Analysis of Scapular Kinematics and Muscle

Activity by Use of Fine-Wire Electrodes

During Shoulder Exercises
Kelly Berckmans,*y PT, Birgit Castelein,y PT, PhD, Dorien Borms,y PT, PhD,
Tanneke Palmans,y Thierry Parlevliet,z MD, and Ann Cools,y PT, PhD
Investigation performed at Ghent University, Ghent, Belgium

Background: During nonoperative or postoperative rehabilitation after sports injuries, exercise selection is often based on min-
imal load on the injured/repaired glenohumeral structures, while optimally activating scapulothoracic muscles. Previous research
explored scapular muscle activity during rehabilitation exercises using surface electromyography (EMG). However, limited
information exists about the deeper lying muscle activity, measured with fine-wire electrodes, even more in combination with
3-dimensional scapular kinematics.
Purpose: To report scapular kinematics synchronously with surface and fine-wire EMG during specific shoulder exercises for
early rehabilitation.
Study Design: Descriptive laboratory study.
Methods: A total of 23 healthy male patients were recruited. Three-dimensional scapular kinematics were measured combined
with EMG recording of 8 muscles during 4 commonly used shoulder exercises (inferior glide, low row, lawnmower, and robbery).
Upper (UT), middle (MT), and lower (LT) trapezius and serratus anterior muscle activities were measured with bipolar surface elec-
trodes. Intramuscular electrodes were placed in the levator scapulae (LS), rhomboid major (RM), pectoralis minor (Pm), and infra-
spinatus (IS) muscles. All data were normalized as a percentage of maximal voluntary isometric contraction (%MVIC). A linear
mixed model with Bonferroni correction was applied for statistical analysis.
Results: Scapular kinematics revealed an anterior tilt position during the inferior glide, low row, and robbery (P \ .05). An upward
rotation position between 20° and 30° was reached in all exercises except low row. Inferior glide (31°) and low row (42°) repre-
sented a significantly increased internal rotation position compared with lawnmower and robbery. Lawnmower and robbery
showed significantly (P \ .05) more MT (lawnmower, 36% MVIC; robbery, 39% MVIC) and RM (lawnmower, 59% MVIC; robbery,
66% MVIC) activation compared with inferior glide and low row. Lawnmower and robbery showed significantly (P \ .05) less Pm
activation (9.5%-12% MVIC). LS was significantly more active during robbery (58% MVIC) compared with inferior glide and low
row (27%-36% MVIC) (P \ .05). IS showed moderate activity (24%-37% MVIC) for all exercises, except low row (13% MVIC).
Conclusion/Clinical Relevance: This study provides new insights about scapular positions and activation of the deeper layer
muscles during 4 commonly used shoulder rehabilitation exercises. The lawnmower showed a favorable position of the scapula
with less Pm activity in contrast to the low row. The inferior glide, lawnmower, and robbery should not be implemented in early
phases of shoulder rehabilitation because of their moderate muscle activity.
Keywords: shoulder; physical therapy/rehabilitation; electromyography; rotator cuff; scapular kinematics

During the early phases of rehabilitation after a sports be a nonspecific response to shoulder dysfunction.9 There-
injury, or the protective rehabilitation phase after shoulder fore, early rehabilitation exercises often focus on strength-
surgery, exercise selection is often based on minimal load on ening the scapular muscles during isometric exercises or
the injured/repaired glenohumeral structures, with optimal in small ranges of motion.10
activation of the scapulothoracic muscles. As an example, Many factors determine the selection of exercises, such
high activity of the healed rotator cuff (RC) muscles is not as the patient’s goals and expectations, the proven effec-
tolerated in the early rehabilitation phase after RC repair. tiveness of a specific program, the tissue irritability, or
Alternatively, exercise therapy is initiated with activation the electromyography (EMG) activity in specific muscles.
of the scapular muscles, as scapular dyskinesis appears to EMG has been widely used to examine muscle activity dur-
ing commonly performed rehabilitation exercises. The sys-
tematic review by Schory et al16 made an overview of 15
The American Journal of Sports Medicine observational studies exploring the muscle activity in the
2020;48(5):1213–1219
DOI: 10.1177/0363546520908604 scapular muscles during exercises, all using surface
Ó 2020 The Author(s) EMG. They proposed a selection of exercises to target

