For the representative subject (subject 1), the V̇O2 response "band" rose to a new "steady state" within 1-2 min of exercise onset, attaining a mean value of 2.50L·min−1; that is, 88% of ∧θL (Fig. 2a, filled circle) and therefore far below V̇O2peak (Fig. 2a, open circle). For the group as a whole, the "steady-state" mean V̇O2 averaged 2.17 ± 0.27 L·min−1, which was not significantly greater than ∧θL (2.05 ± 0.55 L·min−1, P = 0.48). In contrast, Δ[Hb] increased in a steplike fashion with essentially no delay; that is, more rapidly than V̇O2 (Fig. 2a). Also, the "steady-state" Δ[Hb] value was well below the maximal value recorded during the MVC maneuvers (i.e., 100 %). As for VO2, HR also took some minutes to stabilize, and also at a modest submaximal value slightly in excess of 110 bpm (Fig. 2a).
Inspection of the ensemble-averaged single-cycle response (Fig. 2b), revealed a Δ[Hb] oscillation in synchrony with, but lagging, the WR function very slightly. That is, Δ[Hb] rose progressively during the work phase to reach a peak a few seconds into the following recovery phase, and then declined to reach a nadir a few seconds into the following work phase. For the group as a whole, the peaks and nadirs of the Δ[Hb] oscillation averaged 73.2 ± 20.8 and 36.6 ± 25.7%, respectively. In contrast, there was no discernible oscillation of either V̇O2 or HR in synchrony with the WR duty cycle; any such tendency was likely small in relation to the magnitude of the breath-by-breath "noise" (22).
30-s work:60-s recovery protocol
As for the 10-s:20-s protocol, all subjects were able to complete the entire 30min of the test, despite a significantly elevated [L−]c that stabilized after about 10 min to attain a value of 4.9 ± 1.1 mM by the end of the bout (Table 2; Fig. 6). The influence of the intermittent WR forcing could be discerned in the rising phase of the [L−]c response, with successive points (reflecting the timing of blood sampling, at the end of selected work and recovery phases) tending to fluctuate slightly around the best-fit curve (Fig. 6, filled circles).
Systematic oscillatory patterns in synchrony with the intermittent WR protocol were evident in V̇O2, HR, and Δ[Hb], from the onset of the 30-s:60-s intermittent bout (Fig. 3a). The peaks of the V̇O2 oscillation showed a significant increase as the test progressed: for instance, the values for duty cycles 6 and 10 were significantly greater than for duty cycle 2 (Table 3). The peaks then subsequently stabilized at supra-∧θL values, despite [L−]c continuing to increase. The nadirs showed no significant time effect (Table 3), although they consistently exceeded the 20-W baseline value (Fig. 3a). The peaks and nadirs of the Δ[Hb] oscillation both showed a significant time effect (Table 3), although post hoc analysis confined significance to the higher nadir of the final duty cycle relative to that for duty cycle 2 (similar to V̇O2).
The ensemble-averaged single-cycle Δ[Hb] oscillation (Fig. 3b) was essentially similar to that observed for the 10-s:20-s protocol (Fig. 2b), but was significantly larger, yet still submaximal (i.e., 67.1 ± 13.7%; Table 2). The corresponding V̇O2 oscillation was qualitatively similar to that of Δ[Hb], although with a discernibly longer delay (Fig. 3b). The peaks of the V̇O2 oscillation typically exceeded ∧θL (Fig. 3a), averaging 2.89 ± 0.41 L·min−1; this corresponded to 51% of delta.
60-s work:120-s recovery protocol
With this longer duty cycle (i.e., 180 s), there was the first evidence of exercise intolerance (Table 2): i.e., two subjects were unable to complete the full duration (subject 4: 24.8 min; subject 5: 25.0 min) (c.f. (1,33)). [L−]c increased progressively throughout the entire intermittent bout (Fig. 6, open squares) to attain a significantly higher level than for the 30-s:60-s protocol (Table 2).
