Comparison of telmisartan with losartan, valsartan, and candesartan on the risk of new-onset diabetes mellitus.

Evidence remains limited on whether telmisartan with partial peroxisome proliferator-activated receptor-gamma (PPAR-γ) activity lowers new-onset diabetes mellitus (NODM) risk compared with non-PPAR-γ-active angiotensin II receptor blocker (ARB). We therefore evaluated the NODM risk of telmisartan compared with losartan, valsartan, and candesartan using real-world clinical data. We conducted a multicenter, retrospective, treatment-control cohort study using clinical data from 17 institutions between 2005 and 2023, covering 19 283 922 patients converted to a common data model. Adults newly prescribed telmisartan or other ARBs (losartan, valsartan, or candesartan) and taking the medication for ≥6 months were included, while patients with pre-existing diabetes were excluded. Propensity score matching (PSM) was performed, and hazard ratios (HR) were estimated using a Cox proportional hazards model. The primary outcome was incident NODM, defined by a diagnosis of diabetes (ICD-10), initiation of glucose-lowering medications, or HbA1c ≥ 6.5%. A total of 6 868 patients of new-users of telmisartan and 22 500 patients of other ARBs were included. In the overall population, telmisartan was associated with lower NODM risk (HR 0.87, 95% CI 0.80-0.94). After 1:2 PSM, there was no significant difference in the risk of NODM between telmisartan and other ARBs (10.61% [39.50/1 000 person-years] vs. 11.02% [40.42/1 000 person-years]; HR 0.97, 95% CI 0.88-1.08). Secondary analyses were also not significantly different. Overall, this study did not detect sufficient evidence that telmisartan reduces NODM risk compared with losartan, valsartan, or candesartan.