Glenohumeral stability fromconcavity-compression: Aquantitative analysis

Steven B. Lippitt, MD, J. Eric Vanderhooft, MD, Scott L. Harris, MS,John A. Sidles, PhD, Douglas T. Harryman II, MD, andFrederick A. Matsen /11, MO, Seattle, Wash.

The purpose of this research was to determine the degree to which compressionof the humeral head into the glenoid concavity stabilizes it against translatingforces. Ten normal fresh-frozen cadaver glenohumeral ;oints in which the labrumwas preserved were used. A compressive load of 50 N was applied to thehumeral head in a direction perpendicular to the glenoid surface. Increasingtangential forces were then applied until the head dislocated over the glenoidlip. The tangential force at dislocation was examined for eight diHerentdirections, 45° apart around the glenoid. Concavity-compression stability wasthen examined for an increased compressive load of 700 N. Finally, the protocolwith 50 and 700 N of compressive load was repeated after the glenoid labrumwas excised. Concavity-compression of the humeral head into the glenoid is amost eHicient stabilizing mechanism. With the labrum intact the humeral headresisted tangential forces of up to 60% of the compressive load. The degree ofcompression stabilization varied around the circumference of the glenoid withthe greatest magnitude superiorly and inferiorly. This may be attributed to thegreater glenoid depth in these directions. Resection of the glenoid labrumreduced the eHectiveness of compression stabilization by approximately 20%.These results indicate that concavity-compression may be an importantmechanism for providing stability in the mid-range of glenohumeral motionwhere the capsule and ligaments are lax. The eHectiveness is enhanced by thepresence of an intact glenoid labrum. (J SHOULDER ELBOW SURG 7993;2:27-35.)

The mechanisms by which the glenohumeraljoint is stabilized in the mid-range of motionhave not been well-defined. Traditionally, gle­nohumeral stability is thought to be provided bythe ligaments and capsule. Although thesestructures are under tension at the extremes ofmotion, in the mid-range they are lax and can­not exert major stabilizing effects.

Among the possible mechanisms for provid­ing stability in the mid-range of glenohumeral

From the Department of Orthopaedics, University of Wash­Ington, Seattle.

Reprint requests. Steven B. Lippitt, MD, Department of Or­thopaedics, RK-1 0, University of Washington, Seattle, WA98195.

Copyright © 1993 by Journal of Shoulder and Elbow SurgeryBoard of Trustees.

1058-2746/93/$100 +.10 3211/45852

motion, one of the more important may be thecompression of the convex humeral head intothe concave glenoid fossa. The specializedanatomy of the rotator cuff is ideally situated toprovide a compressive load throughout therange of motion of the glenohumeral joint. Thepotential effectiveness of the concavity­compression mechanism is limited by thestrength of the rotator cuff and other compress­ing muscles, as well as by the relatively smallsize of the glenoid fossa, which is only onefourth the size of the articular surface of thehumeral head.

The purpose of this research was to determinethe degree to which concavity compression canstabilize the humeral head against translatingforces. Recognizing that the anatomy of the gle­noid fossa changes around its circumference,we compared the effectiveness of concavity

27

28 Lippitt et 01.

/

Tran lation.......- Force

Force Tran. ducer

Figure 1 Progressively larger translating forces wereapplied manually to humeral head as it was com­pressed into glenoid concavity until humeral head dis­located. Magnitude and direction of these translatingforces were measured by force transducer and spatialsensors.

compression for different directions of trans­lating force. 3 Finally, realizing that the glenoidlabrum appears to contribute to the depth of theglenoid concavity, we compared the effective­ness of concavity compression before and afterresection of the labrum. 12

METHODSWe studied 10 normal fresh-frozen cadaver

glenohumeral joints. The average age of thepatients at death was 73 years (range 55 to89 years). All included shoulder specimenshad intact articular cartilage of the glenoid andhumeral head and were free from tearing orgross degenerative changes in the glenoidlabrum.

The specimens were prepared by sectioningthe scapula from the thorax and dividing thehumerus at mid-shaft. The muscles and tendonsof the rotator cuff, biceps, and deltoid were re­sected. The glenohumeral joint capsule was dis­sected from its humeral insertion and was thenresected just lateral to the glenoid labrum. Thelabral attachment to the glenoid rim was pre­served.

