Immediate influence of split-belt treadmill training on bilateral lower limb muscle synergies in individuals with unilateral cerebral palsy

Typical gait can be characterised by the development of four lower extremity muscle synergy patterns (1). Individuals with spastic cerebral palsy (sCP) often present with altered muscle synergies with respect to typically developing (TD) peers (2). Surgical intervention and biofeedback training does not change the muscle synergy structure in this population (3,4), but the influence of error augmentation strategies, such as split-belt treadmill training, is yet to be explored. Can a single session of split-belt training lead to an immediate change in bilateral muscle synergy structure in individuals with unilateral sCP? Twenty children/adolescents with unilateral sCP (12.8±2.5 years/months; 14 males; 11-GMFCS-I/9-GMFCS-II) walked on a treadmill (Motek-GRAIL) in the following order: 1) tied-belt condition at a comfortable speed (five-minutes), 2) split-belt condition (-40% speed on one leg and +30% on the other, 10-minutes); the shorter step length leg during the first condition walked on the faster belt and 3) tied-belt condition at the previously selected comfortable speed (five-minutes). A two-minute break was provided between each condition. Force plate and surface electromyography (sEMG) data were recorded during the final minute of each condition. sEMG data included 14 lower limb muscles (7 per leg). Non-negative-matrix-factorisation and k-means cluster analysis were used to extract and group the structure of four synergies (weights/activation patterns) for both legs combined, based on 800 strides (40 per participant) (5). To provide a measure of synergy complexity, the walk-DMC (100±10) was calculated using data from 22 TD individuals (9.2±2.3 years/months, 11 males) (6), which was also used as a control reference for the synergy structure. Step length was calculated from the force-plate data and converted into a symmetry index (short step/long step; 1.00=symmetry). All values from the first and last walking conditions were compared. Weights were compared using a repeated measure ANOVA, activation patterns using SPM (paired t-test) and walk-DMC and step length symmetry using a paired t-test (p≤0.013; corrected p-value of 0.050 for four simultaneous comparisons). No significant changes in muscle weights were found, but there were statistically significant differences in all four activation patterns (47%-85% of the gait cycle; p<0.001-0.001; Fig. 1). Fig. 1: Bilateral synergy structureDownload : Download high-res image (297KB)Download : Download full-size image Walk-DMC did not significantly differ (87.2±10.3 to 91.6±11.2), but step length symmetry did (0.93±0.06 to 0.99±0.08; p=0.007). There were significant changes in bilateral synergy activation patterns in individuals with unilateral sCP immediately following ten minutes of split-belt walking, becoming more similar to TD patterns, despite no changes in synergy complexity or weights. This may relate to the immediate improvement in step length symmetry. The unique nature of split-belt treadmill walking likely invokes an interlimb coordination response. Ongoing analysis will explore the underlying mechanism of changes in synergy structure, including per leg, as well as during the split-belt condition.