Analyzing glenohumeral torque-rotation response in vivo.

Background: Because the human shoulder has many degrees of freedom that allow redundant means of producing the same net humerothoracic motion, there are many impediments to objective, repeatable assessment of shoulder function in vivo. Devices designed to date have suffered from poor reliability. In this study we introduce a new device and methods to evaluate human shoulder kinematics and evaluate its reproducibility from subject to subject and from day to day.

Methods: This was a controlled laboratory study. Using electromagnetic motion sensors to record the position and orientation of the thorax, scapula, and humerus, we quantified the kinematic response of twenty four normal shoulders in response to known internal-external torque application. A four-parameter logistic function was selected to characterize the strident features of the torque-rotation relationship.

Findings: Our analysis in conjunction with the measurement technique described herein, allowed the passive glenohumeral internal-external range of motion to be differentiated from other motion components and was determined to within 9.6% of full scale over three repeated trials. Range of motion was the most reliable biomechanical outcome, more so than computed indices of glenohumeral flexibility and hysteresis. The exact profile of the torque-rotation response, and therefore the repeatability of the calculated outcomes, was unique from shoulder to shoulder.

Interpretation: The development of the capacity for precise, non-invasive measurement of shoulder biomechanics over time is a requisite step towards optimizing treatment of shoulder injury and disease. Our current methods are superior to previous attempts at trying to non-invasively evaluate the biomechanics of the glenohumeral joint.