The vertebral border of each scapula was

J. Shoulder Elbow Surg.January / February 7993

potted in a plastic container with plaster ofParis. This container was fixed rigidly to a ver­tically mounted force transducer (Astek Model,FS160A-600, Barry Wright Corporation, Water­town, Mass.) that had 6° of freedom so that theglenoid fossa was facing upward, parallel tothe floor. Once calibrated, the Astek transducerallowed measurement of both the magnitudeand direction of manually applied forces. Byconvention the superior aspect of the glenoidwas designated as 0° and the anterior aspectas 90°.

To maximize the congruity of the glenohu­meral joint, the center of the humeral head ar­ticular surface was aligned symmetrically to thecenter of the glenoid fossa. This roughly cor­responded to placing the humerus in neutralflexion with 45° of abduction and 35° of externalrotation. A 5/8 inch wooden dowel was placedtransversely through the humeral head (abovethe articular surface) to establish an antero­posterior alignment guide. Weights were at­tached to this dowel to compress the humeralhead into the glenoid fossa. Two additional1/4 inch dowels were placed at right anglesinto the humeral head to establish superoinfer­ior and medial-lateral alignment guides. Thestarting position of the humeral head in the gle­noid for each experiment was achieved by com­pressing the head into the glenoid and thenaligning the dowel guides to the superior, an­terior, and vertical axes of the glenoid. The sen­sors of a spatial digitizer (Polhemus, Naviga­tional Sciences, Colchester, Vt.) with 6° of free­dom were attached to the humeral shaft andscapula to allow measurement ofthe translationand rotation of the humeral head with respectto the glenoid fossa. This system has beenshown to be accurate to within 1 mm of trans­lation and 1° of angular motion.9

-11 Data were

recorded during the trials by computer (Or­thokine software, University of Washington, Se­attle, Wash.).

With essentially no load applied, the humeralhead was translated along the 0° axis from thecenter of the glenoid to the superior glenoidedge and was next translated along the 180°axis from the center to the inferior glenoid edge.The process was then repeated in the 45° to 225°plane, the 90° to 270° plane, and the 135° to315° plane. The effective concavity of the gle­noid fossa was demonstrated by measuring the

J. Shoulder Elbow Surg.Volume 2, Number 1, Part 1

Humeral Head

Lippitt et 01. 29

lAbnlm -M....

Cartilage

A Glenoid

e

E 7E

6

CQIE 5III(,)CD

Q.'"o 3

GLENOID CONCAVITY

0"/180" Plane

90"/270" Plane

90·

oi--......-or--_-i--....---.-_-.....,10 10 20

B Translation (mm) from Center of the Glenoid

Figure 2 A Path of center of head as it is translated away from glenoidcenter is shown by "Y" -shaped graph. Effective depth of glenoid concavityin specified direction of translation is defined by maximum lateral displace­ment of humeral head at glenoid edge relative to its starting point centeredin glenoid fossa. B. Effective depth of glenoid in superoinferior directionand anteroposterior direction for typical specimen.

lateral displacement of the humeral head as itwas translated up the contour of the glenoidsurface.

Next, a compressive load of 50 N was ap­plied to the humeral head in a direction per­pendicular to the glenoid surface. To determinethe stability of the head against a defined di­rection of translation, progressively largertranslating forces were applied manually to thehumeral head while the magnitude and direc­tion of these forces were measured by the forcetransducer and spatial sensors (Fig. 1). The

forces were applied tangential to the glenoidsurface until the head dislocated over the gle­noid rim. This dislocation occurred abruptly sothat the maximal force before dislocation waseasily discernible. The translation force at dis­location was recorded for eight different direc­tions at 45° intervals (superior: 0°, anterosu­perior: 45°, anterior: 90°, etc.). To determine theeffect of the magnitud~ of the compressive loadon concavity compression stability, the se­quence was repeated with a 100 N compressiveload (in this sequence the 0°, 90°, 180°, and 270°

30 Lippitt et al.

40% ±16 ()

3

50~ ±19

Figure 3 Stability ratio for 10 glenohumeral joints withlabrum Intact and 50 N compressive load laver­age ± 1 SD). Stability ratio is given as percentage bydefined equation: stability ratio (%) = translatingforce at dislocation/compressive load x 100.

directions were examined). Finally the entireprotocol was repeated after the glenoid labrumwas excised to the level of the articularcartilage.

