- Email: [email protected]
Dynamic inferior stabilizers of the shoulder joint
Clinical Biomechanics 16 (2001) 138±143
www.elsevier.com/locate/clinbiomech
Dynamic inferior stabilizers of the shoulder joint A.M. Halder a, C.G. Halder a, K.D. Zhao a, S.W. O'Driscoll b, B.F. Morrey b, K.N. An a,* a
Orthopedic Biomechanics Laboratory, Mayo Clinic/Mayo Foundation, 200 First Street SW, Rochester, MN, USA b Department of Orthopedics, Mayo Clinic/Mayo Foundation, Rochester, MN, USA Received 4 April 2000; accepted 26 September 2000
Abstract Background. The glenohumeral joint is soft-tissue balanced. However, few studies have focused on its dynamic inferior stabilizers. Objective. The objective of this study was to investigate the dynamic contributions of ®ve shoulder muscles to inferior stability of the glenohumeral articulation in four joint positions. Methods. The anterior, lateral and posterior deltoid, supraspinatus, short head of biceps, coracobrachialis and long head of triceps from ten cadaveric shoulders were tested in 0°, 30°, 60° and 90° of glenohumeral abduction. A constant inferior force of 15 N was applied to the humerus. The tendons were loaded sequentially in proportion to their respective muscle's cross-sectional area. Translations of the humeral head on the glenoid were recorded with a 3-Spacee tracking device. Results. The lateral deltoid (8.2 mm, SD 4.8 mm) was potentially most eective in superior translation of the humeral head followed by the posterior deltoid (7.7 mm, SD 4.8 mm). The coracobrachialis and short head of biceps had considerable capability to translate the humeral head superiorly (2.8 mm, SD 1.3 mm) while the supraspinatus showed the weakest eects (1.3 mm, SD 0.5 mm). Relevance Strengthening exercises of the deltoid may be useful in the treatment of inferior glenohumeral instability, while the supraspinatus seems to be less important for inferior glenohumeral stability than previously assumed. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Shoulder; Glenohumeral joint; Inferior stability; Deltoid; Supraspinatus
1. Introduction The glenohumeral joint is mainly stabilized by muscles, tendons and ligaments. Several studies identi®ed static inferior stabilizers whose individual contributions to stability depend on joint position. In adduction, the superior glenohumeral ligament resists inferior translation of the humeral head. With increasing abduction, the anterior and posterior portions of the inferior glenohumeral ligament are primarily responsible for restraining inferior translation [1]. In external rotation, the coracohumeral ligament stabilizes in the inferior direction [2,3]. In combination with the superior capsuloligamentous structures, scapular inclination contributes to inferior glenohumeral stability [9]. A defect of the ro-
*
Corresponding author. E-mail address: [email protected] (K.N. An).
tator interval capsule [4,5] or loss of negative intraarticular pressure increases inferior translations of the humeral head [6±8]. Few studies focused on dynamic stabilizers of the glenohumeral joint in the inferior direction [10±12] although stabilization by muscle contraction against inferior translation was found to be more eective than negative intraarticular pressure or ligament tension [13]. The supraspinatus and long head of biceps were identi®ed to provide inferior stability [10,12]. A large rotator cu tear involving the supraspinatus may lead to inferior instability [11] but clinically causes superior migration of the humeral head. Muscles translating the humeral head superiorly stabilize the glenohumeral joint in the inferior direction. On the other hand, they may narrow the acromio-humeral interval and thereby expose the rotator cu tendons to increased shear and compressive forces. According to the concept of force couples stabilizing the
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 0 ) 0 0 0 7 7 - 2
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
139
glenohumeral joint, the deltoid is the superior part of the frontal plane force couple counterbalanced by the inferior rotator cu muscles [14]. However, little information exists about the contribution of muscles that translate the humeral head superiorly to inferior glenohumeral stability. However, this would be crucial in conservative treatment of inferior instability as part of multidirectional instability and as an isolated entity. Furthermore, knowledge of the muscles that translate the humeral head superiorly (thereby exposing the rotator cu tendons to increased mechanical wear) would be important in the treatment of rotator cu disease. The objective of this study was to investigate the individual contributions of the anterior, lateral and posterior deltoid, short head of biceps, coracobrachialis and long head of triceps to inferior glenohumeral stability. 2. Methods 2.1. Specimen preparation Ten fresh frozen cadaveric shoulders were studied after ruling out radiological evidence of glenohumeral osteoarthritis and rotator cu tears. During dissection, preparation, and testing specimens were moistened using physiologic saline solution to prevent dehydration. All soft tissues were removed except the anterior, lateral and posterior deltoid, supraspinatus, short head of biceps, long head of triceps and coracobrachialis muscles with tendons. The muscle belly of the deltoid was preserved. The other muscles were elevated from the bone and resected at the musculotendinous junction. Nylon strings were sutured to the ¯at tendons to allow even loading. A ®berglass rod was cemented into the medullary canal of the proximal humerus to control position. A thinner rod was mounted perpendicular to the humeral shaft in the neutral position to control rotation. The scapulae were mounted onto a Plexiglase plate in the shoulder testing device (Fig. 1). 2.2. Testing device The shoulder testing device (Fig. 1) was made of nonmetal materials to avoid interference with the electromagnetic tracking device. A hinged plexiglas arc attached to the humerus allowed control of abduction and rotation. Strings from the tendons were connected through variable-position pulleys to pneumatic actuators (Airpot Corporation, Norwalk, Connecticut, USA). A commercially available computer controlled by Labviewe software (National Instruments Corporation 1994, Austin, Texas, USA) drove electro-pneumatic valves (Proportion-Air 1997, McCordsville, Indiana, USA) to load the pneumatic actuators. Custom-made
Fig. 1. Shoulder testing device. A hinged Plexiglas arc attached to the humerus allowed control of abduction and rotation. Through pulleys that were variable in position, strings from the tendons were connected to pneumatic actuators. A commercially available computer drove electropneumatic valves to load the pneumatic actuators. Custommade load cells on each cylinder veri®ed the applied loads. A 3-Spacee tracking system attached to the specimen measured the three-dimensional positions and orientations of sensors in relation to the source of electromagnetic waves.
load cells on each cylinder veri®ed the applied loads. A 3-Spacee tracking system (Polhemus 1993, Colchester, Vermont, USA) attached to the specimen measured the three-dimensional positions and orientations of two sensors in relation to the source of the electromagnetic waves. 2.3. Mounting the specimen Fixed on the testing device were laser-pointers whose beam intersected at the center of rotation of the hinged arc permitting precise positioning of the glenohumeral articulation. The scapulae were aligned and then rigidly mounted in the shoulder testing device so that the medial margin of the scapula was in line with the vertical axis of the device and the humeral head was in the center of the pivoting arc. The nylon loops sutured to the tendons were connected to the pneumatic actuators by strings that were guided by pulleys. The positions of the pulleys were carefully adjusted so that the strings imitated muscle lines of action by running through the centroids of each muscle [15]. Finally, sensors of the 3Spacee tracking system were attached to the proximal humerus close to the head and the spine of the scapula.
