Functional Biomarkers

1 Two-Day CPET Protocol

Perhaps the most specific biomarker for ME/CFS:

  • Methodology: Maximal exercise testing on consecutive days
  • Finding: 10–25% decline in VO2peak, AT, work capacity on Day 2
  • Specificity: Healthy controls and patients with other conditions reproduce or improve
  • Physiological basis: Reflects post-exertional malaise objectively
  • Limitations: Requires specialized equipment, may exacerbate symptoms

2 NASA Lean Test

Simple orthostatic assessment:

  • Patient leans against wall for 10 minutes
  • Heart rate and blood pressure monitored
  • Identifies POTS and other orthostatic disorders
  • Accessible, low-tech screening tool

3 Cognitive Testing

Standardized neuropsychological assessment:

  • Processing speed measures (e.g., Symbol Digit Modalities Test)
  • Attention tests (e.g., continuous performance tasks)
  • Pattern of deficits may distinguish from depression
  • Sensitive to post-exertional cognitive deterioration

4 Digital Physiological Biomarkers

Mobile health (mHealth) technologies enable continuous or daily physiological monitoring outside clinical settings, offering an alternative biomarker approach: tracking within-person temporal dynamics rather than relying solely on single-timepoint between-group differences. A large intensive longitudinal study using this approach (Aitken et al. 2026, n=4244 Visible app users with Long COVID, ME/CFS, or other energy-limiting conditions) found that 60-second morning PPG assessments of HR, HRV (RMSSD scaled 0–100), and respiratory rate predicted same-day evening symptom reports Multilevel models incorporating both within-person biometric fluctuations and prior-day symptom history achieved AUC values of 0.82 for crash, 0.74 for fatigue, and 0.85 for brain fog using walk-forward cross-validationβ€”higher than models using symptom history alone (AUC 0.78, 0.73, 0.83 respectively). Within-person predictors (daily deviations from individual baselines) were substantially stronger than between-person averages, underscoring the importance of personalized monitoring over population-level thresholds. The study was retrospective in design, analyzing data already collected through a commercial app from self-identified participants; the proportion with clinician-confirmed ME/CFS is not reported. Potential implications for biomarker strategy (pending prospective validation) include:

  • Accessibility: Smartphone-based PPG requires no specialized equipment, which could enable broader deployment if clinical utility is confirmed
  • Within-person design: Each patient serves as their own control, sidestepping the between-person heterogeneity that limits single-timepoint biomarker studies
  • Temporal prediction: Morning biometrics predict evening symptoms within the same day, though whether this time window is sufficient for actionable crash prevention remains untested
  • Incremental value: Biometrics add statistically significant predictive power beyond symptom self-tracking alone, though the magnitude of improvement is modest (AUC improvement of 0.01–0.04 depending on outcome)
  • Engagement burden: Daily monitoring (morning PPG + evening symptom survey) requires sustained cognitive and motor capacity that severe and very severe patients may not have; the monitoring itself could trigger cognitive PEM
  • No demonstrated prevention: Predictive performance does not establish that acting on predictions reduces crash frequency or improves outcomes; this intervention link is untested