PEMF therapy

A PEMF device is a coil that produces brief magnetic pulses. Lie on the mat or strap on the pad, and a changing magnetic field passes through you. Because the field changes, it induces a small electric field in the tissue it passes through. That is the whole physical basis, and it is not in dispute.

What is in dispute is what that does, and whether the machines being sold produce enough of it in the right place.

People ask about this for good reasons. A short paper from 1999 describing three people with ALS/MND who appeared to get better still circulates: old enough to feel like buried knowledge, specific enough to sound technical, short enough to read. The biology is not far-fetched either, since ALS/MND involves mitochondrial dysfunction, oxidative stress, inflammation and breakdown of transport along nerve fibres, and magnetic fields have been reported to affect oxidative and inflammatory processes in cultured cells, and organelle transport in one study of cells from people with a FUS mutation. What that amounts to is set out later, including the findings that run the other way. There is an active research programme with public funding. And unlike a drug in trials, a mat can be ordered this afternoon. A plausible mechanism, an encouraging old report, a live laboratory programme and a purchase button is a combination worth setting out carefully rather than dismissing.

The confusion that matters most is between technologies. A search for evidence about PEMF returns trials of hospital stimulators a thousand times stronger, trials of magnets that do not pulse at all, and studies of the background magnetic fields around power lines. Sellers quote all of it. What follows sorts it into four families, so that when a study appears later in this document you can see which family it belongs to and whether that family is the one under investigation.

Two rules make the rest of this document readable.

Absence of evidence is not evidence of absence. Nothing here shows that PEMF does not work in ALS/MND. Nothing here supports a claim that it does. Those are different statements and both are true. This is stated once, here, and not repeated.

Symptom relief and disease modification are different claims. Evidence that something eases a sore shoulder says nothing about whether it slows motor neuron loss. Pain and cramp are real problems in ALS/MND, so the first is not irrelevant. It is simply not the same question.

Explained: what this word means

Under a strict definition of PEMF, meaning the low-frequency, low-intensity fields a consumer device produces, the direct human evidence in ALS/MND is one uncontrolled report of three people from 1999. These searches found no controlled trial, and no registered trial in the registries searched. Twenty-seven years on, no attempt to repeat it was found.

Randomised studies of other magnetic technologies have been conducted in ALS/MND, including small rTMS and cTBS studies beginning in 2006, neuromuscular magnetic stimulation in 2019, static magnetic stimulation in 2024 and prefrontal rTMS in 2025. Their aims, protocols and results differ, and none tested an ordinary consumer PEMF mat or pad. They are set out here because each is regularly cited as though it were about PEMF, and because an investigation into magnetic therapy that omitted the randomised work in the disease would be incomplete.

The early transcranial studies, and what happened to them

The first controlled result was positive. A 2006 double-blind, placebo-controlled trial randomised 20 people with definite ALS to continuous theta-burst stimulation of the motor cortex or sham, five days a month for six months. Fifteen completed, and the active group showed a modest but significant slowing of the deterioration rate. The authors were careful about it themselves, writing that they could not be sure the effect was attributable to the stimulation given the small number studied.

A 2008 controlled pilot then gave 5 Hz rTMS to 10 people for two weeks and reported differences in quality of life and in dynamometer measures against sham. Those changes were transitory and were no longer significant two weeks after stimulation stopped.

The 2006 signal did not survive testing. The same group ran a longer version in 2009: cTBS again, five days monthly, this time for a full year instead of six months, with twenty people randomised again. It found no significant difference in ALSFRS-R decline between real and sham, and concluded that a larger confirmatory trial did not seem justified, at least at an advanced stage of the disease. It was well tolerated.

A Cochrane review pooled the randomised rTMS work in 2013 and found three randomised placebo-controlled trials with 50 participants in total, all of poor methodological quality and too dissimilar to combine. Its conclusion was that there was insufficient evidence to draw conclusions about either efficacy or safety.

