Event-Related Potentials Linked to Executive Function Reflect Individual and Developmental Differences in Children with Autism Spectrum Disorders

Background: Many children with autism spectrum disorder (ASD) have difficulty with executive function (EF), which involves complex problem solving and goal-directed behavior, in comparison to typically developing children. The N2 event-related potential (ERP) component is generally associated with EF, and may specifically reflect the neural correlates of inhibition and conflict monitoring (Faja et al., 2016). Previous research has shown differences in the scalp distribution of the N2 component, with more widespread posterior and left-lateralized activation for younger children and children with reduced EF (e.g., Lamm et al., 2006). Examining age-related differences in the neural underpinnings of the N2 component in children with ASD may offer insight into the nature of EF challenges in ASD (difference versus delay) and enable interpretation of individual differences relative to age expectations.

Objectives: To explore individual differences and development of EF by examining age-related changes in behavioral performance and N2 responses, particularly their scalp distribution, during a Go/NoGo task.

Methods: Participants included 67 children diagnosed with ASD aged 7 to 11 grouped by age (40 children ages 7-9.49 and 27 children ages 9.5-11 years). EEG was collected while participants completed a cued Go/NoGo task. Neural responses were recorded and N2 mean amplitude and latency was characterized for the go and nogo conditions at Fz and Cz. Twenty-one younger and 16 older children provided usable ERP data. Accuracy was also collected during the task for all participants.

Results: Repeated-measures ANOVA (condition x electrode site x age group) with peak amplitude suggested that the N2 component becomes more focused around Fz with age. Older children with ASD exhibited stronger activation at Fz relative to Cz, in comparison to younger children with ASD, who showed strong activation at Fz and moderate activation at Cz (electrode x age: F(1,35)=4.31, p=0.045). Behaviorally, accuracy was higher for the go than nogo condition (F(1,65)=123.93, p≤0.001). Older children were more accurate during both the go and nogo conditions, as compared to younger children (F(1,65)=13.62, p≤0.001) However, the interaction between age group and go/nogo condition was not significant (F(1,65)=1.79, p=0.186). Better performance across go/nogo conditions related to smaller peak amplitude (greater efficiency) in the older cohort at Fz, r(16)=0.505, p=0.046, but was not significant in the younger cohort.

Conclusions: Individual and developmental differences in scalp activation of the N2 component appear related to neural activation during an EF task. Increased specificity of N2 peak amplitude at Fz with age compared to Cz may indicate greater anterior shift in the neural source of the N2 or a decreased posterior contribution as EF develops in school age children with ASD. Additionally, the relation between greater neural efficiency at Fz and improved behavioral performance across conditions in the older cohort suggests individual differences in EF ability beyond the effects of age can be observed at the neural level.