J Shoulder Elbow Surg (2015) -, 1-9

www.elsevier.com/locate/ymse

Effects of exercise therapy for the treatment

of symptomatic full-thickness supraspinatus
tears on in vivo glenohumeral kinematics
R. Matthew Miller, MSa,b, Adam Popchak, MS, DPT, SCSc, Dharmesh Vyas, MD, PhDc,
Scott Tashman, PhDc, James J. Irrgang, PhD, PT, ATCc, Volker Musahl, MDa,b,c,
Richard E. Debski, PhDa,b,c,*

a
Orthopaedic Robotics Laboratory, University of Pittsburgh, Pittsburgh, PA, USA
b
Department of Bioengineering, Swanson School of Engineering, University of Pittsburgh, Pittsburgh, PA, USA
c
Department of Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, PA, USA

Background: The high incidence of rotator cuff disease combined with high failure rates for nonoperative
treatment of full-thickness rotator cuff tears underlines the importance of improving nonoperative manage-
ment of rotator cuff tears. The study objective was to assess changes in in vivo glenohumeral kinematics of
patients with a symptomatic full-thickness supraspinatus tear before and after a 12-week exercise therapy
program. It was hypothesized that successful exercise therapy would result in improved kinematics
(smaller translations and increased subacromial space).
Materials and methods: Five patients were recruited for the study and underwent dynamic stereoradiog-
raphy analysis before and after a 12-week exercise therapy protocol to measure changes in glenohumeral
joint translations and subacromial space during coronal plane abduction. Strength and patient-reported
outcomes (American Shoulder and Elbow Surgeons; Disabilities of the Arm, Shoulder and Hand; Western
Ontario Rotator Cuff Index) were also evaluated.
Results: After therapy, no subject went on to receive surgery. It was found that the contact path length of
the humerus translating on the surface of the glenoid was reduced by 29% from 67.2%
36.9% glenoid
height to 43.1%
26.9% glenoid height (P ¼ .036) after therapy. Minimum acromiohumeral distance
showed a small increase from 0.9
0.6 mm to 1.3
0.8 mm (P ¼ .079). Significant improvements in
strength and patient-reported outcomes were also observed (P < .05).
Conclusions: Successful exercise therapy for treatment of small full-thickness supraspinatus tears results
in improvements in glenohumeral joint kinematics and patient-reported outcomes through increases in
rotator cuff muscle strength and joint stability. This study may enable identification of prognostic factors
that predict the response of a patient with a rotator cuff tear to exercise therapy.

Support from the Department of Orthopaedic Surgery, the Department of *Reprint requests: Richard E. Debski, PhD, 408 Center for Bioengi-
Bioengineering, The Albert B. Ferguson, Jr., MD Orthopaedic Fund of The neering, 300 Technology Drive, Pittsburgh, PA 15219, USA.
Pittsburgh Foundation (M2012-0035), and the Pittsburgh Chapter of the E-mail address: [email protected] (R.E. Debski).
ARCS Foundation is gratefully acknowledged.
Institutional Review Board approval was provided by the University of
Pittsburgh: PRO1105031.

1058-2746/$ - see front matter Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees.
http://dx.doi.org/10.1016/j.jse.2015.08.048
2 R.M. Miller et al.

Level of evidence: Basic Science Study, Kinesiology.

Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees.
Keywords: Shoulder; rotator cuff tear; kinematics; exercise therapy; biomechanics; glenohumeral joint

Rotator cuff disease is a serious clinical issue, with a terms of decreased joint translation and greater subacromial
high incidence of degenerative rotator cuff tears in persons space and (2) smaller translations and larger acromio-
aged 50 years and older.46,48 Despite the high prevalence of humeral distance (AHD) after exercise therapy would
rotator cuff tears, controversy remains about the optimum correlate with improvements in shoulder strength and
method for management of rotator cuff injury.33 Surgical patient-reported measures of symptoms, activity, and
repair of rotator cuff tears is becoming increasingly participation.
common49 but can sometimes lead to retearing or other
complications, with varying failure rates ranging from 17%
to 69%, depending on initial tear characteristics.11,41 Materials and methods
Nonoperative management of rotator cuff tears remains
the recommendation for initial management of the patient Subject recruitment
with a degenerative rotator cuff tear17,18,24,25,31,32 but is also
associated with failure rates of up to 50%,2,18,24,25 resulting Five subjects (3 women and 2 men; mean age,
in additional temporal and monetary costs of late surgical 60.2
7.6 years; mean body mass index, 32.0
7.8) were
treatment. recruited for the study after providing written informed
It is unclear which factors are most important for consent. Subjects were included in the study if they were
choosing nonoperative vs. surgical treatment.9,43 Previous aged between 45 and 70 years with a symptomatic small,
clinical studies have investigated initial tear size and degenerative full-thickness rotator cuff tear isolated to the
change in tear size during exercise therapy as potential supraspinatus tendon. Subjects with previous shoulder
factors for defining which types of patients are likely to surgeries, injections within 3 months before study partici-
respond better with nonoperative management.13,25,31,40 pation, or exercise therapy within 2 years before the study
However, these studies do not assess changes in gleno- were excluded. Magnetic resonance imaging confirmation
humeral kinematics, which may relate to long-term changes of a supraspinatus tear with a greater percentage of muscle
in joint instability, rotator cuff impingement, and osteoar- than fat (Goutallier grade 2 or less) was also required for
thritis. Studies investigating changes in glenohumeral study eligibility. Subjects who were smokers or had work-
kinematics after a rotator cuff tear in a rat model have related or traumatic injury, diabetes mellitus, or severe
indicated that cuff tears lead to changes in joint kinematics capsular tightness (defined as internal or external rotation
and joint disease.15,38 Therefore, there is a need to better <30 ) were excluded from participating in the study.
understand the impact of glenohumeral joint kinematics as
a measure for determining the successful outcome of Exercise therapy protocol and assessments
nonoperative management of rotator cuff tears. Changes
in glenohumeral kinematics due to rotator cuff disease All 5 patients participated in a standard 12-week exercise
have been reported for both cadaveric26,35,42,44,47 and therapy program for nonoperative management of rotator
in vivo6,14,27 experimental conditions. In particular, previ- cuff tears, with a focus on restoring range of motion
ous in vivo studies have observed changes in glenohumeral (ROM) and strengthening of the rotator cuff and scapular
contact position and subacromial space after rotator muscles. Patients were treated with oral nonsteroidal anti-
cuff surgical repair7 and relationships between changes inflammatories as needed, but no additional use was made of
in kinematics and shoulder strength measured after corticosteroid injections to the subacromial space. This
rehabilitation.34 program consisted of 6 weeks of supervised exercise with a
Although rotator cuff surgery has been shown to single physical therapist (A.P.) for two 45- to 60-minute
improve the in vivo kinematics of the glenohumeral joint, sessions a week. This was followed by an additional 6 weeks
the effect of exercise therapy on the restoration of gleno- of a home exercise program with completion of a daily
humeral kinematics remains unclear. Therefore, the objec- home exercise log to ensure compliance of the patient. An
tive of this study was to assess the effects of a 12-week additional session at the 12-week time point was included
exercise therapy program on glenohumeral kinematics for for data collection purposes and to assess the patient’s status
patients with a symptomatic full-thickness supraspinatus after the completion of the home exercise sessions.
tear. It was hypothesized that (1) successful exercise ther- Initially, isometric and active ROM exercises were used
apy would result in improved glenohumeral kinematics in to strengthen the rotator cuff and scapular muscles.
Effects of PT on shoulder motion 3

