PEMF mats and devices · what sham-controlled trials found
Plate 45
Field record
What a device emitted
23
Sources on the page
7
Goals in people
On this page. Counts, not a grade and not a ranking.
No goal on this page meets the rule for a stronger record. The rule is printed above the goals, and the count it read is printed under each one.
Read this first
Educational use only. This page is not medical advice. Field strengths, frequencies and session lengths are what a study ran, not a setting for you. Anyone with an implanted electronic device, a pregnancy or a fracture under treatment has a question for their own clinician that this page does not answer. Aaron Cuha is a coach, not a doctor.
Educational use only. This database summarizes published research and is not medical advice. Where a dose appears it is a record of what a trial or clinic ran, or what people describe doing, with its source attached. Nothing here is a recommendation for personal use.
PEMF means pulsed electromagnetic fields: a coil carries pulses of current and makes a changing magnetic field that passes into the body. Devices range from coils strapped over a joint, to handheld units, to whole-body mats sold for home wellness. Field strength in the studies here runs from microtesla on mats to 105 millitesla in one handheld device, a thousandfold range.
PEMF began in orthopaedics as bone stimulation for fractures that would not heal. A 2008 meta-analysis noted bone stimulation was then a 500 million dollar market in the United States. Most later trials are in osteoarthritis, pain and osteoporosis, with a few in depression and sleep.
It is not a static magnet, and not transcranial magnetic stimulation at the strengths used in psychiatry. A 2012 review of the whole-body mats found 11 randomised, sham-controlled trials in total, none large.
The proposed story is that pulsed fields nudge cells: bone-forming cells, repair cells and nerves. Laboratory work supports some of that. Whether the field from a given device reaches a given tissue at a dose that matters is the open question, and the trials answer it unevenly.
Fields stimulate bone cells
The Cochrane review of non-healing fractures opens with laboratory studies showing electromagnetic fields can stimulate new bone formation, which is why the clinical trials were run. Its own pooled clinical result was not significant.
Fields ease arthritic joints
Across 15 pooled placebo-controlled trials in osteoarthritis, PEMF improved pain, stiffness and function against placebo, and device settings did not change the result.
Shorter sessions did better
In a meta-analysis of 12 sham-controlled osteoarthritis trials, sessions of 30 minutes or less showed better effects than longer ones, which argues against more being better.
Where the mechanism story is weak
A systematic review of 11 sham-controlled trials of whole-body mats found sporadic positive effects that no independent trial confirmed, and judged the evidence insufficient.
In 55 people with treatment-resistant depression, five weeks of daily PEMF to the head did no better than sham on any measure. Both groups improved.
Cochrane found no clinically important benefit of PEMF for shoulder pain in small trials, on low or very low quality evidence, and no PEMF trial measured adverse events.
Every number here is what a study ran, named beside it. None is a setting for a reader. Two devices both labelled PEMF can differ a thousandfold in field strength, which is the main reason results do not transfer between them.
Each row is a record from a paper. It is not a plan for you.
| Goal | Who | What the trial ran | Source |
|---|---|---|---|
| Knee osteoarthritis | 57 people averaging 61.6 years | A handheld device at 4 to 12 Hz and 105 mT, twice a day for 5 minutes over 18 days, against a placebo device. | PMID 26562074 |
| Shoulder tendinitis | 29 people unresponsive to steroid injection | Active or placebo coils for 4 weeks, then all on active coils for 4 weeks, then 8 weeks off. | PMID 6143039 |
| Whole-body mats | 11 trials of 12 to 71 people | Whole-body mats at 3.4 to 200 microtesla, for up to 12 weeks, against sham. | PMID 21938735 |
| Pain syndrome | 30 people with complex regional pain syndrome | A whole-body device or sham added to 10 consecutive days of rehabilitation. | PMID 29985719 |
| Depression | 55 people with treatment-resistant depression | Five weeks of daily PEMF to the head, active or sham. | PMID 34333176 |
| Insomnia | 101 people with insomnia | 4 weeks of active or placebo impulse magnetic field therapy. | PMID 11697020 |
Stronger records first. Nulls and marketing-ahead goals stay on the same page. Every card is people only; animal work is in its own section below. Nothing here is ranked against the others, and no clinic or device is recommended.
