The ramp and all-out exercise test to determine critical power: validity and robustness to manipulations in body position

Abstract

Purpose: The purpose of the present study was to determine whether a contiguous ramp and all-out exercise test could accurately determine critical power (CP) in a single laboratory visit during both upright and supine cycle exercise.

Methods: Healthy males completed maximal ramp-incremental exercise on a cycle ergometer in the upright (n = 15) and supine positions (n = 8), with task failure immediately followed by a 3-min all-out phase for determination of end-test power (EP). On separate days, participants undertook four constant-power tests in either the upright or supine positions with the limit of tolerance ranging from ~ 2 to 15 min for determination of CP.

Results: During upright exercise, EP was highly correlated with (R² = 0.93, P < 0.001) and not different from CP (CP = 221 ± 40 W vs. EP = 226 ± 46 W, P = 0.085, 95% limits of agreement - 30, 19 W). During supine exercise, EP was also highly correlated with (R² = 0.94, P < 0.001) and not different from CP (CP = 140 ± 42 W vs. EP = 136 ± 40 W, P = 0.293, 95% limits of agreement - 16, 24 W).

Conclusion: The present data suggest that EP derived from a contiguous ramp all-out exercise test is not different from the gold-standard method of CP determination during both upright and supine cycle exercise when assessed at the group level. However, the wide limits of agreement observed within the present study suggest that EP and CP should not be used interchangeably.

Keywords: All-out exercise; Critical power; Exercise testing; Performance; Power–duration relationship.

Fig. 1

Group mean (black circles, with SD displayed as error bars) responses of power output (left) and $\dot{V}$O₂ (right) during the ramp all-out test for negative pacing in the upright group (A, B), even pacing in the upright group (C, D) and the supine group (negative pacing only; E, F), plotted relative to the start of the 3-min sprint phase. Responses were averaged into 30 s epochs to facilitate comparisons. Following task failure during the ramp phase, the ergometer was switched from its cadence-independent mode to its cadence-dependent mode and participants performed a 3 min all-out exercise bout. * indicates significantly different from previous 30 s (P < 0.05)

Fig. 2

A linear regression (solid black line) demonstrating the degree of relative agreement between end-test power and critical power for the upright group (A). A Bland–Altman analysis showing the degree of agreement between the end-test power and critical power for each participant (B, clear circles), the mean bias (solid black line) and the 95% limits of agreement (LoA, dotted black lines) are also displayed

Fig. 3

A Bland–Altman analysis showing the degree of agreement between W′ and the WEP for each participant (clear circles) for the upright group (A) and the supine group (B). The mean bias is displayed as a solid black line, whereas dotted black lines indicate the 95% limits of agreement (LoA)

Fig. 4

A linear regression (solid black line) demonstrating the degree of relative agreement between end-test power and critical power for the supine group (A). A Bland–Altman analysis showing the degree of agreement between the end-test power and critical power for each participant (B, clear circles), the mean bias (solid black line) and the 95% limits of agreement (LoA, dotted black lines) are also displayed