There is strong and consistent evidence of an association between obesity and post-menopausal breast cancer. In spite of these consistent epidemiologic associations, the exact mechanisms by which they are associated are not clear. The menopausal transition is most often characterized by decreased circulating estrogen levels and weight gain and/or changes in adipose distribution. However, hormone therapy (HT), including estrogen, does not favorably alter body composition long-term, nor does it reverse the increased risk of postmenopausal breast cancer. Follicle stimulating hormone (FSH) also increases with the menopausal transition and has emerged as a potential novel factor associated with decreased thermogenesis and increased adiposity in animal models. Therefore, we hypothesize that FSH may play an estrogen-independent role in breast cancer risk via its influence on the development of postmenopausal obesity. To date, there have been no large-scale epidemiologic investigations of these associations. Our study will first evaluate the association of FSH and multiple measures of total and abdominal adiposity in a large sample of postmenopausal women from the Women’s Health Initiative (WHI), a long-term national epidemiologic study. The sample will include 1733 postmenopausal women randomized in WHI HT trials, with repeat DXA scans over 6 years follow-up and 685 postmenopausal women in the observational study (OS) with repeat scans over 20 years follow-up. Additionally, in this sample, we will assess the dynamics of FSH levels in relation to adipose derived regulatory hormones of estrogen, activin and leptin, prior to breast cancer diagnosis, given the importance of feedback and feedforward regulation of FSH. Second, we will compare FSH levels in women who develop breast cancer (N=759 with DXA) to those women who remained breast cancer free (N=2333 non-cases in HT and OS sample above). Finally, we will analyze associations of germline polymorphisms in FSHR and FSHB and serum FSH levels and test for a causal association of FSH on adiposity in a Mendelian randomization framework, hypothesizing that genetic instruments for FSH levels are causally related to postmenopausal adiposity. Our study includes a novel measure of visceral adiposity derived from full body dual xray absorptiometry (DXA) as well as BMI, waist circumference and DXA-derived fat mass and percent body fat, as well as stored blood samples and GWAS data in WHI. To our knowledge, our study is the first 

comprehensive epidemiologic investigation of FSH and obesity, including both cross-sectional and longitudinal designs, and the first to measure and study FSH levels in the years prior to breast cancer diagnosis for study of the potential role of FSH in breast cancer development. Further, we will derive a mechanistic understanding of the influence of the menopausal hormonal milieu, hormone therapy, and genetics on body composition and breast cancer risk. These pieces of information in combination will enable novel and more precise future intervention targeting for breast cancer prevention.