Gait and balance metrics comparison among different fall risk groups and principal component analysis for fall prediction in older people

• DOI: 10.1093/ageing/afaf076
• Publication Date: 2025-03-19T21:54:14Z
• Authors: Lulu Yin; Hyeri Nam; Yaru Wei; Tianyi Feng; Feifei Li; Yushan Wang; Yu Zhang; Lin Wang
• Keywords: Fall prevention

Falls are a leading cause of morbidity and mortality among older adults, often linked to gait balance impairments. To compare metrics across fall risk levels in community-dwelling adults identify principal components predictive risk. Retrospective cohort study. General community. Three hundred were stratified into low, moderate high groups using the STEADI toolkit. Gait compared groups. Principal component analysis (PCA) reduced dimensionality, binary logistic regression assessed value components. High-risk individuals showed slower cadence, shorter step length, wider width, greater variability increased centre pressure (CoP) mass (CoM) sway. PCA identified four seven components, explaining 71.62% 75.88% variance, respectively. Logistic revealed Gait_principal (PC)2 (instability) (OR = 2.545, P < .001), Gait_PC3 (rhythm control) 1.659, .006), Balance_PC1 (CoP sway during single-leg stance) 1.628, .007), Balance_PC2 (CoM velocity variability) 1.450, .032) Balance_PC4 double-leg stance, eyes closed) 1.616, .004) as significant predictors. The model achieved 77.2% accuracy, with sensitivity 73.1% specificity 79.4%. instability, rhythm control postural key predictors Integrating enhances stratification, supporting clinical decision-making.