- Email: [email protected]
Shoulder complex position and glenohumeral subluxation in hemiplegia
857
Shoulder Complex Position and Glenohumeral Subluxation in Hemiplegia Elsie G. Culham, PhD, Renata R. Noce, MSc, Stephen D. Bagg, MD ABSTRACT. Cuiham EG, Noce RR, Bagg SD. Shoulder complex position and glenohumeral subluxation in hemiplegia. Arch Phys Med Rehabil 1995;76:857-64. • Objectives: 1. To determine if scapular and humeral orientation differed between the affected and nonaffected side in two groups of hemiplegic subjects (low tone and high tone). 2. To determine if there was a relationship between these measures and glenohumeral subluxation in either group. Design: Retrospective case-comparison study. Subjects, Setting: Thirty-four hemiplegic subjects, 41 to 89 years of age, participated in the study. Subjects were divided into high-tone (n = 17) and low-tone (n = 17) groups on the basis of Ashworth scoring of muscle tone. Outcome Measures: Linear and angular measures of scapular and humeral orientation were calculated from tridimensional coordinates of bony landmarks collected using an electromagnetic device with subjects in a seated position with arms relaxed by their side. Glenohumeral subluxation was measured from radiographs. Results: The scapula was further from the midline and lower on the thorax on the affected side in the low-tone group (p < .05). Glenohumeral subluxation was greater in the low-tone group (p < .05). The scapular abduction angle (ScAb) was significantly greater on the nonaffected in the low-tone group compared with both the affected side in this group and to the nonaffected side in the high-tone group. In the high-tone group, n o differences were found between the affected and nonaffected side in either the angular or linear measures. There was no significant correlation between scapular or humeral orientation and glenohumeral subluxation in either group (p > .05). Conclusions: This study provided little evidence of a consistent pattern of alteration in shoulder complex orientation, particularly in subjects with increased muscle tone, and no support for the concept of a relationship between scapular and humeral orientation and glenohumeral subluxation.
© 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation Disorders of the shoulder complex are common after cerebrovascular accident (CVA), often leading to pain and impairing return of upper limb function. 1'2 Glenohumeral subluxation, subacromial impingement, shoulder-hand syndrome, brachial plexus injury, and adhesive capsulitis have all been documented in this patient population. 36 Alteration in the alignment of the skeletal components of the shoulder complex have been described in both the flaccid and spastic stages of paralysis after CVA and are believed to contribute to the development of shoulder disorders in the hemiplegic patient. Cailliet 7 and others 8'9 stated that decreased muscle tone in the flaccid stage after stroke results in depression of the scapula on the thorax and rotation of the scapula such that the glenoid fossa assumed a greater downward angulation. It has been proposed that the downward angulation of the glenoid would compromise the "locking mechanism" provided by the normal upward tilt of the glenoid~°; hence contributing to glenohumeral subluxation. Inferior subluxation and downward angulation of the glenoid were proposed to result in an increased abduction angle of the humerus, relative to the scapula, on the affected side] '9 It has also been suggested From the School of Rehabilitation Therapy (Dr. Culham, Ms. Noce) and the Department of Rehabilitation Medicine (Dr. Bagg), Faculty of Medicine, Queen's University, Kingston, Ontario, Canada. Submitted for publication September 1, 1994. Accepted in revised form April 24, 1995. Funded by the Physicians' Services Incorporated Foundation. 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 authors or upon any organization with which the authors are associated. Reprint requests to Elsie G. Culham, PhD, School of Rehabilitation Therapy, Queen's University, Kingston, Ontario, Canada K7L 3N6. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/95/7609-321953.00/0
that functional scoliosis, with the concavity toward the paretic side, contributes to the downward angulation of the glenoid and the increase in the relative abduction angle of the humerus. ~'~ Cailliet 7 reported that scapular depression and downward rotation persisted as spasticity developed because of increased tone in latissimus dorsi, rhomboid, and levator scapulae muscles. Other investigators have postulated that increased tone in the rhomboid and pectoralis minor muscles contributes to downward rotation of the scapula in the spastic stage of recovery. 9'~2 Few studies have reported objective measures of scapular and humeral orientation and related these to degree of glenohumeral subluxation in a hemiplegic population. ~3"14Prrvost and colleagues ~4 used a radiographic technique to measure the orientation of the scapula and humerus in the scapular plane and glenohumeral subluxation bilaterally in 50 subjects who had experienced a CVA. The glenoid fossa was found to be angled downward on both sides, which was consistent with findings in normal shoulders using similar measurement techniques.15'16 The downward orientation of the glenoid was found to be less on the affected side, in direct contrast to the orientation reported in the clinical literature. 7'9 The humerus was significantly more abducted on the paretic side but there was no significant difference in the abduction angle of the humerus relative to the scapula. ~4No significant relationship was found between either scapular or humeral abduction angle and the degree of glenohumeral subluxation. ~4 This lack of relationship was confirmed in a subsequent study of 40 hemiplegic subjects. ~3 Neither the stage of motor recovery nor the degree of muscle tone was reported for subjects in studies by either Prrvost ~4 or Arsenault and associates] 3 Arch Phys Med Rehabil Vol 76, September 1995
858
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham Table 1: Modified Ashworth Scale
2.5 3 4 5
Low tone Normal Slight increase in tone giving a catch when the limb was moved Slight increase in muscle tone, manifested by a catch and followed by minimal resistance throughout the remainder (less than half) of the range of movement More marked increase in tone but limb easily moved Considerable increase in tone, passive movement difficult Limb rigid
tone, resulting in 17 subjects in the low-tone group and 17 subjects in the high-tone group.
Evaluation of Shoulder Complex and Trunk Orientation
The 3-Space I s o t r a k , a a n electromagnetic device, was used to obtain measures of scapular, humeral, and trunk orientation. This device consists of a source, a sensor, and a systems electronic unit (SEU). The SEU contains the circuitry to compute and perform the calculations of the position and Glenohumeral subluxation occurs more commonly in the orientation of the sensor in the magnetic field relative to the flaccid stage of paralysis after C V A . 12'17 It is possible that source. The sensor consists of a pen-shaped stylus, which scapular and humeral orientation also differ depending on was connected to a foot-switch trigger. The stylus tip was the stage of motor recovery and degree of spasticity. This placed over the bony landmark on the subject, and the foot study was designed to address this issue through objective switch was activated to obtain the three-dimensional coordimeasurements of scapular and humeral orientation in two nates of the point relative to the source. Specific landmark locations on the subjects are shown in groups of hemiparetic subjects: a group identified as having low tone and a group with spasticity in the upper extremity figure 1 and included (1) the medial border of the scapula musculature. The orientation of the scapula and humerus in proximal to the inferior angle (IA), (2) the angle of the acrothe scapular plane were compared between the affected and mion process (AA), (3) the junction of the root of the scapular nonaffected limb and the relationship between these mea- spine and medial border of the scapula (R), (4) the spinous sures on the affected side and degree of glenohumeral sub- process of the first thoracic vertebrae (T1), (5) the center of the arm proximal to the olecranon process representing the luxation were determined for both groups of subjects. distal humerus (DH), and (6) the center of the arm 10cm above DH representing the proximal humerus (PH). METHODS The trunk may be rotated either forward or backward Subjects on the affected side in the transverse plane in hemiplegic Thirty-six subjects who had experienced a CVA partici- subjects. 8 Because rotation of the trunk could affect scapular pated in the study. Criteria for inclusion in the study were orientation, a measure of the degree of trunk rotation relative unilateral CVA, Brunnstrom recovery stage I to IV, ~s suffi- to the pelvis was obtained. Four additional points were cient cognitive ability for comprehension of instructions, needed for this calculation: two points located on the pelvis and ability to sit unsupported for 25 minutes. Subjects in 10cm lateral to the midline at the level of the second sacral Brunnstrom recovery stages V and VI were excluded from vertebrae ($2), and two points located on the trunk 8cm the study because they were considered to have reasonably lateral to the midline at the level of the spinous process of good voluntary control of shoulder musculature and there- the first thoracic vertebrae (T1). These points were located fore might not demonstrate effects of paralysis. Subjects using a ruler. A line connecting the two points was perpenwere also excluded from the study if they were medically dicular to the vertebral column at the level of T1 and $2, unstable, had a cardiac pacemaker, or had a history of previ- respectively. Bony landmarks were palpated and marked with erasable ous musculoskeletal or neurological disorders involving the marker with subjects seated on a wooden stool with their upper extremities. Potential subjects were also excluded if they had a clinically apparent scoliosis because this would arms relaxed by their sides. The same investigator deteraffect the measures of scapular orientation. The data from mined the location of the landmarks in all subjects. Once 2 subjects were not used in the analysis because of problems the landmarks were identified, data were collected with the subjects sitting in the same position on the stool. The source maintaining a stable sitting posture during data collection. was located on a height-adjustable platform and was posiEvaluation of Muscle Tone tioned behind the subject so that all landmarks were within A composite score of muscle tone was obtained using the 38cm of the source, the range of greatest accuracy of the 3Modified Ashworth Scale] 9 One additional level was added Space Isotrak (fig 2). With the source in this position, X to this scale to reflect low tone or flaccidity (table 1). The was the anteroposterior axis, Y was the mediolateral axis, shoulder internal rotators, shoulder adductors, and elbow and Z was the vertical axis. The outcome measures were calculated from the threeflexors, groups which commonly exhibit high levels of spasticity after CVA, J,20 were graded using this scale. Reliability dimensional coordinates using software specifically designed of this scale for tone measurement in the elbow flexors and for this purpose. This software has been used previously, shoulder adductor muscle groups has been documented. 21"22 and reliability and accuracy of the outcome measures have The grades of the three muscles were added to obtain a been documented. 23 Six outcome measures were obtained spasticity score, which was used to allocate subjects to a from the coordinates of the bony landmarks, including angulow-tone or high-tone group. Subjects with a total score of lax measures of scapular and humeral orientation, linear mealess than or equal to three were classified as low tone; sub- sures of scapular position on the thorax, and a measure of jects with scores higher than three were classified as high trunk rotation. Arch Phys Med Rehabil Vol 76, September 1995
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
Fig 1 m B o n y landmark locations digitized with the 3Space Isotrak to obtain shoulder complex orientation measures. T1, first thoracic vertebrae; $2, second sacral vertebrae; R, root of the spine of the scapula; IA, inferior angle of the scapula; AA, angle of the acromion process; PH and DH, proximal and distal landmarks located along the long axis of the arm.