How the tier is decided, so you can check it against the counts printed under each one. Stronger: a clinical guideline or a medical society's indication list recommends it in a named setting, or a pooled analysis of randomised trials favours pulsed electromagnetic fields and the source rated the certainty moderate. Thin: the pooled result was null or went against pulsed electromagnetic fields for the goal as named, the source rated it very low, or nothing was pooled at all. Mixed: everything else, including a result that moved one measure and not another.
Mixed, local, or early
15 randomised trials pooled; certainty not graded by the source; result mixed
Who was studied
People with osteoarthritis, mostly of the knee. A 2020 meta-analysis pooled 15 placebo-controlled randomised trials; a 2018 one 12 sham-controlled trials, ten in the knee; a 2019 one 8 trials and 421 people. A 2015 trial of 57 people used a named handheld device.
The 2020 analysis found improved pain (SMD 1.06), stiffness (0.37) and function (0.46). The 2018 analysis found knee pain reduced (SMD -0.54) and function improved, with sessions of 30 minutes or less doing better. The 2019 analysis found function improved but pain, stiffness and total score did not differ significantly from placebo.
Setting in the evidence
Local coils or handheld devices over the knee, against sham devices.
What this does not show
Which device works, or that a whole-body mat does. These were local devices over the joint, and the three analyses disagree on pain.
Mixed, local, or early
19 studies pooled, 1,303 people; certainty not graded by the source; result mixed
Who was studied
Postmenopausal women with osteoporosis in a 2022 meta-analysis of 19 studies and 1303 patients, funded by Chinese science grants.
PEMF added to standard drugs raised bone density and eased pain against drugs alone. Against placebo, PEMF did not change lumbar spine density. Adverse events did not differ.
Setting in the evidence
Clinical PEMF devices alongside or instead of standard osteoporosis drugs.
What this does not show
Fewer fractures. Bone density is a marker, and no fracture outcome is reported in the abstract.
Thin, null, or marketing ahead
4 randomised trials pooled, 125 people; certainty not graded by the source; result null
Who was studied
Adults with fractures of long bones. Cochrane pooled 4 sham-controlled trials and 125 people with fractures that had not healed; a 2008 meta-analysis pooled four trials on 106 such fractures; a 2024 review found 3 trials and 197 people with fresh fractures.
Cochrane found a non-significant effect on healing (risk ratio 1.96) and no pain reduction in two trials. The 2008 analysis found a non-significant relative risk of 1.76. None of the 2024 trials found a significant effect on healing fresh fractures.
Setting in the evidence
Clinical bone stimulator coils over the fracture, against sham.
What this does not show
That PEMF does nothing for bone. Cochrane called the evidence inconclusive and insufficient, not negative, and trials were small.
Thin, null, or marketing ahead
nothing pooled; certainty very low; result mixed
Who was studied
29 people with persistent shoulder tendinitis in a 1984 double-blind trial; Cochrane reviews of electrotherapy for shoulder pain and of 20 small trials and 1239 people with neck pain, which could not pool.
In 1984, active coils beat placebo for the first 4 weeks; once everyone was on active coils the groups no longer differed. Cochrane found no clinically important shoulder benefit from PEMF, and very low quality evidence that PEMF beat placebo for neck pain.
Setting in the evidence
Local coils over the shoulder or neck.
What this does not show
A reliable pain benefit. The certainty is very low.
Thin, null, or marketing ahead
nothing pooled; certainty not graded by the source; no pooled result
Who was studied
A 2012 systematic review found 11 randomised, sham-controlled, double-blind trials of whole-body mats, 12 to 71 people each, in knee and neck arthritis, fibromyalgia, pain, skin ulcers, multiple sclerosis fatigue and wellbeing. A 2020 pilot of a named device enrolled 30 people with a pain syndrome.
Some sporadic positive effects, none independently confirmed; the review judged the evidence insufficient and could not recommend therapeutic use. The pilot found more short-term pain and function improvement with the active device alongside rehabilitation.
Setting in the evidence
Whole-body mats at 3.4 to 200 microtesla.
What this does not show
Any wellness, energy, sleep or longevity benefit from lying on a mat. Long-term use has not been studied.
