Critical shoulder angle

Critical shoulder angle (CSA) is a parameter that, in some studies ¹, correlates with rotator cuff tears (RCT) or glenohumeral osteoarthritis (OA). Other studies have shown that CSA is neither associated with RCT ^5,6 nor OA at 20-year follow-up ⁷. 

The CSA measures acromial cover and the inclination of the glenoid, integrating both anatomical risk factors for RCT and OA into one biomechanical parameter ¹. Moreover, the mean CSA, in combination with age and trauma, has also been demonstrated to predict the integrity of the posterosuperior rotator cuff >.

In a larger CSA, the force vector is mostly dependent on the elevating force of the deltoid muscle, which is directed upward against the rotator cuff, leading to cuff degeneration, tears, and potentially eccentric osteoarthritis ³. The resulting force vector of the deltoid muscle is more balanced on the glenoid surface when the CSA is ~33°, with a potential protective effect against degenerative shoulder conditions. In patients with a smaller CSA, the resulting force vector is unbalanced against the glenoid, favoring concentric OA ⁴. 

Method

CSA is the angle created between the superior and inferior bone margin of the glenoid and the most inferolateral border of the acromion. It is measured on an AP shoulder radiograph ¹.

Interpretation

• normal range: 30-35° ⁸

• increased CSA: >35° ⁸

  • associated with rotator cuff tears (sensitivity 82%; specificity 92%) ^ref

• decreased CSA: <30° ⁸

  • associated with primary concentric glenohumeral OA (sensitivity 78%; specificity 97%) ^ref

  • in patients with anterior glenohumeral instability, associated with type V SLAP lesions (sensitivity 75%, specificity 77%) ⁸

Practical points

• CSA has been shown to be insensitive to minor rotations of the scapula

• scapula malrotation >20°, characterized by an oval glenoid, leads to substantial overlap between the anterior and posterior glenoid rims, and the reproducibility of the CSA decreases ¹

• glenoid morphology, including Walch type A1 through B2, have been shown to have no significant difference in the mean CSA ^ref