Treatments people are trying

If crashes are steps 5 to 7 firing when demand outruns delivery, keeping demand below the threshold should prevent them. Pacing treats nothing; it stops the loop being driven harder.

Keeps effort below the level that triggers post-exertional malaise, using a heart-rate alarm (often set around resting rate plus 15 bpm, or 55-60% of age-predicted maximum) and planned rest. In the model it stops steps 5 to 7 from firing, rather than treating anything.

Everyone with post-exertional malaise. Most useful for people who crash after delays they cannot predict, and for carers who need an objective stop signal.

Few. Heart-rate limits miss cognitive, sensory and orthostatic load, and beta-blockers or ivabradine make the numbers unreliable. Strict tracking can add anxiety.

For bedbound patients, sitting up, talking, light and noise count as exertion; overnight resting heart rate and HRV trends are often more useful than daytime alarms.

OTC · £30-300 one-off for a chest strap or wearable

Standing is the commonest everyday demand on step 5, because low volume (step 3) leaves the brain under-supplied when upright. Less upright time means less under-perfusion.

Counts upright minutes as part of the energy budget: sitting with legs raised, lying for tasks that can be done flat, and short upright spells broken by lying down. It reduces the time the brain and muscles are under-supplied.

Anyone with orthostatic intolerance (light-headedness, brain fog or heart racing on standing), whether or not POTS criteria are met.

Very low. Long-term complete bed rest has its own costs (muscle and bone loss, clots), so keep whatever gentle movement is tolerated lying down.

Usually already in place by necessity; a reclining wheelchair or bed that raises the head slightly at night (some POTS clinics advise 10-15 degrees) may help.

OTC

Directly targets step 3. If low plasma volume limits filling of the heart, more fluid and sodium should raise stroke volume and brain blood flow when upright. It does not correct red-cell volume.

Expands plasma volume so more blood returns to the heart when upright. Typical POTS advice is 2-3 litres of fluid and 6-10 g of salt a day (3-5 g sodium), spread through the day, or oral rehydration solution.

People with orthostatic intolerance, low blood pressure, dark urine or high plasma osmolality, and those who pass a lot of dilute urine.

Not for people with high blood pressure, heart failure or kidney disease. Salt tablets can cause nausea; take with food. Very large water intake without salt can lower blood sodium.

Swallowing and toileting are effortful in severe ME; electrolyte solutions in small, frequent sips and a bedside urinal bottle help.

OTC · £5-30/month

Reduces pooling in leg and gut veins, which raises the return of blood to the heart (step 3) and so brain and muscle perfusion (step 5). Compression halved the fall in cardiac output on tilt in one study.

Squeezes leg and abdominal veins so less blood pools below the heart when upright. Abdominal binders matter as much as stockings, because much of the pooling is in the gut.

Anyone with orthostatic symptoms; third in the patient survey ratings (net ~62%), especially in the POTS-dominant cluster.

Heat, itching and the effort of putting them on. Thigh-high or waist-high class 1-2 (20-30 mmHg) is typical; knee-high adds little. Not for people with peripheral arterial disease.

Only useful when upright; putting on stockings can itself cause a crash, so a carer's help or an easy-on abdominal binder may be better.

OTC · £20-60 per garment

Constricts veins and arteries, partly compensating for low volume and poor venous tone (step 3) and reducing the fall in brain blood flow when upright (step 5).

Tightens arteries and veins for about 4 hours per dose, raising blood pressure and reducing pooling when upright.

People whose problem is low blood pressure or pooling on standing, who are upright for set periods during the day.

High blood pressure when lying down (the main danger), goosebumps, scalp tingling, urinary retention. Not for heart disease, kidney failure or uncontrolled hypertension. Last dose at least 4 hours before lying down for the night.

Usually unsuitable for bedbound patients because of lying-down hypertension.

prescription · NHS prescription charge, or ~£20-40/month private

Strengthens the autonomic signal that tightens veins, addressing preload failure (step 3) and, through it, under-perfusion during exertion (step 5). This fits the subgroup with low filling pressure on invasive testing best.

Boosts acetylcholine signalling in autonomic nerves, which helps tighten veins and return blood to the heart (preload).

POTS or suspected preload failure; possibly those with small-fibre neuropathy on skin biopsy.

Gut cramps, diarrhoea and urgency (about one in five stop for this in POTS series), sweating, increased saliva, muscle twitching. Caution in asthma and slow heart rate.

Bowel urgency is a real burden when getting to the toilet is hard; start at 30 mg or lower.

prescription · NHS prescription charge, or ~£15-30/month private

LIFT trial (pyridostigmine x low-dose naltrexone, n=160, NCT06366724), recruiting, primary completion listed as Sept 2027

Blunt the adrenaline-driven heart-rate rise (step 8) so the heart fills more between beats (step 3). They treat a compensation, which is why higher doses can worsen fatigue.

