A new pathophysiology of atraumatic rotator cuff tears: adduction restriction of the glenohumeral joint.

BACKGROUND: The pathophysiology of atraumatic rotator cuff tears (ATTs) has not been fully understood. Adduction restriction of the glenohumeral joint can cause pain and disability in patients with ATTs. We aimed to use our adduction test (pushing the humerus toward the side in the coronal plane with scapular fixation) to fluoroscopically measure the glenohumeral adduction angle (GAA) and to assess the effectiveness of adduction manipulation.
MATERIALS AND METHODS: Fifty-five patients with ATTs were included in the study. The GAAs of the patients vs. healthy subjects without ATTs were measured fluoroscopically and compared. During the test, patients showed restriction and expressed pain. The visual analog scale (VAS) score, passive range of motion (ROM), and the American Shoulder and Elbow Surgeons score at the initial visit were compared with those after adduction manipulation.
RESULTS: Of the patients, 41 (75%) had positive adduction test results. A higher percentage of positive adduction test results was observed in smaller tears. The average GAA was -21.4° on the affected side, which was smaller than that on the unaffected side, at -2.8° (P < .001), and that in healthy subjects, at 4.8° (P < .001). After manipulation, the GAA was -0.8° (P < .001) and the VAS score, the American Shoulder and Elbow Surgeons score, and all ROM values significantly improved up to the level on the unaffected side.
CONCLUSION: Adduction restriction of the glenohumeral joint was identified in 75% of all the patients with ATTs. Adduction manipulation significantly reduces the VAS score and restores the ROM. Adduction restriction is considered a crucial pathophysiology of ATTs.

An atraumatic rotator cuff tear (ATT) is a very common shoulder disorder, affecting approximately 20% of all persons older than 50 years; however, the pathophysiology of the disorder has not been fully understood.,, Two-thirds of the individuals with ATTs are asymptomatic, but the remaining one-third complain of shoulder pain, stiffness, crepitus, weakness, and instability. If a rotator cuff tear were a cause of these symptoms, the majority of individuals with rotator cuff tears would complain of shoulder symptoms. Other unknown pathophysiologies might change an asymptomatic rotator cuff tear into a symptomatic condition. Recent studies have demonstrated that the size of the rotator cuff tear is not correlated with pain and duration of symptoms.,, Nonoperative treatment is an effective and lasting option for many patients with ATTs.,,,, Moreover, other studies have reported that clinical outcomes of physiotherapy or arthroscopic cuff repair for ATTs are not significantly different.,, These clinical studies suggest that the rotator cuff tear is not a main cause of symptoms but that a different pathophysiology produces the symptoms.

Synovial inflammation,, shoulder stiffness, impingement, muscle weakness, instability, and cuff tear arthropathy have been documented as clinical findings of ATTs. Shoulder stiffness frequently occurs concomitantly with rotator cuff tears. Rotator cuff tears were associated with 56% of patients with mild or moderate shoulder stiffness. Eighty-two percent of patients with ATTs indicated restriction of flexion, external rotation, and internal rotation (hand behind the back) at the initial visit in our previous study. Thickening of the capsule, inferior and middle glenohumeral ligaments, coracohumeral ligament, and subacromial bursa causes shoulder stiffness, adhering to the tendons in patients with ATTs.,, Mobilization and release of tendons must first be performed to assess whether repair is possible and to ensure a repair with minimal tension. The thickening soft tissues surrounding the torn supraspinatus and infraspinatus tendons are extensively removed to pull out the torn tendons and reach the desired insertion at the base of the tuberosity. If the releasing procedure that consequently eliminates joint stiffness can relieve pain and restore shoulder function, the thickening superior part of the glenohumeral joint such as the superior capsule, coracohumeral ligament, and subacromial bursa may be a new pathophysiology of ATTs. Moreover, patients who have undergone a failed rotator cuff repair are often pain free, and their outcome scores are nearly the same as those of patients with a successful repair.,, The outcomes of rotator cuff repair support the importance of the removal of the adhesive soft tissues, suggesting the presence of adduction restriction in the glenohumeral joint.

