Abstract MP57: Objectively-Assessed Sleep Duration And Incident Heart Failure: The Atherosclerosis Risk In Communities Study-Sleep Heart Health Study

Abstract

Introduction: Evidence is mixed on short sleep duration and heart failure (HF) risk. Notably, no studies have investigated objectively-assessed sleep duration or controlled for comorbid sleep apnea in this context. Objective: To examine the independent associations of objective short sleep duration with incident HF, accounting for sleep apnea. Methods: We studied ARIC-SHHS study participants free of HF at visit 4 (1996-98). Sleep duration was measured by at-home polysomnography (PSG) and defined as: 7 to <9 hours (reference), 6 to <7 hours, 5 to <6 hours, and <5 hours. The apnea-hypopnea index (AHI; defined as: <5, 5 to <15, 15 to <30, and ≥30), a measure of sleep apnea, was also measured by at-home PSG. Cox proportional hazards models, with progressive adjustment for comorbidities and AHI, were used to evaluate the association of sleep duration with incident HF (hospitalizations or deaths with HF diagnosis). Results: We included 1,056 participants (mean age 62 years, 50% female, 99% White). During a median follow-up of 22 years, there were 211 HF events. Participants with sleep duration <5 hours had a higher risk of incident HF in unadjusted and age-sex-adjusted models, with corresponding hazard ratios of 2.44 (95% CI : 1.40-4.26) and 1.88 (95% CI : 1.07-3.29) ( Table ). However, further adjustment for clinical factors attenuated the results. The associations for sleep duration appeared more evident when we restricted our analysis to participants without prevalent cardiovascular disease ( Table ). The AHI was not a significant predictor of incident HF in the presence of sleep duration. Conclusions: Objectively-assessed short sleep duration <5 hours was modestly associated with HF risk. Our study suggests the need and value of larger studies exploring objectively-assessed sleep duration in terms of HF risk. Importantly, sleep duration may be objectively measured without PSG, using wearable devices.