Abstract Background/ObjectivesBioimpedance analysis (BIA) is a non-invasive method for estimating fat-free mass (FFM) and fat mass (FM) using resistance and reactance. Its use in children is controversial, as little is known about the accuracy of the measurements (resistance, reactance, and the resulting FFM and FM).Subjects/MethodsWe derived FM and FFM in 2,954 visits from 1,547 children and adolescents between 5 and 21 years of age using BIA measurements and skinfold-based equations. We used Bland-Altman analyses to compare the resulting estimates. Furthermore, we estimated sex- and age-specific percentile curves for height-normalized resistance (Rz/H) and reactance (Xc/H). We used the overall concordance correlation coefficient (OCCC), concordance correlation coefficient (CCC), and intraclass correlation coefficient (ICC) to estimate agreement between the different segments. Bland-Altman analyses were also used to compare Rz/H and Xc/H SDS.ResultsThe FM estimates from BIA were lower than estimates from the skinfold-based calculations, especially for boys. The estimated FFM and FM showed weight-status-dependent discrepancies between the estimation methods. The ICC showed "excellent" (ICC > 0.9) agreement between all segmental measurements. The OCCC and CCC showed varying degrees of concordance. The Rz/H and Xc/H percentile curves showed a similar age progression for all body segments. Rz/H (SDS) and Xc/H (SDS) were lower in participants with obesity.ConclusionBoth measures –- Rz/H and Xc/H – were highly dependent on age and BMI-SDS. Our results suggest that current estimation techniques may be inadequate for children and adolescents with severe obesity, highlighting the requirement for specialized approaches explicitly designed for this population.
Böker et al. (Tue,) studied this question.