J Neurophysiol 95: 3328 –3335, 2006.

First published February 15, 2006; doi:10.1152/jn.01002.2005.

Soleus- and Gastrocnemii-Evoked V-Wave Responses Increase After

Neuromuscular Electrical Stimulation Training

Julien Gondin, Julien Duclay, and Alain Martin

Institut National de la Santé et de la Recherche Médicale, Equipe de Recherche Mixte 207, Motricité Plasticité Laboratory, Faculty of
Sport Sciences, University of Burgundy, Dijon Cedex, France
Submitted 23 September 2005; accepted in final form 11 February 2006

Gondin, Julien, Julien Duclay, and Alain Martin. Soleus- and and muscle activation (Colson et al. 2000; Gondin et al. 2005;
gastrocnemii-evoked V-wave responses increase after neuromuscular Maffiuletti et al. 2002) as well as significant cross-education
electrical stimulation training. J Neurophysiol 95: 3328 –3335, 2006. effect (Hortobagyi et al. 1999; Zhou 2000) have been reported
First published February 15, 2006; doi:10.1152/jn.01002.2005. The
aim of the study was to use combined longitudinal measurements of
after multiple sessions of NMES. Several arguments can be
soleus (SOL) and gastrocnemii evoked V-wave and H-reflex re- advanced to explain the occurrence of neural adaptations.
sponses to determine the site of adaptations within the central nervous Indeed, NMES evokes action potential in both intramuscular
system induced by 5 wk of neuromuscular electrical stimulation nerve branches and cutaneous receptors, thus generating force
(NMES) training of the plantar flexor muscles. Nineteen healthy directly by activation of motor axons (Hultman et al. 1983)
males subjects were divided into a neuromuscular electrostimulated and, indirectly, by reflex recruitment of spinal motoneurons
group (n ⫽ 12) and a control group (n ⫽ 7). The training program (Collins et al. 2001). Moreover, Smith et al. (2003) recently
consisted of 15 sessions of isometric NMES over a 5-wk period. All reported that NMES applied over the quadriceps femoris mus-
subjects were tested before and after the 5-wk period. SOL, lateral cle activated specific supraspinal regions (e.g., primary motor
gastrocnemius (LG), and medial gastrocnemius (MG) maximal H- cortex, primary sensory cortex, cingulate gyrus) in a dose–
reflex and M-wave potentials were evoked at rest (i.e., Hmax and
Mmax, respectively) and during maximal voluntary contraction
response manner. Therefore NMES does not actually bypass
(MVC) (i.e., Hsup and Msup, respectively). During MVC, a supramaxi- the nervous system, at least in an intact system. Nevertheless,
mal stimulus was delivered that allowed us to record the V-wave the sites of adaptations within the central nervous system (i.e.,
peak-to-peak amplitudes from all three muscles. The SOL, LG, and spinal and/or supraspinal) after multiple sessions of NMES
MG electromyographic (EMG) activity as well as muscle activation remain unclear.
(twitch interpolation technique) were also quantified during MVC. In an attempt to investigate neural adaptations to NMES
After training, plantar flexor MVC increased significantly by 22% training, only one longitudinal study has examined spinal
(P ⬍ 0.001). Torque gains were accompanied by an increase in reflex pathways on the basis of the recording of evoked
muscle activation (⫹11%, P ⬍ 0.05), SOL, LG, and MG normalized H-reflex responses in healthy humans (Maffiuletti et al. 2003).
EMG activity (⫹51, ⫹54, and ⫹60%, respectively, P ⬍ 0.05) and Indeed, training-induced modifications in H-reflex amplitude
V/Msup ratios (⫹81, ⫹76, and ⫹97%, respectively, P ⬍ 0.05).
Hmax/Mmax and Hsup/Msup ratios for all three muscles were unchanged
are usually ascribed to changes in alpha motoneuron excitabil-
after training. In conclusion, the increase in voluntary torque after 5 ity and/or changes in presynaptic inhibition of Ia afferents
wk of NMES training could be ascribed to an increased volitional (Zehr 2002). Maffiuletti et al. (2003) reported no change in the
drive from the supraspinal centers and/or adaptations occurring at the resting soleus (SOL) and gastrocnemius maximal H-reflex
spinal level. (Hmax) to maximal M-wave (Mmax) ratios after 4 wk of NMES
training and concluded that the mechanism accounting for the
strength gains could be an increased volitional drive from the
INTRODUCTION supraspinal centers. However, several investigations have in-
dicated that H-reflexes evoked during actual voluntary contrac-
Neuromuscular electrical stimulation (NMES) is commonly tions (Aagaard et al. 2002) and/or during the performance of
used as a means of strength training in healthy humans (Colson the training exercise (Voigt et al. 1998; Zehr 2002) likely
et al. 2000; Gondin et al. 2005; Maffiuletti et al. 2002; Parker represent a more functional estimate of the training-induced
et al. 2003). Many studies have also reported that the applica- adaptations at the spinal cord level. By recording H-reflex
tion of electrical stimulation to a muscle is effective in improv- responses evoked both at rest and during maximal voluntary
ing maximal voluntary strength and functional capacity in contraction (MVC), it would therefore be possible to examine
hypoactive patients with spinal cord injury (Belanger et al. neural adaptations occurring within the spinal locus of the
2000; Kern et al. 2005) or stroke (Glanz et al. 1996; Kimberley motor system.
et al. 2004; Powell et al. 1999). Several authors have suggested During MVC, the study of the so-called V-wave, which is an
that neural factors largely account for NMES training-induced electrophysiological variant of the H-reflex (Hultborn and
strength gains, especially in the case of programs lasting ⬍4 –5 Pierrot-Deseilligny 1979; Upton et al. 1971), may be also used
wk (Enoka 1988; Gondin et al. 2005; Maffiuletti et al. 2002). to reflect the level of efferent neural drive from spinal ␣-mo-
Indeed, increases in surface electromyographic (EMG) activity toneurons (Aagaard et al. 2002; Duclay and Martin 2005;

