Neuromotor frailty phenotype and fall risk in geriatrics with knee osteoarthritis: a cross-sectional analysis of physical therapy performance and psychometric outcomes

Background Falls are a leading cause of morbidity in older adults with knee osteoarthritis (KOA), yet current assessment strategies often lack the specificity needed to identify those at the most significant risk. The integration of physical therapy (PT) outcome measures may enhance fall risk stratification by capturing functional impairments across multiple domains. To compare neuromotor frailty profiles between older adult fallers and non-fallers with KOA using standardized PT assessments, and to evaluate the predictive validity and latent structure of these measures for fall risk classification. Methods In this cross-sectional study, 128 community-dwelling older adults with KOA (64 fallers, 64 non-fallers) completed a battery of PT outcome measures, including the Timed Up and Go Dual Task (TUG-DT), Four Square Step Test (FSST), Falls Efficacy Scale–International (FES-I), Functional Gait Assessment (FGA), Short Physical Performance Battery (SPPB), and others. Group differences were assessed using independent t -tests, logistic regression was employed to examine variables independently associated with fall status, and exploratory factor analysis (EFA) was conducted to examine the underlying constructs. Results Fallers showed significantly poorer performance across all PT measures, including slower TUG-DT (17.90 ± 3.42 vs. 13.42 ± 3.10 s, p < 0.001) and lower FGA scores (16.45 ± 4.16 vs. 20.71 ± 3.87, p < 0.001). TUG-DT (OR = 1.35, 95% CI: 1.18–1.55), FSST (OR = 1.21, CI: 1.05–1.39), and FES-I (OR = 1.11, CI: 1.03–1.19) were independently associated with fall status. EFA identified three distinct factors: dynamic balance, functional mobility, and fear of falling. Conclusion A multidimensional battery of PT outcome measures effectively differentiates fallers from non-fallers and characterizes the neuromotor frailty phenotype in older adults with KOA. These measures demonstrate strong discriminative performance for fall status and provide clinically relevant indicators of fall risk as a multidimensional construct.