Thin, null, or marketing ahead
nothing pooled; certainty not graded by the source; no pooled result
Who was studied
101 people with insomnia in a 2001 trial; 55 people with treatment-resistant depression in a 2021 multicentre sham-controlled trial; 58 in an 8-week study with no control group, several of whose authors declare drug company grants.
In the insomnia trial, 70 percent of the active group reported substantial or complete relief against one placebo patient. In the 2021 trial, PEMF to the head was no better than sham. In the uncontrolled study, depression scores fell from 20.6 to 12.6.
Setting in the evidence
Clinical devices, including PEMF to the head.
What this does not show
A replicated benefit. The positive insomnia result stands alone in the search, and the only sham-controlled depression trial was null.
Thin, null, or marketing ahead
nothing pooled; certainty not graded by the source; no pooled result
Who was studied
Nobody, for these goals. The closest is the 2012 review of whole-body mats, whose trials measured pain, fatigue and wellbeing.
The review found sporadic effects without independent confirmation, and judged the evidence insufficient. The old rat cognition study is in the animal section.
Setting in the evidence
Whole-body mats.
What this does not show
Any lifespan, ageing or energy benefit in people.
Animal work gives PEMF a cell and tissue story, and two of its tendon studies were paid for by a bone stimulator maker. One frequency helped bone and another hurt it. They are here, apart from the people.
Rats, Achilles tendon tears
1 or 3 hours of PEMF a day produced no clear difference in healing, and some tendon properties were worse. Funded by a bone stimulator maker.
Rats, shoulder tendon repair
At 4 weeks the PEMF group had stiffer, stronger repaired tissue, with no difference in joint mechanics or walking. Supported by the same bone stimulator maker.
Diabetic rats, skin wounds
At 5 mT and 25 Hz for 1 hour a day, wounds closed faster on days 10 and 14, and the difference then disappeared.
Old rats, whole-body fields
In rats aged 30 to 32 months, six weeks of whole-body PEMF improved attention and spatial working memory.
Ageing mice, bone
PEMF increased new bone through sensory nerves; removing the nerves removed the effect.
Rats with ovaries removed
40 Hz reduced bone loss more than a standard drug regimen; 25 Hz made the osteoporosis worse.
Dogs with hip arthritis
20 dogs, 50 Hz for 30 minutes twice a week for 5 weeks or nothing. Pain scores improved; joint range did not differ. There was no sham.
Knee osteoarthritis is the best-studied use, and the three meta-analyses agree on function and disagree on pain. These are local devices over the joint, tested against sham devices.
Fracture healing, where PEMF started, pooled as not significant in a Cochrane review and a separate meta-analysis, and a 2024 review of fresh fractures found no healing effect.
The whole-body mats are the version most sold for home wellness and have the thinnest record: 11 small trials, sporadic effects, no independent confirmation, and no long-term safety data.
Device settings vary a thousandfold, and in one rat study changing the frequency reversed the result. A trial of one device says little about another.
The blocks above are the studies. The block below is a judgment about what to do with them. It has to say where it goes past the measurements.
Aaron Cuha’s position
This is opinion, not a finding. Everything else on this page is the evidence.
A local PEMF device over an arthritic knee probably helps function a little against sham, with conflicting results on pain. For fracture healing the pooled trials are not significant, and for whole-body mats sold for wellness, energy or longevity there is no confirmed effect in people.
The reasoning
Where this goes beyond the evidence
Saying the mats have no confirmed effect is a judgment about absence: the trials were small and short, and a larger one might find something. This position also generalises across devices that differ a thousandfold in field strength, which may hide a setting that works. If a well-run trial of a specific mat found a real benefit, this position would be too broad for that device.
Who this is not for
Anyone with a pacemaker or other implanted electronic device, which this page found no study on; anyone with a fracture being treated, whose stimulator is a question for the treating surgeon; anyone pregnant; and anyone who wants this page to replace a clinician.
Aaron Cuha, who built this archive and reads the studies behind the pages. He is a coach, not a doctor, and no medical credential is claimed for him anywhere on this site. More about who writes this
Anyone with an implanted device
No study in the search tested PEMF in people with pacemakers or implanted electronics.
Anyone with a fracture under treatment
Pooled trials did not show faster healing, and a treating surgeon owns that decision.
