GET STARTED
GLP-1s + alcohol / A guide to the evidenceAll research

The science behind
GLP-1s and alcohol.

Could medications known for weight loss change how we treat alcohol use disorder? The research is moving quickly. Here’s what the studies have found, why it matters, and what comes next.

Sunnyside Med ResearchEvidence checked October 30, 2026 · Last updated October 30, 2026First edition · About 30 minutesStart with the latest evidence ↓
A modest red-wine pour beside water and a small conceptual glass sculpture on a sunlit blue and yellow desk
Research into GLP-1s and drinking less.Illustrative scene; the sculpture is not a semaglutide molecular model.
A developing field

Some clinical trials have found improvements in drinking or cravings; other results have been mixed.

Results, in context

Findings differ by medication, formulation and population. The newest study adds patient reports; the trial results show what happened compared with placebo.

Important to know

GLP-1 medications are not FDA-approved for alcohol use disorder. Studied injections and swallowed tablets are different from compounded sublingual products. [18][17]

00Latest evidence

The latest research
on GLP-1s and alcohol.

We highlight important new studies here and keep earlier findings in the study library. The date below tells you when the featured research was published.

Latest original study highlighted
JAMA Psychiatry · October 14, 2026

Patients report drinking less after starting GLP-1s.

44.2%
reported drinking less in the first three months.

Weighted survey estimate among 242 respondents who drank when starting a GLP-1.

A national survey adds patients’ experiences to the evidence. 52.6% reported no change and 3.3% reported drinking more. Among respondents screening positive for potentially unhealthy alcohol use, 54.1% reported drinking less.

What this tells us: many patients noticed a change in their drinking. But the survey relied on people’s memories and did not randomly assign treatment, so it cannot show that GLP-1s caused the change. [1]

Read the research letter
National survey

2,009 adults surveyed; 242 GLP-1 users included in the drinking-change analysis.

Looking back

Participants recalled changes during their first three months of use.

Different from a trial

Patient reports describe what people noticed. A randomized trial can better test whether the medication caused a change.

For clinical trial results, start with the 26-week Lancet study and the other trials that compared medication with placebo. See the trial results.

Also new: An October 23 preprint reviews trials with alcohol and tobacco outcomes. It has not been peer reviewed and is not a new treatment trial. Read the research-watch note.

01The big picture

Can GLP-1s help
people drink less?

A bright blue and yellow sitting room with a water glass and notebook in afternoon light
01 · The big pictureIllustrative imagery

GLP-1s are best known for treating diabetes and helping with weight management. Researchers are now testing whether they can also help people cut back on alcohol or stop drinking. Many studies focus on alcohol use disorder, or AUD.

Early trials have found less heavy drinking, lower alcohol intake or fewer cravings with some medications. Studies of medical records have also found promising patterns. Other results have been mixed, and the studies have measured different things.[2][3][4][5][10][11]

This guide explains what each study found. Drinking less, wanting alcohol less and having fewer alcohol-related hospital visits are all useful outcomes, but they are not the same result.

Randomized trials

People are randomly assigned to medication or a comparison treatment, often a placebo, which contains no active medication. This helps test whether the medication caused a change.

Real-world studies

Researchers look at medical records or ask people about their experiences. These studies can find patterns, but other differences between people may help explain them.

Biology & animal work

Laboratory and animal studies help explain how a medication might affect drinking. Human trials are needed to find out whether it helps people.

What did the study actually measure? Alcohol use disorder, heavy-drinking days, craving, diagnosed recurrence and hospitalization are different outcomes. A lower rate of one does not establish an identical reduction in the others.

02Why GLP-1s?

How might GLP-1s
affect alcohol cravings?

Sculptural blue and yellow forms in a bright research-inspired room
02 · The biologyIllustrative imagery

GLP-1 is a hormone involved in the body’s response to food. Medications that activate its receptor can influence glucose control, appetite and digestion. Researchers are also studying how GLP-1 signals affect brain systems involved in wanting a reward and seeking it out.[18][13]

Animal experiments suggest that GLP-1 signals can affect alcohol-related behavior. A 2025 rodent study examined the lateral septum, a part of the brain involved in reward. It helps explain a possible mechanism, but does not prove that the same thing happens in people.[13]

The medication

What does it act on?

Semaglutide, exenatide, liraglutide and dulaglutide act on the GLP-1 receptor, a target involved in the body’s response to food.

The hypothesis

Could alcohol feel less appealing?

Researchers are testing whether GLP-1 signals change the urge to drink or the response to situations associated with drinking.

The trial

Does drinking actually change?

Trials in people measure whether drinking or cravings change, how much they change and who benefits.

Different GLP-1 medications need their own evidence. Tirzepatide acts at both GIP and GLP-1 receptors. Results for one drug, dose or route cannot be transferred automatically to another. A study of a swallowed semaglutide tablet is not a study of a compounded under-the-tongue tablet.

03What the trials show

What have the trials
found so far?

The studies have tested heavy drinking, alcohol consumption and cravings. Their results differ by medication, study and outcome.

The Lancet · 2026 · Randomized trial

A greater reduction in heavy-drinking days with injected semaglutide.

Reduction in the percentage of days with heavy drinking, baseline to week 26

Semaglutide + CBT
41.1 pp
Placebo + CBT
26.4 pp
01020304050 pp
108 adults with AUD and obesity. Both groups received cognitive behavioral therapy (CBT). Here, pp means percentage points. Bars show reported within-group percentage-point reductions, not the proportion of people who recovered. [2]
13.7

percentage points greater reduction with semaglutide in the adjusted between-group analysis.
95% confidence interval 5.4–22.0 points; p=.0015.

The adjusted estimate is not the raw subtraction of the two bars. This trial used weekly subcutaneous semaglutide up to 2.4 mg, not compounded sublingual semaglutide.

