TL;DR: Two randomized, placebo-controlled human trials now show once-weekly semaglutide reduces alcohol craving and heavy drinking, alongside rodent work mapping the mechanism to GLP-1 receptors in the mesolimbic reward pathway. The effect is real and reproducible across two independent trials, but still early: one ran 48 people for 9 weeks, the other 108 people for 26 weeks, and semaglutide is not approved anywhere for alcohol use disorder.
Why anyone is even asking this question
The “semaglutide alcohol cravings” search came from patient reports before it came from a lab: people started semaglutide or tirzepatide for weight loss or diabetes and noticed, unprompted, that they stopped wanting to drink. Not white-knuckling sobriety, just the pull toward a glass of wine at 6 p.m. going quiet.
That pattern showed up formally in a 2023 mixed-methods study combining machine-learning analysis of roughly 68,250 alcohol-related Reddit posts with a remote observational cohort of 153 current drinkers with a BMI of 30 or higher (Quddos et al., Scientific Reports, 2023). The social-media arm found most alcohol-related posts from GLP-1 users described craving reduction or decreased desire to drink. The observational arm found lower self-reported alcohol intake, fewer drinks per drinking episode, and lower AUDIT scores in the semaglutide and tirzepatide groups relative to controls. That cannot prove causation, but it is exactly the signal that justifies a randomized trial. Two groups have now run one.
The first randomized trial: Hendershot et al., 2025
The first placebo-controlled test of semaglutide specifically for alcohol use disorder came from Hendershot and colleagues, published in JAMA Psychiatry on February 12, 2025 (Vol 82, Issue 4, pages 395-405). It was a phase 2 trial in 48 non-treatment-seeking adults with alcohol use disorder, randomized to once-weekly semaglutide or placebo over a 9-week outpatient titration: 0.25 mg for weeks 1 through 4, 0.5 mg for weeks 5 through 8, then a flexible step to 1.0 mg in week 9.
The primary outcome was laboratory alcohol self-administration, a controlled paradigm where participants can choose to drink or not under observation. Among the 25 participants with complete data, semaglutide was associated with significantly lower grams of alcohol consumed and lower peak breath alcohol concentration relative to placebo.
The secondary and exploratory outcomes told a coherent story. Drinks per drinking day fell significantly on semaglutide, heavy drinking days declined significantly faster over time relative to placebo, and scores on the Penn Alcohol Craving Scale dropped significantly more in the semaglutide arm. Two measures did not reach significance in this small sample: drinks per calendar day and the ratio of drinking to abstinent days. That mixed picture is normal for a 48-person phase 2 trial and worth stating plainly rather than smoothing over. Weight fell about 5% on semaglutide versus roughly flat on placebo, and the trial recorded no serious adverse events and no treatment-related discontinuations.
The second, larger trial: Klausen et al., Lancet
The follow-on question was obvious: does the signal hold in a larger sample, over a longer period, in people who also have obesity? Klausen and colleagues at Mental Health Center Copenhagen answered it in a randomized, double-blind, placebo-controlled trial published in The Lancet (Vol 407, Issue 10540, pages 1687-1698). This is the GLP-1 alcohol use disorder study many clinicians have been waiting on, and its results are now published, not pending.
The trial enrolled 108 adults aged 18 to 70 with a BMI of 30 kg/m squared or higher who reported heavy drinking on at least 6 days per month, randomized 54 to semaglutide 2.4 mg weekly and 54 to placebo, both arms also receiving cognitive behavioral therapy, over 26 weeks.
Klausen et al., The Lancet, 2026. Both groups improved; semaglutide improved significantly more.
The estimated treatment difference was 13.7 percentage points in favor of semaglutide (95% CI, 5.4 to 22.0; P = .0015), hard to write off as noise at this sample size and duration. The trial also reported semaglutide out-performing placebo on total alcohol consumed over 30 days and on drinks per drinking day, alongside larger improvements in liver biomarkers including phosphatidylethanol, gamma-glutamyl transferase, and mean cell volume. Craving scores also improved more on semaglutide, though we could not confirm the exact point estimate for that measure and are not stating a number for it here. Adverse events were transient, mostly mild-to-moderate gastrointestinal complaints, occurring more often in the semaglutide group.
Does tirzepatide reduce drinking too?
The tirzepatide evidence is a step behind semaglutide’s, and it is important not to blur the two. The same 2023 Quddos et al. observational study included tirzepatide users alongside semaglutide users and found a similar pattern of lower self-reported intake and lower AUDIT scores, but that is observational, not randomized.
On the mechanistic side, a rodent study published in eBioMedicine in early 2026 tested tirzepatide directly in an alcohol self-administration and relapse model. Voluntary alcohol consumption fell by more than half in tirzepatide-treated animals relative to controls, and after forced abstinence, treated animals did not show the rebound drinking spike typical of this relapse paradigm. That is a strong preclinical result, but it is rodents, not people. A human trial of tirzepatide in alcohol use disorder, in patients with comorbid schizophrenia, is registered and recruiting (ClinicalTrials.gov NCT06939088). Until that or a similar trial reports, “does tirzepatide reduce drinking” has a promising rodent answer and an unconfirmed human one.
