Collacott EA, Zimmerman JT, White DW, Rindone JP. Bipolar Permanent Magnets for the Treatment of Chronic Low Back PainA Pilot Study. JAMA. 2000;283(10):1322-1325. doi:10.1001/jama.283.10.1322
Author Affiliations: Bio-Electro-Magnetics Institute, Reno, Nev (Dr Zimmerman); and Veterans Affairs Medical Center, Prescott, Ariz (Drs Collacott and Rindone and Mr White).
Context Chronic low back pain is one of the most prevalent and costly medical
conditions in the United States. Permanent magnets have become a popular treatment
for various musculoskeletal conditions, including low back pain, despite little
scientific support for therapeutic benefit.
Objective To compare the effectiveness of 1 type of therapeutic magnet, a bipolar
permanent magnet, with a matching placebo device for patients with chronic
low back pain.
Design Randomized, double-blind, placebo-controlled, crossover pilot study
conducted from February 1998 to May 1999.
Setting An ambulatory care physical medicine and rehabilitation clinic at a
Veterans Affairs hospital.
Patients Nineteen men and 1 woman with stable low back pain of a mean of 19 years'
duration, with no past use of magnet therapy for low back pain. Twenty patients
were determined to provide 80% power in the study at P<.05
to detect a difference of 2 points (the difference believed to be clinically
significant) on a visual analog scale (VAS).
Interventions For each patient, real and sham bipolar permanent magnets were applied,
on alternate weeks, for 6 hours per day, 3 days per week for 1 week, with
a 1-week washout period between the 2 treatment weeks.
Main Outcome Measures Pretreatment and posttreatment pain intensity on a VAS; sensory and
affective components of pain on the Pain Rating Index (PRI) of the McGill
Pain Questionnaire; and range of motion (ROM) measurements of the lumbosacral
spine, compared by real vs sham treatment.
Results Mean VAS scores declined by 0.49 (SD, 0.96) points for real magnet treatment
and by 0.44 (SD, 1.4) points for sham treatment (P
= .90). No statistically significant differences were noted in the effect
between real and sham magnets with any of the other outcome measures (ROM, P = .66; PRI, P = .55).
Conclusions Application of 1 variety of permanent magnet had no effect on our small
group of subjects with chronic low back pain.
Low back pain is a disabling, costly condition that is difficult to
treat effectively. It is estimated that 85% of all people will have back pain
during their lifetime. Annual prevalence reports range from 15% to 45%.1 Currently more than 5 million Americans are disabled
with low back pain.2 The direct cost of treating
low back pain is estimated at $15 billion, with indirect costs as high as
$100 billion annually.3 According to one recent
study, 75% of patients who visit a physician for an acute episode of low back
pain are still symptomatic 1 year later.4 In
view of the significant risks of some conventional treatments, there has been
renewed interest in alternative methods5 for
a variety of conditions. In the United States there has been a dramatic increase
in visits to practitioners of alternative medicine, with expenditures in the
range of $21.2 billion in 1997.6
There currently exists a media campaign promoting the use of permanent
magnets for the treatment of pain,7- 9
which has resulted in large profits: worldwide sales of $5 billion have been
reported.10 Four reports of the use of permanent
magnets for the treatment of pain have been published, only 1 of which used
a double-blind randomized design.10- 13
Vallbona and colleagues12 used active and placebo
permanent magnets in a single 45-minute application to treat muscle pain in
patients with postpolio syndrome. A statistically significant improvement
was reported with the active magnets. Weintraub11
reported a statistically significant reduction in neuropathic pain in patients
with diabetic polyneuropathy using magnetic foot pads for a total of 12 weeks.
Because of public interest, and a lack of adequate scientific documentation,
there has been a call for further clinical trials.14
Two types of permanent magnet therapy are popular: unipolar and bipolar.
The terms unipolar and bipolar refer to the magnetic poles facing the patient's skin. Unipolar magnet
therapy usually uses several discrete, individual magnets aligned with the
same magnetic pole toward the skin. Usually the pole facing the patient is
the biomagnetic north pole or negative pole (hence the term unipolar). The so-called biomagnetic south pole then faces away from
Bipolar magnet therapy usually uses a flexible plastalloy (flexible
plastic containing barium ferrite) sheet of magnetic material that has impressed
upon it an alternating spatial pattern of north and south magnetic domains.
