Robustness of a 3 min all-out cycling test to manipulations of power profile and cadence in humans

The purpose of this study was to assess whether end-test power output (EP, synonymous with 'critical power') and the work done above EP (WEP) during a 3 min all-out cycling test against a fixed resistance were affected by the manipulation of cadence or pacing. Nine subjects performed a ramp test followed, in random order, by three cadence trials (in which flywheel resistance was manipulated to achieve end-test cadences which varied by ~20 r.p.m.) and two pacing trials (30 s at 100 or 130% of maximal ramp test power, followed by 2.5 min all-out effort against standard resistance). End-test power output was calculated as the mean power output over the final 30 s and the WEP as the power–time integral over 180 s for each trial. End-test power output was unaffected by reducing cadence below that of the 'standard test' but was reduced by ∼10 W on the adoption of a higher cadence [244 ± 41 W for high cadence (at an end-test cadence of 95 ± 7 r.p.m.), 254 ± 40 W for the standard test (at 88 ± 6 r.p.m.) and 251 ± 38 W for low cadence (at 77 ± 5 r.p.m.)]. Pacing over the initial 30 s of the test had no effect on the EP or WEP estimates in comparison with the standard trial. The WEP was significantly higher in the low cadence trial (16.2 ± 4.4 kJ) and lower in the high cadence trial (12.9 ± 3.6 kJ) than in the standard test (14.2 ± 3.7 kJ). Thus, EP is robust to the manipulation of power profile but is reduced by adopting cadences higher than 'standard'. While the WEP is robust to initial pacing applied, it is sensitive to even relatively minor changes in cadence