[Causal association between immune cells and sepsis: a based on Mendelian randomization method study].

Objective: To investigate the causal association between immune cell and different types of sepsis by using Mendelian randomization (MR) method, and to find the immune cell phenotypes causally associated with sepsis.

Methods: Summary data for various circulating immune cell phenotypes were obtained from the GWAS catalog (GCST90001391-GCST90002121). Sepsis data were sourced from the UK Biobank database. Single nucleotide polymorphisms (SNP) were used as instrumental variables. The correlation threshold of P < 5×10^-6 was used to identify the strongly correlated instrumental variables, and the code was used to remove the linkage disequilibrium and the instrumental variables with F-value < 10. Inverse variance weighting (IVW) was used as the main research method to evaluate the stability and reliability of the results, including Cochran's Q test, MR-Egger regression and Leave one out. Reverse MR analysis was performed based on the immunophenotypic results of the removal of horizontal pleiotropy, and the immune cell phenotype with one-way causal association was obtained. Odds ratio (OR) and 95% confidence interval (95%CI) were used to represent the effect value of the results.

Results: CD16 on CD14^-CD16⁺; monocyte had horizontal pleiotropy in sepsis (OR = 0.965 4, 95%CI was 0.933 5-0.998 3, P = 0.039 6). There were five immunophenotypes that had reverse causal associations with the types associated with sepsis. After excluding immune cell phenotypes with horizontal pleiotropy and reverse causation, a total of 42 immune cell phenotypes with sepsis, 36 immune cell phenotypes with sepsis (28-day death in critical care), 32 immune cell phenotypes with sepsis (critical care), 44 immune cell phenotypes with sepsis (28-day death), and 30 immune cell phenotypes had potential causal associations with sepsis (under 75 years old). After false discovery rate (FDR) correction, the correlations between BAFF-R on IgD^- CD38br and sepsis (28-day death) were negative and strong (OR = 0.737 8, 95%CI was 0.635 9-0.856 0, P = 6.05×10^-5, P_FDR = 0.044 2).

Conclusions: A variety of immune cell phenotypes may have a protective effect on sepsis, especially BAFF-R on IgD^- CD38br expression is negatively correlated with sepsis (28-day death), which provides a new idea for immune modulation therapy in sepsis.