Laboratory alcohol intake improved.

The 2025 semaglutide trial found lower laboratory consumption and peak breath alcohol. Its primary laboratory analysis included 25 people from 48 randomized.[3]

The swallowed-tablet results were mixed.

The 2026 swallowed-tablet trial found fewer heavy-drinking days on an additional, planned measure. Its main laboratory measure of craving did not significantly improve.[4]

Not every GLP-1 trial was positive.

The 2022 exenatide trial did not find a clear overall benefit on its main heavy-drinking measure. Results in smaller body-size groups differed.[5]

Study populations matter.

A result in people with AUD and obesity does not establish the same benefit for every person who wants to drink less, or for every GLP-1 product.

There is no single “GLP-1 success rate.” These studies used different endpoints, populations and treatment settings. We have not pooled them into a percentage or ranked drugs using separate trials.

04Research timeline

GLP-1 alcohol research,
year by year.

Research notes and books in a sunlit study with blue and yellow accents
04 · Research timelineIllustrative imagery
2022

A randomized exenatide AUD trial reports an overall null primary result, with exploratory subgroup signals.[5]

2024

Large semaglutide health-record analyses add associations with AUD diagnoses and hospitalizations.[8][9]

2025

A small randomized semaglutide trial finds lower laboratory intake and selected drinking outcomes.[3]

2026

Longer injected and initial swallowed-tablet trials expand the evidence. Larger studies remain underway.[2][4][14]

Most human studies of GLP-1s and alcohol are recent. The evidence is growing, but researchers are still learning who benefits and which treatment approaches work best.

Longer studies need to show whether improvements last, what happens after treatment stops and how different products compare. New findings may change what we understand today.

05The study library

The key studies,
explained.

See who took part, what was tested, what changed and what each study could not tell us.

Search all studies

A sunlit research desk with decorative papers, blue glass and yellow stationery
05 · The study libraryIllustrative imagery
Study 1 · Randomized trial
The LancetMay 2026

Fewer heavy-drinking days
with semaglutide plus therapy.

108 adultsAUD + obesity
Weekly injectionUp to 2.4 mg
26 weeksSemaglutide vs placebo
CBT in both armsSingle center

What researchers found

Semaglutide improved the main heavy-drinking outcome more than placebo. The chart above shows the difference. Stomach and digestive symptoms were more common with semaglutide.

What to keep in mind

Participants were seeking treatment and had obesity. The trial does not establish effects in all body sizes, long-term remission, or benefits of other formulations. Therapy was part of both study arms.

Read the original study
Study 2 · Randomized trial
JAMA PsychiatryFebruary 12, 2025

Lower alcohol intake
in a laboratory test.

48 randomized25 with complete lab data
Weekly injectionLow-dose semaglutide
9 weeksLaboratory + weekly measures
Placebo controlledNot seeking treatment

What researchers found

People receiving semaglutide drank less in the laboratory test and had lower peak breath-alcohol levels. Some additional measures, including craving and drinks per drinking day, also improved. [3]

What to keep in mind

The study did not find a statistically clear improvement in average drinks per day or the number of days people drank. The main laboratory analysis used only 25 participants with complete data, and the trial was short.

Semaglutide reduced alcohol intake in the lab

-1-0.50Alcohol consumedPeak breath alcoholEffect-size coefficient β · no effect = 0
Alcohol consumed-0.48 (95% CI -0.85 to -0.11)
Peak breath alcohol-0.46 (95% CI -0.87 to -0.06)

Effect-size coefficient β · no effect = 0

Reported effect-size coefficients β and 95% confidence intervals. Primary laboratory sample: 25 complete cases. These are not percentage reductions.
Read the original study
Study 3 · Randomized trial
American Journal
of PsychiatryJuly 29, 2026

Fewer heavy-drinking days,
but no clear change in the main craving test.

50 randomizedTreatment-seeking AUD
Swallowed tablet3 mg, then 7 mg daily
8 weeksLater outcomes: weeks 5–8
Placebo controlledBMI ≥25

What researchers found

Heavy-drinking days improved on one of the additional outcomes the researchers planned to measure. The main outcome, craving during a laboratory cue test, did not significantly improve. Neither did average drinks per day. [4]

What to keep in mind

Three people in the semaglutide group withdrew; later analyses used complete data from 47 participants. The analysis did not adjust for testing multiple outcomes. Researchers used a commercially sourced swallowed tablet, not a compounded under-the-tongue tablet.

The main craving test did not show a clear benefit

-4-202Cue craving: primaryDifference in 0–20 craving-scale points · no effect = 0
Cue craving: primary-0.614 (95% CI -3.563 to 2.335)

Difference in 0–20 craving-scale points · no effect = 0

Primary laboratory craving: difference −0.614 points (95% CI −3.563 to 2.335; p=.68). The interval includes no effect. Secondary heavy-drinking-day count improved: log-link b −0.580 (95% CI −1.012 to −0.148; p=.008).
Read the original study
Study 4 · Randomized trial
JCI Insight2022

No clear overall reduction
in heavy drinking with exenatide.

127 adultsTreatment-seeking AUD
Exenatide injection2 mg once weekly
26 weeks58 completed
CBT in both armsPlacebo controlled

What researchers found

Exenatide did not significantly improve the main outcome across the full study. An exploratory analysis suggested benefit in people with a body mass index (BMI) above 30. Those with a BMI below 25 had more heavy-drinking days than the placebo group. [5]

What to keep in mind

These smaller groups were exploratory, and many participants left the study. The results do not establish a rule for choosing treatment by BMI. A later blood-marker analysis used the same participants, so it should not be counted as another trial. [7]

Body-size groups had different results

-50-2502550BMI >30 (n=30)BMI <25 (n=52)Difference in heavy-drinking-day percentage points
BMI >30 (n=30)-23.6 (95% CI -44.4 to -2.7)
BMI <25 (n=52)27.5 (95% CI 4.7 to 50.2)

Difference in heavy-drinking-day percentage points

Overall primary outcome: no significant difference. Negative favors exenatide; positive favors placebo. Shown: exploratory subgroup estimates and 95% confidence intervals.
Read the original study
Study 5 · Secondary analysis
JCI Insight2023

Less drinking during
a study on quitting smoking.

