Articles

Copper and Magnetic Bracelets for Arthritis

A recent study published in the journal Complementary Therapies in Medicine shows no benefit from copper or magnetic bracelets for symptomatic treatment of arthritis. While this is a relatively small study, it highlights the lack of evidence to support this billion dollar plus industry.

The study is a double-blind, controlled crossover study involving 45 subjects with osteroarthritis. Each subject wore one of four bracelets – copper, two types of magnetic bracelet, and one demagnetized, in random order each for 16 weeks. It showed no difference among the four groups.

This is only the second published controlled trial looking at copper bracelets for arthritis. The first is from 1976 and showed some benefit. Then there are no published studies (just reviews and comments) for the next 33 years, until this current study.

There have been a number of studies of magnetic bracelets. These are more difficult to interpret – the primary problem with these studies is that it is very difficult to properly blind subjects. It is simply too easy to discover if a bracelet or other magnetic device is actually magnetic. You can partially mitigate this problem by using devices of various magnetic strengths, but even then patients may have an easier time detecting the stronger magnets.

One study found that magnetic bracelets correlated with improved symptoms from hip and knee osteoarthritis. The authors concluded:

“It is uncertain whether this response is due to specific or non-specific (placebo) effects.”

Uncertain indeed. What this study actually shows is how difficult it is to eliminate placebo effects from these studies. There is almost no plausibility to magnetic bracelets improving knee arthritis – the bracelets are simply too remote. None of the hypothesized mechanisms by which static magnets might work have panned out – they do not influence blood flow or inflammation. Most static magnets do not even penetrate to the target tissue, and in this case the magnets are no where near the target tissue.

Studies that use magnetic braces and wraps are also problematic because the braces and wraps themselves can have significant benefit.

A recent systematic review of static magnets to treat pain concluded:

The evidence does not support the use of static magnets for pain relief, and therefore magnets cannot be recommended as an effective treatment. For osteoarthritis, the evidence is insufficient to exclude a clinically important benefit, which creates an opportunity for further investigation.

Consumers using such products are prey to all of these placebo effects, in addition to others, such as regression to the mean. Arthritis, like many illnesses, waxes and wanes. People are likely to seek treatments when their symptoms are bad, and statistically likely to improve to a more average severity. This regression to mean severity is easily interpreted as a response to whatever treatment they initiated when their symptoms were at their worst.

So what we have is a series of implausible claims, negative or insufficient evidence, and a clinical setting where self-deception is likely. This adds up to over a billion dollars of wasted health care dollars.

We have to confront the fact that this is what has and will happen in the absence of adequate regulation – the public will waste money on useless therapies, their attention and resourcesy may be diverted from more effective interventions, and some unscientific interventions may be directly harmful.

There is also no incentive for industry to spend time and money doing quality studies when they don’t have to. Sellers of copper bracelets were happy to do so for decades without studying their claims.

Further, most new health claims are going to turn out to be false. The lower the prior probability (by definition) the more likely the claims are to be false. So accross the board there is a great deal of risk and expense to products with untested or dubious health claims, and very little benefit.

Posted in: Science and Medicine

Leave a Comment (62) ↓