Table I Protocol for 3 phases of exercise therapy program

Week 1: acute phase Weeks 2 and 3: transitional phase Weeks 4 to 6: advanced phase
PROM cane ER Horizontal adduction stretch Continued progression of flexibility and
strengthening from transitional phase
PROM cane IR IR towel stretch Proprioceptive neuromuscular facilitation
patterns
PROM supine flexion Sleeper stretch Lat pull down
ER at 0 with elastic resistance Rhomboids (retractions)
PROM standing extension IR at 0 with elastic resistance Pecs (press, flies)
AAROM supine cane ER ER at 90 with elastic resistance Deltoids (raises)
AAROM standing cane flexion Subscapularis hug with elastic resistance Closed kinetic chain protraction with
rhythmic stabilization
AAROM standing cane abduction Scapular plane abduction ADLs or sport-specific activities
AAROM wall climb/wall walk Prone row into ER
AAROM standing cane extension Prone T’s (horizontal abduction at 90 )
Isometric ER at 0 Prone Y’s (horizontal abduction at 120 )
Isometric IR at 0 Serratus protraction with forward flexion
Side-lying ER in pain-free ROM Wall pushup with plus
Prone glenohumeral extension with ER Lat pull down
Scapular plane abduction Rhythmic stabilization with manual
resistance
Scapular retraction Biceps curl
Manually resisted scapular movements Triceps push down
ER, external rotation; IR, internal rotation; PROM, passive range of motion; AAROM, active assistive range of motion; ADLs, activities of daily living.

Progressive resistance exercises were introduced once the infraspinatus), external rotation at 90 of abduction (biases
patient could actively move the shoulder through a full teres minor), scapular plane abduction (biases supra-
ROM without lag or increased pain. Exercise resistance and spinatus), and internal rotation at 90 of abduction
number of repetitions were determined on the basis of a (involves subscapularis and pectoralis muscles). Patients
modification of the daily adjustable progressive resistance performed 3 trials of each task, and the average value was
exercise program23 (Table I). The specific exercises to taken as a measure of isometric strength. At the same time
target each of the rotator cuff muscles (supraspinatus, points, patients filled out 3 patient-reported outcome mea-
infraspinatus, teres minor, and subscapularis) as well as the sures: the American Shoulder and Elbow Surgeons (ASES)
serratus anterior and middle and lower trapezius were shoulder rating scale20; the Western Ontario Rotator Cuff
selected on the basis of electromyographic evidence of (WORC) Index21; and the Disabilities of the Arm, Shoul-
maximal activity for each muscle.29,36,37 At the conclusion der, and Hand (DASH) outcome measure.16 The ASES,
of each exercise session, cold therapy was applied to WORC, and DASH are commonly used to assess patient-
minimize shoulder pain as needed on the basis of response reported outcomes for a variety of shoulder conditions
to exercise, pain level, and preference of the patient. The including rotator cuff tears. Reliability and validity of
patient’s pain during each session was measured with a 0 to ASES,30 WORC,10 and DASH3 scores for rotator cuff
10 numerical pain rating scale and was used by the physical disease have been previously demonstrated. Changes in the
therapist to adjust the exercise program on an individual ASES, WORC, and DASH scores were compared with
basis. minimal clinically important differences (MCIDs)19 for
During the exercise therapy sessions at 0 and 12 weeks, each outcome measure to determine if improvements were
isometric shoulder strength and patient-reported symptoms, clinically relevant.22,39
activity, and participation were assessed for each patient.
Isometric strength measurements were taken with a hand- Glenohumeral kinematics protocol and
held dynamometer (Lafayette Manual Muscle Testing assessments
System; Lafayette Instrument Company, Lafayette, IN,
USA) for 4 shoulder positions. The dynamometer was A previously established model-based tracking technique
placed distally on the forearm, just proximal to the wrist using dynamic stereoradiography (DSX)8 was used to
joint along the dorsal radius and ulna, in taking the measure glenohumeral kinematics during coronal plane
measurement. Maximal isometric strength was assessed for abduction before and after the exercise therapy program.
external rotation at 0 of coronal plane abduction (biases Before beginning the exercise therapy protocol, subjects
4 R.M. Miller et al.

