Objective: To examine recreational physical activity (PA) and prostate cancer risk in a large cohort of men living in Washington State, focusing on frequency and type of physical activity at various times throughout life.
Methods: In a prospective cohort study, we assessed physical activity in 34,757 men (50-76 years at baseline) using a questionnaire. Men were recruited into the study between 2000 and 2002. Five hundred and eighty-three men developed prostate cancer.
Results: Using Cox proportional hazards regression, PA either in the 10 years before baseline or earlier in life was not associated with prostate cancer risk. However, compared to no activity, >or=10.5 MET-h per week (the median level) of PA was associated with a reduced prostate cancer risk among men who were normal weight (HR = 0.69, 95% CI 0.46-1.0), >or=65 years at diagnosis (HR = 0.75, 95% CI 0.55-1.0) and who had not had a recent PSA (HR = 0.47, 95% CI 0.28-0.81). Greater PA was associated with an increased risk among men who were obese (HR = 1.5, 95% CI 0.95-2.4), and no association among men <65 years or with a recent history of PSA screening (all p for interactions <or=0.02).
Conclusions: PA was not associated with prostate cancer risk, except in subgroups defined by age, obesity, and screening history.