Anyone expecting a mat to replace treatment
A review of 11 whole-body mat trials found the evidence insufficient for any therapeutic use.
Anyone wanting a personal protocol
Field strengths, frequencies and session lengths here are records of what a study ran. They are not instructions for your body.
In sham-controlled osteoarthritis trials, adverse events did not differ from sham. In the osteoporosis meta-analysis they did not differ either.
The 2012 review of whole-body mats found no acute adverse effects reported and no study of long-term use.
Cochrane noted that no PEMF trial for shoulder pain measured adverse events at all.
In rats with ovaries removed, 40 Hz reduced bone loss and 25 Hz made osteoporosis progress. That is an animal result, listed here because it shows settings are not neutral.
This site teaches a blood panel, protein and training before gadgets. PEMF does not move that order. For knee arthritis, the training and weight evidence sits on the training pages.
Disclosed both ways: the owner of this archive lists a PEMF mat in his own protocol at /my-protocols. That is a personal choice recorded there, not a finding on this page.
Nothing on this page is for sale. No seller wrote it, and no device or brand is ranked; the two named devices appear only because a trial named them.
Four kinds of device sold as PEMF. No brand is ranked.
A mat with embedded coils, lain on at home, usually at microtesla strengths.
Evidence match. The 11 trials in the 2012 review, at 3.4 to 200 microtesla.
Mismatch. The review found the evidence insufficient, with no long-term safety data.
A coil strapped over a joint or limb.
Evidence match. Most of the osteoarthritis and shoulder trials.
Mismatch. Settings vary widely between devices, and the meta-analyses disagree on pain.
A hand-held coil held over a body part, at higher field strengths.
Evidence match. The 2015 knee trial at 105 mT.
Mismatch. One partly randomised trial of 57 people.
A prescription coil worn over a fracture for hours a day.
Evidence match. The fracture trials and the Cochrane review.
Mismatch. Pooled healing results were not significant.
Floor habits and the neighbouring pages. None of these is a cart.
Neighbour
Another device, with its own goal-by-goal record.
Neighbour
A chamber with a list of accepted uses.
Floor
The floor habit for joints and strength.
Floor
A marker, or a panel you still need. Not a diagnosis.
Disclosure
Where a PEMF mat appears in a personal plan, labelled as one.
Order
Measure first. Floor habits before gadgets.
Map
Every pillar, this one among them, held to the same rule.
Knowledge stays free. No seller wrote this page. Every source below opens the record that was checked before the sentence was written, with what it found and who funded it.
4 sham-controlled trials, 125 people with fractures that had not healed. The pooled effect on healing was not significant (risk ratio 1.96), with substantial heterogeneity; two trials found no reduction in pain. Inconclusive and insufficient to inform practice. The authors declare no known conflicts.
16 placebo-controlled randomised trials in osteoarthritis, 15 pooled. Against placebo, PEMF improved pain (SMD 1.06), stiffness (0.37) and function (0.46); quality of life did not reach significance. Device settings did not change the result. PubMed tags it as non-government funded; no conflict statement in the record.
12 sham-controlled trials, ten in the knee. PEMF reduced knee pain (SMD -0.54) and improved knee function (-0.34); hand results were larger but from few trials; no benefit in the neck. Sessions of 30 minutes or less did better. Adverse events did not differ from sham. No competing interests declared.
11 randomised, sham-controlled, double-blind trials of whole-body PEMF mats, 12 to 71 people each, at 3.4 to 200 microtesla for up to 12 weeks. Some sporadic positive effects, never independently confirmed. Evidence insufficient; acute adverse effects not reported; long-term effects not studied. PubMed tags it as non-government funded.
55 people with treatment-resistant depression, five weeks of daily active or sham PEMF to the head, multicentre and double-blind. Both groups improved; active was not better than sham on any measure. PubMed tags it as non-government funded; no conflict statement in the record.
Cochrane review of electrotherapies for shoulder pain. In single small trials, PEMF showed no clinically important benefit, on low or very low quality evidence. Adverse events were not measured in any PEMF trial. One author declares an interest as a practising physiotherapist paid to deliver treatment.
