The Reliability Of A Lightweight Portable Metabolic Cart

The use of metabolic carts to measure energy expenditure has broad use in health and sport. Portable metabolic analyzers provide an opportunity to measure energy expenditure in applications outside the constraints of the normal laboratory environment. A recent development in portable metabolic analyzers, the VO2 Master Pro (Device A), has been made broadly available for commercial, personal, and academic use. But there is limited published data regarding the reliability of this device or comparisons to other devices. PURPOSE: To test the reliability of device A as well as to compare it to the Cosmed Quark (Device B). METHODS: Eight cyclists participated in two different trials which consisted of graded cycling stages at 50%, 70%, and 85% of critical power while expired gasses were analyzed by both devices in breath-by-breath mode. Device A was used during the first visit with 10-minute stages and measurements averaged during the final 3 minutes of each stage. Both device A and B were used during the second visit with stages lasting 15 minutes and measures from device A or B were averaged during minutes 7-10 and 12-15. Equivalence testing was done by TOST analysis, with the equivalence bounds set to approximately resting VO2 (150 ml·min-1) of the participants, while tests of difference were done by repeated measures ANOVA using the average of two measures for device A compared to the measure for device B. RESULTS: Device A demonstrated good reliability by TOST analysis within the equivalence bounds set, t-test p = 0.744, lower p = 0.005, upper p = 0.023 with the (mean ± SD) VO2 being 2590.97 ± 761.96 vs. 2611.29 ± 713.56. The mean bias error between the two trials was -20.32 ml·min-1 and mean absolute error was 214.66 ml·min-1. The measure of VO2 ml·min-1 approached statistical significance between device A and B p = 0.057 with a mean of 2601.13 ± 730.34 vs 2773.78 ± 615.03 ml·min-1 respectively. the TOST analysis was unable to reject the null equivalence hypothesis p = 0.003, lower p = 0.667, upper p < 0.001, indicating that the measurement of VO2 was not equivalent between the devices. CONCLUSION: These results indicate that device A demonstrates good reliability across repeated measures within a range of an approximate resting VO2. There may be some difference in the reported VO2 between the two devices.