INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE OXYGEN UPTAKE (O₂) kinetics at the onset of exercise at low-to-moderate intensity is explained by two different hypotheses, i.e., the kinetics of oxygen utilization in muscle tissue (3, 22, 27) and oxygen transport to the tissue (9, 11, 14, 18). The latter hypothesis is derived from the observations that the O₂ kinetics were delayed with slow kinetics of heart rate (HR) or cardiac output () under various conditions, such as prior exercise-to-exercise transition (9), supine position (11), and hypoxia (14, 18), compared with normal stepwise work increase. These manipulations may alter the local blood distribution and/or oxygen flow response, as well as central circulatory responses, i.e., HR and . Shoemaker et al. (24) reported O₂ response was delayed with a delayed artery mean blood velocity as measured with pulsed Doppler methods. However, no previous study that suggests the effect of circulatory response on O₂ could discriminate between central and local circulatory responses. Thus there is no direct evidence that central circulatory response alters O₂ kinetics.

The autonomic nervous system (ANS) is one of the factors influencing oxygen transport (8), because ANS activity controls HR and by the balance of sympathetic and parasympathetic tone. Administration of -adrenergic blocker has been reported to slow down O₂ kinetics (7, 20). Hughson (8) explained that a faster rate of vagal activity withdrawal than sympathetic activation is responsible for faster O₂ kinetics. However, the effect of vagal withdrawal on the O₂ response is just speculation and has not been examined. Although the effect of sympathetic and parasympathetic (vagal) activity seems to be reciprocal, there is a difference in innervation to the control of local blood distribution. Sympathetic nervous activity innervates arterioles, whereas parasympathetic activity does not (21). It is probable that local blood flow response delayed by sympathetic blockade, rather than central circulatory response, altered the O₂ response during -adrenergic receptor blockade.

It has been reported that parasympathetic blockade depressed the HR response to dynamic exercise (15). Manipulation of oxygen transport in previous studies (9, 11, 14, 18) altered whole body circulatory responses, i.e., HR and , as well as local blood flow response. On the other hand, vagal activity has less of an effect on vasomotor tone. Thus the manipulation of vagal activity response enables us to evaluate the single effect of central circulatory response on the O₂ kinetics. If the faster rate of the efferent vagal activity withdrawal than sympathetic activation is one of the factors regulating the O₂ kinetics at the onset of work increase, interruption in vagal withdrawal at the onset of work increase should delay the time course of O₂ increase through slower HR and kinetics. It is essential to clarify the effect of vagal activity withdrawal to examine the role of HR and in modulating O₂ kinetics.

In this study we hypothesized that the delay of efferent vagal activity withdrawal influences O₂ kinetics through central circulatory responses at the onset of work increase. Delayed O₂ kinetics by increased vagal activity would confirm this hypothesis, and in turn the absence of an effect of vagal activity would disprove it. Facial stimulation induces bradycardia through the trigeminal-vagal-cardiac pathways (1, 5, 6, 12). To activate vagal tone, cold water facial stimulation for 2 min at the onset of exercise was performed. Two control trials, i.e., no facial stimulation and 1-min facial stimulation just before the onset of work rate increase, were performed. One-minute facial cooling was performed to consider the effect of different HR baseline values before the step increase of work rate. We continuously measured O₂, HR, and before and during work input, while the vagal activity was facilitated by facial cooling at the onset of exercise.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Eleven healthy males [21.2 ± 2.6 yr, 173.6 ± 5.0 cm, 65.6 ± 4.3 kg, O_{2 max} 3.4 ± 0.3 (SD) l/min] volunteered for this study. The study was explained to each individual, who then signed an informed consent form before participating.

