Novel Hypotheses from Two-Day CPET Findings
The objective demonstration of Day 2 metabolic failure in two-day cardiopulmonary exercise testing (B. Keller et al. 2024) provides unprecedented functional data that suggests several novel therapeutic approaches and previously undocumented biological phenomena. This section explores speculative hypotheses arising directly from these findings.
1 CPET as a Disease-Specific Biomarker
Two-day CPET not only documents PEM objectively but also discriminates ME/CFS from other fatigue conditions. In studies comparing ME/CFS patients to idiopathic chronic fatigue (ICF) patients, ME/CFS patients show deterioration on CPET-2 whereas ICF patients show improvement similar to healthy controls Campen and Visser (2021b) Campen and Visser (2021a). This pattern holds across both sexes, with identical CPET-2 impairment patterns observed in male and female ME/CFS patients Campen and Visser (2021a).
The characteristic failure to reproduce CPET-1 performance on CPET-2 represents a disease-specific signature that distinguishes ME/CFS from other fatigue conditions and potentially from post-COVID conditions with overlapping symptomatology. (Certainty: 0.50. 0.60→0.50: Mancini/Natelson 2026 NH-funded null replication — equiv powered, independent lab, senior investigator — brings evidence to equipoise. One positive study (Keller 2024, CCC, n=84+71) and one equivalently powered null (Mancini 2026, Fukuda+PEM, n=58+25). Elevated RPE is the consistent replicating finding across ALL studies — but RPE is a subjective self-report measure, not an objective metabolic biomarker. The CPET signature hypothesis now rests on partial evidence and faces genuine uncertainty.)
Evidence Base. ME/CFS patients show CPET-2 deterioration across multiple studies (B. Keller et al. 2024) Campen and Visser (2021b) Campen and Visser (2021a), whereas ICF patients show CPET-2 improvement similar to healthy controls Campen and Visser (2021b) Campen and Visser (2021a). Long COVID patients with PEM symptoms do NOT show the same 2-day CPET pattern (Gattoni et al. 2025). However, a 2026 null replication by Mancini, Natelson et al. (Fukuda criteria, n=58 ME/CFS + 25 sedentary controls) found no significant group-average Day 1→Day 2 decline in VO₂ or VO₂ at VT (Mancini et al. 2026), while confirming elevated perceived exertion (Borg RPE) at all workloads — a finding common to all CPET studies. The CPET signature’s reliability may depend on patient selection criteria (CCC vs. Fukuda), max-effort enforcement standards, and whether the signal of interest is group-average metabolic decline or individually elevated RPE.
Clinical Implications. CPET-2 could serve as an objective diagnostic biomarker to distinguish ME/CFS from: (1) idiopathic chronic fatigue, (2) deconditioning, (3) psychiatric fatigue, (4) Long COVID fatigue. This addresses a critical diagnostic gap where current criteria rely on subjective symptom reports.
Research Implications. The pathophysiological basis for CPET-2 deterioration in ME/CFS but not in ICF or Long COVID represents a key differentiating mechanism. Comparative studies of the molecular and physiological responses to Day 1 exercise in these three groups could identify the unique ME/CFS pathway.
Limitations. CPET-2 is not clinically accessible for most patients due to cost, equipment requirements, and the exertional challenge it poses. The protocol excludes the most severely affected patients who cannot complete maximal exercise (Snell et al. 2013). Cross-sectional studies cannot establish whether CPET-2 deterioration is a stable trait or varies with disease course.
Falsifiable Predictions.
- Patients meeting ME/CFS criteria will show CPET-2 deterioration regardless of onset trigger (infectious vs. non-infectious)
- Patients with Long COVID meeting ME/CFS criteria will show CPET-2 deterioration, while Long COVID patients not meeting ME/CFS criteria will not
- CPET-2 deterioration magnitude will correlate with other objective ME/CFS biomarkers (e.g., immune signatures, metabolic profiles) but not with fatigue severity in ICF or Long COVID
2 The Autonomic-Mitochondrial Feedback Loop
Keller et al. identified autonomic dysregulation as the primary mechanism linking Day 2 cardiopulmonary failures (B. A. Keller et al. 2024). Walitt et al. documented central catecholamine deficiency (Walitt et al. 2024). Heng et al. demonstrated cellular ATP depletion (Heng et al. 2025). What if these are not separate phenomena but nodes in a self-amplifying feedback loop?
