Beyond group differences: Exploring the preliminary signals of target engagement of an executive function training for autistic children

Understanding both for whom and how interventions work is a crucial next step in providing personalized care to children with autism spectrum disorder (ASD). Autistic children present with heterogeneity both within core ASD criteria and with respect to co‐occurring mental health challenges, which may affect their ability to benefit from intervention. In a secondary data analysis of a randomized control trial evaluating an executive function (EF) training with 70 7‐ to 11‐year‐old autistic children, we explored: (1) whether co‐occurring attention‐deficit/hyperactivity disorder (ADHD) features or anxiety features at baseline moderated the extent to which children benefited from the EF training. In other words, we asked, “For whom is training effective?” We also explored: (2) the extent to which changes in a brain‐based measure of target engagement predicted the clinical outcomes of the EF training. This is a step towards asking, “How is training effective?” We found that EF training improved behavioral inhibition only for children with clinically significant co‐occurring ADHD features. Anxiety features, while prevalent, did not moderate EF training efficacy. Finally, for the EF training group only, there was a significant correlation between pre‐to‐post change in an EEG‐based measure of target engagement, N2 incongruent amplitude during a flanker task, and change in repetitive behaviors, a behavioral outcome that was reported in the parent RCT to have improved with training compared to waitlist control. This study provides preliminary evidence that EF training may differentially affect subgroups of autistic children and that changes at the neural level may precede changes in behavior. LAY SUMMARY: Understanding both for whom and how interventions work will help us provide personalized care to children with autism spectrum disorder (ASD). Autistic children present with many different strengths and challenges. Co‐occurring mental health challenges may affect how much autistic children benefit from intervention. We analyzed secondary data from a rigorously designed pilot intervention study, a randomized control trial (RCT), that enrolled 70 7‐ to 11‐year‐old autistic children to assess whether a set of computer‐based executive function (EF) training games improved their performance. Executive functions include being able to shift between tasks, inhibit a response, and keep information in working memory. In the current study, we explored: (1) whether children's co‐occurring attention‐deficit/hyperactivity disorder (ADHD) features or anxiety features, measured before the EF training began, affected how much they benefited from the EF training. In other words, we asked, “For whom is training effective?” We also explored: (2) whether children's brain‐based changes in EF predicted their performance in everyday life (e.g., parent report on a survey). This is a step toward asking, “How is training effective?” We found that EF training improved children's inhibition ability, but only for children with clinically significant ADHD features. While many children in our sample also had anxiety features, we found that anxiety levels did not affect how well the EF training worked. Finally, for children who received the EF training, changes in a brain‐based measure of conflict monitoring (i.e., being able to noticing differences in stimuli) predicted changes in children's repetitive behaviors. This study provides early evidence that EF training may be more effective for some autistic children than others, especially those with ADHD features.