The function of neuromuscular compartments in

human shoulder muscles

J. B. Wickham and J. M. M. Brown
J Neurophysiol 107:336-345, 2012. First published 5 October 2011; doi:10.1152/jn.00049.2011

You might find this additional info useful...

This article cites 43 articles, 11 of which can be accessed free at:

/content/107/1/336.full.html#ref-list-1

Updated information and services including high resolution figures, can be found at:
/content/107/1/336.full.html

Additional material and information about Journal of Neurophysiology can be found at:
http://www.the-aps.org/publications/jn

This information is current as of October 22, 2014.

Downloaded from on October 22, 2014

Journal of Neurophysiology publishes original articles on the function of the nervous system. It is published 12 times a year
(monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2012 by the
American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at http://www.the-aps.org/.
J Neurophysiol 107: 336 –345, 2012.
First published October 5, 2011; doi:10.1152/jn.00049.2011.
The function of neuromuscular compartments in human shoulder muscles
J. B. Wickham,1 and J. M. M. Brown2
1
School of Biomedical Sciences, Charles Sturt University, Orange; and 2School of Biomedical Science, University
of Queensland, St. Lucia, Australia
Submitted 18 January 2011; accepted in final form 30 September 2011
Wickham JB, Brown JM. The function of neuromuscular com-
partments in human shoulder muscles. J Neurophysiol 107: 336 –345,
2012. First published October 5, 2011; doi:10.1152/jn.00049.2011.—
The aim of this study was to use a surface electromyographic (sEMG)
technique with a ballistic isotonic shoulder joint adduction movement
to determine the function of the neuromuscular compartments
(NMCs) within the pectoralis major, deltoid, and latissimus dorsi
muscles. Sixteen male subjects (mean age 22 yr) with no known
history of shoulder pathologies volunteered to participate. Timing and
intensity of muscle contraction, recorded with 15 pairs of bipolar
sEMG electrodes, were compared during performance of 40° coronal-
plane ballistic [movement time (MT) ⬍ 400 ms] shoulder joint
adduction movements. The results suggested that heterogeneous
sEMG was present across the breadth of all three muscles, indicating
the presence of individual NMCs with significant (P ⬍ 0.05) differ-
ences observed within the three muscles in NMC onset, duration,
timing of peak NMC intensity, or relative intensity of NMC activa-
tion. For example, within the deltoid NMC activation was closely
related to moment arm (MA) length with the NMC, with the largest
antagonist MA deltoid NMC3 having a late period of activation
[antagonist (Ant)] to slow glenohumeral joint (GHJ) rotation and
maintain its final joint position [with agonist 2 burst (Ag2)]. The most
obvious triphasic EMG patterns (e.g., Ag1-Ant-Ag2) were observed
between the first NMCs activated in the two agonist muscles and the
last NMC activated in the antagonist deltoid muscle. In conclusion,
our findings suggest the presence of in-parallel NMCs within the
superficial muscles of the GHJ and show that biomechanical param-
eters, such as the MA at end-point movement position, influence the
function of each NMC and its contribution to alternating patterns of
agonist and antagonist muscle activity typical of ballistic movement.
muscle segment; triphasic electromyogram
IN BOTH HUMANS AND ANIMALS, skeletal muscles may be parti-
tioned into neuromuscular compartments (NMCs) (Lucas-
Osma and Collazos-Castro 2009; Mu and Sanders 2001; Wid-
mer et al. 1997). A NMC has been defined as the “smallest
portion of a muscle to receive exclusive innervations by a set
of motoneurons” (English et al. 1993). NMCs are generally
thought of as a subvolume within a muscle that completely
encloses the territories of a group of functionally similar
motoneurons (English and Weeks 1984, 1989). Although NMC
boundaries have yet to be fully defined, it has been proposed
that NMCs are generally arranged either in parallel across the
breadth of a muscle (Korfage and Van Eijden 1999) or in series
(Lexell et al. 1994), although other arrangements have been
proposed (Mu and Sanders 2001). Importantly, NMCs may be
independently controlled by the central nervous system (CNS)
to provide fine control of muscle force around a tendon or joint
(Brown et al. 2007; Soderberg and Dostal 1978), with the
Address for reprint requests and other correspondence: J. M. M. Brown,
School of Biomedical Science, Univ. of Queensland, St. Lucia, 4072 Australia
(e-mail: markbrown@uq.edu.au).
336 0022-3077/12 Copyright © 2012 the American Physiological Society
length of their moment arms and the direction of their fiber
orientation (lines of action) determining its subsequent func-
tion at that joint (Ackland et al. 2008; Ettema et al. 1998).
Furthermore, it has been hypothesized that a neural basis for
the independent control of NMCs may be that the descending
supraspinal motor commands are directed to portions of a
muscle’s spinal motoneuronal pool as seen in the flexor digi-
torum superficialis in humans (Butler et al. 2005) and in the
hand muscles of the monkey (Buys et al. 1986; Lemon and
Muir 1983; Maier et al. 1993).
Although evidence of the spatial location of NMCs is accu-
mulating in muscles such as the masseter (Widmer et al. 2003),
Downloaded from on October 22, 2014
temporalis (Korfage and Van Eijden 1999), and vastus lateralis
(Sjostrom et al. 1992), little evidence is currently available
regarding the function and spatial location of NMCs within
muscles producing movements of the glenohumeral joint
(GHJ). Structurally, muscles attaching the trunk and shoulder
girdle to the humerus are characterized by broad origins with
muscle fascicles converging toward a narrower insertion. In-
tuitively, the convergent lines of muscle fascicle application
onto the humerus might favor an in-parallel arrangement of
NMCs, as suggested by previous studies (Paton and Brown
1994, 1995; Wickham and Brown 1998). To test the function
of these potential NMCs, three convergent and predominantly
parallel-fibered GHJ muscles, the pectoralis major, the deltoid,
and the latissimus dorsi, were investigated. To investigate the
presence and function of NMCs a surface electromyographic
(sEMG) technique was utilized. In a previous study sEMG was
used to locate NMCs within the human triceps surae muscle
(Staudenmann et al. 2009). By covering a large portion of that
muscle with miniature surface electrodes it was possible to
locate heterogeneity (e.g., variations in the timing or intensity)
of muscle activation indicative of underlying NMC activity.
Therefore, the first aim of the present study was to utilize a
multichannel sEMG technique to identify individual NMCs by
detecting heterogeneous sEMG activity across the surface of
three superficial muscles controlling movements of the GHJ.
The second aim of the study was to determine the function of
each identified NMC by correlating its timing and intensity of
activation, and also its moment arm length, to movement
variables such as the initial and peak displacement of the upper
limb, peak velocity, and peak acceleration.
To stress the GHJ muscles into robust activity that might
amplify heterogeneity of NMC function, a ballistic isotonic
GHJ adduction task was utilized. Ballistic movements are
controlled by a complex series of agonist (Ag1)-antagonist
(Ant)-agonist (Ag2) muscle activations (Angel 1975; Cheron et
al. 2007) which initially accelerate the limb (Ag1), and then