The associated oscillations in V̇O2, Δ[Hb], and HR were both larger and more obvious than for the 30-s:60-s forcing (Fig. 4; c.f. Fig. 3). The peak-nadir amplitude of the ensemble-averaged Δ[Hb] oscillation was greater than for the 30-s:60-s protocol but still submaximal (83.1 ± 21.2%: Table 2), as was the peak-nadir V̇O2 amplitude (Table 2). The peaks of the V̇O2 oscillation were not significantly different from V̇O2peak (3.71 ± 0.47 L·min−1, P = 0.30) (Table 3). The end-bout HR values were near maximal (Fig. 4a).
Because all subjects were unable to complete this protocol, temporal analysis was confined to duty cycles 1-8 (i.e., these were completed by all six subjects). In contrast to the 30-s:60-s test, the peaks of the V̇O2 oscillation showed a significant progressive increase as the test progressed (Table 3); that is, duty cycles 3-8 were greater than duty cycle 1; duty cycle 8 was greater than duty cycle 2; and duty cycles 6 and 8 were greater than duty cycle 4. This would suggest that, in contrast to the 30-s:60-s test, the V̇O2 peaks did not stabilize after the first few cycles, but continued to increase slightly. However, the V̇O2 nadir values showed no significant change with time (Table 3), although again these values were consistently higher than the 20-W baseline. Similar to the V̇O2 response, the peaks of the Δ[Hb] oscillation showed a significant time effect (Table 3; duty cycle 3 higher than duty cycle 1); this was not the case for the Δ[Hb] nadirs, however (Table 3).
90-s work:180-s recovery protocol
For the longest duty cycle (270 s), none of the subjects was able to complete the full 30 min, with tLIM averaging only 9.1 ± 4.2 min (Table 2). Two subjects (four and six) did not even complete the first work-recovery duty cycle, and a peak [L−]c was therefore not attained. For the remaining four subjects, [L−]c continued to increase throughout to the limit of tolerance to a value not statistically different from that for the 60-s:120-s protocol (Table 2), although at a faster rate (Fig. 6, filled squares). It should be noted, however, that the "loss" of two subjects reduced the statistical power of this comparison.
Despite so few duty cycles for comparison, the patterns for V̇O2, Δ[Hb], and HR were essentially similar to those of the shorter duty-cycle tests; that is, oscillating in synchrony with the WR changes (Fig. 5a).
Owing to the small number of whole duty cycles completed, it was not possible to overlay consecutive duty cycles or to perform repeated-measures ANOVA. Instead, analysis was confined to the second WR duty cycle, because the values were clearly higher than for the first cycle. For the four subjects who were able to complete this second duty cycle, the peaks of the Δ[Hb], V̇O2, and HR oscillations attained maximal values, and the corresponding nadirs were all higher than the 20-W baseline values (Table 3; Fig. 5).
The major finding of this study is that the close intensity-dependent association between the kinetics of blood lactate accumulation and V̇O2 that has been established for constant-load exercise tended also to be evident for the intermittent WR duty-cycle formats explored here. That is, not only did the [L−]c profiles for intermittent exercise of increasing duty-cycle duration bear a close resemblance to those of the moderate, heavy, very heavy, and severe intensity domains for constant-load exercise (Fig. 6) (1,33), but we also found some evidence consistent with the expression of an excess or slow component in the associated V̇O2 response for the 60-s:120-s and 90-s:180-s tests, and to a lesser extent in the 30-s:60-s test.
With very short intermittent duty cycles for which the work and recovery phases are of approximately equal duration (e.g., 1:1 or 1:2), despite WR being higher than WRpeak on the ramp test, there is typically no sustained blood lactate accumulation (1,12,20,33) (Fig. 6). This naturally would not be the case for sustained exercise at such WR.
Interestingly, despite the absence of blood lactate accumulation with such short duty cycles, muscle [L−] has been shown to be elevated (12,33), the femoral venous-arterial [L−] difference to be positive, and muscle glycogen levels to be reduced (33). These observations are suggestive of exercise-related lactate production being matched by tissue clearance (e.g., muscle, liver, heart) (33), as has been reported for sustained moderate constant-load exercise (7,17,34). The transient small increase in blood [L−]c that we (Fig. 6) and others (20) have observed in some instances is reminiscent of the profile reported previously for constant-load exercise at WR below, but approaching, θL (9). This has been attributed to a transient imbalance between lactate production and clearance.
