In 1999, we performed an anatomical morphometric study on a large number of coracoid processes in dry scapulae and a CT morphometric study on some of the specimens [6]. The aim is to determine if anatomical variations of the coracoid process might predispose to the idiopathic impingement syndrome. Gender and side of the scapulae were not taken into account, since these factors did not appear to influence the anatomical features of the coracoid process [7]. A total of 204 scapulae (76 right and 128 left) entered into the study; most (82 %) were from cadavers with the presumed age between 30 and 60 years. Sixty-eight pairs of scapulae were from the same skeletons. None of the scapulae showed degenerative changes of the coracoid process.

We found no significant correlations among length or thickness of the coracoid process, coracoglenoid distance, coracoid slope, and prominence of the coracoid tip beyond the anterior edge of the glenoid. Furthermore, no correlations were found between these anatomical features and the dimensions of the whole scapula as revealed by the length-width ratio. Since the anatomical characteristics, which were found to be extremely variable, cannot be precisely evaluated in radiographs of the shoulder, no information on the dimensions of the coracohumeral space can be obtained by radiographic studies.

The anterosuperior border of the glenoid and the posterolateral margin of the coracoid process enclose an arch-like space of varying shape. We found out that this space may have three configurations: In Type I this space had a “round bracket” configuration, whereas in Type II and III it had the shape of a “square bracket” and a “fish hook,” respectively. Type I configuration was observed in 45 % of scapulae, and Type II and III configuration in 34 % and 21 % of specimens, respectively. Subacromial space presents the shortest dimensions in the Type I scapulae. Furthermore, the scapulae with this configuration were found to have, on CT, the lowest mean values of the coracoglenoid angle (formed by the line originating from the anterior glenoid rim and running tangentially to the most prominent part of the tip of the coracoid; and the line passing through the anterior and posterior glenoid rims) [7] and coracoid overlap (distance by which the coracoid process overlaps the plane of the glenoid fossa) [7]. It has been shown that values of these parameters are related to the width of the coracohumeral space, since it was found to be small in the presence of small coracoglenoid angles and small coracoid overlap [7]. Therefore, Type I configuration implies a short coracohumeral distance.