brake it (Ant), before finally positioning it at its target joint
angle (Ag2). Both high intensities of muscle activation, and
www.jn.org
 NEUROMUSCULAR COMPARTMENTS
complex intermuscular coordinations, are required to complete
ballistic motor tasks in a total movement time (MT) of ⬍400
ms (Brown and Gilleard 1991; Brown and Cooke 1981). It was
expected that if NMCs within the muscles investigated were
independently controlled by the CNS, then significant (P ⬍
0.05) variations in the timing and intensity of NMC sEMG
activity would occur as each NMC was recruited to play its
unique functional role, as determined by its moment arm, in the
ballistic shoulder adduction task.
METHODS
Experimental Procedures
The work described in this study was approved by the Human
Ethics Committee of the University of Wollongong.
Sixteen right-hand-dominant male subjects (mean age 22 yr; range
18 –30 yr), with no known history of shoulder pathologies, volun-
teered to participate in this study after signing an informed consent
approved by the University of Wollongong Human Ethics Committee.
Each subject was seated in a dental chair with his 90° abducted right
upper limb resting on a padded arm rest just distal to the elbow joint
to eliminate excessive valgus forces at the elbow (Fig. 1). The arm rest
was attached by lightweight rope that ran through an overhead
electrogoniometer (10 turn, 20 k⍀ via DC amplifier) before attaching
to adjustable free weights positioned on the floor. No muscular
activity was necessary to maintain the abducted upper limb position,
as the free weight balanced the weight of the resting upper limb.
Each subject was asked to perform a series of ballistic shoulder
adduction movements in response to a verbal “go” signal. From an
initial 90° position of shoulder joint abduction (in the coronal plane)
the GHJ was rapidly (MT ⬍400 ms) rotated, against the resistance
afforded by the free weights, through 40° of adduction as measured by
the electrogoniometer utilizing visual feedback of limb position and
target angle displayed on a cathode ray oscilloscope.
After an initial series of 30 practice trials, a further 10 trials, at an
intertrial interval of at least 30 s, were recorded for analysis. The
subjects were asked to perform each trial “as fast and as accurately as
Fig. 1. Experimental setup. The right arm was firmly secured within an arm rest
that was connected to an overhead electrogoniometer and then to a floor
weight, which was adjusted to equilibrate the weight of the abducted upper
limb. Muscle activity within each neuromuscular compartment (NMC), during
each ballistic 40° glenohumeral joint (GHJ) adduction movement (90° to 50°
of abduction), was detected through 15 pairs of miniature surface electromyo-
graphic (sEMG) bipolar electrodes. Subject feedback on movement accuracy
was provided by a cathode ray oscilloscope (not illustrated). Movement time
was ⬍400 ms.
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
337
possible.” Trials that were not within ⫾5° of the target angle of 50°
of abduction (⬃50% of attempted trials) were rejected from the
analysis.
Electromyographic Technique
To minimize electrode cross talk, miniature gold-plated bipolar
surface electrodes (6.5-mm interelectrode distance; 1.6-mm active
plates) were manufactured by the authors. Fifteen pairs of bipolar
surface electrodes were applied to each subject: four over the deltoid,
five over the latissimus dorsi, and six over the pectoralis major
(Fig. 2). The anatomical criteria utilized to position each bipolar
surface electrode, as determined by anatomical dissection, were
guided by evidence of NMC location suggested in previous studies
(Wickham et al. 2004b). Briefly, these criteria included distinctive
origins and/or insertions, architectural differences within the muscle
mass (strap vs. pennated segments), and intramuscular facial thicken-
ings within the muscle mass (Johnson et al. 1994). Where broad
expanses of seemingly homogeneous tissue were evident, muscle
fascicles with a ⬎10° difference in angle, compared with adjacent
fascicles, were considered potentially independent (Wickham and
Brown 1998).
The miniature surface electrodes were secured to the subject by
double-sided tape after the skin had been shaved, abraded, and washed
with alcohol to reduce skin resistance. Electrode gel was utilized to
Downloaded from on October 22, 2014
improve conductance. Reference surface electrodes (3M pediatric
electrodes) were positioned over the acromion, the clavicle, and the
anterior superior iliac spine. The 15 bipolar surface electrodes, and
their reference electrodes, were connected to differential preamplifiers
and HUMTEC 100 EMG amplifiers (input impedance of 1 ⫻ 1013 ⍀,
CMRR of 110 db, signal-to-noise ratio of 1,000:1, 1000 k gain). The
raw EMG signals were amplified and filtered (10-Hz high-pass and
1-kHz low-pass Butterworth filters) prior to A/D conversion (with the
electrogoniometer output) at a sample rate of 2,000 Hz over a
2,000-ms recording period.
Moment Arm Analysis
Currently there do not exist moment arm data at 90° of cadaveric
shoulder abduction (start position of movement) for each NMC
analyzed here because of technical difficulties related to cadaveric
joint integrity at high angles of shoulder abduction. Moment arm data
are available, however (Wickham et al. 2004b), at 50° abduction,
which was the end position of the movement employed here.
The method utilized to determine moment arm involved the posi-
tioning of a bisected cadaver on a fixture in front of a milling machine
and manipulating the X, Y, and Z slides of the mill to gather
three-dimensional coordinates of origin and insertion points relative to
a reference position and then later using AutoCAD software to obtain
lines of action and moment arm data. It should also be noted that a
fixed center of rotation located in the middle of the humeral head was
used as the axis of rotation, which has also been used in previous
biomechanical models of the shoulder joint (Hogfors et al. 1987).
Data Analysis
Data analysis utilized Digital Signal Processing (DSP) software for
assessment of the timing and intensity of sEMG (Fig. 3) within each
NMC.
Timing analysis. For each trial the beginning (OnD) and end of
GHJ motion (PkD) were determined (Fig. 3) as well as measures of
GHJ peak velocity (PkV) and peak acceleration (PkA) (Fig. 4).
To determine the timing of muscle activity within each NMC, the
raw EMG waveforms were rectified and smoothed with a 20-Hz
low-pass filter. Threshold detectors (10% peak amplitude), combined
with visual analysis, were used to identify the onset (On), peak (Pk),
and cessation (Off) of activity within each muscle segment in relation
338 NEUROMUSCULAR COMPARTMENTS

Fig. 2. sEMG electrode placement. The NMCs of

the 3 shoulder muscles investigated were subjected
to sEMG analysis. Anatomical observations of pre-
vious studies (Paton and Brown 1994, 1995; Wick-
ham and Brown 1998) were used as a guide to
position each electrode pair over a series of in-
parallel NMCs within the pectoralis major (P1–P6),
deltoid (D1–D7), and latissimus dorsi (L1–L6). Be-
cause of subsequent sEMG channel limitations (a
16-channel EMG amplifier was used; 15 sEMG and
1 channel for the movement trace), sEMG from L4,
D2, D4, and D6, used in the above-mentioned for-
mer studies, were excluded from the present analy-
sis. Evidence of significant (P ⬍ 0.05) heterogeneity
in sEMG (timing or intensity) across the breadth of
the muscles investigated was used to identify indi-
vidual NMCs.