Proposed mechanism:
- Central catecholamine deficiency impairs autonomic cardiovascular regulation
- Poor blood flow distribution during exercise causes tissue hypoxia
- Mitochondria operating under hypoxic conditions generate excess ROS
- ROS damages catecholamine synthetic enzymes and depletes Tetrahydrobiopterin (BH4) cofactor
- Further catecholamine reduction worsens autonomic dysfunction
- Cycle amplifies with each exertional episode
This would explain the 13-day recovery period: breaking this vicious cycle requires not just substrate replenishment (hours) but restoration of damaged enzymes, clearance of oxidative damage products, and mitochondrial turnover (days to weeks).
2.1 Testable Predictions
- Catecholamine synthetic enzyme activity should decline further in the 24–72 hours post-exercise
- Tetrahydrobiopterin (BH4) levels should show exercise-dependent depletion with slow recovery kinetics
- Interventions supporting both catecholamine synthesis (Tetrahydrobiopterin (BH4), tyrosine, cofactors) and mitochondrial protection (antioxidants) should show synergistic effects exceeding either alone
- Baseline autonomic function (HRV, baroreflex sensitivity) should predict severity of Day 2 CPET decline
- Serial measurement of oxidative stress biomarkers (isoprostanes, oxidized glutathione) should peak 24–48 hours post-exertion, correlating with symptom severity
2.2 Therapeutic Implications (Speculative)
If the autonomic-mitochondrial feedback loop drives PEM, breaking it might require simultaneous intervention at multiple nodes:
Catecholamine support tier:
- L-tyrosine 1500–3000 mg/day (precursor)
- Sapropterin (Tetrahydrobiopterin, BH4) or methylfolate + B12 (BH4 recycling pathway support)
- Iron, vitamin B6, vitamin C, copper (cofactors for synthetic enzymes)
- Timing: morning administration to support daytime autonomic function
Mitochondrial protection tier:
- MitoQ or ubiquinol 200–400 mg/day (mitochondria-targeted antioxidant)
- N-Acetylcysteine (NAC) 1200–1800 mg/day (glutathione precursor, oxidative stress buffer)
- Alpha-lipoic acid 600 mg/day (mitochondrial antioxidant, Tetrahydrobiopterin (BH4) regeneration support)
- PQQ 20 mg/day (supports mitochondrial biogenesis)
Rationale: If both autonomic and mitochondrial dysfunction must improve simultaneously to break the loop, single-target interventions might fail where combination succeeds. The 13-day recovery period suggests sustained support is needed—acute supplementation around exertion may be insufficient.
Qualification: This is highly speculative and has not been tested. Individual components have varying levels of evidence, but the specific combination and the mechanistic rationale are hypothetical. Safety profile is generally good for listed supplements at suggested doses, but medical supervision is appropriate, especially for patients on other medications.
3 Mitochondrial Turnover Rate Limitation
The 13-day recovery period (B. A. Keller et al. 2024) closely approximates published mitochondrial turnover times in muscle tissue (10–15 days). This is likely not coincidental.
Hypothesis: Exercise-induced ROS damage creates a population of dysfunctional mitochondria that must be removed via mitophagy and replaced via biogenesis. The rate-limiting step is not substrate availability (which recovers in hours) but the physical replacement of damaged organelles.
Implications:
- Why pacing works: Staying below the threshold that causes significant mitochondrial damage prevents the need for prolonged turnover-dependent recovery
- Why GET fails: Repeated exertion before turnover is complete accumulates progressively more damaged mitochondria
- Why baseline function declines: Steady-state mitochondrial dysfunction worsens if damage rate exceeds replacement rate
- Why severity varies: Individual differences in mitophagy/biogenesis capacity determine how quickly patients can recover
Documented in other contexts: Mitochondrial turnover limitation is established in aging, neurodegenerative diseases, and certain myopathies. The novelty here is recognizing it as central to post-exertional malaise.
