Abstract
CONTEXT: Evidence supports a protective effect of menopausal hormone therapy (HT) on bone. However, whether genetic susceptibility modifies the association of HT and fracture risk is not sufficiently explored. OBJECTIVE: The objective was to test an interaction between genetic susceptibility and HT on fracture risk. DESIGN: We constructed two weighted genetic risk scores (GRSs) based on 16 fracture-associated variants (Fx-GRSs) and 50 bone mineral density variants (BMD-GRSs). We used Cox regression to estimate the main effects of GRSs and their interactions with HT on fracture risk. We estimated the relative excess risk due to interaction (RERI) as a measure of additive interaction. We also used the case-only approach to test for a multiplicative interaction. SETTING: Forty US clinical centers. PARTICIPANTS: A total of 9922 genotyped white postmenopausal women (age, 50 to 79) from the Women's Health Initiative HT randomized trials. MAIN OUTCOME MEASURES: Adjudicated fracture incidence. RESULTS: Both GRSs were associated with fracture risk per 1-unit increment in GRS (hazard ratio, 1.04 [95% confidence interval, 1.02 to 1.06] for Fx-GRS and hazard ratio, 1.03 [95% confidence interval,1.02-1.04] for BMD-GRS). We found no evidence for multiplicative interaction for either of the GRS. However, we observed a substantial additive interaction, where the highest quartile of both GRSs and randomization to placebo have excess fracture risk: Fx-GRS P for RERI = 0.047, BMD-GRS P for RERI = 0.046. CONCLUSIONS: These results suggest that HT reduces fracture risk in postmenopausal women, especially in those at highest genetic risk of fracture and low BMD.