Relationship between performance of selected scapular muscles and scapular abduction in standing subjects.

J DiVeta, MS, PT, is Staff Physical Therapist, Sports Medicine and Physical Therapy Clinic, Raleigh Community Hospital, 3410 Executive Dr, Ste 209, Raleigh, NC 27609. He was a master's degree candidate, Program in Physical Therapy, Old Dominion University, Norfolk, VA 23508, when this research was conducted in partial fulfillment of his degree requirements. Address all correspondence to Mr DiVeta at 4501 Hamptonshire Dr, Raleigh, NC 27613 (USA).

M Walker, MS, PT, is Assistant Professor, Program in Physical Therapy, Old Dominion University.

B Skibinski, MS, PT, is Staff Physical Therapist, The Physical Therapy Center, PC, 2501 N Glebe Rd, Ste 102, Arlington, VA 22207. He was a master's degree candidate, Program in Physical Therapy, Old Dominion University, when this research was conducted in partial fulfillment of his degree requirements.

This study was approved by the Human Subjects' Review Committee, Old Dominion University.

This article was submitted July 13, 1989, and was accepted March 21, 1990.

Muscular force production and muscular balance have been postulated to be major factors in postural alignment of a healthy population.[1-3] Kendall and associates[1,2] have proposed that varying degrees of muscle weakness affect postural alignment. In addition to the assumption that muscular weakness would cause postural deviation, Kendall and associates[1,2] have stated that postural deviation can induce muscular weakness.

Posturally induced muscular weakness, or "stretch weakness," has been defined as "the effect on muscles of remaining in a lengthened condition, however slight, beyond the neutral (physiological rest) position."[2(p270)] Kendall and McCreary state

The postural deviation may appear to

be slight, the corresponding changes in

the postural positions of the joints may

appear to be minor, but the muscles

will show significant differences related

to habitual postural position.[2(p271)]

Based on the theory of stretch weakness, a positionally elongated muscle will be relatively weak and a shortened muscle will be relatively strong.[2]

Kendall and associates[1,2] further discussed the importance of muscular balance, in terms of force production, to postural alignment. They state that an imbalance in musculature can lead to postural deviations. For example, they proposed that the postural deviation of forward shoulders can be the result of an imbalance between a shortened or stronger pectoralis minor muscle and an elongated or weaker middle trapezius muscle.[2] Kisner and Colby[3] have stated that weakness of the scapular retractors, such as the upper trapezius, lower trapezius, and rhomboid muscles, causes increased scapular abduction, or a "forward shoulders" posture, during relaxed standing. No research, to our knowledge, has been conducted that demonstrates that scapular abduction in relaxed standing is related to force production of related scapular muscles or a force balance between scapular retractors and protractors.

Electromyographic analysis shows that in relaxed standing muscular activity around the shoulder girdle occurs primarily in the supraspinatus muscle, with slight activity also in the upper fibers of the trapezius muscle.[4] Because the middle trapezius muscle has minimal electrical activity during relaxed standing, we question the assumption that a relationship exists between the strength of that muscle and scapular position in relaxed standing.

Williams and Goldspink[5] demonstrated that cat muscles maintained in lengthened positions for several weeks produced greater tension in the lengthened positions than did control muscles. Peak tension occurred at the muscles' new and longer resting lengths.[5] Neumann et al[6] investigated the effects of pelvic asymmetry, on the performance of the hip abductor muscles. Their results demonstrated that the elongated hip abductor muscles generated as much or greater torque through the range of motion than the control muscles.[6] In another study of muscle performance and posture, Walker et al[7] found that there was no relationship between abdominal muscle performance and position of the lumbar spine and pelvis in relaxed standing.

The purpose of this study was to investigate the relationship between the position of the scapula and 1) the force production of the middle trapezius and pectoralis minor muscles and 2) the ratio of middle trapezius muscle to pectoralis minor muscle force in healthy subjects. We hypothesized that the force production of shoulder girdle muscles would not be related to scapular position during relaxed standing.

