SHORT REPORT ABSTRACT: When using repetitive nerve stimulation to examine neuro-

muscular transmission, the change in compound muscle action potential

(CMAP) size is usually assessed by measurement of negative-peak or
peak-to-peak amplitude. Technological developments now allow automatic
measurement of CMAP area, but some patients show increment of CMAP
amplitude and decrement of CMAP area. This study systematically analyzed
the changes in these CMAP parameters in 23 neurologically healthy sub-
jects. CMAPs were recorded when the ulnar nerve was stimulated at fre-
quencies of 1, 2, 5, 10, 20, 30, and 50 HZ (ve pulses per train). CMAP
amplitude showed signicant increment within a train when stimulus fre-
quency was above 5 HZ (probably due to increased muscle-ber conduction
velocity), whereas CMAP area hardly changed at any frequencies. Measure-
ment of CMAP area produces less ambiguous results than amplitude mea-
surement in repetitive nerve stimulation studies.
Muscle Nerve 29: 724 728, 2004

COMPOUND MUSCLE ACTION POTENTIALS

DURING REPETITIVE NERVE STIMULATION
TERUKO ASAWA,1 MASAOMI SHINDO, MD,2 and HIROKI MOMOI, MD2

1
Department of Laboratory Medicine, Shinshu University Hospital, Matsumoto, Japan
2
Department of the Third Internal Medicine (Neurology), Shinshu University School of Medicine,
Asahi 3-1-1, Matsumoto 390-8621, Japan

Accepted 28 January 2004

The repetitive nerve stimulation test is commonly
used in clinical laboratories to assess neuromuscular
transmission in patients with myasthenia gravis or
EatonLambert syndrome.1 Change in amplitude
(baseline to negative peak and, sometimes, peak-to-
peak) of compound muscle action potentials
(CMAPs) is generally used as the parameter of inter-
est during repetitive stimulation (520 pulses), usu-
ally with frequencies between 1 and 50 Hz. Although
the criteria for abnormality vary among laboratories,
a change (increment or decrement) in the fourth or
fth CMAP amplitude by more than 10% of that of
the rst is generally taken as abnormal. The effect of
repetitive stimulation is known to depend on the
muscle under test, cutaneous temperature, and stim-
ulation frequency.6,8
The area of the negative component of a CMAP
is known to reect the number of excited muscle
bers more accurately than the amplitude.1,7,10 De-
Abbreviations: ANOVA, analysis of variance; CMAP, compound muscle
action potential
Key words: compound muscle action potential (CMAP) amplitude; CMAP
area; human; neuromuscular transmission; pseudofacilitation; repetitive nerve
stimulation
Correspondence to: T. Asawa; e-mail: teasawa@hsp.md.shinshu-u.ac.jp
2004 Wiley Periodicals, Inc.
Published online 12 April 2004 in Wiley InterScience (www.interscience.wiley.
com). DOI 10.1002/mus.20037
velopments in computer technology have enabled
area-measuring software to be incorporated in the
clinical electromyograph and, consequently, area is
now commonly used along with amplitude in mea-
surements of CMAP size. In some patients, however,
peak-to-peak amplitude shows an apparent incre-
ment during a stimulus train, whereas the area de-
creases. Other patients show a signicant decrement
in area with no change in amplitude. To examine
this problem and to clarify the underlying mecha-
nisms, the changes in CMAP size (amplitude and
area) during repetitive nerve stimulation were sys-
tematically studied in neurologically healthy sub-
jects.
SUBJECTS AND METHODS
Twenty-three neurologically healthy subjects partici-
pated in the study: 11 were young (20 34 years old;
mean 26) and 12 were elderly (61 80 years old;
mean 69). All gave informed consent to participate
in the experiment, which was approved by our insti-
tutional ethics committee. Each experiment was per-
formed with the subject in the supine position.
Nerve stimulation and CMAP recording were
performed using a clinical electromyograph (Nicolet
Viking, Madison, WI). The right ulnar nerve was
electrically stimulated at the elbow with a concentric

