Investigating the impact of Glucagon-like peptide-1 therapy on total shoulder arthroplasty outcomes

PURPOSE: The increasing prevalence of Glucagon-like peptide-1 receptor agonists (GLP-1 RA) has raised concerns about their perioperative impact. Although database studies on GLP-1 RAs in total shoulder arthroplasty (TSA) have been published for diabetic and general TSA patients, results are conflicting, and the effects on patients across various obesity classes remain unexplored. This study aims to fill this gap. METHODS: The PearlDiver database was used to identify patients who underwent primary TSA, grouped by GLP-1 use, obesity status, and obesity class. Post-operative outcomes included 90-day readmission and emergency department (ED) visit rates, medical complications (including thromboembolic, infectious, and renal events), and orthopaedic complications (including revision, periprosthetic joint infection, and instability). Cohorts were matched 10:1 by exact matching on age, sex, Elixhauser comorbidity index, diabetes, hypertension and overall obesity status, and multivariable regression was used to adjust for comorbidities that remained imbalanced after matching. RESULTS: Among 95,566 patients, 1,549 (1.62%) were GLP-1 users. Initially, GLP-1 users exhibited higher pre-operative comorbidity burdens and rates of complications. However, after matching and multivariable adjustment for residual imbalance, no significant differences in medical or orthopaedic outcomes persisted. Notably, GLP-1 users demonstrated significantly lower odds of 90-day ED visits (OR 0.75; P < .01) and readmissions (OR 0.71; P < .01). Similar trends were observed within each individual obesity class, but none reached statistical significance, most likely reflecting the small size of these subgroups. CONCLUSION: After accounting for pre-operative comorbidities, our findings suggest that GLP-1 RA use is not associated with worse surgical outcomes in TSA patients and may be associated with lower post-operative healthcare utilisation, although residual confounding cannot be excluded. Further longitudinal studies are warranted to optimise perioperative GLP-1 RA administration.