Accuracy and reliability of the InBody 270 multi-frequency body composition analyser in 10-12-year-old children

The aim of this study was at examining the validity and reliability of a marketed bioimpedance (BIA) scale for body composition assessment, in children engaged in an educational football project (FIFA 11 for Health). One-hundred and twenty-seven children (70 boys and 57 girls; age 10.7±0.5 years, body mass 41.2±9.0 kg, Body mass index 18.5±3.3 kg·m -2 and stature 149±7 cm) were evaluated for total body mass, lean body mass, muscle mass, using BIA (InBody 270, Biospace, California, USA) and dual-energy X-ray absorptiometry (DEXA, Lunar Prodigy, GE Medical Systems, Madison, Wisconsin, USA), at baseline conditions. Data analyses were carried out separately for girls and boys. Nearly perfect associations (r = 0.97−0.99) and excellent absolute (TEM = 0.04−1.9%) and relative (ICC = 0.98−1.00) inter-device reliability were found between DEXA and BIA variables. Fat and lean body mass bias (p < .0001) were practically relevant both for the boys (2.56 and 11.22 kg, respectively) and the girls (2.33 and 10.49 kg, respectively). Muscle mass and body fat were underestimated and overestimated, respectively, for the boys and girls. InBody 270 is a valid BIA system for estimating body composition with an excellent inter-device relative and absolute reliability. However, the remarkable measurements bias of BIA fat and muscle mass values discourage its use for clinical prescription. The BIA body composition biases were sex dependent.