Shoulder impingement: various risk factors for supraspinatus tendon tear

Abstract A total of 680 cases of monolateral shoulder pain and functional impairment were included, and Chi-Squared tests was incorporated to test for possible associations. No relation between impingement syndrome and potential risk factors was found, such as presence of down slopping (P = .083), presence of ossification acromiale (P = .102), presence of calcific tendinitis (P = .144), types of acromion (I [P = .600], II [P = .536], III [P = .633] and IV [P = .832]) and grade of acromioclavicular degenerative changes (mild [P = .077], moderate [P = .111], and severe [P = .700]). However, a significant relationship was uncovered between impingement syndrome and risk factors such as gender (X2 = 7.004, df = 1, P = .08) (where females were more prone), history of shoulder dislocation (X2 = 19.440, df = 1, P = .001), presence of supraspinatus tendon tear or tendinopathy (X2 = 69.344, df = 1, P = .001) and supraspinatus complete tear (X2 = 13.593, df = 1, P = .001). A significant relationship was found between the type of supraspinatus pathology and factors such as gender (female more prone) (X2 = 34.719, df = 3, P = .01), presence of down slopping (X2 = 57.765, df = 3, P = .01), history of shoulder dislocation (X2 = 148.880, df = 3, P = .001), type III of the acromion (X2 = 12.979, df = 3, P = .005), presence of acromioclavicular generative changes mild (X2 = 76.408, df = 3, P = .001) and moderate (X2 = 29.697, df = 3, P = .001), and acromiohumeral distance of ≤3 mm (X2 = 18.915, df = 3, P = .001), 3.1 to 6 mm (X2 = 13.212, df = 3, P = .004), and 9.1–12 mm (X2 = 15.066, df = 3, P = .002). Overall, the Magnetic Resonance Imaging results yielded high sensitivity for detecting full-thickness supraspinatus tears. Considering the findings, this study may help radiologists understand the salient risk factors and identify which factors are mainly responsible for supraspinatus tendon tears and the respective grade of tear (articular partial, bursal partial, complete, or tendinopathy).


