Content uploaded by Iris Bell
Author content
All content in this area was uploaded by Iris Bell on May 08, 2014
Content may be subject to copyright.
Rheumatology 2004;43:577–582 doi:10.1093/rheumatology/keh111
Advance Access publication 20 January 2004
Improved clinical status in fibromyalgia patients
treated with individualized homeopathic remedies
versus placebo
I. R. Bell
1,2,3,4,6,8
, D. A. Lewis II
9
, A. J. Brooks
3
, G. E. Schwartz
3,5,6
,
S. E. Lewis
9
, B. T. Walsh
4
and C. M. Baldwin
3,4,7,8
Objective. To assess the efficacy of individualized classical homeopathy in the treatment of fibromyalgia.
Methods. This study was a double-blind, randomized, parallel-group, placebo-controlled trial of homeopathy. Community-
recruited persons (N ¼ 62) with physician-confirmed fibromyalgia (mean age 49 yr,
S.D. 10 yr, 94% women) were treated in a
homeopathic private practice setting. Participants were randomized to receive oral daily liquid LM (1/50 000) potencies with an
individually chosen homeopathic remedy or an indistinguishable placebo. Homeopathic visits involved joint interviews and
concurrence on remedy selection by two experienced homeopaths, at baseline, 2 months and 4 months (prior to a subsequent
optional crossover phase of the study which is reported elsewhere). Ten der point count and tender point pain on examination by
a medical assessor uninvolved in providing care, self-rating scales on fibromyalgia-related quality of life, pain, mood and global
health at baseline and 3 months, were the primary clinical outcome measures for this report.
Results. Fifty-three people completed the treatment protocol. Participants on active treatment showed significantly greater
improvements in tender point count and tender point pain, quality of life, global healt h and a trend toward less depression
compared with those on placebo.
Conclusions. Thi s study replicates and extends a previous 1-month placebo-controlled crossover study in fibromyalgia that
pre-screened for only one homeopathic remedy. Using a broad selection of remedies and the flexible LM dose (1/50 000 dilution
factor) series, the present study demonstrated that individualized homeopathy is significantly better than placebo in lessening
tender point pain and improving the quality of life and global health of persons with fibromyalgia.
KEY WORDS: Fibromyalgia, Homeopathy, Chronic pain, Global health.
The use of homeopathy as a complementary medical treatment for
a wide range of acute and chronic conditions is increasing [1, 2],
with high levels of patient satisfaction with homeopathic care [3].
Clinicians often report benefit of individualized constitutional
homeopathic remedies in patients having overlapping, polysymp-
tomatic disorders, for example fibromyalgia (FM), chronic fatigue
syndrome and multiple chemical sensitivity with low-level chemical
intolerance, for which conventional medicine has limited options.
Fibromyalgia is a chronic diffuse musculoskeletal pain disorder
involving concomitant fatigue, sleep disturbance and, often,
co-morbid depression [4]. The prevalence in the United States is
2% [5]. Fibromyalgia disproportionately affects women. One
randomized, double-blind crossover study of patients meeting
criteria for a single homeopathic remedy, Rhus toxicodendron,
documented greater improvements over 1 month in number of
painful tender points and better sleep on active versus placebo
treatment [6].
Although systematic reviews of homeopathy have found that
active treatment has an advantage over placebo across various
conditions, investigators have called for greater efforts to replicate
and extend homeopathic studies on specific conventional diag-
nostic entities [7]. The debate over poor reproducibility of findings,
methodological shortcomings, and interpretation of data from
previous studies has been vigorous [8]. The purpose of this study
was to perform a randomized, double-blind, placebo-controlled
feasibility trial of individualized homeopathy in fibromyalgia using
daily LM (1/50 000 dilution factor) potencies.
Methods
Design
A double-blind, parallel group design of randomly assigned active
versus placebo individualized, pragmatic homeopathic treatment
was implemented. Patients had homeopathic visits at a private
clinic in Phoenix, Arizona, at baseline, 2 months, 4 months and
6 months of treatment. They were evaluated with the same battery
of outcome measures during laboratory assessment visits at the
University of Arizona (Tucson) at baseline, 3 months and
6 months. An optional crossover treatment phase of the study
was implemented immediately after the 4-month homeopathic visit
and occurred over months 5 and 6 (post-crossover laboratory and
clinical results are reported elsewhere [9]).
