No causal association between COVID-19 and sepsis: a bidirectional two-sample Mendelian randomization study.

Abstract

Sepsis and COVID-19 have a well-established observable relationship. Whether COVID-19 increases the likelihood of developing sepsis and whether patients with sepsis are at increased risk for COVID-19 infection is unknown. Using a bidirectional 2-sample Mendelian randomization (TSMR) analysis techniques in sizable cohorts, we sought to answer this question. The current study performed Mendelian randomization (MR) on publicly accessible genome-wide association study (GWAS) summary data in order to investigate the causal linkages between COVID-19 and sepsis. A Two-Sample MR(TSMR) analyses was performed. As instrumental variables, a COVID-19 dataset of single nucleotide polymorphisms (SNPs) with significance value smaller than 5*10-8 was employed and Sepsis dataset of SNPs with significance value smaller than 5*10-7was employed. The results suggested that Very severe respiratory confirmed COVID-19(VSRC), hospitalized COVID-19(HC) and Infected COVID-19(IC) had no causal influence on sepsis risk using the inverse variance weighted (IVW) technique (VSRC OR = 1.000, 95% CI, 0.956-1.046, P = 0.996, HC OR = 0.976, 95% CI, 0.920-1.036, P = 0.430, IC OR = 0.923, 95% CI, 0.796-1.071, P = 0.291) and there was no causal effect of sepsis on the risk of VSRC, HC and IC (VSRC OR = 0.955, 95% CI, 0.844-1.173, P = 0.953, HC OR = 0.993, 95% CI, 0.859-1.147, P = 0.921, IC OR = 1.001, 95% CI, 0.959-1.045, P = 0.961). Our findings do not support a causal relationship between COVID-19 and sepsis risk, nor do they suggest a causal link between sepsis and COVID-19. The bidirectional relationship between COVID-19 and sepsis warrants further investigation in large cohorts.