Range of motion (ROM) evaluation is an essential method for assessing ATTs before and after treatment. Conventional evaluation of ROM in the shoulder joint comprises a combined motion of the glenohumeral joint and scapulothoracic joint. To evaluate the glenohumeral condition with ROM precisely, ROM measurement with scapular fixation is recommended., If the scapula remains fixed with upward rotation and pushes the upper arm to the chest in the coronal plane, adduction restriction of the glenohumeral joint can be accurately evaluated. The clinical implications of adduction restriction of the glenohumeral joint were arranged into 3 hypotheses: Adduction restriction of the glenohumeral joint would exist in patients with ATTs, adduction restriction may be correlated with stiffness, and adduction manipulation would lessen pain severity and restore loss of ROM.

There has been no method to assess adduction restriction of the glenohumeral joint. We developed a new method—the adduction test of the glenohumeral joint (Fig. 1, A and B)—to evaluate adduction restriction of the joint, treating patients with adduction restriction via adduction manipulation. This was a prospective, comparative study of patients with ATTs and normal healthy subjects without rotator cuff tears. We aimed to investigate the rate of adduction restriction of the glenohumeral joint in patients with ATTs and to assess the effectiveness of adduction manipulation in patients with adduction restriction.

Figure 1

Adduction test maneuver. (A) Starting position of adduction test. The subject being tested lay in the lateral decubitus position on the examination table. The targeted scapula was rotated upward and fixed manually with 2 hands by 1 examiner. The other examiner moved the humerus in abduction up to 110°. (B) During the adduction test, the upper arm was pushed gently toward the side in the coronal plane and the upper arm easily touched the side; thus, the test result was negative (no restriction of the glenohumeral joint). (C) During the adduction test, the upper arm did not touch the side because of shoulder pain; thus, the test result was positive (restriction of the glenohumeral joint existed).

Materials and methods

Enrollment of healthy subjects

All the participants provided informed consent, and 34 healthy volunteers without a history of shoulder pain and trauma participated. A senior doctor performed the following clinical screening examination: evaluation of ROM, impingement and instability tests, and scapular dyskinesia test. We excluded 2 individuals with ROM reduction compared with that of the contralateral shoulder joint and 2 demonstrating asymptomatic rotator cuff tears on magnetic resonance imaging (MRI) (Echelon RX, 1.5 T; Hitachi, Tokyo, Japan). The remaining 30 subjects were divided into 3 age groups: The first group included 5 men and 5 women in their 20s (average age, 25.4 years), the second group included 5 men and 5 women in their 50s (average age, 56.1 years), and the third group included 5 men and 5 women in their 70s (average age, 73.8 years) (Table I). All the subjects underwent MRI of both shoulder joints to confirm the absence of rotator cuff tears.

Table I

Characteristics of healthy subjects and patients

	Healthy subjects	Patients with rotator cuff tears	Patients with positive adduction test results
n (%)	30	55 (100)	41 (75)
n by age group
20s	10
50s	10
70s	10
Sex
Male	15	26	20
Female	15	29	21
Average age, yr	51.8	72.3	70.7
Average age by age group, yr
20s	25.4
50s	56.1
70s	73.8
Body mass index		24.4	24.8
Body mass index by age group
20s	23.1
50s	26.3
70s	24.2
Side (dominant arm)
Right	30 (27)	34 (33)	30 (30)
Left	30 (3)	21 (4)	11 (2)

Inclusion and exclusion criteria of patients with ATTs

All the patients underwent a routine radiographic evaluation in the anteroposterior view in internal and external rotation, as well as outlet views of the shoulders bilaterally and MRI. The inclusion criteria for patients with symptomatic ATTs were as follows: (1) rotator cuff tear on MRI, (2) persistent symptoms for <3 months, (3) passive ROM of >120° of flexion and >30° of external rotation at the side, and (4) normal radiologic appearance. The study excluded (1) patients with an acute injury; (2) those who had already undergone nonoperative treatment; (3) those with full-thickness tears of the subscapularis and/or teres minor; (4) those with substantial cervical pathology and/or radiculopathy; (5) those with concomitant pathology of the affected shoulder, such as glenohumeral osteoarthritis, calcific tendinitis, a superiorly migrated humeral head, and osteonecrosis of the humeral head; and (6) those with systemic inflammation or metabolic disorders, such as rheumatoid arthritis and diabetes mellitus. A total of 55 patients (26 men and 29 women; average age, 72.3 years) with ATTs who met the inclusion and exclusion criteria were enrolled in the study (Table I). The rotator cuff tears were classified into partial, small, medium, large, and massive tears following the Cofield classification. Of the patients, 41 (75%) had a positive adduction test result (the expression of pain while the upper arm is being pushed toward the side and the upper arm is not touching the side) (Fig. 1, A and C).

