Exploring autism spectrum profiles via network analysis of parent-reported ASSQ patterns

Area of Science:

• Neuroscience
• Psychiatry
• Developmental Psychology

Background:

• Autism spectrum disorder (ASD) presents a heterogeneous behavioral profile.
• Total score analyses may obscure relationships between individual behavioral items.
• Network analysis offers a method to explore the internal structure of ASD behaviors.

Purpose of the Study:

• To apply network analysis to parent-reported Autism Spectrum Screening Questionnaire (ASSQ) data.
• To explore the associative structure among core ASD behavioral domains.
• To identify key behavioral items within the ASD network.

Main Methods:

• Retrospective study of 995 children and adolescents (7-16 years) diagnosed with ASD.
• Utilized ordinal (Gaussian Graphical Model) and dichotomized (Ising model) network strategies on ASSQ data.
• Identified central and bridge nodes, and conducted Network Comparison Tests for sex and age differences.

Main Results:

• The ordinal network model (ASSQ-GGM) showed stable centrality.
• Key central nodes identified: inability to adjust (Q11), unusual posture (Q27), and difficulty with peer relationships (Q15).
• Bridge nodes included unusual facial expression (Q26), inability to adjust (Q11), and lack of empathy (Q12).

Conclusions:

• A network of interrelated ASD-related behaviors was identified using ASSQ items.
• No significant differences in network structure were found based on sex or age.
• Findings are hypothesis-generating, highlighting associative patterns; further longitudinal studies are needed for causal inference.