Control groups in paediatric epilepsy research: do first-degree cousins show familial effects?

To determine whether first-degree cousins of children with idiopathic focal and genetic generalized epilepsies show any association across measures of cognition, behaviour, and brain structure. The presence/absence of associations addresses the question of whether and to what extent first-degree cousins may serve as unbiased controls in research addressing the cognitive, psychiatric, and neuroimaging features of paediatric epilepsies. Participants were children (aged 8-18) with epilepsy who had at least one first-degree cousin control enrolled in the study (n=37) and all enrolled cousin controls (n=100). Participants underwent neuropsychological assessment and brain imaging (cortical, subcortical, and cerebellar volumes), and parents completed the Child Behaviour Checklist (CBCL). Data (based on 42 outcome measures) from cousin controls were regressed on the corresponding epilepsy cognitive, behavioural, and imaging measures in a linear mixed model and case/control correlations were examined. Of the 42 uncorrected correlations involving cognitive, behavioural, and neuroimaging measures, only two were significant (p<0.05). The median correlation was 0.06. A test for whether the distribution of p values deviated from the null distribution under no association was not significant (p>0.25). Similar results held for the cognition/behaviour and brain imaging measures separately. Given the lack of association between cases and first-degree cousin performances on measures of cognition, behaviour, and neuroimaging, the results suggest a non-significant genetic influence on control group performance. First-degree cousins appear to be unbiased controls for cognitive, behavioural, and neuroimaging research in paediatric epilepsy.