INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT HAS BEEN DEMONSTRATED THAT pulmonary O₂ uptake (O₂) kinetics at the onset of supra lactate threshold (LT) leg cycle ergometer exercise in humans can be speeded by a prior warm-up bout of heavy exercise using the same muscles (5, 12, 13, 23). For example, O₂ increases more rapidly in the second bout of two square-wave transitions to work rates of approximately halfway between LT and maximal O₂ (O_2 max) (i.e., ~50%). These investigators proposed that this O₂ acceleration might be due to improved muscle perfusion during the second bout exercise transient consequent to the vasodilating effects of residual lactic acidemia, which is still present at the onset of the second bout (13). A concomitant factor, which could also accelerate these O₂ kinetics, would be a rightward shift of the oxy-hemoglobin (Hb) dissociation curve resulting from the lower pH and increased CO₂ (i.e., Bohr effect) (33).

Recently, Bohnert et al. (4) investigated whether intense warm-up exercise performed at a remote site (the arms) but producing a comparable degree of systemic lactic acidosis could induce the same effect on O₂ kinetics during subsequent heavy leg exercise. The speeding of O₂ kinetics was found to be less prominent after the arm cranking warm-up compared with that after the leg warm-up exercise, although both the arm and leg warm-ups resulted in similar degrees of residual lactic acidosis. On the other hand, a study using the one-legged exercise model indicated that the O₂ response was accelerated only in the case of intense leg exercise preceded by a warm-up using the same leg (37). In an attempt to clarify the differences between this and the previous study (4, 37), we examined whether there is evidence of speeding of O₂ kinetics for high-intensity cycling when preceded by an initial exercise bout performed by a remote muscle group using more rigorous kinetics analysis. If systemic lactic acidosis were to cause vasodilation in all skeletal muscles, the increased total vascular conductance would result in a fall in blood pressure. This is prevented by the sympathetic nervous system, which induces vasoconstriction in inactive muscles to compensate for the vasodilation in working muscles during exercise (e.g., Ref. 28). We therefore hypothesized that warm-up exercise that resulted in vasodilation in a remote muscle group could only have a minor impact on O₂ kinetics during the main exercise.

Near-infrared spectroscopy (NIRS) is a noninvasive method that has been used to evaluate changes in muscle perfusion status; that is, the combined concentration changes of oxy- and deoxy-Hb (i.e., the relative change in total Hb concentration) may be taken to reflect muscle blood volume changes (24, 38). We therefore used NIRS of the vastus lateralis muscle to quantitate the hyperemia induced by a first bout of either intense leg or arm exercise immediately before the second bout of intense leg exercise. In the study of Bohnert et al. (4), each protocol was performed only once, and the kinetic components were evaluated solely by using parameters relating to the increment of O₂ occurring between minutes 3-6 [i.e., O_2(6-3)] and the "partial" O₂ deficit (O₂-def) (4). We, on the other hand, repeated each protocol three times, which allowed us to characterize O₂ kinetics more precisely.

The purpose in this study was, therefore, to examine whether local hyperemia induced by lactic acidosis (i.e., local muscle blood flow response) at the onset of exercise contributed to the speeding of O₂, as has been previously reported (13) for supra-LT cycling, when warm-up exercise was performed by using the same muscle group or muscles at a remote site.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Six healthy male subjects volunteered to participate in this study. Their mean (±SD) age, height, and body weight were 28.7 ± 7.3 yr, 174.5 ± 5.1 cm, and 71.0 ± 7.7 kg, respectively. After the protocols and possible risks associated with participation in the study were explained, subjects signed an informed consent form, which had been approved by the institutional ethics committee.

