Health Education Research, Vol. 18, No. 3, 404-407,
June 2003
© 2003 Oxford University Press
BOOK REVIEW |
Social Epidemiology
Berkman, LF, Kawachi I (Eds) Oxford University Press, New York, 2000, pp. 391.
Division of Cancer Prevention, National Cancer Institute, Rockville, MD
This is a great book, this is a frustrating book. It is a great book because its intent is to push the field of epidemiology and public health as a whole to look beyond individual risk behaviours to examine the broader social determinants of health. This is especially important now when all the rage at NIH—and dollars—is now headed in the opposite direction, towards proteomics and molecular genetics. It is a frustrating book because it is virtually devoid of any social theory that could provide more solid direction about where to look and how to think about such sociological influences. In a book that aims to set the agenda for research on macro-level social factors, it is striking that the only social theorists to appear in the index are fleeting references to Marx and Durkheim. Not once are the words positivism or post-modernism—issues that have raged across the social sciences for decades—mentioned in these pages. The field of epidemiology is notorious for its atheoretical ‘kitchen-sink’ approach to disease investigations and this book does not stray far from the fold.
In claiming to be the first textbook on ‘social’ epidemiology, the purpose of this book is to promote the study of how social factors affect health and well-being. As the pioneering social epidemiologist Dr. Leonard Syme notes in his Foreword, the text seeks to lay out ‘a more appropriate way to study risk factors and diseases that can fundamentally change our approach to the concepts of etiology and intervention.’ These are laudable and long overdue goals, calling for a fundamental re-casting of how we think about the causes of disease, and consequently, disease prevention and health promotion. Berkman and Kawachi are to be highly commended for challenging a deeply-entrenched research enterprise that has rather myopically located the cause of most modern diseases in the individual host and his/her risky lifestyle behaviours (despite the fact that, as Syme makes clear, after 50 years of research using this framework, the best research to date can still only account for at most 40% of the variance in heart disease). Their text makes an undeniably valuable and important contribution to the field as a whole in legitimating research on the social environment. In this, it is basically advocating for US investigators to catch up with our international colleagues, following in the foot steps of the 1980 Black report, the 1987 Epp report, and the 2000 Acheson report, Inequalities in Health. Yet, while there are many tantalizing hints and suggestions, readers looking for new ways of understanding the role of the social structure in shaping health behaviours, specific measures of new or neglected social factors, or more compelling evidence of their impact—beyond the commonplaces of poverty and racism are bad for people’s health—will be hard pressed to find them here.
Epidemiology has traditionally used the three-legged model of agent, host, and environment to explain disease causation. As historians of public health, most notably Rosen, have shown, the focus of attention in the field historically centered on the role of the environment in explaining disease processes. With the rise of modern science, interest shifted to the role and identification of various agents of disease towards the end of the 19th century. Then, with the remarkable declines in infectious disease rates in the first half of the 20th century and subsequent rise in chronic diseases, attention turned to the final leg of the triangle, the role of the host. Neubauer and Pratt have termed this shift the ‘second revolution’ in public health, dating roughly from the time of the release of the Surgeon General’s Report on Smoking and Health in 1964. It is this inordinate pre-occupation with presumed significance of host behaviours that Berkman and Kawachi are calling into question with their plea for a ‘renaissance of interest’ in the role of the environment. But it is at this point that one wishes the editors had asked harder questions and made stronger distinctions.
Conceptually, the editors and contributing authors make no distinction between the physical and social environments, treating them as if their respective roles in causing diseases were analogous, if not identical. Virtually all of the contributing authors blur the distinction, shifting indiscriminately back and forth between discussing factors in the physical environment (housing, downtown car traffic, stores that sell fresh foods, overhead canopies to prevent sunburn, etc.) and factors in the social environment (racism, sexism, social class, etc.). On a minor note, this seems to undermine the primary purpose of the book, which is to call attention to the role of the social environment.
