Levator Scapulae Action during Shoulder Movement: A Possible Mechanism for Shoulder Pain of Cervical Origin

The phenomenon of shoulder pain of cervical JOHN FREDDY BEHRSIN origin being reproduced on shoulder movement is clinically recognized. The action of the shoul­ John Behrsin, B.App. Sc., Post Grad. Dip. Manip. (WAin, is currently working in private practice and der girdle muscles is a hypothetical cause of undertaking M.Sc. studies (part-time) at the University the cervical stress. of Melbourne. This study examined the mode and degree of Levator Scapulae activity during shoulder activ­ KEN MAGUIRE ity. Electromyography and x-rays were used to Ken Maguire, M.B.B.S., B. Med. Sc. (Hons), F.R.A.C.P., measure levator scapulae activity and length. is Consultant Physician at the Royal Perth (Reha­ The results of the study show that levator sca­ bilitation) Hospital, Western Australia. pulae contracts concentrically during the first 90 degrees of shoulder abduction and eccent­ rically during the second 90 degrees. The action of levator scapulae may be re­ sponsible for the application of force on the cervical spine during shoulder abduction. This force might cause cervical joi11t tissue distor­ tion and pain ifa pathological State was present.

The incidence of distally referred mechanism of cervical nerve root ten­ to assess whether a possible mechanism pain arising from irritation of vertebral sion resulting from shoulder move­ for the production of stress and pain structures is a well documented phe­ ment. This mechanism does not ac­ lay in their action. nomenon (Kellgren 1939, Cloward count for those patients who have A number of studies have examined 1959, Hockaday and Whitty 1967, negative nerve root signs and symp­ shoulder girdle muscle activity electro­ Mooney and Robertson 1977, McCall toms of purely cervical joint origin. myographically (Inman et af 1944, et af 1979). The clinical picture of a Muscle actions over a joint produce Yamshon and Bierman 1948, DeFreitas patient presenting with shoulder pain forces which may result in movement et af 1979, 1980, Hagberg 1981) but which is reproduced on certain activi­ but also create compressive forces be­ none are definitive due to methodo­ ties of the shoulder, but which, on a tween the joint surfaces (White and logical flaws. No studies were found more thorough examination, is re­ Panjabi 1978). During active shoulder which examined the mode of contrac­ vealed to have a cervical origin is com­ function, cervical movement may be tion (ie concentric, eccentric or iso­ mon (Cinquegrana 1968, Maigne 1975, prevented by synergistic muscle actions metric) of the shoulder girdle muscles. Maitland 1975, Cyriax 1978, Wells over that area, but unobservable It was felt that further study of the 1982). compression forces may still be pro­ behaviour of the shoulder girdle mus­ Cyriax (1978 p226) and Bogduk duced. Scientific studies (Howe et af cles was required to assess their pos­ (1983) have hypothesized that the ac­ 1977, Shah et af 1978, Rydevik et af sible role in relation to shoulder pain tion of the shoulder girdle muscles on 1984) and clinical observations (White of cervical origin. Levator scapulae was the cervical spine has a role in pro­ and Panjabi 1978, Maitland 1980) in­ selected for this study because it di­ ducing the referred shoulder pain of dicate that compression, especially rectly connects the cervical spine to the cervical origin on shoulder movement. when asymmetrically applied, can pro­ shoulder girdle and because its fibres These hypotheses were made without duce joint tissue or radicular distortion have a discrete pattern of alignment, reference to data. A review of the lit­ and thereby stress and pain. It was allowing for tidy biomechanical as­ erature has not revealed studies which found that there was an apparent lack sessment. have established the mechanics wherein of reliable data on the possible me­ It is generally believed that with the shoulder movement could produce chanical cause ofshoulder pain ofcerv­ cervical spine fixed, levator scapulae stress on the cervical joints and thereby ical orgin produced by shoulder move­ acts to elevate the scapula and down­ cause the referred symptoms in the ment. This led to a decision to examine ward rotate it, ie the glenoid cavity shoulder. Elvey (1980) has described a studies on the shoulder girdle muscles being turned caudally. With the sca-

