Modeling of oxygen uptake kinetics during exercise testing in patients with chronic obstructive pulmonary diseases (COPD)

Abstract

The 6-Minute Walk Test (6MWT) is routinely used to quantify exercise capacity in patients with chronic cardio-pulmonary diseases. Cardiopulmonary exercise testing devices can measure oxygen exchange (VO2) kinetics during exercise. From original VO2 data, curves can be derived and parameters estimated. Simultaneous modeling of multiple kinetics requires nonlinear mixed models methodology. No such approach has been used to analyze multiple VO2 kinetics in both research and clinical practice so far. The aim was to describe functionality of a new statistical package that allows fitting nonlinear mixed models for automated curve fitting. Patients with COPD referred to the Department of Pulmonary Medicine of the University Hospital Basel for a 6MWT gave informed consent to participate to the study. A six-parameter nonlinear regression model describing the 3 main phases of the VO2 kinetics (before, during and after 6MWT) was defined and fitted using the new R package medrc. Our methodology was compared to a curve-by-curve approach. VO2 kinetics were measured in 61 patients with COPD classified into GOLD stages II, III and IV. The model was jointly fitted to the set of 61 oxygen kinetics. Significant differences between disease stages were found regarding maximum oxygen uptake during exercise testing (II vs. IV: p = 0.038), oxygen level after recovery (II vs. IV, p = 0.0013) and inflection point in the recovery phase (II vs. IV, p = 0.088). In comparison with the curve-by-curve approach, the estimates obtained by nonlinear mixed model regression show standard errors consistently smaller. Thus, mixed model methodology resulted in a gain in efficiency and power.