The 3-month laboratory evaluation and 4-month homeopathic
visits were separated in time because of (1) practical considerations
577
Correspondence to: I. R. Bell, Program in Integrative Medicine, The University of Arizona Health Sciences Center, 1249 N. Mountain Avenue,
Tucson, AZ 85719, USA. E-mail: IBELL@U.ARIZONA.EDU
1
Program in Integrative Medicine,
2
Departments of Psychiatry,
3
Psychology,
4
Medicine,
5
Neurology,
6
Surgery, the
7
Arizona Respiratory Center and the
8
Mel and Enid Zuckerman Arizona College of Public Health at the University of Arizona, Tucson, Arizona, and
9
Saybrook Graduate School and Research
Institute, San Francisco, California, USA.
Submitted 9 July 2003; revised version accepted 12 November 2003.
Rheumatology Vol. 43 No. 5 ß British Society for Rheumatology 2004; all rights reserved
of subject time/travel burden (because of the 240-mile/4-h round
trip between Tucson and Phoenix) and (2) the need to ensure
acquisition of follow-up laboratory data prior to the 4-month
homeopathic clinical visit and crossover. As a result, the primary
outcomes reported in this paper derive from baseline and 3-month
Tucson laboratory assessment visits. However, we also include in
the present report the patients’ ratings of treatment helpfulness
obtained at the 4-month clinical follow-up homeopathic visit in
Phoenix, for a fuller picture of evolving outcomes up to and
immediately prior to the optional crossover point.
Patients daily succussed then diluted liquid remedy potencies or
placebo in 4 oz of water, all starting with LM 1 doses (a 1/50 000
ratio dilution in 20% alcohol–water solvent, with succussions) or
placebo. The LM potency was taken orally and gradually raised
over the course of treatment in an individualized manner.
The rationale for LM potencies was two-fold. First, many
fibromyalgia patients in the United States take medications for
symptomatic relief of pain, insomnia and/or depression. Ethical
considerations precluded requiring patients to be completely drug-
free for the study. Homeopaths claim that, unlike other dosing
methods in their field, LM potencies can be given daily for
extended periods and can overcome the presumptive antidoting
effects of conventional drugs [10]. Second, approximately half of
fibromyalgia patients reportedly have co-morbid multiple chemical
sensitivity, including chemical intolerance [11], a condition that
involves reportedly hypersensitive, adverse polysymptomatic reac-
tions to multiple different environmental chemicals, many pre-
scription and over-the-counter drugs and even homeopathic
remedies [12]. LM potencies in homeopathy are touted to lessen
the risk of symptom flares and afford the option of flexible dose
adjustment as needed by the individual patient [10].
Upon baseline enrolment and at 3 months, all patients
completed a set of questionnaires, underwent conventional
medical history and physical examination for tender point pain
rating status by a conventional provider not involved in their
clinical care (rheumatologist or physician’s assistant; the same
individual saw a given patient at baseline and follow-up), and
had laboratory recordings of electroencephalographic (EEG)
and electrocardiographic responses to double-blind olfactory-
administered test doses of their treatment solution and solvent
controls [13, 14].
Classical homeopathic treatment requires selection of a single
homeopathic medicine (remedy) at a time for a given individual,
based on the broad themes and idiosyncratic nuances of the whole
biopsychosocial clinical presentation. Homeopaths must choose
one from over 1300 different possible remedies in the Homeopathic
Pharmacopoeia of the United States (www.hpus.com), though
typically supported now by computer software programs to assist
in narrowing the choices. A major methodological concern of the
European Commission Homeopathic Medical Research Group
consensus panel who reviewed previous clinical trials was the
strong possibility that some remedies in the ‘active’ treatment
group may be incorrectly chosen, especially in a short-term study,
thereby unintentionally placing an unknown subset of the ‘active’
patients on clinically inactive treatment, i.e. essentially a placebo
[15]. Under the latter circumstances, a negative finding of no
apparent difference between ‘active’ and placebo treatment groups
could result from either a true lack of active treatment effects or
simply inaccurate prescribing by the homeopath.
To minimize the latter risk, two experienced homeopaths jointly
interviewed every patient at each visit and had to agree on a
remedy selection with a confidence rating of at least 7 out of 10 for
the patient to enrol. All of the homeopaths in the present study had
similar training in classical homeopathy, at least 5 years experience
in practice, and certification by the Council for Homeopathic
Certification and/or Diplomate in Homeotherapeutics from the
American Board of Homeotherapeutics. Study homeopaths
used widely available homeopathic software programs as part
of their case analyses (MacRepertory and ReferenceWorks,
Kent Homeopathic Associates, Inc., San Rafael, CA, USA and
Cara-Pro, Miccant Ltd, Nottingham, UK). The study was
approved by the Institutional Review Board of the University of
Arizona, which adheres to relevant United States Federal guide-
lines and the Declaration of Helsinki for human subject involve-
ment in research studies. All patients gave written informed
consent for their participation.
