Journal of Electromyography and Kinesiology 17 (2007) 410–419

www.elsevier.com/locate/jelekin

The effect of handedness on electromyographic activity

of human shoulder muscles during movement
Louise Pyndt Diederichsen a,*, Jesper Nørregaard b, Poul Dyhre-Poulsen c, Annika Winther a,
Goran Tufekovic a, Thomas Bandholm a, Lars Raundal Rasmussen a, Michael Krogsgaard d
a
Institute of Sports Medicine, Bispebjerg Hospital, Denmark
b
Department of Rheumatology H, Bispebjerg Hospital, Denmark
c
Department for Medical Physiology, Panum Institute, University of Copenhagen, Denmark
d
Department of Orthopaedic Surgery M, Bispebjerg Hospital, Denmark

Received 1 August 2005; received in revised form 20 January 2006; accepted 13 March 2006

Abstract

The aim of the study was to investigate whether there was a difference in the electromyographic (EMG) activity of human shoulder
muscles between the dominant and nondominant side during movement and to explore whether a possible side-difference depends on the
specific task. We compared the EMG activity with surface and intramuscular electrodes in eight muscles of both shoulders in 20 healthy
subjects whose hand preference was evaluated using a standard questionnaire. EMG signals were recorded during abduction and external
rotation. During abduction, the normalized EMG activity was significantly smaller on the dominant side compared to the nondominant
side for all the muscles except for infraspinatus and lower trapezius (P 6 0.002). In contrast, during external rotation, higher EMG activ-
ity was seen in the supraspinatus, infraspinatus, lower and upper trapezius and latissimus muscles of the dominant side (P 6 0.01). We
demonstrated a side-difference in shoulder muscle activity, which was dependent on the type of motion carried out, suggesting a qual-
itative difference in the activation of muscles during the two types of movement. Dynamic abduction has the characteristics of a dom-
inant arm task (i.e., task performed almost exclusively by the dominant arm) and reduced muscle activity for the dominant side during
abduction indicates a dominance-related advantage in arm dynamics.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Electromyography; Shoulder joint; Handedness

1. Introduction 1998) and authors have reported functional asymmetries

Handedness is a well-known behavioural phenomenon,
defined as the tendency to prefer the use of a consistent
hand in performing selected tasks.
Physiological and anatomical asymmetries at different
levels of the central nervous system controlling the upper
extremity have been established. Handedness-related
asymmetries exist in the motor cortex (Volkmann et al.,
*
Corresponding author. Present address: Department of Medicine,
Svendborg Sygehus, Valdemarsgade 53, DK-5700 Svendborg, Denmark
Or Lodsvej 25A, DK-5270 Odense N, Denmark. Tel.: +45 65 41 26 53;
fax: +45 66 11 74 18.
E-mail address: louise.diederichsen@dadlnet.dk (L.P. Diederichsen).
in some cervical spinal pathways, including the corticospi-
nal tract (Semmler and Nordstrom, 1998; Tan, 1989a; Tan,
1989b). Also, differences in peripheral nervous pathways
have been found such as higher sensory detection thresh-
olds and higher conducting velocities for the motor nerve
in the dominant arm (Friedli et al., 1987; Sathiamoorthy
and Sathiamoorthy, 1990), although results have not been
consistent (Tan, 1985).
In addition to asymmetries in the nervous system, side-
differences exist in the muscles. Long-term preferential use
of muscles of the dominant side of the body may result in
changes of muscle fiber composition with a higher preva-
lence of slow twitch type I fibers (Fugl-Meyer et al.,
1982). The shift towards slow twitch fibers is associated

