Prevalence and clinical predictors of sleep apnea in patients with paroxysmal atrial fibrillation

Background: Recent studies have suggested a strong association between sleep apnea (SA) and atrial fibrillation (AF). Aims and Objectives: In this study we aimed to examine clinical predictors to identify SA and the prevalence and type of SA in patients with paroxysmal AF. Methods: We prospectively studied 578 patients with paroxysmal AF. SA was diagnosed using portable polygraphy two nights at home. Sleep studies were analyzed by an experienced specialist, using standard American Academy of Sleep Medicine definitions. The degree of subjective daytime sleepiness was assessed by Epworth Sleepiness Scale (ESS). Body composition data were collected using a body composition analyzer (Tanita BC-545N). We calculated differences in variables between SA categories by Chi-Square Test or ANOVA as appropriate. Results: 158 women (27 %) and 420 men (73%) were enrolled. Mean age (SD) was 60 (10) years and mean BMI 29 (4) kg/m². 477 (83%) of the patients had an apnea-hypopnea index (AHI) ≥ 5/h, while moderate/severe SA (AHI ≥ 15/h) was diagnosed in 243 patients (42%). The type of SA was predominantly obstructive, with mean AHI 15.3/h (12) (range 0.4/h-85.8/h). Mean central apnea index was 0.8/h (2). AHI increased with age, BMI, weight, neck-and waist circumference, body fat, visceral fat and metabolic age. Mean ESS score was 7 (4). There was no association between ESS and AHI/h, r=0.08. Conclusions: In our AF population the prevalence of SA (AHI ≥ 5/h) was 83 %, which is several times higher than in the general population (16%). The high prevalence of SA detected in this study may indicate that SA is under-recognized in patients with AF. We found a positive correlation between body composition and SA severity.