151 analyzedBaseline drinkers completing follow-up
Dulaglutide injectionUp to 1.5 mg weekly
12 weeksSecondary alcohol analysis
Smoking treatmentVarenicline + counseling

What researchers found

Adjusted weekly alcohol consumption was lower with dulaglutide than placebo (relative effect 0.71, 95% CI 0.52–0.97). [6]

What to keep in mind

This study focused on quitting smoking, not treating AUD. Most participants drank relatively little at the start, and the small heavy-drinking group showed no significant difference. Its Swiss standard glasses contained about 10 g of alcohol, compared with 14 g in a U.S. standard drink.

Weekly alcohol intake compared with placebo

00.51DulaglutideRelative consumption vs placebo · no difference = 1
Dulaglutide0.71 (95% CI 0.52 to 0.97)

Relative consumption vs placebo · no difference = 1

Adjusted relative consumption 0.71 (95% CI 0.52–0.97). Reference 1 means no difference. The result does not mean 29% of patients achieved remission.
Read the original study

06Beyond the trials

What do larger
studies show?

A contemporary sitting room in warm light with a blue chair and yellow accents
06 · Beyond the trialsIllustrative imagery

Studies of medical records can include many more people than an early clinical trial. They help researchers look for patterns in everyday care, including alcohol-related hospital visits that may be too uncommon to study well in a small trial.

People taking a GLP-1 may also differ in their health, access to care, motivation or other treatments. Researchers try to account for those differences, but cannot be sure they have accounted for everything.

Observational · target trial emulation
BMJ OpenJuly 21, 2026

GLP-1 use was linked
with fewer alcohol-related hospital visits.

Among adults with AUD and diabetes or obesity, people starting a newer GLP-1 had a lower observed risk of alcohol-related hospital care than several comparison groups. The study included 40,703 adults; some were included in more than one comparison. [10]

This study cannot prove which treatment works better. Treatment was not randomly assigned, so differences between patients may explain part of the result. It does not establish that GLP-1s outperform approved AUD medications.

Alcohol-related hospital care: comparisons with other medications

00.51Vs sulfonylureasVs other diabetes medsVs other obesity medsHazard ratio · no association = 1
Vs sulfonylureas0.74 (95% CI 0.62 to 0.89)
Vs other diabetes meds0.78 (95% CI 0.65 to 0.92)
Vs other obesity meds0.68 (95% CI 0.54 to 0.85)

Hazard ratio · no association = 1

Hazard ratios and 95% confidence intervals for alcohol-related emergency-department visits or hospitalization. Lower values favor newer GLP-1 initiation observationally; 1 means no association. Comparators differ; these are not absolute risk reductions.
New and repeat AUD diagnoses

Wang and colleagues linked semaglutide use with fewer first-time or repeat AUD diagnoses in people with obesity or diabetes. Medical-record diagnoses do not directly measure how much people drank or whether they recovered. [8]

A Swedish hospitalization signal

Lähteenvuo and colleagues linked semaglutide and liraglutide use with a lower risk of hospitalization for AUD. They compared treatment periods within the same people, but other changing factors could still help explain the findings. [9]

Important mixed evidence
Biological Psychiatry
Global Open ScienceOnline 2025 · 2026 issue

Both groups drank less.
The overall difference was unclear.

Palzes and colleagues compared 1,214 GLP-1-treated and 1,063 untreated adults. Both groups reduced weekly drinking. The overall adjusted difference was −0.43 drinks per week (95% CI −0.87 to 0.01), which was not statistically significant. [11]

A small difference was seen among people whose drinking was lower-risk at the start. Few participants had high-risk drinking, so the study could say less about that group. It did not establish a clear overall benefit.

The larger picture includes mixed results. Some medical-record studies found no clear reduction in alcohol-related admissions. In another study, a difference in weekly drinking was no longer statistically clear after researchers accounted for weight change. These findings help explain why the studies should be read together. [28][32][37]
After treatment stops: A Swedish observational study found lower AUD-admission rates during treatment and for the first six months after stopping; the association was no longer clear at six to twelve months. This does not prove a lasting treatment effect or tell someone when to stop a medication. [36]

Find a study

Search by medication, author, result or study type. Some follow-up analyses use participants from an earlier study.

