signs that this person is unhealthy—no heart disease, with blood pres- sure within reasonable limits. Yet a physician who has accurate information about his alcohol consumption would have to view him as unhealthy. Most Stanford students enjoy excellent physical health. Yet each year a few stu- dents develop potentially serious emotional problems when under the stress of fi- nal exams. Fortunately, the exams last only a few days, and with proper support and ‘counseling the students who experience the emotional cost of the stress triggered by exams recover completely and are ready to resume classes after a short period of rest. There has been careful research on the health outcomes of babies born prema- turely at extremely low birth weight (usually defined as weighing less than 1,000 grams at birth). A two-pound baby weighs less than 1,000 grams. Twenty or 30 years ago such an infant faced a substantial risk of experiencing fairly high levels of per- manent disability as a result of prematurity. Some of these former preemies have significant problems with muscular coordination and mental ability. How should we view a teenager in this situation? Rather than developing physical disabilities as a result of prematurity, many otherwise healthy individuals, at some time during their childhood or early adult- hood, are involved in an accident that damages their spinal cord, leading to perma- nent paralysis from the waist down, known medically as paraplegia. How are we to view a person with this level of disability? Does it matter that this person gradu- ated from a top university and has a successful career as a computer professional, or has superb cardiovascular health based on an excellent diet and regular, vigor- ous exercise? A teenager who is 25 pounds overweight is probably obese (that is, the ratio of a person’s height to weight exceeds the guidelines established by the National In- stitutes of Health). There are few abnormal signs or symptoms due to the obesity for this teenager. The health effects of obesity more typically show up in adulthood. However, this person, in addition to being obese, is consistently sad and discour- aged and has few friends. (In Chapter 4 we will see that having few friends and so- cial contacts is associated with worse health outcomes independent of other factors.) I would be concerned about this person’s health—her emotional health now and the long-term health consequences of obesity and social isolation later.  What Is “Health”? 1b) Finally, we have the 94-year-old. Clearly, she has chronic medical problems that will affect her daily life and may eventually lead to her death. Can we consider a person such as this to be healthy? When students pose this question, I respond: what would this woman say herself? Few 94-year-olds have the luxury of being able to stay in their own home until the end of their life. This particular 94-year-old had a regular circle of friends who visited her every day, took her to church every week, and helped her with shopping and going to the doctor.-As 94-year-olds go, this one was fairly healthy—at least, she thought so. i It should by now be clear that there is a principle we must acknowledge in mea- suring health: How healthy or unhealthy a person is will depend on which definition of health we use. There is an associated principle that we must also consider: How you define health will depend on the level at which you analyze health, with different measures used: — at the level of an individual — at the level of a community or neighborhood — at the level of a country or society Defining Health at the Level of an Individual The World Health Organization (WHO) attempted to define what “health” means for an individual anywhere in the world. The constitution of the WHO was first adopted at the International Health Conference held in New York in 1946. The preamble to that constitution states: “Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.” The intent of the global community adopting this definition was to make it clear that the health of any individual is measured not simply by the presence or absence of disease. Health involves health of the body, health of the mind and the emotions, and health of the social context in which one lives. While this definition has clear value in broadening our understanding of health, it has limited value from a pol- icy context. It defines a level of health that few can ever hope to achieve. I certainly have relatively few moments when |am in such a complete state of well-being. My arm is often bothered by tendonitis; sometimes I have a hard time sleeping because of stressful issues at work. I often wish I could find more time to spend with friends. In addition, it is a rare student in my class who can state that they are in a state of  16 Health Disparities in the United States complete health. The potential danger in setting what for many may be an unre- alistic goal is that it puts us in the position of always falling short, no matter how hard we try. Beyond setting an expectation that few can hope to attain, this definition gives us little guidance on how to measure health. It seems to create a dichotomy: either we are healthy, or we are not. We have no way of stating, for example, that we are “79 percent healthy.” Based on this definition alone, we have no way of following our health over time to determine whether it is improving or declining. Similarly, we have few mechanisms to compare one individual with another. This is not to say that the WHO definition is without merit. On the contrary, by creating an awareness of the many aspects of health, it gives us an opportunity ‘to explore the three axes of health the definition includes: physical health, mental health, and social well-being. I explore each of these aspects of individual health in sequence, and then ask whether there might be a way to combine them into a com- posite measure of health that can be