Alterations of functional connectivities associated with autism spectrum disorder symptom severity: a multi-site...

Area of Science:

• Neuroscience
• Psychiatry
• Medical Imaging

Background:

• Autism spectrum disorder (ASD) is a complex neurodevelopmental condition with significant heterogeneity.
• Accurate ASD severity estimation is crucial for tailored clinical diagnosis and intervention.
• Previous studies on resting-state functional connectivities (RSFCs) and ASD severity were limited by small sample sizes and focus on local connections.

Purpose of the Study:

• To identify a subset of whole-brain RSFCs that are highly predictive of ASD severity using a large-sample dataset.
• To investigate the specific network alterations associated with varying levels of ASD severity.
• To leverage multivariate pattern analysis for robust ASD severity estimation.

Main Methods:

• Utilized multivariate pattern analysis on large-sample resting-state functional magnetic resonance imaging (fMRI) data.
• Applied regression estimation to correlate identified RSFCs with clinically assessed ASD severity.
• Analyzed alterations in both inter-network and intra-network connectivity patterns.

Main Results:

• Achieved a significant Pearson correlation of 0.5 between estimated and observed ASD severity, with a mean absolute error of 1.41.
• Identified specific RSFCs that significantly change with ASD severity.
• Observed abnormalities predominantly in sensorimotor and default mode networks, with inter-hemispheric and left-hemisphere lateralization.

Conclusions:

• The study provides strong evidence for altered RSFCs correlating with ASD severity in a large cohort.
• Identified RSFCs highlight abnormalities in large-scale functional brain networks in individuals with ASD.
• These findings enhance our understanding of the neurobiological underpinnings of ASD severity and heterogeneity.