Subjects performed an incremental exercise test on an electrically braked cycle ergometer (232c-XL, Combi) to estimate their LT, O_2 max, and difference between LT and O_2 max (). Subjects then completed a series of six high-intensity ergometer square-wave exercises. Each series of exercises was a random combination of the following two protocols, each of which was performed three times. Both protocols consisted of a 4-min warm-up at 0 W followed by two consecutive 6-min exercise bouts separated by a 6-min unloaded exercise bout (0 W). In one protocol (L1-ex to L2-ex), both bouts consisted of supra-LT leg cycling. In the other protocol (A1-ex to A2-ex), bout 1 was heavy arm cranking with the use of an arm-cranking ergometer (Monark) and bout 2 was supra-LT leg cycling. For A1-ex to A2-ex, the arm ergometer was fixed on the armrest of the leg cycle ergometer to prevent a subject from having to switch position between arm cranking and leg cycling. The center of the arm crankshaft was adjusted to shoulder level and an appropriate distance from a naturally upright position for each subject. The work intensity corresponding to supra-LT leg cycling was selected as LT+50% (range: 190-240 W). The work rate of arm cranking was adjusted to ~1 W/kg body wt (range: 60-80 W). In a preliminary trial, this particular range was verified to induce at least the same blood lactate levels as heavy cycling (i.e., L1-ex).

During exercise, ventilation and gas exchange were measured with a computerized on-line breath-by-breath system. Inspiratory and expiratory gas volumes were measured with the use of a hot-wire flowmeter (RM-300, Minato Medical Sciences). Respiratory concentrations of O₂, CO₂, and N₂ were analyzed with a mass spectrometer (WSMR-1400, Westron). This system was calibrated by using a 2-liter syringe and two precision-analyzed gas mixtures before each exercise test. The second-by-second time course was calculated for each variable by interpolation of the breath-by-breath data. Data were stored on disk for further statistical analysis. Venous blood was sampled from a subject's heated finger immediately before and after each exercise bout for determination of blood lactate concentration (HEK-30L, Toyobo). The blood lactate analyzer was calibrated with a standard solution of 50 mg/dl lactate.

The NIRS device (NIR-4s, Hamamatsu Photonics) used for this study consisted of a probe and a computerized control system to measure changes in blood Hb concentration as an index of hyperemia in working muscles, as previously described (38). With the use of four wavelengths, Hb-related chromophore concentration changes (in µM/liter) were calculated by using a modified Lambert-Beer law (e.g., Ref. 8). The light source and detector were mounted in a plastic probe such that the path length between them was 45 mm. The probe was placed over the center of the vastus lateralis muscle of the right thigh. This NIRS device measures the magnitude of chromophore concentration change from an arbitrary baseline (assigned a value of 0) at the minute 1 of the 4-min warm-up at 0 W leg cycling for both protocols. That is, all subsequent changes in signal are expressed relative to this initial baseline.

The O₂ time course during cycling (i.e., L1-ex, L2-ex, and A2-ex) was quantified by the following indexes. We used a cluster of three measures to provide a broad description of the O₂ kinetics and compare them to results obtained in previous studies (4, 13). The first measure was the effective speed of the O₂ response, which was evaluated by the time constant ("effective" ) when a monoexponential model was fitted to the data in which the preceding 1 min was included as baseline and the initial 20 s after the onset of exercise (phase 1) was eliminated (33). O₂ at time t after exercise onset is expressed as

(1)

where BL is the baseline value of 0 W cycling, and G, Td, and  are the gain, time delay, and time constant parameters, respectively, of the exponential increases of O₂ after the onset of exercise. The second measure was the increment in O₂ between minutes 3 and 6 of the 6-min exercise bout [O_2(6-3)] used to estimate the magnitude of the O₂ slow component (27). The third measure was the partial O₂-def, which was derived as the integral of the difference between the O₂ attained at the end of exercise bout and the O₂ at t (4, 13).