Clinical and radiologic adduction tests and evaluation

Two examiners were needed to perform the manual adduction test. The subject being tested lay in the lateral decubitus position on the examination table. The targeted scapula was rotated upward and fixed manually with 2 hands by 1 examiner. The other examiner moved the humerus in abduction up to 110° and then gently adducted the glenohumeral joint in the coronal plane until the elbow touched the side (Fig. 1, A and B). Patients who expressed pain and whose upper arms could not touch their sides were considered to have a positive adduction test result. The radiographic adduction test was performed using the same maneuver: The upper arm was pushed toward the side with a pressure of 5 kg using a dynamometer at the distal part of the humerus (Mobie MT-100W; Sakai Medical, Tokyo, Japan). A radiographic cassette was placed on the anterior side of the subject, and the true posteroanterior view of the glenohumeral joint was obtained. The beam was emitted upward at 20° to obtain the true posteroanterior view (Fig. 2, A). The angle between the line connecting the superior to inferior margins of the glenoid and the bone axis of the humerus (passing through the midpoint of 2 transverse lines positioned 7 and 17 cm from the top of the humeral head) was defined as the glenohumeral adduction angle (GAA) (Fig. 2, B). Radiographic adduction of the glenohumeral joint was considered a positive value; parallel glenoid and humeral lines, a value of zero; and abduction, a negative value. The GAA was measured using XTREK FESTA software (J-MAC System, Sendai, Japan). In the healthy group, the GAAs on the dominant and nondominant sides were 4.8° ± 6.6° and 6.3° ± 6.5°, respectively. Because there was no significant difference between the dominant and nondominant sides (P = .45 by Student t test), the GAA on the dominant side was chosen for further evaluation. We measured the GAA bilaterally in the shoulders of the 30 healthy subjects and all patients with ATTs.

Figure 2

(A) Radiographic adduction test. A radiographic cassette (∗) was placed on the anterior side of the subject. The beam was emitted upward at 20° to obtain the true posteroanterior view, and the upper arm was pushed with a force of 5 kg using a dynamometer at the distal part of the humerus (▼). (B) Measurement of glenohumeral adduction angle, defined as the angle between a line connecting the superior to inferior margins of the glenoid (AB) and the bone axis of the humerus (EF) (passing through the midpoint of 2 transverse lines [C, D] positioned 7 and 17 cm from the top of the humeral head). Adduction of the glenohumeral joint was expressed as a positive value; abduction, as a negative value.

We investigated and calculated the intraclass correlation coefficient (ICC) to evaluate the reliability of the radiographic adduction test with respect to both the radiographic technique and measurement accuracy in 5 healthy subjects and 10 patients.

Adduction manipulation

Patients with positive test results were treated with adduction manipulation of the glenohumeral joint to determine whether adduction restriction could be a pathology in patients with ATTs. The manipulation was performed under local anesthesia in an outpatient setting per a standardized protocol. Under ultrasound-guided views, a needle was inserted from the Neviaser portal, and 10 mL of 1% lidocaine was injected into 3 points: 2 points in the glenoid neck and 1 point in the base of the coracoid process (Fig. 3). The physician abducted the affected extremity and then adducted it gently with scapular fixation. After adduction manipulation treatment, poultice preparations of nonsteroidal anti-inflammatory drugs were applied to the patients. Self-stretching and strengthening exercises were undertaken for 3 weeks to maintain ROM under the guidance of a senior researcher.

