Download Effect of rhythmic tetanic skeletal muscle contractions on peak

J Appl Physiol 94: 11–19, 2003.
First published August 16, 2002; 10.1152/japplphysiol.00339.2002.
Effect of rhythmic tetanic skeletal muscle
contractions on peak muscle perfusion
JOHN L. DOBSON AND L. BRUCE GLADDEN
Department of Health and Human Performance, Auburn University, Auburn, Alabama 36849-5323
Submitted 15 April 2002; accepted in final form 8 August 2002
ALTHOUGH IT IS CURRENTLY HELD that vascular conductance in skeletal muscles is accomplished primarily by
local control mechanisms (25), the relative importance
of each form of local control throughout the initiation
and maintenance of rhythmic contractions has not
been well established (8). Similarly, although it is clear
that skeletal muscle blood flow is augmented immediately after the initiation of contractions, controversy
exists as to whether local vasodilation begins instantaneously (7, 34, 41, 46) or only after a latency period
lasting at least 5 s (5, 16, 19, 30, 39). One pertinent
issue that might explain why muscle hyperemia may
precede local vasodilation is whether the actual mechanical action of rhythmic skeletal muscle contraction
acts as an additional form of local blood flow control
during exercise. This form of local control, termed the
skeletal muscle pump, has been implicated not just as
the factor responsible for the onset of active hyperemia
(9, 10, 38, 39, 46) but also as an essential component
needed for the development of peak muscle blood flow
(3, 4, 23, 24, 28).
The compressive force of muscle contraction generates intramuscular pressures in the range of 200–
1,700 mmHg (1, 21, 22, 36), imparting enough kinetic
energy to squeeze blood into the veins, thereby facilitating venous return (2, 29, 44). Furthermore, Pollack
and Wood (37) demonstrated that properly spaced
rhythmic contractions can decrease the extent of venous refilling during muscular relaxation enough to
reduce and maintain lower venous pressures. Subsequently, Folkow et al. (9) introduced the modern concept of the muscle pump by proposing that this gain in
the pressure gradient across a skeletal muscle is sufficient to elevate blood flow during rhythmic contractions. According to Ohm’s law, blood flow through a
tissue is determined by the product of the arterialvenous pressure difference and the conductance of the
tissue’s vascular bed. Therefore, if this principle is
applicable to skeletal muscles during rhythmic contractions, then the kinetic energy imparted to the blood
during contraction, combined with the potentially negative pressures in the venules and small veins during
relaxation (23), may also provide independent facilitation of muscle hyperemia during exercise.
Support for the muscle pump theory has been provided by numerous investigations (10, 28, 38, 40, 41).
The muscle pump has been implicated as the explanation for the significantly higher peak blood flows typi-
Address for reprint requests and other correspondence: J. L. Dobson, Dept. of Exercise Science, Rm. 25—Florida Gymnasium, PO Box
118206, Univ. of Florida, Gainesville, FL 32611 (E-mail: jdobson
@hhp.ufl.edu).
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
blood flow; vasodilation; venous pressure; adenosine; nitroprusside
http://www.jap.org
8750-7587/03 $5.00 Copyright © 2003 the American Physiological Society
11
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Dobson, John L., and L. Bruce Gladden. Effect of
rhythmic tetanic skeletal muscle contractions on peak muscle perfusion. J Appl Physiol 94: 11–19, 2003. First published
August 16, 2002; 10.1152/japplphysiol.00339.2002.—The
purpose of this investigation was to examine the effect of
rhythmic tetanic skeletal muscle contractions on peak muscle perfusion by using spontaneously perfused canine gastrocnemii in situ. Simultaneous pulsatile blood pressures
were measured by means of transducers placed in the popliteal artery and vein, and pulsatile flow was measured with a
flow-through-type transit-time ultrasound probe placed in
the venous return line. Two series of experiments were performed. In series 1, maximal vasodilation of the muscles’
vascular beds was elicited by infusing a normal saline solution containing adenosine (29.3 mg/min) and sodium nitroprusside (180 ␮g/min) for 15 s and then simultaneously
occluding both the popliteal artery and vein for 5 min. The
release of occlusion initiated a maximal hyperemic response,
during which time four tetanic contractions were induced
with supramaximal voltage (6–8 V, 0.2-ms stimuli for 200-ms
duration at 50 Hz, 1/s). In series 2, the muscles were stimulated for 3 min before the muscle contractions were stopped
for a period of 3 s; stimulation was then resumed. The results
of series 1 indicate that, although contractions lowered venous pressure, muscle blood flow was significantly reduced
from 2,056 ⫾ 246 to 1,738 ⫾ 225 ml 䡠 kg⫺1 䡠 min⫺1 when
contractions were initiated and then increased significantly
to 1,925 ⫾ 225 ml 䡠 kg⫺1 䡠 min⫺1 during the first 5 s after
contractions were stopped. In series 2, blood flow after 3 min
of contractions averaged 1,454 ⫾ 149 ml 䡠 kg⫺1 䡠 min⫺1. Stopping the contractions for 3 s caused blood flow to increase
significantly to 1,874 ⫾ 172 ml 䡠 kg⫺1 䡠 min⫺1; blood flow declined significantly to 1,458 ⫾ 139 ml 䡠 kg⫺1 䡠 min⫺1 when
contractions were resumed. We conclude that the mechanical
action of rhythmic, synchronous, maximal isometric tetanic
skeletal muscle contractions inhibits peak muscle perfusion
during maximal and near-maximal vasodilation of the muscle’s vascular bed. This argues against a primary role for the
muscle pump in achieving peak skeletal muscle blood flow.
12
EVIDENCE AGAINST THE MUSCLE PUMP
MATERIALS AND METHODS
Animals, housing, and care. For this study, seven mongrel
dogs (14.1–27.3 kg), of either sex, were acquired from the
J Appl Physiol • VOL
Auburn University Laboratory Animal Health Facility at the
College of Veterinary Medicine. All procedures were reviewed and approved by the Auburn University Institutional
Animal Care and Use Committee (protocol 0308-R-2312).
