Longitudinal Analysis of the White Matter Microstructure of the Arcuate Fasciculus In Autism

Background: A key feature of autism spectrum disorder is impairment in language function and development. Recent findings from cross-sectional diffusion tensor imaging (DTI) studies have found the arcuate fasciculus to have abnormal microstructure, which may partly explain the neurobiological basis for language problems in autism. However, no study to date has examined the white matter microstructure of the arcuate fasciculus longitudinally in autism. Even in typical development, there is little information about development of the arcuate fasciculus from longitudinal studies that follow the same individuals over time. These are essential steps to further our understanding of why language development is aberrant in autism with deficits often persisting into adulthood and adversely affecting adult outcome.

Objectives: The goal of this study was to investigate the development of white matter microstructure of the arcuate fasciculus during late neurodevelopment in autism using longitudinal DTI.

Methods: The data were collected as part of an ongoing longitudinal MRI study on brain development in autism. High resolution DTI was acquired at three time points, each approximately two years apart, from a 3T MRI scanner and analyzed on 27 male individuals with autism (mean age at Time1: 15.98, s.d. 5.59) and 29 typical male control subjects (mean age at Time1: 16.31, s.d. 5.48). The arcuate fasciculus was extracted from the images using a new automated volumetric DTI segmentation algorithm. Derived measures of microstructure (fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)) were computed in the arcuate fasciculus and compared across groups. Longitudinal mixed-effects models were fit to each derived measure as a function of age.

Results: As a group, those with autism displayed a significant increase in MD (p = 0.021), which was due to increases in both AD (p = 0.032) and to a lesser extent, RD (p = 0.059). No significant differences were found in FA between the groups. Both groups displayed a significant negative slope in MD, RD, and AD versus age (p < 0.001) and a significant positive slope in FA (p = 0.016). No significant differences were found between the slopes of the autism and control group for any diffusion measurement.

Conclusions: Rather than “normalization” of the microstructure of the arcuate fasciculus with increasing age in autism, which would require steeper developmental trajectories in order for arcuate fasciculus development in autism to “catch-up” to typical development, the findings suggest a persistent gap arcuate fasciculus development between autism and typical individuals from late childhood into adulthood. The persistent significant delay in microstructural development of the arcuate fasciculus and the neurobiological mechanisms that cause and maintain it may help explain why language deficits persist into adulthood in autism. Understanding factors that cause the gap may lead to new interventions that improve language functioning and overall adult outcome in autism.