BACKGROUND:

Current government physical fitness guidelines state that: 1) the majority of the health benefits from physical activity can be obtained by walking two miles briskly on most days of the week; and 2) the health benefits of physical activity depend principally on the total amount of activity rather than the intensity of the activity. Nevertheless, there are currently no prospective epidemiological studies extant, designed specifically to directly contrast the health benefits and costs of moderate exercise (e.g., walking) versus vigorous exercise (e.g., running).

DESIGN NARRATIVE:

Before the start of the study, 233,000 person-years of follow-up had been accumulated in 56,000 runners (between 1991 and 1997). The runners were resurveyed in 1997 along with 68,000 walkers. The walkers were also solicited through the publication of the questionnaire in Walking magazine followed by a direct mailing of the questionnaire to 425,000 subscribers. Total and cause-specific mortality will be determined from the National Death Index; fatal and nonfatal cancers will be identified from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) and 46 state registries; nonfatal coronary heart disease and injuries will be determined from questionnaires. Survival analyses will be used to test whether runners have greater reduction in heart disease, total mortality, and cancer per unit of exercise. Exercise-related injuries from walking and running will also be examined. Power calculations suggest that detection of differences between runners and walkers, as small as 11% for total mortality, 16% for CHD, 12% for total cancers, and 36% for breast cancer, will be possible. The differences will be adjusted for weekly kilocalories expended by walking and running, for walking and running distance, and for time spent on each activity to test whether these variables account for differences in disease rates between walkers and runners.

By the end of the study, 517,000 person years in 68,000 runners (between 1991 and 2001) will be available for analysis. Survival analysis will be used to test for a dose-response relationship between running mileage and CHD and cancer risk, and whether this relationship is affected by running intensity, running frequency, running history, gender, adiposity, age or medication use. Using conservative rates (25% below published values), statistical power calculations suggest that detectable reduction in coronary heart disease risk as small as 0.71% per mile will be possible, which is far below the estimated reduction from other published studies (2.1%). Additionally, a detectable reduction in breast cancer risk as small as 1.5% per mile run in women is calculated, which is below the 1.7% reduction in risk estimated from other published data.