Abnormalities in Inter-Hemispheric Cortical Thickness Correlations in ASD

Background:  Converging evidence from structural and functional neuroimaging studies points to disrupted connectivity in autism spectrum disorder (ASD), specifically reduced long-range connectivity (Just et al. 2004).  But, the implications of this under-connectivity in terms of brain organization remain unclear.  Correlations in cortical thickness have been shown to be useful in distinguishing controls from individuals with Alzheimer's Disease (He et al. 2008), Schizophrenia (Bassett et al. 2008) and Multiple Sclerosis (He et al. 2009), and offer additional information with respect to cortical organization.  To our knowledge, there are, to date, no studies of cortical thickness correlations patterns in ASD.

Objectives:  We hypothesized that the cortical thickness correlation networks would show prominent changes in ASD, specifically in inter-hemispheric brain regions, as the corpus callosum has been shown to be reduced in ASD (Frazier and Hardan 2009).  Through analysis of the cortical thickness correlation patterns in ASD, we seek to provide additional insight into the impact of the long distance under-connectivity on brain organization.

Methods:  T1-weighted scans were collected from 22 male adults with ASD (34.14 ± 10.67 years) and 22 male controls (31.68 ± 8.75 years).  CIVET, a fully automated structural image analysis pipeline developed at the Montreal Neurological Institute, was used to construct gray- and white-matter surfaces for each subject (Kim et al 2005). The distance between these surfaces was computed at 81,924 points; this was the measure of cortical thickness.  The automated anatomical  labelling (AAL) template, a commonly used atlas (Tzourio-Mazoyer et al. 2002), was used to parcellate cortex into 78 regions, and mean cortical thickness in each region was computed and used in the analysis.  Correlations in cortical thickness between AAL regions were obtained for ASD and controls after controlling for age within each group. The resultant correlation matrices were z-transformed and group comparisons were done for the normal and ASD groups.  

Results:  Analysis of correlations in cortical thickness showed prominent changes in inter-hemispheric brain regions for the ASD as compared to the control group, but no significant difference in correlations for the intra-hemispheric brain regions.  Group comparisons after z-transformation showed significantly increased correlations in several inter-hemispheric cortical regions. 

Conclusions:  Our results of increased correlations in several inter-hemispheric cortical regions in ASD as compared to normals might point to a suboptimal brain organization in ASD. In normal adults, inter-hemispheric cortical regions which are anatomically connected via the corpus callosum exhibit weak correlations in cortical thickness, possibly reflecting hemispheric lateralization. The inter-hemispheric under-connectivity that has been shown in ASD suggests abnormal lateralization. The overall increased cortical thickness correlations across inter-hemispheric cortical regions suggest that this abnormal lateralization takes the form of greater duplication of function between hemispheres.