˙ O2 kinetics in heavy submaximal exercise Acceleration of V by hyperoxia and prior high-intensity exercise MAUREEN MACDONALD,1 PREBEN K. PEDERSEN,2 AND RICHARD L. HUGHSON1 of Kinesiology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1; and 2Department of Physical Education, Odense University, Odense DK-5230, Denmark
1Department
MacDonald, Maureen, Preben K. Pedersen, and Rich˙ O2 kinetics in heavy ard L. Hughson. Acceleration of V submaximal exercise by hyperoxia and prior high-intensity exercise. J. Appl. Physiol. 83(4): 1318–1325, 1997.—We ˙ O2) would change examined the hypothesis that O2 uptake (V more rapidly at the onset of step work rate transitions in exercise with hyperoxic gas breathing and after prior high˙ O2 were determined from intensity exercise. The kinetics of V the mean response time (MRT; time to 63% of total change in ˙ O2) and calculations of O2 deficit and slow component during V normoxic and hyperoxic gas breathing in one group of seven subjects during exercise below and above ventilatory threshold (VT) and in another group of seven subjects during exercise above VT with and without prior high-intensity exercise. In exercise transitions below VT, hyperoxic gas ˙ O2 at the breathing did not affect the kinetic response of V onset or end of exercise. At work rates above VT, hyperoxic gas breathing accelerated both the on- and off-transient MRT, ˙ O2 slow component reduced the O2 deficit, and decreased the V from minute 3 to minute 6 of exercise, compared with normoxia. Prior exercise above VT accelerated the on˙ O2 slow component from transient MRT and reduced the V minute 3 to minute 6 of exercise in a second bout of exercise with both normoxic and hyperoxic gas breathing. However, the summated O2 deficit in the second normoxic and hyperoxic steps was not different from that of the first steps in the same gas condition. Faster on-transient responses in exercise above, but not below, VT with hyperoxia and, to a lesser degree, after prior high-intensity exercise above VT support the theory of an O2 transport limitation at the onset of exercise for workloads .VT. ventilatory threshold; oxygen transport; oxygen utilization; oxygen deficit; oxygen debt
A REDUCTION IN THE MAGNITUDE of the O2 deficit at the onset of exercise during hyperoxic, compared with normoxic, breathing (19, 23) suggested that if one supplied more O2 to working muscle during this transient, it could be used. These data were obtained with Douglas bag collections of mixed expired air during ˙ O2). tests requiring ,70–80% of maximal O2 uptake (V ˙ O2 responses at To date, the only studies to follow the V the onset of exercise during hyperoxia with breath-bybreath techniques found that there was no difference from control (11, 18). It is possible that hyperoxia failed ˙ O2 kinetics because the fractional concento accelerate V tration of O2 was relatively low [inspiratory O2 fraction (FIO2) 5 0.30] (18) or that the work rates studied were of a relatively lower intensity, at which increasing arte˙ O2 kinetics (11, rial PO2 has not been shown to increase V ˙ O2 kinetics were 18), than in those studies in which V accelerated (19, 23). In the face of conflicting evidence ˙ O2 kinetics with hyperoxic for accelerated or unaltered V gas breathing, it is appropriate to set out the potential
1318
explanations. It is possible that, at least over a range of work rates, adequate O2 is delivered at the onset of exercise, and O2 utilization establishes the rate at ˙ O2 increases (8). Alternatively, hyperoxia might which V not supply more O2 over this same range of work rates because of reduced blood flow as noted by some (26), but not all (16), previous investigators. For step increases or decreases in work rates above compared with below the ventilatory threshold (VT), there might be different rate-limiting steps that es˙ O2 (18). In either tablish the time course of change in V case, the mechanisms can be expressed simply as being related to the delivery and distribution of O2 to the working muscles (10, 12, 18) or to utilization of O2 determined by the kinetics of the intramuscular oxidative processes (8, 24). During leg-cycling exercise below VT, there have been numerous examples in which reductions in O2 transport could slow the increase in ˙ O2 at the onset of exercise, including a start from a V baseline of existing mild exercise (12), hypoxia (21), b-adrenergic-receptor blockade (9), and supine compared with upright exercise (15). These examples hold open the possibility that O2 transport might be rate limiting. It has been shown that, when constant-load exercise has been constrained to be only 60–65% of the ˙ O2 kinetics can be altered by work rate at VT, the V manipulation of arterial perfusion pressure (10). Thus, in subjects in the upright and supine positions during ˙ O2 application of lower body negative pressure, the V kinetics were faster than when the sub-VT exercise was conducted in a supine posture (10). However, in healthy individuals, for upright exercise below VT in normoxia, it has never been demonstrated that increasing O2 ˙ O2 kinetics. transport can increase V For work rates above VT, making more O2 available ˙ O2 kinetics. Breathing a 70% inmight accelerate V spired O2-gas mixture could increase dissolved O2 by up to 10 ml O2/l of blood. The examples of smaller O2 deficits with hyperoxia were probably for work rates above VT (19, 23). Two reports (6, 7) have suggested ˙ O2 increased more rapidly in the second of two that V step transitions to work rates about halfway between ˙ O2 and no effect of prior highVT and maximum V intensity exercise for a subsequent subthreshold exercise bout. They attributed this to the existing changes in local muscle environment that promoted both more rapid increases in blood flow and a rightward shift of the O2-hemoglobin dissociation curve that facilitated O2 release at the working muscle. The purposes of this study were threefold. First, we wanted to obtain data from breath-by-breath analysis ˙ O2 with hyperoxic gas breathing during the chalof V lenge of exercise to work rates both below and above
0161-7567/97 $5.00 Copyright r 1997 the American Physiological Society
http://www.jap.org
˙ O2 KINETICS IN HYPEROXIA V VT. Second, we wanted to examine the relationships between the magnitude of the O2 deficit and the slow component of the O2 increase in normoxia and hyperoxia. Third, we wanted to determine whether the second of two step increases in work rate to intensities above VT might be further accelerated by a subject completing this exercise while breathing a hyperoxic gas mixture. These experiments allowed the testing of the hypothesis that, at least for work rates above VT, O2 transport to the exercising muscle acts as the rate˙ O2 at the onset of limiting step for the increase in V exercise. METHODS