Figure 3

Local anesthesia for adduction manipulation. (A) Insertion of needle from Neviaser portal. (B) Injection points for local anesthesia. Before adduction manipulation, a needle was inserted from the Neviaser portal (A), and 10 mL of 1% lidocaine was injected into the glenoid neck (1, 2) and into the base of the coracoid process (3)(B).

Clinical assessment

The body weight and height of the healthy subjects and patients were measured, and the body mass index was calculated. Bilateral GAA and ROM, including flexion, abduction, external rotation at the side, internal rotation with the hand behind the back, external rotation at 90° of abduction, and internal rotation at 90° of abduction, were evaluated in the standing position using calipers before and immediately after manipulation, as well as at 3 weeks and 6 months after manipulation. The visual analog scale score, American Shoulder and Elbow Surgeons score, and Japan Orthopaedic Association score were also recorded. Patients self-assessed their symptoms at the final examination: The outcome was considered excellent in those with no pain or disability in daily and sporting activities; good in those with pain after sporting activities or heavy work; fair in those with pain and disability in daily life; and poor in those with persistent pain and disability or a condition worse than the previously mentioned conditions.

Statistical analyses

Data are expressed as mean (standard deviation) or median (interquartile range). We used the Student t test to analyze age and ROM; the Fisher exact test to assess sex and side; and the Mann-Whitney U test to evaluate the pretreatment period, follow-up period, visual analog scale score, ROM, American Shoulder and Elbow Surgeons score, and Japan Orthopaedic Association score. All the tests were 2-sided, and P < .05 was accepted as statistically significant. All statistical analyses were performed with SPSS software (version 23.0; SPSS Japan, Tokyo, Japan).

Results

Reliability of radiographic adduction angle

The average ICC(1,1), ICC(1,3), and ICC(2,3) measures were 0.93, 0.98, and 0.91, respectively, with a 95% confidence interval from 0.92 to 0.99 for radiographic technique; therefore, interobserver reliability was high. The reliability for measurement accuracy was also high: ICC(1,1), 0.99; ICC(1.3), 0.98; and ICC(2,3), 0.97.

Adduction angle in healthy subjects and patients

The average GAA in the healthy subjects was 5.1° ± 6.3° for those in their 20s, 4.5º ± 6.6° for those in their 50s, and 1.6° ± 8.6° for those in their 70s. A significant difference in the angle was found between healthy subjects in their 20s and those in their 70s (P < .05) (Table II). The GAA in the patients averaged –21.4° ± 10.2° in the affected shoulder and –2.8° ± 7.8° on the unaffected side at the initial visit. The GAA of the bilateral shoulders of the patients was significantly smaller than that of the age-matched healthy subjects in their 70s (P < .001), that is, adduction restriction was significantly more severe in the patients. The average adduction angle in the affected shoulders was –1.0° at 3 weeks (P < .001) and –0.8° at 6 months (P < .001) after adduction manipulation, which was statistically larger than that at the initial visit.

Table II

Radiographic glenohumeral adduction angles in healthy subjects and patients

Group	Glenohumeral adduction angle, °	P value
Healthy subjects (n = 30)	4.8 ± 6.6
20s (n = 10)	5.1 ± 6.3
50s (n = 10)	4.5 ± 6.6	P = .687 vs. 20s
70s (n = 10)	1.6 ± 8.6	P < .05 vs. 20s
Unaffected shoulder (n = 41)	–2.8 ± 7.8	P < .01 vs. 70s
Affected shoulder (n = 41)	–21.4 ± 10.2	P < .001 vs. unaffected shoulder
After manipulation (n = 41)	2.7 ± 6.5	P < .01 vs. affected shoulder
3 weeks after manipulation (n = 41)	–1.0 ± 5.1	P = .076 vs. after manipulation
6 mo after manipulation (n = 41)	–0.8 ± 8.3	P = .834 vs. 3 weeks after manipulation

GAA, glenohumeral adduction angle.

A positive value indicates adduction of the glenohumeral joint, whereas a negative value indicates abduction of the glenohumeral joint.

