Critical evaluation of methods for determination of body fat content in children: back to basic parameters?

Considering the increasing prevalence of obesity among children and of obesity related disorders in the pediatric population, the reliable evaluation of body fat content in children is of critical importance in research and clinical medicine. In this study, we assessed the congruency of different estimates for body fat content in prepubertal children. We determined anthropometric parameters, such as BMI and skinfold thickness, and bioelectrical impedance in 676 prepubertal Caucasian children. We calculated body fat percentage (BF%) from these parameters applying 5 distinct algorithms and established raw centiles for these models. Expectedly, girls had significantly higher BF% regardless of the method applied. There were, however, significant variances in the calculated amount of BF% between the algorithms, with BIA based equations giving highest BF%, while skinfold based equations tended to provide lower BF% values. Direct comparison of the algorithms revealed a high degree of inconsistency and poor agreement in the assessment of body fat with variations of >10% BF%. Great differences in basic parameters, such as DeltaBMI (3.2 kg/m (2)) or Deltaskinfolds (1.75-fold), would be needed to reliably predict correct ranking of 10% difference in body fat with 95% probability. In summary, BF% strongly varies depending on both the method as well as the algorithm used. This questions the applicability of such field methods for the assessment of BF% for comparative analyses and the superiority of information over basic parameters such as BMI.