Public Health and the Risk Factor by William Rothstein

“The acceptance of risk factors has produced changes in public health and medicine as profound as those that resulted from bacteriology and the germ theory of disease…Yet the impact of the risk factor has been much more uneven than the germ theory. The risk factor concept has been controversial because of its statistical methodology, its multifactorial concept of disease etiology, and its effect on the economic interests of commercial, profession, and health organizations.” (p. xi.)

Millions are treated daily with drugs designed reduce cholesterol levels in the bloodstream. Americans pay billions of dollars for these drugs because they have been repeatedly told that these medications will reduce their risks of developing coronary heart disease. Despite these claims, health professionals often do not have sound understanding of the methodological and substantive issues originally involved in correlating these risk factors (such as high blood pressure, cholesterol, smoking, diet, and obesity) to coronary heart disease. The risk factor concept may have revolutionized medicine, but it has produced widely uneven results.

In Public Health and the Risk Factor: The History of an Uneven Medical Revolution, sociologist William Rothstein attempts to explain why the outcome of this revolution has been mixed. By examining the history of the risk factor concept and its application to coronary heart disease, Rothstein slowly builds towards the conclusion that the American medical establishment has misinterpreted the evidence of various studies examining coronary heart disease. These erroneous conclusions have distracted the medical community from the most likely causes of heart disease, obesity and insufficient exercise. Instead, physicians and drug companies have developed expensive treatments, drugs, and surgical procedures which address the symptoms, but not the causes of heart disease.

Life insurance companies (especially the Metropolitan Life Insurance Company), not the medical profession, originally developed the concept of risk factors. Not surprisingly, life insurance companies were especially keen to develop accurate mortality tables. It was critical for life insurance companies to identify criteria to quickly evaluate the general health of their potential customers. It was even more important for companies writing industrial life insurance policies because these policies were small and carried slender profits. Industrial insurance companies only made money by selling the policies in large quantities to poor, urban clients. In the United States, tens of millions of these policies were sold and they became a fixture among urban working classes. In order for industrial insurance to be profitable, it was critical to identify criteria that could accurately predict mortality.

Life insurance companies commissioned studies utilizing statistics and probability. Statistics and probability had been used only sporadically by physicians in clinical studies at the time and they were considered by most physicians to be useless. Insurance companies disagreed. Not only did the companies commission studies, they were uniquely situated to fashion large studies, because they could examine their own clients’ insurance claims.

In 1903, the Specialized Mortality Investigation, a study commissioned by the insurance industry, identified risks factors such as “build, occupation, medical history, and residence” as related to premature mortality. The risk factors provided the companies with easily ascertainable criteria which could be used to identify “substandard risks” or “impaired lives.” (p. 63.) Based on these risks, insurance companies could raise or lower premiums accordingly. One of the key factors identified as having a strong relationship to mortality was an individual’s build. Contrary to the prevailing medical opinion of the time, the study found that overweight people were much more likely to meet an early demise. Other studies quickly determined that blood pressure was also useful in predicting mortality. Blood pressure had never been considered especially important to the medical community before its identification as a risk factor.

Not only did the insurance industry identify risk factors, it educated a generation of physicians “about new developments of diagnosis,” created education programs for the general public and provided nurses to their clients. (p. 74.) Throughout the early twentieth century, insurance companies conducted innovative research designed to improve the accuracy of medical diagnosis. Scores of young doctors were hired by insurance companies to examine potential clients and indoctrinated regarding the importance of risk factors in general health. These early experiences shaped their attitudes regarding the utility of statistics and probability in medical research. On the other hand, the Metropolitan’s health education and visiting nursing programs promoted preventive health care among the general population. The primary motivation behind the Metropolitan’s health programs was financial and not philanthropic. Despite the companies mercenary motivations the programs were overwhelming beneficial and widely praised.

As mortality rates from infectious diseases declined in the twentieth century, the medical profession increasingly focused on preventing chronic diseases (such as coronary heart disease). Unlike the insurance industry, medicine was more skeptical of the utility of statistics and probability in identifying the underlying causes of disease. Additionally, physicians were not adept at interpreting their results. Chronic diseases (such as heart disease) have been especially difficult for the medical profession, because it has been challenging to determine the biological mechanism that causes chronic heart disease. Medicine has not been able to identify the root biological causes of heart disease in the laboratory. Instead, physicians have been forced to diagnose heart disease primarly through large statistical studies.

Rothstein argues that a careful analysis of these studies do not necessarily support the conclusions of their authors. The data from the Framingham Heart Study has been used for years, but it focused on only three specific risk factors (smoking, blood cholesterol, and blood pressure). Rothstein posits that results of that study were flawed because its authors narrowed the concept of risk factors and ignored social characteristics (such as “education, income, occupation, living conditions, health care, marital status, place of birth and family structure”) even though they appeared to play an important role in who developed heart disease. (p. 285.) While Rothstein agrees that smoking and blood pressure appear to be valid risk factors, he questions the benefits of treating high blood pressure in isolation and batters the medical profession for failing to link smoking and heart disease until recently despite clear statistically links. The tobacco industry used medicine’s hesitancy to connect smoking and heart disease to its advantage for years. It was not until the 1990s that the tobacco industry even acknowledged the connection between cigarette smoking, lung cancer and heart disease.

The most problematic risk factor for Rothstein is the correlation between blood cholesterol and heart disease. Blood cholesterol’s role in the development heart disease is uncertain and Rothstein directly questions the findings of that a reduction in cholesterol levels prevents heart disease. The side effects of the drugs and the high cost outweigh any speculative benefits. Additionally, the focus on cholesterol has diverted attention from the most likely causes of heart disease, obesity and insufficient physical exercise. Rothstein’s conclusions fit neatly with the historical evidence provided earlier in the book that heart disease was uncommon before the twentieth century. According to Rothstein, obesity and exercise are the single greatest determining factors for heart disease. Unfortunately, physicians are not well-positioned to alter an individual’s weight or mandate exercise. Instead of focusing on these issues, the medical profession has been sidetracked by prescribing blood pressure and cholesterol medicines.

Rothstein does not appear rely on any specific social theory. If anything, Rothstein’s work is more informed by a sophisticated understanding of statistics, probability, and mathematics. Whether or not that constitutes a type of social theory may be worth debating. Additionally, it is arguable whether his work would benefit from the inclusion of any specific social theory. Unlike Warwick Anderson in his work Colonial Pathologies, Rothstein is not particularly interested in the motivations, morality or ethics of scientists involved in the Framingham Heart Study. Instead, he is focused on analyzing whether their conclusions were supported by the accompanying statistical evidence. Rothstein seeks to undermine the contemporary understanding and treatment of coronary heart disease by questioning the validity of the interpretations of these statistical studies. Because the biological cause of heart disease has not been determined in the laboratory setting, it is possible for statistician or sociologist to ascertain risk factors as well as any physician.