See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/13266063

Changes in surface EMG parameters during

static and dynamic fatiguing contraction

Article in Journal of Electromyography and Kinesiology · March 1999

Impact Factor: 1.65 · DOI: 10.1016/S1050-6411(98)00021-2 · Source: PubMed

CITATIONS READS

250 258

5 authors, including:

Kazumi Masuda Tadashi Masuda

Kanazawa University Fukushima University
66 PUBLICATIONS 867 CITATIONS 90 PUBLICATIONS 1,793 CITATIONS

SEE PROFILE SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Kazumi Masuda
letting you access and read them immediately. Retrieved on: 12 May 2016
 Journal of Electromyography and Kinesiology 9 (1999) 39–46

Changes in surface EMG parameters during static and dynamic

fatiguing contractions
a,*
Kazumi Masuda , Tadashi Masuda b, Tsugutake Sadoyama c, Mitsuharu Inaki d,
Shigeru Katsuta e
a
Doctoral Program in Health and Sport Sciences, University of Tsukuba, 1-1-1-Tennodai, Tsukuba City, Ibaraki 305-8574, Japan
b
National Institute of Bioscience and Human Technology, 1-1 Higashi, Tsukuba City, Ibaraki 305-0046, Japan
c
Department of Kansei Engineering, Faculty of Textile Science and Technology, Shinshu University, 3-15-1 Tokida, Ueda City, Nagano 386-
0018, Japan
d
Seinan Jo Gakuin University, 1-3-5 Ibori, Ogurakita, Kitakyusyu City, Fukuoka 803-0835, Japan
e
Institute of Health and Sport Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba City, Ibaraki 305-8574, Japan

Received 10 March 1997; received in revised form 16 January 1998; accepted 6 May 1998

Abstract

The effect of contraction types on muscle fiber conduction velocity (MFCV), median frequency (MDF) and mean amplitude
(AMP) of surface electromyography was examined in the vastus lateralis of 19 healthy male adults. The subjects performed knee
extension both statically and dynamically until they were exhausted. The static contraction was a sustained isometric extension of
the knee at a joint angle of 90° with 50% of the maximum voluntary contraction (MVC) load. The dynamic contraction was a
repetitive isotonic extension of the knee between the angles of 90° and 180° with the same 50% MVC load at a frequency of 10
times per minute. MFVC during the static contraction significantly decreased during the exercise (p ⬍ 0.01). On the other hand,
MFVC during the dynamic contraction did not significantly change throughout the exercise. MDF decreased and AMP increased
during both types of contractions (p ⬍ 0.01). Because the blood flow within the muscle is maintained during the dynamic contraction
by enhanced venous return from the contracting muscle, these results suggested that MFVC is affected by the metabolic state in
the muscle and the changes in MDF cannot be explained only by that of MFVC.  1999 Elsevier Science Ltd. All rights reserved.

Keywords: Dynamic; Fatigue; Isometric; Median frequency; Muscle fiber conduction velocity; Surface EMG

1. Introduction longed muscle contraction This phenomenon has further

Surface electromyographic (EMG) signals normally
show random waveforms, because they represent a sum
of action potentials from many independently activated
motor units. It has only been possible to estimate a
degree of muscular activity based on the average ampli-
tude of surface EMGs in spite of the efforts of many
researchers who tried to extract information on muscle
functions from EMG waveforms [1]. With the develop-
ment of computer technology, it has become feasible to
apply the technique of frequency analysis to EMG sig-
nals. As a result, the power spectrum of the EMG signal
was found to shift toward the lower band during pro-
* Corresponding author. Tel.: ⫹ 81-298-53-2762; fax: ⫹ 81-2398-
53-6507; e-mail: kazumi@taiiku.tsukuba.ac.jp.
been used to estimate localized muscular fatigue [1].
The shift of the spectrum toward the lower band is
caused by a decrease in muscle fiber conduction velocity
(MFCV) [2–5]. The decrease in MFCV is then due to
an accumulation of metabolic byproducts such as lactic
acid, which reduces intracellular pH and decreases the
excitability of the muscle fiber membrane.
In recent years it has become easy to measure MFCV
using a surface electrode array [5–8], which enables us
to directly examine the relationship between MFCV and
the power spectrum [2,3,9]. Many researchers observed
a positive correlation between the EMG parameters dur-
ing static prolonged contraction [9–13].
Zwarts et al. [12] further reported that the correlation
between MFCV and the power spectrum disappears
under ischemia. They measured the surface EMG para-
meters during the recovery period after an isometric pro-

