Gut microbiota-mediated causal relations between sleep disorders and stroke: a Mendelian randomization study
摘要
Sleep disorders are closely linked to stroke, while the role of gut microbiota in human health and disease is increasingly recognized. However, the causal relationships between sleep disorders, gut microbiota, and stroke remain unclear.
ObjectiveThis study aimed to investigate the causal relationships between sleep disorders, gut microbiota, and stroke (including its subtypes), as well as the potential mediating role of gut microbiota in the association between sleep disorders and stroke.
MethodsUsing a two-sample Mendelian randomization approach, we analyzed causal relationships among various sleep disorders, gut microbiota, and stroke. The inverse variance weighted method served as the primary analysis, supplemented by multiple sensitivity analyses to ensure robustness. Mediation effects of gut microbiota in the sleep disorder-stroke relationship were assessed using the coefficient product method.
ResultsInsomnia was causally associated with transient ischemic attacks (OR = 1.592, 95% CI: 1.116–2.270), excessive daytime sleepiness with ischemic stroke (IS) (OR = 1.887, 95% CI: 1.041–3.420), and sleep apnea with overall stroke risk (OR = 1.247, 95% CI: 1.015–1.532). Several gut microbiota taxa showed causal associations with stroke and its subtypes, with Ruminococcus E sp900314705 increasing the risk of overall stroke, IS, and transient ischemic attacks. Ezakiellaceae partially mediated the causal relationship between daytime sleepiness and IS, accounting for 15.9% of the effect.
ConclusionThe current study established and elucidated causal relationships between sleep disorders and stroke, identified specific gut microbiota associated with stroke risk, and highlighted the mediating role of Ezakiellaceae in the causal relationship between daytime sleepiness and IS. These findings underscore the potential importance of improving sleep quality and modulating gut microbiota for stroke prevention and management.