1050-6411/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jelekin.2006.03.004
 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
with changes in motor unit control properties, which
results in reduced firing rates of motor units on the domi-
nant side (Adam et al., 1998, 1989). In line with these
results, some authors have demonstrated differences in
EMG manifestations of muscle fatigue in the upper trape-
zius muscle with the dominant side showing less fatigue
(Farina et al., 2003).
Although several authors have assessed EMG activity of
the shoulder muscles with different approaches in normal
subjects, effects of side-dominance have remained largely
undetermined (Arwert et al., 1997; Bagg and Forrest,
1986; Inman et al., 1944; Jobe et al., 1983, 1984; Kronberg
et al., 1990; Nuber et al., 1986; Pearl et al., 1992; Ryu et al.,
1988; Saha, 1956; Shevlin et al., 1969). To our knowledge,
the only study on EMG amplitude and side-dominance in
the upper extremities is by Bagesteiro and Sainburg in
which they examined kinematics and EMG from the upper
arm during reaching movements (Bagesteiro and Sainburg,
2002). With equal kinematics, they found side-differences in
torque patterns and EMG profiles. However, the study was
restricted to assessment from just two shoulder muscles.
From a clinical point of view, knowledge of possible
dominant-related differences in EMG activity of shoulder
muscles from healthy subjects could be useful in studies
on motor patterns in subjects with and without shoulder
problems. In some previous studies, results have been
pooled from dominant and nondominant side and in others
Fig. 1. Experimental set-up. Subjects make A: dynamic abduction (0–110) or B: dynamic external rotation ( 80 to 40) in a custom-built shoulder
machine adjustable to different loads and designed to record force and position during motion.
411
information about hand preference has not been included
at all. Some of the diversity in the results of EMG studies
on shoulder patients may be due to the lack of attention to
handedness (Alpert et al., 2000; Ludewig and Cook, 2000;
Michaud et al., 1987; Peat and Grahame, 1977; Reddy
et al., 2000).
In accordance with the evidence of physiological and
anatomical asymmetries associated with side-dominance,
the purpose of the present study was to investigate whether
there was a handedness-related difference in the EMG
activity of different shoulder muscles between the dominant
and nondominant side during different standardized move-
ments in healthy subjects. Moreover, we aimed to explore
whether a possible side-difference may depend on the spe-
cific task.
2. Materials and methods
2.1. Subjects
Twenty healthy subjects without any reported shoulder
problems participated in the study. There were 3 females
and 17 males, aged 23–57 years (median = 37 years). Sev-
enteen out of 20 were right-hand dominant. The hand pref-
erence of each subject was determined using a modified
Edinburgh handedness inventory (Oldfield, 1971), resulting
in a laterality quotient (LQ) ranging from 100 (complete
412 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
left-handedness) to +100 (complete right-handedness). The
mean LQ was 72.3 for the right-handed subjects and 58.5
for the left-handed subjects. Both shoulders were studied.
The study was conducted in accordance with the Decla-
ration of Helsinki. The Committee on Ethics in Science in
Copenhagen approved the study and informed consent was
obtained from each subject.
2.2. Experimental set-up and mechanical recordings
Fig. 1 illustrates the experimental set-up. All subjects
were tested in a custom-built shoulder machine adjustable
to different loads and designed to record force and limb
position during motion. The data were collected with the
subject seated in the upright position against a fixed back
and headrest that did not interfere with scapular motion.
An adjustable support on each side of the body fixed the
hips to prevent changes in posture during the test. The sub-
jects pushed against a lever arm with a single axis of rota-
tion that was adjusted to the rotation axis of the shoulder
joint during abduction and external rotation, respectively.
During abduction the upper arm was held against a sup-
port on the lever arm (Fig. 1A). During external rotation
the forearm was supported by an armrest and held against
a support on the lever arm just proximal to the wrist
(Fig. 1B). All measurements were completed for one shoul-
der first before testing the other shoulder. The order of
shoulder tested was randomised. The target load applied
during movement was adjusted before testing. The force
and position sensors were placed at the rotation centre of
the lever arm. The digital force (N) and position () values
were visualized on a small display at the top of the shoulder
machine.
2.3. Electromyographic recordings
EMG was recorded with bipolar Ag–AgCl surface elec-
trodes (M-00-S, Medicotest, Denmark, spacing 20 mm)
from the superficially located shoulder muscles (serratus
anterior, the anterior and middle part of the deltoid, the
lower and upper part of trapezius and latissimus dorsi).
The electrodes were placed parallel to the muscle fibers on
shaved, abraded ethanol-cleaned skin. For intramuscular
EMG recordings, bipolar wire electrodes (316SS7/44T,
MEDWIRE) were inserted with a hypodermic needle
(21G · 3 1/8) into the infraspinatus and the supraspinatus
muscles during sterile conditions, with an interelectrode dis-
tance of about 20 mm. The electrodes were 200 lm multi-
stranded, Teflon-insulated stainless steel wires with
exposed tips of 3 mm. The proximal tip of the electrode
was formed as a hook to prevent displacement of the elec-
trode within the structure. An indifferent electrode was
placed on the skin over the seventh cervical vertebra. The
surface and the wire electrodes were placed according to
Delagi et al. (1981) and van der Helm (1994). Specifically,
the electrode placement for upper trapezius was 2 cm lateral
to the midpoint of the lead line between the angle of the
acromion and vertebra C7 and for lower trapezius, on a line
perpendicular to the vertebral column at the level of the
inferior angle of the scapula, two fingerbreadths from the
spinous process of that vertebra. The electrode placement
for serratus anterior was anterior to the latissimus dorsi
and below the pectoralis major muscle, when the arm was
flexed. Correct electrode placement was confirmed by visual
monitoring of the EMG signals on a computer screen dur-
ing specific muscle testing (Kelly et al., 1996; Kendall and
Kendall, 1949; Schuldt and Harms-Ringdahl, 1988).
The EMG signals were preamplified by small amplifiers
(fixed gain = 50) connected directly to the electrodes and
the signals were lead to a custom-built amplifier (total
amplification 2000, frequency 10–450 Hz). The EMG, force
and position signals were A/D converted (12 bits accu-
racy), sampled at 1000 Hz (Labview 5.0, National Instru-
ments, Austin, TX) and stored on a personal computer
for later analysis. During recording, the EMG, force and
position signals were displayed simultaneously on the com-
puter screen to verify the digitized recordings.
2.4. Experimental protocol
After familiarization with the experimental set-up and
protocol, the subjects performed three 4 s maximal, volun-
tary isometric contractions (MVC), in four different test
positions: 30 of abduction and 30 of adduction in the
scapular plane, internal and external rotation with the
shoulder in the neutral position and with the elbow in
90 of flexion in all four positions. Before the MVC record-
ings, subjects were instructed to contract as forcefully as
possible and to reach maximal force as rapidly as possible.
During performance of the MVC strong verbal encourage-
ment was given. Each trial was separated with about 1 min
to avoid fatigue.
Further more the experimental protocol consisted of
motions in two directions: (1) abduction in the scapular
plane (0–110) and (2) external rotation with the shoulder
in the neutral position ( 80 to 40). Three to five consecu-
tive trials were performed in each motion at 10% of the
MVC, calculated from the maximum force achieved in
the three trials for the given direction. Each subject was
carefully trained to move their arm through the full range
of motion at a rate equal to two beats of a metronome
set at 92 bpm during abduction and 96 bpm during exter-
nal rotation, to control the speed of motion.
2.5. Signal processing and statistical analysis
The EMG signals were rectified, averaged and normal-
ized to the max EMG (EMGmax) defined as the highest
mean EMG voltage recorded over 0.5 s of the 4 s interval
of the MVCs across all four test positions. Movements per-
formed smoothly and with full range of motion were
included in the analysis. Because of slight individual varia-
tions in the range of motion, the normalized EMG for each
muscle was calculated for every 15 percentiles of the move-
 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419 413

