Comparison of methods for achieving 24-hour energy balance in a whole-room indirect calorimeter.

Abstract

To evaluate and compare methods for achieving 24-hour energy balance in a whole-room indirect calorimeter. Twenty-four-hour energy expenditure (EE) for 34 healthy adults (16 women, 18 men) was measured in a calorimeter during a prestudy day and on a subsequent nonconsecutive assessment day (AD). Several methods for estimating EE on the AD using activity factors or regression equations with data available before the AD [anthropometrics, body composition, resting metabolic rate (RMR), sleeping metabolic rate (SMR) on prestudy day, 24-hour EE on prestudy day] were compared for predictive accuracy. Use of a 24-hour calorimeter stay gave the smallest mean absolute error (119 +/- 16 kcal/d) and smallest single maximum error (361 kcal/d). However, several other methods were only slightly, and not significantly, less accurate (e.g., mean absolute error = 131 +/- 17, 140 +/- 20, and 141 +/- 22 kcal/d and greatest error = 384, 370, and 593 kcal/d for anthropometric, RMR, and SMR regression equations, respectively). Fat-free mass alone and SMR with a simple activity factor were seen to be less accurate. Our results indicate that there may be some improvement in achieving 24-hour energy balance in a metabolic chamber by using a preceding 24-hour calorimeter stay; that only slightly less accurate predictions can be obtained using a combination of anthropometric, body composition, and/or RMR measurements; and that there is little or no advantage in using SMR from a previous overnight calorimeter stay.