0. The 30-second answer: do not confuse “certified medical device” with “strong modern clinical evidence”
Here is the line to underline.
A product being certified as a regulated medical device is not the same thing as that product having been proven effective in a modern randomized controlled trial.
Japan has an official Class II category called the household permanent-magnet magnetic therapy device. Its permitted intended effect includes improvement of stiffness and blood circulation at the site of application.
So a product that fits the category and passes the applicable standards may legally be marketed within that wording.
But the science tells a different story. A 2025 review of static magnetic fields and blood flow found that none of the reviewed human studies showed a statistically significant increase in blood flow attributable to the static magnetic field. One showed a decrease.
So the puzzle is real:
Regulation: “blood-flow improvement” is an accepted intended effect. Current human evidence: increased blood flow from permanent static magnets has not been demonstrated.
This article explains how both statements can be true at the same time.
1. The story starts with a very smooth shirt containing magnets
In September 2024, AEON launched PEACE FIT SilkyFACT × PIP innerwear. The launch material described a smooth functional undershirt carrying 100 mT magnets like those used in PIP Elekiban, positioned around the shoulder, back, or waist, and marketed for improving circulation while being worn.
A normal shopper may naturally reason:
medical device → government checked the effect → the magnet must have been shown in clinical trials to dilate vessels and raise blood flow.
That last step is the trap.
Medical-device certification can be based on conformity with an established device category and technical standard. It does not necessarily mean every new shirt was tested against an identical sham shirt in a fresh large trial.
2. First, split “magnetic therapy” into three different things
A. Static fields from permanent magnets. Magnetic necklaces, patches, and magnet-containing underwear belong here. The field is essentially not changing over time.
B. Time-varying electromagnetic fields. A powered coil creates a changing field. PEMF-type treatments belong here. This is physically different from simply sewing a permanent magnet into clothing.
C. Electrical stimulation. TENS and EMS send electrical current through electrodes. They are not permanent-magnet therapy.
So the fact that medicine can use electromagnetism does not prove that a passive refrigerator magnet has the same effect.
If every magnetic field were automatically restorative, the refrigerator aisle would be a rehabilitation clinic. It is not.
3. What does “medical device” actually certify?
Think of it like a school category.
Medical-device certification is not a gold medal saying “100/100 clinical effectiveness.” It is closer to confirming that the product meets the requirements to belong to a regulated class with a defined intended use.
The current PMDA listing classifies household permanent-magnet therapy devices as Class II and gives the intended effect as improvement of stiffness and blood circulation at the site of application for household use.
The 2024 essential-principles checklist points to JIS T 2007:2024 for conformity and, for benefit-related performance, checks recognized performance items such as maximum magnetic flux density.
Most importantly, the clinical-trial line is marked not applicable: this is not a device requiring a clinical trial.
So a new product in this category is not automatically required to run a new trial with:
- 100 real-magnet users,
- 100 sham-magnet users,
- blinding,
- objective blood-flow measurement,
- pain and function outcomes,
- statistical comparison.
An RCT is basically a fair comparison in which similar people are allocated to different groups, ideally without knowing who receives the real treatment.
4. Why does this category exist? Follow the timeline
1961: Japan explicitly added magnetic therapy devices to the medical-device framework under the new Pharmaceutical Affairs Act.
1980: magnetic therapy-device applications were required to submit data on magnetic characteristics and clinical tests. A related notice said clinical data should in principle come from at least two medical institutions and at least 30 cases per institution for each indication.
That means the category was not created with literally zero clinical data requirements.
But “clinical study in 1980” does not automatically mean a modern randomized, double-blind, sham-controlled, preregistered trial.
1998: a major change arrived through regulatory-relaxation reform. Devices considered relatively low risk and capable of having quality, safety, and effectiveness assured by appropriate standards could become approval-exempt when they conformed to those standards. A specific standard for household permanent-magnet therapy devices was established.
From 2005 onward: the modern Class II third-party certification structure took shape. Today the category refers to JIS T 2007:2024, and fresh clinical trials are not required for ordinary conforming devices.
The system therefore became less like “re-prove permanent magnet physiology from zero every time” and more like “does this new product fit the established box?”
That is a classic way regulatory path dependence can emerge.
5. Does the category itself have convincing modern evidence?
For the mechanism most often advertised—improved circulation from static magnets—the evidence is weak.
The 2025 review found no statistically significant increase in blood flow in the reviewed human studies attributable to a static magnetic field.
The authors also noted limitations: small samples, short exposures, many healthy participants, and limited field orientations.
So the careful conclusion is not “science has proved magnets can never do anything.”
It is:
Current human experimental evidence does not support the claim that static permanent magnets increase blood circulation.
That is the scientifically useful statement.
6. Why can a large market still exist?
From the manufacturer’s side:
an established category exists → build a product to the standard → obtain certification → use the permitted “stiffness/blood circulation” wording → sell patches, necklaces, bands, or clothing.
From the shopper’s side:
“regulated medical device” often sounds like → “this exact product passed a hard modern efficacy trial.”
Those are not the same thing.
That gap is an information asymmetry.
This does not automatically mean the company is selling an illegal fake. A company may be using a legally recognized indication exactly as the system allows.
The problem is that the label can sound scientifically stronger than the evidence required for each individual product.
Also, the timeline does not support the simple claim that PIP created the regulatory category for its product: the category dates to 1961, while PIP Elekiban was launched in 1972.
7. So what actually has evidence for ordinary neck and shoulder pain?
First lesson: the target is not the blood-flow number itself.
