Small trial finds low-dose semaglutide cut lab alcohol intake in adults with alcohol use disorder

A nine-week phase 2 randomized trial found that low-dose semaglutide reduced alcohol consumed during a laboratory drinking task and lowered weekly alcohol craving in adults with alcohol use disorder, but it did not change how often they drank ^[1]. The results were published in JAMA Psychiatry, and the trial is registered as NCT05520775 ^[1].

The trial enrolled 48 non-treatment-seeking adults with alcohol use disorder at a single US academic medical center between September 2022 and February 2024; 504 people were assessed for eligibility ^[1]. Participants were randomly assigned to once-weekly subcutaneous semaglutide or placebo, given at weekly clinic visits. The dose rose over the nine weeks of treatment: 0.25 mg per week for four weeks, 0.5 mg per week for four weeks, then 1.0 mg for one week ^[1]. The group was 71% female (34 participants), with a mean age of 39.9 years (SD, 10.6) ^[1].

For the primary outcome, a laboratory task in which participants could drink alcohol after treatment, semaglutide reduced grams of alcohol consumed (beta, -0.48; 95% CI, -0.85 to -0.11; P = .01) and peak measured breath alcohol concentration (beta, -0.46; 95% CI, -0.87 to -0.06; P = .03) ^[1]. The abstract describes the effect sizes as medium to large ^[1]. That laboratory comparison included 25 participants with complete data, 12 in the placebo group and 13 in the semaglutide group ^[1].

The results were mixed on drinking measured outside the laboratory. Semaglutide did not affect average drinks per calendar day or the number of drinking days ^[1]. It did reduce drinks per drinking day (beta, -0.41; 95% CI, -0.73 to -0.09; P = .04) and weekly alcohol craving on the Penn Alcohol Craving Scale (beta, -0.39; 95% CI, -0.73 to -0.06; P = .01), and it predicted greater reductions in heavy drinking over time relative to placebo (beta, 0.84; 95% CI, 0.71 to 0.99; P = .04) ^[1]. In a subsample of participants who currently smoked cigarettes, semaglutide also predicted greater relative reductions in cigarettes per day (beta, -0.10; 95% CI, -0.16 to -0.03; P = .005) ^[1].

The authors wrote that the 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" ^[1]. One author reported consulting fees from Novo Nordisk outside the submitted work and research-related contracts paid to his institution from Bayer, Boehringer-Ingelheim, Carmot, Diasome, Eli Lilly, Novo Nordisk, Rhythm Pharmaceuticals, and vTv Therapeutics; no other disclosures were reported ^[1].

Why it matters for patients

Semaglutide is not established as a treatment for alcohol use disorder, and this trial does not change that. It was small (48 people), short (nine weeks), conducted at one site, and enrolled people who were not seeking treatment for their drinking ^[1]. Results in a group like that may not carry over to people who are actively looking for help with alcohol use.

The pattern of results also matters for expectations. The largest effects appeared in a laboratory task and in craving scores, while two broader measures — average drinks per calendar day and number of drinking days — showed no significant difference between semaglutide and placebo ^[1]. Side effects, dropouts, and how people fared after treatment ended were not summarized in the published abstract, so those details have not been reported here.

The trial tested semaglutide at 0.25 mg, 0.5 mg, and 1.0 mg once weekly, which the authors describe as low dose ^[1]. It did not compare semaglutide against any medication already used for alcohol use disorder, and no such comparison has been reported in these findings.

What happens next

The authors state that larger clinical trials of GLP-1 receptor agonists for alcohol use disorder are justified by these results ^[1]. Specific follow-up trials, timelines, or regulatory steps have not been disclosed. The trial's registration record is available under ClinicalTrials.gov identifier NCT05520775 ^[1].