Altered Measures of Diffusion at the Grey-White Matter Boundary in Autism Spectrum Disorder
Autism spectrum disorder (ASD) is a highly complex neurodevelopmental condition that is accompanied by neuroanatomical differences on the macroscopic and microscopic level (Ecker, 2016). For example, a recent histological investigation suggests that the grey-white matter boundary is less well defined in ASD, potentially due to an abnormal cell patterning at the grey-white matter (GWM) transition zone (Avino & Hutsler, 2010). Consistent with this finding, it has also been shown in vivo that the GWM tissue contrast at the white matter boundary is significantly reduced in ASD, and in many brain regions that have previously been linked to autistic symptoms and traits (Andrews et al., 2017). However, nobody has yet characterized the GWM boundary based on measures of diffusion in ASD.
Thus, we examined differences in fractional anisotropy (FA) and mean diffusivity (MD) at the GWM boundary in ASD individuals relative to controls. Based on previous reports of a reduced tissue contrast, we also expected significant reductions in measures of diffusion.
92 adults with ASD (53 males, 39 females; diagnosed using ADI-R (Lord et al., 1994)), and 92 matched healthy controls (51 males, 41 females) aged 18-52 years were recruited and assessed at the King`s College London, London, and the University of Cambridge. For all 184 participants, high-resolution structural T1-weighted volumetric images and Diffusion Tensor Imaging (DTI) data were acquired. Cortical surface reconstructions were produced using FreeSurfer v6.0 (http://surfer.nmr.mgh.harvard.edu/). FA and MD measures were derived by mapping the diffusion data onto the white matter surface. Parameter estimates for FA and MD were estimated by regression of a general linear model (GLM) at each vertex along the GWM boundary with (1) diagnostic group, gender, and site as categorical fixed-effects, (2) a group-by-gender interaction, and (3) age and full-scale IQ as continuous covariates. Corrections for multiple comparisons were performed using ´random field theory` (RFT)-based cluster analysis using a p<0.05 (2-tailed) cluster-significance threshold.
We found that MD was significantly decreased in ASD at the GWM boundary underneath the dorsolateral (DLPFC) and medial orbitofrontal cortices, the cingulate cortex (CC), and in the medial somatosensory cortex (SMA). An increase in MD was observed in the right fusiform gyrus. When examining measures of FA, we observed significant decreases in ASD in the right DLPFC, the fusiform gyrus and entorhinal cortex, and in the right occipital lobe. Measures of FA were significantly increased in ASD in the right medial prefrontal cortex, and the right medial and lateral orbitofrontal cortex. In addition, we also observed significant group-by-gender interactions for both FA and MD in many regions with a significant main effect of group.
Our results confirm that the white matter microstructure at the GWM boundary is atypical in ASD, potentially as a result of an abnormal cell patterning at the GWM transition zone. Thus, neuroanatomical features measured at the GWM boundary might be particularly sensitive for identifying and characterizing ASD. Given the significant group-by-gender interactions, it is however likely, that this aspect of the neuropathology of ASD is modulated by biological sex.