Angular measures of scapular and humeral orientation. Before calculation of scapular and humeral abduction angles, the coordinates were rotated around the Z axis by the computer software, so that the spine of the scapula (a line connecting the root of the scapular spine and the angle of the acromion) was parallel to the Y axis of the source in the transverse plane. Thus, the scapular and humeral angles represented orientation of these skeletal structures in the plane of the scapular spine (yz plane) rather than in the true coronal plane. The scapular abduction angle (ScAb) was measured as the lateral angle formed between the medial border of the scapula and the horizontal (Y) axis of the source in the yz plane. The humeral abduction angle (HAb) was represented by the angle that the long axis of the humerus formed with the vertical (Z) axis of the source (fig 3). The abduction angle of the humerus relative to the medial border of the scapula (HRel) was also calculated, a larger angle indicating greater abduction of the humerus relative to the scapula.
859
Linear measures of scapular position. For calculation of linear measures, coordinates were rotated around the Z axis such that a line connecting the two points located 8cm lateral to the midline at the level of T1 was parallel to the Y axis of the Isotrak source. This line, which was perpendicular to the vertebral column, was used to define a coronal axis through the trunk. Thus, the linear measures represented the vertical (Z) and horizontal (Y) position of the scapula on the thorax in the coronal plane (yz plane) as defined by this axis. The horizontal orientation of the scapula on the thorax (Yabs) was measured as the linear distance in centimeters from T1 to the midpoint of the three scapular landmarks, determined by computer software, along the Y axis (fig 4). The vertical orientation of the scapula on the thorax (Zabs) was measured as the linear distance from T1 to the midpoint of the three scapular landmarks along the Z axis (fig 4). Trunk rotation. The lines connecting the two points of the trunk at the level of T1 and the two points on the pelvis at the level of $2 were perpendicular to the vertebral column and represented coronal axes through the trunk and pelvis, respectively. The angle that each of these lines formed with the Y axis of the source in the transverse plane was calculated. To eliminate problems of subject positioning, the pelvic angle was subtracted from the trunk angle, resulting in a measure of the degree of trunk rotation relative to the pelvis in the transverse plane. A positive value indicated forward rotation of the trunk on the affected side. Because measurement of trunk rotation had not been previously obtained using the Isotrak, the intrarater reliability of the measures obtained was determined on 10 healthy subjects. The bony landmarks were palpated, marked, and digitized using the 3-Space Isotrak with subjects positioned as previously described. Data were collected on two occasions 2 days apart. Less than 5 ° of trunk rotation were recorded for all subjects on both occasions. The interclass correlation coefficient for repeated measures was .914. Evaluation of Inferior Humeral Subluxation
A single-phase x-ray machine was used to obtain radiographs of the shoulder complex bilaterally. The radiographs were taken with the subjects' shoulders at a 45 ° angle to the
LA CO MPPTUOTPER
SENSOR~' OIT OCTH SFW STOOL
~
Fig 2--Schematic representation of the Isotrak components and the experimental set up for the measurements of shoulder complex orientation. Arch Phys Med Rehabil Voi 76, September 1995
860
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
I
T1 -~__
___~
0 ~::....
R
Yabs
.-
i
''--.
• r~
;:,i,~:.'iii::'..:~:i:". " . -[ ' :::%::= . . . . '
oJ .,..:.
..... ::.:i:
I
t~
: ~%1:~!!!i~i ~,
c:
D
,
•- :i.~t:~:i:.'~?
HAb IA
Y -axis source Fig 4mLinear measures of scapular position. Yabs, linear distance in centimeters from T1 to the midpoint of the three scapular landmarks along the y axis; Zabs, linear distance in centimeters from T1 to the midpoint of the three scapular landmarks along the z axis.
ScAb
.~.
~
Fig 3--Angular measures of shoulder complex orientation in the plane of the scapular spine. R, root of the scapular spine; IA, inferior angle of the scapula; ScAb, scapular abduction angle; PH and DH, proximal and distal landmarks on the long axis of the arm; HAb, humeral abduction angle. plane of the roentgenogram to provide a more accurate profile of the glenohumeral joint. 14 The subluxation measure was obtained by determining the vertical distance between the center of the head of the humerus and the center of the glenoid fossa (fig 5). The films were placed over a light table. A tracing paper with concentric circles 2mm apart was placed over the film and the humeral head center was marked. The midpoint of the glenoid fossa was located using a ruler. A high degree of correlation (r = .931) between this measure and a three-dimensional measure of subluxation was reported by Pr6vost. 24
Arch
Phys
Med
Rehabil
Vol 76, September
1995
Fig 5--Radiographic measure of glenohumeral subluxation; a, center of the humeral head; b, center of the glenoid fossa.
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
A n"
B
861
8] 7
E
E 4
-
-
¢ 3
-
-
-
1
o
~
2
z~
3
4
s
0
~
2
z5
3
4
5
Fig 6--Histogram showing distribution of Ashworth scores for individual muscle groups in subjects in (A) the low-tone group (n = 17) and (B) the high-tone group (n = 17). IN, shoulder internal rotators; IN, elbow flexors; VT,shoulder adductors. This method was chosen because it is not dependent on anthropometric variation and requires x-ray of only the affected shoulder. 24
Statistical Analysis Dependent student t tests were used to determine whether significant differences existed in scapular and humeral orientation measures between the affected and nonaffected side within groups. Independent t tests were used to compare measures of scapular and humeral orientation on the affected and nonaffected limbs between groups. Pearson correlation coefficients were calculated to assess the relationship between subluxation and scapular and humeral abduction angle within groups. Systat h statistical software was used to analyze the data. All data were analyzed at the .05 level of significance. RESULTS Subjects The mean age of the 17 subjects in the low-tone group was 72.4 years. Eleven subjects were men and 14 were inpatients at the time of testing. The mean time since onset of CVA was 8.7 _ 14.4 months. The high-tone group consisted of 17 subjects with a mean age of 61.3 years and a mean time since onset of CVA of 22.8 __ 23.7 months. Thirteen were men and 4 were inpatients at time of testing. The mean total Ashworth scores for subjects in the low- and high-tone groups were 1.1 _+ 1.3 and 10.3 __ 2.9, respectively. The difference between groups was significant (p < .05; MannWhitney U test). Histograms of Ashworth scores obtained for each muscle group for subjects in each group are provided in figure 6. Figure 7 illustrates the distribution of total tone scores for subjects in each group. Low-Tone Group Results of the scapular and humeral orientation measures of the low-tone group are presented in table 2. The scapula was further from the midline on the affected side, as indicated by the larger value for Yabs (p < .05). The significantly larger value for Zabs indicated a lower position of the scapula on the thorax on the hemiplegic side. The 95% confidence intervals around the difference between the affected and nonaffected side for Zabs and Yabs were .19 to 1.735cm and .05 to
1.29cm, respectively. The scapular abduction angle (ScAb) on the nonaffected side was significantly greater compared to the affected side and to the nonaffected limb of the high-tone group. No difference was found in the humeral abduction angle (HAb) or in the angle of humeral abduction relative to the scapula (HRel). The mean trunk rotation value for subjects in this group was - 0 . 1 ° + 3.1 °.
