Essentials of Complementary and Alternative Medicine (June 1999)

CHAPTER 8. T
HE
 S
AFETY OF
 H
OMEOPATHY
Wayne B. Jonas and Edzard Ernst
Introduction
Risks
 
Direct Risks
 
Indirect Risks
 
Illness Fabrication and Misclassification
Conclusions
Chapter References
INTRODUCTION
Homeopathy involves the application of extremely small doses of substances diluted in a serial fashion and succussed, or agitated, between dilutions. These 
preparations are administered to healthy individuals to illicit symptoms to produce “drug pictures,” which are then matched to those with similar symptoms who are ill.
Originally, the founder of homeopathy, Samuel Hahnemann (1755–1843), began diluting the drugs of his era to reduce their toxic side effects. He and others claimed 
that when these drugs were carefully matched to patients, these serial-agitated dilutions (SADs) lost their toxic effects, yet still retained stimulatory and therapeutic 
effects. It was subsequently discovered that many homeopathic preparations no longer have the possibility of any original molecules left in them. The assumptions 
from this are that they can have no specific effects at all; they are either therapeutic or toxic, and therefore are at least harmless. This attitude is reflected in the 
approach taken by the Food and Drug Administration of the United States, which generally classifies homeopathic preparations as over-the-counter drugs approved 
for sale without the standard toxicity and safety testing required of other medications.
Recent evidence indicates that homeopathic medications may not work in identical fashion to placebo (
1
, 
2
). If these claims are substantiated and homeopathic 
remedies are thought to produce specific effects, then the possibility that they also may produce specific adverse effects exists. In this regard, their evaluation will 
require the same assessment of risk-benefit ratio as any other drug.
RISKS
Direct Risks
Direct risks refer to direct toxic effects from a drug itself or adverse reactive effects to a drug. Homeopathic drug pictures are produced by a process called “provings,” 
in which homeopathic preparations are given to healthy individuals and the symptoms produced are recorded. This is analogous to a Phase I trial and assumes direct 
adverse effects from stimulation by homeopathic remedies. Little animal research has been done to test whether such adverse effects occur in an objective way. 
However, when this has been done, it has not indicated an innocuous nature assumed from low dilutions. Here are some examples:
Stearnes, in 1928, examined the effects of low dilutions of  Naturum muriticum on several generations of guinea pigs. He reported that, compared to a control 
group, guinea pigs given repeated doses of low dilutions of  Naturum muriticum became unhealthy, lost their hair, were less alert, lost weight, and began to stop 
breeding (
3
).
Dwarakanath reported a study in 1979 in which the homeopathy remedy China, in the 200 C and 1000 C potency, produced acute reductions in body 
temperature measured rectally in rats compared to a control group (
4
).
Chandrasekhar reported that high potencies of  Pulsatilla M and 10 M (very low dilutions) administered to rats immediately before and after mating induced 
lymphocytic infiltration of the implanted fetus and a higher frequency of fetal reabsorption compared to controls (
5
).
Kumar reported that high potencies of Caulophytum retarded uterine ovarian cell maturation and stimulated endometrial proliferation in rats. Similar results have 
been reported in other animal experiments (
6
).
These studies are few and often of low methodological quality. Still, they indicate that one cannot assume that high potencies of homeopathic remedies have no direct 
toxic effects. This brings into question the nontoxicity homeopathy in which the perceived minimum toxic dose in a potency is assumed to be safe under all conditions.
Under normal use, homeopathic remedies appear too diluted to bring about toxic effects. This is not true for all dilutions, however. When toxic metals are used as 
homeopathic remedies, unsafe concentrations of arsenic (
7
), cadmium (
8
), and mercury (
9
) can result from the use of low potencies. For example, regular doses of 
Cadmium sulfuricum D
3
 would exceed the daily tolerable dose of cadmium in a human of average size by two- to fourfold. There has been concern about potentially 
carcinogenic effects of low homeopathic potencies of  Aristolochia (
10
). