Abstracts

Rationale: Children with epilepsy are at risk for poor academic performance. Math may be a particularly problematic subject because good performance is dependent on normal attention and executive control, abilities that are weak in many children with seizure disorders. We investigated the impact of sustained attention and inhibition on arithmetic performance. We also examined the contribution of epilepsy-specific factors to specific calculation errors as well as overall calculation performance. Methods: 77 patients with intractable seizures, ages 8 to 18 years, were administered the Applications and Math Computations subtests from the Kaufman Test of Educational Achievement (K-TEA). The patients were also administered a test of intellectual functioning (WISC or WAIS), a continuous performance test (CPT) and the Color-Word Interference-Inhibition (DKEF-IN) and Color-Word Interference /Switching (DKEF-SW) subtests of the Delis Kaplan Executive Functions System. Each error on the computation test was coded as a knowledge error, a procedural slip, a skipped problem or misread operation. Errors were only coded if they occurred on items within the patient's range of age-expected competence. Results: The two KTEA math tests were highly correlated; consequently, only the computation subtest (Comp) scores were used in the analyses. Using multiple regression analysis, KTEA Comp scores were predicted with variables from the CPT and DKEF. Several CPT and DKEF variables were significantly correlated with math performance, but only DKEF-IN met the criteria for inclusion in the regression model, and it accounted for 24% of the variance in Comp scores. When IQ was forced into the model, it accounted for 39% of the variance, but inhibition continued to explain 6% of the variance. COMP errors were primarily due to lack of knowledge (62%). Inattention also contributed in the form of procedural slips (24%), misread operators (7%) and skipped problems (7%). Epilepsy-specific factors-- duration of disorder, seizure frequency over the past 3 months, and number of AEDs--were not directly predictive of math performance. However, the number of AEDs was related to the number of procedural slips (r=.23) and the number of misread operators, (r=.21). Conclusions: The biggest contributor to poor math performance in children with epilepsy is level of intellectual functioning. Ability to inhibit a competing response makes an additional but small contribution. Error analysis, similarly, revealed that the predominant error in math calculations is knowledge-based rather than attentional. Nevertheless, 38% of the errors were attentional, and patients who were taking higher numbers of AEDs were more likely to make those errors.