Introduction
Rotator cuff disease is one of the most common causes of shoulder pain, although there is uncertainty regarding the various options for effective treatment. [1] In recent times, the detection of even small tears has become essential since they have been shown to progressively lead to necessary surgery to mitigate shoulder pain. [2] Traditional magnetic resonance imaging (MRI) and ultrasonography are relatively accurate for the identification of full-thickness tears and a similar sensitivity for detecting partialthickness tears. [3] Impingement is the primary cause of rotator cuff pathologies; the supraspinatus tendon is usually the affected tendon because of anatomical structure, the tendon that passes under the acromion. [4] The humeral head, coracoacromial ligament, acromioclavicular joint, and the surface of the anterior third of the acromion define the subacromial space. The supraspinatus tendons, subacromial bursa, the capsule of the shoulder joint, and long head of the biceps brachii tendon are the tissues that occupy the subacromial space. [5] Subacromial impingement syndrome (SAIS) influences any or all of these structures. It is an encroachment of the subacromial tissues due to the subacromial space's narrowing. The narrowing of space characterizing SAIS has been described by 2 predominant mechanistic theories. [6] The first is extrinsic impingement, where degeneration and inflammation of the tendon occur because of mechanical compression by some external structure to the tendon. The second is intrinsic impingement, where complete-or partial-thickness tendon tears form due to a degenerative process through time with tension overload, the trauma of the tendons, or prolonged overuse. [7] Acromial shapes, abnormal kinematics of the joint, downsloping presence, ossification acromiale, acromioclavicular degenerative changes, and inferior osteophytes are understood risks that can lead to impingement. [3] MRI of the shoulder at 3.0T field strength is highly sensitive and accurate in detecting supraspinatus tendon tears and broadly exceeds the efficacy of arthroscopic analysis. [8] Indeed, even MRI at 1.5T field strength is sensitive enough to detect full-thickness supraspinatus tendon tears, although not as effective in detecting some other tendon pathologies. [8][9][10][11] A partial tear is detected when a tendon defect extends to either the bursal or articular surface or can be intra-tendinous. [12] Initially, MRI sensitivity to detect partial-thickness tears ranged from 35% to 87%. However, with the technological advancement of MRI machines, sensitivity can now be as consistently high as 85%. [12] 2. Materials and methods The study included shoulder MRI patients from January 1, 2017, and May 31, 2020, at King Abdulaziz University Hospital. The study was approved by the King Abdulaziz University, Saudi Arabia, under the Ref. No. 348-20. The study sample included all patients with suspected or diagnosed subacromial impingement syndrome, along with asymptomatic healthy patients. The presence of any risk factors was evaluated in all patients, and the status of the supraspinatus tendon was recorded. We excluded patients with surgical intervention or any adjacent pathology that may affect the subacromial space, such as tumors or fractures.
Two independent radiologists evaluated all analyses with a list of risk factors that evaluated its presence and effect on the supraspinatus tendon. The supraspinatus tendon was assessed with no specific guidelines, but the status of the tendon as found in the MRI images was clearly emphasized to be divided into tendinopathy, articular partial-thickness bursal partial thickness, complete tear, and normal anatomy. The supraspinatus tendon was scanned at parallel and perpendicular perspectives. The risk factors studied included the presence of downsloping, ossification acromiale, acromioclavicular degenerative changes, and inferior osteophytic formations.
MRI examinations were performed with the same protocols, using standard 3T MRI machines (Philips Achieva 3T MRI and Siemens MAGNETOM Trio 3.0T MRI machine manufactured in the Middle East). Examinations were performed on the shoulder using mild external rotation. This anatomic position optimally places the supraspinatus tendon perpendicular and parallel to the oblique sagittal imaging and oblique coronal planes.
Data were reported as mean ± SD for continuous variables, and dichotomous and categorical variables were reported as percentages and frequencies. The Chi-Squared test was used for evaluating relationships between categorical data. A P value of <.05 was set as statistically significant (Fig. 1).
One hundred eighty three cases reported inferior osteophytes size, classified into 5 groups according to size ranging from ≦1 mm to >4 mm. Table 4 shows the relationship between the size of the inferior osteophyte and factors such as impingement syndrome and type of supraspinatus pathology (tear and tendinopathy). It was found that inferior osteophyte size was not significantly associated with the presence of impingement syndrome (P = .367). None of the sizes of osteophyte 1 mm (P = .616), 1.1 to 2 mm (P = .075), 2.1 to 3 mm (P = .794), 3.1 to 4 mm (P = .191), and >4 mm (P = .103) was associated with supraspinatus pathology.

Discussion
The occurrence of tears of the supraspinatus tendon is observed frequently in comparison with that of other rotator cuff tendons. [13,14] Owing to the hypovascularity near the insertion of the supraspinatus tendon, specifically on its articular side, the partial-thickness tear is much higher on the supraspinatus tendon's undersurface. [15] The objective of this study was to evaluate the various risk factors for the supraspinatus tendon in cases of SAIS. Both genders were equally represented in this study. However, the role of gender in developing impingement was statistically significant, with females being more prone to impingement syndrome, partial or complete tendon tear, and tendinopathy. This finding contrasts with the results of several previous studies, which showed that male patients had a comparatively more significant number of shoulder lesions. [16][17][18] However, Razmjou, Lincoln, Macritchie, et al [19] posit that females are more severely affected by shoulder injuries or pathologies despite the higher prevalence in males.
An anterior or lateral down-sloping and a low-lying acromion are essential in developing subacromial impingement. [20] However, in current research, the prevalence of downsloping in the presence of impingement syndrome was found statistically insignificant. Still, more specifically, statistical significance was reached with the presence of supraspinatus tendon tear or tendinopathy. This means the presence of supraspinatus tendon pathology may require the radiologist to look for the existence of down-sloping carefully. By contrast, the presence of downsloping does not mean a significant association with the presence of the SAIS.
Ossification acromiale is effectively evaluated in upper axial images, where a low signal space is identified between the nonfused ossicle and the high-signal marrow of the distal acromion. [21] The double acromioclavicular joint can detect it on the coronal oblique. The ossification acromiale is prevalent in between 1% and 15% of the general population. [22] Previous studies have revealed the considerable risk in developing impingement when this relatively common variant is present. [23,24] However, the current study's findings show no significant relationship with impingement syndrome, supraspinatus tear, or the presence of impingement symptoms. The failure to detect the considerable correlation of this relatively standard variant could be a product of the small sample size, and more studies are recommended to confirm or disconfirm a correlation.
The relationship between the type of acromion and supraspinatus tear was also significant where all the participants with type III had a partial tear (59.3%), complete tear (18.5%), or . This means type III acromion has a higher chance of a concomitant supraspinatus pathology. This is in accordance with one study showing that type III is most significantly associated with full-thickness or a complete tear. [25] However, another study (Kim et al) did not correlate acromion type and rotator cuff tears. The difference in findings between the studies could be because Kim et al [26] carried out minimal participants.
The presence of acromioclavicular generative severity was also significant, where, with mild and moderate degenerative changes, there was a greater chance of having a partial or complete tear. It has been postulated that successful resection of acromioclavicular degeneration is necessary to treat rotator cuff tear. [27] Thus, in the presence of supraspinatus tear, acromioclavicular degeneration should also be monitored to support recovery. Table 3 Relationship between supraspinatus tear and tendinopathy and various risk factors.  Table 4 Relationship between size of inferior osteophyte and other outcomes. Calcific tendinitis was not correlated with impingement syndrome and type of supraspinatus pathology. However, calcific tendinitis and partial supraspinatus tears should be closely monitored, as the symptoms of one may mask the other, preventing detection. [28] Chronic shoulder dislocation had a significant relationship with impingement syndrome: participants with a history of shoulder dislocation had less (39.2%) chance of getting impingement than participants with no history (62.1%). Likewise, participants with a history of shoulder dislocation had a greater chance of not having supraspinatus pathology than participants with no history that observed correlation with supraspinatus tears and tendinopathy. However, further study is required to more fully establish associations because previous studies have not generally considered shoulder dislocation when assessing risk factors for impingement. [29][30] One study has shown that when shoulder instability is concomitant with impingement syndrome, it can cause significantly increased internal rotation. [31] However, the study was limited to radiographical findings, excluding clinical findings.

