Obstructive sleep apnea severity and prevalent atrial fibrillation in a sleep clinic cohort with versus without excessive daytime sleepiness

Background: Obstructive sleep apnea (OSA) is associated with atrial fibrillation (AF) in cardiac cohorts. Less is known regarding the magnitude of this association in a sleep clinic cohort with vs. without excessive daytime sleepiness (EDS). Objectives: To explore the association of OSA severity with AF in a sleep clinic cohort stratified by EDS. Patients and methods: All consecutive adults (n = 3814) admitted to the Skaraborg Hospital, Sweden between Jan 2005 and December 2011 were registered in a local database, and the follow-up ended in December 2018. OSA was defined as an apnea-hypopnea index (AHI) >= 5 events/h. Mild OSA was defined as AHI >= 5 & AHI<15 events/ h; moderate OSA as AHI >= 15 & AHI<30 events/h; and severe OSA as AHI >= 30 events/h. EDS was defined as an Epworth Sleepiness Scale score >= 11. We conducted cross-sectional analyzes of the prevalent AF across the OSA severity categories and logistic regression analyzes stratified by EDS.Results: In all, 202 patients (5.3%) had AF at baseline, 1.6% in no-OSA, 3.9% in mild OSA, 5.2% in moderate OSA, and 7.6% in severe OSA, respectively (p < 0.001). The stratified analyzes revealed that patients with severe OSA without EDS had an increased risk for prevalent AF (OR 2.54, 95% CI 1.05-6.16; p = 0.039) independent of the confounding factors. Conclusions: There was an independent dose-response relationship between OSA and prevalent AF among the non-sleepy phenotype in this sleep clinic cohort. Since adherence to OSA treatment is challenging in the absence of EDS, these patients may have increased risk for adverse cardiovascular outcomes.