Blunts the adrenaline-driven rise in heart rate, letting the heart fill more between beats. Low dose matters: higher doses can worsen fatigue.

POTS with a large, adrenaline-driven heart-rate rise, tremor or palpitations.

Fatigue, low blood pressure, cold hands, vivid dreams; not for asthma. Can mask low blood sugar. Do not stop suddenly.

Fatigue from beta-blockers can be hard to separate from ME; start at 10 mg or less.

prescription · NHS prescription charge

Corrects step 3 quickly and fully for a few days. A reliable short-term response is evidence that volume matters for that patient; its brevity suggests volume is not the whole story.

Puts volume straight into the circulation, temporarily correcting low blood volume and preload. The benefit usually lasts days, not weeks.

Patients with clear hypovolaemia or POTS who still struggle after oral fluids, salt and compression; some use it before a necessary outing or procedure.

Cannula infection, vein damage, fluid overload in heart or kidney disease. Regular infusions often lead to a port or PICC line, which carries a real risk of line infection and clots; avoid long-term lines for this alone.

Home infusion services avoid travel; the trip to a clinic can cost more than the infusion gives.

specialist · £100-250 per private infusion

Saline vs placebo infusion in an ME/CFS PEM study (NCT04740736, n=106) completed July 2025, not yet reported; oral rehydration vs IV saline crossover (NCT02854683) status unknown since 2021

Slowing the heart lets it fill for longer (step 3), and blunts the heart-rate rise of arousal (step 8). It lowered heart rate but did not improve symptoms, which suggests heart rate itself is a marker, not a driver.

Slows the heart's pacemaker without lowering blood pressure, so the heart fills for longer between beats.

Possibly hyperadrenergic POTS (standing noradrenaline above ~600 pg/mL) with distressing palpitations. Do not expect better energy or function.

Flashing lights in vision (phosphenes), slow heart rate, atrial fibrillation; not in pregnancy. Makes heart-rate pacing alarms unreliable.

specialist

Full RECOVER-AUTONOMIC paper or posted results (NCT06305806) not yet available; COVIVA ivabradine RCT (NCT05481177, n=250 planned) completed Sept 2024, not reported

Was meant to expand volume (step 3) by retaining salt. Its failure in a 100-person trial is one of the results the model explains least well: either aldosterone is not the bottleneck, or volume is not the main limit on the crash.

A salt-retaining steroid that makes the kidneys hold sodium and water, to expand blood volume.

Not recommended as an ME/CFS treatment. A POTS specialist may still use it for documented low blood volume or low aldosterone; the trials did not select for those.

Low potassium, high blood pressure, headache, fluid retention; potassium needs checking.

prescription

Targets the brain end of the sickness response (step 7): interferon studies show part of it runs through reduced dopamine signalling in the basal ganglia, and reward-related basal ganglia activity is lower in ME/CFS.

At very low doses it nudges dopamine signalling and may calm microglia, targeting the basal-ganglia dopamine dip in the sickness response.

Prominent brain fog and unrefreshing sleep, once sleep is reasonably steady.

Insomnia, restlessness (akathisia), vivid dreams; rarely movement disorders or impulse-control problems (gambling, shopping). Needs a liquid formulation or tablet splitting to reach 0.25-1 mg. The ME Association advised against use outside trials in 2021.

Start at 0.25 mg or lower; restlessness is hard to bear when bedbound.

prescription · ~£30-80/month private including liquid formulation

PAIS-AriSE (Charité): phase 2b placebo-controlled crossover in ~138 people with post-acute infection syndrome; protocol only (unverified)

Aimed at cognitive symptoms of the sickness response (step 7) and at prefrontal effects of high noradrenaline (step 8). The model gives no reason to expect it to change the crash threshold.

Guanfacine strengthens prefrontal working-memory circuits through alpha-2A receptors; NAC is an antioxidant. Proposed for cognitive symptoms.

Cognitive-dominant illness with adequate blood pressure.

Guanfacine lowers blood pressure and heart rate and causes drowsiness; it is a cousin of clonidine, which reduced daily activity in a large adolescent CFS trial (see clonidine). Do not stop guanfacine suddenly. NAC can cause nausea.

Blood-pressure lowering can worsen orthostatic intolerance.

prescription · ~£50-80/month private (guanfacine); NAC £10-20/month

Better sleep should weaken the feedback from step 8 to step 2. Melatonin fits the subgroup with delayed sleep onset; low-dose sedating drugs act on sleep and pain.

Melatonin shifts the body clock and promotes sleep onset; low-dose trazodone or amitriptyline (10-25 mg) deepen and lengthen sleep and may ease pain.

Broken or delayed sleep; amitriptyline where pain or headache is also present.

Morning grogginess. Amitriptyline: dry mouth, constipation, weight gain, faster heart rate (can worsen POTS). Trazodone: dizziness on standing, rarely priapism. Melatonin is prescription-only in the UK.