Method

Subjects

Sixty subjects (30 male, 30 female) participated in the study. Subjects' ages ranged from 22 to 35 years (X = 26.0, SD = 3.8). All subjects were right-hand dominant, and all measurements were obtained from the right shoulder girdle complex. The sample of subjects was thought to be reflective of the general population in that age range and therefore defined as normal." The subjects had no history of shoulder girdle pathology and presented observed variations in postural alignment of the shoulder girdle region. The procedure was explained to each subject, and each subject signed an informed consent statement before participation in the study.

Instrumentation

Peak muscular force production (in kilograms) was measured with an Accu-Force[registered trademark]II gauge hand-held dynamometer.* This dynamometer is a prototype that is not currently commercially available. The company modified their commercial dynamometer and added a pad for patient comfort.

* Ametek, Inc., Mansfield and Green Div, 8600 Somerset Dr, Largo, FL 34643.

Scapular measurements were obtained by the use of an unmarked section of string to ensure that the measurement would be unbiased. A tape measure was used later to measure each piece of string.

Procedure

The dynamometer was calibrated before each data-collection session. The procedure began by first calibrating a balance scale with a 2-kg weight. Next, the scale was set at 10 kg, and the dynamometer was pressed downward onto the scale until stabilization was attained. At this point, the force recording of the dynamometer was observed. The same procedure was repeated for scale settings of 20, 30, and 40 kg. All dynamometer values were within [plus or minus] 0.5 kg of the scale values.

For the measurement of scapular distance to the spine, the subject assumed a relaxed standing position for 1 minute. The examiner (BS) then palpated and marked the spinous process of the third thoracic vertebrae with an adhesive tag. The subject then stood in a relaxed position for 30 seconds before the same examiner palpated the inferior angle of the acromion. The examiner then used an unmarked section of string to measure the distance from the inferior angle of the acromion to the spinous process of the third thoracic vertebrae. The outstretched piece of string was marked at the site of the palpated inferior angle of the acromion and at the tagged third thoracic spinous process. The linear distance from the third thoracic vertebra to the inferior angle of the acromion was defined as the total scapular distance.

The examiner, while maintaining the position of the string at the inferior angle of the acromion, palpated the medial border of the scapula at the root of the spine. At this time, the examiner measured the length of the scapula, as shown in Figure 2, by placing the string from the inferior angle of the acromion across the spine of the scapula and marking the string at the medial border of the scapula. The distance from the root of the spine to the inferior angle of the acromion was defined as the length of the scapula. The markers were removed and the entire procedure was repeated a second time by the same examiner after a 1-minute rest period. This test-retest procedure allowed the examiners to evaluate the intrarater reliability of scapular measurements. At the completion of the procedure, the marked areas of the string were measured (in centimeters) to determine the total scapular distance.

A normalized value of scapular abduction was calculated to account for subject size. We believe that a measurement of the distance from the spinous process of the third thoracic vertebrae to the medial border of the scapula would not accurately reflect the amount of scapular abduction. For example, a distance of 7 cm would represent a greater abducted position for a smaller person when compared with a larger person. For this reason, scapular abduction values were normalized to body size by dividing the total scapular distance from the third thoracic vertebrae by the length of the scapula.

To determine the force production of the middle trapezius muscle, the subject was stabilized prone on a treatment table in the manual muscle test position described by Daniels and Worthingham.[8] The force pad of the hand-held dynamometer was placed on the lateral angle of the right scapula.[8] The subject was instructed on the performance of an "isometric make test" and permitted one sub-maximal practice trial. The isometric make test has been described by Bohannon[9] as a contraction in which the subject exerts a maximal effort against a stationary force pad. The subject rested for 30 seconds and then exerted a maximal contraction for 4 to 5 seconds against the stabilized hand-held dynamometer. The subject was permitted to rest for 30 seconds before a second maximal contraction was obtained. The force production of the middle trapezius muscle was defined as the peak force recorded by the hand-held dynamometer during an isometric make test in the manual muscle test position described by Daniels and Worthingham.[8] This test position was chosen over the position described by Kendall and McCreary[2] because the authors felt that it would better reflect the force production of the middle trapezius muscle. This reasoning was based on the observation that Kendall and McCreary's test position is dependent on the stabilizing ability of the glenohumeral musculature and also tests across multiple joints.

