Journal of Orthopaedic & Sports Physical Therapy

1999;29(10):574-586

Comparison of 3-Dimensional Scapular

position and Orientation Between-subjects
With and Without Shoulder Impingement
Amy Cole Lukasiewicz, MS, PT1
Philip McClure, PhD, P7; OCSZ
Lori Michener, MEd, P7; ATC3
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.

Neal Praff, PhD, PT4

Brian Senneff, MD5

Study Design: Nonrandomized 2 group post-test only.

T
he term shouZder im-
Objective: To compare scapular position and orientation between subjects with and without pingemat was first in-
im~ineementsvndrome. troduced by Neer in
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

~ackg';ound:Abnormal scapular motion is commonly believed to be a contributingfactor to 1972'"d represents

shoulder impingement syndrome.
mechanical compres
Methods dMeasures: Twenty nonimpaired subjects with a mean age of 34.3 (27.5 years) and 17
patients with impingement syndrome with a mean age of 45.8 (211.0) participated. A 3dimemionaI sion of the rotator cuff and sub-
electromechanical digitizer was used to measure scapular position and orientation in 3 planes. acromial bursa against the anteri-
or undersurface of the acromion
Measurements were taken with the arm at the side, elevated in the scapular plane to horizontal, and at
maximum elevation. Oneway. analysis .
and coracoacromial ligament, es-
of variance was used to compare nonimpaired subjects to the
impingement group and the symptomatic and asymptomatic sides within the impingementgroup. Five pecially during elevation of the
scapular kinematic variables were assessed at each arm position. Orientation was described by posterior
arm (flexion or abduction). Re-
tilting angle, upward rotation angle, and internal rotation angle. Position was described by medial-lateral
search indicates that shoulder im-
Journal of Orthopaedic & Sports Physical Therapy®

position and superior-inferior position and determined by the distance from the scapula centroid to the
pingement is the most common
k e n t h cervical vertebra ( ~ 7 ) .
Results: During scapular plane elevation of the arm, the scapula showed a general pattern of
cause of shoulder pain1J9and is
often treated conservatively with
increasing posterior-tilt angle, increasing upward-rotationangle, and decreasing internal-rotation
angle in both impingement and nonimpaired groups. Also, the scapula moved to a more superior exercises to improve function of
position and a slightly more medial position with increasing arm elevation. Compared to the rotator cuff muscles, as well as
nonimpaired subjects (34.6" 2 9.7), those with impingement demonstrated a significantly lower those controlling the s c a p ~ l a . ~ J ~ J ~
posterior tilting angle of the scapula in the sagittal plane (25.1" t 9.1). Subjects with impingement
Multiple factors are thought to
also demonstrated higher superior-inferiorscapular position with maximal arm elevation (5.2 cm 2contribute to impingement includ-
1.6 below the first thoracic vertebrae) compared to nonimpaired subjects (7.5 cm 2 1.5). ing bony abnormalities, poor rota-
Conclusions: These results suggest that altered scapular kinematics may be an important aspect of
tor cuff muscle function, posterior
the impingement syndrome. I Orthop Spom Phys Ther 1999;29:574-586.
capsular tightness, and improper
Key Words: impingement, kinematics, rotator cuff, scapula, shoulder motor control of scapular motion.
These intrinsic factors are often
Bryn Mawr Spom Medicine, Bryn Maw6 h. combined with external factors
Department of Physical Therapy, Beaver College, Glenside, Pd.
Department of Physical Therapy, MCP-HahnemannUniversify (formerlyAllegheny University of the such as repeated application of
Health Sciences), Philadelphia, Pd. high loads and velocities that can
Department of Physical Therapy, MCP-HahnemannUniversify (formerlyAllegheny Universify of the produce overuse of the rotator
Health Sciences), Philadelphia, Pd. cuff tendons, particularly the su-
Department of Orthopedic Surgery, MCP-Hahnemann University (formerlyAllegheny University of
the Health Sciences), Philadelphia, Pd.
praspinatus.
Send correspondenceto Philip McClure, Department of Physical Therapy, Beaver College, 445 South Abnormal scapular motion is
Easton Road, Glenside, PA 1 9038. E-mail: mcclure@beaver.edu thought to be a potential source
 of mechanical dysfunction leading to impinge- Demographic data of nonimpaired and impingement groups.
ment.9~5~R~10~'4~18~20
and is thought to be due to weak- Nonimpaired (n = 20) Impingement (n = 17)
ness or improper neuromuscular control of the ser- Variable Mean (SD) Mean (SD)
ratus anterior as well as the upper and lower por- Age (years) 34.4 (7.5) 45.9 (11.O)
tions of the trapezius muscle.14A kyphotic thoracic Height (cm) 170.9 (10.4) 172.0 (9.7)
spine and the associated forward head posture are Weight (kg) 69.5 (13.1) 83.4 (18.5)
also believed to produce abnormal scapular motion Hand dominance 15 Right, 5 Left 17 Right
Symptomatic side NA 10 Right, 7 Left
with subsequent impingement because of the multi- Sex 8 Men, 12 Women 12 Men, 5 Women
ple muscular connections between the spine and
scap~la.~~.~.~
A few authors have directly related the problem of
upon correction of faulty and aberrant scapular m e
impingement to scapular position or motion. Solem-
tion patterns, quantitative evidence documenting spe-
Bertoft et allRused magnetic resonance imaging to
cific patterns of abnormal scapular motion in this pa-
study the effect of scapular protraction and retrac-
tient group is still quite limited. Use of 3dimensional
tion on the width of the subacromial space in 4 non-
measures in a group of patients with shoulder im-
impaired subjects. Subjects were supine while the
pingement syndrome could help identify important
scapulae were passively positioned in protraction and
abnormalities that could be specifically addressed by
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.