As duty-cycle duration is lengthened (e.g., 30 s:60 s), a modest oscillatory response has been reported in muscle [L−]; this reflects the magnitude of decrease at end-work matching the subsequent increase at end-recovery (33). Similarly, [L−]c also increases to eventually stabilize (1,33), although an oscillatory pattern is less clear (Fig. 6). Interestingly, in the present study [L−]c stabilized at an average value of 4.9 ± 1.1 mM, which suggests that our subjects were exercising at (or close to) an "intermittent" intensity essentially equivalent to the 4- to 5-mM "maximum lactate steady state" criterion and therefore approximated by critical power (18,30,31). That is, stabilization was reached at (or close to) a WR for which the rate of lactate production in muscle is matched by its rate of clearance/utilization (7,17,34).
With the longer duty-cycle durations (60 s:120 s and 90s:180 s), the compromised exercise tolerance was associated with progressive increases in muscle and blood [L−] throughout the duration of the intermittent bout (1,33) (Fig. 6 and Table 2). Muscle [L−] being higher at the end of the work phase than the recovery phase, the positive femoral venous-arterial [L−] difference and appreciable intramuscular glycogen depletion over the course of the intermittent bout (33) are all indicative of high rates of anaerobic glycolysis.
Pulmonary O2 uptake and muscle O2 consumption
An important corollary of the lactate-based intensity classification for constant-load exercise is that the associated general features of the V̇O2 kinetics (e.g., presence or absence of a V̇O2 sc) are predictable from the [L−]a profile (39). There is little question that work:recovery differences in V̇O2 can be discerned, with magnitude increasing as the duty-cycle period is increased. This has been reported for open-circuit spirometry (1,10,20,26,33) and with breath-by-breath monitoring (5,12,26). However, to our knowledge, this is the first investigation that has attempted to analyze V̇O2 response kinetics to supramaximal intermittent exercise, especially in the context of the V̇O2 slow component.
Given the [L−]c profiles we (Fig. 6) and others have described over the duty-cycle range used here, this would suggest that the corresponding breath-by-breath V̇O2 profiles should reflect the associated kinetic complexities that have been described for sustained supra-θL constant-load exercise. However, unequivocal resolution of this issue is complicated by the challenge of conducting any formal kinetic analysis, owing to, for the shortest duty cycles, the compromised signal-to-noise characteristics resulting from small expected V̇O2 response amplitudes and the brevity of the duty cycle and, for the longer duty cycles, the limited number of dynamic "steady-state" V̇O2 response oscillations (i.e., for which the response amplitude has essentially stabilized). In addition, conventional frequency domain (i.e., Fourier) analysis imposes computational limitations because the on- and off-transients are analyzed in a unitary fashion; however, this is only defensible when dynamic system linearity can reasonably be assumed. The demonstration of some manifestation of an additional superimposed slow component within the nadirs of the V̇O2 oscillations as an incomplete recovery tobaseline, and in the peaks of the V̇O2 oscillations at leastin the 60-s:120-s protocol, challenges this assumption.
10-s work:20-s recovery protocol
We were unable to convincingly demonstrate a work:recovery V̇O2 oscillation for the 10-s:20-s duty cycle (Fig. 2). In our experiments, the lack of a statistically discriminable V̇O2 oscillation is most likely a reflection of breath-by-breath noise obscuring a small, but systematic, underlying oscillatory response. We base this assertion on the demonstration that the SD associated with the mean V̇O2 over the final 3 min of each 10-s:20-s data set was substantially greater than that for the steady-state V̇O2 response to a sub-∧θL constant-load test performed by the same subjects (0.27 vs 0.16 L·min−1, P < 0.05; i.e., noise amplitude having been shown to be independent of exercise intensity (22)).
This view is reinforced by the demonstration of a convincing oscillation in Δ[Hb], with Δ[Hb] increasing during the work phase consistent with increasing O2 extraction, and falling in the subsequent recovery phase (Fig. 2). As the Δ[Hb] signal primarily captures arteriolar, capillary, and venular blood in the field of interrogation (25) and is minimally affected by changes in muscle perfusion (13) and skin perfusion (25), it is thus broadly reflective of the local muscle arteriovenous O2 content difference. This is in contrast to the reports of Christmass et al. (10), for which intramuscular oxygenation status was monitored by the less sensitive Runman approach and was expressed in terms of [HbO2], which has been shown to be highly perfusion-sensitive (13,27).
