Abstract
BACKGROUND: The association between inflammatory biomarkers and individual delirium symptoms remains controversial in observational studies. We investigated the relationship between inflammatory biomarkers and the risk of developing delirium. METHODS: A bidirectional two-sample Mendelian randomization (MR) was performed. Genetic instruments associated with peripheral tumor necrosis factor-a (TNF-a) C-reactive protein (CRP), interleukin (IL)-1α, IL-1β, IL-2, IL-8, IL-6, soluble IL-6 receptor alpha (sIL-6Rα), and soluble gp130 were identified in three different large summary genome-wide association studies (GWAS) conducted in the European population. Summary-level statistics for delirium not induced by alcohol and other psychoactive substances were obtained from the FinnGen consortium (2,612 cases and 325,306 controls). The estimated causal effects were performed using instruments' variants at the genome-wide significant level (P < 5e-8 and P < 5e-6), applying a linkage disequilibrium clumping approach with a threshold of r(2) < 0.001 for each of the exposures. Reverse causation was also performed. The inverse-variance weighted method (IVW), MR-Egger method, weighted median method, MR-Egger regression, and MR Pleiotropy RESidual Sum were used for MR analyses. RESULTS: At the genome-wide significant level (P < 5e-8, r(2) < 0.001), genetically predicted sIL-6Rα was significantly associated with a decreased risk of delirium with less than three single-nucleotide polymorphisms (SNPs) in all three GWAS data sources (OR(Waldratio) = 0.89, 95% CI: 0.79-0.96, P(Waldratio) = 0.0016; OR(IVW) = 0.88, 95% CI: 0.79-0.97, P(IVW) = 0.008; OR(IVW) = 0.88, 95% CI: 0.80-0.96, P(IVW) = 0.004). The causal relationship between sIL-6Rα and delirium became non-significant when a more liberal threshold of P of < 5e-6 was applied (all P(IVW) > 0.05). At the two genome-wide significance levels (P < 5e-8 and P < 5e-6), we found no evidence for the causal effects of peripheral TNF-α, CRP, IL-1α, IL-1β, IL-2, IL-6, IL-8, and soluble gp130 on delirium (all P > 0.05). The MR-Egger intercept and MR-PRESSO results indicated that no SNP had possible pleiotropy (all P > 0.05). Regarding the reverse, no evidence for an effect of delirium on these inflammatory biomarkers could be found (all P > 0.05). CONCLUSION: The results of this MR analysis did not support that peripheral TNF-α, CRP, IL-1α, IL-1β, IL-2, IL-6, sIL-6Rα, soluble gp130, and IL-8 were causally associated with delirium. More research is needed to explore the role of inflammatory factors in the pathogenesis of delirium.