Moderate-vigorous recreational physical activity and breast cancer risk, stratified by menopause status: a systematic review and meta-analysis

Objective: Physical inactivity increases postmenopausal and possibly premenopausal breast cancer risk, although different biologic mechanisms are proposed. Our primary objective was to estimate breast cancer risk associated with high versus low levels of moderate-vigorous recreational activity, separately for premenopausal and postmenopausal women. Methods: We conducted a systematic review of literature published to July 2015. Included reports were cohort or case-control studies relating moderate-vigorous recreational physical activity (metabolic equivalent >= 3.0) to breast cancer incidence, exclusively (>= 90%) in premenopausal or postmenopausal women. We appraised study quality and performed meta-analyses using random effects modeling. Subgroup meta-analyses were based on tumor subtype, race, body mass index, parity, hormone therapy use, family history of cancer, and statistical adjustment for body fatness. Dose-response relations were examined. Results: Pooled relative risks (RRs, 95% CI) for women with higher versus lower levels of moderate-vigorous recreational activity were RR = 0.80 (0.74-0.87) and RR = 0.79 (0.74-0.84) for premenopausal (43 studies) and postmenopausal (58 studies) breast cancer, respectively, with high heterogeneity. Inverse associations were weaker among postmenopausal cohort studies (RR = 0.90 [0.85-0.95]) and studies that statistically adjusted for nonrecreational (eg, occupational, household) activity (RR = 0.91 [0.77-1.06] premenopausal, RR = 0.96 [0.86-1.08] postmenopausal). Risk estimates with versus without body fatness adjustment did not vary by menopause status, although other subgroup effects were menopause-dependent. Among studies of overweight/obese women, there was an inverse association with postmenopausal but not premenopausal breast cancer (RR = 0.88 [0.82-0.95] and RR = 0.99 [0.98-1.00], respectively). Dose-response curves were generally nonlinear. Conclusions: Although risk estimates may be similar for premenopausal and postmenopausal breast cancer, subgroup effects may be menopause-dependent.