THE RHOMBOID MUSCLE group has an important role in

Transcription

1 987 Rhomboid Muscle Electromyography Activity During 3 Different Manual Muscle Tests Jay Smith, MD, Denny J. Padgett, PT, Kenton R. Kaufman, PhD, Shawn P. Harrington, MD, Kai-Nan An, PhD, Steven E. Irby, MMSE ABSTRACT. Smith J, Padgett DJ, Kaufman KR, Harrington SP, An K-N, Irby SE. Rhomboid muscle electromyography activity during 3 different manual muscle tests. Arch Phys Med Rehabil 2004;85: Objective: To determine which of 3 previously published rhomboid manual muscle tests (MMTs) elicits the maximal rhomboid electromyographic activity in an asymptomatic population. Design: Criterion standard. Setting: Motion analysis laboratory at tertiary care medical center. Participants: Eleven male volunteers (age range, 24 40y) without shoulder or neck pain. Interventions: Not applicable. Main Outcome Measures: Peak 1-second normalized electromyographic activity in the rhomboid muscle during 8 different MMT positions, including 3 different rhomboid MMT positions (Kendall, Kendall-Alternative, Hislop-Montgomery). Results: The Kendall MMT (78% maximal voluntary contraction [MVC]) produced higher rhomboid electromyographic activity than the Kendall-Alternative (71% MVC) or the Hislop-Montgomery MMT (52% MVC), but the differences were not statistically significant. The posterior deltoid MMT generated the greatest rhomboid electromyographic activity of all MMTs, and 4% to 30% greater rhomboid electromyographic activity than the 3 rhomboid MMTs (P.0001; posterior deltoid Hislop-Montgomery). Electromyographic profiles of the Kendall and Kendall-Alternative MMTs were similar, whereas the Hislop-Montgomery MMT produced less upper trapezius activity (P.0001 vs Kendall and Kendall-Alternative) and more latissimus dorsi activity (P.0001 vs Kendall-Alternative). The standard MMT positions for the middle trapezius, levator scapula, posterior deltoid, and latissimus dorsi produced the maximal electromyographic activity for their respective target muscles. Conclusions: The posterior deltoid MMT position should be used to produce maximal rhomboid electromyographic activity for normalization purposes during kinesiologic studies. The Kendall and Kendall-Alternative rhomboid MMT are likely to be clinically indistinct. It is unlikely that clinicians can use standard MMT positions to distinguish rhomboid strength from synergists, such as the levator scapula and middle trapezius muscle, for diagnostic purposes. Key Words: Electromyography; Rehabilitation; Scapula; Shoulder. From the Department of Physical Medicine and Rehabilitation, Mayo Clinic Sports Medicine Center (Smith, Harrington) and Department of Orthopedic Surgery, Mayo Clinic Motion Analysis Laboratory (Padgett, Kaufman, An, Irby), Rochester, MN. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Jay Smith, MD, Dept of Physical Medicine and Rehabilitation, 200 First St SW, Rochester, MN 55905, smith.jay@mayo.edu /04/ $30.00/0 doi: /s (03)00618-x 2004 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation THE RHOMBOID MUSCLE group has an important role in normal shoulder function and in the evaluation of shoulder and upper-limb disorders. 1,2 Rhomboid muscle dysfunction has been implicated as a cause of secondary shoulder impingement 1 and can produce interscapular myofascial pain. 1 Additionally, the rhomboid muscles have been examined clinically and electrophysiologically to help differentiate upper-limb nerve injury and nerve pain syndromes. 2 Fine-wire electromyography provides a method with which to examine the neural activation of shoulder girdle muscles, such as the rhomboids, during upper-limb motions and rehabilitative exercises. 3-6 However, clinical interpretation of finewire electromyographic data is not without potential challenges. 4 In particular, normalization of electromyographic signals is necessary to accurately compare quantitative electromyographic values among subjects, muscles, upper-limb motions, and scientific studies. 4 Normalization involves the conversion of absolute electromyographic signals to relative values based on electric activity produced during a reference movement, most commonly an isometric maximal voluntary contraction (MVC; expressed as %MVC). 3,4 Clinically, motions that evoke higher electromyographic activities (%MVC) have been interpreted to be more challenging to a muscle. 3,5-7 Many researchers 5-7 have even proposed criteria that describe electromyographic activity as being low (0% 20% MVC), moderate (21% 40% MVC), high (41% 60% MVC), and very high ( 60% MVC) and have derived activity and exercise recommendations based on this scale. Given the current manner in which clinicians may interpret electromyographic data, it is important to ensure that the reference muscle contraction used for normalization actually produces the greatest electromyographic activity for that muscle. If the normalization procedure is flawed, erroneous clinical conclusions may be made. The role of scapular control in shoulder function remains an active area of kinesiologic research. 1 Although the rhomboid muscles are accessible via fine-wire electromyography, 8,9 review of the literature reveals that 3 different manual muscle tests (MMTs) have been proposed for the rhomboid muscles, with no direct comparisons between the tests in their ability to generate rhomboid muscle electromyographic activity In addition, the few kinesiologic studies that have included the rhomboid muscles did not specify which, if any, of the 3 rhomboid MMTs were used in the normalization process. 3,9 Our primary purpose in this study was to compare the electromyographic activity produced by the 3 published rhomboid MMTs: the Kendall, 10 Kendall-Alternative, 10 and Hislop- Montgomery tests. 11 Establishing the position that consistently produced the maximal electromyographic activity would be a necessary prerequisite for appropriate execution and interpretation of future kinesiologic investigations of scapular stabilizer muscle activity during functional motion.

