RESUMEN
Abstract Objective: The aim of the present study was to evaluate the agreement between the most used skinfold thickness equations with multi-frequency bioimpedance analysis in the prediction of body fat levels in children. Method: A cross-sectional study of healthy Brazilian community-dwelling individuals. The anthropometric assessment included height, body mass, arm circumference, and waist circumference. The percentage of body fat was obtained by measuring skinfold thickness equations and using bioimpedance analysis, and skinfold thickness was measured using a scientific skinfold caliper. Bland-Altman plot analysis was used to verify the agreement between the methods. Results: There were 439 children and adolescents evaluated, with a mean age of 11.6 ± 3.7 years. The mean body fat by bioimpedance analysis was 22.8% ± 10.4%, compared to 22.4% ± 8.8% by Slaughter (1), 20.4% ± 9.2% by Slaughter (2), 19.6% ± 4.4% by Goran, and 24.7% ± 10.0% by Huang equations. Bland-Altman plot analysis revealed limits of agreement greater than 8% between the bioimpedance analysis approach and equations, exceeding the clinically acceptable limit predefined a priori. None of the equations had good agreement with bioimpedance analysis. Conclusion: It was concluded that skinfold thickness and bioimpedance analysis should not be used interchangeably in children and adolescents.