Length of inferior osteophyte
The cases with reported inferior osteophytes were small in number. However, where present, the size of the inferior osteophyte did not correlate with the impingement syndrome or the type of supraspinatus pathology. This appears to contrast with 1 study, which suggests that inferior osteophytes are among the primary causes of impingement. [4] The failure to detect the significant correlation is possibly owing to the small sample size of this relatively normal variant, and future studies are recommended to confirm or disconfirm an association.
Acromiohumeral distance correlated with supraspinatus pathology. The presence of 3 mm distance was 100% correlated with a complete tear. Cases with 6.1 to 9 mm distance had an equally high chance of partial and complete tear, and 9.1 to 12 mm distance had the highest association with tendinopathy. This acromiohumeral distance correlation is well known. [32] However, patients with decreased distance should be closely monitored postoperatively as they pose a higher chance of rotator cuff retear. [33] Further studies should be done to find the association of variants as a secondary cause of impingement that is not correlated with impingement syndrome and is correlated to supraspinatus pathologies-owing to the finding that supraspinatus is not correlated pathology is highly correlated with impingement.

Limitations
This study included all age groups; therefore, an actual risk factor related to age was not found. No randomization between the groups was done. Further, the failure to detect the significant correlation cold is because of the small sample size of variants such as OS acromiale, type 3, and 4 of the acromion. Also, the length of the inferior osteophyte in very few patients was measured.

Conclusion
This study found that females were more susceptible to shoulder impingement syndrome, supraspinatus tendinopathy, and supraspinatus tendon tear. The presence of supraspinatus tendon tear or tendinopathy was a high-risk factor for impingement. The prevalence of down-sloping in the presence of impingement syndrome was found statistically insignificant. Still, more particularly, a statistical significance was found in supraspinatus tendon tear or tendinopathy. The participants with no history of shoulder dislocation had a higher risk of impingement and tendinopathy-however, the less the acromiohumeral distance, the more significant the correlation with complete supraspinatus tear. The present study showed no correlation between ossification acromiale and impingement. Mild acromioclavicular degenerative changes were a risk factor for partial supraspinatus tendon tear, and moderate degenerative changes were a risk factor for complete tear. Calcific tendinitis and inferior osteophytes did not correlate with impingement or tendon pathology. Type of acromion was an associated risk factor, with type III being more susceptible to a partial tear. MRI scans showed a high sensitivity for detecting full-thickness supraspinatus tears.