The present study was conducted in two parts. The first ˙ O2 kinetics in response examined the effect of hyperoxia on V to step changes in work rate to below and above VT. The second investigated the effect of hyperoxia and prior high˙ O2 kinetics after step changes to above intensity exercise on V VT. Subjects Seven subjects participated in each part of the study, with one subject participating in both parts. In the first part, five men and two women volunteered, and in the second part four men and three women volunteered. Each subject signed a consent form approved by the Office of Human Research of the University of Waterloo after reading a description of the methods and possible risks. Preliminary testing of all subjects consisted of an incremental exercise test to exhaustion in normoxia by using a ramp work rate protocol (15 W/min ramp at a cycling frequency of 60 revolutions/min). The gas-exchange data obtained from the normoxic ramp test ˙ O2. These values were used to estimate the VT and the peak V were used in choosing the individual work rates for the step tests. The VT was determined from the point of increased ˙ E)-to-V˙ O2 ratio (V˙ E/V˙ O2) with no change minute ventilation (V ˙ E-to-CO2 output (V˙ CO2) ratio, as previously described in the V (11). The number of tests performed by each subject varied according to the noise observed in the breath-by-breath signal (17). Normally, four tests were sufficient because of the large amplitude of the steps. Experimental Design Part 1 involved testing both below and above VT with step changes in work rate. Subjects cycled at 25 W for 4 min to establish a baseline, before the following work rate changes: a step increase in work rate to below VT (80% VT) for 6 min, a step decrease back to 25 W for 6 min, a step increase in work ˙ O2 for 10 min, rate to a level halfway between VT and peak V and a final step decrease to 25 W for 6 min. Subjects exercised in both normoxia (room air; FIO2 5 0.21) and hyperoxia (FIO2 5 0.70). In part 2 of the study, the focus was on the effect of previous high-intensity exercise on a subsequent identical transition in work rate. After a period of 4 min with subjects pedaling at a baseline work rate of 25 W, the work rate was increased to approximately halfway between the work rates at VT and ˙ O2 for 10 min. This was followed by a step decrease in peak V work rate back to the 25-W baseline for 6 min before a second identical step transition was performed. No warning was given to the subjects before any of the step transitions, although they were made aware of the protocol before the test. Experiments were performed with normoxia (room air,
1319
FIO2 5 0.21) and hyperoxia (FIO2 5 0.70). The four possible conditions were normoxic breathing in the first and second step transitions (NN), normoxic breathing in the first followed by hyperoxic breathing in the second step transition (NH), hyperoxic breathing in the first and second step transitions (HH), and hyperoxic breathing in the first followed by normoxic breathing in the second step transition (HN). Breath-by-Breath Data Breath-by-breath ventilation and gas exchange were measured on a computerized system (First Breath, St. Agatha, ON), which sampled inspired and expired volumes with a volume turbine (VMM-110, Alpha Technologies, Laguna Beach, CA) and fractional concentrations of O2, CO2, and N2 by mass spectrometry (Marquette MGA-1100A, Milwaukee, ˙ O2 and V˙ CO2 were calculated as WI) at a frequency of 200 Hz. V alveolar values with compensation for lung-gas stores and with computation of the effective lung volume (15). For the hyperoxic testing, a large Tissot tank was filled with inspiratory gases from cylinders containing 70% O2-30% N2. The gas was not humidified. The Tissot tank was connected to a Y valve (model 2730, Hans Rudolph, St. Louis, MO) to permit inspiration from the tank. The Y valve was open to room air during the normoxic transitions. The volume turbine was calibrated with a manually pumped syringe and with the equipment configured as it was during testing, including hyperoxic gas, for these tests. The mass spectrometer was calibrated for normoxia and hyperoxia by using two precision gas mixtures that spanned the anticipated fractional gas concentrations in both normoxia and hyperoxia. A calibration procedure was performed to determine the time required for gas transport and mass spectrometric response (lag time) (14). Separate lag times were determined for normoxia and hyperoxia because of the effect of gas density. The lag time for the hyperoxic mixture was ,30 ms slower than for the normoxic mixture (11). Heart rate was measured with an electrocardiograph (7803A, Hewlett-Packard) by using standard bipolar electrode placement. Mean heart rate over each breath was recorded. Data Analysis
˙ O2 from at least three repetiBreath-by-breath data for V tions of an identical test condition for each subject were linearly interpolated between breaths to give values at 1-s intervals. The identical tests were then time aligned, superimposed, and ensemble averaged to give a single data set per subject. The average individual response was fit to a curve by using an exponential model. The curve-fitting procedure involved the calculation of a modeled exponential output for test values of the various parameters by using the leastsquares error approach (13). These modeled outputs were compared with the actual individual averaged data set for that variable. The curve-fitting procedure was iterated until any further changes in the parameters for the model did not result in a reduction in the mean squared error between the curve drawn from the model and the averaged data set. The low-step tests, to work rates below VT, were fit to a two-component model, whereas the high-step tests, to work rates above VT, were fit to a three-component model. Steps from the baseline work rate to a higher work rate were referred to as ‘‘on-transients,’’ whereas steps from a higher work rate to the baseline work rate were referred to as ‘‘off-transients.’’ Work rate transitions to and from a work rate below VT were referred to as ‘‘low steps,’’ and transitions to and from a work rate above VT were referred to as ‘‘high steps.’’
