Heterogeneity of executive function revealed by a functional random forest approach across ADHD and ASD

Area of Science:

• Neuroscience
• Developmental Psychology
• Computational Psychiatry

Background:

• Autism spectrum disorder (ASD) and attention-deficit-hyperactivity disorder (ADHD) share symptoms like hyperactivity and inattention, potentially linked to atypical executive function (EF).
• Existing research shows inconsistent findings regarding shared or distinct EF features across ASD and ADHD, possibly due to the heterogeneity within these conditions.
• This study hypothesizes the existence of nested subgroups within ASD and ADHD exhibiting shared or unique underlying neurobiological mechanisms.

Purpose of the Study:

• To investigate shared and distinct executive function (EF) profiles in adolescents with autism spectrum disorder (ASD) and attention-deficit-hyperactivity disorder (ADHD).
• To identify potential neurobiological underpinnings of these conditions using machine learning and resting-state functional connectivity MRI (rs-fMRI).
• To explore the existence of nested subgroups within ASD and ADHD based on EF performance.

Main Methods:

• A cohort of 130 adolescents (7-16 years) with ASD (n=64) and ADHD (n=66), plus 28 typically developing (TD) controls, underwent cognitive assessments focusing on EF.
• A two-stage machine learning approach, functional random forest (FRF), was employed for classification and subgroup identification using 43 EF variables.
• Resting-state functional connectivity MRI (rs-fMRI) was utilized to assess the neurobiological validity of identified subgroups.

Main Results:

• Executive function (EF) features significantly predicted both hyperactive and inattentive symptoms of ADHD better than chance.
• The FRF successfully identified two distinct subgroups for each symptom domain (hyperactive and inattentive), characterized by severe and mild EF performance.
• Neuroimaging revealed significant differences in functional brain network connectivity within and between identified subgroups and TD participants, but no clear severity-based pattern was observed.

Conclusions:

• The study successfully identified distinct neurocognitive profiles of severe and mild executive function (EF) performance in subgroups of adolescents with ADHD and ASD.
• Functional brain connectivity differences between subgroups did not align with symptom severity, suggesting complex underlying mechanisms beyond observable symptoms.
• These findings highlight the heterogeneity within ASD and ADHD and the potential for distinct neurobiological pathways influencing executive function.