[Causal relationship between sleep and coronary artery disease: a Mendelian randomization study]

Objective: To understand the causal relationship between sleep and coronary artery disease (CAD). Methods: This study included six genome-wide association parts; five for sleep related traits[sleep duration (continuous variable), long sleep duration (binary variable), short sleep duration (binary variable), early-to-bed/up habit (known as 'morningness') and frequently insomnia] and one for CAD. Heterogeneity in dependent instrument approach was used to assess and to remove the pleiotropic instruments. Generalized summary data-based Mendelian randomization was performed to estimate the causal relationships between sleep related traits and CAD. And a conservative Bonferroni was used for statistical tests. Results: Sleep duration was putatively causal for CAD (OR=0.755, 95%CI: 0.658-0.867, P=6.68E-05). Our results also indicated significant causal effects between both short sleep duration and frequently insomnia on CAD, with ORs as 4.251 (95%CI: 2.396-7.541, P=7.51E-07) and 1.814 (95%CI: 1.346-2.446, P=9.25E-05), respectively. There was no convincing evidence of causality between long sleep duration or morningness with CAD. Conclusions: Our findings suggested that both sleep duration and frequently insomnia played causal roles on CAD, indicating that disease models should include sleep duration and insomnia as potential factors for CAD to develop effective interventions.