Semaglutide and Alcohol Cravings: What the Research Shows

It started as an anecdote. People who began taking semaglutide for diabetes or weight loss kept saying the same strange thing: they had stopped wanting to drink. Not because they decided to quit. The desire just quietly went away.

For a couple of years that was all there was. Then the trials arrived. As of September 2026, three randomized controlled studies have tested semaglutide for drinking in humans, and the largest one produced a result that surprised even the researchers who ran it.

Here is what the evidence actually says, what it does not say, and why no doctor can prescribe semaglutide for drinking as an approved treatment yet.

Why would a weight loss drug change how much you drink?

Semaglutide mimics a gut hormone called GLP-1. Most people know it for what it does in the pancreas and stomach: it helps the body release insulin and slows how quickly food leaves the stomach.

But GLP-1 receptors also sit in parts of the brain that handle reward, motivation and the feeling of “I want that.” Animal work pointed this way years before any human trial. Studies in rats found semaglutide reduced alcohol intake and relapse-like drinking in both males and females [1].

That gives a plausible mechanism: if the drug dials down the reward signal for food, it may dial it down for alcohol too. The human trials were built on that idea.

What did the first randomized trial find?

The first one was small and deliberately cautious. Researchers at the University of North Carolina at Chapel Hill, led by Christian Hendershot, enrolled 48 adults with alcohol use disorder who were not seeking treatment. That detail matters: they were not people who had already decided to cut back [2].

Over nine weeks, participants received weekly semaglutide titrated up to 1 mg, or placebo. The headline measure was a supervised laboratory session where participants could drink as much as they wanted.

Results, published in JAMA Psychiatry in February 2025 [3]:

• Grams of alcohol consumed in the lab session fell significantly with semaglutide (effect size beta −0.48, 95 percent confidence interval −0.85 to −0.11)
• Peak breath alcohol concentration fell significantly (beta −0.46, −0.87 to −0.06)
• Weekly alcohol craving dropped significantly compared with placebo
• Among participants who smoked, cigarettes per day fell more on semaglutide

What it did not show: semaglutide did not change average drinks per calendar day, or the number of days people drank. It changed how much they drank when they drank, and how much they wanted to.

The dose used was well below the obesity dose. The trial ran nine weeks. Forty-eight people is a pilot.

What did the 2026 Lancet trial show?

This is the one that changed the conversation.

Researchers at Copenhagen University Hospital, working with scientists from the US National Institutes of Health including the directors of the national alcohol and drug abuse institutes, ran a 26-week randomized, double-blind, placebo-controlled trial in 108 adults who had both alcohol use disorder and obesity, and who were seeking treatment [4].

Everyone got cognitive behavioral therapy. On top of that, half got weekly semaglutide and half got placebo.

The findings, published in The Lancet on May 2, 2026 [5]:

• Heavy drinking days fell by 41.1 percentage points in the semaglutide group against 26.4 points on placebo, a difference of 13.7 percentage points (p=0.0015). Note how large the placebo-plus-therapy improvement was on its own: most of the benefit in both arms came from the cognitive behavioral therapy everybody received
• Total alcohol consumption fell about 1,550 g per 30 days with semaglutide versus 1,026 g with placebo
• Blood biomarkers of alcohol consumption — phosphatidylethanol and gamma-glutamyl transferase — supported the self-reported drinking data
• Body weight, blood pressure and other clinical measures improved more in the semaglutide group
• Side effects were mostly gastrointestinal. They were largely transient, but not trivial: nausea was reported by 57 percent of the semaglutide group against 7 percent on placebo

The number that got clinicians’ attention was the number needed to treat: 4.3. That is a measure of how many people you have to treat for one to benefit. For currently approved alcohol use disorder medications, the comparable figure is 7 or higher [4].

“Very few medications are currently approved for alcohol use disorder, and these are vastly underutilized,” said George Koob, director of the National Institute on Alcohol Abuse and Alcoholism and a study co-author. “A new option that is more accessible and more effective could be a gamechanger for closing the treatment gap” [4].

Two limits to keep in front of you. Everyone in the trial had obesity, so this does not tell you whether semaglutide helps someone with alcohol use disorder at a normal weight. And everyone received therapy, so the honest framing is “semaglutide added to cognitive behavioral therapy,” not “semaglutide instead of treatment.”

Does the pill form work too?

Partly. In July 2026, researchers at the University of Colorado Anschutz Medical Campus published a trial of oral semaglutide in 50 adults with moderate to severe alcohol use disorder who wanted to cut down or stop [6].

Participants took 3 mg daily for four weeks, then 7 mg daily for four weeks, or placebo. The oral form matters because some people who would never inject themselves will take a pill.

The trial missed its primary endpoint. Semaglutide did not significantly reduce laboratory cue-elicited craving at week 6, or drinks per day [7].

But several prespecified secondary outcomes did improve significantly:

• Heavy drinking days went down
• Drinks per drinking day went down
• Day-to-day, real-world alcohol craving went down
• Alcohol-related problems and overall drinking risk level went down
• Cannabis use days went down

An honest reading: the pattern matches the injectable trials, but a missed primary endpoint is a missed primary endpoint. This is phase 2 evidence, and phase 2 results do not always survive contact with a bigger trial.

What do the real-world data say?

