Running Economy: Comparing Alternative Measurement Models

摘要

The oxygen cost of running (OCR) is the oxygen consumed per unit of distance. Running economy (RE) refers to individual differences in OCR. OCR and RE are measured by dividing the oxygen uptake (mlO2) rate during a submaximal run by the speed (s) of that run. The results of these measurements are assumed to be independent of speed. This assumption is invalid unless the intercept (b(sub 0)) of the regression of mlO2 on s is zero. When b(sub 0) not = 0, standard measures are biased. The magnitude of bias is b(sub 0)/s, so bias is contingent on speed. Nevill, Cooke, Holder, Ramsbottom, and Williams (1992) presented evidence that, in general, b(sub 0) not = 0. Instead, b(sub 0) differs between people, so bias is a problem that may obscure important relationships in RE research. This report presents a re-analysis of published data from 5 studies to replicate Nevill et al.'s (1992) findings and examine plausible alternatives to their model. Nevill et al. s findings did replicate, but an alternative mathematical model that replaced differences in b(sub 0) with differences in b(sub 1), the slope of the regression of mlO2 on s, was preferable. The alternative model was slightly more accurate in predicting mlO2 and the variation in individual differences in b(sub 1) was significant while the variation in differences in b(sub 0) was not. The alternative model also was preferable because differences in b(sub 1) were interpretable in bioenergetic terms. A high proportion of negative values made the interpretation of differences in b(sub 0) uncertain. Because the alternative model included a fixed b(sub 0), it implied that RE measures should be based on net oxygen consumption. Subtracting approximately 3.2 ml kg-1 min-1 from measured ml02 would eliminate a speed-related bias. This approach would not be entirely novel. Roughly 8% of RE studies currently use a net uptake approach. The present findings suggest that this practice should become the preferred method of assessing RE.