3.1 Therapeutic Implications (Speculative)
If mitochondrial turnover is rate-limiting, interventions that accelerate both mitophagy (removal) and biogenesis (replacement) might shorten recovery time:
Mitophagy enhancement:
- Urolithin A 500–1000 mg/day: Directly stimulates mitophagy via PINK1/Parkin pathway; human trials show safety and efficacy in improving mitochondrial function in older adults
- Spermidine 1–3 mg/day: Autophagy inducer; safety established in human trials
- Time-restricted eating: If tolerated, 14–16 hour daily fast stimulates autophagy; CAUTION: many ME/CFS patients cannot tolerate fasting due to hypoglycemia symptoms
Mitochondrial biogenesis support:
- NAD+ precursors: NMN 500–1000 mg/day or NR 500–1000 mg/day activate sirtuins and PGC-1\(\alpha\) (master regulator of mitochondrial biogenesis)
- Resistance training: In healthy individuals, resistance exercise stimulates mitochondrial biogenesis; in ME/CFS, would require careful titration below PEM threshold (isometric exercises may be tolerable)
- Cold exposure: Mild cold stimulates PGC-1\(\alpha\); cold showers or cryotherapy if tolerated
Qualification: This approach is speculative. Urolithin A and NAD+ precursors have human safety data but not specifically in ME/CFS. The hypothesis that accelerating turnover would shorten recovery is logical but untested. Paradoxically, stimulating autophagy/mitophagy requires energy, so this approach might initially worsen symptoms in severely affected patients. Starting at very low doses and monitoring carefully would be essential.
4 Pre-Conditioning Hypothesis (Highly Speculative)
A counterintuitive idea emerges from cardiology and neuroscience: ischemic preconditioning. Brief, controlled ischemic episodes protect against subsequent severe ischemia by activating protective cellular programs.
Could analogous “metabolic preconditioning” work in ME/CFS? That is, could carefully controlled, very brief exertional stress—well below the PEM threshold—activate protective adaptations without causing damage?
Theoretical basis:
- Brief ROS bursts activate Nrf2 and other protective transcription factors
- Mild metabolic stress upregulates antioxidant enzymes and heat shock proteins
- Hormetic dose-response: small stress beneficial, large stress harmful
Potential protocol (entirely speculative):
- Very brief activity (30–60 seconds) at 50–60% of anaerobic threshold
- Performed every 48–72 hours initially
- Monitor for any PEM; if occurs, cease immediately and reassess
- Hypothesis: might gradually increase mitochondrial capacity without triggering damage
Major caveats:
- This contradicts pacing principles and could easily cause harm if dose miscalculated
- No evidence this would work in ME/CFS; ischemic preconditioning is mechanistically distinct
- Would only be appropriate for stable mild-to-moderate patients, not severe cases
- Requires extremely careful monitoring and willingness to abandon approach if harmful
Why mention it: Because the two-day CPET shows objective metabolic failure, it also provides an objective outcome measure for testing whether any intervention (including preconditioning) improves function. This hypothesis is offered as an example of testable ideas that emerge from mechanistic understanding, even if it seems counterintuitive.
5 Circadian Optimization of Recovery
Mitophagy and mitochondrial biogenesis are circadian-regulated processes, peaking at specific times of day. What if the prolonged recovery in ME/CFS reflects not just slow turnover but mistimed turnover due to circadian dysregulation?
Known facts:
- Mitophagy peaks during the inactive phase (night in humans)
- PGC-1\(\alpha\) (biogenesis regulator) has circadian expression
- ME/CFS patients have documented circadian abnormalities
- Sleep fragmentation impairs mitochondrial quality control
Hypothesis: If mitochondrial turnover processes are temporally disorganized, damaged mitochondria might persist longer because clearance and replacement occur out of phase with each other or are inefficiently timed.
5.1 Therapeutic Implications (Speculative)
If circadian timing matters for mitochondrial turnover, optimizing the timing of interventions might enhance efficacy:
Circadian stabilization:
- Strict sleep-wake schedule (even on weekends)
- Bright light exposure morning (10,000 lux for 30 min)
- Blue light blocking evening (2–3 hours before bed)
- Melatonin 0.5–3 mg at consistent time (8–9 PM)
- Temperature regulation (cool bedroom, 65–68°F)
Timed supplementation:
- Mitophagy inducers (urolithin A, spermidine): Evening dose to align with natural nocturnal mitophagy peak
- Biogenesis support (NAD+ precursors): Morning dose to support daytime activity
- Antioxidants: Split dose (morning and evening) for continuous protection
Qualification: This is speculative. While chronotherapy principles are established for other conditions (depression, jet lag), application to ME/CFS mitochondrial turnover is hypothetical. The interventions listed are generally safe but untested for this specific purpose.