Very severe patients often react to standard doses; start at a quarter of the usual dose.

prescription · NHS prescription charge

Second melatonin plus zinc RCT (NCT05454683, n=106 planned) completed Sept 2024, not yet reported

Act on the alertness and concentration side of step 7 without touching steps 3 to 6, so the model predicts symptom relief without a higher crash threshold, and a risk of overexertion.

Raises dopamine and noradrenaline, improving alertness and concentration. It does not raise the energy envelope.

Cognitive-dominant illness where specific tasks (work, exams) matter; the cognitive cluster in the patient survey rated stimulants highly.

The main risk is overdoing it: feeling more alert can lead to exertion that triggers a crash. Also raised heart rate (bad for POTS), insomnia, appetite loss, dependence.

Generally unsuitable in severe ME because of heart-rate rise and crash risk.

specialist · NHS prescription charge if prescribed; off-label

Methylphenidate plus nutrient formula RCT (NCT01966276, n=134 planned) completed 2014, results never reported

Proposed to shift autonomic balance toward vagal tone (step 8) and damp immune-to-brain signalling (step 7). Two sham-controlled trials found no difference, so the rationale has not translated into benefit.

A clip on the ear stimulates the vagus nerve, which can shift heart-rate variability toward rest and is thought to dampen inflammatory signalling.

People who want a very low-risk thing to try for sympathetic over-arousal or sleep onset, accepting that sham stimulation did as well in trials.

Skin irritation, dizziness (each ~7% in COVIVA). Not with a pacemaker or implanted stimulator without advice.

Can be used in bed; some find the sensation itself over-stimulating.

OTC · £200-500 one-off

UK vagus nerve stimulation RCT vs usual care (NCT07190105, n=546) finished Dec 2025, not yet reported

Raises dopamine, which could offset the basal ganglia part of the sickness response (step 7). Untested in ME/CFS beyond small open series.

An old antiviral and Parkinson's drug that raises dopamine and blocks NMDA receptors; used for fatigue in multiple sclerosis.

Wait for the placebo-controlled result; the only controlled ME/CFS data were negative.

Insomnia, dizziness, confusion, hallucinations, ankle swelling, livedo (mottled skin); reduce dose with kidney disease.

prescription

Placebo RCT for cognitive impairment in long COVID (NCT06055244, n=64) completed Dec 2025, not yet reported

Lowers sympathetic outflow (step 8). It did not help and trended to fewer steps in an adolescent trial, which argues against high sympathetic tone being a driver.

Damps sympathetic (fight-or-flight) output from the brainstem.

Not recommended for ME/CFS itself. The trial suggests the raised sympathetic tone may be compensating for poor circulation.

Low blood pressure, drowsiness, dry mouth; rebound hypertension if stopped suddenly.

prescription

Meant to strengthen the vagal anti-inflammatory reflex (steps 7 and 8). Failed in a controlled trial.

An Alzheimer's drug that raises acetylcholine in the brain; proposed to boost the anti-inflammatory vagal reflex.

No one with ME/CFS; the large trial was clearly null.

Nausea, vomiting, diarrhoea, weight loss, slow heart rate and fainting.

prescription

Mast-cell mediators dilate vessels and increase leakiness, which would worsen pooling and under-perfusion (steps 3 to 5) and amplify the response to tissue injury (step 6). Plausible for the subgroup with clear mast-cell symptoms.

H1 blockers (loratadine, cetirizine, fexofenadine) and H2 blockers (famotidine) block histamine; cromolyn stops mast cells releasing it in the gut.

Flushing, hives, itch, food or smell reactions, or POTS with allergic-type flares (up to a quarter meet mast-cell activation criteria).

Drowsiness with older antihistamines; H2 blockers reduce stomach acid and can affect absorption of some drugs. Famotidine and cromolyn need a UK prescription; H1 blockers are over the counter.

Non-drowsy H1 blockers first; sedating ones can worsen daytime grogginess.

OTC · £2-10/month (H1); cromolyn ~£60-150/month on private prescription

Proposed to calm microglia through TLR4, which would damp the brain end of the sickness response (step 7) and possibly the immune response to injury (step 6). The first sizeable placebo-controlled trial was null.

At 1-4.5 mg naltrexone briefly blocks opioid receptors and may calm microglia via TLR4; it has also been reported to restore TRPM3 channel function in NK cells in the lab.

People with prominent pain or sleep problems who accept that the trial found no benefit for fatigue or pain; cheap and generally well tolerated.

Vivid dreams, insomnia, headache in the first weeks. Cannot be combined with opioid painkillers.

Start at 0.5-1 mg; sensitive patients react to small doses.

prescription · ~£20-40/month private plus consultation

LIFT trial (NCT06366724, n=160, with pyridostigmine) recruiting, primary completion listed as Sept 2027; RECOVER paediatric trial ongoing

Sigma-1 activity is thought to damp inflammatory signalling inside cells, which would act on the immune response to injury (step 6) and the sickness response (step 7). A 399-person placebo-controlled trial in long COVID was positive; it has not been tested in ME/CFS.

