A 2 minute assessment to get a personalized mental health or alcohol recovery plan.
Two randomized trials have now tested semaglutide against placebo for drinking. Here is exactly what each one measured, what it found, and what it did not.
What You'll Discover:
• What the 2025 JAMA Psychiatry semaglutide trial actually tested.
• What the 2026 Lancet trial added, and why the therapy piece changes how you read it.
• How to translate the effect sizes into something you would notice in a real week.
• The four limits that honest coverage has to include.
• Where naltrexone fits, since it is the option that already has FDA approval.
For close to a decade, doctors kept hearing the same offhand comment from patients taking GLP-1 medications. They were drinking less, and they had not set out to.
That kind of story is a starting point, not evidence. So researchers ran the trials.
Two randomized, placebo-controlled studies have now reported results. They point in the same direction without settling the question, and the details matter more than the headlines suggested.
Why Researchers Started Looking at GLP-1 Medications and Alcohol
GLP-1 receptor agonists were built to manage blood sugar and, later, body weight. Semaglutide is the best known of them, sold as Ozempic and Wegovy.
The reason anyone thought to test them on drinking comes down to receptor location. GLP-1 receptors are not confined to the gut and pancreas. They also appear in the reward-processing regions of the brain.
Alcohol drives a dopamine surge through those same circuits. That overlap is the entire reason the idea was worth a trial, and we go deeper into the biology in our explainer on semaglutide and alcohol cravings.
Animal studies came first, and they were consistent. Rodents given GLP-1 agonists drank less alcohol.
Animal data is where a lot of promising ideas go to die, though. Human trials are what count, and until 2025 there were almost none worth citing.
The 2025 JAMA Psychiatry Trial
The first randomized human trial to report clean results was published in JAMA Psychiatry in early 2025, led by researcher Christian Hendershot.
What It Tested
This was a phase 2, double-blind, randomized trial run at a single academic medical center in the United States, with enrollment running from September 2022 to February 2024.
Forty-eight adults with alcohol use disorder were randomized. They came out of a pool of 504 people who were assessed, which tells you how narrow the eligibility window was.
One design detail sets this trial apart from most alcohol research. Participants were not seeking treatment for their drinking. They were recruited, not walking in asking for help.
Doses were deliberately low. Participants received semaglutide at 0.25 mg per week for four weeks, then 0.5 mg for four weeks, then 1.0 mg for a final week, which sits below the range typically used for weight management.
Total treatment time was nine weeks.
What It Found
Two outcomes moved in the semaglutide group. Drinks per drinking day dropped, and weekly alcohol craving dropped.
Reported as standardized coefficients, drinks per drinking day fell by 0.41 and craving fell by 0.39. Both cleared statistical significance.
The trial also included a laboratory self-administration session, where participants could drink in a controlled, observed setting.
Grams of alcohol consumed and peak breath alcohol concentration both fell in the semaglutide group, with medium to large effect sizes.
That lab result is worth pausing on. Self-reported drinking is famously unreliable, and a controlled session removes most of the guesswork.
The researchers were careful about how they framed all of it. Nine weeks is short, 48 people is small, and they described the study as a signal worth chasing rather than a finished answer.
The 2026 Lancet Trial
The second trial, published in The Lancet in May 2026, was larger, longer, and built very differently.
What It Tested
Researchers at Copenhagen University Hospital enrolled 108 participants between June 2023 and February 2025, split evenly at 54 per arm.
Every participant had moderate to severe alcohol use disorder plus comorbid obesity. Every participant was treatment-seeking, which is the mirror image of the JAMA Psychiatry recruitment approach.
Treatment ran 26 weeks, roughly six months. Participants received a weekly injection of semaglutide or placebo.
All of them were also offered standard cognitive behavioral therapy for alcohol use disorder.
That last piece changes how the results should be read. This was not medication tested alone in a vacuum. It was medication layered on top of behavioral support, which is how alcohol care usually works in practice.
What It Found
The primary outcome was heavy drinking days, a measure the National Institute on Alcohol Abuse and Alcoholism defines in its guidance on alcohol drinking patterns.
Both groups improved. Everyone was getting therapy, so that was expected.
The semaglutide group's heavy drinking days fell 41.1 percentage points from baseline. The placebo group's fell 26.4 percentage points.
The gap between them was 13.7 percentage points in favor of semaglutide, with a p-value of 0.0015.
Translated into days, the semaglutide group averaged roughly five heavy drinking days in the previous 30 by the end of the trial. The placebo group averaged about nine.
Blood-alcohol biomarkers moved in the same direction as the self-reports, which is a meaningful check on the data.
The research team also reported a number needed to treat of 4.3. That means roughly four to five people would need the medication for one additional person to benefit.
Side effects were mostly gastrointestinal, and the authors described them as transient and mild. Weight, blood pressure, and other clinical measures also improved more in the semaglutide group, which was expected.
Two Trials, Side by Side
What Those Numbers Look Like in a Real Week
Percentage points are hard to feel. Days are not.
Going from nine heavy drinking days a month to five means four fewer mornings of regret. Four fewer nights of wrecked sleep, four fewer arguments you half remember.
It also is not zero. Most participants in neither trial stopped drinking entirely, and neither research team claimed they did.
For daily life, the craving finding from the 2025 trial may be the more interesting one. Craving is what makes the decision hard in the first place.
A medication that turns the volume down on craving changes how much willpower has to carry on its own.