Trial / secondaryKlausen 2026 · Injected semaglutide + CBTPeople with AUD and obesity had a greater reduction in heavy-drinking days than with placebo. Trial / secondaryHendershot 2025 · Injected semaglutideSmall AUD trial; laboratory intake improved; some weekly outcomes did not. Trial / secondarySchacht 2026 · Swallowed oral semaglutideThe main laboratory craving test showed no clear benefit; an additional heavy-drinking measure improved. Trial / secondaryKlausen 2022 · ExenatideNo clear overall heavy-drinking benefit; results differed in smaller body-size groups. Trial / secondaryProbst 2023 · DulaglutideAlcohol results from a study on quitting smoking, rather than treating AUD. Trial / secondaryJensen 2025 · Exenatide biomarker follow-upBlood-marker analysis using participants from the 2022 study, not a new trial. Real-worldUscher-Pines 2026 · Patient-reported drinking changesA national survey of remembered drinking changes; treatment was not randomly assigned. Real-worldWang 2024 · Semaglutide and AUD diagnosesMedical-record links with first-time or repeat AUD diagnoses. Real-worldLähteenvuo 2024/2025 · Semaglutide and liraglutideMedical-record links with a lower risk of hospitalization for AUD. Real-worldRodriguez 2026 · Semaglutide, tirzepatide and hospitalizationsSeveral medical-record comparisons; other differences between patients may affect the results. Real-worldPalzes 2025/2026 · Alcohol use in routine careThe overall difference in weekly drinking was not statistically clear. Real-worldCai 2026 · GLP-1s and substance use disordersVeterans with type 2 diabetes; broader substance-use outcomes, observational design. Animal study2025 · Lateral septum and GLP-1 signalingRodent experiments investigate reward circuitry and alcohol intake. Real-worldWium-Andersen 2022 · Early alcohol-related eventsAlcohol-related events were less frequent in the first three months after GLP-1RA initiation; the finding does not establish sustained benefit or reduced alcohol volume. Real-worldQuddos 2023 · Self-reported drinking in obesityTreated respondents reported lower drinking and alcohol-problem scores; self-selection and recalled premedication drinking limit causal conclusions. Real-worldQeadan 2024 · Alcohol-intoxication recordsPrescriptions were associated with fewer alcohol-intoxication events, which is different from drinking half as much. Real-worldMiller-Matero 2024 · Reports during weight-loss treatment45.3% of baseline drinkers reported a lower alcohol-use category; the study did not show a clear advantage for GLP-1 categories over metformin. Real-worldO’Farrell 2025 · Drinking changes in routine careAlcohol intake fell among patients with paired data; substantial missing follow-up and no control group limit attribution. Real-worldFarokhnia 2025 · Alcohol screening scoresGLP-1RA recipients had modestly greater drinking-score reductions; differences were larger in people with baseline AUD or hazardous drinking. Real-worldAl-Moussally 2025 · No clear reduction in admissionsNeither cirrhosis progression nor alcohol-related admission was significantly lower with GLP-1RAs. Real-worldKing 2025 · A small prospective studyAUDIT scores fell, especially among those with very high baseline scores;14 uncontrolled patients cannot establish a medication-specific response rate. Real-worldQuddos 2025 · Alcohol absorption and subjective effectsThe exposed group had a delayed rise in breath alcohol and subjective effects. This investigates mechanism and does not show that drinking while taking these drugs is safe.
Real-worldHenney 2025 · First-recorded AUD diagnosesSemaglutide and tirzepatide were associated with fewer first-recorded AUD diagnoses; liraglutide and dulaglutide comparisons were not significant. A first diagnosis in a health record may reflect newly recognized—not newly developed—AUD. This does not establish that the medications prevented AUD.[44]
Real-worldDai 2025 · Mixed findings in older adultsAlcohol-specific hospitalization was not significantly lower versus DPP-4; no significant differences were seen versus SGLT2. Real-worldAbegaz 2026 · AUD diagnoses in All of UsGLP-1RA use was associated with lower odds of an AUD diagnosis; an odds ratio is not a measured reduction in drinking. Real-worldGougol 2026 · Relapse records in metabolic diseaseGLP-1RA exposure was associated with less coded relapse; liver-decompensation differences were not significant and mortality was similar. Real-worldTyndall 2026 · Drinking frequency in All of UsCurrent users had modestly lower drinking scores, driven mainly by frequency; drinks per occasion and binge measures were not significantly associated. Real-worldBach 2026 · During treatment and after stoppingAUD-admission rates were lower during treatment and the first six months after stopping, but the association was no longer clear at six to twelve months. Real-worldBasri 2026 · Accounting for weight changeAn approximately one-drink-per-week difference became nonsignificant after accounting for weight change. Real-worldJohn 2025 · Liver outcomes and drinking screensGLP-1RA use was associated with fewer liver outcomes and positive drinking screens; the study does not prove alcohol-specific mortality prevention. Real-worldKuo 2025 · Alcohol-related liver diseaseGLP-1RA use was associated with fewer liver outcomes; these are liver-disease outcomes rather than direct measures of drinking. Real-worldFakhoury 2025 · After bariatric surgeryIncretin-based treatment was associated with fewer new AUD diagnoses after bariatric surgery; applicability to people with existing AUD is uncertain.

07Research to watch

What researchers
are studying next.

Books, research notes and blue stationery on a warmly lit desk
07 · What comes nextIllustrative imagery

Researchers are planning larger trials and testing different medications and combinations. The registered trials below have not reported treatment results yet; early reports are labeled separately. Recruitment labels reflect the public records checked for this edition.

Recruiting

Registry checked
October 27, 2026

CRAVE: a larger semaglutide trial in veterans.

The phase 3 trial plans to enroll 622 participants and compare semaglutide with placebo in AUD. Larger confirmatory studies can address questions that smaller early trials cannot. [14]

View NCT07218354
Not yet recruiting

Registry checked
October 27, 2026

GLP-1 plus naltrexone: testing a combination.

A phase 2 study plans to enroll 45 participants. The record lists GLP-1 agonist and oral naltrexone interventions. It has no published efficacy results and does not establish the effectiveness of compounded combinations. [15]

View NCT07249554
Recruiting

Registry checked
October 27, 2026

Tirzepatide in alcohol use disorder.

A phase 2 placebo-controlled study plans to enroll 42 participants. Tirzepatide is a dual GIP/GLP-1 agonist, so this is a distinct medication question rather than a repeat of semaglutide. [16]

View NCT06994338
Design / early report

Alcohol, metabolic health and liver disease.