The Australian Journal of Physiotherapy. Vol. 32, No.2, 1986 101 Levator Scapulae and Shoulder Pain of Cervical Origin

pula fixed, acting unilaterally, levator A Medelec MS6 MkIlI system was and of skinlfat thickness in the EMG scapulae is believed to laterally flex and used to record the EMG activity. The volunteer subjects. These measure­ rotate the vertebrae ipsilaterally: acting apparatus was set at a gain of 500pV/ ments indicated that this site and depth bilaterally, it is believed to assist in division, a sweep speed of 10 millisec/ was the most reliable to ensure elec­ cervical extension (Kendall and Mc­ division and for a frequency range of trode placement into the belly of lev­ Creary 1983). 16-16,000 Hz. The paper recorder was ator scapulae (Behrsin 1984). DeFreitas et at (1979, 1980) studied set at 2 cm/sec. The testing was carried out in an levator scapulae using needle electro­ A single coaxial needle electrode electrically screened room to minimize myography (EMG) during shoulder (2.5cm long in five subjects and 5.0cm extraneous electrical noise. movement but lack of methodological long in the other two) was used to pick In an endeavour to replicate the data, including the exact location of up the muscle activity. The subject was common clinical test procedures, the the EMG electrode, made their results earthed by a reusable rubber surface subjects were examined in standing. questionable. electrode placed over the scapula on Before each recording the subjects were Inman et at (1944) are often cited the test side. requested to relax their arms to obtain for their description of shoulder girdle The needle electrode was inserted in electical silence. muscle function including that of lev­ the right side at a point formed by the The movements of resisted right ator scapulae (ie by Basmajian 1978 insertion of a horizontal line passing shoulder girdle elevation and resisted and Norkin and Levange 1983), but through the Tl spinous process with a shoulder extension/adduction were examination of the original paper re­ vertical line passing one centimetre me­ performed to establish needle location vealed a lack of any data concerning dial to the superior angle of the scapula and function. These responses were levator scapulae activity. (Figure 1). The electrode was inserted noted but not recorded. Inman et al (1944) and Dvir and at a depth between 2.0 and 2.5 cm. The following movements were then Berme (1978) have described scapulo­ This site was based on the results of performed and recorded: thoracic movement during shoulder ac-. cadaveric measurements of trapezius 1. Free right shoulder abduction tivity but a lack of methodology and thickness and levator scapulae position through full range with the arm in neu­ data in the former paper and the small tral rotation. amount of data presented in the latter (ie the results from one subject) make 2. Free right shoulder flexion through these descriptions unreliable. full range with the arm in neutral ro­ This study was undertaken to assess tation. the EMG activity of levator scapulae 3. Maximal resisted right shoulder during shoulder movement and to de­ isometric abduction in five degrees ab­ termine the mode of contraction and duction starting position. length of the muscle during these 4. Free left shoulder abduction movements by way of measurement through full range with the arm in neu­ from x-rays. tral rotation. This data was combined to assess B how levator scapulae behaves during 5. Free right shoulder external ro­ tation in neutral flexion/extension with shoulder movements and whether its action could be responsible for the pro­ the elbow flexed 90 degrees. duction of stress on the cervical spine. The full range movements were per­ formed at a prerehearsed rate of ap­ Method proximately 60 degrees per second. EMG Examination of Levator Scapu­ The resultant EMG recordings were lae during Shoulder Movement then graded using the system described (Method A) Figure 1: Determination of the elec­ by Basmajian (1978). In this system the trode site. Seven volunteers, four male and degree of EMG activity is visually three female, with an average age of A First Thoracic vertebrae graded. The gradings are: 0, where 27 years (range 24-29 years) were stud­ B Scapula there is no activity; 1+ where there is ied electromyographically to observe C Levator Scapulae minimal activity (1-25010 of maximal the pattern of levator scapulae firing D Vertical line one centimetre medial activity); 2 + where there is mild ac­ to the superior angle of the sca­ during shoulder activity. Before comm­ pula tivity (26-50% of maximal activity); 3+ encing the procedure, informed written E Horizontal line through T1 where there is moderate activity (51­ consent was obtained from all subjects. X Electrode site 80% of maximal activity) and; 4 +

