Oxygen Uptake Kinetics During Exercise

REVIEW ARTICLE Sports Med 1999 May; 27 (5): 313-327 0112-1642/99/0005-0313/$07.50/0 Adis International Limited. All rights reserved. Oxygen Uptake Kinetics During Exercise Fan Xu and Edward C. Rhodes School of Human Kinetics, University of British Columbia, Vancouver, British Columbia, Canada Contents Abstract................................................... 313 1. Exercise Oxygen Uptake (V. O2) Kinetics at Different Intensities..................... 314 1.1 Moderate Exercise........................................... 315 1.2 Heavy Exercise............................................. 315 1.3 Severe Exercise............................................ 315 1.4 The Influence of V. O2 Slow Component on Oxygen (O2) Deficit................. 316 2. Mechanism of Exercise V. O2 Kinetics.................................... 316 2.1 Mechanism-Limiting Phase 2 V. O2 Kinetics.............................. 316 2.1.1 O2 Utilisation........................................... 316 2.1.2 O2 Delivery............................................ 317 2.2 Mechanism for V. O2 Slow Component................................ 320 2.2.1 Blood Lactate.......................................... 320 2.2.2 Epinephrine (Adrenaline)................................... 320 2.2.3 Ventilation............................................ 320 2.2.4 Body Temperature....................................... 321 2.2.5 Type IIb Fibre Recruitment.................................. 321 2.2.6 Physical Training and Slow Component of V. O2....................... 321 3. Other Factors Affecting V. O2 Kinetics................................... 322 3.1 Age................................................... 322 3.1.1 Children............................................. 322 3.1.2 Older People........................................... 323 3.2 Pathological Conditions........................................ 323 4. Conclusions.................................................. 324 Abstract The characteristics of oxygen uptake (V. O 2) kinetics differ with exercise intensity. When exercise is performed at a given work rate which is below lactate threshold (LT), V. O 2 increases exponentially to a steady-state level. Neither the slope of the increase in V. O 2 with respect to work rate nor the time constant of V. O 2 responses has been found to be a function of work rate within this domain, indicating a linear dynamic relationship between the V. O 2 and the work rate. However, some factors, such as physical training, age and pathological conditions can alter the V. O 2 kinetic responses at the onset of exercise. Regarding the control mechanism for exercise V. O 2 kinetics, 2 opposing hypotheses have been proposed. One of them suggests that the rate of the increase in V. O 2 at the onset of exercise is limited by the capacity of oxygen delivery to active muscle. The other suggests that the

314 Xu & Rhodes ability of the oxygen utilisation in exercising muscle acts as the rate-limiting step. This issue is still being debated. When exercise is performed at a work rate above LT, the V. O 2 kinetics become more complex. An additional component is developed after a few minutes of exercise. The slow component either delays the attainment of the steady-state V. O 2 or drives the V. O 2 to the maximum level, depending on exercise intensity. The magnitude of this slow component also depends on the duration of the exercise. The possible causes for the slow component of V. O 2 during heavy exercise include: (i) increases in blood lactate levels; (ii) increases in plasma epinephrine (adrenaline) levels; (iii) increased ventilatory work; (iv) elevation of body temperature; and (v) recruitment of type IIb fibres. Since 86% of the V. O 2 slow component is attributed to the exercising limbs, the major contributor is likely within the exercising muscle itself. During high intensity exercise an increase in the recruitment of low-efficiency type IIb fibres (the fibres involved in the slow component) can cause an increase in the oxygen cost of exercise. A change in the pattern of motor unit recruitment, and thus less activation of type IIb fibres, may also account for a large part of the reduction in the slow component of V. O 2 observed after physical training. Physical activity is an effective stimulus for increasing the activity of skeletal muscle and the cardiovascular system. At the onset of exercise, oxygen delivery from the atmosphere to active muscles increases in response to adenosine triphosphate (ATP) production in tissues. When the oxygen supply is insufficient, the generation of ATP has to rely on anaerobic glycolysis, which results in the formation of lactic acid. Associated with the increase in blood lactate levels, fatigue develops progressively and work tolerance is reduced. It has been suggested that oxygen uptake (V. O 2 ) measured at the mouth reflects the oxidative enzyme change in active tissues. [1] Thus, the rate of V. O 2 responses at the onset of exercise is a valuable index that reflects the adjustment of both systemic oxygen (O 2 ) transport and muscle metabolism. Since the first report of the exponential nature of gas exchange responses during a constant-load exercise, [2] exercise V. O 2 kinetics have been extensively studied. To date, the character of V. O 2 response to exercise at different intensities has been clearly described. [1,3-6] Some factors that affect the exercise V. O 2 kinetics, such as physical training, age and diseases have also been investigated. [7-12] However, some questions, such as the control mechanism, continue to be debated. This article reviews the nature of the V. O 2 kinetics at different intensity domains. The possible control mechanism(s) of the different components of the kinetics are discussed. 1. Exercise Oxygen Uptake (V. O 2 ) Kinetics at Different Intensities It has been well documented that the nature of V. O 2 response to exercise is a function of exercise intensity, which can be divided into 3 domains as suggested by Gaesser and Poole, [4] and Whipp. [6] The first domain is moderate exercise. The work rate, which does not induce a significant increase in blood lactate, is defined as moderate intensity. The upper limit of this domain is associated with the individual lactate threshold (LT) or anaerobic threshold (AT). The second domain is heavy exercise. When the exercise intensity is higher than LT, the rate of lactate production exceeds the rate of clearance. Blood lactate levels increase progressively, but lactate levels can be stabilised once again at a elevated new level if the work rate is below the maximum lactate steady-state (MLSS). Thus, the work rate within the range from LT to MLSS is considered as heavy intensity. The third domain is severe intensity. Work rate, which is above MLSS and results in a systemic increase in blood lactate during the exercise, is considered as severe intensity.

