Biomechanical mechanisms of rotator cuff deformation during a simulated Volleyball spike

Abstract

Background: Shoulder pain accounts for hundreds of thousands of surgeries each year and contributes to enormous healthcare spending. Rotator cuff disorders are the primary driver of physician visits for shoulder pain. Despite this, consistently successful treatment for rotator cuff disorders has been elusive. Rotator cuff repair demonstrates high failure rates and recurrent or ongoing pain is common with nonoperative management including physical therapy. Taken together this suggests a lack of complete understanding of the biomechanical demands on the rotator cuff and the role of complex mechanical stress and strain on rotator cuff disorders. Volleyball players’ spiking shoulders are exposed to repetitive spike motions in high humeral elevation angles known to contribute to rotator cuff compression. Additionally, they experience large angular velocities at extremes of axial rotation during arm cocking and acceleration—positions commonly linked to various glenohumeral joint pathologies. Sports biomechanics research of the shoulder is dominated by optical, and to a lesser extent, electromagnetic motion capture which are limited by skin motion artifact and often ignore contributions of the scapula. Deformation of the rotator cuff tendons is related to the position and orientation of their bony origins and insertions—information lost without accurate scapular kinematics. Biplanar videoradiography is a precise tool that has been successfully employed to describe glenohumeral kinematics and elucidate normal mechanisms of rotator cuff compression during reaching tasks. However, high-speed, multiplanar shoulder movements have not been pursued due to limitations in capture frequency and field of view. Detailed glenohumeral kinematics during the multiplanar volleyball spike would provide unique information on normal deformation of commonly injured rotator cuff tendons and critical boundary conditions to constrain future modeling studies of rotator cuff behavior which is infeasible to measure directly in vivo. Objectives: The objectives of this dissertation were three-fold: (1) to determine the effect of velocity control on humerothoracic kinematics of a simulated volleyball spike; (2) to describe glenohumeral kinematics of a simulated volleyball spike; and (3) determine the extent to which foregoing kinematics contribute to compressive deformation of the rotator cuff as compared to scapular plane abduction. Methods: The objectives of this dissertation were approached through two separate studies. First, I needed to determine the feasibility of capturing high-speed, multiplanar glenohumeral kinematics within the constraints of our custom biplanar videoradiography system. To do this, I chose to examine if humerothoracic kinematics during the arm cocking phase of high-speed spiking were similar to those observed at reduced velocities. Optical motion capture was used to obtain humerothoracic kinematics from 11 recreational volleyball players without shoulder pain during simulated volleyball spikes at high speed and under two velocity conditions regulated by a metronome. Elevation, plane of elevation, and axial rotation were extracted at the instant of maximum external rotation and were compared within-subjects across conditions using a repeated measures analysis of variance. A custom biplanar videoradiography system was then employed to assess glenohumeral kinematics of 21 competitive volleyball players during scapular plane abduction and a simulated volleyball spike. Subject-specific bone models were segmented from CT scans of the scapula and humerus. 2D-3D shape matching was employed to iteratively match digitally reconstructed radiographs of the scapula and humerus to biplanar image projections for each frame. Glenohumeral joint rotations, translations, combined supraspinatus/infraspinatus tendon footprint-to-acromion (subacromial), and tendon footprint-to-glenoid (internal) proximities were extracted and described for scapular plane abduction and a simulated volleyball spike. Lastly, glenohumeral kinematics and tendon footprint-to-bone proximities were isolated at the instant of arm cocking of the simulated volleyball spike and compared to scapular plane abduction at an equivalent elevation angle using paired t-tests. This is the first study to employ biplanar videoradiography to directly visualize the GH joint during a dynamic sporting activity. The additional GHER that occurs during late cocking as compared to SAB at an EEA appears to convey a protective mechanism against SA impingement. Results: For the surface motion capture study there was no significant difference in mean humerothoracic elevation (F1.13,10.2 = 0.68, p = 0.45) nor plane of elevation (F2,20 = 0.25, p = 0.78) at the instant of arm cocking across velocity conditions. Mean maximum humerothoracic external rotation was significantly different across velocity conditions (F1.29,12.86 = 5.44, p = 0.03). After adjusting for multiple comparisons, mean maximum ER was significantly less for the 75 BPM (mean maxER = 87.4°) as compared to the 175 BPM and full speed conditions (mean maxER = 100.1° and 107.3°; p = 0.011 and p = 0.04) respectively. In the biplanar videoradiography study, glenohumeral external rotation (ER) was significantly greater (mean difference = 28.94°, p < 0.001) and subacromial proximity was significantly larger (i.e. more space) during arm cocking of the simulated spike as compared to scapular plane abduction (mean difference = 3.12 mm, p < 0.001) at an equivalent elevation angle. Internal impingement proximity (mean difference = -1.94 mm, p = 0.15), plane of elevation (mean difference = -11.27°, p = 0.07), and glenohumeral SI, AP, and ML translations were not significantly different between tasks (Superoinferior mean difference = -0.007 mm, anteroposterior mean difference = 0.293 mm, and medial-lateral mean difference = 0.002 mm; p = 0.98, p = 0.46, and p = 0.99 respectively). Conclusions: Feasibility of using biplanar videoradiography to capture fast speed glenohumeral kinematics needed to be established to justify its associated radiation exposure to participants. I found that humerothoracic elevation and plane of elevation angles at arm cocking were consistent across fast-speed and two velocity-controlled conditions; however, axial external rotation was underrepresented at slower speeds. I felt these results supported the use of biplanar videoradiography for more precise kinematic characterization of the glenohumeral joint. Specifically, elevation angle was unchanged across velocity conditions and it is an established kinematic contributor to rotator cuff compression. Researchers should be aware that glenohumeral external axial rotation may be underrepresented in velocity-controlled conditions. Clinically, movement velocity control via metronome is a low-cost tool to provide progressive exposure to increasing external rotation during simulated spike tasks. This is the first study to employ biplanar videoradiography to directly visualize the glenohumeral joint during a dynamic sport activity. Both tasks involved periods of time where most participants were at risk of both subacromial and internal rotator cuff compression. The additional glenohumeral external rotation during the arm cocking phase of a simulated spike may convey a protective mechanism against subacromial compression as compared to scapular plane abduction.