Purpose: To 1) determine whether critical power (CP), W' and W' reconstitution kinetics (W'REC) are altered by high-intensity team-sport specific intermittent exercise; 2) examine the accuracy with which W' balance (W'BAL) is estimated by the W'BAL model; 3) examine whether physiological variables assessed during ramp incremental exercise and the 3-min all-out test, explain exercise-induced changes in the power-duration relationship. Methods: CP, W', and W'REC were determined during a repeated all-out test, which comprised a 3-min all-out test and a 2-min all-out test, separated by 90 s, in 13 recreationally active team-sport players. The test was performed in a rested condition (control) and immediately following a 40-min (40-IST) and 80-min (80-IST) intermittent sprint test. Results: CP assessed after the 40-IST (222 ± 57 W) and 80-IST (213 ± 49 W) were not different from control (222 ± 52 W; P > 0.05). W'REC was also not significantly different from control (0.08 ± 0.02 kJ·s-2) after the 40-IST (0.07 ± 0.02 kJ·s-2) or 80-IST (0.07 ± 0.02 kJ·s-2; P > 0.05). W' was lower than control (13.5 ± 4.0 kJ) after the 40-IST (10.1 ± 4.0 kJ) and 80-IST (9.1 ± 3.5 kJ) (both P < 0.05). After the 40-IST, W'BAL was not significantly different from the value predicted from the W'BAL model (10.9 ± 2.4 kJ; P > 0.05), but W'BAL was overestimated by the model following the 80-IST (10.7 ± 2.6 kJ; P < 0.05). There were considerable interindividual differences in the changes in the power-duration relationship following exercise, which were not related to physiological variables assessed in the control condition. Conclusions: CP and W'REC are not altered after intermittent exercise, but W' is decreased, with considerable heterogeneity in the changes in the power-duration relationship between individuals. These results have important implications for the application of the W'BAL model to describe intermittent exercise performance.
© Copyright 2026 Medicine & Science in Sports & Exercise. Lippincott Williams & Wilkins. Alle Rechte vorbehalten.