Address for reprint requests and other correspondence: J. Gondin, INSERM The costs of publication of this article were defrayed in part by the payment
ERM 207, Faculté des Sciences du Sport, BP 27877, 21078 Dijon Cedex, of page charges. The article must therefore be hereby marked “advertisement”
France (E-mail: [email protected]). in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3328 0022-3077/06 $8.00 Copyright © 2006 The American Physiological Society www.jn.org
Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
 NEURAL ADAPTATIONS TO NMES TRAINING 3329

Pensini and Martin 2004; Sale et al. 1983). The V-wave wall. To stabilize and to secure the feet in the shoes, straps were firmly
consists of a volley of H-reflex impulses that are allowed to fastened. A strap also secured the feet to the footplate and another
reach the muscle as a result of the removal of antidromic action strap was fastened to avoid any movement of the ankle. Finally, a
potential impulses generated by ␣-motor axon stimulation by strap fixed between the knees and the footplate, surrounding the legs,
maintained the 90° knee joint angle during the evoked contractions.
collision with descending neural drive generated by the volun- Three 2-mm-thick, self-adhesive electrodes were placed over each
tary motor effort (Upton et al. 1971). Moreover, the supra- leg. The cathodes, measuring 25 cm2 (5 ⫻ 5 cm) and exhibiting
maximal level of nerve stimulation used during the recording membrane-depolarizing properties, were placed over the superficial
of V-wave also causes massive excitation of all afferent axons aspect of the SOL muscle, about 5 cm distal from where the two heads
in the peripheral nerve. As a result, the evoked V-wave of the gastrocnemius join the Achilles tendon. The anode measuring
response will recruit both large and small motoneurons, 50 cm2 (10 ⫻ 5 cm) was placed along the middorsal line of the leg,
whereas H-reflex primarily relies on the pool of smaller mo- over both medial and lateral gastrocnemii. A portable battery-powered
toneurons (Aagaard et al. 2002; Duclay and Martin 2005). An stimulator (Sport P, Compex Medical SA, Ecublens, Switzerland) was
increase in V-wave amplitude may result from an enhanced used. Rectangular-wave pulsed currents (75 Hz) lasting 400 ␮s were
neural drive in descending corticospinal pathways, elevated delivered with a rise time of 1.5 s, a steady tetanic stimulation time of
4 s, and a fall time of 0.75 s (total duration of the contraction: 6.25 s).
motoneuron excitability, and/or alterations in presynaptic inhi- Each stimulation was followed by a pause lasting 20 s. Stimulation
bition (Aagaard et al. 2002), i.e., from changes occurring at intensity was monitored on-line and was gradually increased through-
both spinal and supraspinal levels. Because modulations in out the training session to the level of maximally tolerated intensity,
H-reflex responses could be ascribed to spinal mechanisms, a which varied between 32 and 120 mA (mean 74 ⫾ 32 mA), according
concomitant analysis of these two types of evoked responses to individual pain threshold. No subject reported serious discomfort
would therefore allow us to determine the sites mediating from the current. Each session was preceded by a standardized
neural adaptations in response to NMES training. warm-up, consisting of 5 min of submaximal NMES at a freely chosen
The purpose of the present study was therefore to use intensity (5 Hz, pulses lasting 200 ␮s). The average torque produced
combined longitudinal measurements of SOL but also gastroc- during the NMES session was 61 ⫾ 11% of the MVC.
nemii-evoked V-wave and H-reflex responses to examine the
neural adaptations induced by a 5-wk NMES training program Measurements
on the plantar flexor muscles. Because NMES activates both
STUDY DESIGN The EG subjects were tested at baseline and again
motor and sensory fibers (Collins et al. 2001; Hultman et al. 3– 4 days after the last NMES training session. CG subjects received
1983) as well as supraspinal centers (Han et al. 2003; Smith et no exercise training and were tested before and after a 5-wk period.
al. 2003), it can be hypothesized that the strength gains ob- All subjects were asked not to perform any strenuous exercise for ⱖ48
served after short-term NMES training might be associated h before the two testing sessions.
with adaptations at both spinal and supraspinal levels.
TORQUE MEASUREMENTS All measurements were taken on the left
leg muscles. Subjects wore a shoe that was firmly secured to the
METHODS footplate of the isokinetic dynamometer (Biodex, Shirley, NY). Sub-
Subjects jects were examined in the supine position with the hip, knee, and
ankle joints at 90° flexion. To minimize hip and thigh motion during
Nineteen healthy male students gave written informed consent to the contractions, and therefore to avoid the contribution of muscles
participate in this study. They were randomly assigned to a neuro- other than the plantar flexors (e.g., knee extensors, hip flexors), straps
muscular electrostimulated group (EG), composed of 12 subjects (age were fastened across the chest and pelvis. The foot was also secured
21.7 ⫾ 3.4 yr, height 176.6 ⫾ 7.4 cm, weight 70.5 ⫾ 4.7 kg, means ⫾ to the footplate by two straps. One strap was placed around the ankle
SD) or to a control group (CG) (n ⫽ 7, age 26.6 ⫾ 5.5 yr, height and the second strap was placed around the foot, 1–2 cm proximal to
178.3 ⫾ 8.4 cm, weight 71.9 ⫾ 6.6 kg, means ⫾ SD). None of them the metatarsophalangeal joint of the toes. The alignment between the
had engaged in systematic strength training or NMES in the 12 mo center of rotation of the dynamometer shaft and the axis of the ankle
before the experiments began but some were active in recreational joint was checked at the beginning of each session. Particular care was
sports. CG subjects did not engage in any form of resistance training taken in monitoring the posture of the subjects and in avoiding head
exercise and were asked not to begin a new exercise program for the rotations during the test to maintain constant corticovestibular influ-
duration of the study. Approval for the project was obtained from the ences on the excitability of the motor pool and limit afferent feedback
University of Burgundy Committee on Human Research. All proce- from other peripheral receptors, i.e., Golgi tendon organs, cutaneous
dures used in this study were in conformity with the Declaration of and joint afferents (Schieppati 1987; Zehr 2002).
Helsinki.
STIMULATION The posterior tibial nerve was stimulated with a
single rectangular pulse (1 ms) delivered by a Digitimer stimulator
NMES training (model DS7, Hertfordshire, UK). The self-adhesive cathode (8-mm
diameter, Ag–AgCL) was located in the popliteal fossa and the anode,
TRAINING SESSION One week before the stimulation period began, which was a large rectangular electrode (5 ⫻ 10 cm, Compex Medical
the EG subjects participated in one practice session to familiarize SA), was placed on the anterior surface of the knee. Each subject was
themselves with stimulation parameters. The training program con- initially familiarized with several submaximal (range 1–20 mA)
sisted of fifteen 18-min sessions of isometric (bilateral) NMES over a electrical stimuli over a period of nearly 10 min. The stimulation site
5-wk period, with three sessions per week. Forty isometric contrac- resulting in the greatest M-wave amplitude was first located by a
tions were carried out during each training session. The subjects were handheld cathode ball electrode (0.5-cm diameter). Once the stimu-
trained on custom-made equipment especially created for this inves- lation site was determined, the stimulation electrode was firmly fixed
tigation. During the stimulation, they were placed in a supine position to this site with rigid straps and taping.
on a comfortable mattress, with the ankle, knee, and hip joints fixed
at a 90° angle and the head resting on a pillow. Each foot was placed ELECTROMYOGRAPHY Surface EMG activity of the SOL, lateral
in a shoe, which was firmly attached to a footplate, fixed parallel to the gastrocnemius (LG), and medial gastrocnemius (MG) muscles was

