Effects of physical activity on endogenous sex hormones in postmenopausal women

Background: It is hypothesized that physical activity lowers postmenopausal breast cancer risk at least in part via hormonal pathways. Thus, we investigated whether recent physical activity was associated with lower circulating levels of endogenous sex hormones after menopause. Methods: We conducted a cross-sectional study in 1260 postmenopausal women recruited as controls in the population-based breast cancer case-control study MARIE. Serum concentrations of estrone, estradiol, androstenedione, testosterone, and sex-hormone-binding-globulin (SHBG) were measured, and their associations with different physical activity variables were investigated (recent leisure-time physical activity, sports, bicycling, and walking). Regression models were adjusted for potential confounders. As postmenopausal production of sex hormones takes place primarily in adipose tissue, we especially focused on confounding and effect modification by anthropometric factors (BMI, body weight, waist and hip circumference, and waist-to-hip ratio). Results: Sports was inversely associated with levels of estrone and total and free testosterone in adjusted models. After further adjustment for BMI (or any other anthropometric factor), the associations were attenuated but still significant (p-trend for estrone: 0.034, testosterone: 0.015, free testosterone: 0.035). None of the physical activity variables were significantly related to total and free estradiol, androstenedione, or SHBG. Additionally, we did not observe effect modifications by any anthropometric factor. Conclusion: Our data suggest that sports may lead to lower circulating levels of estrone and total and free testosterone in postmenopausal women. This effect might be partly due to favourable effects on body weight and body composition; however, it appears to be independent of anthropometric factors to some amount. This independent effect could be one explanation for the frequently observed reductions in postmenopausal breast cancer risk through physical activity, even after accounting for BMI. However, further studies (especially randomised controlled trials) are needed to confirm our results.