1213
1214 Berckmans et al The American Journal of Sports Medicine

Figure 1. Scapular rotations: (A) downward/upward rotation around the frontal axis (Y-axis); (B) internal/external rotation around
the vertical axis (Z-axis); and (C) anterior/posterior tilt around the horizontal axis (X-axis). Modified and reprinted from Ludewig
PM, Hoff MS, Osowski EE, Meschke SA, Rundquist PJ. Relative balance of serratus anterior and upper trapezius muscle activity
during push-up exercises. Am J Sports Med. 2004;32(2):484-493.12

activity in a specific muscle. However, surface EMG is lim- knowledge about scapular 3D kinematics combined with
ited in use since only superficial muscles can be reached. In scapular muscle activity during exercises is lacking. Kibler
contrast, deeper layer muscles may be explored by using et al10 published a battery of 4 exercises useful during the
fine-wire EMG. Castelein et al5,6 used fine-wire electrodes early stages of rehabilitation. However, neither the activ-
combined with surface EMG to examine the muscle ity of the deeper layer muscles nor the 3D scapular kine-
recruitment during shoulder exercises, offering new matics of those 4 exercises was investigated. In addition,
insights into muscle recruitment of the pectoralis minor the study did not measure the activity of the RC muscles.
(Pm), levator scapulae (LS), and rhomboid major (RM) Therefore, the purpose of this study was to extend the
muscles. However, those studies are limited to a number study of Kibler et al10 and gain new insight into the muscle
of exercises without exploring the 3-dimensional (3D) kine- activity of the deeper layer scapulothoracic muscles.
matics of the scapula.
The knowledge about 3D scapular behavior with respect
to the thorax during shoulder exercises provides an addi-
tional value in the exercise’s choice within a rehabilitation METHODS
program.19,20 The scapula rotates around 3 axes, resulting
in an internal/external rotation, upward/downward rota- Participants
tion, and anterior/posterior tilting (Figure 1). Commonly
reported alterations in patients with injuries are increased A total of 23 healthy male patients (mean age, 22.3 6 1.5
anterior tilt, internal rotation, and decreased upward rota- years; mean height, 180 6 0.06 cm; mean weight, 81.7 6
tion.18 These alterations are often the result of impaired 17.2 kg; mean body mass index, 25.2 6 5.5 kg/m2) voluntar-
muscle function: more anterior tilt as a result of increased ily participated in this study. The patients were included if
tension in the Pm or decreased lower trapezius (LT) activ- the following criteria were met: patients were free of shoul-
ity; increased internal rotation because of lower activity of der or neck problems in the past 6 months, had no fracture
the serratus anterior (SA) muscle; or decreased upward or surgery of the shoulder in the past, and did not perform
rotation as a result of decreased upper trapezius (UT) overhead activities for more than 4 hours a week. All partic-
and SA activity, with increased tension in the LS. ipants signed a written consent before testing. This study
Although a scapula-based program is considered to be was approved by the ethical committee of the Ghent Univer-
relevant in the early phases of a rehabilitation program,4 sity Hospital (UZG 2016/1212).

*Address correspondence to Kelly Berckmans, PT, Department of Rehabilitation Sciences and Physiotherapy, Faculty of Medicine and Health Sciences,
Ghent University, Corneel Heymanslaan 10 - 3B3, Ghent, 9000, Belgium (email: [email protected]).
y
Department of Rehabilitation Sciences and Physiotherapy, Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium.
z
Physical Medicine and Orthopedic Surgery, Faculty of Medicine, University Hospital Ghent, Ghent, Belgium.
Submitted July 15, 2019; accepted January 2, 2020.
The authors declared that they have no conflicts of interest in the authorship and publication of this contribution. AOSSM checks author disclosures
against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or respon-
sibility relating thereto.
AJSM Vol. 48, No. 5, 2020 Scapular Kinematics and EMG During Shoulder Exercises 1215

