A separation between the labrum and the glenoid is often observed clinically. We carried out an anatomical study on the attachment of the labrum to the glenoid in order to investigate whether the separation is a pathological condition or a normal variation. One hundred shoulder joints taken from 50 cadavers, aged between 48 and 98 years, were used in this study. Firstly, the joints were investigated macroscopically. Then, histological sections were made for further detailed investigation with a microscope. Macroscopically, the manner of attachment was classified into two types: Type I; absence of a separation, Type II.; presence of a separation. Separations were frequently observed in the upper part of the glenoid, while scarcely in the lower part. In cases where a separation was found, the labrum covered the glenoid surface and thus a part of the separation was often hidden. Microscopically, Type II was classified into 2 subtypes. Type I: no separation was observed between the glenoid and the labrum, Type II A: a separation was observed but the articular cartilage and the labrum were smoothly connected, Type II B: a separation was observed and the articular cartilage and the labrum were not smoothly connected. When the upper 1/4 of the glenoid was investigated,24 joints were classified as Type I,58 joints as Type II A, and 18 joints as Type II B. No correlation between the types and age or sex was observed. Type I and II A were thought to have no relation to rotator cuff tears, while type II B was. Type I and II A were considered to be normal variations. Type II B was considered to be a pathological condition.
