MUSCLE FUNCTIONS DURING GAIT AS POTENTIAL PREDICTORS OF FALL RISK IN PEOPLE WITH PARKINSON’S DISEASE

Introduction: Parkinson’s disease (PD) is characterized by a heterogeneous constellation of motor and non-motor symptoms that progressively compromise mobility and quality of life. Among these, gait disturbances are particularly disabling and closely associated with an elevated risk of falls, injury, and loss of independence. Efficient gait requires the coordinated activation of agonist and antagonist muscles; however, in PD, this reciprocal inhibition is often disrupted, leading to excessive muscle co-contraction and reduced locomotor efficiency. Aim: This study aimed to investigate whether alterations in lower-limb muscle activation patterns, quantified through the co-contraction index (CI) between the gastrocnemius medialis and tibialis anterior (GM–TA), are associated with fall risk in individuals with PD. CI was expressed as the percentage derived by doubling the ratio of the overlapping area of the EMG signals to the sum of the individual areas of the agonist and antagonist muscles Methods: Patients with idiopathic PD were recruited and classified as fallers (≥1 fall during a 12-month follow-up) or non-fallers. An age-matched group of healthy controls (HC) was also included. Participants performed walking trials under single-task and dual-task conditions along a linear walkway at self-selected speed. Spatiotemporal data were collected using a motion capture system, and surface electromyography (sEMG) recorded muscle activity. The CI GM-TAwas computed and compared between PD fallers and non-fallers. Logistic regression was used to evaluate the predictive value of CI for fall occurrence. Data were processed and analyzed using QTM, Visual3D, SPSS, and MATLAB. Results: Repeated-measures ANOVA revealed a significant group effect, F(1, 37) = 6.211, p = .017, η²ₚ = .144, with PD fallers exhibiting higher CI GM-TA values than non-fallers. No significant main effect of condition (p = .528) or condition × group interaction (p = .350) was observed, indicating consistent performance across single- and dual-task walking. Logistic regression identified CI during single-task gait as a significant predictor of fall status (B = 0.108, p = .033, OR = 1.114, 95% CI [1.009–1.230]), while age and sex were non-significant (p > .05). The model showed good fit (Hosmer–Lemeshow p = .240) and explained 15–20% of the variance in fall status. Discussion/Conclusion: These findings indicate that elevated lower-limb co-contraction, particularly between GM-TA, contributes to reduced gait efficiency and greater fall risk in PD. The lack of task-related modulation further suggests diminished adaptability of motor control under cognitive load. Overall, increased CI GM–TA may represent a potential biomechanical marker of fall vulnerability in PD.