The subjects performed 5 min of exercise after 3 min of 0 W pedaling using an electrically braked cycle ergometer (model 232C, Combi). The work rate of the 5-min exercise corresponded to 80% of the ventilatory threshold of each individual, which was 117.3 ± 20.3 W. The ventilatory threshold was determined as the work rate at which minute ventilation (E), carbon dioxide production (CO₂), and E/O₂ increased without increase of E/CO₂ as a function of O₂. In the trials the work rate was increased in a stepwise fashion without a cue to increase the work rate for the subject. Subjects maintained a constant pedaling rate of 60 rpm during cycling. For facial stimulation a vinyl bag filled with cold water (3-5°C) was applied to the face, i.e., for 1 min before the onset of the work increase (S1 trial) and for 2 min from 1 min before to 1 min after the onset of the work increase (S2 trial). Two experimenters held the bag, which cooled the area just around the forehead and eyes [the most sensitive area for this reflex (12)], and the subject wore a face mask. No cold stimulation was given in the Nr trial. The S1 trial was a pseudo-control trial to cancel the baseline effect. Each subject repeated each trial four times on different days. The testing order was randomized.

Respiratory flow through the mask was measured by a hot-wire flowmeter (RM-300, Minato Medical Sciences). Respired oxygen and carbon dioxide were analyzed with a mass spectrometer (WSMR-1400, Westron). This system was calibrated by using a 2-liter syringe, fresh air, and a precision gas (15% O₂-5% CO₂). was determined continuously by the impedance method (13). Two-band electrodes were placed on the neck and chest. The changes in impedance and electrocardiogram were stored on a hard disk at a frequency of 250 Hz. The beat-by-beat stroke volume change was calculated by the method of Kubicek et al. (13) and was applied to ensemble averaging (19). HR was also calculated from the data of R-R intervals. The product of HR and stroke volume was linearly interpolated once a second. The HR, , O₂, CO₂, and E were averaged for four trials in each subject and trial. The changes in those variables by facial stimulation were regarded as the difference between the mean values for 0-2 min and 2.5-3 min during unloaded pedaling.

Nonlinear least-squares fitting was applied to O₂, , HR, CO₂, and E data from the start of unloaded pedaling to 5 min after the increase by using a computerized regression algorithm where f(t) represents the O₂, , HR, CO₂, and E at time t; G (gain) is the steady-state increase from unloaded pedaling; and TD and  are the time delay and time constant, respectively. The data for 1 min just before the work increase was ignored in the calculation. Mean response time (MRT, sum of time constant and delay) was also used to estimate the kinetics of these variables. In this model, the baseline was obtained during unloaded pedaling. O₂ during the first 15 s after an increase in exercise was ignored in the calculation so that O₂ kinetics at phase 2 (25) were calculated.

The results are expressed as means ± SD. The one-way ANOVA was applied to compare the responses. When a significant effect was noted, the Duncan post hoc test was applied. The significance level was set at P < 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Figure 1 shows data for HR, , and O₂ time courses for one individual. HR decreased during facial stimulation. HR kinetics were clearly delayed in the S2 trial. had large variability but also seemed to be delayed in the S2 trial. The responses of O₂ in three trials were almost the same.

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Heart rate (HR, top), cardiac output (, middle), and oxygen uptake (O2, bottom) kinetics averaged from 4 transitions in 1 subject. HR in S2 trial showed clearly slower kinetics than in Nr and S1 trials. seemed to be slower in S1 and S2 than in Nr trial. O2 in S1 and S2 trial showed slight changes in delay.

Table 1 shows the differences in HR, , O₂, CO₂, and E between baseline (0-2 min of unloaded pedaling) and facial stimulation (2.5-3 min of stimulation in the S1 and S2 trials) or 2.5-3 min of the unloaded pedaling (Nr trial). Facial stimulation significantly decreased HR and and significantly increased CO₂ and E in the S1 and S2 trials. O₂ slightly but significantly decreased in the Nr trial.

View this table: [in this window] [in a new window]   Table 1.   Changes in O2, HR, , CO2, and E from baseline to facial stimulation

Nonlinear regression analysis showed that circulatory responses were significantly influenced by facial stimulation (Table 2). The HR and kinetics had a significantly longer delay and longer time constant in the S2 trial than in the Nr trial. The MRT in the S2 trial was 20 s longer than in the Nr and S1 trials. Facial stimulation did not affect baseline or gain in any variable.

The delay of was significantly longer in the S2 trial than in the Nr and S1 trials. The MRT was longer in the S1 and S2 trials than in the Nr trial. There was no significant difference in the time constant. There was no value for baseline and gain in , because was represented in relative values. The baseline was regarded as 0 and gain as 100 in all subjects.

