Causal relationships between oral-gut microbiome and bone neoplasm-related phenotypes: Insights from bidirectional Mendelian randomization.

Abstract

The human oral and gut microbiota are the 4 largest microbial communities in the body and play crucial roles in maintaining homeostasis and influencing disease. Observational studies have suggested links between these microbiota and bone neoplasm-related phenotypes, but establishing causality has been challenging due to confounding factors and reverse causality. We conducted a bidirectional, 2-sample Mendelian randomization (MR) study to investigate evidence consistent with a potential causal association between the saliva and gut microbiota and various bone neoplasm-related phenotypes. Genetic instruments for saliva and gut microbiota were sourced from large genome-wide association studies. Inverse variance weighted was the primary MR method, supplemented by 4 other MR techniques. Sensitivity analyses, including MR-Egger regression, were performed to assess pleiotropy and heterogeneity. In the forward MR analysis, Veillonella parvula from the saliva microbiota was associated with a decreased risk of bone and connective tissue neoplasms (β: -0.236, 95% CI: [-0.275, -0.197], P = 8.20E-33). MR analyses identified genetically predicted associations between several microbial taxa and bone neoplasm-related phenotypes. Reverse MR analyses showed that genetic liability to bone neoplasm-related phenotypes was associated with variation in the composition of the oral (e.g., Order Bacteroidales, Rothia mucilaginosa) and gut microbiota (e.g., Class Methanobacteria, Genus Eubacterium oxidoreducens group). Sensitivity analyses confirmed the robustness of these findings, as no statistical evidence of substantial heterogeneity or directional horizontal pleiotropy was detected. This study provides genetic evidence supporting a bidirectional causal relationship between specific saliva and gut microbiota and bone neoplasm-related phenotypes. Our findings identify several microbial taxa as potential candidates for future biomarker development and therapeutic investigation in bone neoplasm-related phenotypes. However, these genetically informed associations require further mechanistic, experimental, and prospective clinical validation before clinical application.