SHOULDER
Scapular behavior in shoulder impingement syndrome
Abstract
Hébert LJ, Moffet H, McFadyen BJ, Dionne CE. Scapular behavior in shoulder impingement syndrome. Arch Phys Med Rehabil 2002;83;60-9. Objective: To quantify the contribution of each scapular rotation to the scapular total range of motion (ROM) in both shoulders of persons with a unilateral shoulder impingement syndrome (SIS), to compare 3-dimensional (3D) scapular attitudes of their symptomatic and asymptomatic shoulders in flexion and in abduction, and to characterize the scapular behavior of these subjects by classifying them into subgroups based on scapular tilting differences between their symptomatic and asymptomatic shoulders. Design: Comparisons of 3D scapular attitudes, scapular total ROM, and percentage of contributions of each scapular rotation to the scapular total ROM. Setting: A motricity laboratory. Participants: Fifty-one subjects, including 41 with a SIS (29 had an asymptomatic contralateral shoulder) and 10 healthy subjects. Interventions: The 3D scapular attitudes were calculated with the subjects in a standardized seated position; with the arm at rest; or at 70°, 90°, and 110° of shoulder flexion and abduction. Axial rotation angles were calculated using a fixed set of Cardanic angles. Main Outcome Measures: At 90° of arm elevation, data from 10 shoulders of healthy subjects were used to set up normative values (99% confidence interval of mean 3D scapular attitudes) to compare with 3D scapular attitudes of symptomatic and asymptomatic shoulders of SIS subjects. We analyzed the scapula behavior of subjects with SIS and classified them into subgroups based on scapular anterior tilting asymmetry. Results: In flexion, almost half of the scapular total ROM was provided by anterior tilting (48.2%-51.3%), whereas in abduction, external rotation (40.3%-42.4%) was the main contributor. Scapular total ROM was higher in abduction than in flexion in all arm positions for both shoulder groups (P <.01). Also, 3D scapular attitude patterns of both shoulders of SIS subjects were different from those of healthy subjects. At 90°, scapular asymmetry in anterior tilting allowed us to classify SIS subjects with respect to more (lead) or less (lag) scapular tilting in the affected side (P <.0001) or no difference (P =.11) between the sides (symmetrical). No significant differences (P >.05), except for a small 2° difference in transverse rotation during arm flexion at 110° (P =.002), were observed in 3D scapular attitudes and scapular total ROM between both shoulders of SIS subjects. Patterns of 3D scapular attitudes and scapular total ROM were significantly different between flexion and abduction arm positions (P <.05). Conclusions: The contribution of rotations and scapular total ROM differed according to the plane of arm elevation in SIS subjects. Group analyses revealed no differences in 3D scapular attitudes between symptomatic and asymptomatic shoulders of subjects with unilateral SIS. This could be caused by the use, in SIS subjects, of inappropriate neuromuscular strategies affecting both shoulders. However, individual analyses revealed scapular asymmetry in the sagittal plane, which suggests that SIS subjects with less anterior tilting in the symptomatic shoulder, as compared with the asymptomatic contralateral one, may be at high risk of developing chronic SIS. This last finding provides scientific evidence to focus rehabilitation protocols toward a restoration of anterior tilting. © 2002 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
Abnormalities in posture and scapular motion are considered important risk factors for developing shoulder pathologies.1,2, 3, 4, 5, 6, 7, 8, 9, 10 In patients with shoulder impingement syndrome (SIS), it is well known that an excessive narrowing of the subacromial space and a smaller acromiohumeral distance (AHD) occur during shoulder elevation movements.11, 12,13, 14, 15, 16 However, the specific contribution of the scapula to such impairment is not known.
