le
ier, M
Motoneurone
Motor unit
Synchronization
Cross-correlogram
con
, mo
ync
was assessed using the index k0 , the synchronous impulse probability (SIP), and the synchronous impulse
s activ
can be observed with fatigue and neurological diseases (Baker
et al., 1992; Buchthal and Madsen, 1950; McAuley and Marsden,
2000).
MN synchronous activity is classically attributed to the action of
common inputs and/or to that of pre-synaptically correlated inputs
(Kirkwood and Sears, 1991). The quasi-simultaneous occurrence of
excitatory post-synaptic potentials (EPSPs) generated by branching
common inputs innervating MNs in the same or different motor
2003; Uchiyama and Windhorst, 2007).
Due to the challenge of reliably discriminating single motor unit
action potentials as muscle contraction increases, the influence of
contraction strength on MN synchronous activity is poorly docu-
mented. With stronger contraction, more common and/or synchro-
nized inputs are likely to contribute to the increase in MN drive
and generate more synchronous firings. This effect might be coun-
terbalanced, however, by the impact that MN faster firing rates
may have on synaptic effectiveness (cf. Türker and Powers, 2002).
In previous studies, various indices have been used to assess
MN synchrony with quite discordant findings. Depending on the
* Corresponding author. Tel.: +33 491 164 357.
Clinical Neurophysiology xxx (2010) xxx–xxx
Contents lists availab
ro
.e l
ARTICLE IN PRESSE-mail address: schmied@dpm.cnrs-mrs.fr (A. Schmied).strategies underlying neuromuscular control is still a matter of de-
bate. For some, motoneuronal synchronization is simply a by-prod-
uct of spinal connectivity (De Luca et al., 1993; Freund, 1983). For
others, the occurrence of synchronous discharges looks like an
efficient way for the central nervous system to coordinate the
activity of MNs and optimize motor performance (Baker et al.,
1999; Bremner et al., 1991; Milner-Brown et al., 1975; Santello
and Fuglevand, 2004). However, too high a level of oscillatory syn-
chronization leads to tremor detrimental to motor performance, as
nized at a pre-motoneuronal level are also expected to produce
synchronous MN discharges with looser coupling depending on
the number of synapses between the synchronizing source and
the MNs according to the process of long-term or ‘‘broad peak”
synchronization (Kirkwood et al., 1982). The cortico-spinal path-
way is considered as a major source of synchronizing inputs during
voluntary contraction in humans (Datta et al., 1991; Schmied et al.,
1999), without excluding, however, the contribution of other
sources, such as spinal inhibitory interneurons (Mattei et al.,1. Introduction
Whether or not MN synchronou1388-2457/$36.00  2010 International Federation o
doi:10.1016/j.clinph.2010.02.165
Please cite this article in press as: Schmied A, De
siol (2010), doi:10.1016/j.clinph.2010.02.165frequency (SIF) in cross-correlograms. MU inter-spike interval duration and variability, surface EMG
activity and force output were evaluated concurrently.
Results: Both SIP and SIF increased with contraction strength, whereas k0 remained unaffected. Faster fir-
ing rates and stronger contraction had the greatest effects on SIF.
Conclusions: By testing the same MUs at different force levels, we showed that contraction strength does
influence MN synchrony. The enhancement of MU synchrony with stronger contraction suggests an effi-
cient contribution of more common and/or synchronized inputs.
Significance: Force output must be controlled when assessing MN synchrony. Normalizing MU synchro-
nous activity per reference spike is preferable to minimize the influence of firing rate. This is particularly
relevant for clinical research, in conditions of poorer neuromuscular control.
 2010 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights
reserved.
ity plays a role in the
nuclei produces synchronous discharges tightly coupled within
less than 10 ms according to the process of short-term synchroni-
zation (Sears and Stagg, 1976). Inputs that are themselves synchro-Keywords:
ative impact of MN faster firing rates on synaptic effectiveness.