The measurement for pectoralis minor muscle force was obtained with the subject stabilized supine on a treatment table in the manual muscle test position as described by Kendall and McCreary.[2] The force pad of the hand-held dynamometer was placed on the anterior aspect of the shoulder. The subject performed one practice submaximal isometric make test by thrusting the right shoulder forward while the arm remained in a relaxed position on the treatment table. The subject rested for 30 seconds and then exerted a maximal contraction for 4 to 5 seconds against the stabilized hand-held dynamometer. The subject was permitted to rest for 30 seconds before a second maximal contraction was obtained. The force production of the pectoralis minor muscle was defined as the peak force recorded by the hand-held dynamometer during an isometric make test in the manual muscle test position described by Kendall and McCreary.[2]

The force values obtained from the muscle tests were normalized. The normalized force values were calculated by dividing muscle force by a body mass index. The body mass index was defined as the subject's weight (in kilograms) divided by the subject's height (in meters squared).[10] To examine the "balance" of forces between the middle trapezius and pectoralis minor muscles, a ratio of the two force measurements was made. This ratio was not normalized to body mass index because creating the ratio was a form of normalization and the use of the body mass index was not needed.

Data Analysis

Means, standard deviations, and ranges were calculated for all measured and normalized values. Intrarater reliability for repeated measures of muscular force and scapular measurements was determined by the use of the intraclass correlation coefficient (ICC[1,1]).[11] This form of the ICC was chosen because it uses a one-way analysis of variance (ANOVA) to measure the reliability of a single examiner.[11] The other forms of the ICC, as noted by Riddle et al,[12] require a two-way ANOVA and partition out the variance, which is attributed to the examiner.

The correlation between muscular force measurements and scapular abduction measurements was determined by the use of the Pearson product-moment correlation coefficient.[13] This statistic allowed us to test for a direct relationship, as Kendall and McCreary[2] have stated exists, between the variables.

Reliability

The reliability values (ICCs) ranged from .85 to .96 for all measurements. The intrarater reliability for muscular force measurements tended to be slightly higher than for scapular measurements. The ICC value for scapular length was .85 and for scapular distance was .94. The ICC value for middle trapezius muscle force was .96 and for pectoralis minor muscle force was .96.

The measurements of isometric muscular force used in this study were highly reliable. This finding is in agreement with previous studies, which have shown measurements with a hand-held dynamometer to be reliable when used by one tester during a single session.[9,14,15] The measurement of scapular abduction, as used in this study, was also shown to be reliable.

Results

Table 1 shows the means, standard deviations, and ranges for measurements and normalized values. Pearson product-moment correlation coefficients are presented in Table 2. The correlation between scapular abduction and middle trapezius muscle force was .20. The correlation between scapular abduction and pectoralis minor muscle force was .14. The correlation between scapular abduction and a ratio of middle trapezius muscle force to pectoralis minor muscle force was .01.

Discussion

Force Measurements

Isolated measurement of forces generated by the middle trapezius and pectoralis minor muscles by use of manual muscle test positions and a hand-held dynamometer is questionable. The contributions of the rhomboid muscles during middle trapezius muscle testing and the serratus anterior muscle during pectoralis minor muscle testing cannot be eliminated. Kendall and McCreary,[2] however, emphasize that the manual muscle test can be used to confirm weakness of specific muscles that they postulate to be correlated to postural deviations. They specifically cite the middle trapezius and pectoralis minor muscles as being related to scapular abduction.[2] Because we were testing for a relationship that had been assumed based on manual muscle testing, muscle testing of the middle trapezius and pectoralis minor muscles was chosen.

The muscle force values were normalized to control for variations in subject size. A larger person would be expected to produce more force than a smaller person. Dividing the force values by the body mass index gave us an indication of relative force for each person.

Scapular Measurements

Our search of the literature did not reveal a measurement method for quantifying scapular abduction. Kendall and McCreary[2] suggest use of a descriptive method based on observation. A linear measurement of scapular abduction does not fully measure the complex phenomenon that we call "forward shoulders." This phenomenon may include "winging" of the scapulae and medial (internal) rotation of the humerus along with scapular abduction. Our ability to adequately measure a forward-shoulders posture was limited. We feel, however, that we did measure scapular abduction adequately. Scapular abduction is a main component of a forward-shoulders posture. Kendall and McCreary[2] even use the terms "abducted scapulae" and "forward shoulders" interchangeably. Therefore, a linear measurement of scapular abduction should have been sufficient to detect what Kendall and McCreary[2] define as forward shoulders.

