The association between white matter tract structural connectivity and information processing speed in relapsing-remitting multiple sclerosis

Abstract

Abstract Background Information processing speed (IPS) deterioration is common in relapsing-remitting multiple sclerosis (RRMS) patients [1] and might severely affect quality of life and occupational activity. However, understanding of its neural substrate is not fully elucidated. We aimed to investigate the associations between MRI-derived metrics of neuroanatomical structures, including the tracts, and IPS. Methods Symbol Digit Modalities Test (SDMT), Paced Auditory Serial Addition Test (PASAT), and Color Trails Test (CTT) were used to evaluate IPS in 73 RRMS consecutive patients, all undergoing only interferon beta (IFN-β) therapy during the study. At the same time, 1.5T MRI including diffusion tensor imaging (DTI) data was acquired for each recruited subject. We analyzed volumetric and diffusion MRI measures (FreeSurfer 6.0) including normalized brain volume (NBV), cortical thickness (thk), white matter hypointensities (WMH), volume (vol), diffusion parameters: mean (MD), radial (RD), axial (AD) diffusivities, and fractional anisotropy (FA) of 18 major white-matter (WM) tracts. Multiple linear regression model with interaction resulted in distinguishing the neural substrate of IPS deficit in the IPS impaired subgroup of patients. Results The most significant tract abnormalities contributing to IPS deficit were right inferior longitudinal fasciculus (R ILF) FA, forceps major (FMAJ) FA, forceps minor (FMIN) FA, R uncinate fasciculus (UNC) AD, R corticospinal tract (CST) FA, and left superior longitudinal fasciculus FA (L SLFT). Among volumetric MRI metrics, IPS deficit was associated with L and R thalamic vol. and cortical thickness of insular regions. Conclusion In this study, we showed that disconnection of the selected WM tracts, in addition to cortical and deep gray matter (GM) atrophy, might underlie IPS deficit in RRMS patients but more extensive studies are needed for precise associations.