140
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
The source of the electromagnetic waves was rigidly mounted in line with the vertical axis of the shoulder testing device. 2.4. Testing Because tightness of the ligaments in the extremes of rotation prevents the muscles from generating inferior translations [1,3], the humerus was locked in neutral rotation. Prior to testing, the shoulder joints were vented to exclude the eects of negative intraarticular pressure. A constant inferior force of 15 N was applied to the humerus leading to an inferiorly decentralized position of the humeral head. To ensure identical starting positions for all muscles throughout the experiment, the humerus was reset between each test according to position values provided by the 3-Spacee tracking system. The potentials of the anterior, lateral and posterior deltoid, supraspinatus, short head of biceps/coracobrachialis and long head of triceps muscles to reverse the inferior translation were measured. The short head of biceps and coracobrachialis were loaded simultaneously according to the sum of their cross-sectional areas. The tendons were loaded sequentially in line of muscle action and proportional to their respective cross-sectional areas [16]. This was based on the assumption that maximum generated muscle force is proportional to its cross-sectional area [17]. The experiment started in the hanging arm position with the humeral head in an inferiorly decentralized position and was performed in 30°, 60° and 90° of glenohumeral abduction (Fig. 2). After the position measurements were taken, the glenohumeral joint was disarticulated and the bony landmarks digitized to determine their positions relative to the sensors. 2.5. Data analysis From the digitization data the center of the humeral head, de®ned as the geometric center of its convexity, and the center of the glenoid, de®ned as the intersection of its vertical and horizontal axes, were calculated. Translation distances for each muscle were calculated by comparing the positions of the center of the humeral head before and after muscle loading. 2.6. Statistical analysis Summary statistics are reported as means and standard deviations. Translation distances were ®rst analyzed using two-factor analysis of variance with repeated measures on both factors (muscle and arm position). However, because signi®cant interactions between muscle and arm position were identi®ed, separate oneway repeated measures ANOVAs were run for each arm position. Signi®cant eects were then further analyzed using the Student±Newman±Keuls multiple comparison
Fig. 2. Testing of the supraspinatus muscle in hanging arm position. The humerus was locked in neutral rotation. A constant inferior force of 15 N was applied to the humerus, and the tendon was loaded in line of muscle action proportional to its cross-sectional area by a computer-controlled pneumatic actuator. Positional measurements were taken by a 3-Spacee tracking system.
procedure. All statistical tests were two-sided, with the threshold of signi®cance set at alpha 0.05. All analysis was performed using SAS version 6.12 on a Sun Ultra II computer. 3. Results The anterior, lateral and posterior deltoid, supraspinatus, short head of biceps/coracobrachialis and long head of triceps were tested for their potential to reverse inferior translation of the humeral head due to a constant inferior force applied to the humerus. Results are reported as mean superior translations in millimeters of the four tested joint positions. Generally the largest translation values were seen in 30° and 60° of glenohumeral abduction. In the hanging arm position, the translations were smaller because the superior joint capsule limited inferior translation. In 90° of glenohumeral abduction the smallest translations were detected because the inferior glenohumeral ligament complex was tight (Fig. 3). The lateral deltoid (8.2 mm, SD 4.8 mm) was significantly P < 0:05 most powerful in superior translation of the humeral head throughout all positions although not signi®cantly P > 0:05 dierent from the posterior deltoid (7.7 mm, SD 4.8 mm) in 30°, 60° and 90° of glenohumeral abduction. The lateral deltoid was the only muscle tested, which was consistently able to
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
Fig. 3. Translation values of the tested muscles on average in hanging arm position, at 30°, at 60° and at 90° of glenohumeral abduction. The graph shows the superior translations of the humeral head on the glenoid eected by the tested muscles counteracting a constant inferior force. SSP: supraspinatus, ADLT: anterior deltoid, LDLT: lateral deltoid, PDLT: posterior deltoid, SHB/COB: short head of biceps/ coracobrachialis, TRZ: long head of triceps.
completely reverse inferior subluxation of the humeral head occurring in 30° and 60° of glenohumeral abduction. Apart from a line of action directed straight superiorly, the lateral and posterior deltoid had large cross-sectional areas. The anterior deltoid was signi®cantly P > 0:05 weaker in superior translation (4.9 mm, SD 3.7 mm) than the lateral and posterior deltoid because its line of action was diverted anteriorly by the humeral head and it had a smaller cross-sectional area. The coracobrachialis and short head of biceps had a considerable superior translating eect (2.8 mm, SD 1.3 mm). Their lines of action run obliquely and anteriorly to the glenoid, but they had a relatively large crosssectional area. The long head of triceps showed small superior translations (1.7 mm, SD 0.7 mm) although its line of action runs straight superiorly towards the glenoid, but it had a small cross-sectional area. Finally and surprisingly, the supraspinatus was weakest in superior translation throughout all positions (1.3 mm, SD 0.5 mm), signi®cantly P < 0:05 less effective than all portions of the deltoid. Even the short head of biceps/coracobrachialis showed signi®cantly P < 0:05 larger translation distances in the hanging arm position and at 30° of glenohumeral abduction. Its line of action is nearly horizontal and its cross-sectional area is of medium size. 4. Discussion In these experiments the lateral deltoid was the most capable of reversing the inferior translation of the humeral head, closely followed by the posterior deltoid whereas the anterior deltoid was less eective. As our study showed potential muscle eects, the true in vivo role of the deltoid depends on its recruitment during motion.
