Abstracts

Rationale: Evidence for psychiatric, cognitive, linguistic, and social comorbidities and the association of psychiatric diagnoses with subtle to severe cognitive, linguistic, and social difficulties imply a single psychosocial comorbidity factor (PCF) in children with epilepsy only (CWE) (Hamiwka et al. 2011). The relationship between these individual comorbidities and morphometric abnormalities (Dabbs, et al. 2013) suggest their possible role in the PCF. This study examined if there is an interrelationship between the psychosocial comorbidities in CWE with localization?"related epilepsy (LRE) and regional gray and white matter brain volumes (GMV, WMV), cortical thickness (CT), and sulcal depth (SD). Methods: A surface-based, computational high-resolution 3-D magnetic resonance image analytic technique examined CT and SD and regional parcellations of GMV and WMV were computed on 42 LRE patients, aged 6-16 years, imaged in a 1.5Tesla scanner. Each child had a semi-structured psychiatric interview and cognitive/linguistic testing. Parents completed the Child Behavior Checklist (CBCL) problem and social behaviors, and provided medical chart confirmed seizure information. A principal components analysis on the number of psychiatric diagnoses, CBCL problem scores, CBCL social competence scores, Verbal IQ, Performance IQ and language scores was performed to obtain a single factor, the PCF. General linear models were used to examine associations between the PCF and (i) previously studied regional GMV and WMV (orbital, dorsolateral prefrontal, inferior, posterior frontal and temporal), and (ii) CT and SD (frontal, temporal, parietal, occipital, and limbic regions), controlling for age and sex. For CT and SD, if significant associations were found, post hoc analyses examined which specific subregions contributed to the finding. Results: All CBCL problem scores other than somatic complaints (0.55) loaded highly (.73-.83) on the PCF while the loadings for IQ, language, and CBCL peer interaction and social organization scores ranged from -.49 to -.22 and the number of psychiatric diagnoses loaded at .24. The PCF was significantly negatively associated with frontal, parietal, temporal, and limbic CT (F(1, 30)=4.9-6.6, p < .04- < .01) but not with GMV, WMV, and SD, controlling for age and sex. Posthoc testing (Table 1) revealed significant negative PCF-CT associations in the precentral, middle frontal, and superior frontal; postcentral, superior parietal, and supramarginal; superior and middle temporal and cingulate regions. Demographic and seizure variables were unrelated to these findings. Conclusions: Widespread decreased CT related to a PCF, derived primarily from CBCL problem scores, suggests a role for cortical thinning in the psychosocial comorbidities of CWE. Reduced CT in pediatric new onset seizures (Dabbs et al., 2013) and in children without epilepsy with abnormal CBCL anxiety/depression scores (Ducharme et al., 2014}, attention deficit hyperactivity (Shaw et al., 2006), depression (Luby et al., 2016), and lower IQ (Menary et al., 2013) support this conclusion. Funding: NS 32070, MH 6718