GLP-1 drinkers report lower alcohol use in Harvard survey

GLP-1 users who drank alcohol reported lower consumption in a Harvard-led survey published by JAMA Psychiatry in October, with 44.2% of 242 drinkers saying they drank less.

The finding matters because approved alcohol-use disorder medicines remain underused: Harvard Health Publishing said 3% to 5% of people with a past disorder reported using them, while 54.1% of GLP-1 users who screened at risk for unhealthy drinking reported drinking less.

Researchers from Harvard Medical School, McLean Hospital, Brown, Duke, Brandeis and RAND surveyed RAND’s American Life Panel in March and April 2026. The study was supported by the National Institute on Alcohol Abuse and Alcoholism.

“We wanted to really understand patient perception of this issue, and a large share of people taking GLP-1s are saying they’re drinking less,” Haiden Huskamp, senior author and Harvard Medical School professor of health care policy, said in the Harvard article.

The initial survey had 2,009 respondents. Among them, 368 reported current or previous use of GLP-1 medications, which Harvard described as drugs used to treat diabetes and obesity, and 242 said they drank alcohol when they started taking a GLP-1.

Among those 242 drinkers, 106 said their alcohol use fell. Most others reported no change, and three people said they drank more, according to the Harvard account of the JAMA Psychiatry study.

Researchers also asked one screening question for unhealthy alcohol use: whether respondents had, at any one time in the previous year, consumed enough alcohol to meet the National Institutes of Health threshold for heavy drinking, defined as four drinks for women and five for men.

In that at-risk group, 54.1% reported drinking less, 40.2% reported no change and 5.7%, or three people, reported drinking more. The study applies to adults in this survey who used GLP-1s and were drinkers at the start, not to nondrinkers or to all people with alcohol-use disorder.

The study cannot show that GLP-1 drugs caused lower alcohol intake. It was a self-reported survey, not a randomized controlled trial, and Harvard said the analysis did not measure how large the drinking reductions were.

Huskamp and Alisa Busch, a Harvard Medical School professor of psychiatry at McLean Hospital, cautioned that the survey should not be read as evidence that GLP-1s should be used as approved treatments for alcohol-use disorder. The study screened for possible unhealthy use, not for alcohol-use disorder, which is a clinical diagnosis.

“We weren’t testing GLP-1s for efficacy in reducing alcohol use. We’d need a randomized control trial for that,” Busch said in the Harvard article. “We were trying to understand, in a population at higher risk for harmful drinking, is there a difference in their experience of drinking alcohol compared to individuals who screen negative.”

Harvard said awareness of a possible alcohol-related effect was low: 10% of all survey respondents said they had heard that GLP-1s might reduce drinking. Huskamp said clinical trials are underway and that more evidence may emerge over the next couple of years, but no dated trial readout was provided.

Source: peer-reviewed study in JAMA Psychiatry and Harvard Health Publishing / Harvard Medical School, Nov. 2, 2026.

Atlas360 reports health research and does not provide medical advice.