Hand-to-foot bioelectrical impedance analysis to measure fat mass in healthy children : a comparison with air-displacement plethysmography

Aim: To identify children at risk of overweight, assessing children's body composition in a valid way is crucial. The present cross-sectional study examines the comparability of children's fat mass percentage obtained by hand-to-foot bioelectrical impedance analysis (BIA) versusBodPod air-displacement plethysmography as a reference method. Methods: The body composition of 206 Belgian children (7-13 years; 7.7% overweight) was measured with BodPod and hand-to-foot BIA (Tanita BC418 with built-in formula). Results: Overall, comparability was the highest in girls. Good ranking agreement (rho = 0.819) and interchangeability (intra-class correlation (ICC) = 0.757) were found in girls but not in boys (rho = 0.568; ICC = 0.512). Although bioelectrical impedance resulted in an overall overestimation of only 0.5 fat%, it underestimated and overestimated fat% in children with, respectively, higher and lower fat%. The 95% limits of agreement distance was around 17 fat%, and there was a significant difference from the line of identity (intercept 0; slope 1). Conclusions: Despite good ranking agreement, the use of BIA as an alternative for the BodPod is not recommended because of over-/underestimation and wide limits of agreement. It is not possible to calculate a simple correction factor for the whole fat% range to make the two approaches interchangeable.Aim: To identify children at risk of overweight, assessing children's body composition in a valid way is crucial. The present cross-sectional study examines the comparability of children's fat mass percentage obtained by hand-to-foot bioelectrical impedance analysis (BIA) versusBodPod air-displacement plethysmography as a reference method. Methods: The body composition of 206 Belgian children (7-13 years; 7.7% overweight) was measured with BodPod and hand-to-foot BIA (Tanita BC418 with built-in formula). Results: Overall, comparability was the highest in girls. Good ranking agreement (rho = 0.819) and interchangeability (intra-class correlation (ICC) = 0.757) were found in girls but not in boys (rho = 0.568; ICC = 0.512). Although bioelectrical impedance resulted in an overall overestimation of only 0.5 fat%, it underestimated and overestimated fat% in children with, respectively, higher and lower fat%. The 95% limits of agreement distance was around 17 fat%, and there was a significant difference from the line of identity (intercept 0; slope 1). Conclusions: Despite good ranking agreement, the use of BIA as an alternative for the BodPod is not recommended because of over-/underestimation and wide limits of agreement. It is not possible to calculate a simple correction factor for the whole fat% range to make the two approaches interchangeable.A