Objective To investigate the association between leisure-time physical activity (LPA) and breast cancer risk in consideration of tumour oestrogen/progesterone-receptor (ER/PR) status.

Methods We conducted a population-based prospective cohort study among 53 578 women in the Japan Public Health Center-based Prospective Study. LPA was assessed by self-reported questionnaires. A Cox proportional hazards regression model was used to derive RRs and 95% CIs.

Results From 1990 to 1993 to the end of 2007, 652 cases were identified. The absolute rate of breast cancer was 84 per 100 000 person-years among the sedentary groups (?3 days/month). We observed a statistically significant inverse association between LPA and breast cancer risk (RR ?3 days/week vs ?3 days/month =0.73; 95% CI 0.54 to 1.00; p trend 0.037), particularly in ER+PR+(RR 0.43; 95% CI 0.19 to 1.00; ptrend 0.022) and this inverse trend was apparent among postmenopausal women (RR 0.25; 95% CI 0.06 to 1.06; p trend 0.041). An inverse trend was also observed between daily total physical activity and postmenopausal ER+PR+ risk (p=0.046). Among BMI ?25 kg/m2 group, LPA was associated with decreased risk (RR ?1 days/week vs ?3 days/month =0.65; 95% CI 0.43 to 0.97; p trend 0.033).

Conclusion Active participation in LPA may contribute to a decrease in breast cancer risk, particularly for postmenopausal ER+PR+ tumours.