Symptom clusters in ME/CFS reflect distinct neuroimmune and autonomic pathophysiological mechanisms: a translational model

Abstract Background Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a debilitating multisystem disease characterized by heterogeneous symptom patterns. Previous work suggested that specific symptoms tend to co‑occur, pointing toward underlying biological mechanisms. This study aimed to empirically validate literature‑based, hypothesis‑driven symptom clusters and assess whether they reflect distinct neuroimmune and autonomic pathophysiological pathways. Methods Symptom data from 748 adults with ME/CFS (≥20 years) participating in the APAV‑ME/CFS study were analyzed. Symptoms were assigned to predefined mechanistic groups informed by current pathophysiological hypotheses. Exploratory and Confirmatory Factor Analyses, followed by Structural Equation Modeling (SEM), evaluated the coherence, distinctiveness, and hierarchical structure of each cluster. Robustness was tested using a stratified, randomized training dataset. Results A coherent Brain factor (brain fog, sensory hypersensitivity, visual disturbances, sleep disturbances, headaches) showed excellent fit (RMSEA = 0.021; CFI = 0.996). Gastrointestinal symptoms demonstrated stronger internal consistency than Immune symptoms, and model comparisons supported a two‑factor Gut–Immune structure. Across all analyses, symptom groups emerged as internally consistent and statistically distinct. A higher‑order SEM including a common latent factor yielded excellent fit for the Autonomic symptom complex. Conclusions The findings support ME/CFS as a complex neuroimmune–autonomic multisystem disorder and suggest that symptom clusters align with functional biological systems. Mechanism-aligned symptom subgrouping may enable pathophysiology-guided diagnostics, patient stratification, and targeted therapeutic development. The proposed interpretations of underlying mechanisms derive from the integration of existing literature and were not directly measured in this study. The identified clusters therefore indicate mechanistic alignment rather than direct mechanistic validation.