Association of autoimmune diseases with the occurrence and 28-day mortality of sepsis: an observational and Mendelian randomization study

Abstract Background Observational studies have indicated a potential association between autoimmune diseases and the occurrence of sepsis, with an increased risk of mortality among affected patients. However, whether a causal relationship exists between the two remains unknown. Methods In the Mendelian randomization (MR) study, we accessed exposure Genome-wide association study (GWAS) data from both the MRC Integrative Epidemiology Unit (MRC-IEU) and the FinnGen consortium. GWAS data for sepsis and its 28-day mortality were obtained from MRC-IEU. We employed univariable, multivariable, and reverse MR analyses to explore potential associations between autoimmune disorders and sepsis and its 28-day mortality. Additionally, a two-step mediation MR analysis was performed to investigate indirect factors possibly influencing the relationship between autoimmune disorders and sepsis. Afterward, we conducted an observational analysis to further explore the relationship between autoimmune disease and occurrence as well as 28-day mortality of sepsis using a real-world database (the MIMIC-IV database). A cohort of 2537 patients diagnosed with autoimmune disease were extracted from the database for analysis. Multivariable logistic regression models were used to confirm the association between autoimmune diseases and the occurrence of sepsis, as well as the 28-day mortality associated with sepsis. Results In univariable MR analysis, there appeared to be causal relationships between genetically predicted type 1 diabetes (OR = 1.036, 95% CI = 1.023–1.048, p = 9.130E-09), rheumatoid arthritis (OR = 1.077, 95% CI = 1.058–1.097, p = 1.00E-15) and sepsis, while a potential causal link was observed between celiac disease and sepsis (OR = 1.013, 95% CI = 1.002–1.024, p = 0.026). In a subsequent multivariable MR analysis, only rheumatoid arthritis was found to be independently associated with the risk of sepsis (OR = 1.138, 95% CI = 1.044–1.240, p = 3.36E-03). Furthermore, there was no causal link between autoimmune disorders and 28-day mortality from sepsis. In reverse MR analysis, sepsis was suggested to potentially trigger the onset of psoriasis (OR = 1.084, 95% CI = 1.040–1.131, p = 1.488E-04). In the real-world observational study, adjusting for multiple confounders, rheumatoid arthritis (OR = 1.34, 95% CI = 1.11–1.64, p = 0.003) and multiple sclerosis (OR = 1.31, 95% CI = 1.03–1.68, p = 0.02) were associated with a higher risk of sepsis. In addition, we did not find that autoimmune diseases were associated with 28-day mortality from sepsis. Conclusion Both in observational and MR analysis, only rheumatoid arthritis is highly correlated with occurrence of sepsis. However, autoimmune disease was not associated with an increased 28-day mortality in patient with sepsis. Sepsis may increase the risk of developing psoriasis.