A meta-analysis of literacy and language in children with rolandic epilepsy

Abstract

Aim: Rolandic epilepsy is the most common childhood epilepsy, often presenting with neuropsychological impairments. The aim of the study was to formally assimilate the findings of existing studies varying widely in methodology, thereby confirming the nature and prevalence of impairments in literacy and language.

Methods: Using meta-analytical techniques, we evaluated 22 studies of literacy and/or language skills in children with rolandic epilepsy, published after 2000, among participants with IQs>70 and in which effect sizes could be acquired. Diagnosis required the presence of classical centrotemporal spikes arising from a normal background on electroencephalograms; a clinical history including at least one seizure; and no additional neurological condition. Overall effect size and heterogeneity were measured for single-word reading, phonological processing, and expressive and receptive language.

Results: Mean effect sizes (Cohen's d) ranged from 0.50 (95% confidence interval [CI] 0.23-0.78) for phonological processing, through 0.71 (95% CI 0.52-0.90) for word reading and 0.72 (95% CI 0.34-1.1) for receptive language, to 0.75 (95% CI 0.45-1.05) for expressive language. While group differences for reading measures were consistent, those for language were heterogeneous and varied across studies explained by age and IQ of samples.

Interpretation: The presence of reading and phonological processing deficits in children with rolandic epilepsy highlights the importance of early literacy and language assessment in this population.