retraction with sandbags. The arms of the subject

exercise programs. The purpose of this study was to
were loaded caudally to mimic the weight of the arm
compare the .?dimensional scapular positions and
in upright standing. They found significant narrow-
orientations between nonimpaired subjects and pa-
ing of the subacromial space in the protracted posi-
tients with shoulder impingement syndrome during
tion and also a significant reduction in what they
scapular plane abduction.
termed "acromial angle," which represented relative
anterior tilting in the sagittal plane.'*
Warner et alm used Moir6 topography to study METHODS
scapular position in patients with 22 asymptomatic
shoulders, 22 unstable shoulders, and 7 shoulders Subjects
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

with impingement syndrome. The positions studied

Both shoulders from a total of 37 subjects were
were with the arm at rest and elevated to 90 degrees.
tested. The comparison group consisted of 20 non-
The patients with shoulder impingement showed in-
impaired volunteers (age range 25-54) with asympte
creased scapular elevation and increased scapular
matic shoulders. Seventeen subjects had unilateral,
winging, especially during active motion. The authors
positive impingement criteria at the shoulder joint
hypothesized that the increased scapular elevation
(age range 26-66). The characteristics of each group
was a compensation for abnormal glenohumeral m e
are presented in the table. Prior to participation, all
tion associated with the posterior capsular tightness
subjects were informed of the nature of the study
often seen with impingement. The increased topog-
and signed a consent form approved by the Human
Journal of Orthopaedic & Sports Physical Therapy®

raphy, indicating scapular winging, was interpreted to

Subjects Committee of Allegheny University of the
represent poor serratus anterior function.
Health Sciences.
Greenfield et al-ompared subjects with shoulder
Subjects with impingement were admitted into the
overuse injuries to asymptomatic, nonimpaired s u b
study as they became available from multiple clinics
jects matched by age and sex. They measured head,
throughout the greater Philadelphia region. Nonim-
thoracic spine, and scapular resting posture as well as
paired subjects were recruited throughout the study
shoulder range of motion. Their inclusion criteria
from personal contacts and patients without shoulder
(pain attributed to overuse, insidious onset of symp
impairments. To be included in the impingement
toms, no history of macrotrauma) suggested that
group, subjects must have had at least 3 of the fol-
many patients may have had impingement. They
lowing 6 criteria:
found that subjects with overuse injuries had a more
forward head posture and greater shoulder elevation Positive Neerls sign: The examiner passively flexes
range of motion. However, there were no differences the humerus to end-range with overpressure. The
between groups with respect to thoracic and scapular patient's facial expression and reproduction of the
resting posture. In this study, the scapular measure- pain confirms the presence of impingement.
ments were only 2dimensional and the measure of Positive Hawkins6 sign: The shoulder is passively
"scapular rotation" was indirect. The lack of differ- placed in approximately 90 degrees of flexion and
ences between groups may have been because mea- is passively internally rotated to end-range with ov-
surements in 3 dimensions were not performed and erpressure, reproducing the patient's pain.
only resting posture was assessed. Pain with active shoulder elevation in the scapular
Although much of the nonsurgical treatment for plane.
patients with shoulder impingement is predicated Pain with palpation of the rotator cuff tendons.

