Abstract
Gait analysis provides crucial insights into neuromuscular coordination and postural control, especially in ageing populations and rehabilitation contexts. This study investigates the complexity of muscle activation and ground reaction force patterns during gait by applying detrended fluctuation analysis (DFA) to electromyography (EMG) and force-sensitive resistor (FSR) signals. Data from a two-arm randomised clinical trial (RCT) supplemented with an observational control group were used in this study. Participants performed a single-task walking protocol, with EMG recorded from the tibialis anterior and lateral gastrocnemius muscles of both legs and FSR sensors placed under the feet. Gait cycles were segmented using heel-strike detection from the FSR signal, enabling analysis of individual strides. For each gait cycle, DFA was applied to quantify the long-range temporal correlations in the EMG and FSR time series. Results revealed consistent α-scaling exponents across cycles, with EMG signals exhibiting moderate persistence (α≈0.85-0.92) and FSR signals showing higher persistence (α≈1.5), which is indicative of stable and repeatable gait patterns. These findings support the utility of DFA as a nonlinear signal processing tool for characterising gait dynamics, offering potential markers for gait stability, motor control, and intervention effects in populations practising movement-based therapies such as Tai Chi. Future work will extend this analysis to dual-task conditions and comparative group studies.