Zepbound and Alcohol: The Evidence Ledger of RCTs, Cohorts, and Preclinical Data

At a glance

• Zepbound FDA label is silent on alcohol. No contraindication, no warning, no drug-drug interaction. The only label mentions of “alcohol” refer to the injection-site swab^[1].
• No disulfiram-like enzymatic block. Tirzepatide does not inhibit aldehyde dehydrogenase. Patients on Zepbound will not experience the disulfiram-ethanol flushing reaction.
• One Phase 2 RCT in alcohol use disorder, semaglutide-only. Hendershot 2025 (n=48, 9 weeks) reported reduced grams of alcohol per drinking day and reduced craving scores. The drug was semaglutide 1.0 mg weekly, not tirzepatide^[2].
• Earlier RCT in AUD did not meet its primary endpoint. Klausen 2022 (exenatide, n=127, 26 weeks) was negative overall but showed fMRI alcohol-cue reactivity attenuation in the pre- specified obesity subgroup^[3].
• Tirzepatide-specific human observational data: one published paper. Quddos 2023 in Scientific Reports reported reduced alcohol intake among adults with obesity on semaglutide or tirzepatide^[4].
• Preclinical: tirzepatide blunts alcohol’s interoceptive cues in rats. Windram 2025 showed semaglutide, tirzepatide, and retatrutide all attenuate the discriminative-stimulus effects of alcohol in a drug-discrimination paradigm^[5].
• Largest observational signal is semaglutide and liraglutide. Lähteenvuo 2025 (Swedish nationwide cohort, ~227,000 adults with AUD) reported reduced AUD-related hospitalization with semaglutide and liraglutide. Tirzepatide was not assessed^[6].
• First head-to-head cohort: GLP-1 vs GLP-1/GIP for AUD prevention. Henney 2026 in Diabetes, Obesity & Metabolism used a target-trial emulation framework to compare semaglutide against tirzepatide for incident AUD in adults with obesity and type 2 diabetes. The dual GIP plus GLP-1 agonist signal was at least non-inferior^[7].
• Patient-facing summary: moderation is the default framework. The label permits drinking. Tirzepatide-specific AUD evidence is not yet trial-strength. For the full FDA-label deep-dive, see the sister article Zepbound and Alcohol: Honest Evidence Review.

What the Zepbound FDA label says about alcohol

The current Zepbound prescribing information (DailyMed SetID 487cd7e7-434c-4925-99fa-aa80b1cc776b) was reviewed in full on 2026-05-24^[1]. There is no alcohol-specific contraindication, no boxed-warning mention of alcohol, no Drug Interactions (Section 7) entry naming ethanol or alcoholic beverages, and no Patient Counseling Information bullet about avoiding alcohol. The label’s only “alcohol” references are to the alcohol swab used to prepare the injection site.

Section 5 (Warnings and Precautions) lists labeled risks that can be independently affected by heavy alcohol use: acute pancreatitis (5.5), acute gallbladder disease (5.4), acute kidney injury via volume contraction (5.3), and hypoglycemia with concomitant insulin or insulin secretagogue exposure (5.7). Section 7.1 (Drug Interactions) discusses delayed gastric emptying and its effect on the absorption of co-administered oral medications, not on alcohol per se. For the section-by-section walkthrough of how these labeled risks intersect with alcohol pharmacology, see the companion deep-dive Zepbound and Alcohol: Honest Evidence Review. This article focuses tightly on the published trial and cohort evidence.

The randomized trial evidence ledger

As of May 2026 the trial-level evidence for whether a GLP-1 or GLP-1/GIP receptor agonist reduces alcohol consumption in adults with alcohol use disorder consists of two completed Phase 2 randomized controlled trials, with a third (SEMALCO) in progress. Neither completed RCT used tirzepatide.

• Hendershot 2025 (JAMA Psychiatry, n=48, 9 weeks). Once-weekly semaglutide 1.0 mg vs placebo in adults meeting DSM-5 criteria for alcohol use disorder. The trial reported reductions in grams of alcohol per drinking day and in self-reported craving scores in the semaglutide arm. Sample size was small and the trial duration short, but this is currently the strongest published randomized evidence in the GLP-1 plus AUD literature. The drug used was semaglutide, not tirzepatide^[2].
• Klausen 2022 (JCI Insight, n=127, 26 weeks). Exenatide once weekly vs placebo in adults with AUD. The primary endpoint (heavy-drinking days) was not met in the overall sample. In the pre-specified obesity subgroup, exenatide was associated with attenuated fMRI alcohol-cue reactivity in the ventral striatum and septal area, supporting a CNS-reward mechanism but not a clinical-outcome win^[3].