102 The Australian Journal of PhYSiotherapy. Vol. 32, No.2, 1986 Levator Scapulae and Shoulder Pain of Cervical Origin

where there is strong activity (81-100070 The instantaneous axes of rotation the right arm compared with resisted of maximal activity). Tbis study used for the scapula motion between 0 and abduction. The other four subjects the activity observed during resisted 90 degrees and 90 and 180 degrees of showed the reverse. shoulder abduction were determined by shoulder girdle elevation as the basis During free shoulder abduction on using the method described by White for grading (ie the 100070 value) as this the right side there was a slight ob­ and Panjabi (1978 p479). A common always produced the strongest EMG servable increase in activity during the point (ie the inferior angle of the sca­ response. Due to the limitations of this second half of range (ie 90-180 degrees) pula) on two sequential scapula posi­ system descriptive comments on the to that seen in the early half of range. EMG activity of some movements has tions was joined by a line which was been included in the results. then bisected by another line j!t right EMG activity was next greatest dur­ angles. This procedure was repeated ing flexion. The intensity of activity using several reference points. The in­ varied from 50 to 80 percent of that X-ray Assessment of Levator Scapulae tersection of the bisecting lines was observed during free right shoulder ab­ Length during Shoulder Abduction taken as the instantaneous axis of ro­ duction, except in one subject where it (Method B) tation for that movement. was only minimally active. In three Plain A/P view x-rays of the upper subjects there appeared to be a gradual quadrant were taken to include the right Results slight increase in activity during the glenohumeral joint, the scapulo-thor­ A: The graded responses to the test outer range of flexion. This was not acic joint and the vertebrae from C5 movements as recorded by EMG are observed in the other four subjects. to TlO. X-rays were taken with the shown in Table 1. EMG spikes due to needle movement shoulder actively abducted at 0, 90 and Maximal activity was observed dur­ were most commonly recorded on the 170 degrees with the subject standing. ing resisted shoulder girdle elevation in movements of right shoulder abduction Two subjects were used, one female all seven subjects, while minimal or no and flexion through range. During one aged 24 and one male aged 28 years. activity was observed during resisted subject's recording there appeared an Both had been subjects for the EMG shoulder extension/adduction. underlying interference pattern which study (Method A). Three subjects showed minimally was found to be due to slight loosening Using the spinous process of TI and greater activity on free abduction of of the earth electrode. T4 as common reference points, trac­ ings of the scapula were performed from the x-rays, superimposing the Table 1: three positions. EMG activity in right levator scapulae Measurement from the subjects CI to TI spinous process and the width Movement Grade a Grade 1 + Grade 2 + Grade 3 + Grade 4 + of CI were taken in vivo. These dis­ Right free tances were corrected for the distortion Abduction due to the distance from the x-ray plate (0-90 degrees) 4 3 by multipying by k, using the formula Right free k = T1 to T4 on x-ray Abduction T1 to T4 in vivo (90-180 degrees) 6 and then placed on the tracing of sca­ Right free pula positions. Flexion The relative distance of CI trans­ (0-90 degrees) 5 verse process to superior angle of sca­ Right free pula was then measured for the three Flexion positions and these figures corrected (90-180 degrees) 3 3 by dividing the result by the x-ray dis­ Right res. tortion factor, k. Abduction 7 The values for the different scapula Left free positions were then estim!lted as per­ Abduction· 5 centage values of the resting value ie Right Ext. the length with the shoulder at 0 de­ Rotation· 3 4 grees. These values are assumed to in­ dicate the length of levator scapulae at • Range not included as no significant changes occurred through different shoulder abduction positions. range.

The Australian Journal of PhYSiotherapy. Vol. 32, No.2, 1986 103 Levator Scapulae and Shoulder Pain of Cervical Origin

