Agreement between the respiratory compensation point and various critical power estimates using a step-ramp-step protocol in cyclists.

PURPOSE: Respiratory compensation point (RCP) and critical power (CP) are both used to demarcate the upper boundary of steady-state exercise; however, their agreement may vary depending on the method used to 1) translate RCP to a power output (PO) and 2) derive CP. METHODS: In 19 cyclists, PO at RCP was calculated with or without correction for mean response time (MRT)-determined by two methods-and heavy-domain V̇O2 slow component, then compared against CP estimated from laboratory-based time trials using four models. RESULTS: PO at RCP derived using a step-ramp-step (SRS) protocol (RCPSRS; 246±40W) was significantly lower than uncorrected PO at RCP (303±48W) and PO at RCP estimates only correcting for MRT using the double-linear (278±47W) and moderate step (278±49W) approaches (P<0.001 for all). PO at RCPSRS was less than CP derived from the linear work-time (CPwork: 258±44 W; P=0.007) and linear power vs. 1/time models (CP1/t: 262±47 W; P<0.001) but not different from CP derived from the 2-parameter hyperbolic (CP2hyp; 256±43W; P=0.06) and 3-parameter hyperbolic (CP3hyp; 242±37W; P=0.91) models. Limits of agreement and intraclass correlation coefficients (ICC) indicated good agreement between RCPSRS and all CP models (ICC = 0.90 to 0.93). In contrast, work capacity above CP (W') derived from CP models or the ramp test showed poor concordance with each other (ICC = 0.19 to 0.44). CONCLUSION: Corrected RCPSRS from the single-visit SRS protocol agrees with hyperbolic-model CP values derived from multi-visit testing when estimating the maximal metabolic steady state.