Association of Sleep Duration With Atrial Fibrillation and Heart Failure: A Mendelian Randomization Analysis

Both short (<7 h per night) and long (≥9 h per night) sleep durations are related to atrial fibrillation (AF) and heart failure (HF), but their causality has not been confirmed. We applied Mendelian randomization (MR) approaches to estimate the causal association between genetically determined sleep duration and the risk of AF and HF. We performed two-sample MR analysis to obtain the effect of sleep duration on AF and HF. Instrumental variables were constructed using genetic variants known to be associated with continuous sleep duration, short sleep duration, and long sleep duration. MR estimates of the effect of sleep duration on AF and HF were derived based on two large meta-analyses of genome-wide association studies. The pooled MR estimate demonstrated a significant protective effect of continuous sleep duration on HF [odds ratio (OR) = 0.765, 95% confidence interval (CI) = 0.675-0.867; P = 2.64 × 10^-5] and a suggestive inverse association of continuous sleep duration with AF (OR = 0.893, 95% CI = 0.804-0.991; P = 0.034). In addition, the results showed a suggestive detrimental effect of short sleep duration on the risk of AF (OR = 1.108, 95% CI = 1.017-1.207; P = 0.019) and HF (OR = 1.136, 95% CI = 1.025-1.258; P = 0.015). Conversely, there is no significant evidence for the causal protective effect of long sleep duration on AF (OR = 0.956, P = 0.410) and HF (OR = 0.921, P = 0.202). This MR study indicated that genetically determined continuous sleep duration has a significant protective effect on HF and a suggestive inverse association with AF. Short sleep duration is positively associated with the risk of AF and HF. Nevertheless, there is no significant evidence for the causal protective effect of long sleep duration on AF and HF. Larger intervention studies are required to confirm the effectiveness of improving sleep on reducing the incidence of AF and HF.