Patients on GLP-1 agonists had lower opioid overdose deaths, but does the drug deserve the credit?
STUDY: Lakhani DA et al, Journal of Addictive Diseases, 2026
STUDY TYPE: Retrospective cohort study
FUNDING: National Institute of General Medical Sciences, National Institutes of Health
Background
GLP-1 receptor agonists (semaglutide, liraglutide, and tirzepatide) were developed for diabetes and obesity, but their effects on the brain’s reward system are relevant to addictions. In animal studies, they reduce use of alcohol, cocaine, nicotine, and opioids, and a few human studies are positive in nicotine and alcohol. This study looks at their association with opioids.
The Study
- 473,906 adults with a diagnosis of opioid use disorder, drawn from a large national electronic health record network (TriNetX) spanning 2016 to 2025; 8,737 (1.8%) had a GLP-1 receptor agonist prescription within the month before diagnosis
- Propensity score matching (1:1) balanced 8,052 pairs on age, sex, race, medical comorbidities, psychiatric diagnoses, substance use disorders, and opioid use disorder medications including buprenorphine, methadone, and naltrexone
- Primary outcomes: all-cause mortality and opioid overdose; secondary outcomes: inpatient admissions and emergency department visits, tracked at 6 months, 1 year, 3 years, and 5 years
Results
Mortality was sharply lower in the GLP-1 group at every time point. At 5 years, 4.9% of GLP-1 users had died versus 10.9% of non-users (hazard ratio 0.63). The gap opened early and widened over time.
Opioid overdose was also less common among GLP-1 users: 3.0% versus 4.8% at 5 years (hazard ratio 0.81).
Hospitalizations and emergency department visits followed the same pattern. At 5 years, inpatient admission rates were 31.7% versus 47.2% (hazard ratio 0.73), and emergency department visit rates were 35.7% versus 48.2% (hazard ratio 0.83).
Limitations
Retrospective, cannot test causation (not randomized). Taking a GLP-1 could be a signal of taking better care of health, which itself is protective.
Practice Implications
- This study doesn’t come close to suggesting we should use GLP-1’s in opioid use disorder. It’s the first chapter in a story that will continue to evolve.
— Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