Downloaded from on October 22, 2014

to OnD (Fig. 4). From these data the duration (Dur) of muscle
activation could be determined. The timing of muscle peak intensity
(Pk) was also compared with PkV and PkA within each trial.
Intensity analysis. Each EMG waveform was integrated (iEMG) to
determine the intensity of muscle activation within each NMC. Inte-
gration was performed on the duration of each burst, which repre-
sented an average of 210-ms duration (60% of the overall movement
duration of 354 ms). Full-wave rectification and then integration
(10-ms bins) determined the iEMG for each NMC. The iEMG was the
average intensity of motor unit activation within each NMC. To
normalize the intensity data, the average iEMGs from all NMCs
within each of the three muscles, from each trial, were totaled together
to produce a “muscle” iEMG. Within each muscle, the intensity of
motor unit activation of each NMC was then expressed as a percent-
age of the total intensity of that particular muscle [e.g., normalized
iEMG NMC1 (%) ⫽ (iEMG NMC1/total iEMG of muscle) ⫻ 100].
The intensity of the first burst of the agonist and of the antagonist
NMCs were expressed in this way. Any additional burst (2nd burst) of
activity within the agonist or antagonist NMCs (if present) was
expressed as a percentage of the first burst.
Statistical Evaluation
Repeated-measures two-way analysis of variance (RM ANOVA)
with independent variables of NMC and movement was used. It
should be noted that the original statistical design incorporated two
movements (adduction and extension), although only the adduction
data are reported in the present article. Student-Neumann-Keuls post
hoc analysis techniques were used to evaluate the data when the main
effect of NMC was significant (P ⬍ 0.05) for that particular dependent
variable (e.g., peak, intensity, duration). Only one segment from each
muscle was chosen to perform post hoc comparisons in an effort to
simplify and clarify the data analysis. The earliest activated NMCs of
each muscle [pectoralis major (P)6, latissimus dorsi (L)6, and deltoid
(D)7] were chosen for these comparisons as these were, from previous
experience, most likely to show significant (P ⬍ 0.05) differences
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
compared with the remainder of its muscle mass. Pearson’s correla-
tion coefficients were used to determine the relationship between the
timing of NMC activity and the NMC’s location within the muscle.
Preliminary analysis revealed that performing correlations on individ-
ual data points for each subject instead of group average data gave
similar correlation results compared with group average data. All
correlation coefficients and all other data (On, Pk, Dur, and intensity)
are therefore expressed in the text as group average data. The statis-
tical evaluations were designed to determine whether the timing and
contraction intensity of each NMC varied significantly 1) within each
individual muscle and 2) across all three muscles investigated. The
level of significance was set at P ⬍ 0.05.
RESULTS
Timing of NMC Activation
An overview of the temporal patterns of muscle segment
activation, relative to the movement of the upper limb in
adduction, is provided in Fig. 5. Given the ballistic nature of
the GHJ movement task, differences in the timing (e.g., On,
Off) of NMC activity were measured in milliseconds, re-
sisting statistical interpretation. As a result, regression anal-
ysis, rather than ANOVA, was found to be the more useful
evaluation tool.
The raw data (Table 1) suggested that all NMCs of pectoralis
major and latissimus dorsi were activated before the adduction
movement was initiated (On) and reached peak intensity before
or just after movement onset (OnD). All NMCs within these
muscles had periods of activation (Dur) that approximated 50%
of MT. Their time of peak intensity (Pk) occurred well before
PkV and peak deceleration (PkDc) but generally approximated
PkA (Table 2).
 NEUROMUSCULAR COMPARTMENTS 339

major NMC time of peak intensity (Pk) had the highest

correlation (r2 ⫽ 0.76) with moment arm length.
The NMCs of the antagonist deltoid muscle displayed more
variability in the timing of activity compared with the NMCs
within the two agonist muscles. Review of Table 1 shows that
the NMC with the longest antagonist moment arm (D3) tended
to be activated after movement onset (On) and significantly
(P ⬍ 0.05) later than NMC D7, which had an agonist moment
arm. NMCs with shorter antagonist moment arms (D1 and D5)
were activated together at a time closer to movement onset
(On). Across the breadth of the deltoid, NMC onset times (On)
were highly correlated (r2 ⫽ 0.83) to agonist or antagonist
moment arm length as shown by linear regression analysis.
While all deltoid NMCs reach peak intensity some 112 ms
after movement onset (On), they had significantly (P ⬍ 0.05)
different periods of activation (Dur), with NMC D7 (agonist
moment arm) displaying the longest (P ⬍ 0.05) activation time
(Dur) (Table 1). Interestingly, NMC D3 (longest antagonist
moment arm) was the least variable in its timing of activation
within the deltoid muscle. Compared with the two agonist
muscles, the NMCs of the deltoid reached peak intensity after
(P ⬍ 0.05) PkV and PkA but in closer proximity (P ⬍ 0.05) to

Downloaded from on October 22, 2014

PkDc (Table 2), confirming their predominantly antagonist
function.

Intensity of NMC Activation

There were no significant differences between the intensity
of activation (expressed as % of total muscle intensity) in all
NMCs of the pectoralis major and the latissimus dorsi except
NMC L1, which had a significantly (P ⬍ 0.05) lower (12%)
activation than more inferior NMCs within the latissimus dorsi
(Table 2). The deltoid NMC D3 (with longest antagonist

Fig. 3. Raw sEMG waveforms derived from NMCs within the pectoralis
major (P), deltoid (D), and latissimus dorsi (L). The timing (ms) of each
muscle burst (ON, OFF) was determined in relation to the onset of
displacement (OnD) of the adduction movement. Duration (Dur) was the
difference between ON and OFF. Movement time (MT) was the difference
between the beginning (OnD) and end (PkD) of the GHJ adduction move-
ment. A typical triphasic EMG pattern showing a first antagonist burst
(Ag1), an antagonist burst (Ant), and a second agonist burst (Ag2) was
most obvious between the primary agonist (e.g., L6, P6) and primary
antagonist (e.g., D3) NMCs.

While ANOVA results detected few significant timing dif-

ferences between individual NMCs (Table 2), linear regression
analysis suggested a sequential activation (On) of inferior to
superior NMCs in both the pectoralis major (r2 ⫽ 0.94) and the
latissimus dorsi (r2 ⫽ 0.93). The sequence of NMC activation
in the latissimus dorsi extended into the posterior-medial mus-
Fig. 4. Measures of angular displacement, angular velocity, and angular
cle fibers (NMC D7) of the deltoid (Fig. 6). Within the acceleration from 1 trial. The timing of each NMC (ON and OFF) was
latissimus dorsi NMC onset times (On) were highly correlated compared with the onset of displacement (OnD), peak displacement (PkD),
to moment arm length (r2 ⫽ 0.82), while in the pectoralis peak velocity (PkV), and peak acceleration (PkA) in each trial.