Methods: Pairs of motor units (MUs) were tested at various force levels, in 2-s sequences. MN synchronyInfluence of contraction strength on sing
Annie Schmied a,*, Martin Descarreaux b
a Plasticity and Physiopathology of Movement, UMR 6196 CNRS 31, Chemin Joseph Aigu
bDepartment of Chiropractic-University of Quebec, Trois-Rivières, QC, Canada G9A 5H7
a r t i c l e i n f o
Article history:
Accepted 17 February 2010
Available online xxxx
a b s t r a c t
Objective: The influence of
With stronger contraction
drive and generate more s
Clinical Neu
journal homepage: wwwf Clinical Neurophysiology. Publish
scarreaux M. Influence of contrmotor unit synchronous activity
arseilles 13402 Cedex 20, France
traction strength on motoneurone (MN) synchrony is poorly documented.
re common and/or synchronized inputs might contribute to greater MN
hronous firings. This effect might be counterbalanced, however, by a neg-
le at ScienceDirect
physiology
sevier .com/locate /c l inphed by Elsevier Ireland Ltd. All rights reserved.
action strength on single motor unit synchronous activity. Clin Neurophy-

the steady discharges of two clearly-identifiable motor units over
l Ne
ARTICLE IN PRESS2–3 min. Each motor unit pair was tested at various force levels
during at least two (up to 7) recording periods separated by
2-min resting periods, during which the subjects were asked to re-
lax completely back to zero force level. The force target was ad-
justed in random order between the lowest and the highest
levels of force allowing a reliable discrimination of both MUs.
The force signal was calibrated in Newtons before and after each
recording period. To stay close to the natural conditions of motor
control, a prerequisite is to minimize the possibility of a selective
voluntary modulation of the firing frequency of the MUs tested.
Therefore, the subjects only relied on force signal feedback to
maintain the contraction as steady as possible without any feed-
back of motor unit firing discharge.
2.2. Data recordingmuscle tested and/or, most importantly on the synchronization in-
dex used, with stronger muscle contraction and/or faster MU fring
rates, MN synchrony pattern was reported to be unaffected (Kano-
sue et al., 1979; Christou et al., 2007; Fling et al., 2009; Huesler
et al., 2000), reduced (Nordstrom et al., 1992) or enhanced (Datta
and Stephens, 1990; Schmied et al., 1993, 1994; Fling et al.,
2009;). In view of these discrepancies, further investigation is
clearly needed.
In the present study, the same MUs have been tested at differ-
ent force levels in order to assess the concurrent changes in MN
synchrony. We showed that contraction strength does influence
MN synchrony. As MN synchrony increases with stronger contrac-
tion, it appears that the negative impact of faster firing rates on
synaptic effectiveness (Türker and Powers, 2002) does not prevail
facing the recruitment of more common and/or synchronized in-
puts as MN drive increases.
2. Methods
The experimental procedure was approved by the Ethics Com-
mittee of the University of Marseilles-II (CCPPRB protocol 03005).
Experiments were carried out on eight healthy human subjects
(males, aged 23–33 years, right-handed) who signed an informed
consent form prior to the experimental procedure, as required by
the Declaration of Helsinki.
2.1. Experimental procedure
As described in a previous report (Schmied et al., 1999), the sub-
jects were seated in an adjustable armchair. Their right forearm
was placed in a cushioned groove so that a stereotyped position
could be maintained throughout the experiment and from one
experiment to another. The distal end of the forearm was immobi-
lized in a U-shaped device maintaining the hand in a semi-prone
position, with the wrist flexed at an angle of 10.
