Women's Health

Alcohol on a GLP-1: what the labels say, what the trials found, and what is still anecdote

By the US Health Digest editorial team · Published August 30, 2026 · Every claim linked to its primary source

We searched the full text of twelve current US labels for the string "alcohol". On every FDA-approved GLP-1 used for weight management — Wegovy, Zepbound, Saxenda, Foundayo — and on Ozempic, Mounjaro, Rybelsus and Trulicity, every hit is an alcohol swab, an inactive ingredient, or the retired name for fatty liver disease. Not one carries an alcohol-interaction warning or any drinking advice. Of these twelve, the only labels that tell patients anything about drinking are Soliqua and Xultophy — and both also contain insulin.

This page reports what documents and studies say. It is not advice to drink, not to drink, or to change any medicine.

The label search, product by product

Silence is easy to assert and easy to get wrong, so this was a text search over the complete Structured Product Label — Highlights, Warnings and Precautions, Drug Interactions, Adverse Reactions, Patient Counseling Information, the Medication Guide and the Instructions for Use — not a skim of the warnings section. Every hit was read in context and classified.

An absent warning is not a safety endorsement, and it is not a hidden one either. It means the interaction of these drugs with alcohol was not part of what FDA reviewed for these approvals, so there is no agreed label sentence about it in either direction. The labels are dense with things the agency did review — thyroid C-cell tumours, pancreatitis, gallbladder disease, kidney injury, retinopathy, aspiration under anaesthesia. Drinking is not among them.

The two labels that do name drinking — and why

Soliqua and Xultophy are the exception, and the reason is the other molecule in the vial. Both are fixed-ratio combinations of a basal insulin with a GLP-1 receptor agonist, and both carry alcohol in the row of the section 7.1 interactions table headed "Drugs That May Increase or Decrease the Blood Glucose Lowering Effect of SOLIQUA 100/33" (Xultophy's row carries its own brand name): "Alcohol, beta-blockers, clonidine, and lithium salts." The listed intervention is "Dose adjustment and increased frequency of glucose monitoring may be required when SOLIQUA 100/33 is coadministered with these drugs." Both Medication Guides go further. Under the heading "What should I avoid while taking SOLIQUA 100/33?" the label states: "While taking SOLIQUA 100/33 do not: drive or operate heavy machinery, until you know how SOLIQUA 100/33 affects you. drink alcohol or use prescription or over-the-counter medicines that contain alcohol." Xultophy's is word-for-word the same with its own brand name substituted.

The pattern in one sentence: where insulin enters the product, alcohol enters the label. That is not our inference. The stand-alone insulin glargine label, Lantus, carries the identical interactions row and the identical Medication Guide sentence with no GLP-1 in the product at all. The alcohol language on Soliqua and Xultophy is insulin labelling, and it travels with the insulin. Two caveats. Neither is prescribed for weight management; both are type 2 diabetes medicines. And Xultophy's SPL lists a marketing end date of May 31, 2021 for its national drug code, so we cite it as a labelling document rather than a currently supplied product.

We also widened the search beyond the twelve. DailyMed currently serves 51 US labels for products containing a GLP-1 receptor agonist once generics, authorised generics and repackager listings are counted, and we searched every one of them for the word alcohol on August 30, 2026. Across all 51, once you set aside swabs and wipes, excipients, fatty-liver nomenclature and the section 6.1 note that some patients who developed pancreatitis had a history of cholelithiasis or alcohol abuse — that note recurs on Victoza and on most of the generic liraglutide labels — exactly three documents are left: Soliqua, Xultophy, and one further exception worth stating plainly. A repackager's Byetta listing — SPL version 3, published March 14, 2012, and still served as an active record — asks patients, "Before taking BYETTA, tell your healthcare provider if you have had: pancreatitis. stones in your gallbladder (gallstones). a history of alcoholism. high blood triglyceride levels." That is a disclosure prompt in a pancreatitis-risk list rather than drinking advice, it sits in a fourteen-year-old document, and the current AstraZeneca Byetta label no longer contains it. Outside the two insulin combinations, it is the only patient-facing mention of drinking we found across the 51 — and it asks for a medical history rather than giving guidance.

