Prediction of the metabolic cost of walking with and without loads

Publisher Website
Google Scholar
Abstract
Measurement of the metabolic cost of walking inconveniences subjects, and requires skilled technical support and expensive equipment. These factors have stimulated interest in predictive equations. The present study assessed existing equations. Under each of 17 combinations of gradient (0–6%) and carried load (4·1–37·4 kg), 7–12 men undertook treadmill walking at 1·67 m/s. Measured oxygen consumption and respiratory exchange ratio were used to calculate metabolic rate (MR observed) Metabolic rate was also predicted from the equation of Pandolf el al. (1977) (MR pandolf)and, where appropriate, from another five equations relating to walking without loads. MR observed and MR pandolf did not differ significantly (p>0·05) under any combination of gradient and load. The overall mean MR observed and MR pandolf of 609 W and 602 W, respectively, also did not differ significantly (p>0·05). These variables were highly correlated (r=0·94) with a standard deviation about the prediction error of 47 W. For level walking without loads, the mean predictions from the equations of Pandolf el al. (1977) and Cotes and Meade (1960) did not differ significantly (p>0·05) from the mean MR observed of 428 Watts, but four other equations overestimated by 17–74 W. In conclusion, the Pandolf el al. (1977) equation has given good results across the range of combinations of load and gradient tested, and the errors are considered acceptable for most practical purposes.