The relationship between regression in autism spectrum disorder, epilepsy, and atypical epileptiform EEGs: A meta-analytic review

Background The strength of the relationship between epilepsy and regression in autism spectrum disorder (ASD) has been much discussed but is currently unclear. Methods The authors conducted 2 meta-analyses of published studies to determine if children with ASD who experience regression (ASD-R) epilepsy are more likely to have epilepsy or atypical epileptiform electroencephalograms (aeEEG) compared to individuals with ASD and no regression history (ASD-NR). Results Children with ASD-R were more likely to have epilepsy and atypical aeEEG compared to individuals with ASD-NR. Odds ratio (OR) analyses indicated minor increases in odds of aeEEG regression for ASD-R compared to ASD-NR, OR = 1.29, 95% CI [1.09, 1.52], and small increased odds of reported epilepsy, OR = 1.59, 95% CI [1.31, 1.95], for the ASD-R compared to the ASD-NR group. The operational definition of epilepsy affected reported odds of aeEEG, but not reported epilepsy. The likelihood of aeEEG in children with ASD-R was also moderated by the medication status of participants or whether participants had received sleep EEGs. Across analyses, regression definitions, sample type, diagnostic systems, diagnostic instruments, and source of regression history were unrelated to group differences. Conclusions Altogether, this study indicates that the relationship between regression and epilepsy or aeEEG is relatively weak.