The randomised evidence on drinking

Randomised, and positive: the 2026 Lancet trial

The largest randomised test to date is a 26-week, single-centre, double-blind, placebo-controlled trial in Copenhagen, published in The Lancet in May 2026. It enrolled 108 treatment-seeking participants (53 women, 55 men) with moderate to severe alcohol use disorder and comorbid obesity, randomised 1:1 to once-weekly semaglutide 2.4 mg subcutaneously or saline placebo, with both arms also receiving standard cognitive behavioural therapy. The primary endpoint was the reduction in number of heavy drinking days at 26 weeks, analysed by ANCOVA on the intention-to-treat population with multiple imputation for missing outcomes; 88 of 108 (81%) completed. Heavy drinking days fell by 41.1 percentage points from baseline on semaglutide (95% CI −48.7 to −33.5) versus 26.4 points on placebo (−34.1 to −18.6): estimated treatment difference −13.7 percentage points (95% CI −22.0 to −5.4; p=0.0015). Note what the placebo arm did — it improved substantially too, which is what CBT plus trial participation looks like, and why the between-group difference is the number that counts. The dose was the same 2.4 mg used for weight management; the population was people who had sought treatment for a drinking problem and also had obesity.

Randomised, and smaller: the 2025 JAMA Psychiatry trial

A phase 2, double-blind, parallel-arm trial at a US academic medical centre randomised 48 non-treatment-seeking adults with alcohol use disorder (34, or 71%, female; mean age 39.9) to nine weeks of low-dose semaglutide or placebo. Its primary outcome was not real-world drinking but laboratory alcohol self-administration, measured before and after treatment at the 0.5 mg weekly dose: grams of alcohol consumed fell (β −0.48; 95% CI −0.85 to −0.11; P = .01), as did peak breath alcohol concentration (β −0.46; −0.87 to −0.06; P = .03). Among secondary and exploratory outcomes, semaglutide did not affect average drinks per calendar day or the number of drinking days, but did reduce drinks per drinking day and weekly craving. The authors describe their own result as "initial prospective evidence" justifying larger trials — not as an established treatment effect.

Randomised, and negative on its primary endpoint

The first trial in this line went the other way. A randomised, double-blind, placebo-controlled trial of exenatide 2 mg weekly for 26 weeks plus cognitive behavioural therapy, in 127 treatment-seeking patients, took the same primary outcome as the Lancet trial — reduction in heavy drinking days — and did not meet it. It did find attenuated alcohol cue reactivity on fMRI in the ventral striatum and septal area, and an exploratory subgroup analysis found reductions in heavy drinking days and total alcohol intake in participants with a BMI above 30. Exploratory subgroups generate hypotheses; they do not establish effects.

Everything below that line

A Swedish national register study compared each person's own periods on and off a drug across 227,866 people with alcohol use disorder (36.5% female, 2006–2023). Semaglutide (4,321 users) carried the lowest risk of hospitalisation for alcohol use disorder (adjusted hazard ratio 0.64; 95% CI 0.50–0.83), liraglutide (2,509 users) the second lowest (0.72; 0.57–0.92). That design removes fixed confounders like genetics but not time-varying ones, and cannot show causation. One rung down, a 2023 study in Scientific Reports paired a machine-learning analysis of roughly 68,250 Reddit posts — of 1,580 alcohol-related posts, 71% were classified as craving reduction, decreased desire to drink or other negative effects — with a remote study of 153 self-reported semaglutide or tirzepatide users with a BMI of 30 or above. That is self-report and social-media text: useful for generating hypotheses, useless for estimating an effect size. A 2025 systematic review and meta-analysis pooled 14 studies (4 randomised, 10 observational) and reported a mean AUDIT score reduction of 7.81 points (95% CI −9.02 to −6.60) — but with I² of 87.5%, meaning the included studies disagree severely, and with the two designs pooled together.