˙ O2 KINETICS IN HYPEROXIA V
1320
The two-component model used to fit the responses to the low-step tests had a baseline (G0 ) and two amplitude terms (G1 and G2 ), two time constants (t1 and t2 ), and two time delays (TD1 and TD2 ) as previously described (13)
˙ O2(t) 5 G0 1 G1[1 2 e2 (t 2 TD1)/t1] · u1 1 G2[1 2 e2 (t 2 TD2)/t2] · u2 V where
same gas condition for step 2 and with different previous gas conditions in step 1 were treated as identical and averaged before further statistical analysis. When significant main effects were observed, post hoc testing included comparisons by the Student-Newman-Keuls test. A significance level of P , 0.05 was maintained for all comparisons. RESULTS
u1 5 0
for t , TD1 and u1 5 1 for t $ TD1
u2 5 0
for t , TD2 and u2 5 1 for t $ TD2
˙ O2(t) is the time-dependent variation in V˙ O2. V The data from the high-step tests were fitted to a threecomponent model. The three-component model contained an extra amplitude term (G3 ) and time constant (t3 ) to fit the slower adaptive phase in these tests. In the absence of a rationale for letting the third component begin at some time after the second, we used a model equivalent to that of Linnarsson (18) and had the second and third components start together. This issue is considered further in DISCUSSION ˙ O2(t) 5 G0 1 G1[1 2 e 2 (t 2 TD1/t1] · u1 V 1 G2[1 2 e 2 (t 2 TD2)/t2] · u2 1 G3[1 2 e 2 (t 2 TD3)/t3] · u2 where u1 and u2 were defined above and TD2 5 TD3 u1 5 0 for t , TD1 and u1 5 1 for t ./TD1 u2 5 0 for t , TD2 and u2 5 1 for t ./TD2 The overall time course of the response was determined from mean response time (MRT). The MRT is the time it takes to reach ,63% of the total amplitude of the response from the baseline to the final plateau value. It was calculated as a weighted sum of the time delay and time constant for each component MRT 5 [G1/(G1 1 G2 1 G3)] · (TD1 1 t1) 1[G2 /(G1 1 G2 1 G3)] · (TD2 1 t2) 1[G3 /(G1 1 G2 1 G3)] · (TD3 1 t3)
˙ O2 values of the subjects during The mean peak V testing in normoxia were 45.3 6 1.6 (SE) ml·kg21 · min21 in part 1 and 44.4 6 2.4 ml · kg21 · min21 in part 2. The ˙ O2 values achieved during the below-VT (lowmean V step) and above-VT (high-step) tests in part 1 were ˙ O2 at average work 50.4 6 0.9 and 78.6 6 1.0% of peak V rates of 125 and 215 W, respectively. In part 2, the mean ˙ O2 achieved during the high-step tests was 82.0 6 0.8% V ˙ O2 at an average work rate of 195 W. of peak V ˙ O2 Response Fitting V ˙ O2 Part 1. The effect of hyperoxic gas breathing on V kinetics was examined for work rate transitions to and from below and above VT (Fig. 1, Table 1). The on˙ O2 represented by the MRT were transient kinetics of V not accelerated for low steps with exercise in hyperoxia compared with exercise in normoxia. In contrast, hyperoxic gas breathing resulted in a faster adjustment to ˙ O2 for high steps. No effect of hyperoxic steady-state V gas breathing on the kinetics of the off-transient responses was observed for either low- or high-step transitions. For both on- and off-transients, the MRT was significantly less for low steps compared with high steps regardless of gas-breathing condition. The highstep off-transient responses were faster than the corresponding on-transient responses, for both normoxic and hyperoxic gas-breathing conditions. The low-step off-transient responses were slower than the ontransient responses for hyperoxic gas breathing, and there was no difference between the on- and offtransients during normoxic gas-breathing tests.
˙ O2 reIn addition to the curve-fitting procedure, the V sponses were also analyzed by calculating the O2 deficit and ˙ O2 by determining the slow component response of the V response from 3 to 6 min and from 6 to 10 min after the step change in work rate. The O2 deficit was taken as the ˙ O2 at any time after the difference between the measured V ˙ O2 start of the higher-work-rate exercise and the average V measured during the last minute of exercise. Statistics Statistical analysis was performed by using two-way repeated measures analysis of variance of the main effects. In part 1 the main effects, i.e., inspired gas concentration and work rate, and in part 2 the main effects, i.e., prior exercise and inspired gas concentration, were examined across vari˙ O2. The parameters examined ous parameter estimates for V were those generated from the curve-fitting procedure when applied to the average individual data set for each subject. Significant differences in MRT because of main effects were interpreted as reflecting differences in kinetics, while differences in other parameters were also noted. In part 2, when no significant difference in MRT was observed within step 2 because of the effect of previous gas condition, tests with the
˙ O2) at baseline and during transitions in Fig. 1. Oxygen uptake (V exercise to low and high work rates for 1 subject breathing normoxic gas (dotted line) and hyperoxic gas (solid line). Lines represent average response for 4 identical transitions.