1050-6411/99/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved.
PII: S 1 0 5 0 - 6 4 1 1 ( 9 8 ) 0 0 0 2 1 - 2
40 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46

longed contraction and compared the normal and isch-

emic conditions. The results showed that the median
frequency (MDF) of the power spectrum recovered with
time, but MFCV did not go back to the normal level
before the ischemia. This finding indicates that MDF
does not always change in accordance with MFCV and
suggests that the blood flow in a muscle determines the
relationship between MFCV and MDF.
Except for the study by Zwarts et al. [12], the relation-
ship between MFCV and the power spectrum has been
investigated only under static contraction. It is important
to clarify the changes in EMG parameters during
dynamic contraction for quantitative analyses of fatigue
in sport and labor. Dynamic contraction, which includes
the stretching and shortening of a muscle, should
enhance the blood flow by the enhanced venous return Fig. 1. Schematic illustration of the apparatus for the experiment.
from the contracting muscle. The enhanced blood flow
removes metabolic byproducts and contributes to the
inhibition of the decrease in intracellular pH. The differ- and 2). During the dynamic fatigue test, the frequency
ence in the intracellular state may affect the changes in of the exercise was regulated by a metronome sound at
MFCV and MDF. Therefore, in the present study we six times per minute (Figs. 1 and 2). EMG was recorded
compared the changes in MFCV, MDF and other EMG over the whole region of the change range for the knee
parameters during both static and dynamic fatiguing con- joint. When the joint angle was changed, a motion arti-
tractions. fact prevented derivation of a clear EMG. Muscle fiber
conduction velocity (MFCV) may also be varied during
changes in the joint angle. To avoid these effects, the
2. Methods subjects stopped their leg at 90° of knee joint for 1 s just
before kicking out their feet. This contraction pattern
2.1. Subjects also prevents the effect of inertia during the acceleration
of the weight. EMG signals detected during this brief
The subjects were 19 healthy male adults aged 19–73 stop were analyzed.
years. They were informed of the content and risk of the
experiment in advance and gave written agreement to 2.3. Recording of EMG
voluntarily participate in the experiments. The body
height and body mass are 174.0 ⫾ 4.6 cm and 66.9 ⫾ EMG signals were detected with a linear surface elec-
4.3 kg (mean ⫾ SD), respectively. trode array from the right vastus lateralis (Fig. 3). The
electrode array consisted of 11 stainless-steel contacts
2.2. Experimental protocol 1 mm thick, 10 mm wide and spaced at 5 mm intervals

