Authors :
Presenting Author: Manuela Pisch, PhD – University College London
Aikaterini Vezyroglou, PhD – University College London
Jane Kung, PhD – University College London
Finbar O’Callaghan, Prof – University College London
Christin Eltze, PhD – University College London
Elaine Hughes, PhD – Evelina Hospital London
Manju Kurian, PhD – University College London
Michelle de Haan, PhD – University College London
Rod Scott, PhD – University College London
J. Helen Cross, M.B., Ch.B., PhD – University College London NIHR BRC Great Ormond Street Institute of Child Health
Rationale:
Early-onset epilepsy is associated with a high risk of developmental delay, though outcomes vary and are not well understood. This study aimed to describe neurodevelopmental outcomes and identify predictors in infants with early-onset epilepsy in London and the South-East of England.
Methods:
In this prospective cohort study, 112 infants aged 4 weeks to 12 months with ≥2 unprovoked seizures were recruited from regional hospitals and clinics. Neurodevelopmental assessments were conducted at baseline (T1, mean age = 8.46 ± 3.3 months) and around 12 months (T2), using the Bayley Scales of Infant and Toddler Development (Bayley-III), the Bayley Social-Emotional Questionnaire, and the Vineland Adaptive Behavior Scales. Principal component analysis (PCA) was used to derive a composite neurodevelopmental score. Correlations and multiple linear regressions examined associations with clinical and demographic factors.
Results:
At T1, average scores across domains were below normal. PCA revealed a single factor with high internal consistency (KMO = 0.90). Neurodevelopmental functioning correlated negatively with age at testing (r = -0.38, p < 0.001) but not age of seizure onset. Regression analysis identified normal EEG (β = 1.53, p = 0.005), higher birth weight (β = 1.13, p = 0.019), and younger age at assessment (β = -0.29, p = 0.001) as significant predictors of better functioning. Neurodevelopmental scores were stable between T1 and T2 (e.g., motor t = 9.78, p < 0.001). Clinical variables, including MRI and aetiology, did not significantly predict change over time.