This is the ordinary shape of a small early finding: one encouraging result, then a better test that does not reproduce it. It is worth stating plainly because the 2006 paper still circulates on its own.

The one that showed something

Neuromuscular magnetic stimulation was tested in 22 people with spinal-onset ALS/MND, each with one arm treated and the other shammed in the same session, so the disease's own course cannot explain a difference between the arms. It is one of the strongest designs in the corpus for a localised muscle outcome.

What it found needs its outcomes kept apart. The primary outcome was muscle strength in the treated arm, assessed two ways. On the MRC scale, a manual grading by the examiner, there was no interaction between MRC score and treatment group at any time point. On the handgrip dynamometer, an instrument reading, there was a significant interaction with treatment group at the end of stimulation. So the two measures of its own primary outcome disagreed, and the headline improvement rests on one of them. The named secondary outcomes were CMAP, acetylcholine-receptor currents, gene expression and muscle histology.

Three things about it matter.

It used a hospital stimulator at full output. The paper states no field strength in tesla anywhere. Peak field at the coil face for stimulators of this class is on the order of one to two tesla by manufacturer specification, not from this trial. A consumer mat runs at a few millitesla. That is not a difference of degree.

The effect faded within a month of stopping. The transient improvement in grip strength is more consistent with a short-lived muscle or conditioning effect than with durable reinnervation, although this small trial could not determine the mechanism. That is an interpretation of the pattern, not something the study demonstrated.

CMAP amplitude did not change. CMAP is the summed electrical response of a muscle to stimulation of its nerve. The authors read the absence of change as evidence against reinnervation, and that is a reasonable reading, but CMAP is not a direct count of motor units and a flat CMAP does not settle the question.

The tissue findings carry small denominators that are easy to lose. Of the 22 randomised, 15 underwent needle biopsy; the amount of muscle recovered was not always enough for every planned analysis. The histological analyses used 7 participants and the fast-fibre morphometry used 4. The finding that stimulation counteracted atrophy in fast-twitch fibres rests on those four. In the same four people, slow-twitch fibres were slightly but significantly smaller on the treated side, which the paper reports without explaining. The authors themselves judged the strength gain unlikely to reflect reinnervation. No replication was found in the seven years since.

The two that missed

A phase 2 trial of transcranial static magnetic stimulation in 40 people with ALS/MND missed its primary endpoint: monthly ALSFRS-R decline was 0.90 with real stimulation against 0.94 with sham. A secondary survival figure at eighteen months favoured real stimulation, and the authors' own conclusion is that this supports evaluation in larger and longer studies rather than that the treatment works. It does not share PEMF's pulsing, waveform or repetition frequency.

The 2009 cTBS trial described above is the other. It is counted here rather than twice.

A trial about a different question

A 2025 randomised sham-controlled trial gave 10 Hz rTMS to the prefrontal cortex of 80 people with ALS/MND and cognitive impairment for four weeks. Cognitive scores and caregiver burden improved over the following months, and there was no improvement in ALS/MND severity or in plasma neurofilament.

That is worth knowing and easy to misread. It is evidence about repetitive TMS for one symptom domain in a selected subgroup. It is not evidence about consumer PEMF, and it is not evidence about disease progression.

So: across the randomised magnetic-stimulation work in ALS/MND, the one early positive signal was not reproduced, the trials that followed missed their primary endpoints, the 2019 study moved one of its two primary strength measures and not the other, and the 2025 study improved cognition without changing the disease. None of them used a consumer PEMF device.

Explained: what this word means

Almost everything said about PEMF and ALS/MND traces back to three pages published in 1999: a conference paper by André Bellossi and Richard Berget, in a congress volume from the 2nd World Congress for Electricity and Magnetism in Biology and Medicine, held in Bologna in June 1997.

The full text has been obtained and read. That matters, because for years the only widely circulating version was a transcription posted to a patient forum, and the parameters were doubted because of it. The transcription turns out to have been accurate.