J Orthop Sports Phys Ther .Volume 29.Number 10.0ctober 1999

 5. A history of pain in the C.546 dermatome.
6. Pain with resisted isometric abduction.
Subjects were excluded if they had any of the follow-
ing:
1. Current symptoms related to the cervical spine.
2. Positive tests for shoulder instability:
a. Sulcus sign: A positive sulcus sign is an exces-
sive downward movement of the humeral head
away from the acromion when an inferior pull
is placed on the humerus while the arm is in a
dependent position.
b. Apprehension sign: Performed with the patient
supine, the humerus is externally rotated and
abducted, while stressing the capsule anteriorly.
The patient will report feelings of misgiving
and anxiety, sometimes associated with pain.
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
c. History of shoulder dislocation.
3. Acromioclavicular pain: With the arm flexed to
approximately 90 degrees, pain is reproduced by
passively adducting and internally rotating the hu-
merus across the chest to approximate the acrom-
ioclavicular joint.
To further describe the impingement group, a
shoulder pain and disability index (SPADI) was ad-
ministered to all subjects with impingement." The
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
SPADI questionnaire scales have been found to be a
valid measure of pain and disability for shoulder dys-
function.I7This scale rates pain and disability on a
100-point scale, 0 representing no symptoms of pain
or disability.
Instrumentation
An electromechanical device capable of digitizing
Journal of Orthopaedic & Sports Physical Therapy®
points in 3 dimensions (Metrecorn, Faro Inc, Marys
Lake, Fla) was used to determine Sdimensional posi-
tion of the scapula. The digitizer consists of a linkage
arm instrumented with 6 rotary potentiometers that
is interfaced with a personal computer. A probe at
the end of the linkage is used to locate a point de-
fined in a 3dimensional Cartesian (x,y,z) coordinate
system. These coordinates are stored in a file for s u b
sequent processing. We have shown the accuracy of
this device to be within 2 2 mm for linear measures
over the range used in this study.12
Data Collection
The subject was seated backwards in a sturdy chair
with an adjustable, padded back that allowed for
clearance of the thighs under the back of the chair.
This allowed the subject's back to be exposed while a
hook and loop strap, placed well below the scapulae,
was used to stabilize the anterior aspect of the sub-
ject's trunk against the padded back of the chair.
This provided stabilization, preventing postural sway
576
or weight shifts while elevating the arm. The subject
was seated with the feet flat on the floor, eyes fixed
forward, and arms hanging in a dependent position.
A global reference frame was established within
the stabilization chair that was consistent with the
body anatomical coordinate system of the subject.
Because the subject's torso was firmly stabilized to
the chair, the orientation of the chair (global refer-
ence) and the thorax were considered to be aligned
with the same orientation. The x-axis was a horizon-
tal vector positive to the left, the y-axis was a vertical
vector (along the back of the chair) positive superi-
orly. The z-axis was defined by the cross product of
the x- and y-axes and was positive in the anterior di-
rection.
The examiner palpated 6 bony landmarks and dig-
itized the 6 landmarks with the digitizing probe. The
6 bony landmarks digitized were as follows: (1) the
spinous process of C7, (2) the spinous process of T7,
(3) the medial border of the scapula at the root of
the spine of the scapula, (4) the posterior angle of
the acromion, (5) the inferior angle of the scapula,
and (6) the olecranon process. The positions of
these bony landmarks were digitized with the arm at
rest, abducted in the scapular plane to horizontal,
and at maximal elevation. The scapular plane was de-
termined for the horizontal position by using a goni-
ometer to estimate a position 40" anterior to the
frontal plane of the subject. The Sdimensional Car-
tesian coordinates of the digitized points were stored
in a digital file for subsequent data reduction.
The right arm was measured first, followed by the
left arm. Each subject was asked to elevate the arm
with the forearm in neutral pronation and supina-
tion, elbow extension, and with the thumb pointing
superiorly. The subject was instructed to keep the
palm facing forward. The horizontal shoulder abduc-
tion position was produced by placing a wooden 90"
wedge with a padded arm rest against the subject's
trunk. Once the 90" position was attained, the wedge
was removed and the subject was asked to maintain
that position actively for approximately 30-60 sec-
onds while the skeletal landmarks were digitized.
Two trials at each arm position were performed and
averaged. A 3 k c o n d rest period was allowed be-
tween each trial and arm position.
Data Reduction
Medial-lateral and superior-inferior position were
calculated directly from Cartesian coordinates. The
centroid of the scapula was determined by averaging
the x, y, and z components of the 3 scapular land-
marks. The Cartesian coordinates of the digitized
bony landmarks were also used to calculate relevant
vectors. Angles describing the orientation of the
scapula were calculated using a vector dot product
formula in a custom-written computer program ac-
J Onhop Sports 'Phys Ther .Volume 29. Number 10 .October 1999
 cording to the descriptions provided below. These
positions and angles are depicted in Figure 1. The 5
variables describing scapular position and orientation
were calculated as follows:
a) Medial-lateral position: The horizontal distance
between C7 and the centroid of the digitized
points on the scapula.
b) Superior-inferior position: The vertical distance
between C7 and the centroid of the digitized
points on the scapula.
c) Upward rotation angle: The angle between the
spine and the medial border of the scapula (fron-
tal plane projection).
d) Internal rotation angle: The angle between the
frontal plane and a vector passing through the
root of the spine of the scapula and the posterior
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
angle of the acromion (transverse plane projec-
tion).
e) Posterior-tilt angle: The angle between a vector
paPningthrough and T7 and a passing
through the and the root of the
'pine the (sagittal plane projection).
To represent the humerus, we used the vector
formed between the olecranon and a derived point 2
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
cm directly inferior to the posterior angle of the
acromion. This point was derived by subtracting 2
cm from the y (vertical) component of the posterior
angle of acromion point. The angle between this vec-
tor, representing the humerus and the vertical, was
used to calculate the angle of maximum humeral ele-
vation.
Reliability
Journal of Orthopaedic & Sports Physical Therapy®
A preliminary study was conducted on 21 nonim-
paired subjects to determine the reliability of these
measurements. Each subject was measured in a stabi-
lized, seated position with the arm at rest and also
while elevated 90 degrees in the scapular plane. First,
2 measurement trials were performed. The subject
was then allowed to stand up and walk about for sev-
eral minutes before being retested for 2 more trials
in a similar fashion. The average of the first 2 trials
was compared to the average of the third and fourth
trials to determine reliability. Because only x, y, and z
coordinates of the digitized points were saved at the
time of measurement (no angular positions were cal-
culated), the examiner was not biased by the results
of the first 2 trials while performing the third and
fourth trials. The intraclass correlation coefficients
(ICC 3,l) for the scapular variables ranged from 0.88
to 0.99 and the standard error of measurement was
always less than 2 degrees, indicating satisfactory reli-
ability.'=
J Onhop Sports Phys ThereVolume 29 e Number 10 October 1999
Data Analysis
Means, standard deviations, and ranges were calcu-
lated for all dependent variables: posterior tilting an-
gle, upward rotation angle, internal rotation angle,
medial-lateral position, and superior-inferior position
for each of the 3 arm positions. Comparisons be-
tween the nonimpaired subjects, the asymptomatic
side of patients with impingement, and the sympto-
matic side of patients with impingement were made
using separate 1-way analyses of variance (ANOVA)
for each kinematicdependent variable. Post hoc
analyses were performed using a Tukey's procedure
when a significant difference was found with ANO-
VA. An alpha level of .05 was used for all tests.
RESULTS
Data for scapular angles and positions are shown
in Figures 2 through 6. Prior to our statistical analy-
sis, we evaluated the patterns of scapular motion
visually. The scapula showed a general pattern of in-
creasing posterior-tilt angle (Figure 2), increasing u p
rotation angle (Fiere 3), and decreasing inter-
nal rotation angle ( ~ i G r 4)
e in both the impinge-
ment and nonimpaired groups during scapular plane
elevation of the arm. In addition, the scapula moved
to a more superior position (Figure 5) and a slightly
more medial position (Figure 6) with increasing arm
elevation.
We found significant differences between groups
for posterior tilting at the horizontal and maximal el-
evation positions. Post hoc testing revealed that the
impingement group demonstrated significantly less
posterior tilt at the horizontal and maximal elevation
positions compared to the asymptomatic side and
nonimpaired subjects (Figure 2). No statistically sig-
nificant differences were found between groups on
the measures of scapular upward rotation angle and
internal rotation angle (Figures 3 and 4).
Significant differences were found between groups
for superior-inferior position at the horizontal and
maximum positions of arm elevation. Post hoc test-
ing showed that the symptomatic and asymptomatic
sides of the impingement group demonstrated signif-
icantly greater superior translation at the horizontal
and maximum positions of arm elevation compared
to nonimpaired comparisons (Figure 5). The differ-
ences between the asymptomatic and symptomatic
sides of patients with impingement were not signifi-
cant. Comparisons of medial-lateral position between
groups failed to reveal statistically significant differ-
ences (Figure 6).
The mean (2SD) maximal active humeral eleva-
tion in shoulders with impingement was 126.4"
( 2 13.5) on the symptomatic side and 125.7" ( 2 12.7)
on the asymptomatic side, which is significantly less
compared to 139.5" (212.4) in nonimpaired compar-
577
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

