Different durations within the method of best practice affect the parameters of the speed-duration relationship

The aim of the study was to determine whether estimates of the speed-duration relationship are affected using different time-trial (TT) field-based testing protocols, where exhaustive times were located within the generally recommended durations of 2-15 min. Ten triathletes (mean ± SD age: 31.0 ± 5.7 years; height: 1.81 ± 0.05 m; body mass: 76.5 ± 6.8 kg) performed two randomly assigned field tests to determine critical speed (CS) and the total distance covered above CS (D́). CS and D́ were obtained using two different protocols comprising three TT that were interspersed by 60 min passive rest. The TTs were 12, 7, and 3 min in Protocol I and 10, 5, and 2 min in Protocol II. A linear relationship of speed vs. the inverse of time (s = D́ × 1/t + CS) was used to determine parameter estimates. Significant differences were found for CS (p = 0.026), but not for D́ (p = 0.123). The effect size for CS (d = 0.305) was considered small, while that for D́ was considered moderate (d = 0.742). CS was significantly correlated between protocols (r = 0.934; p < 0.001), however, no correlation was found for D́ (r = 0.053; p = 0.884). The 95% limits of agreement were ±0.28m s^-1 and ±73.9 m for CS and D́, respectively. These findings demonstrate that the choice of exhaustive times within commonly accepted durations results in different estimates of CS and D́, and thus protocols cannot be used interchangeably. The use of a consistent protocol is therefore recommended, when investigating or monitoring the speed-duration relationship estimates in well-trained athletes.