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
3330 J. GONDIN, J. DUCLAY, AND A. MARTIN

recorded bipolarly, during voluntary and electrically evoked contrac- of variation ⬎5%), three additional MVCs were performed. H and V
tions, by silver chloride circular electrodes with a diameter of 20 mm recordings during MVC were performed in a randomized order within
and a recording diameter of 10 mm. For the SOL, recording electrodes and across the testing sessions. Finally, two MVCs of the dorsiflexor
were placed along the middorsal line of the leg, about 5 cm below the muscles separated by ⱖ60 s were performed.
insertion of the gastrocnemii on the Achilles tendon. MG and LG
electrodes were fixed lengthwise over the middle of the muscle belly. Data analysis
These sites were determined in pilot testing by eliciting at a given
intensity the greatest M-wave amplitude for each muscle by tibial MVC, EMG ACTIVITY, AND MUSCLE ACTIVATION Only the highest
nerve stimulation. This procedure was performed to avoid the inner- plantar flexor MVC was considered for analysis. EMG was analyzed
vation zone and therefore to obtain the optimal amplitude of EMG over a 500-ms period before stimulation. The root mean square (RMS)
response (Merletti et al. 2001). Because the electrophysiological EMG values of SOL, LG, and MG muscles were calculated and then
responses induced by tibial nerve stimulation are generated by the normalized: 1) to the corresponding amplitude of the Mmax during
plantar flexors and possibly contaminated by concomitant activation resting conditions to obtain the background level of EMG and 2) to
of the tibialis anterior, the EMG activity of the tibialis anterior was the corresponding amplitude of the Msup during MVC to obtain the
also recorded. For this muscle, the electrodes were positioned on the normalized EMG. Tibialis anterior background EMG values were
line of the fibula at 1/3 of the distance from the tip of the medial calculated by normalizing the RMS values obtained during resting
malleolus (Hermens et al. 2000). The reference electrode was placed conditions to the RMS values obtained during dorsiflexor MVC, and
in a central position on the same leg. The placement of the electrodes were expressed as a percentage. The level of coactivation was calcu-
was marked on the skin with indelible ink, so that they could be lated by normalizing the RMS values of the tibialis anterior when this
exactly repositioned for the posttests. Low resistance (⬍5 k⍀) be- muscle was acting as an antagonist to the RMS obtained when this
tween the two electrodes was obtained by abrading the skin with muscle was acting as an agonist (i.e., during dorsiflexion) and was
emery paper and cleaning with alcohol. EMG signals were amplified expressed as a percentage. Muscle activation was estimated according
with a bandwidth frequency ranging from 15 Hz to 5.0 kHz (common to the following formula: % activation ⫽ (1 ⫺ superimposed doublet/
mode rejection ratio ⫽ 90 dB; impedance ⫽ 100 M⍀; gain ⫽ 1,000). potentiated doublet) ⫻ 100 (Allen et al. 1995).
The EMG and mechanical signals were digitized on-line (sampling
frequency 5 kHz) and stored for analysis with commercially available EVOKED POTENTIALS The average EMG signal for all trials in
software (Tida, Heka Elektronik, Lambrecht/Pfalz, Germany). respective muscles (i.e., SOL, LG, and MG) was used to calculate
peak-to-peak amplitude of 1) H-reflexes (Hmax and Hsup), 2) small
M-waves preceding Hmax and Hsup (i.e., MHmax and MHsup, respec-
Experimental procedure tively), maximal M-waves (Mmax and Msup), and V-wave. Because
The duration of each testing session was between 90 and 120 min the electrical stimulation was optimized for the SOL muscle, LG and
with each subject performing between 12 and 18 maximal efforts. MG H-reflexes were not maximal for all subjects at SOL Hmax
Testing procedure and recordings started by progressively increas- intensity. At this intensity and for the EG subjects, LG and MG
ing the current intensity by 2-mA increments from 0 until there was no H-reflexes were found to be maximal in 22 (about 92%) and 20 (about
further increase in peak twitch torque (i.e., the highest value of the 83%) of 24 sessions and were obtained in the ascending part of the
plantar flexor twitch torque was reached) nor in concomitant peak-to- recruitment curve in two (about 8%) of the 24 sessions, respectively.
