Effects of glycemic control on prevalence of diabetic frozen shoulder

There is controversy regarding the influence of glycemic control in diabetic patients with frozen shoulder. To determine the relationship between glycemic control and the prevalence of frozen shoulder in diabetic patients, we hypothesized that increased glycosylated hemoglobin A1c (HbA1c) levels would correlate with an increased prevalence of frozen shoulder.

A retrospective analysis with statistical review of 201,513 diabetic patients enrolled in a regional health maintenance organization in 2007 was performed. Analysis included determining the relationship between the prevalence of frozen shoulder and the following factors: HbA1c level, type of diabetes treatment, duration of diabetes treatment, and presence of end-stage diabetic manifestations.

There were 1150 diabetic patients with a diagnosis of frozen shoulder. There was no significant relationship between HbA1c level and the prevalence of frozen shoulder. Insulin-dependent patients who used or did not use oral hypoglycemics were 1.93 times more likely than non-insulin-dependent diabetic patients to have frozen shoulder, and that rate increased to 1.96 times more likely when the results were adjusted for HbA1c level. Patients who were taking oral hypoglycemic drugs were 1.5 times more likely to develop frozen shoulder than those who did not use insulin or take oral hypoglycemic drugs. Duration of diabetes was also associated with the development of frozen shoulder, after controlling for insulin use (odds ratio: 1.85 for duration of more than ten years of use compared with less than five years of use). The prevalence of end-stage diabetic manifestations was increased in patients with frozen shoulder as compared with those without frozen shoulder (p < 0.0001).

There was no association found between HbA1c level and the prevalence of frozen shoulder in this diabetic population.