Regional normal-appearing white matter metabolic profiles and cognition in early relapsing-remitting multiple sclerosis: a multi-voxel proton magnetic resonance spectroscopy study
Related Experiment Video
Updated: Aug 22, 2026

Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla
Published on: February 19, 2021
Lorand Sakalaš1,2, Vesna Suknjaja3,4, Jasmina Boban4,5
1Clinic of Neurology, Clinical Center of Vojvodina, Hajduk Veljkova 1-7, Novi Sad, 21000, Serbia. lorand.sakalas@kcv.rs.
Background:
Cognitive impairment in multiple sclerosis (MS) is common but only weakly explained by conventional magnetic resonance imaging. Proton magnetic resonance spectroscopy (¹H-MRS) detects microstructural damage in normal-appearing white matter (NAWM). We examined whether regional NAWM metabolic profiles are spatially patterned in early relapsing-remitting MS (RRMS) and whether they relate to cognition.
Methods:
In this cross-sectional study, 52 patients with early RRMS (Expanded Disability Status Scale 0-3; 51 with usable spectroscopy) and 44 matched controls underwent multi-voxel three-dimensional ¹H-MRS at 3 tesla across six supracallosal NAWM regions. Group differences used Mann-Whitney tests with false discovery rate correction, and a group × region analysis tested for disease-related disruption. NAA/Cr-cognition associations used Spearman and partial correlations and multivariable models.
Results:
NAA/Cr differed between groups in four of six regions, and a group × region interaction (p < 0.001) indicated that the regional NAA/Cr pattern in RRMS was disrupted beyond the physiological anterior-posterior gradient seen in controls. Frontal NAA/Cr correlated with processing speed (ρ = 0.350, p = 0.012) and parietal NAA/Cr with verbal learning (ρ = 0.327, p = 0.019), but neither survived adjustment for age or parsimonious modelling, in which disability was the main independent correlate of processing speed.
Conclusion:
Early RRMS shows spatially organised NAWM metabolic disruption. Regional NAA/Cr-cognition associations are detectable but weak and not independent of age or disability, and frontal NAWM integrity relates more to physical function than cognition, supporting a network-disconnection model with limited cognitive specificity. These findings are preliminary and need longitudinal confirmation.

