The preclinical data is compelling but let's be careful about extrapolation. Here's where we stand in human evidence as of early 2026:

Epidemiological/observational data:

• A large retrospective cohort study using electronic health records (Wang et al., Nature Medicine 2024) found that semaglutide use was associated with significantly lower rates of alcohol use disorder diagnosis (HR 0.44) compared to non-GLP-1 RA anti-obesity medications[4]
• Similar observational signals for reduced opioid use disorder and cannabis use disorder

Prospective trial data:

• Small pilot RCTs of exenatide for alcohol use disorder have shown mixed results — some reduction in drinking days but not consistently significant
• A larger RCT of semaglutide for alcohol use disorder is currently recruiting (NCT06062706)
• No RCT data yet for GLP-1 RAs in nicotine or opioid use disorder

We are at the "promising preclinical + suggestive observational" stage. This is where many promising therapies fail when subjected to rigorous RCTs. The signal is real enough to warrant proper trials, but we should not be prescribing semaglutide off-label for addiction at this point.

I want to share my personal experience — understanding that anecdote is not evidence. I've been on semaglutide for 18 months for obesity. Before starting, I was a moderate drinker (2-3 glasses of wine most nights). Within the first month on semaglutide, my desire for alcohol essentially evaporated. I didn't make a conscious decision to stop — I just stopped wanting it.

I've spoken to others in support groups who report similar experiences. Some describe reduced interest in shopping, gambling, and social media scrolling as well. Whether this is a direct pharmacological effect on reward circuits or an indirect effect of improved mood, better sleep, and less hedonic eating is impossible for me to distinguish subjectively.

The mechanistic story is interesting when you look at it through the lens of dopaminergic circuit modulation. Here's what the neuroscience literature suggests:

1. Tonic vs. phasic dopamine: GLP-1R agonism in the VTA appears to reduce phasic (burst) dopamine signaling in the NAc without dramatically affecting tonic (baseline) dopamine levels[5]. This is significant because phasic dopamine is associated with reward salience and "wanting," while tonic dopamine is more associated with motivation and motor function.
2. Reward prediction error: By blunting phasic dopamine, GLP-1R agonism may reduce the reward prediction error signal — the brain's "this is better than expected" signal that drives reinforcement learning for addictive substances.
3. Hedonic set point: Chronic GLP-1R agonism may gradually recalibrate the hedonic set point, reducing the perceived reward value of highly palatable foods, alcohol, and other reinforcers without inducing anhedonia (which would be a concern).

This pattern — reduced "wanting" without reduced "liking" or general anhedonia — is exactly what you'd want from an anti-addiction pharmacotherapy. It's the holy grail of addiction pharmacology.

I want to inject some caution into this discussion. The pattern of "new drug class shows unexpected broad benefits" has played out before in medicine, and it doesn't always end well. SSRIs were touted for everything from depression to OCD to premature ejaculation to migraine prevention. Some of those indications panned out; others didn't survive rigorous testing.

Specific concerns about the addiction-GLP-1 hypothesis:

1. Confounding in observational studies: Patients who fill GLP-1 RA prescriptions and maintain adherence may be systematically different from those who don't — more engaged in healthcare, more motivated to change behavior, better social support. These same factors protect against substance use disorder.
2. Publication bias: There is massive public interest in GLP-1 RAs right now. Studies showing positive associations get published and amplified; null results may languish.
3. Dose and penetrance: We don't know what dose of semaglutide is needed to meaningfully affect mesolimbic dopamine in humans. Peripheral semaglutide has limited BBB penetration. The CNS concentration may be sufficient for hypothalamic appetite effects but insufficient for VTA reward circuit modulation at standard doses.

I think the hypothesis is worth testing rigorously. But we should resist the temptation to treat GLP-1 RAs as a panacea for all reward-related disorders until the RCT data exist.