BACKGROUND: Predicting resting energy expenditure (REE) in malnourished hospitalized older patients is important for establishing optimal goals for nutritional intake. Measuring REE by indirect calorimetry is hardly feasible in most clinical settings. OBJECTIVE: To study the most accurate and precise REE predictive equation for malnourished older patients at hospital admission and again three months after discharge. DESIGN: Twenty-three equations based on weight, height, gender, age, fat free mass (FFM) and/or fat mass (FM) and eleven fixed factors of kcal/kg were compared to measured REE. REE was measured by indirect calorimetry. Accuracy of REE equations was evaluated by the percentage patients predicted within 10% of REE measured, the mean percentage difference between predicted and measured values (bias) and the Root Mean Squared prediction Error (RMSE). RESULTS: REE was measured in 194 patients at hospital admission (mean 1473 kcal/d) and again three months after hospital discharge in 107 patients (mean 1448 kcal/d). The best equations predicted 40% accuracy at hospital admission (Lazzer, FAO/WHO-wh and Owen) and 63% three months after discharge (FAO/WHO-wh). Equations combined with FFM, height or illness factor predicted slightly better. Fixed factors produce large RMSE's. All predictive equations showed proportional bias, with overestimation of low REE values and underestimation of high REE values. Correction by regression analysis did not improve results. CONCLUSIONS: The REE predictive equations are not adequate to predict REE in malnourished hospitalized older patients. There is an urgent need for either a new accurate REE predictive equation, or accurate easy-to-use equipment to measure REE in clinical practice.
BACKGROUND: Predicting resting energy expenditure (REE) in malnourished hospitalized older patients is important for establishing optimal goals for nutritional intake. Measuring REE by indirect calorimetry is hardly feasible in most clinical settings. OBJECTIVE: To study the most accurate and precise REE predictive equation for malnourished older patients at hospital admission and again three months after discharge. DESIGN: Twenty-three equations based on weight, height, gender, age, fat free mass (FFM) and/or fat mass (FM) and eleven fixed factors of kcal/kg were compared to measured REE. REE was measured by indirect calorimetry. Accuracy of REE equations was evaluated by the percentage patients predicted within 10% of REE measured, the mean percentage difference between predicted and measured values (bias) and the Root Mean Squared prediction Error (RMSE). RESULTS: REE was measured in 194 patients at hospital admission (mean 1473 kcal/d) and again three months after hospital discharge in 107 patients (mean 1448 kcal/d). The best equations predicted 40% accuracy at hospital admission (Lazzer, FAO/WHO-wh and Owen) and 63% three months after discharge (FAO/WHO-wh). Equations combined with FFM, height or illness factor predicted slightly better. Fixed factors produce large RMSE's. All predictive equations showed proportional bias, with overestimation of low REE values and underestimation of high REE values. Correction by regression analysis did not improve results. CONCLUSIONS: The REE predictive equations are not adequate to predict REE in malnourished hospitalized older patients. There is an urgent need for either a new accurate REE predictive equation, or accurate easy-to-use equipment to measure REE in clinical practice.
Authors: Suey S Y Yeung; Marijke C Trappenburg; Carel G M Meskers; Andrea B Maier; Esmee M Reijnierse Journal: Br J Nutr Date: 2020-03-16 Impact factor: 3.718