Implant characteristics affect in vivo shoulder kinematics during multiplanar functional motions after reverse shoulder arthroplasty.

The purpose of this study was to determine how implant characteristics affect in vivo shoulder kinematics after reverse shoulder arthroplasty (RSA). Kinematics of the affected upper limb were measured in 32 participants during five motions (scapular plane abduction, hand-to-head, hand-to-back, internal/external rotation at 90° abduction, and circumduction) using optical motion capture. Shoulder abduction, plane of elevation, and internal/external rotation range of motion (ROM), peak angles, and continuous kinematics waveforms were calculated for each motion. Multiple regression was used to identify associations between kinematics and implant characteristics of lateralization, humeral retroversion, glenosphere size, glenosphere tilt, glenoid eccentricity, and implant neck-shaft angle (135° or 145°). Less humeral retroversion was associated with greater shoulder rotation ROM (p = 0.036) and greater plane of elevation ROM (p = 0.024) during circumduction, while less eccentricity was associated with more posterior plane of elevation during hand-to-back (p = 0.021). The 145° implant was associated with greater internal/external shoulder rotation ROM (p < 0.001), greater internal shoulder rotation (p = 0.002), and greater plane of elevation ROM (p = 001) during the hand-to-back. The 145° implant was also associated with more internal/external rotation ROM (p = 0.043) during shoulder rotation and more abduction ROM during circumduction (p = 0.043). During the hand-to-back motion, individuals having 135° neck-shaft angle implants were more abducted from 21 to 51% of the motion and were less internally rotated from 70 to 100% of the motion, while more lateralization was associated with less internal rotation from 90 to 100% of the motion. Retroversion and implant neck-shaft angle are the primary implant characteristics associated with in vivo shoulder kinematics during complex motions after RSA.