Presuming muscle oxygen consumption (QO2) to oscillate, the outcome for V̇O2 with such a short duty cycle is likely to be complex. On the one hand, a V̇2 oscillation might reasonably be expected to be expressed in V̇O2 (i.e., equivalent to a Φ2 response), with a delay reflecting the muscle-to-lung vascular transit delay (~15-20 s:), which would "push" its rising phase well into the 20-s recovery phase of the duty cycle. Indeed, Margaria et al. (26) and Edwards et al. (12) have reported V̇O2 oscillations for similar (but slightly longer) duty cycles.
However, the extent to which this Φ2 component might actually be offset by any associated cardiodynamic V̇O2 response is difficult to judge; that is, "trimming out" any oscillatory behavior. For example, an increase in pulmonary perfusion at the onset of each 10-s work phase might be expected to induce a concomitant increase in V̇O2 (39) at a time when the Φ2 component has not yet started to increase; the influence of any mechanical compression of the muscle vascular bed consequent to the high required forces for this supramaximal intermittent protocol is unclear, however, but one might expect this to enhance venous return. Certainly, the increase of Δ[HbT] that we saw during the recovery phase relative to the work phase (for those subjects in whom we could discern this signal with confidence) is suggestive of postexercise hyperemia (Fig. 7a). Could this actually induce a flow-mediated V̇O2 increase in recovery? And/or is it possible that the perfusion and arterio-mixed venous O2 content difference responses were each responding to the WR oscillation, but out of phase with each other?
We therefore undertook a simplified Φ2 V̇O2 simulation to gain some insight into the relative influence these two effects might have. This, in turn, allows a comparison of our observed V̇O2 response profiles for all four intermittent protocols (Figs. 2a-5a) with those that might be expected if the response had arisen solely from Φ2 control mechanisms (Fig. 8a). We made the reasonable assumptions, regardless of intensity, of first-order kinetics, a time constant (τ) of 30s, a V̇O2 of 500 mL·min−1 at the 20-W baseline, and a steady-state "gain" (ΔV̇O2/ΔWR) of 10 mL·min−1·W−1 for cycle ergometry (38); and, for simplicity, no perfusion-related delay was assigned. When WR was forced intermittently between 20 and 420 W (as was the case for subject 1), following the first few cycles, asystematic V̇O2 oscillation with a constant peak-to-nadir amplitude of approximately 0.85 L·min−1 ensued (Fig. 8a). This magnitude was only marginally greater than the actual magnitudeof the V̇O2 response "band" seen for subject 1 (Fig. 2), which suggests that any coexisting cardiodynamic influence was in fact not appreciable for this shortest duty cycle.
These several observations, taken with the lack of sustained lactate accumulation (Fig. 6) and the fact that the average V̇O2 response throughout the 30-min intermittent bout was at or just below ∧θL (Fig. 2) coheres well with established views about moderate-intensity constant-load exercise. Furthermore, the findings of Margaria et al. (26), Bogdanis et al. (6), and Trump et al. (35) would suggest that, for short duty-cycle intermittent exercise, aerobic metabolism and PCr hydrolysis contribute to ATP resynthesis to a substantially greater extent than the central roleoriginally proposed for intramuscular oxymyoglobin stores (1).
What we cannot judge is the extent of any O2 stores contribution at this or any longer duty cycle (c.f. (1,33)). That is, whereas NIRS can track changes in the oxygenation status of myoglobin as well as Hb, the influence of myoglobin is assumed to be small because of its lower concentration (25).
30-s work:60-s recovery protocol
With this and other longer duty-cycle formats, any influence of a cardiodynamic component in the overall V̇O2 response might reasonably be expected to be less obvious than for the 10-s:20-s format, given the predictably increasing prominence of the Φ2 component (Fig. 8b-d). This was borne out in the actual 30-s:60-s response profiles for both V̇O2 and Δ[Hb], which, following small increases in the peak-to-nadir amplitude of the oscillations over the first few duty cycles, appeared to attain dynamic "steady states" (Fig. 3). Predictably, also, the V̇O2 oscillation was slightly delayed relative to that of Δ[Hb], consistent with the influence of the limb-lung vascular delay. There was, however, only limited evidence of a marked V̇O2 slow component, despite the peaks of the V̇O2 oscillation (Table 3) being above ∧θL (c.f. "heavy-intensity" constant-load exercise). That is, as predicted from the Φ2 model simulation (Fig. 8b), there was little evidence of an increase in overall V̇O2 system gain over the 30-min duration of the intermittent bout (i.e., excluding the developing phase of the responses over the first few duty cycles), in contrast with heavy-intensity constant-load exercise (38).
