The Respiratory Compensation Point Is Not A Valid Surrogate Of Critical Power

Abstract

There is a growing opinion in the literature that Critical Power (CP) may be confidently estimated from the respiratory compensation point (RCP) obtained during incremental cycling. Opponents of this view argue that any reported correspondence between the RCP and CP is coincidental. It is known that the RCP is labile: i.e., its manifestation defers to higher work rates during fast compared with slow incremental protocols. PURPOSE: We sought to examine the agreement between CP and the work rates corresponding to the RCP obtained during incremental cycling protocols of varying ramp slopes. METHODS: 11 recreationally-active men (23 ± 1 yr; 75 ± 2 kg; 179 ± 2 cm; peak O2 uptake: 53 ± 2 ml·kg-1·min-1) completed 3 separate incremental protocols, where the work rate increment was slow (SR: 15 W·min-1), medium (MR: 30 W·min-1) and fast (FR: 45 W·min-1). Protocol order was randomised. The gas-exchange threshold (GET) and RCP were obtained via computerised analysis of the relationships between O2 uptake and CO2 output, and CO2 output and expired ventilation, respectively. CP was determined via the 3-min “all-out” cycling test. The assumption that the RCP and CP occur at equivalent work rates was assessed by evaluating the concordance correlation coefficient (CCC), and root-mean square error (RMSE), between RCP and CP for each protocol, separately. RESULTS: The GET was similar across ramp protocols (pooled average: 149 ± 8 W). RCP augmented (P < 0.05) with increasing ramp slope (SR = 237 ± 9 W; MR = 253 ± 10 W; FR: 271 ± 10 W). The RCP work rates determined during each ramp protocol were not different from the group-mean CP (249 ± 8 W). The degree to which the relationship between RCP and CP approximated the “line-of-identity” was relatively poor for SR (CCC = 0.589; RMSE = 31 W), MR (CCC = 0.587; RMSE = 32 W) and FR (CCC = 0.438; RMSE = 30 W). CONCLUSIONS: Our findings agree with those of previous studies in that the work rate at RCP augments with increasing ramp slope during incremental cycling. Moreover, despite occurring at similar mean work rates, the relatively poor CCC and large RMSE between CP and RCP for each ramp protocol suggests that the RCP should not be considered a valid surrogate of the CP.