Scientific Integrity in ME/CFS Research

The broader scientific integrity landscape has relevance for evaluating the ME/CFS literature. A meta-analysis by Heathers estimated that approximately 14% of published scientific papers may contain fabricated data (ME/CFS Science 2025). However, the integrity challenges in ME/CFS research differ from this general picture:

• Design flaws rather than fraud: The ME/CFS literature is unreliable primarily because of flawed study design (broad diagnostic criteria, subjective outcomes, non-blinded trials, small samples) rather than data fabrication. Very few ME/CFS papers have been formally retracted (ME/CFS Science 2025).
• Publication bias: Positive results are more likely to be published, creating a distorted picture of treatment efficacy. The FITNET-NHS failure (Section Blinding Failures in ME/CFS Treatment Research) illustrates how replication reveals the extent of this bias.
• Conflict of interest: Researchers invested in the psychogenic model had career incentives to produce supportive evidence; the PACE trial investigators’ resistance to data sharing illustrates this dynamic.
• Replication crisis: The small ME/CFS research community means that many “replicated” findings come from overlapping research networks, inflating apparent replication while lacking true independence. The history of placebo-controlled surgical trials provides useful context for understanding the resistance to rigorous methodology in ME/CFS research (ME/CFS Science 2021). In surgery, the introduction of sham-controlled trials repeatedly demonstrated that widely practised procedures (arthroscopic knee surgery for osteoarthritis, vertebroplasty for compression fractures) were no better than placebo. The same pattern—interventions appearing effective in unblinded studies but failing under rigorous testing—is precisely what has occurred with GET and CBT for ME/CFS. The parallel reinforces that resistance to blinding is not unique to ME/CFS but is a general feature of fields where practitioners have strong prior beliefs about treatment efficacy. These integrity concerns reinforce the need for the methodological improvements outlined throughout this chapter: pre-registered protocols, objective outcomes, PEM-requiring criteria, independent replication, and mandatory data sharing.

References

ME/CFS Science. 2021. “History of Surgery.” 2021. https://mecfsscience.org/history-of-surgery/.

———. 2025. “How Many Scientific Papers Are Fake?” 2025. https://mecfsscience.org/how-many-scientific-papers-are-fake/.