used in a policy context. Physical Health: The Medical Model Sociologist Andrew Twaddle provided a definition of health that the US medical profession relied on for much of the twentieth century. He stated that “health must be understood first as a biophysical state” and that “illness is any state that has been diagnosed as such by a competent professional” (1979, pp. 145-46). Twaddle iden- tified two fundamental dimensions of health according to the medical model: 1. an absence of symptoms—sensations noticed by the patient and interpreted as abnormal 2. an absence of signs—objective characteristics noted by a health professional, of which a patient may often be unaware This approach to defining health tells us what the concept of health is not. A per- son with abnormal signs or symptoms is not healthy. It does not tell us what health is. In the practice of medicine, this might be referred to as a “rule-out” definition. One looks for the presence of abnormal signs or symptoms, and when one deter- mines that there are none, it is possible to rule out ill health. If one does not have ill health, then one is by definition healthy, at least from the medical perspective. There are potential problems, though; with using this approach in isolation. What if the patient and the doctor disagree? Whose definition of “health” then takes precedence? Say, for example, a person is bothered by a headache (an abnor- mal symptom) and goes to a doctor for advice. Knowing that headaches can be  What Is “Health”? 17 either benign, with no specific treatment indicated, or serious, with aggressive ther- apy indicated, the doctor will look for abnormal signs on tests such as a physical examination or a CAT scan. Having looked and found no abnormal signs of illness, the doctor might reassure the person that they are healthy (that is, the doctor has “ruled out” ill health). However, the patient may still feel the headache and may both feel unhealthy and expect to be treated as unhealthy. What are we to make of a condition that has no abnormal symptoms? An impor- tant example of this is high blood pressure, also referred to as hypertension; per- sons with hypertension develop symptoms only after a number of years. Should we consider a person with somewhat elevated blood pressure to be unhealthy, based on our knowledge that their blood pressure will eventually lead to further problems? What might be the consequences of labeling such a person as “unhealthy,” even if they feel fine? Another potential problem with the medical approach to health are the problems that have been identified in the reliability of what would otherwise seem to be ob- jective evidence of abnormal signs of illness. On tests such as EKGs, CAT scans, and certain laboratory tests, different individuals reviewing the tests might inter- pret the same test as either normal or abnormal. There have even been studies in which the same reviewer has been given the same set of tests to interpret on two oc- casions, separated by a period of time, with some of the interpretations differing— even though it was the same test. This phenomenon is thought to contribute to the marked variation in the rates at which certain procedures, such as prostate or heart surgery, are done on similar populations of patients. Wennberg and colleagues doc- umented these “area variations” in a number of contexts, variations that seem to involve doctors in different parts of the country interpreting differently what should be considered a “normal finding” and what constitutes an “abnormal finding” (Wennberg et al. 1982; Wennberg 1993). Health as Functioning at a Normal Level: The Sociocultural Model Sociologist Talcott Parsons looked at health as reflecting the extent to which an individual is able to maintain a normal level of functioning within their social context. He noted that “health may be defined as the state of optimum capacity of an individual for the performance of roles and tasks for which he has been social- ized” (1979, p. 122). Parsons does state that “al/ processes of behavior on whatever level are mediated through physiologic mechanisms.” However, his approach to health focuses more on what a person is able to do with their body than on the  18 Health Disparities in the United States physiological state of that person’s body. Also, rather than comparing a person’s level of functioning to some lofty ideal state, as the WHO definition does, it looks at the person in the context of their own social circumstances. From this perspec- tive, a person with a medical condition that consistently has abnormal signs or symptoms who is nonetheless able to'perform customary roles and tasks should be considered healthy. Thus, from this perspective, health is not the absence of some- thing (signs and symptoms), but rather the presence of something—the ability to function normally. While there are some attractive aspects of adopting this perspective on health, it too has some potential drawbacks when used in isolation to define health. The same level of physical functioning for two individuals may imply different states ‘of health, depending on the social roles and tasks they face. Thus, a concert vio- linist with arthritis in their fingers may be seen as unhealthy and deserving of in- tervention because of the extent to which the arthritis impairs their ability to perform. A house cleaner with the same degree of arthritis in their fingers may warrant little attention, as it is possible to go on cleaning houses despite the dis- comfort of the arthritis. Given the degree to which inequalities continue to exist in educational opportunities, with Inequalities in educational attainment leading to strikingly different socially defined roles and tasks, adopting the sociocultural model of health in isolation may perpetuate inequalities according to race, eth- nicity, or gender. As we will see when we take another look at the health of the 94-year-old de- scribed above as patient 6, the ability to perform common tasks is especially important in evaluating the health of elderly people and of those with physical dis- abilities. There are five common tasks that most of us take for granted; being un- able to perform them directly affects the quality of a person’s daily life. These tasks are commonly referred to as activities of daily living (ADLs). They are: feeding one’s self bathing one’s self dressing one’s self being able to use the toilet without assistance wWN YRbeing able to transfer one’s self without assistance (for example, from a bed into a chair) It should be apparent that a person who is unable to perform any of the ADLs with- out assistance is in a state of relatively poor health from this perspective. A simple summation of the number of ADLs a person requires assistance with is commonly  What Is “Health”? 