peak M-wave amplitudes. Four stimuli were delivered at rest (i.e., The following ratios were then calculated for respective muscles:
without ongoing EMG activity) at each intensity level, at 10-s inter- MHmax/Mmax, MHsup/Msup, Hmax/Mmax, Hsup/Msup, and V/Msup (Du-
vals. The average EMG and mechanical signals obtained at each clay and Martin 2005).
intensity were stored to obtain later H–M recruitment curves for SOL, CONTRACTILE PROPERTIES The torque signals associated with the
LG, and MG muscles. The stimulus intensity needed to obtain the electrically evoked contractions under resting conditions were aver-
maximal SOL H reflex (Hmax) was carefully searched. SOL Hmax aged. The following contractile properties for both single and paired
intensity ranged from 11 to 52 mA. After the H–M calibration stimuli were then analyzed as previously described (Gondin et al.
procedure, four single pulses at supramaximal intensity (150% of the 2005): 1) peak twitch, 2) time-to-peak twitch, and 3) half-relaxation
stimulus intensity) each separated by 8 s were delivered and the time.
recorded M-wave was considered as Mmax for the respective muscles.
Three paired stimuli (10-ms interspike interval), each separated by 6 s STATISTICAL ANALYSIS Normality of the data was checked and
were also evoked at rest. Individual supramaximal intensities were subsequently confirmed using the Kolmogorov–Smirnov test. A sep-
constituted between 36 and 130 mA. Subjects were then instructed to arate two-factor [group (EG vs. CG) ⫻ test (muscle activation vs. H
perform two MVCs of the plantar flexor muscles with paired stimuli superimposed test vs. V superimposed test)] ANOVA with repeated
delivered over the isometric plateau (superimposed doublet) and 3 s measures was used to compare EMG activity obtained during MVC.
after the contraction (potentiated doublet) to assess muscle activation Separate two-factor [group (EG vs. CG) ⫻ session (PRE vs. POST)]
according to the twitch interpolation technique (Allen et al. 1995). To ANOVAs with repeated measures on session were used to compare
ensure reproducible trials, subjects were also provided with visual the MVC, background level of EMG recorded at SOL Hmax intensity,
feedback of the plantar flexor muscle torque on a monitor that was normalized EMG activity and muscle activation obtained during
placed 1 m in front of them. After a 5-min resting period, subjects MVC, twitch contractile properties, V-wave amplitudes, and V/Msup
were then asked to perform five MVCs of the plantar flexor muscles ratio. Separate three-factor [group ⫻ session ⫻ condition (rest vs.
for 3 s separated by 1-min rest periods. Single stimulation at SOL MVC)] ANOVAs with repeated measures on session and on condition
Hmax intensity was superimposed about 1.5 s after the beginning of were performed to compare both the H-reflex (Hmax and Hsup) and
each MVC to record superimposed H-reflex (Hsup) and M-wave at M-wave (Mmax and Msup) amplitudes, the MHmax/Mmax and MHsup/
Hsup (MHsup) (H superimposed tests). After a rest period of 5 min, Msup ratios, and the Hmax/Mmax and Hsup/Msup ratios before and after
subjects were instructed to perform five more MVCs with the same the 5-wk period. When a main effect or a significant interaction was
contraction protocol. During these MVCs, a supramaximal stimulus found, Tukey post hoc analysis was performed. Significance was
allowed us to record the superimposed M-wave (Msup) and the accepted when P ⬍ 0.05. The statistical analyses were performed
V-wave from the three plantar flexor muscles (V superimposed tests). using Statistica software for Microsoft Windows (StatSoft, version
For each subject, the coefficient of variation across trials was calcu- 6.1, Tulsa, OK). All data are expressed as means ⫾ SD within the text
lated. If the V-wave or Hsup amplitudes varied excessively (coefficient and the table and means ⫾ SE in the figures.

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
 NEURAL ADAPTATIONS TO NMES TRAINING 3331

FIG. 1. Normalized EMG activity in soleus (SOL), lateral gastrocnemius (LG), and medial gastrocnemius (MG) muscles obtained during maximal voluntary
plantar flexion before (black bars) and after (white bars) the 5-wk period: A: neuromuscular electrostimulated group (n ⫽ 12). B: control group (n ⫽ 7). All values
are means ⫾ SE. Significantly higher than values before the 5-wk period: **P ⬍ 0.01, ***P ⬍ 0.001.

