Shoulder adhesive capsulitis: epidemiology and predictors of surgery

Abstract

Adhesive capsulitis is characterized by a gradual, painful loss of shoulder motion. This study evaluated patient variables significantly associated with developing adhesive capsulitis compared with a sex-matched control group without adhesive capsulitis. We also sought to determine those factors associated with adhesive capsulitis patients requiring surgical intervention. All patients presenting to our hospital with adhesive capsulitis between 2004 and 2014 were identified. Demographic characteristics were collected, and a sex-matched control group was randomly generated from the electronic medical record and used for comparison. Patients who underwent surgical intervention for adhesive capsulitis were also identified, and factors associated with surgical intervention were identified with logistic regression analysis. Included were 2190 adhesive capsulitis patients with a normal age distribution of 56.4 ± 13.1 years. Most were classified as overweight (30.7%) or obese (27.0%). Compared with controls, adhesive capsulitis patients were more likely to be younger (<50 vs. ≥50 years; odds ratio [OR], 0.69; P < .001), obese (OR, 1.26; P < .001), black/African American (OR, 1.71; P < .001), Hispanic/Latino (OR, 4.85; P < .001), or diabetic (OR, 1.12; P < .001) and less likely to have hypertension (OR, 0.93; P = .006). Overall, 361 subsequently underwent surgical intervention. Older patients, racial minorities, and government-sponsored/uninsured patients were significantly less likely to have surgery for adhesive capsulitis (P < .01), whereas workers' compensation patients were 8 times more likely to receive surgery compared with privately insured patients (P < .001). Obesity and diabetes were significantly associated with adhesive capsulitis and should be considered modifiable patient factors. Additionally, younger patients and racial minorities were also significantly more likely to be diagnosed with adhesive capsulitis. Younger, white, and workers' compensation patients were more likely to receive surgery, whereas patients with government-sponsored or no insurance status were more likely to receive nonoperative treatment.