Postural abnormalities such as abnormal scapular positioning have been related to SIS. An asymmetry of more than 1cm in scapular protraction (a combination of superior translation and external rotation of the scapula) between both shoulders of a subject has been associated with SIS.17 The influence of protraction on the reduction of subacromial space was also confirmed by magnetic resonance imaging (MRI).5 In contrast, the work of Greenfield et al18 supports the association between an abnormal scapular positioning and the presence of SIS. These investigators used 2-dimensional (2D) measurements and found no differences in posture between subjects with shoulder overuse injuries and asymptomatic subjects or between the involved and noninvolved shoulders of patients with SIS with respect to scapular protraction and rotation, mid-thoracic curvature, and scapular resting posture.18 However, as in other studies involving healthy subjects,7,19, 20, 21 one can argue that 2D measurements are limited in describing the complex 3-dimensional (3D) character of scapular posture and motion. Studies using in vivo 3D analysis of movements have clearly shown that scapula moves and rotates around 3 orthonormal axes, and each of these rotations contributes to produce the total range of motion (ROM) of the scapula.22, 23, 24, 25, 26, 27 Each of these rotations, when altered, has an impact on the degree of freedom of the shoulder girdle and will ultimately affect the function of the shoulder joint.
Abnormal scapular motions have also been closely associated with SIS.10, 28, 29, 30, 31 It was suggested that subacromial impingement is caused by a decrease in scapular external rotation, also referred to as scapular upward rotation, a motion that occurs in the frontal plane.28, 29, 30 By using a 3D video system, Yanai and Hay10 analyzed the stroking technique of front-crawl swimming and showed that decreased scapular external rotation, referred to by the investigators as “decreased rotation of the shoulder girdles” or “decreased tilt angle,” was 1 of the 3 factors associated with a high risk of developing SIS in swimmers. The second scapular motion also considered as a risk factor is scapular tilting, a motion that occurs in the sagittal plane. Works published by Lukasiewicz et al31 and Cole et al4 suggest that a decrease in scapular tilting during arm elevation would cause a shoulder impingement. Both studies showed that, when elevating the arm, subjects with SIS had significantly less scapular posterior tilting (tilting of the acromion away from the greater tuberosity of the humerus and driving of the angulus inferior into the thorax) compared with the asymptomatic side31 or nonimpaired subjects.4, 31
Until now, although some indications are available, there is still no consensus on the specific scapular behavior responsible for subacromial impingement. It remains essential to characterize and compare 3D scapular posture and motions of both shoulders of subjects with SIS in different positions. This would provide additional data to understand better the biomechanics of the scapula in SIS. Moreover, it would provide scientific evidence to support current rehabilitation protocols that emphasize restoring scapulothoracic control6, 9, 30 and guide the development of new rehabilitation treatments according to evidence-based practice principles. The aims of this study were (1) to quantify the contribution of each scapular rotation to the scapular total ROM in both shoulders of persons with a unilateral SIS using a standardized protocol27; (2) to compare 3D scapular attitudes between symptomatic and asymptomatic shoulders of these persons at 70°, 90°, and 110° in flexion and in abduction; and (3) to characterize the scapular behavior of these subjects by classifying them into subgroups based on scapular tilting differences between their symptomatic and asymptomatic shoulders and data recorded from healthy subjects.
Methods
Subject selection
Each subject who voluntarily participated in this study had a diagnosis of unilateral SIS. Subjects were prospectively recruited from private clinics and a university hospital. All subjects were screened by the same experienced orthopedic surgeon to rule out fractures, bone abnormalities, calcification, and shoulder instability. A subject was considered eligible if there was at least 1 positive finding for the affected shoulder (shoulder impingement [SHimp]), in each of the 3 following categories: (1) painful movement during active shoulder flexion or abduction; (2) positive Neer32 or Hawkins–Kennedy33impingement signs; or (3) pain on resisted isometric lateral rotation, abduction, or Jobe test.34 The asymptomatic contralateral shoulder (SHctl) was assessed to rule out bilateral SIS and shoulder instability. A subject was considered ineligible if he/she had any of the following criteria: (1) rheumatoid, inflammatory, degenerative, or neurologic diseases; (2) stroke; (3) previous surgery involving the neck or the shoulder; (4) neck pain or restricted neck motion, cervicobrachialgia, or shoulder pain reproduced during active neck movements or surpressions; (5) trapezius myalgia syndrome; (6) shoulder capsulitis, defined as a restriction of active and passive glenohumeral movements of at least 30% of amplitude for 2 or more directions of shoulder movements; and (7) problems of collaboration interfering with tests.