The paper is real and its lead author was a serious researcher. André Bellossi published 85 papers indexed in PubMed between 1969 and 2000 from the Biophysics Laboratory at Rennes, 36 of them on magnetic fields. His own animal work used the same apparatus at 6 mT and 12 Hz, and he ran the same 12 Hz repetition rate in a separate study at 9 mT.

What the protocol was

12 pulses a second, carried on an 85 kHz oscillation, 6 millitesla at the surface of two 12 cm discs, applied to the ankles, groins and spine for two hours, three times a week. The apparatus was a Magnobiopulse. It was discontinued in 1998. Two hours of treatment three times a week would also be a considerable practical undertaking.

The carrier is the parameter that decides whether a modern device is delivering the 1999 exposure at all. Each of those 12 pulses a second was not a plain pulse. It contained a fast 85 kHz oscillation inside it, the way a radio signal carries sound. A device whose specification states only a repetition rate may or may not carry anything inside each pulse, and neither device compared below says. Whether the carrier mattered is unknown, and no test of it was found in these searches, but a device not shown to carry one is not shown to deliver the 1999 exposure, whatever its frequency dial says.

What the paper does and does not show

Three people, uncontrolled, unblinded, with no diagnostic criteria stated, no validated outcome measure, outcomes written down against session numbers rather than dates, no follow-up after the last session, and everyone also doing prescribed rest and home exercise at the same time.

That is not a criticism of the authors, who described what they saw in a conference volume. It is a description of what three uncontrolled narratives can support, which is a reason to run a trial and nothing more. No replication has been found in the twenty-seven years since.

Explained: what this word means

This is the question most likely to be answered by mixing up two populations, so it is split in two.

Symptoms of ALS/MND

These searches found no trial testing PEMF against any ALS/MND symptom. Not cramp, not spasticity, not fatigue, not weakness, not speech or swallowing. Spasticity has been studied with magnetic stimulation in other conditions: a 2022 meta-analysis of eight trials of hospital-strength peripheral magnetic stimulation in spastic paralysis, and a twelve-child sham-controlled PEMF pilot in spastic cerebral palsy that detected no side effects. Neither is about ALS/MND. No study was found of whether PEMF can provoke spasticity, a question raised in community discussion and not found investigated in these searches.

Spasticity and cramp are among the most burdensome symptoms in ALS/MND. That there is nothing here is a finding, and it is the reason the row above reads "not adequately studied" rather than "no benefit".

Musculoskeletal pain, in people who do not have ALS/MND

Here there is something, and it is worth reading closely because it is where the most substantial symptom evidence outside ALS/MND sits. The evidence in other conditions is mixed and condition specific: pooled randomised trials disagree about which outcome moves, and reviews of the same condition disagree with one another.

In knee osteoarthritis, a meta-analysis found that PEMF improved physical function but showed no advantage on pain or stiffness on three pain and stiffness measures and the WOMAC total score. A 2026 meta-analysis of nine trials in 457 people with the same condition reached a different conclusion: objective functional measures did not improve, pain improved only briefly, and the authors doubt the gains reach clinical meaningfulness. In low back pain, a meta-analysis of 14 trials and 618 people found the opposite pattern: pain fell against placebo, and in chronic low back pain specifically, while physical function did not improve and acute low back pain showed no significant effect.

So one knee review moves function and not pain, a later one finds neither convincingly, and the back-pain review moves pain and not function. Three limitations apply. Treatment is local, applied at the site that hurts rather than to the whole body. Follow-up is largely confined to the treatment course, so "short-term" is what the evidence covers and durability is not established. And heterogeneity varies sharply by outcome: I-squared reached 86% for physical function while the pooled pain estimate sat at 31%.

