Abstract
BACKGROUND AND PURPOSE: Cheyne-Stokes respiration (CSR) is frequently observed in patients with acute stroke. There have been conflicting opinions about the associations of CSR with the location and size of the lesion. We aimed to better define the clinical relevance and pathogenesis of CSR in acute stroke. METHODS: We investigated patients who had been admitted with acute ischemic stroke and received an overnight sleep apnea test. We collected data on demographics, risk factors, etiologic subtypes, initial vital signs, clinical course of the stroke, and parameters associated with respiratory events during the sleep apnea test. We performed a multivariate logistic regression analysis to determine the factors associated with CSR. RESULTS: Among 182 patients, 35 patients showed CSR in sleep apnea testing. Large-artery atherosclerosis or cardioembolism, bilateral hemispheric involvement, atrial fibrillation, low left-ventricle ejection fraction (LVEF), and left atrium (LA) enlargement were all associated with the presence of CSR. Multivariate analysis revealed that the previous modified Rankin Scale (mRS) score, bilateral hemispheric involvement, low LVEF, and LA enlargement were significantly associated with CSR. Subgroup analysis with large-artery atherosclerosis without cardiac disease revealed that the previous mRS score is the only independent factor associated with CSR. CONCLUSIONS: CSR frequently occurs in strokes involving large arteries or due to cardioembolism, regardless of the location and severity of the stroke. Predisposing conditions such as preexisting neurologic disability, low LVEF, and LA enlargement are associated with CSR in acute stroke.