Critical power and work-prime account for variability in endurance training adaptations not captured by V̇o 2max .

Responses to exercise at a given percentage of one's maximum rate of oxygen consumption (V̇o 2max ), or percentage of the power associated with V̇o 2max during a graded exercise test (i.e., P GXT ), vary. The purpose of this study was to determine if differences in critical power (P CRIT , maximum metabolic steady state) and work-prime (W', the amount of work tolerated above steady state) are related to training-induced changes in endurance. P CRIT , W', V̇o 2max , and other variables were determined before and after 22 adults completed 8 wk of either moderate-intensity continuous training (MICT) or high-intensity interval training (HIIT) performed at fixed percentages of P GXT . On average, P CRIT increased to a greater extent following HIIT (MICT: 15.7 ± 3.1% vs. HIIT: 27.5 ± 4.3%; P = 0.03), but the magnitude of change varied widely within each group (MICT: 4%-36%, HIIT: 4%-61%). The intensity of the prescribed exercise relative to pretraining P CRIT , not P GXT , accounted for most of the variance in changes to P CRIT in response to a given protocol ( R 2  = 0.61-0.64; P < 0.01). Although P CRIT and V̇o 2max were related before training ( R 2  = 0.92, P < 0.01), the training-induced change in P CRIT was not significantly related to the change in V̇o 2max ( R 2  = 0.06, P = 0.26). Before training, time-to-failure at P GXT was related to W' ( R 2  = 0.52; P < 0.01), but not V̇o 2max ( R 2  = 0.13; P = 0.10) . Training-induced changes in time-to-failure at the initial P GXT were better captured by the combined changes in W' and P CRIT ( R 2  = 0.77, P < 0.01), than by the change in V̇o 2max ( R 2  = 0.24; P = 0.02). Differences in P CRIT and W' account for some of the variability in responses to endurance exercise. NEW & NOTEWORTHY As the highest percentage of V̇O 2max at which steady state conditions can be achieved, a person's critical power (P CRIT ) strongly influences the metabolic strain of a given exercise. In this study we demonstrate that training-induced changes in endurance are more strongly related to the intensity of an exercise training program, relative to P CRIT than relative to V̇o 2max . Thus, exercise may be more homogenously and effectively prescribed in relation to P CRIT than traditional factors like V̇o 2max .