Abstracts

Rationale: We prospectively examined changes in neuropsychological functioning for 36 months following seizure onset in children and examined the degree to which risk factors for deficits and decline varied with age of onset. Methods: We followed 228 children ages 6-14 years who had experienced a first recognized seizure and 125 sibling controls; the two groups did not differ on age or sex. Seizure control, AED use, and syndrome were obtained by chart review and interviews with parents at baseline, 9, 18, 27, and 36 months. Seizure control was classified at 36 months into 3 groups: Single-Seizure (n=56; only 1 seizure), Recurrent (n=143; 2 seizures but not at every clinical follow-up), Persistent (n=29; seizures at every follow-up). AED use was classified at 36 months into 3 groups: Never (n=62; no AED at any time), Partial (n=61; AED at one or more 9-month follow-ups but not all follow-ups), Continuous (n=105; AEDs at all follow-ups). Neuropsychological testing at baseline (M=2.8 months after first seizure) and at 18 & 36 months after onset yielded 4 age-corrected factors: Language, Verbal Memory & Learning, Processing Speed, Attention/Executive/Construction skills (higher scores are better). Epileptic syndrome was defined by ILAE criteria. Mixed models were used to investigate change in factor scores over time by age and by risk factors (AED use, seizure control, epileptic syndrome). Hochberg s step-up Bonferroni method was used to adjust for multiple comparisons. Results: There was an Age x Time x Seizure Control interaction on Processing Speed (p=0.0006); poor seizure control increased the risk of decline in processing speed over time, but only in younger children (Figure 1). In addition, there were Age x Seizure Control interactions on Attention/Executive/Construction (p<0.0001) and on Verbal Memory & Learning (p=0.04), with younger age of onset amplifying the risk associated with poor seizure control in both domains across the 3-year study period. There was an Age x AED interaction on Verbal Memory & Learning (p=0.02); in the older and middle age ranges, the Continuous AED Group had lower memory scores vs. Siblings and vs. the Partial-AED Group; however, in the youngest age range, no AED groups differed significantly (Figure 2). There were no interactions between age and epileptic syndrome. Conclusions: Younger age of seizure onset was clearly associated with increased vulnerability to the effects of undercontrolled seizures on neuropsychological functioning. Unexpectedly, older children seemed to be more vulnerable to adverse cognitive effects of AEDs; this could reflect differences in AEDs used (oxcarbazepine more common in younger vs. older; valproate more common in older vs. younger), but this study was not a randomized controlled trial, did not measure blood levels, and included heterogeneous syndromes; also, AEDs might have different adverse effect profiles at different ages. Thus, seizure control is critical to cognitive development in younger school-age children; the impact of AEDs warrants further examination. Funded by NIH/NINDS R01 NS22416 (Austin)