But more importantly, there is no discussion about whether how we think about the effect of the physical environment might or must be different from how we think about the affect or influence of the social environment. Exposures to any number of physical environmental pollutants are known to cause cancer; are we to think about social factors, such as social inequalities, in precisely the same way in causing illness and disease, a kind of dose-response relationship? Indicative of the scant attention to social theory, most of the contributing authors slide readily into speaking of the social determinants of health without pause, as if there were this direct cause-and-effect relationship and extensive sociological discussions of the complex, problematic nature of the relationship between structure and agency (a la Anthony Giddens) irrelevant. Do ‘social determinants’ cause heart disease in the same way that HIV causes AIDS? The authors never raise the question about whether there is anything theoretically, conceptually, or empirically different between how physical environmental factors impact human health and how social factors do (such as, for example, the capacity for human agency, or the mediating influence of the mind’s interpretation of social events)? Likewise, they pass on the question of whether ‘social forces’ can be measured with ‘tools’ just like, say, electro-magnetic radiation can be quantified. Are ‘factors’ like ‘racism’ or ‘social cohesion’ objective forces with empirical referents, or are they subjective, phenomenological matters of interpretation, social constructions, so to speak? Theoretically, if there are any particular differences or unique challenges in measuring phenomena in the social realm, the authors do not mention the possibility. Methodologically, notwithstanding frequent assertions of the need for a new model, no new alternative research designs are put forward here; the preponderance of evidence cited uses standard case-control and cohort study designs. The editors say that their book is written for graduate students and active research investigators, but for people who are not familiar with the unique moral and interpretive challenges faced by the social sciences, the book may do them a disservice in failing to alert them to current debates. Indeed, they might find themselves in for a rude shock, as it is easy to come away from this book with a rosy, optimistic glow of confidence that factors like racism, social cohesion, social norms, and social capital can be readily measured, thrown into a standard logistic regression model, and voila, heretofore poorly understood and perplexing modern diseases processes will be explained. For those of us who have labored in the social sciences for a while, a more cautionary tale might seem warranted.
The authors frequently cite the work of the sociologist Emile Durkheim and the epidemiologist Geoffrey Rose as the prototype for a new approach to disease investigation. Their work would indeed offer a refreshingly new and welcome approach to thinking about the etiology of disease in society, with far reaching implications for health education. Instead of asking, ‘Why did this individual become sick?’ Durkheim and Rose implore us to ask, ‘Why does this population have this level of disease?’ In his classic work, Durkheim showed that even the most intimate of individual acts, suicide, was consistently correlated with the type of society in which one lived. Individuals may come and go, but societal rates remain constant, solid evidence that something significant is going on beyond individual intra-psychic processes. And hence, by extension, no matter how many suicide-counseling hotlines we might set up, the rates will not change until we change society. So, too, Rose reminds us that, if everyone in our society smoked, we would find the cause of an individual’s heart attack in other factors, since smoking status would not differentiate between those who did and those who did not have heart attacks using standard epidemiological relative risk ratios (even though we know that smoking is the most significant cause). Think about how refreshing it would be to approach currently hot topics like obesity or violence from such a social perspective, asking not why has this individual gained 20 pounds, but why are we, as members of American society, experiencing an epidemic of obesity? And consequently, instead of running around trying to get individuals one-by-one to eat less, we might focus instead on the more widespread, possibly universal, sources of our social malaise. Whatever its specific shortcomings, this is the exciting and worthwhile challenge posed by this book.
According to the editors’ own account, the book is organized into five sections. After an introductory chapter, the first three chapters address socio-economic status, racism, and the size of the gap in income inequalities. The next two chapters focus on workplace stress and unemployment. The next section, on community, deals with social support and that new buzzword, social capital. The fourth section is on psychological factors, with a chapter on depression and another on other affective states (anger, anxiety, etc.). Then there is a large miscellaneous section with six chapters that cross community health education, psychosocial interventions, the social factor-biological mechanism connection, ecological approaches, an overview of multi-level approaches, and a concluding chapter on social policy. The chapters are uniformly well written, well balanced, and thorough. For health educators, the chapters on community health education and psychosocial interventions will contain no surprises. (One might even find that touting church-based interventions or calling for conducting community needs assessments as exemplary of the ‘new’ social epidemiological approach pretty disappointing.) For those who want to know or feel the need to make a more compelling case for plausible linkages connecting social influences with the ultimate outcome of concern, individual bio-physiological breakdowns, the chapter by Dr. Eric Brunner alone is almost worth the price of admission. Likewise, MacIntyre and Ellaway’s discussion and refutation of the ‘ecological fallacy’ and the difference between contextual and compositional influences are excellent.