There is a bias in the O₂ residual in the monoexponential model due to a slow and delayed O₂ during the supra-LT heavy exercise (2, 6, 10). Therefore, the same O₂ data used for the monoexponential model were also fitted with a double-exponential model that incorporated two amplitudes (Gp and Gs), two time constants (p and s), and two time delays (Tdp and Tds)

(2)

The fitting was performed by using a nonlinear regression technique (Marquardt-Levenberg algorithm in sigma plot, Jandel Scientific). Concerns regarding the validity of using the extrapolated asymptotic value for the gain of the slow component (Gs) for comparisons caused us to use the value of the slow exponential function at the end of exercise, defined as A'₂ according to previous studies (3, 21). As the overall kinetic parameter derived from the monoexponential curve fitting, the mean response time ( + Td) was also calculated. To adjust for differences in the absolute work rate for supra-LT cycling among subjects, the gain parameter, G or Gp of primary component by either mono- or double-exponential estimation was divided by the work rate and expressed as G/W or Gp/W, respectively.

Values were expressed as means ± SD excepting any special notions. Differences in the parameters among the three heavy leg exercise bouts (i.e., L1-ex, L2-ex, and A2-ex) were evaluated by ANOVA with repeated measures (SPSS for Windows, SPSS). When a significant difference was detected, this was further examined by post hoc Tukey's test. Significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Blood lactate levels immediately before supra-LT leg exercises were 1.7 ± 0.3, 5.6 ± 0.9, and 6.7 ± 1.4 meq/l in L1-ex, L2-ex, and A2-ex, respectively (Table 1). Blood lactate concentrations in L2-ex and A2-ex were significantly higher than those in L1-ex. Therefore, the residual lactic acidosis induced by the heavy arm cranking was not significantly different from that induced by bout 1 of supra-LT leg exercise (L2-ex: 5.6 ± 0.9 vs. A2-ex: 6.7 ± 1.4 meq/l).

An example of the O₂ kinetics for the two protocols in a representative subject is shown in Fig. 1. Bout 1 of supra-LT leg cycling exercise induced an acceleration of O₂ kinetics for the following leg cycling exercise (as shown in Fig. 1A, right). For the group as a whole, the effective O₂  obtained by using the a monoexponential model showed a significant difference between L1-ex (52.3 ± 8.2 s) and L2-ex (36.8 ± 4.3 s). O_2(6-3) and O₂-def for L2-ex were significantly smaller than those for L1-ex [O_2(6-3): 57 ± 22 vs. 129 ± 51 ml/min; O₂-def: 1,572 ± 268 vs. 2,262 ± 377 ml]. For A2-ex, which was the leg supra-LT exercise after the arm cranking heavy exercise, the O₂ response was closer to that for L1-ex than to that for L2-ex (Fig. 1, A and B, right). On the other hand, by using monoexponential fitting in A2-ex (51.7 ± 9.7 s),  was not significantly different from that in L1-ex (Table 1). O_2(6-3) and O₂-def for A2-ex (84 ± 48 ml/min and 2,096 ± 385 ml, respectively), however, showed an intermediate value between those in L1-ex and L2-ex (Table 1). The gains (G/W) were essentially identical among L1-ex, L2-ex, and A2-ex (Table 1).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   A: response of O2 uptake (O2) to a 2-bout exercise protocol in a representative subject (left) in which both bouts used leg exercise (L1-ex to L2-ex). Response from bout 2 () of the L1-ex to L2-ex protocol is superimposed on that of bout 1 (bold line) (right). B: response of O2 to a 2-bout exercise protocol in the same subject (left) in which the bout 1 used heavy arm cranking and bout 2 used leg exercise (A1-ex to A2-ex). Response from bout 2 () of A1-ex to A2-ex protocol is superimposed on that of bout 1 (bold line) of the L1-ex to L2-ex protocols (right).

