Abstracts

Rationale: While the cognitive consequences of epilepsy are appreciated, the degree to which they are related to underlying neuroanatomy remains to be fully understood. Relatively few studies have examined and compared structure-function relations across epilepsy syndromes. Here we examined associations between brain volumes and full-scale intelligence quotient (FSIQ) in children with idiopathic generalized epilepsy (IGE), idiopathic localization-related epilepsy (ILRE), and healthy controls (HC). In particular, we focused on the degree to which children with IGE or ILRE differ in structure-function relations, relative to HC. Methods: Research participants were children ages 8-18 with IGE (n=31), ILRE (n=35), and HC (n=52). All epilepsy participants were seen and assessed within 12 months of diagnosis, and had normal neurological examinations and no identifiable lesions on clinical MRI. The HC group consisted of age- and gender-matched first-degree cousin controls. All participants completed a test battery assessing broad neuropsychological domains, in addition to undergoing T1 volumetric MRI scans. Grey and white matter volumes were computed using the FreeSurfer image analysis suite. Analyses focused on relations between FSIQ and whole-head measures of cortical grey and white matter volume, as well as caudate and thalamus volume. Results: Children with IGE and ILRE exhibited comparable scaled IQ within the normal range (IGE: M = 100.94, SD = 14.16; ILRE: M = 100.97, SD = 12.80), although both groups were significantly lower than the HC group (M = 107.25, SD = 12.22; p's < 0.05). Total intracranial volume did not differ across groups, and thus was not used as a covariate in further analyses. HC showed significant positive associations between FSIQ and every age-adjusted volume of interest (r's .37 - .43; p's < 0.05). IGE participants showed a similar pattern, with r's ranging from .37 - .49; p's < 0.05. In contrast, ILRE participants did not show this linkage between brain volumes and FSIQ, with the only significant association being that of right caudate and FSIQ (r = .45, p < 0.05). Other correlations within the ILRE group ranged from .018 - .297. Conclusions: Children with new- and recent-onset ILRE differed from a HC group in terms of the relations between grey and white matter brain volumes and IQ in ways that IGE participants did not. This is particularly striking given that the ILRE and IGE groups did not differ in intelligence. Given that IQ is known to be associated with distributed neural systems, it appears that the brains of children with ILRE may rely on different and more diffusely associated networks to achieve similar functional outcomes (in this case, normal-range IQ). In future analyses, we plan to interrogate the nature of these networks in children with ILRE by examining performance in specific neuropsychological domains, as well as by targeting more specific grey and white matter brain regions of interest.