Glucagon‐like peptide‐1 receptor agonists in schizophrenia spectrum disorders: A systematic review and meta‐analysis of efficacy, safety and mental health outcomes

Aims The purpose of this study is to evaluate efficacy, safety and mental‐health outcomes of glucagon‐like peptide‐1 (GLP‐1) receptor agonists in adults with overweight or obesity and antipsychotic‐treated schizophrenia‐spectrum disorders. Methods PubMed/MEDLINE, Embase, PsycINFO, CENTRAL, Virtual Health Library (VHL) (Biblioteca Virtual em Saúde, BVS), Scopus, Web of Science and trial registries were searched on 28 August 2026. Random‐effects models pooled mean difference (MD) or standardized mean difference (SMD) with 95% Wald intervals; Hartung–Knapp–Sidik–Jonkman (HKSJ) intervals were computed in parallel. Missing change‐score SDs were imputed at r = 0.50 (sensitivity r = 0.25 and 0.75). Binary safety outcomes used Mantel–Haenszel risk ratios with a 0.5 continuity correction for zero cells. Results Six trials ( n = 446) were pooled; three ( n = 102) informed the primary outcome. The primary outcome of total symptom severity was imprecise (SMD − 0.21, 95% CI −0.60 to 0.19; HKSJ −1.02 to 0.61). Functioning, Alcohol Use Disorders Identification Test (AUDIT), Schizophrenia Quality of Life Scale (SQLS) and Brief Assessment of Cognition in Schizophrenia (BACS) were likewise imprecise; all intervals included the null. Secondary mental‐health analyses were exploratory and unadjusted for multiplicity. Glucagon‐like peptide‐1 receptor agonists (GLP‐1 RAs) reduced body weight (MD − 7.42 kg, −11.52 to −3.33; I 2 = 94%), body mass index (BMI, −2.77 kg m −2 ), waist circumference (−5.53 cm) and HbA1c (−0.36%). Gastrointestinal events were more frequent; serious adverse events were not. Weight heterogeneity tracked agent, dose, duration and antipsychotic background. Conclusions GLP‐1 RAs were associated with improvements in cardiometabolic measures over 12–36 weeks, although certainty for body weight was limited by heterogeneity and risk of bias. Mental‐health estimates were imprecise and did not allow a clear conclusion regarding benefit or harm. Cardiovascular events were not assessed, and no conclusions can be drawn regarding cardiovascular outcomes in this population.