Informed consent was obtained after the nature of the procedures of the study had been fully explained and understood. This study was approved by the Ethics Committee of the Quebec Rehabilitation Institute.
Study sample
SIS subjects (21 men, 20 women) were 30 to 60 years old (mean ± standard deviation [SD], 44.3 ± 9.2yr) with a mean duration of symptoms of 29.4 ± 22.9 months. Among the 41 subjects, 10 had left SHimp and 31 had right SHimp; 29 impingements were on the dominant side. Within this cohort of 41 subjects, 29 subjects were confirmed to have unilateral SIS, therefore, providing an SHctl for within-subject comparisons (SHctl, 8 left, 21 right). Ten (4 men, 6 women) healthy shoulders (SHhe) were also assessed (5 left, 5 right) and submitted to the same screening as SHctl. Healthy subjects were all right-handed, with a mean age of 34.4 ± 8.4 years.
Data collection
Data was collected at the Human Motricity Analysis Laboratory of the Quebec Rehabilitation Institute using a 3D analysis system.a A 3D scapular attitude was defined as the 3D orientation of the scapula at a specific shoulder position relative to the resting position (ie, the arm at the subject's side at a position of 0° of arm elevation). As shown in figure 1, specifically, the 3D scapular attitudes reflect scapular rotations from resting position around 3 orthogonal axes: the external-internal rotation (E-IR; frontal plane, Zs axis), anteroposterior tilting (APT; sagittal plane, Ys axis), and anteroposterior transverse rotation (APTR; transverse plane, Xs axis) of the scapula.
Fig. 1
Axes and directions of scapular rotations. A local coordinate reference system was determined with P1 as the origin. The 3D scapular attitudes reflect scapular rotations from resting position around 3 orthogonal axes: the E-IR (frontal plane, Zs axis), APT (sagittal plane, Ys axis), and APTR (transverse plane, Xs axis) of the scapula. Scapular attitudes were calculated from a rotation matrix using a ZsXsYs Cardan sequence of rotations. A right-hand rule was used to determine positive rotation. The scapular reference frame (Xs, Ys, Zs) for a right shoulder had the positive Z axis directed posteriorly. The positive Y axis was directed downward along the lateral border of the scapula crossing P1 and P2. The positive X axis, directed obliquely upward toward the left shoulder, was the cross product of the Z and Y axes. Abbreviations: ER, external rotation, IR, internal rotation; ATR, anterior transverse rotation, PTR, posterior transverse rotation.
The calculation method used in this study has been thoroughly described previously27 and shown to be an accurate estimation of 3D scapular attitudes within less than 2° on average. By convention, (1) scapular anterior tilting describes a relative tilting of the scapula when the acromion moves away from the greater tuberosity, and the angulus inferior moves closer to the rib cage; (2) scapular external rotation corresponds to the upward rotation of the angulus inferior in the frontal plane; and (3) posterior transverse rotation moves the spinal edge of the scapula away from the thorax and, therefore, defines the winging of the scapula.
The 3D scapular attitudes were measured in a standardized sitting position with the trunk in an erect position, the back against the chair, the knees and hips flexed at 90°, and the feet flat on the floor. Two adjustable supports fixed to the chair, one posterior at the base of the skull and the other anterior against the thorax, were used to eliminate movements of the seated subject. Scapular attitudes were measured at rest; at 70°, 90°, and 110° of flexion in the sagittal plane; and at 70°, 90°, and 110° of abduction in the frontal plane. The 3D scapular attitudes were measured while the subject was actively maintaining his/her arm against gravity at the desired angle. The shoulder position was monitored with a fluid goniometer (mean error,.96° ±.48°) attached with Velcro® straps to the lateral aspect of the arm.27 Trials showing more than 5° of difference of humeral elevation above or below the target angle (70°, 90°, 110°) were rejected. Two measures of 3D scapular attitudes per shoulder position were recorded.
