Study links tirzepatide and semaglutide to fewer alcohol use disorder diagnoses

Adults with type 2 diabetes who started tirzepatide (sold as Mounjaro and Zepbound) had a 53 percent lower rate of a first diagnosis of alcohol use disorder than similar patients who started a DPP-4 inhibitor, according to a study of electronic health records from more than 120 million US patients ^[1][2][3]. It was the largest reduction among four incretin-based drugs tested, and the authors say randomized evidence is needed before the drugs could be used for that purpose ^[1][3].

The study, led by researchers at the University of Liverpool, emulated four randomized trials. Patients who started tirzepatide, semaglutide (Ozempic, Wegovy, Rybelsus), liraglutide or dulaglutide were each compared with patients who started a DPP-4 inhibitor, a different class of diabetes medicine used as the reference group ^[1][2][3]. Tirzepatide is a dual GLP-1/GIP receptor agonist; the other three are GLP-1 receptor agonists ^[1].

All patients had type 2 diabetes and no prior alcohol use disorder diagnosis. Each drug group was matched 1:1 with comparison patients on characteristics that could affect the results. Researchers then looked for a first-ever alcohol use disorder diagnosis, identified by ICD-10 code F10, during 18 months of follow-up ^[1][2][3].

The matched groups were 7,165 patients on tirzepatide, 20,198 on semaglutide, 6,565 on liraglutide and 19,061 on dulaglutide, each paired with a DPP-4 inhibitor patient ^[1][2][3].

Tirzepatide was associated with a hazard ratio of 0.47 (95 percent confidence interval, 0.29 to 0.75) for a first alcohol use disorder diagnosis compared with DPP-4 inhibitors. Semaglutide was associated with a hazard ratio of 0.68 (0.52 to 0.89). Liraglutide and dulaglutide were not associated with a significant reduction ^[1][2][3].

In head-to-head comparisons, the paper reports tirzepatide having a significant advantage over liraglutide, with a hazard ratio of 0.47 (0.24 to 0.92) ^[1][3]. Other head-to-head results are not stated in the abstract.

Limits of the evidence

This is an analysis of medical records, not a randomized trial, so it can show an association but cannot prove that the drugs caused the difference in diagnoses ^[1][2][3]. Diagnoses were identified from billing codes, so people who drink heavily but were never diagnosed or coded would not be counted.

All patients in the study had type 2 diabetes. Whether the same pattern would appear in people who take these drugs only for weight loss has not been established by this study ^[1][2][3].

The paper's authors call for "robust randomised, controlled evidence" for what they describe as a novel indication, and the paper does not report any such trial ^[1][3]. The paper also lists industry ties among its authors, including a co-author employed by TriNetX LLC and others reporting grants, honoraria or consulting work for Novo Nordisk, AstraZeneca, Boehringer Ingelheim, Sanofi and other companies ^[3].

Why it matters for patients

Alcohol use disorder is not an approved use of these medicines, and the study does not change any prescribing label. The paper describes it as a possible new use that still needs randomized evidence ^[1][3].

For people already taking tirzepatide or semaglutide for diabetes, the results suggest a possible additional benefit that had not been shown in a controlled trial. They do not show that these drugs treat existing alcohol use disorder, and the study did not measure drinking itself — only whether a first diagnosis was recorded ^[1][2][3].

Patients who have concerns about their drinking can raise them with a clinician; the study provides no basis for starting, stopping or switching a GLP-1 medicine for alcohol-related reasons ^[1][2][3].

What happens next

The paper was published online on October 7, 2025, and is scheduled for the January 2026 issue of the journal Diabetes, Obesity and Metabolism (volume 28, issue 1, pages 137 to 150) ^[1][2]. The paper does not list any planned or ongoing randomized trial of tirzepatide for alcohol use disorder, and no regulatory filing or label change related to alcohol use disorder has been reported ^[1][3].