Prior to the experiment, the maximum voluntary con-

traction (MVC) force of the knee extensor muscles at
90° of knee joint was determined using an instrumen-
tation chair (Musculator GT-30, OG-Giken, Okayama,
Japan) within a few trails. The maximal value of MVC
was used as the reference value. The mean ⫾ SD of
MVC was 66.7 ⫾ 11.6 kg. Based on the result of the
MVC test, a load of 50% MVC was determined for each
subject. A weight of 50% MVC was attached to the
ankle of the subject through a steel wire (Fig. 1).
The fatigue test with the 50% MVC weight consisted
of two types of contractions. The first test was an iso-
metric contraction in which the subjects sustained their
knee joint at 90% until their exhaustion. After 30 min
of rest, the subjects performed the second test of repeti-
tive knee extension until they were exhausted. The range Fig. 2. Photograph of the detectors for EMG, muscle temperature and
of motion was from 90° to the full extension (Figs. 1 angle of knee joint placed around the knee.
 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46
Fig. 3. Schematic illustration of the electrode array and raw EMG signals.
(Fig. 3). Ten EMG signals were simultaneously derived
from pairs of contacts adjacent in the array. The ampli-
fier had a gain of 2500 over a frequency range of 5–
1000 Hz. The detected signals were recorded on a digital
tape (5870S, NF Electronic Instruments, Tokyo, Japan)
(Fig. 1). A hum filter was not used to make the sub-
sequent frequency analysis possible.
Signals of goniometer and muscle temperature were
also recorded (Fig. 3). The goniometer was placed on
the lateral side of the knee in order to electrically record
the joint angle (Figs. 1 and 2). The pickup of the ther-
mometer was attached to the skin about 1 cm lateral to
the electrode array.
2.4. Measurement of muscle temperatures
We used two types of thermometers (CTM-201,
TERMO, Tokyo, Japan). One was a conventional ther-
mometer using a thermistor device for measuring the
temperature on the skin surface. The other was a ther-
mometer for measuring the core temperature. It had a
heater to compensate for the heat flow from a human
body and measure the temperature at a deeper position.
The range of the thermometer was set from 20 to 40°C.
The frequency range was from DC to 30 Hz.
41
2.5. Analysis of EMG signal
After the experiment, the recorded signals were digit-
ized with an AD converter (DASBOX-8400A, System
Design Service Co. Ltd, Tokyo, Japan) at a rate of 5 kHz
with a 12-bit resolution. This A/D converter had a
sample/hold circuit for each input channel, which made
an exactly simultaneous sampling possible. The sampled
signals were then transferred to a computer workstation
(VAX Station II, Nippon Digital Equipment Co. Ltd,
Tokyo, Japan) (Fig. 1). The transferred data were auto-
matically processed. MFCV was calculated by the cross-
correlation technique [14].
The cross-correlation was applied to two EMG signals
recorded at positions 15 mm apart. The electrodes for
these two signals were at least 20–30 mm away from
the motor endplate regions. The position of the motor
endplates was identified from the pattern of bi-direc-
tional propagation in motor unit action potentials. Corre-
lation coefficients (CC) were not lower than 0.9 in most
cases. For calculation of MFCV, signals with CC not
lower than 0.85 were adopted. During dynamic contrac-
tion, CCs were slightly lowered and sometimes became
lower than 0.9.
Signals were sampled every 2 s during the static test
and every 6 s during the dynamic test. The number of
42 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46

sample points for each channel was 4096, giving a signal

duration of 0.8192 s. MDF and the average amplitude
(AMP) were calculated from signals that were the sum
of four consecutive channels. This summation gave an
equivalent electrode spacing of 20 mm. Although both
the mean power frequency and MDF have been used as
indices of the EMG power spectrum, MDF is less affec-
ted by noise [15]. AMP was calculated as the root mean
square value of the summed signal.
MFCV, MDF and AMP were normalized by values
obtained at 100% MVC contraction, because the absol-
ute level of each parameter differed among the subjects.
The mean and standard error (SE) were calculated from
the normalized values of 19 subjects.

2.6. Statistical analysis
All results are given in mean ⫾ SE, unless otherwise
noted. The difference between the two types of contrac-
tions was assessed by the two-tailed t-test for paired
observations. Changes in EMG parameters along time
were assessed by regression analysis. The level of sig-
nificance was set at p ⬍ 0.05.
3. Results
3.1. Endurance time
The endurance time of the static contraction ranged
from 54–136 s and had a mean 75.7 ⫾ 20.2 s, while that
of the dynamic contraction ranged from 84 to 258 s with
a mean 149.7 ⫾ 50.9 s. Because the endurance time
varied between subjects, we normalized the time in the
contraction by making the endurance time as 100%. The
endurance time of each subject was divided into 10 equal
segments. The values of MFCV, MDF and AMP were
resampled at 11 points at every 10% time including the
initial and last period of the contraction. When there was
no measurement at the resampled point, an interpolated
value was calculated from the nearest sampled values.
The resampled MFCV, MDF and MDF were averaged
over the 19 subjects.
Fig. 4. Changes in the relative muscle fiber conduction velocity
(MFCV) during static (open circle) and dynamic (closed circle) con-
tractions. MFCV is normalized by the value at 100% MVC. MFCV
showed no significant change during dynamic contraction, while it
decreased significantly during static contraction (p ⬍ 0.01). Values are
mean ⫾ SE (standard error) of the 19 subjects. Asterisks show signifi-
cant differences between the static and dynamic contractions (*p ⬍
0.05; **p ⬍ 0.01).
3.3. Median frequency
Fig. 5 shows the changes in MDF during the two types
of contractions. During the static contraction, MDF
decreased 22.4% from the initial value (p ⬍ 0.01:
regression analysis). A similar tendency was observed
during the dynamic contraction, in which MDF
decreased 15.2% from the initial value (p ⬍ 0.01:
regression analysis). The dynamic contraction showed
significantly higher values of MDF than the static con-
traction at the last stage of the exercise (p ⬍ 0.05).

