Thomas R. Dawber

AN APPROACH TO LONGITUDINAL STUDIES IN A COMMUNITY: THE FRAMINGHAM STUDY

Those concerned with community health have a legitimate interest in all matters related to promoting the health of the community. These interests may vary depending on the purposes for which the health agency was organized, its location, the skills and resources available to it, and many other factors. In general, the major objective of community health activities is to prevent, or improve the medical care of, existing disease. In the past, research engaged in by health agencies has usually been initiated because of some real or suspected problem affecting the health of their community. Because of inability to answer certain crucial questions pertaining to the problem, an investigation is undertaken to attempt to provide the desired information. Studies might be undertaken to assist in administrative planning concerning the number of beds required for some major emergency or to meet changing patterns of disease incidence, or of ways to convert existing facilities for such purposes, e.g., conversion of tuberculosis sanatoria to chronic disease hospitals. Research may be required to discover ways of improving medical care for particular diseases by determining methods required to bring persons with given diseases under medical care early in the course of the illness, in order to minimize disability. This type of research has usually been done with specific uses in mind and can be considered under the category of applied research. In addition to research aimed directly at solving immediate local problems, it is now increasingly common for health agencies to ask how the facilities and resources available to them can best be used for purposes of research to increase general knowledge without necessarily being motivated by the immediate needs of the particular health agency. Studies may be carried out to evaluate the manner in which a particular problem has already been solved or to indicate the desirable approach to a particular health problem. Community health agencies are especially fitted for the conduct of epidemiologic studies which require access to large populations and resources for handling large numbers of subjects. For this reason, a large proportion of community health research projects are of this type. Because of the classic contributions of the epidemiologic method in determining the causes and methods of prevention of infectious and nutritional diseases, this method has been applied in an attempt to unravel the problems surrounding chronic diseases of unknown etiology. Although notable progress is being made, it remains to be demonstrated whether this approach will prove as valuable in solving the problems surrounding chronic disease.

The epidemiology of gallbladder disease: Observations in the Framingham study

Abstract The longitudinal observations for ten years on 5209 men and women aged 30–62 in Framingham, Massachusetts, have been used for a prospective investigation of the epidemiology of gallbladder disease. ‘Definite’ gallbladder disease was diagnosed in 427 persons on the evidence of pathology reports, surgery, X-ray and hospital admissions and 116 were ‘doubtful’, having only the diagnosis suggested by a physician without further verification. The overall incidence of gallbladder disease was about twice as high in women as in men, and it increased with age in both sexes without any evidence of an excess in the forties. Gallbladder disease was not related to ‘fairness’ as measured by national origin. Increase in weight and number of pregnancies were each associated with increased incidence. Despite the presence of cholesterol in many gallstones and the elevation of serum cholesterol often noted with obesity and during pregnancy, no relationship was demonstrated between serum cholesterol level and gallbladder disease. Hemoglobin level was also not related in the population studied. Women with lower systolic blood pressures showed a somewhat reduced risk of gallbladder disease, but the possibility that a mutual relationship to weight could explain this has not been excluded. Slight negative correlations were noted between the disease incidence and height, and alcohol and cigarette consumption. These were small and in general not statistically significant. Dietary fat, protein and cholesterol intake, level of physical activity, age at menopause, marital status and precinct of residence were not related to gallbladder disease. Nor was an excess incidence noted in persons whose spouses had gallbladder disease. Persons entering the study with gallbladder disease showed no excess mortality. It is not yet clear whether such persons develop coronary heart disease at an excess rate. Potent environmental factors influencing the rate of development of gallbladder disease have yet to be identified. On the basis of present evidence, it would appear that more attention should be devoted to the anatomy and pathological physiology of the biliary tract in seeking a better understanding of the pathogenesis of gallbladder disease.

THE CORONARY PROFILE.

Mounting evidence indicates that in recent decades there h a s been an absolute as well a s a relative increase in the occurrence of coronary heart d i sease in this and in other highly developed countries. Moreover, there h a s been an increasing tendency for the d i sease to occur in younger age groups, a t l ea s t among men.’ With the recent advances in the control of infectious diseases, the cardiovascular-renal d i seases in general and coronary heart d i sease in particular have become th i s country’s most serious health problem. International studies, such as those contrasting the coronary heart d i sease experience of the Japanese in Japan, Hawaii, and in the continental United States,l of rural and urban Guatemalans,3 and the newly arrived and long-resident Yemenite J ews in Israe1,4 have all pointed to the importance of environmental factors in th i s disease. The indication that the “way of life’’ of individuals affects their susceptibility to coronary heart d i sease is a more hopeful concept than one that accepts liability to coronary heart d i sease as an inevitable concomitant of aging. We cannot s tay the course of t i m e , but an unhealthful environment is subject to manipulation, an unhealthful way of life is subject to change. Mortality statistics indicate the outside dimensions of the problem and reports on hospital and autopsy series provide important data on selected portions of the population, but community surveys or epidemiological studies are necessary to round out the picture, to show the pattern of the disease, and to indicate where we may mast profitably attack it. Such an epidemiological study h a s been going on in the town of Framingham, Massachusetts s ince 1949.‘ Two thousand two hundred and eighty-two (2,282) men and 2,846 women initially free of coronary heart d i sease ( T A B L E 1) have been followed primarily by means of biennial examinations, each of which consisted of a history, a physical examination, and a battery of laboratory determinations6 L e s s than one half of one per cent of those initially examined have been completely lost to

The use of computers in cardiovascular epidemiology.

Summary o 1. The usefulness of computers in epidemiologic research in the longitudinal study of cardiovascular disease based on the experience of the Framingham Study is discussed. 2. The large size of the populations involved in this type of research, and the numerous and repeated observations required, make machine processing of data desirable and often mandatory. 3. The collection and recording of the observations for computer processing are discussed in detail. The importance of this part of the methodology is stressed since this aspect is the major one with which the epidemiologist is concerned. 4. The application of computers to analysis of data is discussed. Several examples of their value in the Framingham Study have been presented. 5. Future application of computers to cardiovascular epidemiology is considered.