Multimodal approach to identify neuropsychophysiological subgroups in myalgic encephalomyelitis/chronic fatigue syndrome and their relevance for rehabilitation: protocol for a mechanistic cross-sectional and longitudinal study.

INTRODUCTION

Myalgic Encephalomyelitis (ME)/Chronic Fatigue Syndrome (CFS) is a debilitating condition characterized by severe fatigue and post-exertional malaise (PEM). Reported neuropsychophysiological abnormalities suggest ME/CFS is multifactorial, but current knowledge remains fragmented. This study protocol outlines a multimodal investigation designed to (1) compare neuropsychophysiological mechanisms between ME/CFS patients and healthy participants, (2) test an integrative model of ME/CFS, (3) identify neuropsychophysiological subgroups within the patient population, and (4) identify predictors of symptom response during rehabilitation.

METHODS AND ANALYSIS: This study will enroll 115 ME/CFS patients and 55 healthy participants. Groups will be comparable in age, sex, and education level, with a larger patient sample enabling subgroup and longitudinal analyses. A cross-sectional assessment at baseline will be carried out in both groups. Patients will then be evaluated longitudinally throughout a standardized cognitive-behavioral therapy rehabilitation program delivered as routine care. Baseline measures include systemic inflammation and general health biomarkers, measures of autonomic and central nervous system function, neuroinflammation (magnetic resonance spectroscopy, [18F]DPA714 PET in a subsample), serum short-chain fatty acid levels, gut microbiota composition and function, and neuroendocrine and self-reported responses to psychosocial stress. Fatigue severity (physical and cognitive) and PEM will be assessed through validated questionnaires, ecological momentary assessment, and laboratory tasks. These will be re-evaluated during therapy, and all non-neuroimaging measures will be repeated after the rehabilitation program. Statistical analyses will comprise multivariate analysis of variance, general linear models, classification algorithms, structural equation models, least absolute shrinkage selection operator principal component regression (LASSO-PCR), cluster analysis and latent class growth analysis (LCGA).