Oxygen Uptake Kinetics During Exercise 315 1.1 Moderate Exercise Three phases of the exercise V. O 2 kinetics can be identified in the moderate exercise domain (fig. 1). [4,6,13,14] Phase 1 represents the early fast increase in V. O 2, which is usually completed within first 15 to 25 seconds of exercise. This early response is more visible when the exercise is started from rest than from an 0 watt workload [6]. It suggests that in phase 1 the increase in V. O 2 is mainly attributed to the increase in cardiac output and thus pulmonary blood flow. [2,5,6] Any change in venous O 2 content from active muscle may not arrive at the lungs at this transient time and, therefore, does not affect the phase 1 V. O 2 kinetics. However, direct measurement by Casaburi et al. [15] demonstrated that mixed venous O 2 content began to decrease prior to the expected transient delay from the exercising leg to the lungs. Thus, the change in mixed venous O 2 content contributes part of the phase 1 V. O 2 response at the onset of exercise. Phase 2 response reflects the influence of muscle metabolic change on V. O 2 measured at the mouth. Following a short delay of phase 1, V. O 2 increases VO 2 (L/min) 4 3 2 1 1 2 2 Severe 3 Heavy Moderate 4 5 6 10 Time (min) Fig. 1. Schematic of the V. O 2 responses to constant-load exercises at different intensities. The numbers of 1, 2 and 3 indicate the 3 phases of V. O 2 responses. The shaded areas represent the slow component of V. O 2, which is above that predicted from subthreshold V. O 2 work rate relationship (adapted from Gaesser and Poole, [4] with permission). exponentially toward a steady-state level. Neither the slope of the increase in V. O 2 with respect to work rate nor the V. O 2 time constant (τv. O 2 ) has been found to be a function of work rate within this domain, indicating a linear dynamic relationship between the V. O 2 and the work rate. [6,13,14,16] Phase 3 is steady-state V. O 2 levels, which are reached after about 3 minutes. [3,4,6] During phase 3, V. O 2 increases linearly with work rate with a gain of 9 to 11 ml O 2 /watt/min during moderate exercise. [4,6,13,16] In general, the whole response of V. O 2 in the moderate domain can be described by a mathematical equation: V. O 2 (t) = V. O 2 (ss) (1 e (t δ)/τ ) Where V. O 2 (t) is the increase of V. O 2 at time t above the rest or 0 watt background, V. O 2 (ss) is the increase of V. O 2 at steady-state and δ is the time delay response. τ is time constant. [1,4,14] 1.2 Heavy Exercise When exercise intensity is above an individuals LT, the V. O 2 kinetics become more complex. Although the V. O 2 in phase 2 still increases exponentially, [17,18] an additional component is developed slowly after some minutes of exercise. [1,3,4,17] This slow component causes V. O 2 to increase progressively and delays the attainment of the steady-state level. [19,20] The attained value of V. O 2 in steadystate during heavy exercise is also greater than that predicated from the relationship between the work rate and the V. O 2 during moderate exercise (fig. 1). [4,6,13,18,21] The occurrence of the slow component is closely associated with the onset of lactate accumulation. The greater the increase in lactate, the greater the magnitude of the slower component. [6,13,18-21] The slow component is usually expressed as the V. O 2 difference between the sixth minute (or the end-exercise) and the third minute of exercise [ V. O 2 (6-3)]. [22] 1.3 Severe Exercise V. O 2 can not be stabilised during severe exercise, and continues to increase until the point of fatigue

316 Xu & Rhodes (fig. 1). [4,19-21] The maximum level of V. O 2 (V. O 2max ) is attained at the end of exercise. The slow component developed during severe exercise is much greater than that during heavy exercise. The magnitude of this component depends on the duration of the exercise. A value >1.0 L/min has been reported. [6,13,20] 1.4 The Influence of V. O 2 Slow Component on Oxygen (O 2 ) Deficit Following the onset of exercise even of moderate intensity, V. O 2 cannot increase immediately to the steady-state value. During the period of transition the energy demand has to be met partially from the other sources, such as the stores of muscle phosphocreatine (PCr), O 2 stores in the body and minor amounts converted from lactate production. [3] The O 2 equivalent according to the energy produced nonaerobically during exercise plus the decrease in venous O 2 store is defined as the O 2 deficit. [23] Studies have demonstrated that for moderate exercise the O 2 deficit increases with the exercise intensity, and equals the volume of O 2 consumption above the resting volume during the recovery period (O 2 debt). [3,18,24,25] It has also been found that both the V. O 2 -on and -off kinetic curves during moderate exercise are exponential with a similar time constant of about 30 seconds, but in opposite directions. [18] These results indicate that the depletion of muscle PCr and O 2 stores during the onphase are repaid totally in the postexercise period. [18] The O 2 requirement of exercise in this intensity domain is assumed to be the same as the steadystate V. O 2. In this case, the O 2 deficit can be calculated accurately. [1,4] When exercise is performed in the heavy intensity domain, the sustained elevation of blood lactate and the delayed development of the slow V. O 2 component make the calculation of the O 2 deficit more complex. It has been found that the τv. O 2 -off is significant shorter than the τv. O 2 -on. [18,26,27] As a consequence of the faster off-kinetics, the O 2 debt is less than the O 2 deficit. [18,24,25] Thus, the O 2 requirement is uncertain under these conditions even when the slow increase of V. O 2 may approach a final steady-state level. [1,3,4] The energy demand in heavy exercise is underestimated when extrapolated from a linear relationship between V. O 2 and sub-threshold work rates. [28] In contrast with above studies, Engelen et al. [29] reported a symmetry between the exercise and recovery kinetics of V. O 2 at both moderate and heavy exercise intensities. The reason(s) for the observed differences among the various studies is presently unclear. Further study of the relationship between V. O 2 -on and -off kinetics is needed. 