19 used as a measure of health and disability. Often this measure is used to determine a person’s eligibility for extra benefits, such as home health care. Health as a General Feeling of Well-Being: The Psychological Model The first two models of health rely on an assessment of an individual’s health by an independent evaluator. The perceptions and attitudes of the individual in question affect the health assessment, but do not determine it. A third model of health relies on the individuals themselves to provide an assessment of their own health. For ex- ample, it is possible to ask an individual a question such as, On a scale of 1 to 10, how would you rate your overall feeling of well-being today? A number of scales have been developed to measure a person’s health based on that person’s own perceptions. These measures are often time specific; a person may give one answer in the face of an immediate stressor and another answer a short time later after that stressor has been removed. Think, for example, of patient 2 above, the Stanford student during final exams. While this student may report a rather low sense of well-being in the midst of final exams, the same student may report a sub- stantially improved well-being after finals are over and they have had a chance to spend time with family and friends at home. Since 1996 the US Agency for Healthcare Research and Quality (AHRQ) has conducted an ongoing survey of the health status of the US population, re- ferred to as the Medical Expenditure Panel Survey (MEPS). One of the ques- tions they ask of individual respondents is how they would rate their overall health status, based on the following five categories: Excellent, Very Good, Good, Fair, or Poor. Davis and colleagues (2017) have analyzed the pattern of responses to this question for older adults in the US, defined as age 65 or greater. They cate- gorized those who responded either Excellent or Very Good as “healthy” adults, and then followed the trends in the number of healthy adults in the US between 2000 and 2014. Overall, the number of healthy older adults increased from 14.0 million (42.4 percent) to 22.4 million (48.2 percent). The authors also found that, after controlling for age and gender, both African American and Hispanic respondents reported lower rates of health, as did those from lower incomes and those with lower levels of education. The authors report that, “our new finding regarding widening gaps by race/ethnicity and socioeconomic status is concern- ing. This finding supports the notion of ‘2 different Americas’ having formed” (p. 1684).  20 Health Disparities in the United States Another Look at Our Six Patients Having considered three ways to define health (medical, sociocultural, psychologi- cal), we can now take another look at our hypothetical six patients to reconsider who is the healthiest and who is the least healthy. The CEO ofa large company who has between six and 10 alcoholic drinks every 24 hours. From the medical perspective, a perceptive physician would determine that this person is quite unhealthy. While the individual may have few symptoms to complain of at this time, the physician knows that chronic alcoholism will, in all likelihood, lead to serious diseases and has the potential to impair significantly so- cial role functioning. This is why many physicians ask probing questions about alcohol consumption as part of routine health screening, and many validated screen- ing instruments have been developed to look for signs of possible alcohol abuse. From a sociocultural perspective, this person is doing quite well. So long as he can effectively maintain his role as CEO, the alcohol abuse may not impair his health. Once again, however, this person is at high risk of a substantial decline in health if the alcohol abuse begins to impair his functioning at work. In addition, we know little about this person’s personal or family situation. If the alcohol abuse were associated with impaired social relations outside of work, then we would have to consider this person to have serious health impairment. It is difficult to evaluate this person from a psychological perspective. It may well be that the drinking is masking an underlying psychological problem that the per- son may initially fail to acknowledge. Were the person to stop drinking, that psy- chological impairment may become more apparent. It is difficult to predict how this person would rate his own health from this perspective. In summary, the health of the alcoholic CEO is currently impaired in terms of the medical model, and either impaired or significantly threatened from the stand- point of the other models. We should regard this person as generally unhealthy. A student at Stanford University in the middle ofstudying for final exams. This pet- son may report a very high level of stress from the psychological perspective. Stan- ford students often are short on sleep and experience high levels of anxiety during finals. However, finals only last a week, after which students get at least a week of rest with the opportunity for travel and relaxing with family and friends. For most students, the stress is short-term. From a sociocultural perspective this person is doing quite well. Simply being at Stanford suggests that the student is highly func- tional in a social context. Most Stanford students will have friends and social net- works as well that add to their level of overall social functioning. Similarly, most young adults have little in the way of medical abnormalities that show up as abnor-  What Is “Health”? 