Eckey et al. (2025)
- Authors: Martha Eckey, Peng Li, Braxton Morrison, Jonas Bergquist, Ronald W. Davis, Wenzhong Xiao
- Institutes: Computational Research Center for Complex Chronic Diseases, Massachusetts General Hospital, Harvard Medical School; The Myalgic Encephalomyelitis/Chronic Fatigue Syndrome Collaborative Research Centre, Department of Chemistry BMC, Uppsala University; Myalgic Encephalomyelitis/Chronic Fatigue Syndrome Collaborative Research Center, Stanford University School of Medicine.
- Publisher: PNAS
- Link: DOI
Summary
This large-scale survey reinforces the significant symptomatic and therapeutic overlap between ME/CFS and long COVID, suggesting they may share underlying disease mechanisms. By analyzing patient experiences, the research identified several widely-used treatments—such as pacing, electrolytes, and LDN—that are perceived as highly beneficial, while confirming that graded exercise therapy is seen as harmful. The identification of distinct patient subgroups who respond differently to specific medications supports a move towards personalized medicine and provides a data-driven basis for designing more targeted and effective clinical trials for these complex illnesses.
What was researched?
This study analyzed patient-reported outcomes from a large survey of 3,925 individuals to assess the perceived effectiveness of more than 150 different treatments for ME/CFS and long COVID. The research also explored the relationships between patient demographics, symptom profiles, comorbidities, and their responses to these interventions.
Why was it researched?
With no FDA-approved treatments for either ME/CFS or long COVID, patients rely on off-label and palliative therapies. This study was conducted to gather real-world evidence from patient experiences, aiming to identify potentially effective management strategies, highlight symptomatic and therapeutic similarities between the two conditions, and help generate new avenues for future clinical trials.
How was it researched?
This was a patient-reported outcomes study based on data from the online “TREATME” survey. The researchers collected responses from 3,925 patients (2,125 with ME/CFS and 1,800 with long COVID) regarding their symptoms, comorbidities, and perceived effectiveness of various treatments. A “Net Assessment Score” (NAS) was calculated to quantify the benefit of each intervention, and clustering analysis was used to identify patient subgroups based on their unique symptom profiles.
What has been found?
The study revealed a strong correlation () in treatment responses between ME/CFS and long COVID patients, who also shared similar core symptoms like PEM and fatigue. The interventions with the greatest perceived benefits included pacing, fluids/electrolytes 💊, compression stockings, low-dose naltrexone (LDN) 💊, antihistamines 💊, nattokinase/lumbrokinase 💊, and intravenous immunoglobulin (IgG) 💊. Four distinct patient subgroups with unique symptom profiles (e.g., “POTS-Dominant”) were identified, showing varied responses to specific treatments, such as ADHD stimulants 💊 being most effective in the “Cognitive and Sleep Dysfunction” cluster. Graded exercise therapy (GET) was overwhelmingly reported as harmful.
Discussion
The authors note that the study’s findings support the use of large-scale patient surveys for gathering real-world evidence. They acknowledge several limitations, including the self-reported nature of the data, potential for sampling and recall bias, the absence of a true placebo control group, and the difficulty of attributing effects when patients use multiple treatments simultaneously. The reliability for some treatments is also limited by a small number of responses.
Conclusion & Future Work
The researchers conclude that patient-reported data from large surveys provide a valuable foundation of real-world evidence for therapies and can help inform the design of future clinical trials for ME/CFS and long COVID. They suggest future studies could be enhanced by integrating data from health-tracking apps and electronic health records, and by using more targeted surveys to confirm these preliminary findings.