Why GLP-1 drugs curb alcohol cravings: what the research shows

Semaglutide is approved for type 2 diabetes and weight loss, not for alcohol use disorder. That hasn’t stopped researchers from testing it. Over the past two years, several randomized trials have asked whether GLP-1 drugs reduce alcohol craving and drinking. The results are consistent enough to take seriously and mixed enough to demand caution.

Here’s what the trials found, where they fell short, and how the proposed mechanism holds up.

The core finding: semaglutide reduces drinking in randomized trials

The strongest data come from a phase 2 randomized trial of 48 non-treatment-seeking adults with alcohol use disorder. Half received once-weekly semaglutide, titrated from 0.25mg to 1.0mg over nine weeks; half received placebo. In a laboratory alcohol self-administration procedure, the semaglutide group drank less. Grams of alcohol consumed fell significantly (β = −0.48; 95% CI, −0.85 to −0.11; p = .01), and peak breath alcohol concentration dropped too (β = −0.46; 95% CI, −0.87 to −0.06; p = .03).

Craving was tracked weekly in the same trial. It fell significantly with semaglutide versus placebo (β = −0.39; 95% CI, −0.73 to −0.06; p = .01), and so did drinks per drinking day (β = −0.41; 95% CI, −0.73 to −0.09; p = .04). One measure didn’t hold up: drinks per calendar day did not reach significance (β = −0.27; p = .17). People drank less per session. The trial didn’t show they drank on fewer days.

The largest trial: 108 people, 26 weeks, and a caveat about who was studied

The biggest and most positive result comes from a randomized, placebo-controlled trial at Mental Health Center Copenhagen. Researchers enrolled 108 treatment-seeking adults, ages 18 to 70, with a BMI of 30 or higher and an AUDIT score above 15, which indicates likely alcohol dependence. Everyone received cognitive behavioral therapy. Half also got once-weekly subcutaneous semaglutide at 2.4mg, the Wegovy dose, for 26 weeks.

Heavy drinking days dropped by a mean of 41.1 percentage points in the semaglutide group (95% CI, −48.7 to −33.5), against 26.4 percentage points with placebo (95% CI, −34.1 to −18.6). That’s a 13.7 percentage-point advantage for the drug. The semaglutide group also lost more weight (−11.2 kg vs. −2.2 kg) and improved on total alcohol consumption, drinks per drinking day, self-reported craving, and biomarkers of alcohol exposure.

This is the clearest positive signal in the literature so far. It comes with a caveat: every participant had a BMI of 30 or higher. The trial can’t tell you whether semaglutide reduces cravings in someone who drinks heavily but isn’t overweight, because nobody in the study fit that profile.

A messier result: oral semaglutide didn’t reduce craving in the lab

Not every semaglutide trial lines up neatly. A separate phase 2 trial gave 50 treatment-seeking adults with moderate-to-severe alcohol use disorder oral semaglutide (3mg/day for four weeks, then 7mg/day for four weeks) or placebo over eight weeks.

The drug did not significantly reduce laboratory-assessed craving or drinks per day. It did significantly reduce heavy drinking days, drinks per drinking day, and craving reported in real-world, day-to-day settings. An internally mixed result, and note which measure failed: lab-assessed craving is the same type of measure that succeeded in the injectable trial above. The formulations, dosing schedules, and populations all differ, which makes this hard to reconcile cleanly.

Not every GLP-1 drug works the same way

The class-wide picture weakens once you look past semaglutide. A 127-person, 26-week randomized, placebo-controlled trial tested exenatide, a related but distinct GLP-1 drug, in 62 exenatide and 65 placebo patients. Overall, it did not significantly reduce heavy drinking days. Both groups improved at similar rates, with no meaningful difference between them.

There was one exception. In a prespecified subgroup of participants with BMI over 30 (n=30), exenatide reduced heavy drinking days by 23.6 percentage points versus placebo (95% CI, −44.4 to −2.7; p = 0.034). Once again the benefit lands in people who also carry excess weight, not in the trial population as a whole.

Add it up. Semaglutide has three trials with positive signals of some kind. Exenatide has one large trial that was null overall. A 2026 clinical review states the issue directly: “there may be differences between GLP-1 agonists,” and it warns against assuming the whole drug class behaves like semaglutide.

The brain-imaging evidence: this looks like a reward-circuit effect, not sedation

The exenatide trial included a neuroimaging substudy, the closest thing to direct human evidence for how these drugs might work. Researchers used fMRI to measure brain response to alcohol cues and SPECT imaging to measure dopamine transporter availability.

At week 26, alcohol cue-reactivity in the ventral striatum, a region central to drug and alcohol reward, was significantly lower in the exenatide group than in placebo (mean difference = −0.176, SEM = 0.075, p = 0.025). Dopamine transporter availability was also significantly lower in the exenatide group across the striatum (p = 0.044), caudate (p = 0.014), and putamen (p = 0.024).

That pattern fits the proposed mechanism. A systematic review of preclinical research found that GLP-1 receptor agonists like exendin-4 and liraglutide blunt alcohol-induced dopamine release in the nucleus accumbens and ventral tegmental area, the brain’s mesolimbic reward circuit. Injecting exenatide directly into those regions reduced alcohol self-administration in rodents in a dose-dependent way.

