It has long been known that childhood epilepsy is associated with cognitive deficits, academic underachievement, and behavior problems.1 Recent studies characterizing these problems show them to be present at diagnosis, with some occuring before the first recognized seizure.2,3 Although risk factors have been identified, the difficult task of disentangling the myriad contributing and confounding risk factors remains. With regard to neuropsychological and behavioral outcomes, a handful of studies over the past 10 years have attempted to address these questions at or near seizure onset. These studies have contributed to our understanding, but the picture remains incomplete because investigations to date have faced weaknesses in their methodology that limit the yield of the data. These include relatively small samples, exclusion of symptomatic syndromes, lack of control groups, lack of specificity in measurement, and testing many months after seizure onset.
In their study identifying the effects of seizure-related factors on cognition and behavior in a sample of children with new onset epilepsy, Bhise et al.4 addressed many limitations of prior studies. They measured outcomes using standardized, individually administered neuropsychological tests and tested children close to diagnosis. Most importantly, data on cognition and behavior were collected before treatment with anticonvulsant medication. Other aspects of the study were less strong. For example, although they assessed the onset and number of prior seizures, the impact of these seizures was not explored. Of greater concern was the emphasis on comparisons among subgroups despite small subgroup sizes, especially for some tests in which there were many missing observations. The small subgroups could have resulted in erroneous conclusions because of inadequate power. Moreover, the lack of a control group limited the conclusions that could be drawn, especially because the participants were high-functioning. For example, vocabulary scores (a proxy for IQ) were in the high average range for some subgroups whereas scores for other tests were in the average range; if compared with a control group, the latter scores might reflect mild cognitive deficits.
In a recent study by our teams5, we had a unique opportunity to follow a large cohort of children (n=349) from the onset of the first recognized seizure, together with a large control group of healthy siblings. As a whole, children with first recognized seizures exhibited difficulties in multiple neuropsychological functions compared with their healthy siblings. Our large and diverse sample permitted us to make comparisons among subgroups and to control for potentially confounding factors. For example, multiple seizures, epileptiform activity on electroencephalograph, and antiepileptic medications were independently associated with lower neuropsychological scores, and inclusion of both idiopathic and symptomatic/cryptogenic syndromes revealed profiles that could not be appreciated with idiopathic-only samples.
In the report by Bhise et al. little descriptive information was provided on behavior variables. The major finding was that children with primary generalized seizures had more problems than with children with focal seizures. One explanation might be that, in this particular sample, children with primary generalized epilepsy also had lower IQ scores and more memory problems than the children with focal seizures. Lower IQ has been found to be associated with more behavior problems.6 Another possible explanation is that children with primary generalized seizures might have had more prior seizures. Prior seizures, unanalysed in the present study, were shown to be related to behavioral problems in a similar sample.2
In conclusion, we agree with the summary statements by Bhise and colleagues that children at risk for neuropsychological and behavioral problems require prompt evaluation and intervention. Based on recent findings, neuropsychological evaluation would be indicated and beneficial at the time of the first recognized seizure to identify the needed intervention.