The aims of this study were to compare the ultrasonographic (US) thickness of the affected axillary recess capsule (ARC) with that of the unaffected ARC in patients with frozen shoulder (FS), to analyze whether the US measurements of the ARC thickness are correlated with those using MR imaging (MRI), and to assess whether the US thickness of the ARC is correlated with the patterns of range of motion limitation.

Forty-four patients with clinically diagnosed unilateral frozen shoulder and MRI evaluation performed US measurement of ARC. The US measurement of the ARC thickness was performed with the patients in a supine position with their shoulder abducted by 40 degrees. The ARC thickness was also measured by MRI on oblique coronal images by another physician blind to US measurements. With both US and MRI, ARC thickness was determined at the widest portion of the capsule (Fig. 1).

The US thickness of ARC was significantly higher in the affected shoulder (4.4±1.1mm) than in the unaffected one (2.2±0.5mm) (P<0.001) (Fig. 2).

The US thickness of the ARC in the affected shoulder was correlated with that measured by MRI (8.9±1.9mm) (P<0.001, r=0.83). The ARC thickness, whether measured by US or MRI, was not significantly related with the limitation of movement in specific directions.

US can demonstrate the difference in ARC thickness between the affected and unaffected shoulders in patients with unilateral FS. The ARC thickness measured by US is correlated with that measured by MRI.