Two further 2026 reviews pull in opposite directions. The one review of a consumer whole-body mat found here reports that only one of four sham-controlled trials showed it superior to sham, at an intensity of tens of microtesla, far below the devices compared here. A review of neuropathic pain found a pooled effect whose significance disappeared after adjusting for missing studies, with no significant reduction in peripheral neuropathy. Against those, a review of orthopaedic pain found that most included trials reported less pain and better function, though it mixes low-intensity PEMF with hospital-strength peripheral stimulation.

The most relevant symptom evidence from another neurological disease comes from multiple sclerosis. A 2025 meta-analysis of 7 randomised trials involving 327 participants found a small but statistically significant reduction in fatigue and no effect on quality of life or depression. The individual trials do not agree. A 2022 placebo-controlled trial of a whole-body PEMF mat at consumer intensity in 44 people found no difference from placebo on fatigue, walking, depression or quality of life. One 2003 crossover trial reported significantly greater improvement in fatigue and quality of life on the active device, while a 2005 trial found its fatigue reduction was not statistically significant and its authors did not recommend the therapy. Whether those trials sit inside the 2025 pool was not established here. That is a small pooled effect, on fatigue, in a different disease.

In Parkinson's disease the picture is weaker. One randomised trial dominates: 97 people given transcranial PEMF at home for eight weeks. It missed its primary endpoint at p = 0.064 and reports a subgroup at p = 0.049. A separate analysis of the same participants found a tremor effect only in a subgroup. A later study from the same group reported improvements in movement speed but was not randomised and used an untreated control group.

One human result is worth knowing because it is the only one of its kind found here. A 225-person randomised trial of PEMF in diabetic neuropathy found no difference from sham in pain intensity on the Neuropathy Pain Scale or the visual analogue scale, with one of its three pain measures, the global impression of change, favouring PEMF at 44% versus 31%, which the authors called a trend. In a 27-person biopsy subset, 29% of treated participants versus none in the sham group showed an increase of at least half a standard deviation in distal-leg epidermal nerve fibre density. That is a human nerve-regeneration signal. It concerns sensory fibres in the skin rather than motor nerves, and it rests on a small subset of a trial that was otherwise negative.

None of this is evidence about ALS/MND, and a biomarker moving is not a symptom improving. A muscle that looks less denervated under a microscope is not a person who can hold a cup. Two further claims circulate and have no support here: no trial found in this review tested whether PEMF prevents ALS/MND, and no evidence establishes an effective session length for ALS/MND, since the only ALS/MND protocol found, two hours three times a week, was uncontrolled.

Explained: what this word means

There is a real biological rationale here, and it is worth setting out honestly because it is the part of this subject most easily oversold and most easily dismissed.

Several processes that fail in ALS/MND can be changed by magnetic fields in laboratory systems. That is true, it is interesting, and it is why researchers are working on this. It is also a long way from a person.

One finding stands out and needs its label kept on. Motor neurons grown from the skin cells of people carrying FUS mutations recovered the movement of mitochondria and lysosomes along the axon under magnetic field stimulation at 10 Hz; a 2 Hz arm in the same study did nothing. FUS is one genetic subtype among several. This is the closest laboratory finding to the disease in the whole document, and it has been shown in cells carrying that one mutation, not across ALS/MND. The same Dresden group is now building a prototype device on the strength of this result, so the laboratory finding and the research programme described later are one line of evidence, not two.

What follows is the mechanistic case in full, including the parts that run the other way. Each pathway ends with the step nobody has taken, and that is the part to read if you read only one.

Explained: what this word means

Short trials in other conditions have reported few adverse events, and none of those found here was designed to measure safety. A 2026 review of orthopaedic pain trials reported no serious adverse events; that literature is largely musculoskeletal, largely short term, and largely in people without a neurodegenerative condition. ALS/MND-specific and long-term safety evidence is limited. So this is a narrow observation rather than a general assurance, and it does not transfer to ALS/MND.

