Drug Therapies and Incidence of Dementia Among Adults With Overactive Bladder

Importance Concern that anticholinergic medications for overactive bladder (OAB) are associated with increased dementia risk has encouraged a shift toward β3 adrenergic agonists, which are presumed cognitively safer. Whether this presumption is justified is uncertain. Objective To compare the risk of incident dementia among adults initiating high-risk anticholinergic, low-risk anticholinergic, or β3 agonist therapy for OAB and to evaluate the association of treatment duration with risk. Design, Setting, and Participants This retrospective cohort study used Epic Cosmos electronic health record data from 2012 to 2025. Adults aged 40 to 90 years newly treated for OAB within 1 year of diagnosis without preexisting dementia were included. Exposures First prescribed medication class: high-risk anticholinergics (oxybutynin, tolterodine, or solifenacin), low-risk anticholinergics (darifenacin, trospium, or fesoterodine), or β3 agonists (mirabegron or vibegron). Main Outcomes and Measures Incident dementia by diagnosis code or dementia-specific medication. Inverse probability of treatment weighting (IPTW) and Cox proportional hazards models were used, with analyses stratified by age and sex and an exploratory landmark duration-response analysis. Results Among 830 393 patients (585 198 female [70.5%]; median [IQR] age, 67 [57-76] years), incident dementia occurred in 65 134 patients, with a median (IQR) follow-up ranging from 1046 (571-1809) days for β3 agonists to 1568 (857-2614) days for high-risk anticholinergics. After IPTW, high-risk anticholinergics were associated with a lower hazard of dementia than β3 agonists (hazard ratio [HR], 0.90; 95% CI, 0.87-0.93) and low-risk anticholinergics (HR, 0.93; 95% CI, 0.90-0.96); low-risk anticholinergics did not differ from β3 agonists (HR, 0.97; 95% CI, 0.93-1.01). Weighted 10-year dementia risk was highest for β3 agonists (20.0%; 95% CI, 18.8%-21.2%) and lowest for high-risk anticholinergics (18.1%; 95% CI, 17.9%-18.3%). E-values of 1.29 (high-risk anticholinergics vs β3 agonists) and 1.20 (high-risk vs low-risk anticholinergics) suggested that observed differences could be explained by a modest amount of unmeasured confounding. Longer anticholinergic duration was not associated with increased risk. Conclusions and Relevance In this cohort study of patients undergoing treatment for OAB, β3 agonists were not associated with a lower dementia incidence compared with anticholinergics. These findings suggest that previously reported increases in risk may reflect residual confounding rather than a pharmacological effect.