Within-Patient Attractor Migration
The multi-attractor model of ME/CFS (Section Bifurcation Analysis and Disease Subtypes) defines several disease attractor basins: immune-dominant, metabolic-dominant, neurovascular-dominant, and severe/locked. A consequence of the model that has not been explicitly discussed is that patients may migrate between attractor basins over time—a metabolic-dominant patient who develops autoantibodies migrates toward the neurovascular-dominant or severe/locked attractor. This explains the common clinical observation that ME/CFS “changes character” over years (Chu et al. 2019) (Nacul et al. 2020).
The migration dynamics are governed by the acquisition of new locks:
\[ \text{Attractor} &= f(\alpha_\text{CI}(t), S(t), C_\text{pro}(t), overline(\mathcal{M})(t), \text{Ab}(t)) \]
As slow variables evolve (autoantibodies develop, epigenetic consolidation deepens), the attractor landscape itself changes, and the patient’s position migrates across basins.
Longitudinal ME/CFS disease progression follows predominantly directional migration through attractor basins: immune-dominant → metabolic-dominant → severe/locked, driven by progressive acquisition of locks (epigenetic consolidation, autoantibody development). This migration is entropy-increasing (toward deeper, more stable attractors) and explains why ME/CFS “changes character” over years and why late-stage disease is more treatment-resistant than early-stage disease, independent of symptom severity.
Certainty: 0.40. Directly implied by the multi-attractor ODE model and consistent with clinical observations of disease evolution (Chu et al. 2019) (Lacourt, Verson, et al. 2022). The specific directional prediction (toward deeper attractors) follows from the lock-acquisition dynamics. However, prospective longitudinal subtyping data to confirm or deny migration are not available.
Testable predictions:
- Longitudinal subtyping (repeated multi-omics every 6–12 months) will show patients shifting between attractor basins over time.
- Migration is predominantly unidirectional: immune → metabolic → severe (entropy-increasing, toward deeper attractors). Reverse migration (severe → metabolic) should be rare without intervention.
- Specific migration paths predict specific lock acquisitions: metabolic → severe implies epigenetic consolidation occurred during the transition.
- Treatment response at any time point is determined by current attractor (current lock configuration), not original attractor (entry mechanism).
Limitations: No longitudinal multi-omics ME/CFS study exists with the temporal resolution to track attractor migration. Cross-sectional data cannot distinguish migration from initial heterogeneity. The attractor basin definitions are model-dependent and may not correspond to cleanly separable clinical phenotypes.