Validation of End-of-Session D' Balance for Quantifying and Differentiating Exhaustive and Nonexhaustive Intermittent-Training Sessions in Runners.

Purpose: Accurate modeling of intermittent-running sessions by D' balance would aid in understanding their physical demand and subsequent prescription of intermittent sessions. In intermittent cycling, W' balance can be validly quantified by modeling W' depletion and recovery; however, these models have not been validated in intermittent running, which was this project's purpose.

Methods: Critical speed (in meters per second) and D' (in meters) were calculated from the linear relationship between speed (in meters per second) and 1/time from season-best performances over 1500-, 3000-, and 5000-m race distances. Thirty intermittent-running sessions were classified as exhaustive (n = 14) and nonexhaustive (n = 16). D' balance was calculated from time-stamped running-speed data using D' depletion and recovery models, allowing end-of-session D' balance to be found. End-of-session D' balance for exhaustive and nonexhaustive intermittent sessions were compared, and the 95% CI around end-of-session D' balance in exhaustive intermittent sessions was evaluated for inclusion of 0.

Results: A moderate difference (effect size ± 95% CI = 1.01 ± 0.72; P = .01) existed between end-of-session D' balance in exhaustive (mean ± 95% CI = 157 ± 25 m) and nonexhaustive (200 ± 19 m) sessions. The 95% CI around end-of-session D' balance in exhaustive intermittent sessions did not include 0.

Conclusions: End-of-session D' balance did differentiate between exhaustive and nonexhaustive intermittent sessions. However, since D' balance at the end of exhaustive sessions should approximate 0, end-of-session D' balance did not accurately quantify exhaustive sessions as exhaustive. The inaccurate quantification of end-of-session D' balance during exhaustive intermittent running sessions is likely related to the exponential recovery of D' that is reliant on an accurately quantified recovery time constant.