2 988 RHOMBOID MUSCLE ELECTROMYOGRAPHIC ACTIVITY, Smith Table 1: MMT Procedures 10,11 1. Upper trapezius With the patient sitting and facing away from the examiner, manual resistance is applied against the shoulder in a direction of depression and the head in a direction of flexion anterolaterally, while the subject voluntarily attempts to elevate the acromial end of the scapula and extend the neck to approximate the acromion and the occiput. 2. Middle trapezius With the patient prone and the test side elbow extended fully at 90 of shoulder abduction and 90 of shoulder external rotation (thumb-up position), the examiner applies manual pressure downward (to the forearm for leverage) to resist adduction of the scapula and extension of the shoulder. 3. Levator scapulae With the patient prone (head facing the test side) and the test arm adducted against the chest wall, the scapula adducted and elevated with the elbow flexed fully and the shoulder slightly extended. The examiner applies manual pressure with 1 hand against the patient s arm (at the elbow for leverage) in the direction of abduction and with the other hand in the direction of shoulder depression. Weakness is noted as abduction and depression of the scapula (not the arm) and lateral rotation of the inferior angle of the scapula. 4. Rhomboids (Kendall) As for levator scapulae above. 5. Rhomboids (Kendall-Alternative) With the subject in the prone position, the test side arm is placed in 90 of abduction and 90 of internal rotation (thumb-down position). Manual resistance is placed against the forearm in a downward direction. 6. Rhomboids (Hislop-Montgomery) With the subject prone, the test shoulder is internally rotated and the arm adducted across the back so that the dorsum of the hand rests on the sacrum. Manual resistance is applied to the arm just above the elbow in a downward and outward direction. The subject must attempt to lift the hand off the back and to maintain scapular adduction. 7. Posterior deltoid With the patient sitting and facing away from the examiner, the arm is abducted to 90 in slight extension and humerus in slight internal rotation. The examiner applies pressure in the direction of adduction and flexion of the arm against the posterolateral surface of the upper arm. 8. Latissimus dorsi With the patient prone, the arm is medially rotated to 90, adducted against the chest wall and extended, manual pressure is applied (at the forearm) in the direction of abduction and flexion. METHODS Participants Eleven male volunteers (age range, 24 40y) were recruited from the Mayo Clinic to participate in the study. All subjects were free from current or past shoulder or neck pain, demonstrated full pain-free range of motion of both shoulders, and had normal neurologic results based on examinations at the time of the study. The study was approved by the institutional review board at Mayo Clinic, and all subjects completed a verbal and written informed consent process before participating. Test Procedure Each subject completed 3 maximal voluntary muscle contractions in 8 different MMT positions while electromyographic activity was recorded from 6 test muscles (upper trapezius, middle trapezius, rhomboid major [hereafter, rhomboid], levator scapula, posterior deltoid, latissimus dorsi) using fine-wire electrodes. Fine-wire electrodes were placed into each of the 6 muscles, using standard procedures and muscle localizations, 2,8 and proper placement was confirmed by noting increased electromyographic activity during standard MMTs. 10,11 Resting background electromyographic activity was then recorded with the subjects in the prone position for 3 seconds during voluntary relaxation. Thereafter, each subject completed 3 repetitions in each of the 8 MMT positions (table 1, figs 1 4). The test positions for the upper trapezius, middle trapezius, levator scapula posterior deltoid, and latissimus dorsi muscles were performed as described by Kendall. 10 Three different rhomboid MMT positions were used as previously described: (1) Kendall 10 : subject was prone with head turned toward test side, shoulder adducted, slightly extended, and slightly externally rotated with elbow flexed; (2) Kendall- Alternative 10 : subject was prone, test arm placed in 90 of abduction and 90 of internal rotation (thumb down), with elbow extended; and (3) Hislop-Montgomery 11 : subject was prone, head turned to test side, shoulder internally rotated, arm abducted and placed across the back with the elbow flexed, with the back of the hand resting on the back. The MMT order was randomized to control for fatigue. The 3 rhomboid MMT positions were always kept in a group within the randomization scheme, although the order of their completion was random- Fig 1. Rhomboid MMT using the Kendall technique (see table 1).

