Lower Vertebral BMD Assessed Through AI Linked to Faster Cognitive Decline: Study
USA: A study published in Radiology found that lower bone mineral density (BMD) in the lower thoracic vertebrae, assessed through deep-learning analysis of noncontrast chest CT scans, was associated with faster cognitive decline. The association was particularly evident in adults with diabetes, suggesting that vertebral BMD may potentially serve as an imaging marker for cognitive health.
Low bone density and age-related changes in the brain often occur together, but the relationship between skeletal health, white matter (WM) damage, and cognitive deterioration over time remains incompletely understood. Sara Momtazmanesh, Department of Radiology and Radiologic Sciences, Johns Hopkins University, Baltimore, and colleagues investigated whether vertebral BMD measured from routine chest CT could be linked to longitudinal changes in brain white matter and cognition.
The secondary analysis used data from the prospective Multi-Ethnic Study of Atherosclerosis and included adults without clinically recognized cardiovascular disease. Participants underwent noncontrast chest CT, multimodal brain MRI incorporating diffusion tensor imaging and fluid-attenuated inversion recovery, and cognitive assessments. A validated deep-learning algorithm was used to estimate thoracic vertebral BMD at baseline. Researchers then assessed changes in WM hyperintensity (WMH) volume, fractional anisotropy, and cognitive performance while accounting for demographic characteristics, apolipoprotein E-ε4 status, cardiometabolic risk factors, lifestyle factors, and medication use.
The analysis included 715 participants, with a median age of 69 years; 397 were men, and 318 were women. Longitudinal brain MRI data were available for approximately 405–408 participants, while cognitive outcomes were assessed in larger subsets.
Key findings:
- Lower baseline vertebral BMD was associated with a faster decline in overall WM fractional anisotropy, although the association did not remain statistically significant after false discovery rate correction.
- Reduced vBMD was linked to faster fractional anisotropy loss in the anterior limb of the internal capsule.
- Lower vBMD was associated with greater accumulation of WMH in the corpus callosum.
- Participants with lower vertebral BMD experienced faster deterioration in global cognitive composite scores and Cognitive Abilities Screening Instrument scores.
- Diabetes strengthened the association between lower vBMD and WMH progression.
The study was limited by its observational design, potential residual confounding, relatively short follow-up, selection bias, and lack of mechanistic biomarkers. CT-derived vBMD also requires validation across imaging protocols before clinical use. The overall WM fractional anisotropy association did not remain significant after multiple-comparison correction, making the regional findings hypothesis-generating.
The authors concluded that lower thoracic vertebral BMD on noncontrast chest CT was associated with modestly faster regional WM injury and cognitive decline, particularly among people with diabetes. The findings suggest that deep-learning-based opportunistic BMD assessment could potentially provide information about brain aging alongside skeletal health, although independent studies are needed to validate these associations and clarify the underlying mechanisms.
Reference: https://doi.org/10.1148/radiol.260656
Next Story