White Matter Microstructure of Language Pathways in Nonverbal Autism: Insights from Diffusion Tensor Imaging and Myelin Water Imaging

Dominika Slušná¹Jordi Muchart-López²Erick Jorge Canales-Rodríguez³Wolfram Hinzen^4*

• ¹Pompeu Fabra University, Barcelona, Catalonia, Spain
• ²Sant Joan de Déu Hospital, Barcelona, Catalonia, Spain
• ³Laboratoire de Traitement du Signal, École Polytechnique Fédérale de Lausanne, Lausanne, Vaud, Switzerland
• ⁴Unity Productions Foundation, Potomac Mills, United States

Absence of language development is a condition encountered across a large range of neurodevelopmental disorders, including a significant proportion of children with autism spectrum disorder. The neurobiological underpinnings of nonverbal ASD (nvASD) remain poorly understood. This study employed multimodal MRI to investigate white matter (WM) microstructural abnormalities in nvASD, focusing on language-related pathways. We analyzed diffusion tensor imaging (DTI) metrics—fractional anisotropy (FA), mean diffusivity (MD), and radial diffusivity (RD)—alongside myelin water imaging (MWI) metrics, including myelin water fraction (MWF) and intra/extracellular water T2 relaxation time (T2IE). A cohort of ten children with nvASD and ten agematched typically developing controls was examined across eight major language-related tracts and the corticospinal tract (CST) as a motor reference. While DTI and MWI metrics showed no significant inter-group lateralization differences, MWF and T2IE exhibited pronounced lateralization exclusively in the nvASD group. Results also revealed significant microstructural differences in nvASD. MD and RD were the most sensitive DTI parameters, demonstrating widespread increases, whereas FA was less discriminatory. MWF exhibited the largest percentage change relative to controls (25–50%), suggesting a marked reduction in myelin content within affected tracts. Concurrently, widespread increases in T2IE indicate a less densely packed extra-axonal space, consistent with altered axonal integrity and reduced cellular surface area per unit volume. These findings align with prior evidence linking myelin abnormalities to ASD. Notably, microstructural differences were not restricted to language-related tracts but also extended to the CST, suggesting a more extensive WM disruption in nvASD. The absence of significant correlations between MRIderived metrics and clinical measures highlights the complexity of the neurobiological alterations in nvASD. As the observed lateralization patterns may reflect, in part, the influence of methodological variability in tract definition, segmentation strategy, and tractography method, these results should be interpreted with caution. Future studies with larger cohorts and longitudinal designs are required to clarify the developmental trajectory of these microstructural abnormalities, their relationship with language impairment severity, and their potential role as biomarkers for nvASD.