5 Things Addiction Doctors Wish You Knew About GLP-1 Drugs
1. These Drugs Are Not Approved for Addiction — Yet
Semaglutide, tirzepatide, liraglutide, and all other GLP-1 receptor agonists are FDA-approved for type 2 diabetes and/or obesity. None are approved for any substance use disorder. Using them for addiction is off-label at best and investigational at this stage. The Lancet trial (May 2026) was a 108-person study. Phase 3 trials are underway but not complete. The timeline to potential FDA approval for any addiction indication is likely years away.
This matters because insurance coverage, clinical guidelines, and prescribing practices all follow FDA approval. Right now, getting a GLP-1 medication specifically for addiction treatment is not straightforward.
2. They Won't Replace Therapy
Even if GLP-1 medications prove effective for addiction, they will not replace psychosocial treatment. The Lancet trial itself paired semaglutide with cognitive behavioral therapy — it tested whether the drug added benefit on top of CBT, not whether it could replace it. Addiction involves behavioral patterns, environmental triggers, social relationships, and psychological factors that no medication addresses directly.
Dr. Anna Lembke, Stanford psychiatrist and author of Dopamine Nation, has emphasized that medication for addiction works best when combined with behavioral interventions. A pill that reduces craving doesn't teach someone how to rebuild relationships, manage stress, or restructure their daily routine around sobriety.
3. The Evidence Is Strongest for Alcohol
The published clinical trial data (randomized, controlled, prospective) currently exists only for alcohol use disorder. The evidence for opioids, nicotine, cannabis, and stimulants comes from observational studies — large and compelling, but not the gold standard. As Dr. Christian Hendershot of the USC Institute for Addiction Science has noted, the accumulating evidence is promising, but "most people are talking about this from a treatment standpoint" when the data primarily shows association, not proven causation.
4. They Could Work Alongside Existing Medications
GLP-1 agonists work through a different mechanism than any existing addiction medication. This means they could potentially be combined with naltrexone (for AUD or OUD), buprenorphine (for OUD), or nicotine replacement therapy (for smoking) to provide complementary craving reduction. No combination trials have been published yet, but the theoretical rationale is strong, and Dr. Philip Coffin's SF team is already planning to test GLP-1s as an "added agent" alongside buprenorphine.
5. Access and Cost Are Real Barriers
GLP-1 medications currently cost over $1,000 per month without insurance. Even with insurance, coverage is inconsistent and often requires prior authorization. Supply shortages have affected availability. For addiction treatment specifically — a condition that disproportionately affects lower-income populations with limited healthcare access — the cost barrier is especially significant.
If GLP-1 medications are eventually approved for addiction, separate insurance coverage decisions will be required. The compounded versions that exist for weight loss may or may not be available for addiction indications. These practical realities will determine whether the science actually reaches the people who need it.
The Expert Consensus
The scientists studying GLP-1 medications for addiction are excited about the potential but cautious about the hype. Effective addiction treatments exist right now — and they're dramatically underutilized. The biggest barrier to addiction treatment today isn't the lack of a new drug; it's the failure to deliver the treatments we already have. SAMHSA National Helpline: 1-800-662-4357.
Sources
- Lembke A. "Five things to know about GLP-1s and addiction." Stanford Medicine. April 2025.
- Hendershot CS, quoted in "GLP-1s could help curb addiction." NBC News. March 7, 2026.
- Coffin P, quoted in "SF addiction researchers exploring new GLP-1 treatment frontier." SF Examiner. April 9, 2026.
- Klausen MK, et al. The Lancet. 2026;407(10540):1687-1698.