Abstracts

RATIONALE: Cognitive impairment is common in patients who suffer from medically refractory epilepsy. Fully assessing the significance of this impairment requires inferences about the premorbid level of intellectual function (IQ). Since prior testing is rarely available, several clinical and statistical approaches have been used to estimate premorbid IQ in normal, psychiatric, and neurologic populations. This study examines oral reading performance as an estimate of premorbid IQ and its relationship with seizure variables in patients with intractable epilepsy.
METHODS: Records were reviewed from a sample of 146 patients with intractable epilepsy admitted to an epilepsy monitoring unit over a 45 month period. 103 patients met the inclusion criteria of receiving standard IQ (WAIS-R/III) and oral reading measures (WRAT3) with documentation of relevant subject characteristics and seizure variables. A subset of 18 patients (group A) whose measured IQ was [gt]1 SD below predicted IQ was compared with the remaining 85 subjects (group B).
RESULTS: Overall, subjects had a mean age of 36.5 years and level of education of 11.6 years with mean age at seizure onset of 15 years and duration of 21.3 years. There was a moderately positive correlation between current and estimated IQs (r = .66, p [lt] .001). Mean measured IQ of 84.1 fell slightly below predicted IQ of 88.4 (t-test = 4.5, p [lt] .001) with 71% of patients having lower measured than predicted IQ scores. Subjects that fell substantially below expectations ([gt] 1 SD) had a mean measured IQ score of 74 that differed significantly from the remaining subjects with a mean measured IQ score of 86 (t-test = 4.8, p [lt] .001). These groups did not differ in handedness, race, gender, and education. Age at seizure onset and duration as well as etiology (idiopathic, diffuse, or focal) also did not differ between these groups. Subjects in group A with a measured IQ substantially lower than predicted showed a non-significant greater likelihood than group B of mesial temporal lobe epilepsy(78% vs. 57%, Chi-square = 2.8, p = .09). Further, lateralization of seizure focus differed significantly between groups. For group A, the percent of patients with a presumed side of seizure onset (right, left, bilateral, indeterminate) was 28%, 50%, 11%, 11% respectively, while for group B, the corresponding percentages were 47%, 51%, 0%, 2% (Chi-square = 13.6, p = .003).
CONCLUSIONS: Both measured and predicted IQs were below average in this sample of patients with intractable epilepsy. Further, a substantial subset of patients had evidence of cognitive decline from predicted premorbid levels. These results support the hypothesis that epilepsy can cause cognitive decline.