An index called the scapular total ROM was defined as the algebraic sum of the absolute values of all scapular rotations during arm elevation (|E-IR| ± |APT| + |APTR|) relative to the resting position. The scapular total ROM was previously used and considered a representative measure of the total ROM of the scapula.35
Analysis
The contribution of each scapular rotation to the scapular total ROM was calculated as a percentage. A Pearson product- moment correlation (rp) was used to quantify the relationship between each scapular rotation and the scapular total ROM. Significant bivariate associations between each rotation and total ROM were identified for each shoulder group at each shoulder position. Values of rp from.70 to.89 and from.50 and.69 were considered high and moderate correlations, respectively.36 Comparisons between shoulder positions (70°, 90°, 110°) of percentage contributions of each scapular rotation to the scapular total ROM were done with 1-way repeated-measures analysis of variance (ANOVA) for each shoulder group (symptomatic, asymptomatic). A Bonferroni adjustment was applied and the level of significance was fixed to.017 (α/3 comparisons).
Means and SDs of 3D scapular attitudes at rest were calculated and paired t tests were conducted to verify significant differences between symptomatic and asymptomatic shoulders with a level of significance adjusted for multiple tests to.017.
Means and SDs of 3D scapular attitudes and scapular total ROM of both shoulders in arm elevation were also calculated. To identify whether there were significant differences in 3D scapular attitudes and scapular total ROM values between shoulders (SHimp, SHctl), positions (70°, 90°, 110°), and movements (flexion, abduction), as well as to see whether there were significant statistical interactions between these factors 3-way repeated-measures ANOVA (2×3×2 ANOVA) with Newman-Keuls post hoc comparison analyses were used.
Descriptive data were used to compare SHimp, SHctl, and SHhe. Comparisons of data at 90° of flexion and abduction between these shoulders were done by superimposing the 3D scapular attitudes of both SHimp and SHctl to the 99% confidence interval (CI) of the mean 3D scapular attitudes of SHhe. The 90° angle in a seated position was chosen because it represents a condition of maximal mechanical demand for the shoulder, the angle at which several impingement signs are performed, and because it has been shown to be a discriminant angle for AHD between symptomatic and asymptomatic shoulders (unpublished data).
Scapular behavior for the subjects with SIS were also classified into 3 subgroups (lead, symmetrical, lag) based on the difference in magnitude of anterior tilting between both shoulders at 90° of shoulder elevation. This latter analysis was justified by previous studies4, 31 as well as by a work in which we found that the difference in anterior tilting between both shoulders at 90° of arm elevation helped explain the level of disability of subjects with SIS (unpublished data). A similar method of classification based on the visual difference in scapular external rotation between the affected and nonaffected shoulders was also used in patients who had stroke.37
In our study, each SIS subject was classified into the categories of lead, symmetrical, and lag depending on whether the difference in anterior tilting between SHimp and SHctl was superior, equal, or inferior, respectively, to 1 SD of the mean anterior tilting amplitude of SHhe at 90° of arm elevation. Because small groups were compared, a normal distribution could not be assumed, and anterior tilting was converted into ranks to normalize the data for subsequent analyses. Thereafter, in each SIS subgroup, comparisons in anterior tilting between shoulders (SHimp, SHctl) and movements (flexion, abduction) were performed using a 2-way ANOVA applied to the ranks of the data instead of to the data themselves.38 All analyses were conducted with the SPSS software, version 7.5.b
Results
Contribution of each scapular rotation to the scapular total ROM
Figure 2 shows the relative contribution, expressed in percent, of each scapular rotation to the scapular total ROM (100%) with its corresponding rp value.
Fig. 2
Contribution of each scapular rotation of shoulders with impingement syndrome (SHimp, n = 41) and contralateral asymptomatic shoulders (SHctl, n = 29) as a percentage of the scapular total ROM in flexion and abduction. Each bar represents the scapular total ROM (100%) as indicated by the arrow. Pearson's product- moment correlation (rp) between scapular rotations and the scapular total ROM are indicated within the corresponding bar section and significant correlations are labeled with an asterisk.* P <.017; ** P <.001.
