Causal Link between Gut Microbiota, Neurophysiological States, and Bone Diseases: A Comprehensive Mendelian Randomization Study

Abstract

Increasing evidence highlights a robust correlation between the gut microbiota and bone diseases; however, the existence of a causal relationship between them remains unclear. In this study, we thoroughly examined the correlation between gut microbiota and skeletal diseases using genome-wide association studies. Linkage disequilibrium score regression and Mendelian randomization were used to probe genetic causality. Furthermore, the potential mediating role of neuropsychological states (i.e., cognition, depression, and insomnia) between the gut microbiota and bone diseases was evaluated using mediation analysis, with genetic colocalization analysis revealing potential targets. These findings suggest a direct causal relationship between Ruminococcaceae and knee osteoarthritis (OA), which appears to be mediated by cognitive performance and insomnia. Similarly, a causal association was observed between Burkholderiales and lumbar pelvic fractures, mediated by cognitive performance. Colocalization analysis identified a shared causal variant (rs2352974) at the TRAF-interacting protein locus for cognitive ability and knee OA. This study provides compelling evidence that alterations in the gut microbiota can enhance cognitive ability, ameliorate insomnia, and potentially reduce the risk of site-specific fractures and OA. Therefore, strategies targeting gut microbiota optimization could serve as novel and effective preventive measures against fractures and OA.

Keywords: Mendelian randomization; cognitive performance; fracture; gut microbiome; mediation; osteoarthritis.

Figure 1

Comprehensive study design and flowchart. Visual representation of the extensive experimental design used in this study, featuring methodologies such as Mendelian randomization (MR), linkage disequilibrium score regression (LDSC), mediation analysis, and gene colocalization.

Figure 2

Heatmap displaying causalities of 35 families and 131 genera. (a) Causal relationship between 131 gut microbiota genera and bone diseases. (b) Causative roles of 9 phyla, 16 classes, 20 orders, and 35 families within the gut microbiota in bone diseases.

Figure 3

Forest plot representing the results of MR analysis. Individual estimates of the causal relationship between gut microbiota and bone diseases, derived from MR. Each result is represented by boxes and corresponding confidence bars. Only gut microbiota with an inverse variance weight (IVW) < 0.05 were included in the analysis.

Figure 4

A forest plot displaying LDSC results. Results of the LDSC analysis that explored the correlations between neurophysiological states and bone diseases, with the x-axis signifying the genetic correlation (rg). The dots and their corresponding horizontal lines represent specific rg values and their respective 95% confidence intervals. Hollow dots indicate p > 0.05, whereas solid dots denote p < 0.05.

Figure 5

Mediating effects of gut microbiota on bone diseases. β1 signifies the causal influence of Ruminococcaceae or Burkholderiales on potential mediators, particularly insomnia or cognitive performance. β2 depicts the causal effect of these mediators, either insomnia or cognitive performance, on bone diseases. β3 represents the cumulative causal impact of gut microbiota on bone disease.

Figure 6

Results of gene colocalization analysis. Identifiers ebi-a-GCST006572 and ebi-a-GCS007090, sourced from the Integrative Epidemiology Unit OpenGWAS project, correspond to cognitive performance and knee osteoarthritis (OA), respectively. Notably, rs2352974, a shared variant, was present under both conditions. This variant is located within the TRAF-interacting protein (TRAIP) locus on chromosome 3.