Once-Weekly Semaglutide in Adults With Alcohol Use Disorder: A Randomized Clinical Trial

Abstract

Importance: Preclinical, observational, and pharmacoepidemiology evidence indicates that glucagon-like peptide 1 receptor agonists (GLP-1RAs) may reduce alcohol intake. Randomized trials are needed to determine the clinical significance of these findings.

Objective: To evaluate the effects of once-weekly subcutaneous semaglutide on alcohol consumption and craving in adults with alcohol use disorder (AUD).

Design, setting, and participants: This was a phase 2, double-blind, randomized, parallel-arm trial involving 9 weeks of outpatient treatment. Enrollment occurred at an academic medical center in the US from September 2022 to February 2024. Of 504 potential participants assessed, 48 non-treatment-seeking participants with AUD were randomized.

Intervention: Participants received semaglutide (0.25 mg/week for 4 weeks, 0.5 mg/week for 4 weeks, and 1.0 mg for 1 week) or placebo at weekly clinic visits.

Main outcomes and measures: The primary outcome was laboratory alcohol self-administration, measured at pretreatment and posttreatment (0.5 mg/week). Secondary and exploratory outcomes, including prospective changes in alcohol consumption and craving, were assessed at outpatient visits.

Results: Forty-eight participants (34 [71%] female; mean [SD] age, 39.9 [10.6] years) were randomized. Low-dose semaglutide reduced the amount of alcohol consumed during a posttreatment laboratory self-administration task, with evidence of medium to large effect sizes for grams of alcohol consumed (β, -0.48; 95% CI, -0.85 to -0.11; P = .01) and peak breath alcohol concentration (β, -0.46; 95% CI, -0.87 to -0.06; P = .03). Semaglutide treatment did not affect average drinks per calendar day or number of drinking days, but significantly reduced drinks per drinking day (β, -0.41; 95% CI, -0.73 to -0.09; P = .04) and weekly alcohol craving (β, -0.39; 95% CI, -0.73 to -0.06; P = .01), also predicting greater reductions in heavy drinking over time relative to placebo (β, 0.84; 95% CI, 0.71 to 0.99; P = .04). A significant treatment-by-time interaction indicated that semaglutide treatment predicted greater relative reductions in cigarettes per day in a subsample of individuals with current cigarette use (β, -0.10; 95% CI, -0.16 to -0.03; P = .005).

Conclusions and relevance: These findings provide initial prospective evidence that low-dose semaglutide can reduce craving and some drinking outcomes, justifying larger clinical trials to evaluate GLP-1RAs for alcohol use disorder.

Trial registration: ClinicalTrials.gov Identifier: NCT05520775.

Figure 1.. CONSORT Flow Diagram

Figure 2.. Laboratory Self-Administration

Laboratory self-administration, measured in estimated grams of alcohol (A) and peak measured breath alcohol concentration (BrAC; B). Bars depict posttreatment alcohol self-administration following treatment week 8 (0.5 mg/wk), controlling for pretreatment self-administration among those without missing data on the self-administration procedure (n = 25; 12 in the placebo group and 13 in the semaglutide group). C, Mean BrAC measured across 30-minute intervals as a function of treatment condition. D, Descriptive BrAC measurements during self-administration as a function of treatment condition and time point. Bars depict group means and whiskers standard errors.

Figure 3.. Prospective Changes in Weekly Alcohol Outcomes

A, n = 39-48 across weeks (18-24 in the placebo group and 21-24 in the semaglutide group). B, n = 31-48 (15-24 in the placebo group and 16-24 in the semaglutide group). C, n = 39-48 (18-24 in the placebo group and 21-24 in the semaglutide group). D, n = 39-48 (18-24 in the placebo group and 21-24 in the semaglutide group). E, n = 41-48 (19-24 in the placebo group and 22-24 in the semaglutide group). F, n = 41-48 (19-24 in the placebo group and 22-24 in the semaglutide group). Alcohol craving was measured via the Penn Alcohol Craving Scale (0-30). Data points depict group means and standard errors for weekly measurements of the outcome. Medication commenced at week 1. Weekly time points indicate data collected at the end of the week following that week’s injection (eg, week 1 drinking data were measured at week 2 and reflect the week following the week 1 injection).

Figure 4.. Medication Group Differences in Weekly Drinking and Craving as a Function of Treatment Period

A-E, Bars depict group means for changes from baseline on alcohol consumption and craving outcomes during treatment weeks 1-4 (0.25 mg/wk) and 5-8 (0.5 mg/wk) and whiskers indicate standard errors. Effect size estimates (Cohen d) reflect medication vs placebo group differences on the outcomes presented. Conventional benchmarks are d = 0.20 for a small effect, d = 0.5 for a medium effect, and d = 0.8 for a large effect, Numbers of participants across weeks were 43-47 for all models (placebo: 23 in weeks 1-4 and 20 in weeks 5-8; semaglutide: 24 in weeks 1-4 and 23 in weeks 5-8). For visual presentation, panel C excludes the data from 1 participant with an extreme outlier >4 SD above the mean. F, Proportion of participants with zero heavy drinking days by treatment condition and dose (weeks 1-4, 0.25 mg/wk; weeks 5-8, 0.5 mg/wk). ^aP < .005.