INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

SKELETAL MUSCLE BLOOD FLOW increases markedly at initiation of voluntary contractions to make the oxygen delivery meet the regionally elevated metabolic demands (17, 22, 28, 29). However, despite more than a century of research, the exact time course of the increase in muscle blood flow at onset of intense dynamic exercise in humans is still unknown, as the techniques for measuring blood flow have either had unsatisfactory temporal resolution or have at best been semiquantitative (2, 4, 5, 8, 11, 27). An ultrasound Doppler has previously been utilized at the onset of, and during, intermittent static contractions of the quadriceps muscle to detect, on a beat-by-beat basis, a rapid increase in blood velocity (22, 28, 29). However, the inherent variability of this sampling procedure, due to the temporal dissociation between the kicking duty cycle (muscular contraction-relaxation phases) and the cardiac cycle (28, 29), limits the possibility of following the precise transitional changes in blood velocity and flow. In addition, these studies were not designed to investigate the exact time course and the magnitude of increase in blood flow in relation to the intensity of dynamic work. Such data are of value not only to examine how quickly regional blood flow and thus oxygen delivery matches the energy demand after onset of exercise but also to identify the factors that may induce the hyperemia, and especially their time restraints. This is now possible by sampling the blood velocity continuously with an ultrasound Doppler and estimating the blood flow in relation to each kicking duty cycle (16).

It has been suggested that the muscle pump during muscular contractions promotes muscle blood flow by squeezing the blood out of the venous capacitance vessels (15), thereby inducing a lowering of the venous pressure (P_v) and a gain in the pressure gradient (P = P_a  P_v, where P_a is the arterial pressure) across the vascular bed (6), a pressure gradient that, in this context, along with the vascular conductance (VC), determines the arterial inflow (_a = P × VC). This pressure gradient has furthermore been suggested to be enhanced by a negative venular pressure, induced on muscle relaxation by the pulling open of the veins attached to the surrounding tissue (13). Thus, indirectly, the muscle pump also enhances the venous return by promoting the propulsion of blood out of the muscle (3, 4, 6, 7, 13, 15, 21, 24, 26). Moreover, the extent of the effect of the muscle pump depends on the force, frequency, and duration of the muscle contractions. The force determines the compression of the vasculature and degree of venous emptying, whereas the frequency and duration determine the extent of venous refilling (13). However, in this context it should be noted that the muscle contractions may also impose a significant mechanical hindrance to the arterial inflow, where the intramuscular pressure may rise to a level at which the resistance vessels collapse. Recently, it was also proposed that the muscle pump may promote a part of the sudden initial rise in blood flow at onset of exercise, before a further metabolic vasodilatation (21, 26). However, the precise temporal relationship between the arterial blood velocity (inflow) and the variations in intramuscular, arterial, and venous blood pressure has not been described.

To explore these questions, two different experiments were performed. In the first, the temporal course of the initial rise in blood flow was characterized at onset of dynamic muscle contractions at different intensities. In the second, the role and potency of the muscle mechanical factors were studied at onset of exercise at one of these intensities by investigating the temporal relationship between the arterial blood velocity and the fluctuations in intramuscular, arterial, and venous pressure.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Subjects

Experimental Design

Protocol I: Temporal course of changes in blood flow at onset of dynamic exercise. In the sitting position the subjects (n = 7) performed repeated 3-min bouts of one-legged, dynamic knee-extensor exercise (60 cpm) at 10 W (n = 7), 30 W (n = 7), 50 W (n = 5), and 70 W (n = 2), corresponding to 14.1 ± 3.1, 42.2 ± 9.4, 66.9 ± 15.6, and 74.0 ± 5.5% of peak power output. The blood velocity was continuously measured at rest, at onset of passive leg movement, and in the transition to, as well as during, voluntary exercise at each workload. Passive leg movement was included to study the muscle mechanical factors alone, compared with additional metabolic components during voluntary exercise, as well as to control an instantaneous and reproducible start at a fixed contraction rhythm of 60 cpm at onset of the voluntary contractions. This was done by attaching the subject's leg to the knee-extensor lever arm and moving it up to 60 revolutions/min. Each work bout was separated by at least 30 min of rest until the blood velocity spectra and flow had normalized to resting control level. The work bouts were performed in incremental order to avoid a sustained effect of the highest on the lowest bouts. The subjects had before the first incremental exercise bout 1 min of passive leg movement and a warmup for 15-20 min in the knee-extensor ergometer, followed by >30 min of rest. Thus all bouts were preceded by another exercise bout.

