Abstract/Summary

Autism Spectrum Disorders are a set of neurodevelopmental conditions characterised by difficulties in social interaction and communication as well as stereotyped and restricted patterns of interest. With recent advances in neuroimaging techniques and analytical approaches, a considerable effort has been directed towards identifying the neuroanatomical underpinnings of Autism Spectrum Disorders (ASD). Most of the previous studies have treated ASD as a category, using a case-control design to identify the neuroanatomical correlates of ASD. However, it is well�recognised that autistic traits exist in a continuum across the general population, whilst the extreme end of this distribution is diagnosed as clinical ASD. Therefore, we sought to investigate the neural correlates of autistic traits in the clinical and non-clinical population using a dimensional approach. To this end, the proposed research measured the structural brain volumes (using voxel-based morphometry and surface-based morphometry; chapter 1), white matter microstructure properties (using Diffusion Tensor Imaging; chapter 2) and intrinsic functional connectivity (using resting state functional MRI; chapter 3). Previous studies have primarily used case-control design on ASD and applied voxel-based morphometry (VBM) and surface-based morphometry (SBM) analysis. Some of these studies showed widespread grey matter abnormalities including the social brain regions (orbitofrontal cortex, amygdala, superior temporal sulcus and fusiform gyrus) of individuals with ASD as compared to controls. Several studies also used Diffusion Tensor Imaging (DTI) in adults with HFA, which showed white matter microstructure abnormalities with reduced fractional anisotropy (FA) and increased mean diffusivity (MD) in the superior longitudinal, uncinate fasciculus, inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, corpus callosum and cortico�spinal tracts. Resting state functional MRI (rs-fMRI) studies showed atypical functional connectivity in different brain regions/resting state networks including the default mode, executive control, fronto-parietal and visual networks in individuals with ASD. However, this pattern of results is far from unequivocal. In addition, these variations can be accounted by the different analytical approaches used. There is considerable variance inherent in the case-control design due to the sampling of the controls. A dimensional approach avoids this source of variance by sampling across the whole population. High resolution whole brain MPRAGE, DTI and rs-fMRI data were collected from a sample of research volunteers across the population (including those with a clinical diagnosis of ASD) using a 3T MRI Scanner, based at the Centre for Integrative Neuroscience and Neurodynamics (CINN), University of Reading. All volunteers also filled in a questionnaire measuring autism-related traits, such as the Autism Spectrum Quotient. VBM analysis was conducted at a whole-brain level using the Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) incorporated within the SPM analysis package. The regional grey matter volumes were calculated by summing up all voxels in the corresponding tissue maps. In addition, SBM analysis was performed using FREESURFER, a brain imaging analysis suite. The cortical thickness, surface area, volume and gyrification were measured by inflating the whole-brain structural images. Then, a robust regression method was used to test the relationship of the neuroanatomical measures and autism-related traits. Diffusion Tensor Imaging (DTI) data were pre-processed and analysed to measure the Fractional Anisotropy (FA) and Mean Diffusivity (MD) values using tract-based spatial statistics (TBSS), and skeleton�based tracts of interest approach using the FSL analysis package. Correlations and regression analyses were conducted, similar to the VBM. The rs-fMRI data were pre-processed using independent component analysis (ICA) approach and dual regression analysis (based on Beckmann’s eight resting state networks) incorporated in FMRIB FSL software package. Our findings demonstrated widespread grey matter abnormalities including the social brain regions (chapter 1), a partial evidence for white matter microstructure abnormalities (chapter 2) and intrinsic functional connectivity (chapter 3) related to higher autistic traits. Thus, the proposed set of studies addressed the key gap in the literature on grey matter abnormalities, atypical white matter microstructure properties and aberrant intrinsic functional connectivity related to autistic traits. By taking a dimensional approach, this project has the potential to offer new insights into the aetiology of the autistic phenotype.