FIGURE 1. Scapular position and orientation were determined by digitizing the position of the following skeletal landmarks: seventh cervical spinous
process (C7), the seventh thoracic spinous process g7), the root of the spine of the scapula, the posterior angle of the acromion, and the inferior angle
of the scapula.
A) Medial-lateralposition: Calculated as the horizontal distance between C7 and the centroid of the scapula. lncreasing values represent lateral trans-
lation.
B) Superior-inferior position: Calculated as the vertical distance between C7 and the centroid of the scapula. lncreasing values represent inferior
translation.
-.-. .- .-..-. ..

C) Upward rotation angle: Scapulothoracic angle was the angle between the spine and the medial border of the scapula (projected onto the frontal
plane). lncreasing values represent upward rotation. Total arm elevation is the angle between the spine and vector connecting the olecranon and a
derived point 2 cm directly inferior to the posterior angle of the acromion.
Journal of Orthopaedic & Sports Physical Therapy®

Dl Scapular internal rotation angle: The angle between the frontal plane and a vector passing through the root of the spine of the scapula and the
posterior angle of the acromion (projected onto the transverse plane). lncreasing values represent internal rotation.
E) Scapular posterior tilt angle: The angle between a vector passing through C7 and T7 and a vector passing through the inferior angle and the root of
the spine of the scapula (projected onto the sagittal plane). lncreasing values represent posterior tilting.

ison subjects. These absolute values underestimate
the total elevation motion caused by the choice of
landmarks used for digitization, as the olecranon typ
ically was well below the center of the distal humerus
and the humeral head could not be digitized direct-
ly. However, the values are still useful for relative
comparisons. Subjects with impingement had a mean
pain score on the SPADI of 42.7/100 (227.0) with a
range of 0.7 to 90. The mean SPADI disability score
was 30.4/100 (228.5) with a range of 0 to 80.
DISCUSSION
A limitation of this study is that only statically held
positions were studied rather than continuous m e
tion, which may not represent functional movement
patterns. However, the general pattern of motion
found between successive positions of elevation in
this study agrees with other studies where scapular
motion has been as~essed.'~J~ The primary abnormal-
ities found in the shoulders with impingement were
less posterior tilting and excessive superior transla-
tion during scapular plane elevation. The relative
lack of posterior tilting could produce excessive com-
pression in the suprahumeral space as the greater tu-
berosity approaches the anterior aspect of the a c r e
mion. This relative anteriorly tilted position of the
scapula could contribute to impingement. This type
of subtle motion abnormality is probably difficult to
assess in a clinical examination. A lack of posterior
tilting could be caused by an excessively tight pectora-
lis minor muscle or a general lack of scapular mobili-
ty. As a clinical test, Kendallg suggests assessing pec-
toralis muscle-length by having a subject lie supine

J Orthop ~ p o r t s ' ~ Ther

h ~ s .Volume 29. Number 10. October 1999
 Posterior Tilt Angle
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.

Rest Horizontal Max

FIGURE 2. Posterior tilt angle at all 3 arm positions for each group. lncreasing values represent a more posteriorly tilted orientation. Error-ban represent
standard deviation.
Impingement side less posteriorly tilted than nonimpaired subjects and asymptomatic side (P< .05).
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

and noting how high off the table the shoulder gir-
dle rests. Another possible cause of decreased poste-
rior tilting may be inadequate muscle activity, partic-
ularly of the inferior portion of the serratus anterior,
which has been suggested by Perry.14 Based on these
theories, exercise programs for patients with im-
pingement should probably include stretching of the
pectoralis minor as well as strengthening for the ser-
ratus anterior.
We also found excessive superior translation of the
scapula in the impingement subjects. This finding
agrees with the work of Warner,20who used Moire to-

Upward Rotation Angle

Journal of Orthopaedic & Sports Physical Therapy®

Rest Horizontal Max

FIGURE 3. Upward rotation angle at all 3 arm positions for each group. lncreasing values represent a more upwardly rotated orientation. Error-ban
represent standard deviation. No significant differences were found between groups.

J Orthop Sports Phys Ther.Volurne 29. Number 10. October 1999

 Internal Rotation Angle
65
60 T

55 - T T I T I
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.

Rest Horizontal Max

FIGURE 4. Internal rotation angle at all 3 arm positions for each group. Decreasing values represent a more externally rotated orientation. Error-ban
represent standard deviation. No significant differences were found between groups.

pography to document increased scapular elevation humeral motion because of capsular tightness may
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

and scapular winging in subjects with impingement.
Presumably, increased scapular elevation is a com-
pensatory pattern that may be secondary to weakness
and inadequate firing of the intrinsic glenohumeral
muscles (deltoid and rotator cuff). Restricted glen*
also increase scapular elevation. Interestingly, both
the symptomatic and asymptomatic sides of patients
with impingement demonstrated a more superior p*
sition compared to nonimpaired individuals. Perhaps
this reflects a more central phenomenon such as a

-
Superior Inferior Position
Journal of Orthopaedic & Sports Physical Therapy®

I 0 Non-Impaired

Impingement:
Symptomatic
Impingement:
Asvm~tomatic

Rest Horizontal Max

FIGURE 5. Superior-inferior position for all 3 arm positions for each group. Decreasing values represent a more superior position. Error-bars represent
standard deviation.
* Symptomatic and asymptomatic sides of subjects with impingement positioned more superior than nonimpaired subjects (P < .05).

