Cross Validation of Two Regression Equations Developed to Predict Met Levels on the Nustep 4000 Recumbent Stepper

Abstract

In a companion study. Rateike (2000) developed two regression equations to predict metabolic equivalents (METS) above and below 1 15 watts on the NuStep 4000 Recumbent Stepper. The purpose of this study was to cross-validate these predictions in a separate sample. Eighteen patients with either cardiac or pulmonary disease (mean age = 69.1 yrs) were used as subjects. Each subject started at either 25 or 50 watts and increased by 25 or 50 watts per stage until an RPE of 15 was reached. Each stage was 5 minutes in duration and VO2 was measured continuously. A minute rest period separated each stage of exercise. There was no significant difference (p < 0.05) between measured and predicted MET values at any of the stages or overall (predicted = 3.05 METS, measured = 2.95 METS). The correlation between actual and predicted values was r = 0.91, standard error of the estimate (SEE) was 0.38 METS, and total error (TE) was 0.41 METS. These results indicate that the newly developed prediction equations provide excellent estimates of metabolic overload and should allow cardiopulmonary rehabilitation professionals to prescribe workloads on the Nustep with greater accuracy and confidence.