In incremental cardiopulmonary exercise testing, the averaging of data is usually performed to provide group mean data for statistical purposes. They are usually presented as averaged maximum values, or as averaged data at different exercise levels. However, during incremental exercise testing the change in metabolic status may vary between subjects, thus averaging data may not classify the metabolic status accurately. We present an averaging method using a segmented ordinal scale based on individual maximal work performance and the anaerobic threshold (AT). Individual exercise data are grouped into ten classes ranging from unloaded exercise to maximal exercise. The classes are defined in relation to the AT, resulting in an ordinal scale of four classes for exercise data below the AT, one class at the AT and five classes beyond the AT. Resting and unloaded pedalling are treated as separate classes. For evaluation, this method of classification is compared to one based on an absolute scale of oxygen uptake (Cabs) and to another based on a relative scale in 10% steps of maximal oxygen uptake (Crel). Ten healthy male subjects (mean age 23.3 years) performed a ramp cycle ergometer test. When using the Cabs classification method for mean data averaging, mean values for performance at high-intensity exercise were calculated using data from only two of the ten subjects because of variations in individual work capacity. In addition, the AT data were distributed across four classes, thus anaerobic and aerobic exercise data were mixed. Using the Crel classification method enabled data for all ten subjects to be included in the calculation of every data point, but the AT values were still distributed across three classes, resulting in the mixing of anaerobic and aerobic exercise data. However, using the segmented ordinal scale method of classification enabled data from all ten subjects to be included in the calculation of all data points, and it permitted the grouping of the AT values into one class. Thus, this latter method more accurately represents the data of the whole group under study and it allows the metabolic status of the subjects to be taken into consideration.
Read full abstract