Exploring the Acute Effects of Stimulating Different Muscle Combinations During Neuromuscular Electrical Stimulation-Assisted Treadmill Walking in Children with Cerebral Palsy

Progressive gait impairments remain a major challenge for children with cerebral palsy (CP), often limiting independence and quality of life. Although neuromuscular electrical stimulation (NMES) has shown promise for improving gait, most NMES-assisted gait studies in children with CP have targeted only a single muscle. In this exploratory study, we characterized acute biomechanical responses of six children with CP to seven NMES protocols during treadmill walking. The protocols comprised multi-muscle stimulation, single-muscle stimulation, and an individualized physical therapist-derived (PT-derived) condition. Each protocol was compared with a no-stimulation baseline and typically developing (TD) reference data. The primary and secondary outcomes were the Gait Deviation Index (GDI) and Coefficient of Variation (CV), complemented by the targeted outcome measures, including lower-limb joint kinematics using phase-by-phase Dynamic Time Warping (DTW), maximum knee extension, and stride length. We found that the comparison between multi-muscle and single-muscle conditions did not identify a statistically significant difference in GDI. CV analyses showed smaller distances from TD reference values for multi-muscle conditions, although not significant. Stimulation conditions significantly affected maximum knee extension and DTW during loading response at the ankle and pre-swing at the hip; however, no individual condition differed significantly from no stimulation after false discovery rate correction. The PT-derived condition showed favorable directional changes in selected outcomes, but no consistent advantage over the other protocols. These preliminary observations characterize acute response patterns and generate hypotheses for future studies; they do not establish the comparative efficacy of multi-muscle, single-muscle, or individualized NMES protocols.