Longitudinal Changes of the White Matter Microstructural Properties of Autism Spectrum Disorder: A Normative Model Analysis

Cross-sectional research documents atypical age-related development of white matter in autism spectrum disorder (ASD). However, little is known about the developmental changes in white matter microstructural properties in ASD. This study aims to investigate developmental changes in white matter tract microstructural properties in ASD using a longitudinal follow-up design and normative model analysis, and to examine clinical correlates of these changes. We assessed 75 autistic individuals (aged 15.3 ± 4.2 years) with diffusion spectrum imaging at baseline and 4.7 ± 1.9 years later. To measure the magnitude of deviation from the norm, we calculated z-scores for fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), and mean diffusivity (MD) for each of 76 tracts based on a norm established in 680 typically developing individuals. Callosal fibers connecting the temporal poles, hippocampus, and amygdala in ASD individuals showed higher AD, RD, and MD at both time points. Several tracts showed significantly greater increases in FA z-scores from baseline to follow-up, including the right superior longitudinal fasciculus, precentral thalamic radiation, frontal aslant tract, right corticospinal tracts, left arcuate fasciculus, callosal fibers connecting genu, and bilateral thalamic radiation, implying a pattern of greater deviation from the norm at Time 2 than Time 1. Higher autistic severity or social deficits at baseline were related to greater increasing rates in the diffusion metrics of callosal fibers connecting the precuneus (AD, RD, and MD), corticospinal tract (AD), right geniculate fibers (AD and MD), and right medial lemniscus (AD and MD) from baseline to follow-up. Using the normative model method to analyze longitudinal data on white matter microstructures, our findings support persistent alterations in callosal fibers and developmental alterations of several tracts in ASD, which were associated with baseline autistic severity.