Animals were obtained after they were fasted overnight.
The animals were anesthetized with pentobarbital sodium
(30 mg/kg iv) and intubated with an endotracheal tube. The
animals were mechanically ventilated (model 613, Harvard
Apparatus) to maintain normal blood-gas values, and supplemental oxygen was supplied when necessary. Rectal temperature was monitored and maintained at 37°C with a
heating pad placed under the animal and adjusted as needed.
Additional doses of pentobarbital sodium were given as
needed to maintain a deep surgical plane of anesthesia. The
animals were euthanized while still under anesthesia at the
end of the experiments.
Surgical procedure. The left gastrocnemius-plantaris muscle group (GP) was surgically isolated as previously described
(11–13, 15, 43). Briefly, a medial incision was made through
the skin of the left hindlimb from midthigh to the ankle. The
insertion tendons of the sartorius, gracilis, semitendinosus,
and semimembranosus muscles were cut to allow these muscles to be folded back to expose the GP. All branches of the
popliteal vasculature that did not feed directly into or out of
the GP were ligated. Sutures also were tied around the heads
of the muscle groups to occlude any blood flow through the
tissue of the muscle’s origins. Consequently, all inflow into
the GP was isolated to the popliteal artery, and all outflow
from the GP was isolated to the popliteal vein. Complete
arterial isolation was verified by occluding the popliteal artery and observing venous flow. The popliteal vein was cannulated, and the venous outflow was directed through a
flow-through-type transit-time ultrasound flow probe
(6NRB440, Transonic Systems) before being returned to the
animal via a reservoir in the jugular vein. The flow probe was
connected to a flowmeter (T206, Transonic Systems), set on
the “high flow” setting and calibrated with a graduated
cylinder and stopwatch before and during each experiment.
The popliteal vein cannula was also attached to two stopcocks, one for the collection of venous blood samples and the
other to connect to a pressure transducer (model RP-1500,
Narco Biosystems). To avoid interfering with a muscle pump
effect by directly cannulating GP inflow, a branch of the
popliteal artery that was upstream from the GP, but did not
supply the muscle, was cannulated. This arterial cannula
was connected via a y-connector to a pressure transducer
(model RP-1500, Narco Biosystems) and to an infusion pump.
This allowed monitoring of muscle perfusion pressure and
introduction of pharmacological agents (vasodilators) with
minimal intrusion on the integrity of the popliteal artery
itself.
A portion of the calcaneus, with the two tendons of the GP
attached, was cut away for connection to an isometric myograph. The two tendons were clamped around a short metal
rod and connected via a short section of aluminum pipe (29
mm diameter) and a universal joint coupler to the myograph
load cell (Interface SM-250, Narco Biosytems). The universal
joint coupler was used to ensure that the muscle always
pulled directly in line with the load cell, thus alleviating the
application of torque to the load cell. The load cell was
calibrated with known weights before each experiment. The
proximal end of the GP remained attached to its origin. The
GP was covered with a saline-soaked gauze and a thin piece
of plastic to prevent drying and cooling. Both the femur and
tibia were fixed to the base of the myograph via bone nails
and connecting rods. A turnbuckle strut was placed parallel
to the muscle between the tibial bone nail and the arm of the
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
cally reported for muscles in vivo vs. the peak flows
reported for isolated muscles (3, 4, 23, 24, 28). Nevertheless, only Sheriff et al. (39) have provided specific
data to indicate that local muscle hyperemia may occur
independently of vasodilation. In contrast, others (27,
41, 45) have provided evidence that the reduction in
venous pressure that occurs during rhythmic contractions is not sufficient to account for independent facilitation of hyperemia during exercise. The most compelling argument against the effect of the muscle pump,
however, is provided by the proponents of a vascular
waterfall at the level of the muscle arterioles. The
primary concept of the vascular waterfall theory is that
flow through a collapsible tube is described by a different expression of Ohm’s law, wherein the operable
pressure gradient is determined by the difference between the arterial and critical closing pressures of the
tube (35). Accordingly, the presence of a vascular waterfall, or Starling resistance, at the arterioles would
prevent any increase in flow through a muscle due
solely to a decrease in pressure on the venous side (6,
18). Evidence in support of the vascular waterfall, and
against the muscle pump effect, has been provided by
three well-controlled investigations (29, 33, 42), which
demonstrated that reductions in venous pressure did
not have an effect on arterial inflow.
Despite the evidence against a muscle pump effect,
including that from their own investigation, Laughlin
and Schrage (26) concluded that the muscle pump most
likely exists and could possibly be demonstrated by a
study in which instrumentation of the arterial supply
and venous outflow of an isolated muscle is minimal.
Similarly, recent work in our laboratory suggests both
that there is a measurable muscle pump activity during the onset of contractions in the isolated dog gastrocnemius muscle (14) and that cannulation of the
arterial inflow to muscles in situ inhibits the maximal
rate of blood flow (unpublished results). Each of the
investigations that provided evidence that the vascular
waterfall prevents a muscle pump effect (29, 33, 42)
was performed with the use of isolated canine gastrocnemius muscles with cannulation of the arterial supply. Furthermore, although two (Refs. 33 and 42) of
these muscle pump investigations were performed on
muscles with maximally vasodilated vascular beds, the
peak blood flows reported were less than one-half as
great as those measured in our laboratory with a similar muscle preparation (20).
Although it has been postulated that the muscle
pump is necessary to achieve maximal muscle perfusion (23, 24), this idea remains controversial (26).
Therefore, the purpose of the present study was to
determine the effect of rhythmic muscle contractions
on peak skeletal muscle perfusion by using a spontaneously perfused canine gastrocnemius preparation in
situ with no direct cannulation of the muscle’s arterial
supply.
EVIDENCE AGAINST THE MUSCLE PUMP
J Appl Physiol • VOL
internal vasodilator metabolites in response to hypoxia. Our
preliminary data confirmed that this infusion-occlusion technique elicited maximal pharmacological vasodilation and allowed a period of no less than 20 s in which systemic arterial
pressure was not reduced.
