A novel geometric method for determining the time constant for oxygen uptake kinetics

The kinetic response of oxygen uptake (V̇osub2/sub) to transitions of exercise intensity is one of the important parameters of aerobic function. The typical kinetic response between two steady states can be reasonably well fitted with a mono-exponential function that has a time constant τV̇osub2/sub. However, due to the variability of breath-by-breath measures of V̇osub2/subdetermination of τV̇osub2/subhas usually required superimposition of repeated exercise protocols. We developed a novel geometric method of determining τV̇osub2/subfrom the analysis of slopes and intercepts of a plot of cumulative oxygen uptake (cumV̇osub2/sub) versus time for single exercise protocols. We used mathematical modeling to generate 3,600 series of breath-by-breath V̇osub2/submeasures versus time for various exercise protocols. To test whether our geometric method was robust to the presence of real-life variability, we applied random (Gaussian) variation to both the interval between breaths and to the measured values of V̇osub2/sub. Our method derived values for τV̇osub2/subthat were accurate to within 1.5-3.5 s for both on- and off-transits and for models that represented healthy normal subjects as well as persons with cardiovascular disease. The coefficient of variation for multiple iterations of the method waslt;10% as long as the signal-to-noise ratio wasgt;20:1. We present a novel geometric method for deriving the τV̇osub2/subof oxygen uptake kinetics. This method uses analysis of the slopes and intercepts of a plot of cumulative V̇osub2/subversus time and does not require multiple repetitions of the exercise protocol but still gives accurate estimates of τV̇osub2/sub.bNEWamp; NOTEWORTHY/bWe present a novel geometric method for deriving the τV̇osub2/subof oxygen uptake kinetics. This method uses analysis of the slopes and intercepts of a plot of cumulative V̇osub2/subversus time and does not require multiple repetitions of the exercise protocol but still gives accurate estimates of τV̇osub2/sub.