Reported Epilepsy and Abnormal EEG Activity in Individuals with ASD with and without Regression: A Meta-Analysis

Background: A recent meta-analysis reported different rates of regression depending on whether the term regression was operationalized as encompassing mixed, language, social, or language/social (Barger, Campbell, & Donald, 2011). This work is extended by investigating whether children with an autism spectrum disorder with regression (ASD-R) and those without regression (ASD-NR) differ in regards to epilepsy related measures (reported epilepsy and abnormal EEG).  

Objectives:  We performed meta-analyses of the literature comparing children with ASD-R and ASD-NR on (a) reported rates of epilepsy and (b) rates of abnormal EEG readings. 

Methods:  Data regarding reports of epilepsy were extracted from 16 published studies across 2,533 participants. Six studies reported data on mixed, 7 studies reported data on language, no studies reported data on social, and 3 studies reported data on language/social regression. Abnormal EEG readings were extracted from 14 studies across 3,834 participants. Eight studies reported data on mixed, 5 studies reported data on language, no studies reported data on social, and 2 studies reported data on language/social regression. 

Results:  Across all regression types, reports of epilepsy were no more likely for ASD-R (22%) than ASD-NR (18%). This same pattern was found for mixed (ASD-R=34%; ASD-NR=18%), language (ASD-R=13%; ASD-NR=12%), and language/social regression (ASD-R=17%; ASD-NR=12%). Prevalence of social regression could not be calculated due to a lack of data. Furthermore, for children with ASD-R, reported rates of epilepsy were significantly higher for reports of mixed (34%) than reports of language (13%) or language/social (17%) regression. Furthermore, across all regression types, abnormal EEG readings were no more likely for ASD-R (44%) than ASD-NR (48%). This same pattern was found for mixed (ASD-R=48%; ASD-NR=52%), language (ASD-R=32%; ASD-NR=25%), and language/social regression (ASD-R=39%; ASD-NR=35%). Prevalence of social regression could not be calculated due to a lack of data. Finally, for children with ASD-R, reported rates of abnormal EEG readings were higher when mixed regression was reported (48%) than when language (32%) or language/social (39%) regressions were reported. 

Conclusions:  The data reported here indicate no differences between ASD-R and ASD-NR children in regards to reported epilepsy and abnormal EEG readings for any of the regression types. In the ASD-R group, both rates of reported epilepsy and abnormal EEG readings were higher for studies investigating mixed regression compared to language or language/social regressions.