The UK CURB project studies people with AUD, chronic liver disease and obesity. Separately, a 2026 conference abstract from an alcohol-related liver-disease trial reported biomarker improvements but no significant alcohol-consumption difference. Liver outcomes and drinking outcomes must be read separately. [21][20]

Sponsor report · Not yet peer reviewed

An investigational dual agonist has reported early trial results.

In July 2026, Altimmune reported that pemvidutide reduced heavy-drinking days more than placebo in its phase 2 RECLAIM trial. These are sponsor-reported topline findings, not a peer-reviewed paper. Pemvidutide acts on glucagon and GLP-1 receptors and is different from semaglutide. Full results are needed to assess the findings and safety in detail.[45]

Preprint · October 23, 2026

A new review brings trial findings together.

A review posted as a preprint reports 11 trials with alcohol outcomes, including additional outcomes from studies designed for other purposes. That does not mean 11 published trials treating AUD. The underlying studies overlap with this guide, and most outcomes in the review were rated low or very low certainty. We have not independently traced every trial in its full bibliography. [42]

Registry results · Small pilot

An early experiment with only three completed participants.

A single-dose exenatide crossover pilot ended early during COVID-19. Its registry reports laboratory drinking results for three people. It is too small to establish a reliable treatment benefit and is not another published AUD efficacy trial. [43]

What we still need to learn: who benefits most, whether changes last, what happens after stopping, which doses and forms work best, how GLP-1s compare with established AUD treatments, and whether adding naltrexone helps. The completed studies here do not answer all of these questions.

08Before starting

What to discuss
with your clinician.

A calm consultation setting with two people talking in a bright room with Sunnyside colors
08 · Before startingIllustrative imagery

If you’re considering medication, a clinician can help you weigh the options. That conversation should cover your drinking, goals, health, other medications and the support you need.

Describe your drinking and goals.

How much and how often do you drink? Are you hoping to cut back or stop? Include previous attempts and symptoms when you reduce alcohol. These details influence the safest plan.

Plan for withdrawal risk.

Alcohol withdrawal can be dangerous. GLP-1 medication and naltrexone do not treat acute alcohol withdrawal. If you drink heavily or have had withdrawal symptoms, get medical advice before reducing or stopping alcohol. Seizures, hallucinations or severe confusion require emergency care; call 911. [19]

Compare established options.

Ask about FDA-approved AUD medications, including naltrexone, acamprosate and disulfiram, as well as behavioral support. An emerging option should be considered alongside the treatments already available. [19]

Understand the exact formulation.

Ask what product and route you are being offered, whether it is FDA-approved or compounded, and which evidence applies. Compounded drugs do not undergo FDA premarket review for safety, effectiveness or quality. [17]

Discuss your health and medications.

Semaglutide labeling includes a boxed thyroid C-cell tumor warning; personal or family history of medullary thyroid cancer or MEN2 is a contraindication. Discuss pregnancy plans, gastrointestinal conditions, pancreatitis or gallbladder history, diabetes medicines and planned procedures. [18]

Arrange follow-up and side-effect support.

Nausea, vomiting, diarrhea, constipation and abdominal pain are common semaglutide adverse effects. Severe gastrointestinal problems and dehydration can occur. Ask who to contact, what needs urgent attention and how your response will be reviewed. [18]

This is an educational checklist, not a complete prescribing or safety guide. Follow the instructions for the medication actually prescribed to you. A clinician should assess suitability and explain product-specific risks.

09Common questions

Your questions
about GLP-1s and alcohol.

A sunlit library with open books and blue and yellow spines
09 · Common questionsIllustrative imagery
Can GLP-1s help people drink less?

Some randomized trials found improvements, including fewer heavy-drinking days. Other main study outcomes did not significantly improve. The findings are promising, but cannot predict how every person will respond to every product. [2][3][4][5]

Is semaglutide FDA-approved for alcohol use disorder?

No. Its FDA-approved indications concern other conditions. Prescribing a drug outside an approved indication is different from FDA approval of that use. Compounded semaglutide products themselves are not FDA-approved. [18][17]

Is “alcohol noise” a scientific endpoint?

It is a conversational way of describing persistent thoughts or urges about drinking. Studies may instead measure craving on a named scale, drinking quantity, heavy-drinking days or clinical events. Those measures should not be treated as interchangeable.

Does the oral trial prove that an under-the-tongue tablet works?

No. The 2026 trial used swallowed oral semaglutide tablets. An under-the-tongue tablet may deliver medication differently. That trial does not show that a compounded sublingual tablet is absorbed in the same way or has the same effects. [4]

What about tirzepatide, liraglutide and other GLP-1s?

They are different medications, with different research behind them. Findings from medical records and plans for new trials do not show that they all work equally well for AUD. [9][10][16]

Does a hospitalization study show GLP-1s work better than naltrexone?

No. A study of medical records can find a useful link, but differences in illness, prescribing or access to care may help explain it. A randomized trial comparing the medications would give stronger evidence about which works better. [10]

Has semaglutide plus naltrexone been proven as a combination?

The registered combination trial is listed as not yet recruiting in the record checked for this edition. A study plan is not proof of added benefit, and it does not establish outcomes for compounded formulations. [15]

How soon should someone expect a change?

These studies do not establish a reliable personal timeline. They used different drugs, doses, durations and outcomes. Follow-up should assess your own response and tolerability rather than promise a particular week of improvement.

Can someone safely stop heavy drinking after starting medication?

Do not assume that reduced urges remove withdrawal risk. If you may be physically dependent on alcohol, speak with a clinician before abruptly stopping or cutting down. Withdrawal management is a separate clinical need. [19]

How is this guide kept current?

The review date and change log show what was checked for each edition. Important new findings can replace the featured study at the top, while earlier studies remain in the library. Each update needs a review of the original research.

10Sunnyside Med

Medication and support
with Sunnyside Med.

