Context: Previous research has identified specifi c pitching kinematics associated with excessive upper extremity loading. Understanding which of these pathomechanics exist in pitchers who return from injury could create more effi cient treatment techniques, but data comparing previously injured and uninjuredathletes are lacking. Furthermore, most data on pitching mechanics are collected via three-dimensional analysis, which is prohibitively expensive for clinical use. Current advances in two-dimensional camera technology may produce kinematic variables that identify previous injury in pitchers. The purpose of this investigation was to determine the ability of two- or three-dimensional kinematic analyses to predict injury history in high school baseball pitchers. Methods: High school-aged, currently healthy, male baseball athletes (n = 43, age = 15.6 ± 1.3 years, height=178.7±6.1cm, weight=73.5±12.4kg) who had pitched for at least two consecutive seasons participated. Participants were grouped based on self-report of a previous throwing-related upper extremity injury that caused them to limit pitching activity for at least one week (previously injured: n=16, previously uninjured: n=27). Three-dimensional data were recorded using a 12-camera motion-capture system. Two-dimensional data were collected with two commercially available video cameras. Both systems collected pitching biomechanics at 240 Hz. Shoulder abduction, shoulder external rotation, shoulder horizontal abduction, lateral trunk tilt, and elbow fl exion were measured by three-dimensional analysis at three points of the pitching motion: stride foot contact (SFC), maximum external rotation (max-ER), and ball release (BR). Only shoulder abduction at SFC, elbow fl exion at SFC, and trunk lateral fl exion at BR were valid compared to three-dimensional kinematics and used in the two-dimensional analysis. Logistic regressions were used to evaluate how well the dependent variables predicted injury history. Results: The three-dimensional logistic regressionindicated statistically signifi cant improvement over the constant-only model (÷2=10.563, df=4, p=.032). The variables accounted for 30.2% of the variance. In the three-dimensional model, only shoulder abduction at maxER was identifi ed as a statistically significant predictor. Specifi cally, athletes who displayed less abduction at maxER were more likely to be placed in the injured group (Injured: 80.6±9.2°; Uninjured: 88.4±6.2°). Overall, the three-dimensional model correctly classifi ed 78.6% of the athletes into the proper injured and uninjured groups. Analysis of the two-dimensional logistic regression did not indicate a statistically signifi cant improvement over the constant-only model (÷2=2.323, df=3, p=.508), with the included variables only accounting for 8.1% of the variance between groups. Conclusions: Threedimensional analysis of shoulder abduction at maxER successfully grouped high school pitchers with and without injury history. Analysis of two-dimensional angles does not identify previous injury status in high school pitchers. These data indicate that clinicians cannot rely on two-dimensional kinematic analysis to identify injury history. However, they should begin conversations high school pitchers who are visually identifi ed to throw with a reduced shoulder abduction angle at maxER in an attempt to provide appropriate early intervention.
© Copyright 2019 Journal of Athletic Training. National Athletic Trainers' Association. Alle Rechte vorbehalten. / Übersetzt mit https://www.DeepL.com/Translator