ABSTRACT
OBJECTIVE: Osteoporosis (OP), a persistent metabolic bone disorder linked with inflammation, has an undetermined cause. In our research, we employed bidirectional Mendelian randomization (MR) to investigate the interplay between OP and inflammation agents. MATERIALS AND METHODS: We performed two-way pooled-level MR analyses to characterize the causal relationship between 41 circulating inflammatory modulators and OP. Genetic variation data for the 41 regulatory factors associated with inflammation were obtained from genome-wide association studies (GWASs) of human cytokines. Bone mineral density (BMD) was utilized as a phenotype for OP in our approach. The BMD dataset, sourced from the GEFOS consortium, a large GWAS meta-analysis study and UK Biobank, was classified based on varied sections [whole body (TB), lumbar spine (LS), femoral neck (FN), forearm (FA), and heel] and age brackets (0-15 years, 15-30 years, 30-45 years, 45-60 years, and above 60 years). Primary MR analyses were executed using the inverse variance weighting (IVW) method, and sensitivity analyses were performed using the MR-Egger, weighted median, simple model, and weighted model. Cochran's Q test was utilized to evaluate the existence of heterogeneity. We used MR-Egger regression and MR multiplicity of residuals and outliers (MR-PRESSO) to assess pleiotropy. RESULTS: After false discovery rate (FDR) correction, elevated levels of circulating interleukin-8 (IL-8) [ß = 0.072 (0.031-0.114), p < 0.01], macrophage inflammatory protein-1b (MIP-1ß) [ß = 0.008 (0.003-0.013), p < 0.01; ß = 0.026 (0.009-0.042), p < 0.01], and cutaneous T-cell attracting chemokine (CTACK) [ß = 0.037 (0.017-0.056), p < 0.01] was associated with a reduced risk of OP. Reduced levels of hepatocyte growth factor (HGF), IL-1ra, IL-10, macrophage colony-stimulating factor (MCSF), and MIP-1α were associated with a reduced risk of OP [ß = -0.030 (-0.047 - -0.013), p < 0.01; ß = -0.025 (-0.041 - -0.010), p < 0.01; ß = -0.018 (-0.029 - -0.007), p < 0.01; ß = -0.060 (-0.097 - -0.024), p < 0.01; ß = -0.118 (-0.190 - -0.047), p < 0.01]. We observed a significant causal correlation between FN-BMD and MCP-3 (FDR < 0.05). The occurrence of OP may also lead to elevated levels of MCP3 [ß = -0.466 (-0.714 - -0.217), p < 0.01]. The reliability of the results was confirmed by sensitivity analyses. CONCLUSIONS: This study demonstrated the pathogenic role of circulating inflammatory modulators in OP using bidirectional MR analysis. This further deepens the understanding of OP pathogenesis and provides new ideas for therapeutic intervention in OP.
