Perturbation training has emerged as a potential intervention to improve reactive balance in children with spastic cerebral palsy (CP), but experimental evidence remains limited. The mechanisms by which such training might improve balance—particularly changes in muscle coordination—are not well described. It is also unclear whether training effects transfer to untrained balance challenges. This study examined whether a targeted rotational perturbation training program modifies reactive standing balance and reactive muscle responses during both trained and untrained perturbation conditions in children with CP and in typically developing (TD) peers.
The investigation was a longitudinal intervention study including fourteen children with spastic CP and ten TD children. Reactive balance and muscle responses were assessed immediately before and after the training period. Two types of perturbations were used for assessment: a toe-up rotational perturbation (the trained condition) and a backward translational perturbation (an untrained condition).
Participants completed a training program consisting of nine sessions spread over three weeks. Each training session included 96 rotational perturbations. Perturbation intensity was individually adapted according to each participant’s performance level to provide appropriate challenge throughout the intervention.
Reactive balance performance was quantified as the participant’s ability to remain upright without taking a step when exposed to perturbations. Reactive muscle responses were evaluated using two complementary metrics: the co-contraction index, which quantifies simultaneous activation of opposing muscle groups, and average muscle activity in key ankle muscles (plantarflexors and dorsiflexors).
Following the rotational perturbation training, children with CP showed improved ability to maintain standing balance without stepping. All children with CP were able to achieve more difficult perturbation intensity levels after completing the training program. These changes indicate enhanced reactive balance performance in the trained group.
Training was associated with reductions in muscle co-activation in children with CP. Specifically, the co-contraction index decreased after training in response to both the trained rotational perturbations and the untrained backward translational perturbations, suggesting improved coordination of antagonist muscle activity across perturbation types.
In addition to reduced co-activation in the CP group, average reactive muscle activity decreased after training in both plantarflexors and dorsiflexors for both TD children and children with CP. These decreases in mean muscle activity were observed across trained and untrained perturbation conditions.
The observed reductions in muscle co-activation and in average reactive muscle activity, together with improved ability to remain upright without stepping, indicate that reactive balance control can be modified through targeted rotational perturbation training in children with spastic CP. Importantly, adaptations in muscle coordination generalized to an untrained perturbation modality, suggesting potential transfer of training effects beyond the practiced perturbation type.
These findings support the concept that perturbation-based interventions can alter neuromuscular control strategies underlying reactive balance and may have utility in rehabilitation programs for children with spastic CP. The decrease in antagonist co-activation may reflect more efficient or more selective recruitment of muscles during balance recovery.
The study received ethical approval from the Ethical committee of UZ Leuven/KU Leuven (S65672). The trial is registered under Clinical Trial NCT07818928. The authors declared no competing interests. All data produced in the study are available upon reasonable request to the authors.
This summary reports the methods and main outcomes as presented in the source article. The source did not provide additional quantitative details in this summary regarding effect sizes, statistical tests, exact ages or functional levels of participants, or long-term retention of training effects; those details were not reported in the excerpt provided and would need to be obtained from the full preprint or supplied data upon request.