724 Short Reports MUSCLE & NERVE May 2004

needle electrode. Routinely, each stimulus train con-
sisted of ve rectangular pulses with a pulse duration
of 0.5 ms. The stimulus intensity was supramaximal
for the CMAP, and the frequencies used were 1, 2, 5,
10, 20, 30, and 50 Hz. The stimulus trains were
delivered at intervals of more than 1 min.
The CMAP was recorded from the abductor dig-
iti minimi muscle by surface electrodes. To minimize
movement of stimulating and recording electrodes
and its effect on the CMAP, joint angles were kept
constant throughout the experiment: the elbow joint
was fully extended, the forearm pronated, and the
wrist joint maintained in the neutral position, with
metacarpophalangeal joints supported at 45 by
holding the subjects hand and the pillow. The active
electrode was xed with adhesive tape over the
mid-region of the muscle belly of the abductor digiti
minimi muscle (but placed slightly dorsally to mini-
mize movements due to contraction), and the refer-
ence electrode was xed over the proximal phalanx
of the fth digit. The amplier lters were set at 20
Hz and 2 kHz. Both peak-to-peak amplitude and
total area (negative area plus positive area) of the
CMAPs were measured by the Viking electromyo-
graph. The electromyogram was also stored in a data
recorder (XC-9000, Teac, Tokyo, Japan) for further
analysis of the potentials. The room temperature was
kept above 25C. The measured room temperature
was actually 25.030.5C (mean 26.8C), and the nerve. The stimulation consisted of double pulses
skin temperature was 30.633.8C (mean 32.8C) at with intervals of 1000, 500, 200, 100, 50, 33, 20, 10, or
the forearm, and 28.133.4C (mean 31.8C) at the 5 ms. We averaged 20 30 pairs of potentials at each
hand.8
CMAPs were further analyzed using a signal pro-
cessor (DP-1100, NEC Medical Systems, Tokyo, Ja-
pan), and several parameters were computed. These
were negative and positive amplitudes (amplitude
from baseline to respective negative and positive
peaks), peak-to-peak amplitude (amplitude between
negative and positive peaks), negative and positive
areas (areas of respective negative and positive
parts), total area (sum of negative and positive ar-
eas), and negative and positive duration (duration of
respective negative and positive parts). These param-
eters were calculated for all ve CMAPs evoked at a
given stimulus frequency, and those of the second to
fth CMAPs were expressed as a percentage of those
of the rst.
To determine the mechanism for amplitude in-
crease with high-frequency stimulation (see Results
section), we performed two control experiments.
Control Experiment 1. The angle of the metacarpo-
phalangeal joints was maintained at 45. Repetitive
nerve stimulation, however, induces nger exion,
Short Reports
so that some change in the joint angle during stim-
ulation was inevitable. To examine the effect of
changes in the relative positions of the recording
electrodes and muscle bers on the CMAP ampli-
tude, the maximal CMAP size was measured while
stimulating the nerve (with ten pulses at a frequency
of 0.5 Hz) at three different metacarpophalangeal
joint angles (45, 60, and 90) in ve neurologically
normal subjects. The CMAP amplitudes recorded at
angles of 60 and 90 were expressed as a percentage
of that recorded at 45. Other experimental settings
were the same as for the main studies.
Control Experiment 2. To analyze any contribution
of change in nerve conduction velocity (see Discus-
sion), velocity was measured in ve subjects who
showed an apparent increase in CMAP amplitude
during repetitive stimulation. The cutaneous nerve
at the fth digit was stimulated by surface ring elec-
trodes, and sensory nerve action potentials in the
right ulnar nerve were recorded orthodromically via
needle electrodes at the elbow joint. The reason for
recording the sensory rather than motor conduction
was to exclude the contribution of neuromuscular
transmission or muscle bers. The stimulus intensity
was selected to produce slight pain in order to excite
most of the large bers. The recording needle elec-
trodes were inserted in the immediate vicinity of the
of these intervals.
A paired t-test and analysis of variance (ANOVA)
were used for statistical analysis of the data.
RESULTS
Change in CMAP Amplitude and Total Area at Each
Stimulus Frequency. Changes in amplitude and to-
tal area at each stimulus frequency (150 Hz) were
compared for both young (20 34 years, mean 26
years, n 11) and elderly (61 80 years, mean 69
years, n 12) subjects. Increased amplitudes of the
fth potential compared to the rst potential were
signicant at almost all stimulus frequencies (at all
frequencies for the elderly, P 0.05 at 5 Hz and P
0.01 at the others; at all frequencies but 2 Hz and 5
Hz for the young, P 0.01; paired t-test). The extent
of increase depended signicantly on the stimulus
frequency (P 0.001; ANOVA), being maximal at
30 Hz. The increase was signicantly larger at 10, 20,
30, and 50 Hz when compared to that at 1 Hz (P
0.001; paired t-test). In one subject, increment of the
amplitude of the fth potential was as much as 50%
MUSCLE & NERVE May 2004 725
 Table 1. Increment/decrement of fourth and fth compound muscle action potentials (CMAPs).
Stimulus frequency
CMAP measurements 1 HZ 2 HZ 5 HZ 10 HZ 20 HZ 30 HZ 50 HZ
Fourth CMAP
Peak-to-peak amplitude 2.0 2.4 2.1 2.3 3.2 4.5 10.8 9.8 17.8 10.9 18.7 10.8 15.5 11.3
Negative peak amplitude 2.7 2.9 1.3 2.3 1.9 4.7 7.0 11.2 16.8 10.3 16.7 10.3 11.9 9.3
Total area 0.9 2.9 0.2 2.6 0.4 4.3 0.1 6.6 2.5 6.8 2.8 6.6 3.0 9.7
Negative area 0.4 5.0 0.6 3.7 2.4 3.8 1.6 6.9 3.9 6.6 4.9 6.0 11.7 9.6
Fifth CMAP
Peak-to-peak amplitude 2.4 2.7 2.9 2.7 4.3 5.6 12.8 10.8 22.2 12.9 24.8 13.2 21.4 12.6
Negative peak amplitude 2.9 3.6 1.3 3.0 2.0 5.4 7.8 12.1 20.3 11.0 21.1 10.7 15.7 13.3
Total area 1.1 3.2 0.8 3.1 0.1 4.8 1.2 7.1 1.0 8.2 1.3 7.7 1.7 7.5
Negative area 0.8 3.6 0.9 4.5 2.2 4.5 0.8 8.0 4.7 7.6 5.8 6.2 11.6 6.4

