Deltoid fatigue: a longitudinal assessment of reverse shoulder arthroplasty over time

Background Studies evaluating the mid-term performance of reverse shoulder arthroplasty (RSA) have identified a drop in the Constant-Murley score between 6 and 8 years after surgery, which is most affected by a loss of forward elevation and strength. Alterations of the deltoid length and moment arm after RSA lead to nonphysiological stress on the deltoid muscle. Concern has arisen that the long-term implications of increased deltoid work may be causing “deltoid fatigue.” The purpose of this study was to evaluate the long-term effects of RSA on overhead range of motion (ROM) and validate the hypothesis of deltoid fatigue. Methods We performed a retrospective review of 165 RSAs over a 5-year period. Diagnoses were limited to cuff tear arthropathy, osteoarthritis with rotator cuff deficiency, and irreparable rotator cuff tear. All procedures were performed using a single implant system. Patients were evaluated longitudinally at multiple time points. They were required to undergo a minimum of 3 follow-up visits, with at least 1 visit at >5 years. ROM and patient-reported outcome measures were evaluated using linear mixed models for repeated measures to evaluate changes in outcome measures over time. A secondary analysis was performed to assess the influence of patient demographic factors on observed changes in ROM and patient-reported outcome measures. Results Primary RSA shoulders were observed to lose 0.8° of forward elevation and abduction per year starting at 1 year postoperatively (P = .006), without a significant drop at mid-term follow-up. No significant change in external or internal rotation was observed. Male patients and patients with a diagnosis of osteoarthritis with rotator cuff deficiency showed greater baseline overhead ROM at 1 year postoperatively, but the subsequent rates of functional decline were similar regardless of age, sex, or indication. Discussion This study challenges the previous theory of deltoid fatigue resulting in a significant loss of overhead ROM beginning 6-8 years after index arthroplasty. However, a slower progressive decline in overhead ROM in well-functioning RSA shoulders was observed, averaging 0.8° of overhead ROM per year. This progressive deterioration occurs at a slightly greater rate than that observed in the natural shoulder. The observed rate of functional decline was found to be independent of age, sex, and preoperative diagnosis.