Rehabilitation of the Medial

Collateral Ligament-Deficient Elbow:

An In Vitro Biomechanical Study
April D. Armstrong, MD, Cynthia E. Dunning, MSc,
Kenneth J. Faber, MD, Teresa R. Duck, BSc, James A. Johnson, PhD,
Graham J.W. King, MD, London, Ontario, Canada

The purpose of this study was to determine the relative contribution of muscle activity and the
effect of forearm position on the stability of the medial collateral ligament (MCL)-deficient
elbow. Simulated active and passive elbow flexion with the forearm in both supination and
pronation was performed using a custom elbow testing apparatus. Testing was first performed
on intact specimens, then on MCL-deficient specimens. Elbow instability was quantified using
an electromagnetic tracking device by measuring internal– external rotation and varus–valgus
laxity of the ulna relative to the humerus. Compared with the intact elbow, transection of the
MCL, with the arm in a vertical orientation, caused a significant increase in internal– external
rotation during passive elbow flexion with the forearm in pronation, but forearm supination
reduced this instability. Overall, following MCL transection the elbow was more stable with
the forearm in supination than pronation during passive flexion. In the pronated forearm
position simulated active flexion also reduced the instability detected during passive flexion,
with the arm in a varus and valgus gravity-loaded orientation. The maximum varus–valgus
laxity was significantly increased with MCL transection regardless of forearm position during
passive flexion. We concluded that active mobilization of the elbow with the arm in vertical
orientation during rehabilitation is safe in the setting of an MCL-deficient elbow with the
forearm in a fully supinated and pronated position. Splinting and passive mobilization of the
MCL-deficient elbow with the forearm in supination should minimize instability and valgus
elbow stresses should be avoided throughout the rehabilitation period. (J Hand Surg 2000;
25A:1051–1057. Copyright © 2000 by the American Society for Surgery of the Hand.)
Key words: Elbow, rehabilitation, instability, medial, biomechanics.

Chronic medial collateral ligament (MCL) insuf- pitchers and javelin throwers. Sporting events or
ficiency can lead to significant morbidity and may be occupations that place the elbow under repeated val-
career ending for some athletes, particularly baseball gus load may result in microtears and eventual dis-
ruption of the MCL. In the long term patients may
experience chronic pain, weakness, ulnar nerve
From the Bioengineering Research Lab, Hand and Upper Limb
Centre, Lawson Research Institute, St Joseph’s Health Centre, London,
symptoms, and flexion contractures from posterome-
Ontario, Canada. dial olecranon impingement.1 Most competitive ath-
Received for publication November 18, 1999; accepted in revised letes require reconstruction of this ligament to return
form May 16, 2000. to high levels of competition.1,2 Conway et al2 re-
No benefits in any form have been received or will be received from
a commercial party related directly or indirectly to the subject of this ported that only 68% of athletes were able to return
article. to their sporting activity after MCL reconstruction.
Reprint requests: Graham J.W. King, MD, Hand and Upper Limb Valgus elbow instability may also occur as an
Centre, 268 Grosvenor St, London, Ontario, Canada N6A 4L6.
acute traumatic event. Medial collateral ligament in-
Copyright © 2000 by the American Society for Surgery of the Hand
0363-5023/00/25A06-0002$3.00/0 jury in the elbow has been reported to be as high as
doi: 10.1053/jhsu.2000.17819 100% in an acute elbow dislocation.3 Ligamentous

