Validity Of Two Automated Metabolic Gas Analysis Systems At Rest And During Exercise

Abstract

Apart from the conventional Douglas bag method there are numerous automated metabolic carts and portable systems for metabolic analysis (Macfarlane, 2001). Before portable systems can be used in free-living settings, it is important to determine the validity of their metabolic measurements during controlled laboratory exercise. PURPOSE To investigate the validity of a metabolic cart and a portable metabolic analysis system against the Douglas bag criterion. METHODS Eight subjects (5 males and 3 females, 22.3±2.3years) performed a trial at rest and three stages of cycling exercise. Each exercise stage consisted of a 13-minute steady-state measurement of physiological variables using all three systems: the Douglas bag (as criterion), a metabolic cart, and a portable system. Exercise levels were categorized by calculating the MET values of each subject, with moderate = 3.0- 5.9 METs, and vigorous = >6.0 METs. Comparisons of the validity between the two systems against the criterion were made using a RM-ANOVA. RESULTS Comparisons against the criterion showed no significant differences between minute ventilation (VE), oxygen uptake (VO2) and carbon dioxide production (VCO2) measured by the metabolic cart and the portable metabolic analysis system at rest. During moderate and vigorous exercise, no significant differences were found in VE measurements by both machines. However, for VO2 and VCO2, a significant overestimation by of 0.12 I/min (10.7%) and 0.14 l/min (11.6%) during moderate exercise, and 0.27 l/min (13.5%) and 0.39 l/min (14.9%) during vigorous exercise were found in the portable system. CONCLUSIONS The metabolic cart was a valid measure of metabolic activity variables during rest and both levels of exercise. Although the portable system could accurately measure VE across all activity levels, it significantly overestimated VO2 and VCO2 during moderate and vigorous exercise, suggesting further investigations on its performance are warranted. Supported by a HKU CRCG Research Grant