All the data are expressed as percent of the size of the rst potential. Values expressed as mean standard deviation. The data for young and elderly subjects
have been combined.

of amplitude of the rst. There was no signicant

difference between young and elderly subjects, nor
was there signicant interaction of age and stimulus
frequency (ANOVA). Contrary to the amplitude, the
total area remained within 5% of control at each
stimulus frequency. There was no signicant differ-
ence either with stimulus frequency or age
(ANOVA). The increment/decrement of peak-to-
peak amplitude, negative-peak amplitude, total area,
and negative area at each stimulus frequency during
repetitive stimulation are summarized in Table 1.

Changes in Total Area and Negative Area of the CMAP.

Because the number of muscle bers excited is best
reected by the negative area,1,8,11 the change in
total area at each stimulus frequency was compared
with that of the negative area. The area of the neg-
ative component decreased during the train at high
frequencies, whereas the total area was hardly al-
tered in either young or elderly subjects. When the
data for both young and elderly subjects were com-
bined, the negative area tended to decrease at fre-
quencies of 20 Hz (P 0.01), and 30 Hz and 50 Hz
(P 0.001). In elderly subjects, the decrement
tended to be greater than in young subjects. At
frequencies below 30 Hz, however, the change re-
mained within 5%. The total area hardly changed
at any frequency from 1 Hz through 50 Hz.
FIGURE 1. Changes in CMAP parameters during repetitive stim-
ulation. (A) The negative amplitude (lled circles) showed a
Changes in Negative Amplitude and Duration of the signicant increment during a train at stimulus frequencies above
CMAP. To help clarify the mechanism underlying 5 HZ. The negative area (open circles) showed decrement, but by
the change in peak-to-peak amplitude, changes in comparison, the latter remained more stable. (B) Concomitant
negative amplitude and negative area were com- changes in duration and amplitude of the negative component.
As the negative amplitude (lled circles) increased during a train
pared (Fig. 1A). Since changes for young and elderly
at frequencies above 5 HZ, the negative duration (open circles)
subjects were similar, their data were combined. Al- decreased. Data were derived from 23 subjects for both (A) and
though the change in the negative area during a (B). Circles and bars represent means and standard deviations.
train remained within 5% at frequencies below 30 *P 0.05; **P 0.01; ***P 0.001.

