Abstract
Although the potential causal associations between cell-derived signaling molecules and sleep disorder (SD) have been reported, contradictions remain. This study assessed the causal effects and the mediating role of 1400 metabolites among 91 cell-derived signaling molecules and SD from a genetic perspective by performing Mendelian randomization (MR) analyses. Genetic instruments derived from publicly available genome-wide association studies were employed in this study, including 49,880 SD cases and 358,194 controls. Summary statistics of 1400 circulating metabolites were obtained from a cohort of 8299 individuals. The 91 cell-derived signaling molecules were derived from genome-wide association studies data from 11 cohorts comprising 14,824 samples. Multiple statistical analyses were introduced in this study, with inverse variance weighted as the core analysis method, supplemented by 4 additional methods. Besides, various sensitivity analyses were employed to identify horizontal pleiotropy and heterogeneity, thereby evaluating the robustness of the results. Forward MR analysis indicated positive associations between SD and the levels of eotaxin (OR = 1.041, 95% CI: 1.001-1.084, P = .046), CUB domain-containing protein 1 (OR = 1.042, 95% CI: 1.008-1.077, P = .016), interleukin-20 receptor subunit alpha (IL-20RA) (OR = 1.086, 95% CI: 1.031-1.143, P = .002), while the levels of CD40L receptor (OR = 0.968, 95% CI: 0.942-0.994, P = .018), glial cell line-derived neurotrophic factor (OR = 0.947, 95% CI: 0.910-0.986, P = .009) act as the opposite. Reverse MR analysis pointed out that the genetic susceptibility to SD raised interleukin-5 levels. According to the mediation analysis, N-lactoyl-tyrosine levels mediated the increased risks of SD associated with elevated IL-20RA levels, with a mediation effect of 0.009 (95% CI: 0.001-0.018, P = .034), accounting for 11.5% of the total. The study proved the causal associations between 91 cell-derived signaling molecules and SD, confirming that eotaxin, CUB domain-containing protein 1, and IL-20RA may increase the risk of SD, while CD40L receptor and glial cell line-derived neurotrophic factor may act as the opposite. Besides, the study provided abundant evidence for the potential mediating effect of N-lactoyl-tyrosine in the pathway linking IL-20RA and SD risk. To summarize, the findings of this study may benefit the understanding of the pathogenic mechanisms through which cell-derived signaling molecules influence SD.