BackgroundOveractive bladder (OAB) syndrome is a common condition affecting up to 15% of adults. Medical management of OAB has traditionally relied upon treatment with antimuscarinics or the beta-3 agonist mirabegron. Antagonism of muscarinic receptors is hypothesized to impair glucose homeostasis and increase the risk of hyperglycemia. However, the real-world data evaluating the hyperglycemic effects of the antimuscarinics versus mirabegron are limited.ObjectiveOur objective was to compare the risk of hyperglycemia and new-onset diabetes among new users of oral M3-selective antimuscarinic agents or non-selective antimuscarinic agents versus mirabegron.MethodsUsing linked healthcare databases from Ontario, Canada, we conducted a population-based cohort study among all Ontario residents aged ≥ 66 years who were new users of an oral M3-specific antimuscarinic (darifenacin, solifenacin), non-specific antimuscarinic (oxybutynin, tolterodine, fesoterodine, trospium), or mirabegron (reference group) from June 1, 2016, to November 1, 2023. We used three-way matching weights on over 50 covariates to adjust for potential confounding. We used linear mixed models to estimate differences in the annual rate of change in glycated hemoglobin (HbA1c) and Cox proportional hazards models to estimate differences in the time to new-onset diabetes.ResultsAmong the 110,697 patients in the study cohort, 10,285 (9.3%) were new users of a M3-selective antimuscarinic, 34,918 (31.5%) a non-selective antimuscarinic, and 65,494 (59.2%) mirabegron. After applying matching weights, the measured baseline characteristics of groups were similar. The annual rate of increase in HbA1c during follow-up was 0.01% lower (95% confidence interval [CI] − 0.02 to − 0.05) among users of M3-selective antimuscarinics than among users of mirabegron, with no difference in the rate of change in HbA1c between users of the non-selective antimuscarinics versus mirabegron (0.00%; 95% CI − 0.01 to 0.00). The rate of new-onset diabetes did not differ among the study groups: 12.2 per 1000 person-years in M3-selective antimuscarinic users (hazard ratio [HR] 1.09; 95% CI 0.88–1.34), 11.6 per 1000 person-years in non-selective antimuscarinic users (HR 1.04; 95% CI 0.90–1.19), and 11.1 per 1000 person-years in mirabegron users.ConclusionIn a population-based cohort of over 100,000 older adults, new use of an oral M3-selective or non-selective antimuscarinic agent was not associated with an increase in hyperglycemia or new-onset diabetes compared with mirabegron.