The Evaluation of Consistency Between Body Composition Assessments in Pediatric Population Using Pencil Beam and Fan Beam Dual-Energy X-Ray Absorptiometers

• Published in: Journal of clinical densitometry Vol. 13; no. 1; pp. 84 - 95
• Main Authors: Pludowski, Pawel, Jaworski, Maciej, Matusik, Halina, Kobylinska, Maria, Klimek, Paulina, Lorenc, Roman S
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 2010
• Subjects: Absorptiometry, Photon - instrumentation; Adolescent; body composition; Body Constitution - physiology; Bone; Bone Density - physiology; Bone Development - physiology; Calibration; Child; Child, Preschool; cross-calibration; DXA; Endocrinology & Metabolism; Equipment Design; Female; Humans; Lumbar Vertebrae - diagnostic imaging; Male; pediatric; Reproducibility of Results; body composition; cross-calibration; Bone; DXA; pediatric
• ISSN: 1094-6950; 1559-0747
• DOI: 10.1016/j.jocd.2009.11.001

Abstract The replacement of the old dual-energy X-ray absorptiometry system with a novel one should be preceded by a cross-calibration procedure. Therefore, the study was aimed at investigating the consistency of bone and body composition measures performed in pediatric population using pencil beam (DPX-L; GE Healthcare, GE Healthcare, Madison, WI) and fan beam (Prodigy; GE Healthcare, GE Healthcare, Madison, WI) densitometers. The study group consisted of 212 healthy children aged 4–18 yr. Total body (TB) and lumbar spine (S) (L2–L4) measurements were performed using DPX-L and Prodigy during the same visit. Bland-Altman analysis, linear regressions, and paired t -test were performed to evaluate the consistency of measurements and to establish a cross-calibration equation. The average Prodigy values for TB and lumbar spine bone mineral density (BMD) and content (BMC) were 2.7%, 2.4% and 1.6%, 1.6% higher than those of DPX-L, respectively ( p < 0.0001). Prodigy-assessed bone area (BA) was lower by 1.4% for TBBA ( p < 0.0001) and 1.1% for SBA ( p < 0.001). Lean body mass (LBM) from Prodigy was higher by 6.9% ( p < 0.0001), whereas fat mass (FM) was lower by 8.4% compared with those from DPX-L ( p < 0.0001). Bland-Altman analyses revealed the effect of magnitude that was nonlinear (2nd degree polynomial) for TBBMD ( r = 0.32, p = 0.001), TBBMC ( r = 0.51, p < 0.0001), TBBA ( r = 0.34, p < 0.0001), and LBM ( r = 0.56, p < 0.0001), but not for FM ( r = 0.14, not significant [n.s.]). In contrast, in lumbar spine, the magnitude dependence was linear and significant for SBMC ( r = 0.46, p < 0.0001) and SBA ( r = 0.34, p < 0.0001) but not for SBMD ( r = 0.12, n.s.). Both skeletal and body composition variables assessed by DPX-L and Prodigy devices were highly correlated, showing R2 values ranging from 0.976 for FM to 0.994 for SBMC. The results of this study document a necessity for implementation of calculated cross-calibration equations to transform DPX-L–based local pediatric references into a novel Prodigy system.