GLP-1 Receptor Agonists and All-Cause Overdose Risk in Veterans With Type 2 Diabetes and Opioid Use Disorder

Objective: Overdoses remain a major public health challenge. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have been proposed as a treatment for opioid use disorder (OUD) based on preclinical research. Early observational research suggests that GLP-1RAs are associated with reduced opioid overdoses but warrants replication in different patient populations.

Methods: This target trial emulation study used Veterans Health Administration (VA) data to include Veterans with a diagnosis of type 2 diabetes mellitus and OUD who initiated semaglutide and tirzepatide or a comparison diabetes medication (insulin, metformin, a sulfonylurea, a sodium-glucose transport 2 (SGLT2) inhibitor, or a dipeptidyl-peptidase 4 [DPP4] inhibitor) between January 1, 2020, and December 31, 2024. Each set of comparison groups was propensity score matched on relevant variables. Cox proportional hazards regression models were used to compare the time to all-cause overdose events in the 12 months after medication initiation.

Results: After propensity score matching, semaglutide and tirzepatide were associated with significantly lower risk of all-cause overdose compared to insulin (hazard ratio [HR]=0.32; 95% CI=0.14-0.71; P=.006; n=630) and SGLT2 inhibitors (HR=0.25; 95% CI=0.09-0.68; P=.006; n=432). Semaglutide and tirzepatide were not associated with significantly lower risks than metformin (HR=0.75; 95% CI=0.38-1.45; n=1,016), sulfonylureas (HR=0.82; 95% CI=0.33-1.96; n=710), or DPP4 inhibitors (HR=1.20; 95% CI=0.52-2.78; n=858).

Conclusion: Collectively, semaglutide and tirzepatide were associated with lower risk of all-cause overdose compared to insulin and SGLT2 inhibitors, but not metformin, sulfonylureas, or DPP4 inhibitors. These results suggest the possible role of GLP-1RAs in OUD but underscore the need for randomized controlled trials.