Correlation coefficients between ROM and adduction angle

The correlation coefficients between the adduction angle and each ROM were 0.048 in flexion, 0.129 in abduction, 0.23 in external rotation at the side, –0.106 in internal rotation with the hand behind the back, 0.197 in external rotation at 90° of abduction, and 0.319 in internal rotation at 90° of abduction, showing that adduction restriction of the glenohumeral joint was correlated to the loss of each ROM.

Adduction test and tear size

The adduction test findings, tear sizes, treatment options, and outcomes of all patients are summarized in Figure 4. Of the 55 patients, 41 (75%) (20 men and 21 women; average age, 70.7 years) had positive adduction test results and were further evaluated (Table I). The GAA averaged –21.4° in the 41 patients (75%) with positive adduction test results; however, the GAA in the remaining 14 patients, with negative adduction test results, was –0.7° (P < .001). The correlation between the tear size in these 41 patients and the rate of positive test results revealed an inverse relationship: 4 of 4 patients with partial tears (100%), 15 of 17 with small tears (88%), 14 of 17 with medium tears (82%), 3 of 6 with large tears (50%), and 5 of 11 with massive tears (45%). Asymptomatic ATTs were found in 13 of 41 patients (32%) in their contralateral shoulder joints on MRI, wherein the results of the adduction test of these shoulders were negative.

Figure 4

Flowchart showing patients’ treatments and outcomes throughout study.

Adduction manipulation and other treatment

Adduction restriction of the glenohumeral joint, loss of ROM, and pain significantly improved with adduction manipulation (Fig. 4, Table III). Of the 41 patients with positive adduction test results, 36 (88%) improved statistically (excellent and good outcomes) whereas 5 patients had moderate pain after adduction manipulation (fair outcome). Consequently, surgery was performed in 4 patients with negative adduction test results and the 5 patients (fair outcome) whose shoulder pain, muscle weakness, and instability persisted after manipulation; of these patients, 7 underwent arthroscopic cuff repair surgery and 2 underwent reverse total shoulder arthroplasty. Ten patients with negative adduction test results who displayed restriction of scapular movement and tightness of the shoulder muscles were treated with physiotherapy. Consequently, adduction manipulation was effective in 36 patients (65%), physiotherapy was performed in 10 (18%), and operative treatment was performed in 9 (16%).

Table III

Change in pain, clinical scores, and passive range of motion during treatment

Assessment method	Unaffected shoulder	Affected shoulder	P value shows affected shoulder VS unaffected shoulder	After manipulation	P value shows ROM before manipulation vs ROM after manipulation	3 weeks after translation	P value shows affected shoulder VS 3 weeks after translation	6 mo after translation	P value shows 3 weeks after translation and 6 mo after translation
VAS score	0	6.0	<.001			1.5	<.001	1.2	.39
ASES score	100	62.8	<.001			81.3	<.001	92.1	<.01
JOA score	100	72.5	<.001			84.7	<.001	91.5	<.05
Passive ROM
Flexion, °	168	150	<.001	168	<.01	165	<.01	163	>.999
Abduction, °	166	136	<.001	159	<.001	161	<.001	166	>.999
ER1, °	55	41	<.001	50	<.01	50	<.01	56	.87
IR1	T7	T12	<.001	T9	<.001	T7	<.001	T8	.93
ER2, °	87	77	<.05	85	<.05	88	<.05	87	>.999
IR2, °	88	50	<.001	84	<.001	83	<.001	86	>.999

VAS, visual analog scale; ASES, American Shoulder and Elbow Surgeons; JOA, Japan Orthopaedic Association; ROM, range of motion; ER1, external rotation at side; IR1, internal rotation with hand behind back; ER2, external rotation at 90º of abduction; IR2, internal rotation at 90º of abduction.

