| Title: | Bone Density Contributes to Lumbar Disc Herniation: A Mendelian Randomization Study |
| Journal: | Spine Open |
| Published: | 26 May 2026 |
| DOI: | https://doi.org/10.1097/bn9.0000000000000100 |
Publication 19912
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
Study Design: Genetic epidemiology study integrating two-sample univariable and multivariable Mendelian randomization (MR), with observational triangulation in UK Biobank. Objective: To determine whether lifelong liability to higher bone mineral density (BMD) causally increases risk of lumbar disc herniation (LDH), evaluate reverse causation (LDH-BMD), and assess robustness when modeled alongside BMI and smoking liability. Summary of Background Data: Observational studies relating BMD to disc pathology are conflicting and susceptible to confounding by body habitus and activity. MR leverages randomly allocated genetic variants associated with BMD to strengthen causal inference. Methods: Exposure instruments were derived from UK Biobank heel estimated BMD (eBMD) and GEFOS lumbar spine BMD. LDH outcome associations were obtained from FinnGen (ICD-10 M51.1). Primary causal estimates used inverse-variance weighted MR with MR-Egger, weighted median, and weighted mode as robustness checks. Multivariable MR included BMI and smoking. Directionality and validity were evaluated with Steiger testing, MR-PRESSO, MR-Egger intercept, heterogeneity, and leave-one-out analyses. Observational logistic/Cox models in UK Biobank assessed association and incremental model fit, including sensitivity adjustment for occupational activity. Results: Forward MR supported a modest association between higher BMD on LDH (eBMD OR=1.093, 95% CI: 1.054-1.134; lumbar BMD OR=1.115, 95% CI: 1.051-1.182; both P <0.001), with consistent direction across estimators. Reverse MR was null for eBMD (IVW β=−0.003; P =0.895) but showed an estimator-dependent positive lumbar signal interpreted cautiously. In multivariable MR, BMD estimates remained directionally concordant when modeled with BMI and smoking (eBMD OR=1.087, P <0.001; lumbar BMD OR=1.101, P =0.008). Observationally, both eBMD and lumbar BMD improved model fit; activity adjustment attenuated the eBMD association but not lumbar BMD. Conclusion: Across genetic and observational frameworks, higher BMD is modestly associated with increased LDH risk. Findings are consistent with the possibility that bone properties may influence disc failure susceptibility under cumulative loading. </p>
4 Authors
- Temitope M. Adegbenro
- Jun He
- Renato Polimanti
- Konstantinos Margetis
1 Application
| Application ID | Title |
|---|---|
| 814683 | Genomic Studies for Nervous System and Spine Pathologies |
Enabling scientific discoveries that improve human health