After the return of the blood flow and pressure values to
resting levels (⬃15 min), the series 2 protocol was begun. In
series 2, the GP was stimulated for 3 min in the manner
described above. Subsequently, the stimulations were
stopped for an average of 3 s and then immediately restarted.
The purpose of this series of experiments was to compare the
muscle’s blood flow 1) during rhythmic contractions, 2) immediately after contractions were stopped, and 3) immediately after contractions were resumed. If the blood flow were
to decrease immediately after contractions were stopped and
then immediately increase as they were restarted, such evidence would again indicate that there was a muscle pump
effect.
Measurements. Outputs from the flowmeter, pressure
transducers, and load cell were recorded on a strip-chart
recorder (Narcotrace 40, Narco Biosystems) for monitoring
and analysis. Outputs from the load cell, pressure transducers, and flowmeter were also fed into a computerized dataacquisition system (PowerComputing Powerbase 240 Macintosh clone; GW Instruments, Superscope II, InstruNet model
100B A/D converter). The data-acquisition system sampled
all variables at 100 data points per second. Heart rate was
monitored periodically with the arterial pressure tracing.
Arterial and venous blood samples were collected anaerobically in 3-ml plastic syringes both before any experimental
perturbations and after 3 min of rhythmic contractions (see
above stimulation protocol). The blood was analyzed for PO2,
PCO2, and pH with a blood-gas, pH analyzer (Instrumentation Laboratory 1304). The samples were also analyzed for
hemoglobin concentration and percent saturation of hemoglobin with a calibrated CO-oximeter (Instrumentation Laboratory 282) set for dog blood. Oxygen uptake by the GP was
calculated from the blood flows and corresponding arteriovenous oxygen concentration differences. After each experiment, the muscle was removed from the animal, dissected
free of connective tissue, and weighed. The muscle was then
dried in an oven at 80°C to determine its percentage of water.
Data analysis. The data collected from the experiments of
series 1 were divided into three distinct periods for analysis:
hyperemia, contraction, and postcontraction. The hyperemia
period (mean ⫽ 2.0 s) began the moment peak blood flow was
achieved, following the release of the occlusion of the popliteal vasculature, and ended immediately before the first
contraction began. The contraction period (mean ⫽ 4.0 s)
began the moment contractions started and ended 1 s after
the last contraction began. The first 5 s after the contraction
period ended was considered the postcontraction period.
The data collected from the experiments of series 2 were
also divided into three distinct periods for data analysis:
contraction, stop, and resume contraction. The contraction
period (mean ⫽ 5.0 s) began the moment the fifth-to-the-last
contraction began and ended 1 s after the last contraction
began. The stop period (mean ⫽ 3.3 ⫾ 0.7 s) began the
moment the contraction period ended and lasted until the
moment before contractions were resumed. The first five
contractions after the stop period ended were considered the
resume contraction period (mean ⫽ 5.0 s).
Statistics. Differences among the variables measured in
this study were assessed by using a one-way repeated-measures ANOVA. Appropriate post hoc contrasts were used
when necessary to determine where significance occurred.
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
myograph to minimize flexing of the myograph. The sciatic
nerve was exposed, isolated, and ligated near the GP. The
distal 1.5- to 3.0-cm stump of the nerve was pulled through
an epoxy electrode containing two wire loops for stimulation.
Muscle contractions were elicited by stimulating the sciatic
nerve with supramaximal (6–8 V) square-wave pulses (Grass
S48 stimulator). Before each experiment, the GP was set to
optimal length by progressively lengthening the muscle as it
was stimulated at a rate of 0.2 Hz, until a peak in developed
twitch tension (total minus resting tension) was observed.
Experimental protocols. Once surgical isolation of the GP
was complete, all cannulas and equipment were in place, and
optimal length was determined, the muscle was allowed to
rest for 10 min while arterial blood gases and pH and blood
flow were measured. After confirmation that these values
were within normal limits for this experimental model, the
GP was stimulated to contract for 3 min to determine peak
contractile values for muscle blood flow and oxygen uptake.
Isometric tetanic contractions were evoked by stimulation
with trains of stimuli (200-ms duration, 50-Hz frequency) at
a rate of one contraction per second; each stimulus pulse was
of supramaximal voltage (6–8 V) and 0.2 ms in duration.
When the blood flow and pressure values had returned to
resting baseline levels, the muscle was used in the following
two series of experiments.
The general procedure of the experiments of series 1 was to
first elicit peak GP blood flow by dilating the muscle’s vascular bed maximally and then to stimulate the muscle to
contract four times, in the manner described above. The
purpose of these experiments was to compare the blood flow
under three different conditions: 1) during maximal pharmacological dilation, 2) during maximal pharmacological dilation plus high-intensity rhythmic contractions, and 3) immediately after high-intensity contractions but with continued
maximal pharmacological dilation. If the muscle’s blood flow
were to increase immediately after the initiation of contractions and then to decrease immediately after contractions,
such evidence would be interpreted as support for mechanical augmentation of skeletal muscle blood flow and thus a
muscle pump effect.
