Causal relationships between Burkholderiales, Eubacteriaceae and colorectal cancer mediated by plasma metabolites

While Burkholderiales and Eubacteriaceae are associated with colorectal cancer, a causal relationship has not yet been fully established. This study aimed to assess their causal effects and associated plasma metabolites on colorectal cancer. Summary statistics from genome-wide association studies were analyzed for gut microbiota (n=7738), 1400 plasma metabolites (n=8299), and colorectal cancer (n=321040). Genetic correlations were estimated using linkage disequilibrium score regression. Causal relationships were assessed using Mendelian randomization. Furthermore, a two-step mediation analysis was performed to estimate the proportion of the effect of Burkholderiales and Eubacteriaceae on colorectal cancer mediated by plasma metabolites. The inverse variance weighted method was primarily used to estimate the effect. Our results suggested the causal effects of 13 Burkholderiales and eight Eubacteriaceae taxa on colorectal cancer. The positive effect of Oxalobacteraceae, Oxalobacter and Oxalobacter formigenes was potentially mediated by dihomo-linolenoylcarnitine, while the protective effect of Eubacterium siraeum may involve 1-linoleoyl-GPE. Sensitivity analyses indicated no significant heterogeneity or pleiotropy. This study suggests the potential causal effects of the Burkholderiales and Eubacteriaceae taxa on colorectal cancer, with effects potentially mediated by the metabolites dihomo-linolenoylcarnitine and 1-linoleoyl-GPE, respectively. It should be emphasized that these findings are inherently observational in nature and do not establish definitive causality, which is a limitation of the study due to the relaxed statistical threshold for instrumental variable selection and the restriction of the analysis to populations of European ancestry. Further experimental and clinical studies are required to elucidate these mechanistic relationships.