The scapular measurements were normalized to control for differences in subject size. If we had not normalized the distance, a larger person would have appeared to have increased scapular abduction when compared with a smaller person with the same posture.

Correlation Between Scapular Abduction and Muscular Force

The low correlation between the measurements of scapular abduction and muscle performance of the middle trapezius and pectoralis minor muscles demonstrates that a strong, direct relationship between these variables does not exist. Our finding leads us to question the assumption that isometric muscular force production and posture are closely related in healthy subjects. Our findings lead us to question Kendall and McCreary's[2] assumption that a weak middle trapezius muscle leads to an increase in scapular abduction. Our results indirectly also call into question the assumption that posture can induce muscular weakness. The data indicated that the length of the scapular muscles in relaxed standing did not appear to be related to the force production of those muscles when tested in the shortened muscle length test position.

The clinical observation of the relationship between muscular balance, in terms of force production, and postural alignment has also been contradicted by the results of this investigation. Table 2 shows that the ratio of force production of the middle trapezius and pectoralis minor muscles had no relationship to the degree of scapular abduction. If a relationship had existed, the ratio would be expected to decrease in direct proportion to increasing scapular abduction based on the assumption that the lengthened middle trapezius muscle would be weaker and the shortened pectoralis minor muscle would be stronger. The results of this investigation, however, do not demonstrate this relationship.

The assumption that clinical observation of postural malalignment can be used to predict muscular weakness or that clinical assessment of muscular weakness can be used to predict postural deviations in relaxed standing was not supported by the results of our investigation.

A possible limitation of our study may have been the relatively low variance in scapular measurements. For example, scapular distance varied only 9cm. A critical amount of scapular abduction may not have existed to show significant changes in the force production of related scapular muscles. Kendall and McCreary,[2] however, emphasize that the postural deviation only needed to be slight in order to observe marked variations in the muscle. They emphasize that the duration of a postural deviation may be of greater consequence than the severity in leading to weakness. Figure 3 shows that subjects did demonstrate what we consider to be normal variations in posture. We believe that the variation in scapular abduction, although small, represented normal variations within this age group. Given our sample size, if the relationship between variables exists in a healthy population, we believe it should have been evident.

Subjects of both genders participated in this study. We questioned whether considering male and female subjects together in our analysis had an effect on the correlations. After separating the subjects by gender, a post hoc analysis (Pearson product-moment correlation method) was performed. The correlations from this analysis ranged from -.02 to .28. These values are not substantially different from those obtained when the data from all subjects are combined.

Even though the variance in scapular abduction measurements was not large, moderate correlations would be expected if the assumptions of Kendall and McCreary[2] are correct. The extremely low values, however, support the hypothesis that no direct relationship exists between the force production of scapular musculature and the degree of scapular abduction in healthy subjects in this age group.

Clinical implications

Based on the results of this study, we question some procedures currently used in clinical practice. The practice of using postural alignment to make predictions about the weakness or strength of related muscle groups may not be valid. This type of clinical problem solving assumes a direct relationship between isometric muscle performance and posture. The findings of our study contradict this assumption.

The prescription of muscle strengthening exercises to correct a forward-shoulders posture is not supported by this or any other study. Because there was no relationship between the force production of scapular muscles and the position of the scapula in relaxed standing, this treatment approach does not seem to be valid.

Further Research

Investigating the variables of muscle performance and posture using different age groups or other areas of the body would enhance the scope of our knowledge. Another approach would be research designed to evaluate whether muscular force production is significantly different among individuals with varying degrees of postural deviations as classified by observation.