3.2. Muscle fiber conduction velocity

Fig. 4 shows MFCV during the two types of contrac-

tions. MFCV in the static contraction rose about 3.4%
immediately after the beginning of the exercise. It then
decreased 7.4% from the initial value toward the end of
the exercise (p ⬍ 0.01: regression analysis). On the other
hand, MFCV during the dynamic contraction did not sig-
nificantly change throughout the exercise. Consequently,
MFCV in the dynamic contraction showed significantly
higher values than that in the static contraction at the
last stage of the exercise (p ⬍ 0.05).
Fig. 5. Changes in the relative value of median frequency (MDF)
during static (open circle) and dynamic (closed circle) contractions.
MDF is normalized by the value at 100% MVC. MDF decreased sig-
nificantly during both types of contractions (p ⬍ 0.01). Values are in
mean ⫾ SE (standard error) of the 19 subjects. Asterisks show signifi-
cant differences between the static and dynamic contractions (*p ⬍
0.05; **p ⬍ 0.01).
 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46
3.4. Amplitude
Fig. 6 shows the relative values of AMP obtained
from the two types of contractions. AMP increased from
the beginning to the end of the exercise during both
types of contractions (p ⬍ 0.01: regression analysis).
The increment from the initial value was 34.4% in the
static contraction and 48.0% in the dynamic contraction.
The AMP of the dynamic contraction showed higher
values than that of the static contraction throughout the
exercise (p ⬍ 0.05).
3.5. Muscle temperature
Fig. 7 shows the changes in the surface and deep tem-
peratures. The surface and deep temperatures during the
static contraction decreased 0.045°C and 0.007°C,
respectively. On the other hand, those during the
dynamic contraction increased about 0.121°C and
0.251°C, respectively. No significant difference was
observed in the changes of the muscle temperatures per-
haps because of the large variance among the subjects.
During the dynamic contraction, both the surface and
deep temperatures were about 0.41–0.78°C higher than
those during the static contraction throughout the exer-
cise (p ⬍ 0.05).
4. Discussion
4.1. Variability in the endurance time
Differences in endurance time among the subjects are
considered to be caused by individual differences in
Fig. 6. Changes in the relative value of amplitude (AMP) during
static (open circle) and dynamic (closed circle) contractions AMP is
normalized by the value at 100% MVC. AMP increased significantly
during both types of contractions (p ⬍ 0.01). AMP during the dynamic
contraction had a higher value than that during the static contraction
(p ⬍ 0.05). Values are in mean ⫾ SE (standard error) of the 19 sub-
jects. Asterisks show significant differences between the static and
dynamic contractions (*p ⬍ 0.05; **p ⬍ 0.01).
43
Fig. 7. Changes in surface (square) and deep (circle) temperatures
during static (open) and dynamic (closed) contractions. No significant
tendency was observed during both types of contractions. Both surface
and deep temperatures during dynamic contraction had higher values
than those during static contraction (p ⬍ 0.05). Values are mean ⫾
SE (standard error) of the 19 subjects.
resistance to fatigue due to contraction properties of the
skeletal muscles represented by muscle fiber compo-
sition, enzyme activities and differences in metabolic
systems. Since changes in properties of these factors
depend on time elapsed, it may be better to show the
changes in the EMG signals in absolute time. However,
it was more convenient to normalize time in order to
show the changes in the EMG signals more clearly for
the 19 subjects and also for a comparison of differences
in the contraction types.
4.2. Changes in EMG parameters during the static
contraction
MFCV and MDF decreased (Figs. 4 and 5) while
AMP increased (Fig. 6) during sustained static contrac-
tion. The simultaneous decrease in MFCV and MDF has
been reported by several researchers [6,9,11,16,17]. The
rate of the decrement was 7.4% for MFCV and 22.4%
for MDF. The decrement of MDF was, therefore, larger
than that of MFCV. This difference has also been
reported in previous studies [18,19].
