Download Muscle and Joint Loading at the Shoulder During the Bear Hug

Muscle and Joint Loading at the Shoulder during the Bear Hug Rehabilitation Exercise
1
1
Yanagawa, T; +1Torry, M R; 1Shelburne K B, 1Hackett T R; 2Pandy, M G
Steadman♦Hawkins Research Foundation, Vail, CO, 2University of Melbourne, Melbourne, Victoria, AUS
Senior author: [email protected]
INTRODUCTION:
Rotator cuff tears are a common cause of shoulder pain and
dysfunction. Surgical repair attempts to restore the normal anatomic
relationship of the rotator cuff tendons as well as glenohumeral motion.
However, rotator cuff re-tear rates are reported as high as 90%1 with
most studies demonstrating rates from 25% to 35% 2,3. Despite several
reports detailing the potential causes for re-tear which includes anchor
and fixation strength, suture materials, suture pulling through tendon,
how these re-tears occur in vivo is not known.
Rehabilitation after rotator cuff repair consists of motion and
strengthening using several modalities including elastic resistive devices
(ERD), the success of which, may affect clinical outcomes. While many
of these exercises have been well described7, little is known regarding
the actual forces sustained across the glenohumeral joint or sustained by
a repair construct in vivo during these exercises
A number of experimental studies have investigated the contributions
of the individual shoulder muscles to glenohumeral joint stability during
various motions of the upper limb4,5. An important limitation of the in
vitro studies is that the loading applied to the cadavers are substantially
different from that which is present in vivo. The magnitudes of the
muscle forces applied in these experiments were also well below the
values estimated from mathematical models6.
In the present study, we used a computational model of the upper
limb to determine individual muscles forces as well as the contributions
of the individual shoulder muscles to glenohumeral joint loading during
the dynamic bear hug7 rehabilitation exercise. The specific aim was to
determine glenohumeral joint reaction forces as well as the relative
contributions of the rotator cuff muscles to glenohumeral joint load over
the range of the shoulder motion during this exercise.
METHODS:
A 3D computer model of the upper
Figure 1: Dynamic
limb was used to calculate muscle and
Bear Hug with ERD
joint-contact loading at the shoulder
starts with arms
during the bear hug exercise. The
abducted 60 degrees,
model and its construction has been
and the patient
describes previously8. In brief, eighteen
horizontally flexes
muscle bundles were used to represent
until maximal
the lines of action of 11 muscle groups
protraction is
spanning the glenohumeral joint. The
attained.
path of each muscle bundle was
calculated using the obstacle-set
method described by Garner and Pandy
(2001). 3D bone positions measured in
one healthy young male subject who
volunteered and signed informed
consent (age, 34 yrs; height, 179 cm;
weight, 81 kg) performing the Bear
Hug (Figure 1) exercise were used as
input to the model. Steinman pins were
inserted into the clavicle, scapula, and
humerus, and the three-dimensional
coordinates of markers mounted on the
pins were recorded (120 Hz) using a 5-camera motion capture system
(Motion Analysis Corp, Montview, CA). The relative positions of the
bones were measured from 15-75° of horizontal (humeral) adduction
with an external load applied by an elastic band7. Anatomical
differences between the subject and the VHM cadaver meant that the
measured bone positions had to be modified slightly to ensure that the
experimental results were compatible with the kinematic geometry of the
model. To prevent scapulothoracic penetration in the model, an
optimization problem was solved to minimize the differences between
the measured and calculated bone positions subject to two constraints:
(1) that the scapula moves smoothly over the surface of the thorax in the
model; and (2) that the length of the conoid ligament in the model
remains constant during abduction. An optimization problem was
solved to calculate the forces developed by the shoulder muscles for
each prescribed position of the model arm during the bear hug motion.
The problem was to minimize the sum of the squares of all muscle
stresses subject to two constraints: (1) that the arm remains in static
equilibrium at each prescribed angle of humeral abduction; and (2) that
the line of action of the resultant force acting between the humeral head
and the glenoid remains within the glenoid cavity. The optimization
solution was used to determine each muscle’s contribution to
glenohumeral joint loading during the motion. Specifically, each muscle
force vector was resolved into three components: an anterior-posterior
shear force, a superior-inferior shear force, and a compressive force.
RESULTS:
The Anterior Deltoid produced the largest force during the motion
(460 N). Of the cuff muscles, the infraspinatus produced a peak force of
275 N at 15° and at 75° and the subscapularis produced a peak force of
175 N at 15°. All other cuff muscles forces were neglible (<40 N). The
resultant and components of the resultant glenohumeral joint reaction
force are presented in Figure 2. Peak resultant glenohuemral joint
reaction force was 810 N at 15° and was comprised primarily of the
compressive force which exhibited a peak of 790 N at 15°. The largest
posterior shear (275 N) and inferior shear (175 N) forces occurred at
30°.
900
600
300
0
0
10
20
30
40
50
60
70
80
-300
Resultant
Compression
Anterior
Superior
Figure 2: Glenohumeral joint reaction forces
DISCUSSION:
This study estimated shoulder muscle forces during a common
rehabilitation exercise with a 3D computational model validated in
previously published works. Maximal infraspinatus muscle forces were
recorded as high as 275N during the bear hug exercise. Reported pull
out strengths of various suture anchors have ranged from 112-371 N9,10.
The results of this study illustrate the exceptionally high tensile forces
that may occur at the infraspinatus nearing or even potentially exceeding
the repair construct load to failure. Thus, the results of this study warrant
caution to be employed during the rehabilitation phase of rotator cuff
repairs and may bring into question the utility of commonly employed
exercises.
REFERENCES:
1. Calvert et al., (1986) JBJS Br, 68:147-50.; 2. Harryman et al., (1991)
JBJS Am, 73:982-89.; 3. Liu and Baker (1994) Arthroscopy, 23:99104.; 4. McMahon et al., (1995) JSES, 4:199-208.; 5. Thompson et al.,
(1996) Am J Sports Med, 24:286-292.; 6. van der Helm et al., (1994) J
Biomech., 25:527-550.; 7. Decker et al., (1999) AJSM, 27:6:784-791. ;
8. Garner. B.A., Pandy M.G. (2001). Comput Methods Biomech Biomed
Engin, 2, 107-124.; 9. Tingart et al., (2004). Am J Sports Med,
32(6):1466-1473.; 10. Kim et al., (2006). Am J Sports Med, 34, 40714.
ACKNOWLEDGMENTS: Supported by Steadman Hawkins Research
Foundation and the National Football League Charities.
Poster No. 1889 • 55th Annual Meeting of the Orthopaedic Research Society