People taking GLP-1 receptor agonists such as semaglutide for diabetes or obesity showed lower rates of new alcohol and opioid use disorder diagnoses across multiple large studies spanning U.S. veterans, a nationwide Danish registry, and electronic health records from 149 American health systems. A 26-week randomized trial found that adding semaglutide to cognitive behavioral therapy reduced heavy drinking days among participants with alcohol use disorder and obesity. Together, the findings have pushed researchers to ask whether drugs originally designed to manage blood sugar and body weight could become tools against addiction.
Why GLP-1 drugs and addiction risk demand attention now
Alcohol use disorder affects tens of millions of adults in the United States, and opioid overdoses remain a leading cause of preventable death. Standard treatments for both conditions, including medications such as naltrexone, acamprosate, methadone, and buprenorphine, help many patients but leave large gaps in care. Relapse rates remain high, a substantial share of people with addiction never receive any treatment at all, and side effects or stigma can limit uptake of existing medications.
The rapid growth in prescriptions for GLP-1 receptor agonists for weight loss and diabetes has created a natural experiment: millions of people are now taking these drugs, and their medical records can reveal patterns that clinical trials have not yet fully tested. Clinicians began to notice anecdotal reports of reduced interest in alcohol, nicotine, and even compulsive behaviors among patients using semaglutide or related medications. Those observations prompted researchers to mine large datasets for signals that GLP-1 drugs might influence substance use.
The central scientific question is whether GLP-1 agonists reduce cravings for alcohol and opioids by acting directly on the brain’s reward pathways, independent of their effects on appetite and weight. Animal research has long suggested that GLP-1 receptors are present in the mesolimbic dopamine system, the circuitry that drives desire for food, drugs, and other rewards. If the drugs dampen reward-seeking behavior broadly rather than just suppressing hunger, that would help explain why some people taking semaglutide report less interest in drinking or using opioids even when weight loss is not the primary goal. None of the large observational studies published so far measured dopamine activity or other direct markers of reward-circuit function in human participants, so the mechanism remains an inference drawn from preclinical work and patient-reported outcomes rather than a confirmed biological pathway.
Converging evidence from veterans, Danish registries, and U.S. health records
The strongest signal comes from several independent datasets that point in the same direction. A U.S. Veterans Affairs cohort published in The BMJ compared veterans with type 2 diabetes who started GLP-1 receptor agonists against those who started SGLT2 inhibitors, a different class of diabetes drug. The GLP-1 group had lower rates of incident alcohol and opioid use disorders as well as fewer overdoses. Because SGLT2 inhibitors also improve metabolic health but do not act on brain reward circuits, the comparison helps isolate the effect of GLP-1 receptor activation from the general benefits of better glucose control.
A nationwide Danish registry study covering 2009 through 2017 took a similar approach, comparing new users of GLP-1 agonists with new users of DPP-4 inhibitors. After adjustment for age, sex, comorbidities, and other factors, GLP-1 users had fewer alcohol-related hospital events. The researchers also ran a self-controlled case series analysis, which compares each patient’s own risk before and after starting the drug, reducing the chance that healthier or more motivated patients simply chose GLP-1 therapy and would have had fewer alcohol problems regardless.
Beyond these comparisons, other real-world analyses have focused on semaglutide specifically. One study using a large electronic health record network found that patients prescribed semaglutide had lower rates of both new and recurrent alcohol use disorder diagnoses over 12 months, even after propensity-score matching to balance baseline health characteristics. Another investigation drawing on Oracle Health Real-World Data from 149 U.S. health systems examined people with co-occurring alcohol- and opioid-related diagnoses between 2014 and 2024. That analysis incorporated overdose outcomes and stratified results by whether patients were already receiving medications for opioid or alcohol use disorder, attempting to separate the added contribution of GLP-1 therapy from existing addiction care.
A nested case-control study conducted within the All of Us Research Program further broadened the picture by including people with type 2 diabetes or obesity who used GLP-1 receptor agonists. In that cohort, GLP-1 exposure was associated with lower odds of new substance use disorder diagnoses, including both alcohol and opioid use disorder. The consistency across these datasets, each with different populations, comparison groups, and analytic designs, strengthens the case that the association is not simply an artifact of one health system or one set of statistical assumptions.