The Journal of Hand Surgery 1051

1052 Armstrong et al / Rehabilitation MCL Deficient Elbow

healing occurs in the majority of patients and symp- lary canal of the humerus. Similarly, the pronator
tomatic valgus instability is uncommon.3,4 Long- teres origin was simulated by a cable passing through
term follow-up studies of elbow dislocations man- a Delrin sleeve in the medial epicondyle and the
aged nonsurgically show a 17 % incidence of valgus humeral medullary canal.
instability on manual stress testing.4 These patients The specimens were mounted into a specialized
are often asymptomatic since the general population testing apparatus by clamping the humerus in neutral
does not typically place their elbow in a provocative rotation so that when the elbow was flexed to 90°, the
valgus-loaded position unless required by their ac- forearm was perpendicular to the floor. Full uncon-
tivity or occupation. Josefsson et al3 showed that strained motion of the elbow with the arm oriented
surgical repair of the MCL in simple acute elbow vertically were used in this study, along with varus
dislocations does not provide superior results to non- and valgus gravity-loaded positions (Fig. 1).11,12
surgical treatment. Each stainless steel cable was then secured to an
The optimal method to rehabilitate the elbow fol- associated pneumatic actuator of the testing device.
lowing an acute MCL tear or an MCL reconstruction The muscle lines of action were maintained using a
has received little attention. It has been reported that tendon alignment unit.9 The actuator pressure was
muscles about the elbow joint have the potential to governed by electromechanical pneumatic propor-
provide stability by compressing the articular sur- tional pressure controllers, with the input signal con-
faces.5 We hypothesized that the dynamic stabilizers trolled using a custom-designed LabVIEW program
are important in the rehabilitation of the MCL-defi- (National Instruments, Austin, TX).13 The relative
cient elbow. The optimal position of the forearm muscle force level for each motion was estimated
during rehabilitation also needs to be addressed to by combining the results of electromyography
determine whether supination or pronation provides and cross-sectional area data.6,8 Active testing was
better stability. We hypothesized that supination of achieved by using an algorithm that produced
the forearm would provide better stability than pro- timed loading to the actuators, specific for the func-
nation. The purpose of this study was 2-fold: to tion simulated. The soft tissues and skin of the elbow
determine the relative contribution of muscle activity and forearm were maintained intact and sutured
and to determine the effect of forearm position on the closed between stages of the testing protocol.
stability of the MCL-deficient elbow. Specimens were kept moist by irrigation with
0.9% normal saline and all testing was performed
Materials and Methods at 22° ⫾ 2°C.
Before testing 10 preconditioning cycles were con-
Ten fresh cadaveric upper extremities (mean age, ducted by passively moving the forearm through a
69 ⫾ 10 years) were amputated at the midhumerus full arc of flexion and extension cycles. This was
and stored at –20°C. The specimens were thawed repeated for 5 active preconditioning cycles. For
overnight at room temperature. Stainless steel cables each testing sequence full passive supinated then
(0.8 mm diameter) were sutured into the distal ten- pronated elbow flexion was performed with the el-
don of each muscle or muscle group with a #2 bow oriented in the vertical gravity-dependent posi-
Ethibond suture (Johnson & Johnson, Canada). The tion (Fig. 1). For each cycle the same tester grasped
biceps brachii, brachialis, and brachioradialis were the wrist and hand to passively rotate the forearm
identified as the principle elbow flexors and the tri- into full supination or pronation until a definite end
ceps brachii as the principal elbow extensor.6,7 The point of range of motion was reached. The passive
biceps brachii and the pronator teres were identified testing cycles were then followed by simulated active
as the principal forearm supinator and pronator, re- supinated then pronated elbow flexion with the hu-
spectively.7,8 The lines of action of the biceps, tri- merus oriented in a vertical gravity-dependent posi-
ceps, and brachialis were simulated by running the tion. Pretesting active cycles were performed before
cables through a line-of-action approximating each data collection to ensure that maximal forearm supi-
muscle’s centroid.9,10 The brachioradialis origin was nation and pronation was created with the simulated
simulated using a Delrin sleeve (Plastifab Industries, muscle activity. Passive elbow flexion was per-
Canada) inserted into the proximal portion of the formed with the humerus also oriented in the varus or
lateral supracondylar ridge to ensure replication of valgus gravity-dependent positions with the forearm
the muscle’s moment arm throughout elbow motion. in pronation or supination.13
Cables were run through the sleeve and the medul- Testing was first conducted with the forearm in-
 The Journal of Hand Surgery / Vol. 25A No. 6 November 2000 1053

Figure 1. Specialized testing

apparatus. Specimens were
mounted into a specialized
testing apparatus that allowed
for vertical (shown), varus,
and valgus gravity-loaded ori-
entations. Unconstrained mo-
tion of the ulna with respect to
the humerus was measured via
receivers rigidly fixed to the
distal medial edge of the ulna
and a transmitter fixed to the
base of the testing apparatus.
Cables were sutured into mus-
cle group(s) that were then se-
cured to an associated pneu-
matic actuator to allow for
simulated active motion of the
elbow.