RESULTS Examples of SOL, LG, and MG V-waves and M-waves

No significant baseline differences were observed between evoked during MVC before and after the 5-wk training period
the two groups for the set of the dependent variables. are shown in Fig. 2. SOL, LG, and MG V-wave amplitude
increased significantly in the EG after training by 75 ⫾ 75%
(P ⬍ 0.05), 80 ⫾ 56% (P ⬍ 0.05), and 84 ⫾ 88% (P ⬍ 0.01),
MVC, EMG activity, and activation level
The plantar flexor MVC torque increased significantly by
22 ⫾ 19% after training in the EG (from 106 ⫾ 14 to 128 ⫾
15 Nm, P ⬍ 0.001), whereas no significant change occurred in
the CG (118 ⫾ 18 vs. 114 ⫾ 10 Nm).
In the EG, muscle activation increased significantly by 11 ⫾
15% after training (from 90 ⫾ 11 to 99 ⫾ 4%, P ⬍ 0.05),
whereas no significant change was found in the CG between
the two testing sessions (96 ⫾ 4 vs. 94 ⫾ 11%).
No main effect or significant interaction was found for SOL,
LG, MG, or tibialis anterior background EMG level obtained at
SOL Hmax intensity.
There was no main effect or significant interaction for SOL,
LG, or MG EMG values recorded during muscle activation
assessment, H and V superimposed tests.
SOL, LG, and MG normalized EMG increased significantly
after training in the EG by 51 ⫾ 44% (P ⬍ 0.001), 54 ⫾ 41%
(P ⬍ 0.01), and 60 ⫾ 59% (P ⬍ 0.01), respectively (Fig. 1A).
No significant changes were observed in the CG before and
after the 5-wk period (Fig. 1B).
There was no main effect or significant interaction for the
level of coactivation (0.29 ⬍ P ⬍ 0.49). The level of coacti-
vation before and after the 5-wk period was 8.9 ⫾ 4.9 and
8.5 ⫾ 2.4% in the EG and 7.0 ⫾ 1.6 and 8.3 ⫾ 3.6% in the CG,
respectively.

Evoked potentials
There was a significant condition effect (P ⬍ 0.05) for both
H-reflex and M-wave amplitudes in all three muscles (data not
shown). Hsup and Msup during MVC were significantly higher
than Hmax and Mmax at rest for the SOL (P ⬍ 0.05 and P ⬍ FIG. 2. Recordings from one representative subject of the evoked V-wave
0.001, respectively), LG (P ⬍ 0.001 and P ⬍ 0.05, respec- responses and the associated superimposed M-wave (Msup) before (solid line)
tively), and MG (P ⬍ 0.05 and P ⬍ 0.01, respectively). There and after the 5-wk neuromuscular electrical stimulation (NMES) training
program (dotted line) in the soleus (A), the gastrocnemius lateralis (B), and the
was no session or group effect and no significant interaction for gastrocnemius medialis (C) muscles. Initial spikes represent the stimulus
H-reflexes and M-waves obtained at rest and during MVC in artifact. Note that the superimposed M-wave did not change, whereas the V
any of the three muscles (data not shown). response was higher after training in all 3 muscles.

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
3332 J. GONDIN, J. DUCLAY, AND A. MARTIN

respectively (Fig. 2). V-wave amplitude showed no changes in electrical stimulation intensity used in the present investigation
the CG in any of the three muscles between the two testing was optimized to record the SOL maximal H-reflex. In the
sessions (data not shown). SOL muscle, we observed that this small M-wave represented
nearly 10 –15% of the maximal M-wave during both resting
Reflex ratios and active conditions, which is in agreement with the values
reported in the literature (Duclay and Martin 2005; Maffiuletti
No main effect or significant interaction was found for et al. 2001; Scaglioni et al. 2003). Despite the optimization of
MHmax/Mmax and MHsup/Msup ratios (Table 1), suggesting that stimulus intensity for the SOL muscle, gastrocnemii H-reflexes
stimulus conditions were stable. No main effect or significant were found to be maximal or obtained in the ascending part of
interaction was observed between Hmax/Mmax and Hsup/Msup the recruitment curve in about 90% of the trials. Our results
ratios (Fig. 3) in any of the three muscles. On the other hand, also indicated that the ratio between the small M-wave accom-
SOL, LG, and MG V/Msup ratios increased significantly in the panying H-reflex and the corresponding maximal M-wave
EG after training by 81 ⫾ 89% (P ⬍ 0.01), 76 ⫾ 52% (P ⬍ either at rest or during MVC was almost identical after the
0.05), and 97 ⫾ 67% (P ⬍ 0.01), respectively (Fig. 3, A–C), 5-wk period. We can thus suppose that the same proportion of
whereas no difference was noted in the CG before and after the ␣-motoneurons was activated by the stimulation during both
5-wk period (Fig. 3, D–F). testing sessions in all three muscles. Moreover, background
EMG activity in all three muscles, which could also induce
Contractile properties changes in H-reflex amplitude (Schiepatti 1987), was no dif-
No significant interaction (0.24 ⬍ P ⬍ 0.85) or main effect ferent before and after the 5-wk period. All these findings
(0.21 ⬍ P ⬍ 0.95) was noted for the contractile properties clearly indicated that H-reflex recording conditions were sim-
associated with single and paired stimuli (data not shown). ilar over the two testing sessions in all three muscles.

DISCUSSION Effects of NMES training on MVC, EMG activity,

and muscle activation
This study demonstrated that the significant increase in
plantar flexor muscles MVC observed after 5 wk of NMES An average increase of 22% in plantar flexion MVC was
training was accompanied by a significant increase in 1) observed after 5 wk of NMES training. This finding is in agree-
muscle activation and 2) normalized EMG activity and V-wave ment with several studies using similar training programs on the
response (both raw and normalized to the maximal M wave) in same muscular group (Maffiuletti et al. 2002; Martin et al. 1993).
all three muscles. The data also indicated that H-reflex re- In the present study, both muscle activation and normalized
sponses obtained both at rest and during MVC were not EMG of three plantar flexor muscles increased significantly
significantly affected by the NMES training program. after the treatment, therefore confirming the occurrence of
neural adaptations as a result of a short-term NMES training
Methodological considerations (Gondin et al. 2005; Maffiuletti et al. 2002). Similarly, Maf-
fiuletti et al. (2002) reported a higher EMG activity of the
According to Zehr (2002), certain methodological require- agonist muscles and an enhancement of the voluntary activa-
ments must be respected to ensure accurate interpretation of tion of the plantar flexor muscles after 4 wk of NMES. We
H-reflex in exercise study. Usually, H-reflexes should be recently demonstrated that neural adaptations accounted for the
evoked at a stimulation level sufficient to provide a corre- larger proportion of initial strength increment after multiple
sponding small M-wave, to help ensure stimulus constancy sessions of NMES (Gondin et al. 2005). Such adaptations
(Schiepatti 1987). Moreover, H-reflexes should be obtained in within the nervous system could thus be ascribed to changes
the ascending part of the recruitment curve to be sensitive to occurring at the spinal and/or supraspinal level. Electrically
excitation and/or inhibition mechanisms (Crone et al. 1990; evoked reflexes (H and V) represent a good tool to investigate
Pierrot-Deseilligny and Mazevet 2000). For these reasons, the whether adaptation occurs at the spinal level.
TABLE 1. Ratios between small M-wave accompanying H-reflex and corresponding maximal M-wave for soleus, lateralis gastrocnemius,
and medialis gastrocnemius muscles at rest (MHmax/Mmax) and during MVC (MHsup/Msup) before (PRE) and after 5-wk period (POST) for
neuromuscular electrostimulated group (EG) and control group (CG)