B: Tracings of the subject's scapular distances are set out in Table 2. These Discussion positions during shoulder abduction results are given as the actual distances DeFreitas et at (1979, 1980) found taken from x-rays are presented in Fig­ (ie they have been recorrected from the in their EMG study of levator scapulae ure 2 along with the instantaneous axes x-ray data) and as percentage values of that it was strongly active in abduction of motion. the resting length. and elevation, moderately active in The data indicates that levator scap­ shoulder flexion and minimally active The x-ray distortion was calculated ulae undergoes shortening during the in scapular retraction and shoulder ex­ at 1.03 for the male subject and 0.93 first half of abduction but undergoes tension. Those results are in agreement for the female. lengthening during the second 90 de­ with those ofthis current study in which The results for the C1 transverse grees of shoulder abduction in these some of the methodological flaws of process to superior angle of the scapula two subjects. DeFreitas et at (1979 1980) were rec­ tified. This study also found that lev­ ator scapulae activity increased in the A B outer range of flexion and abduction in some subjects, a finding not previ­ j' .-.-: ously noted. " "' \ Inman et at (1944) and Dvir and j -- --, Berme (1978) described different phases .----, \ of scapulo-thoracic movement during -- I '. I shoulder abduction and flexion. These ._._., I i I phases are a result of interaction be­ .I /1 j I II tween shoulder girdle muscle activity, .I i I sternoclavicular and acromioclavicular ! I •I /1 ! I joint motions, and ligament tension. i I .' I The phases, briefly, are: I ~/ I Phase One, setting (between 0 and I I 30 degrees abduction and 0 and 60 in ! i , , i ,I flexion), in which the scapula is aligned ,CE / for optimal glenohumeral movement. I 1 / I Phase Two (between 30 and 90 de­ 1 / I I grees abduction and 60 and 90 degrees / I I \ \ in flexion) in which the scapula rotates 1 '. /( I around an axis approximately passing / I \~./ , through the root of the spine of the ',-_/ scapula and the sternoclavicular joint. Phase Three in which the scapula as a whole rotates about an axis at the Legend for Figure 2 acromioclavicular joint. Phase Four in which the scapula and A Scapular positions for subject A clavicle move as a unit about the long B Scapular positions for subject B axis of the clavicle. Phases three and four account for the ranges of 90 to _____ Scapula at 0 degrees shoulder abduction 180 degrees in flexion and abduction. This current study found that the Scapula at 90 degrees shoulder abduction movement of the scapula relative to Scapula at 170 degrees shoulder abduction the spine in the first 90 degrees of shoulder abduction was quite different Instantaneous axis of rotation for 0 - 90 degrees abduction to that described by Inman et at (1944) and Dvir and Berrne (1978). Rather Instantaneous axis of rotation for 90 - 170 degrees abduction than rotating about an axis through the root of the spine of the scapula, the rotation occurred through a point Figure 2: closer to the centre of the scapula near Tracings from x-rays of subject's scapular positions during shoulder abduction. its medial border. Also, contrary to the

104 The Australian Journal of Physiotherapy. Vol. 32, No.2, 1986 Levator Scapulae and Shoulder Pain of Cervical Origin

Table 2: It is likely that the interaction of Relative distance of C} transverse process to superior angle of scapula other synergistic muscles is important in considering the production of shoul­ em % of Resting Length der pain of cervical orgin on shoulder Subject A Degrees 13.4 100 o movement. The anatomical alignment 90 Degrees 12.0 89 of levator scapulae and its eccentric 170 Degrees 14.5 108 behaviour during shoulder movement indicate that it may have a significant Subject 8 Degrees 14.0 100 o role in the phenomenon. 90 Degrees 13.1 94 170 Degrees 14.2 101

Dvir and Berme (1978) paper, there was an actual increase in levator sca­ was significant movement of the root pulae length at full elevation compared of the spine of the scapula during the to that at rest (ie 108070). As passive first 90 degrees. tension due to a muscle being stretched During the second 90 degrees of ab­ can produce force (DeLauter 1982), this duction the root of the spine of the force should be considered additive to scapula was found to move downwards that being produced by the eccentric and laterally. This is in agreement with contraction of levator scapulae. Dvir and Berme (1978), but the deg~ As levator scapulae exerts an increas­ of the movement was not as great as ing amount of force on the cervical their paper seemed to indicate. spine during outer range shoulder ab­ These differences in scapular m

The Australian Journal of Physiotherapy. Vol. 32, No.2, 1986 105 Levator Scapulae and Shoulder Pain of Cervical Origin

