GLP-1 and Addiction: The Emerging Signal From Alcohol, Nicotine, and Opioid Data

The most talked-about off-label GLP-1 story of 2026 isn't about weight or diabetes — it's about addiction. A study of more than 600,000 U.S. veterans found that people taking GLP-1 medications had a 14% to 25% lower risk of developing substance use disorders across alcohol, opioids, cannabis, cocaine, and nicotine.¹

This is not a randomized controlled trial. It is observational data — large, well-designed observational data, but observational nonetheless. The distinction matters because observational studies cannot prove causation. However, when the same signal appears across multiple substances, multiple study designs, and is supported by a plausible biological mechanism, the evidence base starts to become difficult to dismiss.

The Biological Mechanism

GLP-1 receptors are not limited to the gut and pancreas. They are expressed throughout the brain, including in the mesolimbic dopamine pathway — the same neural circuit that drives addiction. When GLP-1 agonists bind to receptors in the nucleus accumbens and ventral tegmental area, they appear to modulate dopamine signaling in ways that reduce reward-driven behavior.⁴

In plainer terms: the same mechanism that reduces "food noise" — the constant preoccupation with eating — appears to reduce the neurological pull of alcohol, nicotine, and other substances. Patients frequently report this anecdotally: "I just don't think about drinking as much."

The Evidence by Substance

Alcohol

The strongest clinical data is here. A 2022 RCT tested exenatide (an older GLP-1) in people with alcohol use disorder. It showed no overall effect — but a subgroup with comorbid obesity had reduced consumption.² A 2025 JAMA Psychiatry trial tested semaglutide specifically in adults with AUD. Over 9 weeks, semaglutide significantly reduced alcohol cravings, drinking frequency, and amount consumed versus placebo.³

Novo Nordisk has announced a trial studying GLP-1 effects on alcohol consumption in people with alcohol-related liver disease.⁶

Opioids

In rodent models, GLP-1 agonists reduced self-administration of heroin, fentanyl, and oxycodone, and reduced relapse-like behavior. Human clinical trial data is limited but the VA cohort study showed significant risk reduction.⁴

Nicotine

Preclinical data shows GLP-1 agonists reduce nicotine self-administration and relapse in animal models. Early human trials suggest reduced daily cigarette consumption and, importantly, prevention of the weight gain that typically follows smoking cessation.⁴

Cannabis and Cocaine

A 2026 study linked GLP-1 medications to 14% lower cannabis use disorder risk. Cocaine use reductions of 20–26% have been observed in preclinical models. Clinical trial data remains sparse.¹

Eli Lilly is testing brenipatide, an experimental compound, specifically on alcohol, tobacco, and opioid use disorders.⁶

What This Means

The addiction signal is the most important unanswered question about GLP-1 drugs. If confirmed by large RCTs, it would represent a paradigm shift in addiction medicine — a single drug class treating obesity, diabetes, cardiovascular disease, and multiple substance use disorders. The mechanism is biologically plausible, the observational data is large and consistent, and multiple trials are now underway.

But we are not there yet. The history of medicine is full of promising observational signals that failed in randomized trials. Until those trials report, the addiction application remains intriguing and unproven.