Protocol II: Temporal relationship between blood velocity and intramuscular, arterial, and venous pressure at onset of dynamic exercise. Intramuscular pressure was measured with a Millar Micro-Tip catheter transducer (2-Fr, diffused semiconductor, model SPC-320, Millar Instruments, Houston, TX) inserted under sterile conditions via a venflon in the quadriceps muscle of the subjects (n = 7), about one-half the distance between the pubic bone and patella. The insertion was performed at an angle of ~30-45° with respect to the skin and with a length giving a location and fixed depth approximately central in the muscle. The signal was amplified by a Millar transducer control unit (Millar Instruments). The intramuscular pressure transducer was specifically calibrated in relation to a water column of different heights, and the measured pressure related to this external pressure. Intra-arterial and venous blood pressure (Dialogue 2000, Danica Elektronik, Copenhagen, Denmark) were measured via catheters (20G, Ohmeda, Wiltshire, UK) placed in the proximal direction in the femoral artery and vein, 2-5 cm below the inguinal ligament. The knee-extensor force (strain gauge) and the femoral arterial blood velocity determined by the ultrasound Doppler (model CFM 800, Vingmed Sound, Horten, Norway) were measured simultaneously and continuously. The subjects performed the exercise with their thighs in the horizontal position and their upper body slightly bent upward (160°). The exercise began by five to seven passive leg movements at 60 revolutions/min. They thereafter started the voluntary exercise at a workload of 47.1 ± 1.8 W (65.4 ± 3.6 % of peak power output), at which they exercised for 5-8 min. The load was thereafter increased with 5- to 10-W increments every 30 s up to their maximum workload of 72.8 ± 3.8 W.

Instrumentation and Methodological Considerations

The femoral artery was insonated (direction of the ultrasound waves at the site of measurement) at a fixed perpendicular angle. Two-dimensional longitudinal images were captured and stored, with 25 frames/s in the image buffer and on magneto-optical discs. The diameter was subsequently determined along the central path of the ultrasound beam where the best spatial resolution is achieved. The systolic and diastolic diameters were separately measured over the cardiac cycle guided by the electrocardiogram (ECG). A diameter [D_{(systole+diastole)}] based on the relative time periods of the systolic (one-third) and diastolic (two-thirds) blood-pressure phases was assumed to be the most representative diameter size and was utilized to determine the cross-sectional area, A = r², where r is the radius of the vessel (16).

The blood velocity was measured with the Doppler probe stabilized in a fixed position at an insonation angle as low as possible, during simultaneous vessel visualization. This procedure allows centering and size adjustment of the sample volume in relation to the vessel diameter, so that the sample volume covers the width of the vessel and the parabolic velocity profile, and also allows direct correction for the angle of insonation by positioning the rotatable axis of the sample volume parallel with the direction of the flow and vascular walls. It also enables optimization of the Doppler velocity recording on the basis of a direct feedback from the sample volume size and positioning in the vessel as well as the Doppler signal intensity (16). Slight turbulence and velocity irregularities occurring at the pulsating vascular walls were reduced by low-velocity rejection filtration.

The Doppler blood velocity spectra and the strain-gauge kicking-force tracings were continuously sampled and transferred with a frequency of 100 Hz via the analog-to-digital converter to the PC. This continuous sampling procedure allows a direct quality control of each velocity spectra after the experiments, where insonation failures can be detected and excluded from the flow analysis (16). It also eliminates the possible interference in the size of the averaged mean blood velocity and corresponding flow value that may occur when the blood flow velocity is averaged for each cardiac cycle triggered by the ECG (22, 28, 29). The blood velocity and flow were analyzed in relation to the muscle contraction force (strain-gauge) profile (16). A cuff below the knee around the calf muscles was temporarily inflated to a suprasystolic (>240 mmHg) blood pressure before the flow measurements to eliminate blood flow contributions to the lower leg. The volume of blood flow (in l/min) ( = 6 × 10⁴ × v_mean × r²) was calculated over the parabolic velocity profile by multiplying the cross-sectional area [A = r² (in m²)] of the artery with the angle-corrected, time- and space-averaged, and amplitude (signal intensity)-weighted mean blood velocity (v_mean, in m/s). The constant 6 × 10⁴ is the conversion factor from meters per second to liters per minute.

