Trajectory of weight regain after cessation of glucagon-like-peptide-1 receptor agonists: a systematic review and nonlinear meta-regression

Glucagon-like peptide 1 receptor agonists (GLP-1RAs) have emerged as breakthrough weight loss agents (1,2) . However, discontinuation is common (3) , and clinical trials have demonstrated significant weight regain following cessation (4,5) . In this systematic review, we aimed to characterise the trajectory of weight regain after GLP-1RA cessation. This study followed Cochrane and PRISMA guidelines. We searched MEDLINE, Embase, Cochrane Library, Scopus and Web of Science from inception to 26/12/2024 for studies reporting weight outcomes after GLP-1RA cessation in overweight/obese adults. Eligible studies were randomised controlled trials (RCTs) or observational studies; participants had treatment for ≥8 weeks and a post-cessation follow-up ≥4 weeks, and studies could not include co-treatment with other FDA-approved weight loss drugs. Weight regain was the primary outcome. Randomised studies with n≥100, which reported on-treatment weight loss ≥3kg with ≥2 post-cessation datapoints or 1 datapoint after 12 weeks, were included in a meta-regression analysis in which percentage weight-regain was modelled as a function of time. Exponential recovery model fit was assessed with root mean square error (RMSE) and Akaike information criterion (AIC). Inverse-variance weighting was applied using the standard error of the mean of each observation. Secondary outcomes included HbA1c and systolic blood pressure (SBP). Risk of bias was assessed using the Cochrane RoB 2 tool for RCTs and the Cochrane ROBINS-I tool for non-randomised studies. The study protocol is registered with PROSPERO (CRD420250631751). We identified 44 relevant studies; most were assessed to have moderate risk of bias. Treatment duration spanned 10-104 weeks and post-cessation follow-up spanned 4-104 weeks. Weight consistently rebounded after cessation of GLP-1RAs. 6 trials with 3236 participants were included in the exponential recovery model to characterise weight regain. Weight regain was estimated to plateau at 75.6% (95% CI 68.5-82.7) of weight lost on treatment. The rate constant was 0.0302 per week (95% CI 0.0204-0.0399), corresponding to a 23.0 week half-life. At 1 year after cessation, an estimated 40.2% of the on-treatment weight loss remained. Post-cessation HbA1c was reported by 21 studies. Virtually all observed post-cessation increases in HbA1c. In most cases, around half the initial reduction appeared to be regained by 8 to 12 weeks after cessation. The extent and timing of this varied significantly between studies. Post-cessation SBP was reported by 18 studies. After cessation of treatment, SBP typically rebounded towards the pre-treatment baseline, with most studies reporting that around 70 to 80% of the initial reduction was regained within 12 weeks. GLP-1RA cessation is associated with a predictable and decelerating pattern of weight regain, which appears to plateau below pre-treatment levels, suggesting that partial weight-loss benefit may persist long-term but is substantially attenuated.