tact. The MCL complex and joint capsules were then rior and posterior capsule were sectioned to generate
sectioned and the testing sequence was repeated. A complete valgus elbow instability. No repair of any
straight midline posterior skin incision, elevating the of the sectioned tissues was performed before test-
medial skin flap to expose the medial aspect of the ing.
elbow, was used. The ulnar nerve was isolated and Joint kinematic data were collected using the
protected. The origin of the common flexor–pronator Flock of Birds electromagnetic tracking system
muscle group was divided from the medial epicon- (Ascension Technology, Burlington, VT).11 A trans-
dyle and elevated to expose the MCL complex. The mitter was positioned and secured to the base of the
anterior and posterior bundle of the MCL and ante- testing apparatus. To quantify forearm motion, a
1054 Armstrong et al / Rehabilitation MCL Deficient Elbow

receiver was rigidly fixed to the distal–medial edge

of the ulna. Care was taken to ensure that there was
no muscle impingement and that the receiver did not
hinder forearm rotation.
The motion of the ulna was expressed relative to
the humerus using an anatomically derived coordi-
nate system in each bone.13 These coordinate sys-
tems were defined from osseous landmarks digitized
with a stylus attached to a second Flock of Birds
receiver.13 After completion of each testing sequence
the elbow and wrist were disarticulated and the an-
atomic landmarks of the humerus and ulna were
determined from the average of 3 successive digiti-
Figure 2. Passive elbow flexion in the pronated forearm
zations. The humeral coordinate system was estab-
position. The mean and standard deviation of I-E ulnar
lished from 3 landmarks: the center of curvature of rotation is shown for the intact and MCL-deficient elbow
the capitellum, the center of the trochlea, and the during passive flexion with the forearm pronated. Tran-
center of the humeral shaft. The ulnar coordinate section of the MCL significantly increased the magnitude
system was calculated from the center and plane of of I-E rotation measured with the forearm pronated during
the greater sigmoid notch of the proximal ulna and passive elbow flexion (p ⬍ .0001).
the ulnar styloid. A least-square circle-fitting routine
was used to determine the centers of these structures.
The relative motion of the ulna with respect to the
humerus was analyzed using the Euler Z-X-Y se- Small but statistically significant differences in I-E
quence.14 rotation were found at 90° (p ⫽ .004) and 120° (p ⫽
Elbow instability was quantified by internal– ex- .006) compared with the intact elbow during passive
ternal (I-E) rotation of the ulna relative to the hu- elbow flexion with the forearm in full supination; the
merus. The effects of active and passive motion and intact elbow was more stable. Overall, following
forearm pronation and supination on elbow stability, MCL transection, the elbow was more stable with the
for the intact elbow and after sectioning of the MCL, forearm in supination than pronation during passive
were analyzed. The varus–valgus (V-V) laxity was flexion (p ⬍ .0001).
calculated by measuring the difference in the valgus Active elbow flexion significantly increased the
orientation of the ulna relative to the humerus with rotational stability in the MCL-deficient elbow com-
the arm in the varus and valgus gravity-dependent pared with passive flexion with the forearm pronated
positions. The maximum V-V laxity was defined (p ⬍ .0001) (Fig. 4). With the arm in the varus and
as the largest difference in V-V laxity measured valgus gravity-loaded positions there was a signifi-
throughout the arc of elbow flexion. The effect of cant increase in maximum V-V laxity measured fol-
forearm pronation and supination on maximum V-V lowing MCL transection (p ⬍ .0001) during passive
laxity after sectioning of the MCL were analyzed. elbow flexion (Fig. 5). This instability was not af-
One-way and two-way repeated-measures ANOVA fected by forearm position (p ⫽ .20). The elbow
and Student-Newman-Keuls multiple comparison became so unstable in the valgus-loaded position
procedure, with ␣ ⫽ .05, were used. following sectioning of the MCL that the elbow
could not be moved using simulated active motion
Results and only passive motion tests could be conducted.
Compared with the intact elbow transection of the
Discussion
MCL significantly increased the magnitude of I-E
rotation measured with the forearm pronated during The main rehabilitative goal for patients with an
passive elbow flexion (p ⬍ .0001) (Fig. 2). With the MCL disruption or reconstruction of the elbow is to
forearm supinated during passive flexion, however, return to their previous level of activity and to pre-
there was no significant difference in the I-E rotation vent any further degenerative sequelae. Therefore,
measured between the intact and MCL sectioned the rehabilitation protocol followed by the therapist
elbows at 30° (p ⫽ .90) or 60° (p ⫽ .20) (Fig. 3). for these injuries needs to facilitate an environment
 The Journal of Hand Surgery / Vol. 25A No. 6 November 2000 1055