57 people with knee osteoarthritis averaging 61.6 years, a named handheld device at 4 to 12 Hz and 105 mT, twice a day for 5 minutes over 18 days, against a placebo device. Pain, stiffness and disability scores fell more than with placebo. The authors call it partly randomised. No funding or conflict statement in the PubMed record.
29 people with shoulder tendinitis that had not responded to steroid injection. In the first 4 weeks, the 15 on active coils improved more than the 14 on placebo; once all were on active coils, the groups no longer differed. PubMed tags it as non-government funded; no conflict statement in the record.
30 people with complex regional pain syndrome, a named whole-body device or sham added to 10 days of rehabilitation, double-blind. Pain and function improved more in the short term with the active device. A pilot. No funding or conflict statement in the PubMed record.
101 people with insomnia, 4 weeks of active treatment or placebo. 70 percent of the active group reported substantial or complete relief against one placebo patient with very clear relief. No adverse effects reported. One trial, not replicated in the search. No funding or conflict statement in the PubMed record.
8 randomised trials, 421 people with knee osteoarthritis, classical PEMF against placebo. Physical function improved (WMD -5.28), but pain, stiffness and total WOMAC score did not differ significantly. No funding or conflict statement in the PubMed record.
11 randomised trials; four on 106 delayed or ununited fractures pooled to a non-significant relative risk of 1.76. Single trials found benefit in some settings and none in others. Notes that bone stimulation is a 500 million dollar market in the United States. No funding or conflict statement in the PubMed record.
3 randomised trials, 197 people with fresh fractures. None found a significant effect on bone healing; pain results conflicted. Does not support PEMF for healing acute fractures. The authors declare no conflicts.
19 studies, 1303 postmenopausal women. PEMF added to standard drugs raised bone density and eased pain against drugs alone. Against placebo, PEMF did not change lumbar spine density. Adverse events did not differ. Funded by Chinese national and city science grants.
20 small trials, 1239 people with neck pain, too varied to pool. Very low quality evidence that PEMF was more effective than placebo. Funded in part by the NIH; no known conflicts.
58 people with treatment-resistant depression, 8 weeks of PEMF to the head, no control group. Depression scores fell from 20.6 to 12.6. Without a sham arm, the drop cannot be separated from expectation or time. Several authors declare grants and honoraria from drug companies.
Rats with complete or partial Achilles tears, 1 or 3 hours of PEMF a day or none, to 6 weeks. No clear differences in healing; some tendon properties were worse. Funded by Orthofix, a bone stimulator maker, and the NIH.
70 rats, shoulder tendon repair with or without PEMF. At 4 weeks the PEMF group had stiffer, stronger repaired tissue; joint mechanics and walking did not differ. Funded by Orthofix, a bone stimulator maker.
Diabetic rats with a 2 cm skin wound, PEMF at 5 mT and 25 Hz for 1 hour a day. Wound closure was faster on days 10 and 14, then the difference disappeared. PubMed tags it as non-government funded.
Male rats aged 30 to 32 months, whole-body PEMF for six weeks, 3 times a week, 24 minutes a session. Attention in a recognition test and spatial working memory improved. PubMed tags it as non-government funded.
Ageing male mice. PEMF increased new bone formation through sensory nerves releasing semaphorin 3A; removing those nerves or the gene removed the effect. Funded by Chinese science grants; no competing interests declared.
Rats with ovaries removed. PEMF at 40 Hz reduced bone loss more than a standard drug, vitamin D and calcium; PEMF at 25 Hz made osteoporosis progress further. No funding or conflict statement in the PubMed record.
20 dogs with hip osteoarthritis, PEMF at 50 Hz for 30 minutes twice a week for 5 weeks or no treatment, owners not blinded to a sham. Pain scores improved; joint range and thigh size did not differ. Most authors declare no financial interests.
Educational use only. This page is not medical advice. Field strengths, frequencies and session lengths are what a study ran, not a setting for you. Anyone with an implanted electronic device, a pregnancy or a fracture under treatment has a question for their own clinician that this page does not answer. Aaron Cuha is a coach, not a doctor.
Educational use only. This database summarizes published research and is not medical advice. Where a dose appears it is a record of what a trial or clinic ran, or what people describe doing, with its source attached. Nothing here is a recommendation for personal use.