Preliminary experiments indicated that maximal GP vasodilation could be elicited by infusing the muscle at 9 ml/min
for 60 s with a normal saline solution containing sodium
nitroprusside (20.0 ␮g/ml, Sigma Chemical S-0501), to activate the cGMP pathway, and adenosine (3.25 mg/ml, Sigma
Chemical A-9251), to activate the cAMP pathway. During
that required duration of infusion, however, a sufficient
amount of adenosine and sodium nitroprusside escaped into
the general circulation to reduce the systemic arterial pressure by an average of 10 mmHg. Because a decrease in
perfusion pressure alters the pressure gradient across the
GP, an alternate means of eliciting maximal GP conductance
had to be devised. Attempts were then made to match the
peak blood flows of maximal pharmacological vasodilation by
occluding the popliteal artery for 5–10 min. The hyperemia
that resulted from the release of the occlusion neither was as
great as that from maximal pharmacological vasodilation nor
was the peak blood flow maintained long enough for reliable
study. After further preliminary experimentation, it was
determined that the desired vasodilation could be accomplished by infusing the pharmacological agents for 15 s and
then stopping the infusion while simultaneously occluding
both the popliteal artery and vein for 5 min. This infusionocclusion technique sequestered the pharmacological agents
from the systemic circulation and simultaneously subjected
the GP to the required quantity of vasodilators, while concurrently initiating the accumulation of the muscle’s own
13
14
EVIDENCE AGAINST THE MUSCLE PUMP
Table 1. Resting and peak steady-state values
during muscle contraction
Venous outflow, ml 䡠 kg⫺1 䡠 min⫺1
Arterial pressure, mmHg
Venous pressure, mmHg
Conductance,
ml 䡠 min⫺1 䡠 mmHg⫺1
PaO2, Torr
CaO2, vol%
V̇O2, ml 䡠 kg⫺1 䡠 min⫺1
PCO2, Torr
pH
[HCO3⫺], mM
Peak force, kN/kg wet muscle
Rest
Contractions
197 ⫾ 26
130 ⫾ 7
4.6 ⫾ 0.9
1,374 ⫾ 132
133 ⫾ 6
8.1 ⫾ 1.9
0.12 ⫾ 0.03
93 ⫾ 2
19.1 ⫾ 0.8
3.7 ⫾ 0.5
35.1 ⫾ 2.2
7.43 ⫾ 0.01
22.6 ⫾ 0.3
0.92 ⫾ 0.15
93 ⫾ 2
19.1 ⫾ 0.8
189.1 ⫾ 10.8
35.1 ⫾ 2.2
7.43 ⫾ 0.01
22.6 ⫾ 0.3
4.6 ⫾ 0.46
Values are means ⫾ SE; n ⫽ 7 dogs. The conductance data from
the contraction period actually represent virtual venous outflow
muscle conductances. PaO2 , arterial PO2 ; CaO2 , arterial O2 concentration; V̇O2, O2 uptake; [HCO3⫺], HCO3⫺ concentration.
RESULTS
Table 1 shows mean control values for major variables that were measured during rest and the peak
steady state of contractions. Some of the results of
series 1 are shown in Table 2. The purpose of the
experiments of series 1 was to compare GP blood flow
and pressures before, during, and after rhythmic contractions were stimulated throughout a period of maximal muscle vasodilation. In series 1, mean GP blood
flow increased to more than 10 times the resting level
within an average of 2.6 ⫾ 0.4 s after the release of the
popliteal occlusion (Table 2, hyperemia period). Stimulation of rhythmic contractions then caused an immediate 15 and 18% decrease in GP blood flow and conductance (Table 2, contraction period), respectively,
followed by a return of both to precontraction levels
immediately after the contractions ended (Table 2,
postcontraction period). It is important to note that the
contractile-induced conductances reported herein actually represent virtual muscle conductances because
extravascular compression during muscle contractions
affects both intravascular pressure and resistance.
During the postcontraction period of series 1, the
cessation of compression of the compliant vasculature,
along with the increase in mean muscle outflow,
Table 2. Series 1: effect of rhythmic skeletal muscle contractions on hyperemia
during maximal muscle vasodilation
Venous outflow, ml 䡠 kg 䡠 min
Arterial pressure, mmHg
Venous pressure, mmHg
Conductance, ml 䡠 min⫺1 䡠 mmHg⫺1
⫺1
⫺1
Hyperemia
Contraction
Postcontraction
2,056 ⫾ 246
122 ⫾ 8
9.6 ⫾ 1.4
1.58 ⫾ 0.27
1,738 ⫾ 225*
122 ⫾ 8
8.7 ⫾ 1.2
1.30 ⫾ 0.21*
1,925 ⫾ 225†
119 ⫾ 8
11.0 ⫾ 1.7†
1.51 ⫾ 0.24†
Values are ⫾ SE; n ⫽ 7 dogs. The conductance data from the contraction period actually represent virtual venous outflow muscle
conductances. * Significant difference between the hyperemia and contraction periods, P ⬍ 0.05. † Significant difference between the
contraction and postcontraction periods, P ⬍ 0.05. There were no significant differences between the hyperemia and postcontraction periods.
J Appl Physiol • VOL
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
The level of statistical significance was set at 0.05. Data are
reported as means ⫾ SE.
caused sufficient blood to accumulate in the popliteal
vein to significantly increase the venous pressure. Figure 1 illustrates responses during each individual contraction as well as the mean response across the entire
period of four contractions. Within each distinct muscle
contraction, GP outflow, venous pressure, and virtual
venous outflow conductance rapidly increased to peak
levels along with the force of contraction; during muscle relaxation, GP outflow, venous pressure, and virtual venous outflow conductance values first plummeted to their nadir and then gradually began to rise
toward their initial, precontraction levels. Although
the highest outflow, venous pressure, and conductance
values observed in series 1 occurred at the time of peak
force during the contraction of the muscle, the averages
of these variables across the entire cycle of four contractions (contraction period) were lower than the corresponding averages during the hyperemia and postcontraction periods (see Fig. 1).
As is shown in Table 3, the results of series 2 are
similar to those of series 1. Rhythmic contractions
significantly reduced GP blood flow and virtual venous
outflow conductance by 22 and 23%, respectively.
There were no significant differences in the blood flows
of either the hyperemia or postcontraction periods of
series 1 vs. the stop period of series 2, indicating that
the degree of muscle vasodilation was similar between
the two series of experiments. GP outflow, venous
pressure, and virtual venous outflow conductance were
elevated throughout the period when contractions were
stopped, and then these values quickly returned to
lower steady-state levels as contractions were resumed. The mean duration of contraction stoppage
during the experiments of series 2 was 3.3 ⫾ 0.7 s (Fig.
