Validation of quantitative computed tomography-derived areal bone mineral density with dual energy X-ray absorptiometry in an elderly Chinese population

Background The performance of computed tomography X-ray absorptiometry (CTXA) against the dual energy X-ray absorptiometry (D)(A) as standard has not been studied in Chinese population. The aim of this study was to evaluate the precision of this measurement and validate the value of quantitative computed tomography (QCT) by comparing CTXA results with DXA results in an elderly Chinese population. Methods One hundred and three females of 46 to 76 years old and 49 males of 52 to 76 years old were recruited from the Prospective Urban Rural Epidemiology study. All subjects underwent hip scans by both QCT and DXA on the same day. For precision determination, 30 subjects had duplicate DXA hip scans. The hip QCT data of a subset of 27 subjects were separately analyzed by two observers and reanalyzed by one observer at a different time. The inter- and intra-observer variations of CTXA measurement were assessed, and the difference and correlation between CTXA and DXA results were analyzed. Results The inter- and intra-observer variations of CTXA were 0.070 and 0.024 g/cm(2) in the femoral neck (FN), and 0.030 and 0.012 g/cm(2) in the total hip (TH), which were comparable to the DXA inter-scan variations (0.013 g/cm(2) for FN and 0.014 g/cm(2) for TH). The results of CTXA bone mineral density (BMD) were highly correlated with those of DXA (R-2=0.810 for FN and R-2=0.878 for TH). The BMD values of CTXA in FN and TH were lower than those of DXA by 21.0% and 17.8% (P<0.05), respectively. However, after appropriate transformation, the difference was eliminated and a comparable T score could be obtained. Conclusions CTXA shows good agreement with DXA for the measurement of BMD in the proximal femur, which makes QCT suitable for the quantification of bone mineral content in the hip and helpful for the diagnosis of osteoporosis.