2. Mechanism of Exercise V. O 2 Kinetics 2.1 Mechanism-Limiting Phase 2 V. O 2 Kinetics Although V. O 2 kinetics have been studied for almost a century, the question, regarding what mechanism(s) controls the rate of the V. O 2 responses at the onset of exercise, remains unanswered. Some researchers suggest that the V. O 2 kinetics are mainly determined by the rate of O 2 delivery to the active muscles. [30,31] Others suggest that the capacity of muscle O 2 utilisation sets the limit for the V. O 2 responses. [32] 2.1.1 O 2 Utilisation There is some evidence, obtained from both animal and human experiments, to support the hypothesis that the rate of adjustment of O 2 delivery is sufficient to meet the metabolic requirement of active muscle during submaximal exercise. For example, it has been found that cardiac output kinetics in normal people are faster than that of V. O 2 at the onset of exercise. [27,33,34] In addition, during the stimulation at 70% of V. O 2max the O 2 partial pressure (PaO 2 ) in isolated dog muscle is always above 2mm Hg, the critical level for maintaining aerobic ATP turnover. [35] Furthermore, Whipp and Mahler [32] investigated the control model of muscle respiration by using isolated frog muscle and found that the time constant of muscle O 2 consumption (V. O 2mus ) behaved in a first-order linear manner after a brief stimulation. The V. O 2mus kinetics in recovery was independent of the intensity of stimulation. The V. O 2mus changed in parallel with the level of creatine (Cr) and PCr when ATP levels

Oxygen Uptake Kinetics During Exercise 317 were constant. Thus, the kinetics of V. O 2mus reflect the dynamics of PCr in active muscle. [6,32] A study in humans using a biopsy technique has demonstrated that the PCr levels in active muscle decreases at the beginning of exercise at 65% of V. O 2max. After this initial decline, PCr levels remains unchanged. [36] It has also been shown that the amount of PCr depletion is significantly correlated with the mean response time of V. O 2 at the onset of submaximal exercise. [10] Since it is difficult to directly measure muscle O 2 utilisation in humans, the measure of the dynamic changes in PCr levels by nuclear magnetic resonance (NMR) spectroscopy is a common approach to estimate the metabolic rate in the active muscles. Most NMR studies [37-39] have demonstrated that the pattern of the PCr dynamic adjustments both at the onset and during the recovery after moderate exercise is first-order exponential. Recently, Barstow et al. [40] compared PCr and V. O 2 kinetics, which were measured during plantar flexion and the cycling exercises, respectively. The 2 exercise tasks were performed by the same participants. They found that the time constants for PCr kinetics was similar to that for V. O 2 kinetics in phase 2. They concluded that phase 2 V. O 2 kinetics reflect the kinetics of muscle PCr and thus the rate of muscle O 2 utilisation. Further support for this conclusion can be found in the data presented by McCreary et al. [41] Their data showed that PCr and phase 2 V. O 2 kinetics, measured simultaneously during the same moderate exercise task (ankle plantar flexion), occurred at similar rate. The problems associated with exercise with small muscle group are the small amplitude of the V. O 2 response and the high breathto-breath noise, which would affect the exponential curve fitting. The noise has to be minimised by averaging several repetitions of exercise transition. [42] In the study of McCreary et al., [41] for example, participants performed 12 square-wave repetitions over 4 to 5 sessions. In a different approach, Grassi and his colleagues [43] modified the thermodilution technique, which allowed them to measure leg blood flow (Q leg) during the non steady state period of exercise. Six trained participants in their study performed upright bicycle exercise at workloads less than LT. The alveolar V. O 2 (V. O 2alv ), Q leg and arteriovenous O 2 difference (a-vo 2 difference) were measured simultaneously. The V. O 2 in active legs (V. O 2leg ) was calculated by the product of the Q lleg and the a-vo 2 difference. They found that in phase 2 the V. O 2alv, Q leg and V. O 2leg increased exponentially at similar rates. However, in phase 1 (the first 10 to 15 seconds of the exercise), the V. O 2alv and Q leg kinetics were significantly faster than that of V. O 2leg while a-vo 2 difference actually decreased at the same time. Based on these results, the authors concluded that in the early phase of exercise, leg V. O 2 kinetics are not limited by the O 2 delivery to the exercising muscle. This conclusion is supported by the study of Williamson et al. [44] They applied positive pressure to the lower limbs to decrease leg perfusion during bicycle exercise and found that V. O 2 kinetics were not significantly different between the testing and control conditions. These investigators suggested that the O 2 transport is not a key factor in the control of the V. O 2 kinetics at the onset of moderate exercise in normal individuals. The body can increase cardiac output or O 2 extraction in order to compensate for a mild reduction in muscle blood flow. [44] In other studies, [45,46] it also showed that an increase in leg perfusion induced by prior heavy exercise did not alter the V. O 2 kinetics in subsequent moderate exercise. 2.1.2 O 2 Delivery The theory of O 2 transport limitation was initially proposed by Hughson and Morrissey. [30] They demonstrated that the V. O 2 kinetics in the transition from rest to exercise at 40% of AT was faster than the V. O 2 kinetics during the transition from 40 to 80% of AT. If V. O 2 kinetics had behaved as a linear system, the time constant of the response would have been unchanged whether or not the workload was increased from the same baseline. Thus, their result implies that other factors may be involved in the control of V. O 2 dynamic responses at the onset of exercise. In a further investigation, [47] they demonstrated that the slower V. O 2 kinetics during the transition from 40 to 80% of AT was linked to the