2). Figure 2 also illustrates that the relationship between GP force production and virtual venous outflow
conductance in series 2 was similar to that observed in
series 1. Although peak muscle virtual venous outflow
conductance was achieved at the peak of muscle force
development during the resume contraction period, the
highest mean conductance observed in series 2 occurred during the stop contraction period.
A comparison of blood flow values during each contraction within the two series of experiments (Table 4)
reveals that there was a significant reduction in the
venous outflow after the initial contraction. Additionally, there was an 18% reduction in GP virtual venous
15
EVIDENCE AGAINST THE MUSCLE PUMP
outflow conductance between the first and final three
contractions in series 2, along with small but statistically insignificant decreases during the first three contractions of series 1. There were no significant changes
in venous pressure from one contraction to the next in
either series.
The average weight of the muscles was 87.7 ⫾ 7.7 g.
The percentage of muscle tissue that was water at the
end of the experiments averaged 76.0 ⫾ 0.5%, which is
similar to that found in previous studies in this preparation (11–13).
DISCUSSION
Table 3. Series 2: effect of suddenly stopping and then restarting rhythmic skeletal muscle contractions
Venous outflow, ml 䡠 kg 䡠 min
Arterial pressure, mmHg
Venous pressure, mmHg
Conductance, ml 䡠 min⫺1 䡠 mmHg⫺1
Force, kN/kg wet muscle
⫺1
⫺1
Contraction
Stop
Resume Contraction
1,454 ⫾ 149
132 ⫾ 9
8.3 ⫾ 2.2
1.03 ⫾ 0.17
3.97 ⫾ 0.4
1,874 ⫾ 172*
132 ⫾ 8
10.5 ⫾ 2.3*
1.34 ⫾ 0.21*
1,458 ⫾ 139†
134 ⫾ 9
9.4 ⫾ 2.3
1.02 ⫾ 0.15†
4.05 ⫾ 0.4
Values are ⫾ SE; n ⫽ 6 dogs. The conductance data from the contraction and resume contraction periods actually represent virtual venous
outflow muscle conductances. * Significant difference between stop and contraction periods, P ⬍ 0.05. † Significant difference between the stop
and resume contraction periods, P ⬍ 0.05. There were no significant differences between the contraction and resume contraction periods.
J Appl Physiol • VOL
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 1. Effect of rhythmic skeletal muscle contractions on selected
cardiovascular variables during maximal muscle vasodilation
(n ⫽ 7 dogs). Each recording illustrates data averaged from the
experiments of series 1. The gray lines illustrate the mean values
during the contraction period. The conductance data from the
contraction period actually represent virtual venous outflow muscle conductances.
Skeletal muscle contractions generate high intramuscular pressures that impede the flow of blood into
the arteriolar vasculature, while at the same time
expelling blood from the venous vasculature. According
to the muscle pump hypothesis, the net effect of these
two events is an overall increase in the perfusion of the
muscle. It has been proposed that the muscle pump
mechanism is necessary to achieve maximal muscle
blood flow during high-intensity rhythmic contractions
(23, 24, 26). The results of the present investigation, by
contrast, indicate that the overall effect of rhythmic
contractions is to actually reduce mean muscle blood
flow in a maximally vasodilated muscle. To our knowledge, this is the first investigation to demonstrate that
the mechanical action of rhythmic tetanic skeletal contractions, on average, reduces peak muscle perfusion.
Our interest in performing this investigation was
initiated by the recent review by Laughlin and Schrage
(26), in which the authors concluded that the muscle
pump effect most likely exists and may be demonstrated in situ using a preparation that does not interfere with the muscle’s vascular mechanics. Recent
work in our laboratory also had indicated that cannulation of the arterial inflow to muscles in situ inhibits
the maximal rate of blood flow (unpublished results).
Consequently, venous outflow was assessed in this
investigation using a large-bore flow-through-type flow
probe, and the arterial pressure was sampled without
direct cannulation of the muscle’s inflow. As a result,
our peak contractile flows were in the upper range of
those typically reported (23, 26).
To ensure that any potential increases in muscle
blood flow during the contractions of series 1 were not
due to an increase in muscle metabolism, it was nec-
16
EVIDENCE AGAINST THE MUSCLE PUMP
Fig. 2. Effect of rhythmic skeletal muscle contractions
on virtual venous outflow muscle conductance during
peak contractile-induced perfusion of the muscle (n ⫽ 6
dogs). The recordings are illustrations of data averaged
from the experiments of series 2. According to the timeline provided, the initial contraction period ended at
second 1.0, the stop contraction period ended at second
4.3, and the resume contraction period ended at second
9.3. The numbers above the schematic are the means ⫾
SE of the conductances from each period of data collection.
transiently interrupted. This evidence suggests that
we were able to augment the perfusion gradient in both
series of experiments in the manner required to generate a muscle pump effect.
The statistical similarity between the blood flows
observed during the hyperemia and postcontraction
periods of series 1 indicates that the extent of muscle
vasodilation was consistent between the initial and
final periods of data collection. The vascular beds of the
muscles used in series 1 were, therefore, most likely
fully dilated throughout the duration of the contraction
period. Similarly, the immediate decrease in muscle
blood flow after the stop period of series 2, coupled with
the statistical similarity between the contraction and
resume contraction periods, indicates that flow inhibition during both sets of contractions probably occurred
despite a consistent level of muscle vasodilation
throughout the series. Consequently, the only experimental factor that affected muscle perfusion in these
experiments was the increase in vascular resistance
elicited by extravascular compression during rhythmic
contractions. Our results indicate that maximal muscle blood flow may not be achieved while all of the
motor units of the muscle are contracting in a rhythmic
and synchronous manner. Furthermore, if it were possible to elicit the increases in vascular conductance
Table 4. Contraction comparison data
1st Contraction
2nd Contraction
3rd Contraction
4th Contraction
1,958 ⫾ 279
1,723 ⫾ 220*
1,617 ⫾ 190*
1,627 ⫾ 176*
1.25 ⫾ 0.29
1.07 ⫾ 0.18
1.13 ⫾ 0.19
5th Contraction
Series 1
Venous outflow, ml 䡠 kg⫺1 䡠 min⫺1
Conductance,
ml 䡠 min⫺1 䡠 mmHg⫺1
1.39 ⫾ 0.31
Series 2 (resume contraction)
Venous outflow, ml 䡠 kg⫺1 䡠 min⫺1
Conductance,
ml 䡠 min⫺1 䡠 mmHg⫺1
1663 ⫾ 182‡
1486 ⫾ 164†
1353 ⫾ 123†
1363 ⫾ 116†
1412 ⫾ 134†
1.10 ⫾ 0.21
1.02 ⫾ 0.21
0.91 ⫾ 0.19†
0.90 ⫾ 0.15†
0.90 ⫾ 0.16†
Venous outflow values are ⫾ SE for 7 (series 1) and 6 (series 2) dogs. Virtual venous outflow conductance values are ⫾ SE for 5 dogs.