Recruitment of participants
Volunteer non-pregnant female and male patients with fibro-
myalgia were recruited from the greater Tucson and Phoenix
communities by media announcements, newspaper advertisements,
flyers in local health-food stores and word-of-mouth in patient
support organizations. Prospective patients had to report a prior
physician diagnosis of fibromyalgia, stable conventional medica-
tion doses for at least 2 months prior to enrolment (steroid drugs
were an exclusion criterion), score to criteria for fibromyalgia on a
15-item, 4-point Likert symptom screening questionnaire and
have their fibromyalgia diagnosis confirmed on rheumatological
physical examination using the 1990 American College of
Rheumatology criteria [16]. Two patients with physician diagnoses
of fibromyalgia, with random assignments to placebo, had fewer
than 11/18 positive tender points on initial examination, but both
met the diagnostic cut-off on the second rheumatological exam-
ination. All prospective participants underwent a semi-structured
clinical interview for psychiatric and substance abuse disorders
prior to enrolment.
To minimize confounds in the psychophysiological component
of the study, patients could not have a history of alcohol or drug
abuse, current narcotic analgesic, benzodiazepine or antihyperten-
sive medication use or nasal trauma. For patient safety, anaphy-
laxis history, diabetes, serious neurological, heart, lung, liver or
kidney disease, psychosis and active suicidality were also exclusion
criteria.
Treatment, blinding and randomization
After each visit, the homeopathic office sent a fax to Hahnemann
Laboratories, San Rafael, CA, with current remedy selection and
dose prescription. Homeopaths were instructed to treat each
participant as if they were receiving active treatment; they were
permitted to change remedy prescriptions and potencies at any visit
or between visits if clinically indicated. Hahnemann Laboratories
dispensed a 16 oz glass bottle monthly (or as needed) of either the
active liquid homeopathic remedy in the prescribed LM potency or
placebo. All bottles contained the same amount of 20% alcohol-
distilled water solvent. Patients began the study on LM 1 potency.
The active and placebo bottles were indistinguishable and were
all labelled with date, subject number and bottle number [all
patients received bottles in order LM 1, LM 2, LM 3, etc., where
LM 2 ¼ (1/50 000)
2
dilution factor].
Treatment bottles were mailed directly from the pharmacy
to each patient, with a split sample bottle of the same material
mailed directly to the local research pharmacist. Contents of the
bottles were filled in accord with a randomized assignment
in blocks of six to either active or placebo group, generated by
www.randomizer.org. The randomization was recorded by the
study methodologist (AJB), who sent the sequence to the
pharmacist at the start of the study. Only the methodologist in
Tucson and Hahnemann Laboratories’ pharmacist in California
had access to the randomization code during the study. The
methodologist was available to break the code of individual
patients for emergency clinical intervention. This type of situation
occurred in only one patient, who dropped out of the study
because of her concern about perceived worsening emotional and
physical symptoms and a request to the principal investigator for
578 I. R. Bell et al.
immediate open label treatment under her own physician’s care.
This individual turned out to be assigned to placebo. All clinicians
and research staff interacting with and assessing patients were kept
blinded as to group assignments, including dropouts, for the
duration of the study.
Measures
At baseline and 3 months, all patients completed an expectation
rating of benefit from treatment, the McGill Pain Questionnaire
(short form) [17], Appraisal of Fibromyalgia quality of life scale
[18], global self-rated health scale (5-point Likert ratings of current
health, health compared with peers, health compared with 6 months
ago) [19], and Profile of Mood States (POMS) scale (Educational
and Industrial Testing Service, San Diego, CA). Symptom criteria
for a chronic fatigue syndrome diagnosis and Bell Chemical
Intolerance Index [20] ratings were obtained at baseline. On
follow-up visits, patients completed the Patient Satisfaction Scale
regarding the homeopaths involved in their treatment [21] and a
0- to10-point Likert scale on helpfulness of the treatment.
Analysis
This study was designed as a feasibility or pilot study rather than a
definitive clinical trial, with adequate power planned to detect a
large effect in the outcome variable likely to be most sensitive, i.e.
tender point pain on palpation (a type of ‘stress test’ of pain
reactivity, as opposed to a pain rating on a standardized
questionnaire, such as the McGill Pain Scale, completed while at
rest). The previous fibromyalgia study [6] was performed within
subjects, with a total sample size of 30; it did not specify dropout
rate, standard deviations or confidence intervals to permit
statistical power analysis. Since a fairly large effect size (d) was
likely to be clinically important, we used an estimate that was large
for planning purposes. With d ¼ 0.8, assuming a dropout rate of
approximately 15% and ¼ 0.05, two-tailed, a sample size of 30 per
group enrolled would yield a statistical power of 0.8 for the tender
point pain outcome (nQuery Advisor 1997).
