Semaglutide Alcohol Use Disorder Trial: Results

The 2025 semaglutide alcohol use disorder trial found less drinking in a laboratory test and lower cravings. It did not clearly change how often people drank. That is an encouraging result from 48 people, not proof of lasting recovery.

Alcohol use disorder, or AUD, means difficulty controlling drinking despite harm. Christian Hendershot's team tested whether semaglutide could change drinking even when participants weren't trying to cut back.

What the semaglutide alcohol use disorder trial asked

Semaglutide copies GLP-1, a gut hormone involved in appetite and blood sugar. Researchers wanted to know whether it also changed alcohol consumption and craving.

The Hendershot trial in JAMA Psychiatry, 2025, took place at the University of North Carolina at Chapel Hill. Its registry record, NCT05520775, identifies the study and its planned measures.

Who took part and what they received

Researchers assigned 24 adults to semaglutide and 24 to placebo, an inactive comparison treatment. Eligible ages were 21–65; the average age was about 40, and 34 participants were women. Most had overweight or obesity.

Participants had AUD but weren't seeking alcohol treatment or actively trying to reduce drinking. Exclusions included diabetes, serious medical illness and body mass index below 23. Body mass index is a weight-to-height measure.

The nine-week study used weekly injections: 0.25 mg for four weeks, then 0.5 mg for four weeks. A final 1 mg dose was optional depending on tolerability and physician judgment. These were study doses, not a prescribing schedule for AUD.

What researchers measured and what changed

The main test compared alcohol consumed during monitored, two-hour drinking sessions before treatment and after week eight. Researchers measured grams consumed and peak breath alcohol level. The final session occurred while participants were receiving 0.5 mg weekly.

Laboratory consumption was lower with semaglutide than placebo. After accounting for each person's pre-treatment session, the semaglutide group drank about 25 grams less alcohol during the session than the placebo group (a U.S. standard drink holds 14 grams), though the uncertainty range ran from about 6 to 45 grams. The paper called this a medium-to-large effect. That does not mean 48% less drinking. Peak breath alcohol levels were also lower.

The published laboratory analysis included just 25 people who drank during both laboratory sessions: 13 receiving semaglutide and 12 receiving placebo. Others attended but chose not to drink, and that choice wasn't linked to treatment group. That smaller sample matters when judging the result.

What changed in weekly drinking

Weekly reports showed fewer drinks per drinking day, lower craving and a greater decline in heavy-drinking days with semaglutide. Heavy drinking meant at least four drinks for women or five for men in a day. These were additional, exploratory measures, intended to generate questions for later trials.

Average drinks per calendar day, the one weekly drinking measure the trial registered in advance, and the number of drinking days did not clearly differ. Thirteen participants smoked; cigarette counts fell more with semaglutide, but that subgroup was too small to establish a quitting benefit.

Side effects in each group

About seven in ten semaglutide participants reported nausea. The paper's Supplement 2, eTable 1, gives these selected results. An adverse event is a health problem reported during a study. Each group contained 24 people; counts represent people reporting the problem, not episodes.

Most symptoms were mild. The supplement's severity rows list moderate problems in 8 semaglutide and 4 placebo participants, and severe ones in 2 and 1; the paper notes that “severe” describes how much a symptom interfered with daily life and is not the same as a serious adverse event. No medication-related discontinuations occurred; one placebo participant left during a COVID-19 infection. A small, short study cannot rule out uncommon harms.

For the separate question of drinking during treatment, semaglutide and alcohol explains the overlapping risks.

What earlier and later studies add

The earlier Klausen trial in JCI Insight, 2022, studied exenatide, another GLP-1 medicine, in 127 adults for 26 weeks. Both groups received counseling. Exenatide did not outperform placebo on heavy-drinking days overall. Exploratory findings suggested fewer heavy-drinking days in participants with obesity, but more in those with body mass index below 25. Those subgroup findings weren't the main result.

As of October 29, 2026, the day this article was written, later studies have added evidence without settling every question. A May 2026 Lancet trial enrolled 108 treatment-seeking adults with AUD and obesity. Both groups received cognitive behavioral therapy, counseling that works on drinking triggers and responses. After 26 weeks on a higher weekly dose (2.4 mg) than the 2025 trial used, heavy-drinking days fell from baseline by 41.1 percentage points with semaglutide versus 26.4 with placebo; the trial's statistically adjusted estimate of the difference between groups was 13.7 percentage points.

As of October 29, 2026, the day this article was written, an oral-semaglutide trial published online July 29, 2026 adds a third result. It followed 50 treatment-seeking adults with moderate to severe AUD for eight weeks. Heavy-drinking days improved, but the main laboratory craving measure and drinks per calendar day did not clearly differ from placebo.

As of October 29, 2026, the day this article was written, the same research group's smoking trial, published May 26, 2026, found no clear benefit on its two planned main laboratory smoking tests or on cigarettes per day; semaglutide and smoking explains that separate result.

What the headlines leave out

As of October 29, 2026, the day this article was written, a September 21, 2026 editorial by Singh and Das emphasizes the difference between reduced drinking and lasting abstinence, meaning not drinking at all. Benefits after stopping treatment and effects on long-term health remain unsettled. A claim that Ozempic ends alcohol addiction goes beyond these studies.

The original trial lasted weeks and excluded people seeking alcohol treatment. It cannot establish lasting recovery, the best AUD dose or equivalence between the tested medicine and a compounded preparation.

What larger trials are testing now

As of October 29, 2026, the day this article was written, the VA's CRAVE record, updated September 21, lists recruiting status and a planned 622 participants. It is a phase 3 trial, a larger study intended to confirm benefit. Its main measure is a two-level drop in a drinking-risk scale, rather than a laboratory drinking session.

Tirzepatide acts on GLP-1 and GIP, a second gut-hormone signal. As of October 29, 2026, the day this article was written, its AUD trial record, last updated October 14, 2025, lists 42 planned participants and recruiting status, with no posted results. Semaglutide findings cannot establish tirzepatide's alcohol effects.

What the labels and existing treatment say

As of October 29, 2026, the day this article was written, neither the Ozempic nor Wegovy label lists AUD in section 1. Ozempic's label, “Revised: 05/2026,” includes “improve glycemic control,” meaning blood-sugar control. Wegovy's label, “Revised: 06/2026,” includes “reduce excess body weight.”

Semaglutide and tirzepatide prescribed through Promise are dispensed by a licensed U.S. compounding pharmacy as compounded medications, which are different from an FDA-approved product: the formulations offered here are not FDA-approved. A licensed provider may still prescribe — that decision is between you and your doctor.

NIAAA's treatment guide describes naltrexone, acamprosate and disulfiram, alongside counseling, as established options. Early semaglutide findings don't replace that care. Alcohol withdrawal, symptoms caused by a sharp reduction after prolonged heavy drinking, can be life-threatening and needs medical assessment.

What belongs in the visit conversation

A weight-management prescription and alcohol-use treatment address different needs. Drinking patterns, previous withdrawal and current medicines help a clinician assess both. Questions about NAD for addiction involve a different compound and a separate evidence base.

At Promise, a licensed U.S. provider reviews every request; not everyone qualifies. The prescriber sets any dose and follow-up plan.