You Don’t Have to Want to Quit Drinking for This Drug to Help You Drink Less

Nobody has to hit zero to survive this. For most of American addiction treatment’s history, saying that sentence out loud in a treatment group could get you labeled in denial. Now it’s the plain-language summary of a randomized controlled trial published in one of medicine’s most conservative journals.

The recovery field spent ninety years insisting there was one destination: abstinence, full stop. The data increasingly says staying alive on the way there was always the actual point — and a growing set of tools finally treats it that way.

What the trial actually found

Researchers at Mental Health Center Copenhagen, working with NIH collaborators, ran a 26-week randomized, double-blind, placebo-controlled trial in 108 adults who had both alcohol use disorder and a BMI of 30 or higher — people seeking treatment, not people trying to lose weight as a side effect. Half got weekly 2.4mg injections of semaglutide, the same GLP-1 drug sold as Wegovy and Ozempic; half got placebo. Everyone also got up to ten sessions of cognitive behavioral therapy. The results, published in The Lancet and covered in an NIH release, showed the semaglutide group’s heavy-drinking days dropped by roughly 41 percentage points over the trial, compared to about 26–27 points on placebo. Total alcohol intake over a 30-day measurement window fell from roughly 78 ounces to about 23 ounces — a drop of more than 70% — corroborated by blood-alcohol biomarkers, not just self-report, which matters because self-reported drinking data is notoriously unreliable in exactly the direction that would make a company’s drug look better than it is.

Here’s the plain-language version of why this drug might work on drinking at all: GLP-1 medications were built to blunt appetite by dialing down reward signaling in the brain — the same dopamine circuitry that makes a second drink feel necessary instead of optional. Semaglutide doesn’t appear to know the difference between craving a plate of food and craving a drink, and that overlap, which started as a side observation in diabetes and weight-loss trials, is now its own field of study.

Dr. Joseph Schacht at the University of Colorado, quoted in coverage of the trial, said something that would have been treatment-group heresy a decade ago: this data suggests semaglutide may help people reduce heavy and harmful drinking even when they aren’t trying to quit alcohol entirely. That’s not a footnote. That’s the headline the field spent decades not being allowed to write.

The street-level version of the same idea

If a weekly injection is the high-tech version of “less still counts,” managed alcohol programs are the version that’s existed for years with far less funding and far more controversy. The model is blunt: instead of demanding abstinence from people with severe, long-standing alcohol use disorder — often people who are also homeless, who face seizures or delirium tremens if they stop drinking without medical supervision — a managed alcohol program provides regulated, measured doses of alcohol on a fixed schedule, alongside housing, healthcare, and social support. NEJM Catalyst’s overview of the model documents outcomes that read like a rebuke to the abstinence-only assumption: fewer emergency room visits, fewer police interactions, more stable housing, all without requiring the one thing traditional treatment insisted on first.

The two cities running this experiment right now are heading in opposite directions. San Francisco shut down its Managed Alcohol Program in January 2026 after it became a viral cost-efficiency target — roughly $5 million a year for 55 clients drew public criticism as an unsustainable per-person cost, and the mayor ended it. Meanwhile, in Thunder Bay, Ontario, the Kwae Kii Win Managed Alcohol Program just opened an expanded facility, run by NorWest Community Health Centres and Shelter House, betting the other direction: that the model is worth building out further, not winding down. Both cities are looking at the same evidence base and the same population. One read it as too expensive to keep. The other read it as too effective to cut.

Both cities are looking at the same evidence base and the same population.

We have been here before, and the field eventually caught up

This isn’t the first time addiction treatment has had to relitigate what counts as success. In the 1930s, Bill Wilson and a small group in Akron, Ohio built Alcoholics Anonymous around total, lifelong abstinence — a model that filled a genuine vacuum in a field with almost no other tools, and one that became the dominant framework for treating alcohol and later drug addiction across the country for the rest of the century. For decades, saying “moderation” instead of “sobriety” in a clinical setting could get you dismissed as not yet ready for real treatment. Moderation-management approaches existed on the margins as early as the 1950s but stayed there — under-published, under-funded, treated as a bargaining stage on the way to admitting abstinence was the only real answer. It’s only since the 1990s that research reducing drinking, not just eliminating it, became a legitimate measure of whether a medication or a program was working. Semaglutide’s trial data and Thunder Bay’s expanded managed-alcohol facility are both downstream of that shift — proof it’s still happening, not evidence it already finished.

The part that isn’t proven yet

None of this means every harm-reduction tool for alcohol has the same evidence behind it. The Sinclair Method — taking naltrexone about an hour before drinking to blunt the reward response, aiming to gradually extinguish the urge to drink over months — gets promoted heavily in recovery-adjacent corners of the internet as a gentler alternative to abstinence. It isn’t recommended as a first-line approach in any major evidence-based clinical guideline for alcohol use disorder, and the research behind it is thinner than its online popularity suggests. That’s worth saying plainly, in the same piece that’s making the case for harm reduction, because the credibility of “less can count as a win” depends on being honest about which tools have real trial data behind them and which ones are running on word of mouth.

If you’re in recovery and reading this, you already know the argument you’re not supposed to make out loud: that “I drank less this month than last month” should count for something, even if it isn’t the finish line anyone told you to aim for. The Lancet just ran a controlled trial that agrees with you. Thunder Bay is betting a building on it. Ninety years of one answer is a long time to wait for the field to say what a lot of people in recovery already knew.

Rize’s alcohol substance hub tracks this research as it develops; see our harm reduction coverage for the broader shift beyond alcohol.