EG CG

PRE POST PRE POST

Soleus
MHmax/Mmax 0.12 ⫾ 0.11 0.10 ⫾ 0.05 0.12 ⫾ 0.06 0.11 ⫾ 0.04
MHsup/Msup 0.14 ⫾ 0.12 0.12 ⫾ 0.15 0.10 ⫾ 0.06 0.12 ⫾ 0.13
Lateralis gastrocnemius
MHmax/Mmax 0.40 ⫾ 0.34 0.40 ⫾ 0.32 0.37 ⫾ 0.33 0.36 ⫾ 0.26
MHsup/Msup 0.40 ⫾ 0.40 0.36 ⫾ 0.31 0.37 ⫾ 0.41 0.36 ⫾ 0.43
Medialis gastrocnemius
MHmax/Mmax 0.26 ⫾ 0.32 0.27 ⫾ 0.26 0.27 ⫾ 0.35 0.30 ⫾ 0.37
MHsup/Msup 0.30 ⫾ 0.23 0.28 ⫾ 0.32 0.28 ⫾ 0.26 0.25 ⫾ 0.33

Values are means ⫾ SD. EG, n ⫽ 12; CG, n ⫽ 7.

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
 NEURAL ADAPTATIONS TO NMES TRAINING 3333

FIG. 3. Soleus (SOL), lateral gastrocnemius (LG), and medial gastrocnemius (MG) Hmax/Mmax, Hsup/Msup, and V/Msup ratios obtained at rest (Hmax/Mmax)
and during maximal voluntary plantar flexion (Hsup/Msup and V/Msup) before (black bars) and after the 5-wk period (white bars). A–C: for the neuromuscular
electrostimulated group (n ⫽ 12). D–F: for the control group (n ⫽ 7). All values are means ⫾ SE. Significantly higher than values before the 5-wk period: *P ⬍
0.05, **P ⬍ 0.01.

Effects of NMES training on H-reflex responses al. 1998) likely represent a more functional estimate of the
In the present study, SOL and gastrocnemii resting Hmax/ training-induced adaptations at the spinal cord level. For ex-
Mmax ratios did not change significantly in response to multiple ample, Aagaard et al. (2002) showed no changes in resting
sessions of NMES. These results are in accordance with those SOL H-reflex amplitude after 14 wk of heavy weight-lifting
recently reported (Maffiuletti et al. 2003) after 4 wk of NMES exercises, whereas the H-reflex recorded during maximal iso-
training of the plantar flexor muscles. Nevertheless, several metric ramp contractions significantly increased with training.
authors have suggested that evoked reflex responses obtained However, these authors evoked SOL H-reflexes at rest and
during actual voluntary contraction (Aagaard et al. 2002) during MVC with stimulation intensity that produced an M-
and/or during the performance of the training exercise (Voigt et wave corresponding to 10 and 20% of the maximal M-wave,
J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org
Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
3334 J. GONDIN, J. DUCLAY, AND A. MARTIN