Table 1 abduction 3 ± 2% and 3 ± 2%, adduction 2 ± 1% and

The mean maximal muscle force (MVCs) of 20 subjects expressed in
Newton (mean ± SD)
Dominant shoulder Nondominant shoulder
Abduction 303 ± 50 283 ± 61a
Adduction 317 ± 38 308 ± 40
External rotation 140 ± 36 144 ± 38
Internal rotation 223 ± 43 238 ± 48a
a
Statistically significant.
ment cycle (i.e., 0–15 percentiles = phase 1, 16–30 percen-
tiles = phase 2, etc.) and then the data for each phase of
the movement cycle from each muscle were averaged across
all subjects. The first phase of motion (0–15 percentiles)
indicated the beginning of the movement (0). The last
phase of motion (91–100 percentiles) was excluded from
the calculations since the subjects in that phase gradually
reversed the direction of movement, bringing the arm back
to the start position. Data for each phase and group did
not diverge from a normal distribution; therefore EMG
data from dominant and nondominant side were compared
using three-way ANOVA for the whole movement cycle
with factors of side (dominant or nondominant), phase of
movement and person. Side-difference in EMG for each
phase within the cycle was analysed using a paired t-test.
MVCs from the dominant and nondominant side were also
compared for the different directions using a paired t-test.
Individual coefficients of variation (CV, in %) for force
were calculated for the three MVCs in each of the four
positions. A significance level of 0.05 was considered as sig-
nificant for all tests, however corrected for number of tests
(Bonferroni) when side-differences in EMG for each phase
within the cycle were analysed.
3. Results
3.1. Muscle strength
The average isometric MVC data are given in Table 1.
Comparison of MVCs for the two sides revealed a signifi-
cant increased strength on the dominant side for abduction
(P = 0.004) and decreased strength for internal rotation
(P = 0.05). There was no significant difference in strength
between the dominant and nondominant side for adduc-
tion (P = 0.2) and external rotation (P = 0.5). CVs for
the three MVCs in each of the four different positions were:
3 ± 2%, external rotation 4 ± 2% and 4 ± 2% and internal
rotation 3 ± 2% and 3 ± 2%, for the dominant and the
nondominant side, respectively (mean ± SD).
3.2. EMG amplitude
3.2.1. EMG during MVC
EMG obtained during the MVCs was examined across
all four positions and EMGmax was found for every muscle
for each subject and used in the normalization process of
the dynamic EMG. The averaged EMGmax of the muscles
for all subjects across all four positions are shown in Table
2. There was no significant side-difference in EMGmax.
3.2.2. EMG during movement
In general, there was a side-difference in the EMG
amplitude of the shoulder muscles between the dominant
and the nondominant side during both abduction and
external rotation.
3.3. Abduction
The averaged EMG data for the eight shoulder muscles
during abduction across all subjects are illustrated in Fig. 2
(mean ± SE). All the shoulder muscles were active on both
sides during abduction in the scapular plane. Particularly
the anterior deltoid and serratus anterior were highly acti-
vated with EMG amplitudes greater than the peak EMG
for the same muscles obtained during the MVC trials. Also,
the middle part of deltoid and upper trapezius were sub-
stantially active during arm elevation. A uniform shape
of the activity curves for the two sides, with gradually
increased EMG activity during arm elevation, was
observed for all of the muscles but the supraspinatus and
infraspinatus. These two muscles showed increased EMG
activity in the initial part of the abduction followed by a
continuous decline on both sides.
The EMG activity during the whole arc of motion was
significantly smaller in the dominant arm compared to
the nondominant arm for all the shoulder muscles except
infraspinatus and lower trapezius (supraspinatus
P < 0.001, infraspinatus P = 0.9, lower trapezius P = 0.2,
serratus anterior P < 0.001, anterior deltoid P < 0.001,
middle deltoid P = 0.002, latissimus P < 0.001 and upper
trapezius P = 0.001). When each phase (i.e., 0–15 percen-

Table 2
Mean rectified EMG (EMGmax) for 8 shoulder muscles obtained during MVCs across four positions (abduction, adduction, external and internal rotation)
for 20 subjects expressed in lV (mean ± SD)
Supraspinatus Infraspinatus Lower Serratus Anterior Middle Latissimus Upper trapezius
trapezius anterior deltoid deltoid dorsi
Dominant 322 ± 108 353 ± 171 225 ± 124 151 ± 108 281 ± 139 228 ± 77 136 ± 64 274 ± 126
shoulder
Nondominant 337 ± 153 333 ± 97 204 ± 127 164 ± 123 253 ± 116 203 ± 106 149 ± 88 276 ± 142
shoulder
No significant differences between dominant and nondominant shoulder (P < 0.05).
414 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419