Discussion

Numerous studies regarding rotator cuff tears have been published in the past 30 years; however, the pathophysiology of the ATT remains unclear. Therefore, there are no established guidelines for the optimal management of this disorder. Recent studies have demonstrated that the severity of rotator cuff tears does not correlate with the duration of symptoms and pain.,, Clinical outcomes of physiotherapy and surgical repair are not significantly different in terms of pain and disability.,, Patients who have a retear of the rotator cuff are often pain free, and the outcome scores are similar to those of patients with a successful repair.,, These studies have suggested the existence of a pathophysiology besides rotator cuff tears but have provided us with no indication as to why asymptomatic rotator cuff tears become symptomatic or to identify the crucial pathophysiology of ATTs. We have verified that adduction restriction of the glenohumeral joint is present in 75% of patients with ATTs. ROM of the shoulder joint is dependent on the glenohumeral and scapulothoracic joints, and the scapula rotates downward to compensate for the adduction restriction during conventional ROM measurement techniques. In turn, it conceals the adduction restriction of the glenohumeral joint. The restriction, new shoulder stiffness, signifies an important pathophysiology of ATTs. Adduction manipulation turned symptomatic ATTs asymptomatic; therefore, the adduction restriction may be why asymptomatic rotator cuff tears become symptomatic. To our knowledge, this is the first report concerning adduction restriction of the glenohumeral joint in ATTs that leads to loss of all ROMs in the shoulder joint.

What is the lesion responsible for adduction restriction? We infer that the thickness and adhesion of the superior soft tissues of the glenohumeral joint result in adduction restriction, including the coracohumeral ligament, subacromial bursa, and superior capsule., The coracohumeral ligament envelops the supraspinatus, infraspinatus, and subscapularis tendons. Furthermore, the ligament is divided into 2 parts: One part spreads the fibers over the rotator interval to the posterior portion of the greater tuberosity, and the other envelops the entire supraspinatus tendon and the superior portion of the subscapularis. The thickness of the latter part in the ligament may contribute to adduction restriction of the glenohumeral joint. A dynamic MRI study revealed increased blood flow at the rotator interval and subacromial bursa in patients with ATTs. The increasing blood flow in the superior portion of the joint induces inflammation and proliferation of the coracohumeral ligament, superior capsule, and bursal wall in patients with frozen shoulder and ATTs.,, Thus, abnormal blood flow may relate to shoulder pain and consequently develop into adduction restriction and initiate loss of ROM.,, A direct cause of adduction restriction is assumed to be the adhesion around the supraspinatus tendon and muscle, and adduction manipulation may alleviate the adhesion to reduce pain and restore ROM of the shoulder joint.

The correlation between the tear size and rate of positive adduction test results shows an inverse proportion: The rate of positive adduction test results is higher in partial and smaller tears. This result corresponds with the findings of a dynamic MRI study that reported that a greater increase in abnormal blood flow exhibited a higher rate of smaller tears. With enlargement of the tear, the soft tissues surrounding the supraspinatus tendon do not proliferate sufficiently because of a smaller volume of abnormal blood flow. The prevalence of positive adduction test results is higher in smaller tears because the anatomic magnitude of adduction restriction may be larger in smaller tears.

The outcomes of operative treatment can be explained by the adduction restriction. The outcomes between patients with successful rotator cuff repairs and those with unhealed repairs are similar.,, The soft tissues around the supraspinatus and infraspinatus tendons are dissected before tendon repair, and the dissection contributes to complete removal of the adduction restriction. Such débridement for massive rotator cuff tears is clinically relevant in the treatment. Shoulder surgeons have apparently focused on tendon repair. Extensive release of the soft tissues around the torn tendons is considered to assess the extent of retraction and the tendon quality, but the release is not recognized as a treatment option for symptomatic ATTs. An arthroscopic technique dissecting the coracohumeral ligament, superior capsule, bursa, and rotator interval in patients with frozen shoulder has been described. This advanced surgical technique should also be useful for arthroscopic cuff repair in patients with ATTs.