The mechanism: why a diabetes and weight-loss hormone touches drinking at all
GLP-1 receptors are not confined to the pancreas and gut. They sit throughout the mesolimbic reward pathway, the dopaminergic circuit that alcohol, drugs of abuse, and palatable food all engage, including the ventral tegmental area (VTA), where dopamine neurons projecting to the nucleus accumbens originate, and the nucleus accumbens itself, which translates that dopamine signal into wanting and reward.
Rodent microinjection studies have mapped this at the level of specific brain nuclei, not just the whole animal. Vallöf and colleagues showed that infusing exendin-4 directly into the nucleus accumbens shell blocked alcohol-induced locomotor stimulation and conditioned place preference, and separately reduced intake in rats drinking heavily over 12 weeks. Allingbjerg et al. (Experimental and Clinical Psychopharmacology, 2022) found exendin-4 infused into the nucleus accumbens, ventral hippocampus, or lateral septum reduced alcohol self-administration comparable to a full-body injection, while infusion into the dorsal striatum, which has little GLP-1 receptor expression, did almost nothing. That specificity argues the effect works through the receptor, not as a side effect of nausea or reduced appetite.
The working model is that GLP-1 receptor activation in these regions dampens dopaminergic signaling from the VTA to the nucleus accumbens, turning down the reward value assigned to alcohol, the same way it has been shown to blunt the reward value of palatable food and, in earlier rodent work, psychostimulants and nicotine.
How the evidence stacks up, honestly
What tier of evidence each claim sits on is the actual substance of the story, not a caveat tacked onto the end of it.
| Claim | Evidence tier | What it means |
|---|---|---|
| Semaglutide reduces heavy drinking days and craving | Two independent RCTs (n=48 and n=108) | Strongest tier available for a repurposed indication; still phase 2 to phase 3 scale |
| Semaglutide reduces lab-measured alcohol self-administration | Single RCT, small completer sample (n=25) | Statistically significant, but a small sample within a small trial |
| Semaglutide/tirzepatide reduce self-reported drinking in obesity | Observational cohort and social-media analysis (Quddos et al.) | Hypothesis-generating, not causal; self-selected sample |
| Tirzepatide reduces drinking and relapse | Rodent study only (eBioMedicine, 2026) | Strong preclinical signal; human RCT not yet reported |
| Mechanism via VTA/nucleus accumbens GLP-1 receptors | Rodent microinjection studies | Well-mapped in animals; human confirmation is thinner |
What this means if you are already on a GLP-1
If you already take semaglutide or tirzepatide and notice your desire to drink has faded, the trial data suggests you are not imagining a placebo effect. Two separate randomized trials found the same direction of effect using objective, validated measures, not just self-report. For how the drug reaches a steady effect, our guide to semaglutide dosing and titration covers the same escalation schedule used in the Hendershot trial. For the basics of the compound, see what is semaglutide, and for the gastrointestinal tradeoffs, our GLP-1 side effects management guide covers what to expect.
For molecules compared head to head, see the semaglutide versus tirzepatide comparison, and for where retatrutide and other incretin-pathway compounds sit relative to this same reward-circuit question, see next-generation weight-loss peptides for 2026.
Related reading
Frequently asked questions
Does semaglutide reduce alcohol cravings?
Yes, in two independent randomized, placebo-controlled trials. A 2025 phase 2 trial found semaglutide significantly reduced Penn Alcohol Craving Scale scores versus placebo over 9 weeks, and a 2026 Lancet trial in 108 people found significantly greater reductions in heavy drinking days over 26 weeks. Both effects were statistically significant, though both trials are still small to moderate in size relative to the large multi-site trials used to approve a drug for a new indication.
Does tirzepatide reduce drinking the same way semaglutide does?
The evidence for tirzepatide is earlier-stage. A 2026 rodent study found tirzepatide cut voluntary alcohol consumption by more than half and blocked relapse-like drinking after abstinence, and an observational human study found lower self-reported drinking in tirzepatide users. No randomized human alcohol-use-disorder trial for tirzepatide has reported results yet, though one is registered and recruiting.
Is Ozempic or Wegovy approved for alcohol use disorder?
No. Semaglutide is approved for type 2 diabetes (as Ozempic and Rybelsus) and for chronic weight management (as Wegovy). Alcohol use disorder is not an approved indication for semaglutide, tirzepatide, or any other GLP-1 or dual-agonist medication anywhere in the world as of this writing.
What is the mechanism behind GLP-1 drugs and reduced drinking?
GLP-1 receptors sit in the mesolimbic reward pathway, including the ventral tegmental area and nucleus accumbens, the circuit that assigns reward value to alcohol, food, and other reinforcing substances. Rodent studies show that activating GLP-1 receptors directly in these regions reduces alcohol-seeking behavior and blunts the dopamine signaling that normally reinforces drinking, and this appears specific to GLP-1-receptor-rich regions rather than a general side effect of the drug.
How strong is the evidence, really?
Strong enough to justify the larger trials now underway, not strong enough to call this settled. The two human RCTs total 156 participants across 9 and 26 weeks, a real and replicated signal, but a fraction of the patient-years behind semaglutide’s diabetes and obesity approvals. Treat it as an active, increasingly well-replicated research finding, not an approved treatment claim.