Abstract
Open-circuit whole-room indirect calorimetry is a reliable method for assessing cumulative energy expenditure and substrate oxidation over a prolonged period without disrupting normal activities such as eating and sleeping. Over the past three decades, efforts to improve the time resolution of indirect calorimetry have increased, aiming to synchronize energy metabolism with rapidly changing behaviors and physiologic signals. As such, various new algorithms have been proposed: trend identification, total variation denoising, wavelet de-noising, Kalman estimation, and deconvolution with a regularization parameter. The ability to estimate the time course of energy metabolism differs among algorithms. In this review on the methodology of open-circuit whole-room indirect calorimetry, we discuss similarities and critical differences in the theoretical background of various algorithms and compare their performance in simulations. In terms of correlation with the true value and mean squared errors, deconvolution with a regularization parameter outperformed the other algorithms. As further improvements in the algorithmic approaches are expected, we propose a practical guideline to support ongoing development: newly developed algorithms should be benchmarked against existing algorithms, and the source code should be made publicly available to facilitate transparent evaluation and future comparisons.