The delay of O₂ was significantly longer in the S2 trial than in the Nr and S1 trials. However, the MRT and time constant did not show any significant differences (P = 0.38, P = 0.36 in ANOVA, respectively). There was no difference in the kinetic variables in CO₂ and E.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The main finding in our study is that increased vagal activity after the onset of stepwise exercise slows central circulatory responses but not O₂ response to the onset of work increase. HR time course clearly showed an effect of interruption in vagal activity withdrawal, i.e., vagal activation by facial stimulation just before and after the onset of exercise. O₂ MRTs were not significantly affected, even with slowed HR and kinetics by vagal activation at the onset of work increase. These findings disprove our hypothesis that central circulatory response plays a role in adjustment of O₂ kinetics, indicating that an adjustment of oxygen distribution to working muscles and/or some intramusclar oxidative processes maintains O₂ kinetics during disorder of vagal withdrawal response. It is suggested that only a change in vagal activity response cannot alter the O₂ response to exercise.

Oxygen transport to active muscles has been reported to be one of the factors regulating the O₂ kinetics at the onset of exercise (9, 11, 14, 18). O₂ kinetics were decelerated by reduced oxygen transport in the supine position (11) and by hypoxia (14, 18). Also, Hughson and Morrissey (9) reported delayed O₂ kinetics at the onset of second work increase from prior exercise. This has been explained by oxygen transport control by the ANS activity with faster vagal activity withdrawal than sympathetic activation (8). However, the effect of vagal activity on O₂ response has not been studied. Whereas sympathetic nervous and vagal activities regulate many organs reciprocally, not all the effect is reciprocal, e.g., the modulation of vasomotor tone. Thus the single role of central circulatory response on O₂ response has not been established. In the present study we found no difference in O₂ response between control and vagal disorder trials. It is concluded that central circulatory response cannot alter the O₂ response during vagal disorder.

We speculated that vagal activity withdrawal alters O₂ kinetics through central circulatory responses at the onset of exercise, because vagal activity is related to chronotropic heart control. The O₂ kinetics at the onset of stepwise work increase have been reported to be decelerated with sympathetic activity abolishment by a -adrenergic blocker (7, 20). In the present study, O₂ kinetics did not slow down despite slowed circulatory response. The discrepancy between previous studies and the present study by manipulating ANS activity may be explained by two factors. First, a -adrenergic blocker has affected both vasomotor and stroke volume responses in the previous studies. The arterioles of skeletal muscle are innervated by sympathetic cholinergic fibers (4). Splanchnic, renal, and cutaneous blood flow are also controlled by sympathetic noradrenergic nerve fibers. These organs are highly compliant and play an important role in regulating distribution of blood (21). The regulation of blood distribution through both muscular and nonmuscular organs is related to sympathetic nervous activity. Also, stroke volume is determined by filling pressure and sympathetic stimulation. Vagal activity cannot control the stroke volume enough. Consequently, the effect of vagal activity on regulations of the distribution of blood and is less in the present study. The manipulation of vagal activity could not control the sympathetic nervous activity so that distribution of blood and stroke volume responses might be modulated by the sympathetic nervous activity to maintain the O₂ response. On the other hand, in a previous study (7), the O₂ response was delayed despite no change in HR kinetics. This might be attributed to a change in responses of local blood distribution and stroke volume. The manipulation of sympathetic activity might change the blood distribution and stroke volume, resulting in a delay of O₂ response without altered HR response.

Second, -adrenergic blocker affects certain metabolic processes. Indeed, the preexercise O₂ value after -adrenergic administration has been reported to be slightly lower than that in placebo controls (7). In the present study, there was no change in the baseline or gain of O₂ in the S1 and S2 trials. We found no evidence that a facial stimulation itself alters oxygen utilization in tissue. Previous studies that suggested oxygen transport as a factor affecting the O₂ kinetics used manipulations such as supine position (11) and hypoxia (14, 18), which may change the local blood distribution, i.e., oxygen distribution, as well as central circulation. It is plausible that the difference between the present and previous studies is due to differences in sympathetic nervous control of stroke volume and blood distribution and in metabolic process.

