Association of Pulmonary Function and Foot Biomechanics: A Multimodal Analysis

Abstract

Chronic respiratory impairment necessitates systemic adaptations as well; however, this relationship is poorly understood. There are a few studies indicating the association between pulmonary function, and the study was designed to investigate associations between pulmonary function and foot biomechanics using multimodal statistical and machine learning approaches, with the Chippaux Smirak Index (CSI) as the primary outcome. Subject-level demographic, pulmonary, and biomechanical data were analysed without reporting descriptive tables of variables. Baseline (demographic + pulmonary) and multimodal (demographic + pulmonary + biomechanical) models were evaluated using linear regression, Random Forest regression, canonical correlation analysis, and structural pathway analysis. Multimodal integration improved CSI prediction, with Random Forest regression indicating near-complete variance (R²=0.9988). Biomechanical indices dominated model performance, while pulmonary variables contributed indirectly via cross-domain interactions. Canonical and structural analyses reveal interactions between pulmonary and foot biomechanics and indicate that foot indices may be complementary markers for respiratory analysis.