LAUSR

Background In the throwing shoulder, repetitive motion increases proximal humerus and glenoid retroversion. However, clinical factors that affect this adaptation are poorly understood. Objective Evaluate and compare the degree of glenoid retroversion of the throwing and non-throwing shoulders of elite baseball players, and clarify their correlations with clinical factors such as age, position, or period of playing baseball. Design Retrospective study. The subject group in this study consisted of 291 elite baseball players over 13 yrs who visited our clinic for medical check and performed bilateral shoulder three-dimensional computed tomography from December 2011 to December 2014. We reviewed player's hand dominance, age at pitching onset, number of years of playing baseball, position, and demographic data (height, weight). Patients (or Participants) 291 elite baseball players. Players were subdivided into different age groups (juvenile: 13–15, B adolescent: 16–18, youth: 19–25) and their position groups. Methods We measured glenoid version of both throwing and non-throwing shoulders by modified method of Friedman's using developed 3D analysis software. Main Outcome Measurements Each subject underwent three-dimensional (3D) CT, which include bilateral scapulas and proximal humeri. We developed the 3D analysis software to quantitatively measure the glenoid version angle. Results The average glenoid retroversion was significantly larger in the throwing than the non-throwing shoulder. The glenoid retroversion of the non-throwing shoulder became smaller in older age groups (Pearson correlation with age: −0.125; p=0.034). Also, glenoid retroversion of the non-throwing shoulder decreased gradually in proportion to period of playing. The difference between the throwing and non-throwing shoulder became larger in older age groups (correlation with age: 0.121; p=0.039). The difference in retroversion of both shoulders was positively correlated with age and period of playing. Similarly, height and weight also showed significant results (correlation with height: 0.116; p=0.048, with weight: 0.144; p=0.014). But, the glenoid retroversion of throwing shoulder had no statistically significant tendency (p>0.05). There was no difference in glenoid retroversion between different player positions (p>0.05). Conclusions In throwing athletes, glenoid retroversion of the non-throwing shoulder decreased with age, displaying a similar natural course as by normal development. Glenoid retroversion of the throwing shoulder showed no tendency to change and it seems likely that the developmental anteversion course is offset by the retroversion force of the throwing mechanism. The difference between shoulders increases with the age and the duration of the period of playing baseball.