The mechanisms that are actually documented

Four things on these labels would matter to somebody who drinks. All four come from the label text, not from us reasoning about pharmacology.

• Hypoglycaemia, and specifically with insulin or a sulfonylurea. Wegovy's section 5.4 states: "In glycemic control clinical trials, the risk of hypoglycemia was increased when semaglutide injection or tablet was used concomitantly with insulin or an insulin secretagogue (e.g., sulfonylurea)." Zepbound, Saxenda and Foundayo carry the same framing; the Ozempic, Rybelsus and Mounjaro sections are titled "Hypoglycemia with Concomitant Use of Insulin Secretagogues or Insulin". Alcohol is named in that context on exactly the two labels that contain insulin.
• Delayed gastric emptying. Every one of these labels states it plainly — "WEGOVY delays gastric emptying" — in the context of a warning about pulmonary aspiration during general anaesthesia or deep sedation, which candidly adds that "Available data are insufficient to inform recommendations to mitigate the risk". One small human study has tested whether that changes how alcohol is absorbed. In a pilot study of 20 adults with obesity published in Scientific Reports in October 2025 — half taking a GLP-1 receptor agonist, half not — everyone drank a fixed challenge dose, and the GLP-1 group showed a delayed rise in breath alcohol concentration and delayed subjective effects, which was not explained by nausea. The authors present it as preliminary evidence of a peripheral mechanism. Twenty people is a hypothesis rather than a number anyone can apply to themselves, and a dedicated pharmacokinetic study in people taking tirzepatide for weight is registered but not yet enrolling. That measurement is a different thing from the breath-alcohol result in the JAMA Psychiatry trial, which reflects participants choosing to drink less.
• Pancreatitis. Every label carries an acute pancreatitis warning — Wegovy's opens "Acute pancreatitis, including fatal and non-fatal hemorrhagic or necrotizing pancreatitis, has been observed in patients treated with GLP-1 receptor agonists, including WEGOVY". Alcohol is named in none of those warnings. It appears only in the Victoza, Soliqua and Xultophy adverse-reaction sections, describing pre-existing risk factors carried by some patients who developed pancreatitis in the trials.
• Dehydration. The kidney-injury warnings on Zepbound, Saxenda, Foundayo and the semaglutide labels all attribute the reported cases to one route: "The majority of the reported events occurred in patients who experienced gastrointestinal adverse reactions leading to dehydration such as nausea, vomiting, or diarrhea". See our guides to eating on a GLP-1 and side effects in women.

One further instruction catches alcohol without naming it. Both oral semaglutide labels tell patients to take the tablet on an empty stomach with up to 4 ounces of water, and add: "Do not take RYBELSUS or OZEMPIC tablets with other liquids besides water" and "After taking RYBELSUS or OZEMPIC tablets, wait at least 30 minutes before eating food, drinking beverages or taking other oral medications". The Wegovy tablet carries the same two sentences. A drink is a beverage; the instruction concerns absorption of the drug rather than alcohol, but on the tablets it is the only place the timing of any drink is addressed.

Where this lands for women

Here the ground has shifted, and much of what circulates online is out of date. The Dietary Guidelines for Americans, 2025–2030, released January 7, 2026, dropped the numeric, sex-specific limits earlier editions carried. Its alcohol section now reads, in full: "Consume less alcohol for better overall health." and "People who should completely avoid alcohol include pregnant women, people who are recovering from alcohol use disorder or are unable to control the amount they drink, and people taking medications or with medical conditions that can interact with alcohol. For those with a family history of alcoholism, be mindful of alcohol consumption and associated addictive behaviors." There is no "one drink a day for women" in the current edition, and the document does not say which medications it means.

