Background: Autism is accompanied by highly individualized patterns of neurodevelopmental differences in brain anatomy. This variability makes the neuroanatomy of autism inherently difficult to describe at the group level. Here, we examined inter-individual neuroanatomical differences using a dimensional approach that decomposed the domains of social communication and interaction (SCI), restricted and repetitive behaviors (RRB), and atypical sensory processing (ASP) within a neurodiverse study population. Moreover, we aimed to link the resulting neuroanatomical patterns to specific molecular underpinnings.
Methods: Neurodevelopmental differences in cortical thickness and surface area were correlated with SCI, RRB and ASP domain scores by regression of a General Linear Model in a large neurodiverse sample of N=288 autistic and N=140 non-autistic individuals, aged 6-30, recruited within the EU-AIMS Longitudinal European Autism Project (LEAP). The domain-specific patterns of neuroanatomical variability were subsequently correlated with cortical gene expression profiles via the Allan Human Brain Atlas.
Results: Across groups, behavioral variations in SCI, RRB and ASP were associated with interindividual differences in CT and SA in partially non-overlapping fronto-parietal, temporal, and occipital networks. These domain-specific imaging patterns were enriched for genes (i) differentially expressed in autism, (ii) mediating typical brain development, and that are (iii) associated with specific cortical cell types. Many of these genes were implicated in pathways governing synaptic structure and function.
Conclusions: Our study corroborates the close relationship between neuroanatomical variation and interindividual differences in autism-related symptoms and traits within the general framework of neurodiversity, and links domain-specific patterns of neuroanatomical differences to putative molecular underpinnings.
Keywords: MRI; autism spectrum disorder; autism symptomatology; brain structure.
Copyright © 2024. Published by Elsevier Inc.