An SSRI antidepressant that also acts on the sigma-1 receptor, which is thought to damp inflammatory signalling inside cells.

Long COVID with prominent fatigue, after discussing interactions with a doctor. Only tested for 60 days, and not in ME/CFS that did not follow COVID.

Nausea, sleep changes, agitation in the first weeks. Strong interactions: it raises caffeine, theophylline, tizanidine, clozapine and some blood thinner levels. Serotonin syndrome risk with other serotonergic drugs. Taper rather than stop suddenly.

Start at a low dose; sensitive patients often react to standard SSRI doses.

prescription · NHS prescription charge if prescribed; off-label

ESPRIT placebo RCT in long COVID (NCT07359482, n=160) recruiting

Restrains over-active immune cells and boosts cellular clean-up, which could act on a primed immune state (step 1) and on the response to tissue injury (step 6). Evidence is uncontrolled; a controlled trial ends in November 2026.

Blocks mTOR, which increases cellular clean-up (autophagy) and restrains over-active immune cells.

Worth revisiting when the Mount Sinai placebo-controlled trial reports.

Mouth ulcers, raised cholesterol and blood sugar, impaired wound healing, infection risk; needs blood monitoring. Many drug interactions (grapefruit, some antifungals and antibiotics).

specialist · ~£30-100/month private plus monitoring

Mount Sinai placebo-controlled sirolimus trial in long COVID (NCT06960928, n=80), ends Nov 2026

Broad immune suppression, testing whether an ongoing immune process (steps 1 and 6) drives the illness. Open-label only, with serious toxicity.

A chemotherapy drug that kills dividing immune cells, including antibody-making cells.

No one outside a trial; the investigators themselves advise against routine use.

Nausea in ~90%, infertility, bladder damage and bladder cancer, other cancers, infections; 11 serious adverse events in 8 of 40 patients.

trial

Removes long-lived plasma cells that rituximab does not reach, testing whether persistent antibodies (step 1, and the IgG effect on endothelium in step 4) sustain the illness. Open-label pilot only.

An anti-CD38 antibody that removes long-lived plasma cells, the source of persistent antibodies that rituximab cannot reach.

Only within trials; an NK-cell threshold is being used to select patients.

Infusion reactions, infections (shingles prophylaxis usually needed), low antibody levels. Very expensive cancer drug.

trial

ResetME randomised placebo-controlled phase II, Haukeland (n=66, CTIS 2024-520094-13-00), started 2025, ongoing

Would act on autoimmunity (step 1) and transiently expands volume (step 3). Controlled results conflict.

Pooled antibodies from donors that dampen autoimmune activity; the infusion also expands blood volume.

Proven small-fibre neuropathy or immune deficiency are separate indications a neurologist or immunologist may treat; not for ME/CFS alone.

Headache, aseptic meningitis, clots, kidney injury, infusion reactions. Very expensive and in short supply.

Severe patients rated IVIG highly in the patient survey, but the controlled data do not support it.

specialist · ~£2,000-5,000 per infusion privately

RECOVER-AUTONOMIC IVIG arm in long-COVID POTS (NCT06305793, n=200) finished Mar 2026, not yet reported; a subcutaneous immunoglobulin trial in post-COVID POTS (NCT06524739, n=16) was stopped early with uninterpretable results

Blocks inflammasome activity and white-cell movement, which would damp the immune response to tissue injury (step 6). A long COVID trial was null.

An old gout drug that blocks inflammasome activity and white cell movement.

Not for long COVID or ME/CFS outside a trial; the one blinded trial was null.

Diarrhoea and nausea are common; toxic in overdose; dangerous interactions with clarithromycin, some antifungals and statins; avoid with kidney or liver disease.

prescription

RECLAIM (metformin and colchicine vs usual care, NCT07280572, n=1,250 planned) enrolling

Lowers all IgG, testing whether antibodies (step 1, and the IgG effect on endothelium) sustain the illness. Its long COVID trial was null.

Lowers all IgG antibodies, including autoantibodies, by blocking the receptor (FcRn) that recycles them.

Not for long COVID, POTS or ME/CFS outside a trial. It is licensed for myasthenia gravis and CIDP.

Infections, headache, infusion reactions; weekly IV infusions; very expensive.

Weekly infusions are a heavy load.

specialist

Proposed to release anti-inflammatory and repair signals, which would act on steps 1 and 6. Early and small.

Infuses donor blood-forming or mesenchymal stem cells, which are thought to release anti-inflammatory and repair signals rather than replace tissue.

Only within registered trials. Private stem cell clinics offering this for ME/CFS or long COVID are selling an unproven product.

Infusion reactions, infection from poorly made products, and at unregulated clinics serious harms reported with other conditions. Large costs abroad.

