Atrophy Patterns in Patients With Multiple Sclerosis With Cognitive Impairment, Fatigue, and Mood Disorders

Cognitive impairment, fatigue, and mood disorders are common in multiple sclerosis (MS), but their interplay and neurologic substrates are not well understood. This study aims to identify atrophy patterns in patients with MS with varying levels of cognitive impairment, fatigue, anxiety, and depression.

A cross-sectional cohort study enrolling patients with relapsing-remitting MS meeting the 2017 McDonald criteria and healthy controls (HCs), with similar age, sex, and education, was conducted. Participants completed cognitive assessments across 5 domains and inventories for fatigue, anxiety, and depression. Disability was quantified using the Expanded Disability Status Scale (EDSS). K-means clustering grouped patients with MS based on cognitive performance, fatigue, anxiety, and depression, accounting for disability. Group surface analysis from FreeSurfer, corrected for multiple comparisons, was used to assess cortical atrophy (p < 0.05). Deep gray matter and infratentorial atrophy were defined by a regional volumetric z-score <-1.5, with intergroup comparisons using Bayesian multinomial logistic regression (95% CI).

One hundred and two patients with relapsing-remitting MS (67 women; mean age 37 ± 10 years; mean disease duration 9.8 ± 6.5 years; mean EDSS score 2.2 ± 1.73) and 98 HCs (63 women; mean age 36 ± 12 years) were included. K-means clustering identified 4 MS patient clusters: K1 (n = 26): cognitively preserved (CP) without fatigue, anxiety, and depression; K2 (n = 31): CP with fatigue, anxiety, and depression; K3 (n = 18): cognitively impaired (CI) without fatigue, anxiety, and depression; K4 (n = 27): CI with fatigue, anxiety, and depression. Dimension-1 (cognitive tests) explained 34.7% of the variability; Dimension-2 (fatigue, anxiety, and depression) explained 18.7%. EDSS score correlated more with Dimension-2 (loading = 0.32) and was higher in K2 and K4 (p < 0.05). K2 and K3 had similar cortical atrophy, but K3 showed more unilateral hippocampus (log odds = 2.47, p = 0.023) and amygdala (log odds = 1.79, p = 0.003) atrophy, compared with K1. K4 had widespread bilateral cortical atrophy, with more bilateral thalamus (log odds = 1.28, p = 0.023), amygdala (log odds = 2.57, p = 0.003), and basal ganglia (log odds = 1.44, p = 0.01) atrophies. Cerebellar atrophy was significant in K2 (log odds = 17.68, p < 0.001) and K4 (log odds = 18.05, p < 0.001), compared with K1.

Cognitive impairment, fatigue, anxiety, and depression in patients with MS can coexist or present independently. Our findings suggest a stronger association between cognitive impairment and deep gray matter atrophy while fatigue, anxiety, and depression are more strongly associated with cerebellar atrophy. The accumulation of cognitive impairment, fatigue, anxiety, and depression is associated with increasing global cortical and deep gray matter atrophy.