High-Tone Group Results of the scapular and humeral orientation measures in the high tone group are presented in Table 3. No differences were found between the affected and nonaffected side in either the angular or linear measures of scapular and humeral orientation. The mean trunk rotation angle was - 1 . 7 ° + 4.3 °. Three of the 17 subjects had greater than 5 ° of trunk rotation; the trunk was rotated posteriorly on the affected side relative to the pelvis. High- Versus Low-Tone Group No significant differences were found in the angular measures of shoulder complex position between the affected limbs of the high- and low-tone subjects. Comparison of the same measures on the nonaffected side between groups 17 16 15 14 13 12 11
10
,.=
E
9
0 1 22.53' ~ Ashworth total score
~'~ ~ 10 11 t 9 . ~ ' -= ~o 14 15
Fig 7 - - I - I i s t o g r a m demonstrating the distribution of total tone score for subjects in the low- and high-tone groups, m, low tone (n : 17); Fq, high tone (n = 17).
Arch Phys Med Rehabil Vol
76, S e p t e m b e r 1995
862
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham Table 4: Association Between Subluxation and Scapular and Humeral Orientation Measures
Table 2: Scapular and Humeral Orientation, Low-Tone Group (n = 17) Variable ScAb HAb HRel Yabs (cm) Zabs (cm)
Affected Side 86.4 ° 8.7 ° 12.3 ° 12.0 9.2
(+5.5) (-+5.9) (_+7.7) (_+1.5) (+_1.7)
Nonaffected Side 91.1 ° 10.8 ° 9.7 ° 11.3 8.3
(___4.9) (+_5.7) (_+5.4) (+_1.0) (_+1.6)
p
Group
Variable
r
p
.003* .100 .126 .041" .021"
Low tone (n = 17)
ScAb HAb HRel ScAb HAb HRel
-.121 - . 145 -.026 -.221 -.089 .164
.643 .578 .921 .393 .734 .528
Values given as mean (+SD). * Significant difference at p < ,05.
r = Pearson correlation coefficient.
showed a significantly greater scapular abduction angle in the low-tone group (91.1 ° versus 86.4°; p < .05. The mean trunk rotation measure was not significantly different between the high- and low-tone group (p > .05).
Scapular Orientation and Subluxation There was a significantly greater glenohumeral subluxation on the affected side in the low-tone group than in the high-tone group. Mean values were .52 + .38cm for the lowtone group and .21 _+ .41cm for the high-tone group (p < .05). No significant relationship was identified between scapular abduction angle and glenohumeral subluxation in either the low-tone group (r = - . 121; p = 0.6) or in the hightone group (r = -.221; p = 0.4). Similarly, no significant relationship was found between either the absolute (HAb) or relative (HRel) humeral abduction angle and subluxation in either the low-tone or high-tone group. (table 4). DISCUSSION
Scapular Orientation The scapula was lower on the thorax and further from the midline on the affected side compared with the nonaffected side in the low-tone group. Decreased tone in the trapezius and rhomboid muscles in the flaccid stage after a stroke may have contributed to these findings. However, the differences between the affected and nonaffected limbs' sides were of small magnitude, suggesting that scapular position on the thorax is not greatly affected by flaccid paralysis of these muscle groups. Because the linear measures are dependent on stature, no comparison of these measures between the high- and low-tone groups was performed. The scapular abduction angle was greater on the nonaffected side in the low-tone group compared with the affected side. This value was also greater than the angle measured on either side in the high-tone group. Thus, there is little evidence from this study that low tone in the scapular musculature in the affected limb contributes to significant scapular Table 3: Scapular and Humeral Orientation, High-Tone Group (n = 17) Variable ScAb HAb HRel Yabs(cm) Zabs (cm)
Affected Side 87.0 ° 10.7 ° 13.7 ° 12.3 8.9
(_+6.5) (_+4.7) (_+6.2) (_+1.3) (_+1.5)
High tone (n = 17)
Nonaffected Side 86.4 ° 10.4 ° 14.0 12.1 8.7
Arch Phys Meal Rehabil Vol 76, September 1995
(_+6.0) (_+5.6) (_+8.7) (_+1.5) (-4-1.1)
p .752 .875 .883 .576 .643
downward rotation on the hemiparetic side as proposed by Cailliet ~ and others. 8'9 Trunk alignment may be a possible explanation for the greater abduction angle on the nonaffected side in subjects with low tone. Bobath ~ stated that hemiplegic subjects tend to sit with more weight on the nonaffected buttock because of sensory deficits and loss of trunk control. The transference of weight to the normal hip is more apparent in subjects with low tone and is postulated to lead to lateral flexion of the trunk toward the hemiparetic side. A small shift in trunk alignment could alter scapular position on either side or both sides in hemiparetic subjects. The mean difference in scapular abduction angle between the two sides was less than 5 °. The degree of trunk asymmetry needed to cause such a small difference would be minimal and may have been undetectable on clinical examination. That no difference in scapular abduction angle was found between the affected and nonaffected sides in subjects with high tone may be related to better postural control in this group. Fourteen of the 17 subjects in the high-tone group were living in the community and had completed a rehabilitation program. The mean age of subjects in this group (61.3 years) was significantly younger than the mean age of subjects in the low-tone group (72.4 years). It is possible that subjects in the high-tone group had better sitting posture and greater ability to sit with weight equally distributed on both buttocks. The findings in the high-tone group do not support clinical descriptions available in the literature. Cailliet ~and o t h e r s 9"12 suggested that during the spastic stage the scapula is lower on the thorax and rotated downward, due to increased tone in the rhomboid muscles, which overwhelm the scapular elevators (trapezius and serratus anterior muscles). Increased tone in the latissimus dorsi and pectoralis minor was also postulated to contribute to the decreased abduction angle of the scapula. It is recognized that an increase in tone in the upper extremity musculature, as measured in this study, does not necessarily indicate an increase in tone in the scapular retractor or pectoralis minor muscles. Because no difference was found in scapular orientation between the affected and nonaffected sides, it would appear that if tone was increased in the rhomboids and pectoralis minor muscles, a downward rotation of the scapula on the affected side did not result, contrary to the suggestions in the literature. The findings in the current study concerning scapular orientation in the low- and high-tone groups are not consistent with those reported previously. Prtvost j4 found that the mean scapular abduction angle in 50 hemiplegic subjects was significantly greater on the affected side than on the nonaffected
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
863
side (mean difference 3.5°).14 Similarly, Arsenault j3 reported a significantly greater scapular abduction angle on the affected side in hemiplegic subjects without glenohumeral subluxation. The authors stated, however, that the difference was small (4 °) and deemed it clinically unimportant. Neither the degree of muscle tone nor the stage of recovery was documented, making comparison difficult.