The poorly discernible additional V̇O2 slow component for this duty-cycle format, despite the presence of sustained lactate accumulation, may reflect in part the duration of the work phase (i.e., 30 s) being appreciably shorter than the V̇O2 sc latency for constant-load exercise and also its slow kinetics (2,29,38). There is the possibility that the early WR duty cycles exert a "priming" effect that serves to reduce the prominence of the V̇O2 sc in subsequent cycles (possibly through vasodilatation, influences on O2 unloading from hemoglobin, or altered fiber-type recruitment), as has been reported for paired bouts of high-intensity exercise with longer work:recovery durations (i.e., 360 s:360 s) (16,19,24) but, interestingly, not for subsequent repeats of high-intensity exercise (15). Whereas vasodilation may have improved O2 delivery on subsequent duty cycles, our NIRS technique does not have the capability to resolve this. Interestingly, the lack of any sustained changes in the Δ[Hb] oscillation (Table 3) might suggest a limited effect of an acidosis-induced Bohr shift of the HbO2 dissociation curve.
Such "priming" exercise has been reported to increase the amplitude of the fundamental component for supra-θL exercise (reviewed in (19,38,39)). However, the amplitude of the ensemble-averaged single-cycle V̇O2 oscillation (Fig. 3b) was actually considerably smaller, not larger, than that predicted by our simple first-order Φ2 model; for instance, for subject 1, 1.28 versus 2.35 L·min−1, respectively (Fig. 3b vs Fig. 8b). The causes of this discrepancy are unclear. Slowing of the Φ2 V̇O2 kinetics might be expected to have such an effect; however, the Φ2 on- and off-transient time constants for heavy-intensity constant-load exercise are broadly similar to those for moderate cycle-ergometer exercise (2,25). We also return to the possibility of complex cardiodynamic kinetics, as discussed previously. The work:recovery Δ[HbT] fluctuationwas more marked than seen for 10-s:20-s exercise (Fig. 7b), at a time when there was a greater likelihood of the Φ2 V̇O2 increase starting to develop before the end of the 30-s work phase. As importantly, could V̇O2 in recovery represent offsetting effects from the subsequent Φ2 decrease and a cardiodynamic component related to postexercise hyperemia (see 10-s:20-s above)? Certainly, although the nadir values for V̇O2 did not progressively increase, they were higher than the 20-W baseline; however, given the short duration of the recovery (60 s), we could not be certain to what asymptotic value V̇O2 was projecting towards.
60-s work:120-s and 90-s:180-s recovery protocols
The maximal or near-maximal V̇O2 and Δ[Hb] responses observed (Table 3, Figs. 4 and 5) and progressively increasing [L−]c profiles for both the 60-s:120-s and 90-s:180-s duty cycles (Fig. 6) are characteristic of constant-load WR that exceed critical power. In contrast to the 30-s:60-s test, however, there was clear evidence of a developing V̇O2 slow component in the 60-s:120-s test, with the peaks of the V̇O2 oscillation not actually stabilizing until the fifth cycle of the 30-min duration (although one might argue that this was due to the peaks becoming truncated by the attainment of V̇O2peak). The mechanisms responsible for the V̇O2 sc remain to be fully elucidated, but have variously been proposed to include a lactic acidosis-induced Bohr shift of the HbO2 dissociation curve, increased work of the cardiac and respiratory muscles, and progressive recruitment of fast-twitch (Type II) muscle fibers (14,19,39). Indeed, the progressive depletion of glycogen in this demanding test (33) might result in the recruitment of proportionally greater numbers of fast-twitch muscle fibers, which have been shown to be less aerobically efficient and have slower kinetics than slow-twitch fibers (3,11). Furthermore, heart rate can clearly be seen to progressively increase throughout the test period (Fig. 4a). As discussed previously, interpretation of the Δ[HbO2] signal is complex, but interestingly the Δ[Hb] peaks did not continue to increase beyond the third cycle (Table 3), questioning the importance of the facilitation of O2 unloading caused by the presumed decline in pH.