Check these before starting

Implanted electronic devices. Pacemakers, implanted defibrillators, cochlear implants, neurostimulators and intrathecal pumps. A pulsed magnetic field can interfere with them. Whether it applies to a given product, and at what distance, is a question for that product's manual and its maker. If you have an implanted electronic device, do not proceed without asking your clinician and the manufacturer.

Anything near the head or upper neck. The 1999 protocol did not treat the head; it used the ankles, groins and spine. Reasoning from it to cranial application goes beyond what its own source did, which is the reason for raising it here. There is also one small observation from a different population: in a seven-person post-concussion pilot, two of seven worsened and five improved. Seven people, after brain injury, not ALS/MND. That does not establish an ALS/MND-specific risk and it is not a contraindication; it is the only report of worsening under transcranial PEMF found in these searches. Larger transcranial PEMF trials exist, such as a 97-person randomised trial in Parkinson's disease. None found here was designed to measure safety, though that trial described its adverse events as benign, mild and transient, with no difference between groups. That is tolerability in another disease, not reassurance about ALS/MND. A reason to ask rather than to assume in either direction.

Read your own device's manual, because warnings vary. Biomag's published contraindication list has eighteen items, given without rationale or severity grading. Several matter to this readership: it includes myasthenia gravis, seizure disorders and pacemakers, and the maker's own page adds that the list is indicative and that applications should be discontinued in case of unexpected reactions. For one of the devices looked at here, the product page reviewed did not state contraindications. Its other manuals and supporting documentation were not obtained, so nothing follows about what they contain. A manufacturer's list is a manufacturer's list. It is not a class-wide medical rule, and Compass does not turn it into one.

If you feel worse, that is a reason to stop and ask

Consumer guidance often frames early discomfort as a sign that treatment is working: a detoxification response, a healing reaction, or pain moving as deeper problems surface. One widely read guide states that most adverse reactions are mild and temporary and can be managed by simply continuing the therapy. Compass does not repeat that advice. No test distinguishes a harmless adjustment from a real adverse effect or from the disease progressing, so an interpretation that always points toward continuing is not a safe rule. In a progressive condition it is also the interpretation most likely to delay a conversation that should happen.

Three claims from that guidance are not repeated here, because no independent support for them was found: a 5 percent adverse-reaction rate offered as clinical experience rather than measurement; the claim that PEMF acts on platelets comparably to aspirin, with a suggestion that someone taking an anticoagulant might lower their dose; and the claim that no peer-reviewed, substantial study shows PEMFs can damage the body. The mechanism table above lists cell findings in the harm direction at overlapping field strengths but under different exposures. Those do not show that anyone has been harmed. They do show the claim is stated too broadly.

If you already own one

Nothing here says stop, and nothing here says continue. Tell your clinic that you are using it, particularly before any change to medication and before joining a trial, since trials ask about other treatments. Do not judge its effect from a period of stability: in one dataset of 3,132 people, roughly a quarter had no net ALSFRS-R decline over one six-month interval, whatever they were doing. If anything gets worse, stop and ask.

What people report, on safety

One community report describes external electromagnetic muscle stimulation provoking severe leg spasticity, after which the person stopped. That is a different technology and it is one account. It is recorded because a targeted search for this question returned nothing at all.

The occupational exposure question

This is raised often and is weaker than it sounds. A 2021 meta-analysis of 27 publications found a small association between occupational extremely-low-frequency magnetic field exposure and ALS/MND risk, relative risk 1.20 (1.05 to 1.38), with substantial heterogeneity and acknowledged publication bias. Against that: a residential dose-response meta-analysis was null; a 2026 Swiss cohort of 3.6 million adults found no association between residential exposure and ALS/MND; a cohort of 37,986 UK electricity workers followed 1987 to 2018 found motor neurone disease mortality at national rates, though the same paper notes positive associations for some categories of recent exposure; a 2025 systematic review of animal studies found too few studies to draw any conclusion about causation in motor neuron disease and no effect of magnetic-field treatment on progression in motor neuron disease models; and exposure did not modify age of onset or progression in 1,098 people with ALS/MND.

