W' reconstitution accelerates more with decreasing intensity in the heavy versus the moderate intensity domain

Introduction: The purpose was to investigate the effect the recovery intensity domain on W' reconstitution. We used the W'BAL model as a framework and tested its predictive capabilities (W'PRED) across the different intensity domains. Methods: Twelve young men (51.7 ± 5.9 mL•kg•min) completed a ramp incremental test, three to five constant power output (PO) tests to determine critical power (CP) and W' and minimally two trials to verify the maximal lactate (La) steady state. During four experimental trials, subjects performed two work bouts (WB1 and WB2) at P6 (i.e., PO that predicts exhaustion within 6 min) separated by a recovery interval at CP - 10 W, Δgas exchange threshold (GET)-CP, GET and 50% GET, respectively. WB1 was designed to deplete 75% W' and the recovery time varied in order to replenish 50% W'. WB2 was performed to exhaustion (W'ACT). W'PRED was compared with W'ACT to evaluate the accuracy of the W'BAL model. Excess post-exercise oxygen consumption (EPOC) was calculated as the difference between the measured and predicted oxygen uptake during recovery. Results: W'ACT averaged 49%±24%, 69%±24%, 81%±28% and 93%±21% for respectively CP - 10 W, ΔGET-CP, GET and 50% GET (P=0.002). W'PRED overestimated W'ACT in CP-10W (34%±32%, P=0.004) and underestimated W'ACT in 50% GET (24%±28%, P=0.013). EPOC was lowest in CP - 10 W (P<0.01) and higher in GET compared to ΔGET-CP (P=0.01). Conclusion: We demonstrated that W'PRED overestimated and underestimated W'ACT in the heavy and moderate intensity domain, respectively. Therefore, the practical applicability of a single recovery time constant, which only relies on the difference between the recovery PO and CP, is questionable.