Two people talking in a bright consultation space with blue and yellow details
10 · AI-generated illustration; not actual Sunnyside clinicians or patients. Care is provided online.Illustrative imagery

If you want to change your relationship with alcohol, it helps to have a plan for the everyday moments: dinner out, the end of a long day, or the evening that doesn’t go as planned.

Sunnyside Med is a 100% online medical program for eligible adults who want to drink less or quit. Clinical care, the Sunnyside app, real coaches, community and live member meetings give you practical support as you work toward your goal.

What these studies did not test: These studies tested specific medications, doses and treatment settings, not Sunnyside Med’s complete program. They do not establish the effectiveness of compounded formulations.

Important naltrexone safety: Tell your clinician about any opioid use, including opioid pain or cough medicines, tramadol, methadone and buprenorphine. Starting naltrexone too soon after opioid use can cause sudden, severe withdrawal. Your clinician must determine when it is safe to start. Never try to overcome its opioid-blocking effects by taking more opioids; this can cause a fatal overdose. Discuss liver problems and all medications before treatment. [41]

Compounded medications are not FDA-approved. FDA does not review compounded medications for safety, effectiveness or quality before marketing. Medications are prepared and dispensed by a partner pharmacy, not Sunnyside. A licensed clinician determines whether treatment is appropriate; a prescription is not guaranteed. [17]

  • Clinical care

    A licensed clinician reviews your health, medications and goals and determines whether a prescription is appropriate.

  • Real human coaching

    Talk through challenges, plan for difficult moments and get encouragement from a real coach.

  • Daily habit support

    Make a weekly plan, track your drinking and notice the patterns you want to change.

  • Community and live meetings

    Connect with people working toward their own drinking goals and join live member meetings for support.

11In the news

GLP-1s and alcohol,
in the news.

Read and watch how journalists are covering the research, the people behind it and the questions still being tested.

These are selected reports about the field, not coverage or endorsements of Sunnyside. For the findings themselves, read the original studies above. Some publishers require a subscription.

12Sources & updates

Read the original studies.

An open book and research volumes on a sunlit shelf
12 · SourcesIllustrative imagery

This guide brings together key clinical trials, observational studies and emerging research on GLP-1s and alcohol, with an emphasis on randomized trials in people. It is a curated guide, not an exhaustive systematic review. It also covers medical-record studies, surveys, animal research and ongoing trials. Studies with mixed results or no clear benefit are included.

The charts use numbers from the linked studies. Notes explain what was measured, how many people were included and how uncertain the estimates are. Different studies have not been combined into a single success rate.