and thus the required motion was performed more slowly

(same number of images captured during a 4-second period
at 25 Hz) to help minimize pain. For these patients, the
post-therapy trial was also captured at 25 Hz. The 3D
kinematics of the glenohumeral joint was obtained with a
well-established technique that has been previously vali-
dated for the glenohumeral joint.8 Briefly, volumetric
models for the humerus and scapula were generated from
the 3D computed tomography scan by segmenting each
bone from surrounding tissues using commercial software
(Mimics). Digitally reconstructed radiographs of the 3D
bone models were simulated using the known projection
geometry of the biplane imaging system. Six degree-of-
freedom pose of the bone models was manipulated using
custom image correlation tracking software until the digi-
tally reconstructed radiographs matched the DSX images
simultaneously in both views for every frame. The resulting
dynamic, 3D positions of each bone were combined to
generate glenohumeral kinematics. The 3D reconstructed
Figure 1 Dynamic stereoradiography system setup. The patient model for the humerus was assigned a local coordinate
sits with the glenohumeral joint centered under the focal point of system based on International Society of Biomechanics
the two x-ray sources with back straightened. A laser pointer is
standards.45 The local coordinate system of the scapula was
strapped to the patient’s hand and used with a strip of tape on a flat
modified from the International Society of Biomechanics
surface aligned with the patient’s sagittal plane to assist the patient
in visually maintaining the affected limb in the coronal plane standards to create a glenoid-based coordinate system with
throughout the path of motion. origin at the midpoint between the most anterior-posterior
(AP) points of the glenoid rim. Axes were aligned with
the most superior-inferior (SI) and AP points along the rim
underwent computed tomography (0.625  0.625  1.25 of the glenoid,5 with positive directions pointing superior
mm voxels) of the affected shoulder. These computed and anterior. Translations and rotations of the humerus with
tomography images were segmented using the Mimics 14 respect to the scapula were then calculated using a Euler
software package (Materialise, Leuven, Belgium) to rotation sequence (Y-X-Y)45 and expressed in the glenoid
generate 3-dimensional (3D), subject-specific models of the coordinate system, with an accuracy of
0.4 mm and
humerus and scapula. For collection of x-ray images using
0.5 .8 The glenohumeral contact center location was
the DSX system, subjects were seated with the affected estimated for each motion frame using the centroid deter-
glenohumeral joint positioned at the center of the system mined from the 3D distance between the 2 bone surfaces, as
(Fig. 1). Subjects performed 3 arm abduction trials starting previously described.1 Minimum AHD was calculated as
in a resting position with the affected arm at the side and the absolute smallest distance between points on the surface
then raising the arm in the coronal plane to the maximum of the humeral head (including greater and lesser tuberos-
attainable elevation during a period of 2 seconds. A ities) and the undersurface of the acromion within a
metronome was used to assist in timing of the motion, and 200-mm2 area defined by a specific region of interest on
subjects were allowed a maximum of 2 practice trials each bone.1,4
before collection of data. For each arm abduction trial, Kinematic variables of interest included the average
subjects performed 3 abduction/adduction cycles, with data contact center of the humeral head on the glenoid surface5
collected on the second cycle. Position in the coronal plane and the minimum AHD.6,7,14 From the glenohumeral con-
was maintained by use of a laser pointer strapped to the tact data, the contact path length (i.e., translation of the
hand on the affected arm; subjects were asked to keep the humerus on the glenoid through the range of abduction)
laser dot within a properly aligned vertical stripe of tape. was calculated as the change in frame-by-frame position of
Subjects were instructed to keep their backs straight and not the joint contact center and was normalized to glenoid
lean to compensate for loss of humerothoracic elevation height.7 Normalized ranges of AP and SI translation of the
due to the rotator cuff tear. joint were calculated as the largest difference in AP and SI
High-speed dynamic stereo x-ray images (75 kVp, position of the joint contact center over the entire contact
125 mA, 2 ms pulse width) of the glenohumeral joint were path length. The absolute minimum AHD and average
collected at 50 Hz during a 2-second period starting just minimum AHD across the full ROM were determined for
before the patient’s raising the arm and continuing until the each subject. Similarly, the absolute and average minimum
point at which the arm was maximally elevated. Two AHDs were also determined for the abduction range of
patients had severe shoulder pain at their first DSX session, 40 to 60 . This subset of the full ROM was chosen as an
Effects of PT on shoulder motion 5

Table II Individual differences in kinematics after 12 weeks of exercise therapy (average)

Kinematic Parameter Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Pre Post Pre Post Pre Post Pre Post Pre Post
       