Travel to clinics abroad is a heavy exertion load.

trial

Cord blood RCT in long COVID (NCT06492798, n=76 planned) primary completion July 2025, status unknown, not reported

Tested whether autoantibody-producing B cells sustain the illness (step 1). The 151-person phase III trial was null, which fits the model's view that immune changes are mostly downstream.

Removes B cells, the precursors of antibody-producing cells.

No one, for ME/CFS. The earlier positive Norwegian phase II results did not survive a proper trial.

Infusion reactions, infections, low antibody levels, rare brain infection (PML).

specialist

Tested whether circulating receptor autoantibodies drive vascular dysfunction (steps 1 and 4). Sham-controlled trials were null.

Filters antibodies (immunoadsorption) or all plasma proteins (plasma exchange) out of the blood over several sessions. BC007 is an infused drug meant to neutralise the same autoantibodies.

Not recommended outside a trial; removing antibodies did not predict who improved even in the open-label work.

Central line infection and clots, low blood pressure, bleeding, low antibody levels afterwards. Several days of hospital procedures.

Travel and sessions are a heavy exertion load.

specialist · €10,000-20,000 per course privately (Germany)

Would act on step 1 if intermittent herpesvirus reactivation keeps priming the immune system. Fits the herpes open question; controlled results are mixed.

Blocks herpesvirus DNA copying. Valganciclovir covers HHV-6 and CMV better; valacyclovir mainly covers HSV and EBV lytic replication.

Illness that began with glandular fever or another proven herpesvirus infection, ideally with evidence of active replication rather than old antibody titres.

Valganciclovir: low white cells and platelets, needs regular blood counts, and is toxic to sperm and pregnancy. Valacyclovir: headache, nausea, kidney effects in dehydration.

Regular blood monitoring needs home phlebotomy.

specialist · Valacyclovir ~£15-30/month; valganciclovir several hundred £/month

Valacyclovir + celecoxib pilot RCT in long COVID (NCT06316843, n=59) completed Oct 2024, not yet reported; Valacyclovir + celecoxib (+ Paxlovid) placebo-controlled long-COVID trials planned (Bateman Horne Center; NCT07597902, not yet recruiting)

Stimulates the TLR3 antiviral pathway, aimed at an abnormal antiviral state (step 1). One borderline phase III result, manufacturer-run.

A synthetic double-stranded RNA that stimulates the TLR3 antiviral pathway.

Only via US expanded-access or trials; not available in the UK.

Flushing and flu-like reactions, liver enzyme rises; twice-weekly IV infusions for months, usually through a line.

Infusion schedule is a heavy load.

trial

Tested whether an activated endogenous retrovirus drives steps 1 and 6. Null in 203 patients.

An antibody against the envelope protein of HERV-W, an ancient retrovirus sequence in human DNA that viral infection can switch on.

No one outside a trial; not available, and the trial was null.

Infusion reactions; monthly IV infusions.

trial

Tested whether persisting SARS-CoV-2 drives step 1. Three placebo-controlled trials were null.

Paxlovid blocks the SARS-CoV-2 protease; monoclonal antibodies bind the spike protein. Both test whether persisting virus drives symptoms.

Not for established long COVID or ME/CFS. Treating acute COVID early may lower later long-COVID risk (small, early evidence) - that is a different question.

Paxlovid: taste disturbance, diarrhoea, many serious drug interactions via ritonavir.

prescription

Supplies cofactors for mitochondrial energy production, which would help tissue cope with under-perfusion (step 5). The model puts the limit in delivery rather than in mitochondria, so it predicts at most a small effect.

Supplies two cofactors of the mitochondrial energy chain.

Anyone wanting a low-risk supplement; expect at most a small effect.

Mild stomach upset, insomnia if taken late; CoQ10 can reduce warfarin effect.

OTC · £30-60/month

Aims to support energy metabolism in under-supplied tissue (step 5). The model predicts a small effect for the same reason as CoQ10.

A Krebs-cycle intermediate thought to shift the NAD+/NADH balance and support energy metabolism.

People who can afford it and want a low-risk trial of a supplement; give it 6-12 weeks.

Mild stomach upset. Mostly a cost issue.

OTC · ~£100-200/month at 2 g/day

Creatine and ribose top up energy buffers in tissue under strain (step 5). Like CoQ10, these act after the delivery problem the model puts first.

Creatine tops up the phosphocreatine energy buffer; D-ribose is a sugar used to build ATP; carnitine carries fats into mitochondria; NAD precursors raise a key energy cofactor. Cocktails combine these with B vitamins and antioxidants.

People who want to self-experiment cheaply; there is no good reason to spend much on them.

Generally low. Creatine raises blood creatinine, which can be misread as kidney damage (tell the GP; avoid with kidney disease). D-ribose can lower blood sugar; carnitine can cause fishy body odour and gut upset.

OTC · £5-80/month

MINIRICO, nicotinamide riboside plus mind-body therapy in long COVID (NCT05703074, n=310), completed Sept 2024, not yet reported

Changes cellular energy signalling and has proposed anti-inflammatory effects (steps 5 and 6). Its long COVID prevention signal was in acute infection; trials in established illness are null.