141
Traditionally the deltoid was described as an essential mover of the glenohumeral joint eecting abduction, ¯exion and extension [18]. In abduction, the deltoid becomes more ecient [19,20] and shows maximum electromyograhic activity at higher degrees [21]. Thus, removal of the deltoid leads to a progressive decrease in strength during abduction [22] and a decrease in range of motion [23]. However, only few studies have focused on the stabilizing function of the deltoid. In a cadaver study the passive stabilizing capability of the deltoid was found to be insigni®cant [24] but the tests were performed with low loads at the extremes of glenohumeral motion, when the ligamentous constraints provided sucient inferior stability. In our preliminary experiments, removal of the deltoid caused inferior subluxation of the humeral head in the mid-range of motion after venting of the joint capsule. Kuechle [25] found that the lateral deltoid has the largest elevator moment arm in the scapular plane, which implies the likelihood to be recruited for abduction function and superior translation of the humeral head action as well. In an electromyographic study, an active stabilizing eect of the deltoid was neglected because the deltoid was found to be inactive with the hanging arm carrying a weight [26], but in this position the superior joint capsule and inclination of the glenoid stabilize the glenohumeral joint inferiorly. On the other hand, a recent electromyograhic study of ®ve parts of the deltoid showed activity not related to the generation of an abduction moment and a stabilizing eect was suggested [27]. This is con®rmed by the ®nding that electromyographic activity of the anterior and lateral deltoid during abduction and ¯exion was decreased in patients with shoulder instability [28]. Finally, the deltoid was reported to be the superior part of the frontal plane force couple being counterbalanced by the inferior rotator cu muscles. Thereby it may actively stabilize the glenohumeral joint [14,29]. The short head of biceps and coracobrachialis showed considerable superior translating eects in our study. Although the long head of the biceps was found to be a superior stabilizer [30] and the whole biceps to be an inferior stabilizer [12], the results of electromyographic studies are still controversial. While several studies reported biceps activity related to shoulder ¯exion and abduction [26,32]; some did not detect it [31,33]. Our study con®rms its potential as an inferior stabilizer, but its actual physiologic role in shoulder function needs further clari®cation in future in vivo studies. The long head of triceps showed a minor superior translating eect. Only a few studies investigated the activity of the triceps reporting little or no activity during shoulder movements [34]. Therefore, it is considered to be primarily an elbow mover and its eects on the glenohumeral joint to be secondary.
142
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
The supraspinatus is an important mover of the glenohumeral joint. Speci®cally it participates in the initiation of abduction [35,36] and in rotation [37]. Although some investigators have demonstrated a major stabilizing eect of the supraspinatus in the anterior [38] and inferior directions [12], others reported a limited contribution to joint stability [36,39]. Our study showed that the supraspinatus had a minor inferior stabilizing eect. Moreover, our ®ndings support the biomechanical suspension bridge model which assumes that normal glenohumeral joint kinematics could be restored even without complete repair of a supraspinatus tendon defect [14]. The functional result depends on the integrity of the muscular force couples in the frontal and horizontal planes.
proximation of the varied lines of action of a muscle, this appears to be a legitimate biomechanical model for muscle forces. The eects of the ligamentous tightness in the extremes of rotation and the eects of negative intraarticular pressure were excluded by testing in neutral rotation and venting the joint capsules. Thereby, we were able to focus on the muscular eects. However, in the hanging arm position and 90° of glenohumeral abduction, the ligaments limited translations. Despite these limitations, the results were reproducible with a high degree of accuracy providing a basis of con®dence in this methodology.
4.1. Limitations
In our study the deltoid was potentially most capable of translating the humeral head superiorly while the supraspinatus was least eective. Therefore, deltoid strengthening exercises may be useful in the treatment of inferior instability. The biceps, long head of triceps and coracobrachialis had a considerable capability to translate the humeral head superiorly, while the supraspinatus was less eective than previously assumed. Further studies should clarify their in vivo contributions to shoulder stability.
As simulation of muscle force and direction was crucial for the result, we aimed at the highest accuracy possible using computer-controlled pneumatic actuators and low-friction pulleys for muscle loading. Nevertheless, loading was based on the assumption that maximum muscle force is proportional to its cross-sectional area, which was derived from the literature. In each position, the muscle loading was performed sequentially, disregarding eventual muscle interactions. Furthermore, as there is a high degree of variability of individual muscle activity as a function of joint position, and consistent electromyograhic data are lacking, muscle loading was performed with maximum force. Finally, the applied constant force does not occur naturally, but represented the shear force component generated by muscles or gravity. However, the study was designed to report a ranking of individual muscles according to their potential to stabilize the glenohumeral joint inferiorly in case of maximum muscle activity in selected joint positions. In combination with the results of future in vivo studies that will investigate the activity of individual muscles at speci®c joint positions, relative muscle contributions to inferior glenohumeral joint stability can be estimated. By applying a constant inferior force of 15 N, the humeral head was translated inferiorly before muscle loading. As a result, the line of muscle action was changed. Inferior subluxation and a major change in line of muscle action were prevented by leaving the muscle belly of the deltoid intact. The supraspinatus was aected most of all among the muscles tested. And, although the supraspinatus should be even more powerful at pulling the humeral head superiorly when starting from an inferior position, it was least eective. To enhance the accuracy of the model, the strings representing lines of muscle action were adjusted according to the centroids of the tested muscles as derived from the literature. Although strings are merely an ap-
5. Conclusions
Acknowledgements The ®rst author was supported by the Max-Biedermann-Institut, Berlin, Germany. The authors wish to acknowledge D. Larson from the section of Biostatistics for the statistical calculations and L. Berge from the Orthopedic Biomechanics Laboratory for the drawing of the ®gures.
References [1] Warner JP, Deng XH, Warren RF, Torzilli PA. Static capsuloligamentous restraints to superior±inferior translation of the glenohumeral joint. Am J Sports Med 1992;20:675±85. [2] Ovesen J, Nielsen S. Experimental distal subluxation in the glenohumeral joint. Arch Orthop Trauma Surg 1985;104:78±81. [3] Itoi E, Berglund LJ, Grabowski JJ, Naggar L, Morrey BF, An KN. Superior±inferior stability of the shoulder: the role of the coracohumeral ligament and the rotator interval capsule. Mayo Clin Proc 1998;73:508±15. [4] Harryman DT, Sidles JA, Harris SL, Matsen FA. The role of the rotator interval capsule in passive motion and stability of the shoulder. J Bone J Surg (Am) 1992;74-A:53±66. [5] Field LD, Warren RF, O'Brien SJ, Altchek DW, Wickiewicz T. Isolated closure of rotator interval defects for shoulder instability. Am J Sports Med 1995;23:557±63. [6] Gibb TD, Sidles JA, Harryman DT, McQuade KJ, Matsen FA. The eect of capsular venting on glenohumeral laxity. Clin Orthop 1991;268:120±7.