The effect also looks anatomically specific rather than like a general sedative or appetite-suppressing side effect. In the same review, GLP-1 agonist infusion into other dopamine-relevant regions, including the dorsal striatum, basolateral amygdala, and hypothalamic nuclei, did not consistently reduce alcohol intake in the same animal models. It varied by sex, with stronger effects in male rodents, and by drinking phenotype, working in high-alcohol-drinking rats but not low-alcohol-drinking ones.

Keep the limits in view. This is largely animal data plus one human imaging substudy, and the imaging subset was smaller than the full trial. No study has used GLP-1 receptor blockade or genetic knockout in humans to confirm the mechanism actually drives the behavioral effect rather than just correlating with it.

What’s still missing

A few gaps cap how far this evidence stretches right now:

No approval, and none imminent without more trials. A 2026 clinical review is blunt about the state of the evidence: the one large exenatide trial was null overall, and retrospective studies of semaglutide and liraglutide are “promising associations” that carry real risk of selection bias and unmeasured confounding. The authors conclude that “additional clinical trials are needed… to open the pathway for FDA approval and payer coverage.” No GLP-1 drug is FDA-approved for alcohol use disorder today.

Durability is unknown. The longest randomized trial ran 26 weeks. Alcohol use disorder is chronic, and 26 weeks is a short window for judging whether an effect holds up. The preclinical review adds a specific worry: in rodents, GLP-1 agonist suppression of alcohol-related behavior can fade with repeated dosing, likely through receptor desensitization or compensatory changes in the brain. Some semaglutide effects in animals didn’t persist beyond about two days post-injection. Whether that pattern appears in humans on chronic weekly dosing is an open question.

Obesity may be doing some of the work. The largest and most positive semaglutide trial required a BMI of 30 or higher. So did the one subgroup where exenatide worked. It’s not established whether these drugs reduce alcohol craving in people who drink heavily but don’t carry excess weight.

Frequently Asked Questions

Is semaglutide (Ozempic/Wegovy) FDA-approved to treat alcohol cravings or alcohol use disorder? No. Semaglutide’s only FDA-approved uses are type 2 diabetes (Ozempic, Rybelsus) and chronic weight management (Wegovy). Any use for alcohol cravings or alcohol use disorder is off-label. A 2026 review explicitly calls for more trials “to open the pathway for FDA approval,” which confirms that pathway doesn’t exist yet.

How exactly does semaglutide reduce the urge to drink, biologically? The leading explanation involves the brain’s dopamine reward circuit. In animal studies, GLP-1 receptor activation in the nucleus accumbens and ventral tegmental area blunts the dopamine surge that alcohol normally triggers. In humans, an exenatide neuroimaging substudy found lower fMRI reactivity to alcohol cues in the ventral striatum and reduced dopamine transporter availability after 26 weeks of treatment. This looks like reward-blunting specific to alcohol-related brain circuitry, not general sedation or appetite suppression.

Does this work for everyone, or mainly people who are also overweight? The evidence is strongest in people with both alcohol use disorder and obesity. The 108-person Copenhagen trial, the largest and most positive study to date, enrolled only participants with a BMI of 30 or higher. The exenatide trial’s overall result was null, and only its BMI-over-30 subgroup showed a significant benefit. Whether GLP-1 drugs reduce cravings in normal-weight people with alcohol use disorder is unproven.

Which GLP-1 drug has the best evidence for alcohol cravings: semaglutide or others like exenatide? Semaglutide has the more consistent record, with positive signals across three separate randomized trials. Exenatide’s one large randomized trial was null overall, with benefit confined to a subgroup with obesity. A 2026 review notes there may be real differences between individual GLP-1 drugs, so results from semaglutide shouldn’t be assumed to apply to the whole class.

Are the craving reductions large and durable, or modest and short-term? Modest and short-term, based on what’s been measured. The statistically significant effects reported so far are real but not dramatic; the Copenhagen trial’s 13.7 percentage-point between-group difference in heavy drinking days is the largest effect size reported. Trial durations range from 8 to 26 weeks, which is short relative to a chronic condition. Animal data raise a specific concern about tolerance or desensitization with repeated dosing, and that hasn’t been ruled out in humans.

The bottom line

Semaglutide has a real, statistically significant effect on alcohol craving and heavy drinking in multiple randomized trials, and a plausible brain mechanism backs it up. That’s more than can be said for most off-label uses circulating online. But the effect isn’t uniform across the GLP-1 class, it hasn’t been confirmed in people without obesity, the longest trial ran 26 weeks, and no regulatory body has approved any GLP-1 drug for this purpose. If you’re on semaglutide for diabetes or weight loss and notice reduced interest in drinking, that lines up with the research. It is not, on current evidence, a reason to use these drugs specifically to treat alcohol use disorder outside of a clinical trial or your doctor’s guidance.

If you’re weighing GLP-1 therapy for weight management and cost is a factor, see our breakdowns of semaglutide cost per month and GLP-1 options without insurance. For background on how compounded versions differ from brand-name drugs like Wegovy, see our guide to compounded semaglutide safety and our comparison of Wegovy vs. compounded semaglutide.

This article is for informational purposes only and is not medical advice. Talk to your healthcare provider before starting, stopping, or changing any medication.