Nobody told them it would work. Nobody was studying that. The patients just started noticing that the urge to drink was gone, the cigarettes had lost their pull, and the craving that had followed them for years had gone quiet. Their doctors were prescribing Ozempic for blood sugar. Something else was happening.
What researchers at Washington University School of Medicine found when they went looking for an explanation has become one of the most significant surprises in addiction medicine in a generation. In an analysis of more than 600,000 US veterans with type 2 diabetes, the WashU Medicine team discovered that GLP-1 receptor agonists, the drug class behind Ozempic, Wegovy, Mounjaro, and Zepbound, reduced the risk of developing substance use disorders across every major addictive substance studied, including alcohol, cannabis, cocaine, nicotine, and opioids. Not one. All of them. The results were published on March 4, 2026, in The BMJ. The word Dr. Ziyad Al-Aly used to describe the finding was “revelation.”
The Drug That Was Never Supposed to Touch Addiction
The GLP-1 class of medications did not begin as a weight loss revolution. They began as a treatment for type 2 diabetes, a way to stimulate insulin release, slow digestion, and help the body manage blood sugar. Ozempic arrived in 2017. Wegovy, the higher-dose version approved specifically for obesity, came in 2021. What followed was one of the fastest adoptions of any drug class in modern pharmaceutical history, driven largely by patients reporting dramatic weight loss and, strangely, a quieting of what they described as “food noise,” the constant mental chatter about eating that drives overeating.
The addiction signal emerged from the edges. Patients began telling their doctors they had stopped drinking without really trying. Others found they had no interest in cigarettes. Some reported that the pull of substances they had struggled with for years had simply diminished. These were anecdotes. Medicine does not run on anecdotes.
So Al-Aly and his collaborators at the VA Saint Louis Health Care System ran the numbers. They pulled the electronic health records of 606,434 US veterans with type 2 diabetes, split them into those taking GLP-1 drugs and those taking a different class of diabetes medication called SGLT2 inhibitors, and followed them for up to three years. The comparison was designed to isolate the effect of the GLP-1 medication itself. What they found was not subtle.
The Numbers That Stopped the Room
Compared to patients on non-GLP-1 diabetes medications, those taking GLP-1 drugs were 14 percent less likely to develop any new substance use disorder over the following three years. Break that number down by substance, and the picture becomes even sharper. Risk of developing alcohol use disorder fell by 18 percent. Cannabis by 14 percent. Cocaine and nicotine each by 20 percent. Opioids by 25 percent.
For patients who already had a substance use disorder when the study began, the results were more striking. After three years on a GLP-1 medication, emergency department visits related to substance use dropped by 30 percent. Hospitalizations fell by 25 percent. Overdose dropped by 40 percent. Drug-related deaths fell by 50 percent.
Translated into human terms: for every 1,000 people taking a GLP-1 drug for diabetes, seven fewer people developed a new substance use disorder. Twelve fewer experienced a serious harm event related to addiction. These are not marginal statistical improvements. In a country where the opioid crisis has reshaped the economic and social fabric of entire regions, a 50 percent reduction in drug-related deaths from a medication already being prescribed to millions of people is not a footnote. It is a headline that has not yet been heard loudly enough.
Why One Drug Works on Everything
The mechanism matters. In addiction medicine, treatments have historically been substance-specific. A nicotine patch helps with smoking. Naltrexone targets alcohol and opioids. Methadone is for opioid use disorder. There is no approved medication that works across all addictive substances. The pharmaceutical logic of addiction treatment has always been: find the receptor, block or activate it, and address that specific drug.
What the GLP-1 finding suggests is something different. GLP-1 receptors exist not just in the pancreas and gut, where they manage blood sugar, but in the brain, specifically in regions that govern reward processing, motivation, and craving. The hypothesis Al-Aly has advanced is that these drugs are not acting against alcohol, opioids, or nicotine specifically. They are acting against the craving itself. The underlying biological signal that drives addiction across substances, whatever that common pathway is, appears to be something GLP-1 drugs can dampen.