˙ O2 KINETICS IN HYPEROXIA V
1321
˙ O2 responses for step transitions below and above VT Table 1. Parameter estimates from fitting V in normoxia and hyperoxia G0 , ml/min
G1 , ml/min
TD1 , s
t1 , s
G2 , ml/min
TD2 , s
t2 , s
G3 , ml/min
t3 , s
MRT, s
On-transient ,VT Normoxia Hyperoxia .VT Normoxia Hyperoxia
734 6 22 742 6 21
387 6 42 402 6 52
3.4 6 0.6 11.2 6 1.1 3.6 6 0.6 12.1 6 1.7
739 6 24 723 6 20
614 6 94† 3.4 6 0.8 10.8 6 2.3 539 6 92† 3.6 6 0.3 8.2 6 1.6
561 6 64 578 6 66
23.5 6 1.2 23.6 6 1.3
19.8 6 2.6 18.0 6 2.6*
779 6 60† 17.8 6 0.6† 17.6 6 1.0† 1,180 6 129*† 16.6 6 0.8† 23.6 6 2.0*†
31.3 6 1.3 31.4 6 1.4 516 6 32 104 6 17 239 6 112* 101 6 10
53.9 6 6.2† 44.1 6 5.2*†
Off-transient ,VT Normoxia Hyperoxia .VT Normoxia Hyperoxia
1,682 6 116‡ 2324 6 79‡ 1.7 6 0.4 10.9 6 3.5 1,715 6 125‡ 2388 6 113‡ 2.4 6 0.6 12.2 6 3.4
2620 6 92‡ 16.6 6 1.7‡ 28.5 6 0.6‡ 2599 6 120*‡ 14.9 6 1.5*‡ 33.3 6 2.4*‡
34.6 6 0.7 35.4 6 1.2‡
2,641 6 173†‡ 2503 6 95†‡ 1.6 6 0.2 14.7 6 8.2 21,084 6 132†‡ 18.9 6 0.4 21.9 6 1.3†‡ 2265 6 62†‡ 82.7 6 8.8† 43.5 6 1.1†‡ 2,676 6 195†‡ 2435 6 63‡ 2.2 6 0.4 11.2 6 7.4 21,218 6 131†‡ 16.8 6 0.9* 21.7 6 0.6† 2260 6 34†‡ 78.1 6 6.7† 40.9 6 1.4†‡
Values are means 6 SE for at least 4 repetitions by 7 subjects. VT, ventilatory threshold; G0 , 2-component model baseline; G1 and G2, 2-component 2 amplitude terms; G3 , 3-component model amplitude term; TD1 and TD2 : 2-component model time delays; t1 and t2 , 2-component model time constants; t3 , 3-component model time constant; MRT, mean response time. * Significantly different from normoxia for same step change in work rate, P , 0.05. † Significantly different from ,VT step in same gas-breathing condition, P , 0.05. ‡ Significantly different from the on-transient for same work rate and same gas-breathing condition, P , 0.05.
Part 2. In this part of the study, when normoxic and hyperoxic gas breathing were balanced between the step 1 and step 2 transitions, it was found that there was no effect of prior gas breathing on the MRT of the second of two exercise transitions (P . 0.05). Consequently, results were pooled and analyzed according to the gas condition within the exercise transition being examined (Fig. 2, Table 2). As in part 1, during high˙ O2 on-response was faster step exercise transitions the V in hyperoxia than in normoxia, and this was true for both steps 1 and 2 (Table 2). Similarly, there were no ˙ O2 kinetics after effects of hyperoxia on off-transient V ˙ O2 on-transient kinetics the step 1 or step 2 test. The V were significantly accelerated as a result of prior exercise in both the normoxic and hyperoxic gas-
˙ O2 at baseline and during transitions in exercise to high Fig. 2. V work rates for 1 subject breathing normoxic gas (dotted line) and hyperoxic gas (solid line). Lines represent average response for 4 identical transitions.
breathing tests. Prior high-intensity exercise had no effect on any of the off-transient MRT. In step 1, during normoxic gas breathing the off-transient responses were faster than the corresponding on-transient responses, and no difference was observed between the on- and off-transients during hyperoxic gas breathing. ˙ O2 kinetics were During the step 2 transition, slower V observed during the off-transient in hyperoxia than during the corresponding on-transient, whereas there were no differences between the on- and off-transient responses in normoxic gas breathing.