In an attempt to prevent desiccation and tomaintain constant friction between the glenoidand the humeral head, the articular surfaceswere kept moist by periodic irrigation with nor­mal saline solution.

Because the same specimens were testedmany times under different conditions (groups:50 N versus 100 N compressive load, with andwithout labrum), the repeated-measures anal­ysis of variance (ANOYA) was used for statis­tical analysis. Intragroup comparisons of thestability for the different directions of transla­tion force (0°,45°,90°, etc.) were analyzed withthe Dunnett ttest. Statistical significance was setat the p < 0.05 level.

RESULTSGlenoid concavity and eHective gle­

noid depth. The effective depth of a glenoidin a specified direction was determined by plot­ting the lateral displacement of the humeralhead as it was translated in the specified di­rection from the glenoid center up the surfaceof the glenoid concavity. This plot created a"Y" -shaped graph (Fig. 2A and B), which rep­resents the changing slope of the glenoid con­cavity in contact with the humeral head. Theeffective depth of the concavity is the maximum

J. Shoulder Elbow Surg.January / February 1993

Table I Translating force resisted forcompressive loads

Compressive lood Tronslating 'orceDegree (N) (N)

0 50 29 ::':: 7100 51 ± 9

90 50 17 ± 6100 29 ± 5

180 50 32 ± 4100 56 ± 12

270 50 17 ± 6100 30 ± 12

lateral translation of the humeral head at theglenoid edge compared with the starting pointat the glenoid center (Fig. 2A).

In this group of shoulders the average effec­tive glenoid depths varied around the perimeterof the glenoid as shown for a typical shoulderin Fig. 2B. For all of the shoulders the averageglenoid depths were greater superiorly (4.8mm) and inferiorly (4.9 mm) than anteriorly (2.2mm) and posteriorly (2.1 mm).

Stability from concavity compression.The following data show that the effectivenessof concavity compression was affected by themagnitude of the compressive load, the direc­tion of the translating force, the presence of theglenoid labrum, and the effective depth of theglenoid concavity.

Magnitude of the compressive loadFifty N compressive loads stabilized the hu­meral head against translating forces up to 32N, depending on the direction of the translatingforce (see following). Increasing the compres­sive load to 100 N increased humeral head sta­bility against even greater translating forcesmeasuring of up to 56 N (See Table I).

To facilitate the analysis of the effectivenessof concavity compression under different con­ditions, a "stability ratio" was calculated assuggested by Fukuda et al.e For conveniencethis ratio was expressed as a percentage in­stead of a decimel by the following equation:stability ratio (%) = translating force at dislo­cation / compressive load x 100.

Directional effect of the translatingforce. The stability ratio varied around theperimeter of the glenoid (Fig. 3). For all prep­arations the stability ratios were significantlygreater (p < 0.05) superiorly (0°) and inferiorly

J. Shoulder Elbow Surg.Volume 2, Number 1, Part 1

Table II Change in the stability ratio with removal of the labrum

Lippitt et 01. 31

Stability ratio at 50 N compressive load Stability ratio at 100 N compressive load

Labrum Intad Labrum excised Stability ratio I Labrum excised IDegree (%) (%) pValue (%) (%) p Value

0 59 ± 13 47 ± 8 0.012 51 ± 9 45 ± 7 0.02845 38 ± 11 30 ± 10 0.03290 35 ± 11 28 ± 6 0.062 29 ± 5 26 ± 5 0.021

135 46 ± 6 39 ± 8 0.008180 64 ± 8 41 ± 13 0.0001 56 ± 12 40 ± 13 0.0002225 50 ± 19 30 ± 10 0.001270 33 ± 12 25 ± 11 0.017 30 ±12 23 ± 9 0.01315 40 ± 16 32 ± 14 0.032

Data reported as average (n = 10) ± 1 standard deviation.Statistical significance p < 0.05.

(180°) than anteriorly (90°) or posteriorly (270°).In contrast, there was no significant differencein the stability ratio (p < 0.05) between the su­perior and the inferior directions nor betweenthe anterior and the posterior directions.