In both shoulders, in the critical zone of impingement at 90° and 110° of arm elevation, about half of the scapular total ROM in flexion was provided by anterior tilting (SHimp, 48.9%, 49.4%; SHctl, 48.2%, 51.3%), whereas, in abduction, the major contributor to the total scapular motion was external rotation (SHimp, 40.3%, 43.3%; SHctl, 41.3%, 42.4%). Accordingly, the highest rp values were observed in anterior tilting in flexion (SHimp, rp =.73-.86, P <.017; SHctl, rp=.77-.88, P <.017) and in external rotation in abduction (SHimp, rp =.76-.86, P <.017; SHctl, rp =.89-.91, P <.017).
Comparisons of scapular attitudes of both shoulders in subjects with SIS
At rest, scapular attitudes between SHimp and SHctl did not differ statistically (P >.06; fig 3).
Fig. 3
Mean ± 1 SD scapular attitudes at rest of SHimp and SHctl of 29 subjects. Paired t tests were conducted to verify significant differences between SHimp and SHctl for each scapular rotation at rest.
During arm elevation, no significant differences were found between the shoulders of subjects with SIS for scapular total ROM, anterior tilting, and external rotation (table 1). A significant difference was found in transverse rotation (P =.0002). Post hoc comparisons showed a higher amplitude of transverse rotation at 110° in flexion (mean difference, 2.0°; P <.01) in SHimp (12.0°) than in SHctl (10.0°). As shown in figure 4, a progressive increase in scapular ROM was observed as the arm elevation increased from 70° to 110° both in flexion and abduction.
Fig. 4
3D scapular attitudes of SHimp (n = 41) and SHctl (n = 29) at different flexion and abduction positions. Each bar graph indicates the mean scapular rotation measured and 1 SD (T bar). The large translucent rectangle indicates the scapular total ROM. Mean scapular total ROM values and SDs are indicated on the top of this rectangle. An asterisk indicates a significant difference between SHimp and SHctl (P <.05).
| 3D Scapular Attitudes | ||||
|---|---|---|---|---|
| Factors | Scapular Total ROM | Tilting | External Rotation | Transverse Rotation |
| Main shoulder effect | .58 | 0.3 | .31 | .0002 |
| Main position effect | <.0001 | <.0001 | <.0001 | <.0001 |
| Main movement effect | <.0001 | <.01 | <.0001 | <.0001 |
| Shoulder × position interaction | .19 | .11 | .83 | .79 |
| Shoulder × movement interaction | .06 | 0 | .98 | .87 |
| Position × movement interaction | .02 | <.0001 | <.0001 | .001 |
| Shoulder × position × movement interaction | .92 | .37 | .61 | 0.9 |
| * SHimp and SHctl shoulders of subjects with SIS. † Three shoulder positions: 70°, 90°, and 110°. ‡ Two shoulder movements: flexion and abduction. | ||||
The plane of arm elevation also had an impact on scapular attitudes. As reported in table 1, a significant difference in all components of 3D scapular attitudes and total ROM between flexion and abduction was observed (main movement effect,P <.01). Comparable profiles of change of scapular attitudes and scapular total ROM across positions were observed for both shoulders. This was supported by the fact that shoulder by position interactions were not significant. However, the profiles of change in anterior tilting differed across movements for the 2 shoulder groups, as shown by a significant shoulder by movement interaction (P =.0002), whereas a tendency for scapular total ROM was observed (P =.06). Finally, scapular total ROM and 3D scapular attitudes showed different profiles of change between positions across movements (shoulder position by movement interaction).
Comparisons of scapular attitudes between SIS and healthy subjects
Figure 5 compares scapular rotations (at 90° of flexion and abduction) of the 29 SIS subjects who had both shoulders assessed to the 99% CI of the mean scapular attitude of the shoulders of the 10 healthy subjects.