˙ O2 Slow Component O2 Deficit and V Low steps. There was no difference in the O2 deficit calculated for hyperoxic and normoxic tests to below ˙ O2 VT. There was not a significant slow component of V in either hyperoxic or normoxic low-step transitions (Table 3). These results were consistent with those obtained by curve fitting (Table 1). High steps. Also consistent with the curve-fitting results presented above were the observations of a significantly smaller O2 deficit in high-step transitions with hyperoxia compared with normoxia in both parts 1 and 2 (Table 3). However, calculated O2 deficit in the second normoxic and hyperoxic steps was not different from step 1 in the same gas conditions. ˙ O2 kinetics, from the V˙ O2 slow The indication of V component between minutes 3 and 6 of the high-step tests, was consistent with the curve-fitting results (Tables 1 and 2). There was significantly less slow component in hyperoxia compared with normoxia, and the step 1 transition had a greater O2 slow component than that in step 2 (Table 3). Only for step 2 in normoxia ˙ O2 was there a significant further slow component in V between minutes 6 and 10 of the step tests. These results indicate that most effects of the experimental
˙ O2 KINETICS IN HYPEROXIA V
1322
˙ O2 kinetics for step transitions above VT in normoxia and hyperoxia Table 2. Parameter estimates for V without and with prior exercise G0 , ml/min
G1 , ml/min
t1 , s
TD1 , s
G2 , ml/min
TD2 , s
t2 , s
17.5 6 1.1 20.6 6 1.2
18.8 6 1.2 20.0 6 2.7
G3 , ml/min
t3 , s
MRT, s
On-transient Step 1 Normoxia Hyperoxia Step 2 Normoxia Hyperoxia
787 6 36 822 6 40*
451 6 38 2.9 6 0.7 7.2 6 2.0 777 6 115* 3.0 6 0.4 13.4 6 1.3
842 6 43† 871 6 43*†
606 6 112 552 6 90*
960 6 108 850 6 125
2.6 6 0.5 13.2 6 3.0† 2.9 6 0.8 10.4 6 2.7
913 6 126 1,226 6 159
17.0 6 0.8† 18.6 6 0.9 16.6 6 0.9† 21.4 6 1.9
536 6 85 106 6 7 53.4 6 2.6 347 6 125* 87 6 10* 40.5 6 2.6* 346 6 81† 113 6 12 46.2 6 3.0† 133 6 68*† 98 6 12* 37.8 6 3.3*†
Off-transient Step 1 Normoxia Hyperoxia Step 2 Normoxia Hyperoxia
2,718 6 199‡ 2527 6 139‡ 1.5 6 0.4 2,760 6 200‡ 2501 6 78‡ 2.2 6 0.5
6.5 6 3.0 3.8 6 1.3
21,120 6 127‡ 19.3 6 1.1 21,281 6 139‡ 16.9 6 0.9
21.7 6 1.6 2232 6 41‡ 24.3 6 1.0‡ 2148 6 57‡
125 6 11.2 45.1 6 1.8‡ 127 6 12 41.4 6 3.7
2,722 6 194‡ 2477 6 78‡ 1.9 6 0.4 2,780 6 199‡ 2679 6 210‡ 2.0 6 0.3
4.0 6 1.5‡ 21,252 6 91‡ 18.9 6 1.1 6.2 6 3.5 21,139 6 162‡ 20.0 6 3.8
25.6 6 1.3‡ 2167 6 42‡ 25.0 6 2.4‡ 2164 6 50‡
127 6 13 44.6 6 2.7 137 6 16‡ 43.4 6 3.0‡
Values are means 6 SE for at least 4 repetitions by 7 subjects. Steps 1 and 2: without and with prior exercise, respectively. * Significantly different from normoxia for same step change in work rate, P , 0.05. † Significantly different from step 1 in the same gas-breathing condition, P , 0.05. ‡ Significantly different from on-transient for same step change in work rate and same gas-breathing condition, P , 0.05.
manipulations were complete within the first 6 min of exercise. DISCUSSION
Previous evidence that O2 transport might limit the ˙ O2 at the onset of exercise came rate of increase in V from experiments in which the magnitude of the O2 deficit was reduced at the onset of moderately heavy exercise while a subject breathed a hyperoxic gas mixture (19, 23) and in which incomplete recovery after prior exercise at a similar high intensity might have facilitated blood flow and O2 release from hemoglobin in a second bout of exercise (6, 7). This study has combined these methodologies in an attempt to examine the hypothesis that O2 transport acts as the rate˙ O2 at the onset of limiting step for the increase in V exercise above VT. Both hyperoxia and prior exercise ˙ O2 kinetics (i.e., faster MRT, caused an acceleration of V
˙ O2 slow component Table 3. O2 deficit and V for parts 1 and 2 O2 Deficit, liters
˙ O2 Slow V Component (3–6 min), ml/min
˙ O2 Slow V Component (6–10 min), ml/min
Part 1 Low normoxia Low hyperoxia High normoxia High hyperoxia
0.50 6 0.06 0.50 6 0.07 1.72 6 0.22 1.37 6 0.21*
10 6 6‡ 4 6 11‡ 87 6 22 43 6 21*
30 6 17‡ 11 6 14‡
Part 2 Step 1 Normoxia Hyperoxia Step 2 Normoxia Hyperoxia
1.55 6 0.20 1.18 6 0.15*
93 6 13 47 6 10*
9.3 6 13.2‡ 14 6 15‡
1.46 6 0.09 1.15 6 0.13*
46 6 8† 30 6 7*†
31 6 4 12 6 11‡
Values are means 6 SE; n 5 7 subjects. * Significantly different from normoxia for same step, P , 0.05. † Different from step 1 in same gas-breathing condition. ‡ Not different from zero, P , 0.05.
