Interaction Between Spasticity and the Initial Response After Balance Perturbations in Children With Spastic Cerebral Palsy

Association Between the Initial Response After Perturbation to Standing Balance and Spasticity Measurements (1) and Kinematic Response Strategy to Restore Balance (2) in Children With Spastic Cerebral Palsy

The aim of this study was to investigate the link between spasticity and the initial response after standing balance perturbations in children with spastic cerebral palsy. Reactive balance performance was tested using a moving platform. The investigators provided two types of perturbations, (1) backward translations and (2) rotations towards dorsiflexion, of different magnitudes. Spasticity was assessed using instrumented clinical tests of spasticity as the pendulum test and isolated passive joint rotations. Kinematics and EMG were measured simultaneously.

Study Overview

• Status: Completed
• Conditions: Cerebral Palsy (CP); Typically Developing Children

Detailed Description

Background:

Spasticity is a common impairment following an upper motor neuron lesion, such as cerebral palsy (CP). Children with CP suffer a lot from balance impairments, which will impair their participation in daily life activities. When clinically assessing spasticity, a passive muscle is stretched and spasticity is scored based on the observed resistance against this stretch. Similar muscle stretches also occur when standing balance is perturbed, as for example when standing on a departing bus or walking on uneven terrain.

Maintaining balance involves complex sensorimotor transformations to activate the muscle to produce the required balance correcting response. Sensorimotor processing refers to how the nervous system translates incoming sensory information about body motion into motor commands to activate muscles. In both healthy animals and humans, sensorimotor processes underlying reactive standing balance can be explained by delayed feedback from CoM kinematics. The sensitivity to the CoM disturbance have been shown to be relatively constant within one subject, but change with age, age-related cognitive decline, sensory deficits, and neurological impairments.

Both kinematic and muscle responses to perturbations of standing balance are impaired in children with CP. Children with CP have a higher chance of losing their balance, change to a stepping strategy at lower perturbation levels, and have increased muscle (co-)activation. However, it is unknown whether the underlying sensorimotor transformations are altered in CP and how these changes contribute to the altered balance responses observed in CP. Therefore, the first aim was to understand alterations in sensorimotor processing underlying reactive standing balance control in children with spastic CP.

The increased muscle co-activation that is observed in CP might be a functional compensation strategy to improve balance control. For example, muscle co-activation will increase joint stiffness and therefore resisting movement of the body with respect to the feet during backward translational perturbations. In this case, co-activation will help to maintain balance. However, it is not clear whether children with CP use the increased co-activation as compensation strategy to improve balance control or whether the muscle co-activation is a consequence of impaired balance control. When standing balance is perturbed using rotational perturbations, muscle co-activation will hinder balance control. Increased muscle co-activation and the following increased joint stiffness will couple body movement to platform movement resulting in body tilt. Hence, increasing joint stiffness might not be beneficial during rotational perturbations. Our second aim was to investigate whether children with CP use increased muscle co-activation as a compensation strategy to improve balance control or whether muscle co-activation causes balance control impairments by combining translational and rotational perturbations of standing balance.

Muscle stretches occur during isolated joint rotations, as clinically applied when assessing joint hyper-resistance, and during perturbations of standing balance. Children with CP respond to both isolated joint rotations and perturbations of standing with increased excitation of the stretched muscle, combined with increased agonist-antagonist co-activation. Although the striking similarities between both reactive muscle responses to muscle stretches, little is known about the relation between joint hyper-resistance and reactive balance. Therefore our third aim was to explore the relation between muscle responses to instrumented tests of joint hyper-resistance and translational and rotational perturbations of standing balance.

Methods:

Twenty children with CP and twenty age-matched typically developing (TD) children participated in this study. Children with CP were diagnosed by a neuro-pediatrician and met the following inclusion criteria: (1)aged between 5 and 17 years old; (32) Gross Motor Classification Scale I-III; (3) able to stand independently for at least 10 minutes; (4) no orthopedic/neurological surgery in the previous year; and (5) no botulinum neurotoxin injections in the previous six months.

An instrumented spasticity assessment (hereafter called isolated joint rotations) using the method described by Bar-On and colleagues was performed. Participants lay supine and were asked to relax. The lower leg was supported by a customized frame that allowed ankle rotation. A researcher applied five times a passive rotation of the ankle joint, as fast as possible from a plantar flexed position to the end range of motion towards dorsiflexion.