6 Exercise Metabolomics-Guided Personalization
The two-day CPET provides a standardized exertional challenge. What if we performed detailed metabolomics immediately after Day 1 exercise to identify which specific substrates, cofactors, or metabolites are depleted in individual patients, then targeted repletion before Day 2?
Undocumented phenomenon: No study has performed comprehensive metabolomics in the immediate post-exercise period (0–6 hours) in ME/CFS to identify acute depletions.
Hypothesis: Individual patients may have distinct metabolic bottlenecks:
- Patient A: carnitine depletion (impaired fatty acid oxidation)
- Patient B: glutathione depletion (oxidative stress overwhelm)
- Patient C: tryptophan/kynurenine pathway derangement
- Patient D: purine nucleotide depletion (ATP synthesis substrate limitation)
Targeted repletion based on individual metabolic signatures might prevent Day 2 deterioration more effectively than generic interventions.
6.1 Research Protocol (Proposed)
- Baseline metabolomics: Plasma/serum immediately before CPET-1
- Post-exercise metabolomics: 30 min, 2 hours, and 6 hours after CPET-1
- Identify depletions: Metabolites showing >30% decline post-exercise
- Cluster analysis: Identify metabolic subgroups
- Targeted repletion trial: Provide individualized supplementation between Day 1 and Day 2
- Outcome: Measure whether Day 2 deterioration is reduced
Qualification: This is a proposed research direction, not an established finding. Metabolomics is expensive and not clinically available. However, if successful, it could guide development of standardized metabolic phenotyping that eventually becomes clinically accessible.
7 Vagal Stimulation for Recovery Acceleration
The autonomic nervous system has two branches: sympathetic (“fight or flight”) and parasympathetic (“rest and digest”). The parasympathetic branch, mediated by the vagus nerve, promotes:
- Anti-inflammatory signaling (cholinergic anti-inflammatory pathway)
- Enhanced mitochondrial biogenesis
- Improved heart rate variability
- Activation of repair/regeneration programs
ME/CFS patients show reduced vagal tone (low HRV, poor parasympathetic modulation). What if enhancing vagal activity could accelerate recovery from exertion?
Evidence level: Vagal nerve stimulation (VNS) is FDA-approved for epilepsy and depression. Non-invasive VNS devices are available. VNS has been shown to reduce inflammation and improve mitochondrial function in other contexts. However, it has not been tested specifically for ME/CFS post-exertional recovery.
7.1 Therapeutic Approach (Speculative)
Proposed protocol:
- Device: Transcutaneous auricular vagal nerve stimulation (taVNS) or transcutaneous cervical VNS
- Timing: Initiated within 1–2 hours of unavoidable exertion
- Duration: 30–60 minutes daily for 3–5 days post-exertion
- Parameters: Device-specific; typically 20–30 Hz stimulation
- Goal: Enhance parasympathetic tone during critical recovery period
Non-device alternatives:
- Deep breathing exercises (5–6 breaths per minute activates vagal reflexes)
- Humming or singing (stimulates vagus)
- Cold water face immersion (dive reflex)
- Specific yoga practices (if tolerable)
Qualification: This is moderately speculative. VNS devices have established safety profiles and known anti-inflammatory effects. The hypothesis that vagal stimulation could accelerate ME/CFS recovery is logical but unproven. Non-device alternatives are essentially free and safe, making them reasonable to try. Device-based VNS should be discussed with physicians and might not be covered by insurance for this indication.
8 Blood Flow Redistribution Training
Keller et al. concluded autonomic dysregulation affects blood flow and oxygen delivery (B. A. Keller et al. 2024). Standard autonomic training focuses on heart rate or blood pressure. What if we could specifically train better blood flow distribution to working tissues during activity?
Potential approaches (all speculative):
- Biofeedback: Real-time muscle oxygenation monitoring (NIRS - near-infrared spectroscopy) paired with activity; patient learns to maintain tissue oxygenation
- Blood flow restriction training: Paradoxically, very light exercise with partial blood flow restriction might train compensatory mechanisms; used in rehabilitation but untested in ME/CFS
- Postural countermeasures: Physical medicine approaches from POTS treatment (leg crossing, muscle tensing) might improve orthostatic blood redistribution
Undocumented: Muscle/brain tissue oxygenation during and after exercise has not been systematically measured in ME/CFS using NIRS or similar techniques. This would reveal whether oxygen delivery failure is indeed occurring and where (central vs peripheral).
