Abstract 13423: Sleep Disruption is Associated With Incident Atrial Fibrillation Independent of Obstructive Sleep Apnea

Introduction: Obstructive sleep apnea (OSA) is an established risk factor for atrial fibrillation (AF), but the mechanism is unknown. Repeated hypopneas may induce mechanical stress but OSA also causes poor sleep which has been linked to other CVD. We sought to determine if sleep disruption independent of OSA is associated with AF. Methods: We compared sleep quality to AF in three population-based cohorts. (1) In the Health eHeart Study (HeH), an internet-based cohort study, we compared self-reported frequent nighttime awakening (waking at night ≥ 3x / week) to prevalent AF using logistic regression. (2) In the Cardiovascular Health Study, (CHS) we used Cox proportional hazards models to assess self-reported frequent nighttime awakening at baseline as a predictor of incident AF. (3) Last, we used statewide medical records in the California Healthcare Cost and Utilization Project (C-HCUP) from 2005 to 2009 to compare the diagnosis of insomnia to incident AF, again with Cox proportional hazards models. Results: Sleep disruption was associated with prevalent AF (HeH) and incident AF (CHS, C-HCUP) before and after adjustment for OSA and other potential confounders ( Figure ). (1) In HeH, 553/4603 (12%) had AF, and frequent nighttime awakening was associated with increased odds of AF. (2) In CHS, over a median 11.6 years (IQR 6.2 - 16.4) of follow-up 1593 (28%) developed AF; frequent nighttime awakening was associated with increased risk of incident AF. (3) Among 13,967,949 patients in C-HCUP over the 5-year study period, of which 375,318 (3%) developed AF over a median of 3.8 years (IQR 1.8 - 4.3), insomnia was associated with an increased risk of incident AF. Conclusions: Sleep disruption was associated with an increased risk of prevalent and incident AF after adjustment for OSA and other confounders in 3 diverse cohorts. Factors inherent to sleep may be important in AF pathogenesis and sleep quality may therefore be a modifiable risk factor for AF.