Statistical Analysis

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Temporal Changes

Blood velocity and flow. Femoral artery blood velocity and flow were not different (P = NS) before the start of the different exercise interventions, with mean values of 0.062 ± 0.0046 m/s and 0.35 ± 0.028 l/min (~3.5 ml · min¹ · 100 g¹), respectively. With passive leg movements, both increased (P < 0.008) with the very first passive relaxation, reaching a plateau (P = NS) within four to five passive duty cycles at a 3.3-fold higher level compared with rest and with a mean blood flow of 1.17 ± 0.11 l/min (~37 ml · min¹ · 100 g¹). The level of blood flow was similar (P = NS) regardless of the following exercise intensity, as well as during the full minute of passive leg movement during the warmup session.

1

1

1

1

1

1

1

1

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Blood flow (mean ± SE) in the femoral artery in the transition from rest to passive exercise and to voluntary, 1-legged, dynamic knee-extensor exercise at 10 W (n = 7), 30 W (n = 7), 50 W (n = 5), and 70 W (n = 2). Voluntary exercise was initiated at time 0. Increase in blood flow was significant (P < 0.05) at all time points from rest to passive and voluntary exercise, as well as from passive to voluntary exercise, at 10, 30, and 50 W, and there was an equally apparent further elevation in blood flow for the 2 subjects at 70 W.

Mathematical modeling. Curve fitting of the increase in blood velocity and flow for the full-kicking duty cycles at onset of the voluntary dynamic contractions was best described by a one-, two-, or three-component exponential model. In the transition from passive leg movement to voluntary exercise, blood flow increased according to the equation (t) = _passive + ₁ · [1  e ^{(t  TD₁/₁)}] + ₂ · [1  e^{(t  TD₂/₂)}] + ₃ · [1  e^{(t  TD₃/₃)}]. In this equation, _passive represents the blood flow during passive leg movement; ₁, ₂, and ₃ represent the magnitude of response for each component toward the stable steady-state asymptotic blood flow, leveling off at _passive + ₁ + ₂ + ₃; the time constants ₁, ₂, and ₃ reflect the rapidity of the rise for each exponential component, where smaller time constants correspond to faster rises, and TD₁, TD₂, and TD₃ represent the onset latency time of each component, respectively.

1

1

1

1

1

1

1

1

Temporal Relationship

Rest and passive movement. The intramuscular pressure at rest was stable at ~25.8 ± 5.3 mmHg and changed for all subjects at the start of passive leg movement as a function of the force; it was highest (68.6 ± 15.7 mmHg) during passive pulling (muscle elongation) and lowest (6.4 ± 8.6 mmHg) during passive pushing (muscle shortening) of the leg (P < 0.002). Analogous with protocol I, femoral arterial mean blood velocity and flow increased (P < 0.05) with 52.5 ± 22.2% in the first pushing phase of the first duty cycle of the passive leg movements and increased (P < 0.05) further successively for each subsequent duty cycle to level off (P = NS) at a 4.4-fold higher magnitude (compared with rest) between the fourth and fifth passive leg movement at a blood flow of 1.01 ± 0.092 l/min (~34 ml · min¹ · 100 g¹).

Voluntary contraction. At the onset of the voluntary contractions (47.1 ± 1.8 W), the amplitude of the oscillations in intramuscular pressure increased (P < 0.004) further compared with rest and during the passive leg movements (Figs. 2 and 3) in relation to the knee-extensor force over the contraction-relaxation cycle. The mean blood velocity and flow peaked with an amplitude increase (P < 0.006) of 28.1 ± 6.6% with the first relaxation, analogous with protocol I, giving a TD₁  0.3-0.5 s. The blood flow furthermore increased (P < 0.05) for the consecutive kicking duty cycles to a mean value of 1.65 ± 0.24 l/min (~ 55 ml · min¹ · 100 g¹), 2.0 ± 0.4 l/min (~67 ml · min¹ · 100 g¹), and 2.2 ± 0.32 l/min (~ 74 ml · min¹ · 100 g¹) after 1, 3, and 5 s (duty cycles), respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Mean blood flow, heart rate (HR), and mean arterial blood pressure (MAP), as well as nadir (nd)-to-peak (pk) variations and mean values of intramuscular (IMP) and venous pressure (VP) at rest; during passive exercise; during voluntary, 1-legged, dynamic knee-extensor exercise at 47.1 ± 1.8 W [65.4 ± 3.6% of maximum work rate; i.e., submaximal voluntary exercise (Vol. Sub-Max)]; and at peak exercise effort of 72.8 ± 3.8 W. Significant differences (P < 0.05): * from rest, ** from rest and passive, and *** from rest, passive, and submaximal voluntary exercise, respectively.

