Can GLP-1 Medications Treat Addiction? What the Research on Ozempic Shows

Why would a GLP-1 affect addiction at all?

Because GLP-1 receptors are not confined to the gut and pancreas. They also appear in brain regions that govern reward and motivation.

GLP-1 stands for glucagon-like peptide-1, a hormone the gut releases after eating. It prompts insulin release and signals fullness. GLP-1 receptor agonists, the drug class that includes semaglutide, imitate that hormone more strongly and for far longer.

Semaglutide is sold as Ozempic and Wegovy. Tirzepatide, sold as Mounjaro and Zepbound, works on the GLP-1 receptor and on a second gut-hormone receptor, GIP. Liraglutide is an older, once-daily member of the class. The two randomized trials described below used semaglutide, and the large veterans study covered the class as a whole, so whether tirzepatide behaves the same way in addiction is an open question rather than a safe assumption.

Those regions are part of the brain's reward system, where the neurotransmitter dopamine helps signal that something is worth pursuing. Addictive substances act on that system. The working hypothesis is that GLP-1 receptor activity dampens the response: in animal studies, GLP-1 receptor agonists reduce the dopamine release triggered by alcohol and other drugs, and reduce how much of those substances the animals will self-administer. If that holds up, one drug class could in principle act on several addictions through a shared pathway.

That is the extent of what the mechanism work supports: a plausible pathway seen in animals, which the trial literature describes as preclinical evidence. It has not been demonstrated in humans, and rodent findings routinely fail to reproduce in people.

Does Ozempic reduce alcohol cravings?

Early randomized trials suggest semaglutide can reduce alcohol craving and heavy drinking, but it is not an approved treatment for alcohol use disorder.

In the two trials run so far, craving and heavy drinking both fell on semaglutide, the drug in Ozempic and Wegovy. Both were small, so this is an early signal in the populations studied, not evidence that it works as a treatment.

In 2025, JAMA Psychiatry published a phase 2 randomized controlled trial led by Christian Hendershot. Forty-eight adults with alcohol use disorder, none of them seeking treatment, received low-dose semaglutide or placebo for nine weeks, with neither participants nor researchers knowing which. In a laboratory drinking session after treatment, the semaglutide group consumed less alcohol, with medium-to-large effect sizes. Over the nine weeks they also reported less craving, drank less on the days they drank, and showed greater reductions in heavy drinking over time. The drug did not change how many days they drank. Forty-eight people over nine weeks is enough to justify larger trials and not enough to change practice.

In May 2026, The Lancet published a longer one. At a single centre in Copenhagen, 108 adults with moderate to severe alcohol use disorder, all of them seeking treatment, were randomized to weekly semaglutide at 2.4mg or placebo for 26 weeks, on top of standard cognitive behavioral therapy for everyone. Both groups improved, which is what happens when everyone receives therapy. The semaglutide group improved more: the share of days that were heavy drinking days fell by 41 percentage points from baseline, against 26 on placebo, a difference of about 14 points. In a 30-day month that works out to roughly 12 fewer heavy drinking days versus roughly 8. The authors also report substantial effects on secondary alcohol-related and physical outcomes. Side effects were mostly mild to moderate gastrointestinal ones, transient, and more frequent on semaglutide.

One enrolment detail changes how that result should be read. Every participant had obesity as well as alcohol use disorder; that was a criterion for entry. The trial therefore tells us the medication helped that specific group. It cannot tell us whether people with alcohol use disorder and no obesity would respond similarly, better or worse.

What did the largest study find?

Lower rates of substance use disorders, and lower rates of overdose and death, across a very large population, in a study design that cannot establish cause.

In March 2026 The BMJ published an analysis by a team at Washington University in St. Louis using US Department of Veterans Affairs records, drawn from a base population of 606,434 veterans with type 2 diabetes and followed for up to three years.