Conclusion

The measurements of scapular distance and isometric scapular muscle force were shown to be reliable when taken by a single examiner in the manner used in this study. Using these measurements, we found no direct correlation between isometric force production of the middle trapezius muscle and scapular abduction. We found no direct correlation between isometric force production of the pectoralis minor muscle and scapular abduction. In addition, we found no direct correlation between a ratio of middle trapezius muscle to pectoralis minor muscle force production and scapular abduction. Based on these results, there does not appear to he a relationship between isometric muscle performance or a balance of antagonist muscle performance and position of the scapula in relaxed standing in healthy subjects.

Acknowledgments

We thank JL Echternach, EdD, PT, for his assistance in the preparation of this manuscript and SD Finucane, PT, and DL Riddle, MS, PT, for their help during the design of the investigation. We also thank SK DiVeta for her work on the illustration.

Table 2. Pearson Product-Moment
Correlations Between Scapular Abduction
and Middle Trapezius Muscle Force and
Pectoralis Minor Muscle Force
Correlation                                r
Scapular abduction vs middle
  trapezius muscle force                  .20
Scapular abduction vs pectoralis
  minor muscle force                      .14
Scapular abduction vs ratio of middle
  trapezius muscle force to
  pectoralis minor muscle force           .01

References

1. Kendall HO, Kendall FP, Boynton PA. Postture and pain. Baltimore, Md: Williams & Wilkins; 1977.

2. Kendall FP, McCreary EK. Muscles: Testing and Function. 3rd ed. Baltimore, Md: Williams & Wilkins; 1983.

3. Kisner C, Colby LA. Therapeutic Exercise: Foundations and Techniques. 6th ed. Philadelphia, Pa: FA Davis Co; 1987.

4. Basmajian JV, DeLuca CJ. Muscles Alive: Their Functions Revealed by Electromyography. 5th ed. Baltimore, Md: Williams & Wilkins; 1985

5. Williams PE, Goldspink G. Changes in sarcomere length and physiological properties in immobilized muscle. J Anat. 1978; 127:459-468.

6. Neumann DA, Soderberg GL, Cook TM. Comparison of maximal isometric hip abductor muscle torques between hip sides. Phys Ther 1988;68:496-502.

7. Walker ML, Rothstein JM, Finucane SD, Lamb RL. Relationships between lumbar lordosis, pelvic tilt, and abdominal muscle performance Phys Ther. 1987;67:512-516.

8 Daniels L, Worthingham C. Muscle Testing: Techniques of manual Examination. 4th ed. Philadelphia, Pa: WB Saunders Co; 1980:94-96.

9. Bohannon RW. Make tests and break tests of elbow flexor muscle strength. Phys Ther, 1988;68:193-194.

10. Pollock ML, Schmidt DH, Jackson AS. Measurement of cardiorespiratory fitness and body composition in the clinical setting. Compr Ther. 1980:6:12-27.

11. Shrout PE, Fleiss JL. intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420-428.

12. Riddle DL Finucane SD, Rothstein JM, Walker ML Intrasession and intersession reliability of hand-held dynamometer measurements taken on brain-damaged patients. Phys Ther 1989;69:182-194.

13. Kerlinger FN. Foundations of Behavioral Research. 3rd ed. New York, NY: Holt, Rinehart & Winston General Book; 1986.

14. Bohannon RW, Andrews AW. Interrater reliability of hand-held dynamometry. Phys Ther. 1987;67:931-933.

15. Bohannon RW. Test-retest reliability of hand-held dynamometry during a single session of strength assessment. Phys Ther, 1986;66:206-209.

Commentary

DiVeta and colleagues are to be commended on their efforts to examine one of the hypotheses of the Kendalls concerning the relationship between alignment and muscle strength. Because the Kendalls have provided the profession with the best-stated and most clearly explained theories relating faults in posture to muscle performance, they have made it possible to study the accuracy of their ideas. Careful scrutiny of the relationship among alignment, muscle length, and movement is particularly important because I believe these elements are factors in musculoskeletal pain syndromes and could provide guidelines for preventive programs. Therefore, critical analysis that provides an accurate and fair examination of the hypotheses is not only useful but necessary.

