What Is This?
Ozempic was approved for type 2 diabetes in 2017. Wegovy — the same molecule at a higher dose — was approved for obesity in 2021. Neither was designed to treat addiction. Neither was tested for addiction. And yet, within months of the first major prescribing wave, something strange started happening in the clinical reports: people were spontaneously stopping drinking. Stopping gambling. Losing interest in cigarettes. Reporting that the compulsive pull toward food, alcohol, and other substances had simply gone quiet — without specifically trying.
These weren't side effects listed on the label. They were surprises.
The drug class involved is GLP-1 receptor agonists — synthetic versions of glucagon-like peptide-1, a hormone naturally released from the gut after eating. Semaglutide (Ozempic, Wegovy), liraglutide (Victoza, Saxenda), and the newer tirzepatide (Mounjaro, Zepbound) are the major players. Their stated mechanism is metabolic: they slow gastric emptying, reduce appetite, and improve insulin sensitivity. This is why they cause weight loss.
But GLP-1 receptors aren't only in the gut and pancreas. They're distributed throughout the brain — including in the mesolimbic dopamine system: the nucleus accumbens, the ventral tegmental area, the prefrontal cortex. This is the reward circuitry. The same architecture that drives every addiction.
When GLP-1 agonists bind to receptors in this system, they appear to reduce the dopamine response to rewarding stimuli. The "hit" from a drink, a cigarette, a gambling win, a binge-eat — the motivational salience of these experiences — gets turned down. Not eliminated, but quieter. And for people who have spent years fighting the pull of these behaviours with willpower, therapy, and substitutes, "quieter" is transformative.
Why Does It Matter?
- The clinical evidence is building fast. A Phase 2 double-blind RCT published in JAMA Psychiatry (2025) tested once-weekly semaglutide in adults with alcohol use disorder. The result: semaglutide significantly reduced laboratory alcohol self-administration, drinks per drinking day, and craving scores compared to placebo. This is a controlled trial, not anecdote.^1 Separately, rodent studies have shown that GLP-1 agonists "significantly reduce intake, preference, and relapse-related behaviors associated with alcohol, cocaine, nicotine, and highly palatable foods."^2 Multiple Phase 2 trials for nicotine dependence and opioid use disorder are currently running.
- It's the first plausible mechanism for treating multiple addictions with a single intervention. Current addiction pharmacology is substance-specific: naltrexone for alcohol and opioids, varenicline for nicotine, methadone for opioids. None cross over. GLP-1 agonists appear to work across the board because they're not targeting the substance — they're targeting the underlying reward architecture that all addictive behaviours hijack. This is a category-level shift if it holds up in larger trials.
- It reframes what addiction actually is. The dominant model of addiction is the disease model: a chronic brain disorder characterised by compulsive substance use despite harmful consequences. GLP-1 evidence adds a metabolic dimension. The gut-brain axis that regulates hunger and satiety is also regulating reward sensitivity. Addiction may be, in part, a dysregulation of the same signalling pathways that govern eating. The drugs work across both domains because those domains share deeper architecture than the disease model suggested.^3
- The anhedonia risk is the serious open question. If GLP-1s globally dampen reward sensitivity, what else gets turned down? Reports from patients on semaglutide describe reduced enjoyment of food, alcohol, and substances — but also, for some, blunted enjoyment of other pleasurable activities: hobbies, socialising, sex. The clinical term is anhedonia: the inability to experience pleasure. Some patients report feeling "flattened." This is not universal, and the clinical trials haven't found severe anhedonia at scale, but it's being actively studied. A drug that extinguishes addiction by extinguishing the capacity for reward is not the same as a drug that extinguishes addiction while preserving the capacity for healthy pleasure. That distinction matters enormously.^4
- The market implications are enormous and already moving. Novo Nordisk (semaglutide) and Eli Lilly (tirzepatide) are running multiple addiction indication trials. Alcohol use disorder affects ~400 million people globally. Nicotine addiction ~1.3 billion. The current approved medications for AUD reach roughly 10% of those who could benefit. An FDA-approved semaglutide indication for AUD would create a market that dwarfs the existing weight loss opportunity.
Key People & Players
Carolina Haass-Koffler (Brown University School of Public Health) — Leading neurobiological mechanistic research on GLP-1s and addiction. Her lab is specifically investigating how GLP-1 agonists influence addictive behaviours at the biological level — mapping the receptor distribution in reward circuits and what happens when you activate them pharmacologically.^5
Christian Hendershot (University of Southern California) — Lead investigator on the JAMA Psychiatry 2025 semaglutide/AUD Phase 2 trial. The most rigorous controlled human evidence to date.
Novo Nordisk — Manufacturer of semaglutide. Running multiple clinical trials for addiction indications. The financial incentive to establish these indications is substantial.
Nora Volkow (NIDA) — Director of the National Institute on Drug Abuse. The leading US voice on addiction neuroscience, has publicly described GLP-1 addiction evidence as one of the most significant findings in the field in years.
The Current State
The field is at the transition from "compelling signals" to "controlled trials." The animal data is robust. The anecdotal reports are consistent. The Phase 2 human trials for alcohol use disorder are showing significant effects. Phase 3 trials are the next gate — larger, longer, powered to demonstrate clinical outcomes.
What's not yet established:
- Long-term safety for addiction indications (current obesity/diabetes trials run longer, but addiction patients have different profiles)
- Whether the effect persists after stopping GLP-1 treatment (early data suggests it may not — the quieting of craving may require ongoing drug exposure)
- The anhedonia question at scale
The regulatory timeline: FDA approval for an addiction indication (alcohol use disorder is the lead candidate) is likely 3–5 years away at current trial speeds. But prescriptions off-label for addiction are already happening, driven by the anecdotal reports and the Phase 2 data.
The deeper shift:
The GLP-1 addiction story is also a story about how scientific progress works. The most important finding wasn't discovered — it was reported by patients. People started telling their doctors "I've stopped drinking" or "I don't want cigarettes anymore" and clinicians paid attention. The formal trials followed the anecdotes. This is how some of the most important pharmacological discoveries happen: not through hypothesis-driven experiments but through unexpected patient reports that the model doesn't predict, and careful researchers who take them seriously.
Best Resources to Learn More
- JAMA Psychiatry 2025 — semaglutide in alcohol use disorder (RCT) — The primary controlled trial data.^6
- Endocrine Society: GLP-1s show promise in treating alcohol and drug addiction — Clinical summary without the jargon.^7
- Brown University SPH: A turning point in addiction psychiatry? — Good profile of Haass-Koffler's mechanistic work.^8
- Dopamine Nation by Anna Lembke — The best book on dopamine and addiction before the GLP-1 story broke. Required background for understanding what these drugs are doing at the circuit level.^9
- PMC: Mechanisms of GLP-1 in Modulating Craving and Addiction — The neurobiological mechanism review for those who want the full science.^10