The Relationship between Circulating Immune Cell Phenotypes and Sepsis: A Mendelian Randomization Study

Objective: The role of immune cells in sepsis remains unclear, and there is some controversy. Here, we aim to systematically assess whether distinct immune cell phenotypes impact the susceptibility to sepsis. Methods: In this study, we harnessed publicly available summary-level data from genome-wide association studies (GWASs). The selection of genetic variations strongly associated with 731 phenotypes of circulating immune cells served as instrumental variables (IVs). Using a two-sample Mendelian randomization (MR) analysis, we investigated the relationships between different immunophenotypes and the occurrence of sepsis, as well as the 28-day mortality. The MR study utilized the inverse variance weighting (IVW) method as the main analytical approach. In addition, we incorporated four other MR methods for supplementary causal inference, including weighted median (WME), MR-Egger regression, simple mode, and weighted mode. Furthermore, the robustness of the results was affirmed through multiple sensitivity analyses. Results: The results of the IVW method indicated that a total of 36 immunophenotypes are associated with the risk of sepsis. We also identified 34 immunophenotypes with a causal association with the 28-day mortality. Interestingly, before multiple testing corrections, 11 immunophenotypes were determined to have consistent causal relationships with both the occurrence of sepsis and the 28-day mortality. Notably, after false discovery rate (FDR) correction, four immunophenotypes were found to be significantly correlated with susceptibility to sepsis: CD45RA- CD4+ %CD4+ (odds ratio [OR], 1.355; 95% confidence interval [CI], 1.139~1.611; P < 0.001, PFDR = 0.192), HLA DR on HLA DR+ NK (OR, 0.818; 95% CI, 0.726~0.922; P = 0.001, PFDR = 0.192), IgD+ CD24+ %B cell (OR, 0.626; 95% CI, 0.473~0.828; P = 0.001, PFDR = 0.192), and TD DN (CD4- CD8-) AC (OR, 0.655; 95% CI, 0.510~0.840; P < 0.001, PFDR = 0.192). Following FDR correction, only one immunophenotype was confirmed to be negatively correlated with the 28-day mortality: CD39 on CD39+ CD8br (OR, 0.820; 95% CI, 0.737~0.912; P < 0.001, PFDR = 0.184). Conclusion: This study, for the first time, has uncovered indicative evidence of a causal relationship between circulating immune cell phenotypes and varying degrees of sepsis through genetic means. These findings underscore the significance of immune cells in the pathogenesis of sepsis.