Baseline anticholinergic burden from medications predicts incident fatal and non-fatal stroke in the EPIC-Norfolk general population

Background: Stroke is primarily a disease of older age, with a substantial impact on global mortality and morbidity. Medications with anticholinergic effects are widely used, but no studies have been conducted to examine the relationship between anticholinergic burden (ACB) and stroke in a general population. 
Method: The sample was drawn from the EPIC-Norfolk cohort. Baseline assessments were carried out during 1993–97 and participants were followed up until March 2016. Participants were divided into four groups according to their total ACB score at baseline; these groups were those with a total ACB score of 0, 1, 2–3 and >3. After exclusion, Cox proportional hazards models were constructed to determine the associations between the ACB score groups and the risk of incident stroke and stroke mortality. Sensitivity analysis and propensity score matched analyses were performed. 
Results: In total 25 639 participants attended the first health check; 3917 participants were excluded, leaving 21 722 participants to be included. Participants had a mean age [standard deviation (SD)] of 58.9 (9.2) years (54.4% women). Of these, 2131 suffered incident stroke and 562 died from stroke. Mean follow-up was approximately 18 years for both outcomes. In the fully adjusted model, those with an ACB of >3 had 59% relative risk of incident stroke {hazard ratio [HR] [95% confidence interval (CI) 1.59 [1.34–1.89]} and 86% relative risk of stroke mortality [1.86 (1.37–2.53)] compared with those in ACB 0 category. Sensitivity analyses and propensity score matched analyses showed similar results. 
Conclusions: Our results provide an incentive for the cautious use of medications with anticholinergic properties, to help reduce the global burden of stroke.