either group. Thus, no support is provided for the concept that downward rotation of the scapula contributes to glenohumeral subluxation as postulated by Cailliet ~ and Davies. 9 These results are in agreement with those reported by Prdv o s t j4 and Arsenault. ~3 The authors in both of these studies concluded that there was no relationship between scapular abduction angle and degree of glenohumeral subluxation. Cailliet ~ stated that the humeral abduction angle increases as the humerus subluxes on the glenoid fossa. However, no Humeral Orientation The absolute humeral abduction angle (HAb) and the hu- correlation was found between humeral abduction angle and meral abduction angle relative to the scapula (HRel) were not glenohumeral subluxation in either group in the current significantly different between the affected and nonaffected study. Prtvost ~3 reported a significant correlation between sides in either the low-tone or high-tone groups. Cailliet ~ both humeral abduction angle and humeral abduction angle and Davies 9 stated that with flaccid paralysis, the humeral relative to the scapula and glenohumeral subluxation. Howabduction angle increases relative to the downwardly rotated ever, the correlations explained only 6% and 9% of the scapula. The relative humeral abduction angle was greater variability in the data, respectively. Arsenault ]3 found a sigon the affected side (12.3 ° versus 9.7 °) in the low-tone group nificantly greater humeral abduction angle on the affected in the current study, but the difference was not significant. side compared with the nonaffected side in a group of hemiPr6vost 14 reported a small (2.3 °) but significantly greater plegic subjects with glenohumeral subluxation. However, humeral abduction angle on the affected side compared with the mean difference was only 3 ° and the authors concluded the nonaffected side in hemiplegic subjects. Similarly, Arse- that there was no systematic relationship between humeral nault ~3 found a greater humeral abduction angle on the af- abduction angle and glenohumeral subluxation. fected side compared with the nonaffected side in a group CONCLUSIONS of hemiplegic subjects with subluxation. However, the magDownward rotation and depression of the scapula on the nitude of the differences was small (3 °) and deemed clinithorax and humeral abduction relative to the scapula have cally unimportant by the authors.~3 been described in both the flaccid and spastic stages of recovery after cerebrovascular accident. It has been postulated that Trunk Rotation these alterations in alignment contribute to glenohumeral In the present study, no pattern of trunk rotation was subluxation and possibly to pain and functional impairment observed in either the low- or high-tone group. Of the 17 in the upper extremity. Rehabilitation efforts are often disubjects in the high-tone group, 14 had less than 5 ° of rota- rected toward restoration of correct alignment of the scapula tion of the trunk relative to the pelvis, within the range found on the thorax through stretching of soft tissues, and decreasin the healthy subjects. However, three subjects demon- ing tone and increasing strength of weak musculature. The strated greater than 5 ° of posterior trunk rotation of the af- results of this study indicate that the scapula on the affected fected side ( - 12.88°; -8.19°; -5.98°). Ryerson and Levit 8 side was lower on the thorax and further from the midline described a hemiplegic posture in the spastic stage in which in hemiplegic subjects with low tone. Low tone in the perithe trunk is rotated posteriorly on the affected side. The scapular musculature may contribute to the lower position scapula was described as downwardly rotated and elevated on the thorax and increased distance from the midline. The on the thorax in subjects with this trunk posture. Examina- scapular abduction angle was greater on the nonaffected side tion of individual data, in the current study, found no consis- of the low-tone group compared with the affected and to the tent pattern of scapular orientation in the subjects with poste- nonaffected side in the high-tone group. Unequal weight rior rotation of the trunk on the affected side. distribution on the buttock resulting in asymmetry in trunk alignment is a possible explanation for the difference in Glenohumeral Subluxation scapular abduction angle. The differences between the sides There was a significantly greater glenohumeral subluxa- for all measures was small, albeit statistically significant. tion on the affected side in the low-tone group compared to The importance of the differences in terms of clinical manthe high-tone group. This finding is consistent with reports agement is questionable. No differences were found in scapin the literature. 17'25 Basmajian and Bazant j° postulated that ular or humeral orientation between the affected and nonafin the flaccid stage the loss of tone in the supraspinatus fected side in the subjects with increased tone. Glenohumeral muscle allows the humeral head to sublux inferiorly. Paraly- subluxation was significantly greater on the affected side in sis of this muscle was shown by Chaco and Wolf j7 to be an the low-tone group compared with the high-tone group. No important factor in the onset of subluxation in hemiplegia relationship was found between scapular or humeral orientaduring the flaccid stage. The same authors demonstrated that tion and glenohumeral subluxation in either group. The lack when spasticity developed, the supraspinatus muscle began of a consistent pattern of shoulder complex orientation in to respond to loading, and subluxation did not occur. A either group indicates the need for detailed assessment and correlation between a decrease in subluxation and an in- individualized rehabilitation programs. crease in muscle tone has been reported. ~2'25'26 References No significant correlation between scapular abduction 1. Cailliet R. Shoulder pain. Philadelphia: Davis, 1991. angle and degree of glenohumeral subluxation was found in 2. O'Sullivan SB. Stroke. In: O'Sullivan SB, Schmitz TJ, editors. Physical Arch Phys Med Rehabil Vol 76, September 1995
864
3. 4. 5.
6. 7. 8.
9. 10. 11. 12.
13.
14.
15.
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
rehabilitation: assessment and treatment. Philadelphia: Davis, 1988:335-79. Van Ouwenaller C, Laplace PM, Chantraine A. Painful shoulder in hemiplegia. Arch Phys Med Rehabil 1986;67:23-6. Hakuno A, Sashika H, Ohkawa T, Itoh R. Arthrographic findings in hemiplegic shoulders. Arch Phys Med Rehabil 1984;65:706-11. Tepperman P, Greyson ND, Hilbert L, Jimenez J, Williams JI. Reflex sympathetic dystrophy in hemiplegia. Arch Phys Med Rehabil 1984; 65:442-7. Najenson T, Pikielny SS. Malalignment of the glenohumeral joint following stroke. Ann Phys Med 1965;8:96-9. Cailliet R. The shoulder in hemiplegia. Philadelphia: Davis, 1980. Ryerson S, Levit K. The shoulder in hemiplegia. In: Donatelli R, editor. Physical therapy of the shoulder. New York: Churchill Livingstone, 1991:117-45. Davies PM. Steps to follow. A guide to the treatment of adult hemiplegia. Berlin: Springer-Verlag, 1985. Basmajian JV, Bazant FJ. Factors preventing downward dislocation of the adducted shoulder joint. J Bone Joint Surg 1959;41-A: 1182-6. Bobath B. Adult hemiplegia: evaluation and treatment. London: Williams Heinemann Medical Books, Ltd., 1990. Van Langenberghe HVK, Hogan BM. Degree of pain and grade of subluxation in the painful hemiplegic shoulder. Scand J Rehab Med 1988;20:161-6. Arsenault AB, Bilodeau M, Dutil E, Riley E. Clinical significance of the V-shaped space in the subluxed shoulder of hemiplegics. Stroke 1991; 22:867-71. Pr6vost R, Arsenault B, Dutil E, Drouin G. Rotation of the scapula and shoulder subluxation in hemiplegia. Arch Phys Med Rehabil 1987;68:786-90. Freedman L, Munro RR. Abduction of the arm in the scapular plane:
Arch Phys Med Rehabil Vol 76, September 1995
16. 17. 18. 19. 20. 21. 22.
23. 24. 25. 26.
scapular and glenohumeral movements. J Bone Joint Surg 1966;48A:1503-10. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg 1976;58-A:195-201. Chaco J, Wolf E. Subluxation of the glenohumeral joint in hemiplegia. Am J Phys Med 1971;50:139-43. Brunnstrom S. Movement therapy in hemiplegia. New York: Harper and Row, 1970. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth Scale of muscle spasticity. Phys Ther 1987;67:206-7. Sawner K, LaVigne J. Brunnstrom's movement therapy in hemiplegia. A neurophysiological approach. New York: Lippincott, 1992:1-276. Sloan RL, Sinclair W, Thompson J, Taylor S, Pentland B. Inter-rater reliability of the modified Ashworth Scale for spasticity in hemiplegic patients. Int J Rehabil Res 1992; 15:158-61. Lee K, Carson L, Kinnin E, Patterson V. The Asworth Scale: A reliable and reproducible method of measuring spasticity. J Neuro Rehab 1989;3:205-9. Culham E, Peat M. Spinal and shoulder complex posture. I: Measurement using the 3Space Isotrak. Clin Rehabil 1993;7:309-18. Pr6vost R, Arsenault AB, Dutil E, Drouin G. Shoulder subluxation in hemiplegia: A radiologic correlational study. Arch Phys Med Rehabil 1987;68:782-5. Finch L, Harvey J. Factors associated with shoulder-hand syndrome in hemiplegia: clinical survey. Physiother Canada 1983;35:145-8. Moskowitz H, Goodman CR, Smith E, Balthazar E, Mellins HZ. Hemiplegic shoulder. N Y State J Med 1969; 15:548-50.
Suppliers a. Polhemus Navigation Sciences Division, McDonnel Douglas Electronics Company, 1 Hercules Drive, PO Box 560, Colchester, VT 05446. b. Systat, Inc., 1800 Sharman Ave., Evanston, IL 60201-3793.
