The critical shoulder angle: can it be sufficient to reflect the shoulder joint without the humeral head?

We hypothesized that a new method considering the humeral head would distinguish rotator cuff tears (RCTs) and osteoarthritis (OA) better than the critical shoulder angle (CSA).

A total of 1011 patients were tested in this study and divided into 4 groups: those with RCTs (n = 493), those with OA (n = 73), those with anterior instability (n = 361), and those with adhesive capsulitis (n = 84). The CSA and new radiologic parameters including the humeral head were measured in the true anterior-to-posterior view: the Y angle connecting the lower end of the glenoid (LG), the center of the humeral head (CH), and the upper end of the glenoid (UG); the G angle connecting UG, CH, and the lateral tip of the acromion; the YG angle connecting LG, CH, and the lateral tip of the acromion; and the R angle connecting UG, LG, and CH.

The CSA and G angle were the largest in the RCT group (34.2° and 70.4°, respectively; P < .001) and the smallest in the OA group (29.8° and 61.7°, respectively; P < .001). The Y angle was the largest in the OA group (82.8°, P < .001). The R angle in the RCT group (52.9°) was significantly larger than that in the OA group, which was the smallest among the groups (48.0°; P < .001). The CSA was correlated with the G and YG angles in the RCT group, whereas the CSA was correlated with the Y, G, and R angles in the OA group (P < .05). The CSA showed the highest correlation with the size of RCTs (correlation coefficient = 0.138).

The Y, G, and R angles reflected the lesions of RCTs or OA. The CSA showed good correlations with the new radiologic parameters, and it had the highest correlation coefficient with the size of RCTs.