The eHect of excision of the glenoidlabrum. Excision of the glenoid labrum de­creased the effectiveness of concavity compres­sion as indicated by diminished stability ratios(Table II). This was true for all directions of dis­placement and for both magnitudes of com­pressive loading.

The percentage of labral contribution to thestability ratio (SR) for a given direction ofdisplacement was calculated by the formula:labral contribution (%) = (SR labrum intact) ­(SR labrum excised) /SR labrum intact. This la­bral contribution to stability through concavity­compression was variable, averaging 20%, anddid not differ for the different directions oftranslation at the 95% confidence level. How­ever, there was a trend toward greater contri­bution of the labrum to the stability ratio in theinferior and posteroinferior directions (average37%) as compared with all other directions onthe glenoid (average 18%).

Relationship between stability ratioand effedive glenoid depth. As discussedpreviously, the effective depth of the glenoidconcavity was greater in the superoinferior thanit was in the anteroposterior plane. The increased stability ratios in the superoinferior di­rections correlated with the increased effectivedepth of the glenoid concavity in this plane. In­deed, both the depth and stability in the su­peroinferior direction were approximately

twice the respective values in the anteroposte­rior direction. This relationship of depth to con­cavity-compression stability was also demon­strated by comparing the stability ratios afterresection of the glenoid labrum, which de­creased the effective depth of the glenoid con­cavity. This relationship between effectivedepth and stability ratio appeared to be linear(Fig. 4).

DISCUSSIONThe glenohumeral joint provides the greatest

range of motion of any joint in the body but alsois one of the most susceptible to instability.1S, 21Because the glenohumeral ligaments and cap­sule must have sufficient laxity to allow the im­mense range, they can provide stability onlywhen they become tight at the extremes of gle­nohumeral motion. 16, 24, 26, 27 Other factors mustcontribute to glenohumeral stability in mid­range glenohumeral positions, where the armis most commonly positioned.

The glenoid articular geometry may seem tooshallow to provide a significant stabilizing rolefor the humeral head. Though the glenoid fossais only one fourth the size of the articular surfaceof the humeral head, it is not flat. Radiographsare misleading because they show only thebony contour of the glenoid in relation to thehumeral head. Flatow et al.7demonstrated thatalthough the subchondral bone of the glenoidis relatively flat, the cartilage surface is thickerperipherally than in the center, producing a ra­dius of curvature that matches the humeral headvery closely.

Radiographic studies by Howell et al. 13 and

32 Lippitt et 01.

70

60~

~ 50

.2'Iii 40a:

~30

.Q20III

iii10

0

0

J. Shoulder Elbow Surg.January / February 7993

DEPTH GLENOID CONCAVITY vs.

STABILITY RATIO

o 5 1 1 5 2 2.5 3 3 5 4 4.5 5

Effective Depth Glenoid Concavity (mm)

Figure 4 Average data for labrum-intact and labrum-exCised specimenswith 50 N compressive load for superior, Inferior, anterior, and posteriordirections. linear regression by least-squares method provides formula:stability ratio (%) = 10.3 (effective depth glenoid concavity In mm) + 9.7.

by Poppen and Walker18 demonstrate that thehumeral head essentially remains centered onthe glenoid throughout active motion in the fron­tal and horizontal planes, respectively. Thesestudies support a "ball and socket" kinematicmodel of the humeral head rotating in a con­gruent glenoid concavity during active motion.

In our study we demonstrated the effectiveglenoid depth of the concavity by observing thelateral displacement of the humeral head as itwas translated from the glenoid center to theglenoid rim in various directions. The depth ofthe concavity was the greatest in the superiorand inferior directions. This is consistent withthe results of Howell and Galinat/ 2 who iden­tified the glenoid fossa as deeper in the su­peroinferior direction, averaging approxi­mately 9 mm as compared with the antero­posterior direction, which averages 5 mm. Thismay be explained by understanding the pear­shaped anatomy of the glenoid fossa. Becausethe superoinferior dimension is wider than theanteroinferior direction, the glenoid fossa mustbe deeper in this direction, given that the gle­noid radius of curvature has minimal deviationfrom sphericity (Fig. 5).7

The functional importance of this concavitywas demonstrated in the present study by com­pressing the humeral head into the glenoidfossa and applying a tangential translationforce until dislocation occurred. The maximumstability ratio (translation force divided by the

compressive load) ranged from 33% to as muchas 64% for the labrum-intact glenoid speci­mens. The greater stability ratio for the superiorand inferior directions (64%) as compared withthe anterior and posterior directions (33% to35%) was related to the greater effective depthin this plane (approximately 4.8 mm versus 2.2mm). In fact, an almost linear relationship ex­isted between the effective glenoid depth andits stability ratio for any given direction.