Supraspinatus Infraspinatus
70 40
P < 0.001 P = 0.9

EMG amplitude (%MVC)

EMG amplitude (%MVC)
60 35

50 30

40 25

30 20

Dominant
20 15
Nondominant

10 10

Lower trapezius Serratus anterior

60 P < 0.001
P = 0.2
120

EMG amplitude (%MVC)

EMG amplitude (%MVC)

50
100
40
80
30
60

20 40

10 20

0 0

Anterior deltoid Middle deltoid

120 P < 0.001 P = 0.002
EMG amplitude (%MVC)

100
EMG amplitude (%MVC)

100
80
80
60
60

40
40

20 20

0 0

Latissimus dorsi Upper trapezius

50
P < 0.001 P = 0.001
EMG amplitude (%MVC)
EMG amplitude (%MVC)

80
40
60
30

40
20

10 20

0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Position (phase) Position (phase)

Fig. 2. The averaged normalized EMG activity (mean ± SE) for the 8 shoulder muscles during abduction in the scapular plane (n = 20). The position is
indicated for every 15 percentile of the movement cycle (15 percentiles = 1 phase), the first phase of motion (0–15 percentiles) presents the beginning of the
movement, with the arm at the side of the body as start position (0). When analysing the whole arc of motion, there was significant difference (P < 0.05) in
EMG activity between the dominant and the nondominant side for all the analysed muscles but two (infraspinatus and lower trapezius).

tiles, 16–30 percentiles, etc.) of the abduction cycle was 3.4. External rotation
analysed, the EMG activity in the anterior deltoid muscle
was significant smaller during the first phase (P = 0.003) EMG muscle activity during external rotation is pre-
of abduction for the dominant arm, and in the upper trape- sented in Fig. 3. Of all the muscles tested, particularly the
zius muscle there was less activation during the second infraspinatus muscle showed significant activity during
phase (P = 0.008) of motion. external rotation, but also the lower trapezius and supra-
 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419 415

Supraspinatus Infraspinatus
50 P = 0.01 80 P < 0.001

EMG amplitude (%MVC)

EMG amplitude (%MVC)

40
60

30
40
20

Dominant 20
10
Nondominant

0 0

Lower trapezius Serratus anterior

60 P < 0.001 25 P = 0.5

EMG amplitude (%MVC)

EMG amplitude (%MVC)

50 20

40
15
30
10
20

5
10

0 0

Anterior deltoid Middle deltoid

20 30
EMG amplitude (%MVC)

P = 0.5 P = 0.9
EMG amplitude (%MVC)

25
15
20

10 15

10
5
5

0 0

Latissimus dorsi Upper trapezius

30 20
P < 0.001 P < 0.001
EMG amplitude (%MVC)

EMG amplitude (%MVC)

25
15
20

15 10

10
5
5

0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Position (phase) Position (phase)

Fig. 3. The averaged normalized EMG activity (mean ± SE) for the 8 shoulder muscles during external rotation (n = 20). The position is indicated for
every 15 percentile of the movement cycle (six phases), the first phase of motion (0–15 percentiles) presents the beginning of the movement, with the arm
fully internal rotated as start position ( 80). When analysing the whole arc of motion, there was significant difference (P < 0.05) in EMG activity between
the dominant and the nondominant side for the supraspinatus, infraspinatus, lower and upper trapezius and latissimus dorsi muscles.