General Design TABLE 1

Description: Positions of the Rehabilitation Exercises
The 3D kinematics regarding the scapular rotations were Including Pictures of the Exercises
measured and synchronized with EMG data of 7 scapulo-
thoracic muscles (UT, middle trapezius [MT], LT, SA, LS, Exercise Description
RM, Pm) and the musculus infraspinatus (IS) on the domi-
Inferior glide The patient sits with the arm 90°
nant side (side of writing hand) during the commonly used abducted and fist supported. The
shoulder rehabilitation exercises as described by Kibler patient is asked to press with the
et al.10 The exercises are described in Table 1. The patients fist in the direction of an arm
randomly performed each exercise 5 times with 30 seconds adduction for 5 seconds.
of rest between the trials and 3 minutes of rest between
the exercises. Two exercises were performed isometrically
(inferior glide and low row), and 2 were dynamic (lawnmower
and robbery). The isometric exercises were performed for a 5-
second duration. The dynamic exercises were performed for 3
seconds during concentric movement, followed by a 5-second
isometric hold. In view of comparative analysis of all 4 exer-
cises, the 5-second isometric hold position at the end of the
dynamic exercises was used for data analysis. Low row The patient stands in front of
a bench with palm of the hand
facing posteriorly. The patient is
asked to push with the
Test Procedure metacarpal part of the hand
against the edge of the table for 5
A shoulder clinical evaluation was performed before preparing seconds.
the patient for the measurements. This examination was com-
posed of 5 tests to make sure the participants did not suffer
from a shoulder disorder: Jobe, Neer, Hawkins, painful arc,
or external rotation against resistance.14 None of the patients
reported any painful item during this examination.
The preparation for the electromyographic measure-
ments consisted of shaving, scrubbing, and cleaning the
patient’s skin with alcohol to reduce impedance (\10 kX).
A medical doctor specializing in fine-wire EMG (T. Parlev-
liet) placed the intramuscular paired hooked fine-wire elec- Lawnmower The patient moves the arm into
trodes (wire length, 125 mm; Chalgren) with ultrasound retraction and external rotation
guidance into the following deeper layer muscles: LS, RM, with the elbow in the direction of
IS, and Pm. A single-use 25-gauge hypodermic needle the back pocket while performing
(length of the needle varying between 30 and 50 mm) was a trunk rotation and holds this
end position for 5 seconds.
inserted in the muscle belly according to the location
described by Delagi et al.8 For the other 4 scapulothoracic
muscles (UT, MT, LT, and SA), surface electrodes (Ambu
Blue Sensor N – ECG electrodes; N-00-S/25) were placed
with a 1-cm interelectrode distance according the instruc-
tions of Basmajian and De Luca.1 Both surface and fine-
wire electrodes were connected to a direct transmission sys-
tem (DTS) EMG sensor (Noraxon USA Inc) with shielded
cables. The cables were looped and taped on the skin to
avoid movement artifacts or being accidently removed dur- Robbery The patient moves both arms into
ing the exercises. The EMG data were collected by a Tele- retraction and external rotation,
Myo DTS Desk Receiver (Noraxon USA Inc) at a sampling with the elbow in the direction of
rate of 3000 Hz. the back pocket and holds this
A warm-up consisting of 15 repetitions of internal and end position for 5 seconds.
external rotation in a neutral and 90° of abduction position
and 15 repetitions of circumduction was followed by the
performance of maximal voluntary isometric contractions
(MVICs). The protocol of Castelein et al7 was used for the
scapulothoracic muscles (see Appendix Table A1, available
in the online version of this article). For the MVIC of the
IS, the patient was asked to perform an external rotation
against resistance when sitting in an upright position
1216 Berckmans et al The American Journal of Sports Medicine