A diabetes drug that changes cell energy signalling (AMPK) and has proposed antiviral and anti-inflammatory effects.

Not for treating established long COVID or ME/CFS. Taking it during acute COVID to lower long-COVID risk is a separate question this entry does not cover. People with diabetes take it for that.

Diarrhoea, nausea and stomach pain are common; lower vitamin B12 with long use; rarely lactic acidosis, mainly with kidney disease or dehydration.

Gut side effects and appetite loss are a problem for patients who struggle to keep weight on.

prescription

RECLAIM (metformin and colchicine vs usual care, NCT07280572, n=1,250 planned) enrolling

Light absorbed by mitochondria can release nitric oxide, dilating local vessels (step 4) and raising local energy supply (step 5). No ME/CFS trials.

Light at 600-850 nm is absorbed by cytochrome c oxidase in mitochondria, which can release nitric oxide and raise ATP locally; near-infrared reaches a few centimetres into tissue.

Anyone wanting a very low-risk option that can be done lying down.

Eye safety (goggles for near-infrared), skin warmth; avoid over the eyes and in light-sensitive patients.

Light-sensitive patients may not tolerate red panels; near-infrared is largely invisible.

OTC · £100-1,500 one-off

Proposed to restore cholinergic signalling, which would act on vagal tone (step 8) and vessel function (step 4). Case series only.

Proposed to free up nicotinic acetylcholine receptors thought to be blocked after COVID, restoring cholinergic signalling.

Long-COVID-onset illness; only with a doctor's agreement, given the cardiovascular effects.

Nausea, dizziness, palpitations, sleep disturbance, skin reaction; dependence. Some users report crashes. Not for people with heart disease, pregnancy or a history of nicotine dependence.

Start with a fraction of a 7 mg patch.

OTC · £15-30/month

Dissolves extra oxygen in plasma, which reaches tissue that red cells under-supply (steps 4 and 5), and at full dose may stimulate vessel growth if capillary loss matters (step 4). A sham-controlled ME/CFS trial is under way.

Breathing 100% oxygen at twice atmospheric pressure dissolves extra oxygen in plasma; repeated sessions may stimulate new vessel growth.

Mild to moderate patients who can manage 40+ trips; the positive trials used 40 sessions, not 10.

Ear and sinus barotrauma, temporary short-sightedness, confinement anxiety; rarely oxygen seizures. Not with untreated pneumothorax.

Daily travel to a centre for 8 weeks is usually impossible for severe patients and could cause a crash on its own.

specialist · £50-150/session; ~£2,000-6,000 for 40

Sham-controlled ME/CFS RCT at Shamir (Assaf-Harofeh), Israel: 60 sessions, n=74, NCT07621068, completion ~2027

Pushes blood back to the heart (step 3) and raises shear stress on vessel walls, which can improve endothelial function (step 4). Case series only in this illness.

Cuffs on the legs inflate in time with the heartbeat, pushing blood back to the heart and raising shear stress on vessel walls; usually 35 one-hour sessions.

Mild patients near a centre; cardiology departments use it for angina.

Leg bruising and skin abrasion; not for people with leg clots, aortic aneurysm or severe valve disease.

Thirty-five sessions of travel make it impractical for severe patients.

specialist · ~£3,000-6,000 per course privately

EXPECT sham-controlled trial, NCT05668039 (status unknown, primary completion Apr 2024, results not posted)

The same rationale as clinical HBOT (steps 4 and 5), at roughly a third to two-thirds of the dissolved oxygen. In the HOT-LoCO trial the sham arm at a similar pressure improved as much as the oxygen arm, so this dose may behave like sham.

Raises pressure to 1.3-1.5 ATA while the user breathes concentrated oxygen through a mask, delivering roughly a third to two-thirds of the dissolved oxygen of clinical HBOT depending on mask seal.

Patients who cannot travel and can afford rental, accepting that the dose is untested and may sit below a threshold.

Ear and sinus pain, confinement, noise. Fire risk with oxygen: no electronics, heat packs or synthetic fabrics inside, and keep the concentrator outside the chamber. Getting in and out takes effort.

Use a non-rebreather mask with reservoir and check the oxygen at the mask; stop if overnight heart rate or HRV worsens.

unregulated · ~£400-1,000/month rental; £5,000+ to buy

In rodent and cell studies it acts on vessel lining and vessel growth (step 4) and tissue repair (step 6). Human data are three small studies, so the model rationale is untested in people.

A synthetic peptide that in animals activates endothelial nitric oxide and VEGFR2 signalling, speeds vessel growth around injuries, and protects the gut lining.

No one outside a study. Interesting mechanism for the vessel links, but there is no human efficacy data and products are of unknown content.

