A national survey found that 44.2% of drinkers taking GLP-1 medications reported drinking less, though clinical trials are needed to assess the effect.
Nearly half of drinkers taking GLP-1 medications said they cut back on alcohol, according to a national survey by researchers at Harvard Medical School and McLean Hospital. The finding adds population-level evidence to reports of reduced drinking among people taking the drugs for diabetes or obesity. It doesn’t establish that GLP-1s cause the change, or show how large the reported reductions were. That distinction matters: a pattern in patient reports can help frame a research question, but it can’t answer whether the medication itself changed drinking behavior.
What the survey found
The study, published in October in JAMA Psychiatry, drew on RAND’s American Life Panel. Researchers surveyed 2,009 adults in March and April 2026. Among them, 368 reported current or past GLP-1 use, and 242 said they drank alcohol when they started taking the medication. The last group is the most relevant for the drinking comparison, since respondents who weren’t drinking at the start couldn’t report cutting back from that baseline.
Of those 242 respondents, 106, or 44.2 percent, reported drinking less while on GLP-1s. Most of the others reported no change; three said they drank more. Among respondents classified as at risk for unhealthy drinking, 54.1 percent reported a decline, 40.2 percent reported no change, and 5.7 percent—three people—said they drank more. The pattern is notable, especially among the higher-risk group, but the figures describe what participants recalled and reported. They don’t show how many drinks people stopped having, how long any change lasted, or whether it affected health.
A signal, with limits
The survey used one screening question to identify possible unhealthy alcohol use: whether participants had consumed enough at one time in the past year to meet the heavy-drinking threshold—four drinks for women and five for men, as defined by the National Institutes of Health. That screen can flag a reason for closer attention, but it doesn’t diagnose alcohol use disorder, a clinical condition involving significant disruption to health or daily life. The distinction matters because a single episode above a threshold and a persistent disorder aren’t interchangeable measures.
Nor did the survey measure how much respondents reduced their drinking. The researchers asked about people’s reported experience, not whether GLP-1 treatment reduces alcohol consumption under controlled conditions. “We’d need a randomized control trial for that,” said Alisa Busch, a study co-author and professor of psychiatry at McLean Hospital. Without that comparison, other changes around the time someone starts medication—such as changes in routine or health priorities—can’t be separated from a drug effect.
“Finding that about half of GLP-1 users who drink reported reductions in their alcohol use after starting the medication is pretty striking,” said senior author Haiden Huskamp, the Henry J. Kaiser Professor of Health Care Policy at Harvard Medical School.
That’s a fair description of the survey’s headline. It’s also a reason to be precise about what the result means: almost half reported a reduction, while most of the remainder reported no change and a small number reported drinking more. The result points to a question worth testing, rather than settling it.
The researchers caution that the results don’t establish GLP-1s as a treatment for alcohol use disorder. Their study focused on self-reported drinking changes, including among people who screened positive for possible unhealthy use; it wasn’t designed to test the drugs’ effectiveness as treatment. A medication’s possible effect on cravings, if confirmed, would still need to be distinguished from evidence that it can treat a diagnosed disorder.
From treatment access to unexpected effects
The research grew out of work by Huskamp and Busch on the low use of approved medications for alcohol use disorder. Although medications such as disulfiram have been available for decades, only 3 to 5 percent of people with a past alcohol abuse disorder reported using medication in a recent national survey. Among people seeking treatment, just 10 percent had received an approved drug. Those figures describe a gap between available medication and reported use; they don’t explain why individuals do or don’t receive it.
“Far and away most people with alcohol-use disorder receive no treatment, period,” Busch said. Her earlier work examined which patient and clinician characteristics might help explain whether these medications are prescribed. That background helps explain the researchers’ interest in a possible effect from drugs prescribed for other conditions: an unexpected association could point toward new questions about treatment, even before it supports any change in care.
As accounts accumulated of people on GLP-1s experiencing fewer cravings to drink, the researchers wanted to see whether those reports appeared across a broader group. The survey offers an early measure of how patients describe their drinking, while leaving questions about cause, scale and clinical relevance open. Here’s the thing: a self-reported pattern can make a clinical question harder to ignore, but it can’t tell patients or clinicians what to expect from treatment.
What comes next
Awareness also appears limited: only 10 percent of all respondents said they had heard that GLP-1s might reduce drinking. Several clinical trials are underway to examine the question. Their results should provide a clearer test of whether the survey’s reported pattern reflects an effect of the medication. If you’re working in this space, the key issue will be whether controlled research can measure a change in drinking and clarify which patients, if any, experience it.
This is more significant than it looks as a research signal, given the reported gap in treatment for alcohol use disorder. But the survey doesn’t justify treating GLP-1s as an established option for that condition. For now, the useful takeaway is narrower: many respondents who drank when they began GLP-1 treatment later said they drank less. A clearer answer will depend on the trials.
Still an open question.
This research was funded by the National Institute on Alcohol Abuse and Alcoholism.
Discussion
Sign in to join the discussion.