3 RHOMBOID MUSCLE ELECTROMYOGRAPHIC ACTIVITY, Smith 989 Fig 2. Rhomboid MMT using the Kendall-Alternative technique (see table 1). Fig 4. Posterior deltoid MMT (see table 1). This position produced the greatest rhomboid muscle electromyographic activity of all muscle tests examined. ized between subjects. All data were collected during a single testing session for each subject to minimize test-retest variability. Electromyographic Signal Collection and Processing Electromyographic data were collected using an MA-100 electromyography system a with a personal computer (PC) based data acquisition system. The system consisted of MA- 110 electromyograph preamplifiers located at the fine-wire electrode insertion sites. These provided a nominal signal gain of 380% 3%, a common mode rejection rate of 100dB at 60Hz, an input impedance of 1T, and a bandwidth of 10Hz to 30kHz ( 3dB). Each electrode was connected to the electromyograph backpack unit via a single, thin, super-flexible, 3 32-in diameter coaxial cable so to transfer information from the lightweight backpack to the PC-based acquisition system. The backpack has 10 electromyography channels ( Hz), each with individual gain calibration and an adjustable low-pass filter. The electromyographic data for this study was bandpass filtered from 30 to 1300Hz using a scaling factor Fig 3. Rhomboid MMT using the Hislop-Montgomery technique (see table 1). of 5. Computerized data collection was performed with commercially available CODAS software, b which provides realtime data display and recording capabilities. Electromyographic data postprocessing was first performed with WinDaq software, b and then completed with custom programming developed with Matlab, version 6.0, software. c Each subject completed 2 voluntary, 4-second maximal isometric muscle contractions for each muscle, using standard positions. After subtracting the resting electromyographic signal, the peak 1-second electromyographic activity for each muscle during the maximal contractions was calculated. The mean peak electromyographic activity elicited from 2 MMT trials was calculated for each muscle. To compare electromyographic activity from subject to subject, the mean peak values were normalized to the maximum electromyographic value in that specific muscle across all MMTs. Statistical Analysis The outcome of primary interest was the mean peak normalized electromyographic activity for each muscle. The effect of the muscle and testing procedure on normalized electromyographic activity was evaluated using analysis of variance (ANOVA) with repeated measures. 12 Initially, a 2-factor model was developed, but a significant muscle by test interaction was observed (F 6.55, P.0001). Consequently, a separate 1-factor repeated-measures ANOVA model was run for each test, comparing all muscles. For completeness, an additional 1 factor repeated-measures ANOVA model was run for each muscle, comparing all tests. Significant main effects were further analyzed using the Ryan-Einot-Gabriel-Welsch multiple comparisons test. All statistical tests were 2-sided and P values less than.05 were considered significant. All analyses were done with SAS, version 8.02, d and S-Plus, version e RESULTS Figures 5 and 6 show the mean peak 1-second normalized electromyographic activity for each of the 6 test muscles during each of the 8 MMTs. Note that the data in both figures are the same but the presentation mode differs to facilitate reader interpretation. Among the 8 MMTs, the traditionally described MMT positions for the middle trapezius, levator scapula, posterior deltoid, and latissimus dorsi produced the largest mean peak normalized electromyographic activity for each respective

4 990 RHOMBOID MUSCLE ELECTROMYOGRAPHIC ACTIVITY, Smith (F 12.22, P.0001 for Hislop-Montgomery Kendall-Alternative), and qualitatively although not statistically significant less activity in the middle trapezius, posterior deltoid, and levator scapula muscles. Fig 5. Normalized mean peak electromyographic (EMG) activity. Mean peak 1-second electromyographic activity for each of 6 test muscles during the 8 MMTs. See Methods for normalization procedure. Abbreviations, LD, latissimus dorsi; LS, levator scapula; MT, middle trapezius; PD, posterior deltoid; RH-HM, rhomboid Hislop- Montgomery; RH-K, rhomboid Kendall; RH-KA, rhomboid Kendall- Alternative; UT, upper trapezius. muscle. In contrast, 2 of the 6 test muscles, using previously described MMTs, did not produce the maximal electromyographic