The relationship between critical power and running performance

Abstract

Critical power is a theoretical concept that presumes there is a certain work‐rate which may be maintained without exhaustion. The extent to which critical power predicts running performance over varying distances has not been determined, and so the aim of this study was to correlate measurements of critical power in the laboratory to running performances in the field at 40 m and 1, 10 and 21.1 km in a group of 17 male longdistance runners (mean ± s.d. age = 31.7 ± 7.3 years). Each subject ran to exhaustion on the treadmill in the laboratory at six different speeds, ranging from 17 to 25 km h^‐1. Least squares analyses were used to fit an exponential decay to the relationship between the running speed (y) versus time to exhaustion (x). Critical power was calculated as the running speed (y) coinciding with the asymptote or C parameter of the y = A‐e^(‐Bx) + C relationship. The VO₂ max was also measured in all subjects. For the data in the field, each subject was timed over 40 m and 1 km and participated in 10‐ and 21.1‐km races. The mean critical power of die subjects in this study was 18.5 ± 1.6 km h^‐1. The test‐retest correlation coefficient for the determination of critical power was r = 0.99. The mean VO₂ max, measured in a progressive exercise protocol starting at 13 km h^‐1 and increasing by 1 km h^‐1 every minute, was 59.2 ± 4.6 ml O₂ kg^‐1 min^‐1. The 40‐m times ranged from 5.57 to 6.95 s, the 1‐km times from 2:46 to 3:55 min:s, the 10‐km times from 30:43 to 42:02 min:s and the 21‐km times from 67:00 to 95:45 min:s. Critical power predicted running times over 1 km (r = ‐0.75, P< 0.001), 10 km (r = ‐ 0.85, P< 0.00001) and 21.1 km (r = ‐ 0.79, P< 0.001) in this heterogeneous group of runners. The correlation coefficients for VO₂ max and running performances were similar to the above at all distances. Even in the best relationship between critical power and the time taken to run 10 km, only 72% of the variation in the 10‐km running time could be accounted for by differences in critical power. While the test of critical power may be repeatable and correlate significantly with VO₂ max (r = 0.77, P< 0.001), the measurement lacks the degree of specificity required to predict running performance in trained subjects of varying ability.

Keywords:

• Critical power
• running performance
• VO₂ max

Notes

Author to whom all correspondence should be addressed.