Neurobiomechanical Characterization of Feedforward Phase of Gait Initiation in Chronic Stroke: A Linear and Non-Linear Approach

Postural control arises from the complex interplay of stability, adaptability, and dynamic adjustments, which are disrupted post-stroke, emphasizing the importance of examining these mechanisms during functional tasks. This study aimed to analyze the complexity and variability of postural control in post-stroke individuals during the feedforward phase of gait initiation. A cross-sectional study analyzed 17 post-stroke individuals and 16 matched controls. Participants had a unilateral ischemic stroke in the chronic phase and could walk independently. Exclusions included cognitive impairments, recent surgery, and neurological/orthopedic conditions. Kinematic and kinetic data were collected during 10 self-initiated gait trials to analyze centre of pressure (CoP) dynamics and joint angles (−600 ms to +50 ms). A 12-camera motion capture system (Qualisys, Gothenburg, Sweden) recorded full-body kinematics using 72 reflective markers placed on anatomical landmarks of the lower limbs, pelvis, trunk, and upper limbs. Ground reaction forces were measured via force plates (Bertec, Columbus, OH, USA) to compute CoP variables. Linear (displacement, amplitude, and velocity) and non-linear (Lyapunov exponent—LyE and multiscale entropy—MSE) measures were applied to assess postural control complexity and variability. Mann–Whitney U tests were applied (p < 0.05). The stroke group showed greater CoP displacement (p < 0.05) and reduced velocity (p = 0.021). Non-linear analysis indicated lower LyE values and reduced complexity and adaptability in CoP position and amplitude across scales (p < 0.05). In the sagittal plane, the stroke group had higher displacement and amplitude in the head, trunk, pelvis, and limbs, with reduced LyE and MSE values (p < 0.05). Frontal plane findings showed increased displacement and amplitude in the head, trunk, and ankle, with reduced LyE and MSE (p < 0.05). In the transverse plane, exaggerated rotational patterns were observed with increased displacement and amplitude in the head, trunk, pelvis, and hip, alongside reduced LyE convergence and MSE complexity (p < 0.05). Stroke survivors exhibit increased linear variability, indicating instability, and reduced non-linear complexity, reflecting limited adaptability. These results highlight the need for rehabilitation strategies that address both stability and adaptability across time scales.