respectively. Thus the proportion of motor axons directly least partially, by neural mechanisms at the spinal level. To
recruited was probably different between active and resting compare H-reflex and V-wave modulations, we have to take
conditions, thus indicating that H-reflex modulations observed into account that these two types of evoked responses could
in this study may be partially attributed to recording method- recruit different portions of the spinal motoneuron pool be-
ologies. Moreover, Lagerquist et al. (2006) recently demon- cause of the large difference in stimulation intensities (i.e.,
strated no change in the soleus Hmax/Mmax ratio obtained at a submaximal vs. supramaximal, respectively). Indeed, V-wave
10% EMG background contraction after a 5-wk isometric response will recruit both large and small motoneurons,
plantar flexor voluntary strength training program (i.e., with a whereas H-reflex primarily relies on the pool of smaller mo-
training duration similar to our NMES program). toneurons (Aagaard et al. 2002; Duclay and Martin 2005). The
In the current study, not only SOL but also gastrocnemii training effect could therefore be attributed to spinal elements
H-reflexes recorded during actual MVC were used for the first that affect motoneuron firing without affecting the efficacy of
time to evaluate the modulation of the spinal loop after NMES the elements of the H-reflex pathways.
resistance training. Although NMES applied over the triceps
surae muscle evokes action potentials in both intramuscular Functional implications
nerve branches (Hultman et al. 1983) and cutaneous receptors,
thus resulting in reflex recruitment of spinal motoneurons In the current work, the voluntary strength gains induced by
(Collins et al. 2001), the Hsup/Msup ratio of all three muscles our NMES training were associated with neurological changes
was unaffected by the NMES training program. It can therefore occurring within the spinal and/or supraspinal locus of the
be suggested that the elements of the H-reflex pathway are not motor system. Thus the benefits of NMES might be useful in
directly involved in the chronic adjustments in response to the design of rehabilitation programs to improve recovery of
NMES training. upper-limb function (Wade 1989) and to increase cortical
activation in subjects with cerebral damage. However, the
Effects of NMES training on V-waves short-term NMES effects we observed in healthy subjects may
not be the same in a patient population and further studies are
This study is the first to demonstrate a significant increase in needed to examine the effects of multiple sessions of NMES in
SOL and gastrocnemii V/Msup ratios with NMES training. patients with CNS disorders. Moreover, it is crucial to inves-
Although there are discrepancies in the literature concerning tigate whether such neural drive improvement is maintained
the specific response to NMES versus voluntary actions (Bax et after the interruption of NMES training.
al. 2005), similar findings have been reported after voluntary In conclusion, the present study demonstrated that the sig-
resistance training programs (Aagaard et al. 2002; Sale et al. nificant increase in both plantar flexor MVC and associated
1983). For example, a 55% increase in SOL V-wave amplitude normalized EMG activity observed after 5 wk of NMES
was reported after 14 wk of heavy-resistance strength training training could be attributed to an enhanced neural drive in
(Aagaard et al. 2002). An increase in both V-wave amplitude descending corticospinal and/or extrapyramidal pathways from
and normalized EMG such as those obtained in our study for supraspinal centers and/or neural mechanisms occurring at the
both SOL and gastrocnemii may result from an enhanced spinal level.
neural drive in descending corticospinal and/or extrapyramidal
pathways, elevated motoneuron excitability, and/or alterations ACKNOWLEDGMENTS
in presynaptic inhibition (Aagaard et al. 2002). The authors thank the subjects for participation in this study. We also
Thus the observed increase in SOL and gastrocnemii V- express our gratitude to Dr. Nicola Maffiuletti for critical reading of the
wave amplitude could be attributable to an increase in mo- manuscript, to L. Rochette for helpful contribution to data analyses, and to M.
toneuron firing frequency and/or recruitment. Both mecha- Bouley for the English revision of the manuscript.
nisms result indeed in a direct proportional increase in the
REFERENCES
probability of antidromic collision arising from an enhanced
volitional drive from higher motor centers. In support of this Aagaard P, Simonsen EB, Andersen JL, Magnusson P, and Dyhre-Poulsen
hypothesis, recent functional neuroimaging studies have dem- P. Neural adaptation to resistance training: changes in evoked V-wave and
H-reflex responses. J Appl Physiol 92: 2309 –2318, 2002.
onstrated significant cerebral cortex activation during electri- Allen GM, Gandevia SC, and McKenzie DK. Reliability of measurements of
cally evoked contractions of human wrist extensor (Han et al. muscle strength and voluntary activation using twitch interpolation. Muscle
2003) and knee extensor muscles (Smith et al. 2003). Such Nerve 18: 593– 600, 1995.
activation of selected brain regions during NMES might be Bax L, Staes F, and Verhagen A. Does neuromuscular electrical stimulation
strengthen the quadriceps femoris? A systematic review of randomised
involved in the increased volitional drive from the supraspinal controlled trials. Sports Med 35: 191–212, 2005.
centers. In the same way, several studies (Khaslavskaia and Belanger M, Stein RB, Wheeler GD, Gordon T, and Leduc B. Electrical
Sinkjaer 2005; Khaslavskaia et al. 2002; Thompson et al. stimulation: can it increase muscle strength and reverse osteopenia in spinal
2005), using transcranial magnetic stimulation, found in- cord injured individuals? Arch Phys Med Rehabil 81: 1090 –1098, 2000.
creased amplitude of motor evoked potentials in the tibialis Collins DF, Burke D, and Gandevia SC. Large involuntary forces consistent
with plateau-like behavior of human motoneurons. J Neurosci 21: 4059 –
anterior after a single session of NMES. Nevertheless, further 4065, 2001.
research is warranted to determine the relative contribution of Colson S, Martin A, and Van Hoecke J. Re-examination of training effects
motor unit recruitment versus discharge rate in the NMES by electrostimulation in the human elbow musculoskeletal system. Int
training-induced V-wave enhancement. J Sports Med 21: 281–288, 2000.
Crone C, Hultborn H, Mazieres L, Morin C, Nielsen J, and Pierrot-
Although the lack of change in Hsup/Msup ratio indicates that Deseilligny E. Sensitivity of monosynaptic test reflexes to facilitation and
the spinal reflex was unchanged after the present NMES inhibition as a function of the test reflex size: a study in man and the cat. Exp
training program, the V-wave increment could be explained, at Brain Res 81: 35– 45, 1990.

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.
 NEURAL ADAPTATIONS TO NMES TRAINING 3335