Unregulated grey-market products of unknown purity and dose; contamination risk with injectables. Theoretical concern that pro-angiogenic effects could feed tumours. In the US it is not legal to compound (removed from FDA Category 2 in April 2026, not added to the approved list).

unregulated · ~£30-80/month grey market

Aimed at fibrin amyloid deposits blocking capillaries (step 4). The microclot finding itself is contested and anticoagulation carries bleeding risk; trials have not shown benefit.

Antiplatelet and anticoagulant drugs, or filtering the blood, aimed at dissolving amyloid fibrin 'microclots' seen under the microscope.

No one outside trials. The microclot test is not standardised, and haemostasis experts judge these treatments not justified (Hunt 2024).

Serious bleeding, including brain bleeds, with combined antithrombotics; line complications with apheresis. Statins and simple antiplatelets have no trial in ME/CFS.

unregulated · Apheresis abroad ~€1,000-2,000/session

Feeds butyrate-producing gut bacteria, which are depleted early in the illness and may help regulate immune tone (step 1). The model treats the gut as a modifier, not a driver.

Feeds the bacteria that make butyrate and other anti-inflammatory metabolites.

Anyone without IBS flares from fibre; increase slowly.

Bloating and wind, especially with IBS (about 37% of ME/CFS patients); low-FODMAP fibres such as oats, kiwi and psyllium are gentler.

Avoid if weight is falling or eating is effortful: calories come first.

OTC

Aimed at gut-driven immune tone (step 1). Supplemented strains do not colonise, and the model gives the gut a minor role.

Adds live bacteria, usually Lactobacillus and Bifidobacterium strains, which do not colonise long term.

People who want to try, mainly for gut symptoms; not expected to change ME/CFS itself.

Bloating; rarely bloodstream infection in people with central lines or severe immune suppression (relevant for those with ports for IV saline).

Avoid with a central line or port.

OTC · £10-30/month

Several placebo RCTs have not reported: ME/CFS (NCT04741841, n=74, completed 2021); long COVID (NCT05975034, n=240 planned, status unknown; NCT05813899, stopped early); PURE-LC (NCT06643299, n=194) completed July 2026

Tests whether the gut drives immune priming (step 1). The only randomised trial was small and null.

Transfers stool bacteria from a screened donor by colonoscopy or capsules.

Only within trials.

Colonoscopy risks and preparation (hard for severe patients), transmission of infections or resistant bacteria from unscreened donors. DIY FMT is dangerous.

Bowel preparation alone can cause a crash.

trial

Comeback Study, Norway (n=80, NCT03691987): protocol published 2024, results not yet reported

Replaces a hormone only when testing shows it is low (step 2). The model gives no reason to expect benefit from replacement in people with normal levels.

Replaces a hormone that is low.

Only when blood tests show a true deficiency (raised TSH or low free T4; low morning testosterone on repeat testing with symptoms). Worth a GP test, not empirical treatment.

Excess thyroid hormone causes fast heart rate, atrial fibrillation and bone loss, and worsens POTS. Testosterone: clots, high red cell count, fertility suppression, acne; not in women outside specialist care.

prescription

Would correct a vasopressin deficit if one exists (step 2), expanding volume (step 3). Only worth considering after stimulated copeptin shows a deficit; the risk is low sodium.

A synthetic form of antidiuretic hormone (ADH) that makes the kidneys hold water, expanding volume.

People who pass large volumes of dilute urine, especially at night, with high plasma and low urine osmolality on paired tests or low copeptin.

Low blood sodium, which can cause confusion, seizures and death; needs sodium checks within days of starting and after dose changes. Do not combine with high water intake.

Home blood tests are needed for safe sodium monitoring.

specialist · NHS prescription charge

Would correct mildly low cortisol (step 2) and suppress inflammation (step 6). Benefits in trials were small and short, with adrenal suppression; the model suggests testing the pituitary-adrenal axis first.

Replaces or boosts the stress hormone cortisol, which is mildly blunted in ME/CFS, and suppresses inflammation.

Only for proven adrenal insufficiency (a separate diagnosis needing a short Synacthen test), which is more common than expected in post-viral fatigue.

Adrenal suppression, bone loss (spine density -2% in 12 weeks at 25-35 mg), weight gain, high blood sugar, mood change, infections. Short courses of prednisolone sometimes give a temporary lift followed by a dip.

Ask for a morning cortisol and Synacthen test if there is weight loss, low sodium, salt craving or collapse.

prescription

Two long-COVID placebo RCTs not yet reported: prednisolone plus B vitamins (PreVitaCOV, NCT05638633, n=321, completed Jan 2025) and methylprednisolone (NCT05986422, stopped early, n=96)

Can help with coping and the sleep and arousal side of the illness (step 8). The model does not predict that it changes the underlying steps.

NICE 2021 allows CBT only to help people cope with a long-term illness, not as a treatment that reverses ME/CFS.

People who want help with sleep, anxiety, grief or adjusting to illness, from a therapist who accepts the illness is physical.

