BACKGROUND:

Epidemiologic investigations of the influence of physical activity on the incidence of specific chronic diseases encounter problems of confounding or interaction with eating behavior, tobacco use, other social habits, and a host of personal characteristics. These influences of life styles on chronic diseases are not mutually exclusive, nor do they operate independently. Rather, they blend, amplifying or diminishing the effects of one another.

DESIGN NARRATIVE:

The study used data collected since 1960 as part of the College Alumni Health Study -- a cohort of 50,000 former students who attended Harvard College, 1916-1950, or the University of Pennsylvania, 1931-1940, and who reported by questionnaire on their health habits and health status in post-college years through 1988. Over 17,000 Harvard alumni have been followed from 1962 through 1988 for both non-fatal and fatal chronic diseases.

These predictor (risk factors) and outcome (specific diseases) data were analyzed to test the hypothesis that different kinds and amounts (frequencies, intensities, durations, and constancies) of physical activity affect differently the incidence rates of hypertension, NIDDM, and CHD. Continuities and changes in specific life styles and chronic diseases have been measured continually through return-mail questionnaires in 1962, 1966, 1972, I977, and 1988. Cause-specific mortality has been monitored continuously over this span of time, 1962-1988.

Using the data, the investigators computed relative and attributable risks of developing hypertension, NIDDM, and CHD that related to alumni patterns of physical activity, body size, alcohol consumption, cigarette use, and other personal characteristics and ways-of-living. In this discrete population, they searched for the relative importance of physical activity and other potential predictors in influencing the occurrence of these hypertensive-metabolic-atherosclerotic diseases.