Shoulder Complex Position and Glenohumeral Subluxation in Hemiplegia Elsie G. Culham, PhD, Renata R. Noce, MSc, Stephen D. Bagg, MD ABSTRACT. Cuiham EG, Noce RR, Bagg SD. Shoulder complex position and glenohumeral subluxation in hemiplegia. Arch Phys Med Rehabil 1995;76:857-64. • Objectives: 1. To determine if scapular and humeral orientation differed between the affected and nonaffected side in two groups of hemiplegic subjects (low tone and high tone). 2. To determine if there was a relationship between these measures and glenohumeral subluxation in either group. Design: Retrospective case-comparison study. Subjects, Setting: Thirty-four hemiplegic subjects, 41 to 89 years of age, participated in the study. Subjects were divided into high-tone (n = 17) and low-tone (n = 17) groups on the basis of Ashworth scoring of muscle tone. Outcome Measures: Linear and angular measures of scapular and humeral orientation were calculated from tridimensional coordinates of bony landmarks collected using an electromagnetic device with subjects in a seated position with arms relaxed by their side. Glenohumeral subluxation was measured from radiographs. Results: The scapula was further from the midline and lower on the thorax on the affected side in the low-tone group (p < .05). Glenohumeral subluxation was greater in the low-tone group (p < .05). The scapular abduction angle (ScAb) was significantly greater on the nonaffected in the low-tone group compared with both the affected side in this group and to the nonaffected side in the high-tone group. In the high-tone group, n o differences were found between the affected and nonaffected side in either the angular or linear measures. There was no significant correlation between scapular or humeral orientation and glenohumeral subluxation in either group (p > .05). Conclusions: This study provided little evidence of a consistent pattern of alteration in shoulder complex orientation, particularly in subjects with increased muscle tone, and no support for the concept of a relationship between scapular and humeral orientation and glenohumeral subluxation.
© 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation Disorders of the shoulder complex are common after cerebrovascular accident (CVA), often leading to pain and impairing return of upper limb function. 1'2 Glenohumeral subluxation, subacromial impingement, shoulder-hand syndrome, brachial plexus injury, and adhesive capsulitis have all been documented in this patient population. 36 Alteration in the alignment of the skeletal components of the shoulder complex have been described in both the flaccid and spastic stages of paralysis after CVA and are believed to contribute to the development of shoulder disorders in the hemiplegic patient. Cailliet 7 and others 8'9 stated that decreased muscle tone in the flaccid stage after stroke results in depression of the scapula on the thorax and rotation of the scapula such that the glenoid fossa assumed a greater downward angulation. It has been proposed that the downward angulation of the glenoid would compromise the "locking mechanism" provided by the normal upward tilt of the glenoid~°; hence contributing to glenohumeral subluxation. Inferior subluxation and downward angulation of the glenoid were proposed to result in an increased abduction angle of the humerus, relative to the scapula, on the affected side] '9 It has also been suggested From the School of Rehabilitation Therapy (Dr. Culham, Ms. Noce) and the Department of Rehabilitation Medicine (Dr. Bagg), Faculty of Medicine, Queen's University, Kingston, Ontario, Canada. Submitted for publication September 1, 1994. Accepted in revised form April 24, 1995. Funded by the Physicians' Services Incorporated Foundation. 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 authors or upon any organization with which the authors are associated. Reprint requests to Elsie G. Culham, PhD, School of Rehabilitation Therapy, Queen's University, Kingston, Ontario, Canada K7L 3N6. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/95/7609-321953.00/0
that functional scoliosis, with the concavity toward the paretic side, contributes to the downward angulation of the glenoid and the increase in the relative abduction angle of the humerus. ~'~ Cailliet 7 reported that scapular depression and downward rotation persisted as spasticity developed because of increased tone in latissimus dorsi, rhomboid, and levator scapulae muscles. Other investigators have postulated that increased tone in the rhomboid and pectoralis minor muscles contributes to downward rotation of the scapula in the spastic stage of recovery. 9'~2 Few studies have reported objective measures of scapular and humeral orientation and related these to degree of glenohumeral subluxation in a hemiplegic population. ~3"14Prrvost and colleagues ~4 used a radiographic technique to measure the orientation of the scapula and humerus in the scapular plane and glenohumeral subluxation bilaterally in 50 subjects who had experienced a CVA. The glenoid fossa was found to be angled downward on both sides, which was consistent with findings in normal shoulders using similar measurement techniques.15'16 The downward orientation of the glenoid was found to be less on the affected side, in direct contrast to the orientation reported in the clinical literature. 7'9 The humerus was significantly more abducted on the paretic side but there was no significant difference in the abduction angle of the humerus relative to the scapula. ~4No significant relationship was found between either scapular or humeral abduction angle and the degree of glenohumeral subluxation. ~4 This lack of relationship was confirmed in a subsequent study of 40 hemiplegic subjects. ~3 Neither the stage of motor recovery nor the degree of muscle tone was reported for subjects in studies by either Prrvost ~4 or Arsenault and associates] 3 Arch Phys Med Rehabil Vol 76, September 1995
858
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham Table 1: Modified Ashworth Scale
2.5 3 4 5
Low tone Normal Slight increase in tone giving a catch when the limb was moved Slight increase in muscle tone, manifested by a catch and followed by minimal resistance throughout the remainder (less than half) of the range of movement More marked increase in tone but limb easily moved Considerable increase in tone, passive movement difficult Limb rigid
tone, resulting in 17 subjects in the low-tone group and 17 subjects in the high-tone group.
Evaluation of Shoulder Complex and Trunk Orientation
The 3-Space I s o t r a k , a a n electromagnetic device, was used to obtain measures of scapular, humeral, and trunk orientation. This device consists of a source, a sensor, and a systems electronic unit (SEU). The SEU contains the circuitry to compute and perform the calculations of the position and Glenohumeral subluxation occurs more commonly in the orientation of the sensor in the magnetic field relative to the flaccid stage of paralysis after C V A . 12'17 It is possible that source. The sensor consists of a pen-shaped stylus, which scapular and humeral orientation also differ depending on was connected to a foot-switch trigger. The stylus tip was the stage of motor recovery and degree of spasticity. This placed over the bony landmark on the subject, and the foot study was designed to address this issue through objective switch was activated to obtain the three-dimensional coordimeasurements of scapular and humeral orientation in two nates of the point relative to the source. Specific landmark locations on the subjects are shown in groups of hemiparetic subjects: a group identified as having low tone and a group with spasticity in the upper extremity figure 1 and included (1) the medial border of the scapula musculature. The orientation of the scapula and humerus in proximal to the inferior angle (IA), (2) the angle of the acrothe scapular plane were compared between the affected and mion process (AA), (3) the junction of the root of the scapular nonaffected limb and the relationship between these mea- spine and medial border of the scapula (R), (4) the spinous sures on the affected side and degree of glenohumeral sub- process of the first thoracic vertebrae (T1), (5) the center of the arm proximal to the olecranon process representing the luxation were determined for both groups of subjects. distal humerus (DH), and (6) the center of the arm 10cm above DH representing the proximal humerus (PH). METHODS The trunk may be rotated either forward or backward Subjects on the affected side in the transverse plane in hemiplegic Thirty-six subjects who had experienced a CVA partici- subjects. 8 Because rotation of the trunk could affect scapular pated in the study. Criteria for inclusion in the study were orientation, a measure of the degree of trunk rotation relative unilateral CVA, Brunnstrom recovery stage I to IV, ~s suffi- to the pelvis was obtained. Four additional points were cient cognitive ability for comprehension of instructions, needed for this calculation: two points located on the pelvis and ability to sit unsupported for 25 minutes. Subjects in 10cm lateral to the midline at the level of the second sacral Brunnstrom recovery stages V and VI were excluded from vertebrae ($2), and two points located on the trunk 8cm the study because they were considered to have reasonably lateral to the midline at the level of the spinous process of good voluntary control of shoulder musculature and there- the first thoracic vertebrae (T1). These points were located fore might not demonstrate effects of paralysis. Subjects using a ruler. A line connecting the two points was perpenwere also excluded from the study if they were medically dicular to the vertebral column at the level of T1 and $2, unstable, had a cardiac pacemaker, or had a history of previ- respectively. Bony landmarks were palpated and marked with erasable ous musculoskeletal or neurological disorders involving the marker with subjects seated on a wooden stool with their upper extremities. Potential subjects were also excluded if they had a clinically apparent scoliosis because this would arms relaxed by their sides. The same investigator deteraffect the measures of scapular orientation. The data from mined the location of the landmarks in all subjects. Once 2 subjects were not used in the analysis because of problems the landmarks were identified, data were collected with the subjects sitting in the same position on the stool. The source maintaining a stable sitting posture during data collection. was located on a height-adjustable platform and was posiEvaluation of Muscle Tone tioned behind the subject so that all landmarks were within A composite score of muscle tone was obtained using the 38cm of the source, the range of greatest accuracy of the 3Modified Ashworth Scale] 9 One additional level was added Space Isotrak (fig 2). With the source in this position, X to this scale to reflect low tone or flaccidity (table 1). The was the anteroposterior axis, Y was the mediolateral axis, shoulder internal rotators, shoulder adductors, and elbow and Z was the vertical axis. The outcome measures were calculated from the threeflexors, groups which commonly exhibit high levels of spasticity after CVA, J,20 were graded using this scale. Reliability dimensional coordinates using software specifically designed of this scale for tone measurement in the elbow flexors and for this purpose. This software has been used previously, shoulder adductor muscle groups has been documented. 21"22 and reliability and accuracy of the outcome measures have The grades of the three muscles were added to obtain a been documented. 23 Six outcome measures were obtained spasticity score, which was used to allocate subjects to a from the coordinates of the bony landmarks, including angulow-tone or high-tone group. Subjects with a total score of lax measures of scapular and humeral orientation, linear mealess than or equal to three were classified as low tone; sub- sures of scapular position on the thorax, and a measure of jects with scores higher than three were classified as high trunk rotation. Arch Phys Med Rehabil Vol 76, September 1995
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
Fig 1 m B o n y landmark locations digitized with the 3Space Isotrak to obtain shoulder complex orientation measures. T1, first thoracic vertebrae; $2, second sacral vertebrae; R, root of the spine of the scapula; IA, inferior angle of the scapula; AA, angle of the acromion process; PH and DH, proximal and distal landmarks located along the long axis of the arm.
