Atypical Cortical Gyrification but Not Autism Spectrum Disorder Diagnosis Predicts Differences in White-Matter Wiring

Background: Evidence suggests that Autism Spectrum Disorder (ASD) is accompanied by neurodevelopmental differences in brain structure and connectivity. More specifically, it is thought that the brain in ASD is over-connected locally, and under-connected globally (Belmonte et al., 2004). While evidence for global under-connectivity is growing (Koshino et al., 2008; Pugliese et al., 2009), there is currently a lack of neuroimaging studies examining in vivo proxy measures for local grey-matter wiring (Ecker et al., 2013). Furthermore, it is currently unknown how proxy measures of grey-matter connectivity relate to differences in white-matter wiring in ASD.

Objectives: Here, we therefore examined (1) vertex-wise differences in local gyrification (Schaer et al., 2008) as a proxy measure for atypical microstructural grey-matter connectivity between individuals with ASD and controls, and (2) established the relationship between atypical local gyrification and local differences in white-matter connections.

Methods: Structural MRI and DTI data was collected on 51 well-characterized male adults with an ADI-R confirmed diagnosis of ASD (mean age 26+7 years, FSIQ 112+13), and 49 matched typically developing male controls (mean age 28+6 years, FSIQ 115+10). Surface reconstructions for all participants were performed using FreeSurfer software on the basis of high-resolution T₁-weighted inversion recovery images. Between-group differences in local gyrification index (LGI) were examined using a GLM on the vertex level. The individual’s structural MRI data (both surfaces and volumes) were subsequently co-registered with the DTI data in order to create regions-of-interest for automated fibre tracking using the surface-based clusters of significant between-group differences in LGI. Between-group differences in diffusion measures were then examined for tracts originating or terminating in the ROI. Last, we established the relationship between variations in LGI and diffusion measures within and between groups.

Results: Individuals with ASD had significantly increased local gyrification in a large left-hemisphere cluster centered on the central sulcus, including the primary and pre-motor cortex and the primary somatosensory cortex (t_max = 3.33, n_vertices = 8996, p_cluster = 0.039) relative to controls. Within the cluster of between-group differences in LGI, individuals with ASD also had a significant increase in axial diffusivity (diffusivity along axons), particularly in short intra-regional U-shaped fibers. Last, we found that the degree of axial diffusivity of tracts within the surface-based ROI was predominantly correlated with the degree of cortical folding, rather than diagnostic category.

Conclusions: Our study demonstrates that the degree of cortical gyrification of the outer grey-matter surface of the brain is regionally enhanced in individuals with ASD, and that grey-matter differences - but not diagnostic status (i.e. having a diagnosis of ASD) - significantly predicted the white-matter connectivity of these regions. Thus, white-matter differences in ASD should not be interpreted in isolation (i.e. without consideration of grey-matter differences) and may be secondary to an abnormal development of grey-matter.