|Gribok, Andre - University Of Tennessee|
|Reed, Hoyt - Us Army Medical Research Institute|
|Buller, Mark - Us Army Medical Research Institute|
Submitted to: Physiological Measurement
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/12/2013
Publication Date: 5/29/2013
Citation: Gribok, A., Reed, H., Buller, M., Rumpler, W.V. 2013. On the accuracy of instantaneous gas exchange rates, energy expenditure, and respiratory quotient calculations obtained in indirect whole room calorimetry. Physiological Measurement. 34:737-755.
Interpretive Summary: Room calorimetry at the Beltsville Human Nutrition Research Center is based on the measurement of changes in the concentration of oxygen and carbon dioxide in the air enclosed in a 20000 l room. This measurement is inherently noisy due to the high sensitivity of the analysis required and the small changes in concentration affected by a human subject living in the measurement room. Traditionally, the data collection in room calorimeters has been of sufficiently long time periods to buffer the noise through the use of averaging. However, this approach has limited time periods to which meaningful changes in energy expenditure or fuel use can be examined from 4 to 24 hour. Recent interest in metabolic flexibility created the need to examine changes in fuel use over periods of time in the minutes not hours. In this paper we examine the use of two smoothing techniques, one previously published and a proposed new approach, on the sensitivity of room calorimeters to examine much shorter periods of time. These techniques will be of interest to scientists interested in understanding metabolism, obesity, and the prevention of unhealthy weight gain.
Technical Abstract: The molar balance equations of indirect calorimetry are treated from the point of view of cause-effect relationship where the gaseous exchange rates representing the unknown causes heed to be inferred from a known noisy effect – gaseous concentrations. Two methods of such inversion are analyzed. The first method is based on the previously published regularized deconvolution of the molar balance equation and the second one, proposed in this paper, relies on regularized differentiation of gaseous concentrations and subsequent use of the derivatives in the molar balance equations. The performance of both methods is tested on simulated as well as on real-world data sets. In addition, the uncertainties introduced by both methods are analyzed, and their dependence on the regularization parameter, selected through discrepancy principle, is studied. The gas exchange rate recovery demonstrates that both methods successfully solve the inverse problem delivering virtually identical results with similar uncertainties when the regularization parameters are optimally selected. However, to accurately estimate such a vital parameter as respiratory quotient, both techniques require to oversmooth the carbon dioxide production rate to avoid amplification of high-frequency components. Also, it is demonstrated through simulations as well as through analytical derivations that the uncertainty in respiratory quotient estimation depends reciprocally on the absolute level of gaseous exchange rates. To take this dependency into account for the real-world data, the analytical approach to quantify the uncertainty is proposed and its performance is analyzed.