spinatus were quite active. The EMG amplitude in the ser-
ratus anterior was low during all parts of the motion. The
course of the activity curves at the two sides was almost
identical for supraspinatus, infraspinatus, lower trapezius,
latissimus and upper trapezius, except for the last part of
external rotation where a tendency towards reduced ampli-
tude was seen for some of the muscles on the dominant
side.
In contrast to the EMG activity during abduction, a
general pattern of greater EMG activity was seen in
the shoulder muscles of the dominant side during exter-
nal rotation, except for the serratus anterior that was
almost inactive (Fig. 3). Analysis of the EMG activity
during the full range of motion revealed a significant dif-
ference between dominant and nondominant side for the
supraspinatus (P = 0.01), infraspinatus (P < 0.001), lower
416 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
trapezius (P < 0.001), upper trapezius (P < 0.001) and
latissimus (P < 0.001) muscles. Analysis of each phase
of motion exhibited a significantly higher EMG activity
on the dominant side in the infraspinatus for phase 2
(P = 0.001), phase 3 (P < 0.001) and phase 4 (P = 0.001).
4. Discussion
We studied differences in EMG activity of the shoulder
muscles for the dominant and nondominant side during
movements in two directions: abduction in the scapular
plane and external rotation. There was a substantial
asymmetry in the EMG activity of the tested muscles
between the two sides. In addition, this side-difference in
muscle activity was dependent on the type of motion car-
ried out. During abduction, the shoulder muscles exhib-
ited lower EMG activity on the dominant side than the
nondominant side. During external rotation the opposite
was found; higher EMG activity on the dominant side
compared to the nondominant side.
The activation pattern of the shoulder muscles from our
study is summarized in brief. During abduction, all the
shoulder muscles were active. The anterior and middle part
of the deltoid muscle, upper trapezius and the serratus ante-
rior displayed increased activity during arm elevation. The
rotator-cuff muscles, i.e., supraspinatus and infraspinatus
showed a somewhat different activity pattern with a rise in
amplitude in the beginning of the motion followed by a grad-
ual reduction related to abduction angle. During external
rotation, the infraspinatus in particular was substantially
active, but also the lower trapezius and supraspinatus
showed a general pattern of continuous rise in activity with
increased degree of motion. The EMG amplitude of serratus
anterior, anterior and middle part of deltoid, latissimus and
upper trapezius was low throughout the motion. In the pres-
ent study, the EMG analysis was limited to EMG amplitude
and therefore we were not able to detect motor unit recruit-
ment, which requires a different approach.
Our observations of the activation pattern of the different
muscles acting on the shoulder joint are consistent with pre-
vious studies assessing EMG of shoulder muscles in healthy
subjects (Arwert et al., 1997; Bagg and Forrest, 1986; Inman
et al., 1944; Jobe et al., 1984; Jobe et al., 1983; Kronberg
et al., 1990; Nuber et al., 1986; Pearl et al., 1992; Ryu
et al., 1988; Saha, 1956; Shevlin et al., 1969). Although these
studies have provided important contributions to the func-
tion of various shoulder muscles during both isometric
and dynamic conditions as during sports activities, they
did not explore the possible influence of handedness on
EMG activity. Additionally, studies of shoulder patients
have shown motor patterns unlike the patterns found in
healthy subjects, however the results are ambiguous – hence
both hypo- and hyper-activity of shoulder muscles have
been reported (Alpert et al., 2000; Ludewig and Cook,
2000; Michaud et al., 1987; Peat and Grahame, 1977; Reddy
et al., 2000). Some of the diversity in these studies may be
due to the lack of attention to handedness.
In the study of Bagesteiro and Sainburg, side-differences
in coordination were investigated through analysis of kine-
matics and EMG during reaching movements (Bagesteiro
and Sainburg, 2002). When the kinematic pattern was sim-
ilar between sides, they demonstrated substantial differ-
ences in torque patterns and corresponding EMG
profiles; the dominant arm movements were consistently
performed with less muscle torque at both the elbow and
shoulder joint and with corresponding smaller EMG mus-
cle activity compared to the nondominant arm in the two
shoulder muscles that were assessed (pectoralis major and
posterior deltoid). In our study, eight shoulder muscles of
the dominant and nondominant side were examined during
two standardized movements, abduction and external rota-
tion. Our results confirmed the observations of Bagesteiro
and Sainburg. In addition, our results showed that
decreased EMG activity was consistently found for all
the shoulder muscles on the dominant side during arm ele-
vation. Although previously understated, our findings of
interlimb asymmetries in EMG activity of the shoulder
muscles were not surprising since there is overwhelming
evidence for physiological and anatomical side-differences
in the nervous system. Still, it is not possible from the pres-
ent study to identify the sites of neural adaptation respon-
sible for the handedness-related difference in muscle
activation. However, is seems unlikely that the diversity
occurs at the level of the muscles, as long as the differences
alters between abduction and external rotation, pointing at
a more centrally derived mechanism.
Contrary to the finding of lower EMG activation of the
shoulder muscles on the dominant side than on the non-
dominant side during arm elevation we found a general
pattern of higher activity during external rotation in the
shoulder joint. This finding of a task dependent side-differ-
ence in the activity pattern of the muscles suggests a sub-