318 Xu & Rhodes slower heart rate responses. Since the stroke volume usually reaches the maximum level at about 40 to 50% of V. O 2max, the slower heart rate kinetics may reflect the slower response of cardiac output. The authors suggested that the slower heart rate kinetics during work-to-work transition may be due to a reduced influence of parasympathetic neural activity on the control of heart rate at the onset of exercise. Furthermore, Hughson and his colleagues [48,49] found that V. O 2 kinetics during exercise were slowed by administration of β-blockers in healthy participants. The slowed V. O 2 kinetics were secondary to the slower cardiac output responses in the presence of β-blockade compared with the value in control. Their results are in agreement with the observation of other investigators. [50] Along with cardiac output, the sufficient O 2 supply to exercising muscles is also dependent on the local blood distribution and the arterial O 2 content. It has been found that hypoxia, which reduces arterial partial O 2 pressure, slows V. O 2 kinetics and results in a greater O 2 deficit compared with the normal breathing. [12,29,51] In order to ensure that exercise performed during both air breathing and hypoxic gas breathing would be within the moderate intensity domain, the work rate used for both conditions was chosen to require a steady-state V. O 2 below the hypoxic AT in all participants. The heart rate at rest and during exercise was higher in participants who were hypoxia than in controls, but the rate of the increase in heart rate was significantly slower in participants with hypoxia. [12,51] The slowed V. O 2 and heart rate responses observed in hypoxic breathing are consistent with the results reported by Hughson and Morrissey [30] as previously noted. V. O 2 kinetics during exercise are also reported to be slower in supine than in upright positions. [52,53] Gravitational factors, which reduce the blood flow to the exercise legs, may contribute to the slowed response rate during supine exercise. Recently, the relationship between muscle blood flow and muscle O 2 uptake during exercise has been investigated by Grassi et al. [43] and Hughson et al., [54] respectively. Although different techniques were employed in the 2 studies, both demonstrated that V. O 2mus kinetics are closely matched by the rate of the increase in muscle blood flow. However, there is a major difference between the 2 studies regarding the early responses of V. O 2mus and muscle blood flow during exercise. Grassi et al. [43] measured the exercising leg blood flow with an invasive thermodilution method and showed that there were 2-phase responses for both V. O 2mus and blood flow in the transition period of exercise. In phase 1, muscle blood flow increased more rapidly than did V. O 2mus. In contrast, Hughson et al. [54] calculated blood flow to the forearm from the product of blood velocity and cross sectional area obtained with Doppler techniques. They did not find the biphasic changes in either muscle blood flow or V. O 2mus in unsteadystate periods of exercise. Whether the discrepancy between the 2 studies is due to the method of blood flow measurement or to the muscle mass involved in exercise is unknown at this time. To date, studies, using a positive intervention to examine the theory of O 2 transport limitation, have produced mixed results. Some studies support the concept that V. O 2 kinetics are limited by O 2 delivery, [54-57] while others do not. [24,25,45,46,58] Linnarsson et al. [57] found that exercise during hyperoxia resulted in a smaller O 2 deficit and less depletion of PCr than that with air breathing. Hughson and Imman [55] reported a faster rate of V. O 2 responses during arm exercise when the circulation in nonexercise legs was occluded. In another study, Hughson et al. [56] demonstrated that the slower kinetics during supine exercise were improved by applying negative pressure to lower limbs. Recently, more direct evidence has been provided in the study of Hughson et al. [54] They investigated the forearm blood flow and muscle O 2 uptake in 10 individuals during rhythmic handgrip exercises with the arm either above or below the heart level and found that the V. O 2mus increased proportionally to the increase in blood flow. The flow and the V. O 2mus kinetics at the onset of exercise were significantly faster in the position below than above heart levels. These data suggest that the O 2 delivery system can play an

Oxygen Uptake Kinetics During Exercise 319 important role in the control of the V. O 2 kinetics at the onset of exercise. In contrast with above studies, Ren et al. [24] and Sahlin et al. [25] compared the O 2 deficit during constant-load exercises at 2 conditions in which the rest-to-work transition was completed either directly or gradually in a stepwise manner. They failed to demonstrate any significant difference in O 2 deficit across a range of work rates from 20 to 80% of V. O 2max between the 2 conditions. Since the O 2 transport system has more time for adaptation during exercise with slow transition, a smaller O 2 deficit should be expected in this condition. Their results are contrary to the concept that the increase in V. O 2 during exercise is limited by the O 2 transport capacity. Yoshida et al. [58] designed another experiment to examine the theory of O 2 transport limitation. The experiment consisted of 3 repeated 1-leg exercises at a constant work rate. Each lasted for 5 minutes and was followed by 5 minutes rest. In the third repetition, the exercise was performed by the previous non-working leg. They hypothesised that when exercise is switched from one leg to another, the PCr and O 2 stores in the previous non-working leg are still in the resting state and, therefore, any increase in V. O 2 kinetics will only be attributed to the improvement of central circulatory function induced by prior exercise bout. However, their data did not support the hypothesis. They found that there was a faster increase in both V. O 2 and cardiac output at the onset of the second repetition than at the onset of the first. However, when the exercise was switched to the non-working leg in the third repetition, the V. O 2 kinetics became