Inconsistency between the body fat percentages estimated through anthropometric measurements and manual bioimpedance in children and adolescents

The rate is assessed as per the body fat percentages (%BF) estimated through anthropometric measurements and manual bioimpedance, according to gender and adiposity rate within the paediatric population. The %BF is assessed by anthropometry using the Slaughter equation for cutaneous skinfolds and manual bioimpedance (Omron model HBF-306INTTM) to 321 students (160 females and 161 males) between 10-18 years. The association degree among the methods was measured using the Pearson correlation coefficient (r). The consistency among methods was analysed through the intra-class correlation coefficient (ICC) and the consistency correlation coefficient of Lin (CCC), grouping the population by gender and adiposity rate. The reference method was the cutaneous skinfolds assessed by the Slaughter equation. Optimum and high adiposity rates were recorded for the studied sample. The r between the bioimpedance and the anthropometry was 0.689, which corresponds to moderate. However, it decreased according to the gender and adiposity rate (females: optimum 0.409 and high 0.610; males: optimum 0.335 and high 0.507). The ICC was 0.668, which also decreased by gender and adiposity rate (females: optimum 0.323 and high 0.578; males: optimum 0.229 and high 0.506). CCC was poor (<0.90) by gender and adiposity rate. The study shows differences for both methods in this age range, which is accentuated by gender and adiposity rate. Therefore, the use of manual bioimpedance as an interchangeable method with the anthropometric measurements in children and adolescents is not suggested as it underestimates the %BF.