The role of the glenoid labrum as a stabilizerof the glenohumeral joint was popularized byBankart. 1 The labrum is not only a significantsite of attachment of the glenohumeral liga­ments but also provides a fibrous extension tothe glenoid rim, which further enhances thedepth of the fossa. 5

• 6.16 Howell et a1. 12• 14 pre­

sented the concept of the II glenoid-Iabral"socket, in which the circular, pliable, fibrous la­brum contributed approximately 50% to the to­tal depth of the glenoid. In our study the stabilityratio was significantly less after the glenoid la­brum was excised for each direction tested. Onthe average, the labrum contributed approxi­mately 20% to the concavity-compression sta­bilization of the glenohumeral joint.

Cooper et al.3 investigated the anatomy of theglenoid labrum in cadaver shoulders and founddistinct morphologic differences around theglenoid rim. The superior and anterosuperiorlabrum appeared meniscal with a loose attach­ment to the glenoid, at times allowing the la-

J. Shoulder Elbow Surg.Volume 2, Number 7, Part 7

Po teriorU teriO

'

Lippitt et 01. 33

Figure 5 Width of glenoid in superoinferior direction (Ws,) is greater thanwidth of glenoid in anteroposterior direction (WAP). For given radius ofcurvature (rJ, increase in width results in increase in depth. Thus depth ofglenoid in superoinferior direction (ds,) is greater than depth of glenoid inanteroposterior direction (dAP )'

brum to be quite mobile. In contrast, the inferiorportion of the labrum represented an immobile,intimately attached fibrous extension of the ar­ticular cartilage. They concluded that this firmattachment of the inferior labrum to the glenoidrim provided evidence of its role as a stabilizingbumper against translation of the humeralhead. Our results are consistent with this con­cept, showing a highly suggestive trend towardgreater contribution of the labrum to inferiorand posterior inferior stability than that of otherdirections on the glenoid.

The quality and size of the labrum may varyamong shoulders. DePalma et al.5 reported theattritional changes that occur in the glenoid la­brum with increasing age, such as detachmentand fraying. Our experiment used older shoul­der specimens, ranging from 55 to 89 years ofage, though we included only those withoutgross evidence of degenerative causes. It ispossible that in younger shoulders with morerobust labra, the labra may demonstrate aneven greater effect on glenohumeral stability.

Clinical support exists for the concept of con­cavity-compression as a stabilizing mechanismfor the glenohumeral joint. Glenoid rim frac­tures involving significant loss of the articularsurface (glenoid concavity) can be associatedwith glenohumeral instability.23 Repair of gle­noid labrum and glenohumeral ligament de­tachment (Bankart lesion) back to the glenoidrim has been documented as an effective treat­ment for traumatic anterior instability.20.25 Ananatomic glenohumeral ligament repair notonly allows firm attachment of the ligaments but

also helps restore the concavity to the glenoidfossa.

Fukuda et al.8 demonstrated the principle ofconcavity-compression in various designs ofglenohumeral joint prosthetic replacement.Their study reported the amount of force re­quired to cause subluxation and dislocation ofthe prosthetic glenohumeral joint under knownaxial load. Among the designs, the glenoidcomponents with the largest curvature had thehighest resistance to subluxation of the corre­sponding humeral head component. In allcases, the anteroposterior subluxation forceswere lower than the corresponding superoin­ferior subluxation forces. This was attributed tothe variation of constraining wall height pro­vided by the glenoid surface. They related thepresence of an axial force as the most importantfactor in maintaining joint stability.