J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org

340 NEUROMUSCULAR COMPARTMENTS

and the time between peak NMC intensity (Pk) and PkV (r2 ⫽
0.42), PkA (r2 ⫽ 0.50), and PkDc (r2 ⫽ 0.38).

DISCUSSION

Overview of Aims
The first aim of the present study was to utilize a multichan-
nel sEMG technique to locate individual NMCs by detecting
heterogeneous sEMG activity across the surface of three su-
perficial muscles controlling movements of the GHJ. The
second aim of the study was to determine the function of each
identified NMC by correlating its timing and intensity of
activation to movement variables such as NMC moment arm
length and the timing of the movement’s peak displacement,
peak velocity, peak acceleration, and peak deceleration.

Use of Ballistic GHJ Movements

The ballistic (MT ⬍400 ms) GHJ adduction movement used
here to stress the GHJ muscle NMCs into robust activity was
utilized since it produced a complex series of agonist (Ag1)-
antagonist (Ant)-agonist (Ag2) muscle activations (Angel

Downloaded from on October 22, 2014

1977, 1981; Berardelli et al. 1984; Cheron et al. 2007; Gottlieb
et al. 1989), from which it was expected that heterogeneity in
the timing and intensity of individual NMCs might be detected.
This unique sequence of muscle activity, termed the “triphasic
EMG pattern” (MacKinnon and Rothwell 2000), is well estab-
lished across the joint as movement time decreases below 400
ms (Brown and Gilleard 1991). It was assumed that the
complexity of the ballistic movement studied here would
require each NMC to adopt a functional role consistent with its
unique moment arm—a functional role supported by a unique
pattern of motoneuron activation.
Table 1. Activation, time of peak intensity, and duration
of sEMG activity within each NMC

On Pk Dur
Mean SD Mean SD Mean SD
Fig. 5. Timing of NMC myoelectric activity in the pectoralis major (P), deltoid
(D), and latissimus dorsi (L). Vertical dotted lines indicate the onset (0 ms) and P6 ⫺118 20 ⫺10 27 185 33
end (354 ms) of the GHJ adduction movement as measured by an electrogo- P5 ⫺116 22 ⫺26 20 176 35
niometer. Heavy lines between diamonds indicate Ag1; light unbroken arrows P4 ⫺104 21 ⫺18 23 161 32
indicate Ag2; dotted lines between diamonds indicate Ant; triangles indicate P3 ⫺96 20 ⫺17 15 150 38
peak of myoelectric activity in that NMC. Ag2 activity was highly variable and P2 ⫺87‡ 17 2 34 163 62
is only shown if present in ⬎50% of trials examined. Group mean data from P1 ⫺88‡ 14 15 31 178 49
16 subjects are shown. D1 ⫺23*‡ 84 103‡ 51 205* 113
D3 40*‡ 27 114‡ 23 177* 41
moment arm) had a significantly (P ⬍ 0.05) higher intensity of D5 ⫺21*‡ 51 103‡ 62 277*‡ 85
activation (46%) compared with NMCs D1, P1, and P6. D7 ⫺86‡ 32 135‡ 108 390‡ 160
L1 ⫺108 23 67 80 257 80
Within the deltoid, the intensity of NMC activation (as % of L2 ⫺107 18 25 58 216 50
total muscle intensity) was moderately correlated to moment L3 ⫺120 17 5 71 215 80
arm (r2 ⫽ 0.62) although not related to the time between peak L5 ⫺130 18 ⫺6 82 193 27
NMC intensity (Pk) and PkV (r2 ⫽ 0.17), PkA (r2 ⫽ 0.15), or L6 ⫺135 27 19 94 201 50
PkDc (r2 ⫽ 0.16). In the agonist pectoralis major, the intensity
Values (in ms) are means and SD of group mean data from 16 subjects for
of NMC activation was not related to moment arm (r2 ⫽ 0.16), activation (On), time of peak intensity (Pk), and duration (Dur) of surface
although stronger relationships existed between moment arm electromyographic (sEMG) activity within each neuromuscular compartment
and the time between peak NMC intensity (Pk) and PkV (r2 ⫽ [NMC; pectoralis major (P), deltoid (D), and latissimus dorsi (L)]. For both On
0.71), PkA (r2 ⫽ 0.79), and PkDc (r2 ⫽ 0.71). Finally, the and Pk, values represent time either before (negative values) or after (positive
values) the movement onset, with duration representing the length of the first
intensity of NMC activation within the agonist latissimus dorsi EMG burst. Within-muscle analysis significant difference (P ⬍ 0.05): *to
muscle was not related to moment arm length (r2 ⫽ 0.22), NMC D7. Between-muscle analysis significant difference (P ⬍ 0.05): ‡to
while only weak relationships existed between moment arm NMC L6. Average movement time approximated 354 ms.

J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org

 NEUROMUSCULAR COMPARTMENTS 341

Table 2. NMC intensity time between peak intensity and the three kinematic variables of peak velocity, peak acceleration, and peak
deceleration

NMC Intensity, %total Peak Intensity to Peak Peak Intensity to Peak Peak Intensity to Peak
muscle intensity Velocity, ms Acceleration, ms Deceleration, ms
Mean SD Mean SD Mean SD Mean SD

P6 13‡ 6 ⫺101 29 ⫺8 26 ⫺203 33

P5 16 6 ⫺120 25 ⫺19 20 ⫺222 31
P4 20 4 ⫺112 22 ⫺12 21 ⫺214 26
P3 22 10 ⫺113 13 ⫺13 14 ⫺215 17
P2 16 6 ⫺92 32 9 35 ⫺195 31
P1 13‡ 6 ⫺79 31 20 28 ⫺181 35
D1 8*‡ 7 8‡ 49 108‡ 47 ⫺94‡ 52
D3 46*‡ 8 18‡ 16 119‡ 22 ⫺84‡ 12
D5 25 9 18‡ 40 119‡ 42 ⫺84‡ 40
D7 22 5 38‡ 102 139‡ 106 ⫺64‡ 98
L1 12† 6 ⫺25 79 75†‡ 77 ⫺127 80
L2 16 5 ⫺53 83 47 79 ⫺155 87
L3 26 11 ⫺79 77 21 77 ⫺182 76
L5 22 7 ⫺99 83 1 82 ⫺202 85
L6 24 6 ⫺76 97 17 93 ⫺178 103

Values are means and SD of group mean data from 16 subjects for NMC intensity (as % of total muscle intensity for each of the 3 muscles) and the time
between peak intensity (Pk) and the 3 kinematic variables of peak velocity, peak acceleration, and peak deceleration. Note that a negative value signifies that
the segment’s peak intensity occurred before the kinematic variable (peak velocity, peak acceleration, or peak deceleration) and a positive value indicates that

Downloaded from on October 22, 2014

the segment’s peak intensity occurred after the kinematic variable. Within-muscle analysis significant difference (P ⬍ 0.05): *to NMC D7, †to NMC L6.
Between-muscle analysis significant difference (P ⬍ 0.05): ‡to NMC L6. Average movement time approximated 354 ms.