There is no completed Phase 2 or Phase 3 randomized trial of tirzepatide (Zepbound or Mounjaro) for alcohol use disorder published as of May 2026. For the broader GLP-1 plus AUD trial landscape (including the in-progress SEMALCO Phase 2 protocol and the three FDA-approved AUD medications naltrexone, acamprosate, and disulfiram), see the dedicated article GLP-1 Drugs and Alcohol Use Disorder.

Tirzepatide-specific human data: what is published

The single peer-reviewed human study that includes tirzepatide directly on an alcohol-consumption endpoint is Quddos 2023 in Scientific Reports^[4]. The design was observational using a large US claims and EHR network. Adults with obesity on semaglutide or tirzepatide were compared against matched non-GLP-1 controls. Alcohol consumption was self-reported and assessed via the AUDIT-C questionnaire. Both semaglutide and tirzepatide groups had lower follow-up AUDIT-C scores than controls. The signal direction is consistent with the Hendershot 2025 RCT, but observational design and self-reported outcome measures limit causal inference.

Beyond Quddos, the published tirzepatide-and-alcohol literature consists of case reports and case series — not formal cohort or RCT evidence. The 2025 systematic review and meta-analysis by de Faria Moraes in Drug and Alcohol Dependence includes the available GLP-1 receptor agonist evidence on alcohol consumption and liver-related outcomes; the pooled estimate supports a class-level reduction in alcohol intake, with the caveat that most contributing studies used semaglutide or liraglutide^[9].

Preclinical: tirzepatide attenuates alcohol’s interoceptive cues

The Windram 2025 study in Psychopharmacology tested semaglutide, tirzepatide, and retatrutide in a rat drug- discrimination paradigm trained on alcohol’s interoceptive stimulus. All three incretin-based agents attenuated the discriminative-stimulus effects of alcohol in both male and female rats^[5]. Drug-discrimination is a translatable preclinical readout for the perceptual or subjective effects of a substance; attenuation in this paradigm is consistent with a CNS mechanism that reduces the salience or reward value of alcohol. This is preclinical evidence, not a clinical endpoint, but it is the most direct mechanistic evidence to date that tirzepatide engages the same alcohol-related neural pathway implicated for semaglutide.

The large observational signals

Two recent cohort studies have produced the largest observational signals in the GLP-1 plus AUD literature.

• Lähteenvuo 2025 (JAMA Psychiatry). Swedish and Finnish nationwide registries of adults with a recorded diagnosis of alcohol use disorder (~227,000 individuals) were analyzed using a within-individual design comparing periods on vs off semaglutide or liraglutide. Both drugs were associated with reduced AUD-related hospitalization and reduced all-cause hospitalization. Tirzepatide was not separately analyzed^[6].
• Henney 2026 (Diabetes, Obesity & Metabolism). A target-trial emulation using a global federated EHR network (TriNetX) compared GLP-1 receptor agonists (semaglutide, liraglutide, dulaglutide) against the dual GIP/GLP-1 agonist tirzepatide for incident alcohol-use-disorder diagnosis in adults with obesity, with and without type 2 diabetes. The tirzepatide signal was at least non-inferior to the GLP-1-only signal on the AUD-prevention endpoint and trended favorably on some secondary measures. This is the first published cohort comparison that includes tirzepatide on an AUD endpoint^[7].

Both cohorts are observational and subject to confounding by indication, residual confounding, and differential follow-up. The target-trial emulation framework in Henney 2026 mitigates some of these limitations but does not eliminate them. The signals are directionally consistent with the RCT and preclinical evidence, but they are not a substitute for prospective randomized data.

