Characteristics And Biomechanical Factors Associated With Patellofemoral Pain In Runners

PURPOSE: To classify runners with active PFP using a combination of factors including anthropometric, demographic and biomechanical as well as the interactions between them. METHODS: A cross-sectional study with 76 rearfoot strike runners (38 PFP; 38 Healthy Controls (CON)). Age, sex, and body mass index (BMI) were included. Biomechanical variables included peak vertical ground reaction force (vGRF), vertical average loading rate (VALR), vertical instantaneous loading rate (VILR), braking ground reaction force impulse (BGRFI), contact time, step rate, and center of mass excursion (COMEX). Interactions among these factors were analyzed using a CART analysis. Prevalence Ratios (PR) with 95% confidence intervals were calculated for each terminal node. A receiver operating characteristic (ROC) curve was created to verify the accuracy of the model. RESULTS: The CART model correctly classified 32 (84.2%) runners with PFP, and 30 (78.9%) who were CON. The total correct classification was 81.6%, and the area under the ROC curve (accuracy) was 0.83 (95% CI, 0.74-0.93; SE, 0.04; P .012 BW*s, Contact time > .29 s. 4.16 (1.04 to 16.60)* Node 8 Braking Ground Reaction Force Impulse > .012 BW*s, Contact time 66.9 BW/s, Age > 27.5 y 9.86 (1.16 to 83.34)* Protective Profile Node 5 Braking Ground Reaction Force Impulse > .012 BW*s, Contact time < = .29 s, Average Vertical Loading Rate < = 66.9 BW/s 0.06 (0.02 to 0.23)* PR prevalence ratio, CI confidence interval. *Statistically significant (P < .05).