Abstracts

Rationale: Cognitive impairment in epilepsy reflects the effects of underlying etiology, medications, and seizures. Most evidence comes from testing done during the active phase of epilepsy. It is unclear whether impairment persists beyond the active phase, particularly in childood onset epilepsy.Methods: We analyzed cognitive function in 613 children and adolescents enrolled in a prospective community-based study, using information from all available medical records, parental interviews and a standardized neuropsychological assessment that was performed for study purposes ~8-9 years after initial diagnosis in 335 cases and also in 215 sibling controls. Cognitive function was classified as normal (consistent with IQ >=80); borderline (BL, IQ 70-79); very mild mental retardation (Mild-MR, IQ 60-69), MR (IQ <60) neurologically devastated (ND), and impaired but not further classifiable (I-NFC). Results: Median follow-up is 10.5 years; 7.7% were followed <5 years. Overall cognitive function was normal in 451 (73.6%), BL in 31 (5.1%), Mild-MR in 21 (3.4%), MR in 45 (7.3%), ND in 29 (4.7%), and I-NFC in 36 (5.9%). As expected, remote symptomatic etiology, epileptic encephalopathy, remission for < 5 years, and being on AEDs at the time cognition was assessed were all strongly associated with worse cognitive function (all p<0.0001). We considered just those study subjects with idiopathic or cryptogenic epilepsy for whom we had performed standardized neuropsychological assessments on both the subject and a sibling control and both were of normal intelligence (IQ>=80; N=145 pairs). Cases had a slightly lower IQ (case-control Full Scale IQ difference Δ = -3.4, p=0.008; Verbal IQ Δ = -2.4, p=0.07; Performance IQ Δ = -3.6, p=0.008); however, this was mainly due to slower Processing Speed Δ = -5.7, p=0.0003. Case-control differences for other factor scores were: Verbal Comprehension Δ = -2.7, p=0.05; Perceptual organization Δ= - 0.3, p=0.87; and Freedom from distractibility Δ = -2.5, p=0.09. WRAT reading (Δ= -2.5, p=0.02), spelling (Δ = -3.1, p=0.009), and arithmetic (Δ = -2.9, p=0.07) were slightly lower in cases than controls. There were no differences on measures of verbal or nonverbal memory and attention. In just pairs for which the case was >=5 years seizure-free and off medication (N=90 pairs), the same patterns with similar effects sizes were found (Δ for Processing Speed = -6.0, p=0.004).Conclusions: The frequency and degree of cognitive co-morbidity in young people with epilepsy is strongly associated with etiology, syndrome, medications, and remission status. In people with idiopathic or cryptogenic epilepsy, whose general cognitive function was in the “normal” range and who were in remission and off medication, only very small effects on general cognitive indices (IQ) and school achievement (WRAT) could be documented 8-9 years after initial diagnosis. The largest effect was seen for processing speed and could potentially reflect a subtle residual comorbidity of epilepsy even after the active phase of the disorder has passed.