The V̇O2 slow component that has been discerned (without a delay) during very heavy and severe sustained constant-load exercise recovery transients (29) also appeared to be manifest in the present study (Figs. 4 and 5). Although the observed V̇O2 response profiles were qualitatively consistent to those resulting from our simplified Φ2 model (Fig. 8c and d), the observed V̇O2 nadir values were consistently higher than the 20-W baseline, suggesting an additional slow or excess V̇O2 component.
Presumably as a result of this additional recovery component in V̇O2, the peak-to-nadir amplitudes of the ensemble-averaged V̇O2 oscillations for the two longest duty cycles were far less for the representative subject than for the Φ2 model prediction. Thus, for subject 1, the actual V̇O2 oscillation amplitude for the 60-s:120-s protocol, was 2.50 L·min−1 (Fig. 4b), whereas the model amplitude was 3.41 L·min−1 (Fig. 8c). Similarly, for the 90-s:180-s protocol, the actual V̇O2 amplitude was 2.72 L·min−1 (Fig. 5b), whereas the model amplitude was 3.84 L·min−1 (Fig. 8d). We suggest that this is unlikely to reflect the assumptions of an invariant gain in the Φ2 model, however. The fundamental gain for suprathreshold cycling is not appreciably different than that for moderate exercise (2,19,29), and the weight of available evidence suggests that it is also similar above critical power and below ((2,29); c.f. (19)).
One of the major questions arising from this research is whether our measures of V̇O2 kinetics weresensitive enough to discriminate the existence of a slow component. Our results are certainly suggestive of anexcess component during the recovery periods, and theincreasing peaks of the V̇O2 oscillations in the 60-s:120-s test would imply that our tests and analysis were sufficiently sensitive. However, the relatively short durations of the work and recovery periods for the 30-s:60-s tests may have constrained the time within which a V̇O2 slow component could become manifest. We also acknowledge the relatively small subject numbers used predispose towards a higher probability of a Type II statistical error. In addition, our ability to better interpret the V̇O2 oscillatory profiles would have been enhanced with some knowledge of venous return or pulmonary blood flow dynamics. Finally, P31 nuclear magnetic resonance spectroscopical monitoring of intramuscular [PCr] might have allowed cardiodynamic contributions to be discriminated from an underlying Φ2 component (c.f. (32)).
Also, we were unable to take account of functional heterogeneities in muscle energetics (19,38,39). That is, our analyses are based on the probably oversimplistic assumption that the spectrum of mechanisms contributing to muscle ATP resynthesis for any particular duty cycle are distributed relatively homogenously within the involved musculature. However, the spatial and temporal features of muscle fiber-type recruitment during intermittent exercise are unknown.
These intermittent-exercise experiments thus demonstrate interesting parallels with respect to the conventional kinetics-based characterizations of exercise intensity (29,38). Thus, for short duty cycles eliciting V̇O2 no greater than θL, there was no evidence of sustained lactate accumulation or of additional excess V̇O2. A departure from constant-load exercise was found, however, in the lack of a demonstrable V̇O2 oscillation in synchrony with the work rate: our Φ2 model predicted this, and we also found evidence of a fluctuating intramuscular V̇O2 extraction. Whereas suboptimal signal:noise characteristics might have been contributory, we speculate that a significant complication was introduced by the kinetics of "cardiodynamic" gas exchange, which, under these particular non-steady-state conditions, served to offset expression of the primary V̇O2 response. The association of blood lactate accumulation with manifestations of an additional excess V̇O2 component for the longer duty cycles, and at V̇O2s that were clearly supra-θL, similarly coheres with what is known for constant-load exercise. And although we could not formally demonstrate that this additional component had slow kinetics (as is the case for supra-θL constant-load exercise (2,23,29), it is tempting to speculate that blood lactate accumulation (or whatever it might be serving as a proxy for) is necessary for its induction. In conclusion, in contrast to classical markers of exercise intensity based on simple scalar constructs such as work rate, heart rate, V̇O2, and rating of perceived exertion, the association of V̇O2 kinetics and blood lactate accumulation profiles would seem to provide a functionally rigorous classification of intermittent exercise intensity, as in the case for constant-load exercise.