We compared active and placebo groups with one-way analyses
of variance and
2
tests for differences in baseline demographics
and clinical status. For subsequent analyses of covariance (SPSS
version 11.0, Chicago, IL, USA: GLM procedures), we used
baseline values of a given outcome variable and variables on which
the groups differed at a P < 0.10 level or better as covariates
(despite randomization). Primary outcome variables included
tender point count, mean tender point pain on palpation, McGill
Affective and Sensory Pain Ratings and Appraisal of Fibromyalgia
score. Secondary outcome variables were changes in POMS fatigue
and depression subscales and global health self-ratings. Groups
were compared using general linear model statistics, first without
and then adjusted with appropriate covariates as detailed above,
including follow-up scores for the active and placebo groups.
Intent to treat (ITT) analyses were conducted for treatment
completers (those with 3-month follow-up data) and for the full
randomized sample, using mean substitution for values of isolated
scale items missing at random and last observation carried forward
(i.e. baseline value) for 3-month values of dropouts. Analysis of the
ITT treatment completer and ITT last observation carried forward
datasets produced the same results. Many statisticians disagree
with use of last observation carried forward to generate an ITT
dataset when a subject has only a baseline value [22]. Thus, only
ITT results for all patients with 3-month follow-up data are shown.
Data were analysed with and without the two individuals who
carried a physician diagnosis of fibromyalgia but whose tender
point counts were initially below criterion. The main findings
remained when these individuals were excluded; conse-
quently, results are reported with all subjects included. Statistical
significance was set at P < 0.10 to examine for trends.
Results
Sample characteristics
After telephone screening, 90 fibromyalgia patients were judged
potentially eligible for the study (Fig. 1). From those patients,
62 were randomized, meeting homeopathic agreement for remedy
selection. Persons who chose not to participate typically cited
FIG. 1. Patients entered, randomized and withdrawn from the study.
Individualized homeopathy in fibromyalgia 579
reluctance to make the required trips between Tucson and Phoenix
or unwillingness to complete the extensive questionnaire and
laboratory components of the study.
Active and placebo groups did not differ in demographic
characteristics (Table 1), duration of fibromyalgia, chronic fatigue
syndrome diagnostic criteria, chemical intolerance index scores,
baseline POMS fatigue scores, global ratings of health or
expectation ratings of possible benefit from treatment. Groups
had the same number of tender points, but there was a trend for the
active group to have more tender point pain on palpation
examination at baseline. The active group was significantly more
depressed and angry–hostile on the POMS and used more
antihistamine and/or expectorant drugs than did the placebo
group (Table 1). Thus, POMS depression and anger–hostility as
well as baseline values of the relevant outcome variable were
covariates in analyses comparing active and placebo group
outcomes. Homeopathic remedy choices over the whole sample
were highly individualized to the same degree in both groups
(homeopaths prescribed 41 different remedies for 62 participants)
(supplementary data, Table 3). Only two remedies, Calcarea
carbonica and Rhus toxicodendron, each were chosen for four
patients.
Treatment outcomes
A total of 53 patients completed the 4 months of the study to the
point of optional crossover (14.5% dropout rate). Although the
study requirements had been explained thoroughly prior to
enrolment, the primary reasons for the nine dropouts nonetheless
related to time and travel demands of the study, or exces-
sive experience of scalp pain during EEG laboratory hook-up
procedures. Dropout rates and baseline patient demographic
characteristics of dropouts did not differ between active and
placebo groups. No patient reported an adverse drug reaction to a
treatment solution as a reason for dropping out. The 3-month
ratings on the Patient Satisfaction Scale for the homeopaths did
not differ between groups. Both groups progressed comparably in
LM doses (mean LM dose 2.4,
S.D. 0.9 at follow-up). However,
consistent with the homeopaths’ possible perception of a lack of
expected improvements over time and consequent decisions to
change remedy selections for placebo-treated patients, the average
number of remedies recommended by the homeopaths was
significantly higher in the placebo group (mean 1.7,
S.D. 0.7) than
in the active treatment group (mean 1.3,
S.D. 0.5) [F(1,60) ¼ 5.5,
P ¼ 0.023].
For treatment completers, Table 2 shows that the active group
exhibited a significantly greater improvement in tender point count
and tender point pain on palpation, Appraisal of Fibromyalgia
scores and global health ratings, with trends toward lower POMS
depression, POMS anger–hostility and McGill Affective Pain
scores compared with placebo at 3 months. McGill Sensory Pain
ratings did not differ significantly between groups at 3 months.
