Improvement in the accuracy of dual energy x-ray absorptiometry for whole body and regional analysis of body composition: validation using piglets and methodologic considerations in infants.

Previously, we conducted dual energy x-ray absorptiometry (DXA) (Hologic QDR-1000/W) scans and carcass analysis of piglets to evaluate the Pediatric Whole Body software (PedWB) (V5.35) for use in infants. A software upgrade designed for infant whole body (InfWB) (V5.56) led to a reassessment of DXA by: 1) reanalysis of the original scans using InfWB software and 2) comparison of InfWB-estimates of bone mineral content (BMC) and lean and fat mass with chemical analysis. Other assessments included 1) methods of regional analysis and 2) artifacts and the Infant Table Pad in the scan field. The mean coefficients of variation for InfWB whole body measures in small piglets (n = 10, weight 1575 +/- 73 g) and large piglets (n = 10, weight 5894 +/- 208 g) were less than 2.6% except for fat mass which was higher (8.0% versus 6.3% and 6.6% versus 3.5%, respectively) compared with PedWB. In large piglets InfWB produced good estimates of BMC, lean and fat masses. In small piglets, fat mass by InfWB was correlated with chemical analysis, but not by PedWB. There was improvement in the estimation of BMC with InfWB, from 27 +/- 2.2 g to 32 +/- 2.3 g (carcass ash = 38 +/- 3.3 g). Femur BMC analysis by InfWB was precise and was accurate when compared with chemical analysis. Artifacts in the DXA scan field (diapers and blankets) resulted in an increase of the DXA-estimated fat and lean masses. The Infant Table Pad increased the estimate of fat mass in a small piglet by 50%, thus further study is required before it is used routinely. Improvements of the DXA technology have resulted in a more accurate tool, if scanning procedures are carefully implemented.