Identification of modifiable gait parameters predictive of simulated tibial stress-fracture risk in men and women.

Objective: This exploratory study investigated the ability of biomechanical gait parameters to predict and stratify simulated tibial stress-fracture (SF) risk in young, healthy adults. Methods: We analyzed 30 modifiable gait parameters (both kinematic and kinetic) for 21 men and 20 women when they ran with external loads (0, 11.3, or 22.7 kg) and the associated tibial SF risks during a simulated 10-week basic combat training in the U.S. military. To identify informative parameters and predict SF risk of the tibia, we separately developed personalized linear mixed-effects models for men and women. To stratify the SF risk, we developed logistic regression (LR) models. Results: Out of all the 30 parameters, ankle joint reaction force was the most predictive for both men and women, followed by hip external rotation angle for men and hip adduction moment for women. Using these parameters, LR models yielded area under the curve values of 0.81-0.95. For kinematic parameters alone, ankle dorsiflexion angle was the most predictive for both men and women, followed by hip external rotation angle for men and hip flexion angle for women. The associated LR models yielded reasonable performance for women (0.72-0.82) but weaker performance for men (0.53-0.71). Conclusion: We identified modifiable gait parameters predictive of simulated tibial SF risk in men and women and developed models based on kinematic and kinetic parameters that adequately predicted at-risk individuals. In the future, upon independent validation, this work has the potential to help identify recruits who may benefit from individualized training regimens and proactive interventions to reduce injury risk.