˙ O2 slow component smaller O2 deficit, and reduced V from 3 to 6 min of exercise) for work rates above VT. When the two manipulations were combined, there was ˙ O2 response, indicating that further speeding of the V the two stimuli may work independently to increase O2 transport at the onset of exercise. For work rates below VT, there was no significant effect of hyperoxia on the ˙ O2 at the onset of exercise. time course of increase in V ˙ O2 was faster after The postexercise rate of decrease in V the below-VT exercise than after the above-VT exercise, yet there was no difference between normoxia and hyperoxia. Methodological Considerations There are two components of the design of this study that merit further consideration before physiological interpretation of the data can be discussed. The first is the selection of the work rates for the study, and the second is the model used to fit the experimental data. We selected work rates that were constant across the normoxia or hyperoxia treatments even though the ˙ O2 are often found to be increased peak work rate and V by up to 10% by hyperoxia (25). It was felt that this would not affect the kinetics responses because in each of the below- and above-VT cases, the work rates were clearly within the domains required. Furthermore, it ˙ O2 are altered has been observed that kinetics of V across a range of work rates, being slower for above-VT work rates, as we found; except for minor differences in percentage of peak work rate, there is little effect on the ˙ O2 at the onset of exercise (4, 28). In this time course of V study, our focus was on the effects of hyperoxia and/or ˙ O2 response to a fixed work rate. prior exercise on the V ˙ O2 response at the onset and end of Fitting of the V exercise has been done by a variety of methods, primarily on the basis of the observation of an exponential, or near-exponential, change (18). At the below-VT work rate, we used a two-component exponential, in which the first component accounted for the rapid increase in
˙ O2 KINETICS IN HYPEROXIA V ˙ O2 at the onset or decrease in V˙ O2 at the end of V exercise, because the return of venous blood is altered with the change in work rate. The second component represents the major change to the new steady-state value and is described by a single time constant with time delay (15). The major difference between studies ˙ O2 response to work has been in the description of the V rates above VT. We used a three-component exponential model. The first component of the above-VT model was identical to that in the below-VT exercise. The second and third components represent a relatively rapid and a more slowly developing component, respectively. This interpretation is consistent with other research (28) ˙ O2 slow-compoand with our observations from the V nent analysis (Table 3). In accord with Linnarsson (18), we had both phases 2 and 3 start at the same time point. In contrast, some researchers have allowed the time delay of the third phase to vary, demonstrating that this component of the response started 90–120 s after the onset of the exercise (2, 22). Curve fitting of physiological responses can incorporate physiological correlates, or it can simply attempt to minimize the error of the distribution of the data points about the line of best fit. In this case, it is not possible to identify specifically the physiological mechanism responsible ˙ O2 response (28). for the slow component of the V Therefore, we justify our selection of the model with a common time delay for phases 2 and 3 by the absence of a statistically significant improvement in fit between our model and that with an extra parameter for a third time delay. Whichever model is selected, the physiological interpretation as presented below is not altered. Effect of Hyperoxia In the steady-state phases of the present experiments, there were only minor differences in measured ˙ O2 between normoxia and hyperoxia. In the baseline V period before step 1 of part 2 of the study (Table 2), ˙ O2 with there was a small but significantly greater V hyperoxia compared with normoxia. Later within this same protocol, during the baseline period before step 2, ˙ O2 was also higher with hyperoxia. Some previous V research (25) has found an increase in fat utilization with hyperoxic breathing, although this was not suggested in the present study because the respiratory ˙ O2 could exchange ratio was unaltered. The elevated V have also been a consequence of inadequate time for equilibration to hyperoxia. Careful calibration of the equipment in this study was performed. We cannot, however, totally disregard the possibility that small errors in our measurements might have caused this statistical finding. ˙ O2 on-transient MRT and the smaller O2 The faster V ˙ O2 slow component observed for the deficit and V above-VT steps in hyperoxia, compared with normoxia, are in agreement with previous studies using Douglas ˙ O2 bags and measurements of O2 deficit to estimate V ˙ kinetics (19, 23). The faster VO2 responses at the same absolute work rate are consistent with the pattern to be expected when the relative work rate is reduced by
1323
hyperoxia. Closer examination of the parameter estimates for the exponential curve fitting (Tables 1 and 2) shows that the increase during phase 3 was less in the hyperoxia tests than in normoxia. Therefore, to attain ˙ O2 values at the end of exercise, a greater slow similar V increase (see G3 and t3 in Tables 1 and 2) occurred in normoxia. This observation is also supported by a greater slow component between minutes 3 and 6 during normoxia compared with hyperoxia (Table 3). This observation supports the findings of Paterson and Whipp (22). Hyperoxic gas breathing was used in this study in an attempt to increase the transport of O2 to working muscle and therefore to determine whether O2 trans˙ O2 response port is a limiting factor in the transient V after a step increase in work rate above and below VT. Others have shown that hyperoxia does not accelerate ˙ O2 for step transients below VT (11, 18). the kinetics of V The mechanism responsible for this could be that adequate O2 is already being delivered at the onset of this light exercise, and hence O2 delivery would not be limiting, or that hyperoxia does not, in fact, provide more O2 at the onset of exercise. During steady-state exercise, blood flow to working muscles is reduced (26) or unaltered (16) with hyperoxic gas breathing. It is not known whether hyperoxic breathing resulted in decreased blood flow to the exercising muscle in this study. A reduction in blood flow, to counter the increase in blood O2 content, would serve to maintain the O2 delivery to the working muscle at approximately the same level as in normoxia. Another unknown is the mean capillary PO2. Knight et al. (16) identified an apparent nonlinear relationship between their estimated value for mean capillary PO2 and peak leg ˙ O2 with hyperoxia. Among other mechanisms muscle V that might account for this, Knight et al. included heterogeneity of flow distribution to the exercising muscle because of hyperoxia-induced vasoconstriction. ˙ O2 kinetics were The observation that on-transient V accelerated with hyperoxic gas breathing for step transitions to levels above VT indicates that the rate of O2 transport was probably accelerated by hyperoxia during these step transitions. Effect of Prior Exercise Prior high-intensity exercise was also used in this study as a means of attempting to increase the O2 transport to the working muscle. A prior bout of high˙ O2 onintensity exercise (.VT) accelerated the V transient kinetics in normoxia and in hyperoxia, as evidenced by a smaller MRT. These findings are in agreement with previous research using a similar work rate protocol (6, 7), which found that, for exercise in normoxia at work rates above VT, prior exercise acceler˙ O2 kinetics of a subsequent exercise bout. ated the V Metabolic acidosis as a result of the prior exercise bout has been suggested as a possible explanation for ˙ O2 kinetics of the second exercise bout (6, 7). the faster V These previous reports postulated that the prior exercise elevated the concentration of blood lactate and decreased the pH in the working muscle, thereby