The investigators have been thorough in their attempt to consider variations that could exist because of body size. They have normalized their data by using a ratio of the amount of scapular abduction, measured from T3 to the tip of the acromion, to the scapular width. They also normalized the dynamometer force measurements by using the ratio of the force production by a body index, They used these ratios for their correlational analysis. Unfortunately, they did not report intraclass correlation coefficients (ICCs) for these ratio variables, but reported only those for their direct measurements. Ratio variables are known to be less reliable than nonratio data. Thus, the reliability of the data used for correlational analysis may be somewhat less than is suggested by the ICCs obtained from their direct measurements and may further detract from demonstrating a correlation between force production and scapular abduction.

Although my own clinical experience raises questions about the exact nature of the "stretch weakness" concept provided by the Kendalls, I do not believe this article is a fair assessment of their hypothesis. My major concern is the one expressed by the authors themselves, namely, that there was not enough variation in the scapular position of the subjects examined. The ratio variable that they used for analyzing scapular abduction had a mean of 1.6 with a standard deviation of less than 0.1 and a range of only 1.4 to 1.8. This truncated range makes it essentially impossible to demonstrate a correlation because of the lack of variability.

As I understand the Kendalls' hypothesis, the scapula must be in a faulty position, with the muscle beyond its physiological resting length, for weakness to occur. The hypothesis does not state that a linear relationship exists between the length of the muscle and its strength. This study examined the hypothesis that, given a normalized scapular position, the more abducted the scapula, the less force the middle trapezius muscle would generate. Such a hypothesis would also imply that all individuals would be able to generate the same force, given certain correction factors, in their middle trapezius muscle at a standardized position and that variation in force from that position would be entirely the effect of the change in muscle length.

Stretch weakness, as defined by the Kendalls, would only become a factor in the scapula that is abducted beyond the ideal position. The population sample used in this study in which the standard deviation of the scapular position was less than 2 cm and the normalized scapular abduction standard deviation was less than 0.1 cannot be considered as a sample that includes a meaningful number of individuals with faulty scapular position. The means for scapular and force measures were near the low end of the range, which further indicates that the sample did not represent a good normal distribution. I believe the design for a "fair" analysis of stretch weakness would test subjects having a wide range of scapular positions and thus would include those with extreme scapular abduction. This design would avoid the truncated range problem present in this study. I agree totally with the authors when they stated that their study supported the hypothesis that no direct relationship exists between the force production of the middle trapezius muscle and the degree of scapular abduction in healthy subjects in this age group. I must disagree, however, that they can generalize these findings to subjects with a markedly abducted scapula. I must also disagree with their conclusion that the results of this study contradict the relationship between faulty scapular position and muscle strength.

Clarification of other concepts and hypotheses concerning muscle length and strength could be useful and hopefully encourage further studies. Another hypothesis concerning a finding of stretch weakness arises from two of the studies cited in this article.[1,2] That hypothesis, which Gossman and colleagues[3] called "length-associated changes," states that a muscle that has added sarcomeres by remaining in a lengthened position for a prolonged period of time is strong, but tests weak in a shortened position such as is used in manual muscle testing. DiVeta and colleagues did not mention any control for the amount of abduction of the scapula during the force production measurement. When testing patients using the Kendall method, I am careful to start the test with the scapula in abduction. Most often, when individuals who have forward shoulders are in the prone position, their shoulders drop into the forward position and the scapulae are abducted. The test position used in this study does not correct for this problem as well as the position used by the Kendalls or direct attention to scapular position. Thus, the possibility exists that the subjects were tested with the middle trapezius muscle in a lengthened position rather than in a shortened position. Control of the muscle length during testing would be important when assessing the length-associated change hypothesis.

As stated by the investigators, during relaxed standing, alignment is maintained by passive properties of the musculotendinous complex and not by active muscle contraction; thus, the concept that strength is required to hold segments in their proper position while relaxed is not the main issue. I do not believe that this is the rationale behind the Kendalls' hypothesis. Rather, I believe it is inarguable that faults in alignment are associated with alterations in muscle length and that, once established as the familiar length, concepts such as the equilibrium-point theory or mass-spring model can explain the persistent return to a constant muscle length.[4] One of the factors that would affect this equilibrium point is the passive tension of the muscle. The passive tension of the muscle is, in part, a function of the number of myofilaments, which does relate to force production. A variety of mechanisms underlying the alteration in length can be given, including weakness of the muscles, as is seen in neurological conditions. Although the Kendalls identify the fault in a lengthened muscle by testing its strength, the emphasis of their corrective program is on using the muscle in its normal range or in a shortened range and on not performing resistive exercises in the already lengthened range. This study was a reasonable attempt to answer an important issue but did not provide the required sample. Thus, I do not believe the definitive statements made in the conclusion are appropriate. Additional investigations of the relationship between muscle length and strength are necessary. The insights gained from this study and the commentary should provide a useful basis for an even more critical analysis of this important question.