Elevation range shared between trials 21 -56 34 -92 31 -51 44 -73 52 -77


Minimum AHD over full ROM (mm) 0.1 0.3 0.8 1.4 0.3 0.1 0.3 0.8 1.4 0.3
Minimum AHD from 40 to 60 abduction (mm) 0.1 0.3 0.7 1.4 0.3 0.0 0.8 1.6 1.6 2.0
Average AHD over full ROM (mm) 0.3 0.9 2.5 3.0 1.1 0.4 1.2 1.1 2.2 2.2
Average AHD from 40 to 60 abduction (mm) 0.2 0.5 1.1 1.7 0.5 0.3 0.8 1.5 1.8 2.2
SI contact center range (% glenoid SI height) 13.6 15.0 23.8 20.8 7.8 7.0 32.4 35.8 8.4 7.3
AP contact center range (% glenoid AP width) 5.7 4.6 4.7 4.5 1.1 0.9 3.4 4.5 1.2 1.3
Contact path length (% glenoid height) 53.2 23.9 104.8 81.4 26.0 28.3 107.5 61.2 43.2 20.5
Pre, before therapy; Post, after therapy; AHD, acromiohumeral distance; ROM, range of motion; SI, superior-inferior; AP, anterior-posterior.

additional parameter to measure AHD at lower joint angles sessions as determined by the supervising physical therapist
and provided a measure of the minimum AHD just as the (i.e., did not complete all exercises during scheduled ses-
greater tuberosity begins to roll underneath the acromion sions and did not follow the daily adjustable progressive
during abduction (as determined during the tracking pro- resistance exercise program as the other 4 subjects did).
cess). For each subject, the average values of these kine- However, this subject was satisfied with the clinical out-
matic variables across the 3 abduction trials were calculated comes and did not elect to undergo surgical repair after
and used for subsequent statistical analysis. completing the study. The maximum elevation attained
Because of the large variation in ROM between subjects during testing varied substantially between subjects,
resulting from shoulder pain and loss of rotator cuff func- resulting in different ranges of glenohumeral elevation for
tion, pretherapy and post-therapy comparisons were made comparison between subjects (Table II). Subjects 1 and 3
on an individual basis using the largest shared ROM had significant shoulder pain and were unable to achieve
between data collection sessions. For example, for a subject much greater than 50 of glenohumeral elevation at the
with a pretherapy ROM of 20 to 60 and a post-therapy pretherapy time point. Subject 2 had minimal shoulder pain
ROM of 20 to 80 , the shared ROM over which kinematic and nearly normal ROM and thus was able to achieve
variables were calculated was 20 to 60 . The 3-trial maximum elevation >90 before therapy. Subjects 4 and 5
average values for each kinematic variable were averaged had shoulder pain during testing at the pretherapy session
across all subjects both before and after therapy and were but were still able to attain elevations of at least 70 before
used for statistical analysis. Paired t tests were performed to therapy.
compare the contact center path length, path range, and All 5 subjects showed improvements in isometric
minimum AHD variables before and after completion of shoulder strength and patient-reported outcomes after
exercise therapy (0- and 12-week time points). Paired t tests 12 weeks of exercise therapy (Table III). Strength measures
were also used to compare the 4 strength measurements and showed significant increases of 54%, 31%, 74%, and 54%
shoulder questionnaire scores, with the exception of the for external rotation at 0 abduction (P ¼ .005), internal
ASES survey, which used a Wilcoxon signed rank test rotation at 0 abduction (P ¼ .036), external rotation at 90
because the data were nonparametric. Because of the small abduction (P ¼ .009), and scaption at 90 abduction
sample size, Spearman r correlation coefficients were (P ¼ .024), respectively. Average improvements in the
calculated to determine relationships between changes in ASES (P ¼ .043), DASH (P ¼ .047), and WORC (P ¼ .02)
the patient-reported outcome scores, changes in shoulder scores were all greater than the MCID for ASES (6.4),39
strength, and changes in kinematic variables from before DASH (10.2),39 and WORC (245.26)22 outcomes. At the
therapy to after therapy. The significance level for all tests individual patient level, all subjects showed improvements
was set at P < .05. in the ASES score, and 4 of 5 showed improvements in the
DASH and the WORC scores greater than the MCID for
each. The subject who did not show improvements in the
Results DASH or WORC scores greater than the MCID already had
very good scores before therapy (DASH, 9.2 vs. 4.2;
All 5 subjects successfully completed the 12 weeks of WORC, 613 vs. 388).
therapy, and none had sought surgical treatment at Overall, 4 of 5 subjects showed individual decreases
24 months of follow-up. One subject showed poor of at least 20% in glenohumeral contact path length
compliance with the home exercise protocol and did not after therapy. Path length decreased by an average of
give maximal effort during supervised exercise therapy 36% after 12 weeks of exercise therapy (before therapy,
6 R.M. Miller et al.

rotation at 0 abduction (r ¼ -0.10; P ¼ .87), external

Table III Average differences in clinical outcomes after
12 weeks of exercise therapy (average
standard deviation)
rotation at 0 abduction (r ¼ -0.50; P ¼ .39), or external
rotation at 90 abduction (r ¼ -0.70; P ¼ .19). The change
Clinical measure Before therapy After therapy in minimum AHD for the elevation range of 40 to 60 was