TABLE 2 a DTS force sensor (Noraxon USA Inc), placed between

Placement of Reflective Spheres the patient’s hand and the bench, was used to determine
for 3D Scapular Kinematicsa when the patient started and ended pushing against the
bench. For the isometric end position of the dynamic lawn-
Segment Placement of the Marker mower and robbery, the windows of data were based on
Head LFHD-RFHD-LBHD-RBHD manually placed markers by the investigator at the start
Thorax Processus spinosus C7 and T8–processus xiphoideus of the 5-second holding position and at the end of the move-
(most caudal point on the sternum) –processus ment. For all exercises, the first and last repetition of 5
coracoideus–deepest point IJ-LAC-RAC were discarded. The average of the 5-second isometric
Scapula TS–AI–angulus acromialis (placed on cluster; most hold of the second to fourth repetition was used for further
laterodorsal point of scapula)–scapula cluster data analysis with respect to the 3D scapular position and
(upper, front, and back cluster marker) EMG activity for each muscle for all 4 exercises.
Arm Upper arm cluster (PM-PL-DM-DL)–ME–LE–lower The MyoResearch 3.10 Master edition (Noraxon USA
arm cluster marker (PM-PL-DM-DL)–processus
Inc) software was used for the signal processing of MVIC.
styloideus ulna and radius
Pelvis LSIAS-RSIAS-LSIPS-RSPIPS-LIC-RIC
Cardiac artifacts were subsequently removed from the sig-
nals, followed by a 30-Hz high-pass Butterworth filter,
a
AI, angulus inferior; DL, distal-lateral; DM, distal-medial; IJ, a full-wave rectification, and smoothing (root mean square
incisura jugularis; LAC, left acromioclavicular joint; LBHD, left of 100 milliseconds). For the MVIC analysis, a marker was
back head; LE, lateral epicondyle; LFHD, left front head; LIC, manually placed at the start and after 3 seconds for each
left iliac crest; LSIAS, left spina iliac anterior superior; LSIPS, repetition. The mean amplitude of the 3 seconds between
left spina iliac posterior superior; ME, medial epicondyle; PL, 2 markers was computed for each muscle at each position.
proximal-lateral; PM, proximal-medial; RAC, right acromioclavic- For all muscles except the Pm and IS, the greatest mean
ular joint; RBHD, right back head; RFHD, right front head; RIC, EMG value over positions 1 to 4 (see Appendix Table A1,
right iliac crest; RSIAS, right spina iliac anterior superior; RSIPS,
available online) was taken determining the MVIC used
right spina iliac posterior superior; TS, trigonum spinae.
for the normalization of the EMG activity of the exercises.
Concerning the IS and Pm, the EMG value of position 5
with the arm against the body and elbow 90° flexed in the (see Appendix Table A1, available online) was used for
neutral position.2 All positions were held for 5 seconds and the MVIC of Pm and the seated position with resistance
repeated thrice. against external rotation for the MVIC of IS.
After performing the MVIC, reflective spheres 12 mm in Subsequently, all values for EMG activity were
diameter were subsequently placed on bony landmarks at expressed as a percentage of MVIC (%MVIC). The values
the dominant side, head, thorax, and pelvis of the patient of EMG were categorized to facilitate the interpretation
for the measurement of the 3D kinematic data (Table 2). as low (\20% MVIC), moderate (20%-50% MVIC), or high
The kinematic data were collected by 6 Oqus 31 and 2 (.50% MVIC) activity.3
Oqus 4 motion capture cameras supported by the Qualysis
Track Manager (QTM; Qualysis AB) software. The marker
capture frequency was set at 300 Hz. Both EMG and 3D Statistical Analysis
kinematics were synchronously collected with QTM soft-
ware during exercise performance. The statistical software SPSS version 25.0 (IBM Corp) was
used for all statistical analysis. Means and standard devi-
ations were calculated for the 3 scapular rotations (X-axis,
Y-axis, and Z-axis) and normalized EMG activity for the 8
Signal Processing and Data Analysis
muscles investigated (UT, MT, LT, SA, LS, RM, Pm, and
Visual3D (Version 6 Professional, C-motion Inc) was used IS) during all 4 exercises. A linear mixed model for
for the EMG and kinematic data processing of the different repeated measures with post hoc Bonferroni correction
exercises. The kinematic patterns are described by the for pairwise comparison was used. Based on the residuals,
upper limb and pelvis model (Ghent University).13,15,21,22 data were normally distributed. The significance level was
In this upper limb and pelvis model, the joint coordinate set at alpha = .05. Interaction effects exercise*scapular
system for the shoulder girdle was based on the Interna- rotation and exercise*muscle were of primary interest. In
tional Society of Biomechanics recommendation with a gle- case of absence of significant interaction effects, main
nohumeral joint center estimation from scapular bony effects for ‘‘scapular rotation,’’ ‘‘muscle,’’ and ‘‘exercise’’
landmarks described by Meskers et al.13 This model were further explored.
described the movements of the scapula around 3 different
axes: horizontal axis (X-axis) resulting in anterior tilt or
posterior tilt; frontal axis (Y-axis) resulting in upward RESULTS
rotation or downward rotation; and vertical axis (Z-axis)
Scapular Kinematics
resulting in an external rotation or internal rotation of
the scapula. All EMG signals were filtered with a high- Results of descriptive analysis are displayed in Table 3. A
pass Butterworth filter at 30 Hz and full-wave rectified. significant interaction effect for exercise*scapular rotation
For the isometric exercises (inferior glide and low row) (P  .001) was found.
AJSM Vol. 48, No. 5, 2020 Scapular Kinematics and EMG During Shoulder Exercises 1217