Shirley A Sahrmann, PhD, PT, FAPTA Associate Director for Research Program in Physical Therapy Washington University School of

Medicine 660 S Euclid Ave St Louis, MO 63110

References

1. Williams PE, Goldspink G. Changes in sarcomere length and physiological properties in immobilized muscle. J Anat. 1978; 127:459-468.

2. Neumann DA, Soderberg GL, Cook TM. Comparison of maximal isometric hip abductor muscle torques between hip sides. Phys Ther. 1988;68:496-502.

3. Gossman MR, Sahrmann SA, Rose SJ. Review of length-associated changes in muscle: experimental evidence and clinical implications. Phys Ther. 1982;62:1799-1808.

4. Brooks VB. The Neural Basis of Motor Control New York, NY: Oxford University Press Inc; 1986.

Author Response

We thank Dr Sahrmann for her thoughtful commentary. We have considered her comments carefully, and we hope that our response will help to clarify points that were not clear in our article. We believe that open discussion is valuable for the advancement of the profession.

Intraclass correlation coefficient (ICC) values for measurements of scapular length, scapular distance, and muscle force were calculated so that the reliability of these measurements could be assessed. As Dr Sahrmann points out, we neglected to test the reliability of the normalized values of scapular abduction and force that were derived from these measurements. The reliability of the normalized values could be lower, higher, or the same as the measurements from which they were derived. Therefore, we should have evaluated them directly. We thank Dr Sahrmann for calling this omission to our attention, and we have since examined the reliability of the normalized values.

The ICC value for scapular abduction was .78. This value is moderately less than the values for scapular length (ICC = .85) and scapular distance (ICC = .94). Although this value represents a decrease in reliability, we feel that the measurement of scapular abduction is reliable enough to demonstrate a strong correlation between the variables, if one exists. The ICC values for normalized middle trapezius and pectoralis minor muscle force were both .96. These values were the same as the ICC values of the force measurements. We conclude from this that the values we used to test for a relationship between scapular position and muscle force were reliable.

We have noted that measured scapular position did not vary greatly even though the subjects displayed significant variations in observed postural alignment. Given the large sample size (N = 60), we do not consider our measurements to represent only a truncated portion of a much larger range of scapular abduction values from the general population. Rather, we consider that our values represent a normal range of measurements for scapular abduction, and this range is normally very small. The scapula can only deviate so far. The point that this lack of variation makes correlation with the wide range of muscle performance values unlikely is not a fault of this study; it is a condition of the phenomenon we are examining, and it should make even more clear the highly questionable nature of attempting to correlate scapular position with muscle performance in healthy subjects.

The hypothesis that we have tested is based directly on what Kendall and McCreary[1] have stated. They state that posturally induced weakness will result from "the effect on muscles of remaining in a lengthened condition, however slight, beyond the neutral (physiological rest) position."[1(p270)] In addition, they state that

the postural deviation may appear to

be slight, the corresponding changes in

the postural positions of the joints may

appear to be minor, but the muscles

will show significant differences related

to habitual postural position.[1(p271)]

We found that this was not the case. Sahrmann seems to be agreeing with us that such variations in position are unlikely to be major factors in muscular force production.

Sahrmann believes that this study was not a fair analysis of "stretch weakness." In fact, the study was not designed to analyze "stretch weakness"; it was designed to test for a relationship between scapular position and the performance of selected scapular muscles in a healthy population of young adults. This relationship was not found in this sample. We believe that this is because the concept of stretch weakness, on which the relationship is presumed, is flawed. We invite Dr Sahrmann and others to look for this relationship in other populations. To date, no research has demonstrated that such a thing as stretch weakness exists in any population.