slower while the cardiac output and heart rate were still at similar levels compared with the second bout. An alternative explanation of the data is that the residual vasodilation in the previous working leg may offset the effect of increased cardiac output on V. O 2 kinetics at the onset of third repetition. It is well documented that mild lactic acidosis causes vasodilatation and facilitates haemoglobin dissociation. [59,60] These effects would increase O 2 delivery to the exercising muscle. [59,60] However, Gerbino et al., [45] and MacDonald et al. [46] recently demonstrated that the rate of the increase in V. O 2 during moderate exercise, but not during heavy exercise, was independent of prior blood lactate levels. Muscle blood flow, however, was not assessed in their study. Exercise training can be considered as another positive intervention. The effect of long term endurance training on exercise V. O 2 kinetics has been established in several investigations. [61-64] However, the mechanism for the training-induced adaptation of V. O 2 kinetics has not been fully understood. It is well known that long term endurance training can improve cardiovascular function and muscle oxidative capacity, as evidenced by the increase in cardiac output, [65] and the increases in mitochondrial density and aerobic enzyme activity. [66] These factors might have contributed to the acceleration of V. O 2 kinetics with training. Recently, Phillips et al. [10] demonstrated that an improvement of V. O 2 kinetics was observed only few days after training. At that time, there was no significant change in mitochondrial enzyme activity. A further improvement of the V. O 2 kinetics and a significant increase in the mitochondrial enzyme activity were evident with the extension of the training. Thus, it seems that at least in the early period of the training, muscle oxidative capacity does not play a role for the improvement of V. O 2 kinetics in phase 2. Because of technique difficulty the noninvasive measurements of muscle PCr and blood flow can only be performed in small muscle groups during exercise. Since the physiological stress produced by small muscle exercise differs with that of large muscle exercise, extrapolation of the results obtained from small muscle exercise must be done with caution. [31] Slower V. O 2 kinetics have been reported during arm compared with leg exercises. The reasons for the difference are likely to reflect the extra muscle activity employed for stabilisation and the transient increase in lactate at the onset of arm exercise. [26,67] At the present time, there is not a definite answer for the mechanism(s) limiting V. O 2 kinetics during exercise.

320 Xu & Rhodes 2.2 Mechanism for V. O 2 Slow Component For exercise above LT, the slow component of V. O 2 may reflect an inadequate O 2 supply in the active muscles, which results in an increase in anaerobic ATP regeneration and the accumulation of blood lactate. [45,46,60,68] However, the mechanism of the slow component of V. O 2 has not been determined. Some potential mechanisms have been investigated by several researchers and are outlined in the following sections. 2.2.1 Blood Lactate Numerous studies have demonstrated that the slow component of V. O 2 is closely linked with blood lactate levels. The slow component of V. O 2 occurs at approximately same time as the increase in lactate levels. [3,6,19] The magnitude of the V. O 2 slow component and the level of blood lactate is highly correlated. [6,19-21] Recently, Barstow [3] provided indirect evidence against a direct coupling of lactate production and the onset of the slow component V. O 2. He demonstrated that the slow component V. O 2 was initiated 80 to 110 seconds into exercise. At that time, however, femoral vein blood lactate levels had already been elevated. Brooks [69] suggests that more than 70% of the lactate formed during sustained exercise is oxidised within active muscles, while the remainder is removed through hepatic gluconeogensis or by other means. The biochemical process of the lactate oxidation may cause a slow increase in O 2 consumption during exercise. However, some studies have demonstrated that the V. O 2 during exercise is unaffected when blood lactate levels are elevated to a considerable level by either infusing L-(+)-lactate in dog gastrocnemius [68,70] or infusing epinephrine in humans. [71] The study by Engelen et al. [29] also showed that a 57% elevation of blood lactate levels induced by hypoxic breathing during heavy exercise affected neither the time constant nor the amplitudes of the slow component V. O 2. It appears that lactate itself is not the cause of the slow V. O 2 component during exercise. Recently, Stringer et al. [72] and Wasserman et al. [60] suggested that during intense exercise, lactic acidosis could play an essential role in promoting oxyhaemoglobin dissociation which, in turn, increases O 2 transport to muscle and maintains the PaO 2 above critical levels. Stringer et al. [72] also suggested that the increase in O 2 delivery induced by lactic acidosis accounted for 62% of the slow component of V. O 2. Further support for this hypothesis is provided by the study of Belardinelli et al., [73] in which they found a high correlation between the magnitude of the slow component and the amount of the decrease in oxyhaemoglogin saturation in vastus lateralis muscle (measured by near-infrared spectroscopy) during cycling exercise. However, Gaesser and Poole, [4] and Whipp [1] argued that this hypothesis is not sufficient to explain why during heavy exercise that V. O 2 exceeds the value predicted from the submaximal V. O 2 -work rate relationship. 2.2.2 Epinephrine (Adrenaline) Adrenaline is another possible mechanism behind the V. O 2 slow component. Plasma adrenaline levels increase during exercise in a similar manner as blood lactate. The thresholds of lactate and epinephrine occur at approximately similar work rates. [74] It has been reported that epinephrine infusion in individuals at rest increases V. O 2 by about 20%. [75] The increase in V. O 2 is due to the integrated influence of adrenaline on the circulatory, respiratory and metabolic systems. However, unlike the significant effect of adrenaline on basal metabolic rate, studies did not demonstrate an increase in V. O 2 when plasma adrenaline levels were elevated significantly (4- to 6-fold higher than the value in controls) by the epinephrine infusion during exercise. [4,71,76] These results imply that adrenaline is not responsible for the slow component of V. O 2 during exercise. 2.2.3 Ventilation High intensity exercise is accompanied by hyperventilation which serves to maintain effective gas exchange in the lungs. The increased respiratory work may be a mechanisms of the slow component of V. O 2. However, the estimates of the O 2 cost of ventilation vary considerably between studies. [77,78] Shephard, [78] for example, reported that the O 2 cost for unit increase in ventilation was 4.3 ml when the ventilation was above 89 L/min,