Strength for ER at 0 54.2
31.3 83.9
22.3 negatively correlated with strength measures for external
abduction (N) rotation at 0 abduction (r ¼ -0.90; P ¼ .037) but was
Strength for IR at 0 93
45.2 121.8
36.9 not significantly correlated with internal rotation at
abduction (N) 0 abduction (r ¼ -0.60; P ¼ .28), scaption at 90 abduction
Strength for ER at 90 46.7
34.1 81.2
20.3
(r ¼ -0.60; P ¼ .28), or external rotation at 90 abduction
abduction (N)
Strength for scaption at 40.1
27.4 61.9
20.2
(r ¼ -0.80; P ¼ .10). No correlations were observed
90 abduction (N) between changes in any kinematic variables with the
ASES score) 50.7
19.3 88.1
16.3 change in ASES, DASH, or WORC scores.
DASH score 35
18.2 5.6
3.4
WORC score 1198.8
347.6 344.8
393.2
ER, external rotation; IR, internal rotation; ASES, American Shoulder
Discussion
and Elbow Surgeons; DASH, Disabilities of the Arm, Shoulder, and
Hand; WORC, Western Ontario Rotator Cuff Index. The primary findings of this study were that (1) patients
All differences were statistically significant between therapy time showed improvements in patient-reported outcomes and
points.
) Data were nonparametric and used Wilcoxon signed rank test for
shoulder strength and (2) an overall decrease in the trans-
comparisons. lation of the glenohumeral joint occurred with a 12-week
exercise therapy program for treatment of a full-thickness
supraspinatus tear. The decrease in contact path length was
substantial, indicating greater joint stability during arm
67.2%
36.9% of glenoid height; after therapy, abduction and supporting the hypothesis that exercise
43.1%
26.9% of glenoid height; P ¼ .036; Fig. 2). therapy improves joint kinematics. This improvement in
However, whereas the overall distance traveled by the stability likely resulted from compensatory increases in
humerus on the glenoid surface was shorter after exercise rotator cuff and scapular stabilizer muscle strength brought
therapy, the overall SI and AP range of translation was on by the 12 weeks of exercise therapy. The finding that
similar (P ¼ .88 and P ¼ .89, respectively) to baseline isometric shoulder strength increased for all test positions
values. Before therapy, the SI range was 17.1%
10.0% indicates that exercise therapy was successful in improving
glenoid height and the AP range was 3.5%
2.3% glenoid muscle strength to compensate for loss of rotator cuff
AP width. After therapy, the SI and AP ranges were function, one of the primary goals of exercise therapy.12 In
17.2%
12.0% glenoid height and 3.2%
1.9% glenoid addition, whereas the contact path length was found to
AP width, respectively. decrease after exercise therapy, the SI and AP range of the
Three of 5 subjects showed increases of at least 0.4 mm contact path did not change. This shows that the excursion
in absolute and average minimum AHDs after exercise of the humerus on the glenoid surface remained within the
therapy compared with before therapy, although these same SI/AP area but traveled a greater distance within that
changes were not statistically significant in any of the area before exercise therapy, i.e., glenohumeral joint
subjects. The largest individual increase in subacromial translation is not more restricted after exercise therapy but
space observed over the full ROM was approximately translates less within a given space. This may indicate that
0.7 mm (Table II, patient 2). Over the full range of gle- if it is left untreated, joint instability could lead to greater
nohumeral elevation, the absolute minimum AHD was wear on joint cartilage, possibly inducing pain and leading
0.6
0.6 mm before therapy and 0.7
0.8 mm after to a higher risk of osteoarthritis.
therapy (P ¼ .31). The average minimum AHDs over the The contact path length results of this study differ from
full ROM before and after therapy were 1.5
0.9 mm and those of a previous study comparing changes in kinematics
1.5
1.1 mm, respectively (P ¼ .81). In considering only after rotator cuff repair with contralateral and healthy
the range between 40 and 60 , the absolute minimum shoulders using biplane radiography.7 No differences were
AHD before therapy was 0.7
0.6 mm, which increased to found in contact path length or SI contact range between
1.1
0.9 mm after therapy (P ¼ .099). Similarly, there was repaired and contralateral shoulders or between repaired
an increase from 0.9
0.6 mm to 1.3
0.8 mm for the shoulders and shoulders from healthy individuals, sug-
average minimum AHD between 40 and 60 (P ¼ .079). gesting that rotator cuff repair restores the glenohumeral
In comparing magnitude of changes in kinematic joint kinematics to nearly normal. However, the overall
parameters with the percentage strength increase between contact path length measured by Bey et al7 was only 20% to
0 and 12 weeks, the change in path length showed 25% of the glenoid SI height, which was 50% of the contact
nonsignificant negative correlations with strength measures path length measured after therapy in this study (and
for scaption at 90 abduction (r ¼ -0.10; P ¼ .87), internal approximately one third of the path length measured before
Effects of PT on shoulder motion 7

Figure 2 Representative contact path kinematics for a single patient before therapy (A) and after therapy (B). The contact center is
represented by a black circle, and the path followed by the contact center throughout the range of abduction is represented by the white line.
Qualitatively, there is a significant reduction in contact path length after therapy.