4.3. Changes in EMG parameters during the dynamic
contraction
Different changes in EMG parameters were observed
during the dynamic contraction (Figs. 4 and 5). MFCV
did not change throughout the exercise during the
dynamic contraction. Consequently, during the dynamic
contraction, MFCV had significantly higher values than
those during the static contraction at the last stage of the
exercise (Fig. 4). On the other hand, MDF during the
dynamic contraction decreased with time, as in the case
of the static contraction. However, at the last stage of the
44 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46
exercise, MDF during the dynamic contraction showed
significantly higher values than those during the static
contraction. MDF decreased 15.2% from the initial value
during the dynamic contraction, while it decreased
22.4% during the static contraction. The difference of
7% in MDF between the two types of contractions corre-
sponded to the difference in MFCV. Therefore, 7% of
the total decrease in MDF during the static contraction
may be caused by the decrease in MFCV.
4.4. Factors affecting the EMG parameters during the
two types of contractions
The most plausible factor that caused the difference
between the static and dynamic contractions is the blood
flow in the contracting muscle. The intramuscular press-
ure during the static contraction prevents the blood flow,
and the metabolic byproducts, such as lactic acid,
accumulate in the muscle. On the other hand, the
dynamic contraction, which includes the stretching and
shortening of the muscle, maintains the blood flow by
enhanced venous return from the contracting muscle, and
the blood flow removes the metabolic byproducts.
It should be noted that the contraction force and mus-
cle temperature may affect MFCV and cannot be neg-
lected. In this study, we recorded the amplitude of EMG
and the muscle temperature and confirmed that the
strength of contraction or the muscle temperature is not
a primary factor that caused the difference in MFCV and
MDF during the different types of contractions.
The subjects performed the static contraction first
throughout the present experiment. This procedure might
have caused some bias on the results. However, since
changes in the position of the electrode deteriorate the
reproducibility of electromyograms (EMG), all measure-
ments had to be performed in a day for respective sub-
jects. Moreover, if a dynamic contraction test is perfor-
med in advance, it requires considerable time to recover
from fatigue. We would like to take this point into con-
sideration in the next experiment.
4.5. Effect of blood flow
The blood flow may affect MFCV and MDF by
changing the intramuscular pH or the concentration of
potassium ion (K+). These ions influence the excitability
of the muscle fiber membrane and affect MFCV
[4,11,20]. In fact, Zwarts and Arendt-Nielsen [13]
reported that the peripheral occlusion occurs in a con-
tracting muscle during an isometric contraction with a
load of 50% MVC or greater. Jennische [21] showed in
an animal experiment that the clearance of lactic acid
and K+ from the working muscle enhances the recovery
of MFCV.
4.6. Effect of lactic acid
Under ischemia, oxygen supply to an active muscle
is prevented, and lactic acid rapidly accumulates due to
enhanced glycolytic metabolism. The accumulation of
lactic acid decreases pH in a contracting muscle. The
accumulation also inhibits the excitability of the muscle
membrane [4,13]. The produced H+ is consumed chemi-
cally and metabolically in the muscles. The buffering
system by bicarbonate (HCO3) has a significant role in
the homeostasis of the pH10. This buffering system of
HCO3 depends on the intracellular pressure of CO2
(PCO2). Because intracellular PCO2 increases under the
ischemic condition, this buffering system cannot work
effectively during the prolonged static contraction [22].
On the other hand, during dynamic contraction, the
oxygen supply is maintained through the blood flow, and
the buffering system of HCO3 works effectively because
an excessive increase in PCO2 does not occur [22]. Fur-
thermore, a decrease in pH can be inhibited by the
removal of H+ with the blood flow [23]. Therefore, the