Figure 3. Passive elbow flexion in the supinated forearm

Figure 5. Maximum V-V laxity during passive elbow
position. The mean and standard deviation of I-E ulnar
flexion. The mean and standard deviation of maximum
rotation is shown for the intact and MCL-deficient elbow
V-V laxity is shown for the intact and MCL-deficient
during passive flexion with the forearm supinated. At 30°
elbow during passive flexion with the forearm in pronation
(p ⫽ .90) and 60° (p ⫽ .20) there was no significant
and supination. The maximum V-V laxity measured was
difference in I-E rotation measured between the intact and
significantly greater after the MCL was transected (p ⬍
MCL-sectioned elbows. Small but statistically significant
.0001). There was no statistically significant difference in
differences in I-E rotation were found at 90° (p ⫽ .004)
stability with the forearm held in pronation or supination
and 120° (p ⫽ .006), with the intact elbow being more
(p ⫽ .20).
stable.

that minimizes stress on the healing tissues on the prevent contractures, particularly flexion contrac-
medial side of the elbow. tures.15 Others immobilize the elbow initially for up
Some physiotherapy protocols for MCL deficiency to 2 weeks and then begin active range of motion
focus initially on stretching and passive range of with no emphasis on stretching or passive mo-
motion of the elbow. It is thought that full range of tion.2,3,16,17 The results of our study suggest that
motion should be achieved before strengthening to active motion of the MCL-deficient elbow, even in
the early stages of healing, is protective and provides
stability similar to that of the intact joint. In contrast,
passive motion may cause significant instability, par-
ticularly if the forearm is maintained in pronation.
Josefson et al3 showed no significant difference
between their surgical and nonsurgical groups with
respect to functional outcome for MCL-deficient el-
bows after simple acute elbow dislocations. Patients
were immobilized in a cast at 90° for 2 weeks fol-
lowed by active motion of the elbow without force.
We believe the use of active motion to rehabilitate
these patients allowed for good results to be observed
in the nonsurgical group.
The optimal method to rehabilitate the MCL-defi-
cient elbow has received little attention. Therapy
protocols have been suggested and followed, but
Figure 4. Active versus passive elbow flexion in the
pronated forearm position. The mean and standard devia-
with little scientific evidence. Our results have
tion of I-E ulnar rotation is shown for the MCL-deficient shown that active mobilization of the elbow with the
elbow during active and passive flexion. Active elbow arm in the vertical orientation is safe in the setting of
flexion significantly increased the stability in the MCL- an MCL-deficient elbow with the forearm in a supi-
deficient elbow compared with passive flexion with the nated or pronated position. We feel that muscle ac-
forearm pronated (p ⬍ .0001). tivation may apply a compressive force to the elbow,
1056 Armstrong et al / Rehabilitation MCL Deficient Elbow

