Abstracts

Rationale: Epilepsy can affect both episodic and semantic memory function in adults. Although it has been demonstrated that children with epilepsy have episodic memory impairments, semantic memory has not yet been systematically investigated in pediatric epilepsy. Smith, Elliott and Lach (Child Neuropsychology, 2006, 12(3):151-64) reported that children and adolescents subjectively described difficulties in the semantic memory realm. However, Vargha-Khadem et al. (Science, 1997, 277(5324):376-80) described that early onset structural lesions in the mesial temporal lobe resulted in selective impairments in episodic memory with preserved development of semantic memory. The present study examined both episodic and semantic memory performance in children with temporal lobe epilepsy (TLE). Methods: Participants were being investigated as candidates for epilepsy surgery, and had well documented unilateral seizure foci in the temporal lobe (34 left, 21 right). Mean age was 13.3 years (range 5.0-17.9); mean age of seizure onset was 7 years (range 0-17). There were no differences between the groups with respect to age, age at seizure onset, number of antiepileptic drugs, or sex. Four measures of verbal semantic knowledge were administered: vocabulary (defining words), naming pictures, animal fluency, and general information. Two verbal episodic memory tasks were also given: delayed story recall and delayed recall of a list or words (learned over five presentation trials). Performance on all tasks was converted to z-scores based on age norms. Results: Although children with LTLE had somewhat lower scores on the tasks than those with RTLE, no significant differences were found relating to laterality. Performance in both groups was significantly lower than the level expected based on population age norms on vocabulary, information, naming, story recall and word list recall. Episodic and semantic memories were equivalently affected. Older age at seizure onset was associated with better performance on naming (r=.37, p=.014) and story recall (r=.36, p=.012). Duration of epilepsy was negatively correlated with naming (r=-.31, p=.044) and vocabulary (r=-.371, p=.007), but with neither episodic memory measure. Conclusions: These findings demonstrate that children with TLE have impairments in both episodic and semantic memory. The deficits appear equivalent in magnitude. Although the tasks emphasized verbal material, both LTE and RTE had a negative impact on the development of these skills. This is the first systematic investigation of semantic memory in pediatric epilepsy. The results add to our knowledge of memory in this population by showing that temporal-lobe seizures adversely affect the encoding or retrieval of information in semantic memory. The contrast in results with respect to Vargha-Khadem et al.’s (1997) findings of intact semantic memory after mesial temporal lobe lesions may reflect differential involvement of the lateral and the mesial temporal lobe structures between the two studies, or the negative and perhaps more widespread influence of ongoing seizures in the present sample.