Oxygen Uptake Kinetics During Exercise 321 whereas Aaron et al., [77] recently reported that the O 2 cost was 3 ml O 2 /L for ventilation in the range from 117 to 147 L/min. Thus it is difficult to determine the true energy cost of ventilatory work during exercise. Data from Aaron et al, [77] Gaesser and Poole [4] and Womack et al. [71] suggest that the O 2 cost of the increase in ventilation in their studies could account 18 to 23% of the total V. O 2 slow component. 2.2.4 Body Temperature During exercise increased metabolic activity within working muscles will result in an increase in muscle temperature. The increase in body temperature may also lead to an increase in O 2 consumption due to the Q 10 effect (the effect of temperature on reaction rate where the ratio of the reaction rates at 2 temperatures 10 K apart is the Q 10 ). [6] A study with isolated mitochondria has shown that when the muscle temperature is increased by 6 C, the efficiency of mitochondrial respiration decreases by 20%. [79] However, studies of humans have produced different results. V. O 2 during submaximal exercise has been reported to be either increased [80,81] or unchanged [82] by the elevation of body temperature. The studies, which reported an increase in V. O 2 with body temperature, also demonstrated that V. O 2 and body temperature did not change with the same time course during exercise. The increase in V. O 2 lagged behind the increase in body temperature. Recent data presented by Koga et al. [83] demonstrated directly that a rise of 3.5 C in muscle temperature without a concomitant change in body core temperature prior to exercise did not affect the slow component of V. O 2 during heavy exercise. Therefore, it seems that some factors other than the elevation of body temperature may have contributed to the slow increase in V. O 2 during exercise. 2.2.5 Type IIb Fibre Recruitment A study by Poole et al. [68] has suggested that the slow component of V. O 2 is associated with some factors within the exercise muscle itself. It is well known that skeletal muscles contain 2 basic types of fibres, type I and type II. The type II fibres can be further divided into type IIa and type IIb according to their oxidative capacity. The aerobic potential of type IIb fibres is much lower than that of type I fibres. [79,84] Type IIb fibres consume more O 2 than type I fibres at a given force production. [84] Studies have shown that the recruitment of type IIb fibres increases in proportion to the exercise intensity. [85] Measurements using the integrated electromyogram (iemg) in working muscle also demonstrated a positive correlation between the increase in iemg signals and the increase in V. O 2 during a constant-load exercise. [86] This result suggests that the increase in the recruitment of motor units may contribute to the slow component of V. O 2 although it does not adequately indicate that the type IIb fibres have been recruited at this time. Xu and Montgomery [87] examined the effect of prolonged running on V. O 2 at subsequent exercise. The participants (n = 14) performed two 90-minute runs on a 400-metre track at velocities of 65 and 80% of V. O 2max. Before and after each 90-minute run, steady-state V. O 2 during moderate treadmill exercise was measured. When the post-test was conducted after each 90-minute run, there was a significant increase in V. O 2. The increase in V. O 2 after the 90-minute run at 80% of V. O 2max was greater than that after the run at 65% of V. O 2max. The authors suggest that muscle fatigue induced by prolonged exercise, and thereby an increase in the recruitment of fibres for force maintenance, may have been one of the factors attributing to the increase in V. O 2 on the post-test. Recently, Barstow et al. [88] reported a negative correlation between the magnitude of V. O 2 slow component and the percentages of type I fibres in exercising muscles (vastus lateralis), indicating that fibre type distribution has a great influence on the slow component of V. O 2 during exercise. Based on these results, it appears that the recruitment of type IIb fibres may be a major contributor for the slow increase in V. O 2 during heavy exercise. [1,3,4,70] 2.2.6 Physical Training and Slow Component of V. O 2 An alternative way to investigate the mechanism of V. O 2 slow component is to examine the training responses of its potential contributors. Several studies [71,89,90] have demonstrated that 6 to 8 weeks of

