Bone mineral density measured by a portable X-ray device agrees with dual-energy X-ray absorptiometry at forearm in preschool aged children

Abstract

Dual-energy X-ray absorptiometry (DXA) measures of bone mineral density (BMD) are generally not feasible in fieldwork. The present study determined the agreement between BMD measured by DXA and portable peripheral DXA in preschool aged children. Fifty-seven children (4.2 ± 1.0 yr) had their nondominant distal forearm scanned using a peripheral DXA scanner (PIXI; GE Medical Systems Lunar, Madison, WI) at their daycare and a DXA (4500A Discovery Series; Hologic Inc., Bedford, MA) at our research clinic. Correlation analysis, one-way analysis of variance, and Bland-Altman plots were performed to examine the agreement between measurements. Data were also divided into tertiles for cross-classification analysis and calculation of kappa coefficients. Distal forearm BMD measured by PIXI was significantly correlated with DXA measures of total forearm BMD (r > 0.51; p < 0.001), proximal 1/3 BMD (r > 0.41; p < 0.001), mid-BMD (r > 0.37; p < 0.001), and ultradistal (UD) BMD (r > 0.57; p < 0.001). Cross-classification in the same or adjacent tertile between measures (UD forearm: 96.5%; UD radius: 94.4%; total forearm: 87.7%; total radius: 84.2%) resulted in weighted kappa coefficients of 0.46, 0.58, 0.42, and 0.43, respectively. Bland-Altman plots further clarified these agreements as all had low bias (UD forearm: bias = 0.003 ± 0.002; UD radius: -0.015 ± 0.021; total forearm: -0.062 ± 0.027; total radius: -0.077 ± 0.026). These results demonstrate that portable DXA measures of forearm BMD agree moderately with DXA.

Keywords: DXA; PIXI; peripheral DXA; radius; ulna.