That is the same reason people describe naltrexone the way they do. We compare the two mechanisms in our breakdown of whether GLP-1 medications reduce alcohol cravings.
The Four Limits That Matter
None of these are minor, and any coverage that skips them is selling something.
Sample sizes are small. Forty-eight and 108 participants are early-stage numbers. Effects frequently shrink when trials scale up to hundreds or thousands of people.
The Lancet population all had obesity. Every single participant. Whether the same benefit appears in people at a normal weight is genuinely unknown, and an earlier GLP-1 trial found its strongest signal specifically in the subgroup with obesity.
Both trials were short. Nine weeks and 26 weeks. Drinking patterns run in years, and no long-term data has been published.
Neither trial supports self-treatment. These are prescription medications with real side effects and real interactions. We cover the safety picture in detail in our piece on whether GLP-1 medication for alcohol is safe.
There is one more limit that gets less attention. Both trials measured drinking, not the life around it.
Nobody has published data on whether people in these trials kept the gains after the injections stopped. That is the question that actually determines whether a medication changes a life.
What Is Approved Today and What Is Not
This is the part that gets blurred in most coverage, so it is worth stating flatly.
Semaglutide is not FDA-approved for alcohol use disorder. It is approved for type 2 diabetes and for weight management. Any use aimed at drinking is off-label.
Naltrexone is FDA-approved for alcohol use disorder, and has been since 1994. It is a 50mg oral tablet that blocks the opioid receptors alcohol acts on, which dulls the reward and, over weeks, the craving.
That gap in status is not a technicality. On one side sits three decades of accumulated trial data across tens of thousands of participants. On the other sit two encouraging studies with fewer than 160 people between them.
The full side-by-side is in our guide to naltrexone versus GLP-1 medication for alcohol use disorder.
Where the Two Approaches Meet
Some people end up on both, under physician supervision, and the reasoning is straightforward.
Naltrexone acts on the opioid and dopamine reward pathway that makes the next drink appealing once you have started the first. GLP-1 medications appear to work earlier in the sequence, on the appetitive pull that makes you want the first one at all.
Different points on the same circuit. That is why the combination is being discussed at all.
Choose Your Horizon supports both paths. Naltrexone-assisted care is the core of what we do, and physician-guided GLP-1 options are available for people who are appropriate candidates for them.
Whether combining them makes sense for you is a decision for a prescribing physician with your full health history in front of them, not something to settle from a news article.
What You Can Do While the Science Catches Up
Waiting for a phase 3 readout is a poor plan if drinking is affecting your life right now.
The evidence-backed option available today is naltrexone paired with actual support. That is the combination the trial literature has backed for thirty years.
It is also structurally the same setup the Lancet trial used, with medication added to behavioral care rather than replacing it. That design choice was not an accident.
An integrated solution means both halves working together. Medication that lowers the biological pull, plus coaching, tracking, and clinical follow-up that handle the parts no pill reaches.
You do not need a diagnosis or a rock-bottom story to qualify for help. Wanting to drink less is reason enough.
It also helps to know what you are actually measuring. Both trials tracked heavy drinking days and drinks per drinking day rather than pass or fail abstinence.
Those are the numbers clinical, neurological, and behavioral science has settled on because they reflect how change usually happens. Fewer heavy nights first, then fewer drinking nights, then a different relationship with the whole thing.
Tracking your own version of those numbers for two weeks tells you more than any headline will.
The Honest Summary
Two randomized trials, both positive, both small, both short. That is where semaglutide and alcohol stand as of today.
The direction is encouraging enough that the National Institutes of Health, the Department of Veterans Affairs, and academic centers are all funding larger studies. That is real institutional momentum, and it does not happen around weak signals.
Encouraging is not the same as proven, and it is not the same as approved. Anyone telling you otherwise has gotten ahead of the data.
If you want help with drinking today, the approved and well-studied option already exists. It works for a great many people, and it does not require waiting on a trial that reports in 2028.
Frequently Asked Questions
Does semaglutide reduce alcohol cravings?
In the 2025 JAMA Psychiatry trial, weekly alcohol craving fell significantly in the semaglutide group compared with placebo. That was one trial of 48 people over nine weeks, so the finding is promising rather than settled.
Is Ozempic approved for alcohol use disorder?
No. Semaglutide is approved for type 2 diabetes and weight management only. Using it for drinking is off-label, which is legal under a physician's judgment but not the same as FDA approval.
How much did the Lancet trial actually reduce drinking?
Heavy drinking days fell 41.1 percentage points from baseline on semaglutide versus 26.4 points on placebo, a 13.7 point difference. In practical terms, that was about five heavy drinking days per month instead of nine.
Why did both trials focus on people with obesity?
The Lancet trial required comorbid obesity by design, partly because an earlier GLP-1 study found its clearest benefit in that subgroup. It means the results cannot simply be assumed to apply to people at a normal weight.
Can you take naltrexone and a GLP-1 medication at the same time?
Some people do, under physician supervision, because the two act on different points of the reward pathway. That decision belongs with a prescribing doctor who knows your full medical history.
What would it take for a GLP-1 to be approved for drinking?
Large phase 3 trials with hundreds of participants and endpoints the FDA accepts, followed by regulatory review. Several are underway now, and realistic timelines run in years.
Find out whether naltrexone fits your situation. Take the online Alcohol Use Assessment and a physician will review your history and your goals with you.
