The duration of predicting trials influences time to fatigue at critical power

The present study compared times to fatigue at CP which had been calculated from relatively long duration predicting trials. With eight recreationally active males (mean age +/- SE = 18.6 +/- 2.1 years) having first cycled to fatigue on five occasions at different fixed work rates, CP was calculated in three ways: 1. using data from the three lowest exercise intensities (CPl); 2. using data from all five exercise intensities (CPa); and 3. using data from the highest three exercise intensities (CPh). Although CP was calculated using a linear and a three-parameter non-linear model, there were insufficient suitable data to complete the latter analysis. After three days and over an eight day period, the subjects cycled for up to 60 minutes at each of the three CPs calculated using the linear model. Analysis revealed that despite high linearity with the five-point work-time regression (average r2=0.996), CPl, CPa and CPh significantly differed to each other (268 +/- 17.5W; 285 +/- 12.1W; 321 +/- 8.8W respectively; p<0.05). Significant differences were also found between times to fatigue at CPl, CPa and CPh (42.9 +/- 3.9, 39.9 +/- 4.6 and 34.4 +/- 2.7 minutes respectively; p<0.05). The data show that when CP is calculated using a linear work-time regression, time to fatigue at CP is significantly influenced by the duration of predicting trials. Moreover, exercise at CP could only be maintained for 43 minutes despite CP being calculated from predicting trials averaging between 10 and 25 minutes.