Correlations of dual-energy X-ray absorptiometry, quantitative computed tomography, and single photon absorptiometry with spinal and non-spinal fractures.

Controversy continues as to which method of measuring bone mineral density (BMD) best detects osteoporosis and best correlates with fractures of the spine, hip and elsewhere. To answer these questions the prevalence of fractures was carefully determined among 90 subjects (70 with osteoporosis, 6 with mild primary hyperparathyroidism, 1 with osteomalacia and 13 normals) and simultaneous measurements were made using spinal computed tomography (QCT), spinal anteroposterior (AP) and supine lateral dual X-ray absorptiometry (DXA), femoral neck and total hip DXA, and distal third radial DXA and single photon absorptiometry (SPA). The DXA measurements which had the greatest sensitivity in detecting osteoporosis (defined as a BMD lower than -2.5 SD of peak bone mass at age 30 years) were the supine lateral spine DXA (84%) and femoral neck DXA (75%); less sensitive were the DXA measurements of the distal third of the radius (61%) and AP spine (51%). DXA measurements of the femoral neck and distal third of the radius were more useful than spinal measurements in detecting the osteopenia of mild primary hyperparathyroidism. Vertebral compression fractures (VCF) correlated well with spinal QCT (r = 0.38) and lateral spine DXA (r = -0.41), but poorly with AP spine DXA (r = -0.17) and distal third radial DXA (r = -0.02). Non-spinal fractures correlated best with the distal third radial DXA (r = -0.42). In conclusion, spinal QCT, supine lateral spine DXA and femoral neck DXA are the best BMD methods to screen for osteoporosis, whereas AP spine DXA is a poor screening method in women over 60 years of age. Spinal QCT and lateral spine DXA correlate well with VCFs, whereas correlations of VCFs with AP spine DXA, femoral neck DXA and distal third radial DXA are poor.