Glucagon-like peptide-1 receptor agonists for antipsychotic-associated cardio-metabolic risk factors: A systematic review and individual participant data meta-analysis

Dan Siskind, Margaret Hahn, Christoph U. Correll, Anders Fink-Jensen, Anthony W. Russell, Nikolaj Bak, Brian V. Broberg, Julie Larsen, Pelle L. Ishøy, Tina Vilsbøll, Filip K. Knop, Steve Kisely, Bjørn H. Ebdrup

Research output: Contribution to journalArticleResearchpeer-review

69 Citations (Scopus)


Aims: To evaluate if glucagon-like peptide-1 receptor agonists (GLP-1RAs) reduce antipsychotic-associated body weight gain in patients with schizophrenia, when compared to controls. Materials and methods: We systematically searched PubMed/EMBASE/PsycINFO/Cochrane using the search terms ‘(antipsychotic and GLP-1RA)’. Individual participant data from studies randomizing patients to GLP-1RA or control were meta-analysed. The primary outcome was difference in body weight between GLP-1RA and control; secondary outcomes included cardio-metabolic variables and adverse drug reactions (ADRs). Multiple linear regression was conducted including sex, age, psychosis severity, metabolic variable, ADRs, and GLP-1RA agent. Results: Three studies (exenatide once-weekly = 2; liraglutide once-daily = 1) provided participant-level data (n = 164, age = 40.0 ± 11.1 years, body weight = 105.8 ± 20.8 kg). After 16.2 ± 4.0 weeks of treatment, body weight loss was 3.71 kg (95% CI = 2.44-4.99 kg) greater for GLP-1RA versus control (p < 0.001), number-needed-to-treat ≥5% body weight loss = 3.8 (95% CI = 2.6-7.2). Waist circumference, body mass index, HbA1c, fasting glucose and visceral adiposity were each significantly lower with GLP-1RA. Sex, age, psychosis severity, nausea, any ADR, and GLP-1RA agent did not significantly impact outcomes. Body weight loss with GLP-1RAs was greater for clozapine/olanzapine-treated patients (n = 141) than other antipsychotics (n = 27) (4.70 kg, 95% CI = 3.13-6.27 vs. 1.5 kg, 95% CI = −1.47-4.47) (p < 0.001). Nausea was more common with GLP-1RAs than control (53.6% vs. 27.5%, p = 0.002, number-needed-to-harm = 3.8). Conclusion: GLP-1RAs are effective and tolerable for antipsychotic-associated body weight gain, particularly clozapine/olanzapine-treated patients. With few included patients, further studies are required before making routine use recommendations for GLP-1RAs.

Original languageEnglish
Pages (from-to)293-302
Number of pages10
JournalDiabetes, Obesity and Metabolism
Issue number2
Publication statusPublished - Feb 2019
Externally publishedYes


  • antipsychotics
  • body weight loss
  • cardiovascular risk
  • GLP-1RAs
  • obesity
  • schizophrenia

Cite this