Non-linear interactions among hip and foot biomechanical factors predict foot pronation during walking in women

Abstract

Background

Interactions between hip and foot biomechanical factors may result in different magnitudes of foot pronation during walking.

Objective

To investigate non-linear interactions between hip and foot biomechanical factors and their capability to predict foot pronation during walking and identify the profiles of biomechanical factors that predict greater and lower pronation.

Methods

This is a cross-sectional study. Fifty-one women were classified into greater and lower foot pronation during walking. Biomechanical factors measured: (1) foot-ankle varus alignment, (2) hip passive stiffness, (3) isokinetic eccentric strength of hip external rotators, and (4) foot abduction angle. Classification and regression trees (CART) were used to investigate non-linear interactions that predict greater and lower foot pronation.

Results

Four main profiles of biomechanical factors were identified as related to greater and lower foot pronation. Profiles for greater pronation were: (1) varus >25.83°; (2) interaction between varus ≤25.83° and hip stiffness ≤0.09 Nm/rad kg^-1; (3) interaction between varus ≤25.83°, hip stiffness >0.09 Nm/rad kg^-1, and foot abduction >19.58° The profile for lower pronation involved an interaction among varus ≤25.83°, hip stiffness >0.09 Nm/rad kg^-1, and foot abduction ≤19.58° The model had 61 % sensitivity and 96 % specificity, with the total prediction of 78 %. The area under the ROC curve was 0.79 (p = 0.001).

Conclusion

Foot-ankle varus, hip passive stiffness, and foot abduction predicted greater and lower foot pronation. Non-linear interactions between hip and foot factors influence the magnitude of foot pronation during walking. The observed profiles help identify which combinations of biomechanical factors should be assessed in individuals with increased or reduced pronation.