A significantly higher proportion of patients in the active group
experienced at least a 25% improvement in tender point pain on
examination (13/26, 50%) versus placebo (4/27, 15%) (Fisher’s
exact test, two-tailed, P ¼ 0.008). At the 4-month homeopathic
visit, patients on active rated the helpfulness of the treatment (7.8,
S.E. 0.6) significantly greater than did those on placebo (5.3, S.E. 0.5)
(P ¼ 0.004).
Discussion
The findings demonstrate that the active group on individualized
homeopathy showed a greater reduction in tender point count and
TABLE 1. Baseline descriptive characteristics of participant sample as randomized, means (SD) unless otherwise stated
Individualized homeopathy
(n ¼ 30)
Placebo
(n ¼ 32)
Age (yrs) 49.1 (9.9) 47.9 (10.8)
Number of women 29 (97%) 29 (91%)
Ethnicity (no. white) 24 (80%) 29 (91%)
Marital status (no. married) 18 (60%) 20 (63%)
Education (no. with some college or more) 25 (83%) 29 (91%)
Duration of fibromyalgia (yr) 14.8 (14.0) 11.9 (11.4)
Meet Chronic Fatigue Syndrome Diagnostic Criteria (no. with CFS) 25 (83%) 28 (88%)
Bell Chemical Intolerance Score 7.3 (2.6) 7.0 (3.1)
Severity of illness—baseline clinical global impression (0–7) (rheumatologist) 2.7 (0.8) 2.8 (0.6)
Severity of illness—baseline clinical global impression (0–7) (homeopaths’ ave.) 4.0 (0.7) 4.1 (0.7)
Patient expectation of benefit from treatment (0–10) 8.1 (1.9) 8.5 (1.8)
Rheumatologist expectation of benefit from treatment (0–10) 3.8 (1.6) 4.0 (1.4)
Ave. homeopath expectation of benefit from treatment (0–10) 6.8 (1.4) 6.8 (0.98)
Non-narcotic pain medications 18 (60%) 17 (53%)
Serotonin re-uptake inhibitor drugs 7 (23%) 5 (16%)
Muscle relaxant drugs 3 (10%) 5 (16%)
Antihistamine or expectorant use* 10 (33%) 0
Individualization ratio of initial homeopathic remedies (unique no. chosen/no. patients) 24/30 (0.80) 25/32 (0.78)
Tender point count (0–18) 16.8 (1.8) 16.4 (2.6)
Tender point pain on palpation exam (0–180)
a
97.7 (35.0) 82.0 (33.1)
McGill Affective Pain (0–12) 4.2 (2.4) 4.2 (2.8)
McGill Sensory Pain (0–33) 15.6 (5.5) 16.2 (6.2)
Appraisal of fibromyalgia (7–35) 22.4 (5.3) 21.4 (5.0)
POMS fatigue (0–28) 12.1 (7.7) 14.1 (6.6)
POMS depression (0–60)
b
9.5 (12.3) 4.6 (5.1)
POMS anger–hostility (0–48)
c
5.0 (7.3) 1.9 (3.3)
Global Health Rating (3–15) 7.1 (2.3) 7.3 (2.9)
*P<0.001.
a
Main effect for group at baseline, P ¼ 0.08.
b
Main effect for group at baseline, P ¼ 0.04.
c
Main effect for group at baseline, P ¼ 0.03.
580 I. R. Bell et al.
tender point pain, better fibromyalgia-related quality of life,
improved global health and a trend toward less affective dis-
turbance. Notably, Jensen et al. [23] previously found that myalgic
pain ratings on palpation were a better indicator of fibromyalgia-
related disability than tender point count. Other less sensitive
outcome measures such as the McGill Pain Scale short-form did
not reach significance at P < 0.05 with the present sample size.
Although regression to the mean might account for some of the
apparent improvement in the active group [24], the improved
status of the active group compared with the placebo group at
3 months for tender point pain, tender point count, global health
and fibromyalgia-related quality of life (Appraisal of Fibromyalgia
Scale) remained after covarying for the baseline value of the
relevant dependent variable, as well as baseline differences in
depression and anger–hostility. These data constitute a replication
and extension of the earlier study by Fisher et al. [6] showing
individualized homeopathic treatment superior to placebo in the
treatment of fibromyalgia.
