Measured and predicted resting energy expenditure in clinically stable patients

Resting energy expenditure (REE), measured by bedside indirect calorimetry, was compared to estimated REE by the Harris-Benedict and Kleiber predictors in 200 clinically stable hospitalized patients (100 males, 100 females) and 72 healthy control subjects (20 males, 52 females).Mean predicted values were not significantly different from measured REE for the male patients and control subjects, but measured REE was significantly overestimated by the Kleiber formula in female patients and controls (p<0.01). In comparison to control subjects, a substantially larger range of individual differences between measured REE and resting energy expenditure as estimated by the Harris-Benedict and Kleiber formulae existed among the male and female patient samples. Measured REE was over or underestimated by greater than 10% via the Harris-Benedict predictors in 40% of the patients but only 20% of the healthy controls. The Kleiber formulae were inappropriate for 46% of the individual patients and 33% of the normal subjects.Since no method exists for identifying the clinically stable patient for whom REE cannot be estimated via these commonly employed predictors, the bedside measurement of resting energy expenditure is the most appropriate method for deriving caloric expenditure and designing subsequent caloric provision regimens for adequate and safe nutritional repletion or maintenance.