Migraine as Slow-Motion ME/CFS: Cumulative Metabolic Reserve Erosion
This is the seventh and final article in a series on the energy biology linking ADHD, autism, and ME/CFS. This one examines migraine — not as a pain condition, but as a disease of progressive mitochondrial depletion that slowly erodes the metabolic reserve over decades.
1 The epidemiological signal
A Taiwan national cohort study followed 6,902 migraine patients and 27,608 matched controls from 2006 to 2010 (Lau et al. 2015):
- CFS incidence: 52.72 per 10,000 person-years in migraineurs vs 28.85 in controls
- Incidence rate ratio: approximately 1.5×
- Age ≥65: IRR = 2.11 — more than double the risk
- Dose-response relationship: risk scaled with migraine frequency
The risk increases with age. It increases with frequency. It accumulates.
Limitation: This study used ICD codes from the Taiwan National Health Insurance database. CFS diagnosis via administrative codes has low specificity — many coded cases may not meet ME/CFS research criteria (ICC or CCC). The association may partially reflect shared diagnostic patterns or medication side effects (migraine prophylactics commonly cause fatigue) rather than a pure biological relationship.
2 What happens during a migraine
Every migraine begins with cortical spreading depression (CSD) — a wave of intense neural depolarisation that sweeps across the brain surface at 2–5 mm/minute. During CSD:
- Massive ion flux occurs (potassium floods out of cells, sodium and calcium flood in)
- Neurons depolarise completely and fire simultaneously
- ATP consumption spikes to restore ion gradients via Na⁺/K⁺-ATPase
- The cortex goes electrically silent in the wave’s wake — because every neuron has exhausted its energy reserves
CSD is one of the most energy-demanding events in neurobiology. The recovery — restoring ion gradients, rebuilding ATP stores, repairing oxidative damage — requires hours of mitochondrial work.
3 The chronic deficit between attacks
Migraineurs are not metabolically normal between attacks. Multiple lines of evidence show chronic mitochondrial impairment even in the interictal period (Wang et al. 2023):
- Elevated blood lactate (indicating anaerobic glycolysis compensation)
- Lower N-acetylaspartate (NAA) on MR spectroscopy (a marker of neuronal energy capacity and mitochondrial function)
- Decreased activities of:
- NADH dehydrogenase (Complex I)
- Citrate synthase (Krebs cycle entry enzyme)
- Cytochrome c oxidase (Complex IV — the terminal electron acceptor)
- Calcium dysregulation creating feed-forward oxidative damage
Between migraines, the brain is not “normal.” It is operating with reduced mitochondrial capacity. Under the cumulative depletion model, each attack depletes reserves and each recovery is slightly incomplete.
4 The cumulative depletion model
This is where migraine differs from ADHD and ASD as a metabolic reserve reducer. ADHD and ASD represent static baseline deficits — the generator was built smaller. Migraine represents progressive erosion — the generator was normal-sized but is being damaged over time.
Each cortical spreading depression event:
- Creates massive, transient ATP demand
- Generates reactive oxygen species during the recovery phase
- Oxidatively damages mitochondrial membranes and mtDNA
- Requires mitochondrial repair and biogenesis for full recovery
- In a brain whose mitochondria are already impaired, full recovery may not occur
The result: a ratchet effect. Each CSD episode depletes and damages. Recovery is incomplete. Baseline capacity drops slightly. Next episode starts from a lower floor.
Over 10 years of episodic migraine: hundreds of CSD events. Over 30 years: thousands. The cumulative effect is measurable in the interictal metabolic markers cited above.
5 Why risk amplifies with age
The Lau et al. finding that CFS risk is 2.11× at age ≥65 (vs 1.5× overall) makes immediate sense under this model:
- A 25-year-old with 10 years of episodic migraine has experienced perhaps 100–500 CSD events (rough illustrative estimate based on attack frequency)
- A 65-year-old with 50 years of migraine has experienced perhaps 1,000–5,000 CSD events (same logic, longer timeframe)
- More events = more cumulative damage = lower remaining reserve = lower threshold for decompensation
Normal ageing also reduces mitochondrial function (declining Complex I activity, accumulating mtDNA mutations, reduced mitophagy). Migraine accelerates this process. The curves converge at older ages.
7 Implications for migraine patients
1. Aggressive prevention is metabolic reserve preservation.
Every prevented CSD event is energy the brain gets to keep. Every prevented migraine is insurance against future post-infectious collapse. This reframes migraine prevention from “comfort measure” to “neuroprotective intervention.”
CGRP monoclonal antibodies (erenumab, fremanezumab, galcanezumab) are prescription biologics that reduce migraine frequency by 50–75% in responders. Under the metabolic reserve model, they may be preserving reserve as a secondary benefit — though this has never been measured.
2. Mitochondrial support has a specific rationale in migraine.
CoQ10 (100–300 mg), riboflavin (400 mg), and magnesium (400–600 mg) have evidence as migraine prophylactics. The metabolic reserve model explains why: they support electron transport chain function, partially offsetting the cumulative CSD damage. They are literally reserve-building supplements.
3. For ME/CFS patients with pre-existing migraine history:
The mitochondrial deficit is both older and potentially larger than ME/CFS illness duration alone would suggest. A patient with 20 years of migraine before ME/CFS onset has a pre-existing deficit that predates and compounds the post-infectious damage. Treatment strategies should acknowledge this longer history of cumulative depletion.
Note: The dosages and medications named above reflect published research findings on migraine prophylaxis, not personalised treatment recommendations. Any intervention should be discussed with a clinician familiar with the individual’s full medical context.
8 The complete picture: Architecture C
Across this seven-article series, a unified framework has emerged:
| Condition / predisposing factor | Mechanism of reserve reduction | Type |
|---|---|---|
| ADHD | Prefrontal hypometabolism + dopaminergic inefficiency (Zametkin et al. 1990) (Berthier et al. 2025) | Static baseline deficit |
| ASD | Systemic mitochondrial ETC dysfunction + low BH4 (Frye et al. 2024) (Fanet et al. 2021) | Static baseline deficit |
| hEDS/POTS | Cerebral hypoperfusion from connective tissue laxity | Static delivery deficit |
| Migraine | Cumulative CSD damage to mitochondria | Progressive erosion |
| GCH1 rs841 | Reduced BH4 production | Static genetic deficit |
| Iron deficiency | Impaired complex I/II + dopamine synthesis | Modifiable deficit |
Each condition reduces the brain’s metabolic reserve through a different mechanism. The convergent result: less buffer to absorb an immune insult. Neurodivergent traits in childhood predict chronic disabling fatigue in adolescence (Quadt et al. 2024). The convergent outcome when a virus arrives: ME/CFS.
The framework generates clear research priorities:
- Measure pre-illness metabolic reserve in at-risk populations
- Track post-infectious outcomes stratified by reserve markers
- Test whether correcting modifiable reserve reducers (iron, BH4, cerebral perfusion) prevents ME/CFS or improves outcomes
These studies haven’t been done. But for the first time, the hypothesis is clear enough to test.
What would refute this model? If pre-illness metabolic reserve markers (ferritin, BH4, cerebral perfusion) do NOT predict post-infectious ME/CFS risk after controlling for inflammatory markers, the metabolic reserve hypothesis would be falsified in favor of a simpler inflammation model. If correcting modifiable deficits (iron, BH4, perfusion) in at-risk populations has no effect on ME/CFS incidence, the framework loses its clinical utility. The model makes specific, testable predictions — and it can fail.
Part 7 (final) of a series on the energy biology linking ADHD, autism, and ME/CFS.