Comparison and cross-validation of cycle ergometry estimates of VO2max

The purpose of this study was to examine the accuracy and quality of a series of cycle ergometry tests used to estimate maximal aerobic capacity (VO2max). One-hundred nine males and 71 females participated in five tests: a maximal exercise test on the treadmill, three Air Force Cycle Ergometry Tests (AF1, AF2, AF3), and a Progressive cycle ergometry test (PROG). The VO2max value measured during the treadmill test was compared with the VO2max estimates from each ergometry test. The AF1, AF2, and AF3 results were used to determine reliability. The mean estimated VO2max for each, except the PROG, was significantly different (P < 0.05) from the measured VO2max. The AF1 and AF-avg tests underestimated VO2max by 8.0 and 6.5 mL.kg-1.min-1, respectively, values which were 17.3 and 14.9% lower than the measured VO2max. Correlation coefficients between estimated VO2max values and the measured VO2max ranged from 0.59 to 0.80 with SEE ranging from 1.8 to 2.2 mL.kg-1.min-1. The PROG had the greatest sensitivity (82.2%), while the AF2 had the greatest specificity (70.6%). Additionally, 23.4% of the VO2max estimates from the PROG were within +/- 5% of the measured VO2max compared with 9.9% for the average of the Air Force tests. The intraclass correlation coefficient for the three AF tests or the reliability for a single AF test was 0.26. In sum, the Air Force test provides an estimate of VO2max, and hence aerobic capacity, which is unreliable and underestimates the true VO2max by approximately 15%.