Quantitative interpretation and modeling of continuous nonprotein respiratory quotients

The respiratory exchange ratio (RER), which is the ratio of total carbon dioxide produced over total oxygen consumed, serves as a qualitative measure to determine the substrate usage of a particular organism on the whole body level. Quantification of RER by its direct conversion into %glucose- (%G_ox) and %lipid oxidation (%L_ox) at a given timepoint can be done by utilizing nonprotein respiratory quotient tables. These tables, however, are limited to specific increments, and intermediate RER values are not covered by these tables. RER data are mostly continuous, which requires faithful interpolation, which we aimed for here. We first determined, statistically and schematically, that linear interpolation would lead to incorrect values. Therefore, we constructed a new mathematical model as an interpolating strategy to translate continuous RER values into correct values of %G_ox and %L_ox. We validated our new mathematical model against the original table by Péronnet and Massicotte (Can J Sport Sci 16: 23-29, 1991), against a linear interpolation of these data, as well as against a model based on an exponential approach using a dataset of a nutritional intervention study in mice. This showed that our model outperforms the other methods, providing more accurate data. We conclude that applying our mathematical model will lead to an increase in data quality and offer a very simple, straightforward approach to obtain the best %G_ox and %L_ox levels from continuous RER values.NEW & NOTEWORTHY With the here proposed mathematical model, we provide a new tool to convert continuous RER data into more accurate estimations of %G_ox and %L_ox. It circumvents the use of nonprotein respiratory quotient tables and thereby aids and simplifies by automating the conversions. The model can further be implemented into software commonly used for indirect calorimetry measurements and thereby provides %G_ox and %L_ox data in real-time during a running experiment.