Before any of these trials, large database studies were already pointing the same direction. A 2024 analysis of electronic health records covering 83,825 patients with obesity found semaglutide was associated with a 50 to 56 percent lower risk of both new and recurrent alcohol use disorder compared with other anti-obesity medications [1].

A 2026 study in The BMJ emulated eight parallel trials in US Department of Veterans Affairs records to look at whether starting a GLP-1 drug was associated with lower risk of new alcohol, cannabis, cocaine, nicotine, opioid and other substance use disorders, and better outcomes in people who already had one [8].

And a Swedish national cohort of nearly 100,000 people with depression or anxiety, published in The Lancet Psychiatry in 2026, found that during periods of semaglutide use, the risk of hospital care or sick leave for substance use disorder was 47 percent lower [9].

These are all observational. People who get prescribed semaglutide differ from people who do not, in ways no statistical adjustment fully captures. Effect sizes in health-record studies are typically larger than in randomized trials, and that pattern held here.

What about smoking and other substances?

The smoking evidence is more mixed. A phase 2a trial published in JAMA Network Open in 2026 tested semaglutide by itself in adults who smoke daily and were not necessarily trying to quit [10].

Semaglutide did not significantly increase people’s ability to resist smoking in the laboratory, which was the primary measure. It did significantly reduce weekly cigarette craving, body weight and hemoglobin A1c. Weekly cigarettes per day did not change significantly.

Several larger semaglutide smoking trials are now recruiting, including studies in people with type 2 diabetes and in psychiatric populations.

For opioids, a randomized trial protocol was published in October 2025 but has not reported results [11]. Observational work has linked semaglutide to lower opioid overdose risk in people with type 2 diabetes and opioid use disorder.

Is anyone filing for an addiction approval?

Not yet, and this is an important practical point.

Novo Nordisk’s alcohol-related research has been framed around liver disease, not addiction. The company ran a phase 2 trial of 270 people with alcohol-related liver disease, testing semaglutide, cagrilintide and an experimental agent alone and in combination. The primary endpoint is a liver fibrosis score; change in alcohol consumption is a secondary endpoint. The trial completed in January 2026 and results were not public as of September 2026 [12].

When the Lancet trial came out, commentators noted that there is no obvious regulatory template for approving a metabolic drug as an addiction medicine, and that the next move is the company’s.

Until something is filed and approved, prescribing semaglutide for drinking is off-label. Insurance will not cover it for that reason, and a prescriber has to justify the decision on their own clinical judgment.

What should you do with this information?

A few things are worth saying plainly.

This is not a reason to start or stop any medication on your own. Semaglutide has real side effects, real contraindications and a real cost. The trials described here ran under medical supervision with careful monitoring.

If you are already taking semaglutide and have noticed you want to drink less, that is consistent with what the research shows. It is not imaginary and it is not unusual.

If you are struggling with drinking, there are treatments available today that are approved, studied and often underused. Naltrexone, acamprosate and disulfiram all have FDA approval for alcohol use disorder. A 2026 experimental drug is not a substitute for talking to a healthcare provider about options that exist now.

If you have both obesity and a drinking problem, the Lancet trial is the closest thing to evidence for your situation, and it is worth raising with your provider specifically because that was the population studied.

What comes next

The researchers behind the Lancet trial said their next step is to examine effects over a longer period and in a larger population. That likely means a phase 3 program, which typically takes years.

Watch for: whether Novo Nordisk or another sponsor files for an alcohol use disorder indication, whether the Danish result replicates in people without obesity, and whether the alcohol-related liver disease trial results shift the picture when they publish.

Until then, the summary is short. The effect looks real, the trials are still small, the mechanism makes sense, and nobody is approved to prescribe it for this.

Sources

1. Associations of semaglutide with incidence and recurrence of alcohol use disorder in real-world population — Nature Communications, 2024
2. Semaglutide Shows Promise in Reducing Cravings for Alcohol, Heavy Drinking — UNC Health, February 2025
3. Once-Weekly Semaglutide in Adults With Alcohol Use Disorder: A Randomized Clinical Trial — JAMA Psychiatry, 2025;82(4):395-405
4. Adding weekly GLP-1 to cognitive behavioral therapy further reduces heavy drinking — National Institutes of Health, April 30, 2026
5. Once-weekly semaglutide versus placebo in patients with alcohol use disorder and comorbid obesity — The Lancet, May 2, 2026;407(10540):1687-1698
6. New Research: Oral Semaglutide Reduced Heavy Drinking in Adults with Alcohol Use Disorder — American Psychiatric Association, July 29, 2026
7. Oral Semaglutide for Alcohol Use Disorder: A Randomized Clinical Trial — American Journal of Psychiatry, 2026;183(9):636-645
8. GLP-1 receptor agonists and risk of substance use disorders among US veterans with type 2 diabetes — The BMJ, March 2026
9. Association between GLP-1 receptor agonist use and worsening mental illness in people with depression and anxiety in Sweden — The Lancet Psychiatry, 2026;13(4):327
10. Once-Weekly Semaglutide in Adults With Daily Cigarette Use: A Randomized Clinical Trial — JAMA Network Open, 2026;9(5):e2614898
11. Efficacy of semaglutide in abstinence from illicit and nonprescribed opioids: trial protocol — Addiction Science and Clinical Practice, October 2025
12. Effects of NNC0194-0499, Cagrilintide, and Semaglutide on Liver Damage and Alcohol Use in People With Alcohol-related Liver Disease — ClinicalTrials.gov