Why the mechanism is plausibly class-level

GLP-1 receptors are expressed in the ventral tegmental area and nucleus accumbens, brain regions central to drug and alcohol reward. GLP-1 receptor activation in these regions attenuates alcohol-induced dopamine release in rodent models. Tirzepatide adds GIP receptor co-activation. The Windram 2025 rat study showed that tirzepatide attenuates alcohol’s interoceptive effects to a similar degree as semaglutide^[5], supporting a class-level CNS mechanism. The Henney 2026 cohort then provides the first human-population signal that the dual GIP plus GLP-1 mechanism is at least non-inferior to pure GLP-1 on AUD outcomes^[7]. The class-level interpretation is plausible but not RCT-confirmed for tirzepatide specifically.

What this means for someone on Zepbound who drinks

The published evidence does not support changing alcohol-use guidance for the typical Zepbound patient beyond what the FDA label and standard moderation principles already imply. The label permits drinking; the patient-reported reduced-tolerance experience is real and plausibly mediated by tirzepatide’s pronounced gastric-emptying delay and by a class-level reduction in alcohol reward signaling.

• Most Zepbound users without diabetes can drink moderately. The hypoglycemia warning in Section 5.7 applies specifically to patients with type 2 diabetes on concomitant insulin or a sulfonylurea, which is uncommon in the Zepbound weight- management population^[1].
• Reduced tolerance is the most common patient experience. Many patients report that one drink feels like two or three. The mechanism is plausibly altered ethanol-absorption kinetics from delayed gastric emptying plus blunted reward signaling.
• Binge patterns stack risk on labeled adverse events. Acute pancreatitis (Section 5.5) and gallbladder disease (Section 5.4) are labeled risks of tirzepatide; alcohol is an independent risk factor for both, with pancreatitis particularly tied to binge drinking^[1][8].
• Patients with active alcohol use disorder should talk to the prescriber. The emerging evidence suggests Zepbound may help, but tirzepatide-specific RCT data in AUD are not yet available and this is not an FDA-approved indication. Naltrexone, acamprosate, and disulfiram remain the standard of care.

For the practical timing and dose framework (when in the titration ladder GI tolerability is at its worst, why the first 4 weeks and the week after each escalation matter most), see How quickly does tirzepatide work for weight loss? For week-by-week side-effect expectations, the GLP-1 side-effect timeline tool gives a structured view.

How this article differs from the FDA-label deep-dive

This article is the evidence-ledger summary: what the randomized trials, observational cohorts, and preclinical studies actually show about Zepbound (tirzepatide) and alcohol. For the section-by-section walkthrough of the Zepbound FDA label, the SURMOUNT adverse-event tables that govern the practical risk-stacking with alcohol, and the dual-GIP-plus-GLP-1 gastric-emptying pharmacology, see the companion deep-dive Zepbound and Alcohol: Honest Evidence Review of FDA Label, SURMOUNT Data, and AUD Trials. For the cross-class GLP-1 plus AUD context (semaglutide RCTs, exenatide history, the in-progress SEMALCO trial, and how GLP-1 compares to the three FDA-approved AUD medications) see GLP-1 Drugs and Alcohol Use Disorder. For the tirzepatide-specific drinking guidance article see Can you drink alcohol while taking tirzepatide?

Verdict

The Zepbound FDA label does not prohibit alcohol and does not identify a drug-drug interaction with ethanol. The trial evidence for whether tirzepatide reduces alcohol consumption in adults with alcohol use disorder is thinner than the social-media narrative suggests: one Phase 2 RCT exists in the GLP-1 plus AUD space (Hendershot 2025) and it used semaglutide, not tirzepatide. The tirzepatide-specific human evidence is limited to a single observational study (Quddos 2023), preclinical rat-model evidence (Windram 2025), and one target-trial-emulation cohort (Henney 2026). All three signals point in the same direction as the semaglutide literature but none is RCT-strength for tirzepatide. For most adults on Zepbound for weight management, moderate drinking is permitted by the label; patients with active alcohol use disorder, a history of pancreatitis or gallbladder disease, or concurrent insulin or sulfonylurea use should have an explicit conversation with their prescriber.

This article summarizes published evidence and does not constitute medical advice. Decisions about alcohol use while on Zepbound should be made with the prescribing clinician, particularly for patients with type 2 diabetes on insulin or a sulfonylurea, a history of pancreatitis or gallbladder disease, or a history of alcohol use disorder.