Assessement of resting energy expenditure of obese patients: Comparison of indirect calorimetry with formulae

Abstract

Determining the caloric expenditure of overweight (Body mass index - BMI from 25-29.9 kg/m(2)) or obese (BMI> or =30 kg/m(2)) patients is a difficult task. The importance of the measurement using indirect calorimetry (IC) is to avoid the administration of an inadequate caloric load, which can increase clinical morbidity. The aim of this study was to compare the resting energy expenditure (REE) obtained by IC with that estimated by the Harris-Benedict (HB) and the Ireton-Jones (IJ) equations and a fixed amount of 21 kcal/kg of body weight, using actual, average and adjusted weights. A total of 71 IC measurements were prospectively performed in 44 patients using the "DELTATRAC II" calorimeter. In 53 of the measurements the patients were on mechanical ventilation and 18 were performed using the canopy, in the fasting state and 24h after achieving calculated caloric requirements. The best REE prediction value in the fasting state was obtained with the HB equation using the actual body weight (HB/ABW), 1873+/-484 kcal/day (p=0.49) when compared to the REE of 1798+/-495 kcal/day measured by IC. After achieving caloric needs, the best REE prediction values were using the HB/ABW equation, 1873+/-484 kcal/day (p=0.56) and the IJ equation with the adjusted body weight, 2103+/-580 kcal/day (p=0.19) when compared to 1948+/-507 kcal/day measured by IC. Although, when the above results were plotted on the Bland and Altman limits of agreement test their average bias were low, but the ranges of their absolute values from the observed average agreement were large, > or =531 kcal. From these data we can conclude that even though the average values obtained with the equations utilized to predict the REE for this population, showed good correlation with those by IC, as they had the best comparative absolute results, however they demonstrated an unacceptable variability when matched to the measured REE values.