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143 [7] Itoi E, Motzkin NE, Browne AO, Homeyer P, Morrey BF, An KN. Intraarticular pressure of the shoulder. Arthroscopy 1993;9:406±13. [8] Wuelker N, Korell M, Thren K. Dynamic glenohumeral joint stability. J Shoulder Elbow Surg 1998;7:43±52. [9] Itoi E, Motzkin NE, Morrey BF, An KN. Scapular inclination and inferior stability of the shoulder. J Shoulder Elbow Surg 1992;1:131±9. [10] Pagnani MJ, Deng XH, Warren RF, Torzilli PA, Altcheck DW. Eect of lesions of the superior portion of the glenoid labrum on glenohumeral translation. J Bone J Surg (Am) 1995;77-A:1003± 10. [11] Hsu HC, Luo ZP, Co®eld RH, An KN. In¯uence of rotator cu tearing on glenohumeral stability. J Shoulder Elbow Surg 1996;6:413±22. [12] Soslowsky LJ, Malicky DM, Blasier RB. Active and passive factors in inferior glenohumeral stabilization: a biomechanical model. J Shoulder Elbow Surg 1997;6:371±9. [13] Warner JJ, Bowen MK, Deng X, Torzilli PA, Warren RF. Eect of joint compression on inferior stability of the glenohumeral joint. J Shoulder Elbow Surg 1999;8:31±6. [14] Burkhart SS. Fluoroscopic comparison of kinematic patterns in massive rotator cu tears. A suspension bridge model. Clin Orthop 1992;234:144±52. [15] Johnson GR, Spalding D, Nowitzke A, Bogduk N. Modelling the muscles of the scapula morphometric and coordinate data and functional implications. J Biomech 1996;29:1039±51. [16] Veeger HE, Yu B, An KN, Rozendal RH. Parameters for modeling the upper extremity. J Biomech 1997;30:647±52. [17] Tirosh R. 1 kgf/cm2 ± the isometric tension of muscle contraction: implications to cross-bridge and hydraulic mechanisms. Adv Exp Med Biol 1984;170:531±9. [18] Inman VT, Saunders JB, Abbott LC. Observations on the function of the shoulder joint. J Bone J Surg (Am) 1944;16A:268±71. [19] Otis JC, Jiang CC, Wickiewicz TL, Peterson MG, Warren RF, Santner TJ. Changes in the moment arms of the rotator cu and deltoid muscles with abduction and rotation. J Bone J Surg (Am) 1994;76:667±76. [20] McMahon PJ, Debski RE, Thompson WO, Warner JJ, Fu FH, Woo SL. Shoulder muscle forces and tendon excursions during glenohumeral abduction in the scapular plane. J Shoulder Elbow Surg 1995;4:199±208. [21] Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination in the normal shoulder. An electromyographic study. Clin Orthop 1990;257:76±85. [22] Colachis SC, Strohm BR, Brechner VL. Eects of axillary nerve block on muscle force in the upper extremity. Arch Phys Med Rehabil 1969;50:647±54.
143
[23] Markhede G, Monastyrski J, Stener B. Shoulder function after deltoid muscle removal. Acta Orthop Scand 1985;56:242±4. [24] Motzkin NE, Itoi E, Morrey BF, An KN. Contribution of passive bulk tissues and deltoid to static inferior glenohumeral stability. J Shoulder Elbow Surg 1994;3:313±9. [25] Kuechle DK, Newman SR, Itoi E, Morrey BF, An KN. Shoulder muscle moment arms during horizontal ¯exion and elevation. J Shoulder Elbow Surg 1997;6:429±39. [26] Basmajian JV, Latif MA. Integrated actions and functions of the chief ¯exors of the elbow. J Bone J Surg (Am) 1957;39-A:1106±18. [27] Michiels I, Bodem F. The deltoid muscle: an electromyographical analysis of its activity in arm abduction in various body postures. Int Orthop 1992;16:268±71. [28] Kronberg M, Brostrom LA, Nemeth G. Dierences in shoulder muscle activity between patients with generalized joint laxity and normal controls. Clin Orthop 1991;269:181±92. [29] Lucas D. Biomechanics of the shoulder joint. Arch Surg 1973;107:425±32. [30] Warner JJ, McMahon PJ. The role of the long head of biceps brachii in superior stability of the glenohumeral joint. J Bone J Surg (Am) 1995;77-A:366±72. [31] Furlani J. Electromyographic study of the m. biceps brachii in movements at the glenohumeral joint. Acta Anat Basel 1976;96:270±84. [32] Sakurai G, Ozaki J, Tomita Y, Nishimoto K, Tamai S. Electromyographic analysis of shoulder joint function of the biceps brachii muscle during isometric contraction. Clin Orthop 1998;354:123±31. [33] Yamaguchi K, Riew KD, Galatz LM, Syme JA, Neviaser RJ. Biceps activity during shoulder motion: an electromyographic analysis. Clin Orthop 1997;336:122±9. [34] Basmajan JV, Bazant FJ. Factors preventing downward dislocation of the adducted shoulder joint. J Bone J Surg (Am) 1959;41A:1182±6. [35] Howell SM, Imobersteg AM, Seger DH, Marone PJ. Clari®cation of the role of the supraspinatus muscle in shoulder function. J Bone J Surg (Am) 1986;68-A:398±404. [36] Thompson WO, Debski RE, Boardman ND, Taskiran E, Warner JJ, Fu FH, Woo SL. A biomechanical analysis of rotator cu de®ciency in a cadaveric model. Am J Sports Med 1996;24:286± 92. [37] Ihashi K, Matsushita N, Yagi R, Handa Y. Rotational action of the supraspinatus muscle on the shoulder joint. J Electromyogr Kinesiol 1998;8:337±46. [38] Malicky DM, Soslowsky LJ, Blasier RB, Shyr Y. Anterior glenohumeral stabilization factors: progressive eects in a biomechanical model. J Orthop Res 1996;14:282±8. [39] Sharkey NA, Marder RA. The rotator cu opposes superior translation of the humeral head. Am J Sports Med 1995;23:270±5.
