Analyzing the causal relationship between gut microbiotas, blood metabolites, and COVID-19 susceptibility: A...

Gut microbiota and blood metabolites play crucial roles in the progression and outcomes of COVID-19, but the causal relationships and mechanisms remain unclear. Our aim is to use two-sample Mendelian randomization (MR) to explore the causal relationships between gut microbiota, COVID-19 susceptibility, and potential mediating blood metabolites. We utilized summary statistics from the largest genome-wide association studies (GWAS) to date on gut microbiota (n = 18,340), blood metabolites (n = 115,078), and COVID-19 susceptibility (cases n = 60,176 and controls n = 1310,725 from the COVID-19 Host Genetics Initiative meta-analysis). We conducted bidirectional MR analyses to explore the causal relationships between gut microbiota and COVID-19 susceptibility and performed two-step MR to identify potential mediating blood metabolites. Five analytical methods were used to assess two-sample causal relationships, with inverse variance weighted (IVW) being the primary method. Sensitivity analyses were also conducted to ensure the robustness of the main MR results. Using the IVW method, we found causal relationships between 3 types of gut microbiota and 34 blood metabolites with COVID-19 susceptibility. In the two-step MR, the non-oxidative branch of the Pentose phosphate pathway was shown to reduce Sebacate (C10-DC) levels, and the species Parabacteroides goldsteinii was negatively correlated with Acetoacetate levels. Sebacate (C10-DC) levels were negatively associated with COVID-19 susceptibility, while Acetoacetate levels were positively associated with COVID-19 susceptibility. Furthermore, these causal relationships remained significant after correcting for false discovery rates (all q-values < 0.05). Heterogeneity and pleiotropy tests showed no statistical significance (P > .05). Mediation analysis indicated that the abundance of the non-oxidative branch of the Pentose phosphate pathway and COVID-19 susceptibility was mediated by Sebacate (C10-DC) levels (mediation proportion of 15.8%), and the abundance of P goldsteinii and COVID-19 susceptibility was mediated by Acetoacetate levels (mediation proportion of 31.7%). The current MR study provides evidence supporting the causal relationships between several specific gut microbiotas and COVID-19 susceptibility, as well as potential mediating blood metabolites. Our findings warrant further validation through larger epidemiological studies.