The Intrinsic Geometry of the Cerebral Cortex In AUTISM – the Relationship BETWEEN Cortical Folding and White MATTER Wiring

Background: Evidence suggests that the brain in Autism Spectrum Disorder (ASD) undergoes a period of precocious accelerated growth during early postnatal life, followed by an atypically slow or arrested growth during childhood.  The early overgrowth is idiosyncratic for different lobes of the brain with frontal and temporal lobes being more affected than occipital and parietal lobes.  Such differential growth will not only affect the overall size of the brain in ASD but also influence the way the brain is shaped and ‘wired’.  Despite growing evidence for atypical structural and functional connectivity coming from DTI and fMRI studies, it is currently unknown how the intrinsic cerebral organization of the cortex affects local and global wiring in ASD.

Objectives: Our objective was therefore to examine how the intrinsic geometry of the cortex affects local and global wiring of the brain in ASD.  Furthermore, we aimed to demonstrate that regional patterns of cerebral connectivity affect autistic symptoms.

Methods: Structural MRI data was collected on 34 well-characterized male adults with an ADI-R confirmed diagnosis of ASD (mean age=26yrs, mean FSIQ=112), and 34 age/FSIQ matched neurotypicals.  Surface reconstructions for all participants were performed using FreeSurfer software on the basis of high-resolution structural T₁-weighted inversion recovery images.  The intrinsic geometry of the grey matter surface was examined using geodesic distances (i.e. shortest paths linking to points on a surface) and geodesic circles.  This allowed us to estimate intra- and inter-regional wiring costs for different brain regions.

Results: Overall, the intrinsic geometry of the cortex in ASD differed significantly from neurotypicals in terms of local and global wiring costs predominantly in anterior temporal, prefrontal and central regions (pre-/post-central gyrus).  In ASD, we observed significant increases in local connectivity around the temporal pole, the pre-central gyrus, and the dorsolateral prefrontal cortex.  At the same time, these regions displayed significantly reduced inter-areal connectivity (i.e. enhanced wiring costs) in ASD relative to controls.  Significant differences in regional cortical geometry were accompanied by differences in white matter connectivity, and were correlated with autistic symptoms in the social domain.

Conclusions: Our results confirm the hypothesis that the brain in adults with ASD is abnormally shaped and wired, which most likely results from an altered trajectory of brain development during early childhood.  Atypical patterns of connectivity predominantly affect phylogenetically younger brain regions maturing later during development, and are associated with white matter deficits.  Taking together these findings suggest that regional differences in cortical folding in addition to differences in brain volume might underlie the social deficits observed in ASD.