Content area
Abstract
Abstract
Although the association of sleep disordered breathing (SDB) and nocturnal cardiac arrhythmia is well-established, there is a limited understanding of the relationship between SDB with risk of incident arrhythmias in community based samples based upon objectively identified cardiac arrhythmia not limited to self-report or adjudication of hospital-based arrhythmic events. We hypothesize that SDB is associated with incident clinically significant cardiac arrhythmia on overnight polysomnogram (PSG) ECG after considering potential confounders.
In this multi-center prospective study (Outcomes of Sleep Disorders in Older Men Study), 980 participants without baseline PSG-identified atrial fibrillation (AF) or pacemakers who completed baseline and follow-up exam (follow-up 7.8 ± 2.5 years) were identified. PSG ECG processing was performed via the Somte® software and visual inspection/scoring for arrhythmia subtype (AF and non-sustained ventricular tachycardia (NSVT), the latter defined as >3 consecutive ventricular ectopic beats, >120bpm. Logistic regression models were used to analyze the SDB severity predictors, i.e. the Apnea Hypopnea Index (AHI) as a continuous variable and dichotomized at 15, and hypoxemia (% total sleep time with SaO2 < 90% dichotomized at 10%, TSTO2>10%) with odds ratios and 95% confidence intervals presented. Models were adjusted for age, race, body mass index (BMI), self-reported coronary disease and hypertension, and cardiac medications (beta blockade and/or calcium channel blockade medications).
Participants were 74.6 ± 4.7 years, 87.1% Caucasian with BMI=27.1 ± 3.6 kg/m2). The AHI (continuous and dichotomized) was not associated with incident AF and/or NSVT in the fully adjusted models. In contrast, hypoxemia (TSTO2>10%) was associated with incident AF (2.53, 1.21-5.26), NSVT (2.26, 1.13-4.54) and composite AF or NSVT (2.27, 1.30-3.98) in the unadjusted model with attenuation of this association in the fully adjusted model: incident AF (1.81, 0.80-4.08), NSVT (2.11, 0.99-4.50) and composite AF or NSVT (1.90, 1.03-3.49).
/b> /b> Incident AF or NSVT identified on overnight PSG was associated with baseline overnight hypoxemia but not AHI. These findings support the utility of nocturnal hypoxia as a marker, and potential risk factor, for nocturnal cardiac arrhythmia development.