J Orthop Sports Php Ther-Volume 29. Number 10. October 1999

 -
Medial Lateral Position
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
Rest Horizontal
FIGURE 6. Medial-lateral position for all 3 arm positions for each group. Decreasing values represent a more medial position. Error-ban represent
standard deviation. No significant differences were found between gr&qi
preferential motor pattern or generalized capsular
tightness. Babyar2 also found excessive scapular eleva-
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
tion in a group of subjects with a history of various
shoulder problems, although she only studied the
symptomatic side. Her study suggested that the
amount of scapular elevation could be reduced after
a single session, emphasizing motor control of shoul-
der p o s i t i ~ n . ~
Many authors have previously stated that shoulder
impingement may be caused by decreased scapular
upward r~tation.~.~." If this were true, the subacromi-
Journal of Orthopaedic & Sports Physical Therapy®
a1 space would become more narrow as the greater
tuberosity moved closer to the acromion, and the
soft tissues that lie in this space might be pinched.
That explanation is not supported by our data be-
cause no significant difference was noted for scapu-
lar upward rotation angle between or within groups.
Subjects with impingement had slightly less upward
rotation compared to the nonimpaired group or the
asymptomatic side. Because we used a relatively small
sample with substantial measurement variation, the
lack of statistical significance may be due to a lack of
power (type I1 error). We considered a 5" difference
between groups to be clinically meaningful. Using
the obtained standard deviation (9.2), the power in
this study was 0.42 to detect a 5" difference between
groups (a,= .05) in scapular upward rotation angle.
A sample size of 50 subjects per group would have
been required to achieve a power level of 0.80 to de-
tect a 5" difference in upward rotation angle between
groups. Despite a potential type I1 error, we do not
believe a difference of less than 5" between groups,
as found in this study, is clinically important.
J Orthop Sports Phys Ther .Volume 29. Number 10. October 1999
Impingement:
Symptomatic
Impingement:
&vmtomatic
Max
Detailed inspection of our data revealed that in
the nonimpaired comparison group, the left side was
more upwardly rotated at rest (14.8" 2 8.7) when
compared to the right (8.5" 2 4.9). However, the
amount of upward rotation from the rest position to
maximum elevation was not statistically different be-
tween sides. When assessing scapular motion patterns
clinically, it is common practice to compare the
symptomatic side to the asymptomatic side. Our data
suggest that this practice is sound, given that the t c ~
tal excursion from rest to maximal elevation was sim-
ilar between sides in the nonimpaired subjects and
on the asymptomatic side of impingement subjects.
However, the rest position must be considered be-
cause of potential differences between right and left
sides. Our data suggest that the right side may tend
to start less upwardly rotated. All subjects were right-
hand dominant except for 5 subjects in the compari-
son group. These subjects were not obviously differ-
ent from the right-hand dominant subjects, but our
sample of left-hand dominant subjects was too small
to draw conclusions regarding the effect of limb
dominance on the patterns of scapular motion.
A difficulty in measuring scapular motion is the
lack of discrete landmarks for palpation. The bony
landmarks chosen in this study were occasionally dif-
ficult to palpate. As the arm elevation increased and
the musculature became more active, the landmarks
became increasingly more difficult to palpate. The
landmarks that were most difficult to locate were the
inferior angle of the scapula and the posterior angle
of the acromion at levels greater than 90" of arm ele-
vation. The amount of subcutaneous fat on a particu-
 lar subject also influenced the palpation accuracy.
Subjects were also required to actively hold the arm
in space for 30-60 seconds while the digitization
took place. As the subject's shoulder musculature be-
gan to fatigue, the arm elevation may have de-
creased.
Our reliability coefficients suggest good consisten-
cy, and our data regarding the amount and general
pattern of scapular motion are similar to other stud-
ies,1°.15which helps validate our method. Ultimately,
our method of measuring scapular motion should be
validated against radiographic measures or methods
where the measurement tool is directly fixed to the
bone.
Ludewig et al used a similar digitizing instrument
to document Sdimensional scapular motion in 25
asymptomatic subjects.1° They found the same gener-
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
al pattern of motion for scapular angles with eleva-
tion that we have reported, although they found
slightly different amounts of motion. Their mean val-
ues for total scapular motion (0" humeral elevation
to 140" humeral position) were as follows: 15" poste-
rior tilting, 34" upward rotation, - 13" internal rota-
tion. We found mean ( S D ) values of 22.8" (28.9)
posterior tilting (Figure 2), 28.2" (28.4) upward ro-
tation (Figure 3), and -7.1" (27.8) internal rotation
(Figure 4) for the same motions in nonimpaired sub-
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
jects. We believe one explanation for the difference
in these values lies in the different methods used to
determine and control humeral position. Ludewig et
a1 only had subjects elevate to 140" of humeral eleva-
tion as measured by a gravity-referenced pendulum
potentiometer attached to a humeral cuff whereas
our subjects raised their arms to maximum. Despite
our relatively low values for humeral elevation, our
subjects probably raised their arms higher than those
Journal of Orthopaedic & Sports Physical Therapy®
in the Ludewig et al study because our humeral ele-
vation angles are most likely to be artificially low. Lu-
dewig et all0 also required subjects to begin at a true
0" position whereas our subjects simply assumed a
rest position that was probably slightly abducted. Our
initial values for scapular upward rotation and inter-
nal rotation angle are approximately 10" higher than
