The purpose of this work was to propose an objective radiographic evaluation of the antero-inferior gleno-humeral ligament for comparison with the clinical assessment proposed by Gagey.

A test radiogram was obtained from 32 healthy volunteers (15 men, 17 women, mean age, 29 years, age range 21-54 years) free of shoulder disease. The dynamic test image consisted in a strictly AP view of the shoulder in forced abduction in neutral rotation. The subjects were awake and in dorsal decubitus. Displacements of the scapulo-thoracic articulation were limited by a counter-force applied to the acromion, in accordance with the method described by Gagey. Bilateral images were obtained for comparison. Several angles were measured between the humeral shaft and the scapula to search for the most reliable and reproducible measurement.

Three series of angles were measured between the axis of the humeral shaft and the scapula. The mean angle between the axis of the humeral shaft and line drawn from the lower rim of the glenoid cavity to the lateral border of the scapular tubercle was 130.3° (range 110-148°) on the dominant side and 131.5° (108-148°) on the non-dominant side; giving 38° variability on the dominant side and 40° variation on the non-dominant side and a standard deviation of 10.4° on the dominant side and 11.5° on the non-dominant side. The mean difference in gleno-humeral abduction was 3.8° (range 0-14°) between the dominant and non-dominant side.

Among the different angles measured between the scapula and the humerus, the angle between the axis of the humeral shaft and the line drawn from the lower rim of the glenoid cavity to the lateral border of the scapular tubercle was the most reliable and reproducible. Inter-observer measurements were well correlated. We observed that the variability in the radiographic values of the scapulo-humeral angle was much greater than the clinical values described by Gagey who, finding very constant values during forced abduction, described “invariable” scapulo-humeral abduction of the shoulder. Our study demonstrates that scapulo-humeral abduction is not an invariable parameter. More interestingly, the difference in amplitude between the dominant and non-dominant sides showed very strong interindividual correlation. Interobserver variability was low and reproducibility was good.

This dynamic radiographic test enables a precise quantified assessment of pure gleno-humeral abduction which depends on the antero-inferior gleno-humeral ligament. This test is reliable and reproducible. Variations in the length of the antero-inferior gleno-humeral ligament evaluated radiographically were greater than described clinically. We did not find any difference in pure gleno-humeral abduction greater than 14° between the dominant and non-dominant sides in healthy subjects.