GLP-1 Drugs Reduce Alcohol Consumption, But Similar Diabetes Drugs Do Not | Research

The takeaway

A large VA study of over 15 years of health records plus animal experiments confirmed that GLP-1 receptor agonists significantly reduce alcohol drinking, while DPP-4 inhibitors — which also raise GLP-1 levels — have no effect on alcohol intake.

P<0.0001 in hazardous drinkers
GLP-1 RA users with baseline hazardous drinking showed a 1.00-point greater reduction in AUDIT-C scores versus DPP-4I users, with no effect from DPP-4 inhibitors at all

What the researchers found

GLP-1RA recipients showed significantly greater reductions in AUDIT-C drinking scores compared to both unexposed individuals (DiD: 0.09, P=0.0025) and DPP-4I recipients (DiD: 0.11, P=0.0002). The effects were dramatically larger in people with baseline alcohol use disorder (GLP-1RA vs DPP-4I: DiD 0.65, P<0.0001) and hazardous drinking (GLP-1RA vs DPP-4I: DiD 1.00, P<0.0001).

DPP-4 inhibitor recipients showed no difference from unexposed individuals. Animal experiments confirmed this: neither linagliptin nor omarigliptin reduced binge-like drinking in mice or alcohol self-administration in alcohol-dependent rats, despite successfully lowering blood glucose — confirming target engagement. This convergent human-animal evidence indicates the alcohol-reducing effect requires direct GLP-1 receptor agonism.

Why it matters

Alcohol use disorder affects millions of people and has very few effective medications. This is among the strongest evidence to date that GLP-1 drugs could be repurposed for AUD treatment. The finding that DPP-4 inhibitors don't work — despite also raising GLP-1 levels — is mechanistically important: it tells us the alcohol-reducing effect requires pharmacological-level receptor activation, not just modest increases in endogenous GLP-1.

How the study worked

The study combined a large retrospective cohort analysis using 2008-2023 VA electronic health records with reverse translational animal experiments. The human component compared propensity-score-matched groups of GLP-1RA recipients, DPP-4I recipients, and unexposed individuals using changes in AUDIT-C scores. The animal component tested two DPP-4 inhibitors (linagliptin and omarigliptin) in established mouse binge-drinking and rat alcohol self-administration models.

What this study cannot tell us

The human component is observational (not a randomized trial), so confounding factors may influence results despite propensity-score matching. AUDIT-C is a self-reported measure of drinking, which can be subject to reporting bias. The VA population is predominantly male and older, which may limit generalizability. The study could not determine optimal GLP-1RA dosing for alcohol reduction or whether the effect persists after stopping the medication.

How to read the evidence

This is a strong study combining a large human cohort analysis (VA records spanning 15 years) with confirmatory animal experiments, published in one of the top medical journals (JCI). While the human data is observational rather than from a randomized trial, the convergent cross-species evidence and rigorous methodology make this compelling evidence.

When this study was published

Published in 2025, this is a cutting-edge study directly informing the rapidly evolving field of GLP-1 drugs for addiction treatment.

The bigger picture

This study published in the Journal of Clinical Investigation represents a major step toward repurposing GLP-1 drugs for addiction medicine. It builds on preclinical evidence that semaglutide reduces alcohol intake in animals and provides the largest clinical evidence base to date. The mechanistic dissection — showing that DPP-4 inhibitors don't work — narrows the therapeutic target and could guide future drug development. Multiple clinical trials of GLP-1 drugs for AUD are now underway.

Questions still open

What is the optimal GLP-1 RA dose and duration for treating alcohol use disorder specifically?
Does the alcohol-reducing effect persist after GLP-1 RA therapy is discontinued?
Which GLP-1 RA (semaglutide, liraglutide, tirzepatide, etc.) is most effective for reducing alcohol consumption?

Common questions

Can I ask my doctor for a GLP-1 drug to help me drink less?

GLP-1 drugs are not yet FDA-approved for alcohol use disorder, but clinical trials are underway. This large study provides strong evidence that they reduce drinking, especially in people with AUD or hazardous drinking. If you're already being considered for a GLP-1 drug for diabetes or weight loss and also struggle with alcohol, discussing this potential added benefit with your doctor is reasonable.

Why don't DPP-4 inhibitors reduce drinking if they also increase GLP-1?

DPP-4 inhibitors work by preventing the breakdown of GLP-1, which modestly raises its levels. But GLP-1 receptor agonists provide much stronger, sustained receptor activation. This study shows that the modest GLP-1 increase from DPP-4 inhibitors isn't enough to affect alcohol consumption — you need the pharmacological-strength receptor activation that only direct GLP-1 RA drugs provide.

Read the original research

Glucagon-like peptide-1 receptor agonists, but not dipeptidyl peptidase-4 inhibitors, reduce alcohol intake.

The Journal of clinical investigation, 135(9)

View on PubMed ↗Publisher’s paper ↗

Citation

Farokhnia, Mehdi; Tazare, John; Pince, Claire L; Bruns, Nicolaus; Gray, Joshua C; Lo Re, Vincent; Fiellin, David A; Kranzler, Henry R; Koob, George F; Justice, Amy C; Vendruscolo, Leandro F; Rentsch, Christopher T; Leggio, Lorenzo. (2025). Glucagon-like peptide-1 receptor agonists, but not dipeptidyl peptidase-4 inhibitors, reduce alcohol intake.. The Journal of clinical investigation, 135(9). https://doi.org/10.1172/JCI188314