Neurometabolic brain changes in MS associated with fatigue and anxiety

Multiple sclerosis (MS) is a disease characterised by progressive neurodegeneration and demyelination, manifesting as focal white matter lesions (WML) and extending into normal-appearing white matter (NAWM) and grey matter (NAGM). In this exploratory, hypothesis-generating study, we investigated regional neurometabolic abnormalities within the frontal and parietal regions of these tissue types and examined their preliminary associations with clinical measures, using in vivo magnetic resonance spectroscopic imaging (MRSI).

Neurometabolic differences between the frontal and parietal regions of WML, NAWM, and NAGM in people with relapsing-remitting MS (pw-RRMS; n=34; FSS≥ 4 inclusion criteria, indicating clinically significant fatigue) and healthy controls (HCs; n=10) were evaluated using MRSI at 3T. The acquired spectra were analysed with LCModel, and multi-voxel segmentations were performed using in-house pipelines. Metabolite levels were expressed as ratios to total creatine (tCr). Z-scores were derived by normalising to healthy controls. Neurometabolic changes among regions and tissues were statistically tested using linear mixed model (LMM) analysis with interaction effects followed by post hoc subgroup testing. Clinical parameters were assessed at the time of imaging. Expanded Disability Status Scores (EDSS), fatigue and anxiety scores were linked with neurometabolites using Pearson's correlations in an exploratory analysis.

The LMM indicated interaction effects of region and tissue on all metabolite z-scores (p < 0.001). Exploratory post hoc voxel-level comparisons showed higher mean z-scores for glutamate (Glu) and glutamine+glutamate (Glx) ratios in frontal and parietal regions across all tissue types in pw-RRMS relative to HCs (all p < 0.0001). tCr ratios of Glu and Glx in NAWM and WML were higher in the frontal than in the parietal regions in pw-RRMS. Parietal glutathione (GSH) and myo-inositol ratios were higher than frontal metabolites across all tissue types. In exploratory, uncorrected analyses, Glu and GSH ratios showed positive correlations with anxiety and fatigue (r= 0.35 and r= 0.33) in frontal-WML and parietal-NAGM, respectively, and total N-acetylaspartate (tNAA) showed negative correlations with disability in frontal and parietal regions (r= -0.34 to -0.46). These symptom-correlation analyses were not corrected for multiple comparisons; the observed coefficients are modest and should be interpreted as preliminary, hypothesis-generating findings requiring confirmation in larger, independent cohorts.

In this exploratory analysis, pw-RRMS exhibited region-specific neurochemical alterations associated with MS-related symptoms. As the cohort was intentionally enriched for clinically significant fatigue (FSS≥4), the observed associations with fatigue may not readily generalise to the broader RRMS population. Accordingly, these findings should be interpreted as hypothesis-generating and require confirmation in larger, independent cohorts before any biomarker or therapeutic implications can be established.