Abstract
OBJECTIVE: Long-term weight regain limits the population-level benefits of obesity interventions. We tested whether the polygenic risk score of BMI (PRS(BMI)) modifies weight trajectories following modest weight loss. METHODS: The analytic sample included 9897 postmenopausal women from the Women's Health Initiative Dietary Modification Trial (6132 European American; 3749 African American). PRS(BMI) was derived from a trans-ancestry GWAS of ~2 million participants. Longitudinal weight change (7 years) was modeled using weighted GEE. RESULTS: In European Americans, the PRS(BMI) × randomization × time interactions approached significance at the 95th percentile (p = 0.052) and 85th percentile (p = 0.07). No interaction was observed in African Americans. In analyses restricted to European Americans who lost ≥ 5% of initial weight by year 1 (20%; n = 1273), women in the ≥ 95th percentile of PRS(BMI) regained nearly twice as much per year as those with average risk (0.94 vs. 0.48 kg/year, p = 0.0016). CONCLUSIONS: A high PRS(BMI) was associated with faster weight regain following modest weight loss in European American women. While further validation is required in a diverse population, these results suggest the potential for genetics to inform targeted strategies for sustaining long-term weight management. TRIAL REGISTRATION: ClinicalTrials.gov identifier: 75N92021D00001, 75N92021D00002, 75N92021D00003, 75N92021D00004, and 75N92021D00005.