Purpose: To verify the accuracy of predicting performance in the severe-intensity domain by means of end-test power output (EP) and the work performed above EP (WEP) obtained from a 3-min all-out test in competitive cyclists.
Methods: Ten welltrained cyclists performed a ramp incremental test and a 3-min all-out familiarization test. Subsequently, they performed a 3-min all-out experimental test to obtain EP and WEP and a 10-min time trial (TT). The actual 10-min-TT mean power output was then compared with the power output predicted as P = WEP/Tlim + EP, where Tlim corresponds to 600 s. The ramp-test peak power output (PPO) was compared with PPO predicted as [Formula: see text], where S represents the ramp slope (0.5 W/s).
Results: The actual (347 ± 30 W) and predicted (376 ± 48 W) 10-min TT mean power output were correlated (r = .87, P = .001) but significantly different (P < .01). The coefficient of variation (CV) between the predicted and actual performance was 5.6% ± 4.4%. The error of prediction was positively correlated to EP (r = .80, P = .005) and negatively correlated to WEP (r = -.71, P = .021). No significant difference was found between the 10-min-TT mean power output and EP (351 ± 53 W). The actual (438 ± 30 W) and predicted (472 ± 41 W) ramp PPO were correlated (r = .88, P < .001) but significantly different (P < .001). The CV between the predicted and actual PPO was 5.2% ± 3%. The error of prediction was positively correlated to EP (r = .63, P = .051).
Conclusions: EP and WEP obtained from a 3-min all-out test overestimate severe-intensity performance in competitive cyclists.
Keywords: critical power; cycling; performance prediction; power-duration relationship.