Reactive balance was tested on a Caren platform. Participants stood barefoot on the platform and were secured with a safety harness. Instructions were to stand upright and maintain balance without stepping unless stepping was necessary to avoid failing. The protocol consisted of six increasingly difficult perturbations levels for the backward translations and four increasingly difficult perturbation levels for the toe-up rotational perturbations. Withing each level, eight perturbations were administered. When participants stepped I more than 3 trials within a level, the participant did not continue to the next level.

Trajectories of reflective skin markers were captured by infrared Vicon Cameras and activity of the lower leg muscles (lateral gastrocnemius, medial gastrocnemius, soleus and tibialis anterior) was measured simultaneously through surface electromyography.

• Study Type: Observational
• Enrollment (Actual): 40

Contacts and Locations

Study Locations

• Belgium
  • Vlaams-Brabant
    • Leuven, Vlaams-Brabant, Belgium, 3000
      • UZ Leuven

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child
• Accepts Healthy Volunteers: Yes

• Sampling Method: Non-Probability Sample

Study Population

Children with spastic cerebral palsy, who have routine follow-up care at the CP reference center of the university hospitals Leuven.

Typically developing children are recruited through the social and professional network of involved researchers and thesis studies, who approach participants through flyers and social media.

Description

Children with CP:

Inclusion Criteria:

• Diagnosis of cerebral palsy
• Spasticity as defined by clinical assessment
• Aged between 5-17 years old
• Gross motor classification scale I-III
• Able to stand independently for at least 10 minutes

Exclusion Criteria:

• Orthopedic/neurological surgery in the previous year
• Botulinum neurotoxin injections in the past 6 months
• Presence of ataxia or dystonia
• Cognitive problems that impede measurements
• Severe co-morbidities

Typically developing children:

Inclusion criteria:

• Aged between 5 and 17 years old
• Good health

Exclusion criteria:

• Presence of neuro-musculoskeletal or vestibular diseases
• Lower limb injuries during the past 6 months
• Irritated skin or open wounds where sensors will be placed (CP en TD)

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort
Children with cerebral palsy; Children with CP aged between 5 and 17
Typically developing children; Typically developing children aged between 5 and 17