  1. Uscher-Pines L, et al. Patient-Reported Changes in Alcohol Use Following GLP-1 Receptor Agonist Initiation. JAMA Psychiatry. Published online October 14, 2026 · National survey · DOI 10.1001/jamapsychiatry.2026.3242
  2. Klausen MK, et al. Once-weekly semaglutide versus placebo in patients with alcohol use disorder and comorbid obesity. The Lancet. May 2026 · Randomized trial · DOI 10.1016/S0140-6736(26)00305-3
  3. Hendershot CS, et al. Once-Weekly Semaglutide in Adults With Alcohol Use Disorder. JAMA Psychiatry. Published online February 12, 2025 · Randomized trial · DOI 10.1001/jamapsychiatry.2024.4789
  4. Schacht JP, et al. Oral Semaglutide for Alcohol Use Disorder. American Journal of Psychiatry. Published online July 29, 2026 · Randomized trial · DOI 10.1176/appi.ajp.20260003
  5. Klausen MK, et al. Exenatide once weekly for alcohol use disorder investigated in a randomized, placebo-controlled clinical trial. JCI Insight. 2022 · Randomized trial · DOI 10.1172/jci.insight.159863
  6. Probst L, et al. Effects of dulaglutide on alcohol consumption during smoking cessation. JCI Insight. 2023 · Secondary analysis of a randomized trial · DOI 10.1172/jci.insight.170419
  7. Jensen ME, et al. Blood phosphatidylethanol measurements indicate GLP-1 receptor stimulation causes delayed decreases in alcohol consumption. Alcohol, Clinical and Experimental Research. 2025 · Secondary biomarker analysis of the 2022 exenatide cohort · DOI 10.1111/acer.70041
  8. Wang W, et al. Associations of semaglutide with incidence and recurrence of alcohol use disorder in real-world population. Nature Communications. 2024 · Observational cohorts · DOI 10.1038/s41467-024-48780-6
  9. Lähteenvuo M, et al. Repurposing Semaglutide and Liraglutide for Alcohol Use Disorder. JAMA Psychiatry. Online November 13, 2024; January 2025 issue · Observational study · DOI 10.1001/jamapsychiatry.2024.3599
  10. Rodriguez PJ, et al. Association between GLP-1 receptor agonists and alcohol-related hospitalisations among adults with alcohol use disorder. BMJ Open. July 21, 2026 · Multi-target trial emulation · DOI 10.1136/bmjopen-2025-109259
  11. Palzes VA, et al. GLP-1 Receptor Agonists and Alcohol Use: A Real-Word Observational Study in a Large, Integrated Health Care System. Biological Psychiatry Global Open Science. Online November 24, 2025; March 2026 issue · Observational study · DOI 10.1016/j.bpsgos.2025.100659
  12. Cai M, et al. GLP-1 receptor agonists and risk of substance use disorders among US veterans with type 2 diabetes: cohort study. The BMJ. 2026 · Observational cohort · DOI 10.1136/bmj-2025-086886
  13. Edvardsson CE, et al. An inhibitory GLP-1 circuit in the lateral septum modulates reward processing and alcohol intake in rodents. 2025 · eBioMedicine · Preclinical rodent research · DOI 10.1016/j.ebiom.2025.105684
  14. ClinicalTrials.gov. CRAVE: Cessation or Reduction of Alcohol Consumption in Veterans. Registry NCT07218354 · Recruiting · Record updated September 21, 2026; checked October 27, 2026
  15. ClinicalTrials.gov. Combination Therapy for Alcohol Use Disorder. Registry NCT07249554 · Not yet recruiting · Record updated October 1, 2026; checked October 27, 2026
  16. ClinicalTrials.gov. Tirzepatide for Alcohol Use Disorder. Registry NCT06994338 · Recruiting · Record updated October 14, 2025; checked October 27, 2026
  17. U.S. FDA. Concerns with Unapproved GLP-1 Drugs Used for Weight Loss. Official drug-safety information · Checked October 27, 2026
  18. U.S. FDA. Wegovy (semaglutide) prescribing information. FDA label revised February 2026 · Approved indications, boxed warning and safety information
  19. NYSDOH AIDS Institute Clinical Guidelines Program. Treatment of Alcohol Use Disorder. Clinical guideline · Withdrawal assessment and established treatment options
  20. Shah V, et al. Effects of zalfermin, cagrilintide, and semaglutide on non-invasive fibrosis biomarkers in alcohol-related liver disease. Hepatology, Liver Meeting abstract 0087. 2026 conference abstract · Not a full published AUD efficacy report
  21. NIHR. CURB trial: GLP-1 receptor agonist for AUD in people with chronic liver disease and obesity. Research funding and trial-design record · Results pending
  22. Wium-Andersen IK et al. Use of GLP-1 receptor agonists and subsequent risk of alcohol-related events. A nationwide register-based cohort and self-controlled case series study. Basic & clinical pharmacology & toxicology. Online 2022-08-30; November 2022 issue · Observational · DOI 10.1111/bcpt.13776
  23. Quddos F et al. Semaglutide and Tirzepatide reduce alcohol consumption in individuals with obesity. Scientific reports. Published 2023-11-28 · Observational · DOI 10.1038/s41598-023-48267-2
  24. Qeadan F et al. The association between glucose-dependent insulinotropic polypeptide and/or glucagon-like peptide-1 receptor agonist prescriptions and substance-related outcomes in patients with opioid and alcohol use disorders: A real-world data analysis. Addiction (Abingdon, England). Online 2024-10-16; February 2025 issue · Observational · DOI 10.1111/add.16679
  25. Miller-Matero LR et al. Alcohol Use and Antiobesity Medication Treatment. JAMA network open. Publisher online 2024-11-26; November 2024 issue · Observational · DOI 10.1001/jamanetworkopen.2024.47644
  26. O’Farrell M et al. Glucagon-like peptide-1 analogues reduce alcohol intake. Diabetes, obesity & metabolism. First published 2025-01-02; March 2025 issue · Observational · DOI 10.1111/dom.16152
  27. Farokhnia M et al. Glucagon-like peptide-1 receptor agonists, but not dipeptidyl peptidase-4 inhibitors, reduce alcohol intake. The Journal of clinical investigation. Online 2025-03-06; May 2025 issue · Observational · DOI 10.1172/jci188314
  28. Al-Moussally F et al. Association of Glucagon-Like Peptide-1 Receptor Agonist with Progression to Liver Cirrhosis and Alcohol-Related Admissions in Patients with Alcohol Use Disorder and Diabetes: A Retrospective Cohort Study. Drugs. Online 2025-04-13; June 2025 issue · Observational · DOI 10.1007/s40265-025-02177-x
  29. King AC et al. GLP-1 RA Medication Associations With Hazardous Alcohol Drinking Reductions in Patients With Overweight or Obesity: A Prospective Observational Study. Journal of addiction medicine. Online 2025-08-12; May–June 2026 issue · Observational · DOI 10.1097/adm.0000000000001565
  30. Quddos F 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. Published 2025-10-15 · Observational · DOI 10.1038/s41598-025-17927-w
  31. Henney AE et al. Relative efficacy of GLP-1 and GLP-1/GIP receptor agonists in the prevention of alcohol-use disorders using a target trial emulation approach. Diabetes, obesity & metabolism. Online 2025-10-07; January 2026 issue · Observational · DOI 10.1111/dom.70169
  32. Dai H et al. GLP-1 Receptor Agonists and Substance Use Disorders in Older Adults With Type 2 Diabetes: A Target Trial Emulation. Obesity (Silver Spring, Md.). Online 2025-09-15; September 2026 issue · Observational · DOI 10.1002/oby.70024
  33. Abegaz TM et al. Association between GLP-1 receptor agonist use and substance use disorders among individuals with type 2 diabetes or obesity: a nested case-control study in the All of Us research program. Frontiers in psychiatry. Published 2026-03-10 · Observational · DOI 10.3389/fpsyt.2026.1766770
  34. Gougol A et al. Real-World Alcohol Use Disorder Outcomes in Patients With Concurrent Metabolic Dysfunction: GLP-1 Receptor Agonists Versus FDA-Approved AUD Medications. Alimentary pharmacology & therapeutics. Online 2026-03-11; June 2026 issue · Observational · DOI 10.1111/apt.70596
  35. Tyndall B et al. Association of GLP-1 Receptor Agonist Prescriptions and Alcohol Consumption in the National Institutes of Health’s All of Us Cohort. Alcohol, clinical & experimental research. July 2026 issue; exact online day not independently established · Observational · DOI 10.1111/acer.70357
  36. Bach P et al. Association of GLP-1 receptor agonists with alcohol use disorder-related and substance use disorder-related hospital admissions during treatment and after discontinuation: a Swedish register-based within-individual observational study. The lancet. Psychiatry. August 2026 issue; exact first-online day unresolved · Observational · DOI 10.1016/s2215-0366(26)00172-0
  37. Basri T et al. Alcohol Use Following GLP-1 Receptor Agonist Therapy in the Cooper Center Longitudinal Study. Mayo Clinic proceedings. Online 2026-07-22 · Observational · DOI 10.1016/j.mayocp.2026.07.014
  38. John BV et al. Association of Glucagon-Like Peptide-1 Receptor Agonists With Liver-Related Outcomes and All-Cause Mortality in Patients With Harmful Alcohol Use: A Target Trial Emulation Study. The American journal of gastroenterology. Online 2025-06-10; March 2026 issue · Observational · DOI 10.14309/ajg.0000000000003585
  39. Kuo CC et al. Impact of GLP-1 Receptor Agonists on Alcohol-Related Liver Disease Development and Progression in Alcohol Use Disorder. Alimentary pharmacology & therapeutics. Online 2025-01-31; April 2025 issue · Observational · DOI 10.1111/apt.70007
  40. Fakhoury B et al. Incretin-Based Therapies and Post-Bariatric Surgery Alcohol Use Disorder. JAMA network open. December 2025 issue; exact electronic day not independently verified · Observational · DOI 10.1001/jamanetworkopen.2025.49086
  41. DailyMed, National Library of Medicine. Naltrexone hydrochloride tablets USP 50 mg: prescribing information. Prescribing information · Opioid contraindications, withdrawal and pain-medication risks · Checked October 28, 2026
  42. Taylor G, et al. GLP-1 receptor agonists for treatment of tobacco and alcohol dependence: a systematic review and meta-analysis of randomised controlled trials. October 23, 2026 · Preprint; not peer reviewed · Secondary synthesis, not a new treatment trial
  43. ClinicalTrials.gov. The Effects of Exenatide, a GLP-1 Agonist, on Alcohol Self-Administration in Heavy Drinkers. Registry-posted results February 9, 2022 · Three completed participants · Early-terminated pilot; not a peer-reviewed paper
  44. Sung CH, Wei LC. First-Recorded Alcohol Use Disorder Diagnosis as an Ascertainment Outcome: Implications for GLP-1 Target Trial Emulations. Pharmacoepidemiol Drug Saf. 2026;35:e70472. doi:10.1002/pds.70472. ↗First published August 27, 2026 · Methodological correspondence; no new participants · Full letter checked October 30, 2026
  45. Altimmune. RECLAIM phase 2 topline results announcement. July 28, 2026. ↗Sponsor report; not peer reviewed · Pemvidutide is a glucagon/GLP-1 dual agonist, not semaglutide · Checked October 30, 2026
  46. Lilly. RENEW-ALC-1 (NCT07219966): A Study of Brenipatide in Participants With Moderate-to-Severe Alcohol Use Disorder. ↗Live sponsor record checked October 30, 2026 · Phase 3; completed enrollment; planned enrollment 1,100; approximately 56 weeks participation · Enrollment complete does not mean study complete; not efficacy results.
  47. Lilly. RENEW-ALC-2 (NCT07219953): A Study of Brenipatide in Participants With Alcohol Use Disorder. ↗Live sponsor record checked October 30, 2026 · Phase 3; enrolling; planned enrollment 1,100; approximately 56 weeks participation · Investigational brenipatide versus placebo; not efficacy results.
  48. Eli Lilly and Company. Clinical Development Pipeline: brenipatide (LY3537031). ↗Sponsor pipeline checked October 30, 2026 · Investigational dual GIP/GLP-1 receptor agonist; a different molecule from semaglutide or tirzepatide · Pipeline stage is not evidence of treatment benefit.

