Abstracts

To evaluate the spectrum of EEG abnormalities in children with autistic spectrum disorder and their association with a family history of seizures or autism. Patients seen at two large university-based pediatric neurology practices with a diagnosis of autistic spectrum disorder (ASD) and who had an EEG were identified by computer database review. Children with a known genetic disorder, cerebral malformation or significant CNS injury were excluded. EEGs were considered to show epileptiform activity if spikes, sharp waves or electrographic seizures were reported. Detailed family history was documented for 88. 142 children with a neurologic diagnosis of ASD were identified in whom at least one EEG had been performed. 48 (33%) had clinical seizures or epilepsy, and an additional 19 had questionable seizures. 66 (46%) had epileptiform EEGs, including 25 with known seizures. 47 had normal EEGs and 29 had only non-epileptiform abnormalities on EEG (primarily background slowing or asymmetry). Among children with well-defined EEG patterns, 55% had a single epileptic focus and the remainder had multifocal or generalized epileptiform activity. Children with known seizures were more likely to have abnormal EEGs than those without (54% vs 44%). Children with reported developmental regression were also more likely to have epileptiform EEGs than those with no known regression (59% vs 41%). A family history of seizures was found at similar rates in those with and without epileptiform EEGs (26% vs 23%). However, a family history of ASD was more common in those with epileptiform EEGs (20% vs 9%). There appeared to be a trend for a family history of ASD to be more prominent in children with generalized or multifocal spikes (41%), than in those with a single epileptic focus (6%). Seizures occur in a significant minority of children with ASD, and epileptiform EEGs are even more frequent. The lack of association between family history of seizures and epileptiform EEGs may reflect a lower threshold for obtaining EEGs in these children. Our finding that epileptiform EEG activity was more common in children with developmental regression is consistent with those of other investigators, and emphasizes the overlap between this diagnosis and Landau-Kleffner syndrome. The increased frequency of a family history of ASD in children with epileptiform EEGs is a novel finding. Family and twin studies strongly implicate a genetic etiology for at least a subset of children with ASD. If epileptiform EEG patterns are more common in families with a genetic predispostion for autism, EEG may serve as a marker for a subset of children in whom further genetic investigation is warranted. This is particularly true if our finding of a more specific association between family history of ASD and a generalized or multifocal epileptiform EEG patterns holds true in further studies.