Bidirectional Causal Relationship Between Sleep Duration, Insomnia, and Cerebral Infarction: A Two-Sample Mendelian Randomization Study

Abstract

Objective: Observational studies suggest that abnormal sleep duration and insomnia are associated with stroke, but residual confounding factors, reverse causal bias, and the nonlinear characteristics of exposure-response relationships limit precise causal inference. This study employed a two-sample Mendelian randomization (MR) approach to systematically investigate the bidirectional causal relationship between sleep duration, insomnia, and cerebral infarction, clarifying their genetic causal effects. Methods: Genome-wide association study (GWAS) summary statistics were retrieved from the IEU OpenGWAS platform for sleep duration (ieu-a-1088; 128,266 participants of European descent), insomnia (ukb-b-3957; 462,341 participants of European descent), and cerebral infarction (ukb-d-I63; 361,194 participants of European descent). The inverse variance weighted (IVW) method was used as the core analytical approach, combined with the maximum likelihood method, MR-Egger regression, weighted median method, and weighted mode method for sensitivity analysis. Cochran’s Q test quantified heterogeneity across genetic instruments, while the MR-Egger intercept test was applied to detect directional horizontal pleiotropy. Results: After data harmonization, 26, 38, 24, and 128 genetic variants were included in the analyses of sleep duration → cerebral infarction, insomnia → cerebral infarction, cerebral infarction → sleep duration, and cerebral infarction → insomnia, respectively. IVW analysis revealed no significant causal effect of genetically predicted sleep duration on cerebral infarction [OR=1.002, 95% confidence interval (CI): 0.998-1.006, P=0.252], nor did genetic susceptibility to insomnia significantly affect the risk of cerebral infarction (OR=0.994, 95% CI: 0.988-1.001, P=0.100). In reverse analyses, genetic susceptibility to cerebral infarction showed no significant causal relationship with sleep duration (β=0.394, 95% CI: -1.620-2.408, P=0.701) or the insomnia phenotype (β=0.057 SD, 95% CI: -0.245-0.360, P=0.711). The results of various supplementary analyses were consistent with the core IVW results, demonstrating stable conclusions. Neither the Cochran’s Q test nor the MR-Egger intercept test detected significant heterogeneity or directional horizontal pleiotropy. Conclusion: This bidirectional MR study indicates that there is currently no reliable genetic evidence supporting a bidirectional causal relationship between sleep duration, insomnia, and cerebral infarction. Their clinical correlation may be largely mediated by non-genetic confounding factors.