Índice de adiposidade corporal como método para avaliação do porcentual de gordura em crianças e adolescentes

Abstract

It is important to clarify the association of lean and fat mass among children and adolescents, considering the search for sports performance or with the increased prevalence of overweight and obesity. The body adiposity index (BAI) uses anthropometry to estimate the percentage of body fat (F%). It is a valid equation for some population groups. However, for the pediatric population, the BAI presents weaknesses, and it is necessary to adjust its aritmetic structure for use in children and adolescents. The aim of this study was to analyze the as a method of measurement of fat percentage in children and adolescents, as well as to adjust the BAI for its use in pediatrics. The sample consisted of 247 children and adolescents, 44 female (11.5 ± 1.80 years) and 203 males (13.6 ± 2.60 years). F% was evaluated by Dual Energy X-Ray Absorptiometry (DEXA) and IAC. The analyses of F% by DEXA and BAI were strongly correlated (boys: r = 0.71, p<0.0001; girls: r = 0.72, p<0.0001). Linear regression analyses showed that IAC is significant for estimating F% in subjects of both sexes (p<0.0001). However, although for boys, the F% analyses performed by BAI and DEXA showed no differences when compared (p= 0.2), there were significant differences for the analyses in females (p<0.05). In addition, the BAI showed a significant proportion bias for both sexes (p<0.0001), which suggests its inefficiency in the analysis of F% by the BAI, thus, the aritmetic adjustment of the BAI was necessary and successful for the male subjects, reaching the mathematical model named in the present study of adjusted body adiposity index (BAIADJ). BAI and BAIADJ correlated with DEXA significantly (r≤700, p <0.001). BAI (95% CI β: [1.35; 1.90], p <0.0001) and BAIADJ (95% CI β: [1.40; 1.90], p < 0.0001) were pointed out as good prdtors of F%. However, BAI showed a significant difference in relation to DEXA (p=0.04). While there was no difference between BAIADJ and DEXA (p=0.1). In addition, there was a bias in proportion to BAI (p <0.05), but not to BAIADJ (p>0.05). Thus, we conclude that F% analyses by BAI and DEXA correlate, however the BAI is not effective in estimating F% in pediatrics of both sexes. We also concluded that the BAIADJ developed by the present study proved to be an effective tool for an analysis of F% in male children and adolescents, demonstrating significant degrees of agreement and validity in relation to the analyses of F% performed in DEXA.