therapy). This may indicate that whereas exercise therapy therapy. No correlations were observed between kinematic
improves joint stability and reduces translation of the gle- parameters and patient-reported outcome scores, which is
nohumeral joint compared with pre-exercise levels, it does consistent with what has been previously reported.7
not restore kinematics to the levels of an intact or repaired A limitation of this study is the small sample size.
rotator cuff. Although only 5 patients were tested, all patients success-
In this study, any increases observed in subacromial fully completed the exercise therapy program and fit spe-
space after therapy were all submillimeter in magnitude cific recruitment criteria representing the general
and in a range that was similar to the accuracy of the population of individuals with degenerative rotator cuff
system (
0.4 mm). Therefore, any changes in subacromial tears that are treated with exercise therapy. Overall, the
space due to exercise therapy would likely not be clinically group chosen was sufficient to determine statistically sig-
relevant, indicating that exercise therapy does not appre- nificant differences in important kinematic parameters. Post
ciably increase the minimum AHD in the glenohumeral hoc power analysis indicated that a sample size of 12
joint. The absolute minimum AHDs measured in this study subjects would be required to achieve statistical signifi-
were found to be 2 to 3 mm smaller than those of healthy cance in the minimum AHD parameter, and 9 subjects
controls14 or individuals after rotator cuff repair,7 indicating would be required to achieve statistical significance in
that subacromial impingement may still occur in patients correlations between the strength parameters and path
after completion of exercise therapy, potentially leading to length. In addition, we did not measure tear size, so we are
later shoulder pain. unable to compare our results to previous studies that found
The finding that increases in minimum AHD were not changes in tear size during exercise therapy. Future studies
correlated with improvements in shoulder strength suggests will require expert ultrasound operators to track tear
that whereas exercise therapy for treatment of impingement propagation over time for comparison with changes in joint
syndrome often leads to improvements in patient-reported kinematics as a potential factor for determining the success
outcomes and reduction of pain,24 these improvements may or failure of exercise therapy. Last, the model tracking
not be due to increased subacromial space for the affected technique used does not account for cartilage on the hu-
rotator cuff tendons. However, additional studies with merus or glenoid surface, which can lead to overestimations
larger groups of patients are required to better assess re- of contact path length and range.28 However, because this
lationships between glenohumeral joint kinematics and study was primarily interested in differences in joint con-
changes in traditional patient outcomes after exercise tact from before to after exercise therapy, the lack of
8 R.M. Miller et al.

cartilage in the model should not have affected the main 4. Bey MJ, Brock SK, Beierwaltes WN, Zauel R, Kolowich PA,
findings of this study. Lock TR. In vivo measurement of subacromial space width during
shoulder elevation: technique and preliminary results in patients
following unilateral rotator cuff repair. Clin Biomech 2007;22:767-73.
http://dx.doi.org/10.1016/j.clinbiomech.2007.04.006
Conclusion 5. Bey MJ, Kline SK, Zauel R, Kolowich PA, Lock TR. In vivo mea-
surement of glenohumeral joint contact patterns. EURASIP J Adv Signal
Process 2010;2010:162136. http://dx.doi.org/10.1155/2010/162136
Despite exercise therapy’s not improving glenohumeral
6. Bey MJ, Kline SK, Zauel R, Lock TR, Kolowich PA. Measuring
joint kinematics to the level of healthy individuals or to dynamic in-vivo glenohumeral joint kinematics: technique and pre-
that of individuals after rotator cuff tear repair, this study liminary results. J Biomech 2008;41:711-4. http://dx.doi.org/10.1016/
shows that exercise therapy leads to satisfactory im- j.jbiomech.2007.09.029
provements in patient-reported outcomes, strength, and 7. Bey MJ, Peltz CD, Ciarelli K, Kline SK, Divine GW, van
Holsbeeck M, et al. In vivo shoulder function after surgical repair of a
glenohumeral contact path length. These changes are
torn rotator cuff: glenohumeral joint mechanics, shoulder strength,
generally sufficient to restore function and to minimize clinical outcomes, and their interaction. Am J Sports Med 2011;39:
shoulder pain to satisfactory levels for patients. How- 2117-29. http://dx.doi.org/10.1177/0363546511412164
