Widespread White Matter Diffusivity Changes in Autism: A Tract Based Statistical Analysis of Diffusion Tensor Imaging Study

Background: Characterizing the features of white matter structural connectivity is of central importance in autism brain science. Many studies have shown differences in white matter between children with autism spectrum disorders (ASD) and children with typical development (TD). Diffusion tensor imaging (DTI) techniques can be used to quantify white matter integrity in vivo in terms of fractional anisotropy (FA) and mean, axial and radial diffusivity (MD, RD, AD). Here we applied tract-based spatial statistics (TBSS) to compare children with autism spectrum disorders (ASD) and children with typical development (TD) regarding the orientation, coherence and magnitude of their white matter tracts. We hypothesized that tissue pathophysiology documented in autism, particularly neuroinflammation, would lead to lower FA and greater diffusivity in ASD  

Objectives: To characterize white matter microstructure as part of an ongoing multimodal investigation of the relationships among structural, functional and pathophysiological features of brain in autism.

Methods: Magnetic resonance volumetric and diffusion tensor imaging were performed on 50 males, 23 ASD (10.43y± 1.77, NVIQ 99.77±17.11), 27 TD (9.94y± 2.49, NVIQ 113.34±16.59). ASD diagnosis was ascertained using ADOS and clinical assessment. DTI metrics (FA, MD, AD, RD) were generated by FSL (FMRIB Software Library V5.0.6), and then aligned and projected onto the mean FA skeleton which represents the centers of all the tracts. Between group voxelwise statistical analysis was applied to the resultant data. To test for (Group X Age) interaction we performed GLM analysis using age as a covariate  

Results: Our results revealed significantly increased right and left hemispheric white matter MD and RD in the ASD group as compared to the TD group in the anterior thalamic radiation, cortico-spinal tract, forceps major, inferior fronto-occipital fasciculus, inferior and superior longitudinal fasciculus (as well as the temporal part of the superior longitudinal fasciculus), the uncinate fasciculus, cingulate gyrus and the hippocampal cingulum. AD was significantly higher in the ASD group only in the left hemispheric white matter in the anterior thalamic radiation, cortico-spinal tract, inferior fronto-occipital fasciculus, inferior and superior longitudinal fasciculus as well as the temporal part of the superior longitudinal fascicules and the uncinate fasciculus. There was no significant difference between the groups regarding FA. When we added age as covariate using GLM, we found that MD, AD and RD retained significance in the above-mentioned tracts, and that FA was did not achieve significance  

Conclusions: Our findings suggest widely distributed differences between ASD and TD in the white matter fiber tracts with many fiber tracts involved. Such differences may be related to specific genetic or functional vulnerability of these tracts, but the distribution of this increased diffusivity appears more consistent with a downstream effect of a more widespread abnormality in the brain that may involve both the axons and the tissue milieu around the tract fibers, including the extracellular matrix, which could impact diffusivity measures.  Follow-up multimodal analyses are planned to assess the correlation of these changes with measures of functional connectivity in the same subjects.