PURPOSE: To evaluate the utility of dual x-ray absorptiometry (DXA) in children with bone mineral disorders. MATERIALS AND METHODS: In phase 1, radial DXA was compared with single-energy photon absorptiometry (SPA) (n = 117). In phase 2, radial and lumbar bone mineral density (BMD) measured with DXA and second metacarpal cortical thickness were compared (254 examinations, 224 children). RESULTS: For radial BMD, DXA and SPA correlated well (r = .956) and SPA-equivalent values could be calculated from DXA measurements (mean residual error = 0.024 g/cm2). After controlling for age, sex, weight, and height, partial correlations were very small for lumbar BMD with radial BMD (r = .186) and lumbar BMD with cortical thickness (r = .158), and slightly better for radial BMD with cortical thickness (r = .544). Z scores also correlated poorly with no meaningful correlation for lumbar BMD with radial BMD (r = .07) CONCLUSION: In children with bone mineral disorders, radial DXA and SPA measurements correlate well. However, lumbar BMD, radial BMD, and cortical thickness correlate poorly and lumbar BMD frequently does not identify abnormality in patients with abnormal radial BMD. Lumbar BMD alone is not adequate for evaluation of bone mineral status in these patients.
PURPOSE: To evaluate the utility of dual x-ray absorptiometry (DXA) in children with bone mineral disorders. MATERIALS AND METHODS: In phase 1, radial DXA was compared with single-energy photon absorptiometry (SPA) (n = 117). In phase 2, radial and lumbar bone mineral density (BMD) measured with DXA and second metacarpal cortical thickness were compared (254 examinations, 224 children). RESULTS: For radial BMD, DXA and SPA correlated well (r = .956) and SPA-equivalent values could be calculated from DXA measurements (mean residual error = 0.024 g/cm2). After controlling for age, sex, weight, and height, partial correlations were very small for lumbar BMD with radial BMD (r = .186) and lumbar BMD with cortical thickness (r = .158), and slightly better for radial BMD with cortical thickness (r = .544). Z scores also correlated poorly with no meaningful correlation for lumbar BMD with radial BMD (r = .07) CONCLUSION: In children with bone mineral disorders, radial DXA and SPA measurements correlate well. However, lumbar BMD, radial BMD, and cortical thickness correlate poorly and lumbar BMD frequently does not identify abnormality in patients with abnormal radial BMD. Lumbar BMD alone is not adequate for evaluation of bone mineral status in these patients.