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Balance performance	The number (#) of completed perturbation levels (backward translational and toe-up rotational) without stepping.	Cross-sectional data collection at a single time point (baseline measurement).
Maximal horizontal center of mass displacement	The maximal horizontal Center of Mass (CoM) displacement (cm) is computed in a period of 1.5s following perturbation onset. CoM position is computed by consequently applying OpenSim's Inverse Kinematics and Body Kinematics tools with reflective marker trajectories as input. CoM displacement was calculated relative to the ankle.	Cross-sectional data collection at a single time point (baseline measurement).
Goodness of fit between predicted and reconstructed muscles activity - R²	Computational model; Sensorimotor transformations were evaluated by reconstructing measured EMG trajectories (V) by delayed feedback from CoM kinematics. EMGrecon=e0+⌊kd∗dCoM(t -τ)+kv∗vCoM(t -τ)+ka∗aCoM(t -τ)+ks ∗aCoM Init(t -τ)⌋ with EMG recon = reconstructed muscle activity E0 = baseline muscle activity (during quiet standing) dCoM, vCoM, aCoM = CoM displacement, velocity and acceleration kd, kv, ka = feedbakc gains or weights τ = common time delay of 100 ms To test whether CoM feedback can explain reactive muscle activity, the goodness of fit was assessed between predicted and reconstructed muscle activity using the coefficient of determination (r²). R² was calculated as the squared correlation coefficient.	Cross-sectional data collection at a single time point (baseline measurement)
Goodness of fit between predicted and reconstructed muscles activity - VAF	Computational model; Sensorimotor transformations were evaluated by reconstructing measured EMG trajectories (V) by delayed feedback from CoM kinematics. EMGrecon=e0+⌊kd∗dCoM(t -τ)+kv∗vCoM(t -τ)+ka∗aCoM(t -τ)+ks ∗aCoM Init(t -τ)⌋ with EMG recon = reconstructed muscle activity E0 = baseline muscle activity (during quiet standing) dCoM, vCoM, aCoM = CoM displacement, velocity and acceleration kd, kv, ka = feedbakc gains or weights τ = common time delay of 100 ms To test whether CoM feedback can explain reactive muscle activity, the goodness of fit was assessed between predicted and reconstructed muscle activity using the Variance Accounted For (VAF). VAF was calculated as the uncentered r²	Cross-sectional data collection at a single time point (baseline measurement)
Goodness of fit between predicted and reconstructed muscles activity - RMSE	Computational model; Sensorimotor transformations were evaluated by reconstructing measured EMG trajectories (V) by delayed feedback from CoM kinematics. EMGrecon=e0+⌊kd∗dCoM(t -τ)+kv∗vCoM(t -τ)+ka∗aCoM(t -τ)+ks ∗aCoM Init(t -τ)⌋ with EMG recon = reconstructed muscle activity E0 = baseline muscle activity (during quiet standing) dCoM, vCoM, aCoM = CoM displacement, velocity and acceleration kd, kv, ka = feedbakc gains or weights τ = common time delay of 100 ms To test whether CoM feedback can explain reactive muscle activity, the goodness of fit was assessed between predicted and reconstructed muscle activity using the Root Mean Square Error (RMSE).	Cross-sectional data collection at a single time point (baseline measurement)
Sensitiviy of reactive muscle activity to CoM perturbations - gains	Computational model; Sensorimotor transformations were evaluated by reconstructing measured EMG trajectories (V) by delayed feedback from CoM kinematics. EMGrecon=e0+⌊kd∗dCoM(t -τ)+kv∗vCoM(t -τ)+ka∗aCoM(t -τ)+ks ∗aCoM Init(t -τ)⌋ with EMG recon = reconstructed muscle activity E0 = baseline muscle activity (during quiet standing) dCoM, vCoM, aCoM = CoM displacement, velocity and acceleration kd, kv, ka = feedbakc gains or weights τ = common time delay of 100 ms To test the sensitivity of the reactive muscle activity to the CoM perurbations, the gains (kd,kv,ka) were assessed. Gains indicate the sensitivity of the muscle response to CoM perturbations.	Cross-sectional data collection at a single time point (baseline measurement)
Mean reactive muscle activity	Average reactive muscle activity for; lateral gastrocnemius; medial gastrocnemius; soleus; tibialis anterior was computed in three time bins. Reactive muscle activity was calculated by subtracting baseline activity, i.e. average muscle activity in the 100 ms preceding perturbation onset, from the filtered and scaled EMG. The first time bin lasted from platform onset to 150 ms after perturbation onset. The second time bin lasted from 150 ms to 250 ms after perturbation onset. The third time bin lasted from 250 ms to 400 ms after perturbation onset.	Cross-sectional data collection at a single time point (baseline measurement).
Mean center of mass movement	Average horizontal center of mass movement (cm) was computed in three time bins. The first time bin lasted from platform onset to 50 ms after perturbation onset. The second time bin lasted from 50 ms to 150 ms after perturbation onset. The third time bin lasted from 150 ms to 300 ms after perturbation onset.	Cross-sectional data collection at a single time point (baseline measurement).
Mean ankle kinematics	Average ankle kinematics (°) was computed in three time bins. The first time bin lasted from platform onset to 50 ms after perturbation onset. The second time bin lasted from 50 ms to 150 ms after perturbation onset. The third time bin lasted from 150 ms to 300 ms after perturbation onset.	Cross-sectional data collection at a single time point (baseline measurement).
Co-Contraction Index (CCI) during translational and rotational perturbations and during isolated joint rotations	The co-contraction index was calculated as the minimum tibialis anterior and respectively lateral gastrocnemius, medial gastrocnemius and soleus filtered and scaled EMG averaged over the time interval of interest. CCI was calculated during backward translational, toe-up rotational perturbations and during isolated joint rotations.	Cross-sectional data collection at a single time point (baseline measurement).

Collaborators and Investigators

• Sponsor: Universitaire Ziekenhuizen KU Leuven
• Collaborators: KU Leuven
• Investigators: Principal Investigator: Kaat Desloovere, Prof. dr., Department of Rehabilitation Sciences, KU Leuven, Belgium

Study record dates

Study Major Dates

• Study Start (Actual): March 13, 2020
• Primary Completion (Actual): August 30, 2021
• Study Completion (Actual): August 30, 2021

Study Registration Dates

• First Submitted: May 8, 2026
• First Submitted That Met QC Criteria: May 22, 2026
• First Posted (Actual): June 1, 2026

Study Record Updates

• Last Update Posted (Actual): June 1, 2026
• Last Update Submitted That Met QC Criteria: May 22, 2026
• Last Verified: April 1, 2026

More Information

Terms related to this study

• Keywords: Balance control; neurological impairment; Reactive balance; spastic cerebral palsy; muscle co-activation; Sensorimotor processing; Center of mass feedback
• Additional Relevant MeSH Terms: Brain Diseases; Central Nervous System Diseases; Nervous System Diseases; Brain Damage, Chronic; Pathological Conditions, Signs and Symptoms; Signs and Symptoms; Neurologic Manifestations; Cerebral Palsy
• Other Study ID Numbers: S63321

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No