Sex-specific Knee Biomechanics Associated with Patellofemoral Pain in Amateur Runners During Running

Chen Yang,Bing Yu,Hui Liu,Feng Qu
DOI: https://doi.org/10.1249/01.mss.0000683000.43664.f2
2020-01-01
Abstract:Inconsistent associations have been reported for ground reaction force (GRF) variables and running injuries when grouping all injuries together. However, previous work has shown more consistent associations when focusing on specific injuries. PURPOSE: To establish general and injury-specific associations between GRFs and five common running injuries. METHODS: 126 runners presenting with patellofemoral pain (PFP), tibial bone stress injury (TBSI), plantar fasciitis (PF), Achilles tendinitis, or iliotibial band syndrome and 70 healthy controls (CON) completed an instrumented treadmill assessment at a self-selected speed. All were rearfoot strikers. Injured/control groups were matched for gender and running speed. Vertical average and instantaneous load rates (VALR, VILR) were calculated over the first 15% of stance. Peak vertical, posterior, medial (mGRF), and lateral forces were calculated over all of stance. Mean comparisons were made between the general injury (INJ) and CON groups, and specific injuries and CON. Further, optimal cutoff values were established for variables using ROC curves. Area under the curve (AUC) and odds ratios (ORs) were used to evaluate their diagnostic value. RESULTS: Mean differences for variables between CON and injury groups are presented below. Only VALR, VILR and mGRF showed significant differences between groups (p<0.05). Cutoffs for VALR had superior diagnostic value for PFP and PF (OR=5.8-10.6, AUC=0.70-0.73) compared to INJ (OR=3.8, AUC=0.65). This trend held for VILR (PFP: OR=5.1, AUC=0.72; INJ: OR=3.5, AUC=0.64) and mGRF (TBSI: OR=8.8, AUC=0.74; INJ: OR=3.0, AUC=0.60). CONCLUSION: VALR, VILR and mGRF variables showed significant associations with running injury, even when grouped across injuries. However, associations were driven by PFP, PF and TBSI. Diagnostic value of variables was greatly improved when separating specific from general injuries.
What problem does this paper attempt to address?