Causal effect of negative emotions and insomnia on sepsis and its associated clinical indicators: A Mendelian randomisation and mediation analysis

Abstract

Abstract Background Negative emotions and insomnia (NEI) are associated with changes in inflammatory factors, which play a role in sepsis. Methods We performed Mendelian randomisation (MR) analysis of genome-wide association study (GWAS) data of NEI and sepsis to investigate the causal effect of NEI on sepsis. We employed linkage disequilibrium score regression (LDSC) to calculate the genetic correlation between NEI and sepsis. Inverse variance weighting (IVW) was primarily used for investigating causality, while the weighted median and MR-Egger methods ensured the reliability of the findings. To assess heterogeneity, we employed RadialMR and Cochran’s Q test, and we used MR-Egger regression and Mendelian randomisation pleiotropy residual sum and outlier analyses to evaluate the bias of gene polymorphism. Mendelian mediation analysis was conducted to quantify the intermediate effect of inflammatory factors in mediating the relationship between NEI and sepsis, including the percentage of this mediating effect. Results LDSC analysis revealed a genetic correlation between NEI and sepsis. Two-sample MR analysis revealed a causal relationship between NEI and sepsis (odds ratio = 1.21, 95% confidence interval: 1.08–1.36, p = 1.07×10− 3), with no significant heterogeneity and pleiotropy bias. Mendelian mediation analysis revealed an intermediate effect of NEI on sepsis mediated by chitinase 3-like 1 (CHI3L1) (0.12, 10.31%). Conclusions Our findings prove the causal relationship between NEI and sepsis. We identified CHI3L1 as a potential mediator, offering insight into the pathogenesis of sepsis.