Identification of Individual NMCs
Evidence was found that individual NMCs were present, in
a parallel spatial arrangement, across the breadth of the three
muscles investigated based upon the presence of heterogeneous
sEMG (Table 2). Of most importance were the results of the
linear regression analysis, which indicated that the onset (On)
of NMC activation was highly coordinated. As seen in Fig. 6,
a strong relationship existed between the onset (On) of NMCs,
in both the pectoralis major (r2 ⫽ 0.94) and the latissimus dorsi
(r2 ⫽ 0.93), and their anatomical position on the thorax. As
shown by the regression analyses the ballistic adduction GHJ
movement was initiated by two synchronous waves of mo-
toneuron activation, both anterior and posterior, that com-
menced at the lower NMCs of P6 and L6 and then proceeded
to the upper NMCs of P1 and L1. Furthermore, the posterior
wave of activation then continued onto the antagonist deltoid
muscle (D7), which seemingly acted more like a seventh NMC
of the latissimus dorsi. The caudal initiation of the movement
in both muscles coincides with the more “optimal” lines of
action apparent in the lower segments of both these agonist
muscles (Wickham et al. 2004b). Specifically, NMCs P6 and
L6 have the most vertically oriented lines of action, compared
with the abducted humerus, and would be well suited to
effectively adduct (rotate) the humerus, with less translation
(stabilization) typical of the more horizontal upper fibers, at
movement initiation. In support of this assumption were the
results of Brown et al. (2007), which showed that at 20° of
shoulder abduction, performing a similar adduction task, NMC
L3 was now the first to be activated. The early involvement of
NMC L3, and not NMC L6, can also be explained by a more
efficient rotatory component to its moment arm at this partic-
ular position of shoulder abduction.
In regard to the NMCs of the deltoid muscle, significant
(P ⬍ 0.05) differences in onsets (On) were much more appar-
ent than those seen within the two agonist muscles. These
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
NMC onset times also appeared to reflect the magnitude and
direction of their antagonist moment arms. Together, these data
suggest heterogeneous sEMG activity across all three investi-
gated muscles consistent with the presence of individual NMCs
(Staudenmann et al. 2009).
Function of Individual Muscles
The data suggested that each GHJ muscle had a unique role
in producing the ballistic shoulder adduction movement. The
strong correlation (r2 ⫽ 0.84) between agonist muscle moment
arm length and NMC onset time in the latissimus dorsi com-
bined with early NMC activation (Table 1), and a lack of a
relationship between its timing and intensity and PkV, PkA, or
PkDc, suggested that its primary role was to strongly initiate
the movement. In contrast, the pectoralis major, with its
slightly later onsets (Table 1) and stronger temporal relation-
ships with PkV, PkA, and PkDc, appeared better suited to
contribute to the control of the adduction movement after its
initiation. In contrast, most of the deltoid muscle, with its
antagonist moment arms (except NMC D7), acted to decelerate
the upper limb at the target angle.
Function of Individual NMCs
Prime movers. It has been shown, both within single muscles
(Wickham and Brown 1998) and across a muscle group
(Brown et al. 2007), that individual NMCs may be classified
according to their function during a motor task. Prime mover
NMCs have been shown to have agonist moment arms and
early activations to initiate and drive the motor task (Brown et
al. 2007; Wickham and Brown 1998) and, from the results of
this study (Table 3), to have a relatively high contribution to
the development of the adductor force impulse at the shoulder.
This assertion is consistent with the findings of other authors
who have reported that the first “peak” of muscle activation
was always within motor units close to the optimal line of
342 NEUROMUSCULAR COMPARTMENTS
Fig. 6. Onset (ON) of latissimus dorsi (L), deltoid (D), and pectoralis major (P)
NMCs during the ballistic shoulder adduction task. Note the synchronized
waves of agonist NMC activation (arrows) sweeping superiorly from the
inferior L6 and P6 NMCs toward and even into the deltoid (D7) (note the high
R2 values). The late onset of D3, after movement initiation (0 ms), along with
its large abduction moment arm (43 mm), suggests a primary antagonist role
for this NMC.
action (for the movement) within the muscle (Flanders and
Soechting 1990).
In this experiment, all NMCs of the latissimus dorsi (L1 to
L6) (Paton and Brown 1995), as well as those within the sternal
head of the pectoralis major (P3 to P6) (Paton and Brown
1994), were activated together and significantly (P ⬍ 0.05)
earlier than the other segments investigated. What functional
conclusions may be drawn from this finding? Reference to our
own segmental moment arm data for the end position of the
adduction movement (Table 3) shows that all NMCs within
both the latissimus dorsi and the pectoralis major were shoul-
der adductors. This finding was in agreement with classical
descriptions (Standring 2008) of whole muscle function and
our own observations of each NMC’s line of action. With their
collective times of activation being generally more than 100 ms
before the adduction movement commenced, more than one
electromechanical delay period (Gabriel and Boucher 1998;
Vos et al. 1991), these NMCs could be confirmed as the “prime
movers” for this motor task.
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
Further observations revealed that the prime mover seg-
ments (L1 to L6, P3 to P6) all reached their maximal myo-
electric activity at about the time the upper limb commenced
the adduction movement (OnD) and well before peak move-
ment velocity (PkV), a result in agreement with other authors’
findings (Brown and Cooke 1990). Therefore, these segments
were capable of both initiating the motor task and then con-
tributing to achieving the target movement angle (50° of
abduction).
In contrast, the duration (Dur) of NMC activation and the
relative intensity of that activation showed little variability
between the agonist segments. What was apparent from these
data was the remarkable consistency in the length of activation
(Dur), and the intensity of activation, once the NMC had been
called into action by the CNS. This was not a surprising result
given that there is a linear relationship between the duration of
Ag1 (first burst of the prime mover) and the duration of the
acceleratory phase of the movement (Cooke and Brown 1994).
Furthermore, it was also apparent that rapid movements are
controlled by periods of NMC activity that are shorter than the
duration of the movement (MT) itself, a finding consistent with
results from previous investigations of ballistic motor tasks
(Angel 1977, 1981; Gottlieb et al. 1989).
Downloaded from on October 22, 2014
Synergists. Those agonist muscle segments with smaller
agonist moment arms and/or lower relative contributions to the
generation of the adductor moment at the GHJ and activated
after the prime mover segments were termed “synergists”
(Wickham and Brown 1998). On the basis of our data it was
apparent that synergist NMCs were capable of both helping to
drive and guide the shoulder joint to the target movement
angle.
NMCs P1 and P2 (clavicular head), as well as NMC D7,
were assessed as having an agonist function for this motor task.
With moment arms for adduction (Table 3), their “synergist”
status was confirmed by their significantly (P ⬍ 0.05) later
(about ⫺86 ms before movement onset) activation compared
with the prime mover segments.
Although both shared a synergist function during the adduc-
tion motor task, in many ways the activities of NMCs D7 and
P1/P2 were dissimilar. Although having similar activation
times (⬃86 ms before movement onset), D7 was shown to
have a significantly (P ⬍ 0.05) longer period of activation
(Dur), a later peak (Pk) of myoelectric activity compared with
movement onset (OnD), PkV, and PkA. Although NMC D7
had an agonist moment arm for GHJ adduction (Wickham and
Brown 1998), it was small (⬃8 mm) compared with the other
agonist segments. Similarly, NMC D7’s line of action was
close to the neutral axis between adduction and abduction,
suggesting a greater joint stabilization, rather than joint rota-
tion, function. These factors may explain why the activation of
NMC D7 was, in part, more closely associated with the
antagonist NMCs (D1, D3, D5) of the deltoid than the synergist
NMCs within the pectoralis major. Clearly, the CNS has
considerable flexibility to vary the activation of synergist
muscle segments, wherever they occur across the joint, to best
meet the demands of the motor task.
Antagonists. From their significantly (P ⬍ 0.05) later times
of activation and their antagonist moment arms (Fig. 6; Table
3), it is possible to conclude that deltoid NMCs D1, D3, and D5
were all antagonist muscle segments. These NMCs provided an
important braking force to help the limb reach its final target
 NEUROMUSCULAR COMPARTMENTS 343