activity in targeted muscles. Quantitatively, the largest activity in the rhomboid was produced by the posterior deltoid MMT (fig 4), and the largest activity in the upper trapezius was produced by the levator scapula MMT. Figures 5 and 6 also show that the upper trapezius MMT actually produced more electromyographic activity in the levator scapula than in the upper trapezius, although the quantitative values were close. Similarly, the posterior deltoid MMT produced the same activity in the posterior deltoid, middle trapezius, and rhomboid muscles, although the middle trapezius MMT did produce higher middle trapezius muscle electromyographic activity than the posterior deltoid MMT. The 3 rhomboid MMTs were of particular interest for an purposes in this study. None of the 3 elicited the maximal activity in the rhomboid muscle, which was actually achieved by the posterior deltoid MMT (F 5.75, P.0001 for posterior deltoid Hislop-Montgomery; posterior deltoid vs Kendall or Kendall-Alternative, not significant). Of the 3 rhomboid MMTs, the Kendall position produced the highest rhomboid activity (mean peak, 78% MVC) and the Hislop-Montgomery position produced the least activity (mean peak, 52% MVC), although the differences between the tests were not statistically significant. The Kendall position produced near equal activity in the levator scapula, whereas the Kendall-Alternative position actually produced greater activity in the upper trapezius, middle trapezius, levator scapula, and posterior deltoid compared with the activity produced in the rhomboid muscle itself. As shown in figures 5 and 6, there was no substantial difference in the electrophysiologic profiles of the Kendall and Kendall- Alternative tests. On the other hand, the Hislop-Montgomery position produced more activation in the latissimus dorsi and levator scapula muscles than in the rhomboid muscle; equal activation in the middle trapezius, posterior deltoid, and rhomboids; and less activation in the upper trapezius. As compared with the other rhomboid MMTs, the Hislop-Montgomery position produced significantly less upper trapezius activity (F 8.51, P.0001 for Hislop-Montgomery Kendall or Kendall-Alternative), significantly more latissimus dorsi activity DISCUSSION Our primary purpose in this study was to determine which of the 3 previously published rhomboid MMTs elicited the maximal electromyographic activity from the rhomboid muscle (rhomboid major as examined in this study). The results quantitatively indicate that the Kendall technique elicited more rhomboid electromyographic activity than either the Kendall- Alternative or the Hislop-Montgomery tests, although the differences between the tests did not reach statistical significance. Surprisingly, the posterior deltoid MMT elicited the most rhomboid electromyographic activity of the 8 MMTs, producing activity that was on average 4% greater than the Kendall position, 11% greater than the Kendall-Alternative position, and 30% greater than the Hislop-Montgomery position. In the latter case, the difference was statistically significant (F 5.75, P.0001). Our findings may be explained in part by the positioning of the patient. The posterior deltoid MMT is performed with the subject in a sitting position and the arm abducted. 10 Previous research has shown significant rhomboid activity during active arm abduction, because the rhomboid and other scapular stabilizer muscles attempt to control the scapula s rotational position in the gravitational field. 2,9 Therefore, the posterior deltoid MMT challenges the rhomboid as both a scapular rotator and scapular retractor. On the contrary, the 3 rhomboid MMTs we used primarily challenged the rhomboid in its role as a scapular retractor. 2,9 Although the electrophysiologic profiles of the Kendall and Kendall-Alternative tests were similar, the Hislop-Montgomery test elicited a distinctly different profile, with more latissimus dorsi activity (F 12.22, P.0001) and less upper trapezius activity (F 10.33, P.0001). These differences can be explained by the positioning of the test arm in near maximal adduction and internal rotation during the Hislop-Montgomery test, 11 a position that would inhibit the upper trapezius but facilitate the latissimus dorsi. 2 Although these differences were statistically significant, they are unlikely to be clinically significant, in our opinion. Fig 6. Normalized mean peak electromyographic activity. Figure shows the data presented in figure 1 in bar graph format.

