Fatigue in multiple sclerosis patients with different clinical phenotypes: a clinical and magnetic resonance imaging study.

BACKGROUND AND PURPOSE: The prevalence of fatigue and its relation with clinical, neuropsychological and brain magnetic resonance imaging (MRI) variables in a large cohort of multiple sclerosis (MS) patients was investigated.

METHOD: The Modified Fatigue Impact Scale and its subdomains were collected from 725 healthy controls and 366 MS patients [238 relapsing-remitting (RRMS) and 128 progressive (PMS)]. For the Modified Fatigue Impact Scale global and subdomains, MS patients were classified as fatigued (F-MS) or non-fatigued (NF-MS) according to cut-off values provided by logistic regression models with a specificity of 90% (i.e. a 10% false-positive rate in classifying healthy controls). MS patients underwent neurological, neuropsychological and MRI evaluations. Clinical and MRI measures were compared between F-MS and NF-MS patients using age-, sex- and phenotype-adjusted linear models. Heterogeneities between phenotypes were tested with specific interaction terms.

RESULTS: Global fatigue affected 174 (47.5%) MS patients, being more prevalent in PMS (PMS 64.1% vs. RRMS 38.7%, P < 0.001). For all dichotomizations, F-MS were older (P from <0.001 to 0.012) and more depressed (P < 0.001) than NF-MS patients. Compared to NF-MS, cognitive F-MS patients had lower education (P = 0.035). Compared to NF-MS, patients with global and physical fatigue had higher Expanded Disability Status Scale only for RRMS (P < 0.001). Only RRMS patients with physical fatigue had lower brain (P = 0.05), white matter (P = 0.039) and thalamic volumes (P = 0.022) compared to NF-MS patients.

CONCLUSIONS: In MS, fatigue is associated with older age, lower education and higher depression. Only in RRMS, fatigue is associated with Expanded Disability Status Scale and brain atrophy. A plateauing effect of disability and structural damage can explain the lack of associations in PMS.