Methodological effects on the VO2-power regression and the accumulated O2 deficit

The VO2-power regression and O2 demand predicted for a supra-VO2peak intensity (i.e., 432 W) were determined in seven well-trained male cyclists (mean +/- SD: VO2peak = 5.29 +/- 0.51 l.min-1), using five incremental exercise protocols. These protocols were either continuous (CON) or discontinuous (DISCON), and comprised five to eight work bouts ranging in intensity between 40% and 85% VO2peak; the work bouts differed in duration (4-15 min), and the VO2 was measured during the 4th minute (CON4, DISCON4), from min 4 to 6 (DISCON6), 8 to 10 (DISCON10), or 13 to 15 (DISCON15) of each work bout. The y-intercepts of the VO2-power regressions were not different (P > 0.05), whereas the slope was higher (P < or = 0.01) when determined using DISCON10 (12.7 +/- 0.9 ml.min-1.W-1) and DISCON15 (12.5 +/- 0.9 ml.min-1.W-1) compared with DISCON6 (12.2 +/- 1.0 ml.min-1.W-1), DISCON4 (11.6 +/- 1.1 ml.min-1.W-1) or CON4 (11.9 +/- 0.7 ml.min-1.W-1). The O2 demand (at 432 W) was also higher (P < or = 0.01) for DISCON10 (6.05 +/- 0.29 l.min-1) and DISCON15 (6.05 +/- 0.28 l.min-1) compared with DISCON6 (5.88 +/- 0.31 l.min-1), DISCON4 (5.70 +/- 0.31 l.min-1) and CON4 (5.82 +/- 0.25 l.min-1). This demonstrates that the O2 demand predicted for high power outputs depends on the incremental protocol used.