View larger version (59K): [in this window] [in a new window]   Fig. 3.   Femoral arterial blood velocity (vmean), knee-extensor force (F), IMP, arterial (BPa) and venous (BPv) blood pressure, and HR in the transition from rest to passive exercise to voluntary, 1-legged, dynamic knee extensor exercise at 50 W from 1 representative subject. Onset of voluntary exercise is marked by a "1." BPM, beats/min.

Arterial pressure. Mean arterial blood pressure increased (P < 0.05) initially within the time frame of the first to third contraction by 8.9 ± 1.9 and 6.4 ± 2.3 mmHg compared with rest and passive leg movements, respectively. This was followed by a temporary decrease (P < 0.002) of 8.1 ± 1.0 mmHg compared with during the first three contractions and a tendency (P = NS) to decrease by 2.0 ± 2.0 and 4.2 ± 2.1 mmHg below the mean arterial pressure level at rest and during passive leg movements, respectively. The decrease in the arterial blood pressure was observed (P < 0.002) with a time delay of 4.2 ± 0.5 s after the onset of voluntary exercise and reached its nadir and turning point within 6.5 ± 0.6 s, indicating a period of vasodilatation, and analogous with protocol I, an observable onset latency time point (TD₂) for the start of phase 2. At steady-state voluntary exercise, mean arterial pressure increased (P < 0.02) compared with rest and passive leg movements, whereas it was unaltered (P = NS), with only a slight tendency to increase, with passive leg movement compared with rest (Fig. 2).

Venous pressure. The oscillations in venous pressure on the femoral level, which increased (P < 0.001) with the onset of passive leg movements, giving an elevation of mean venous pressure with 1.3 ± 0.1 mmHg compared with rest, were further potentiated (P < 0.025) at the onset of the voluntary contractions in phase with the intramuscular pressure variations (see below). The mean venous pressure increased (P < 0.025) by 7.6 ± 1.7 and 6.4 ± 1.6 mmHg compared with rest and passive leg movement, respectively (Fig. 2). The arterial-venous pressure difference on the femoral level was, however, unaltered (P = NS) with the first contractions of phase 1.

Temporal variation. The blood velocity altered from being three phasic and dependent on the cardiac cycle and pulse pressure at rest (Fig. 3) to being directly related to the elongation and shortening of the muscle during the passive leg movements and voluntary contractions (Figs. 3 and 4). The arterial pulse pressure had a superimposed influence on the primary effects of the variations in intramuscular pressure during the contraction and relaxation phases, respectively (Fig. 4). When a series of (12) kicking duty cycles for each subject were studied, the intramuscular and venous pressures were significantly greater (P < 0.001, P < 0.04) during the contraction phase than during the relaxation phase, respectively, but were unaffected (P = NS) by the arterial pulse pressure during each condition (Table 2). The arterial pressure was furthermore increased (P < 0.03) during the contraction phase compared with the relaxation phase at the second dicrotic notch and at diastole, both by approximately 6-7 mmHg. The peak systolic blood pressure was similar (P = NS) under both conditions (Table 2).

View larger version (40K): [in this window] [in a new window]   Fig. 4.   Femoral arterial vmean, knee-extensor F, IMP, and BPa and BPv during steady-state voluntary, 1-legged, dynamic knee-extensor exercise at 50 W, from 1 representative subject. Temporal variation in arterial vmean and inflow for successive kicking duty cycles in relation to IMP variations is shown, along with superimposed influence of BPa and BPv. The § indicates an insonation failure, excluded from subsequent blood flow analysis.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

This study demonstrates that it is possible in humans to continuously measure the arterial inflow to dynamically contracting muscle in the transition from rest to exercise at different intensities up to levels approaching peak power output. Moreover, the temporal resolution of the ultrasound Doppler technique makes it possible to follow each phase of a duty cycle, demonstrating a very close temporal relationship between the changes in arterial blood velocity, and thus inflow, to the contraction- and relaxation-induced variations in intramuscular, arterial, and venous pressures.