Angular measures of scapular and humeral orientation. Before calculation of scapular and humeral abduction angles, the coordinates were rotated around the Z axis by the computer software, so that the spine of the scapula (a line connecting the root of the scapular spine and the angle of the acromion) was parallel to the Y axis of the source in the transverse plane. Thus, the scapular and humeral angles represented orientation of these skeletal structures in the plane of the scapular spine (yz plane) rather than in the true coronal plane. The scapular abduction angle (ScAb) was measured as the lateral angle formed between the medial border of the scapula and the horizontal (Y) axis of the source in the yz plane. The humeral abduction angle (HAb) was represented by the angle that the long axis of the humerus formed with the vertical (Z) axis of the source (fig 3). The abduction angle of the humerus relative to the medial border of the scapula (HRel) was also calculated, a larger angle indicating greater abduction of the humerus relative to the scapula.
859
Linear measures of scapular position. For calculation of linear measures, coordinates were rotated around the Z axis such that a line connecting the two points located 8cm lateral to the midline at the level of T1 was parallel to the Y axis of the Isotrak source. This line, which was perpendicular to the vertebral column, was used to define a coronal axis through the trunk. Thus, the linear measures represented the vertical (Z) and horizontal (Y) position of the scapula on the thorax in the coronal plane (yz plane) as defined by this axis. The horizontal orientation of the scapula on the thorax (Yabs) was measured as the linear distance in centimeters from T1 to the midpoint of the three scapular landmarks, determined by computer software, along the Y axis (fig 4). The vertical orientation of the scapula on the thorax (Zabs) was measured as the linear distance from T1 to the midpoint of the three scapular landmarks along the Z axis (fig 4). Trunk rotation. The lines connecting the two points of the trunk at the level of T1 and the two points on the pelvis at the level of $2 were perpendicular to the vertebral column and represented coronal axes through the trunk and pelvis, respectively. The angle that each of these lines formed with the Y axis of the source in the transverse plane was calculated. To eliminate problems of subject positioning, the pelvic angle was subtracted from the trunk angle, resulting in a measure of the degree of trunk rotation relative to the pelvis in the transverse plane. A positive value indicated forward rotation of the trunk on the affected side. Because measurement of trunk rotation had not been previously obtained using the Isotrak, the intrarater reliability of the measures obtained was determined on 10 healthy subjects. The bony landmarks were palpated, marked, and digitized using the 3-Space Isotrak with subjects positioned as previously described. Data were collected on two occasions 2 days apart. Less than 5 ° of trunk rotation were recorded for all subjects on both occasions. The interclass correlation coefficient for repeated measures was .914. Evaluation of Inferior Humeral Subluxation
A single-phase x-ray machine was used to obtain radiographs of the shoulder complex bilaterally. The radiographs were taken with the subjects' shoulders at a 45 ° angle to the
LA CO MPPTUOTPER
SENSOR~' OIT OCTH SFW STOOL
~
Fig 2--Schematic representation of the Isotrak components and the experimental set up for the measurements of shoulder complex orientation. Arch Phys Med Rehabil Voi 76, September 1995
860
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
I
T1 -~__
___~
0 ~::....
R
Yabs
.-
i
''--.
• r~
;:,i,~:.'iii::'..:~:i:". " . -[ ' :::%::= . . . . '
oJ .,..:.
..... ::.:i:
I
t~
: ~%1:~!!!i~i ~,
c:
D
,
•- :i.~t:~:i:.'~?
HAb IA
Y -axis source Fig 4mLinear measures of scapular position. Yabs, linear distance in centimeters from T1 to the midpoint of the three scapular landmarks along the y axis; Zabs, linear distance in centimeters from T1 to the midpoint of the three scapular landmarks along the z axis.
ScAb
.~.
~
Fig 3--Angular measures of shoulder complex orientation in the plane of the scapular spine. R, root of the scapular spine; IA, inferior angle of the scapula; ScAb, scapular abduction angle; PH and DH, proximal and distal landmarks on the long axis of the arm; HAb, humeral abduction angle. plane of the roentgenogram to provide a more accurate profile of the glenohumeral joint. 14 The subluxation measure was obtained by determining the vertical distance between the center of the head of the humerus and the center of the glenoid fossa (fig 5). The films were placed over a light table. A tracing paper with concentric circles 2mm apart was placed over the film and the humeral head center was marked. The midpoint of the glenoid fossa was located using a ruler. A high degree of correlation (r = .931) between this measure and a three-dimensional measure of subluxation was reported by Pr6vost. 24
Arch
Phys
Med
Rehabil
Vol 76, September
1995
Fig 5--Radiographic measure of glenohumeral subluxation; a, center of the humeral head; b, center of the glenoid fossa.
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
A n"
B
861
8] 7
E
E 4
-
-
¢ 3
-
-
-
1
o
~
2
z~
3
4
s
0
~
2
z5
3
4
5
Fig 6--Histogram showing distribution of Ashworth scores for individual muscle groups in subjects in (A) the low-tone group (n = 17) and (B) the high-tone group (n = 17). IN, shoulder internal rotators; IN, elbow flexors; VT,shoulder adductors. This method was chosen because it is not dependent on anthropometric variation and requires x-ray of only the affected shoulder. 24
Statistical Analysis Dependent student t tests were used to determine whether significant differences existed in scapular and humeral orientation measures between the affected and nonaffected side within groups. Independent t tests were used to compare measures of scapular and humeral orientation on the affected and nonaffected limbs between groups. Pearson correlation coefficients were calculated to assess the relationship between subluxation and scapular and humeral abduction angle within groups. Systat h statistical software was used to analyze the data. All data were analyzed at the .05 level of significance. RESULTS Subjects The mean age of the 17 subjects in the low-tone group was 72.4 years. Eleven subjects were men and 14 were inpatients at the time of testing. The mean time since onset of CVA was 8.7 _ 14.4 months. The high-tone group consisted of 17 subjects with a mean age of 61.3 years and a mean time since onset of CVA of 22.8 __ 23.7 months. Thirteen were men and 4 were inpatients at time of testing. The mean total Ashworth scores for subjects in the low- and high-tone groups were 1.1 _+ 1.3 and 10.3 __ 2.9, respectively. The difference between groups was significant (p < .05; MannWhitney U test). Histograms of Ashworth scores obtained for each muscle group for subjects in each group are provided in figure 6. Figure 7 illustrates the distribution of total tone scores for subjects in each group. Low-Tone Group Results of the scapular and humeral orientation measures of the low-tone group are presented in table 2. The scapula was further from the midline on the affected side, as indicated by the larger value for Yabs (p < .05). The significantly larger value for Zabs indicated a lower position of the scapula on the thorax on the hemiplegic side. The 95% confidence intervals around the difference between the affected and nonaffected side for Zabs and Yabs were .19 to 1.735cm and .05 to
1.29cm, respectively. The scapular abduction angle (ScAb) on the nonaffected side was significantly greater compared to the affected side and to the nonaffected limb of the high-tone group. No difference was found in the humeral abduction angle (HAb) or in the angle of humeral abduction relative to the scapula (HRel). The mean trunk rotation value for subjects in this group was - 0 . 1 ° + 3.1 °.