The honest reading is that this signal should be disclosed and not presented as an established hazard. It is also a question about chronic workplace exposure, not therapeutic use.

Explained: what this word means

This section is not evidence, and nothing in it changes any grade above. It is here because what people report matters, because a safety signal can appear in a report before it appears in a study, and because community discussion is where most people first meet this subject.

A read-only pass over one closed ALS/MND Facebook group and public forums (other groups, Reddit, Inspire and ALSForums) collected a few dozen records on 1 September 2026. Four things came out of it.

The reports were mixed. Some people described improvement; more described nothing, or continuing to progress. Neither number means anything on its own. Nobody was randomised or followed up, and people post selectively, so unusually good and unusually bad experiences are both over-represented.

The positive reports were confounded. The two clearest accounts of improvement each involved several treatments at once. In one, a second body area was being treated with a different device alongside a non-medical explanation of the recovery. In the other, a package of several simultaneous therapies was delivered by a paid provider. Neither can be attributed to a magnetic field, and none of the accounts included an objective measure or durable follow-up.

Much of the conversation was about something else. A substantial share of the records concerned Rife or Spooky2 configurations rather than consumer PEMF mats or pads. These are not equivalent to a consumer PEMF mat or pad, and what a given record describes depends on the device and accessory used, which the record often did not say. All of them were excluded from the PEMF assessment. ALSUntangled, the clinician group that reviews alternative treatments in ALS/MND, has published a review of the Rife machine; that review, not this document, is where the question belongs.

Almost nobody said what they actually did. The records reviewed generally did not report waveform, and most stated neither a frequency nor an intensity. That is not a criticism of anyone writing in a support group. It is the reason none of these accounts can be compared with a research exposure, with a device specification, or with each other.

How to read any single account

This is a condition that fluctuates and plateaus. In one dataset of 3,132 people, roughly a quarter had no net ALSFRS-R decline over one six-month interval. That is a measured score trajectory over one window. It does not demonstrate biological disease stability, and it is not remission. A period of stability after starting something is expected some of the time, whether or not the something did anything. Reports made after one day or three days are very early observations, too early to separate a treatment effect from the ordinary fluctuation above.

Community reports cannot establish that PEMF works, and they cannot establish that it does not. No effectiveness rate is calculated from them here, and none should be.

Nothing from the closed group is quoted on this page. No names, profile links, photographs or dates of diagnosis appear anywhere. Practitioner testimonials and manufacturer material are counted separately from what patients and carers said, because they are a different kind of object.

Explained: what this word means

The programme most directly connected to this question

A research group at the Helmholtz-Zentrum Dresden-Rossendorf published the FUS-ALS motor neuron work and holds roughly 2 million euro to build a magnetic pulse therapy prototype, running for three years from July 2024. This is the directly connected translational programme these searches identified. Its team anticipates a first clinical trial in 2027, dependent on further experiments and funding.

The stages are worth keeping apart, and all of it is as reported by HZDR rather than independently verified. Completed and peer-reviewed: the FUS-ALS motor neuron work above, by the same group. Reported by HZDR as completed: a mouse study with Hannover Medical School in which, in the team's own unpublished statement, progression slowed and lifespan lengthened by about ten percent. Its methods and results cannot yet be independently assessed; publication is expected by the end of 2026. Reported as submitted: an ethics application to TU Dresden, with the Department of Neurology at the Carl Gustav Carus University Hospital in Dresden as clinical partner, and a transcriptome manuscript in July 2026. Not found: any trial registration.

