Public Health and the Risk Factor: A History of an Uneven Medical Revolution
http://www.100md.com
《新英格兰医药杂志》
"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. . . . 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, professional, and health organisations." This excerpt from the preface provides an excellent summary of this book.
William Rothstein, professor of sociology at the University of Maryland, explains how "the risk factor" arose in life insurance and from developments in population statistics and probability theory. Since the end of the 19th century, major U.S. life-insurance companies have collected sociodemographic data and health data about millions of their policyholders, followed these persons for long periods, and used the data to calculate premiums and benefits.
Initially, the companies used information on the results of urinalysis (to detect kidney disease and diabetes), "build" (i.e., weight in relation to height), medical history, occupation, and place of residence, because their records showed that these factors were strongly associated with mortality rates. Later, blood pressure and smoking status were added. By conducting medical examinations and taking measurements for life-insurance companies, physicians became familiar with the concept of risk factors and incorporated it into their clinical practice.
Risk factors are identified through correlations with diseases, rather than from laboratory evidence of biologic mechanisms. Statistical inference is used to examine associations between multiple risk factors and the probability of disease. The scientific credibility of risk factors accrues from repeated demonstration of the associations in different populations and in different settings, dose-response effects, and reductions in disease after changes to the risk factors.
The second half of the book is about the rise and fall of the epidemic of coronary heart disease (CHD) in the 20th century. Rothstein examines the evidence for the main risk factors for CHD, using the standard criteria for assessing epidemiologic results — chance, bias, confounding, reverse causation, and possible true causation. He relies heavily on life-insurance findings, because they meet many of these criteria. He is relatively skeptical about randomized clinical trials owing to concern about the generalizability of the findings.
Tobacco smoking and high blood pressure meet the criteria for risk factors for CHD and other diseases. The diet–heart hypothesis is where confusion sets in. The evidence is not strong. Advice from the medical profession fluctuates. Rothstein believes that, rather than cholesterol or saturated fat, the relevant risk factor is total caloric intake. The life-insurance data have for many decades demonstrated associations between overweight and CHD and diabetes, yet reducing population levels of caloric intake is not in the interest of the food industry or within the expertise of the medical profession.
In the last 20 pages of the book Rothstein claims that "personal risk factors," such as cigarette smoking and high blood pressure or lipid levels, cannot account for the epidemic of CHD. (In my view, his brief analysis is flawed by an assumption that long latency times are needed.) Rather, he argues that "social and cultural factors" are important determinants of CHD but does not explain how they might account for the major epidemic of the 20th century. This weak ending of the book is disappointing. Nevertheless, I strongly recommend this book to everyone interested in the interface of public health and clinical medicine and in the epidemiology of CHD.
Annette J. Dobson, Ph.D.
University of Queensland
Brisbane, Queensland 4072, Australia
a.dobson@sph.uq.edu.au(Rochester Studies in Medi)
William Rothstein, professor of sociology at the University of Maryland, explains how "the risk factor" arose in life insurance and from developments in population statistics and probability theory. Since the end of the 19th century, major U.S. life-insurance companies have collected sociodemographic data and health data about millions of their policyholders, followed these persons for long periods, and used the data to calculate premiums and benefits.
Initially, the companies used information on the results of urinalysis (to detect kidney disease and diabetes), "build" (i.e., weight in relation to height), medical history, occupation, and place of residence, because their records showed that these factors were strongly associated with mortality rates. Later, blood pressure and smoking status were added. By conducting medical examinations and taking measurements for life-insurance companies, physicians became familiar with the concept of risk factors and incorporated it into their clinical practice.
Risk factors are identified through correlations with diseases, rather than from laboratory evidence of biologic mechanisms. Statistical inference is used to examine associations between multiple risk factors and the probability of disease. The scientific credibility of risk factors accrues from repeated demonstration of the associations in different populations and in different settings, dose-response effects, and reductions in disease after changes to the risk factors.
The second half of the book is about the rise and fall of the epidemic of coronary heart disease (CHD) in the 20th century. Rothstein examines the evidence for the main risk factors for CHD, using the standard criteria for assessing epidemiologic results — chance, bias, confounding, reverse causation, and possible true causation. He relies heavily on life-insurance findings, because they meet many of these criteria. He is relatively skeptical about randomized clinical trials owing to concern about the generalizability of the findings.
Tobacco smoking and high blood pressure meet the criteria for risk factors for CHD and other diseases. The diet–heart hypothesis is where confusion sets in. The evidence is not strong. Advice from the medical profession fluctuates. Rothstein believes that, rather than cholesterol or saturated fat, the relevant risk factor is total caloric intake. The life-insurance data have for many decades demonstrated associations between overweight and CHD and diabetes, yet reducing population levels of caloric intake is not in the interest of the food industry or within the expertise of the medical profession.
In the last 20 pages of the book Rothstein claims that "personal risk factors," such as cigarette smoking and high blood pressure or lipid levels, cannot account for the epidemic of CHD. (In my view, his brief analysis is flawed by an assumption that long latency times are needed.) Rather, he argues that "social and cultural factors" are important determinants of CHD but does not explain how they might account for the major epidemic of the 20th century. This weak ending of the book is disappointing. Nevertheless, I strongly recommend this book to everyone interested in the interface of public health and clinical medicine and in the epidemiology of CHD.
Annette J. Dobson, Ph.D.
University of Queensland
Brisbane, Queensland 4072, Australia
a.dobson@sph.uq.edu.au(Rochester Studies in Medi)