Methods: We reviewed English-language published papers on physical activity and cancer through Medline searches for epidemiologic studies, and through references on individual reports. We reviewed general texts on effects of exercise on human biology and applied the concepts to the biology of cancer in humans to describe potential mechanisms for a physical activity-cancer association.

Results: Considerable epidemiologic evidence has accrued linking increased physical activity with reduced occurrence of cancers of the breast and colon. The association between physical activity and cancers of other sites is unclear. Potential mechanisms for the association between physical activity and reduced risk for breast and colon cancer are varied: they range from bias due to physical activity's strong correlations with other health factors (e.g., diet, smoking, alcohol use, use of medications) to the metabolic effects resulting from increased physical activity and fitness, such as reduced obesity, hormonal and reproductive effects, mechanical effects, and enhancement of the immune system.

Conclusions: The elucidation of biologic mechanisms for an association between physical activity and cancer may provide biological support for the association. It will contribute information to determine the type, frequency, and duration of exercise needed to maximize protection. This information will be needed before large-scale community interventions are begun, in order to choose the correct interventions for the desired effect of reduced incidence of the most common cancers.