those reported by Ludewig et al.1° Another differ-
ence is that we projected vectors onto various planes
to calculate angles whereas Ludewig et all0 used Eu-
ler angles based on an axis system embedded within
the scapula.
Although the mean (2SD) pain rating for subjects
with impingement was 42.7 (227.0) and the average
disability rating was 30.4 (228.5), the data show a
high degree of variation. The range of pain scores
was from 0.06 to 90 and the range of the disability
scale was 0 to 80. With this variability, the impinge-
ment group may have had subgroups that could r e p
resent a spectrum of severity from a full rotator cuff
tear to various degrees of inflammation of the tissue
in the subacromial space. Increasing the number of
impingement subjects to allow a breakdown of the
group into subgroups, or using a minimum score for
pain and disability, may have narrowed the distribu-
tion of data and increased the likelihood of showing
a significant difference in scapular motion between
groups.
CONCLUSIONS
We compared Sdimensional scapular position and
orientation between subjects with impingement syn-
drome and nonimpaired subjects at 3 successive stat-
ic positions: rest, elevation to horizontal, and maxi-
mal elevation. The general pattern of scapular mo-
tion between the 3 statically held positions was p r e
gressive upward rotation, external rotation, and
posterior tilting. The scapula also translated superior-
ly and slightly medial during arm elevation. The
chief abnormalities identified in the symptomatic
shoulders of subjects with impingement were lack of
posterior tilting and excessive superior translation.
These differences may be a consequence of impinge-
ment rather than a causal factor. The findings of this
study may have implications for the nonsurgical treat-
ment of impingement syndrome.
ACKNOWLEDGMENTS
We thank John Ashenbrenner, PT,for his assis-
tance in obtaining appropriate patients for this study
and Dr Andrew Karduna for his editorial assistance.
REFERENCES
Altchek DW, Warren R, Wickiewicz TL, Skyhar MJ, Ortiz
G, Schwartz E. Arthroscopic acromioplasty: technique
and results. j Bone joint Surg Am. 1990;72A:1198-1207.
Babyar S. Excessive scapular motion in individuals recov-
ering from painful and stiff shoulders: causes and treat-
ment strategies. Phys Ther. 1996;76:226-238.
Cailliet R. Neck and Arm &in. 3rd ed. Philadelphia, Pa:
FA Davis; 1991.
Culham EG, Peat M. Functional anatomy of the shoulder
complex. ) Orthop Sports Phys Ther. 1993;18:342-350.
Greenfield B, Catlin PA, Coats PW, Green E, McDonald
JJ, North C. Posture in patients with overuse injuries and
healthy individuals. j Orthop Sports Phys Ther. 1995;21:
287-295.
Hawkins RJ, Kennedy JC. Impingement syndrome in ath-
letes. Am j Sports Med. 1980;8:151-158.
Kaebeatse M, McClure PW, Pratt N. The effect of thoracic
posture on scapular position, shoulder elevation muscle
force and shoulder elevation range of motion. Arch Phys
Med Rehabil. 1999;80:945-950.
Kelly MJ, Clark WA. Orthopedic Therapy of the Shoulder.
Philadelphia, Pa: JB Lippincott Co; 1995:65-103, 231-
232.
Kendall FP, McCreary EK, Provance PG. Muscles-Testing
and Function. 4th ed. Baltimore, Md: Williams & Wilkins;
1993.
Ludewig PM, Cook TM, Nawoczenski DA. Three-dimen-
sional scapular orientation and muscle activity at selected
J Orthop Sports Phys Ther-Volume 29. Number 10.0ctober 1999
 positions of humeral elevation. / Orthop Sports Phys Ther.
1996;24:57-65.
Master FA, Arntz CT. Subacromial impin ement. In: Rock-
(i
wood C, Matsen F, e d ~ .The Shoulder.Vo 2. Philadelphia,
R: WB Saunders; 1990.
Michener L, Silfies S, Hutchinson D, McClure P. Accuracy
of the metrecom skeletal analysis system for linear and
angular measurements [abstract]. 1 Orthop Sports Phys
Ther. 1997;25:69. Abstract 29.
Neer CS II. Anterior acromioplasty for chronic impinge-
ment in the shoulder: a preliminary report. J Bone Joint
Surg Am. 1972;54A:41.
Perry J. Muscle control of the shoulder. In: Rowe CR, ed.
The Shoulder. New York, NY: Churchill Livingstone Inc;
1988:l-17.
Poppen NK, Walker PS. Normal and abnormal motion of
the shoulder. / Bone Joint Surg Am. 1976;58A:195-201.
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Journal of Orthopaedic & Sports Physical Therapy®
J Orthop Sports Phys There Volume 29. Number 10 October 1999
16. Portney LC, Watkins MP. Foundations of Clinical Re-
search: Applications to Practice. East Norwalk, Conn: Ap-
pleton & Lange; 1993.
17. Roach KE, Budiman-Mak E, Songsiridej N, Lertratanakul
Y. Development of a shoulder pain and disability index.
Arthritis Care and Research. 1991;4:143-149.
18. Solem-Bertoft E, Thuomas K, Westerberg C. The influence
of scapula retraction and protraction on the width of the
subacromial space. Clin Orthop. 1993;296:99-103.
19. Tibone JE, Shaffer B. Shoulder pain: when is it impinge-
ment?/ Musculo Med. 1995;12(4):65-77.
20. Warner JP, Micheli LJ, Arslanian LE, Kennedy J, Kennedy
R. Scapulothoracic motion in normal shoulders and
shoulders with glenohumeral instability and impingement
syndrome: a study using Moir6 topographic analysis. Clin
Orth. 1992;285:191-199.
583
 Invited Commentary
The article by Lukasiewicz and her colleagues6r e p
resents the first published work that demonstrates re-
liable and quantifiable Sdimensional scapular posi-
tion and orientation in a specific, clinically defined
population. This work is important and clinically rel-
evant, particularly in light of our current health care
environment that demands maximum outcomes with
minimal intervention.
One primary finding reported by Lukasiewicz et
a16 is the significantly decreased scapular posterior
tilting on the affected side in members of the im-
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
pingement group at elevated arm positions. The au-
thors suggest that this could contribute to compres-
sion and impingement beneath the anterior acromi-
on, one of the primary sites of impingement. Be-
cause posterior tilting of the scapula elevates the
anterior acromion, this motion may be more critical
than upward scapular rotation for obtaining ade-
quate clearance of the rotator cuff tendons under
the acromial surface during elevation of the arm. Ac-
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
romial contact with underlying tissues during humer-