TABLE 3 the inferior glide as well as the low row represent signifi-
Descriptive Statistics of the Scapular Positions During cantly (P  .001) more internal rotation.
the Isometric Phase of the Rehabilitation Exercisesa

X-Axis (°) Y-Axis (°) Z-Axis (°) EMG Activity of Scapulothoracic Muscles
AT (–)/PT (1) DR (–)/UR (1) IR (–)/ER (1)
The results showed a significant interaction effect for posi-
Inferior glide –7.19 6 6.96 23.91 6 14.52 –31.30 6 6.95 tion*EMG muscle activity (P  .001). The results of the
Low row –8.83 6 5.49 3.08 6 5.13 –42.32 6 15.04 pairwise comparison with the post hoc analysis are pre-
Lawnmower 0.915 6 4.89 29.24 6 8.47 –14.97 6 9.90 sented in Table 4. With respect to deep scapular muscles,
Robbery –7.04 6 5.79 30.32 6 9.66 –13.34 6 12.49
LS and RM activity ranged between 27% and 66% MVIC.
a Pm activity was moderate (34%-44% MVIC) during low
AT, anterior tilt; DR, downward rotation; ER, external rota-
tion; IR, internal rotation; PT, posterior tilt; UR, upward rotation. row and inferior glide compared with the lawnmower
(9.53% MVIC) and robbery (11.85% MVIC).

The isometric phase of inferior glide, low row, and rob-

bery resulted in a significantly (P \ .05) increased anterior DISCUSSION
tilt position of the scapula compared with a slight posterior The purpose of this study was to explore the differences
tilt position for the lawnmower. The low row displayed sig- between previously published exercises regarding electro-
nificantly (P  .001) less upward rotation compared with myographic muscle activity and 3D kinematics of the scap-
the other 3 exercises, where an upward rotation between ula.10 Combining scapular 3D kinematics measurements
20° and 30° was reached. Besides the decreased upward with fine-wire EMG measurements of the scapulothoracic
rotation, the low row resulted in a significantly (P  muscle revealed new insights about the previously recom-
.001) increased internal rotation relative to the other 3 mended exercises by Kibler et al10 for early rehabilitation
exercises. In comparison with the lawnmower and robbery, after shoulder injury.

TABLE 4
Detail of the Comparative Statistical Analysis for EMG Measurements Based on Exercisesa