Popular patterns in some bipolar magnetic designs are stripes, concentric
circles, squares, and triangles. In each case, the adjacent shape is of the
opposite magnetic polarity relative to its contiguous neighbor. Bipolar magnet
therapy thus uses magnetic material arranged in an alternating pattern, so
that both north and south poles face the skin.
The length of time the magnets must be worn to obtain maximum pain relief
has not been established. Vallbona et al12
reported pain relief from a 45-minute application while Weintraub10 suggested continuous treatment.
Therapeutic permanent magnets are popular for a variety of musculoskeletal
Low back pain was selected for study because it is 1 of the most common problems
for which magnets are used. The objective of this study was to evaluate the
effectiveness of a commonly used bipolar magnet, used repetitively for a fixed
period, in the treatment of chronic low back pain.
This was a randomized, double-blind, placebo-controlled, crossover pilot
study. The trial was carried out in the physical medicine and rehabilitation
outpatient facility of a Veterans Affairs hospital in Prescott, Ariz. The
institutional review board of the Veterans Affairs hospital approved the study
and all subjects gave written informed consent prior to participation. Randomization
of devices was achieved using a computer-generated list of random numbers.
The order of treatment (magnet first and placebo second or vice versa) also
Participants were recruited from the primary care and physical medicine
and rehabilitation clinics of a Veterans Affairs hospital. Participants were
recruited from February 1998 through March 1999. A total of 24 individuals
met the study criteria. Two individuals did not wish to participate and 2
of the remaining 22 failed to complete the protocol because of time conflicts,
leaving 20 subjects who completed the study. Their general characteristics
are given in Table 1.
All patients were examined by the same investigator (E.A.C.). Evaluations
included radiography of the lumbosacral spine for all subjects. All imaging
studies were consistent with spondylosis and some studies revealed additional
diagnoses (Table 1). However,
the source of subjects' existing back pain was thought to be degeneration
of the "3-joint complex"15 (intervertebral
disks and facet joints) in all cases. Subjects were required to have stable
low back pain of at least 6 months' duration. None had a new neurological
deficit. Exclusion criteria were past use of magnet therapy for low back pain;
secondary gain issues such as recent application for disability or ongoing
litigation; pregnancy; cardiac pacemaker; skin lesions over the site of pain;
and difficulty understanding the directions.
The visual analog scale (VAS) was chosen as the primary outcome measure
and used to quantify pain intensity. The VAS, shown to be a reliable and valid
measure,16,17 consists of a standard
10-cm line with verbal anchors indicating "none" at one end (0) and "severe"
at the other (10). Participants were told to estimate their current level
of pain by an appropriate mark on the line, with severe indicating the worse
imaginable pain. The McGill Pain Questionnaire was used to measure the affective
component of pain.18 The Pain Rating Index
(PRI) of the McGill Pain Questionnaire was the sole measure analyzed rather
than the subscales.19,20 The PRI
uses 20 category scales of verbal descriptors to demonstrate the sensory and
affective components of pain, with a minimum score of 0 and a maximum score
(most severe pain) of 78. Lacking a reliable measure of the physical response
to pain, formal measurements of the range of motion (ROM) of the lumbosacral
spine were obtained for all subjects by the same investigator (D.W.W.). The
degree of flexion/extension of the lumbosacral spine was obtained using the
The trial lasted from February 9, 1998, to May 21, 1999. All subjects
followed the treatment protocol for 2 weeks: 1 week with the magnets and 1
week with the sham devices. There was a 1-week washout period between the
2 treatment weeks. The protocol consisted of application of the devices 6
h/d, 3 d/wk (Monday, Wednesday, and Friday of each treatment week). Therefore,
all participants were exposed to a total of 18 hours of treatment for both
real and sham devices.
Both the real and sham devices consisted of a flexible rubberlike compound.
