Rotational perturbation training improves muscle coordination during reactive standing balance in trained and untrained conditions in children with spastic cerebral palsy

Perturbation training seems a promising tool to improve reactive balance in children with cerebral palsy. However, experimental evidence remains limited. In particular, little is known about how perturbation training alters muscle coordination, limiting insight into the mechanisms underlying training-induced improvements in balance control. Furthermore, it is unclear whether adaptations in balance control generalize to untrained balance tasks. Here, we investigated the effect of rotational perturbation training on reactive balance performance and reactive muscle activity during trained and untrained perturbation conditions in children with cerebral palsy and typically developing children. We performed a longitudinal intervention study in fourteen children with cerebral palsy and ten typically developing children. Reactive balance was assessed before and after the intervention using toe-up rotational and backward translational perturbations of standing balance. The training intervention consisted of nine sessions over three weeks, with each session including 96 rotational perturbations. Perturbation intensity was adapted to the participants performance level. Reactive balance performance was quantified as the ability to stay upright without stepping. Reactive muscle responses were evaluated using the co-contraction index and average muscle activity. Rotational perturbation training improved the ability of children with cerebral palsy to maintain standing balance without stepping. All children were able to perform more difficult perturbation levels after training. Muscle co-activation decreased after training in response to both trained rotational and untrained translational perturbations in children with cerebral palsy. In addition, average reactive muscle activity in both plantarflexors and dorsiflexors decreased after training in both typically developing children and children with cerebral palsy. Our findings suggest that reactive balance control can be modified through perturbation training, and that training-induced adaptations generalize to untrained perturbation conditions.