Purpose: To investigate the shape of the relation between amount of leisure-time physical activity (LTPA) and myocardial infarction (MI) risk.
Methods: Data were from a case-control study in a Washington State health maintenance organization, 1986 to 2002. Participants had no prior cardiovascular disease and good self-reported health before selection. Telephone interviews asked 697 nonfatal MI cases and 3,397 control subjects about 26 types of LTPA. Models adjusted for age, sex, year, treated hypertension, family history of heart disease, smoking, alcohol, aspirin, race, retirement, income, and education.
Results: Some LTPA was reported by 90% of control subjects and 84% of cases. Compared with no LTPA, participation in LTPA was associated with lower risk of MI (adjusted odds ratio [OR] = 0.67, 95% confidence interval [CI]: 0.52, 0.86). Among active participants, LTPA time was associated with risk of MI (OR = 0.66 for high versus low quartile, 95% CI: 0.51, 0.86). Risk of MI decreased with increasing total or nonstrenuous LTPA time up to the median, beyond which we could not detect an association between LTPA time and MI risk.
Conclusions: Time engaged in LTPA, even nonstrenuous LTPA, was associated with lower risk of MI, and the shape of this relationship was nonlinear.