Higher Genetic Bone Mineral Density Tracked With Greater Spinal Stenosis Risk
Genetically proxied bone mineral density was associated with a higher likelihood of spinal stenosis in a combined Mendelian randomization and cohort analysis that probed whether skeletal biology contributes to narrowing of the lumbar canal. The signal was not subtle in the genetic analyses. Across the main instruments, higher heel estimated BMD was linked to stenosis before and after MR-PRESSO outlier removal, and lumbar spine BMD showed an even stronger association in the primary inverse-variance weighted model.
The biologic premise is straightforward even if the clinical translation is not. Bone that is denser, or genetics that favor higher mineralization, may also favor structural remodeling around the spine, including hypertrophy of bone and other degenerative changes that reduce canal diameter. In that setting, stenosis can emerge less as a pure disc problem and more as a consequence of cumulative skeletal architecture. The study does not claim that higher BMD causes stenosis in a simple one-way fashion; it points to a shared remodeling pathway that may shape susceptibility over time.
One result cut against a clean causal story. Reverse Mendelian randomization also linked stenosis liability with higher lumbar BMD, raising the possibility of local remodeling or disease-site contamination rather than a pure upstream effect. In other words, the lumbar spine is not just a passive observer in these analyses. It is also the anatomic site where degenerative change, loading, and imaging artifacts can blur the boundary between cause and consequence.
Genetic Signals Held After Adjustment for Body Size Traits
The main two-sample Mendelian randomization framework used a large set of independent variants, with the final instrument set including 462 independent SNPs and a mean F-statistic of 46.3, indicating strong genetic instruments. The primary stenosis outcome came from FinnGen R12, which included 26,090 cases and 353,224 controls. A separate decompression surgery genome-wide association dataset was generated in UK Biobank and included 1,215 cases and 155,913 controls.
Multivariable Mendelian randomization adjusted for body mass index and height, two traits that matter in any discussion of the lumbar spine. Larger body size changes mechanical loading, alters degenerative risk, and can confound any link between bone density and stenosis. Even after those adjustments, the associations between higher genetically proxied BMD and stenosis persisted. Heel eBMD remained associated with stenosis, and lumbar spine BMD retained a stronger effect estimate in the inverse-variance weighted analysis.
The reported effect sizes were modest but consistent enough to merit attention. For heel eBMD, stenosis odds were elevated before outlier removal and remained elevated after MR-PRESSO correction, with odds ratios around 1.09 to 1.10. Lumbar spine BMD produced a larger odds ratio of 1.23, with the confidence interval staying well above unity. For a trait as polygenic and clinically heterogeneous as stenosis, even that pattern is informative: the relationship looks biological, not random.
Surgery Did Not Mirror the Genetic Stenosis Signal
The surgical endpoint did not behave in the same way as the disease endpoint. Heel eBMD was not associated with lumbar decompression in MRlap, and lumbar BMD was not significant there either. Observationally, baseline heel eBMD showed a small association with incident stenosis, with a hazard ratio of 1.04, but that did not extend to decompression surgery.
This divergence matters because surgery is a downstream clinical decision, not a direct surrogate for disease presence. A person can have symptomatic stenosis without undergoing decompression, and another can undergo surgery for reasons that involve access, symptoms, imaging interpretation, comorbidities, or local practice patterns. Decompression reflects the pathway from anatomy to symptoms to operative selection. Stenosis, by contrast, is the structural disease state. The study makes clear that genetics appears linked more tightly to the substrate than to the final surgical act.
A single sentence captures the practical split: genetic liability mapped to stenosis, not to decompression.
That difference fits bedside reality. Surgeons do not operate on a genotype. They operate on symptoms, disability, imaging, and the cumulative judgment that conservative treatment has failed. A genetic association with stenosis can therefore be real even when the same association does not survive the extra layers of triage that lead to lumbar decompression.
Why Lumbar Remodeling May Be the More Relevant Biology
The stronger lumbar spine BMD association, paired with the reverse MR findings, points toward local skeletal remodeling as the more plausible mechanism. Degenerative lumbar stenosis often develops through a combination of facet joint hypertrophy, ligamentum flavum thickening, disc degeneration, and bony overgrowth. Higher bone density may mark a musculoskeletal milieu in which these structural changes are more likely to accumulate, especially at the spine where load-bearing is greatest.
Heel eBMD is useful because it is less anatomically entangled with the target disease site, yet even there the association with stenosis persisted. Lumbar spine BMD, on the other hand, is easier to contaminate by local degeneration and disease-related imaging changes. That is why the reverse MR result deserves respect. It suggests that lumbar BMD may be acting partly as a readout of the stenotic process itself rather than only as a predisposing trait.
The observational UK Biobank analysis was directionally consistent but weaker than the genetic results. Baseline heel eBMD was only modestly associated with incident stenosis and not with decompression. Cohort associations like these can be diluted by measurement noise, treatment decisions, and the long latency between bone phenotype and clinical presentation. Genetic analyses reduce some of that noise, but they do not erase the complexity of the spine.
What the Data Support, and What They Do Not
The most defensible conclusion is narrow. Higher genetically proxied BMD appears linked to susceptibility to lumbar spinal stenosis, while the connection to lumbar decompression is not supported by the same methods. The study supports a remodeling biology in which skeletal traits help shape stenosis risk, but it does not convert BMD into a surgical prediction tool.
No external funding supported the work, and the analysis used deidentified data under exempt institutional review board status. Those details do not change the science, but they frame the study as a genetically driven investigation rather than a commercialized clinical program.
For spine care, the message is restrained but useful. Higher bone density should not be interpreted as universally protective, at least not in the lumbar canal. In this context, denser bone may travel with the very remodeling forces that crowd neural structures and set the stage for symptomatic stenosis. The unanswered question is how much of that signal reflects true causal bone biology versus the local degenerative changes already taking place inside the spine.