High-Tone Group Results of the scapular and humeral orientation measures in the high tone group are presented in Table 3. No differences were found between the affected and nonaffected side in either the angular or linear measures of scapular and humeral orientation. The mean trunk rotation angle was - 1 . 7 ° + 4.3 °. Three of the 17 subjects had greater than 5 ° of trunk rotation; the trunk was rotated posteriorly on the affected side relative to the pelvis. High- Versus Low-Tone Group No significant differences were found in the angular measures of shoulder complex position between the affected limbs of the high- and low-tone subjects. Comparison of the same measures on the nonaffected side between groups 17 16 15 14 13 12 11
10
,.=
E
9
0 1 22.53' ~ Ashworth total score
~'~ ~ 10 11 t 9 . ~ ' -= ~o 14 15
Fig 7 - - I - I i s t o g r a m demonstrating the distribution of total tone score for subjects in the low- and high-tone groups, m, low tone (n : 17); Fq, high tone (n = 17).
Arch Phys Med Rehabil Vol
76, S e p t e m b e r 1995
862
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham Table 4: Association Between Subluxation and Scapular and Humeral Orientation Measures
Table 2: Scapular and Humeral Orientation, Low-Tone Group (n = 17) Variable ScAb HAb HRel Yabs (cm) Zabs (cm)
Affected Side 86.4 ° 8.7 ° 12.3 ° 12.0 9.2
(+5.5) (-+5.9) (_+7.7) (_+1.5) (+_1.7)
Nonaffected Side 91.1 ° 10.8 ° 9.7 ° 11.3 8.3
(___4.9) (+_5.7) (_+5.4) (+_1.0) (_+1.6)
p
Group
Variable
r
p
.003* .100 .126 .041" .021"
Low tone (n = 17)
ScAb HAb HRel ScAb HAb HRel
-.121 - . 145 -.026 -.221 -.089 .164
.643 .578 .921 .393 .734 .528
Values given as mean (+SD). * Significant difference at p < ,05.
r = Pearson correlation coefficient.
showed a significantly greater scapular abduction angle in the low-tone group (91.1 ° versus 86.4°; p < .05. The mean trunk rotation measure was not significantly different between the high- and low-tone group (p > .05).
Scapular Orientation and Subluxation There was a significantly greater glenohumeral subluxation on the affected side in the low-tone group than in the high-tone group. Mean values were .52 + .38cm for the lowtone group and .21 _+ .41cm for the high-tone group (p < .05). No significant relationship was identified between scapular abduction angle and glenohumeral subluxation in either the low-tone group (r = - . 121; p = 0.6) or in the hightone group (r = -.221; p = 0.4). Similarly, no significant relationship was found between either the absolute (HAb) or relative (HRel) humeral abduction angle and subluxation in either the low-tone or high-tone group. (table 4). DISCUSSION
Scapular Orientation The scapula was lower on the thorax and further from the midline on the affected side compared with the nonaffected side in the low-tone group. Decreased tone in the trapezius and rhomboid muscles in the flaccid stage after a stroke may have contributed to these findings. However, the differences between the affected and nonaffected limbs' sides were of small magnitude, suggesting that scapular position on the thorax is not greatly affected by flaccid paralysis of these muscle groups. Because the linear measures are dependent on stature, no comparison of these measures between the high- and low-tone groups was performed. The scapular abduction angle was greater on the nonaffected side in the low-tone group compared with the affected side. This value was also greater than the angle measured on either side in the high-tone group. Thus, there is little evidence from this study that low tone in the scapular musculature in the affected limb contributes to significant scapular Table 3: Scapular and Humeral Orientation, High-Tone Group (n = 17) Variable ScAb HAb HRel Yabs(cm) Zabs (cm)
Affected Side 87.0 ° 10.7 ° 13.7 ° 12.3 8.9
(_+6.5) (_+4.7) (_+6.2) (_+1.3) (_+1.5)
High tone (n = 17)
Nonaffected Side 86.4 ° 10.4 ° 14.0 12.1 8.7
Arch Phys Meal Rehabil Vol 76, September 1995
(_+6.0) (_+5.6) (_+8.7) (_+1.5) (-4-1.1)
p .752 .875 .883 .576 .643
downward rotation on the hemiparetic side as proposed by Cailliet ~ and others. 8'9 Trunk alignment may be a possible explanation for the greater abduction angle on the nonaffected side in subjects with low tone. Bobath ~ stated that hemiplegic subjects tend to sit with more weight on the nonaffected buttock because of sensory deficits and loss of trunk control. The transference of weight to the normal hip is more apparent in subjects with low tone and is postulated to lead to lateral flexion of the trunk toward the hemiparetic side. A small shift in trunk alignment could alter scapular position on either side or both sides in hemiparetic subjects. The mean difference in scapular abduction angle between the two sides was less than 5 °. The degree of trunk asymmetry needed to cause such a small difference would be minimal and may have been undetectable on clinical examination. That no difference in scapular abduction angle was found between the affected and nonaffected sides in subjects with high tone may be related to better postural control in this group. Fourteen of the 17 subjects in the high-tone group were living in the community and had completed a rehabilitation program. The mean age of subjects in this group (61.3 years) was significantly younger than the mean age of subjects in the low-tone group (72.4 years). It is possible that subjects in the high-tone group had better sitting posture and greater ability to sit with weight equally distributed on both buttocks. The findings in the high-tone group do not support clinical descriptions available in the literature. Cailliet ~and o t h e r s 9"12 suggested that during the spastic stage the scapula is lower on the thorax and rotated downward, due to increased tone in the rhomboid muscles, which overwhelm the scapular elevators (trapezius and serratus anterior muscles). Increased tone in the latissimus dorsi and pectoralis minor was also postulated to contribute to the decreased abduction angle of the scapula. It is recognized that an increase in tone in the upper extremity musculature, as measured in this study, does not necessarily indicate an increase in tone in the scapular retractor or pectoralis minor muscles. Because no difference was found in scapular orientation between the affected and nonaffected sides, it would appear that if tone was increased in the rhomboids and pectoralis minor muscles, a downward rotation of the scapula on the affected side did not result, contrary to the suggestions in the literature. The findings in the current study concerning scapular orientation in the low- and high-tone groups are not consistent with those reported previously. Prtvost j4 found that the mean scapular abduction angle in 50 hemiplegic subjects was significantly greater on the affected side than on the nonaffected
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
863
side (mean difference 3.5°).14 Similarly, Arsenault j3 reported a significantly greater scapular abduction angle on the affected side in hemiplegic subjects without glenohumeral subluxation. The authors stated, however, that the difference was small (4 °) and deemed it clinically unimportant. Neither the degree of muscle tone nor the stage of recovery was documented, making comparison difficult.
