Associations of coffee consumption and caffeine intake with mammographic breast density

Abstract

Purpose: Previous studies suggest that coffee and caffeine intake may be associated with reduced breast cancer risk. We investigated the association of coffee and caffeine intake with mammographic breast density by woman’s menopausal status and, in postmenopausal women, by hormone therapy (HT). Methods: This study included 4130 cancer-free women within the Nurses’ Health Study and Nurses’ Health Study II cohorts. Percent breast density (PD) was measured from digitized film mammograms using a computer-assisted thresholding technique and square root-transformed for the analysis. Average cumulative coffee/caffeine consumption was calculated using data from all food frequency questionnaires preceding the mammogram date. Information regarding breast cancer risk factors was obtained from questionnaires closest to the mammogram date. We used generalized linear regression to quantify associations of regular, decaffeinated, and total coffee, and energy-adjusted caffeine intake with percent density. Results: In multivariable analyses, decaffeinated coffee was positively associated with PD in premenopausal women (2+ cups/day: β = 0.23, p trend = 0.03). In postmenopausal women, decaffeinated and total coffee were inversely associated with PD (decaffeinated 2+ cups/day: β = − 0.24, p trend = 0.04; total 4+ cups/day: β = − 0.16, p trend = 0.02). Interaction of decaffeinated coffee with menopausal status was significant (p-interaction < 0.001). Among current HT users, regular coffee and caffeine were inversely associated with PD (regular coffee 4+ cups/day: β = − 0.29, p trend = 0.01; caffeine 4th vs. 1st quartile: β = − 0.32, p trend = 0.01). Among past users, decaffeinated coffee was inversely associated with PD (2+ cups/day β = − 0.70, p trend = 0.02). Conclusions: Associations of decaffeinated coffee with percent density differ by woman’s menopausal status. Associations of regular coffee and caffeine with percent density may differ by HT status.