The arterial blood velocity and inflow are significantly elevated already at the first relaxation phase of a passive leg movement, plateauing at a 3.3- to 4.4-fold higher level compared with rest within four to five passive duty cycles, i.e., an elevation due purely to muscle mechanical factors. On initiation of voluntary exercise, arterial blood velocity and inflow may be either blocked (zero) or retrograde with the first contraction. A marked elevation in blood flow follows with the first relaxation phase, due to refilling of the vascular bed, a process just as apparent as for the purely passive leg movements. Furthermore, the mean arterial blood velocity in the femoral artery (internal diameter of ~ 10 mm) increased by ~50% by the end of the first passive or voluntary duty cycle compared with a rest velocity of ~0.06 m/s. Thus this arterial inflow corresponds well with the ~50-ml volume of blood contained in the knee-extensor muscles, with a muscle volume of ~3,000 ml, and assuming a capillarization of ~1.5 %, along with the volume contributions in venules and small arteries (18). Moreover, the squeezing of the blood volume out of the muscle during the contraction and the refilling of the vasculature during the relaxation seem sufficient to induce the initial elevation in blood flow during the first relaxation phase. Depending on power output, this process contributes to the elevation in blood flow during the first seconds of work with a 5.5- to 12.5-fold elevation compared with rest. The mean arterial-venous pressure gradient on the femoral level was unaltered with the first three contractions, as the mean pressures of both were slightly elevated. It must be emphasized, however, that these pressures are valid for the level of measurement in the vascular tree and may not represent the absolute true values in the microvasculature. However, their changes may respectively give information about the general temporal oscillations during exercise and be used as a guideline for the events occurring. Thus, in light of the marked oscillations in venous pressure with values close to zero at a location proximal to the venous valves, there is a possibility that a small elevation in the arterial-venous pressure gradient still occurs on the microvascular level, as previously suggested (6, 13, 21). There is a close temporal relationship between the marked oscillations in intramuscular pressure and the initial amplitude increase in arterial blood velocity already with the first relaxation(s) during passive as well as voluntary exercise. This relationship emphasizes the importance of purely muscle-mechanical factors for facilitating the first stage of the hyperemic response during the very first phase (onset latency of ~ 0.3-0.5 s). The arterial blood velocity and inflow is increased with the successive duty cycles during the first 5 s of exercise, whereas the oscillations in intramuscular pressure variations stabilize after the first contraction at a fixed amplitude. This suggests that factors other than muscle mechanical cause the amplification of the hyperemic response.

The arterial blood pressure drops after the first contractions, with an onset latency of 4.2 ± 0.5 s, reaching its nadir after 6.5 ± 0.6 s and thereafter becoming elevated again. Thus there is a distinct separate second phase in the hyperemic response, which is induced by vasodilatation. In light of the successive increases in blood flow for each duty cycle after the first contraction, and the very minor changes in blood pressure, it appears that the vasodilation is present already after the second contraction. The temporary slight drop in blood pressure before its elevation probably reflects the fact that the sympathetic drive for a brief period is lagging behind. A third phase in the elevation of blood flow was observed after ~30 s at the highest intensity. In previous studies with a similar exercise model, Eriksen et al. (12) have shown that there is no detectable delay between the initial increase in cardiac output and femoral arterial inflow. However, the magnitude of total increase in femoral arterial inflow may markedly exceed the increase in cardiac output (12), suggesting an immediate redistribution of blood flow from other vascular beds to ensure an adequate delivery to the contracting muscles.

The increase in arterial blood velocity and inflow at the onset of the voluntary contractions was related to the intensity and was best described by a one-component (~15 % of peak power output), two-component (~40-70% of peak power output), or three-component (75% of peak power output) exponential function. This resembles a similar mathematical model for the onset of intermittent static contractions-relaxations (each of 2-s duration) at 10% of MVC previously described by Shoemaker et al. (22). The phasic appearance of different components with different onset latencies (TD₁ ~ 0.3-0.5 s, TD₂ ~ 4.2 s, TD₃ ~ 30 s), rates of rise, and potencies further demonstrated the existence and role of several factors for elevating the blood flow during submaximal exercise. The muscle pump alone may, however, be sufficient for the initial elevation at very light workloads. The magnitude and rapidity of rise in blood flow for the first phase at onset of voluntary exercise increased with exercise intensity, where an asymptotic leveling off in blood flow was reached within 10 s for the exercise at ~15% of peak power output. The slightly slower but more potent second phase induced a further elevation of the blood flow to its final level within 50-60 s and 90-150 s at ~40 and 70% of peak power output, respectively. The third and slowest component induced a blood flow increase with a leveling off within ~150 s at 75 % of peak power output. The time for the blood flow to reach 50% of the increase in leg blood flow was in the range 2.2-8.9 s, i.e., for higher work intensity, the time was longer.

