BACKGROUND: The incidence of adhesive capsulitis (AC) is higher in patients with diabetes mellitus. While AC is usually treated nonoperatively, diabetic patients are more likely to require more extensive treatments such as manipulation under anesthesia or arthroscopic capsular release. Despite the recent surge in popularity of GLP-1 agonists ("GLP-1s") for the treatment of type 2 diabetes (T2DM), there is a lack of literature describing the effect of GLP-1 use on the incidence and management of AC in patients with T2DM. We aimed to compare T2DM patients with and without GLP-1 use, and determine differences in rates of AC, rates of operative AC management, and rates of 90-day postoperative complications among operatively managed AC patients. METHODS: A retrospective cohort analysis of the PearlDiver database from 2010 to 2022 was performed to identify T2DM patients with AC. Exclusion criteria included patients <
18 year old, type 1 diabetes, <
2-year follow-up, and medical conditions that were a contraindication for GLP-1 use. Four of the following cohorts were identified: 1) T2DM taking GLP-1s, 2) T2DM not taking GLP-1s, 3) T2DM with AC taking GLP-1s, and 4) T2DM with AC not taking GLP-1s. The primary outcome was the proportion of T2DM patients who developed AC. Secondary outcomes included rates of operative management and 90-day postoperative medical complications. A 1:1 propensity score match was performed, controlling for age, gender, Charlson Comorbidity Index, obesity, tobacco use, hypothyroidism, metformin use, insulin use, and the presence of complicated diabetes. Chi-square and Multivariable Linear Regression analyses were performed. RESULTS: 100,000 T2DM patients taking GLP-1s and 100,000 T2DM patients not taking GLP-1s were randomly sampled. After matching, GLP-1 patients were more likely to develop AC (odds ratio [OR] = 1.28, P <
.001). 253,717 T2DM patients with AC were then identified, of which, 46,156 (18.2%) were taking GLP-1s. After matching, GLP-1 users had higher odds of requiring operative management for AC compared to non-GLP-1 users (OR = 1.18, P <
.001). The odds of undergoing manipulation under anesthesia were significantly higher among GLP-1 users (OR = 1.20, P <
.001)
however, there was no difference in the odds of undergoing capsular release (OR = 0.68, P = .29). DISCUSSION AND CONCLUSIONS: Among a national cohort of T2DM patients, patients taking GLP-1 agonists had higher odds of developing AC and requiring operative management for AC when compared to those not taking GLP-1s. These results can be useful when counseling T2DM patients who develop AC while taking GLP-1s to better inform these patients of the increased chance of requiring more extensive treatment.