Conclusion References Howe J, Loeser J and Calvm W (1977), Mechan­ Basmajlan J (1978). Muscle Ailve, 2nd ed. WIl­ osensltlvlty of dorsal root ganglia and chrom­ The EMG results and estimated hams and WIlkms, BaltImore. cally mjured axons; A phySiOlogIcal baSIS for lengths of levator scapulae during dif­ Behrsin J (1984), The acllon of Levator Scapulae the radicular pam of nerve root compreSSion, ferent parts of active shoulder abduc­ durmg shoulder movement: A dJssectlOo, EMG Pain, 3, 25-41. and x-ray study. Graduate dIploma 10 mamp­ Inman V, Saunders J and Abbott L (1944), Ob­ tion range indicate that levator sca­ ulatIve therapy (W.A.I.T.) Anatomy project, servatlons on the functIOn of the shoulder Jomt, pulae contracts concentrically during (unpubhshed data). The Journal ofBone and Jomr Surgery, 26 (I), the first 90 degrees and eccentrically Bogduk N (1983), Neurology of the neck/shoulder 1-30. complex, 10 Proceedmgs of MTAA Neck and Kellgren J (1939), On the dlstnbutIon of pam during the second 90 degrees. Shoulder Symposium, Bnsbane. ansmg from deep somatIC structures With charts Cmquegrana 0 (1968), Chromc cervIcal radICuhlls of segmental pam areas. Clinical SCience. 4, 35­ Due to the different amount of force and Its relatIOnship to "chrome bursitIS", Amer­ 46. Ican Journal of PhYSical Medlcme, 47 (I), t}­ Kendall PF and McCreary EK {I983), Muscles; generated by these two types of con­ 30. Testmg and FunctIon, 3rd edition, Wllhams and traction, it appears that levator sca­ Cloward R (1957), Cervical dIscography - a con­ Wllkms, Baltimore. pulae exerts greater force during the tnbutlOn to tbe ello!ogy and mechamsm of neck, Matgne R (1975), Pseudo tendlmtls ofthe shoulder shoulder and arm pam. Annals ofSurgery, 150, and cervical spme, Excerpta MedIca 217, Sect outer range ofshoulder adbuction. This 1052-1064. 33, 2743. force can act on the cervical spine via Cyriax J (1978), Textbook of Orthopaedic Med­ Mattland G (1975), Vertebral mampulatlOn, 3rd Icme, Vol. I., Ballhere Tmdall, London. edItIOn, Butterworths, London. the muscle's cervical attachment. DeFreItas V, Vitll M and Furlam J (1979), EMG MaItland G (1980), The hypotheSIS of addmg analySIS of the levator scapulae and rhombOJ­ compreSSIOn when exammmg and treating syn­ It is suggested that the potential stress deus major muscles m movements of the shoul­ OVIal jomts, Journal ofSports PhYSical Therapy, on the cervical spine due to levator der, Electromyography and Clmlcal Neuro­ 2 (I), 7-14. phYSIOlogy, 19, 335-342. McCall I, Park Wand O'Bnens J (1979), Induced scapulae action, which increases in DeFreItas V, VItti M and Furlanl J (1980), EMG pam referred from postenor lumbar elements m outer abduction, may, in the presence study of levator scapulae and rhombOIdeus ma­ normal subjects, Spme, 4 (5), 441-446. of pathological changes, cause pain. Jor muscles ill movements of the shoulder and Mooney V and Robertson J (1976), The facet arm, Electromyography and C/lmcal Neuro­ syndrome, ClImcal OrthopaedICS and Related Levator scapulae action may be a po­ phYSIOlogy, 20, 205-216. Research, 115, 149-156. tential mechanism for shoulder pain of DeLauter B (1982), Therapeullc exercIse to de­ Norkm C and Levange P (1983), Jomt Structure cervical origin during active shoulder velop strength and endurance, In F Kottke, G and FunctIOn, FA Davis and Co, Phlladelphia. Stillwell and J Lehman (Eds), Krusen's Hand­ Rydevlk B, Brown M and Lundborg G (1984), movement. book of PhYSIcal Medlcme and RehabllllatlOn, Pathoanatomy and pathophysiology of nerve 3rd ed., WB Saunders, Philadelpla. root compreSSIOn, Spine, 9 (I), 7-15. DVlr Z and Berme N (1978), The shoulder complex Shah J, Hamson Wand Jayson M (1978), The In elevatIon of the arm; A mechamsm approach, dlstnbutlon of surface stram in the cadaveric Journal of B/Omechamcs, 11, 219-225. lumbar spIlle, Journal ofBone and JOint Surgery Elvey R (1980), BrachIal plexus tensIOn tests and 60 B (2), 246-251. the pathoanatomlcal orgin of arm pam, m As­ pects ofManlpulatlve Therapy, Lmcoln InstItute Wells P (1982), CervIcal dysfunctIon and shoulder Acknowledgements problems, PhYSIOtherapy, 68 (3), 66-73. The authors wish to thank Professor of Health ScIences, Melbourne. Hagberg M (1982), Work, load and fatIgue 10 Wlute A and PanJabl M (1978), ClImcal lJlOme­ Lance Twomey for invaluable assist­ repetitlve arm elevations, Ergonomics, 24 (7), chames of the Spme, JB LIppmcott, Phlladel­ ance with the preparation of the ma­ 543-555. plua. Hockaday J and Whitty C (1967), Patterns of Yamshon L and BIerman W (1948), Kinesiologlc muscript, and Mr Bill Fitzsimmons for referred pain III the normal subject, Bram, 90 electromyography: The Trapezius, ArchIVes of the preparation of the x-rays. (3),481-495. PhySIcal Medlcme, Oct, 647-651.

106 The Australian Journal of PhySiotherapy. Vol. 32, No.2, 1986