Abstracts

Rationale: The cognitive abilities of children with benign idiopathic epilepsies can be adversely affected even in the context of average intelligence. Traditional neuropsychological approaches to examine cognition in epilepsy involve examination of diverse individual test measures cross-sectionally or prospectively, an approach that does not provide information regarding the interrelationship between different abilities or how such interrelationships might evolve over time in children with epilepsy compared to healthy controls. In this investigation we used graph theory techniques to investigate the development of cognitive landmarks in children with epilepsy compared to controls.Methods: A total of 178 subjects (age range 8-18), including 104 with new and recent-onset epilepsy and 74 healthy and normally developing first-degree cousin controls. A comprehensive battery of neuropsychological tests was administered to all youth that yielded 20 test measures assessing the domains of intelligence, academic achievement, language, memory, executive function, and cognitive/psychomotor speed. Differences between test scores at both time points (baseline and two-year follow-up) were calculated and normalized to the baseline evaluation in order to focus on the development of cognitive landmarks in children with epilepsy and healthy controls, therefore constructing a symmetric weighted graph of 20 nodes representing the actual development of cognitive abilities between groups. In addition to graph analyses we characterized the development of intelligence, executive function, language, and processing speed using traditional analytic methods for comparison of inferences derived by the two approaches.Results: Traditional analyses showed that children with epilepsy exhibited poorer test scores at baseline than controls that improved in a parallel fashion, therefore maintaining cognitive differences two years later (Figure 1). Graph analyses, however, showed that subjects with epilepsy presented with both lower integration and segregation in the development of their cognitive networks as shown by higher path length and lower clustering, respectively, compared to healthy controls, which also presented a superior level of cognitive development organization (Figure 2). In terms of hubs (Figure 2), both groups presented the test of category switching (CATSWS) while also showing a test of verbal intelligence as hub: controls showed verbal reasoning (IQSIMS) and participants with epilepsy showed vocabulary (IQVOCS).Conclusions: The group of patients showed abnormal global topology development of their cognitive networks while maintaining hubs with functional similarity to those in controls. Such topological abnormalities could be the underlying cause affecting cognitive improvement in participants with epilepsy. In conclusion, graph analyses of neuropsychological measures appear to capture a dynamic and changing complexity in the interrelationships among diverse cognitive skills, the maturation of the cognitive network over time, and the nature of cognitive differences between normally developing children and participants with epilepsy.