From cortical to subcortical: aberrant structural brain organization in autism spectrum disorder acrossdevelopment

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by communication difficulties, social interaction impairments, and stereotyped patterns of behavior. Prior studies have shown that ASD is associated with differences in neuroanatomy in the cerebral cortex and the subcortical regions as well as the connectivity among these regions. However, findings have been mixed due to the varying age group sampled and the methods used to measure these brain structures. In view of the heterogeneous findings in ASD, three cross-sectional design studies were conducted in this thesis to examine brain structural pathologies that may be related to the clinical and behavioural phenotype of the disorder across development. In the childhood and adolescent sample, two studies were carried out. The first one examined cortical thickness using a vertex-wise approach. Results revealed thinner cortex in the occipital, parietal and frontal regions, and thicker cortex in the inferior parietal and caudal anterior cingulate regions. These regions also showed age-related differences that deviated markedly from the typical developmental trajectories observed in the control group. Some of these regions with significant differences in cortical thickness were found to be associated with clinical symptoms in ASD. The second study in the childhood and adolescent sample examined the volume of subcortical structures and CSF using a spatially non-biased parcellation approach. It was found that intracranial volume was enlarged in children with ASD, accompanied by smaller bilateral cerebellum and left thalamus. These regions showed an age-related increase in volume in children with ASD, whereas the typically developing children showed a general age-related decrease in volume of the same regions. The volumes of the cerebellum, thalamus and basal ganglia structures were associated with relatively weaker motor control in ASD, and in particular greater volume of the left thalamus rather than age predicted worse motor performance in the clinical group. The third study was carried out in a large adult sample. The cerebellar white matter system, that interconnects cortical and subcortical targets, was examined. Using a diffusion-tensor imaging tractography approach, the cerebellar input and output white matter pathways were dissected. Both the input and output pathways were observed to be disrupted in ASD, supporting the hypothesis that ASD may be a “disconnectivity disorder”. Lower fractional anisotropy of the left middle cerebellar peduncles was associated with increased difficulties in communication and social interaction, and lower fractional anisotropy in the right superior cerebellar peduncle was linked to worse motor performance in adults with ASD. Therefore, my studies confirmed differences in neuroanatomy of cortical and subcortical regions with altered brain developmental trajectories in children and adolescence with ASD, and revealed disrupted cerebellar network system in adults with ASD. Dysmaturation of cortical and subcortical regions as well as cerebellar white matter pathways may contribute to clinical and motor phenotype of the disorder. Lastly, postmortem and early life imaging studies, together with evidence that prenatal stressors during 21 to 32 weeks of gestation may increase incidence of ASD, lead me to speculate whether the abnormalities reported here may have origins prior to 31 weeks of gestation.