Optimization of VO2 and VCO2 outputs for the calculation of resting metabolic rate using a portable indirect calorimeter

This study aimed to compare the Cosmed K5 portable indirect calorimeter, using the mixing chamber mode and face mask, with a stationary metabolic cart when measuring the resting metabolic rate (RMR) and to derive fitting equations if discrepancies are observed. Forty--three adults (18--84 years) were assessed for their RMR for two 30--min consecutive and counterbalanced periods using a Cosmed K5 and an Oxycon Pro. Differences among devices were tested using paired sample Student's t--tests, and correlation and agreement were assessed using Pearson's correlation coefficients, intraclass correlation coefficient and Bland--Altman plots. Forward stepwise multiple linear regression models were performed to develop fitting equations for estimating differences among devices when assessing oxygen uptake (VO(2)diff, mL center dot min-1) and carbon dioxide production (VCO2diff, mL center dot min-1). Furthermore, the Oxycon Pro was tested before being confirmed as a reference device. Significant differences between devices were found in most metabolic and ventilatory parameters, including the primary outcomes of VO2 and VCO2. These differences showed an overestimation of the Cosmed K-5 in all metabolic outcomes, except for Fat, when compared to the Oxycon Pro. When derived fitting equations were applied (VO2diff - 139.210 + 0.786 [weight, kg] + 1.761 [height, cm] --0.941 [Cosmed K-5 VO2, mL center dot min(-1)]; VCO(2)diff - 86.569 + 0.548 [weight, kg] + 0.915 [height, cm] --0.728 [Cosmed K-5 VCO2, mL center dot min(-1)]), differences were minimized, and agreement was maximized. This study provides fitting equations which allow the use of the Cosmed K5 for reasonably optimal RMR determinations.