5 RHOMBOID MUSCLE ELECTROMYOGRAPHIC ACTIVITY, Smith 991 Similar to most kinesiologic electromyography research, interpretation of our results requires that a clinician determine the purpose of the MMT. If it is to serve as a reference movement for normalization of electromyographic activities, then the primary goal should be to choose an MMT position that provides the most neural activation of that muscle. 2,4 Accordingly, our results suggest that the posterior deltoid MMT position serves as a better reference position for rhomboid muscle normalization than the 3 previously published rhomboid MMTs. As we discussed in the introduction, choosing the appropriate reference position is necessary to minimize misinterpretation of the %MVC values generated in a kinesiologic electromyography study. For example, if the reference position does not elicit high target muscle electromyographic activity, then the %MVC values derived during test motions, such as in rehabilitative exercises, will be artificially high. These erroneously high results may then cause clinicians to make incorrect clinical decisions regarding the appropriateness of particular movements or in prescribing rehabilitative exercise. Although most clinicians are aware of the imprecision of correlating electromyographic data with muscle tension 2 and the lack of clear definitions between acceptable and dangerous levels of electromyographic activity, 4 clinical recommendations continue to be based on experimentally derived electromyographic values. 3,5-7 A second purpose of MMT is to isolate a muscle from its synergists to assess the ability of that muscle to generate force for diagnostic purposes. 4,10,11,17 In the case of the rhomboid muscles, clinical isolation would be useful to detect rhomboid weakness in cases of scapular instability with secondary rotator cuff impingement, to evaluate the etiology of interscapular pain, or to assist in the differentiation of C5 radiculopathies from brachial plexopathies or more distal peripheral neuropathies. Although it was not our primary purpose to examine the ability of various MMTs to isolate the rhomboids, the data did offer some insight into the ability to clinically isolate the rhomboid muscle from synergists, such as the middle trapezius and posterior deltoid. As shown in figures 5 and 6, all 3 published rhomboid MMTs elicited similar or greater electromyographic activity in the levator scapula, middle trapezius, and/or posterior deltoid muscles compared with the activity elicited in the rhomboid muscle itself. Consequently, our results suggest that the 3 rhomboid MMT positions we used cannot be considered to quantitatively isolate the rhomboid muscle from the posterior deltoid (C5-6, axillary nerve), levator scapula (C5, dorsal scapular nerve), or the middle trapezius (CN XI). In fact, none of the 8 MMT positions we used provided meaningful isolation of the rhomboid from its synergists, with the notable exception that the MMT of the middle trapezius elicited 22% more activity in the middle trapezius than in the rhomboid. The inability to effectively clinically isolate the medial scapular stabilizers (rhomboids, levator scapula, trapezius) from each other has been previously suggested 2,9,18,19 and is confirmed by our results. Reassuringly, our results indicate that the standard MMT positions for the middle trapezius, levator scapula, posterior deltoid, and latissimus dorsi 10,11 do elicit the maximal electromyographic activity in these muscles (figs 5, 6). Therefore, in the context of standardized MMT positions, these positions represent reasonable reference positions for the purposes of electromyographic normalization procedures. What is unknown is whether nonstandard shoulder girdle or arm positions would elicit even greater activity from any of these target muscles. 5 This study has 2 primary limitations. First, only standard reference MMT positions were used because these are the most common positions used to normalize electromyographic data. 4 As discussed, it is possible that nonstandard positions may produce greater activity in target muscles such as the rhomboids. 