Autism spectrum disorders: a meta-analysis of executive function

Area of Science:

• Neuroscience
• Psychology
• Developmental Psychology

Background:

• Evidence regarding executive dysfunction (EF) in autism spectrum disorders (ASD) is inconsistent.
• Understanding EF's role is crucial for effective ASD diagnosis and intervention strategies.

Purpose of the Study:

• To meta-analyze EF performance in ASD, examining subdomain fractionation, clinical utility of measures, and moderator influences (age, gender, diagnosis).
• To assess the clinical sensitivity of EF measures for ASD diagnosis and treatment.

Main Methods:

• Searched Embase, Medline, and PsychINFO for peer-reviewed studies (1980-June 2016) comparing EF in ASD with neurotypical controls.
• Utilized a random-effects model with subgroup analysis for moderators. Primary outcome: Hedges' g effect size. Clinical sensitivity: overlap percentage (OL%).
• Followed PRISMA guidelines for reporting. Included 235 studies with 14,081 participants (6,816 ASD, 7,265 controls).

Main Results:

• A moderate overall effect size (Hedges' g=0.48) for reduced EF in ASD was observed, consistent across domains.
• Most moderator comparisons were non-significant; EF dysfunction effect size decreased post-DSM-III introduction.
• Few EF measures demonstrated clinical sensitivity; fractionation of EF into subdomains was not supported.

Conclusions:

• Confirms broad, development-stable executive dysfunction in ASD.
• Current EF measures lack diagnostic sensitivity and do not effectively fractionate into specific subdomains.
• Prioritizes development of feasible EF measures with enhanced clinical sensitivity for ASD diagnosis and treatment.