Sex-specific numbers survive at NIAAA, but as thresholds for risk patterns rather than as a permitted allowance. On its drinking patterns page, binge drinking is the pattern bringing blood alcohol concentration to 0.08% or higher, which "For a typical adult… corresponds to consuming five or more drinks (male), or four or more drinks (female), in about two hours"; heavy drinking is defined "For women, consuming four or more on any day or eight or more drinks per week"; and a standard drink is "any beverage containing 0.6 fl oz or 14 grams of pure alcohol". The phrase "moderate drinking" does not appear on that page at all. NIAAA's clinician resource now tells prescribers to "Advise some patients not to drink at all, including those who are managing health conditions that can be worsened by alcohol… are taking medications that could interact with alcohol". Whether a GLP-1 is such a medication is exactly the question the labels leave open.

What is still anecdote

The reports that started all of this remain reports. A 2024 paper in the Journal of Studies on Alcohol and Drugs characterises the phenomenon carefully: the rise of these drugs has been accompanied by "widespread reports of unintended reductions in alcohol use (and other addictive behaviors) during treatment", and those accounts are "now a primary reason for interest" in the question. Volume is not evidence. Nobody posts about the weeks their drinking did not change, nobody knows the denominator, and people starting a weight-management drug are usually changing several things at once.

Turning that into a finding requires trials in the population actually asking the question — people prescribed these drugs for weight, not people recruited for alcohol use disorder — powered for drinking outcomes measured prospectively rather than recalled. Several trials are running, though none yet in that population. The largest is the VA's CSP #2041 "CRAVE" trial, a phase 3, double-blind, placebo-controlled study of semaglutide in US veterans with an estimated 622 participants, which began enrolling in July 2026 with primary completion scheduled for March 2029; its primary outcome is a two-level reduction in WHO risk drinking level over the last 28 days of treatment. NIDA's phase 2 STAR trial (63 participants) is active and not recruiting, with primary completion listed for March 2027. Until those results land, the honest description of the reduced-drinking effect in a general weight-management population is: plausible, supported by randomised trials in a different population, and unquantified for anyone else.

Questions for your prescriber

• Which molecule am I actually on, and does its current label say anything at all about alcohol? (On the weight-management products, as of today, it does not.)
• Am I also taking insulin or a sulfonylurea? If so, what does my label say about combined hypoglycaemia risk, and does that change anything about how I should be monitored?
• If I am on tablets rather than an injection, how does the water-only, wait-30-minutes dosing instruction fit around when I eat and drink?
• Do I have any of the pancreatitis or liver risk factors that would make this a bigger conversation than the label alone suggests?
• I have noticed my desire to drink has changed since starting. Is that worth recording in my notes, and does it change anything you want to monitor?
• What symptoms should make me contact you rather than wait until the next check-in?

Common questions

Does the Ozempic or Wegovy label say you cannot drink alcohol?

No. As of August 30, 2026, a full-text search of the current Wegovy label returns 11 occurrences of the word alcohol, all of which are alcohol swabs in the injection instructions or the retired name for fatty liver disease in the MASH indication. The Ozempic injection label returns 8, all alcohol swabs. The Rybelsus and Ozempic tablets label returns zero. None of the three contains an alcohol-interaction warning, an alcohol entry in Warnings and Precautions, or any drinking advice. That is a statement about what these documents contain, not a statement that drinking is safe for any individual.

Is the reduced desire to drink real, or is it just what people say online?

Both things are true at once. The online reports are extremely widespread and are not evidence. Separately, three randomised placebo-controlled trials have now tested GLP-1 receptor agonists in people with alcohol use disorder. A 2026 Lancet trial of semaglutide 2.4 mg in 108 people with alcohol use disorder and obesity met its primary endpoint, with heavy drinking days falling 13.7 percentage points more than on placebo. A 2025 JAMA Psychiatry trial in 48 people found reduced laboratory alcohol self-administration and reduced craving on low-dose semaglutide. An earlier exenatide trial in 127 people did not meet its primary endpoint. None of those trials enrolled people taking these drugs purely for weight management, so none of them measures what happens in that group.