ever, it is unknown if patient-reported outcomes remain 8. Bey MJ, Zauel R, Brock SK, Tashman S. Validation of a new model-
improved during the long term or revert to pretherapy based tracking technique for measuring three-dimensional, in vivo
glenohumeral joint kinematics. J Biomech Eng 2006;128:604-9. http://
levels. Therefore, our future directions are to determine
dx.doi.org/10.1115/1.2206199
the long-term effects of exercise therapy on gleno- 9. Boorman RS, More KD, Hollinshead RM, Wiley JP, Brett K,
humeral joint kinematics for tasks that mimic common Mohtadi NG, et al. The rotator cuff quality-of-life index predicts the
activities of daily living. Additional individuals with a outcome of nonoperative treatment of patients with a chronic rotator
rotator cuff tear, including those who fail to improve cuff tear. J Bone Joint Surg Am 2014;96:1883-8. http://dx.doi.org/10.
2106/JBJS.M.01457
with exercise therapy, must be studied to evaluate
10. de Witte PB, Henseler JF, Nagels J, Vliet Vlieland TP, Nelissen RG.
specific factors associated with success or failure of The Western Ontario rotator cuff index in rotator cuff disease patients:
exercise therapy. Our ultimate goal is to develop an a comprehensive reliability and responsiveness validation study. Am J
algorithm to identify those individuals with a rotator cuff Sports Med 2012;40:1611-9. http://dx.doi.org/10.1177/036354651
tear who will best respond to nonoperative treatment that 2446591
11. Duquin TR, Buyea C, Bisson LJ. Which method of rotator cuff repair
includes exercise therapy. This study provides pre-
leads to the highest rate of structural healing? A systematic review.
liminary data on the use of in vivo glenohumeral kine- Am J Sports Med 2010;38:835-41. http://dx.doi.org/10.1177/03635
matics as a factor for predicting the successful outcome 46509359679
of exercise therapy for the treatment of rotator cuff tears. 12. Escamilla RF, Yamashiro K, Paulos L, Andrews JR. Shoulder muscle
activity and function in common shoulder rehabilitation exercises.
Sports Med 2009;39:663-85. http://dx.doi.org/10.2165/00007256-
200939080-00004
13. Fucentese SF, von Roll AL, Pfirrmann CW, Gerber C, Jost B. Evo-
Disclaimer lution of nonoperatively treated symptomatic isolated full-thickness
supraspinatus tears. J Bone Joint Surg Am 2012;94:801-8. http://dx.
doi.org/10.2106/JBJS.I.01286
The authors, their immediate families, and any research 14. Giphart JE, van der Meijden OA, Millett PJ. The effects of arm
foundation with which they are affiliated have not elevation on the 3-dimensional acromiohumeral distance: a biplane
received any financial payments or other benefits from fluoroscopy study with normative data. J Shoulder Elbow Surg 2012;
any commercial entity related to the subject of this 21:1593-600. http://dx.doi.org/10.1016/j.jse.2011.11.023
15. Hsu JE, Reuther KE, Sarver JJ, Lee CS, Thomas SJ, Glaser DL, et al.
article.
Restoration of anterior-posterior rotator cuff force balance improves
shoulder function in a rat model of chronic massive tears. J Orthop Res
2011;29:1028-33. http://dx.doi.org/10.1002/jor.21361
16. Hudak PL, Amadio PC, Bombardier C. Development of an upper
extremity outcome measure: the DASH (disabilities of the arm,
References shoulder and hand) [corrected]. The Upper Extremity Collaborative
Group (UECG). Am J Ind Med 1996;29:602-8.
1. Anderst WJ, Tashman S. A method to estimate in vivo dynamic 17. Itoi E. Rotator cuff tear: physical examination and conservative
articular surface interaction. J Biomech 2003;36:1291-9. http://dx.doi. treatment. J Orthop Sci 2013;18:197-204. http://dx.doi.org/10.1007/
org/10.1016/S0021-9290(03)00157-X s00776-012-0345-2
2. Baydar M, Akalin E, El O, Gulbahar S, Bircan C, Akgul O, et al. The 18. Itoi E, Tabata S. Conservative treatment of rotator cuff tears. Clin
efficacy of conservative treatment in patients with full-thickness ro- Orthop Relat Res 1992:165-73.
tator cuff tears. Rheumatol Int 2009;29:623-8. http://dx.doi.org/10. 19. Jaeschke R, Singer J, Guyatt GH. Measurement of health status.
1007/s00296-008-0733-2 Ascertaining the minimal clinically important difference. Control Clin
3. Beaton DE, Katz JN, Fossel AH, Wright JG, Tarasuk V, Trials 1989;10:407-15.
Bombardier C. Measuring the whole or the parts? Validity, reliability, 20. King GJ, Richards RR, Zuckerman JD, Blasier R, Dillman C,
and responsiveness of the Disabilities of the Arm, Shoulder and Hand Friedman RJ, et al. A standardized method for assessment of elbow
outcome measure in different regions of the upper extremity. J Hand function. Research Committee, American Shoulder and Elbow Sur-
Ther 2001;14:128-46. geons. J Shoulder Elbow Surg 1999;8:351-4.
Effects of PT on shoulder motion 9