www.elsevier.com/locate/clinbiomech
Dynamic inferior stabilizers of the shoulder joint A.M. Halder a, C.G. Halder a, K.D. Zhao a, S.W. O'Driscoll b, B.F. Morrey b, K.N. An a,* a
Orthopedic Biomechanics Laboratory, Mayo Clinic/Mayo Foundation, 200 First Street SW, Rochester, MN, USA b Department of Orthopedics, Mayo Clinic/Mayo Foundation, Rochester, MN, USA Received 4 April 2000; accepted 26 September 2000
Abstract Background. The glenohumeral joint is soft-tissue balanced. However, few studies have focused on its dynamic inferior stabilizers. Objective. The objective of this study was to investigate the dynamic contributions of ®ve shoulder muscles to inferior stability of the glenohumeral articulation in four joint positions. Methods. The anterior, lateral and posterior deltoid, supraspinatus, short head of biceps, coracobrachialis and long head of triceps from ten cadaveric shoulders were tested in 0°, 30°, 60° and 90° of glenohumeral abduction. A constant inferior force of 15 N was applied to the humerus. The tendons were loaded sequentially in proportion to their respective muscle's cross-sectional area. Translations of the humeral head on the glenoid were recorded with a 3-Spacee tracking device. Results. The lateral deltoid (8.2 mm, SD 4.8 mm) was potentially most eective in superior translation of the humeral head followed by the posterior deltoid (7.7 mm, SD 4.8 mm). The coracobrachialis and short head of biceps had considerable capability to translate the humeral head superiorly (2.8 mm, SD 1.3 mm) while the supraspinatus showed the weakest eects (1.3 mm, SD 0.5 mm). Relevance Strengthening exercises of the deltoid may be useful in the treatment of inferior glenohumeral instability, while the supraspinatus seems to be less important for inferior glenohumeral stability than previously assumed. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Shoulder; Glenohumeral joint; Inferior stability; Deltoid; Supraspinatus
1. Introduction The glenohumeral joint is mainly stabilized by muscles, tendons and ligaments. Several studies identi®ed static inferior stabilizers whose individual contributions to stability depend on joint position. In adduction, the superior glenohumeral ligament resists inferior translation of the humeral head. With increasing abduction, the anterior and posterior portions of the inferior glenohumeral ligament are primarily responsible for restraining inferior translation [1]. In external rotation, the coracohumeral ligament stabilizes in the inferior direction [2,3]. In combination with the superior capsuloligamentous structures, scapular inclination contributes to inferior glenohumeral stability [9]. A defect of the ro-
*
Corresponding author. E-mail address: [email protected] (K.N. An).
tator interval capsule [4,5] or loss of negative intraarticular pressure increases inferior translations of the humeral head [6±8]. Few studies focused on dynamic stabilizers of the glenohumeral joint in the inferior direction [10±12] although stabilization by muscle contraction against inferior translation was found to be more eective than negative intraarticular pressure or ligament tension [13]. The supraspinatus and long head of biceps were identi®ed to provide inferior stability [10,12]. A large rotator cu tear involving the supraspinatus may lead to inferior instability [11] but clinically causes superior migration of the humeral head. Muscles translating the humeral head superiorly stabilize the glenohumeral joint in the inferior direction. On the other hand, they may narrow the acromio-humeral interval and thereby expose the rotator cu tendons to increased shear and compressive forces. According to the concept of force couples stabilizing the
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 0 ) 0 0 0 7 7 - 2
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
139
glenohumeral joint, the deltoid is the superior part of the frontal plane force couple counterbalanced by the inferior rotator cu muscles [14]. However, little information exists about the contribution of muscles that translate the humeral head superiorly to inferior glenohumeral stability. However, this would be crucial in conservative treatment of inferior instability as part of multidirectional instability and as an isolated entity. Furthermore, knowledge of the muscles that translate the humeral head superiorly (thereby exposing the rotator cu tendons to increased mechanical wear) would be important in the treatment of rotator cu disease. The objective of this study was to investigate the individual contributions of the anterior, lateral and posterior deltoid, short head of biceps, coracobrachialis and long head of triceps to inferior glenohumeral stability. 2. Methods 2.1. Specimen preparation Ten fresh frozen cadaveric shoulders were studied after ruling out radiological evidence of glenohumeral osteoarthritis and rotator cu tears. During dissection, preparation, and testing specimens were moistened using physiologic saline solution to prevent dehydration. All soft tissues were removed except the anterior, lateral and posterior deltoid, supraspinatus, short head of biceps, long head of triceps and coracobrachialis muscles with tendons. The muscle belly of the deltoid was preserved. The other muscles were elevated from the bone and resected at the musculotendinous junction. Nylon strings were sutured to the ¯at tendons to allow even loading. A ®berglass rod was cemented into the medullary canal of the proximal humerus to control position. A thinner rod was mounted perpendicular to the humeral shaft in the neutral position to control rotation. The scapulae were mounted onto a Plexiglase plate in the shoulder testing device (Fig. 1). 2.2. Testing device The shoulder testing device (Fig. 1) was made of nonmetal materials to avoid interference with the electromagnetic tracking device. A hinged plexiglas arc attached to the humerus allowed control of abduction and rotation. Strings from the tendons were connected through variable-position pulleys to pneumatic actuators (Airpot Corporation, Norwalk, Connecticut, USA). A commercially available computer controlled by Labviewe software (National Instruments Corporation 1994, Austin, Texas, USA) drove electro-pneumatic valves (Proportion-Air 1997, McCordsville, Indiana, USA) to load the pneumatic actuators. Custom-made
Fig. 1. Shoulder testing device. A hinged Plexiglas arc attached to the humerus allowed control of abduction and rotation. Through pulleys that were variable in position, strings from the tendons were connected to pneumatic actuators. A commercially available computer drove electropneumatic valves to load the pneumatic actuators. Custommade load cells on each cylinder veri®ed the applied loads. A 3-Spacee tracking system attached to the specimen measured the three-dimensional positions and orientations of sensors in relation to the source of electromagnetic waves.
load cells on each cylinder veri®ed the applied loads. A 3-Spacee tracking system (Polhemus 1993, Colchester, Vermont, USA) attached to the specimen measured the three-dimensional positions and orientations of two sensors in relation to the source of the electromagnetic waves. 2.3. Mounting the specimen Fixed on the testing device were laser-pointers whose beam intersected at the center of rotation of the hinged arc permitting precise positioning of the glenohumeral articulation. The scapulae were aligned and then rigidly mounted in the shoulder testing device so that the medial margin of the scapula was in line with the vertical axis of the device and the humeral head was in the center of the pivoting arc. The nylon loops sutured to the tendons were connected to the pneumatic actuators by strings that were guided by pulleys. The positions of the pulleys were carefully adjusted so that the strings imitated muscle lines of action by running through the centroids of each muscle [15]. Finally, sensors of the 3Spacee tracking system were attached to the proximal humerus close to the head and the spine of the scapula.