P Valueb

Muscle Inferior Glide Low Row Lawnmower Robbery Pairwise Comparisons

UT 9.32 6 5.16 7.39 6 4.17 20.60 6 12.23 23.49 6 11.46 IG-LR: .999 LR-LM: .287
IG-LM: .521 LR-RB: .079
IG-RB: .160 LM-RB: .999
MT 18.38 6 9.87 25.66 6 12.39 36.21 6 17.36 38.70 6 22.18 IG-LR: .999 LR-LM: .608
IG-LM: .031 LR-RB: .251
IG-RB: .009 LM-RB: .999
LT 12.31 6 5.83 13.26 6 11.38 29.20 6 14.90 24.42 6 12.15 IG-LR: .999 LR-LM: .075
IG-LM: .048 LR-RB: .495
IG-RB: .355 LM-RB: .999
SA 20.27 6 13.88 18.35 6 17.92 20.42 6 13.42 14.85 6 11.93 IG-LR: .999 LR-LM: .999
IG-LM: .999 LR-RB: .999
IG-RB: .999 LM-RB: .999
LS 27.35 6 22.62 36.37 6 27.98 49.39 6 34.89 58.05 6 40.23 IG-LR: .917 LR-LM: .282
IG-LM: .004 LR-RB: .004
IG-RB: \.001 LM-RB: .952
RM 27.51 6 14.10 27.02 6 14.95 58.68 6 25.17 66.12 6 39.66 IG-LR: .999 LR-LM: \.001
IG-LM: \.001 LR-RB: \.001
IG-RB: \.001 LM-RB: .999
Pm 44.36 6 37.65 34.06 6 38.43 9.53 6 10.99 11.85 6 22.20 IG-LR: .618 LR-LM: \.001
IG-LM: \.001 LR-RB: .002
IG-RB: \.001 LM-RB: .999
IS 23.79 6 15.21 12.73 6 8.18 36.58 6 30.67 30.13 6 21.23 IG-LR: .617 LR-LM: .001
IG-LM: .279 LR-RB: .036
IG-RB: .999 LM-RB: .999

a
Values are expressed in %MVIC 6 SD. EMG, electromyography; IG, inferior glide; IS, infraspinatus; LM, lawnmower; LR, low row; LS,
levator scapulae; LT, lower trapezius; MT, middle trapezius; MVIC, maximal voluntary isometric contraction; Pm, pectoralis minor; RB, rob-
bery; RM, rhomboid major; SA, serratus anterior; UT, upper trapezius.
b
P \ .05 indicates statistical significance.
1218 Berckmans et al The American Journal of Sports Medicine