thereby augmenting stability. Morrey et al18 studied our statistical analysis which shows that overall, fol-
the transmission of axial force across the radiohum- lowing MCL transection, the elbow was more stable
eral joint during simulated active motion of the el- with the forearm in supination than pronation during
bow. These investigators showed that the force trans- passive flexion (p ⬍ .0001). External rotation of
mission at the radiohumeral joint during simulated the ulna occurs with forearm supination and internal
biceps muscle load, to produce elbow flexion, was rotation occurs with pronation.23 Forearm supination
greater with the forearm in pronation compared with creates an external rotation moment on the ulna that
supination. Palmer et al19 have shown that proximal further allows the medial side of the elbow to effec-
radial migration can occur with forearm pronation. tively close. Passive mobilization of the MCL defi-
The results of these studies further support the con- cient elbow with the forearm in pronation should be
cept that muscle activation allows for an axial com- avoided, while splinting or passive mobilization of
pressive load across the joint to enhance elbow sta- the MCL-deficient elbow with the forearm in supi-
bility. nation should minimize instability and allow optimal
The radial head has been shown to be an important MCL healing.
secondary restraint to valgus instability.13,18,20 Ra- Forearm position and muscle activation did not
dial head excision in the setting of MCL insuffi- influence the gross instability observed in the MCL-
ciency dramatically increases valgus instability. Ra- deficient elbow with the arm in valgus gravity-loaded
dial head excision without MCL insufficiency has orientation, suggesting that this position should be
only a small effect on valgus stability.13,18 Knowing avoided during rehabilitation.
that muscle activation affects the radiohumeral artic- A weakness of our study is that the effect of
ulation and likely contributes to elbow stability, it is neutral forearm rotation on the MCL-deficient elbow
speculated that the stability afforded by muscle ac-
was not examined. The load control testing system
tivation would be reduced in the setting of radial
used in this study was not able to reliably position the
head excision in association with MCL insufficiency.
forearm in neutral rotation due to changes in muscle
This needs to be addressed in future biomechanical
moment arms throughout elbow flexion. A displace-
studies.
ment control system is needed to maintain neutral
Dynamic stability of the elbow has been examined
rotation and should be performed in a future study.
previously in pitchers through the use of electromyo-
Another weakness of this study is the fact that this
graphic analysis of the elbow musculature during
is a biomechanical study and any extrapolation to the
throwing.21,22 Hamilton et al22 showed that during
the late cocking and acceleration phase of throwing, clinical setting should be made with caution. Take
when the elbow is under its maximal valgus load, for instance the MCL insufficiency model. All me-
decreased firing was found in the flexor pronator dial valgus elbow stabilizers to the elbow were di-
musculature in athletes with MCL-deficient elbows vided in this study with no subsequent repair. In the
compared with athletes with normal elbows. This clinical setting there would be a spectrum of insta-
suggests that in extreme positions of valgus, the bility ranging from partial to complete tears of the
stability afforded by the surrounding musculature is flexor pronator muscle group, capsule, or MCL. It
not effective in the MCL-deficient elbow. was our intent in this study to model the most un-
During passive elbow flexion with the forearm stable situation to better determine the contribution
supinated there was no significant difference be- of muscles to stability of the MCL-deficient elbow.
tween the intact and MCL-deficient elbows at 30° Our data demonstrate that even in the worst valgus
and 60° flexion; however, there was a statistically elbow injury important compensations can be made
significant difference at 90° and 120°. This suggests by the remaining musculature and soft tissue com-
that the stability afforded by the supinated position is plexes. If only partial tears of the medial structures
better at lower flexion angles. Since the magnitude of have occurred then one could assume that stability to
rotational instability found at the higher flexion an- the elbow would be less compromised and the effects
gles with the forearm supinated were considerably of forearm rotation and muscle activation would be
less than the magnitude of rotational instability found less pronounced. It also would be expected that any
at these flexion angles with forearm pronation, how- repairable tissue at the time of a soft tissue recon-
ever, the investigators feel that in the clinical setting struction would also be dealt with appropriately. The
supinated positioning in the MCL-deficient elbow is fact that the worst-case scenario was analyzed also
still safer than pronation. This is also supported by could be considered a strength of this study since
 The Journal of Hand Surgery / Vol. 25A No. 6 November 2000 1057