Do the US drinking guidelines still say one drink a day for women?

Not in the current edition. The Dietary Guidelines for Americans, 2025–2030, released January 7, 2026, removed the numeric daily limits and now advises consuming less alcohol for better overall health, and names people taking medications that can interact with alcohol among those who should avoid it entirely. NIAAA still uses sex-specific thresholds to define binge and heavy drinking for women — four or more drinks in about two hours, and four or more on any day or eight or more per week — but those are definitions of risk patterns, not an allowance.

What this page does not claim

It does not say drinking on a GLP-1 is safe, or unsafe, for any individual. It does not tell anyone to start, stop or change a medication or a drinking pattern. It does not claim these drugs treat alcohol use disorder — no US label carries that indication, and the trials above are phase 2 and phase 3 research, not approvals. And it does not extend to compounded copies, whose contents are covered by none of these labels. Related reading: GLP-1s, mood and antidepressants, what happens in the telehealth medical review, GLP-1s in menopause, and the GLP-1 FAQ for women.

Sources

1. DailyMed. WEGOVY (semaglutide) injection and tablets — current prescribing information. National Library of Medicine; SPL version 19, published June 30, 2026 (document effective June 18, 2026). Sections 2.1, 5.2, 5.4, 5.5, 5.10 and the Instructions for Use.
2. DailyMed. ZEPBOUND (tirzepatide) injection. SPL version 38, published May 6, 2026. Sections 5.3, 5.5, 5.7, 5.9, 11 and the Instructions for Use.
3. DailyMed. MOUNJARO (tirzepatide) injection. SPL version 38, published May 6, 2026.
4. DailyMed. OZEMPIC (semaglutide) injection. SPL version 20, published June 10, 2026.
5. DailyMed. RYBELSUS and OZEMPIC tablets (semaglutide). SPL version 14, published August 19, 2026. Section 2.1, Important Administration Instructions.
6. DailyMed. SAXENDA (liraglutide) injection. SPL version 22, published June 15, 2026. Sections 5.2, 5.4, 5.6, 5.9.
7. DailyMed. VICTOZA (liraglutide) injection. SPL version 31, published November 17, 2025. Section 6.1, pancreatitis cases in glycemic control trials.
8. DailyMed. FOUNDAYO (orforglipron) tablets. SPL version 10, published August 13, 2026. Sections 2.1, 5.2, 5.4, 5.5, 5.9, 11.
9. DailyMed. TRULICITY (dulaglutide) injection. SPL version 62, published August 10, 2026.
10. DailyMed. BYETTA (exenatide) injection. SPL version 26, published September 8, 2025.
11. DailyMed. BYETTA (exenatide) injection [Physicians Total Care, Inc.]. Repackager listing, SPL version 3, published March 14, 2012; document effective March 12, 2012. Medication Guide and Patient Counseling Information, alcoholism among the pancreatitis risk factors to disclose. Cited as the one non-insulin exception found in the 51-label sweep; it is not the current Byetta label.
12. DailyMed. SOLIQUA 100/33 (insulin glargine and lixisenatide) injection. SPL version 25, published March 20, 2026. Section 7 Drug Interactions table and the Medication Guide, "What should I avoid while taking SOLIQUA 100/33?"
13. DailyMed. XULTOPHY 100/3.6 (insulin degludec and liraglutide) injection. SPL version 14, published November 17, 2025. Same two locations. The SPL's marketing act for its listed national drug code records an end date of May 31, 2021.
14. DailyMed. LANTUS (insulin glargine) injection. Cited to show that the alcohol row in the Drug Interactions table and the Medication Guide sentence are standard basal-insulin labelling, present with no GLP-1 in the product.