Because of the wealth of experimental evidence for the O₂ slow component in the supra-LT exercise domain (11, 30), we further fitted O₂ data to a double-exponential model to describe more precise kinetics (Fig. 2). Figure 2 shows the mono- and double-exponential curve-fitting procedures in L1-ex to L2-ex protocol in a typical subject. It is clear from the residuals at the foot of each panel that the double-exponential model provided a superior fit compared with the mono-exponential model, which appeared to generate bias in the residuals. Indeed, the monoexponential model caused serious bias in O₂ values throughout the exercise bouts, especially in L1-ex. The p of the primary O₂ component in the double-exponential model was similar in all three supra-LT leg exercises (L1-ex: 22.4 ± 3.8 s; L2-ex: 23.5 ± 4.1 s; and A2-ex: 27.7 ± 8.1 s). However, the gain of the primary O₂ component in L1-ex (8.6 ± 0.7 ml · min¹ · W¹) was significantly smaller than that in L2-ex (9.7 ± 0.7 ml · min¹ · W¹) and in A2-ex (9.9 ± 0.8 ml · min¹ · W¹). The absolute magnitude of the slow O₂ component in the double-exponential model, A'₂, was significantly smaller in L2-ex than in L1-ex, as seen in the conventional measure, O_2(6-3) (Table 1).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Example of the O2 response to L1-ex to L2-ex in a representative subject. Solid and bold curves indicate the estimated model derived from the application of mono- and double-exponential fitting, respectively, to the second-by-second data (point) from 20 s after the onset of exercise. The residuals were expressed as 10-s averaging values (bottom) by either mono- (solid line) or double-exponential (bold line) curve fit, which indicates that the double-exponential model provides a qualitatively superior fit without any bias compared with the monoexponential fit.

The response of total Hb in the right vastus lateralis, as measured by NIRS, during the two different protocols is shown in Fig. 3. The change of blood Hb concentration relative to the initial value during warm-up 0-W cycling has been assumed to be directly proportional to the change in blood volume in the working muscle (24, 38). Therefore, we reported the mean relative change of Hb concentration over the minute preceding the onset of supra-LT leg exercises as being reflective of baseline hyperemia (Table 1). The mean of the relative change of Hb concentration in L2-ex (386 ± 180 µM/l) was significantly higher than that in L1-ex (29 ± 19 µM/l) and A2-ex (89 ± 79 µM/l). When the relative concentration changes observed before bout 1 of exercise were used as baseline values, the increment () of total Hb concentration change in the vastus lateralis before the bout 2 of supra-LT leg exercise was significantly greater in L2-ex (367 ± 174 µM/l) compared with that in A2-ex (48 ± 73 µM/l).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Example in a representative subject (A) and group means ± SE (B) of the relative change of blood Hb concentration in the vastus lateralis muscle every 10 s, as measured by near-infrared spectroscopy during the two protocols. , L1-ex to L2-ex; , A1-ex to A2-ex.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The differences in "effective" , as determined by the monoexponential model, as well as O_2(6-3) between L1-ex and L2-ex, were similar to that reported in previous studies (4, 5, 12, 13, 23). However, when the preceding exercise was performed by a different muscle group (i.e., arm cranking), the O₂ response during the following supra-LT leg exercise was very close to the first bout of leg exercise (i.e., L1-ex). This finding suggests that more rapid O₂ dynamics during the second supra-LT leg exercise was not simply due to systemic lactic acidosis, as lactic acid levels were similar after either the arm cranking exercise or the first bout of leg exercise.

It is recognized that the O₂ slow component in the supra-LT exercise domain is evident and induces a certain bias when monoexponential fitting is applied (2, 6, 10). Therefore, we fitted the O₂ data to a double-exponential model to describe the kinetics more precisely (Fig. 2). The estimated kinetic parameters from a double-exponential fitting showed that the main difference of the overall O₂ kinetics between L1-ex and L2-ex was attributed to the difference of the slow component but not to the primary component. Indeed, the slow component [i.e., A'₂ and/or O_2(6-3)] was significantly reduced in L2-ex compared with that in L1-ex. In contrast, the  in the primary component (p) was essentially the same, whereas its gain (Gp/W) increased in L2-ex. The double-exponential model can explain the difference of the effective  derived from the overall O₂ kinetics by the monoexponential model. Our results using both mono- and double-exponential models were identical to those of a recent study that characterized the O₂ kinetics to double bouts of heavy leg exercise (5). However, it should be emphasized that O₂ during the first half of 6 min in L2-ex was consistently higher than that in L1-ex, although precise characterization of the "facilitated" O₂ phenomenon has yet to be achieved. The higher O₂ in L2-ex reduced the calculated partial O₂-def (Table 1).

