ME/CFS-Specific Research Methods
Why ME/CFS research is methodologically hard: case definition heterogeneity, lack of objective endpoints, PEM as a constraint on study participation (selection bias), the 2-day CPET reliability debate, subjective outcomes in unblinded trials, appropriate control group selection, biomarker overfitting in small samples, statistical power with heterogeneous populations, and the problem of replicating findings across diagnostic criteria.
ME/CFS research faces a unique constellation of methodological challenges that collectively undermine replicability. This chapter examines eight interlocking problems, provides evidence-based guidance for evaluating the existing literature, and recommends standards for future research design. Each section is self-contained — a reader interested in a specific methodological issue can read the relevant section without needing the others — but the cross-cutting synthesis at the end argues that the problems are mutually reinforcing and require systemic solutions.
For patients: read the subjective-outcomes and biomarker-overfitting sections to judge the strength of claims you read online or in the news. Skim the methodological detail.
For caregivers: read the subjective-outcomes and cross-cutting-synthesis sections to understand why some treatments lack solid evidence. The technical methods can be skimmed.
For clinicians: read the subjective-outcomes, control-group-selection, and biomarker-overfitting sections to judge the quality of treatment studies you may rely on.
For researchers: read the full methodological framework — it establishes the conventions used to evaluate every study cited across this book.
1 Contents
- Case Definition Heterogeneity
- PEM as a Source of Selection Bias
- The 2-Day CPET Reliability and Validity Debate
- Subjective Outcomes in Unblinded Trials
- Appropriate Control Group Selection
- Biomarker Overfitting in Small Samples
- Statistical Power with Heterogeneous Populations
- Replication Across Diagnostic Criteria
- Cross-Cutting Synthesis