GLP-1 Drugs Cut Addiction Risk by Half: The Smoking-to-Gambling Link

Last Updated: March 2025

In a retrospective cohort study of 603,000 patients with type 2 diabetes, those prescribed GLP-1 receptor agonists showed a 50% to 56% reduction in substance use disorder encounters compared to insulin users, according to 2024 research from Washington University School of Medicine published in Addiction. The data spans nicotine, alcohol, opioids, cannabis, and stimulants—suggesting these medications act on fundamental reward circuitry rather than individual substance pathways.

GLP-1 drugs were engineered to mimic a gut hormone that regulates blood sugar and appetite. But their effect on dopamine signaling in the mesolimbic pathway—the brain's reward center—appears to dampen the neural noise that drives compulsive behavior across multiple addiction types. Clinicians are now reporting patients who quit smoking without trying, lost interest in their nightly wine, or stopped compulsive online shopping while taking semaglutide or tirzepatide for weight loss.

The Mechanistic Case: How GLP-1s Alter Reward Processing

GLP-1 receptors densely populate the ventral tegmental area and nucleus accumbens, regions that process reward salience and motivation. When nicotine, alcohol, or cocaine hits these circuits, dopamine floods the synapse. The brain learns: this matters, do it again. GLP-1 receptor activation appears to modulate this dopamine response, reducing the perceived value of the reward.

Animal models demonstrate this clearly. Rodents given liraglutide self-administer 40% less cocaine than controls, per 2023 research from the University of Pennsylvania published in Molecular Psychiatry. The effect persists across alcohol, methamphetamine, and nicotine. The mechanism isn't sedation or nausea—the rats remain active and eat normally. The drugs seem to specifically blunt the motivational pull of addictive substances.

In humans, the effect manifests as what patients describe as reduced "mental chatter" around the addictive behavior. Dr. Anna Lembke, addiction psychiatrist at Stanford University, told National Geographic in 2024: "These drugs appear to quiet the preoccupation phase of addiction—the rumination, the planning, the constant cognitive pull toward the substance or behavior."

This preoccupation phase consumes enormous psychological bandwidth. Addiction isn't just about physical withdrawal. It's the intrusive thoughts, the bargaining with yourself, the mental energy spent on obtaining and using. If GLP-1s reduce that cognitive load, they're targeting a core feature of addictive disorders that most pharmacotherapies don't touch.

The Clinical Evidence: Substance by Substance

The Washington University study examined electronic health records from the U.S. Department of Veterans Affairs between October 2017 and September 2023. They compared outcomes in 221,111 patients prescribed GLP-1 agonists versus 381,889 prescribed insulin, controlling for demographics, comorbidities, and baseline substance use.

Substance	Hazard Ratio (GLP-1 vs Insulin)	Risk Reduction
Tobacco	0.44	56%
Alcohol	0.50	50%
Opioids	0.60	40%
Cannabis	0.54	46%
Stimulants	0.52	48%

These aren't trivial effect sizes. For context, varenicline (Chantix), the most effective smoking cessation medication, shows roughly 50% to 60% improved quit rates versus placebo in controlled trials. The GLP-1 data comes from real-world observational records—messier, but arguably more representative of how these drugs perform outside trial conditions.

The alcohol findings align with earlier Danish registry data from 2022 examining 227,866 patients. Those prescribed semaglutide showed a 14% reduction in alcohol-related healthcare contacts compared to SGLT2 inhibitor users. A small randomized trial at the National Institute on Alcohol Abuse and Alcoholism, currently enrolling, is testing exenatide in 100 patients with moderate-to-severe alcohol use disorder.

For opioids, the signal is weaker but present. The 40% risk reduction in the VA study is notable given that opioid use disorder often involves powerful physical dependence mechanisms beyond dopamine signaling alone. Preliminary work from researchers at the University of Texas is exploring whether GLP-1s could augment buprenorphine treatment, potentially improving retention rates that typically hover around 50% at one year.

Beyond Substances: Behavioral Addictions

The dopamine-modulating mechanism suggests GLP-1s might affect behavioral addictions—gambling, compulsive shopping, problematic internet use—where no chemical substance is involved. Case reports are accumulating. One patient on tirzepatide for diabetes told his physician he'd lost interest in the sports betting that had previously consumed several hours daily and considerable money. Another stopped compulsive skin-picking that had persisted for fifteen years.

Researchers at Icahn School of Medicine at Mount Sinai are launching a trial examining semaglutide for gambling disorder. The hypothesis: if the drugs reduce reward sensitivity broadly, they should blunt the dopamine spike from a winning bet the same way they blunt the spike from a cigarette. Gambling disorder affects approximately 1% of U.S. adults, with limited effective pharmacological treatments. Current options include naltrexone, which blocks opioid receptors, and shows modest efficacy.

The behavioral addiction angle also raises questions about romantic obsession, compulsive sexual behavior, and even intense infatuation—states that hijack reward circuitry similarly. If GLP-1s dampen these experiences, are we medicalizing normal human passion? Or offering relief to people whose compulsions cause genuine dysfunction?