Among veterans with no prior substance use disorder, starting a GLP-1 was associated with a 14% lower rate of developing one, including 18% lower for alcohol, 20% for cocaine and 14% for cannabis. Among veterans who already had a substance use disorder, GLP-1 use was associated with fewer substance-related emergency visits and hospital admissions, and with lower rates of drug overdose, substance-related death, and suicidal ideation or attempts.

Those mortality figures are also the ones that most require care, because this is an observational study. Nobody was randomized. Researchers compared veterans who started a GLP-1 with veterans who started a different diabetes drug, an SGLT-2 inhibitor, and used statistical weighting to make the two groups comparable. That is a stronger design than comparing users with non-users, but it still cannot rule out that the groups differed in ways that explain the gap. The population is narrow as well: veterans with type 2 diabetes, already in regular care.

Do GLP-1s help with nicotine, opioids or stimulants?

Less is known, and in two of the three cases we will not guess.

Nicotine. In the BMJ cohort, starting a GLP-1 was associated with a 20% lower rate of developing nicotine use disorder, and the JAMA Psychiatry trial found an exploratory signal of fewer cigarettes per day among the participants who smoked. Our source set does not include a completed randomized smoking-cessation trial, so we are treating this as unresolved.

Opioids. The BMJ cohort found a 25% lower rate of developing opioid use disorder, the largest reduction for any substance in that study, and lower overdose risk among veterans who already had a substance use disorder. Those are observational findings. We could not source a completed randomized outpatient trial in people with opioid use disorder, and outpatient is where the difficult part of opioid treatment lives. This is a gap, not a negative result.

Stimulants. The BMJ cohort found a 20% lower rate of developing cocaine use disorder. That is a finding about prevention, not treatment; our source set does not include a completed trial in people who already have a stimulant use disorder, and preventing a habit from forming and dismantling an established one are different problems. "GLP-1s work on addiction" is too coarse a claim to be useful until that distinction is tested.

The evidence at a glance

Every row is off-label. No GLP-1 medication is approved for any of these uses.

What are the risks and the limits?

The medications carry the same risks they carry for any other use, and the off-label status is the central limit.

No GLP-1 medication is approved by the FDA to treat any addiction. As of August 2026, Ozempic is indicated for glycemic control in type 2 diabetes and for cardiovascular and kidney outcomes in adults with type 2 diabetes. Wegovy is indicated for weight reduction, for cardiovascular risk reduction in adults with obesity or overweight and established cardiovascular disease, and, as an injection, for noncirrhotic MASH with moderate to advanced fibrosis. The two labels are not interchangeable, and neither mentions addiction. Off-label prescribing is legal, but it means no regulator has assessed this use.

On the Ozempic label, the most common adverse reactions reported in patients treated for type 2 diabetes are nausea, vomiting, diarrhea, abdominal pain and constipation. The labels also carry warnings for uncommon but serious events including acute pancreatitis and acute gallbladder disease. The labels state these warnings and the contraindications that rule some people out for the drug itself, without limiting them to a particular indication.

The evidence base is also young. Two modest randomized trials and one large observational study make a promising start rather than a foundation, and effects appear to vary by substance and by group.

Bottom line

GLP-1 medications are not approved to treat addiction. Early randomized evidence suggests semaglutide may reduce alcohol craving and heavy drinking, but larger trials are needed before it can be considered an established treatment.

Addiction has resisted pharmacology for a century. There are a handful of medications for alcohol and opioid use disorder, all underused, and essentially nothing for stimulants. A drug class that reduced craving across substances would matter, which is exactly why it deserves scrutiny rather than enthusiasm. Until that evidence arrives, interest is the appropriate posture, not conviction.

Measured is a weight-management practice. We prescribe GLP-1 medications within their approved indications, and we do not prescribe them to treat addiction. If you are struggling with alcohol or another substance, the useful step is a conversation with a clinician who knows your history, alongside the treatments that already have evidence behind them.

This article is educational and is not medical advice. It does not recommend any medication or dose for any purpose. Talk to a qualified healthcare professional about your own situation.