This cross-sectional observational study is a level 3 study. The thickness of CHL and CHD, measured by USG, were compared between SIS patients with and without SSC lesion. This study was approved by University Hospital Ethical Review Board for Medical Research Involving Human Subjects, and each participant gave his/her informed consent.
Subjects
Based on previous studies, a total sample size of 40 participants was calculated to provide 80% power with detection of a difference of more than 1.0 mm CHD between 2 groups [7]. The inclusion criteria of the participants were age of 20–60 years old and positive unilateral shoulder results on at least 3 of 5 tests: 1) Neer's test, 2) Hawkins' test, 3) the Empty can test, 4) the pain or weakness with resisted ER test, and 5) tenderness in the tendon of the rotator cuff [14, 15]. Participants with a history of shoulder dislocation, fracture or surgery, history of direct contact injury to the neck or upper extremities within the past month, glenohumeral joint instability (positive apprehension test, sulcus sign), neurologic disorder (upper motor neuron diseases, cervical radiculopathy), passive ER ROM < 30°, or pain (visual analogue scale, VAS > 5) during the experimental tasks were excluded. After the impingement tests were performed to ensure that the participants met our inclusion criteria, all participants were assessed with 3 special SSC tests, namely, the lift-off test, belly-press test (Napoleon sign) and bear-hug test [16, 17] for group allocation. Patients with positive results on at least 2 of the 3 special tests were allocated to the SSC lesion (SSCL) group. SSC tears and cuff lag signs were also excluded as they may falsely affect the CHD due to resting anterior positioning of the humerus. To clarify SSC lesion without tears, integrity of the biceps pulley was confirmed by USG.
Instrumentation and procedures
The T3300 ultrasound system (BenQ, Taipei, Taiwan), a portable ultrasonography machine, was used to assess the thickness of the CHL, SSC tendon, and supraspinatus (SSP) tendon, as well as the CHD and acromiohumeral distance (AHD). A L154BH linear array probe with a range of 4–15 MHz was used to collect data. A B-mode USG with preset musculoskeletal (MSK) examining parameters (gain, 64; dynamic range, 75; QScan, 4; persistence, 3; gray map, 3; chroma map, 0; and steering angle, 0) was used to collect data. Depending on the depth of the structure we wanted to measure, tissue harmonic imaging (THI) was turned on at 5.5MHz for better resolution and turned off at 12MHz for better penetration. Previous studies have shown relatively higher accuracy and reliability in measurement of shoulder pathologies [18.19].
The experimental procedure is illustrated in Fig. 1. The characteristics of the participants were collected by one assessor, including age, gender, height, weight, dominant side, involved side, duration of symptom, pain (VAS), occupation ratio, Flexilevel scale [20] and internal rotator strength.
USG measurements
We measured the following outcomes: (1) CHL thickness, (2) CHD, (3) SSP thickness, (4) SSC thickness and (5) AHD. Each outcome was measured in 3 trials and the mean of the 3 trials was used for data analyses. All of the measurements were measured with THI turned on except for that of CHD, due to the deeper anatomical structure.
For measurement of the CHL thickness, the position of the linear probe was on the lateral border of the coracoid process to obtain a longitudinal image of the CHL. Each participant was instructed to lie in supine position and relax while the examiner maintained the elbow of the participant at flexion of 90° and the shoulder under maximal ER without shoulder abduction or flexion (arm by side) (Fig. 2). Maximal ER of the shoulder was achieved when the examiner could not further externally rotate the shoulder of the participant. The thickness of the CHL at a 2-mm distance from the coracoid process was measured. The intraclass correlation coefficient (ICC) of CHL thickness was 0.854 with 0.2 mm standard error of measurement (SEM).
CHD was measured with the probe positioned on the lateral border of the coracoid process to obtain images of the coracoid process and humeral head in 4 different shoulder rotation positions: (1) shoulder neutral rotation (CHD-NR), (2) external rotation (CHD-ER) and (3) shoulder internal rotation with maximal forward flexion and full adduction (CHD-IRFA, with the arm adducted across the chest reaching for the opposite shoulder) and (4) shoulder internal rotation (CHD-IR) (Fig. 2) [7, 19]. Participants were asked to sit with their arms by their sides and to perform the 4 different positions respectively. The measurements were repeated for 3 trials with repositioning of the arm to a neutral position for intervals of 10 seconds. The distance measured was that between the coracoid process and the lesser tuberosity of the humerus. The ICCs of CHD were 0.996, 0.969, 0.893 and 0.930 with 0.1 mm, 0.3mm, 0.5 mm and 0.4 mm SEMs respectively.
SSP tendon thickness was evaluated with the patient's palm placed over his/her iliac wing, or “back pocket”, with the elbow flexed and directed medially [21]. The transducer was placed over the anterior aspect of the shoulder, perpendicular to the supraspinatus tendon and just anterior of the anterior–lateral margin of the acromion. A transverse glide was then performed at the site to determine the exact position where the observer judged that the tendon thickness was at its maximum. The thickness of the SSP tendon was measured 2 cm away from the biceps long head tendon [22]. For measuring SSC tendon thickness with a short axis of view, the probe was positioned horizontally on the bicipital grove. The participant’s forearm was placed with the elbow flexed to 90° in slight internal rotation, with the palm facing upward and medially. Then the patient was asked to rotate the forearm externally, keeping the palm up and the elbow strictly close to the iliac crest. The ICC of SSP tendon thickness was 0.943 with 0.2 mm SEM.
AHD was measured under 0° and 60° of scapular plane shoulder elevation. The transducer was placed on the most anterior aspect of the acromion edge, with the long axis of the transducer placed in the plane of the scapula and parallel to the flat surface of the acromion [21]. The participants sat in an upright position without back support and with their feet flat on the floor, holding their shoulders back and looking straight ahead, to achieve retracted shoulders and extension in the thoracic and cervical spine. AHD was measured at 0° and 60° of active shoulder elevation in the scapular plane. A 60° scapular elevation of AHD was then measured with a goniometer placed on the patient's arm to determine 60° of active shoulder elevation, with the thumb pointing up. The ICCs of AHD were 0.963 and 0.982 with 0.1 mm and 0.1 mm SEMs respectively.
Statistical analysis
All data were analyzed in SPSS 25.0 (IBM, USA), with an α level 0.05. The Shapiro–Wilk test was used for the test of normality. Demographic data were calculated by independent t-test for normally distributed data or Mann–Whitney U test for non-normally distributed data. Correlation of CHL and CHD was calculated by Pearson correlation coefficient for normally distributed data or Spearman correlation coefficient for non-normally distributed data. Group differences in CHL and CHD were tested by independent t-test for normally distributed data or Mann–Whitney U test for non-normally distributed data.