The line isn't always clear. But for someone whose gambling has destroyed their finances, or whose drinking has cost them custody of their children, philosophical concerns about dampening life's intensities seem less pressing than the practical question: does it work?

The Prescribing Reality: Off-Label Use and Access Barriers

GLP-1 agonists are FDA-approved for type 2 diabetes and obesity, not addiction. Off-label prescribing is legal but complicates insurance coverage. Monthly costs range from $900 to $1,300 for semaglutide and tirzepatide without insurance. Prior authorization requirements typically mandate BMI above 30 or above 27 with weight-related comorbidities.

This creates an access problem. Addiction treatment already struggles with inadequate insurance coverage. Methadone clinics operate on shoestring budgets. Community mental health centers can't afford $1,000 monthly per-patient drug costs. If GLP-1s prove genuinely effective for addiction—and that's still an "if" pending randomized controlled trials—payer coverage policies will need dramatic revision.

Some addiction specialists are already prescribing off-label for patients who meet obesity criteria and have co-occurring addiction. The weight loss becomes the documented indication; reduced drinking or smoking is the unstated target. This workaround is ethically ambiguous and excludes patients without obesity.

The pharmaceutical companies are watching closely. Novo Nordisk and Eli Lilly have not publicly announced addiction-focused trials for their blockbuster drugs, but investor calls hint at interest. Expanding indications to alcohol use disorder alone would open a market of 14.5 million U.S. adults with the condition, per NIAAA 2023 estimates.

What We Don't Know: The Gaps in Evidence

Observational data can't prove causation. Patients who get prescribed GLP-1s might differ systematically from insulin users in ways the statistical models don't capture. Maybe they have better access to care, higher health literacy, or stronger support systems. The VA study controlled for numerous confounders, but residual confounding always remains possible.

We also don't know optimal dosing for addiction. Weight loss trials use semaglutide 2.4 mg weekly or tirzepatide 10 to 15 mg weekly. Are lower doses sufficient for addiction effects? Higher doses necessary? The dose-response relationship is unclear.

Duration of treatment is another unknown. Obesity treatment involves indefinite use—stop the drug, weight returns. Does addiction treatment follow the same pattern? If someone quits smoking on semaglutide but stops the medication, do cravings return? Recovery frameworks emphasize long-term remission. A medication that requires lifelong use creates different considerations than one that facilitates initial abstinence then can be discontinued.

Side effects matter more in addiction treatment than obesity treatment. Nausea, the most common GLP-1 side effect, affects 15% to 40% of users depending on the drug and dose. In weight loss trials, that's often tolerable or even welcomed. In someone with alcohol use disorder who's already managing nausea, vomiting, and gastrointestinal distress, it's a harder sell.

Rare but serious risks include pancreatitis, gallbladder disease, and possibly thyroid tumors (seen in rodents, not definitively in humans). For obesity, where cardiovascular benefits are well-established, the risk-benefit calculation is clear. For addiction treatment, we're still gathering the benefit data.

The Controlled Trial Pipeline

Several randomized trials are underway or planned. The NIAAA exenatide trial for alcohol use disorder aims to report results in 2026. Researchers at the University of North Carolina are planning a semaglutide trial for cocaine use disorder. A team at Johns Hopkins is exploring tirzepatide for cannabis use disorder.

These trials will provide cleaner evidence than observational studies. But they're small—typically 50 to 200 participants—and short, often 12 to 16 weeks. That's standard for addiction medication trials, but it means questions about long-term efficacy and safety will persist even after positive results.

The research infrastructure for addiction medication trials is underfunded compared to other therapeutic areas. Recruiting people with active addiction is challenging. Retention is difficult. Placebo response rates are high, making it hard to detect drug effects. These aren't problems unique to GLP-1s, but they slow the evidence generation process.

Clinical Implications: What Prescribers Should Consider

The current evidence doesn't support prescribing GLP-1s solely for addiction in patients without diabetes or obesity. The observational data is compelling but preliminary. We need randomized trials showing efficacy, safety, and optimal management strategies.

For patients already on GLP-1s who report reduced substance use, that's useful information. It suggests the effect is real and might influence treatment planning. If someone with diabetes and alcohol use disorder is choosing between a GLP-1 and insulin, the addiction data might tip the decision.

Addiction treatment providers should track this literature. The potential is significant enough to warrant attention, even while maintaining appropriate skepticism. The field has been burned before by overhyped pharmacological solutions. Baclofen for alcohol use disorder generated enthusiasm in the mid-2000s but failed to deliver in large trials. Modafinil for cocaine addiction similarly disappointed.

GLP-1s have a plausible mechanism and increasingly robust observational data. That puts them ahead of those prior candidates. But mechanism and observational data don't guarantee clinical utility. The trials will tell.

Sources

1. Wang W, Volkow ND, Berger NA, Davis PB, Kaelber DC, Xu R. Association of s