al elevation always remains on the anterior undersur-
face of the a c r ~ m i o nFurthermore,
.~ surgical tech-
niques for impingement have evolved from lateral ac-
romioplasty to the current method of anterior
acromioplasty. Although upward scapular rotation re-
mains important to shoulder function, a lack of ade-
quate posterior tilting may create mechanical com-
pression beneath the anterior acromion. The finding
Journal of Orthopaedic & Sports Physical Therapy®
by Lukasiewicz et al of decreased posterior tilting in
subjects with impingement is consistent with results
of a similar investigation of symptomatic subjects5
If we accept a lack of adequate posterior tilting of
the scapula as an important variable contributing to
impingement symptoms, then we need methods to
assess and treat perceived dysfunction associated with
a lack of posterior tilting. Lukasiewicz et al note the
difficulty in assessing decreased posterior tilting in a
clinical examination. Beyond examining for extreme
dysfunction of the inferior angle moving posteriorly
off the thorax during elevation of the arm (anterior
tipping), clinical assessment of this motion has not
been thoroughly described. As we try to improve in
patient evaluation, I believe we need to move beyond
assessing only "rest position" of the scapula on the
thorax. Evidence accumulated from Lukasiewicz et
al, as well as from other recent studies4~%uggestthat
scapular position differences in subjects with im-
pingement are not significant when the arms are not
elevated. Regarding clinical treatment approaches, I
believe studies such as presented by Lukasiewicz et al
provide a scientific basis for refining clinical treat-
ment approaches beyond generalized exercises to im-
prove strength and flexibility of the scapular muscu-
lature. As the authors suggest, their investigations
supports exercise programs that incorporate pectoral-
is minor stretching and serratus anterior strengthen-
ing. Finding the best approach to activate and
strengthen the serratus anterior through a therapeu-
tic exercise program, however, remains a challenge.
The finding of excessive superior translation in the
subjects with impingement is more challenging to in-
terpret. It is interesting to note that the symptomatic
subjects demonstrated increased superior translation
on both the symptomatic and asymptomatic sides,
while the differences in posterior tipping were specif-
ic to the symptomatic side. In the absence of associ-
ated decreases in upward scapular rotation, could the
increased superior scapular motion be an attempt to
increase the available subacromial space? I realize
that this and other questions relating to cause and
effect relationships require speculation beyond avail-
able data. Longitudinal studies would allow us to de-
termine whether or not any kinematic patterns are
predictive of the subsequent development of im-
pingement symptoms.
Without performing a statistical comparison for
significance, demographic data for the 2 subject
groups suggest that the impingement group was a p
proximately 10 years older. Culham and Peat2 found
greater anterior tilting in the scapular "rest position"
in older women when compared to younger women.
Did Lukasiewicz et al assess age as a possible covari-
ate in their study? The symptomatic group consisted
of approximately 70% men, while the nonimpaired
group consisted of only 40% men. Given the associa-
tion between exposure to overhead activity and in-
creased prevalence of shoulder ~ymptoms,~ do the
authors know if the disproportion of men and wom-
en between the samples was related to exposure, or
some other factor?
Lukasiewicz et a1 also discuss the magnitude of an-
gular differences that might be clinically meaningful.
Previous anatomical studies have associated 3-5 de-
gree changes in acromial slope with an increased oc-
currence of rotator cuff tears and impingement syn-
drome.'~~ Additionally, biomechanical studies of simu-
lated shoulder elevation have reported soft tissue
contact of the rotator cuff tendons with the under-
surface of the acromion during elevation of the hu-
m e r u ~It
. ~seems conceivable that inflammation and
anatomical factors such as acromial shape or slope,
J Orthop Sports Phys Ther *Volume 29. Number 10 October 1999
 combined with even subtle decreases in the available
subacromial space due to kinematic changes, may
contribute to the initiation or progression of shoul-
der impingement syndromes.
Finally, the authors raise an interesting point regard-
ing the possibility of subgroups of subjects with im-
pingement within a "spectrum of severity." Subgroups
may also exist with respect to the site of impingement
Although the supraspinatus tendon is the most f i e
quently affected by impingement, any of the rotator
cuff tendons, as well as the long head of the biceps,
may be involved. In addition to the undersurface of
the acromion, several structures including the cora-
coacromial ligament, coracoid process, or undersurface
of acromioclavicularjoint may be the superior struc-
tures contributing to impingement Different levels of
symptom-severityand different sites of impingement
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
may relate to unique kinematic abnormalities, which
make it more difficult to ascertain overall group diier-
ences between asymptomatic and symptomatic subjects.
As our ability to identify subgroups of patients with im-
pingement improves, the understanding of relation-
ships between kinematic abnormalities and impinge-
ment syndromes, and ultimately the specificity of thera-
peutic interventions, should continue to evolve.
I congratulate the authors on the publication of
this interesting article and their use of a sound, sci-
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
entific approach to address a question of substantial
clinical relevance. I look forward to additional stud-
ies that might refine clinical practice for the treat-
ment of shoulder pain.
Paula Ludewig, PhD, F T
Assistant Professor
Journal of Orthopaedic & Sports Physical Therapy®
Author Response
We thank Dr Ludewig for her comments on our
findings. Our symptomatic and comparison groups
were dissimilar regarding age and sex, which may
have affected our results. We are currently conduct-
