Modeling the relationship between velocity and time to fatigue in rowing

Several mathematical models describe the relationship between velocity and time to fatigue. The purposes of this study were to evaluate different critical velocity (V(critical)) models applied to rowing ergometry and to investigate prediction of performance time in a 2000-m race based on results from shorter trials. Sixteen men performed seven rowing ergometer tests. Velocity and time to fatigue data from the 200-m (approximately 0.5 min) to 1200-m (approximately 3 min) trials and from the 200-m to 2000-m (approximately 6.5 min) trials were fit to a two-parameter hyperbolic model, a three-parameter hyperbolic model, and a three-parameter exponential model. Including data from the 2000-m trial generally resulted in higher R2 and smaller SEE. V(critical) from the three versions of the two-parameter model were 4.71 +/- 0.28 m x s(-1), 4.80 +/- 0.27 m x s(-1), and 5.04 +/- 0.24 m x s(-1) (P < 0.001). The two three-parameter models had high R2 (0.991 and 0.990, respectively) and generated parameter estimates that appeared reasonable. Time for a 2000-m race was predicted better using the two-parameter model (r > 0.974) than using the three-parameter models (r = 0.820-0.870). It is necessary to include the relatively long 2000-m predicting trial to describe accurately the velocity-time relationship in rowing. The two-parameter model may be useful in predicting time for a 2000-m race but is not otherwise appropriate for modeling when predicting trials of