Grassi et al. (3) reported that the bulk delivery of oxygen to the working muscle does not limit O₂ kinetics. They suggested the metabolic control as a rate-limiting step for O₂ response. Also, Yoshida et al. (27) reported faster O₂ response in the second bout of exercise than in the first bout, suggesting metabolic control. They indicate metabolic control as a rate-limiting step for O₂ response. On the other hand, our data suggest regulation mechanisms, that is, mechanisms not necessarily rate limiting for O₂ response. Namely, O₂ response might not change when a blood distribution response is faster than the normal condition. Thus the present study, which suggests the effect of blood distribution on O₂ response, does not contradict the metabolic control hypothesis. Hughson et al. (10) and Shoemaker et al. (24) suggested inadequate blood flow contributed to the delayed phase 2 of O₂ at the onset of exercise. The exercise mode in our study is different from these studies, in which arm (10) or knee extension exercises (24) were used. Nevertheless, regulation of blood distribution by sympathetic activity during delayed vagal withdrawal response is supported by these observations. Also, the role of sympathetic nervous activity in tissue O₂ has been reported (16). It has been found that -adrenergic tone contributes to the steady-state critical oxygen extraction during hypovolemia in dogs. Although the previous study was not about kinetics, a sympathetic contribution to the O₂ was clearly revealed. However, it has not been examined whether a manipulation of the sympathetic nervous activity makes the O₂ response faster than normal condition. This is necessary to certify the blood distribution as the "rate limiting" step for O₂ response.

It has been suggested that tachycardia induced by a dynamic exercise is mediated by rapid vagal activity withdrawal and slow sympathetic increase (15, 21). A study on HR variability revealed that vagal activity decreases to a trough below the ventilatory threshold, and sympathetic nervous activity increases above the ventilatory threshold (26). Plasma catecholamine concentration and muscle sympathetic nerve activity increase as a function of work rate above the lactate threshold (17, 23). Below the ventilatory threshold, vagal activity gradually decreases, while sympathetic activity shows only a slight increase. Namely, during normal conditions, vagal activity withdrawal plays a larger role in HR increase than sympathetic activation during moderate exercise below the ventilatory threshold such as the work intensity in this study. Our findings demonstrate that delayed vagal activity withdrawal alters the central circulatory responses below the ventilatory threshold, where vagal activity withdrawal might be an important factor in increasing HR.

The difference in MRT between Nr and S2 trials is clearly smaller than that in HR. The less effect of facial stimulation on than HR was clear, although the MRT should be compared carefully. Vagal activation can depress the HR increase because the sinoatrial node, which relates to chronotropic HR control, is innervated by vagal and sympathetic nerve fibers (4). On the other hand, stroke volume is determined by the amount of stretch of the ventricular wall and the degree of sympathetic nervous stimulation, which determine the force of ventricular contraction (4). Namely, vagal activity can alter response only chronotropically. The effect of vagal activity on stroke volume is less than the effect on HR. It is speculated that a faster rise in stroke volume by sympathetic nervous activity and the amount of stretch of the ventricular in S2 trial than Nr trial might contribute less MRT change in than in HR. These considerations suggest that the effect of delayed HR kinetics due to vagal activity is reduced in kinetics by a change in the force of ventricular contraction through the sympathetic activity. This regulation of stroke volume diminishes the effect of slow vagal activity withdrawal on kinetics. Furthermore, the change in kinetics by vagal disorder is lessened by blood distribution change to maintain O₂ kinetics. In other words, sympathetic nervous activity plays a major role in maintaining and O₂ responses through changes in ventricle contraction and vasomotor tone regulation during vagal withdrawal disorder, then compensates for the effect of delayed vagal withdrawal response.

Facial cold stimulation, known as a part of the diving reflex, induces bradycardia through trigeminal-vagal-cardiac pathways (1, 5, 6, 12). For noninvasive and nonpharmacological intervention to delay the withdrawal of vagal activity, the diving reflex without breath holding was used. Facial stimulation decreased HR by 10 beats/min and by 19.6%. This confirms the increased vagal activity by facial stimulation in the present study. Increased vagal activity had no effect on O₂ before the onset of exercise. However, the stimulation induces hyperpnea, indicated by 2 l/min increases in E. This hyperpnea would induce hypocapnia indicated by increased CO₂, which would affect the O₂ kinetics through the increase in blood pH, resulting in a change in the dissociation curve of hemoglobin. Yet, the comparison between the S1 and S2 trial is adequate, because both trials included facial cooling. The comparison in O₂ response between the S1 and S2 trial was similar to the comparison between the Nr and S2 trials.