15. Klausen MK, Justesen SK, Pedersen JN, et al. Once-weekly semaglutide versus placebo in patients with alcohol use disorder and comorbid obesity: a randomised, double-blind, placebo-controlled trial. Lancet. 2026;407(10540):1687–1698. PMID 42070571. Registered as NCT05895643.
16. Hendershot CS, Bremmer MP, Paladino MB, et al. Once-Weekly Semaglutide in Adults With Alcohol Use Disorder: A Randomized Clinical Trial. JAMA Psychiatry. 2025;82(4):395–405. PMID 39937469; full text at PMC11822619. Registered as NCT05520775.
17. Klausen MK, Jensen ME, Møller M, et al. Exenatide once weekly for alcohol use disorder investigated in a randomized, placebo-controlled clinical trial. JCI Insight. 2022;7(19):e159863. PMID 36066977.
18. Lähteenvuo M, Tiihonen J, Solismaa A, et al. Repurposing Semaglutide and Liraglutide for Alcohol Use Disorder. JAMA Psychiatry. 2025;82(1):94–98. PMID 39535805. Observational within-individual cohort, not a trial.
19. Quddos F, Hubshman Z, Tegge A, et al. Semaglutide and Tirzepatide reduce alcohol consumption in individuals with obesity. Scientific Reports. 2023;13(1):20998. PMID 38017205. Social-media text analysis plus a remote self-report study; not randomised.
20. Quddos F, Fowler M, de Lima Bovo AC, et al. A preliminary study of the physiological and perceptual effects of GLP-1 receptor agonists during alcohol consumption in people with obesity. Scientific Reports. 2025;15(1):32385. PMID 41093891. Pilot, n=20 (1:1 GLP-1 receptor agonist vs control), fixed alcohol challenge dose, breath alcohol and subjective effects; the authors describe it as preliminary.
21. Eshraghi R, Ghadimi DJ, Montazerinamin S, et al. Effects of glucagon-like peptide-1 receptor agonists on alcohol consumption: a systematic review and meta-analysis. EClinicalMedicine. 2025;90:103645. PMID 41324012. Pools randomised and observational designs; I² = 87.5%.
22. Bremmer MP, Hendershot CS. Social Media as Pharmacovigilance: The Potential for Patient Reports to Inform Clinical Research on Glucagon-Like Peptide 1 (GLP-1) Receptor Agonists for Substance Use Disorders. Journal of Studies on Alcohol and Drugs. 2024;85(1):5–11. PMID 37917019.
23. U.S. Department of Agriculture and U.S. Department of Health and Human Services. Dietary Guidelines for Americans, 2025–2030, page 5, "Limit Alcoholic Beverages". Released January 7, 2026; PDF creation date January 6, 2026. Listed as the current edition on ODPHP's Healthy People resource page.
24. National Institute on Alcohol Abuse and Alcoholism. Understanding Alcohol Drinking Patterns — binge, high-intensity and heavy drinking definitions and the standard drink. Retrieved August 30, 2026.
25. National Institute on Alcohol Abuse and Alcoholism. Core Resource on Alcohol — The Basics: Defining How Much Alcohol is Too Much. Retrieved August 30, 2026.
26. U.S. Department of Veterans Affairs Office of Research and Development. CSP #2041 — Cessation or Reduction of Alcohol Consumption in Veterans (CRAVE), phase 3, estimated enrollment 622, recruiting; start July 28, 2026, primary completion March 31, 2029. Record retrieved August 30, 2026.
27. National Institute on Drug Abuse. Semaglutide Therapy for Alcohol Reduction (STAR), phase 2, 63 participants, active and not recruiting; primary completion March 31, 2027. Record retrieved August 30, 2026.
28. Impact of GLP-1 Receptor Agonist Therapy on Alcohol Pharmacokinetics, phase 1, estimated enrollment 5, not yet recruiting; primary outcomes ethanol Cmax, Tmax and elimination rate in adults on a stable dose of tirzepatide. Record retrieved August 30, 2026.