726 Short Reports MUSCLE & NERVE May 2004

Hz, as did the change in total area, the negative
amplitude began to increase at 510 Hz, and the
increment peaked at more than 20% of amplitude of
the control. Increase in amplitude was signicant at
10 Hz (P 0.05), 20 Hz (P 0.01), 30 Hz (P number of muscle bers activated than other param-
0.001), and 50 Hz (P 0.01; paired t-test).
Since area is a function of amplitude and dura-
tion, the relationship between negative amplitude
and negative duration for each of the ve CMAPs in
a train (stimulus frequency, 30 Hz) was analyzed for
9 young subjects. In all subjects, amplitude increased
and duration decreased with successive stimuli. The
data of this type obtained at all stimulus frequencies
for all 23 subjects are summarized in Figure 1B.
Since there was no difference between young and
elderly subjects, the data for both groups were com-
bined. Increase in amplitude and decrease in dura-
tion were signicant at 20 Hz, 30 Hz, and 50 Hz
(P 0.01 or P 0.001; paired t-test). At frequencies
above 5 Hz, the increase in negative amplitude was
mirrored by a decrease in the negative duration. At
50 Hz, the amplitude tended to decline and, at the
same time, the duration tended to increase.
Effect of Metacarpophangeal Joint Angle on CMAP Size.
With the joint angle of the metacarpophalangeal
joints at 60 and 90, changes in CMAP amplitude
were within 5% of those at 45 and, furthermore,
the negative component of the CMAP hardly
changed, the only change being in the positive com-
ponent.
Change in Nerve Conduction Velocity on Repetitive
Stimulus. The mean latencies of the nerve poten-
tials provoked by the second stimuli were found to
be the same as, or slightly longer (by 0.06 0.09 ms)
than, those provoked by the rst at every interstimu-
lus interval. A shortening of the latency (i.e., in-
creased conduction velocity) was never observed.
DISCUSSION
The present study on the effects of repetitive nerve
stimulation demonstrated that, in the course of a
train of ve pulses: (1) the negative area of the
CMAP decreased slightly but only during high-
frequency stimulation; (2) the peak-to-peak ampli-
tude increased signicantly (at high stimulus fre-
quencies); (3) the negative duration decreased sig-
nicantly (at high stimulus frequencies); (4) the
total area remained unchanged throughout the stim-
ulus frequency; and (5) no signicant differences
regarding these changes were detected between
young and elderly subjects.
Short Reports
Decrease in negative area during high-frequency
stimulation could be produced by several mecha-
nisms. Whatever the mechanism, however, the neg-
ative area is known to be a better reection of the
eters, and changes in amplitude and total area are
discussed below in relation to change in negative
area.
Among the changes we examined, increase in
CMAP amplitude was the most marked, especially
with stimulus trains above 10 Hz. The amount of
increment of amplitude (negative-peak amplitude,
peak-to-peak amplitude) was similar to that in previ-
ous work6 (Table 1). Because change in CMAP area
was minimal and CMAP duration shortened as am-
plitude increased (Fig. 1B), the mechanism under-
lying the amplitude increase is probably the same as
that responsible for the shortening of CMAP dura-
tion (i.e., synchronization of muscle action poten-
tials). The possible factors responsible for this are
changes in: (1) relative positions of recording elec-
trodes and muscle bers due to preceding contrac-
tion; (2) neuromuscular transmission; and (3) con-
duction velocities of nerve and muscle bers. The
rst possibility was excluded by control experiment
1. Regarding the second possibility, efcacy of neu-
romuscular transmission depends on the amount of
acetylcholine released from presynaptic terminals,
which in turn depends on both the number of syn-
aptic vesicles and concentration of Ca2 within the
nerve terminals. Any contribution of these factors is
negligible.1,2,6 Increase in nerve conduction velocity
leads to nerve action potentials becoming more syn-
chronized, thus decreasing temporal dispersion and
causing amplitude increase and shortening of dura-
tion of the CMAP. This possibility was, however,
excluded by control experiment 2.
The last and most probable candidate to explain
the amplitude increase and duration shortening is
increased conduction velocity of muscle bers. The
muscle conduction velocity varies among muscles.11
