Our aim was to compare the outcome of arthroscopic release for frozen shoulder in patients with and without diabetes. We prospectively compared the outcome in 21 patients with and 21 patients without diabetes, two years post-operatively. The modified Constant score was used as the outcome measure. The mean age of the patients was 54.5 years (48 to 65; male:female ratio: 18:24), the mean pre-operative duration of symptoms was 8.3 months (6 to 13) and the mean pre-operative modified Constant scores were 36.6 (standard deviation (sd) 4.6) and 38.4 (sd 5.7) in the diabetic and non-diabetic groups, respectively. The mean modified Constant scores at six weeks, six months and two years post-operatively in the diabetics were 55. 6 (sd 4.7), 67. 4 (sd 5.6) and 84. 4 (sd 6.8), respectively; and in the non-diabetics 66.8 (sd 4.5), 79.6 (sd 3.8) and 88.6 (sd 4.2), respectively. A total of 15 (71%) of diabetic patients recovered a full range of movement as opposed to 19 (90%) in the non-diabetics. There was significant improvement (p < 0.01) in the modified Constant scores following arthroscopic release for frozen shoulder in both groups. The results in diabetics were significantly worse than those in non-diabetics six months post-operatively (p < 0.01) with a tendency towards persistent limitation of movement two years after operation. These results may be used when counselling diabetic patients for the outcome after arthroscopic treatment of frozen shoulder.

Cite this article: Bone Joint J 2014;96-B:1355–8.