21. Kirkley A, Alvarez C, Griffin S. The development and evaluation of a 36. Reinold MM, Escamilla RF, Wilk KE. Current concepts in the sci-
disease-specific quality-of-life questionnaire for disorders of the ro- entific and clinical rationale behind exercises for glenohumeral and
tator cuff: the Western Ontario Rotator Cuff Index. Clin J Sport Med scapulothoracic musculature. J Orthop Sports Phys Ther 2009;39:105-
2003;13:84-92. 17. http://dx.doi.org/10.2519/jospt.2009.2835
22. Kirkley A, Griffin S, Dainty K. Scoring systems for the functional 37. Reinold MM, Wilk KE, Fleisig GS, Zheng N, Barrentine SW,
assessment of the shoulder. Arthroscopy 2003;19:1109-20. http://dx. Chmielewski T, et al. Electromyographic analysis of the rotator cuff
doi.org/10.1016/j.arthro.2003.10.030 and deltoid musculature during common shoulder external rotation
23. Knight KL. Knee rehabilitation by the daily adjustable progressive exercises. J Orthop Sports Phys Ther 2004;34:385-94. http://dx.doi.
resistive exercise technique. Am J Sports Med 1979;7:336-7. org/10.2519/jospt.2004.34.7.385
24. Kuhn JE. Exercise in the treatment of rotator cuff impingement: a 38. Reuther KE, Thomas SJ, Tucker JJ, Sarver JJ, Gray CF, Rooney SI,
systematic review and a synthesized evidence-based rehabilitation et al. Disruption of the anterior-posterior rotator cuff force balance
protocol. J Shoulder Elbow Surg 2009;18:138-60. http://dx.doi.org/10. alters joint function and leads to joint damage in a rat model. J Orthop
1016/j.jse.2008.06.004 Res 2014;32:638-44. http://dx.doi.org/10.1002/jor.22586
25. Kuhn JE, Dunn WR, Sanders R, An Q, Baumgarten KM, Bishop JY, 39. Roy JS, MacDermid JC, Woodhouse LJ. Measuring shoulder function:
et al. Effectiveness of physical therapy in treating atraumatic full- a systematic review of four questionnaires. Arthritis Rheum 2009;61:
thickness rotator cuff tears: a multicenter prospective cohort study. J 623-32. http://dx.doi.org/10.1002/art.24396
Shoulder Elbow Surg 2013;22:1371-9. http://dx.doi.org/10.1016/j.jse. 40. Safran O, Schroeder J, Bloom R, Weil Y, Milgrom C. Natural history
2013.01.026 of nonoperatively treated symptomatic rotator cuff tears in patients 60
26. Lee YS, Lee TQ. Specimen-specific method for quantifying gleno- years old or younger. Am J Sports Med 2011;39:710-4. http://dx.doi.
humeral joint kinematics. Ann Biomed Eng 2010;38:3226-36. http:// org/10.1177/0363546510393944
dx.doi.org/10.1007/s10439-010-0074-7 41. Seida JC, LeBlanc C, Schouten JR, Mousavi SS, Hartling L,
27. Massimini DF, Boyer PJ, Papannagari R, Gill TJ, Warner JP, Li G. In- Vandermeer B, et al. Systematic review: nonoperative and operative
vivo glenohumeral translation and ligament elongation during treatments for rotator cuff tears. Ann Intern Med 2010;153:246-55.
abduction and abduction with internal and external rotation. J Orthop http://dx.doi.org/10.7326/0003-4819-153-4-201008170-00263
Surg Res 2012;7:29. http://dx.doi.org/10.1186/1749-799X-7-29 42. Su WR, Budoff JE, Luo ZP. The effect of anterosuperior rotator cuff
28. Massimini DF, Warner JJ, Li G. Glenohumeral joint cartilage contact tears on glenohumeral translation. Arthroscopy 2009;25:282-9. http://
in the healthy adult during scapular plane elevation depression with dx.doi.org/10.1016/j.arthro.2008.10.005
external humeral rotation. J Biomech 2014;47:3100-6. http://dx.doi. 43. Tanaka M, Itoi E, Sato K, Hamada J, Hitachi S, Tojo Y, et al. Factors
org/10.1016/j.jbiomech.2014.06.034 related to successful outcome of conservative treatment for rotator cuff
29. McCann PD, Wootten ME, Kadaba MP, Bigliani LU. A kinematic and tears. Ups J Med Sci 2010;115:193-200. http://dx.doi.org/10.3109/
electromyographic study of shoulder rehabilitation exercises. Clin 03009734.2010.493246
Orthop Relat Res 1993:179-88. 44. Thompson WO, Debski RE, Boardman ND 3rd, Taskiran E,
30. Michener LA, McClure PW, Sennett BJ. American Shoulder and Warner JJ, Fu FH, et al. A biomechanical analysis of rotator cuff
Elbow Surgeons Standardized Shoulder Assessment Form, patient deficiency in a cadaveric model. Am J Sports Med 1996;24:286-92.
self-report section: reliability, validity, and responsiveness. J Shoulder 45. Wu G, van der Helm FC, Veeger HE, Makhsous M, Van Roy P,
Elbow Surg 2002;11:587-94. http://dx.doi.org/10.1067/mse.2002. Anglin C, et al. ISB recommendation on definitions of joint coordinate
127096 systems of various joints for the reporting of human joint motion–Part
31. Moosmayer S, Lund G, Seljom US, Haldorsen B, Svege IC, Hennig T, II: shoulder, elbow, wrist and hand. J Biomech 2005;38:981-92. http://
et al. Tendon repair compared with physiotherapy in the treatment of dx.doi.org/10.1016/j.jbiomech.2004.05.042
rotator cuff tears: a randomized controlled study in 103 cases with a 46. Yamaguchi K, Ditsios K, Middleton WD, Hildebolt CF, Galatz LM,
five-year follow-up. J Bone Joint Surg Am 2014;96:1504-14. http://dx. Teefey SA. The demographic and morphological features of rotator
doi.org/10.2106/JBJS.M.01393 cuff disease. A comparison of asymptomatic and symptomatic
32. Pappou IP, Schmidt CC, Jarrett CD, Steen BM, Frankle MA. AAOS shoulders. J Bone Joint Surg Am 2006;88:1699-704. http://dx.doi.org/
appropriate use criteria: optimizing the management of full-thickness 10.2106/JBJS.E.00835
rotator cuff tears. J Am Acad Orthop Surg 2013;21:772-5. http://dx. 47. Yamaguchi K, Sher JS, Andersen WK, Garretson R, Uribe JW,
doi.org/10.5435/JAAOS-21-12-772 Hechtman K, et al. Glenohumeral motion in patients with rotator cuff
33. Pedowitz RA, Yamaguchi K, Ahmad CS, Burks RT, Flatow EL, tears: a comparison of asymptomatic and symptomatic shoulders. J
Green A, et al. American Academy of Orthopaedic Surgeons Clinical Shoulder Elbow Surg 2000;9:6-11.
Practice Guideline on: optimizing the management of rotator cuff 48. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D,
problems. J Bone Joint Surg Am 2012;94:163-7. Shitara H, et al. Prevalence and risk factors of a rotator cuff tear in the
34. Peltz CD, Haladik JA, Hoffman SE, McDonald M, Ramo N, general population. J Shoulder Elbow Surg 2010;19:116-20. http://dx.
Moutzouros V, et al. Associations among shoulder strength, gleno- doi.org/10.1016/j.jse.2009.04.006
humeral joint motion, and clinical outcome after rotator cuff repair. 49. Zhang AL, Montgomery SR, Ngo SS, Hame SL, Wang JC,
Am J Orthop (Belle Mead NJ) 2014;43:220-6. Gamradt SC. Analysis of rotator cuff repair trends in a large private
35. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. insurance population. Arthroscopy 2013;29:623-9. http://dx.doi.org/
J Bone Joint Surg Am 1976;58:195-201. 10.1016/j.arthro.2012.11.004