Accuracy and preference of measuring resting energy expenditure using a handheld calorimeter in healthy adults

Background: Accurate estimates of energy expenditure are required in clinical nutrition in order to determine the requirements of individuals and to inform feeding regimes. Calorimetry can provide accurate measurements but is often impractical in clinical or community settings; prediction equations are widely used to estimate resting energy expenditure (REE) but have limited accuracy. A portable, self-calibrating, handheld calorimeter (HHC) may offer an alternative way of determining REE. The aim of the study was to evaluate whether estimates of REE derived using an HHC are closer to accurate measurements than values calculated using selected prediction equations. Method: REE was measured in 36 healthy adults aged 21-58 years using a flow-through indirect calorimeter (FIC) and HHC. Estimated REE was calculated using three predictive equations (Harris & Benedict; Schofield; Henry). Differences in REE between the ‘gold standard’ values derived using the FIC and those derived using the HHC and equations were examined using paired t-tests and Bland Altman plots. Results: Mean REEHHC was significantly lower than mean REEFIC (1089±249 vs 1489±214 kcal, P=0.000) and also significantly lower than mean values calculated using all three equations. The mean difference between REEHHC and REEFIC (400±217 kcal) was significantly greater than the mean differences between the values calculated using the three prediction equations (65±117 [Harris-Benedict], 63±122 [Schofield], 20±120 [Henry] kcal, P=0.000). Conclusion: The HHC provides estimates of REE in healthy people that are less accurate than those calculated using the prediction equations and so does not provide a useful alternative.