Harvard Alumni Study
The Harvard Alumni Study was one of the first epidemiologic studies to examine the relationship between physical activity and incidence of heart attack. The study commenced in 1962 with questionnaires sent to a large cohort of college-educated men.
Male surviving alumni who enrolled in Harvard University between 1916 and 1950 were mailed brief questionnaires in 1962 and 1966, and 16,936 men responded (71% response rate). Physical activity records were examined for the periods when each subject was a student and in middle age, using college archives on athleticism and personal background for data on student years. Returned surveys provided data on activity level, tobacco use, parental disease history and medical care for physician-diagnosed diseases.  Follow-up surveys of respondents, mailed in 1972, 1988 and 1993, collected data on chronic diseases and death. 
To quantify the physical activity level of the respondents, an index was created that assigned values to each activity based on energy expenditure. Physical activity responses on the questionnaires included “light sports (e.g., bowling, baseball, biking, boating, dancing, golf, and light housework),” “strenuous sports (e.g., basketball, running, mountaineering, skiing, swimming and tennis),” total stairs climbed, and blocks walked per day. Based on the responses from the survey, each participant was assigned an estimated value of kilocalories burned per week.
At the first follow-up interval in 1972, 695 deaths attributed to all causes were recorded. In the 6-10 years of follow-up, 357 men had nonfatal heart attacks and 215 men died of heart attack.  The relative risk for first heart attack for those who climbed less than 50 stairs a day compared to those who climbed fifty or more was 1.25 (p= .008). For men who walked less than five city blocks daily, their relative risk for first heart attack was 1.26 when compared to men who walked more than five blocks a day (p= .016). Men who burned less than 2000 kilocalories per week had a relative risk of 1.64 (p< .001) compared to men who burned more than 2000 kilocalories weekly. Comparisons of light sports play yielded a relative risk of 1.08 with an insignificant p-value (.501). 
In the 1993 survey, 2,135 subjects had experienced coronary heart disease (87.1% response rate). In this analysis, the researchers found increased physical activity level was associated with lower coronary heart disease risk when considered singly (RR for those who burned more than 8400 kilojoules per week= 0.73, 95% CI= 0.63-0.84, p <.001).  When considered along with age, BMI, alcohol intake, hypertension, diabetes, smoking status and early parental death, higher levels of physical activity were also apparently protective (RR for those who burn more than 4200 kilojoules per week= 0.81, 95% CI: 0.71-0.92, p= .003) .
The Harvard Alumni Study was pioneering in the field and used a simple questionnaire at low cost. Additional strengths of the study include the lengthy follow-up period, high response rate, and completeness of data. Limitations include the potential bias of self-administered questionnaires. (HB)
 Paffenbarger RS, Wing AL, Hyde RT (1978). Physical activity as an index of heart attack risk in college alumni. American Journal of Epidemiology, 108 (3), 161-175.
 Sesso HD, Paffenbarger RS, Lee I-M (2000). Physical activity and coronary heart disease in men: the Harvard alumni study. Circulation, 102, 975-980.