Ozempic Wasn’t Built for Alcohol Use Disorder. New Data Says It’s Helping Anyway.

Researchers from Truveta, Johns Hopkins, Duke, and Providence pulled electronic health records for 40,703 adults who had both alcohol use disorder and either diabetes or obesity, then compared what happened to the ones prescribed a GLP-1 receptor agonist — semaglutide (Ozempic, Wegovy) or tirzepatide (Mounjaro, Zepbound) — against those on other medications for the same conditions. The results, published in BMJ Open: people on GLP-1 drugs had a 22-26% lower hazard of alcohol-related hospitalization compared to those on other diabetes medications, and a 32% lower hazard compared to those on other anti-obesity medications.

A drug class built to treat diabetes and obesity may be doing more for alcohol use disorder, almost by accident, than most of the medications actually designed for it.

GLP-1 receptor agonists work by mimicking a gut hormone that signals fullness to the brain — that’s why they’re so effective for weight loss. The same brain circuitry involved in that fullness signal, it turns out, overlaps significantly with the reward pathways that drive craving for alcohol and other substances, which is the biological plausibility behind why a drug for blood sugar and appetite might also dial down the urge to drink. This isn’t the first hint of that connection — animal studies and smaller human trials have pointed the same direction for a couple of years — but 40,703 people is a large enough real-world sample to take seriously, not just file under “interesting.”

Study co-author Hemalkumar B. Mehta of Johns Hopkins framed the stakes plainly: alcohol use disorder, he noted, remains one of the most common and most undertreated chronic conditions in the country — a point SAMHSA’s own newly released 2025 survey underscores, counting 25.7 million Americans who met criteria for alcohol use disorder last year, the overwhelming majority of whom will never see a dedicated AUD medication. Co-author Jay B. Lusk of Duke was the one urging caution on methodology — this is observational data, drawn from people who were prescribed GLP-1 drugs for diabetes or weight loss, not from a randomized trial designed to test alcohol outcomes specifically, so it can’t yet prove the drugs caused the drop rather than correlating with people who were already managing their health more closely.

What it can do is give providers a real, low-risk reason to pay attention: for a patient already on a GLP-1 drug for diabetes or weight management who also has alcohol use disorder, this data is a reason to ask about drinking patterns at the next visit rather than treating the two conditions as unrelated boxes on a chart. A dedicated randomized trial testing GLP-1s specifically for alcohol use disorder is the obvious next step. But the population that would benefit most doesn’t have to wait for it — millions of them are already taking the drug for something else.