Agreement of anthropometric equations with the 4-component model in the prediction of body fat in obese schoolchildren

Aim: To evaluate the agreement between body fat determined by the 4-component (4C) model and anthropometric equations in obese schoolchildren. Methods: The sample included 61 obese children (body mass index = p 95) of both sexes, aged 8 to 13 years, from families with a history of chronic non-transmissible disease, recruited from a school in a district of the city of Santiago. The multidimensional determination of body composition (not carried out in a Chilean population) included isotope dilution, air-displacement plethysmography and radiographic absorption. Body fat was calculated using the Fuller 4C model. The results of this model were compared with the anthropometric equations generated by Slaughter, Ellis, Huang, Johnston and Deurenberg. This comparison was performed using Lin's agreement coefficient and the Bland and Altman method. Results: The anthropometric equation that best correlated with the four reference components was Slaughter for males (r = 0.94; p < 0.0001) and Ellis equation (r = 0.96; p < 0.0001) in the females. The Slaughter equation underestimated body fat in males by -0.93 (-6.81 4.96) kg. In females, all of the anthropometric equations underestimated body fat, while Ellis showed the closest difference of 1.12 (-2.92 5.17) kg. Conclusions: The Slaughter equation for obese boys and the Ellis equation in obese girls allowed an adequate estimation of total body fat, applicable to clinical and public health situations.