Re: Snake Oil

[Stan Dodds <dodds@RICE.EDU>]

Ray Rogoway asks:

> I think it's our responsibility to deal with "JUNK SCIENCE" in the
> classroom. Does anyone have any data that supports magnetic therapy?

I have not done a literature search, but there is at least one
controlled study of the phenomenon: Vallbona et al, "Response of Pain
to Static Magnetic Fields in Postpolio Patients: A Double-Blind Pilot
Study", Arch. Phys. Med. Rehabil. 78,1200 (1997). The magnets had
concentric circles of opposite polarity magnetization, with field
strengths of several hundred gauss at the surface of the
device. Identical but non-magnetized devices were used as
control. Subjects, about 50, were selected for a particular type of
pain, and randomly assigned. Pain scoring was subjective, using an
apparently standard questionnaire. Overall the design seems
reasonable, and the results are clearly positive.

There are obvious objections, of course. The study is very small,
patients could probably figure out whether they had magnets or not
(conflating placebo effect with magnetic effects), there is no
indication of the spatial dependence of the field. The authors seem
aware of these problems, and do not make sweeping claims, nor do they
claim a mechanism.

As in many areas of biomedicine, the situation appears hazy to this
outsider. Some devices, as pointed out by Andy Dougherty and Ron
Parks, don't produce any field within the body, and hence can't
work. Very strong fields, as in MRI, have no apparent effects at
all. One controlled study seems to show a positive effect, but other
less rigorous work is mixed.

I think this is an area that has a lot of junk, but it could be
science if the marketing hype could be stripped away. Magnets either
do or do not help pain, a question that can be decided by sufficiently
clever experiments. If there is an effect, one could pursue a
mechanism. It might, therefore, be a very good area for us to deal
with in a classroom.

Stan