ABSTRACT
BACKGROUND/OBJECTIVES: Accuracy of impedance analysis depends on population-specific prediction equations. The aim of this study was to propose new prediction equations to estimate body composition using impedance analysis for healthy Tunisian children and validate them using the deuterium oxide dilution (D2O) technique. SUBJECTS/METHODS: Body composition was estimated by the Tanita impedance analyzer. The validity of this system was investigated in 134 school children aged between 8 and 11 years. The validity of other published equations was also tested in our sample. Total body water (TBW) and Fat-Free Mass (FFM) were determined using the D2O technique. Participants were divided equally into development (n = 67) and validation groups (n = 67) to develop prediction equation using linear regression models. RESULTS: The comparison between body composition obtained by Tanita system and by D2O technique illustrated a significant difference (p < 0.01). Compared to D2O technique, Tanita impedance analysis underestimates fat mass and overestimates FFM and TBW. The prediction equations for TBW and FFM were developed with sex, age, weight and resistance index as possible predictor variables. The selected equations presented the highest adjusted coefficient of determination (R2), the lowest standard error of the estimate (SEE) value and the lowest p-values. The pure error was 1.263 for the TBW equation and 1.646 for the FFM. The Bland Altman plot illustrated the good level of concordance between the TBW and FFM predicted by the new equations and measured by isotope dilution. CONCLUSIONS: Our study provides valid prediction equations for estimation of TBW, FFM from impedance analysis measures for Tunisian children. These equations are applicable to children aged between 8 and 11 years.