either group. Thus, no support is provided for the concept that downward rotation of the scapula contributes to glenohumeral subluxation as postulated by Cailliet ~ and Davies. 9 These results are in agreement with those reported by Prdv o s t j4 and Arsenault. ~3 The authors in both of these studies concluded that there was no relationship between scapular abduction angle and degree of glenohumeral subluxation. Cailliet ~ stated that the humeral abduction angle increases as the humerus subluxes on the glenoid fossa. However, no Humeral Orientation The absolute humeral abduction angle (HAb) and the hu- correlation was found between humeral abduction angle and meral abduction angle relative to the scapula (HRel) were not glenohumeral subluxation in either group in the current significantly different between the affected and nonaffected study. Prtvost ~3 reported a significant correlation between sides in either the low-tone or high-tone groups. Cailliet ~ both humeral abduction angle and humeral abduction angle and Davies 9 stated that with flaccid paralysis, the humeral relative to the scapula and glenohumeral subluxation. Howabduction angle increases relative to the downwardly rotated ever, the correlations explained only 6% and 9% of the scapula. The relative humeral abduction angle was greater variability in the data, respectively. Arsenault ]3 found a sigon the affected side (12.3 ° versus 9.7 °) in the low-tone group nificantly greater humeral abduction angle on the affected in the current study, but the difference was not significant. side compared with the nonaffected side in a group of hemiPr6vost 14 reported a small (2.3 °) but significantly greater plegic subjects with glenohumeral subluxation. However, humeral abduction angle on the affected side compared with the mean difference was only 3 ° and the authors concluded the nonaffected side in hemiplegic subjects. Similarly, Arse- that there was no systematic relationship between humeral nault ~3 found a greater humeral abduction angle on the af- abduction angle and glenohumeral subluxation. fected side compared with the nonaffected side in a group CONCLUSIONS of hemiplegic subjects with subluxation. However, the magDownward rotation and depression of the scapula on the nitude of the differences was small (3 °) and deemed clinithorax and humeral abduction relative to the scapula have cally unimportant by the authors.~3 been described in both the flaccid and spastic stages of recovery after cerebrovascular accident. It has been postulated that Trunk Rotation these alterations in alignment contribute to glenohumeral In the present study, no pattern of trunk rotation was subluxation and possibly to pain and functional impairment observed in either the low- or high-tone group. Of the 17 in the upper extremity. Rehabilitation efforts are often disubjects in the high-tone group, 14 had less than 5 ° of rota- rected toward restoration of correct alignment of the scapula tion of the trunk relative to the pelvis, within the range found on the thorax through stretching of soft tissues, and decreasin the healthy subjects. However, three subjects demon- ing tone and increasing strength of weak musculature. The strated greater than 5 ° of posterior trunk rotation of the af- results of this study indicate that the scapula on the affected fected side ( - 12.88°; -8.19°; -5.98°). Ryerson and Levit 8 side was lower on the thorax and further from the midline described a hemiplegic posture in the spastic stage in which in hemiplegic subjects with low tone. Low tone in the perithe trunk is rotated posteriorly on the affected side. The scapular musculature may contribute to the lower position scapula was described as downwardly rotated and elevated on the thorax and increased distance from the midline. The on the thorax in subjects with this trunk posture. Examina- scapular abduction angle was greater on the nonaffected side tion of individual data, in the current study, found no consis- of the low-tone group compared with the affected and to the tent pattern of scapular orientation in the subjects with poste- nonaffected side in the high-tone group. Unequal weight rior rotation of the trunk on the affected side. distribution on the buttock resulting in asymmetry in trunk alignment is a possible explanation for the difference in Glenohumeral Subluxation scapular abduction angle. The differences between the sides There was a significantly greater glenohumeral subluxa- for all measures was small, albeit statistically significant. tion on the affected side in the low-tone group compared to The importance of the differences in terms of clinical manthe high-tone group. This finding is consistent with reports agement is questionable. No differences were found in scapin the literature. 17'25 Basmajian and Bazant j° postulated that ular or humeral orientation between the affected and nonafin the flaccid stage the loss of tone in the supraspinatus fected side in the subjects with increased tone. Glenohumeral muscle allows the humeral head to sublux inferiorly. Paraly- subluxation was significantly greater on the affected side in sis of this muscle was shown by Chaco and Wolf j7 to be an the low-tone group compared with the high-tone group. No important factor in the onset of subluxation in hemiplegia relationship was found between scapular or humeral orientaduring the flaccid stage. The same authors demonstrated that tion and glenohumeral subluxation in either group. The lack when spasticity developed, the supraspinatus muscle began of a consistent pattern of shoulder complex orientation in to respond to loading, and subluxation did not occur. A either group indicates the need for detailed assessment and correlation between a decrease in subluxation and an in- individualized rehabilitation programs. crease in muscle tone has been reported. ~2'25'26 References No significant correlation between scapular abduction 1. Cailliet R. Shoulder pain. Philadelphia: Davis, 1991. angle and degree of glenohumeral subluxation was found in 2. O'Sullivan SB. Stroke. In: O'Sullivan SB, Schmitz TJ, editors. Physical Arch Phys Med Rehabil Vol 76, September 1995
864
3. 4. 5.
6. 7. 8.
9. 10. 11. 12.
13.
14.
15.
THE SHOULDER COMPLEX IN HEMIPLEGIA, Culham
rehabilitation: assessment and treatment. Philadelphia: Davis, 1988:335-79. Van Ouwenaller C, Laplace PM, Chantraine A. Painful shoulder in hemiplegia. Arch Phys Med Rehabil 1986;67:23-6. Hakuno A, Sashika H, Ohkawa T, Itoh R. Arthrographic findings in hemiplegic shoulders. Arch Phys Med Rehabil 1984;65:706-11. Tepperman P, Greyson ND, Hilbert L, Jimenez J, Williams JI. Reflex sympathetic dystrophy in hemiplegia. Arch Phys Med Rehabil 1984; 65:442-7. Najenson T, Pikielny SS. Malalignment of the glenohumeral joint following stroke. Ann Phys Med 1965;8:96-9. Cailliet R. The shoulder in hemiplegia. Philadelphia: Davis, 1980. Ryerson S, Levit K. The shoulder in hemiplegia. In: Donatelli R, editor. Physical therapy of the shoulder. New York: Churchill Livingstone, 1991:117-45. Davies PM. Steps to follow. A guide to the treatment of adult hemiplegia. Berlin: Springer-Verlag, 1985. Basmajian JV, Bazant FJ. Factors preventing downward dislocation of the adducted shoulder joint. J Bone Joint Surg 1959;41-A: 1182-6. Bobath B. Adult hemiplegia: evaluation and treatment. London: Williams Heinemann Medical Books, Ltd., 1990. Van Langenberghe HVK, Hogan BM. Degree of pain and grade of subluxation in the painful hemiplegic shoulder. Scand J Rehab Med 1988;20:161-6. Arsenault AB, Bilodeau M, Dutil E, Riley E. Clinical significance of the V-shaped space in the subluxed shoulder of hemiplegics. Stroke 1991; 22:867-71. Pr6vost R, Arsenault B, Dutil E, Drouin G. Rotation of the scapula and shoulder subluxation in hemiplegia. Arch Phys Med Rehabil 1987;68:786-90. Freedman L, Munro RR. Abduction of the arm in the scapular plane:
Arch Phys Med Rehabil Vol 76, September 1995
16. 17. 18. 19. 20. 21. 22.
23. 24. 25. 26.
scapular and glenohumeral movements. J Bone Joint Surg 1966;48A:1503-10. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg 1976;58-A:195-201. Chaco J, Wolf E. Subluxation of the glenohumeral joint in hemiplegia. Am J Phys Med 1971;50:139-43. Brunnstrom S. Movement therapy in hemiplegia. New York: Harper and Row, 1970. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth Scale of muscle spasticity. Phys Ther 1987;67:206-7. Sawner K, LaVigne J. Brunnstrom's movement therapy in hemiplegia. A neurophysiological approach. New York: Lippincott, 1992:1-276. Sloan RL, Sinclair W, Thompson J, Taylor S, Pentland B. Inter-rater reliability of the modified Ashworth Scale for spasticity in hemiplegic patients. Int J Rehabil Res 1992; 15:158-61. Lee K, Carson L, Kinnin E, Patterson V. The Asworth Scale: A reliable and reproducible method of measuring spasticity. J Neuro Rehab 1989;3:205-9. Culham E, Peat M. Spinal and shoulder complex posture. I: Measurement using the 3Space Isotrak. Clin Rehabil 1993;7:309-18. Pr6vost R, Arsenault AB, Dutil E, Drouin G. Shoulder subluxation in hemiplegia: A radiologic correlational study. Arch Phys Med Rehabil 1987;68:782-5. Finch L, Harvey J. Factors associated with shoulder-hand syndrome in hemiplegia: clinical survey. Physiother Canada 1983;35:145-8. Moskowitz H, Goodman CR, Smith E, Balthazar E, Mellins HZ. Hemiplegic shoulder. N Y State J Med 1969; 15:548-50.
Suppliers a. Polhemus Navigation Sciences Division, McDonnel Douglas Electronics Company, 1 Hercules Drive, PO Box 560, Colchester, VT 05446. b. Systat, Inc., 1800 Sharman Ave., Evanston, IL 60201-3793.