Duclay J and Martin A. Evoked H-reflex and V-wave responses during Maffiuletti NA, Pensini M, Scaglioni G, Ferri A, Ballay Y, and Martin A.
maximal isometric, concentric, and eccentric muscle contraction. J Neuro- Effect of electromyostimulation training on soleus and gastrocnemii H- and
physiol 94: 3555–3562, 2005. T-reflex properties. Eur J Appl Physiol 90: 601– 607, 2003.
Enoka RM. Muscle strength and its development. New perspectives. Sports Martin L, Cometti G, Pousson M, and Morlon B. Effect of electrical
Med 6: 146 –168, 1988. stimulation training on the contractile characteristics of the triceps surae
Glanz M, Klawansky S, Stason W, Berkey C, and Chalmers TC. Func- muscle. Eur J Appl Physiol Occup Physiol 67: 457– 461, 1993.
tional electrostimulation in poststroke rehabilitation: a meta-analysis of the Merletti R, Rainoldi A, and Farina D. Surface electromyography for non-
randomized controlled trials. Arch Phys Med Rehabil 77: 549 –553, 1996. invasive characterization of muscle. Exerc Sport Sci Rev 29: 20 –25, 2001.
Gondin J, Guette M, Ballay Y, and Martin A. Electromyostimulation Parker MG, Bennett MJ, Hieb MA, Hollar AC, and Roe AA. Strength
training effects on neural drive and muscle architecture. Med Sci Sports
response in human femoris muscle during 2 neuromuscular electrical stim-
Exerc 37: 1291–1299, 2005.
ulation programs. J Orthop Sports Phys Ther 33: 719 –726, 2003.
Han BS, Jang SH, Chang Y, Byun WM, Lim SK, and Kang DS. Functional
magnetic resonance image finding of cortical activation by neuromuscular Pensini M and Martin A. Effect of voluntary contraction intensity on the
electrical stimulation on wrist extensor muscles. Am J Phys Med Rehabil 82: H-reflex and V-wave responses. Neurosci Lett 367: 369 –374, 2004.
17–20, 2003. Pierrot-Deseilligny E and Mazevet D. The monosynaptic reflex: a tool to
Hermens HJ, Freriks B, Disselhorst-Klug C, and Rau G. Development of investigate motor control in humans. Interest and limits. Neurophysiol Clin
recommendations for SEMG sensors and sensor placement procedures. J 30: 67– 80, 2000.
Electromyogr Kinesiol 10: 361–374, 2000. Powell J, Pandyan AD, Granat M, Cameron M, and Stott DJ. Electrical
Hortobagyi T, Scott K, Lambert J, Hamilton G, and Tracy J. Cross- stimulation of wrist extensors in poststroke hemiplegia. Stroke 30: 1384 –
education of muscle strength is greater with stimulated than voluntary 1389, 1999.
contractions. Motor Control 3: 205–219, 1999. Sale DG, MacDougall JD, Upton AR, and McComas AJ. Effect of strength
Hultborn H and Pierrot-Deseilligny E. Changes in recurrent inhibition training upon motoneuron excitability in man. Med Sci Sports Exerc 15:
during voluntary soleus contractions in man studied by an H-reflex tech- 57– 62, 1983.
nique. J Physiol 297: 229 –251, 1979. Scaglioni G, Narici MV, Maffiuletti NA, Pensini M, and Martin A. Effect
Hultman E, Sjoholm H, Jaderholm-Ek I, and Krynicki J. Evaluation of of ageing on the electrical and mechanical properties of human soleus motor
methods for electrical stimulation of human skeletal muscle in situ. units activated by the H reflex and M wave. J Physiol 548: 649 – 661, 2003.
Pfluegers Arch 398: 139 –141, 1983. Schieppati M. The Hoffmann reflex: a means of assessing spinal reflex
Kern H, Salmons S, Mayr W, Rossini K, and Carraro U. Recovery of excitability and its descending control in man. Prog Neurobiol 28: 345–376,
long-term denervated human muscles induced by electrical stimulation. 1987.
Muscle Nerve 31: 98 –101, 2005. Smith GV, Alon G, Roys SR, and Gullapalli RP. Functional MRI
Khaslavskaia S, Ladouceur M, and Sinkjaer T. Increase in tibialis anterior determination of a dose-response relationship to lower extremity neuro-
motor cortex excitability following repetitive electrical stimulation of the muscular electrical stimulation in healthy subjects. Exp Brain Res 150:
common peroneal nerve. Exp Brain Res 145: 309 –315, 2002. 33–39, 2003.
Khaslavskaia S and Sinkjaer T. Motor cortex excitability following repeti- Thompson AK, Doran B, and Stein RB. Short-term effects of functional
tive electrical stimulation of the common peroneal nerve depends on the electrical stimulation on spinal excitatory and inhibitory reflexes in ankle
voluntary drive. Exp Brain Res 162: 497–502, 2005. extensor and flexor muscles. Exp Brain Res 1–11, 2005.
Kimberley TJ, Lewis SM, Auerbach EJ, Dorsey LL, Lojovich JM, and Upton AR, McComas AJ, and Sica RE. Potentiation of “late” responses
Carey JR. Electrical stimulation driving functional improvements and evoked in muscles during effort. J Neurol Neurosurg Psychiatry 34: 699 –
cortical changes in subjects with stroke. Exp Brain Res 154: 450 – 460, 2004. 711, 1971.
Lagerquist O, Zehr PE, and Docherty D. Increased spinal reflex excitability Voigt M, Chelli F, and Frigo C. Changes in the excitability of soleus muscle
is not associated with neural plasticity underlying the cross-education effect. short latency stretch reflexes during human hopping after 4 weeks of
J Appl Physiol 100: 83–90, 2006. hopping training. Eur J Appl Physiol Occup Physiol 78: 522–532, 1998.
Maffiuletti NA, Martin A, Babault N, Pensini M, Lucas B, and Schieppati Wade DT. Measuring arm impairment and disability after stroke. Int Disabil
M. Electrical and mechanical H(max)-to-M(max) ratio in power- and Stud 11: 89 –92, 1989.
endurance-trained athletes. J Appl Physiol 90: 3–9, 2001. Zehr PE. Considerations for use of the Hoffmann reflex in exercise studies.
Maffiuletti NA, Pensini M, and Martin A. Activation of human plantar flexor Eur J Appl Physiol 86: 455– 468, 2002.
muscles increases after electromyostimulation training. J Appl Physiol 92: Zhou S. Chronic neural adaptations to unilateral exercise: mechanisms of cross
1383–1392, 2002. education. Exerc Sport Sci Rev 28: 177–184, 2000.

J Neurophysiol • VOL 95 • JUNE 2006 • www.jn.org

Downloaded from journals.physiology.org/journal/jn (087.223.122.019) on March 4, 2022.