Fig. 5
Scapular rotations at 90° of flexion and abduction of SHimp (●) and SHctl (□) of 29 subjects with SIS who were bilaterally assessed. The grey band indicates the 99% CI of the mean scapular rotation of 10 SHhe. The number of SHimp and SHctl located above, within, and below the 99% CI of SHhe is indicated in each condition.
Classification of subjects with SIS in subgroups
Figure 6 shows the classification of subjects with SIS into 3 subgroups: lead, symmetrical, and lag, based on the magnitude of the difference in APT between SHimp and SHctl at 90° of flexion and abduction.
Fig. 6
Classification of subjects with a SIS as lead, symmetrical, or lag depending on whether the difference in anterior tilting between the SHimp (●) and SHctl (□) was superior, equal, or inferior, respectively, to 1 SD of the mean anterior tilting of SHhe at 90° of arm elevation. Subject numbers are indicated on the horizontal axis. The grey band indicates the 99% CI of the mean amplitude of anterior tilting of SHhe.
Discussion
This study highlighted 3 important aspects of the scapular behavior of persons with SIS. First, the contribution of each individual scapular rotation to the scapular total ROM differed according to the arm movement performed. The main contributor to flexion was anterior tilting (sagittal plane), whereas it was external rotation (frontal plane) in abduction. This finding reinforces the importance of considering task requirements such as the plane of arm elevation when assessing the impact of daily activities in persons with SIS and when comparing results from different studies.
Second, when considering the whole group of 29 subjects with a confirmed unilateral SIS, no significant differences in 3D scapular attitudes were found between the symptomatic and contralateral asymptomatic shoulders. However, when looking at individual behaviors, a large proportion of both symptomatic and asymptomatic shoulders had scapular rotation magnitudes outside the ROM of healthy subjects. Therefore, both shoulders of SIS subjects had a similar behavior that was different from healthy subjects. This raises the question as to the extent to which abnormal scapular kinematics precede clinical signs and symptoms of SIS.
Third, when looking thoroughly at individual scapular behaviors, it was found that tilting asymmetry at the critical amplitude for shoulder impingement of 90° can be used to characterize the scapular behavior of subjects with unilateral SIS into 3 subgroups. This classification contributed to highlight differences that were hindered by group comparisons. This last finding supports the idea that scapular measurements in the sagittal plane are useful to categorize subjects with SIS into different classes of disability.
Our results about the main contributors to scapular total ROM in flexion and abduction confirm our previous findings obtained in a smaller group of SIS subjects.35 To our knowledge, no other studies have characterized the contribution of scapular rotations to a total scapular ROM within the flexion and abduction planes in subjects with SIS. In studies with healthy subjects, there is no consensus as to the most important contributor to the total ROM of the scapula during arm flexion.23, 26, 27 Posterior transverse rotation (referred to as “spinal tilt”),26 external rotation,23 and anterior tilting27 have all been identified as the major contributors to scapular total ROM in flexion. In abduction, the pattern of a progressive increase in external rotation and anterior tilting is in agreement with other studies performed with healthy22, 23, 24, 25, 26and SIS subjects.31 A maximal rotation of the scapula must be achieved to maintain an optimal contact between the glenoid fossa and the humeral head.39 For abduction tasks, it is therefore not surprising to observe that, as reported in studies involving healthy subjects, external rotation (also referred to as “latero-rotation,” “upward rotation,” “lateral rotation”) is responsible for most of the ROM of the scapula.22, 23, 24, 25, 26, 27
One unexpected finding in our study was the absence of significant differences in 3D scapular attitudes between symptomatic and asymptomatic shoulders of subjects with SIS when using group analyses. In fact, only a small significant difference of 2° in transverse rotation was found at 110° of arm flexion. No such results were previously reported. Previous results rather suggested that SIS may be caused by decreased scapular external rotation28, 29, 30 and anterior tilting.4, 31Our study showed no significant differences in external rotation for which the largest mean difference between shoulders was 1.4°. However, although no trend analysis was performed, we observed that, in abduction, SHimp showed less anterior tilting than SHctl at all arm positions. Mean scapular anterior tilting in SHimp was 9.5° ± 3.7°, 14.1° ± 4.5°, and 20° ± 6.1° compared with 11.5° ± 4.5°, 17.7° ± 5.6°, and 25.4° ± 6.4° in SHctl at 70°, 90°, and 110° of abduction, respectively. These results agree with previous work showing that subjects with shoulder impingement during elevation of the arm in the scapular plane had a lower anterior tilting angle (referred to as “posterior tilting” by others) than nonimpaired subjects31 with similar magnitudes to ours.