1324
˙ O2 KINETICS IN HYPEROXIA V
promoting changes in osmolality and acidity that result in vasodilation in the working muscles. Other vasoactive substances might also still be elevated after 6 min of recovery. An additional factor that could promote O2 delivery to the working muscle would be a rightward shift of the O2-hemoglobin dissociation curve resulting from the accumulation of H1 and increased CO2 (5). These results support the theory that O2 transport is ˙ O2 ona limiting factor to high-intensity-exercise V transient kinetics, although changes in the intracellular metabolic environment that might promote a more rapid increase in oxidative metabolism have not been investigated. The time course of increase in oxidative phosphorylation at the onset of exercise has been estimated by a range of techniques, yet it is uncertain over what range of work rates the time course of the adaptive process remains constant (14, 20, 28). With cycling exercise, ˙ O2 measured there are definitely slower kinetics for V at the mouth at high work rates (28). Results from recent studies with nuclear magnetic resonance spectroscopy indicate both slower changes (20) and no differences in the rates of change in phosphocreatine ˙ O2 (30) at higher work rates. Measurement of muscle V as the product of blood flow and arteriovenous O2 content difference also showed a slower response at higher work rates (14). None of the experimental conditions resulted in high-step kinetics that were as fast as the low-step kinetics (Tables 1 and 2). This indicates either that O2 transport was not the only limiting factor for the high steps or that the experimental conditions did not completely alleviate the O2 transport deficit. O2 Deficit and O2 Debt The literature contains reports of the magnitude of O2 debt exceeding, equaling, or being less than the magnitude of the O2 deficit. Part of this discrepancy arises from the differences in exercise work rates studied, and part arises from differences in baseline state after the exercise. In an earlier study (1) in which debt exceeded deficit, the intensity of exercise studied was high, and the baseline after exercise was considered to be basal rest. In many studies in which lighter intensities of exercise (,VT) were studied with either rest or light exercise baselines, O2 deficit and O2 debt were normally about equal (9, 27, 31). The present results for below VT in both normoxia and hyperoxia, in which the time course of recovery (MRT) and the magnitude of change are almost the same as those at the onset of exercise, are in agreement with this. When the O2 deficit is increased, by b-adrenergic-receptor blockade (9) or by higher intensity exercise, as shown in the present study by the slower MRT, the O2 debt can be found to be smaller than O2 deficit when the recovery baseline is mild exercise. The most probable explanation for this is that lactate produced during the ontransition to compensate for the inadequate O2 delivery is metabolized as a substrate during exercise and in the mild exercise recovery (3). That is, it is not necessary to repay this component of the debt (11).
Wilson et al. (29) have shown that the ATP-to-ADP ˙ O2 as a ratio can be altered at a given steady-state V function of intracellular PO2. This could account for the different concentrations of lactate observed in hypoxia vs. normoxia or hyperoxia (19). Recently, Hughson et al. (14) speculated that altered intracellular PO2 at the ˙ O2 response during onset of exercise might modify the V the non-steady-state transition. The intracellular PO2 at this time would be a function of O2 delivery (arterial PO2 and blood flow) and O2 utilization. The present results are consistent with this hypothesis for work rates .VT. ˙ O2 slow-component values were The O2 deficit and V smaller than the corresponding measurements of Gerbino et al. (7). These differences are likely to be because of the higher work rates used in the high-step transitions of the previous study. Although in both studies subjects were assigned a high work rate that was ˙ O2, the approximately halfway between VT and peak V ˙ O2) measured values in the previous study (92% peak V were considerably higher than in the present study ˙ O2). These differences in protocol (79 and 82% of peak V may also explain the observation that prior highintensity exercise did not influence the magnitude of the O2 deficit in the present study, in contrast to the findings of Gerbino et al. (7). The major contribution to O2 deficit occurs before 3min. However, we measured O2 deficit over the full 10 min of exercise. The lack of effect of prior exercise could have resulted from small differences in the early response or in the apparent plateau value with the longer exercise duration in this study. Conclusions
˙ O2 kinetics observed for the on-transient The faster V during step changes above VT with hyperoxia, and to a lesser degree with prior exercise, provide evidence that ˙ O2 the supply of O2 contributes to the control of tissue V for this relatively high exercise intensity. The obser˙ O2 kinetics for steps below VT vation of no change in V may indicate that O2 transport is not a limiting factor in this light-exercise-intensity condition. However, in the absence of definitive data concerning O2 transport in the critical adaptive phase, it is not possible to rule out the alternative hypothesis. Off-transient kinetics above VT did not change significantly because of prior exercise or hyperoxic gas breathing, indicating that regulatory mechanisms for O2 supply and utilization may be different in on- and off-transients above VT. This research was supported by the Natural Sciences and Engineering Research Council (NSERC) of Canada. M. MacDonald is an NSERC Graduate Scholarship recipient. Address for reprint requests: R. L. Hughson, Dept. of Kinesiology, Univ. of Waterloo, Waterloo, Ontario, Canada N2L 3G1 (E-mail:
[email protected]). Received 24 September 1996; accepted in final form 11 June 1997. REFERENCES ˚ strand, P. O., and B. Saltin. Oxygen uptake during the first 1. A minutes of heavy muscular exercise. J. Appl. Physiol. 16: 971– 976, 1961. 2. Barstow, T. J., and P. A. Mole. Linear and nonlinear characteristics of oxygen uptake kinetics during heavy exercise. J. Appl. Physiol. 71: 2099–2106, 1991.
