Symptom-based clusters in people with ME/CFS: an illustration of clinical variety in a cross-sectional cohort.
Vaes, Anouk W, Van Herck, Maarten, Deng, Qichen et al. · Journal of translational medicine · 2023 · DOI
Quick Summary
This study looked at whether people with ME/CFS can be grouped into different subtypes based on their specific combination of symptoms. Researchers asked 337 people with ME/CFS to report their symptoms and found 13 major groups, where people in each group shared similar patterns of symptom severity and frequency. While fatigue and post-exertional malaise (PEM) appeared in all groups, other symptoms varied—meaning people with ME/CFS experience the illness quite differently from one another.
Why It Matters
Understanding that ME/CFS presents as multiple symptom subtypes could enable personalized treatment approaches tailored to individual symptom profiles, potentially improving clinical outcomes. This research provides empirical evidence for clinical heterogeneity that many patients recognize intuitively, validating the need for precision medicine approaches in ME/CFS rather than one-size-fits-all treatments.
Observed Findings
Fatigue and post-exertional malaise were reported across all 13 major symptom-based clusters
45 symptom-based clusters were initially identified, with 13 clusters containing 10 or more patients
The 13 largest clusters were defined by distinct patterns of symptom severity and frequency, not just presence/absence
Symptom patterns identified in the Dutch cohort (n=337) were reproducible in an independent validation sample (n=252)
11% of ME/CFS patients could not be reliably classified into any of the 13 largest clusters
Inferred Conclusions
ME/CFS patients form identifiable symptom-based subgroups with reproducible patterns, supporting the concept of clinical heterogeneity in the disease
Personalized or tailored treatment approaches based on symptom cluster classification may be more appropriate than uniform treatment protocols
The universal presence of fatigue and PEM across all clusters suggests these are core ME/CFS features, while other symptoms vary considerably between individuals
Remaining Questions
Are these symptom clusters stable over time, or do patients move between clusters as their illness evolves?
Do these symptom-based clusters correlate with different biological markers or underlying pathophysiological mechanisms?
What This Study Does Not Prove
This study does not prove that these clusters represent distinct biological subtypes or disease mechanisms—only that symptom patterns group together statistically. It does not establish causation or the stability of clusters over time, nor does it prove these groupings will predict treatment response (though the authors suggest this as a future direction). The 11% of patients unclassified into the 13 largest clusters indicates the model may not capture all clinically meaningful variation.