* Significant difference from the 1st contraction of series 1, P ⬍ 0.05. † Significant difference from the 1st contraction of series 2, P ⬍ 0.05.
‡ Significant difference between the 1st contraction and the stop period (1,791 ⫾ 185) of series 2, P ⬍ 0.05.
J Appl Physiol • VOL
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
essary to elicit maximal dilation of the muscle’s vascular bed. Therefore, we used adenosine to activate the
cAMP pathway and sodium nitroprusside to activate
the cGMP pathway, along with hypoxia to ensure full
vasodilation. This combination allowed us to maintain
perfusion pressure throughout each period of data collection, while at the same time achieving mean blood
flows equal to or greater than the highest values obtained by simply increasing the infusion rate of vasodilators (see MATERIALS AND METHODS). In addition, the
peak blood flows observed in this investigation during
maximal vasodilation were more than 2.5 times
greater than those reported in previous investigations
that have examined the muscle pump effect using
similar muscle preparations in the dog (33, 42).
The stimulation pattern we utilized was selected
because it elicits the highest blood flows in this muscle
preparation (20) and because the stimulus frequency
and train duration are within the range recommended
by Folkow et al. (9) to optimize the muscle pump effect.
More specifically, the contractions were spaced such
that they continually expelled blood from the muscle’s
compliant region fast enough to prevent full refilling,
and yet the duration of relaxation was sufficient to
allow hyperemia to occur. Indeed, we observed a 21%
increase in venous pressure when contractions were
EVIDENCE AGAINST THE MUSCLE PUMP
J Appl Physiol • VOL
in net venous outflow during rhythmic contractions, we
conclude that the net inflow of blood must also have
been reduced during rhythmic contractions. Furthermore, we reason that if the inflow of blood had increased during the second, third, fourth, and so forth
contractions within a series, then the venous outflow
should have also increased with each subsequent contraction. Nevertheless, because we do not know how
the inflow varied with each contraction, we must use
caution when interpreting these results, particularly
as to the effect of the initial contraction on arterial
inflow.
Care must also be taken when interpreting the results of series 1 because the extent of GP vasodilation
observed may have exceeded the greatest dilation that
could be achieved in response to high-intensity rhythmic contractions. It is also possible that the insertion of
a relatively noncompliant catheter into the venous
drainage of the muscle may have inhibited the muscle
pump effect. Still, we anticipated this problem and,
therefore, made attempts to alleviate any possible vascular interference by using a catheter with a bore size
greater than that of the lumen of the popliteal vein.
Consequently, the use of isometric tetanic muscle contractions was probably the most significant limitation
to this investigation.
Much of the support for the muscle pump effect has
developed from the observation that peak blood flows
typically reported for muscles in vivo are significantly
higher than the peak flows reported for isolated muscles (3, 4, 23, 24, 28). Because locomotion involves both
sequential muscle fiber recruitment and muscle shortening, whereas synchronous isometric contractions do
not, Laughlin and Schrage (26) hypothesized that the
disparity in peak blood flows typically reported in vivo
vs. in situ is due to the possibility that the muscle
pump is less efficient in isolated muscle preparations.
Although it seems likely that the muscle pump effect
would be optimized by contractions that progressed in
a wavelike manner from arterial to venous sites within
the muscle, we are not aware of any specific data
indicating that this actually occurs during locomotion.
Although we cannot be sure how blood flow through the
muscle was specifically affected by the use of synchronous tetanic contractions, it is clear that we were able
to use contractions to decrease the blood volume in the
muscle’s compliant region enough to reduce venous
pressures proximal to the muscle. Because the muscle
pump is thought to operate by simple mechanical augmentation of the muscle’s perfusion gradient, and
given that we were able to produce this effect using the
animal’s own muscles and vasculature, we conclude
that our results should be indicative of muscle blood
flow during peak perfusion in vivo. Laughlin and
Schrage also concluded that, if the muscle pump effect
does exist, then it should be possible to demonstrate it
with an isolated muscle preparation. We found no
muscle pump effect despite the fact that we obtained
muscle flows that were within the highest range reported for dogs in vivo (31) and were also well within
the range of the those others reported in vivo that were
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
observed during high-intensity rhythmic contractions
without the mechanical compression of the vasculature, we would expect the perfusion, and perhaps the
peak performance, of the muscle to increase.
Although it was beyond the scope of our investigation to determine whether a Starling resistor at the
level of the arterioles prevented a muscle pump effect,
our results do support previous findings that reductions in venous pressure do not enhance blood flow
through skeletal muscle tissue (29, 33, 42). It is interesting, however, in light of the similarities between our
experimental preparations, that the results obtained
by Naamani et al. (33) did not demonstrate that rhythmic contractions reduce muscle perfusion during vasodilation. Because the duty cycle they (33) used was
slightly greater than ours, the contractile-induced reduction in muscle blood flow observed in the present
investigation was most likely not due to duration of
contraction. It is possible that the higher absolute
forces generated in the present investigation occluded
the intramuscular vasculature more completely and
thus inhibited blood flow to a greater extent. It is
important to mention that the data we used for the
comparison of peak forces produced were taken from
the original thesis written by Naamani (32) because we
believe there was an error in the forces published in
the journal article (33).
