Abstract TU207: Independent and Joint Association of GLP-1 Therapy and Physical Activity with Weight Loss: Real-World Evidence from All of Us

Abstract

Background: Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are increasingly prescribed for weight management, while physical activity remains a central component of lifestyle management. Evidence on their comparative and combined effects on weight in real-world populations is limited. Methods: We analyzed data from All of Us Research Program (v8), containing electronic health record and Fitbit step counts. We included 316 adults (age ≥18) with obesity (BMI ≥ 30 kg/m 2 ) without type 2 diabetes (T2D), weight available at baseline and 1-year follow-up, and ≥3 months of persistent GLP-1 use. Comparative analyses used a covariate-adjusted regression model (2,279 non–GLP-1 users) and a propensity score matching (293 samples per group). The primary outcome was percent change in body weight, with GLP-1 use (yes/no) and physical activity (average daily steps >5,000) as main predictors. We further examined four prespecified contrasts to evaluate the main effects of GLP-1 use and step counts, as well as their interaction. Additional analyses included step counts as a continuous variable, subgroup analyses by obesity class, and assessment of Fitbit-related selection bias by comparing the cohort to those without Fitbit data. Results: Mean age was 53.3 years (SD 12.6); 73% female, 70% white, and 88.4% non-Hispanic. Mean one-year weight loss was 5% (SD = 8.7%); 47% achieved ≥ 5% loss and 25% achieved ≥ 10%. The greatest reduction was observed among individuals with higher step counts in addition to GLP-1 therapy (mean: 6.1%, 95% CI: 4.3-8.0) after propensity score matching on age, sex, race and baseline BMI. GLP-1 use and step counts were independently associated with weight loss, with no significant interaction effect. Higher step counts, higher baseline BMI and White race (vs. Black) were significant predictors of weight loss. In contrast to clinical trials, real-world data show higher BMI is linked to less weight loss, possibly due to lower activity or unmeasured lifestyle factors. No major selection bias was detected from Fitbit data availability. Conclusions: Both GLP-1 therapy and physical activity were independently associated with clinically meaningful weight reduction at one year. Physical activity may add further benefit to GLP-1 use. This would support the complementary roles of lifestyle and pharmacological strategies in obesity management.