Oral Semaglutide for Alcohol Use Disorder: A Randomized Clinical Trial

When a Weight-Loss Medicine Meets Alcohol Use Disorder

A new randomized trial of oral semaglutide found fewer heavy-drinking days and fewer alcohol-related consequences—but missed its primary craving outcome. The mixed result is precisely why the study deserves attention without hype.

An unexpected therapeutic bridge

GLP-1 receptor agonists entered public consciousness as treatments for diabetes and obesity. Yet clinicians and patients began reporting something more difficult to measure: some people described less “noise” around food, alcohol, or other rewards. Animal studies and observational data added biological plausibility, but anecdotes cannot establish a treatment effect.

That is why the randomized clinical trial by Joseph Schacht and colleagues, published online in the American Journal of Psychiatry on July 29, 2026, matters. It tested oral semaglutide in people actively seeking help for alcohol use disorder rather than relying on social-media reports or retrospective medical records.

What the researchers did

The phase 2, double-blind trial enrolled 50 adults with moderate-to-severe alcohol use disorder. Participants were randomized to oral semaglutide or matching placebo for eight weeks. The semaglutide group received 3 mg daily for four weeks and then 7 mg daily for four weeks.

The prespecified primary outcome was alcohol cue–elicited craving in a laboratory at week six. Key secondary outcomes included heavy-drinking days and drinks per day during the final four weeks. The researchers also examined everyday craving, drinks per drinking day, alcohol-related consequences, World Health Organization drinking-risk categories, and cannabis-use days.

A promising—and genuinely mixed—result

Semaglutide did not significantly reduce the primary laboratory craving outcome compared with placebo. It also did not significantly reduce overall drinks per day. Those negative findings must lead any responsible account of the study.

Several other outcomes nevertheless favored semaglutide. The model-estimated effect showed fewer heavy-drinking days, fewer drinks per drinking day, and lower naturalistic craving. Alcohol-related consequences declined faster, and more participants reduced their World Health Organization risk-drinking level by at least one category. Cannabis-use days also declined, although that finding was exploratory.

This pattern raises an intriguing question: might a treatment alter behavior in daily life even when a laboratory cue test does not capture the change?

That is a hypothesis, not a conclusion. Laboratory and real-world outcomes measure different things, and a small study can produce chance-positive secondary findings.

Why this is not ready for routine prescribing

Fifty participants and eight weeks of treatment are enough to detect a signal—not enough to establish durable effectiveness, identify who benefits, compare semaglutide with approved alcohol-use-disorder medications, or characterize uncommon harms.

Multiple outcomes were tested, increasing the possibility that some positive results occurred by chance. The study also cannot tell us whether benefits persist after the medicine is stopped.

Semaglutide is not FDA-approved for alcohol use disorder. Current evidence-based options—including naltrexone, acamprosate, disulfiram in selected circumstances, behavioral treatments, mutual-help approaches, and treatment of co-occurring psychiatric illness—remain essential.

A medication associated with weight loss also raises questions about access, cost, contraindications, adverse effects, and equitable use.

What the study changes today

It should not trigger indiscriminate off-label enthusiasm. It should trigger larger, longer and adequately powered trials with clearly prioritized outcomes, objective alcohol biomarkers, detailed safety analyses, and comparisons with established treatments.

It also supports a broader scientific idea: appetite, reinforcement, craving, and substance use may share more neurobiological machinery than our clinical silos imply.

The right headline is not “Ozempic cures alcoholism.” It is more modest—and more interesting:

A metabolic medicine has produced another randomized signal in alcohol use disorder, and addiction researchers now have a compelling reason to keep looking.

Schacht JP, Sakai JT, Raymond K, Shelton R. Oral Semaglutide for Alcohol Use Disorder: A Randomized Clinical Trial. American Journal of Psychiatry. Published online July 29, 2026:appiajp20260003. doi:10.1176/appi.ajp.20260003. PMID: 42522065.