When the blood flows during each contraction within
a series of experiments are compared, we see that the
venous outflow during the first contraction is greater
than the outflow of each subsequent contraction. This
indicates that either the contractile-induced inhibition
of muscle perfusion was additive throughout the first
two contractile cycles or that the volume of blood contained within the muscle’s compliant vasculature was
greatest before the first contraction. When the venous
outflow values were converted into volumes of blood
flowing from the muscle, the change between the first
and subsequent contractions was only ⬃0.4 ml per
contraction. On the basis of this evidence, it seems
likely that the perfusion of the muscle was inhibited as
completely by the first contraction as it was by each
contraction thereafter, yet the venous outflow during
the first contraction was greater due to compression of
the compliant region (see Table 4). More specifically,
the muscle’s compliant vasculature had swelled with
blood during the hyperemia and stop periods of series 1
and 2, respectively, and then this blood was rapidly
expelled when contractions were initiated. During the
second, third, fourth, and so forth contractions, the
rhythmic compression of the vasculature prevented the
compliant region from fully refilling.
In terms of limitations to the present investigation,
we measured muscle venous outflow and therefore our
results do not provide specific data to indicate how the
flow of blood into the muscle is affected by rhythmic
contractions. According to the muscle pump theory,
intense rhythmic contractions enhance venous outflow
by emptying the muscle’s compliant region, and so the
net flow of blood out of the muscle should be at least as
great as the net inflow. Because we observed decreases
17
18
EVIDENCE AGAINST THE MUSCLE PUMP
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
We appreciate the assistance provided by Drs. David Pascoe, Dean
Schwartz, Lawrence Wit, and Robert Keith.
25.
REFERENCES
1. Ameredes BT and Provenzano MA. Regional intramuscular
pressure development and fatigue in the canine gastrocnemius
muscle in situ. J Appl Physiol 83: 1867–1876, 1997.
2. Anrep GV and Saalfeld EV. The blood flow through the skeletal muscle in relation to its contraction. J Physiol 85: 375–399,
1935.
3. Armstrong RB, Delp MD, Goljan EF, and Laughlin MH.
Distribution of blood flow in muscles of miniature swine during
exercise. J Appl Physiol 62: 1285–1298, 1987.
4. Armstrong RB, Delp MD, Goljan EF, and Laughlin MH.
Progressive elevations in muscle blood flow during prolonged
exercise in swine. J Appl Physiol 63: 285–291, 1987.
5. Armstrong RB, Vandenakker CB, and Laughlin MH. Muscle blood flow patterns during exercise in partially curarized
rats. J Appl Physiol 58: 698–701, 1985.
6. Braakman R, Sipkema P, and Westerhof N. Two zero-flow
pressure intercepts exist in autoregulating isolated skeletal
muscle. Am J Physiol Heart Circ Physiol 258: H1806–H1814,
1990.
7. Corcondilas A, Koroxenidis GT, and Shepherd JT. Effect of
brief contraction of forearm muscles on forearm blood flow.
J Appl Physiol 19: 142–146, 1964.
8. Delp MD. Control of skeletal muscle perfusion at the onset of
dynamic exercise. Med Sci Sports Exerc 31: 1011–1018, 1999.
9. Folkow B, Gaskell P, and Waaler BA. Blood flow through
limb muscles during heavy rhythmic exercise. Acta Physiol
Scand 80: 61–72, 1970.
10. Folkow B, Haglund U, Jodal M, and Lundgren O. Blood flow
in the calf muscle of man during heavy rhythmic exercise. Acta
Physiol Scand 81: 157–163, 1971.
11. Gladden LB. Net lactate uptake during progressive steadylevel contractions in canine skeletal muscles. J Appl Physiol 71:
514–520, 1991.
12. Gladden LB, Crawford RE, and Webster MJ. Effect of blood
flow on net lactate uptake during steady-level contractions in
canine skeletal muscle. J Appl Physiol 72: 1826–1830, 1992.
13. Gladden LB, Crawford RE, and Webster MJ. Effects of
lactate concentration and metabolic rate on net lactate uptake by
canine skeletal muscle. J Appl Physiol 266: R1095–R1101, 1994.
14. Gladden LB, Grassi B, Kelley KM, Aschenbach WG, Hamann JJ, Evans RK, Pattillo RE, and Hogan MC. Blood flow
J Appl Physiol • VOL
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
during and between tetanic contractions in canine skeletal muscle (Abstract). Med Sci Sports Exerc 31: S353, 1999.
Gladden LB and Yates JW. Lactic acid infusion in dogs: effects
of varying infusate pH. J Appl Physiol 54: 1254–1260, 1983.
Gorczynski RJ, Klitzman B, and Duling BR. Interrelations
between contracting striated muscle and precapillary microvessels. Am J Physiol Heart Circ Physiol 235: H494–H504, 1978.
Hamann JJ, Valic Z, Buckwalter JB, and Clifford PS.
Muscle pump does not enhance blood flow in exercising skeletal
muscle. J Appl Physiol 94: 6–10, 2003.
Jackman AP and Green JF. A theoretical description of arterial pressure-flow relationships with verification in the isolated
hindlimb of the dog. Ann Biomed Eng 18: 89–101, 1990.
Joyner MJ and Proctor DN. Muscle blood flow during exercise: the limits of reductionism. Med Sci Sports Exerc 31: 1036–
1040, 1999.
Kelley KM, Hamann JJ, Aschenbach WG, and Gladden LB.
Canine gastrocnemius muscle in situ: V̇O2 max (Abstract). Med
Sci Sports Exerc 28: S62, 1996.
Kirkebo A and Wisnes A. Regional tissue fluid pressure in rat
calf muscle during sustained contraction or stretch. Acta Physiol
Scand 114: 551–556, 1982.
Kjellmer I. An indirect method for estimating tissue pressure
with special reference to tissue pressure in muscle during exercise. Acta Physiol Scand 62: 31–40, 1964.