Capturing the contradictory reactions provoked by this book, the chapter by Dr. Nancy Krieger on racism and other forms of discrimination is both deeply commendable for forcing us to confront the uncomfortable fact that racism is probably the most significant factor shaping the distribution of disease in our society and almost equally deflating in demonstrating how far we have to go in understanding it. Much as I admire her work, is a six-item, yes/no, self-report scale with no psychometrics the best that the field of epidemiology has to offer to assess this important and powerful social factor? Likewise, when the book notes in passing that strong secular trends have mitigated or overwhelmed the effects of planned interventions, one wishes that the author had paused to tell us why or how such strong secular trends came about—isn’t that what this book is supposed to be about, or at least to provide us with a framework for thinking about how to capture and assess such powerful social influences? In its best light, perhaps this dismal state of affairs should be read as a comment on the dire need for this book and its call for a renewed research emphasis on social influences. But it is hard for me to shake the feeling that the current methods and tools of epidemiology will never rise much beyond this level, because human behaviour is more complex than the established epidemiological research designs can accommodate. It may take a much more radical re-thinking of basic scientific assumptions and models to make further progress in understanding why American society has such unusually high levels of obesity, violence, drug use, teen pregnancy, child abuse, infant mortality, crime, and glaring inequalities in health. Along these lines, the selection of contributors might be faulted for its Harvard bias, leaving one wondering why topics like powerlessness, community organizing, homelessness, community-based public health, community competence, participatory action research, and the like were passed over.
So how should we think about the impact of social factors on human health? If one believes that individuals do have some degree of choice or control over whether they decide to eat a big dessert or not, then clearly the use of the term social ‘determinants’ is too strong a word. If there are such things as human agency, free will, autonomy and responsibility, then social factors do not cause us to do anything, in the sense of ineluctably force or make happen, like the chemical reaction that will inevitably ensue when sodium and chlorine are mixed together in a test tube. But it seems to me that an integrated interpretive probabilistic model might move us in the right direction. Here, we can think of various hypothesized unhealthy social influences as ‘insults,’ or assaults, first on our senses, then on our bio-physiological responses. An African-American is turned down for a job. How one reacts to it depends on one’s interpretation of events. For example, is it an insult for a man to open a door for a woman, or not to open the door? Well, it depends. But if one perceives it as an egregious act, then it is likely (although, again, it is important to emphasize, not evitable) that some amount of various stress hormones will be released into one’s blood stream. These chemicals will bump into the body’s biological cells. In a probabilistic model, it may only take one ‘hit,’ one ‘insult,’ to trigger changes in the cell’s functioning to cause it to become cancerous, or it may take many, or it may never happen. But eventually and cumulatively these insults are more likely to wear down the cell’s resiliency. Similarly, social insults may eventually wear down the individual’s psychological resiliency of ignoring or rising above their sting, until something snaps, or we sublimate our feelings through solace in food, or alcohol, or drugs, or sex. Nothing inevitable, or deterministic, but the odds increase with each new insult. The challenge posed by Berkman and Kawachi’s book is whether we can think creatively about and adequately capture the range, type, and level of such social insults on a societal level. The book points to many intriguing suggestions, for example, levels of trust, social stratification by honorific status (prestige, respect), social cohesion, relative deprivation, social norms, social integration, financial strain, social control, job insecurity, social disorganization, investment in human capital, social solidarity, civic participation, cultural consonance, fatalism, cultural assets, material insecurity, and others. As health educators, we want to know, for example, why different colleges have different drinking rates and different levels of drinking problems. We want to know if and why levels of stress or depression in our society are going up (or down) over time. We want to know how to capture the impact of the fast food industry’s deliberate decision to ‘supersize’. The immeasurable contribution of Social Epidemiology is in advocating that such questions be asked.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||