The effect of body mass index on internal rotation and function following anatomic and reverse total shoulder arthroplasty

Background: The exact relationship between body mass index (BMI) and internal rotation (IR) before and after total shoulder arthroplasty has not been studied to date. The purpose of this study was to determine the effects of BMI on the preoperative and postoperative shoulder range of motion and function in anatomic (aTSA) and reverse total shoulder arthroplasty (rTSA), and specifically how IR affects patient ability to perform IR-related activities of daily living (ADLs). Methods: Patients from a prospective multicenter international shoulder arthroplasty registry who underwent primary rTSA (n=1171) and primary aTSA (n=883) were scored preoperatively and at latest follow-up (2-10 years, mean = 3 years) using the Simple Shoulder Test, University of California–Los Angeles shoulder score, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form, Constant score, and Shoulder Pain and Disability Index patient-reported outcome measures (PROMs). Measured active abduction, forward flexion, IR, and active and passive external rotation were recorded, and BMI was evaluated as a predictor of motion and patient-reported outcomes. Patient responses to questions regarding the difficulty level of IR-related ADLs were studied. The relationships between BMI, IR, and ability to perform IR-related ADLs were quantified through analysis of variance with post hoc comparisons by Tukey honestly significant difference tests, where significance was denoted as P <.05. Results: BMI was found to be inversely correlated with IR in patients undergoing both aTSA and rTSA, both preoperatively (P <.001 and P =.002) and postoperatively (P <.001 and P <.001). BMI affected the range of motion parameters of forward flexion abduction and external rotation but to a lesser extent than that of IR. Nonobese patients demonstrated significantly greater IR than overweight, obese, and morbidly obese patients postoperatively for aTSA (P <.001). For rTSA, nonobese patients had a significantly greater postoperative IR than obese and morbidly obese patients (P <.001 and P =.011, respectively). For both aTSA and rTSA patients, mean IR scores significantly differed between patients reporting normal function vs. patients reporting slight difficulty, considerable difficulty, or inability to perform IR-related ADLs. Increasing IR demonstrated a significant, positive correlation with all PROMs for both aTSA and rTSA patients (Pearson correlation, P <.001). Conclusions: BMI is an independent predictor of IR, even when controlling for age, gender, glenosphere size, and subscapularis repair. BMI was inversely correlated with the degree of IR, and decreased IR significantly negatively affected the ability to perform IR-related ADLs. Clinical relevance: Increasing BMI adversely affects shoulder ROM, particularly IR. IR is correlated with the ability to perform ADLs requiring IR in both aTSA and rTSA patients.