Reliability Analysis of the 3-min All-Out Exercise Test for Cycle Ergometry

Abstract

The 3-min all-out exercise test (3 MT) is purported to estimate critical power (CP) and anaerobic work capacity (W') and serve as an exercise mode for measuring maximum oxygen uptake (VO(2max)). Reliability analysis of the 3 MT has been confined to CP, not W', and verification of "true" VO(2max)was exclusive to a graded exercise test (GXT). We conducted a reliability analysis of the 3 MT and compared VO(2max)values from the 3 MT with a GXT and an exhaustive square-wave verification bout. Upon completion of a custom GXT and square-wave verification protocol, 11 subjects of various aerobic powers completed two 3 MTs (separate visits). CP, W', average power during 150 s from the 3 MT, and VO(2max)values were assessed using typical error (TE), coefficient of variation (CV), and intraclass correlation (α). CP (W) (trial 1 = 206 ± 47, trial 2 = 206 ± 42) did not differ between 3 MT trials (P = 0.37) and was reliable (TE = 15 W, CV = 7%, α = 0.93). W' was less reliable (TE = 2864 J, CV = 28%, α = 0.76) but did not alter power-duration estimates from the two 3 MTs (P > 0.05). Variability for VO(2max)(TE (mL·kg(-1)·min(-1)) between the GXT and the verification bout (1.16) was more consistent than the first (2.03) or second (2.69) 3 MT. The 3 MT yields reliable estimates of CP and consistent estimates of the power-duration relationship. Power for 150 s, in comparison with W', is a more reliable metric of short-term power performance. The square-wave protocol is recommended over the 3 MT for verifying true VO(2max).