2] mal symptoms or signs. So, even though this person might report experiencing high stress and anxiety right now, we have little to worry about regarding the stu- dent’s overall health. A teenager born weighing two pounds who exhibits problems with coordination and cognition. We have little to go on to evaluate the health of this person. Certainly from a strictly medical perspective this person has signs of potentially significant health impairment. The physical and cognitive difficulties many fprmer premature infants face as teenagers often require extra help and support. They may be less able to participate fully in academic, athletic, and social activities. One might expect such a child to feel discouraged or depressed about the life they face. Later in this chapter I return to this issue by looking at research that has actually followed for- mer premature infants into adolescence, and has asked these types of questions of the children, their teachers, and their parents. A paraplegic person who bikes to work each morning as an engineer at a major high- tech company. This person has evidence of a health impairment on only one level— the use of his legs. He appears to have compensated for this impairment quite effec- tively through the use of a mechanical assist device—the hand-crank bike. I would expect him to have excellent upper-body strength (from all that cranking) and su- perb cardiovascular fitness (from biking to work every day rather than driving a car). From a sociocultural perspective this person seems to be doing superbly. Engineers at high-tech companies are generally highly educated, tend to face interesting and challenging work, and enjoy generous incomes. There is no reason to think that this person’s family life has been impaired by paralysis. Unless he finds the paralysis itself so discouraging that it affects his own perceptions of his quality of life, one would expect this person to report fairly high perceptions of his own health. An isolated physical impairment does not necessarily make someone unhealthy, and in the face of other signs of positive health takes on less significance in this regard. A teenager who is 25 pounds overweight and is sad and discouraged because she has few friends. Here is someone we should worry about. This person is showing signifi- cant health impairment on all fronts. The obesity itself will very likely be associ- ated with other serious medical problems, either now or in the near future (diabe- tes and high blood pressure, for example). With few friends this person is showing evidence of poor social role functioning. The sadness and discouragement reported by this teenager is a clear statement of her own perception of poor emotional health. Any health professional who encounters this person would be quite concerned, and would want to evaluate possible interventions on a number of levels. A 94-year-old who has lived in the same house for 20 years and who has the begin- nings of kidney failure and is in the early stages ofAlzheimer disease. At first, most of  22 Health Disparities in the United States my students place this person squarely in the “unhealthy” category, many at the bot- tom of the list. In response, I ask the students to consider if their own grandparent were in this situation, how that grandparent would respond to the question, “How are you doing?” This person represents an actual patient I took care of until she eventually died of her worsening kidney failure. There was treatment that would have kept her alive longer, but she refused it. She said she was happy being who she was, and she wanted to die happy. And she did. For her, the impaired memory and physical weakness from her kidneys was of little importance. What was important was that at 94 she still lived in the house she had lived in for decades and that she needed no help with her ADLs. She cooked for herself, she dressed herself, and she bathed herself. Her many neighbors and friends helped her with shopping, visited her regularly, and took her to church every week. From this person’s own perspec- tive, life didn’t get any better. I knew she had signs of poor memory and an abnor- mal kidney test, but I believe I enjoyed our visits together almost as much as she did. So who is the healthiest? For me, the 94-year-old and the bike-to-work engineer are at the top of the list. Who is the least healthy? From my perspective, it has to be either the CEO or the overweight and depressed teenager. For the Stanford stu- dent I’ll need to wait until after finals to be sure. The former preemie who now has substantial impairments may yet surprise us. Health as a Multidimensional Concept It seems apparent that each model of health we have considered tells us something about a person’s state of health, but tells us little about a person’s overall state of health. Rather than being one-dimensional, health has multiple dimensions, three of which we have considered so far. Wolinsky (1988) suggests that we think of health as a three-dimensional concept. As represented in Figure 2.1, each of the three axes represents one of the three ways of approaching health we have discussed: health as the absence of disease, health as social role functioning, and psychological health. Wolinsky suggests that we dichotomize each dimension into “well” and “ill.” Overall health is measured by the ratio of “well” dimensions to “ill” dimensions. Someone who is well on