Conventional MRI Markers and Cognitive Dysfunction in Multiple Sclerosis: An Exploratory Pilot Study

Cognitive impairment is a common and disabling manifestation of multiple sclerosis (MS), yet its structural correlates remain incompletely understood in routine clinical practice. Although advanced neuroimaging has improved our understanding of the underlying mechanisms, its limited accessibility restricts widespread clinical implementation. We aimed to investigate whether conventional MRI markers are associated with domain-specific cognitive impairment in patients with MS.

In this exploratory cross-sectional study, 25 adults with clinically confirmed multiple sclerosis underwent cognitive assessment using the standard Persian Montreal Cognitive Assessment (MoCA), version 7.1, and conventional 1.5-T brain MRI. Conventional MRI measures included lesion burden, lesion topography, and visually rated cortical atrophy. Associations between MRI markers and MoCA-derived cognitive domains were assessed using Spearman rank correlation with false discovery rate adjustment for multiple testing.

Twenty-five patients with clinically confirmed MS completed cognitive and MRI assessments. Mean age was 42.4 ± 9.1 years, 56% were women, and the mean MoCA score was 22 ± 7. Periventricular (88%) and juxtacortical/cortical (84%) lesions were the most frequent MRI findings, total T2/FLAIR lesion burden showed only weak correlations with cognitive performance (ρ range -0.02 to 0.29; all FDR-adjusted p > 0.65). The largest lesion-based association was observed between infratentorial lesions and concentration (ρ = -0.48, unadjusted p = 0.018; FDR-adjusted p = 0.405). Among atrophy measures, GCA showed the strongest association with visuospatial performance (ρ = 0.51, unadjusted p = 0.012; FDR-adjusted p = 0.405), with additional moderate trends for concentration (ρ = 0.40, p = 0.052) and executive function (ρ = 0.40, p = 0.052). No MRI marker remained statistically significant after correction for multiple comparisons.

Although no MRI-cognitive association remained statistically significant after correction for multiple comparisons, several unadjusted associations showed moderate effect sizes and may warrant further investigation. The strongest exploratory associations involved infratentorial lesions and concentration (ρ = -0.48) and global cortical atrophy and visuospatial performance (ρ = 0.51). The absence of consistent relationships between overall lesion burden or individual regional MRI markers and cognitive performance suggests that focal structural measures alone may not fully capture the neural substrates of cognitive dysfunction in MS. This pattern provides a rationale for further investigation of distributed, network-level mechanisms of cognitive dysfunction, while the present findings should be considered hypothesis-generating rather than confirmatory. Conventional MRI may therefore provide complementary structural information in the clinical assessment of cognitive dysfunction in MS. Larger, adequately powered studies using comprehensive neuropsychological assessment and quantitative, network-sensitive MRI measures are needed to determine whether these observed patterns are reproducible and clinically meaningful.