Abstract 4142226: Quantifying the influence of baseline body weight on sex differences in semaglutide-associated weight loss

Background: Males tend to lose less weight than females with semaglutide, a GLP-1 receptor agonist, for unknown reasons. One hypothesis is that lower average body weight in females may result in greater drug exposure, leading to more weight loss. Objective: Leveraging a multi-state network of clinico-genomic cohorts, we aimed to quantify the extent to which sex differences in baseline body weight could explain semaglutide-associated weight loss differences between males and females. Methods: We utilized electronic health record data and clinical-grade sequencing from multiple U.S. cohorts (n>100k). Among individuals starting semaglutide with BMI ≥27 kg/m 2 (consistent with guidelines), we analyzed weight measurements over a 12-month use period. A BMI polygenic score (PGS) using 27k variants (PGS001228) was calculated. Percent weight change was modeled using mixed effects models with quadratic and cubic terms for time and interaction terms with baseline weight, sex, and confounders (semaglutide dose, age, ancestry, BMI PGS, comorbidities, medications). We also conducted a mediation analysis to test whether baseline weight mediated the relationship between sex and weight loss. Results: Among 959 females and 352 males, 7,381 weight measurements occurred over a median of 8.9 months (IQR: 6.2-10.8). Median baseline weight was 102 kg (IQR: 90-119) for females and 114 kg (IQR: 100-133) for males. Median weight loss at 6- and 12-months was 5.2% (IQR: 1.9-9.1) and 6.5% (IQR: 2.0-11.8) for females and 3.3% (IQR: 1.0-6.5) and 4.5% (IQR: 1.3-7.7) for males. After adjustment, male sex was associated with a lower rate of weight loss (β=+0.95% [per 6 months], p=0.003). Other factors associated with lower weight loss were BMI PGS (β=+1.28% [top vs bottom quintile], p=0.002), type 2 diabetes (β=+0.88%, p=0.006), hypertension (β=+0.72%, p=0.025), obstructive sleep apnea (β=+0.65%, p=0.032), and insulin use (β=+0.76%, p=0.035). In the mediation analysis, the proportion of the sex-weight loss association that could be explained by baseline weight was 8.4% (95% CI: 2.9-20.6) and 16.0% (95% CI: 10.7-26.2) at 6- and 12-months. Conclusions: Baseline body weight explained only a small fraction of the observed sex difference in semaglutide-associated weight loss, suggesting that other factors (e.g., hormonal) may play a more significant role. Findings also highlight the impact of traditional metabolic risk factors and BMI-associated genetic variants on weight loss with semaglutide.