Intervertebral Disk Degeneration and Bone Mineral Density: A Bidirectional Mendelian Randomization Study

Abstract

This study aimed to investigate the causal relationship between bone mineral density (BMD) and intervertebral disk degeneration (IVDD) using a two-sample bidirectional Mendelian randomization analysis. Summary-level data from the Genome-Wide Association Study (GWAS) were used. Instrumental variables (IVs) for IVDD were selected from the large-scale Genome-Wide Association Study (GWAS) (20,001 cases and 164,682 controls). Bone mineral density (BMD) at five different sites (heel (n = 426,824), total body (TB) (n = 56,284), forearm (FA) (n = 8143), femoral neck (FN) (n = 32,735), and lumbar spine (LS) (n = 28,498)) was used as a phenotype for OP. Bidirectional causality between IVDD and BMD was assessed using inverse variance weighting (IVW) and other methods. Related sensitivity analyses were performed. Myopia was also analyzed as a negative control result to ensure the validity of IVs. Heel bone mineral density (heel BMD), total body bone mineral density (TB-BMD), femoral neck bone mineral density (FN-BMD), and lumbar spine bone mineral density (LS-BMD) have a direct causal relationship on intervertebral disk degeneration (IVDD) [heel BMD-related analysis: beta = 0.06, p = 0.03; TB-BMD-related analysis: beta = 0.18, p = 8.72E-08; FN-BMD-related analysis: beta = 0.15, p = 4.89E-03; LS-BMD-related analysis: beta = 0.16, p = 1.43E-04]. There was no evidence of a significant causal effect of IVDD on BMD. In conclusion, our study found a significant positive causal effect of lower BMD on IVDD, and we identified significant causal effects of heel, TB-, FN-, and LS-BMD on IVDD, but there was no evidence of a significant causal effect of IVDD on BMD.

Keywords: Bone mineral density; Genome-wide association study; Intervertebral disk degeneration; Mendelian randomization; Osteoporosis.

Fig. 1

Schematic diagram of assumptions and process analysis for bidirectional MR analysis A [1] the genetic IVs used are strongly associated with exposure; [2] the selected IVs are not associated with potential confounders; and [3] the IVs can only influence the outcome risk through exposure. B Steps in bidirectional MR analysis: osteoporosis was used as an exposure in the first step, while IVDD was studied as an outcome, and vice versa in the second step. The arrows represent the direction of the causal relationship between the two of the results. IVs instrumental variables, OP osteoporosis, IVDD intervertebral disk degeneration, TB total body, FA forearm, FN femoral neck, LS lumbar spine, BMD bone mineral density, MR Mendelian randomization