Objectives Our objective was to investigate the actual incidence and clinical determinants of cough leading to discontinuation of ACE-inhibitors. Cough is the most frequent reason to stop ACE-inhibitor treatment.
Methods We studied 27,492 ACE-inhibitor naïve patients randomized to the ACE-inhibitor perindopril or placebo using individual data of 3 clinical trials. Multivariate logistic regression analysis was used to study the incidence of cough in relation to baseline clinical characteristics including racial background.
Results In 27,492 patients with cardiovascular disease, 1076 patients discontinued ACE-inhibitor perindopril due to cough (3.9%), 703 patients during run-in period of 4 weeks and 373 patients during a mean four years of follow-up. Significant determinants of cough were female gender (OR 1.92 95% CI 1.68-2.18), age above 65 years (OR 1.53 95% CI 1.35-1.73), and concomitant use of lipid-lowering agents (OR 1.37; 95% CI 1.18-1.59). A simple clinical risk score composed of these 3 predictors of cough mounted to an odds ratio of 4.4 (95% CI 3.1-5.4) in the subjects with highest score (i.e. all determinants present). Racial background was not related to a differential incidence of cough in patients of Caucasian or Asian descendent (OR 1.11 95% CI 0.92-1.39). Conclusion This large combined analysis of randomized clinical trials in 27,492 patients showed an overall lower incidence of cough leading to discontinuation of ACE-inhibitors (3.9%) as compared to literature. Clinical determinants of such cough are older age, female gender and concomitant use of lipid-lowering agents. In contrast, racial differences were not related to the incidence of cough.