Quantitative Anisotropy of the White Matter Microstructure Underlying Language Tracts in Autism

Autism is a neurodevelopmental condition that involves differences in social communication and restricted behaviors, often accompanied by altered structural and functional brain connectivity. This preliminary study examined differences in structural connectivity of language-related white matter tracts between autistic and neurotypical (NT) children, exploring the impact of such differences on language comprehension.

Diffusion-weighted imaging data were acquired from 25 children (13 autistic and 12 NT) on a Siemens 3T MAGNETOM Prisma MRI scanner. Data were processed using DSIstudio. Using deterministic tractography, we mapped 16 white matter fiber tracts and assessed diffusion metrics. Correlational tractography examined whole-brain white matter differences, controlling for age, gender, and full-scale IQ. Reading comprehension was assessed using the Gray Oral Reading Test-Fifth Edition, and regression analyses were conducted to examine associations with quantitative anisotropy (QA).

Autistic children showed significantly higher QA in 6 of the 16 language-related tracts and higher isotropic diffusion in 2 tracts, suggesting denser fiber bundles with more surrounding free water. In the autistic group, lower QA, particularly in the left arcuate fasciculus, was associated with better reading comprehension, whereas, in NT children, higher QA was associated with better reading. These findings suggest that autistic children may rely on different, more localized white matter pathways for reading compared with the more integrated language networks observed in NT children. Divergent QA-reading comprehension relationships across groups suggest that QA may serve as a candidate biomarker for stratifying language-based interventions.Statement of Importance:This study demonstrates that autistic children exhibit higher Quantitative Anisotropy (QA) in language-related white matter tracts and elevated isotropic diffusion, indicating dense fiber bundles with increased extracellular fluid. Importantly, while higher QA in the left arcuate fasciculus associated with better reading comprehension in neurotypical children, lower QA was associated with better comprehension in autistic children. These findings suggest that autistic individuals have distinct, localized neural pathways for reading rather than integrated canonical networks. QA serves as a promising biomarker to guide individualized language interventions and track neurostructural plasticity.