ing a trial where subjects with impingement are be-
ing matched with asymptomatic subjects by age and
sex for comparison. We reanalyzed the data related
to posterior tilting and did not find age to be a signif-
icant covariate (ie, adjusting for age differences did
not change our results). As Dr Ludewig points out,
Culham and Peat found more anterior tilting in old-
er women compared to younger women.' They did
not, however, account for differences in thoracic pos-
ture. Because we studied younger subjects, con-
strained thoracic posture by strapping the trunk to a
J Onhop Sports Phys Ther.Volume 29 Number 10. October 1999
Program in Physical Therapy
Box 388 Mayo
The University of Minnesota
Minneapolis, MN 55455
REFERENCES
1. Aoki M, lshii S, Usai M. Clinical appl~cationfor measuring
the slope of the acromion. In: Post M, Morrey B, Hawkins
R, eds. Surgery of the Shoulder. St Louis, Mo: Mosby-Year-
book; 1990:20&203.
2. Culham E, Peat M. Functional anatomy of the shoulder
complex. / Orthop Sports Phys Ther. 1993;18:342-350.
3. Flatow EL, Soslowsky LJ, lcker JB, et al. Excursion of the
rotator cuff under the acromion: patterns of subacromial
contact. Am / Sports Med. 1994;22:779-788.
4. Greenfield B, Catlin PA, Coats PW, Green E, McDonald JJ,
North C. Posture in patients with overuse injuries and
healthy individuals. / Orthop Sports Phys ner. 1995;21:
287-295.
5. Ludewig PM, Cook TM. Shoulder kinematics in persons
with shoulder impingement symptoms [abstract]. I Orthop
Sports Phys Ther. 1998;27:68.
6. Lukasiewicz AC, McClure P, Michener L, Pratt N, Sennett
B. Comparison of 3-dimensional scapular position and
orientation between subjects with and without shoulder
impingement. / Orthop Sports Phys Ther. 1999;29:574-
583.
7. National Institute for Occupational Safely and Health.
Shoulder musculoskeletal disorders: evidence for work-re-
latedness. In: Bernard BP, ed. Musculoskeletal Disorders
and Workplace Factors: A Critical Review of Epidemiologic
Evidence for Work-Related Musculoskeletal Disorders of
the Neck, Upper Extremi?: and Low Back. 2nd ed. Cin-
cinnati, Ohio: US Department of Health and Human Ser-
vices, Public Health Service, Centers for Disease Control
and Prevention; l997:1-72.
8. Zuckerman JD, Kummer FJ, Cuomo F, Simon J, Rosenblum
S, Katz N. The influence of coracoacromial arch anatomy
on rotator cuff tears. / Shoulder Elbow Surg. 1992;1:4-14.
fixed surface, and referenced scapular position to
the thorax, our measurements are substantially differ-
ent than theirs. If scapular position is measured rela-
tive to a global coordinate system (as in Culham and
Peat') rather than using the thoracic spine as a posi-
tional reference, thoracic flexion will result in great-
er anterior tilt as it will tilt the scapula forward along
with the thorax. We recently published a study docu-
menting the effects of thoracic posture on scapular
kinematics during humeral elevation and we con-
sider thoracic posture an important variable influenc-
ing scapular motion and shoulder f u n c t i ~ n . ~
The higher proportion of men in the symptomatic
group could be related to greater exposure to over-
head activities. Both groups, however, represented
585
 samples of convenience rather than true random
samples. While exposure to overhead activity is clear-
ly a risk factor, we are unaware of data documenting
greater exposure in men, although this is a possibili-
ty. In all likelihood, exposure to overhead activities is
directly related to specificjob duties, which we did
not account for in our study.
We strongly agree with the need to assess scapular
position and orientation during motion as opposed
to just resting posture. The difficulty lies in how to
accurately assess scapular motion with simple tech-
niques that are feasible for routine clinical use. Ease
of measurement is probably one reason resting p o s
ture is frequently measured despite the lack of docu-
mented association with shoulder symptoms.
It is possible that the increased scapular elevation
in the symptomatic group represents an attempt to
Downloaded from www.jospt.org at USD Univ Libraries on June 14, 2014. For personal use only. No other uses without permission.
increase subacromial space, yet our belief is that this
represents muscular substitution for rotator cuff
weakness. In other work, we found a significant in-
crease in scapular elevation when subjects raised
their arms from a flexed thoracic posture compared
to a more erect p o ~ t u r eWe
. ~ attribute this finding to
substitution from the trapezius and levator scapula
due to excessive shortening (active insufficiency) of
the deltoid and supraspinatus induced by the flexed
Copyright © 1999 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
thoracic position. In another study, we found a signif-
icant decrease in scapular elevation after a &week
strengthening program in asymptomatic subjects
judged to have a protracted shoulder posture.' The
Journal of Orthopaedic & Sports Physical Therapy®
strengthening program emphasized shoulder exter-
nal rotation and scapula retraction. We are preparing
to conduct a trial where we will measure scapular
kinematics before and after a rehabilitation program
in patients with impingement. This may help address
the issue.
We thank Dr Ludewig for highlighting some im-
portant issues related to the understanding of im-
pingement and the complexities involved with study-
ing scapular kinematics. We are aware of her excel-
lent work in this areaSand look forward to contin-
ued dialogue on this topic.
Amy Cole Lukasiewicz, MS, PT
Philip McClure, PhD, PT, OCS
Lori Michener MEd, PT, ATC
Neal Pratt, PhD, PT
Brian Sennett, MD
REFERENCES
Culham E, Peat M. Functional anatomy of the shoulder
complex. I Orthop Sports Phys Ther. 1993;18:342-350.
Kebaetse M, McClure PW, Pratt NE. Thoracic position ef-
fect on shoulder range of motion, strength, and three-di-
mensional kinematics. Arch Phys Med Rehabil. 1999;80:
945-950.
Ludewig PM, Cook TM. Shoulder kinematics in persons
with shoulder impingement symptoms [abstract].1 Orthop
Sports Phys Ther. 1998;27:68.
Wang CH, McClure PW, Pratt NE. Stretching and strength-
ening exercises: their effect on three-dimensional scapular
kinematics. Arch Phys Med Rehabil. 1990;80:923-929.
J Onhop .Spo&i Phys Ther-Volume 29 Number 10-October 1999