4,5 However, because our primary purpose was to compare the electromyography profiles of the 3 published rhomboid MMT positions, we did not explore nonstandard positions. Second, as is typical of electromyography-based investigations, interindividual variability may limit the power of the study to detect statistically significant differences. Our primary goal was to determine which MMT elicited the greatest quantitative rhomboid electromyographic activity that would be desired for normalization purposes. The determination of whether statistically significant differences existed between various MMT was secondary, and therefore prestudy power calculations were not performed. However, poststudy power calculations were completed to determine the power of the study to demonstrate statistically significant differences between MMTs. Based on the observed data for the 11 subjects, there was 80% power to detect differences in mean peak normalized electromyographic activity between any 2 of the 8 MMTs, ranging from.18 to.22 for the latissimus dorsi, middle trapezius, posterior deltoid, and upper trapezius, and from.72 to.73 for the levator scapula and the rhomboids. The calculations suggest that our study was underpowered for determining statistically significant differences between the 3 rhomboid MMTs. However, it did have sufficient power to demonstrate the superiority of the posterior deltoid MMT in eliciting rhomboid muscle electromyographic activity. CONCLUSIONS Of the positions tested in this study, the posterior deltoid MMT elicited the most rhomboid muscle electromyographic activity. Because both the posterior deltoid and rhomboid muscles are of interest in kinesiologic shoulder research, the posterior deltoid MMT would thus suffice to normalize both the posterior deltoid and rhomboid muscles. The electrophysiologic similarities between the Kendall and Kendall-Alternative MMTs suggest that these 2 rhomboid MMTs are clinically indistinct. Clinically, based on our results, it is unlikely that clinicians can use MMTs to distinguish the rhomboid muscles from synergists, such as the middle trapezius and levator scapula muscles, for diagnostic purposes. The standard MMT positions for the middle trapezius, levator scapula, posterior deltoid, and latissimus dorsi elicited the maximal electromyographic activities in their respective muscles among the MMT positions we examined. References 1. Kibler W. The role of the scapula in athletic shoulder function. Am J Sports Med 1998;26: Basmajian J, DeLuca C. Muscles alive: their functions revealed by electromyography. 5th ed. Baltimore: Williams & Wilkins; p Moseley JB Jr, Jobe FW, Pink M, Perry J, Tibone J. EMG analysis of the scapular muscles during a shoulder rehabilitation program. Am J Sports Med 1992;20: Kelly B, Kirkendall DT, Levy AS, Speer KP. Current research on muscle activity about the shoulder. Instr Course Lect 1997;46: Townsend H, Jobe FW, Pink M, Perry J. Electromyographic analysis of the glenohumeral muscles during a baseball rehabilitation program. Am J Sports Med 1991;19: McCann PD, Wootten ME, Kadaba MP, Bigliani LU. A kinematic and electromyographic study of shoulder rehabilitation exercises. Clin Orthop 1993;Mar(288):

6 992 RHOMBOID MUSCLE ELECTROMYOGRAPHIC ACTIVITY, Smith 7. Jobe FW, Moynes DR, Tibone JE, Perry J. An EMG analysis of the shoulder in pitching. A second report. Am J Sports Med 1984;12: Perotto AO, Delagi EO, editors. Anatomical guide for the electromyographer: the limbs and trunk. 3rd ed. Springfield: CC Thomas; De Freitas V, Vitti M, Furlani J. Electromyographic study of levator scapulae and rhomboideus major muscles in movements of the shoulder and arm. Electromyogr Clin Neurophysiol 1980;20: Kendall F, McCreary E, Provance P. Muscles: testing and function. 4th ed. Baltimore: Williams & Wilkins; p Hislop H, Montgomery J. Muscle testing and function. 7th ed. Philadelphia: WB Saunders; p Winer B. Statistical principles in experimental design. New York: McGraw-Hill; p Ryan T. Multiple comparisons in psychological research. Psychol Bull 1959;56: Ryan T. Significance tests for multiple comparisons of proportions, variances, and other statistics. Psychol Bull 1960;57: Einot I, Gabriel K. A study of the powers of several methods of multiple comparisons. J Am Stat Assoc 1975;70: Welsch R. Stepwise multiple comparison procedures. J Am Stat Assoc 1977;72: Kelly B, Kadrmas W, Speer K. A manual muscle examination for rotator cuff strength: an electromyographic investigation. Am J Sports Med 1996;24: Brandell B, Wilkinson D. An electromyographic study of manual testing procedures for the trapezius and deltoid muscles. Physiother Can 1991;43: Inman V, Saunders M, Abbot L. Observations on the function of the shoulder joint. J Bone Joint Surg 1944;27:1-30. Suppliers a. Motion Lab Systems, Old Hammond Hwy, Baton Rouge, LA b. Dataq Instruments, 150 Springdale Dr, Ste 220, Akron, OH c. The MathWorks, 3 Apple Hill Dr, Natick, MA d. SAS Institute Inc, 100 SAS Campus Dr, Cary, NC e. Insightful Corp, 1700 Westlake Ave N, Ste 500, Seattle, WA,