The Framingham Study
The Framingham Study, the prototypical prospective study at the origins of formal CVD epidemiology, was born in the mind of Joseph Mountin, Assistant Surgeon General and head of the Division of Chronic Diseases of the U.S. Public Health Service. He saw cardiovascular diseases emerging after World War II and replacing infectious diseases as major killers and was in a central position of influence as head of State Services of the USPHS and Deputy Surgeon General. He was also experienced in disease surveillance and mass interventions from directing the Malaria Corps of World War II, and he had envisioned and founded the Centers for Disease Control. Having proposed a two-pronged community-wide program, one observational and one with intervention, he was invited by Harvard’s David Rutstein to establish the epidemiological study in Framingham, MA, an independent community near Boston. Gilcin Meadors, recent Johns Hopkins graduate in public health, was appointed Framingham’s first Medical Director. (NHLBI Archives of the Framingham Study)
Soon after the establishment of the new National Heart Institute in 1948, Felix Moore, its chief of biometrics, was sent to report on the Framingham Study and recommend whether it should be incorporated into the new National Heart Institute (NHI) as part of its research mission. When he strongly recommended this course, NIH director, C.J. Van Slyke charged Moore to provide estimates of the required sample size for a definitive epidemiological study having reasonable likelihood of establishing, in a 20-year period, the relationship of given characteristics to risk of death from heart attack. This resulted in a plan to recruit a two-thirds sample of Framingham men and women, 5000 to 6000 people, ages 30-59. (Moore interview)
The basic scientists who directed NIH tended to regard the new extramural research program as depriving funds for “basic research,” and they particularly looked askance at funding for population studies. This entrenched view at NIH was overcome, in part, by arguments of Van Slyke’s chief cardiological advisor, Boston cardiologist Paul Dudley White, who had become interested, as a student of Sir James MacKenzie in the United Kingdom, in the natural history of heart diseases outside the clinical setting.
Soon after the administrative move of the Framingham Study to the NHI, Roy Dawber, a young Public Health Service internist, was appointed medical director to bring the medical community of Framingham into support for the PHS project. He came with an early set of prior epidemiological hypotheses, arguments, and assumptions to what was to become one of the best-known epidemiological studies of all time. Dawber states: “The task of epidemiology… is to determine to what degree an observed relationship may be the result of chance and at what point the relationship is sufficiently strong that it may well be involved in causality.” (1).
“The characteristics of persons who already have the disease are not necessarily the same as those that predispose to the disease. Observations of population characteristics must be made well before disease becomes overt if the relationship of these characteristics to the development of the disease is to be established with reasonable certainty.”
If the relationship is one that fits what is known about the disease and has a logical explanation, it is worth exploring further, regardless of the strength of the relationship. If, however, the relationship is very powerful, it deserves careful scrutiny even though the alleged relationship may be unexplained at the time.
The Framingham Study was among several of the early cohort study groups that published in 1957 a landmark monograph on the predictive power of blood pressure, blood cholesterol level, and cigarette smoking habit for heart and blood vessel diseases (2). In fact, the term “risk factor” is attributed to the investigators of Framingham, who went on to elaborate many central concepts and practical tools in the identification and prevention of elevated cardiovascular risk.
The Framingham Study, with congruent findings from other studies of healthy cohorts in the U.S. and abroad, sparked a revolution in understanding of the individual and the mass causes as well as the preventability of heart attack and stroke. Together they provided a sound basis for successful medical action and health promotion policy to reduce the death rate from these diseases.
The Framingham Study remains the responsibility of the National Heart, Lung, and Blood Institute and is now carried out under contract by researchers at the Boston University School of Medicine. It has been enlarged twice, in 1971 with the “Offspring Study” of original participants’ children and their spouses and in the late 1990s with the “Omni Study” of minorities. Every other year, after an extensive baseline examination, subjects undergo repeat testing that includes medical history, a blood profile, echocardiogram, and bone, eye, and other specialized tests.
The Framingham Study today still involves some of the early investigators, including William Kannel and Philip Wolf, along with generations of new researchers. It has new emphases on the risk of particular disease manifestations such as heart failure, peripheral artery disease, stroke types, and arrhythmias. It explores new risk characteristics such as the apolipoproteins and their regulating genes, homocysteine, blood clotting factors, and inflammation. And its scope is widened to the study of a whole set of chronic conditions such as obesity, diabetes, and cardiac enlargement, and other diseases including osteoporosis, cancer, and Alzheimer’s disease (3). (Henry Blackburn)
Among an extensive literature of some 1,000 articles are the following classic references to The Framingham Study.
1. Dawber, TR. The Framingham Study. The Epidemiology of Atherosclerotic Disease. Harvard University Press. Cambridge, MA. 1980. 258 pp.
2. Dawber, TR, FE Moore, Jr., and GV Mann. Coronary heart disease in the Framingham Study. Amer. J. Public Health. 1957:47:4-24.
3. Kannel, WB. Clinical misconceptions dispelled by epidemiological research. The Ancel Keys Lecture. Circulation. 1995;92:3350-3360.