stantial qualitative difference in the two types of
movement. It should be noted that dominant arm advanta-
ges do not apply to all tasks or all aspects of a task (Beu-
kelaar and Kroonenberg, 1983; Healey et al., 1986; Plato
et al., 1984). Healey et al. examined an extensive range of
tasks through a questionnaire and found that a group of
factors accounted for 80% of the variance in hand prefer-
ence among the 110 subjects tested (Healey et al., 1986).
Some tasks were performed almost exclusively by the dom-
inant arm, whereas others were significantly less dependent
on hand preference. The dominant arm tasks were charac-
terized by (1) relatively fine motor performance like writing
or (2) associated with ballistic activities that involve move-
ment of the proximal musculature requiring precision in
interlimb coordination. Dynamic abduction in the scapular
plan (as in our study) has the characteristics of a dominant
arm task. The decreased muscle activity on the dominant
side during arm elevation suggests a dominance-related
advantage since the movement can be done with less energy
consumption.
Conversely, The external rotation in our study did not
match the features of a dominant arm task (Beukelaar
 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
and Kroonenberg, 1983). The movement was neither char-
acterized by fine motor performance nor was it ballistic in
nature, since the movement was limited to the shoulder
joint with the elbow fixed and the forearm supported in
an armrest. The increased muscle activation on the domi-
nant side during external rotation suggests that when less
strongly lateralized tasks are performed with the dominant
side there may be increased cocontraction. The conse-
quence of increased activity during external rotation in
both agonists and antagonists would probable be an
increased control of the movement. Further studies, which
focus on weakly lateralized tasks, are needed before a
definitive conclusion can be reached on this point.
Long-term preferential use of selected muscles during
common tasks in daily living (like arm abduction, as in
our study) can be viewed as a moderate form of exercise
and may thus have the capacity to induce changes at differ-
ent levels within the nervous system resulting in reduced
EMG activity, maybe in part due to decreased co-contrac-
tion. In two studies, EMG was measured during reaching
movements and it was shown that co-contraction (activa-
tion of antagonists around a joint simultaneously with the
agonists) decreased gradually over the course of learning
a novel motor task (Osu et al., 2002; Thoroughman and
Shadmehr, 1999). Furthermore, Carroll et al. investigated
the effect of resistance training on the functional properties
of the corticospinal pathways in humans (Carroll et al.,
2002). They compared the effects of resistance training on
the EMG responses to transcranial magnetic and electrical
stimulation and found that for a specific level of muscle
activity, the magnitude of the EMG responses to both
forms of transcranial stimulation were smaller following
resistance training. Although they addressed resistance
training, one might expect the same neural adaptation fol-
lowing a moderate form of exercise due to long-time prefer-
ential use of one side.
The reported side-difference in EMG activity of the
shoulder muscles could theoretically be due to crosstalk
from the activity of the heart. Still that would give the same
bias in the EMG activity for the muscles during both
abduction and external rotation, which was not the case.
Furthermore, the change in activity probably would have
been presented just in the muscles next to the heart and
not to all the muscles, as in the current study. The different
muscle activation pattern during abduction could be
related to the increased strength on the dominant side,
although this was unlikely the case, since the EMG data
were normalized. Further, with the use of surface elec-
trodes there could be a risk of crosstalk from nearby mus-
cles especially for the upper and lower part of the trapezius
and for the serratus anterior. The electrode placements for
the upper and lower part of the trapezius and for the serra-
tus anterior were chosen to minimize the risk of crosstalk
from muscles such as the rhomboids and latissimus dorsi.
In the present study, MVC was used as normalisation
method. The shoulder muscles were tested according to
the recommendations for different shoulder muscles
417
(Inman et al., 1944; Kelly et al., 1996; Kendall and Ken-
dall, 1949; Schuldt and Harms-Ringdahl, 1988). However,
the EMG activity of the anterior deltoid and the serratus
anterior muscles was higher during dynamic abduction
than values from the MVC trials. The maximal voluntary
contraction is the most commonly used normalisation
method for EMG, although it is a well known phenomenon
that EMG amplitudes greater than that obtained at 100%
MVC can be obtained during dynamic contractions (Bur-
den and Bartlett, 1999; Morris et al., 1998). Since other
normalisation methods have similar or other technical lim-
itations it is difficult to bypass the problem.
Several factors regarding our subject sample should be
considered. Previous studies have found evidence for differ-
ent muscular activation patterns in elderly individuals, most
likely due to losses of motor units in older subjects (Brown
et al., 1988; Valour et al., 2003). Consequently, older adult
(age > 60 years) were excluded from our study. Further,
both males and females were included in our study.
Although there are no data identifying gender differences
for neither hand preference nor muscle activation patterns
of the shoulder, generalization of the study results to women
might be uncertain due to the small women sample (Clark
et al., 2003; Levander and Schalling, 1988). Furthermore,
because of the few left-handed subjects in the current study
we cannot compare left and right-handed subjects. Hand
preference may relate to unique motor patterns, which