It is not evidence that magnetic stimulation helps anyone with ALS/MND. HZDR reports no human exposed so far, and no registration for this programme was found in the registries searched (ClinicalTrials.gov, DRKS, CTIS and EudraCT, on 2 September 2026). An ethics submission is not a registration, and neither is a grant. The device's field strength, frequency, waveform, target and schedule are all unpublished. The cell work it builds on used about 10 millitesla at 10 Hz for more than seven hours a session around a culture dish. That figure overlaps the consumer range on paper, and the overlap establishes nothing: waveform, carrier, coil geometry, session length, tissue depth and the induced field can all differ, and a mat on a bed for twenty minutes is not that exposure.

Adjacent magnetic stimulation work, which is not about PEMF

Several related studies are registered, and none of those found uses a consumer PEMF device. Four representative records follow, statuses last checked on 2 September 2026: three use repetitive TMS and one a static magnet, all hospital technologies applied to the head. None will directly answer whether consumer PEMF devices work, although they may provide indirect information about magnetic stimulation biology or safety.

• QuARTS-ALS (NCT05983211), at Sunnybrook in Toronto, is listed as recruiting for its second stage: a safety and feasibility study of theta-burst rTMS in an estimated 15 people. Its registry record was last updated in September 2024.
• iFIRST-ALS (NCT06819358), at Peking University Third Hospital, is listed as not yet recruiting. It plans imaging-guided rTMS for postural and gait problems in an estimated 45 people. Its registry record was last updated in February 2025.
• tSMS and biomarkers in ALS/MND (NCT06834269), at Campus Bio-Medico University in Rome, is listed as recruiting an estimated 60 people for transcranial static magnetic stimulation, the larger study the 2024 trial's authors called for. Its registry record was last updated in February 2025.
• Brain stimulation and exercise in MND (NCT07067229), at Chulalongkorn University in Bangkok, is listed as recruiting an estimated 100 people for personalised, standard or sham rTMS combined with exercise. Its registry record was last updated in September 2025.

A registry status is only as current as the sponsor's last update. The two older records predate this check by more than a year, and one lists a primary completion date that has passed without an update.

And a different technology entirely

Focused ultrasound comes up in the same conversations because both are described as non-invasive brain treatments. They are not related. Details are in the panel; the short version is that ultrasound uses pressure waves rather than magnetic fields, and a four-person study opened the blood-brain barrier in ALS/MND temporarily and reversibly without testing whether it changed the disease. A registered ALS/MND pilot of transcranial pulse stimulation, which is ultrasound despite the name, will be found by anyone searching for pulsed therapy; it is not PEMF and its results will not transfer.

What would change this assessment

In short: one proper trial would change this. Nothing else on the list would, on its own.

In rough order of how much each would move it.

1. Any controlled trial of PEMF in ALS/MND, at any endpoint. The largest single lever, and the one thing whose absence defines the current position.
2. A replicated, longer neuromuscular magnetic stimulation trial, three months or more, with motor unit number and single-fibre jitter as co-primary endpoints. This would test the reinnervation question directly instead of inferring it from CMAP, which is what the 2019 trial measured. It would say nothing about PEMF mats.
3. Any demonstration of collateral sprouting under magnetic exposure. No such demonstration was found in this search.
4. A living-animal test of the FUS-ALS transport finding. The 1995 rat study found no change in a healthy-nerve assay at 2 Hz and the 2023 cell study found rescue at 10 Hz; they are not comparable, so the animal question is open rather than answered.
5. A head-to-head comparison of carrier-based against carrier-free PEMF, which would settle whether the 85 kHz carrier is load-bearing or incidental.
6. A therapeutic-protocol study in an ALS/MND animal model. The only long-term exposure study in SOD1 mice found here was an environmental design and was null.
7. Publication of the Dresden prototype's parameters, and of its completed mouse study.
8. Independent replication of the 75 Hz cellular work by a group with no manufacturer funding, and more decisively any application of it to a motor neuron model at all.
9. Manufacturer disclosure of carrier, waveform, rise time and depth-resolved intensity. This would not change the evidence. It would make the comparison above answerable.

What would not change it: another positive cell-culture paper, another testimonial, or a device that can be set to 12 Hz.

Explained: what this word means