Programmes that frame ME/CFS as fear of activity or push increasing activity can cause harm; check the approach before starting. Sessions are themselves exertion.

Short sessions by phone or message, at home, if at all.

prescription · Free on NHS referral; £60-120/session privately

Breathing retraining could reduce over-breathing on standing (step 2 into 5). Exercise components risk driving steps 5 to 7 in people with post-exertional malaise, so the model favours programmes built around pacing.

Supervised programmes of breathing, light exercise, symptom management and psychological support, delivered in groups, online or in clinic.

Long COVID without post-exertional malaise. People with PEM or ME/CFS should ask whether the programme is symptom-led and stops when symptoms rise; fixed activity targets are graded exercise under another name.

PEM and lasting worsening if the programme pushes through symptoms. The trials were unblinded and most did not screen for or measure PEM.

Not suitable for severe patients beyond gentle, symptom-led input at home.

specialist · Free on NHS referral to a long COVID service

GLP-1 receptors on immune cells, vessels and brain could act on steps 1, 4 and 7. Against that, these drugs raise heart rate, promote sodium loss and reduce thirst, which works against step 3.

GLP-1 receptors sit on immune cells, vessel lining and brain; proposed anti-inflammatory and endothelial effects on top of weight loss.

Overweight patients with stable weight could ask about the Charité study; otherwise wait for controlled results.

Nausea and vomiting, muscle loss (up to ~40% of weight lost), higher heart rate, reduced thirst and sodium loss (the opposite of volume treatment), gallstones, pancreatitis. Grey-market retatrutide is unlicensed.

Not for severe patients who are losing weight or muscle.

prescription · ~£120-250/month private

Charité semaglutide observational study (planned 2026); RECOVER and Scripps long-COVID trials

Aims to reset sympathetic over-activity (step 8) and its effect on brain control (step 2). Case series only.

Local anaesthetic injected next to the sympathetic nerve chain in the neck, meant to reset sympathetic over-activity.

Only within trials; the most-cited series comes from the clinic selling the procedure.

Hoarse voice, droopy eyelid, difficulty swallowing for hours; rarely injection into a blood vessel causing seizures, or bleeding in the neck.

specialist · ~£800-2,000 per block privately

Two placebo-controlled long-COVID RCTs recruiting: STAR-CO (NCT07374562, n=40, single-blind) and CeASE cervicothoracic sympathetic block vs saline (NCT07468604, n=78)

No model rationale. It is said to contain very highly diluted antibodies to a brain protein, and there is no accepted way for doses this low to act.

A Russian tablet said to contain highly diluted antibodies to a brain protein (S100); there is no accepted way for doses this low to act.

Not available in the UK; wait for independent replication.

Few reported in the trial.

unregulated

Aimed at cognitive symptoms through brain serotonin receptors (step 2 and step 7). A long COVID trial was null.

An antidepressant with extra serotonin receptor effects, licensed for depression and claimed to help thinking speed.

Not for brain fog alone. Depression alongside long COVID or ME/CFS is a separate reason a doctor may prescribe it.

Nausea (common), itch, sleep changes; serotonin syndrome risk with other serotonergic drugs.

prescription

Proposed to calm brain inflammation (step 7). A long COVID trial was null.

Very low doses of lithium, far below psychiatric doses, are proposed to calm brain inflammation.

Not supported; the only placebo trial was null.

Low at these doses, but higher doses need blood monitoring for kidney and thyroid effects. Over-the-counter lithium orotate is unregulated.

unregulated

Targets brain fog directly. The model places brain fog downstream of brain blood flow and the sickness response (steps 5 and 7), so it predicts little benefit without addressing them.

Computer-based brain training and coaching, sometimes with weak electrical current applied to the scalp (tDCS), aimed at brain fog.

Not expected to fix brain fog. Practical strategies (notes, reminders, doing hard tasks at your best time) are a different, low-cost thing.

Cognitive effort can trigger PEM. tDCS can cause skin burns and headache; home kits are unregulated.

Screen time and concentration are exertion for severe patients.

OTC

Aimed at inflammation and scarring after injury (step 6). Trials in long COVID were null.

LAU-7b is an oral drug aimed at lipid-driven inflammation; deupirfenidone is an anti-scarring drug developed for lung fibrosis.

No one; neither is available and both trials were null.

Not relevant outside trials.

trial

Fixed increases in exertion repeatedly drive steps 5 to 7, the sequence the model holds responsible for the crash, so the model predicts harm in people with post-exertional malaise.

Fixed, stepwise increases in exercise, based on the now largely refuted theory that deconditioning and fear of activity maintain the illness.

No one with post-exertional malaise. Gentle, symptom-led movement within the energy envelope is a different thing.

Worsening that can be prolonged or permanent, including moving from moderate to severe. A 2026 Charité rehab study left 42 of 97 worse even with pacing elements.

Severe patients have the most to lose; refuse any programme that sets activity targets regardless of symptoms.

specialist