Impedance index or standard anthropometric measurements, which is the better variable for predicting fat-free mass in sick children?

Aim: To compare the predictive value of impedance index (ZI, height(2)/impedance) with anthropometric measurements for estimating fat-free mass (FFM). Methods: FFM of 120 white paediatric children (46 males, 74 females), aged 2.5-18 years was measured by using dual energy X-ray absorptiometry. Weight, height, mid-upper arm circumference (MUAC), skinfold thickness (biceps, triceps, subscapular and suprailiac) and bioelectrical impedance were also obtained. Stepwise multiple regression analysis and residual plots were performed to determine the most significant variables to predict FFM. Results: The single best predictor of FFM was ZI, which explained 96.2% of the variance in FFM (r = 0.981, SEE = 2.15 kg). Addition of weight to the model containing ZI increased the explained variance of FFM to 96.6% (r = 0.983, SEE = 2.03 kg). BMI and MUAC were the poorest predictors of FFM: r = 0.422, SEE = 10.2 kg and r = 0.621, SEE = 8.93 kg, respectively. Conclusion: Impedance index is a more significant single predictor of FFM than other anthropometric measurements. The predictive accuracy of bioelectrical impedance analysis-based prediction equations for FFM was improved by addition of weight.