A randomized trial adds experimental weight
Observational studies, however carefully adjusted, cannot prove causation. To move beyond association, researchers have begun testing GLP-1 drugs directly in people with addiction. A randomized, double-blind, placebo-controlled trial registered as NCT05520775 and reported in The Lancet enrolled adults with moderate-to-severe alcohol use disorder plus obesity. Participants received either semaglutide, titrated to 2.4 mg weekly, or placebo, with all participants also attending manualized cognitive behavioral therapy for 26 weeks.
According to the National Institutes of Health summary, those assigned to semaglutide had a greater reduction in heavy drinking days than those receiving CBT plus placebo. Secondary endpoints, including standardized craving scales, Alcohol Use Disorders Identification Test (AUDIT) scores, and liver enzyme measures, also tended to favor semaglutide. Adverse events resembled the known gastrointestinal side effects seen when the drug is used for obesity.
Crucially, the trial population all had obesity in addition to alcohol use disorder. That design choice reflects the drug’s current approval status and safety data but leaves open the question of whether semaglutide would help people with AUD who are not living with obesity. It also does not address whether lower doses than those used for weight management might still confer anti-craving benefits.
What mechanisms might explain the signal?
A recent review in Frontiers in Psychiatry outlines several plausible mechanisms by which GLP-1 receptor agonists could influence addiction risk. In animal models, GLP-1 signaling within the ventral tegmental area and nucleus accumbens appears to blunt dopamine release in response to addictive substances, dampening their rewarding impact. GLP-1 receptors in the prefrontal cortex may also modulate executive control and decision-making, potentially making it easier to resist urges.
Peripheral effects might contribute as well. By slowing gastric emptying and altering blood glucose dynamics, GLP-1 drugs can reduce the rapid spikes and crashes that sometimes trigger alcohol use in people who drink to manage energy or mood. Weight loss and improved metabolic health could indirectly decrease substance use by lessening depression, pain, or sleep problems that sometimes drive self-medication. Disentangling these overlapping pathways will require studies that directly measure brain activity, reward sensitivity, and subjective craving over time.
Limitations, cautions, and what comes next
Despite the excitement, the current evidence has important caveats. Most of the large datasets involve people with diabetes or obesity, so the findings may not generalize to those without metabolic disease. Confounding by indication remains a concern: people willing and able to start GLP-1 therapy may differ in motivation, access to care, or socioeconomic status from comparison groups in ways that are hard to fully adjust away.
Moreover, lower rates of new diagnoses in health records do not necessarily mean lower rates of actual substance use. People on GLP-1 therapy may see their clinicians more often for metabolic care, but they might also attribute symptoms to the new drug rather than to alcohol or opioid use, potentially shifting how problems are documented. Conversely, improved overall health could make clinicians more likely to screen for and address addiction, increasing detection. Each of these forces could bias results in different directions.
From a clinical standpoint, GLP-1 receptor agonists are not benign. They can cause nausea, vomiting, diarrhea, and in rare cases pancreatitis or gallbladder disease. They are expensive, often require prior authorization, and may not be covered for indications beyond diabetes or obesity. Using them off-label for addiction would raise ethical and equity questions if only well-resourced patients could access them.
Next steps include larger randomized trials in people with alcohol or opioid use disorder, including those without obesity, and head-to-head comparisons with existing medications. Researchers are also exploring whether different GLP-1 agents, or dual agonists that also target GIP receptors, have stronger or weaker effects on craving and relapse. If the early signals hold up, future guidelines may consider GLP-1 drugs as adjuncts to behavioral therapy and standard pharmacotherapy, rather than as stand-alone cures.
For now, the emerging data offer cautious optimism: medications developed to treat diabetes and obesity may unexpectedly expand the toolkit for addressing some of the most stubborn and deadly forms of addiction. But translating that promise into practice will require rigorous trials, careful attention to safety, and policies that ensure any benefits are available to the people who need them most.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
