Semaglutide and Tirzepatide Cut All-Cause Overdose Risk in Veterans with Type 2 Diabetes and OUD

Overdoses, particularly from opioids, represent a critical public health crisis. Current treatments for opioid use disorder (OUD) face significant access barriers and efficacy limitations. Preclinical research has suggested that glucagon-like peptide-1 receptor agonists (GLP-1RAs) may offer a novel therapeutic avenue for OUD by influencing reward pathways. Early observational studies hinted at a reduced opioid overdose risk with GLP-1RAs, but robust replication in diverse patient populations, such as Veterans with type 2 diabetes mellitus (T2DM) and OUD, was needed to solidify this potential link.

This target trial emulation study utilized Veterans Health Administration (VA) data, enrolling Veterans diagnosed with type 2 diabetes mellitus and opioid use disorder who initiated either semaglutide or tirzepatide, or a comparison diabetes medication (insulin, metformin, sulfonylurea, SGLT2 inhibitor, or DPP4 inhibitor). The study period spanned from January 1, 2020, to December 31, 2024. Each comparison group was meticulously propensity score matched on relevant variables to minimize confounding. Cox proportional hazards regression models were then employed to assess the time to all-cause overdose events over a 12-month follow-up period after medication initiation.

After rigorous propensity score matching, the analysis revealed that semaglutide and tirzepatide were associated with a significantly lower risk of all-cause overdose compared to certain comparator medications. Specifically, when compared to insulin, the GLP-1RAs showed a substantial reduction in risk: HR=0.32 (95% CI=0.14-0.71; P=.006; n=630). A similar protective effect was observed against SGLT2 inhibitors: HR=0.25 (95% CI=0.09-0.68; P=.006; n=432). However, this significant reduction was not uniform across all comparisons.

Semaglutide and tirzepatide did not demonstrate a significantly lower risk of all-cause overdose when compared to 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). These findings suggest a differential impact based on the comparator drug class.