favorable treatment options were found even in the 8. Amis AA, Dowson D, Wright V. Muscle strengths and
face of total valgus elbow instability. musculo-skeletal geometry of the upper limb. Eng Med
1979;8:41– 48.
Another strength would be that an elbow simulator
9. An KN, Hui FC, Morrey BF, Linscheid RL, Chao EY.
was used to generate active motion. Laboratory test- Muscles across the elbow joint: a biomechanical analysis.
ing is an important step in leading to improved J Biomech 1981;14:659 – 669.
clinical results and we feel that we have been able to 10. Murray WM, Delp SL, Buchanan TS. Variation of muscle
simulate active elbow motion to further our knowl- moment arms with elbow and forearm position. J Biomech
edge. It is important to gain quantitative data to 1995;28:513–525.
support techniques used clinically. Our kinematics 11. Milne AD, Chess DG, Johnson JA, King GJW. Accuracy
of an electromagnetic tracking device: a study of the op-
have been shown to be reproducible in the past.12
timal operating range and metal interference. J Biomech
Differences in muscle activation may influence our 1996;29:791–793.
results; however, one would still expect to see a 12. Johnson JA, Rath DA, Dunning CE, Roth SE, King GJW.
stabilizing influence of muscle activation based on Simulation of forearm and elbow motion in-vitro using a
vectors of forces across the elbow. load controlled testing apparatus. J Biomech 2000;33:635–
Active mobilization of the elbow during rehabili- 639.
tation appears to be safe in the setting of an MCL- 13. King GJW, Zarzour ZDS, Rath DA, Dunning CE, Patter-
son SD, Johnson JA. Metallic radial head arthroplasty
deficient elbow. Splinting and passive mobilization improves the valgus stability of the medial collateral defi-
of the MCL-deficient elbow should be performed cient elbow. Clin Orthop 1999;368:114 –125.
with the forearm maintained in supination. This re- 14. Morrey BF, Tanaka S, An K-N. Valgus stability of the
habilitation approach may be useful following acute elbow: a definition of primary and secondary constraints.
MCL disruptions, surgical procedures that violate the Clin Orthop 1991;265:187–195.
MCL, or following reconstruction of the MCL for 15. Wilk KE, Arrigo C, Andrews JR. Rehabilitation of the
elbow in the throwing athlete. J Orthop Sports Phys Ther
chronic instability.
1993;17:305–317.
16. Jobe FW, Stark H, Lombardo SJ. Reconstruction of the
References ulnar collateral ligament in athletes. J Bone Joint Surg
1. Lee ML, Rosenwasser MP. Chronic elbow instability. Or- 1986;68A:1158 –1163.
thop Clin North Am 1999;30:81– 89. 17. Kuroda S, Sakamaki K. Ulnar collateral ligament tears of
2. Conway JE, Jobe FW, Glousman RE, Pink M. Medial the elbow joint. Clin Orthop 1986;208:266 –271.
instability of the elbow in throwing athletes: treatment by 18. Morrey BF, An KN, Stormont TJ. Force transmission through
repair or reconstruction of the ulnar collateral ligament. the radial head. J Bone Joint Surg 1988;70A:250 –256.
J Bone Joint Surg 1992;74A:67– 83. 19. Palmer AK, Glisson RR, Werner FW. Ulnar variance de-
3. Josefsson PO, Gentz C-F, Johnell O, Wendeberg B. Sur- termination. J Hand Surg 1982;7:376 –379.
gical versus non-surgical treatment of ligamentous injuries 20. Hotchkiss RN, Weiland AJ. Valgus stability of the elbow.
following dislocation of the elbow joint: a prospective J Orthop Res 1987;5:372–377.
randomized study. J Bone Joint Surg 1987;69A:605– 608. 21. Glousman RE, Barron J, Jobe FW, Perry J, Pink M. An
4. Josefsson P0, Johnell O, Gentz CF. Long-term sequelae of electromyographic analysis of the elbow in normal and
simple dislocation of the elbow. J Bone Joint Surg 1984; injured pitchers with medial collateral ligament insuffi-
66A:927–930. ciency. Am J Sports Med 1992;20:311–317.
5. King GJW, Morrey BF, An K-N. Stabilizers of the elbow. 22. Hamilton CD, Glousman RE, Jobe FW, Brault J, Pink M,
J Shoulder Elbow Surg 1993;2:165–174. Perry J. Dynamic stability of the elbow: electromyographic
6. Funk DA, An KN, Morrey BF, Daube JR. Electromyo- analysis of the flexor pronator group and the extensor
graphic analysis of muscles across the elbow joint. J Or- group in pitchers with valgus instability. J Shoulder Elbow
thop Res 1987;5:529 –538. Surg 1996;5:347–354.
7. Williams PL. Skeletal system. In: Gray’s Anatomy. 38th 23. Connor PM, An KN. Biomechanics of total elbow arthro-
ed. New York: Churchill Livingstone, 1995:622– 646. plasty. Semin Arthroplasty 1998;9:25–31.