Abstract
Sex differences in thoracolumbar vBMD were apparent before menopause, and disappeared after menopause, likely attributable to an accelerated and more profound vBMD decline in women at the thoracic spine. In patients without advanced spine degeneration, the overall volumetric changes in the vertebral body appeared subtle.
Introduction
Osteoporosis remains a significant global health challenge, yet longitudinal data on bone mineral density (BMD) changes across the entire spine is limited. Traditional clinical assessments often rely on Dual-energy X-ray Absorptiometry (DXA), which provides 2D "areal" density measurements. However, DXA can be influenced by patient size and degenerative changes, potentially masking underlying bone loss.
This study explores "opportunistic Quantitative Computed Tomography" (QCT), utilizing routine clinical CT scans—originally performed for other indications like oncology or trauma—to monitor bone health. By leveraging deep learning for automated analysis, researchers aimed to map sex-specific aging patterns in vertebral volume and volumetric BMD (vBMD) across the full thoracolumbar spectrum.
Methods
The researchers conducted a retrospective longitudinal study following 168 adults (51 women, 117 men) at the Technical University of Munich. The dataset included 1,554 MDCT scans, totaling 22,720 individual vertebrae, with a mean follow-up period of 9.0 years.
To process this extensive data, a convolutional neural network (CNN) framework was employed for automated labeling and segmentation of vertebral bodies. Hounsfield Units (HU) were converted into vBMD values using asynchronous calibration and CNN-based corrections for contrast media. To ensure data quality, human radiologists performed semiquantitative grading to identify fractures and degenerative changes, such as osteophytes, which could otherwise skew density and volume measurements.
Results
The Vanishing Advantage
A key finding was the shift in bone density advantage between sexes. In the 40–60 age group, women exhibited significantly higher trabecular vBMD than men. However, this advantage vanished in the 60–80 age group. This transition is attributed to an accelerated decline in women, likely coinciding with menopause.
Thoracic Vulnerability
The study highlighted that the thoracic spine (T1–T11) in women declines more rapidly than in men ($p < 0.01$ to $p < 0.0001$). While clinical focus is often on the lumbar region, these results suggest the thoracic spine is a critical area for monitoring fracture risk. A strong correlation ($r = 0.85$) was found between thoracic and lumbar vBMD, supporting the potential for thoracic-specific clinical thresholds.
Volume vs. Degeneration
Initial observations suggested an increase in vertebral body volume with age (approx. 1.0–1.5 $cm^3$). However, when vertebrae with bony degenerations were excluded, the residual volume increase was minimal (<5%). This indicates that perceived vertebral "growth" is primarily a result of age-related wear-and-tear rather than a physiological compensatory mechanism.
Discussion
This research demonstrates the power of opportunistic QCT in providing high-resolution, longitudinal insights into skeletal aging. The findings emphasize the distinct vulnerability of the thoracic spine in post-menopausal women and clarify that age-related volume increases are largely degenerative.
While the study is limited by its clinical cohort and predominantly male sample, it paves the way for using "waste" clinical data to enhance osteoporosis screening. Future efforts should focus on establishing standardized thoracic vBMD thresholds to improve fracture risk assessment across the entire spine.
Summary updated