The relationship of peripheral to axial bone density.

Bone measurements were made in young normal women (n = 125) using single-photon absorptiometry (SPA with 125I) of the radius shaft and ultradistal radius, dual-photon absorptiometry (DPA with 153Gd) of the lumbar spine and the proximal femur, and 125I-based quantitative computed tomography (QCT) of the distal radius. Measurements of the appendicular and axial sites also were made in 21 women with spinal osteoporosis. The standard error of estimate in predicting axial densities from peripheral measurements was about 0.11 g/cm2 (10-13%) with all methods. The osteoporotic-normal difference was 21% for SPA of the radius shaft and 125I-QCT of the distal radius, 30% for the lumbar spine and 34% for the femoral neck. Prediction of spine and femur density, and diagnostic sensitivity, with 125I-QCT were equivalent to that obtained with conventional SPA, but the precision of 125I-QCT (0.8%) was 2-3 times better than both SPA and DPA.