Summary: While most studies describe shoulder asymmetry as difference in height of shoulders, these findings suggest there are 2 distinct regions (lateral and medial) of shoulder height asymmetry.
Introduction: Shoulder balance is often described as “level” shoulders. We hypothesize that there are 2 distinct features of shoulder asymmetry (Clavicle Tilt and Trapezial Prominence) that correspond to different underlying radiographic parameters.
Methods: A consecutive series of 113 pre‐operative patients with right main thoracic Lenke 1 and 2 AIS curves were investigated to evaluate the correlations between clinical and radiographic findings of shoulder imbalance. The following parameters were defined and evaluated from pre‐op clinical photographs: clavicle angle (tilt), trapezial angle, ratio of left to right trapezial area (Figure). These were compared to radiographic measures of T1 tilt, first rib angle, magnitudes of curves, and thoracic and lumbar apical translation to C7 plumb line and CSVL.
Results: The average age of patients was 14.1 yrs. There were 82 Lenke 1 and 31 Lenke 2 curves with an average thoracic Cobb of 52.1 degrees. The clinical clavicle angle ranged from ‐10.7 to +6.0 degrees (+ = high on left) and had the modest correlation with the proximal thoracic curve size, T1 tilt, coronal balance, and thoracic apical deviation (r value, 0.43, 0.42, 0.41 and 0.41, respectively). In contrast, medial shoulder trapezial prominence as measured clinically by both the trapezial angle (range: ‐12.7 to +6.6) and trapezial area ratio (natural log of ratio range; ‐1.15 to +0.80) correlated well with the radiographic measures of T1 tilt and the 1st rib angle (trap angle: 0.70 and 0.68; trap area ratio: 0.58 and 0.60, respectively).
Conclusion: Shoulder imbalance is often reported as differences in the “height” of the shoulders. Our analysis suggests there are 2 distinct regions (lateral and medial) of shoulder height asymmetry. Medial differences reflected in trapezial prominence relate principally to deformity created by upward tilted proximal ribs and thus T1 tilt. While, lateral differences in shoulder symmetry as reflected in the clavicle angle correlate only weakly with radiographic measures of spinal deformity. This suggests correction of trapezial prominence may be more predictable compared to clavicle angulation following scoliosis surgery.
Measures of shoulder imbalance on clinical photos