Table 3. Architectural measures and moment arm data gathered from one cadaver for each NMC of pectoralis major, deltoid, and
latissimus dorsi as determined from cadaveric dissection

MA at 50°
Abduction, mm
ADD ABD Volume, cm3 Muscle Length, cm CSA, cm2 Proportion Total Muscle CSA, % CSA Adjusted, cm2 Force, N

P6 44.3 26.6 20.3 1.3 11.0 1.7 49.4

P5 41.7 35.5 19.2 1.8 15.6 2.3 69.8
P4 45.3 33.9 19.4 1.7 14.7 2.2 65.8
P3 37.1 27.3 17.5 1.6 13.2 2.0 59.1
P2 31.5 50.8 17.1 3.0 25.1 3.8 112.6
P1 13.5 36.3 14.9 2.4 20.5 3.1 91.9
D1 6.2 55.3 15.6 3.5 15.2 4.2 126.1
D2 31.6 47.5 14.6 3.3 14.0 3.9 115.8
D3 42.6 62.1 12.4 5.0 21.5 6.0 178.6
D4 37.4 59.3 14.0 4.2 18.2 5.0 151.2
D5 27.0 42.6 12.1 3.5 15.1 4.2 125.6
D6 14.9 18.9 11.0 1.7 7.4 2.0 61.4
D7 4.4 26.7 13.4 2.0 8.5 2.4 71.0
L1 66.1 23.9 19.2 1.2 13.2 1.2 36.6
L2 65.8 42.9 23.1 1.9 19.7 1.8 54.6
L3 66.1 41.5 23.9 1.7 18.3 1.7 50.7
L4 60.1 44.4 25.6 1.7 18.3 1.7 50.7
L5 59.0 47.3 28.3 1.7 17.7 1.6 49.1
L6 57.8 26.7 22.1 1.2 12.7 1.2 35.2

Downloaded from on October 22, 2014

Values are architectural measures and moment arm data gathered from 1 cadaver for each NMC of pectoralis major (P), deltoid (D), and latissimus dorsi (L)
as determined from cadaveric dissection. The proportion of total muscle CSA% refers to cross-sectional area (CSA) with the NMC now expressed as a % of the
total muscle mass for that muscle. Adjusted CSA has been obtained by dividing a representative figure (multiple cadavers used) for the particular muscle by the
% of total muscle mass obtained in this study, with the size of the cadaver used in this study approximating values obtained by other authors (Veeger et al. 1991).
The force data were obtained by multiplying the adjusted CSA (cm2) by a constant (30 N) (Enoka 1994). Moment arm (MA) data, as derived from Wickham
et al. (2004b), determined the length (mm) of the NMC abduction (ABD) or adduction (ADD) moment arm at 50° of shoulder abduction in the coronal plane.