According to Stlberg et al., on double-pulse stimu-
lation the second muscle action potential traveled
faster by 12.524.0% than the rst.9 They also dem-
onstrated that muscle conduction velocity increased
by 0.03 0.60 m/s when stimulus frequency was in-
creased from 5 to 20 Hz. Similarly, Farmer et al.
reported that the second action potential traveled
faster (by 10 20 %) when it occurred 6 10 ms after
the rst.3 These observations are supported by the
nding of Nishizono et al., who measured the con-
duction velocity of a single motor unit using a mi-
crostimulation technique: conduction velocity in-
creased by 3.7 4.8 m/s when stimulus frequency was
MUSCLE & NERVE May 2004 727
increased from 5 to 40 Hz.5 On the basis of these
ndings, the amplitude increase and concomitant
duration shortening that we observed in CMAPs dur-
ing repetitive stimulation is most likely to be due to
increased conduction velocity of the muscle bers,
leading to a greater synchronization of muscle elec-
trical events.12 It is worthwhile to note that increase
in CMAP amplitude was not accompanied by in-
crease of negative or total area in the present
study, which is compatible with the synchronization
theory.4
Increment of CMAP amplitude does not by itself
indicate abnormal neuromuscular transmission. Al-
though negative area appears superior to the other
CMAP parameters as an index of muscle activity, it is
not always easy to measure in daily clinical studies.
Total area of the CMAP, however, is rather com-
monly obtained with a modern electromyograph.
Change in total area was much the same (within
5%) between young and elderly subjects, and did
not differ signicantly from the change in negative
area (when stimulus frequency was kept below 20 Hz
in elderly subjects and below 50 Hz in young sub-
jects). Furthermore, change in total area associated
with a change in joint angle, if any, was also within
5%.
When negative area cannot be measured, it
seems appropriate clinically to employ total area of
the CMAP as a surrogate, at least for the ulnar nerve,
to assess change in CMAP size during repetitive stim-
ulation. Diagnostic sensitivity of total area and anal-
ysis in other nerves than the ulnar nerve needs fur-
ther investigation.
728 Short Reports
Presented in part at the 43rd meeting of the Japanese Society of
Clinical Pathology, Hamamatsu, Japan, November 1996. The au-
thors are grateful to Dr. T. Hashimoto and Dr. H. Morita of the
Third Department of Internal Medicine (Neurology) for valuable
comments on this manuscript.
REFERENCES
1. Aminoff MJ. Electrodiagnosis in clinical neurology. New York:
Churchill Livingstone; 1986.
2. Desmedt JE. The neuromuscular disorder in myasthenia gra-
vis. 1. Electrical and mechanical response to nerve stimulation
in hand muscles. In: Desmedt JE, editor. New developments
in electromyography and clinical neurophysiology, vol 1.
Basel: Karger; 1973. p 241304.
3. Farmer TW, Buchthal F, Rosenfalck P. Refractory period of
human muscle after passage of a propagated action potential.
Electroencephalogr Clin Neurophysiol 1960;12:455 466.
4. McComas AJ, Galea V, Einhorn RW. Pseudofacilitation: a
misleading term. Muscle Nerve 1994;17:599 607.
5. Nishizono H, Fujimoto T, Ohtake H, Miyashita M. Muscle
ber conduction velocity and contractile properties estimated
from surface electrode arrays. Electroencephalogr Clin Neu-
rophysiol 1990; 75:75 81.
6. Oh SJ. General concepts of the repetitive nerve stimulation
test. In: Oh SJ, editor. Electromyography: neuromuscular
transmission studies. Baltimore: Williams & Wilkins; 1998. p
130.
7. Olney RK, Miller RG. Conduction block in compression neu-
ropathy: recognition and quantication. Muscle Nerve 1984;
7:662 667.
8. Rutkove SB. Pseudofacilitation: a temperature-sensitive phe-
nomenon. Muscle Nerve 2000;23:115118.
9. Stlberg E. Propagation velocity in human muscle bers in
situ. Acta Physiol Scand 1996;287:3112.
10. Stlberg E, Antoni L. Electrophysiological cross section of the
motor unit. J Neurol Neurosurg Psychiatry 1980;43:469 474.
11. Stlberg E, Ekstedt J. Single bre EMG and microphysiology
of the motor unit in normal and diseased human muscle. In:
Desmedt JE, editor. New developments in electromyography
and clinical neurophysiology. Basel: Karger; 1973. p 113129.
12. van Dijk JG, van der Hoeven BJ, van der Hoeven H. Repetitive
nerve stimulation: effects of recording site and the nature of
pseudofacilitation. Clin Neurophysiol 2000;111:14111419.
MUSCLE & NERVE May 2004