˙ O2 KINETICS IN HYPEROXIA V 3. Belcastro, A. N., and A. Bonen. Lactic acid removal rates during controlled and uncontrolled recovery exercise. J. Appl. Physiol. 39: 932–936, 1975. 4. Casaburi, R., T. J. Barstow, D. M. Robinson, and, K. Wasserman. Influence of work rate on ventilatory and gas exchange kinetics. J. Appl. Physiol. 67: 547–555, 1989. 5. Cochrane, J. E., and R. L. Hughson. Computer simulation of O2 transport and utilization mechanisms at the onset of exercise. J. Appl. Physiol. 73: 1–7, 1992. 6. 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 lacate threshold. J. Physiol. (Lond.) 417: 138, 1989. 7. Gerbino, A., S. A. Ward, and B. J. Whipp. Effects of prior exercise on pulmonary gas-exchange kinetics during highintensity exercise in humans. J. Appl. Physiol. 80: 99–107, 1996. 8. Grassi, B., D. C. Poole, R. S. Richardson, D. R. Knight, B. K. Erickson, and P. D. Wagner. Muscle O2 uptake kinetics in humans: implications for metabolic control. J. Appl. Physiol. 80: 988–998, 1996. 9. Hughson, R. L. Alteration in the oxygen deficit-oxygen debt relationships with b-adrenergic receptor blockade in man. J. Physiol. (Lond.) 349: 375–387, 1984. 10. Hughson, R. L., J. E. Cochrane, and G. C. Butler. Faster O2 uptake kinetics at onset of supine exercise with than without lower body negative pressure. J. Appl. Physiol. 75: 1962–1967, 1993. 11. Hughson, R. L., and J. M. Kowalchuk. Kinetics of oxygen uptake for submaximal exercise in hyperoxia, normoxia, and hypoxia. Can. J. Appl. Physiol. 20: 199–211, 1995. ˙ O2 12. Hughson, R. L., and M. A. Morrissey. Delayed kinetics of V in the transition from prior exercise. Evidence for O2 transport ˙ O2 kinetics. A review. Int. J. Sports Med. 11: limitation of V 94–105, 1983. 13. Hughson, R. L., D. L. Sherrill, and G. D. Swanson. Kinetics ˙ O2 with impulse and step exercise in man. J. Appl. Physiol. of V 64: 451–459, 1988. 14. Hughson, R. L., J. K. Shoemaker, M. E. Tschakovsky, and ˙ O2 on blood flow J. M. Kowalchuk. Dependence of muscle V dynamics at onset of forearm exercise. J. Appl. Physiol. 81: 1619–1626, 1996. 15. Hughson, R. L., H. C., Xing, C. Borkoff, and G. C. Butler. Kinetics of ventilation and gas exchange during supine and upright cycle exercise. Eur. J. Appl. Physiol. 63: 300–307, 1991. 16. Knight, D. R., W. Schaffartzik, D. C. Poole, M. C. Hogan, D. E. Bebout, and P. D. Wagner. Effects of hyperoxia on maximal leg O2 supply and utilization in men. J. Appl. Physiol. 75: 2586–2594, 1993.
1325
17. Lamarra, N., B. J. Whipp, S. A. Ward, and K. Wasserman. Effect of interbreath fluctuations on characterizing exercise gas exchange kinetics. J. Appl. Physiol. 62: 2003–2012, 1987. 18. Linnarsson, D. Dynamics of pulmonary gas exchange and heart rate changes at start and end of exercise. Acta Physiol. Scand. 415, Suppl. S1–S68, 1974. 19. Linnarsson, D., J. Karlsson, L. Fagraeus, and B. Saltin. Muscle metabolites and oxygen deficit with exercise in hypoxia and hyperoxia. J. Appl. Physiol. 36: 399–402, 1974. 20. McCully, K., S. Iotti, K. Kendrick, Z. Wang, J. O. Posner, J. Leigh, and B. Chance. Simultaneous in vivo measurements of HbO2 saturation and PCr kinetics after exercise in normal humans. J. Appl. Physiol. 77: 5–10, 1994. 21. Murphy, P. C., L. A. Cuervo, and R. L. Hughson. A study of cardiorespiratory dynamics with step and ramp exercise in normoxia and hypoxia. Cardiovasc. Res. 23: 825–832, 1989. 22. Paterson, D. H., and B. J. Whipp. Asymmetries of oxygen uptake transients at the on- and offset of heavy exercise in humans. J. Physiol. (Lond.) 443: 575–586, 1991. 23. Pedersen, P. K. Oxygen uptake kinetics and lactate accumulation in heavy submaximal exercise with normal and high inspired oxygen fractions. In: Biochemistry of Exercise, edited by H. G. Knuttgen, J. A. Vogel, and J. Poortmans. Champaign, IL: Human Kinetics, 1987, p. 415–420. 24. Pendergast, D. R., D. Shindell, P. Cerretelli, and D. W. Rennie. Role of central and peripheral circulatory adjustments in oxygen transport at the onset of exercise. Int. J. Sports Med. 1: 160–170, 1980. 25. Welch, H. G., and P. K. Pedersen. Measurement of metabolic rate in hyperoxia. J. Appl. Physiol. 51: 725–731, 1981. 26. Welch, H. G., F. B. Petersen, T. Graham, K. Klausen, and N. Secher. Effects of hyperoxia on leg blood flow and metabolism during exercise. J. Appl. Physiol. 42: 385–390, 1977. 27. Whipp, B. J., C. Seard, and K. Wasserman. Oxygen deficitoxygen debt relationships and efficiency of anaerobic work. J. Appl. Physiol. 28: 452–456, 1970. 28. Whipp, B. J., and K. Wasserman. Oxygen uptake kinetics for various intensities of constant-load work. J. Appl. Physiol. 33: 351–356, 1972. 29. Wilson, D. F., M. Erecinska, C. Drown, and I. A. Silver. Effect of oxygen tension on cellular energetics. Am. J. Physiol. 233 (Cell Physiol. 2): C135–C140, 1977. 30. Yoshida, T., and H. Watari. 31P-nuclear magnetic resonance spectroscopy study of the time course of energy metabolism during exercise and recovery. Eur. J. Appl. Physiol. 66: 494–499, 1993. 31. Yoshida, T., and B. J. Whipp. Dynamic asymmetries of cardiac output transients response to muscular exercise in man. J. Physiol. (Lond.) 480: 355–359, 1994.