Gerbino et al. (13) hypothesized that a facilitated O₂ response during L2-ex was caused by lactic acidosis because it was not observed when exercise was below LT. It was speculated that the lactic acidemia-induced vasodilation induced an increase in muscle blood flow as well as faster O₂ extraction during a preliminary trial using NIRS (31). The present study evaluated their hypothesis by using NIRS in an essentially similar fashion to measure the relative hyperemic status in the exercising muscles (i.e., vastus lateralis muscle). Although Hb measurement by NIRS might be influenced by unquantified movements of fluid in and out of the vascular space, the consensus is that changes in Hb concentration provide a reliable indicator of changes in blood volume within the interrogation site of the working muscles and also reflect blood flow status (24). Therefore, we reported the mean relative change of total Hb concentration over the minute preceding the onset of supra-LT leg exercise as reflective of the baseline hyperemia (Table 1 and Fig. 3). We found that the mean of the relative change of Hb concentrations in the vastus lateralis muscle just before bout 2 of supra-LT leg in the L1-ex to L2-ex protocols was significantly elevated from the initial baseline compared with that observed before leg exercise in the A1-ex to A2-ex protocols. This result suggests that hyperemia induced by a warm-up exercise using the same muscle group as in the main exercise bout may contribute to the facilitation of O₂ kinetics, with the systemic lactic acidosis having a less important role to play.

Bohnert et al. (4) found a less prominent but significant positive effect of prior arm exercise for O₂ kinetics during the following supra-LT leg cycling, whereas we found no such significant effect. The slow-component parameters [e.g., A'₂ and O_2(6-3)] and O₂-def in A2-ex in this study showed intermediate but not significantly different values from those found in L1-ex and L2-ex, even though it was apparent that the time-serial change of O₂ during A2-ex was very close to that in L1-ex but not to that in L2-ex (Fig. 1). Although the reasons for this discrepancy are not clear, they may be related to differences in the amount of physiological stress induced by arm exercise in both studies. We adjusted the work rate for arm cranking to individual body weights, whereas Bohnert et al. (4) deducted the rates directly from an arm-cranking ramp-exercise test. Compared with the group mean value of "arterialized" venous blood lactate (6.0 meq/l) induced by arm cranking and measured at the onset of leg exercise in the Bohnert et al. study, the corresponding lactate concentrations in this study, which were derived from "venous" blood, were an average of 6.7 meq/l. This indicates that the arm-cranking exercise in this study may be more stressful physiologically compared with that in the Bohnert et al. study. Such factors may explain the different results between the two studies.

The physiological mechanism(s) explaining the facilitation of O₂ kinetics during repeated bouts of identical supra-LT exercise can be linked to the etiology of the O₂ slow component during supra-LT exercise because the main difference between bouts 1 and 2 was attributable to the slow component (Table 1 and Figs. 1 and 2). The cause(s) of the O₂ slow component is currently a topic of considerable interest (e.g., Refs. 11, 33). Poole et al. (26) demonstrated that ~80 to 90% of the pulmonary O₂ slow component could be accounted for by the associated increase in leg O₂. Consequently, most authors ascribe the O₂ slow component to factors related to the exercising limb rather than to elsewhere in the body. Therefore, current proposed mechanisms to explain the facilitated O₂ in the second bout should be ascribed to the improved O₂ transport to the exercising musculature, greater O₂ utilization in exercising muscle(s), and/or their interaction.

