Severe Central Sleep Apnea Is Associated with Atrial Fibrillation in Patients with Left Ventricular Systolic Dysfunction

Background The results of previous studies investigating the association between atrial fibrillation (AF) and central sleep apnea (CSA) in patients with left ventricular (LV) systolic dysfunction are contradictory. Methods We prospectively enrolled 267 patients in this cross‐sectional study with LV ejection fractions ≤50%, who were screened for sleep disordered breathing using cardiorespiratory polysomnography after patients with predominantly obstructive sleep apnea or insufficient sleep studies had been excluded. Results AF at study entry was found in 70 of 267 patients (26%). CSA with an apnea/hypopnea index (AHI) ≥15/hour was present in 116 patients (43%) and 67 patients (25%) had severe CSA with an AHI > 30/hour. Univariate analysis revealed a significant association between AF and severe CSA, age, male gender, arterial hypertension, left atrial diameter, brain natriuretic peptide, chronic kidney disease, New York Heart Association class, digitalis, and the lack of angiotensin‐converting enzyme inhibitors or angiotensin receptor blockers. Multivariate analysis revealed a significant association between AF and severe CSA (odds ratio [OR]: 5.21; 95% confidence interval [CI]: 1.67–16.27, P = 0.01), age (OR: 1.22 per 5‐year increase; 95% CI: 1.05–1.40, P = 0.01), left atrial diameter (OR 1.61 per 5‐mm increase; 95% CI: 1.22–2.01, P < 0.01), and digitalis (OR: 2.7; 95% CI: 1.26–5.79, P = 0.01). Conclusions AF is associated with severe CSA but not with moderate CSA in addition to age, use of digitalis, and left atrial size in patients with LV systolic dysfunction. Future studies evaluating the potential benefit of adaptive servo‐ventilation therapy to prevent AF or to decrease the AF burden in heart failure patients should therefore focus on patients with severe central sleep apnea.