140
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
The source of the electromagnetic waves was rigidly mounted in line with the vertical axis of the shoulder testing device. 2.4. Testing Because tightness of the ligaments in the extremes of rotation prevents the muscles from generating inferior translations [1,3], the humerus was locked in neutral rotation. Prior to testing, the shoulder joints were vented to exclude the eects of negative intraarticular pressure. A constant inferior force of 15 N was applied to the humerus leading to an inferiorly decentralized position of the humeral head. To ensure identical starting positions for all muscles throughout the experiment, the humerus was reset between each test according to position values provided by the 3-Spacee tracking system. The potentials of the anterior, lateral and posterior deltoid, supraspinatus, short head of biceps/coracobrachialis and long head of triceps muscles to reverse the inferior translation were measured. The short head of biceps and coracobrachialis were loaded simultaneously according to the sum of their cross-sectional areas. The tendons were loaded sequentially in line of muscle action and proportional to their respective cross-sectional areas [16]. This was based on the assumption that maximum generated muscle force is proportional to its cross-sectional area [17]. The experiment started in the hanging arm position with the humeral head in an inferiorly decentralized position and was performed in 30°, 60° and 90° of glenohumeral abduction (Fig. 2). After the position measurements were taken, the glenohumeral joint was disarticulated and the bony landmarks digitized to determine their positions relative to the sensors. 2.5. Data analysis From the digitization data the center of the humeral head, de®ned as the geometric center of its convexity, and the center of the glenoid, de®ned as the intersection of its vertical and horizontal axes, were calculated. Translation distances for each muscle were calculated by comparing the positions of the center of the humeral head before and after muscle loading. 2.6. Statistical analysis Summary statistics are reported as means and standard deviations. Translation distances were ®rst analyzed using two-factor analysis of variance with repeated measures on both factors (muscle and arm position). However, because signi®cant interactions between muscle and arm position were identi®ed, separate oneway repeated measures ANOVAs were run for each arm position. Signi®cant eects were then further analyzed using the Student±Newman±Keuls multiple comparison
Fig. 2. Testing of the supraspinatus muscle in hanging arm position. The humerus was locked in neutral rotation. A constant inferior force of 15 N was applied to the humerus, and the tendon was loaded in line of muscle action proportional to its cross-sectional area by a computer-controlled pneumatic actuator. Positional measurements were taken by a 3-Spacee tracking system.
procedure. All statistical tests were two-sided, with the threshold of signi®cance set at alpha 0.05. All analysis was performed using SAS version 6.12 on a Sun Ultra II computer. 3. Results The anterior, lateral and posterior deltoid, supraspinatus, short head of biceps/coracobrachialis and long head of triceps were tested for their potential to reverse inferior translation of the humeral head due to a constant inferior force applied to the humerus. Results are reported as mean superior translations in millimeters of the four tested joint positions. Generally the largest translation values were seen in 30° and 60° of glenohumeral abduction. In the hanging arm position, the translations were smaller because the superior joint capsule limited inferior translation. In 90° of glenohumeral abduction the smallest translations were detected because the inferior glenohumeral ligament complex was tight (Fig. 3). The lateral deltoid (8.2 mm, SD 4.8 mm) was significantly P < 0:05 most powerful in superior translation of the humeral head throughout all positions although not signi®cantly P > 0:05 dierent from the posterior deltoid (7.7 mm, SD 4.8 mm) in 30°, 60° and 90° of glenohumeral abduction. The lateral deltoid was the only muscle tested, which was consistently able to
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
Fig. 3. Translation values of the tested muscles on average in hanging arm position, at 30°, at 60° and at 90° of glenohumeral abduction. The graph shows the superior translations of the humeral head on the glenoid eected by the tested muscles counteracting a constant inferior force. SSP: supraspinatus, ADLT: anterior deltoid, LDLT: lateral deltoid, PDLT: posterior deltoid, SHB/COB: short head of biceps/ coracobrachialis, TRZ: long head of triceps.
completely reverse inferior subluxation of the humeral head occurring in 30° and 60° of glenohumeral abduction. Apart from a line of action directed straight superiorly, the lateral and posterior deltoid had large cross-sectional areas. The anterior deltoid was signi®cantly P > 0:05 weaker in superior translation (4.9 mm, SD 3.7 mm) than the lateral and posterior deltoid because its line of action was diverted anteriorly by the humeral head and it had a smaller cross-sectional area. The coracobrachialis and short head of biceps had a considerable superior translating eect (2.8 mm, SD 1.3 mm). Their lines of action run obliquely and anteriorly to the glenoid, but they had a relatively large crosssectional area. The long head of triceps showed small superior translations (1.7 mm, SD 0.7 mm) although its line of action runs straight superiorly towards the glenoid, but it had a small cross-sectional area. Finally and surprisingly, the supraspinatus was weakest in superior translation throughout all positions (1.3 mm, SD 0.5 mm), signi®cantly P < 0:05 less effective than all portions of the deltoid. Even the short head of biceps/coracobrachialis showed signi®cantly P < 0:05 larger translation distances in the hanging arm position and at 30° of glenohumeral abduction. Its line of action is nearly horizontal and its cross-sectional area is of medium size. 4. Discussion In these experiments the lateral deltoid was the most capable of reversing the inferior translation of the humeral head, closely followed by the posterior deltoid whereas the anterior deltoid was less eective. As our study showed potential muscle eects, the true in vivo role of the deltoid depends on its recruitment during motion.
