QCT-based geometric, densitometric, and biomechanical properties of the aging human thoracolumbar spine

Vertebral fractures represent a growing health issue in our aging societies. We performed a large-scale quantitative assessment of the geometric, densitometric, and biomechanical characteristics of the aging thoracolumbar spine using QCT and homogenized finite element (hFE) analysis. A total of 9298 intact vertebral bodies from T1 to L5 were evaluated across 629 radiologically healthy post-mortem spines (194 female, 435 male). We quantified vertebral height, cross-sectional area (CSA), BMC, BMD, synthetic areal BMD, and hFE-derived yield strength and applied a mixed linear regression model to investigate their dependence on age, sex, BMI, and spinal level. We found that vertebral height was greatest in the lumbar region, while CSA increased caudally and showed positive associations with both age and BMI. Both BMD and yield strength declined with advancing age, with significantly steeper reductions observed in females and in individuals with lower BMI. Interestingly, not yield force, but yield strength was found to be minimal at the thoracolumbar junction (T12-L1), which coincides with the highest occurrence of vertebral fractures. Despite inter-individual variability, BMD demonstrated strong segmental correlations, particularly between adjacent vertebrae, with correlation strength increasing with anatomical proximity. We established a robust power-law relationship (R2 = 0.91) between BMD and yield strength, which generalized across all vertebral levels. Collectively, these findings delineate spatial and demographic patterns in vertebral strength degeneration and provide normative QCT-based reference data for biomechanical modeling. The derived regression framework enables non-invasive estimation of vertebral yield strength, supporting future applications in fracture risk prediction and spine biomechanics research.