Measurement of trabecular bone mineral density in the thoracic spine using cardiac gated quantitative computed tomography

OBJECTIVES: To develop a method and evaluate the performance of thoracic bone mineral density (BMD) measurement using cardiac gated quantitative computed tomography (QCT). METHODS: A total of 762 participants (57% female) with a mean age of 61 years had a CT examination of the heart using prospective cardiac gating. A subset of 443 participants had replicate CT examinations of the heart. Another, nonindependent subset of 464 participants had CT examination of the abdomen. A QCT calibration phantom was included in all scans. Trabecular BMD was measured in the thoracic (T6-T11) and lumbar (T11-L4) spine. Tests of calibration and refinement and simple correlations between replicate thoracic BMD measurements and between thoracic and lumbar BMD measurements were calculated. RESULTS: There was high correlation between replicated thoracic BMD measurements in men (r = 0.995, P < 0.0001) and in women (r = 0.995, P < 0.0001). There was high correlation between thoracic and lumbar BMD in men (r = 0.90, P < 0.0001) and in women (r = 0.94, P < 0.0001). The mean BMD was higher in the thoracic spine than the lumbar spine in men (137.58 mg/cm3 vs. 126.94 mg/cm3, P < 0.0001) and in women (152.07 mg/cm3 vs. 133.44 mg/cm3, P < 0.0001). In both genders, thoracic and lumbar BMD was inversely associated with age (all P < 0.05). CONCLUSIONS: Cardiac gated CT, primarily intended for measurement of coronary vascular calcium, can be used to measure thoracic BMD with high precision. Thoracic BMD measurements using this method are highly correlated with QCT measurements in the lumbar spine.