Abstract

In order to identify injury-proneness in female competitive gymnasts, 20 measures of flexibility, hypermobility, spinal posture and anthropometry were performed on 40 competitive gymnasts and injury scores were derived from the severity and extent of previous gymnastic injury and inherent hypermobility traits. Results were compared between contrasting groups of "low" and "high" injury gymnasts respectively (both N = 10). Nine variables demonstrated significant differences between the "low" and "high" injury risk status groups namely, weight (p less than 0.001), height (p less than 0.001), age (p less than 0.001), mesomorphy (p less than 0.01), Quetelet Index (p less than 0.01), shoulder flexion (p less than 0.05) and lumbar extension (p less than 0.05), standing lumbar curvature and total peripheral flexibility score (both p less than or equal to 0.05). Multiple regression analysis was applied to determine the relative contribution of these variables to the estimation of injury-proneness as evidenced by previous history of injury and hypermobility traits. Using 9 independent variables, multiple regression yielded a multiple correlation coefficient (R) = 0.840, accounting for over 70% of the observed variance (R2 = 0.706) in injury scores among the total group of gymnasts. However, a subset of five variables, (weight, mesomorphy, standing lumbar curvature, age and height) yielded a multiple correlation coefficient (R) = 0.834 accounting for almost 70% of the observed variance (R2 = 0.696). This was not significantly different from the larger subset. Using injury classification system of "low", "medium", and "high" risk categories, comparisons were made between predicted and observed injury scores in the respective risk categories. In "high" risk and "low" risk gymnasts, injury scores could be classified correctly with 70% and 79% accuracy respectively, so that relative risk status could be determined from simple physical tests which may be employed by practitioners in the field.