This edition

October 27, 2026 · First edition. Added the October 14 patient-report survey as the latest-evidence feature; reviewed the Lancet semaglutide trial, swallowed oral trial, earlier intervention studies and selected real-world evidence. Checked the public recruitment records shown above.

October 28, 2026 · Source and release audit. Expanded the searchable index with 19 original observational studies, including mixed, null and discontinuation findings. Added six news links, an explicitly unreviewed October 23 preprint, a very small registry-results pilot and naltrexone/opioid safety information. Product-marketing and AI claims reviews informed this revision; human clinical review remains pending.

October 30, 2026 · Trial-watch addition. Added a prominent near-top callout for Lilly’s two phase 3 brenipatide alcohol studies. Kept enrollment status, planned size and the distinction from semaglutide separate from published results. The featured original study and earlier findings are unchanged.

October 30, 2026 · Weekly evidence update. Clarified that first-recorded AUD diagnoses do not establish prevention. Added a separately labeled July RECLAIM sponsor-results report, distinct from peer-reviewed evidence and semaglutide research. Reviewed new literature, safety information, corrections and trial-watch records; earlier trial charts and the featured survey are unchanged. Human clinical review remains pending.

Search coverage: The October 28 audit screened public bibliographic records for original human alcohol outcomes and checked 27 distinct trial registrations. Selected original abstracts and full texts were inspected. This is not a registered systematic review. Some full texts and supplements were inaccessible, and the ancillary trials cited by the October 23 preprint have not all been independently traced.

Next revision: assess new primary publications, corrections and trial results; replace the lead feature when a finding materially changes the picture; retain earlier evidence and record substantive changes here. Clinical review is pending.

Educational information, not individual medical advice. Medication is not appropriate for everyone. Decisions require a licensed clinician’s assessment. No GLP-1 medication is FDA-approved for alcohol use disorder as of this evidence check. Compounded medications are not FDA-approved and are not established as equivalent to the drug products studied here.

About this guide. Human clinical review is pending; no clinical approval is claimed. Study findings are drawn from the cited primary sources; some full-text access was limited. Images are AI-generated illustrations, not actual studies, molecular structures, patients, clinicians or results. Publication names identify sources and do not imply endorsement.