The authors would like to thank Professor Brian J. Whipp for his constructive contributions to the study, and also the subjects for their time and commitment.
1. Åstrand, I., P. -O. Åstrand, E. H. Christensen, and R. Hedman. Myohaemoglobin as an oxygen-store in man. Acta Physiol. Scand.
2. Barstow, T. J., and P. A. Molé. Linear and non-linear characteristics of oxygen uptake kinetics during heavy exercise. J. Appl. Physiol.
3. Barstow, T. J., A. M. Jones, P. H. Nguyen, and R. Casaburi. Influence of muscle fiber type and pedal frequency on oxygen uptake kinetics of heavy exercise. J. Appl. Physiol.
4. Beaver, W. L., K. Wasserman, and B. J. Whipp. On-line computer analysis and breath-by-breath graphical display of exercise function tests. J. Appl. Physiol.
5. Billat, V. L., J. Slawinski, V. Bocquet, et al. Intermittent runs atthe velocity associated with maximal oxygen uptake enables subjects to remain at maximal oxygen uptake for a longer timethan intense but submaximal runs. Eur. J. Appl. Physiol.
6. Bogdanis, G. C., M. E. Nevill, L. H. Boobis, and H. K. Lakomy. Contribution of phosphocreatine and aerobic metabolism to energy supply during repeated sprint exercise. J. Appl. Physiol.
7. Brooks, G. A. Lactate shuttles in Nature. Biochem. Soc. Trans.
8. Cathcart A. J., N. Herrold, A. P. Turner, J. Wilson, and S. A. Ward. Absence of long-term modulation of ventilation by dead-space loading during moderate exercise in humans. Eur. J. Appl. Physiol.
9. Cerretelli P., and P. E. Di Prampero. Gas exchange in exercise. In: Handbook of Physiology. The Respiratory System. Gas Exchange
. Bethesda, MD: Am. Physiol. Soc. Sect. 3, Vol. IV, Chapter 16, 1987, pp. 297-340.
10. Christmass, M. A., B. Dawson, and P. G. Arthur. Effect of work and recovery duration on skeletal muscle oxygenation and fuel use during sustained intermittent exercise. Eur. J. Appl. Physiol. Occup. Physiol.
11. Crow, M. T., and M. J. Kushmerick. Chemical energetics of slow- and fast-twitch muscles of the mouse. J. Gen. Physiol.
12. Edwards, R. H. T., L. -G. Ekelund, R. C. Harris, et al. Cardiorespiratory and metabolic costs of continuous and intermittent exercise in man. J. Physiol.
13. Ferrari, M., T. Binzoni, and V. Quaresima. Oxidative metabolism in muscle. Philos. Trans. R. Soc. Lond. B. Biol. Sci.
352: 677-683, 1997.
14. Gaesser, G. A., and D. C. Poole. The slow component
of oxygen uptake kinetics in humans. Exerc. Sport Sci. Rev.
15. Gausche, M. A., T. Harmon, N. Lamarra, and B. J. Whipp. Pulmonary O2
uptake kinetics in humans are speeded by a bout of prior exercise above, but not below, the lactate threshold. J. Physiol.
16. Gerbino, A., S. A. Ward, and B. J. Whipp. Effects of prior exercise on pulmonary gas-exchange kinetics during high-intensity exercise in humans. J. Appl. Physiol.
17. Gladden, L. B. Lactate metabolism: a new paradigm for the third millenium. J. Physiol.
18. Hill, D. V. The critical power concept: a review. Sports Med.
19. Jones, A. M., J. S. Pringle, and H. Carter. Influence of muscle fibre type and motor unit recruitment on VO2
kinetics. In: Oxygen Uptake Kinetics in Health and Disease
, A. M. Jones and D. C. Poole (Eds.). London, UK: Routledge, 2005, pp. 261-293.
20. Keul, J., E. Doll, D. Keppler, and H. Reindell. Intervall-training und anaerobe Energiebereitstellung. Sportarzt. Sportmed.
12: 493-496, 1967.
21. Kowalchuk, J. M., H. B. Rossiter, S. A. Ward, and B. J. Whipp. The effect of resistive breathing on leg muscle oxygenation using near-infrared spectroscopy
during exercise in men. Exp. Physiol.
