Context:

Dynamic malalignment (DM), abnormal muscle activation, and static malalignments all might lead to patellofemoral pain (PFP) but have not been examined using a multifactorial approach.

Objective:

To determine which measures of static malalignment, DM, and muscle-onset times best predict PFP.

Design and Setting:

Between-subjects, laboratory.

Subjects:

2 groups (PFP and uninjured) of 16 subjects each.

Interventions:

EMG and 3-D kinematic data were recorded during a step-down. Five static-alignment assessments were performed.

Measurements:

Three discriminant analyses using injury as the grouping variable and static measures, joint angles, and EMG onsets as the predictor variables. A final combined discriminant analysis using the most predictive variables from each set.

Results:

The static-alignment discriminant function was most predictive (81.3% correct), followed by the kinematic (69%) and the EMG (67%) functions. The final discriminant function included iliotibial-band flexibility, navicular drop, pronation, knee flexion, hip adduction, gluteus medius, and vastus medialis obliquus onset time and correctly classified 92.3% of PFP subjects.

Conclusions:

PFP can most accurately be predicted when multiple measures of lower extremity function are considered together.