The real devices were impregnated with active magnetic material; sham devices
were identical, but had been demagnetized using a magnetizer/demagnetizer
(Master Magnetics, Colorado Springs, Colo). After demagnetization the sham
devices showed no detectable magnetism by gaussmeter. The magnetic strength
measured on the cloth surface of the real devices approximated 300 G (range,
282-330 G). The devices were trapezoid in shape (19 × 11.5 × 14
cm) and about 2 mm thick. The surface applied to the participant's skin was
covered with cloth fabric, while the outer surface was covered with smooth
gold foil. The magnets were of the bipolar variety, with a magnetic design
of multiple triangles of alternating north and south magnetic polarity. An
abdominal binder, connected via Velcro straps, held the device in place. Subjects
were advised not to manipulate the devices.
Participants completed VAS measurements before and after each treatment
(ie, before and after either the real or sham device was applied). Measurements
for ROM and PRI were obtained before and after the first day's treatment and
at the end of each week. The same examiner (D.W.W.) applied all the devices
and obtained all ROM measurements. Subjects were advised to avoid new treatments
during the time of the study, and they were instructed not to alter their
usual pattern of medication use.
Statistical analysis of the data was performed using SigmaStat software
(San Rafael, Calif). A 2-point difference in the VAS scores was selected as
a clinically significant reduction in pain intensity. The power calculation
was performed for the primary outcome measure (VAS). Analysis showed that
enrolling 20 patients would provide an adequate sample size to achieve a power
of 80%, assuming an SD of 1.5. The difference between baseline (preapplication)
and posttreatment VAS scores and ROM measurements was analyzed using repeated
measures analysis of variance and paired t test,
respectively. The difference between baseline and posttreatment PRI scores
(nonparametric) was analyzed and compared between treatments using the Wilcoxon
signed rank test. Statistical significance was set at P<.05.
The mean (SD) cumulative baseline VAS score for all participants was
4.8 (2.2). The mean cumulative baseline score for the sham treatment was 5.0
(2.4) and for the magnet treatment, 4.7 (2.9). Table 2 shows the difference between baseline and posttreatment
VAS scores, by day, for magnet and sham treatments. Table 3 shows the difference between baseline and posttreatment
ROM and PRI measures for day 1. Table 4 compares cumulative change from baseline (day 1 vs posttreatment
day 3) for VAS, ROM, and PRI. No statistically significant differences between
magnet and sham treatment were found with any outcome measure. No adverse
effects were reported by any of the participants.
This study found no immediate or cumulative difference in the outcome
measures of low back pain comparing real with demagnetized (sham) therapeutic
magnets. Neither magnet nor sham treatment achieved the prospectively determined
reduction in pain intensity (2-point reduction in the VAS score). To our knowledge,
this is the only randomized, double-blind, placebo-controlled study reporting
the use of permanent magnets on more than a single occasion and for more than
45 minutes. Our protocol treatment was 3 times a week for a total of 18 hours
for both treatments. This was a pilot study and was not intended to prove
or disprove the effectiveness of magnet therapy in general. Additional studies
using different magnets (unipolar and bipolar), treatment times, and patient
populations are needed.
A stronger magnet may be necessary to penetrate to the source of chronic
low back pain. Expanding the treatment time in an ambulatory setting would
require manipulation of the devices by the subjects, which would increase
the chances of detecting whether the magnet/sham was magnetized. If a device
with opposite polarity or of a certain design was found to be ineffective,
it could serve as a placebo.
Our study population was small and had special requirements. The subjects
had to travel to the clinic twice daily at specific times, conditions that
eliminated many employed and younger individuals. Other potential subjects
lived too far away or had no transportation. Since we recruited from the Veterans
Affairs clinic population there were too few women. Thus, it is difficult
to generalize these results to the population at large with chronic low back
Our results did not support the findings of Vallbona et al12
and Weintraub.10 However, there were considerable
differences in the study designs and populations, including the cause of pain.
The patients of Vallbona and colleagues had muscle pain while Weintraub's
subjects had neuropathic pain. The source of pain in our participants would
appear to be deeper than that of the former, and may explain the lack of beneficial
effect from the magnets used (300 G).