The strengths of the current study include a longer duration of
treatment than in the previous fibromyalgia study [6] (3 months
versus 1 month), enrolment of persons needing a wide range of
different individualized remedies rather than only one (for fidelity
to typical homeopathic practice), requirement for agreement of
two homeopaths on each remedy selection with high confidence
(thereby limiting concerns that the active group could have
received non-active treatment), use of daily, flexibly dosed LM
potencies to obviate homeopathic methodological concerns from
prior studies such as remedy antidoting or aggravations, and
inclusion of continuous rather than categorical outcome variables
for sensitivity to change.
Weaknesses of the present study include a comparatively small
group sample size, providing adequate power for detecting change
primarily in tender point pain but not necessarily other outcome
measures, and lack of objective measures directly related to
fibromyalgia status (none are available in this field). In view of
the travel and laboratory session demands, some loss of data from
drop-outs might have been avoided by pursuing relevant follow-up
outcome measures at times separate from those of the laboratory
sessions. Nonetheless, the findings were robust for changes in
tender point pain, and other types of objective measures, i.e. EEG
variables during olfactory laboratory administration of the
homeopathic remedies, did differentiate active from placebo
treatment and exceptional clinical responders from all other
participants [13, 14].
The most marked divergence between active and placebo treated
groups occurred in the pain variable involving central nervous
system activation or evocation with stimuli (pressure on tender
points), the main variable for which the study was properly
powered to avoid Type II error. Convergent evidence identifies the
central nervous system as a key mediator of the pain in
fibromyalgia [25]. The reductions in tender point pain on
examination were clinically meaningful, and, together with the
associated changes in EEG alpha cordance (derivative of absolute
and relative EEG that correlates with functional neuroimaging
scans) in exceptional clinical responders observed in this study [14],
raise the possibility of remedy-related attenuation in central
processing of painful stimuli. Consistent with homeopathic
theories of healing [26], the active remedy group tended to become
less, while the placebo group became more, depressed, in addition
to the changes in the physical pathology (though overall depression
levels were fairly low at baseline). Other outcome variables
were statistically significant, but appear less significant in
magnitude clinically. Within homeopathic thinking, however,
the remedy is not chosen for the diagnosis of ‘fibromyalgia’, but
for the unique person who has the fibromyalgia [26]. Con-
sequently, individualized homeopathy is expected clinically to
mobilize changes in multiple domains [27], in some cases leading
to gradual improvements in other aspects of health before
changes in pain [28].
This is the second study in which homeopathy performed better
than placebo in treating patients with fibromyalgia [6]. Given
the lack of definitive conventional treatments for fibromyalgia, the
lack of improvement in pain over the natural history of the
condition [29] and the high rates of utilization of complementary
medicine by fibromyalgia patients [30], homeopathy emerges as a
potentially low-risk, evidence-based option in an integrated
package of care. Homeopaths claim that patients need at least
1 month of active treatment for every year of illness. With that
reasoning, the present sample would have required a 12-month,
not a 3 to 4 month, trial to assess optimal benefits. In the double-
blind optional crossover phase of this study, persons who stayed
with active and placebo group assignments for the full 6 months
maintained their divergence on the outcome variables [9].
Well-designed randomized controlled trials on larger samples, for
longer periods of time, are now indicated, especially in view of
emerging basic scientific evidence that homeopathic remedies
have physical–chemical properties that differ from those of
placebo [31–33].
Supplementary Data
Supplementary data are available at Rheumatology online.
TABLE 2. Outcomes after 3 months (active n ¼ 26; placebo n ¼ 27). Means (standard deviation, S.D.) for actual follow-up values and mean group
differences for 3-month follow-up scores (95% confidence interval, CI), unadjusted and covariate adjusted, are shown, using the SPSS GLM
UNIANOVA statistical procedure. Adjusted values reflect analysis of covariance using the SPSS GLM statistical procedure, covaried for baseline value
of each dependent measure, baseline POMS depression and POMS anger–hostility scores. (Significant differences between active and placebo for
adjusted values in follow-up scores also remained significant after re-analysing the data without the 10 patients on active/Verum who reported baseline
use of antihistamine or expectorant drugs)
Mean (
S.D.)
follow-up
(active)
Mean (S.D.)