The subjects were asked to produce isometric contraction of
their wrist extensor muscles by pushing with the back of their
hand on a vertical metal bar mechanically connected to a force
transducer device (FT03, Grass, W. Warwick, RI USA). The exten-
sion force signal was displayed on-line on an oscilloscope facing
the subject (2 ms  1 N per division). The subjects had to maintain
the force level on a target materialized by two markers (height
0.3 cm, length 1 cm) located on both sides at mid-screen level of
the oscilloscope. The target was adjusted by the experimenters in
such a way that extension force was set high enough to record
2 A. Schmied, M. Descarreaux / ClinicaSurface EMG activity was recorded on the right extensor carpi
radialis (ECR) muscles. The synchronous activity of motor units
tested in the ECR muscles may be influenced by the degree of
Please cite this article in press as: Schmied A, Descarreaux M. Influence of contr
siol (2010), doi:10.1016/j.clinph.2010.02.165co-activation of antagonistic muscles (Sturm et al., 2000). There-
fore EMG activity on the wrist flexor muscles was recorded concur-
rently as a control for the lack of consistent changes in the amount
of co-activation of antagonistic muscles across the various force
levels tested. EMG activities were recorded by pairs of non-polari-
sable single-use surface electrodes (Ag–AgCl) placed 2 cm apart
over the belly of the ECR and wrist flexor muscles.
The discharges of two single motor units were recorded concur-
rently on the right ECR muscle by means of two sterilized single-
use tungsten microelectrodes (impedance 12 MX tested at 1 kHz;
Frederick Haer and Co., Bowdoinham, ME, USA). The microelec-
trodes were inserted between the ECR surface EMG electrodes
and moved in minute steps until a stable recording of clearly-iden-
tifiable single motor unit action potentials was obtained from each
microelectrode. The quality of the recording was monitored upon
discriminating motor unit action potentials on-line through dual-
window, time–amplitude discriminators (BAK Electronics, Inc.,
Mount Airy, MD, USA).
The surface electrodes and microelectrodes were connected to
amplifiers via probes (P5, Grass Instruments Co., Quincy, MA,
USA) with an isolated ground for optimum protection of the test
subjects (current leakage less than 3 mA). EMG and motor unit
activities were amplified and filtered (band-pass: 30 Hz–1 kHz,
300 Hz–3 kHz, respectively). Force, EMG and single motor unit sig-
nals were digitized (sampling rates: 1, 5 and 30 kHz, respectively)
and stored on a computer with an acquisition device (1401 plus
interface, Spike2-5 software) from the Cambridge Electronic De-
sign company (CED, Cambridge, UK).
At the end of the experiment, the microelectrodes were with-
drawn and the subject was asked to produce 2–3 bouts of maximal
voluntary isometric wrist extension and flexion forces for a few
seconds, under strong verbal encouragements of the experiment-
ers. The highest level of EMG activity assessed in root mean square
(RMS) values in these bouts was subsequently used as a reference
to normalize wrist extensor and flexor EMG activity as a percent-
age of maximal voluntary muscle contraction (% MVC) during each
of the recording sequence performed in the same experiment.
2.3. Single motor unit firing pattern analysis
MU action potentials were re-discriminated off-line and ana-
lysed by Spike 2 software (CED Company, Cambridge UK). To check
for the absence of abnormally short inter-spike intervals (ISIs),
which may reveal the contribution of other MUs to the recorded
signal, the firing behaviour of each MU was plotted on an instanta-
neous frequency curve, as shown in Fig. 1D and E. The presence of
abnormally low instantaneous frequency values corresponding to
ISIs twice as long as the average was carefully monitored to ensure
that no spike had been missed in the discrimination process.
Each MU was identified on the basis of the signature of its over-
all electrical activity or motor unit potential (MUP), which repre-
sents the sum of action potentials of the muscle fibre population
innervated by the same MN. The corresponding MUP was extracted
by spike-triggered averaging of surface EMG activity of the ECR
muscles using events corresponding to discriminated MU action
potential as triggers throughout each of the recording periods.
Their reproducibility of size and shape of the MUPs extracted
was examined to ensure that the same MUs were being tested
throughout all of the recording periods at different levels of force.
A MU firing pattern was characterized on the basis of ISI dura-
tions measured in successive, non-overlapping, 5-s steps at each
contraction level. ISIs longer than 300 ms (about 4–5 times the
mean) were taken to result from pauses in MU tonic activity and
urophysiology xxx (2010) xxx–xxxwere not included in the calculation of inter-spike interval statis-
tics. The mean value (ISI mean) and standard deviation (ISI SD)
of the ISIs were calculated across each of the 5-s steps, and ISI var-
action strength on single motor unit synchronous activity. Clin Neurophy-