Previous studies (7, 20) examining the effects of ANS activity on gas exchange kinetics found no significant changes in E kinetics, as in the present study. A change in vagal activity itself would not affect E kinetics. The altered CO₂ kinetics could change the E kinetics. It is difficult to interpret the E kinetics in this study, because of the carbon dioxide store. It has been reported that CO₂ kinetics were delayed during adrenergic blockade (7). This was explained by the delayed metabolic production due to the slow rise in O₂ and lower , resulting in a greater carbon dioxide store. Although rose more slowly in the S1 than in the Nr trial in the present study, CO₂ kinetics were not delayed. The effect of vagal activity on CO₂ kinetics remains to be clarified.

In a linear system, a change in baseline does not affect the kinetics. However, the baseline of HR affects the kinetics (9). This means that the kinetics is not a linear system (2). S1 trials were performed to evaluate whether the different baseline in S2 trials before the exercise onset altered the kinetics in Nr trial. The S1 trial showed similar MRT in HR response and O₂ to Nr. This validates the consideration above of no concern of HR in O₂ response during vagal disorder. However, was slower in the S1 than in the Nr trial. We have no basis for an argument on this finding. It could be speculated that vagal activity withdrawal with work increase and recovery from facial stimulation overshoot the HR response, whereas this effect might alter the sympathetic nervous response to work onset and alter the response.

In the present study, the cold stimulation was not exposed throughout the 5 min of exercise, because the subject felt cold or pain when stimulated on the face for >2 min. Thus, in the S2 trial, the cessation of facial stimulation may have affected the kinetics and model fitting. HR seemed to rise rapidly on removal of cold stimulation probably because of the rapid vagal activity withdrawal rather than sympathetic activity, because the HR increase by sympathetic activity may be slower than that by vagal withdrawal (15). Mathematically, the monocomponent fitting is thought to incompletely describe the HR behavior, because there were at least two components: HR increase by exercise input and decrease by vagal activation. However, a more suitable fitting, if any, would be complicated to use. Whether more accurate data could be obtained is questionable. Although a rapid change was seen when facial stimulation was ceased, the kinetics were delayed in this study. This finding does not contradict the delayed HR in exponential fitting.

We must consider the result of MRT carefully. Although repeated trials improved the signal-to-noise ratio of data, the data have large variability around the response because it was measured by the impedance method. This is the method frequently used to continuously measure during the transient phase from rest to cycle exercise. However, this method produces errors resulting from the effect of respiration movement (13). We cannot ignore this error around the response. Yet, one cannot deny the sympathetic contribution to the response, if the error had induced the slower response in the S2 than in the Nr trial. The delayed HR response during vagal stimulation should be regulated by the blood distribution and/or faster stroke volume response induced by sympathetic nervous activity to maintain O₂ response. We can propose no other mechanisms than the change in response of blood distribution and stroke volume to maintain O₂ response during delayed vagal regulation, even assuming that response is not delayed in the S2 compared with the Nr trial.

In summary, we examined the single effect of vagal withdrawal on central circulation and O₂ responses. The delayed vagal withdrawal slows down central circulatory responses, but not O₂ response to the onset of work increment. The maintenance of O₂ kinetics during acute bradycardia may either reflect the fact that some intramuscular process limits O₂ kinetics or alternatively that increased sympathetic vasoconstriction at some remote site defends exercising muscle blood flow. It was speculated that the sympathetic nervous activity, which has less effect on O₂ response under normal conditions, maintained the O₂ response during delayed vagal withdrawal. This study clearly discriminated between central and local circulatory responses and showed that the O₂ kinetics are altered by local blood distribution, rather than the central circulatory response during disorder of vagal withdrawal response to the onset of exercise.

Perspectives