141
Traditionally the deltoid was described as an essential mover of the glenohumeral joint eecting abduction, ¯exion and extension [18]. In abduction, the deltoid becomes more ecient [19,20] and shows maximum electromyograhic activity at higher degrees [21]. Thus, removal of the deltoid leads to a progressive decrease in strength during abduction [22] and a decrease in range of motion [23]. However, only few studies have focused on the stabilizing function of the deltoid. In a cadaver study the passive stabilizing capability of the deltoid was found to be insigni®cant [24] but the tests were performed with low loads at the extremes of glenohumeral motion, when the ligamentous constraints provided sucient inferior stability. In our preliminary experiments, removal of the deltoid caused inferior subluxation of the humeral head in the mid-range of motion after venting of the joint capsule. Kuechle [25] found that the lateral deltoid has the largest elevator moment arm in the scapular plane, which implies the likelihood to be recruited for abduction function and superior translation of the humeral head action as well. In an electromyographic study, an active stabilizing eect of the deltoid was neglected because the deltoid was found to be inactive with the hanging arm carrying a weight [26], but in this position the superior joint capsule and inclination of the glenoid stabilize the glenohumeral joint inferiorly. On the other hand, a recent electromyograhic study of ®ve parts of the deltoid showed activity not related to the generation of an abduction moment and a stabilizing eect was suggested [27]. This is con®rmed by the ®nding that electromyographic activity of the anterior and lateral deltoid during abduction and ¯exion was decreased in patients with shoulder instability [28]. Finally, the deltoid was reported to be the superior part of the frontal plane force couple being counterbalanced by the inferior rotator cu muscles. Thereby it may actively stabilize the glenohumeral joint [14,29]. The short head of biceps and coracobrachialis showed considerable superior translating eects in our study. Although the long head of the biceps was found to be a superior stabilizer [30] and the whole biceps to be an inferior stabilizer [12], the results of electromyographic studies are still controversial. While several studies reported biceps activity related to shoulder ¯exion and abduction [26,32]; some did not detect it [31,33]. Our study con®rms its potential as an inferior stabilizer, but its actual physiologic role in shoulder function needs further clari®cation in future in vivo studies. The long head of triceps showed a minor superior translating eect. Only a few studies investigated the activity of the triceps reporting little or no activity during shoulder movements [34]. Therefore, it is considered to be primarily an elbow mover and its eects on the glenohumeral joint to be secondary.
142
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143
The supraspinatus is an important mover of the glenohumeral joint. Speci®cally it participates in the initiation of abduction [35,36] and in rotation [37]. Although some investigators have demonstrated a major stabilizing eect of the supraspinatus in the anterior [38] and inferior directions [12], others reported a limited contribution to joint stability [36,39]. Our study showed that the supraspinatus had a minor inferior stabilizing eect. Moreover, our ®ndings support the biomechanical suspension bridge model which assumes that normal glenohumeral joint kinematics could be restored even without complete repair of a supraspinatus tendon defect [14]. The functional result depends on the integrity of the muscular force couples in the frontal and horizontal planes.
proximation of the varied lines of action of a muscle, this appears to be a legitimate biomechanical model for muscle forces. The eects of the ligamentous tightness in the extremes of rotation and the eects of negative intraarticular pressure were excluded by testing in neutral rotation and venting the joint capsules. Thereby, we were able to focus on the muscular eects. However, in the hanging arm position and 90° of glenohumeral abduction, the ligaments limited translations. Despite these limitations, the results were reproducible with a high degree of accuracy providing a basis of con®dence in this methodology.
4.1. Limitations
In our study the deltoid was potentially most capable of translating the humeral head superiorly while the supraspinatus was least eective. Therefore, deltoid strengthening exercises may be useful in the treatment of inferior instability. The biceps, long head of triceps and coracobrachialis had a considerable capability to translate the humeral head superiorly, while the supraspinatus was less eective than previously assumed. Further studies should clarify their in vivo contributions to shoulder stability.
As simulation of muscle force and direction was crucial for the result, we aimed at the highest accuracy possible using computer-controlled pneumatic actuators and low-friction pulleys for muscle loading. Nevertheless, loading was based on the assumption that maximum muscle force is proportional to its cross-sectional area, which was derived from the literature. In each position, the muscle loading was performed sequentially, disregarding eventual muscle interactions. Furthermore, as there is a high degree of variability of individual muscle activity as a function of joint position, and consistent electromyograhic data are lacking, muscle loading was performed with maximum force. Finally, the applied constant force does not occur naturally, but represented the shear force component generated by muscles or gravity. However, the study was designed to report a ranking of individual muscles according to their potential to stabilize the glenohumeral joint inferiorly in case of maximum muscle activity in selected joint positions. In combination with the results of future in vivo studies that will investigate the activity of individual muscles at speci®c joint positions, relative muscle contributions to inferior glenohumeral joint stability can be estimated. By applying a constant inferior force of 15 N, the humeral head was translated inferiorly before muscle loading. As a result, the line of muscle action was changed. Inferior subluxation and a major change in line of muscle action were prevented by leaving the muscle belly of the deltoid intact. The supraspinatus was aected most of all among the muscles tested. And, although the supraspinatus should be even more powerful at pulling the humeral head superiorly when starting from an inferior position, it was least eective. To enhance the accuracy of the model, the strings representing lines of muscle action were adjusted according to the centroids of the tested muscles as derived from the literature. Although strings are merely an ap-
5. Conclusions
Acknowledgements The ®rst author was supported by the Max-Biedermann-Institut, Berlin, Germany. The authors wish to acknowledge D. Larson from the section of Biostatistics for the statistical calculations and L. Berge from the Orthopedic Biomechanics Laboratory for the drawing of the ®gures.
References [1] Warner JP, Deng XH, Warren RF, Torzilli PA. Static capsuloligamentous restraints to superior±inferior translation of the glenohumeral joint. Am J Sports Med 1992;20:675±85. [2] Ovesen J, Nielsen S. Experimental distal subluxation in the glenohumeral joint. Arch Orthop Trauma Surg 1985;104:78±81. [3] Itoi E, Berglund LJ, Grabowski JJ, Naggar L, Morrey BF, An KN. Superior±inferior stability of the shoulder: the role of the coracohumeral ligament and the rotator interval capsule. Mayo Clin Proc 1998;73:508±15. [4] Harryman DT, Sidles JA, Harris SL, Matsen FA. The role of the rotator interval capsule in passive motion and stability of the shoulder. J Bone J Surg (Am) 1992;74-A:53±66. [5] Field LD, Warren RF, O'Brien SJ, Altchek DW, Wickiewicz T. Isolated closure of rotator interval defects for shoulder instability. Am J Sports Med 1995;23:557±63. [6] Gibb TD, Sidles JA, Harryman DT, McQuade KJ, Matsen FA. The eect of capsular venting on glenohumeral laxity. Clin Orthop 1991;268:120±7.
