Foot and Hip Contributions to High Frontal Plane Knee Projection Angle in Athletes: A Classification and Regression Tree Approach

Study Design Cross-sectional. Objective To investigate predictors of increased frontal plane knee projection angle (FPKPA) in athletes. Background The underlying mechanisms that lead to increased FPKPA are likely multifactorial and depend on how the musculoskeletal system adapts to the possible interactions between its distal and proximal segments. Bivariate and linear analyses traditionally employed to analyze the occurrence of increased FPKPA are not sufficiently robust to capture complex relationships among predictors. The investigation of nonlinear interactions among biomechanical factors is necessary to further our understanding of the interdependence of lower-limb segments and resultant dynamic knee alignment. Methods The FPKPA was assessed in 101 athletes during a single-leg squat and in 72 athletes at the moment of landing from a jump. The investigated predictors were sex, hip abductor isometric torque, passive range of motion (ROM) of hip internal rotation (IR), and shank-forefoot alignment. Classification and regression trees were used to investigate nonlinear interactions among predictors and their influence on the occurrence of increased FPKPA. Results During single-leg squatting, the occurrence of high FPKPA was predicted by the interaction between hip abductor isometric torque and passive hip IR ROM. At the moment of landing, the shank-forefoot alignment, abductor isometric torque, and passive hip IR ROM were predictors of high FPKPA. In addition, the classification and regression trees established cutoff points that could be used in clinical practice to identify athletes who are at potential risk for excessive FPKPA. Conclusion The models captured nonlinear interactions between hip abductor isometric torque, passive hip IR ROM, and shank-forefoot alignment. J Orthop Sports Phys Ther 2012;42(12):996–1004, Epub 18 September 2012. doi:10.2519/jospt.2012.4041