Comparisons of Predictive Equations for Resting Energy Expenditure in Patients with Cerebral Infarct during Acute Care

Abstract

Estimation of resting energy expenditure (REE) is essential in planning nutrition support. Several equations are used for this estimation in the clinical setting. The purpose of this study was to compare the predictive accuracy of existing equations for REE in patients with cerebral infarct during acute care. We assessed the Harris-Benedict, Mifflin, Owen, Japanese simplified, Wang, and Cunningham equations. The Owen and Japanese simplified equations use sex and weight as explanatory variables, the Harris-Benedict and Mifflin equations include sex, weight, age, and height, and the Wang and Cunningham equations use fat-free mass (FFM) measured using bioelectrical impedance technology. Actual REE values were measured by indirect calorimetry on days 2 and 7 and were then averaged. Applying analysis of variance, predictive accuracy was assessed by comparing the predicted and actual values. A total of 30 patients were analyzed. Actual REE values ranged from 796 to 1637 kcal (mean, 1109). The standard deviation of these values was the smallest with the Harris-Benedict equation (99), followed by the Cunningham (165), and Wang (181) equations. The Mifflin equation underestimated REE in females, whereas the Owen and Japanese simplified equations tended to overestimate it. Based on our results, the Harris-Benedict equation provides the most accurate prediction of REE. In addition, the Cunningham and Wang equations may be useful in long-term care settings involving patients at risk of malnutrition resulting in uneven loss of FFM relative to weight.