A.M. Halder et al. / Clinical Biomechanics 16 (2001) 138±143 [7] Itoi E, Motzkin NE, Browne AO, Homeyer P, Morrey BF, An KN. Intraarticular pressure of the shoulder. Arthroscopy 1993;9:406±13. [8] Wuelker N, Korell M, Thren K. Dynamic glenohumeral joint stability. J Shoulder Elbow Surg 1998;7:43±52. [9] Itoi E, Motzkin NE, Morrey BF, An KN. Scapular inclination and inferior stability of the shoulder. J Shoulder Elbow Surg 1992;1:131±9. [10] Pagnani MJ, Deng XH, Warren RF, Torzilli PA, Altcheck DW. Eect of lesions of the superior portion of the glenoid labrum on glenohumeral translation. J Bone J Surg (Am) 1995;77-A:1003± 10. [11] Hsu HC, Luo ZP, Co®eld RH, An KN. In¯uence of rotator cu tearing on glenohumeral stability. J Shoulder Elbow Surg 1996;6:413±22. [12] Soslowsky LJ, Malicky DM, Blasier RB. Active and passive factors in inferior glenohumeral stabilization: a biomechanical model. J Shoulder Elbow Surg 1997;6:371±9. [13] Warner JJ, Bowen MK, Deng X, Torzilli PA, Warren RF. Eect of joint compression on inferior stability of the glenohumeral joint. J Shoulder Elbow Surg 1999;8:31±6. [14] Burkhart SS. Fluoroscopic comparison of kinematic patterns in massive rotator cu tears. A suspension bridge model. Clin Orthop 1992;234:144±52. [15] Johnson GR, Spalding D, Nowitzke A, Bogduk N. Modelling the muscles of the scapula morphometric and coordinate data and functional implications. J Biomech 1996;29:1039±51. [16] Veeger HE, Yu B, An KN, Rozendal RH. Parameters for modeling the upper extremity. J Biomech 1997;30:647±52. [17] Tirosh R. 1 kgf/cm2 ± the isometric tension of muscle contraction: implications to cross-bridge and hydraulic mechanisms. Adv Exp Med Biol 1984;170:531±9. [18] Inman VT, Saunders JB, Abbott LC. Observations on the function of the shoulder joint. J Bone J Surg (Am) 1944;16A:268±71. [19] Otis JC, Jiang CC, Wickiewicz TL, Peterson MG, Warren RF, Santner TJ. Changes in the moment arms of the rotator cu and deltoid muscles with abduction and rotation. J Bone J Surg (Am) 1994;76:667±76. [20] McMahon PJ, Debski RE, Thompson WO, Warner JJ, Fu FH, Woo SL. Shoulder muscle forces and tendon excursions during glenohumeral abduction in the scapular plane. J Shoulder Elbow Surg 1995;4:199±208. [21] Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination in the normal shoulder. An electromyographic study. Clin Orthop 1990;257:76±85. [22] Colachis SC, Strohm BR, Brechner VL. Eects of axillary nerve block on muscle force in the upper extremity. Arch Phys Med Rehabil 1969;50:647±54.
143
[23] Markhede G, Monastyrski J, Stener B. Shoulder function after deltoid muscle removal. Acta Orthop Scand 1985;56:242±4. [24] Motzkin NE, Itoi E, Morrey BF, An KN. Contribution of passive bulk tissues and deltoid to static inferior glenohumeral stability. J Shoulder Elbow Surg 1994;3:313±9. [25] Kuechle DK, Newman SR, Itoi E, Morrey BF, An KN. Shoulder muscle moment arms during horizontal ¯exion and elevation. J Shoulder Elbow Surg 1997;6:429±39. [26] Basmajian JV, Latif MA. Integrated actions and functions of the chief ¯exors of the elbow. J Bone J Surg (Am) 1957;39-A:1106±18. [27] Michiels I, Bodem F. The deltoid muscle: an electromyographical analysis of its activity in arm abduction in various body postures. Int Orthop 1992;16:268±71. [28] Kronberg M, Brostrom LA, Nemeth G. Dierences in shoulder muscle activity between patients with generalized joint laxity and normal controls. Clin Orthop 1991;269:181±92. [29] Lucas D. Biomechanics of the shoulder joint. Arch Surg 1973;107:425±32. [30] Warner JJ, McMahon PJ. The role of the long head of biceps brachii in superior stability of the glenohumeral joint. J Bone J Surg (Am) 1995;77-A:366±72. [31] Furlani J. Electromyographic study of the m. biceps brachii in movements at the glenohumeral joint. Acta Anat Basel 1976;96:270±84. [32] Sakurai G, Ozaki J, Tomita Y, Nishimoto K, Tamai S. Electromyographic analysis of shoulder joint function of the biceps brachii muscle during isometric contraction. Clin Orthop 1998;354:123±31. [33] Yamaguchi K, Riew KD, Galatz LM, Syme JA, Neviaser RJ. Biceps activity during shoulder motion: an electromyographic analysis. Clin Orthop 1997;336:122±9. [34] Basmajan JV, Bazant FJ. Factors preventing downward dislocation of the adducted shoulder joint. J Bone J Surg (Am) 1959;41A:1182±6. [35] Howell SM, Imobersteg AM, Seger DH, Marone PJ. Clari®cation of the role of the supraspinatus muscle in shoulder function. J Bone J Surg (Am) 1986;68-A:398±404. [36] Thompson WO, Debski RE, Boardman ND, Taskiran E, Warner JJ, Fu FH, Woo SL. A biomechanical analysis of rotator cu de®ciency in a cadaveric model. Am J Sports Med 1996;24:286± 92. [37] Ihashi K, Matsushita N, Yagi R, Handa Y. Rotational action of the supraspinatus muscle on the shoulder joint. J Electromyogr Kinesiol 1998;8:337±46. [38] Malicky DM, Soslowsky LJ, Blasier RB, Shyr Y. Anterior glenohumeral stabilization factors: progressive eects in a biomechanical model. J Orthop Res 1996;14:282±8. [39] Sharkey NA, Marder RA. The rotator cu opposes superior translation of the humeral head. Am J Sports Med 1995;23:270±5.