Hormone-associated dietary patterns and premenopausal breast cancer risk.

Abstract

Purpose: Circulating levels of sex steroid hormones have previously been associated with premenopausal breast cancer risk. Few studies have considered the association between dietary patterns and premenopausal hormone levels. Our objective was to derive dietary patterns associated with premenopausal hormone levels and investigate the association between pattern scores and premenopausal breast cancer risk.

Methods: Using reduced rank regression among a subset of participants from the Nurses' Health Study II (NHSII) (n = 8,962), we identified dietary patterns correlated with premenopausal levels of five sex steroid hormones measured in the follicular and luteal phases. Then, in the full NHSII cohort (n = 90,341), we used Cox proportional hazards models to calculate hazard ratios (HRs) for breast cancer risk associated with each dietary pattern score.

Results: Dietary patterns were identified for luteal estradiol, luteal free estradiol, follicular estrone, luteal estrone, and free testosterone. However, these patterns explained a low percent variation in individual hormone levels, ranging from 2.5-4.1%. During 24 years of follow-up, 1,956 premenopausal breast cancer cases were ascertained. Dietary patterns associated with luteal free estradiol (HR for fifth versus first quintile = 1.29; 95% CI = 1.11-1.49; P_trend < 0.01) and follicular estrone (HR for fifth versus first quintile = 1.28; 95% CI = 1.10-1.49; P_trend < 0.01) were positively associated with premenopausal breast cancer risk.

Conclusion: Our findings indicate that while some dietary factors may marginally influence premenopausal hormone levels, the relation between sex steroid hormones and premenopausal breast cancer risk is likely not driven by diet. Future studies should consider other mechanisms through which diet may impact breast cancer risk, including inflammatory processes.