follow-up
(placebo)
Unadjusted differences
in follow-up scores
(95% CI)
(active – placebo)
Adjusted differences
in follow-up
scores (95% CI)
(active – placebo)
Tender point count (0–18) 14.8 (3.9) 16.1 (2.7) –1.3 (–3.2 to 0.56) –1.9 (–3.5 to –0.24)**
Tender point pain on palpation exam (0–180) 71.3 (36.3) 82.8 (36.0) –11.0 (–31.0 to 8.9) –22.6 (–38.3 to –6.9)***
McGill Affective Pain (0–12) 3.3 (2.9) 3.5 (2.7) –0.14 (–1.7 to 1.4) –1.0 (–2.2 to 0.16)*
McGill Sensory Pain (0–33) 12.9 (7.4) 12.4 (6.9) 0.48 (–3.6 to 4.5) –1.2 (–4.1 to 1.7)
Appraisal of fibromyalgia (7–35) 19.2 (5.7) 19.9 (5.3) –0.62 (–3.6 to 2.4) –2.1 (–4.0 to –0.28)**
POMS fatigue (0–28) 10.0 (7.0) 13.4 (8.1) –3.4 (–7.6 to 0.73) –2.9 (–6.6 to 0.88)
POMS depression (0–60) 7.3 (9.5) 8.1 (10.4) –0.82 (–6.3 to 4.7) –4.4 (–8.8 to 0.06)*
POMS anger–hostility (0–48) 2.9 (4.2) 3.7 (6.5) –0.74 (–3.8 to 2.3) –2.4 (–5.1 to 0.34)*
Global Health Rating (3–15) 8.2 (2.9) 7.7 (3.0) 0.47 (–1.2 to 2.1) 1.5 (0.14 to 2.8)**
*P 0.10, **P < 0.05, ***P < 0.01.
Individualized homeopathy in fibromyalgia 581
The authors have declared no conflicts of interest.
Acknowledgements
We thank the patients who participated; Mary Gorman, PA,
who performed some of the rheumatology examinations; the
homeopaths who provided the treatment—Todd Rowe, MD,
Edward Kondrot, MD, Yolande Grill, HMA (all of the Desert
Institute of Classical Homeopathy, Phoenix, AZ); Michael
Quinn and colleagues of Hahnemann Laboratories (San Rafael,
CA) for implementing the randomization procedures, blinding
and dispensing of the homeopathic medicines and placebo;
Deborah Noah HMA and Nancy Tichenor RN HMA for
coordination of homeopathic patient visits; and Diana
Christeson for pharmacy preparation of the sniff bottles.
This study was supported by NIH grants R21 AT00315 (IRB),
K24 AT00057 (IRB), P20 AT00774 (GES), P50 AT00008 from
the National Institutes of Health National Center for
Complementary and Alternative Medicine (NCCAM) and NIH
HL53938–07S1 (CMB). Its contents are solely the responsibility
of the authors and do not necessarily represent the official views
of NCCAM or NIH.
References
1. Thomas KJ, Nicholl JP, Coleman P. Use and expenditure on
complementary medicine in England: a population-based survey.
Complement Ther Med 2001;9:2–11.
2. Eisenberg DM, Davis RB, Ettner SL et al. Trends in alternative
medicine use in the US, 1990–1997. Results of a follow-up national
survey. J Am Med Assoc 1998;280:1569–75.
3. Goldstein MS, Glik D. Use of and satisfaction with homeopathy in a
patient population. Alt Ther Health Med 1998;4:60–5.
4. Friedberg F, Jason LA. Chronic fatigue syndrome and fibromyalgia:
clinical assessment and treatment. J Clin Psychol 2001;57:433–55.
5. Lawrence RC, Helmick CG, Arnett FC et al. Estimates of the
prevalence of arthritis and selected musculoskeletal disorders in the
United States [comment]. Arthritis Rheum 1998;41:778–99.
6. Fisher P, Greenwood A, Huskisson EC, Turner P, Belon P. Effect of
homeopathic treatment on fibrositis (primary fibromyalgia). Br Med J
1989;299:365–6.
7. Linde K, Clausius N, Ramirez G et al. Are the clinical effects of
homeopathy placebo effects? A meta-analysis of placebo-controlled
trials. Lancet 1997;350:834–43.
8. Vickers AJ. Clinical trials of homeopathy and placebo: analysis of a
scientific debate. J Alt Complement Med 2000;6:49–56.
9. Bell IR, Lewis DA II, Brooks AJ, Schwartz GE, Lewis SE, Caspi O,
Cunningham V, Baldwin CM. Individual differences in response to
randomly-assigned active individualized homeopathic and placebo
treatment in fibromyalgia: implications of a double-blind optional
crossover design. J Altern Complement Med 2004; in press.
10. De Schepper L. LM potencies: one of the hidden treasures of the sixth
edition of the Organon. Br Homoeopath J 1999;88:128–34.
11. Slotkoff AT, Radulovic DA, Clauw DJ. The relationship between
fibromyalgia and the multiple chemical sensitivity syndrome. Scand J
Rheumatol 1997;26:364–7.
12. Sherr J. The Dynamics and Methodology of Homeopathic Provings,
2nd edn. Malvern: Dynamis Books, 1994.
