BACKGROUND: The energy requirement of a patient receiving nutrition support is typically estimated by calculating the basal energy expenditure (BEE) using the Harris-Benedict equations and multiplying by stress and activity factors. Because fat-free mass (FFM) and fat mass (FM) are important determinants of BEE, we hypothesized that body composition estimates derived from bioelectrical impedance analysis (BIA) could be used to develop predictive equations for resting energy expenditure (REE) that were more accurate than those calculated using the Harris-Benedict equations. METHODS: Seventy-six adults referred to the nutrition support service were studied. REE was measured by indirect calorimetry, and single-frequency BIA was used to estimate FFM and FM. Using the first 20 male and 20 female patients, predictive equations for REE were developed by multiple regression analysis, using BIA-derived body composition values, age, and gender. The next 36 patients were used to compare the accuracy of these equations with the Harris-Benedict equations in estimating REE. RESULTS: Using BLA-derived body composition values, gender, and age, predictive equations were developed for REE that explained approximately 65% of the variance. Inclusion of other BIA or anthropometric parameters did not improve the equations. When compared with the Harris-Benedict equations, the equations developed in this study were significantly more accurate, providing an REE estimate that was closer to the measured value in about 75% of patients. CONCLUSIONS: These results indicate that BLA-derived body composition estimates may be used to more accurately predict the energy requirements of patients receiving nutrition support than calculations based on the Harris-Benedict equations.
BACKGROUND: The energy requirement of a patient receiving nutrition support is typically estimated by calculating the basal energy expenditure (BEE) using the Harris-Benedict equations and multiplying by stress and activity factors. Because fat-free mass (FFM) and fat mass (FM) are important determinants of BEE, we hypothesized that body composition estimates derived from bioelectrical impedance analysis (BIA) could be used to develop predictive equations for resting energy expenditure (REE) that were more accurate than those calculated using the Harris-Benedict equations. METHODS: Seventy-six adults referred to the nutrition support service were studied. REE was measured by indirect calorimetry, and single-frequency BIA was used to estimate FFM and FM. Using the first 20 male and 20 female patients, predictive equations for REE were developed by multiple regression analysis, using BIA-derived body composition values, age, and gender. The next 36 patients were used to compare the accuracy of these equations with the Harris-Benedict equations in estimating REE. RESULTS: Using BLA-derived body composition values, gender, and age, predictive equations were developed for REE that explained approximately 65% of the variance. Inclusion of other BIA or anthropometric parameters did not improve the equations. When compared with the Harris-Benedict equations, the equations developed in this study were significantly more accurate, providing an REE estimate that was closer to the measured value in about 75% of patients. CONCLUSIONS: These results indicate that BLA-derived body composition estimates may be used to more accurately predict the energy requirements of patients receiving nutrition support than calculations based on the Harris-Benedict equations.