The magnitude and direction of the nor­mal compressive forces acting at the gleno­humeral joint have been estimated by variousauthors.2,4, 14, 17-19,23,28 Walker and Poppen l9,28showed the resultant joint contact force in­creased linearly with progressive arm abduc­tion to a maximum of 0.89 times body weight at90° abduction. The great majority of the com­pressive forces are provided by the rotator cuffmuscles: subscapularis, supraspinatus, infra­spinatus, and teres minor, and by the long headof the biceps and the deltoid and pectoral mus­cles in certain positions.

The importance of concavity compression asa stabilizing mechanism for the glenohumeraljoint may be demonstrated in the normal indi-

34 Lippitt et al.

vidual. When the subject is completely relaxed,an examiner can easily translate the humeralhead anteriorly or posteriorly with respect tothe glenoid. If the subject then contracts theshoulder muscles (e.g., by slightly abducting theshoulder), this anteroposterior excursion is vir­tuallyeliminated.

CLINICAL RELEVANCEOur study investigated the stabilizing effect

of concavity-compression in the glenohumeraljoint in vitro. The re:;ults indicate that compres­sion of the humeral head into the concave gle­noid fossa can stabilize the articulation effec­tively. This mechanism can act throughout therange of motion, including in the mid-range,where the capsule and ligaments are lax. Con­cavity-compression stabilization is increasedwith increasing compressive loads and in di­rections with greater effective glenoid depth(superior and inferior). This mechanism is sig­nificantly compromised by the absence of anintact glenoid labrum.

These results may be relevant to clinical trau­matic instability. They suggest that detachmentor surgical resection of the glenoid labrum maydecrease shoulder stability, giving rise to sub­luxation or dislocations. They suggest that re­establishment of the effective glenoid depththrough anatomic repair may be an importantelement of the surgical repair for recurrent gle­nohumeral instability.

These results may also be relevant to clinicalatraumatic multidirectional instability, wherethe joint is characteristically unstable in the mid­range of motion. Flatness of the glenoid artic­ular surface could predispose a joint to this typeof instability, allowing relatively easy transla­tion in multiple directions as a result of a lackof effective glenoid depth. They also suggestedthat rotator-cuff strengthening may enhancestability by providing for increased compressiveloads.

REFERENCESBankart ASB The pathology and treatment ot recurrentdislocation of the shoulder. Br J Surg 1938;26:23-9

2 Cain PR, Mutschler TA, Fu FH, Lee SK Antenor Instabilityof the glenohumerallolnt a dynamiC model. Am J SportsMed 1987,15144-8

3 Cooper DE, Arnoczky SP, O'Bnen SJ, Warren RF, D,­Carlo E, Allen AA. Anatomy, histology, and vosculantyof the glenOid labrum an anatomical study. J Bone JOintSurg [Am] 1992,74A 46-52.

J. Shoulder Elbow Surg.January / February 1993

4 DeDuca CJ, Forrest WJ Force analySIS ot Ind,vldua!muscles acting simultaneously on the shoulder lolnt durIng Isometnc abduclion J Biomech 1973,6 385-93

5. DePalma AF, Callery G, Bennett GA Vanatlonal anatomy and degenerative leSions of the shoulder 10lnt InBlount WP, ed AAOS Instructional Course Lectures1949;6.255-81

6 Detnsac DA, Johnson LL ArthroscopIc shoulder anm·omy Thorofare NJ Slack Publ, 1986

7 Flotow EL, Soslowsky LJ, Ateshlan GA, et:l. ShoulderjOint anatomy and the effect of subluxatlons and sizemismatch on patterns of glenohumeral contact OrthopTrans 1991,15'803-4

8 Fukuda K, Chang·Mlng C. Cofield RH, Chao EY B,o,mechanical analYSIS of stability and fixation strength oftotal shoulder prostheSIS Orthopedics 1988,11 141-9

9 Harryman II DT, Sidles JA, Clark JM, McQuade KJ, GlbbTD, Matsen III FA Translation of the humeral head onthe glenOid With passive glenohumeral motion J BoneJoint Surg [Am] 1990,72A 1334-43

10 Harryman II DT, Sidles JA, Matsen III f-A fhe humeralhead translates on the glenOid With passive motion InHowklns RJ, Morrey BF, Post M, ed Surgery of the shou:·del' St LOUIS Mosby Year Book, Inc, 1990 186-90