13. Bell IR, Lewis DA II, Lewis SE, Schwartz GE, Scott A, Brooks AJ,
Baldwin CM. Electroencephalographic alpha sensitization in indi-
vidualized homeopathic treatment of fibromyalgia. Int J Neurosci
2004; in press.
14. Bell IR, Lewis DA II, Schwartz GE, Lewis SE, Caspi O, Scott A,
Brooks AJ, Baldwin CM. Electroencephalographic cordance patterns
distinguish exceptional clinical responders with fibromyalgia to
individualized homeopathic medicines. J Altern Complement Med
2004; in press.
15. Kron M, English JM, Gaus W. Guidelines on methodology of clinical
research in homeopathy. In: Ernst E, Hahn EG (eds). Homeopathy: a
Critical Appraisal. Oxford: Butterworth-Heinemann, 1998: 9–47.
16. Wolfe F, Smythe HA, Yunus MB et al. The American College of
Rheumatology 1990 criteria for the classification of fibromyalgia:
report of the multicenter criteria committee. Arthritis Rheum
1990;33:160–72.
17. Melzack R. The short-form McGill Pain Questionnaire. Pain
1987;30:191–7.
18. Walker EA, Keegan D, Gardner G, Sullivan M, Katon WJ, Bernstein
D. Psychosocial factors in fibromyalgia compared with rheumatoid
arthritis: I. Psychiatric diagnoses and functional disability. Psychosom
Med 1997;59:565–71.
19. Bell IR, Warg-Damiani L, Baldwin CM, Walsh M, Schwartz GE.
Self-reported chemical sensitivity and wartime chemical exposures in
Gulf War veterans with and without decreased global health ratings.
Mil Med 1998;163:725–732.
20. Szarek MJ, Bell IR, Schwartz GE. Validation of a brief screening
measure of environmental chemical sensitivity: the chemical odor
intolerance index. J Environ Psychol 1997;17:345–51.
21. DiMatteo MR, Hays R. The significance of patients’ perceptions of
physician conduct: a study of patient satisfaction in a family practice
center. J Commun Health 1980;6:18–34.
22. Lavori PW. Clinical trials in psychiatry: should protocol deviation
censor patient data? Neuropsychopharmacology 1992;6:39–48.
23. Jensen B, Wittrup IH, Rogind H, Danneskiold-Samsoe B, Bliddal H.
Correlation between tender points and the Fibromyalgia Impact
Questionnaire. J Musculoskelet Pain 2000;8:19–29.
24. Vickers A, Altman DG. Analysing controlled trials with baseline and
follow up measurements. Br Med J 2001;323:1123–4.
25. Stevens A, Batra A, Kotter I, Bartels M, Schwarz J. Both pain and
EEG response to cold pressor stimulation occurs faster in fibromyal-
gia patients than in control subjects. Psychiat Res 2000;97:237–47.
26. Vithoulkas G. The Science of Homeopathy. New York: Grove
Weidenfeld, 1980.
27. Bell IR. Evidence-based homeopathy: empirical questions and
methodological considerations for homeopathic clinical research.
Am J Homeopath Med 2003;96:17–31.
28. Bell IR, Koithan M, Gorman MM, Baldwin CM. Homeopathic
practitioner views of changes in patients undergoing constitutional
treatment for chronic disease. J Alt Complement Med 2003;9:39–50.
29. Baumgartner E, Finckh A, Cedraschi C, Vischer TL. A six year
prospective study of a cohort of patients with fibromyalgia. Ann
Rheum Dis 2002;61:644–5.
30. Pioro-Boisset M, Esdaile JM, Fitzcharles MA. Alternative medicine
use in fibromyalgia syndrome. Arthritis Care Res 1996;9:13–17.
31. Rey L. Thermoluminescence of ultra-high dilutions of lithium chloride
and sodium chloride. Physica A 2003;323:67–74.
32. Bell IR, Lewis D, Brooks AJ, Lewis S, Schwartz GE. Gas discharge
visualization evaluation of ultramolecular doses of homeopathic
medicines under blinded, controlled conditions. J Alt Complement
Med 2003;9:25–38.
33. Elia V, Niccoli M. Thermodynamics of extremely diluted aqueous
solutions. Ann NY Acad Sci 1999;879:241–8.
Rheumatology
Key messages
Individualized homeopathy has efficacy
in treatment of fibromyalgia.
Daily LM potencies minimize methodo-
logical concerns about antidoting homeo-
pathic remedies.
To avoid Type II error, homeopathy
trials must evaluate both disease-specific
and global outcomes.
582 I. R. Bell et al.
