Neurostructural and Neurophysiological Correlates of Multiple Sclerosis Physical Fatigue: Systematic Review and Meta-Analysis of Cross-Sectional Studies

Abstract

Fatigue is one of the most debilitating symptoms for people with multiple sclerosis (PwMS). By consolidating a diverse and conflicting evidence-base, this systematic review and meta-analysis aimed to gain new insights into the neurobiology of MS fatigue. MEDLINE, ProQuest, CINAHL, Web of Science databases and grey literature were searched using Medical Subject Headings. Eligible studies compared neuroimaging and neurophysiological data between people experiencing high (MS-HF) versus low (MS-LF) levels of perceived MS fatigue, as defined by validated fatigue questionnaire cut-points. Data were available from 66 studies, with 46 used for meta-analyses. Neuroimaging studies revealed lower volumetric measures in MS-HF versus MS-LF for whole brain (-22.74 ml; 95% CI: -37.72 to -7.76 ml; p = 0.003), grey matter (-18.81 ml; 95% CI: -29.60 to -8.03 ml; p < 0.001), putamen (-0.40 ml; 95% CI: -0.69 to -0.10 ml; p = 0.008) and acumbens (-0.09 ml; 95% CI: -0.15 to -0.03 ml; p = 0.003) and a higher volume of T1-weighted hypointense lesions (1.10 ml; 95% CI: 0.47 to 1.73 ml; p < 0.001). Neurophysiological data showed reduced lower-limb maximum voluntary force production (-19.23 N; 95% CI: -35.93 to -2.53 N; p = 0.02) and an attenuation of upper-limb (-5.77%; 95% CI:-8.61 to -2.93%; p < 0.0001) and lower-limb (-2.16%; 95% CI:-4.24 to -0.07%; p = 0.04) skeletal muscle voluntary activation, accompanied by more pronounced upper-limb fatigability (-5.61%; 95% CI: -9.57 to -1.65%; p = 0.006) in MS-HF versus MS-LF. Results suggest that MS fatigue is characterised by greater cortico-subcortical grey matter atrophy and neural lesions, accompanied by neurophysiological decrements, which include reduced strength and voluntary activation. Prospero registration Prospero registration number: CRD42016017934.

Keywords: Fatigue; Multiple sclerosis; Neuroimaging; Neurophysiology; Neurostructural.

Fig. 1

PRISMA flow chart for literature search and study selection

Fig. 2

Summary of results of meta-analyses comparing neuroimaging and neurofunctional data for MS-HF versus MS-LF. Data are presented as standardised mean difference and 95% confidence intervals. The upper figure presents summary data for neuroimaging variables and the lower figure presents summary data for neurofunctional variables, with the abscissas representing a decrease or increase for MS-HF in comparison with MS-LF. TBV, total brain volume; BPF, brain parenchymal fraction; GMV, gray matter volume, WMV, white matter volume, T1-WLV, T1-weighted lesion volume, T2-WLV, T2-weighted lesion volume, NAA/Cr, N-acetylaspartate to creatine ratio Cho/Cr, choline to creatine ratio, UL, upper-limb; LL, lower-limb; MEP, motor evoked potential; CMCT, central motor conduction time; SICI, short-interval intracortical inhibition, ICF, intracortical facilitation; MVC, maximum voluntary contraction force

Fig. 3

Methodological quality of the included studies evaluated using the Cross-Sectional/Prevalence Study Quality Scale, recommended by the Agency for Healthcare Research and Quality (AHRQ). Scores of 0–3 indicate “low quality”, 4–7 “moderate quality” and 8–11 “high quality”