Search Abstracts | Symposia | Slide Sessions | Poster Sessions
White matter microstructure is associated with language abilities in 4-year-old Cantonese-speaking children
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Huan REN1, Xiaocong Chen1, Jueyao Lin1, Jiayi Lu1, Lu Li1, Zhengqin Liu1, Changsheng Li1, Han Zhang2, Caicai Zhang1; 1The Hong Kong Polytechnic University, 2ShanghaiTech University
Introduction: Prior to the onset of formal schooling, most typically developing children make remarkable progress in language learning, achieving major milestones in higher-order language skills, such as grammar and narrative (Sakai, 2005). Neurobiologically, this period coincides with rapid white matter microstructural maturation. Diffusion magnetic resonance imaging (dMRI) studies have identified several white matter tracts associated with language abilities, including the dorsal arcuate fasciculus for syntax, the ventral inferior fronto-occipital fasciculus for semantics, the corpus callosum for interhemispheric integration, and the corona radiata connecting cortex to subcortical structures involved in attention and working memory (Friederici, 2026; Mathur et al., 2026). However, most prior studies in children have focused on a priori selected tracts and on lower-level language skills such as vocabulary, leaving the broader white matter correlates of preschool language, particularly higher-order skills like narrative comprehension underexplored. Moreover, most studies have focused on fractional anisotropy (FA) as a single metric, leaving the roles of diffusivity-based metrics largely unspecified. Here, we used whole-brain tract-based spatial statistics (TBSS) to examine how distributed white matter systems support language development across multiple levels and domains. Methods: Using dMRI, this study examined the association between microstructural features (FA, MD, AD and RD) and multidimensional language skills in 54 typically developing 4-year-old Cantonese-speaking children (26 males). Language measures included expressive grammar, receptive grammar, expressive vocabulary, and narrative comprehension from the Hong Kong Test of Preschool Oral Language (TOPOL). MRI data were preprocessed using MRTrix3 and FSL. Whole-brain TBSS was performed to examine voxel-wise correlations between FA, MD, AD, RD and each language area, while controlling for age and sex. Cluster-based permutation testing (p < 0.05, corrected) was used. Significant clusters were anatomically localized using the JHU white matter tract atlas. Results: No significant clusters emerged for FA or RD. MD was negatively correlated with narrative comprehension (one cluster, 1,850 voxels, p = 0.042). AD showed negative correlations with narrative comprehension (one cluster, 65 voxels, p = 0.049) and expressive grammar (one cluster, 15,643 voxels, p = 0.024). Significant clusters were localized to the splenium of the corpus callosum and the corona radiata. Conclusion: This study linked white matter microstructure with multidimensional language abilities in 4-year-olds using whole-brain TBSS. Lower MD and AD in the splenium of the corpus callosum and the corona radiata were associated with better narrative comprehension and expressive grammar, whereas FA and RD showed no significant associations. These negative MD/AD correlations align with normative developmental trajectories showing rapid diffusivity decline during early childhood and with previous preschool findings (Kim et al., 2026; Walton et al., 2018). The absence of FA findings may reflect high FA variability at this age, whereas null RD associations suggest that individual differences at age 4 are more related to axonal organization (AD) than to myelination (RD). Beyond classic dorsal language pathways, the involvement of corpus callosum and corona radiata indicates that higher-order language abilities draw on a more distributed white matter network that potentially includes domain-general cognitive processes.
Topic Areas: Language Development/Acquisition,