makes it more difficult to ascertain differences in activation
patterns between the dominant and the nondominant side.
In conclusion, the present study demonstrated a hand-
edness related asymmetry in EMG activity of the muscles
acting on the shoulder joint, supporting previous findings
of anatomical and physiological asymmetries at different
levels of the nervous system. Furthermore, we established
a task dependent side-difference in the activity pattern of
the muscles suggesting a substantial qualitative difference
in the two types of movement, i.e., arm elevation present-
ing a dominant arm task. Reduced muscle activity on the
dominant side during abduction and thereby less energy
consumption indicates dominance-related advantages in
arm dynamics. However, with less muscle activity espe-
cially of the antagonists, the stability of the shoulder joint
might be decreased, which might make the shoulder more
vulnerable to injuries during dominant arm tasks. Further
studies are needed to examine this hypothesis.
From a clinical point of view, this knowledge of side-
asymmetries should be taken into account in future inves-
tigations on EMG activity in shoulder muscles of both
healthy subjects and patients with shoulder disorders, and
information about test side and hand preference should
be included.
Acknowledgements
This study was supported by Direktør E. Danielsen og
Hustrus Fond, C.C. Klestrup og Hustru Henriette Klet-
rups Legat, Beckett-Fonden and Smedemester Niels Han-
418 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
sen og Hustru Johanne, F. Frederiksens Legat. The
authors thank Dr. Peter Magnusson for assisting with
manuscript preparation.
References
Adam A, De Luca CJ, Erim Z. Hand dominance and motor unit firing
behavior. J Neurophysiol 1998;80:1373–82.
Alpert SW, Pink MM, Jobe FW, McMahon PJ, Mathiyakom W.
Electromyographic analysis of deltoid and rotator cuff function
under varying loads and speeds. J Shoulder Elbow Surg
2000;9:47–58.
Arwert HJ, de Groot J, Van Woensel WW, Rozing PM. Electromyog-
raphy of shoulder muscles in relation to force direction. J Shoulder
Elbow Surg 1997;6:360–70.
Bagesteiro LB, Sainburg RL. Handedness: dominant arm advantages in
control of limb dynamics. J Neurophysiol 2002;88:2408–21.
Bagg SD, Forrest WJ. Electromyographic study of the scapular rotators
during arm abduction in the scapular plane. Am J Phys Med
1986;65:111–24.
Beukelaar LJ, Kroonenberg PM. Towards a conceptualization of hand
preference. Br J Psychol 1983;74:33–45.
Brown WF, Strong MJ, Snow R. Methods for estimating numbers of
motor units in biceps-brachialis muscles and losses of motor units with
aging. Muscle Nerve 1988;11:423–32.
Burden A, Bartlett R. Normalisation of EMG amplitude: an evaluation
and comparison of old and new methods. Med Eng Phys
1999;21:247–57.
Carroll TJ, Riek S, Carson RG. The sites of neural adaptation induced
by resistance training in humans. J Physiol 2002;544:641–52.
Clark BC, Manini TM, The DJ, Doldo NA, Ploutz-Snyder LL. Gender
differences in skeletal muscle fatigability are related to contraction type
and EMG spectral compression. J Appl Physiol 2003;94:2263–72.
Delagi E, Perotto A, Lazatti J. Anatomic Guide for the electromyogra-
pher. Second, ed. Springfield (IL): Charles C Thomas; 1981 (Ref
Type: Serial (Book, Monograph)).
Farina D, Kallenberg LA, Merletti R, Hermens HJ. Effect of side
dominance on myoelectric manifestations of muscle fatigue in the
human upper trapezius muscle. Eur J Appl Physiol 2003;90:480–8.
Friedli WG, Fuhr P, Wiget W. Detection threshold for percutaneous
electrical stimuli: asymmetry with respect to handedness. J Neurol
Neurosurg Psych 1987;50:870–6.
Fugl-Meyer AR, Eriksson A, Sjostrom M, Soderstrom G. Is muscle
structure influenced by genetical or functional factors? A study of three
forearm muscles. Acta Physiol Scand 1982;114:277–81.
Hagberg C, Hagberg M. Surface EMG amplitude and frequency
dependence on exerted force for the upper trapezius muscle: a
comparison between right and left sides. Eur J Appl Physiol Occup
Physiol 1989;58:641–5.
Healey JM, Liederman J, Geschwind N. Handedness is not a unidimen-
sional trait. Cortex 1986;22:33–53.
Inman VT, Saunders JB, Abbott LC. Observations of the function of the
shoulder joint. J Bone Joint Surg 1944;26:1–30.
Jobe FW, Tibone JE, Perry J, Moynes D. An EMG analysis of the
shoulder in throwing and pitching. A preliminary report. Am J Sports
Med 1983;11:3–5.
Jobe FW, Moynes DR, Tibone JE, Perry J. An EMG analysis of the
shoulder in pitching. A second report. Am J Sports Med
1984;12:218–20.
Kelly BT, Kadrmas WR, Kirkendall DT, Speer KP. Optimal normaliza-
tion tests for shoulder muscle activation: an electromyographic study. J
Orthop Res 1996;14:647–53.
Kendall HO, Kendall FP. Muscle testing and function. Baltimore: Wil-
liams and Wilkins; 1949.
Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination
in the normal shoulder. An electromyographic study. Clin Orthop
1990:76–85.
Levander M, Schalling D. Hand preference in a population of Swedish
college students. Cortex 1988;24:149–56.
Ludewig PM, Cook TM. Alterations in shoulder kinematics and associ-
ated muscle activity in people with symptoms of shoulder impinge-
ment. Phys Ther 2000;80:276–91.
Michaud M, Arsenault AB, Gravel D, Tremblay G, Simard TG. Muscular
compensatory mechanism in the presence of a tendinitis of the
supraspinatus. Am J Phys Med 1987;66:109–20.
Morris AD, Kemp GJ, Lees A, Frostick SP. A study of the reproducibility
of three different normalisation methods in intramuscular dual fine
wire electromyography of the shoulder. J Electromyogr Kinesiol
1998;8:317–22.
Nuber GW, Jobe FW, Perry J, Moynes DR, Antonelli D. Fine wire
electromyography analysis of muscles of the shoulder during swim-
ming. Am J Sports Med 1986;14:7–11.
Oldfield RC. The assessment and analysis of handedness: the Edinburgh
inventory. Neuropsychologia 1971;9:97–113.
Osu R, Franklin DW, Kato H, Gomi H, Domen K, Yoshioka T, Kawato
