Deviation from normative brain development is associated with symptom severity in autism spectrum disorder

Area of Science:

• Neuroscience
• Developmental Psychology
• Biomedical Engineering

Background:

• Autism spectrum disorder (ASD) is a complex neurodevelopmental condition with significant heterogeneity.
• Understanding how brain development deviates in ASD and relates to behavioral outcomes is crucial.
• This study hypothesizes a link between the degree of atypical brain development and ASD symptom severity.

Purpose of the Study:

• To investigate the relationship between deviations in brain development and ASD symptom severity.
• To establish a normative model of brain development using machine learning.
• To characterize individual differences in ASD based on brain development trajectories.

Main Methods:

• Compared anatomical and diffusion MRI metrics between ASD (n=247) and typically developing children (TDC) (n=220) aged 6-25.
• Developed a machine learning model to predict 'brain age' from MRI metrics in TDC, creating a normative development model.
• Calculated a developmental deviation index (DDI) in ASD individuals by comparing brain age to chronological age and correlated DDI with ASD symptom severity.

Main Results:

• Machine learning accurately predicted age in both TDC (r=0.88) and ASD (r=0.85) samples, with diffusion metrics being key predictors.
• A significant correlation (r=-0.2) was found between DDI derived from fractional anisotropy and ASD symptom severity.
• Individuals with advanced brain age exhibited lower ASD symptom severity, whereas those with delayed brain age showed higher severity.

Conclusions:

• The degree of deviation from typical brain development is associated with ASD symptom severity, contributing to observed heterogeneity.
• The developed approach allows for individual characterization relative to normative brain development.
• This method aids in identifying distinct developmental subtypes within ASD, enhancing understanding of its heterogeneity.