With respect to O₂-transport limitation, the augmenting blood flow in electrically stimulated isolated canine gastrocnemius muscle has been reported not to affect the O₂ kinetics at the onset of exercise of ~60 to 70% O_2 max, which is presumably corresponding to heavy exercise domain (14). This was true even when the exercise muscle received the blood flow equal to that required during the steady state from the onset of exercise. Furthermore, they demonstrated that increased O₂ driving pressure from capillary to muscle cell by using RSR-13 (the drug to induce the rightward shift of the oxy-Hb dissociation curve) had no effect on O₂ kinetics (15). However, when electrical stimulation was applied to induce peak exercise in the same experimental model, O₂ kinetics were facilitated by faster O₂ delivery (16). This indicates that convective O₂ delivery to muscle, together with the inertia of oxidative metabolism, seems to contribute to determining O₂ kinetics during intense exercise. The same group also showed a consistently faster blood flow on transient compared with that for muscle O₂ at the onset of cycling exercise (sub-LT) in humans (17). Furthermore, Williamson et al. (35) showed no effect of lower body positive pressure (i.e., impaired muscle perfusion) on O₂ kinetics at the onset of heavy leg cycling exercise. These studies indicate that O₂ delivery to muscular sites plays no major role; rather, intrinsic inertia of oxidative metabolism in muscle cell may be the primary locus regulating O₂ kinetics at the onset of, at least, heavy exercise.

In contrast, the recent study by MacDonald et al. (22) provided evidence that improved O₂ delivery could accelerate O₂ kinetics in a second bout of handgrip exercise after an identical exercise. Muscle O₂ in the forearm during the first 30 s in bout 2 of exercise was elevated compared with that in bout 1 with a concomitant increase in blood flow to the forearm. However, no kinetic parameters were measured because of the difficulty of frequent blood sampling. In addition, the same group demonstrated that enhanced O₂ delivery by hyperoxic breathing has a significant role to accelerate O₂ kinetics during heavy leg cycling exercise (23). However, the reason(s) for this contradiction regarding the role of circulation to exercising muscle in determining O₂ kinetics remains unclear. There are, of course, several possible explanations based on differences among studies. For instance, chosen species, whether a study was in vivo or not, exercise mode (voluntary contraction vs. electrically stimulation), muscle mass exercising, blood sampling and blood flow measurement site (vein vs. artery), and/or method for measuring blood flow (Doppler ultrasound vs. thermodilution) varied from one study to the next. It appears there is not yet sufficient evidence to establish whether O₂ delivery or intracellular processes are most important in controlling the O₂ time course at this exercise intensity.

With respect to O₂ utilization within the muscle and its causing the O₂ slow component, it would seem that O₂ utilization is dependent on the degree of the recruitment of less efficient type II muscle fibers during heavy exercise (11). The amount of slow component in response to a single bout of heavy exercise was significantly associated with the percentage of type II fibers in the vastus lateralis muscle in populations with a range of percentages of different fiber types (3). These data suggest the importance of activation pattern of muscle fibers within the exercising muscle. Type II fibers have a relatively greater glycolytic capacity, greater O₂ diffusion distances, and presumably slower O₂ kinetics (7, 30). Thus the change of the O₂ slow component may be influenced by the relative amount of recruitment of type II fibers. In addition to the recruitment pattern of the different fiber types, attention should be given to the importance of microcirculatory changes to modulate the fibers' O₂ (25).