M. Short- and long-term changes in joint co-contraction associated
with motor learning as revealed from surface EMG. J Neurophysiol
2002;88:991–1004.
Pearl ML, Perry J, Torburn L, Gordon LH. An electromyographic
analysis of the shoulder during cones and planes of arm motion. Clin
Orthop 1992:116–27.
Peat M, Grahame RE. Electromyographic analysis of soft tissue lesions
affecting shoulder function. Am J Phys Med 1977;56:223–40.
Plato CC, Fox KM, Garruto RM. Measures of lateral functional
dominance: hand dominance. Hum Biol 1984;56:259–75.
Reddy AS, Mohr KJ, Pink MM, Jobe FW. Electromyographic analysis of
the deltoid and rotator cuff muscles in persons with subacromial
impingement. J Shoulder Elbow Surg 2000;9:519–23.
Ryu RK, McCormick J, Jobe FW, Moynes DR, Antonelli DJ. An
electromyographic analysis of shoulder function in tennis players. Am
J Sports Med 1988;16:481–5.
Saha AK. Studies on electromyographic changes of muscles acting on the
shoulder joint complex. Calcutta Med J 1956;53:409–13.
Sathiamoorthy A, Sathiamoorthy SS. Limb dominance and motor
conduction velocity of median and ulnar nerves. Indian J Physiol
Pharmacol 1990;34:51–3.
Schuldt K, Harms-Ringdahl K. Activity levels during isometric test
contractions of neck and shoulder muscles. Scand J Rehabil Med
1988;20:117–27.
Semmler JG, Nordstrom MA. Hemispheric differences in motor cortex
excitability during a simple index finger abduction task in humans. J
Neurophysiol 1998;79:1246–54.
Shevlin MG, Lehmann JF, Lucci JA. Electromyographic study of the
function of some muscles crossing the glenohumeral joint. Arch Phys
Med Rehabil 1969;50:264–70.
Tan U. Velocities of motor and sensory nerve conduction are the same for
right and left arms in right- and left-handed normal subjects. Percept
Mot Skills 1985;60:625–6.
Tan U. Spinal motor lateralization assessed by recovery curve of H reflex
from wrist flexors in right-, and left-handed normal subjects. Int J
Neurosci 1989a;48:309–12.
Tan U. The Hoffmann reflex from the flexor pollicis longus of the thumb in
left-handed subjects: spinal motor asymmetry and supraspinal facilita-
tion to Cattell’s intelligence test. Int J Neurosci 1989b;48:255–69.
Thoroughman KA, Shadmehr R. Electromyographic correlates of learn-
ing an internal model of reaching movements. J Neurosci
1999;19:8573–88.
Valour D, Ochala J, Ballay Y, Pousson M. The influence of ageing on the
force–velocity–power characteristics of human elbow flexor muscles.
Exp Gerontol 2003;38:387–95.
van der Helm FC. Analysis of the kinematic and dynamic behavior of the
shoulder mechanism. J Biomech 1994;27:527–50.
Volkmann J, Schnitzler A, Witte OW, Freund H. Handedness and
asymmetry of hand representation in human motor cortex. J Neuro-
physiol 1998;79:2149–54.
 L.P. Diederichsen et al. / Journal of Electromyography and Kinesiology 17 (2007) 410–419
Louise Pyndt Diederichsen received her MD
from the University of Copenhagen, Denmark
in 1994. She has currently done her PhD at The
Sports Medicine Research Unit at Bispebjerg
Hospital in Copenhagen, Denmark. Her main
research interest is the neuromuscular interplay
between the shoulder muscles and ligaments in
the glenohumeral joint in the dynamic stability
of the shoulder among normal individuals and
individuals suffering from shoulder pain.
Jesper Nørregaard received his MD from the
University of Copenhagen, Denmark in 1987.
He is a specialist in internal medicine and
rheumatology and has made a medical disser-
tation on widespread muscle pain (fibromyal-
gia). He is now working as a chief consultant at
Hørsholm sygehus and is chairman of the
Danish Society of Rheumatology.
Poul Dyhre-Poulsen is MD from University of
Copenhagen, Denmark. He is Associate Pro-
fessor at Institute of Medical Physiology,
Panum Institute, University of Copenhagen.
His research field is motor control of human
movement and he focuses his research on
influences of proprioceptive information on
control of movement in human subjects.
Annika K.N. Winther received her MD from
the University of Copenhagen in 2004. She has
currently finished her internship and is plan-
ning to specialize in orthopaedic surgery. Her
main research area is the influence of acute
pain upon the activity of the shoulder muscles
during dynamic movement among normal
individuals.
419
Goran Tufekovic completed his master in Sci-
ence and Sport at the University of Copenha-
gen, Denmark 2003. The main focus of his
studies is the interaction of heavy resistance
training with the use of dietary supplementa-
tion of protein, carbohydrates and creatine.
Currently he is working as strength trainer and
physical fitness consultant to different sports
teams, including Team Danmark, the national
organisation for elite Olympic athletes.
Thomas Bandholm received his M.Sc. from the
Institute of Exercise and Sport Sciences in
2004. He is currently a PhD-student at the Gait
Analysis Laboratory, Hvidovre University
Hospital in Copenhagen, Denmark. His main
interest is motor control with an emphasis on
rehabilitation biomechanics.
Lars Raundahl Rasmussen became a certified physiotherapist from the
School of Physiotherapy in Holstebro, Denmark in 1997. He recieved his
masters degree in exercise physiology from the University of Copenhagen,
Denmark in 2005. He is currently in charge of physical training, testing
and development in FC. Midtjylland (first leaque male soccer team in
Denmark) and Skjern Haandbold (first leaque team handball club in
Denmark).
Michael R. Krogsgaard is MD since 1980 from
the University of Copenhagen. He received a
gold medal from Aarhus University in 1983
and obtained his PhD from Copenhagen Uni-
versity in 1992 on a thesis regarding cortico-
steroid induced bone loss. Since 1992 he has
been specialist in orthopaedic surgery, and
since 1996 consultant and associate professor
at Copenhagen University Bispebjerg Hospital,
Denmark. He works with neuromuscular
interplay in joints and other research fields.