The two-day CPET protocol (B. A. Keller et al. 2024) provides evidence of functional deterioration (reduced VO2max, workload, and ventilatory threshold on Day 2). However, a 2026 null replication by Mancini, Natelson et al. (n=58 ME/CFS, Fukuda criteria, n=25 sedentary controls) found no significant Day 1→Day 2 group-average decline in peak VO₂ or VO₂ at VT (Mancini et al. 2026). This null result raises critical questions about replicability: the Keller (2024) and Mancini (2026) studies differ on patient criteria (CCC vs. Fukuda), max-effort enforcement, and COI status. Both studies concur that ME/CFS patients have elevated perceived exertion at all workloads and lower maximal HR. The mechanistic hypotheses built upon CPET findings have not been directly tested:
- The “autonomic-mitochondrial feedback loop,” “mitochondrial turnover rate limitation,” and “circadian recovery gating” are interpretive models imposed on the 13-day recovery observation—the recovery duration alone cannot distinguish between these and simpler explanations (e.g., delayed inflammatory resolution, psychological anticipation).
- All supplement protocols proposed (autonomic-mitochondrial co-support, accelerated mitochondrial turnover, chronotherapy) are speculative combinations that have not been tested even in pilot studies; individual components have varying evidence levels, but the specific combinations and rationales are hypothetical.
- The “metabolic preconditioning” hypothesis directly contradicts established pacing principles and carries risk of harm; it is included as a theoretical exercise, not a recommendation.
- Quantitative model parameters (13-day recovery, mitochondrial half-life estimates) are drawn from healthy-population physiology and may not apply to ME/CFS patients with altered mitochondrial dynamics.
9 Summary Table: Novel Hypotheses from CPET Findings
Table CPET-Derived Hypotheses: Untested Mechanistic Inferences from Functional Data summarizes the mechanistic hypotheses and treatment implications emerging from two-day CPET evidence, ranked by likelihood and therapeutic potential.
| Hypothesis | Evidence | Therapeutic Potential | Key Prediction | Nearest-Term Test |
|---|---|---|---|---|
| Autonomic-mitochondrial feedback loop | Moderate | High | Synergy between catecholamine support + antioxidants exceeds either alone | 3-month trial: tyrosine + BH4 + MitoQ + N-Acetylcysteine (NAC) vs. components |
| Mitochondrial turnover rate limitation | Moderate–High | Moderate–High | Urolithin A + NAD+ precursors shorten recovery time | Repeat 2-day CPET after 12 weeks urolithin A/NMN |
| Circadian recovery gating | Low–Moderate | Moderate | Evening mitophagy enhancers + morning biogenesis support outperform mistimed dosing | Crossover trial: timed vs. untimed supplementation |
| Exercise metabolomics-guided therapy | Moderate | Very High | Individual metabolic signatures predict treatment response | Metabolomics at 0, 0.5, 2, 6h post-CPET; cluster patients |
| Vagal stimulation for recovery | Low–Moderate | Moderate | taVNS post-exertion reduces PEM severity and shortens duration | Post-exertion VNS vs. sham; symptom tracking 7 days |
| Blood flow redistribution training | Low | Low–Moderate | NIRS-guided biofeedback improves tissue oxygenation during activity | NIRS monitoring during standardized activity ±biofeedback training |
| Metabolic preconditioning (hormesis) | Very Low | Low (High Risk) | Brief sub-threshold stress improves Day 2 CPET metrics | NOT RECOMMENDED without extensive safety data |
Evidence level definitions:
- Very Low: Purely theoretical; no supporting evidence in ME/CFS
- Low: Mechanism plausible; analogous evidence from other conditions
- Low-Moderate: Mechanism plausible; some supportive but indirect ME/CFS evidence
- Moderate: Mechanism supported by multiple ME/CFS findings; direct intervention untested
- Moderate-High: Strong mechanistic support; similar interventions show benefit
- High: Direct evidence from ME/CFS studies
Therapeutic potential considers both magnitude of potential benefit and safety/accessibility profile.