“Moving beyond food noise to drug noise,” Al-Aly said in the WashU Medicine release, “GLP-1s are quieting the roar of addiction.”
The published study appears in The BMJ, one of the world’s oldest and most rigorously peer-reviewed medical journals. The research was funded by the United States Department of Veterans Affairs.
What This Means for 48 Million Americans
The scale of the addiction crisis in the United States does not require elaboration for anyone who has lived through the last two decades. Roughly 48 million Americans met the criteria for a substance use disorder in 2022, according to federal data. Opioids alone killed more than 80,000 people in 2023. Alcohol use disorder affects an estimated 29 million adults. Tobacco remains the leading preventable cause of death in the country. The existing treatment infrastructure, detox programs, behavioral therapy, and medication-assisted treatment reach a fraction of those who need them. Dr. Lorenzo Leggio of the National Institute on Drug Abuse has noted that 98 percent of people with alcohol use disorder do not receive any of the FDA-approved medications for the condition.
GLP-1 drugs are already being prescribed to tens of millions of Americans for diabetes and obesity. If the addiction-reducing effect holds up in randomized controlled trials, which several research groups, including the VA itself, are now actively conducting, the implications extend far beyond adding a new line to a drug’s label.
The same pharmaceutical market structures that have kept life-saving medications out of reach for millions of patients will shape who can access GLP-1 drugs for addiction treatment and at what price. Ozempic costs roughly $900 per month without insurance in the United States. Wegovy runs higher. The patients most likely to benefit from GLP-1 addiction treatment, low-income, uninsured, in the communities most devastated by opioids, are precisely the patients least likely to afford it.
The Steelman: Why This Is Not a Cure
The study has real limitations that honest reporting requires acknowledging.
This is an observational cohort study, not a randomized controlled trial. The veterans in both groups were not randomly assigned to their medications; they were prescribed them for clinical reasons that may have differed in ways the analysis could not fully control for. Patients prescribed GLP-1 drugs may be systematically different from patients prescribed SGLT2 inhibitors in ways that affect addiction risk independent of the medication.
Lorenzo Leggio of the National Institute on Drug Abuse said explicitly that the findings are not sufficient to justify choosing GLP-1 drugs over currently approved addiction treatments. The Science.org coverage of the study noted that at least one separate trial using a different GLP-1 drug did not show a significant effect on alcohol use. The cross-substance signal is striking, but AAAS researchers emphasized that randomized controlled trials remain necessary before clinical guidelines should change.
None of that diminishes what 606,000 veterans just told us. It means the signal is real enough to require urgent investigation, not cautious waiting.
The Quiet That Nobody Expected
Al-Aly’s framing of the finding is worth sitting with. Patients taking GLP-1 drugs for food describe a quieting of noise. The persistent, intrusive mental preoccupation with eating, the craving that overrides intention, goes quiet. What the study of 600,000 veterans suggests is that the same quieting applies to drug cravings. That the noise addiction makes the pull toward the bottle, the cigarette, the pill runs through some of the same biological circuitry as the noise hunger makes.
If that is true, the implications reach beyond pharmacology. Addiction has been framed, culturally and politically, as a failure of will. The person who cannot stop drinking lacks discipline. The person who cannot quit opioids has made bad choices. That framing has shaped policy, criminal justice, and public sympathy for decades.
A drug that quiets craving across substances by acting on shared brain circuitry does not describe a moral failure. It describes a biological one that medicine may now have a tool to address. The breakthroughs arriving in medicine in 2026 are arriving faster than policy or public understanding can absorb them. This one, if it holds, will matter to more people than almost any other.
They went in for diabetes treatment. They came out without the craving that had defined their lives. Nobody planned it. Nobody predicted it. The data did not care.