position (Berardelli et al. 1996; Brown and Cooke 1990;
MacKinnon and Rothwell 2000). However, the significant
(P ⬍ 0.05) difference in onset times between NMCs D1 and
D5, as opposed to NMC D3, suggested a contrast in function.
It is clear that NMC D3 played the “primary antagonist” role
(Wickham et al. 2004a) in that its late activation and large
antagonist moment arm provided the major braking force for
the movement, the true “Ant” component of the triphasic EMG
pattern (MacKinnon and Rothwell 2000). Its lower standard
deviations (Table 1) also indicate a more defined role with less
flexibility in timing, compared with other antagonist segments,
if it is to play this important role. Differences in reciprocal
innervation among the antagonist NMCs of the deltoid suggest
that certain NMCs of the antagonist deltoid (ones with optimal
lines of action and large moment arms) may receive “stronger”
neural inhibition that delayed the onset of the primary antag-
onist NMC of D3 (Wickham et al. 2004a). The smaller antag-
onist moment arms, and earlier onsets for NMCs D1 and D5
(Table 1), indicated a “secondary antagonist” role to help guide
the limb during the adduction motion. In this way, NMCs D1
and D5 appeared to be working as “synergistic antagonists” to
help D3 control the adduction motor task.
In contrast to onset time (On), the antagonist NMCs
within the deltoid (D1, D3, D5) all reached maximal activity
(Pk) significantly (P ⬍ 0.05) later, or ⬎100 ms after move-
ment onset, and within one electromechanical delay period
of peak movement deceleration. This finding was consistent
with other results (Karst and Hasan 1987; MacKinnon and
Rothwell 2000) that have suggested that the antagonist
(Ant) muscle burst is always activated after the initial ago-
nist (Ag1).
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
Roles in accessory movements. It is also interesting to
speculate upon the role of the individual NMCs in controlling
the small accessory movements (e.g., gliding, translations) that
occur whenever the shoulder joint rotates. For example, shoul-
der abduction and adduction are accompanied by lateral and
medial humeral rotations that in part help to minimize com-
pression under the coracohumeral ligament (Yanai et al. 2006).
On the basis of their moment arms, NMCs D4 –D7 (lateral
rotators) and D1, D2, and those within the latissimus dorsi and
pectoralis major (medial rotators) appear well suited to assist in
these complementary motions. Likewise, the posterior glide of
the humeral head during shoulder extension/lateral rotation
would be promoted by NMCs D4 –D6, which all combine
moment arms for shoulder extension with the capacity to draw
the humerus posteriorly. It seems probable that the presence of
independent NMCs, within shoulder muscles like the deltoid,
each with unique moment arms, provide the necessary motive
drive to accomplish both the pure and accessory motions of the
shoulder joint.
Clearly, our results show that NMCs within three super-
ficial shoulder joint muscles may be functionally classified
as prime mover, synergist, primary antagonist, and second-
ary antagonist segments on the basis of their timing of
activation, moment arms, and relative contribution to the
production of joint torque around the joint. It was apparent
that the “function” of each muscle NMC determined its
“pattern” of activation and that each NMC’s function was
determined by its moment arm, its spatial location on the
trunk and shoulder girdle, and its ability to contribute to, or
resist, the adductor force moment around the shoulder joint.
The results of this study suggest that the CNS controls
344 NEUROMUSCULAR COMPARTMENTS
individual NMCs, rather than whole muscles. Whether that
level of fine control is present in all human skeletal muscles
and how that level of fine muscle control is controlled (e.g.,
through the intensity and spatial distribution of local muscle
spindle populations) are questions that remain unanswered
at this time.
Limitations
Several limitations in the experimental design need to be
recognized and discussed. First, our analysis of NMC function
relied upon moment arm data that were available only for the
final position of the shoulder (50° of abduction) (Wickham et
al. 2004b) because of technical difficulties related to joint
integrity at higher angles of shoulder abduction.
We do not believe that this limitation would influence the
primary outcomes detailed here as, on the basis of previous
work at 20° and 50° of shoulder abduction (Wickham 2002),
only moment arm length, rather than direction, changes at
greater angles of shoulder elevation.
Second, while the utilization of a ballistic task was generally
advantageous, the compressed nature of the subsequent NMC
sEMG waveforms produces dynamic movements shorter than
400 ms (Brown and Gilleard 1991; Brown and Cooke 1981).
As a result, variations in the timing of adjacent NMC activity
were often only in the order of 10 –20 ms, which made
ANOVA analyses problematic given the inherent variability of
sEMG. Therefore, the selection of the ballistic task itself
somewhat negated our ability to find significant (P ⬍ 0.05)
differences between individual NMCs where significant differ-
ences might have been present if movement time had been
somewhat longer. However, we believe that the regression
analyses more than compensated for this limitation.
Finally, the problem of cross talk between adjacent bipolar
electrode pairs needs to be addressed. If detectable levels of
cross talk were present, the ability to differentiate between the
activities of adjacent NMCs would be compromised. Consid-
erable effort was made to minimize the effects of electrode
cross talk. The bipolar electrodes were specifically designed to
have small active plates (1.6 mm in diameter) and interelec-
trode distances (6.5 mm). This electrode design, using the cross
talk “rule of thumb” (Basmajian and DeLuca 1985), would
suggest meaningful pickup of muscle fiber activity only within
⬃6.5 mm of each bipolar electrode pair. Adjacent bipolar
electrode pairs were spaced ⬎1 cm from each other so as to
minimize, as much as practicable, electrode cross talk. In
addition, regardless of possible electrode cross talk, the results
presented here provide strong evidence of the sequential acti-
vation of adjacent NMCs to control the adduction movement.
Conclusion
The results of this investigation have suggested that the
pectoralis major, deltoid, and latissimus dorsi are composed of
NMCs that have a parallel spatial arrangement as evidenced by
heterogeneity of sEMG activity across the breadth of these
muscles. The NMCs controlling a ballistic GHJ adduction
movement appear to be highly coordinated in both intensity
and timing of activation and may be functionally classified as
prime mover, synergist, primary antagonist, or secondary an-
tagonist NMCs based on their moment arms and periods of
activation.
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
Our findings are important in that they show, for the first
time, how the CNS controls multiple NMCs within a group of
GHJ muscles, how they are temporally activated, and how their
intensity of activation is modulated in relation to their moment
arms and consequent functional role in the movement.
ACKNOWLEDGMENTS
The authors thank Prof. Brian Key for thoughtful suggestions regarding the
manuscript. The authors also thank Monty Brown and Annette Wickham for
assistance with preparation of the illustrations.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
AUTHOR CONTRIBUTIONS
Author contributions: J.B.W. and M.J.B. conception and design of research;
J.B.W. performed experiments; J.B.W. analyzed data; J.B.W. and M.J.B.
interpreted results of experiments; J.B.W. prepared figures; J.B.W. and M.J.B.
drafted manuscript; J.B.W. and M.J.B. edited and revised manuscript; J.B.W.
and M.J.B. approved final version of manuscript.
REFERENCES
Downloaded from on October 22, 2014
