Cross-validation of prediction equations for resting energy expenditure in young, healthy children

Objective To examine the accuracy of several prediction equations for resting energy expenditure (REE) in children. Design REE was measured in 113 prepubertal children (60 girls and 53 boys aged 3.9 to 7.8 years old, weighing 14.7 to 30.0 kg) using Indirect calorimetry and compared with values estimated from the prediction equations of Altman and Dittmer, The Food and Agriculture Organization/World Health Organization/United Nations University (FAO/WHO/UNU), Maffeis et al, and Harris and Benedict. Statistical analysis Measured REE (MREE) was compared with predicted REE (PREE) by means of regression analysis. Prediction equations were considered accurate if the regression of MREE vs FREE was not significantly different from the line of identity (slope=1.0; intercept=0). Precision was assessed by lie multiple correlation coefficient of the regression of MREE vs FREE. Results MREE was 938+/-119 kcal/day, and FREE was 1,057+/-224 kcal/day for the Altman and Dittmer equations, 956+/-84 kcal/day for the FAO/WHO/UNU equations, 948+/-64 kcal/day for the equations of Maffeis et al, and 954+/-102 kcal/day for the Harris-Benedict equations. The regression of MREE vs PREE was significantly different from the line of identity for all prediction equations except the FAO/WHO/UNU equations (slope=0.96, P=.735, intercept=-15 kcal/day P=.885 for girls and slope=1.08, P=.635; intercept=-62 kcal/day; P=.635 for boys). None of the equations was precise for MREE vs FREE (for all. R(2)<.6). For the FAO/WHO/UNU equations, less than half of the predictions were within +/-50 kcal/day but 99% were within 200 kcal/day. Conclusion Most prediction equations for REE in children do not accurately or precisely estimate REEs. The exception is the FAO/WHO/UNU equations, which are reasonably accurate and precise for practical purposes.