IMPORTANCE: Preliminary studies suggest that glucagon-like peptide-1 receptor (GLP-1) agonists, used to treat type 2 diabetes and obesity, may decrease alcohol consumption.
OBJECTIVE: To test whether the risk of hospitalization due to alcohol use disorder (AUD) is decreased during the use of GLP-1 agonists compared with periods of nonuse for the same individual.
DESIGN, SETTING, AND PARTICIPANTS: This cohort study was an observational study conducted nationwide in Sweden using data from January 2006 to December 2023. The population-based cohort was identified from registers of inpatient care, specialized outpatient care, sickness absence, and disability pension. Participants were all residents aged 16 to 64 years who had a diagnosis of AUD.
EXPOSURES: The primary exposure was use of individual GLP-1 agonists (compared with nonuse of GLP-1 agonists), and the secondary exposure was medications with indication for AUD.
MAIN OUTCOMES AND MEASURES: The primary outcome was AUD hospitalization analyzed in a Cox regression within-individual model. Secondary outcomes were any substance use disorder (SUD)-related hospitalization, somatic hospitalization, and suicide attempt.
RESULTS: The cohort included 227 866 individuals with AUD; 144 714 (63.5%) were male and 83 154 (36.5%) were female, with a mean (SD) age of 40.0 (15.7) years. Median (IQR) follow-up time was 8.8 (4.0-13.3) years. A total of 133 210 individuals (58.5%) experienced AUD hospitalization. Semaglutide (4321 users) was associated with the lowest risk (AUD: adjusted hazard ratio [aHR], 0.64; 95% CI, 0.50-0.83; any SUD: aHR, 0.68; 95% CI, 0.54-0.85) and use of liraglutide (2509 users) with the second lowest risk (AUD: aHR, 0.72; 95% CI, 0.57-0.92; any SUD: aHR, 0.78; 95% CI, 0.64-0.97) of both AUD and SUD hospitalization. Use of any AUD medication was associated with a modestly decreased risk (aHR, 0.98; 95% CI, 0.96-1.00). Semaglutide (aHR, 0.78; 95% CI, 0.68-0.90) and liraglutide (aHR, 0.79; 95% CI, 0.69-0.91) use were also associated with decreased risk of somatic hospitalizations but not associated with suicide attempts (semaglutide: aHR, 0.55; 95% CI, 0.23-1.30; liraglutide: aHR, 1.08; 95% CI, 0.55-2.15).
CONCLUSIONS AND RELEVANCE: Among patients with AUD and comorbid obesity/type 2 diabetes, the use of semaglutide and liraglutide were associated with a substantially decreased risk of hospitalization due to AUD. This risk was lower than that of officially approved AUD medications. Semaglutide and liraglutide may be effective in the treatment of AUD, and clinical trials are urgently needed to confirm these findings.