Comparing our results with the Kibler et al10 study, could have appeared during the performance of a move-
comparable activity levels were found for the UT, LT, ment when fine wires are used. However, while performing
and SA, illustrating the similarity between studies regard- the exercises none of the patients were limited in arm
ing common outcome variables. Interestingly, new insights motions because of discomfort.
were noted in the results of the activity in the deeper layer Second, although the procedure was performed as stan-
muscles (Pm, RM and LS, and the IS) and the 3D kinemat- dardized as possible, some inconsistencies may have influ-
ics of the scapula. enced our results. The same medical doctor (T. Parlevliet)
The MT (36%-39%) and RM (59%-66%) showed their performed all fine-wire applications, and the same
greatest activity at the isometric phase of the dynamic researcher (K.B.) performed the resistance during the
exercises. The greater activity of the RM compared with MVIC for all participants. Furthermore, all participants
the MT can be clarified by the moderate upward rotation had to perform the exercises with the same time cues to
position during the exercises. Possibly, MT would be standardize the movement as much as possible (see Appen-
more active in higher degrees of upward rotation of the dix Table A2, available online). Notwithstanding, the per-
scapula.11 The LS and RM revealed high activity of 50% formance of the exercises could not be totally
to 66% during the isometric phase of the dynamic exer- standardized. To minimize the differences, each patient
cises. We asked patients to retract their scapula during was placed in the same position at the beginning of the
this phase, which explains why the RM shows this amount exercise as well as informed in the same way regarding
of activity. The LS seems to help the RM in this function. how he was supposed to execute the movement.
The muscle activity of the Pm was the greatest (34%- Future studies should include other exercises to gain
44%) during the inferior glide and low row, both isometric more information about synchronized measured 3D scapu-
exercises against a fixed resistance. Similar results were lar position and EMG activity. As we only measured the IS,
reported by Castelein et al6 during a series of closed- because of technical limitations, the other RC muscles of
kinetic chain exercises with fixed hands. the shoulder girdle were not explored in this study. As
In the protective phase of nonoperative or postoperative the muscles could be reached with fine-wire EMG,
rehabilitation programs, less activity of the IS is recom- researchers should consider exploring the other RC
mended.17 Therefore, the inferior glide, lawnmower, and muscles in the future as well.1,8
robbery exercises should be avoided because of the moder-
ate activity of the IS. The low row had the least activity
(13%) for the IS, making the exercise suitable in the early Clinical Implications
phase of rehabilitation of shoulder disorders.
All 4 exercises except the lawnmower showed an ante- Based on the results of this study, several clinical implica-
rior tilt of the scapula. For the inferior glide and low row, tions can be drawn to guide the clinician in the exercise
the position of the scapula can be explained by the higher choice for scapular rehabilitation.
amount of EMG activity of the Pm. However, for the rob-
(1) In patients who exhibit insufficient posterior tilt and/or
bery this position cannot be explained by increased Pm
hyperactivity of the Pm, all exercises, except the lawn-
activity, but possibly rather by the instruction to bring
mower, are contraindicated based on increased ante-
the elbows down during this exercise. The upward rotation
rior tilt of the scapula and/or increased activity in
is related to the position of the arm; namely, greater
Pm. Only the lawnmower leads to a posterior tilt of
upward rotation values are described in exercises with
the scapula and reveals EMG activity of only 9.5% in
a relative arm elevation. Therefore, the low row showed
the Pm.
the least degree of upward rotation. The lawnmower and
(2) For patients with insufficient upward rotation, and
robbery revealed lesser degrees of internal rotation of the
possibly hyperactivity in LS, the exercises are not rec-
scapula (15°) compared with the inferior glide and low
ommended, given the moderate to great activity of LS
row. Those results can be elucidated by the greater activity
for all 4 exercises.
of the RM, as this muscle is expected to participate in this
(3) For patients with increased internal rotation, the
movement.
lawnmower and robbery may be advised since these
exercises place the scapula in a more external position
Strengths and Limitations compared with the common resting position in the
scapular plane (approximately 30° anterior of the fron-
The major strength of this study lies in the complex meth-
tal plane).18 The other 2 exercises are contraindicated.
ods used to analyze muscular activity simultaneously with
3D scapular kinematic behavior during commonly used In summary, the lawnmower contains scapular positions
exercises. The combination of fine wire with surface elec- in the direction of posterior tilt, upward rotation, and
trodes synchronously measured with the 3D kinematics external rotation combined with less Pm activity, com-
successfully provides us new information about scapular pared with the other exercises. On the other hand, the
behavior during the exercises. However, some limitations low row resulted in an anterior tilt scapular position with
should be noted. less upward rotation and great internal rotation in combi-
First, only healthy patients were included in this study. nation with a moderate activity (635%) of the LS and Pm.
Therefore, some caution should be taken when generaliz- Three of the exercises (inferior glide, lawnmower, and rob-
ing the results. We can also assume that some discomfort bery) should not be included in the early phase of
AJSM Vol. 48, No. 5, 2020 Scapular Kinematics and EMG During Shoulder Exercises 1219