The examiners in this study did control for the amount of scapular abduction in the muscle test position by supporting the upper extremity at the midaxillary line. This position ensured that all muscles were tested in what would be considered a standardized anatomical alignment.

It is interesting that Sahrmann uses the phenomenon of length-associated changes in muscle to support the existence of stretch weakness. Neither study she cited found weakness at resting length resulting from adaptive lengthening.[2,3] Williams and Goldspink,[2] using animals, found that lengthened muscles produced greater active tension than either control muscles or shortened muscles, even when compared at the resting length of the control muscles. When the muscles were tested at shorter positions, a decrease in tension became apparent because of a shift of the length-tension curve of the adaptively lengthened muscles. It has not been shown that postural deviations cause enough adaptive lengthening in muscle so as to significantly shift the length-tension curve.

Neumann et al,[3] in their study of hip abductor muscle length and torque production, found that posturally lengthened muscles produced more torque at all points in the range measured, although this increase in torque was not statistically significant except at 0 and - 10 degrees of hip abduction. No weakness in these lengthened muscles was ever measured; it was merely hypothesized to exist if one extrapolated the hip position to beyond 52 degrees of abduction. In addition, it is not clear at what angle beyond 52 degrees this weakness would become statistically significant, if at all.

It should be emphasized that even a significant shift in the length-tension curve, as Williams and Goldspink[2] found, does not automatically mean a significant shift in the torque curve of an intact muscle. Biomechanical factors such as the angle of muscle insertion and limitations in joint range of motion would lessen the clinical significance of any length-tension shift.

Although length may be a factor in muscle performance, the clinical significance of posturally lengthened muscle has not been demonstrated. The current study does not support the idea that posturally induced changes in muscle length have any significant effect on muscular force production.

Dr Sahrmann states that the rationale behind the Kendalls' hypothesis is not that strength is required to hold segments in their proper position. With due respect, we would like to suggest that through her own work in the area, Dr Sahrmann has moved beyond the original hypotheses proposed by Kendall and McCreary.[1] Kendall and McCreary state, "Muscle weakness allows separation of the parts to which the muscle is attached because the strength to maintain good alignment is lacking."[1] We believe that this statement is a clear indication that Kendall and McCreary believe that muscular strength is directly related to ideal postural alignment. They justify this statement by observing the effect that paralysis has on muscle function.[1]

Sahrmann similarly states that muscular weakness, as seen in neurological conditions, is one of the mechanisms underlying the change in muscle length. We are concerned that a connection has been made between pathological changes seen in neuro-muscular diseases and variations seen in a nonpathologic population. It is clear that severe atrophic paralysis of muscles as seen in poliomyelitis or peripheral nerve disease can induce changes in skeletal alignment.[4] However, we do not believe that it is valid to correlate the mechanism for alignment changes seen with paralysis and severe atrophy to those seen in a healthy population. It appears that both Kendall and McCreary[1] and Dr Sahrmann have tried to justify their ideas through this connection. As a result, Kendall and McCreary[1] have recommended strengthening exercises for the middle trapezius muscle to address the postural fault of forward shoulders. We feel that it is of utmost importance to clarify ideas such as these and examine them fully, especially since clinical practice is directly affected by them.

In summary, we believe that this study used a large sample of young adults with average variations in scapular position. The measurements used were found to be reliable. Using these measurements, we found no direct correlation between isometric muscle performance or a balance of antagonist muscle performance and position of the scapula in healthy subjects. We look forward to seeing more studies in this area of research.

Martha L Walker

Jack DiVeta

Bernard Skibinski

References

1. Kendall FP, McCreary EK. Muscles: Testing and Function. 3rd ed. Baltimore, Md: Williams & Wilkins; 1983.

2. Williams PE, Goldspink G. Changes in sarcomere length and physiological properties in immobilized muscle. J Anat. 1978;127:459-468.

3. Neumann DA, Soderberg GL, Cook TM. Comparison of maximal isometric hip abductor muscle torques between hip sides. Phys Ther. 1988;68:496-502.

4. Adams RD, Victor M. Principles of Neurology. 3rd ed. New York, NY: McGraw-Hill Book Co; 1985.