Frontiers in Nutrition (Mar 2022)

Development and Cross-Validation of a Predictive Equation for Fat-Free Mass in Brazilian Adolescents by Bioelectrical Impedance

  • Roberto Fernandes da Costa,
  • Analiza M. Silva,
  • Kalina Veruska da Silva Bezerra Masset,
  • Tatianny de Macêdo Cesário,
  • Breno Guilherme de Araújo Tinoco Cabral,
  • Gerson Ferrari,
  • Gerson Ferrari,
  • Paulo Moreira Silva Dantas

DOI
https://doi.org/10.3389/fnut.2022.820736
Journal volume & issue
Vol. 9

Abstract

Read online

The bioelectrical impedance analysis (BIA) is one of the most commonly used techniques for assessing body composition in a clinical setting and in field approaches, as it has the advantages of easy application, fast, and non-invasive, in addition to its relatively low cost. However, the available predictive equations need to be valid for the evaluated subjects. The aim of this study was to verify the validity of several published BIA equations in estimating fat-free mass (FFM) among Brazilian adolescents, in addition to developing and cross-validating a BIA equation to estimate FFM appropriate for Brazilian adolescents. This is a cross-sectional study with 257 adolescents (128 girls) aged 10–19 years, randomly divided into two groups, namely, development (n = 172) and cross-validation (n = 85). The standard technique for assessing FFM was dual X-ray absorptiometry (DXA). The paired t-test, multiple regression, and the Bland-Altman plots were used to test the validity of the proposed models and to perform cross-validation of the model. The equation derived in this study was as follows: FFM = −17.189 + 0.498 (Height2/Resistance) + 0.226 Weight + 0.071 Reactance − 2.378 Sex + 0.097 Height + 0.222 Age; r2 = 0.92; standard error of the estimate = 2.49 kg; the new equation for FFM showed better agreement when compared with that of the equations developed in other countries. In conclusion, the newly developed equations provide a valid FFM estimation and are recommended for Brazilian adolescents with similar characteristics.

Keywords