Laughlin MH. Skeletal muscle blood flow capacity: role of
muscle pump in exercise hyperemia. Am J Physiol Heart Circ
Physiol 253: H993–H1004, 1987.
Laughlin MH and Armstrong RB. Muscular blood flow distribution patterns as a function of running speed in rats. Am J
Physiol Heart Circ Physiol 243: H296–H306, 1982.
Laughlin MH, Korthuis RJ, Duncker DJ, and Bache RJ.
Control of blood flow to cardiac and skeletal muscle during
exercise. In: Handbook of Physiology. Exercise: Regulation and
Integration of Multiple Systems. Control of Respiratory and Cardiovascular Systems. Bethesda, MD: Am. Physiol. Soc., 1996,
sect. 12, chapt. 16, p. 705–769.
Laughlin MH and Schrage WG. Effects of muscle contraction
on skeletal muscle blood flow: when is there a muscle pump? Med
Sci Sports Exerc 31: 1027–1035, 1999.
Leyk D, E␤feld D, Baum K, and Stegemann J. Early leg
blood flow adjustment during dynamic foot plantar flexions in
upright and supine body positions. Int J Sports Med 15: 447–
452, 1994.
Mackie BG and Terjung RL. Influence of training on blood
flow to different skeletal muscle fiber types. J Appl Physiol 55:
1072–1078, 1983.
Magder S. Venous mechanics of contracting gastrocnemius
muscle and the muscle pump theory. J Appl Physiol 79: 1930–
1935, 1995.
Marshall JM and Tandon HC. Direct observations of muscle
arterioles and venules following contraction of skeletal muscle
fibers in the rat. J Physiol 350: 447–459, 1984.
Musch TI. Skeletal muscle blood flow in exercising dogs. Med
Sci Sports Exerc 20: S104–S108, 1988.
Naamani R. The mechanical effects of contractions on blood
flow to the muscle (Abstract). Masters Abstracts Int 30: 1321,
1992. (UMI Dissertation Services AAT MM67692)
Naamani R, Hussain SN, and Magder S. The mechanical
effects of contractions on blood flow to the muscle. Eur J Appl
Physiol 71: 102–112, 1995.
Naik JS, Valic Z, Buckwalter JB, and Clifford PS. Rapid
vasodilation in response to a brief tetanic muscle contraction.
J Appl Physiol 87: 1741–1746, 1999.
Permutt S and Riley RL. Hemodynamics of collapsible vessels
with tone: the vascular waterfall. J Appl Physiol 18: 924–932,
1963.
Petrofsky JS, Phillips CA, Sawka MN, Hanpeter D, and
Stafford D. Blood flow and metabolism during isometric contractions in cat skeletal muscle. J Appl Physiol 50: 493–502,
1981.
Pollack AA and Wood EH. Venous pressure in the saphenous
vein at the ankle in man during exercise and changes in posture.
J Appl Physiol 1: 649–662, 1949.
94 • JANUARY 2003 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
used as indirect evidence for a muscle pump effect (3, 4,
23, 24, 28). Despite these arguments, we cannot be
absolutely certain that our results preclude operation
of the muscle pump under muscle contraction patterns
in vivo. However, it is important to note that data from
the companion manuscript (17) provide evidence that
the muscle pump also does not elevate hindlimb blood
flow during the spontaneous contractions of rhythmic
locomotor exercise in intact, awake dogs.
In conclusion, given the limitations discussed above,
the results of the present investigation suggest that
the mechanical action of rhythmic, synchronous, maximal tetanic skeletal muscle contractions reduces peak
muscle perfusion. Therefore, maximal skeletal muscle
vascular conductance under such conditions is accomplished exclusively via dilation of the muscle’s vascular
bed, most likely in response to endothelial-derived nitric oxide and/or a combination of accumulated metabolites (25). The hemodynamic benefit of rhythmic contractions during maximal muscle vasodilation may be
to compress the compliant region, thereby facilitating
venous return and possibly reducing tissue edema by
decreasing capillary filtration pressures (29, 33).
EVIDENCE AGAINST THE MUSCLE PUMP
38. Radegran G and Saltin B. Muscle blood flow at onset of
dynamic exercise in humans. Am J Physiol Heart Circ Physiol
274: H314–H322, 1998.
39. Sheriff DD, Rowell LB, and Scher AM. Is rapid rise in
vascular conductance at onset of dynamic exercise due to muscle
pump? Am J Physiol Heart Circ Physiol 265: H1227–H1234,
1993.
40. Shiotani I, Sato H, Sato H, Yokohama H, Ohnishi Y,
Hishida K, Kinjo K, Nakatani D, Kuzuya T, and Hori M.
Muscle pump-dependent self-perfusion mechanism in legs in
normal subjects and patients with heart failure. J Appl Physiol
92: 1647–1654, 2002.
41. Shoemaker JK, Tschakovsky ME, and Hughson RL. Vasodilation contributes to the rapid hyperemia with rhythmic contractions in humans. Can J Physiol Pharmacol 76: 418–427, 1998.
19
42. Shrier I and Magder S. Maximal vasodilation does not eliminate the vascular waterfall in the canine hindlimb. J Appl
Physiol 79: 1531–1539, 1995.
43. Stainsby WN and Otis AB. Blood flow, blood oxygen tension,
oxygen tension, oxygen uptake, and oxygen transport in skeletal
muscle. Am J Physiol 206: 858–866, 1964.
44. Stegall HF. Muscle pumping in the dependent leg. Circ Res 19:
180–190, 1966.
45. Toska K and Eriksen M. Peripheral vasoconstriction shortly
after onset of moderate exercise in humans. J Appl Physiol 77:
1519–1525, 1994.
46. Tschakovsky ME, Shoemaker JK, and Hughson RL. Vasodilation and muscle pump contribution to immediate exercise
hyperemia. Am J Physiol Heart Circ Physiol 271: H1697–H1701,
1996.
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
J Appl Physiol • VOL
94 • JANUARY 2003 •
www.jap.org