rehabilitation because of the moderate activity of the IS. 10. Kibler WB, Sciascia AD, Uhl TL, Tambay N, Cunningham T. Electro-
However, the clinician should take into account that this myographic analysis of specific exercises for scapular control in early
phases of shoulder rehabilitation. Am J Sports Med. 2008;36(9):
study only focused on local scapular muscle activity,
1789-1798.
whereas in the clinical setting the patient should always 11. Kinney E, Wusthoff J, Zyck A, et al. Activation of the trapezius muscle
be managed from a biopsychosocial context. during varied forms of Kendall exercises. Phys Ther Sport. 2008;
9(1):3-8.
12. Ludewig PM, Hoff MS, Osowski EE, Meschke SA, Rundquist PJ. Rel-
ACKNOWLEDGMENT ative balance of serratus anterior and upper trapezius muscle activity
during push-up exercises. Am J Sports Med. 2004;32(2):484-493.
The authors thank the master’s students who helped dur- 13. Meskers CGM, Van Der Helm FCT, Rozendaal LA, Rozing PM. In vivo
estimation of the glenohumeral joint rotation center from scapular
ing the measurements and all the volunteers who partici-
bony landmarks by linear regression. J Biomech. 1997;31(1):93-96.
pated in this study. 14. Michener LA, Walsworth MK, Doukas WC, Murphy KP. Reliability and
diagnostic accuracy of 5 physical examination tests and combination
of tests for subacromial impingement. Arch Phys Med Rehabil.
REFERENCES 2009;90(11):1898-1903.
15. Robertson DGE, Dowling JJ. Design and responses of Butterworth
1. Basmajian JV, De Luca CJ. Muscles Alive: Their Functions Revealed and critically damped digital filters. J Electromyogr Kinesiol. 2003;
by Electromyography. Williams & Wilkins; 1985. 13(6):569-573.
2. Boettcher CE, Ginn KA, Cathers I. Standard maximum isometric vol- 16. Schory A, Bidinger E, Wolf J, Murray L. A systematic review of the
untary contraction tests for normalizing shoulder muscle EMG. exercises that produce optimal muscle ratios of the scapular stabil-
J Orthop Res. 2008;26(12):1591-1597. izers in normal shoulders. Int J Sports Phys Ther. 2016;11(3):321-336.
3. Borms D, Ackerman I, Smets P, Van Den Berge G, Cools AM. Biceps 17. Sgroi TA, Cilenti M. Rotator cuff repair: post-operative rehabilitation
disorder rehabilitation for the athlete. Am J Sports Med. 2017; concepts. Curr Rev Musculoskelet Med. 2018;11(1):86-91.
45(3):642-650. 18. Struyf F, Nijs J, Baeyens JP, Mottram S, Meeusen R. Scapular posi-
4. Bury J, West M, Chamorro-Moriana G, Littlewood C. Effectiveness of tioning and movement in unimpaired shoulders, shoulder impinge-
scapula-focused approaches in patients with rotator cuff related ment syndrome, and glenohumeral instability. Scand J Med Sci
shoulder pain: a systematic review and meta-analysis. Man Ther. Sports. 2011;21(3):352-358.
2016;25:35-42. 19. Turgut E, Duzgun I, Baltaci G. Effects of scapular stabilization exer-
5. Castelein B, Cagnie B, Parlevliet T, Cools A. Serratus anterior or pec- cise training on scapular kinematics, disability, and pain in subacro-
toralis minor: which muscle has the upper hand during protraction mial impingement: a randomized controlled trial. Arch Phys Med
exercises? Man Ther. 2016;22:158-164. Rehabil. 2017;98(10):1915-1923.e23.
6. Castelein B, Cagnie B, Parlevliet T, Cools A. Superficial and deep 20. Turgut E, Pedersen Ø, Duzgun I, Baltaci G. Three-dimensional scapular
scapulothoracic muscle electromyographic activity during elevation kinematics during open and closed kinetic chain movements in asymp-
exercises in the scapular plane. J Orthop Sports Phys Ther. 2016; tomatic and symptomatic subjects. J Biomech. 2016;49(13):2770-2777.
46(3):184-193. 21. Van Der Helm FCT. A standardized protocol for motion recordings of
7. Castelein B, Cagnie B, Parlevliet T, Danneels L, Cools A. Optimal nor- the shoulder. Proc First Conf ISG. 1996;7-12. https://www.research
malization tests for muscle activation of the levator scapulae, pector- gate.net/profile/Frans_Van_Der_Helm/publication/238167085_A_
alis minor, and rhomboid major: an electromyography study using standardized_protocol_for_motion_recordings_of_the_shoulder/links/
maximum voluntary isometric contractions. Arch Phys Med Rehabil. 5445ac010cf22b3c14ddec09/A-standardized-protocol-for-motion-
2015;96(10):1820-1827. recordings-of-the-shoulder.pdf. Accessed February 17, 2020.
8. Delagi EF, Iazzetti J, Perrotto A, Morrison D. Anatomical Guide for the 22. Wu G, Van Der Helm FCT, Veeger HEJ, et al. ISB recommendation on
Electromyographer: The Limbs and Trunk. Charles C Thomas; 1994. definitions of joint coordinate systems of various joints for the report-
9. Kibler WB, John MM. Scapular dyskinesis and its relation to shoulder ing of human joint motion—Part II: Shoulder, elbow, wrist and hand.
pain. J Am Acad Orthop Surg. 2003;11(2):142-151. J Biomech. 2005;38(5):981-992.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.