Exploring the bone-vascular axis: AI-augmented chest CT analysis in COPD highlights association between vertebral bone density and arterial calcifications.

OBJECTIVES

Artificial intelligence (AI) tools enable automated assessment of vertebral bone mineral density (BMD), coronary artery calcification (CAC), and aortic calcification on CT scans, offering multimorbidity analysis in chronic obstructive pulmonary disease (COPD) patients. We aimed to investigate the bone-vascular axis with a focus on COPD severity and inhaled corticosteroid (ICS) use.

METHODS

Low-dose chest CTs of 540 patients from the COSYCONET study were analyzed using AI-based tools. Total thoracic calcification (TTC) was defined as the sum of CAC and aortic calcification volumes. The BMD was measured for T12 in Hounsfield units. Group comparisons and correlation analyses were stratified by COPD severity (GOLD 0-2 vs. GOLD 3-4) and ICS treatment. The modification effect was assessed with multivariable and interaction models.

RESULTS

Higher CAC and TTC were significantly associated with increased age, smoking, BMI, and impaired physical function, while gait speed, female sex, and BMD were inversely associated. With no significant difference between GOLD stages, BMD showed a consistent negative correlation with both CAC (ρ = -0.247 and -0.243) and TTC (ρ = -0.286 and -0.304). Interaction analyses revealed no significant modification effect for COPD severity; a marginally significant trend toward a stronger association among ICS users for CAC (β = -0.175, p = 0.042), but no significant effect for TTC.

CONCLUSION

The decreased vertebral BMD was associated with increased vascular calcification, independent of COPD severity and marginally dependent on ICS therapy. These findings support the concept of "bone-vascular axis", likely driven by shared degenerative mechanisms.