322 Xu & Rhodes endurance training results in a significant reduction in the slow component of V. O 2 when the post-training exercise is performed at the same absolute work rate as the pretraining condition. However, the relative intensity (expressed as %V. O 2max or %LT) for the post-training exercise has been altered as a consequence of the increases in both V. O 2max and LT. Recently, Womack et al. [71] investigated the time course of the training adaptation in V. O 2 slow component. Participants in their study trained 4 days per week for 6 weeks on a cycle ergometer. The exercise V. O 2 for work rate corresponding to 60% of the difference between the pretraining LT and V. O 2max was examined before and after each of the 6 weeks of training. They found that V. O 2 slow component (expressed as the difference of V. O 2 between the end and the third minute of exercise) was reduced significantly after 2 weeks of training. After that, no further reduction was observed over the remaining training period. Furthermore, the training had no effect on the V. O 2 at the third minute of exercise. The V. O 2 at the third minute was similar before and after each week of training. Therefore, the reduction in the V. O 2 slow component resulted from the reduction in the end-exercise V. O 2. Investigators have found that the adaptation of V. O 2 slow component with training was accompanied by decreases in blood lactate, ventilation per minute (V. E), plasma adrenaline level and less of an increase in rectal temperature. [90] Statistical analysis has shown that the reduction in V. O 2 slow component was significantly correlated with the reductions in lactate (r = 0.64) and V. E (r = 0.51), but not correlated with the decreases in plasma adrenaline level (r = 0.13) or rectal temperature (r = 0.15). [90] However, the data presented by Womack et al. [71] suggest that neither lactate nor V. E are the major candidate for the training-induced reduction in V. O 2 slow component. First, their study demonstrated that the pattern of adaptation in blood lactate levels differed with that of V. O 2 slow component. Following the initial reductions in both lactate levels and V. O 2 slow component at week 2 of training, lactate decreased continuously throughout the remaining 4 weeks without further changes in V. O 2. Finally, in order to determine the relative contribution of adrenaline to the training-induced reduction in V. O 2 slow component, participants performed an additional constant-load exercise after a few days of the test at week 6 (using the same work rate as previous tests), during which epinephrine was infused. The epinephrine infusion produced a 16-fold increase in the end-exercise plasma epinephrine concentration compared with the value at the week-6 test. The end-exercise blood lactate and V. E during the epinephrine infusion test were also increased by 2.4 mmol/l and 7 L/min, respectively. However, the V. O 2 expressed as either the slow component or the end-exercise values was not different between the epinephrine infusion and the week-6 tests. Therefore, the reduction of V. O 2 slow component with training is attributed to factors other than the decreases in lactate, V. E and adrenaline. Gaesser and Poole [4] suggested that endurance training may change the pattern of motor unit recruitment, and thus less type IIb fibres are recruited during posttraining exercise. This may contribute to the reduction in V. O 2 slow component with training. However, the precise relationship between type IIb fibre recruitment and the slow component of V. O 2 has not been established. More studies are needed in this area. 3. Other Factors Affecting V. O 2 Kinetics 3.1 Age 3.1.1 Children Compared with adults, children have lower bodyweight, smaller size, lower V. O 2max and higher basic metabolic rates. [12,91] However, these factors do not appear to affect V. O 2 kinetics during moderate exercise. Dynamic responses of V. O 2 during moderate exercise were the same in children and adults despite slower kinetics for heart rate in the children. [91] Even under hypoxic conditions, the V. O 2 kinetics were affected in children and adults to the same degree. [12] This suggests that the control mechanism for the V. O 2 kinetics during moderate exercise is same between children and adults. [12] However, there are some differences for V. O 2 res-

Oxygen Uptake Kinetics During Exercise 323 ponse during heavy exercise between children and adults. Armon et al. [92] found that almost 50% of the children in their study did not develop the slow component of V. O 2 during high-intensity exercise. The V. O 2 response curve for these children could be adequately described by a single exponential model. For the children who experienced the slow component of V. O 2, the magnitude of the slow component expressed in both L/min and ml/kg/min was smaller than in the adults, and was not increased with exercise intensity. However, over a range of work rates, the O 2 cost of exercise (expressed as V. O 2 /watt) was greater in children than adults. The mechanisms for these differences between the 2 groups are not fully understood. It has been known that children have a lower capacity of anaerobic glycolysis than adults, and hence they are unable to produce a high level of lactate during high intensity exercise. [93] The absence of the V. O 2 slow component in children may relate to their lower ability to produce lactate. [92] However, many studies [3,4,6,70,71] have indicated that the blood lactate level is not the key factor for the slow component of V. O 2 in adults, so the relationship between the reduced V. O 2 slow component and the reduced blood lactate responses in children is still unknown. 3.1.2 Older People Aging is associated with decline in cardiorespiratory function and muscle oxidative capacity. [65,94] These factors will influence V. O 2 kinetics during moderate exercise. Studies have demonstrated that phase 2 V. O 2 kinetics at the onset of cycling exercise are slowed with age. [7,8] The slower V. O 2 kinetics in older people are associated with slower heart rate responses. [95] This implies that the rate of V. O 2 adjustment at the onset of exercise may be limited by the decreased central circulatory function in older people. However, unlike in cycling exercise, V. O 2 and heart rate kinetics have been reported to be similar between older and young people during ankle plantar flexion. [95] This finding is consistent with the concept that the central O 2 transport capacity is not a limiting factor in small-muscle exercise. [31] On the other hand, as the major muscle group used in walking, the ankle plantar flexors may be used more frequently than other muscles in the routine activities of older people. The oxidative potential, and thus the capacity of O 2 utilisation in this muscle group, may be maintained to a better extent than in other muscle groups. [95] This may also contribute to the faster V. O 2 kinetics observed during ankle plantar flexion. Multiple linear regression analysis [8] has demonstrated that a large part of the variability