22. Lamarra, N., B. J. Whipp, S. A. Ward, and K. Wasserman. Effect of inter breath fluctuations on characterizing exercise gas exchange kinetics. J. Appl. Physiol.
23. Linnarsson, D. Dynamics of pulmonary gas exchange and heartrate changes at start and end of exercise. Acta Phys. Scand. Suppl.
24. Macdonald, M., P. K. Pedersen, and R. L. Hughson. Acceleration of V̇O2
kinetics in heavy submaximal exercise by hyperoxia and prior high-intensity exercise. J. Appl. Physiol.
25. Mancini, D. M., L. Bolinger, H. Li, K. Kendrick, B. Chance, and J. R. Wilson. Validation of near-infrared spectroscopy
in humans. J. Appl. Physiol.
26. Margaria, R., R. D. Oliva, P. E. Di Prampero, and P. Cerretelli. Energy utilization in intermittent exercise of supramaximal intensity. J. Appl. Physiol.
27. McCully, K. K., and T. Hamaoka. Near-infrared spectroscopy
: what can it tell us about oxygen saturation in skeletal muscle? Exerc. Sport Sci. Rev.
28. Morton, R. H., and V. Billat. The critical power model for intermittent exercise. Eur. J. Appl. Physiol.
29. Özyener, F., H. B. Rossiter, S. A. Ward, and B. J. Whipp. Influence of exercise intensity
on the on-and off-transient kinetics of pulmonary oxygen uptake in humans. J. Physiol.
30. Poole, D. C., S. A. Ward, G. W. Gardner, and B. J. Whipp. Metabolic and respiratory profile of the upper limit for prolonged exercise in man. Ergonomics
31. Pringle, J. S. M., and A. M. Jones. Maximal lactate steady state, critical power and EMG during cycling. Eur. J. Appl. Physiol.
32. Rossiter, H. B., S. A. Ward, J. M. Kowalchuk, F. A. Howe, J. R. Griffiths, and B. J. Whipp. Dynamic asymmetry of phosphocreatine concentration and O2
uptake between the on- and off-transients of moderate- and high-intensity exercise in humans. J. Physiol.
33. Saltin, B., and B. Essén. Muscle glycogen, lactate, ATP, and CP in intermittent exercise. In: Muscle Metabolism during Exercise, Advances in Experimental Medicine and Biology, vol 11
, B. Pernow and B. Saltin (Eds.). New York: Plenum Press, 1971, pp. 419-424.
34. Spriet, L. L., R. A. Howlett, and G. J. Heigenhauser. An enzymatic approach to lactate production in human skeletal muscle during exercise. Med. Sci. Sports Exerc.
35. Trump, M. E., G. J. Heigenhauser, C. T. Putman, and L.L.Spriet. Importance of muscle phosphocreatine during intermittent maximal cycling. J. Appl. Physiol.
36. Wasserman, K., and R. Casaburi. Acid-base regulation during exercise in humans. In: Pulmonary Physiology and Pathophysiology of Exercise
, B. J. Whipp and K. Wasserman (Eds.). New York: Dekker, 1991, pp. 405-448.
37. Wasserman, K., J. E. Hansen, D. Y. Sue, R. Casaburi, and B. J. Whipp. Principles of Exercise Testing and Interpretation
, Philadelphia: Lippincott, Williams & Wilkins, 1999, p. 19.
38. Whipp, B. J., and H. B. Rossiter. The kinetics of oxygen uptake: physiological inferences from the parameters. In: Oxygen Uptake Kinetics in Health and Disease
, A. M. Jones and D. C. Poole (Eds.). London, U.K.: Routledge, 2005, pp. 64-94.
39. Whipp, B. J., H. B. Rossiter, and S. A. Ward. Exertional oxygen uptake kinetics: a stamen of stamina? Biochem. Soc. Trans.
30: 237-247, 2002.
40. Whipp, B. J., S. A. Ward, and K. Wasserman. Respiratory markers of the anaerobic threshold. Adv. Cardiol.
Keywords:©2006The American College of Sports Medicine
EXERCISE INTENSITY; BLOOD LACTATE; V̇O2 KINETICS; NEAR-INFRARED SPECTROSCOPY; SLOW COMPONENT