One explanation for the absence of significant differences in 3D scapular attitudes could be methodologic, such as the presence of an asymmetry in posture at rest. In our calculation method, 3D scapular attitudes were determined using the posture at rest as the reference point. Thus, it is essential to show that resting posture was symmetrical before assuming that similar 3D scapular attitudes obtained in arm elevation had not been calculated from different reference points between both shoulders of SIS. As seen in figure 3, only small differences were observed at rest between SHimp and SHctl, and 3D scapular attitudes were not significantly different. Therefore, the lack of statistical 3D scapular attitude differences between both shoulders of SIS subjects during arm elevation cannot be due to a posture asymmetry at rest. This agrees with another study reporting no statistical differences in scapular protraction, midthoracic curvature, and posture at rest within and between subjects with shoulder overuse injuries and healthy subjects.18
Another explanation for such a result could be poor posture during arm elevation, which affects the kinematic data of both shoulders.8, 40 Kebaetse et al8 reported decreased scapular upward rotation (called “external rotation” in our study) and posterior tilting (called “anterior tilting” in our study) in the slouched position. Culham and Peat40 also showed that as the slope of the upper thoracic axis increased, the scapula exhibited more anterior tilt (called “posterior tilt” in our study). Thus, even though subjects with SIS had a symmetrical posture at rest, this posture may be further altered and restrict the degrees of freedom of the scapula during arm motion. One may hypothesize that such postural abnormalities combined with other factors such as dominance, differences in bone morphology, and ergonomic exposures may increase the risk of developing SIS.
A third explanation for similar 3D scapular attitudes between shoulders of unilateral SIS subjects could be perturbations in the neuromuscular control. In fact, bilateral abnormalities in scapular kinematics were found in SIS shoulders (symptomatic, asymptomatic), as compared with shoulders of healthy subjects. From these results, we speculate that inappropriate neuromuscular strategies affecting both shoulders might have been used. Indeed, it was shown that during active arm elevation in the scapular plane, both the symptomatic and asymptomatic shoulders of patients with SIS had a superior scapular position compared with nonimpaired subjects.31 Scapular movements are accomplished by a number of muscles attached to the shoulder complex (humerus, scapula, clavicula, thorax), neck, and trunk. Synergist and antagonist patterns of muscles linking these bones are complex, and muscles in the shoulder and neck region may be organized functionally in such a way that activation on the symptomatic side also affects the contralateral side.41 There is evidence that unilateral arm activity increases the tension level of shoulder and neck muscles on the other side of the body,41 and contralateral coactivation has been shown during maximal activity in shoulder and neck muscles.42 One way to confirm this assumption would have been to record, during unilateral shoulder elevation, 3D scapular attitudes of the contralateral scapula and bilateral muscle activation using electromyography.
In this study, it was expected that 3D scapular attitudes at 90° of arm elevation would have had the capacity to discriminate symptomatic shoulders from asymptomatic and healthy ones because, in a previous study, we showed that the AHD measured by MRI at this specific arm position was sensitive enough to allow such a discrimination (unpublished data). However, as shown in figure 5, this was not the case. Scapular rotations of both shoulders were randomly distributed within and outside the 99% CI of healthy shoulders, and no specific 3D scapular attitudes patterns for each shoulder emerged from this analysis. On the other hand, data dispersion of SHimp and SHctl indicates that an important between-subject variability characterizes these shoulders even in the SHctl in which no clinical signs and symptoms were found. Therefore, the major conclusion of this analysis was that different strategies were used by subjects with SIS. Classifying SIS subjects into subgroups may allow better comparisons between healthy and pathologic shoulders because, as we observed, opposite scapular behaviors may hinder scapular kinematic abnormalities in SIS when only group analyses are performed.