The aim of the present study was not to ascertain which substance(s) cause hyperemia in skeletal muscle but instead to give very precise information about the time frame within which various factors may act. Even though vasodilatation may not induce the elevation in blood flow with the first duty cycle, it must be initiated during the second to fourth second to allow for the marked hyperemia that occurs. Adenosine, NO, and acetylcholine are all potent and endogenous vasodilators and candidates for this hyperemic response, but their time frame of action and source of release in relation to skeletal muscle contractions and blood flow regulation still need to be clarified in vivo in humans. The rapid and potent initial increase in blood flow, despite no simultaneous increase in femoral arterial diameter, stresses that its size is not the determining factor for delivery during submaximal exercise up to intensities near peak level in humans. It also emphasizes that the cause of the blood flow elevation must be attributed to the effect of the muscle mechanical factors or vasodilators on the microvasculature with its relatively larger surface area and greater vasodilatory capacity, rather than an effect on the major feeding artery. The vasodilators of interest are therefore limited to these with a short half-life, i.e., less than the time of recirculation.

The study furthermore emphasizes that the minimum and maximum intramuscular pressure variations over the relaxation and contraction cycles correspond to the highest and the lowest phase of the mean blood velocity, respectively, with the arterial pressure as a superimposed influence. Blood flow is thus promoted during the muscle relaxation phase and further enhanced by the arterial pulse pressure. At the submaximal workload studied (~65% of peak power output), blood flow is impeded by mechanical hindrance during the muscle contraction phase, unless the perfusion pressure is in its peak value range, where the flow restriction may be overcome. As intramuscular pressure may vary with the tension of the muscle fibers, recording depth, and fiber geometry (9, 14, 19, 20), the intramuscular probe was inserted in the same location and muscle type in all the subjects, as well as at a fixed and specific depth approximately central in the muscle. The recorded intramuscular pressure at baseline rest was similar to previous measurements by others (14, 23, 25, 30). Even though it is beyond the focus of this study to clarify regional differences in intramuscular pressure within the muscle, it is, however, interesting to speculate that such variations may as previously has been suggested possibly be the cause of regional differences and heterogeneity in blood flow within a muscle group (23).

In addition, one could argue that the stretch of the muscle during passive leg movement could evoke a reflexive contraction via the stretch reflex, affecting the intramuscular pressure recordings. In this study we did not elucidate this phenomenon using electromyography. However, such a mechanism seems unlikely to have affected our results because the intramuscular pressure oscillations during passive exercise were extremely stable, despite the variations in the applied pulling force, which naturally was greatest in the initial acceleration phase of the ergometer lever arm during the first duty cycles from the resting condition up to 60 revolutions/min. The oscillations in intramuscular pressure were equally stable as the applied pulling force was decreased, and started to stabilize and resemble the force during the steady-state voluntary exercise. Moreover, in a parallel study we have measured heat storage in the muscle during 60 s of passive leg movements, without any elevation in the muscle temperature. In contrast, heat is gained within the first one to three voluntary contractions. Thus the passive leg movements seem a valid comparison to the voluntary exercise, possessing the additive metabolic contributions.

In conclusion, the study demonstrates a very close temporal relationship between the arterial inflow and the variations in intramuscular, arterial, and venous pressure at onset of, and during, dynamic exercise, where the intramuscular pressure variations may promote as well as impede the muscle perfusion. The arterial blood pressure is an additional superimposed influence, although it is of less importance. The venous pressure may also still oscillate to possibly increase the pressure head and thus the driving force on the microvascular level. Blood flow to the exercising muscle is optimized when the heart beat and arterial systolic blood pressure is timed to occur during the relaxation. Depending on intensity, the initial hyperemia shows a one-, two-, or three-phasic appearance, where the arterial inflow increases with the first muscle relaxation(s) because of refilling of the arterial and the venous vasculature immediately after the emptying with the first contraction(s). The elevation of blood flow is most rapid during the first phase (0.5-5 s), initially facilitated by the muscle pump. The muscle mechanical component is followed by a second and more potent vasodilatory phase, observed with an onset latency of ~4 s but most probably initiated already during the second to fourth contraction of the muscle mechanical stage of phase 1. At greater intensities, a third phase is identified with an onset latency of ~30 s.