摘要：In the epidemiological imagination, the Framingham Heart Study has attained iconic status, both as the prototype of the cohort study and as a result of its scientific success. When the Public Health Service launched the study in 1947, epidemiological knowledge of coronary heart disease was poor, and epidemiology primarily involved the study of infectious disease. In constructing their investigation, Framingham’s initiators had to invent new approaches to epidemiological research. These scientific goals were heavily influenced by the contending institutional and personal interests buffeting the study. The study passed through vicissitudes and stages during its earliest years as its organizers grappled to define its relationship to medicine, epidemiology, and the local community. THE FRAMINGHAM HEART STUDY has attained iconic status in the epidemiological imagination. Initiated in 1947, the study endures, having left in its wake more than 1200 peer-reviewed articles. A pioneering effort in the epidemiological investigation of chronic noninfectious disease, it has made rich methodological use of the term “risk factor,” 1– 3 which it popularized. 4 Framingham’s results, reinforced by those of similar contemporary investigations, 5– 7 mapped the relations of coronary heart disease (CHD) to factors such as serum cholesterol, blood pressure, and cigarette smoking. Unlike those other epidemiological studies, however, it analyzed such factors in women, who constituted more than half of its participants. The fact that Framingham has followed its cohort for more than half a century has allowed investigation of a spectrum of diseases, including those of old age. According to Mervyn Susser, Framingham “is the epitome of successful epidemiological research, productive of insights and applications . . . [and] the prototype and model of the cohort study.” 8 (p31) Few modern textbooks in epidemiology miss the opportunity to refer to it. 9– 12 In 1947, epidemiological knowledge of CHD was spotty; morbidity incidence and prevalence rates from unbiased samples were almost nonexistent. Mortality statistics, collected by the government 13, 14 and the insurance industry, 15 revealed the weight of heart disease, variously defined. The newly formed National Heart Institute (NHI) reported that, by 1948, 44% of deaths in the United States could be attributed to cardiovascular disease, an increase of 20% since 1940. 16 The causes of cardiovascular disease, unfortunately, were poorly understood. Heart disease experts, however, increasingly stressed the role of arteriosclerosis in the development of CHD 17 and the importance of environmental factors, as opposed to aging, in the etiology of arterial disease. 18 In 1946, epidemiology’s prime focus remained the study of infectious disease. At the US Public Health Service (PHS), Joseph Mountin, director of the Bureau of States Services, had just created for that purpose the Communicable Disease Center. 19, 20 But Mountin, a master of public health policy, also recognized the significance of the epidemiological transition (and the importance of earlier work on chronic disorders conducted by the PHS). 21– 23 After World War II, he championed control programs for chronic, noninfectious diseases, including community-based screening and diagnostic interventions. 19, 24, 25 In pressing for heart disease control efforts, Mountin added an epidemiological investigation. 2 He subsequently selected Gilcin Meadors, a young PHS officer, to initiate that epidemiological research, an investigation that evolved into the Framingham Heart Study. 2 He also approached local public health departments and academic experts to advance and bolster this plan. In launching what became the Framingham Study, the government flagged CHD as a problem of national significance. However, the PHS, by inviting multiple parties to participate in the investigation, left its objectives open to contestation and negotiation. Subsequently, competing institutional interests buffeted the study, shaping its research design and scientific goals. I describe the vicissitudes and stages through which the study passed during its earliest years, as organizers grappled with its relationship to medicine and the local community. The sway of competing interests was the greater because the science supporting the research was so weak. How did one design and analyze an epidemiological study of chronic noninfectious disorders? Previous experience was sparse. 2 The solution was especially difficult because, unlike tobacco and lung cancer research, one had to incorporate into the study multiple hypothetically causal variables. 26– 28