intramuscular pH and the excitability of the muscle
membrane are stable during dynamic contraction, and
consequently, MFCV does not decrease with time.
4.7. Effect of K+ concentration
The other metabolic factor that may cause the differ-
ence between the static and dynamic contractions is K+
concentration ([K+]) in the serum. The conduction of
excitation along the muscle fiber is affected by the gradi-
ent of potentials across the muscle fiber membrane. The
serum [K+] increases during sustained muscle contrac-
tion [23–25] because the blood flow in a muscle is pre-
vented by the contraction. Then the gradient across the
muscle fiber membrane decreases and the excitability of
the muscle fiber membrane also decreases [26]. The
change of Na+/K+ conductance becomes slower [26] and
the muscle fiber conduction velocity decreases. During
dynamic contraction the blood flow removes K+ from
the contracting muscle. Consequently, the excitability of
the muscle fiber membrane is not inhibited and MFCV
does not decrease.
4.8. Effect of amplitude
The difference in AMP between the static and
dynamic contractions is not related to the difference in
MFCV. Although there was a significant difference in
the absolute value of AMP throughout the contraction
(Fig. 6), the ratio of change from the beginning to the
end of the exercise showed no significant difference
between the two types of contractions. For example, both
in the static and dynamic contractions, AMP at 80% of
time increased about 30% from the initial value, when
MFCV showed significant difference between the two
 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46
types of contractions. In addition, at the beginning of the
exercise AMP showed a significant difference between
the contractions, but there was no difference in MFCV.
The increase in AMP is, therefore, not directly related
to the dissociation between MFCV and MDF.
4.9. Effect of muscle temperature
The different changes in MFCV during the two types
of contractions are not due to muscle temperature. It is
reported that MFCV increases with muscle temperature
[27–29]. It is, therefore, possible that muscle temperature
influences the change in MFCV. In this study, the sur-
face and deep temperatures during the static contraction
decreased 0.045°C and 0.007°C, respectively, whereas
those during the dynamic contraction increased 0.121°C
and 0.251°C, respectively. Those changes were, how-
ever, statistically not significant.
Morimoto et al. [27] reported that the rate of increase
in MFCV is 0.2 m/s/°C, while the rate reported by Troni
and Contegiacomo [29] is 0.13 m/s/°C. Based on the rate
reported by Morimoto et al. [27] MFCV decreases
0.0090 or 0.0014 m/s during the static contraction and
increases 0.0242 or 0.0502 m/s during the dynamic con-
traction. The estimated change in MFCV becomes
smaller, if we use the rate reported by Troni and Conteg-
iacomo [29].
The muscle temperature during the dynamic contrac-
tion was always higher, about 0.41–0.78°C, than that
during the static contraction (p ⬍ 0.05). This difference
may be caused by the order of the contractions and insuf-
ficient rest between contractions. However, it equally
influences both MFCV and MDF. The dissociation
between MFCV and MDF is, therefore, not due to the
difference in muscle temperature.
4.10. Relationship between MFCV and MDF
The decrease in MDF without change in MFCV dur-
ing the dynamic contraction indicates that MFCV is not
a unique factor for determining the change in MDF. The
shift of MDF to lower frequencies is caused by the pro-
longation of the duration of MUAPs, which has been
considered to be caused by the decrease in MFCV [9–
11,13]. However, the dissociation of MFCV and MDF
during the dynamic contraction indicates that the change
in MDF is not explained by the relationship between
MFCV and the duration of the MUAPs.
MFCV and the duration of the MUAPs can be inde-
pendent from each other, and MFCV does not change,
even if the duration of the MUAPs is prolonged and
MDF decreases. Prolongation of the duration of MUAPs
without the change in MFCV means that the depolariz-
ation zone along the muscle fiber becomes longer.
Another possible reason which could explain the dis-
