Between-Seasons Variability of Cyclists' Peak Performance: A Longitudinal Analysis of "Real-World" Power Output Data in Male Professional Cyclists

Purpose: The record power profile (RPP) has gained popularity as a method of monitoring endurance cycling performance. However, the expected variation of cyclists' performance between seasons remains unknown. We aimed to assess the between-seasons variability of peak performance (assessed through the RPP) in male professional cyclists. Methods: The study followed a longitudinal observational design. Sixty-one male professional cyclists (age 26 [5] y) with power output data from both training sessions and competitions were analyzed for a median of 4 consecutive seasons (range 2-12). The highest mean maximum power values attained for different durations (from 10 s to 30 min), as well as the resulting critical power, were determined for each season. Within-cyclist variability between seasons was assessed, and the upper threshold of expected changes (ie, twice the normal coefficient of variation) was determined. Results: All mean maximum power values showed an overall high agreement and low variability between seasons (intraclass correlation coefficient [ICC] = .76-.88 and coefficient of variation [CV] = 3.2%- 5.9%), with the lowest variability observed for long efforts (>1 min). Critical power showed an ICC and CV of .79 (95% CI, .70-.85) and 3.3% (95% CI, 3.0%-3.7%), respectively. Upper thresholds of expected variation were <12% for short efforts (<_1 min) and <8% for long efforts. Conclusions: "Real-world" peak performance assessed through the RPP shows a low variability between seasons in male professional cyclists-especially for long efforts-with expected variation being around 6% and 3% for short (<_1 min) and long efforts, respectively, and with changes >12% and >8%, respectively, being infrequent for these effort durations.