Microstructural Features Linking White Matter and Mathematics in Adolescence

Abstract

Abstract Mathematics is a complex skill requiring the coordination of distributed gray matter brain regions connected by white matter tracts. Diffusion tensor imaging (DTI) studies have revealed a network of white matter tracts that support math processing, but the specific microstructural features driving this relationship remain unclear. Other magnetic resonance imaging (MRI) methods—neurite orientation dispersion and density imaging (NODDI), inhomogeneous magnetization transfer (ihMT), multicomponent driven-equilibrium single-pulse observation of T1 and T2 (mcDESPOT), and g-ratio imaging—can probe microstructural features like axon packing, fiber orientation, and myelin more specifically than DTI. We applied these methods alongside DTI to evaluate links between white matter microstructure and math in a longitudinal cohort of 33 6–16 year olds (66 datasets total). Partial correlations between metrics of white matter microstructure and math skill, controlling for age and gender, were carried out in the left superior longitudinal and inferior longitudinal fasciculi, corticospinal tract, and the splenium. Cross-sectionally, fiber coherence of the corticospinal tract and superior longitudinal fasciculus correlated to mathematics performance. Longitudinally, change in markers of axonal packing and myelin were linked to changes in both math skill and fluency in a regionally-specific manner, with links to myelin-sensitive metrics most prevalent. Notably, decreases in myelin were linked to improvements in mathematics over time, suggesting ongoing refinement of the math network. Findings presented here did not survive multiple comparisons corrections, but provide insight for future work elaborating upon these associations in larger samples.