11 Harryman II DT, Sidles JA, Horns SL, Matsen III FA Therole of the rotator Interval capsule In passive motion andstability of the shoulder. J Bone JOint Surg [Am]1992,74A53-66

12 Howell SM, Gallnat BJ. The glenold·labral socket a canstrained articular surface Clin Orthop 1989;243 122-5

13 Howell SM, Gallnat BJ, RenZI AJ, Marone PJ Normu:and abnormal mechaniCS of the glenohumeral 10lnt Inthe honzontal plone J Bone JOint Surg [Am:1988;70A227-32

14 Howell SM, Kraft TA. fhe role of the supraspinatus andinfraspinatus muscles In glenohumeral kinematics of an·tenor shoulder Instability On Orthop 1991,263 12834

15 Inman VT, Sounders JB, Abbott LC, et 01 Observatlowof the funclion of the shoulder 10lnt J Bone JOint Surq[Am] 1944,26Al-30 .

16 Moseley HF, Overgaard K. The antenor capsular mech·anlsm in recurrent antenor dislocation of the shouldermorphological and clinical studies With speCial referencE'to the glenOid labrum and the gleno-humerailigamentsJ Bone JOint Surg [Br] 1962,44B 913-27

17 Perry J. Anatomy and biomechaniCS of the shoulder If'

thrOWing, swimming, gymnastics, and tennis CI,n SpOilsMed 1983,2:247·70

18 Poppen NK, Walker PS. Normal and abnormal motionof the shoulder J Bone JOint Surg [Am] 1976,58A 195·201

19 Poppen NK, Walker PS. Forces at the glenohumerallolnrIn abduction Clin Orthop 1978; 135: 165-70

20 Rowe CR, Patel D, Southmayd WW The Bankart pro·cedure' a long-term end-result study J Bone JOint Surg[Am] 1978;60A 1-16

21 Saha AK DynamiC stabdity of the glenohumeral 10lntActa Orthop Scand 1971;42491·505

22 Saha AK MechaniCS of elevation of the gienohumeluljoint. Acta Orthop Scand 1973;44.668-78

23 Steinman S, Bigllani LU, Mcilveen SJ GlenOid fracturesassoCiated With recurrent anterior dislocation of the

J. Shoulder Elbow Surg.Volume 2, Number 7, Pari 7

shoulder. Presented at the Fifty-seventh annual meetingof the Amencan Academy of Orthopedic Surgeons, NewOrleans, Louisiana, February 8-13, 1990

24 Terry GC, Hammon D, France P, Norwood LA The sta­biliZing function of passive shoulder restraints Am JSports Med 1991,19:26-34

25. Thomas SC, Matsen III FA. An approach to the repairof avulsion of the glenohumeral ligaments in the man­agement of traumatic anterior glenohumeral Instability.J Bone JOint Surg [AmI 1989;71 A:506-13

Lippitt et 01. 35

26. Townley CO. The capsular mechanism in recurrent dis­!ocallon of the shoulder. J Bone Joint Surg [Am]1950,32A:370-80.

27 Turkel SJ, Panlo MW, Marshall JL, Girgis FG Stabilizingmechanisms preventing anterior dislocation of the gle­nohumeral joint J Bone Joint Surg [Am] 1981 ,63A 1208­17

28. Walker PS, Poppen NK. Biomechanics of the shoulderjoint during abduction in the plane of the scapula. BullHosp Jt Dis Orthop Inst 1977;38 107-11.

Bound Volumes Available to SubscribersBound volumes of the 1993 issues of JOURNAL OF SHOULDER AND ELBOW SURGERY

are available to subscribers (only) from the publisher, at a cost of $46.50 fordomestic, $55.76 for Canadian, and $52.50 for international subscribers, forVol. 1 (January-December); shipping charges are included. Each bound volumecontains subject and author indexes, and all advertising is removed. Copiesare shipped within 60 days after publication of the last issue in the volume. Thebinding is durable buckram, with the journal name, volume number, and yearstamped in gold on the spine. Payment must accompany all orders. ContactMosby, Subscription Services, 11830 Westline Industrial Dr., St. Louis, MO63146-3318, USA; phone (314)453-4351 or (800)325-4177, ext. 4351.

Subscriptions must be in force to qualify. Bound volumes are notavailable in place of a regular subscription.