45
sociation between MFCV and MDF is a non-uniform
change of the MFCVs of different active motor units [3].
To clarify the mechanism, further investigation is
necessary, for example, by measuring MFCV and the
extent of the depolarization zone during an electrical
stimulation of the muscle. Ischemia condition during the
dynamic contraction is also necessary to clarify the
mechanism of changes in the EMG parameters.
5. Conclusions
MDF decreased and AMP increased during both the
static and dynamic contractions. These results agreed
with the previous studies. On the other hand, MFCV sig-
nificantly decreased during the static contraction, but did
not decrease during the dynamic contraction. This dis-
crepancy is caused by the difference in blood flow,
which is maintained during the dynamic contraction and
removes the metabolic byproducts. The dissociation
between MFCV and MDF during the dynamic contrac-
tion indicates that MFCV is not a unique factor for
determining the changes in the power spectrum. To
make clear the effect of blood flow, it is necessary to
perform the same measurements under ischemic con-
ditions during the dynamic contraction in the next
experiments.
References
[1] Basmajian JV, De Luca CJ. Muscles alive, 5th ed. Baltimore:
Williams & Wilkins, 1985:125–7.
[2] Gamet D, Duchéne J, Goubel F. Reproducibility of kinetics of
electromyogram spectrum parameters during dynamic exercise.
Eur J Appl Physiol 1996;74:504–10.
[3] Merletti R, Roy SH. Myoelectric manifestations of muscle
fatigue: a simulation study. Proc ISEK 1996;176–7.
[4] Mortimer JT, Magnusson R, Petersén I. Conduction velocity in
ischemic muscle: effect on EMG frequency spectrum. Am J Phy-
siol 1970;219:1324–9.
[5] Rau G, Disselhorst-Klug C, Silny J. Noninvasive approach to
motor unit characterization: muscle structure, membrane dynam-
ics and neuronal control. J Biomech 1997;30:441–6.
[6] Lindström L, Magnusson R, Petersén I. Muscular fatigue and
action potential conduction velocity changes studied with fre-
quency analysis of EMG signals. Electromyogr Clin Neurophy-
siol 1970;10:341–51.
[7] Lynn PA. Direct on-line estimation of muscle fiber conduction
velocity by surface electromyography. IEEE Trans Biomed Eng
1979;BME26:564–71.
[8] Masuda T, Sadoyama T. Skeletal muscles from which the propa-
gation of motor unit action potentials is detectable with surface
electrode array. Electroencephalogr Clin Neurophysiol 1987;
67:421–7.
[9] Sadoyama T, Masuda T, Miyano H. Relationships between mus-
cle fiber conduction velocity and frequency parameters of surface
EMG during sustained contraction. Eur J Appl Physiol
1983;51:247–56.
[10] Arendt-Nielsen L, Forster A, Mills KR. EMG power spectral shift
46 K. Masuda et al. / Journal of Electromyography and Kinesiology 9 (1999) 39–46
and muscle-fiber velocity conduction during human muscle
fatigue. J Physiol 1984;353:54.
[11] Arendt-Nielsen L, Mills KR, Forster A. Changes in muscle fiber
conduction velocity, mean power frequency, and mean EMG
voltage during prolonged submaximal contractions. Muscle
Nerve 1989;12:493–7.
[12] Zwarts MJ, Van Weerden TW, Haenen HTM. Relationship
between average muscle fiber conduction velocity and EMG
power spectra during isometric contraction, recovery and applied
ischemia. Eur J Appl Physiol 1987;56:212–6.
[13] Zwarts MJ, Arendt-Nielsen L. The influence of force and circu-
lation on average muscle fiber conduction velocity during local
muscle fatigue. Eur J Appl Physiol 1988;58:278–83.
[14] Nishizono H, Saito Y, Miyashita M. The estimation of conduction
velocity in human skeletal muscle in situ with surface electrodes.
Electroencephalogr Clin Neurophysiol 1979;46:659–64.
[15] De Luca CJ, Knaflitz M. Surface electromyography: What’s new?
Torino: CLUT, 1992:23–38.
[16] Arendt-Nielsen L, Mills KR. Muscle fiber conduction velocity,
mean EMG voltage and force during submaximal fatiguing con-
tractions of human quadriceps. Eur J Appl Physiol 1988;58:20–5.
[17] Eberstein A, Beattie B. Simultaneous measurement of muscle
conduction velocity and EMG power spectrum changes during
fatigue. Muscle Nerve 1985;8:768–73.
[18] Arendt-Nielsen L, Mills KR. The relationship between mean
