Characterizing predictors of response to behavioral interventions for children with autism spectrum disorder: A meta-analytic approach

Abstract

A comprehensive understanding of specific factors contributing to variability in responsiveness of children with autism to interventions is paramount for making evidence-based clinical and policy decisions. This meta-analysis examined child and family characteristics, as well as intervention design factors, associated with outcomes of behavioral interventions for children with autism. A systematic review identified 95 studies published between 1987 and 2024, encompassing 6780 children on the autism spectrum and 2150 independent effect sizes. Results indicated that stronger post-intervention effects were observed across intervention approaches for children with higher cognitive, language, and other developmental abilities, greater adaptive functioning, and fewer autism-related features. Additionally, interventions of longer duration and greater total hours were associated with stronger post-intervention outcomes. In contrast, intervention approach (Early Intensive Behavioral Intervention, Naturalistic Developmental Behavioral Interventions, or Developmental Interventions), delivery agent, and child age at intervention onset did not significantly predict the strength of post-intervention outcomes. While study methodology and reporting quality were marginally associated with predictive strength, adjusting for these factors had minimal impact on the reported findings. The insights from this meta-analysis have significant implications for the development of personalized intervention models for children with autism. These models have the potential to optimize outcomes and offer critical guidance for decision-making in both the service and policy levels, ensuring efficient and equitable allocation of resources.