Endoscopic Sleeve Gastroplasty Versus Oral Semaglutide for Obesity: A Real-World Comparative Cohort Study

Abstract

Aims Endoscopic sleeve gastroplasty(ESG) and semaglutide are established therapies for obesity. However, comparative real-world data remain limited. We compared the effectiveness and safety of ESG versus oral semaglutide 14mg in adults with obesity. Methods We conducted a retrospective comparative study between January 2024 and April 2025. Adults aged18–65 years with BMI≥30kg/m2 or ≥27kg/m2 with obesity-related comorbidity treated with either ESG or oral semaglutide 14mg daily and having 6-month follow-up were included. Primary endpoint was percentage total body weight loss(%TBWL) at 6 months. Secondary outcomes included responder rates, adverse events and 12-month weight outcomes. Analyses included Welch’s two sample t-tests, ANCOVA adjusted for age, sex, baseline BMI and diabetes, inverse probability of treatment weighting(IPTW),and1:1 propensity score matching. Results A total of 150 patients were included(ESG=50;semaglutide=100). Baseline characteristics were broadly similar between groups. At 6-months, mean%TBWL was significantly higher with ESG than semaglutide(12.72±5.67%vs8.67±3.84%;p=0.0001). This difference remained after covariate adjustment(adjusted mean difference=-4.04%;p=0.0001),IPTW analysis(p<0.001) and propensity-matched analysis(p=0.021). Responder rates ≥10%TBWL were achieved in 70% with ESG versus 43% with semaglutide(RR0.62;95%CI;0.46–0.82;p=0.0009), and ≥15%TBWL in 36%versus7%(RR0.20;95%CI;0.09–0.44;p=0.0006). There were no major adverse events. At 12-months, mean%TBWL was 11.92±6.93 with ESG and 10.91±4.66 with semaglutide(p0.41). During follow-up, 18patients discontinued semaglutide and 5 patients in ESG group initiated semaglutide. Conclusions In real-world retrospective cohort, ESG was associated with greater short-term weight loss and higher responder rates than oral semaglutide. At 12-months, weight loss was sustained in both groups without a significant between-group difference. These findings are preliminary and should be interpreted with caution in absence of randomized trial design.