Harvard-Led Study Finds 44.2% of GLP-1 Users Report Drinking Less

Nearly half of adults who drank alcohol when they started taking GLP-1 drugs said they later drank less, according to a Harvard-led study that adds survey evidence to a pattern that had largely been described through patient anecdotes.

The study, published in JAMA Psychiatry in October, examined responses from a national panel of U.S. adults and found that 44.2% of GLP-1 users who were drinkers when they began the medication reported lower alcohol consumption after starting treatment. Most of the rest said their drinking had not changed, and three people said they drank more.

The findings come as GLP-1 medicines, which are used to treat diabetes and obesity, continue to gain wide use. Their spread has drawn attention not only because of weight loss effects, but also because some patients have reported changes in cravings and other behaviors. The new research does not prove that the drugs reduce alcohol use, but it offers a broader look at how patients describe their own experience.

Researchers from Harvard Medical School and McLean Hospital worked on the survey with collaborators from Brown University, Duke University, Brandeis University and the RAND Corporation in Boston. The work was supported by the National Institute on Alcohol Abuse and Alcoholism.

The survey used RAND’s American Life Panel, a group of adults across the country who are provided internet-connected devices to complete research questionnaires. Researchers surveyed participants in March and April. Of 2,009 people who completed the initial survey, 368 said they currently used or had previously used a GLP-1 drug. Among them, 242 said they drank alcohol when they started taking the medication.

The researchers also looked more closely at respondents considered at risk for unhealthy alcohol use. To identify that group, they used a single screening question asking whether, at any one time in the previous year, the respondent had consumed enough alcohol to meet a heavy-drinking threshold defined by the National Institutes of Health: four drinks for women or five drinks for men. In that subgroup of GLP-1 users, 54.1% said they were drinking less after starting the medication, 40.2% reported no change and 5.7% — three individuals — said they were drinking more.

Haiden Huskamp, a senior author of the study and the Henry J. Kaiser Professor of Health Care Policy at Harvard Medical School, said the result stood out because earlier evidence had been anecdotal. Alisa Busch, a co-author and professor of psychiatry at McLean Hospital who is also an associate professor of health care policy at Harvard Medical School, said the research was not designed to test whether GLP-1 drugs work as a treatment for alcohol-related problems.

That distinction is central to how the study should be read. The researchers said they were not testing the drugs for efficacy in reducing alcohol use, and they said a randomized controlled trial would be needed to answer that question. The survey was observational and based on self-reports. It also did not measure the size of any decline in drinking, only whether respondents believed they were drinking less, more or about the same.

The authors also cautioned that the study screened for potentially unhealthy alcohol use, not alcohol use disorder, which is a clinical diagnosis tied to significant harm or disruption in a person’s life. That means the findings should not be taken as evidence that GLP-1 medicines should be added to the set of approved treatments for alcohol use disorder.

The study grew out of earlier work by Busch and Huskamp on the underuse of medications that are already approved to treat alcohol use disorder. The researchers noted that such drugs have been available for decades, including disulfiram, which has been in use since the 1940s. Even so, uptake has remained low. In a national survey cited by the researchers, only 3% to 5% of people with a past alcohol abuse disorder reported using those medications, and only 10% of people seeking treatment had received approved drugs.

That background helps explain why reports about GLP-1s and alcohol drew the team’s attention. If a widely used class of metabolic drugs turns out to affect drinking behavior, the implications could reach beyond addiction medicine. Over time, and only if clinical trials confirm a real effect, lower alcohol consumption among some GLP-1 users could alter demand patterns for beer, wine and spirits. For beverage companies, distributors and retailers, that raises the possibility of a structural shift in consumption tied to a medication trend rather than to prices, marketing or broader economic conditions. The current study does not show such a market effect, and it does not quantify how much less alcohol respondents may have consumed.

Another finding in the study was how little public awareness there appears to be around the idea. Only 10% of all respondents said they had heard that GLP-1 drugs might reduce drinking. That suggests patient reports and discussion within medical circles have not yet translated into broad public understanding.

The researchers said stronger evidence may not be far off. Clinical trials are already underway to test whether GLP-1 drugs have a measurable effect on alcohol use. Those trials are expected to address questions the survey could not answer, including causation, the degree of change in consumption and whether any effect is large enough to matter in clinical practice.