The outcomes that matter are pain, function, movement, daily-life interference, and recurrence.
| Method | Evidence for pain/function | Evidence for blood flow | Practical role |
|---|---|---|---|
| Exercise / strengthening | Consistently supported, though certainty varies | Active muscle blood flow rises | Core treatment |
| Short movement breaks | Promising evidence in sedentary workers | Muscle activity raises flow | Very practical |
| Topical NSAIDs | Supported for acute musculoskeletal pain; neck-pain RCTs exist | Not a circulation treatment | Short-term pain relief |
| Heat / bathing | Direct neck-pain trials are limited | Direct local heat measurably raises muscle flow | Cheap adjunct |
| Manual therapy + exercise | Supported better than passive care alone | Not the main mechanism | Adjunct |
| Acupuncture | Several RCTs/meta-analyses show benefit | Circulation is not the main evidence base | Chronic-pain adjunct |
| Massage | Effect versus placebo is uncertain | Temporary changes possible | Comfort adjunct |
| TENS | Evidence insufficient for chronic neck pain | Not established as a circulation solution | Lower priority |
| Permanent magnets | Human blood-flow increase not demonstrated | Not supported at present | Low priority |
Exercise is boring in the same way brushing your teeth is boring: unfortunately, boring things can work.
A 2024 review in office workers found that strengthening neck, shoulder, and scapular muscles reduced pain and disability, although certainty was low. A 2026 review of micro-exercises in sedentary workers also suggested reductions in neck/shoulder pain.
8. What about patches? Drug patches are not the same as magnet patches
Topical NSAIDs such as diclofenac have a pharmacological mechanism.
A Cochrane review supports topical NSAIDs for acute musculoskeletal pain. A randomized double-blind study of 72 people with acute neck pain found diclofenac gel superior to placebo for movement pain, rest pain, and function. Another randomized trial in 153 people with upper-trapezius myofascial pain favored a diclofenac patch over a control patch for several outcomes.
The important part is not “something was stuck to the shoulder.”
It is what active mechanism was stuck to the shoulder.
A drug patch and a passive magnet are not evidence-equivalent just because both use adhesive.
9. Public baths and ordinary baths: this part is much easier physiologically
Heat really does heat you. A shocking development for science.
In a 2011 PET study, direct local heating raised calf muscle temperature from about 33.4°C to 37.4°C and increased muscle blood flow from about 1.4 to 2.3 mL/100 g/min.
So:
direct local heating → vascular response → measurable increase in local muscle blood flow
has direct human evidence.
But the same experiment also showed that whole-body heating did not automatically increase blood flow in a calf muscle that was not directly heated.
So “I feel warm” and “every deep muscle in my body now has massively increased flow” are not the same claim.
If your shoulders are actually in warm water, they are being directly heated. That makes the mechanism much more straightforward than passive static magnets.
However, a 2026 systematic review of heat for nonspecific neck pain found only two eligible studies, totaling 67 participants. Pain and disability improved, but the evidence base was too small for a strong efficacy conclusion.
Therefore:
Heat increases local circulation: well supported physiologically. Heat alone cures chronic neck pain: not established.
10. Hot-spring “indications” need the same careful reading
Japan’s Ministry of the Environment lists general bathing indications for therapeutic hot springs, including chronic muscle/joint pain or stiffness, cold sensitivity, peripheral circulatory disorders, recovery from fatigue, and health promotion.
But the Ministry also explicitly explains that hot-spring effects arise from a combination of temperature, physical factors, chemical components, local climate, changes in daily life, and psychological responses. Effects cannot simply be assigned to one mineral component.
That means an ordinary public bath can still provide a large part of the thermal, buoyancy, rest, and relaxation experience.
Mineral-specific extra effects may exist in some settings, but “hot spring always beats ordinary hot water” is too broad a claim.
11. Massage, manual therapy, acupuncture, and TENS
Massage: a 2024 Cochrane review included 33 studies and 1,994 analyzed participants. Compared with placebo, massage may make little or no difference in pain in subacute/chronic neck pain; certainty was low. High-dose massage subgroups looked better, but the overall evidence remains uncertain.
Manual therapy: a 2025 umbrella review found relatively strong support for manual therapy combined with exercise for nonspecific neck pain. Passive care seems more defensible when it helps people move, not when it replaces movement.
Acupuncture: a 2025 meta-analysis of 26 trials and 3,520 participants found greater pain reduction versus inert treatments; 13 trials were judged at low risk of bias. It is reasonable as an adjunct for some chronic cases, not a universal first-line answer.
TENS: the Cochrane review included seven studies and 651 participants. Evidence versus sham was very low certainty, so benefit remains uncertain.
“Electric” sounds scientific. “Magnetic” sounds scientific. Science does not grade the packaging. It grades the comparison.
12. The teacher-big-sister spending order
For ordinary nonspecific neck/shoulder stiffness:
- Move first. Break long static postures and move the neck, shoulders, and shoulder blades within a comfortable range.
- For persistent symptoms, add strengthening. Consistency beats exotic technique.
- Use heat if it feels good. A home bath, public bath, or heat pack is a reasonable low-cost adjunct.
- For stronger acute pain, topical NSAIDs may be an option after checking suitability with a pharmacist or clinician.
- Add exercise-based manual therapy or, in chronic cases, acupuncture if needed.
- Do not infer strong clinical efficacy from the words “medical device” alone when looking at expensive magnetic products.
And the four questions to remember are:
- What exact device category is this?
- Is the claim an accepted regulatory indication or a result proven for this exact product?
- What did human trials actually measure?
- Is there a cheaper intervention with stronger evidence?
Once you ask those four questions, the supermarket shelf gets much less magical.
And no, you do not need to stand in front of the refrigerator to recover.
The refrigerator has one job. It keeps the food cold.