The results of our study showed the nonhomogeneity of our subjects in terms of scapula behaviors according to the direction of scapular tilting asymmetry. Classification of subjects as lead, symmetrical, or lag showed that scapular measurements in the sagittal plane were useful in categorizing SIS subjects. Moreover, we have found that all subjects in the lag subgroup had a high level of disability (>40%) as measured with the Shoulder Pain and Disability Index43 (SPADI, unpublished data). In fact, the mean SPADI score for subjects classified with a lag in flexion was 43.3% (41.4%-45.8%), and it was 49% (41.4%-67.9%) for those classified with a lag in abduction. In contrast, the lead and symmetrical subgroups had a mean SPADI score that did not exceed 40% (lead: 25.4% in flexion, 20.2% in abduction; symmetrical: 37.2% in flexion, 35.1% in abduction). Lag subjects were found to have less anterior tilting in the SHimp than in the SHctl. A lack of anterior tilting of the angulus inferior during arm elevation may keep the anterior undersurface of the acromion at a close distance from the greater tuberosity of the humerus, which may contribute to produce a subacromial shoulder impingement in lag subjects.
In light of these results, magnitude and direction of anterior tilting asymmetry of the scapula during abduction of the arm appear to be indicative of either a proper or an undesirable compensation, which has an impact on the level of disability. Specific efforts should be made to identify subjects at risk of SIS who show a lack of anterior tilting and a high level of disability. Indeed, several factors may predispose a subject to a lack of scapular anterior tilting, such as increase slouched posture,8 posterior capsule tightness,44 excessively tight pectoralis minor muscle,31 or improper control of the serratus anterior.3, 45 Immediate rehabilitation directed at specifically treating such impairments in SIS subjects may lead to a better functional outcome.
Our results are restricted to static conditions and cannot be generalized for dynamic activities. Because no electromyographic data were recorded, the interpretation of scapular kinematics with respect to neuromuscular control was restricted. The lack of information about abnormal superior translation of the humeral head, specifically in the symptomatic shoulder, does not allow us to draw any conclusion about the impact of this parameter with regard to excessive narrowing of the subacromial space during arm elevation. Future studies should verify to what extent the asymmetry in anterior tilting between both shoulders of SIS subjects is a good indicator of their functional performance. Simple clinical indicators highly correlated with the complex 3D measurements that we used should be developed to allow clinicians to assess scapular kinematic behavior accurately.
Conclusion
Abnormal scapular kinematics associated with SIS is difficult to compare between studies unless the same task requirements, such as identical plane of arm elevation, are defined. When referring to group analyses, 3D scapular attitudes in symptomatic and asymptomatic shoulders of subjects with unilateral SIS are similar, and both shoulders of subjects with SIS differ from the shoulders of healthy subjects. This finding suggests that alterations in 3D scapular attitudes may be a causal factor to SIS because it precedes clinical signs and symptoms in asymptomatic shoulders. When subgroup analyses are performed, different scapular behaviors in the sagittal plane are observed. Indeed, scapular tilting asymmetry may be a good indicator of SIS severity. SIS subjects showing less anterior tilting in the SHimp as compared with the SHctl shoulder have a poor shoulder function and may therefore be at high risk for developing chronic SIS. Although further investigation is required to verify this assumption, present data provide scientific evidence to focus rehabilitation protocols toward a restoration of the control of the scapula in the sagittal plane, especially restoring anterior tilting.
Acknowledgements
We thank R. Lirette, MD, an orthopedic surgeon at the CHUQ–Laval University, for his expertise and active collaboration with the selection of subjects, and Guy St-Vincent, an engineer, for his important contribution to both the development of the 3D analysis and data collection. We also thank Johanne Tardif, PT, from the CHUQ–Laval University for her collaboration in the standardization of the clinical tests used.
Nenhum comentário:
Postar um comentário