Ackland D, Pak P, Richardson M, Pandy M. Moment arms of the muscles
crossing the anatomical shoulder. J Anat 213: 383–390, 2008.
Angel RW. Antagonist muscle activity during rapid arm movements: central
versus proprioceptive influences. J Neurol Neurosurg Psychiatry 40: 683–
686, 1977.
Angel RW. Electromyographic patterns during ballistic movements in normals
and hemiplegic patients. In: Progress in Clinical Neurophysiology, edited by
Desmedt JE. Basel: Karger, 1981, p. 347–357.
Angel RW. Myoelectric patterns associated with ballistic movement: effect of
unexpected changes in load. J Hum Mov Stud 1: 96 –103, 1975.
Basmajian JV, DeLuca C. Muscles Alive; Their Functions Revealed By
Electromyography. Baltimore: Williams and Wilkins, 1985.
Berardelli A, Hallett M, Rothwell JC, Agostino R, Manfredi M, Thompson
PD, Marsden CD. Single-joint rapid arm movements in normal subjects
and in patients with motor disorders. Brain 119: 661– 674, 1996.
Berardelli A, Rothwell JC, Day BL, Kachi T, Marsden CD. Duration of the
first agonist EMG burst in ballistic arm movements. Brain Res 304:
183–187, 1984.
Brown J, Wickham J, McAndrew D, Huang XF. Muscles within muscles:
coordination of 19 muscle segments within three shoulder muscles during
isometric motor tasks. J Electrocardiol 17: 57–73, 2007.
Brown JMM, Gilleard W. Transition from slow to ballistic movement:
development of triphasic electromyogram patterns. Eur J Appl Physiol 63:
381–386, 1991.
Brown SH, Cooke JD. Movement-related phasic muscle activation. I. Rela-
tions with temporal profile of movement. J Neurophysiol 63: 455– 464,
1990.
Brown SHC, Cooke JD. Amplitude and instruction dependent modulation of
movement-related electromyogram activity in humans. J Physiol 316: 97–
107, 1981.
Butler TJ, Kilbreath SL, Gorman RB, and Gandevia SC. Selective recruit-
ment of single motor units in human flexor digitorum superficialis muscle
during flexion of individual fingers. J Physiol 567: 301–309, 2005.
Buys EJ, Lemon RN, Mantel WH, Muir RB. Selective facilitation of
different hand muscles by single corticospinal neurones in the conscious
monkey. J Physiol 381: 529 –549, 1986.
Cheron G, Cebolla A, Bengoetxea A, Leurs F, Dan B. Recognition of the
physiological actions of the triphasic EMG pattern by a dynamic recurrent
neural network. Neurosci Lett 414: 192–196, 2007.
Cooke JD, Brown SH. Movement-related phasic muscle activation. III. The
duration of phasic agonist activity initiating movement. Exp Brain Res 99:
473– 482, 1994.
English AW, Weeks OI. Compartmentalisation of single muscle units in cat
lateral gastrocnemius. Exp Brain Res 56: 361–368, 1984.
English AW, Weeks OI. Electromyographic cross-talk within a compartmen-
talised muscle of the cat. J Physiol 416: 327–336, 1989.
 NEUROMUSCULAR COMPARTMENTS
English AW, Wolf SL, Segal RL. Compartmentalisation of muscles and their
motor nuclei: the partitioning hypothesis. Phys Ther 73: 857– 867, 1993.
Enoka R. Neuromechanical Basis of Kinesiology. Champaign, IL: Human
Kinetics, 1994, p. 466.
Ettema G, Styles G, Kippers V. The moment arms of 23 muscle segments of
the upper limb with varying elbow and forearm positions: implications for
motor control. Hum Mov Sci 17: 201–220, 1998.
Flanders M, Soechting JF. Arm muscle activation for static forces in
three-dimensional space. J Neurophysiol 64: 1818 –1837, 1990.
Gabriel DA, Boucher JP. Effects of repetitive dynamic contractions upon
electromechanical delay. Eur J Appl Physiol Occup Physiol 79: 37– 40,
1998.
Gottlieb GL, Corcos DM, Agarwal GC. Strategies for the control of
voluntary movements with one mechanical degree of freedom. Behav Brain
Sci 12: 189 –250, 1989.
Hogfors C, Sigholm G, Herberts P. Biomechanical model of the human
shoulder. I. Elements. J Biomech 20: 157–166, 1987.
Johnson G, Bogduk N, Nowitzke A, House D. Anatomy and actions of the
trapezius muscle. Clin Biomech (Bristol, Avon) 9: 44 –50, 1994.
Karst GM, Hasan Z. Antagonist muscle activity during human forearm
movements under varying kinematic and loading conditions. Exp Brain Res
67: 391– 401, 1987.
Korfage J, Van Eijden T. Regional differences in fibre type composition in
the human temporalis muscle. J Anat 194: 355–362, 1999.
Lemon RN, Muir R. Cortical addresses of distal muscles: a study in the
conscious monkey using the spike-triggered averaging techniques. Exp
Brain Res Suppl 7: 230 –238, 1983.
Lexell J, Jarvis J, Currie J, Downham D, Salmons S. Fibre type composi-
tion in rat tibialis anterior and extensor digitorum longus muscles. J Anat
185: 95–101, 1994.
Lucas-Osma A, Collazos-Castro J. Compartmentalization in the triceps
brachii motoneuron nucleus and its relation to muscle architecture. J Comp
Neurol 516: 226 –239, 2009.
MacKinnon CD, Rothwell JC. Time-varying changes in corticospinal excit-
ability accompanying the triphasic EMG pattern in humans. J Physiol 528:
633– 645, 2000.
Maier M, Bennett K, Hepp-Reymond M, Lemon R. Contribution of the
monkey corticomotoneuronal system to the control of force in precision
grip. J Neurophysiol 69: 772–785, 1993.
Mu L, Sanders I. Neuromuscular compartments and fiber-type regionalization
in the human inferior pharyngeal constrictor muscle. Anat Rec 264: 367–
377, 2001.
J Neurophysiol • doi:10.1152/jn.00049.2011 • www.jn.org
345
Paton ME, Brown JMM. An electromyographic analysis of functional
differentiation in human pectoralis major muscle. J Electromyogr Kinesiol
4: 161–169, 1994.
Paton ME, Brown JMM. Functional differentiation within latissimus dorsi.
Electromyogr Clin Neurophysiol 4: 161–169, 1995.
Sjostrom M, Lexell J, Downham D. Differences in fiber number and fiber
type proportion within fascicles. A quantitative morphological study of
whole vastus lateralis muscle from childhood to old age. Anat Rec 234:
183–189, 1992.
Soderberg GL, Dostal WF. Electromyographic study of three parts of the
gluteus medius muscle during functional activities. Phys Ther 58: 691– 696,
1978.
Standring S. Gray’s Anatomy: The Anatomical Basis of Clinical Practice.
New York: Churchill Livingstone Elsevier, 2008, p. 1576.
Staudenmann D, Kingma I, Daffertshofer A, Stegeman D, van Dieen J.
Heterogeneity of muscle activation in relation to force direction: a multi-
channel surface electromyography study on the triceps surae muscle. J
Electromyogr Kinesiol 19: 882– 895, 2009.
Veeger HEV, VanDerHelm FCT, VanDerWoude LHV, Pronk GM,
Rozendal RH. Inertia and muscle contraction parameters for musculoskel-
etal modelling of the shoulder mechanism. J Biomech 24: 615– 629, 1991.
Vos EJ, Harlaar J, van Ingen Schenau GJ. Electromechanical delay during
knee extensor contractions. Med Sci Sports Exerc 23: 1187–1193, 1991.
Wickham J. Muscles Within Muscles: The Neuromotor Activation Patterns of
Intramuscular Segments (PhD Thesis). Wollongong, Australia: University
of Wollongong, Department of Biomedical Science, 2002, p. 320.
Wickham J, Brown J, Green R, McAndrew D. Muscles within muscles: the
classical triphasic EMG burst and its applicability to segments of large
Downloaded from on October 22, 2014
radiate agonist/antagonist muscles. J Musculoskelet Res 8: 107–117, 2004a.
Wickham J, Brown J, McAndrew D. Muscles within muscles: anatomical
and functional segmentation of selected shoulder joint musculature. J
Musculoskelet Res 8: 57–73, 2004b.
Wickham JB, Brown JMM. Muscles within muscles: the neuromotor control
of intra-muscular segments. Eur J Appl Physiol 78: 219 –225, 1998.
Widmer C, Carrasco D, English A. Differential activation of neuromuscular
compartments in the rabbit masseter muscle during different oral behav-
iours. Exp Brain Res 150: 297–307, 2003.
Widmer CG, Klugman D, English AW. Anatomical partitioning and nerve
branching patterns in the adult rabbit masseter. Acta Anat (Basel) 159:
222–232, 1997.
Yanai T, Fuss F, Fukanaga T. In vivo measurement of subacromial impinge-
ment: substantial pressure develops in abduction with large internal rotation.
Clin Biomech (Bristol, Avon) 21: 692–700, 2006.