power frequency of the EMG spectrum and muscle fiber conduc-
tion velocity. Electroencephalogr Clin Neurophysiol 1985;
60:130–6.
[19] Broman H, Bilotto G, De Luca CJ. Myoelectric signal conduction
velocity and spectral parameters: influence of force and time. J
Appl Physiol 1985;8:1428–37.
[20] Radicheva N, Gerilovsky L, Gydikov A. Changes in the muscle
fiber extracellular action potentials in long-lasting (fatiguing)
activity. Eur J Appl Physiol 1986;55:545–52.
[21] Jennische I. Relation between membrane potential and lactate in
gastrocnemius and soleus muscle of the cat during tourniquet
ischemia and postischemic reflow. Pflügers Arch 1982;
394:329–32.
[22] Hultman E, Sahlin K. Acid–base balance during exercise. In: Wil-
more JH, Keogh JF, editors. Exercise and sport sciences reviews,
vol. 8. New York: Academic Press, 1980:41-128.
[23] Sjøgaard G. Muscle fatigue. In: Hebbelinck M, Shephard RJ, edi-
tors. Medicine and sport science, vol. 23. Basel: Karger,
1987:99–109.
[24] Kiens B, Saltin B. Endurance training of man decreases muscle
potassium loss during exercise. Acta Physiol Scand 1986;
20A:126.
[25] Madsen K, Pedersen PK, Djurhuus MS, Klitgaard NA. Effects of
detraining on endurance capacity and metabolic changes during
prolonged exhaustive exercise. J Appl Physiol 1993;75:1444–51.
[26] Milner-Brown HS, Miller RG. Muscle membrane excitation and
impulse propagation velocity are reduced during muscle fatigue.
Muscle Nerve 1986;9:367–74.
[27] Morimoto S, Umazume Y, Masuda M. Properties of spike poten-
tials detected by a surface electrode in intact human muscle. Jpn
J Physiol 1980;30:71–80.
[28] Stålberg E. Propagation velocity in human muscle fibers in situ.
Acta Physiol Scand 1966;70:3–112.
[29] Troni W, Contegiacomo V. The effect of temperature on conduc-
tion velocity in human muscle fibers. J Electromyogr Kinesiol
1991;1:281–7.
Kazumi Masuda received the master’s degree
in health and sport sciences from the University
of Tsukuba, Japan, in 1996. He is currently in
the final year of the doctoral program at the Uni-
versity of Tsukuba. His major research interests
focus on exercise-induced skeletal muscle plas-
ticity, especially in mechanisms of oxygen
transport into muscle tissue and energy pro-
duction.
Tadashi Masuda received the ME degree in
information engineering in 1978, and the PhD
degree in mathematical engineering in 1987
from the University of Tokyo, Japan. In 1978
he joined Industrial Products Research Institute,
Tsukuba Science City, Japan, which is reor-
ganized as the National Institute of Bioscience
and Human Technology in 1993. He is currently
a chief of the Physiological Informatics
Division. His research interests are in the instru-
mentation and analysis of myoelectric and
myomagnetic signals for the diagnosis of human
motor functions.
Tsugutake Sadoyama was born in Nagoya,
Japan in 1941. He received the BPhysEd degree
in 1966 from Tokyo University of Education
and the PhD degree in 1993 of from Tsukuba
University, Japan. He is currently an associate
professor of applied physiology in the Depart-
ment of Kansei Engineering in the Faculty of
Textile Science and Technology at Shinshu Uni-
versity. His current research and teaching inter-
ests concentrate on the area of psychophysiol-
ogy in ergonomics.
Mitsuharu Inaki received the PhD degree in
health and sport sciences from the University of
Tsukuba, Japan, in 1994. After working as an
assistant at the University of Tsukuba from 1994
to 1997, he has been affiliated with Seinan Jo
Gakuin University in Kyushu, Japan. His major
interests include the energy metabolism using
31
P-NMR during exercise.
Shigeru Katsuta received the PhD degree in
medicine from Kyushu University, Japan, in
1974. He worked as a guest professor in Wash-
ington State University from 1974 to 1975.
After working as an associate professor at Kyu-
shu University, he has been affiliated with the
Exercise Physiology Laboratory at the Institute
of Health and Sport Sciences, the University of
Tsukuba, since 1979. His major research inter-
ests include the plasticity of skeletal muscle with
exercise. He is currently the chairperson of the
board of trustees in the Japanese Society of
Physical Education.