In this historical cohort study, 10,701 consecutive adults undergoing their first diagnostic polysomnogram between July 1987 and July 2003 were included. During follow-up up to 15 years, incident resuscitated or fatal sudden cardiac death (SCD) was assessed in relation to the presence of OSA, physiological data including the apnea-hypopnea index (AHI), and nocturnal oxygen saturation (O2sat) parameters.Authors found that, during an average follow-up of 5.3 years, 142 patients had resuscitated or fatal SCD (annual rate 0.27%). A multivariate analysis was performed and independent risk factors for SCD were found to be age, hypertension, coronary artery disease, cardiomyopathy or heart failure, ventricular ectopy or non-sustained ventricular tachycardia, and lowest nocturnal O2sat (per 10% decrease, hazard ratio [HR]:1.14;p=0.029). SCD was best predicted by age >60 years (HR:5.53), followed by nocturnal O2sat <93% (HR:2.93), and lowest nocturnal O2sat <78% (HR:2.60) and apnea-hypopnea index >20 (HR:1.60, all p<0.0001).
Comments. In this large population based study, OSA diagnosis independently predicted incident SCD. The authors correlated increasing risk of SCD with the severity of OSA (AHI>20) and the severity of nocturnal hypoxemia, indicating good biological plausibility. It was however difficult to determine how well the patients were treated with therapy such as CPAP during the follow-up period.