At the onset of the second exercise bout after a warm-up involving the same muscle group, the matching of O₂ delivery to O₂ demand, i.e., blood flow-O₂ in microcirculation within the contractile units, may be more rapidly established, i.e., there might be greater O₂ availability. Total Hb concentration measured by NIRS in this study supports this hypothesis indirectly. Therefore, a more aerobic metabolic profile, i.e., greater recruitment of type I fibers, may exist during the second bout of exercise. It would reduce the need for recruitment of additional type II fibers throughout bout 2 of exercise compared with during the bout 1. The hypothesis of a more homogeneous blood flow-O₂ in the second bout leading to a reduced recruitment of type II fibers may also explain the decreased amount of slow component [A'₂ and O_2(6-3)] and an increased gain (Gp/W) of primary component in this study. The delivery and distribution of O₂ to working muscles, i.e., the interaction (matching or mismatching) of peripheral microcirculatory blood flow and O₂ demand within the muscles, might be one of the principal rate-limiting steps in a normal heavy exercise situation. Accordingly, manipulated increases of bulk O₂ delivery to an exercising limb and O₂ diffusion into the muscle cell (14, 15) may have no significant effect on muscle O₂ kinetics as far as blood flow-O₂ match at the microcirculatory level. In one of our recent studies, pulmonary O₂ and femoral artery blood flows were measured simultaneously during repeated bouts of heavy exercise by using relatively large working muscles (bilateral knee extension). That study demonstrated that faster kinetics of blood flow to exercising limb (i.e., bulk O₂ delivery) was not seen, even when O₂ was significantly accelerated in exercise bout 2 (9). This seems to indicate the importance of blood flow-O₂ matching (i.e., postexercise hyperemia), as opposed to improved bulk O₂ delivery regarding faster blood flow during exercise.

Another likely explanation of the O₂ slow component is related to the factor(s) involved in the metabolism of the muscle cell. A recent study that used the repeated-stimulation protocol in frog lumbrical myocytes indicated that intracellular factors rather than O₂ availability may determine the on-kinetics of oxidative phospholylation, and a prior contractile period results in more rapid on-kinetics (19). Furthermore, Howlett et al. (20) demonstrated that dichloroacetic acid (DCA) infusion decreased the reliance on substrate level for phosphorylation during the transition from rest to 65% O_2 max exercise in humans. Elevated pyruvate dehydrogenase activation by the infused DCA increased the provision of substrate for oxidative metabolism resulting in decreased phosphocreatine utilization and in an improved cellular energy state during the first 2 min after the onset of exercise. In other words, a delayed increase in pyruvate dehydrogenase activity resulted in insufficient substrate for the tricarboxylic acid cycle in the initial phase of exercise (1, 18). Therefore, there appears to be some limitation to O₂ utilization in exercising muscle in the initial phase of a first exercise bout. However, a more recent study with DCA infusion by the same group showed no effect on skeletal muscle metabolism if the exercise intensity was 90% O_2 max and not 65% (29).

Elevated muscle temperature is another possibility that might cause the O₂ slow component by decreasing the phosphorylation potential and increasing the rate of mitochondrial respiration, i.e., a Q₁₀ effect (36). However, at least during high-intensity exercise (at 50%), external heating of exercising limbs by ~3°C before the onset of heavy cycling exercise had no significant effect on the O₂ response in the initial phase (21). The Q₁₀ effect is, therefore, unlikely to be a dominant factor in determining O₂ kinetics during heavy exercise.

In summary, this study demonstrated that O₂ did not accelerate at the onset of square-wave high-intensity leg exercise transitions after a high-intensity warm-up exercise involving a different muscle group, even if systemic residual lactic acidosis was comparable to that measured after a warm-up exercise involving the same muscle group. The degree of hyperemia in working leg muscles was clearly greater at the onset of exercise bout 2 when preceded by a warm-up exercise using the same muscle group. These data suggest that the facilitated phenomenon of the O₂ kinetics during supra-LT exercise preceded by an intense warm-up may not be solely due to lactic acidosis. Rather, mechanisms that induce hyperemia in the site of muscular work seem necessary to induce this phenomenon. We speculate that such mechanisms improve the matching of O₂ delivery to O₂ demand just before bout 2. Further studies are needed to elucidate the cause of O₂ kinetics during the repeated-bout protocol, which may lead to further understanding of the regulation of O₂ transport and utilization during high-intensity exercise.