Glucocorticoid injection and physiotherapy are 2 widely used conservative treatments for ATTs. Subacromial bursa and/or intra-articular injection of glucocorticoids achieves pain relief and functional improvement for 3 months., Injections of glucocorticoid agents have an anti-inflammatory effect and inhibit proliferation of the capsule. A sufficient volume of local anesthetic may occlude abnormal blood vessels formed in the superior part of the glenohumeral joint and decrease the development of adduction restriction in the initial phase of ATTs. Physiotherapy, including joint mobilization and manipulation, as well as specific exercise regimens, is a first-line treatment for ATTs. There is substantial evidence to support the use of physiotherapy as a first-line management, especially in individuals older than 60 years with chronic and degenerative full-thickness rotator cuff tears. The goal of physiotherapy is to relieve pain, increase ROM, promote healing, strengthen the periscapular musculature, and improve the stabilizing function of the rotator cuff., Many prognostic factors concerning physiotherapy have been evaluated, and it has been shown that (1) smoking and pain catastrophizing and (2) disability and pain catastrophizing are two of the best models; however, none of the models had a satisfactory performance. Despite numerous studies on physiotherapy for ATTs, it has remained unclear which exercise regimen is effective for each patient owing to limited knowledge of the crucial pathophysiology. Adduction restriction of the glenohumeral joint results in loss of all ROMs. Therefore, joint mobilization or manipulation can eliminate the restriction and increase the shoulder joint ROM. In our study, 36 of 41 patients with positive adduction test results had excellent or good outcomes with adduction manipulation, and 10 patients with negative adduction test results were treated with physiotherapy. If physiotherapy were effective for adduction restriction, these 36 and 10 patients (84% of the total) would achieve pain relief, restoration of ROM, and improvement in functional outcomes. A systematic review that investigated the effectiveness of physiotherapy in 2010 shoulders with rotator cuff tears reported the following outcomes: 78% showed pain reduction, 81% showed improved ROM, 85% had better strength, and 84% had improved functional outcomes. The improvement rate in the study is in agreement with that in our study. A further comparative study between physiotherapy and adduction manipulation is needed in the future.

Our study verifies the clinical effectiveness of the adduction test and manipulation in patients with coexisting ATTs and adduction restriction. It is noteworthy that 36 of 41 patients (88%) who underwent adduction manipulation had relief from shoulder pain and showed restoration of ROM for a short period. However, 5 patients (12%) showed mild improvement and had residual symptoms that were attributed to rotator cuff tears. The remaining symptoms caused by rotator cuff tears, such as muscle weakness, instability, and pseudoparalysis, are apparently exposed by adduction manipulation, and arthroscopic rotator cuff repair or reverse total shoulder arthroplasty is consequently the optimal management technique. Rotator cuff tear is a solely morphologic diagnosis made based on MRI or ultrasound sonography; however, the clinical findings are numerous and varied.,,29, 30, 31, Symptoms due to rotator cuff tears are frequently concealed by adduction restriction. Physicians must therefore exclude this restriction using the adduction test. Because adduction restriction is identified as a new pathophysiology of ATTs, we should expect that the examination, diagnosis, and subsequent management of ATTs would be more simplified.

There are several limitations to this study. One important limitation is that responsible lesions were not definitely detected using imaging. Therefore, we rigorously observed the stiff soft tissues in the superior part of the glenohumeral joint on MRI. Further anatomic and biomechanical studies are needed to verify the relationship between coracohumeral ligament thickening and adduction restriction. Another limitation was the short follow-up period. Although the symptoms of patients apparently improved after the manipulation, the shoulders will carry a risk of tear enlargement. A follow-up period of up to 2 years is required to assess the effectiveness of adduction manipulation. The final limitation is that we do not know whether physiotherapy is effective for treating the adduction restriction. A comparative study of physiotherapy and adduction manipulation in ATT patients should therefore be conducted in the future.

Conclusion

Seventy-five percent of the patients with ATTs exhibited adduction restriction of the glenohumeral joint. The adduction angles in the affected shoulders were significantly smaller than those in the contralateral shoulders and in the shoulders of age-matched healthy subjects. The prevalence of a positive adduction test result was higher for smaller tears. Adduction manipulation of the glenohumeral joint greatly reduces pain intensity and restores ROM and shoulder joint function for a short period. Adduction restriction is considered a new pathophysiology of ATTs.

Acknowledgments

The authors would like to express their appreciation to Dr. Kiyohisa Ogawa, my senior doctor, who gave us excellent advices on this study. And also we thank all physical therapists of the Department of Rehabilitation in Kuwano Kyoritsu Hospital for support of this study.

Disclaimer

The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.