in V. O 2 kinetics among individuals can be accounted for by individual fitness levels. The changes in lifestyle, especially the reduction in daily activities, may cause the decline in fitness level in older people. A study by Babcock et al. [96] has shown that the kinetics of V. O 2 during cycling exercise were faster after 24- week endurance training in older males when compared with the pretraining condition. The posttraining τv. O 2 in older individuals is similar to that reported in untrained individuals. It is well known that long term endurance training can produce significant cardiorespiratory and skeletal muscle adaptations in both the young and in adults. [65,97] However, little is known regarding their relative contributions to the improvement of V. O 2 kinetics with training. Recently, Phillips et al. [10] reported that the improvement in V. O 2 kinetics in healthy young people occurs more quickly after training, and before changes in mitochondrial enzyme activity. It remains unclear whether this finding holds true for older individuals. To date, little is known about the effect of endurance training on V. O 2 slow component during heavy exercise in older people. Belman and Gaesser [98] reported a reduction in end-exercise V. O 2 after an 8-week training programme in elderly men and women. However, the slow component of V. O 2 was not assessed in their study. 3.2 Pathological Conditions Pathological conditions that affect the cardiopulmonary function and the muscle metabolic capacity would be expected to alter V. O 2 kinetics. As compared with normal individuals, slower phase 2 V. O 2 responses and smaller increases in phase 1 V. O 2 have been demonstrated in patients with congenital heart disease, [11] and also in patients with chronic

324 Xu & Rhodes obstructive pulmonary disease. [99] The high pulmonary vascular resistance, which reduces the capacity for increasing pulmonary blood flow at the onset of exercise, is the major cause for the reduction in phase 1 V. O 2 responses in these patients. [13] On the other hand, the prolonged phase 2 kinetics probably reflect the combination of abnormal pulmonary blood flow and reduced muscle oxidative capacity (as a consequence of long term deconditioning). [3,10,99] Zhang et al. [100] suggest that the exercise V. O 2 kinetics in heart failure patients and normal individuals is affected by the same mechanism without the consideration of any heart disease or the fitness level. Thus, the noninvasive measurement of V. O 2 kinetics during submaximal work provides a useful tool in the assessment of working tolerance in patients with cardiopulmonary diseases. 4. Conclusions The nature of V. O 2 kinetics is a function of exercise intensity. During exercise at a work rate below LT, V. O 2 increases exponentially toward a steadystate level within 3 minutes. The τv. O 2 is relative unchanged with exercise intensity. For exercise at intensities above LT, the V. O 2 kinetics become more complex. An additional component is developed slowly and delays the attainment of steady-state. Moreover, the steady-state V. O 2, when attained, is greater than that predicated from the sub-lt work- V. O 2 relationship. The available evidence suggests that the rate of V. O 2 adjustment at the onset of exercise reflects the capacity of central O 2 delivery and muscle O 2 utilisation. However, it is still unclear which mechanism plays a predominant role in the control of exercise V. O 2 kinetics. To further gain insight of the nature of V. O 2 kinetics, it is necessary to develop more sensitive technology to measure the time course of muscle blood flow responses, and the rate of muscle metabolism during exercise, especially exercises involving large muscle groups. The physiological factors that determine the V. O 2 slow component during heavy exercise remain to be resolved. The magnitude of the slow component is highly correlated with the increase in blood lactate levels, both before and after endurance training. However, there is no proof to indicate the cause and effect relationship between the 2 variables. Other factors, such as the increased ventilatory work and the elevation of body temperature, have been suggested to only account for a minor part of the V. O 2 slow component. Another potential contributor for the slow component of V. O 2 is the recruitment of type IIb fibres. These fibres have lower efficiency of O 2 utilisation than type I fibres. During high intensity exercise the reduction in ph accompanying the lactic acidosis may impair muscle contractile function such that more type IIb fibres are recruited to maintain a constant power output. Thus, the O 2 cost of exercise may increase. The alteration of motor unit recruitment pattern and, thereby, less activation of type IIb fibres may also account for a large portion of the reduction in slow component of V. O 2 observed after endurance training. However, because of the lack of adequate evidence to indicate that the recruitment of type IIb fibres, the increase in blood lactate levels or the increase in V. O 2 occur at the same work rate, this hypothesis has not been established. In children, the V. O 2 responses rate at the onset of exercise below LT are not significantly different from those in adults. However, during exercise at work rate above LT children have demonstrated smaller or no slow component of V. O 2 as compared with adults. The reason for the absence of the slow component in children is unknown at the present time. On the other hand, V. O 2 kinetics in older people have been demonstrated to be slower than the young. The slowed V. O 2 responses at the onset of exercise may be associated with age-related changes in physiological function. Exercise training can improve the V. O 2 kinetics in both old and young. However, the mechanism for the training adaptation is not fully understood. These unsolved questions require further research. References 1. Whipp BJ. The slow component of O 2 uptake kinetics during heavy exercise. Med Sci Sports Exerc 1994; 26: 1319-26 2. Krogh A, Lindhard J. The regulation of respiration and circulation during the initial stages of muscular work. J Physiol (Lond) 1913; 47: 112-36