Earlier Therapy Start Linked to Bigger Motor Gains in Infants With Cerebral Palsy

Infants with unilateral cerebral palsy who began an intensive motor therapy at a younger age showed larger gains in gross motor function and in parent-rated functional performance, according to an exploratory secondary analysis of a randomized controlled trial published in Frontiers in Neurology. The study, led by researchers at UCLouvain in Brussels, pooled data from 38 infants aged 6 to 20 months who all received 50 hours of Hand-Arm Bimanual Intensive Therapy Including Lower Extremities, known as HABIT-ILE, and were assessed immediately before the intervention and again three months later. The findings suggest that within this narrow early window, timing may matter for some outcomes but not for others.

HABIT-ILE is an intensive intervention grounded in motor skill learning principles, delivered as day camps that continuously stimulate both the upper and lower extremities along with posture. The baby version of the therapy used in this trial adapts the approach for children as young as six months, spreading 50 hours of practice over two weeks at five hours per day, excluding naps. Previous randomized trials of baby and early HABIT-ILE suggested that the optimal window for maximizing gains in hand function in children with unilateral cerebral palsy lies between 6 and 36 months of age, but it remained unclear whether starting earlier within that period is associated with greater improvement.

The research team addressed two related questions. First, is younger age at therapy initiation associated with greater improvements in infants with unilateral cerebral palsy aged 6 to 20 months? Second, which baseline characteristics are associated with greater improvements following the intervention? The infants came from a two-arm parallel-group randomized trial conducted in Brussels in 2021 and 2022, in which 48 infants were pair-matched randomized to either immediate or delayed treatment. For the secondary analysis, the team pooled data from all infants, since every child eventually received the therapy, and used the assessments taken immediately before and three months after HABIT-ILE to measure change during the intervention period. Changes occurring during the three months before therapy in the delayed group were analyzed separately to characterize spontaneous evolution without intervention.

Of the 48 infants in the primary trial, 38 completed the two key assessments and were included in the secondary analysis. Eight dropped out and two were excluded because they did not participate sufficiently in the assessments, possibly due to limited engagement associated with cognitive impairments. One additional infant from the delayed group refused to complete the Mini-Assisting Hand Assessment at all three time points despite having the required motor abilities, so his data were excluded from the analysis of characteristics. The final sample had a mean corrected age of 14.4 months, with 29 boys and 19 girls in the full trial cohort, and a distribution of lesion types dominated by grey matter lesions, which accounted for roughly three quarters of participants.

The outcome measures captured different dimensions of motor ability. The Mini-Assisting Hand Assessment, the primary outcome, evaluates how well the more affected hand is used as an assisting hand during bimanual activities, and was videotaped and blindly scored by certified examiners masked to group allocation and assessment timing. The Gross Motor Function Measure-66 assesses gross motor function and was scored in the same blinded fashion. The Canadian Occupational Performance Measure, a semi-structured interview with parents, quantified the child’s performance and satisfaction relative to individually defined functional goals.

The regression results revealed a striking split across outcomes. Age at therapy onset showed no statistically significant association with changes in Mini-AHA scores or in the COPM satisfaction score. However, age at initiation was negatively associated with changes in the GMFM-66, with a regression coefficient of −0.52 and a 95 percent confidence interval of −0.75 to −0.29, meaning that younger infants demonstrated larger gains in gross motor function. A similar but smaller negative association emerged for the COPM performance score, with a coefficient of −0.17. Crucially, complementary analyses during the control period, when the delayed group received no therapy, found no significant association between age and change on any of these measures, suggesting that the age-related pattern was specific to the intervention period rather than a reflection of ordinary development.

The analysis of baseline characteristics used multiple linear regression with prematurity, lesion side and lesion type as fixed factors, and baseline Mini-AHA units and GMFM-66 scores as covariates. Baseline gross motor function emerged as the most consistent correlate: infants with lower baseline GMFM-66 scores showed larger increases in GMFM-66 during the intervention, and the same pattern held for both the COPM performance and satisfaction scores. The authors noted a moderate positive correlation between age and baseline GMFM-66, meaning younger children tended to start with fewer gross motor skills and therefore had more room to acquire new abilities, which may partly explain why younger infants and those with lower baseline function showed the largest changes.

One exploratory finding concerned lesion type. The regression model for Mini-AHA changes was statistically significant overall, and lesion type showed an association with change in manual ability, with infants who had periventricular white matter lesions showing larger estimated changes than those with grey matter lesions, an adjusted between-group difference of 13.18 Mini-AHA units. However, the distribution of lesion types in the sample was small and highly uneven, with only five infants with brain maldevelopments and eight with periventricular lesions, so the authors emphasized that this result should be treated as hypothesis-generating and requires confirmation in larger, more balanced samples. Prematurity, lesion side and baseline manual performance were not significantly associated with Mini-AHA changes, a result that contrasts with findings in preschool and school-aged children, where poorer baseline hand function has sometimes predicted greater benefit from upper limb interventions.

The authors were careful to frame the entire analysis as exploratory. The sample size was calculated for the primary outcomes of the original randomized trial, not for these secondary questions, the statistical approach and the baseline-characteristic analyses were not prespecified and were defined after the trial was completed, and no adjustment for multiple comparisons was applied across the overall set of exploratory tests. Subcategories such as lesion type were underrepresented, limiting precision and generalizability, and the analyses cannot determine whether the age-related pattern in gross motor gains reflects differential responsiveness to HABIT-ILE, natural developmental variability, or a combination of both. Ceiling effects were considered unlikely to explain the findings, since no participant had a maximum GMFM-66 score at baseline and treatment goals with baseline scores of 10 out of 10 were not selected.

The broader implication is that the impact of intensive therapy is not confined to the youngest infants. Most infants in the study exceeded the minimal clinically important difference for the GMFM-66, even among the older children, echoing earlier reports of large gross motor gains in toddlers up to five years of age. Within the 6-to-20-month window, however, the divergence between outcomes is intriguing: timing appeared linked to gross motor and functional performance changes but not to changes in assisting hand use, which the authors suggest may reflect distinct maturation timelines of the neural pathways governing gross motor and upper limb function. The team calls for studies identifying characteristics associated with greater change following other early interventions, so that results can be pooled in meta-analyses and rehabilitation strategies can eventually be tailored to each child’s specific profile.

Subject of Research: Therapy onset timing and baseline characteristics associated with improvement after HABIT-ILE in infants with unilateral cerebral palsy

Article Title: Therapy onset and characteristics associated with improvement following HABIT-ILE in infants with unilateral cerebral palsy: secondary analysis of a randomized controlled trial

Article References: Carton de Tournai, A., Herman, E., Ebner-Karestinos, D., Gathy, E., Araneda, R., Renders, A., De Clerck, C., Kilcioglu, S., Vandermeeren, Y., & Bleyenheuft, Y. (2026). Therapy onset and characteristics associated with improvement following HABIT-ILE in infants with unilateral cerebral palsy: secondary analysis of a randomized controlled trial. Frontiers in Neurology, 17, 1882719. https://doi.org/10.3389/fneur.2026.1882719

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1882719

Keywords: cerebral palsy, HABIT-ILE, infants, neurorehabilitation, gross motor function, Mini-AHA, GMFM-66, early intervention, neuroplasticity, randomized controlled trial, motor learning, therapy timing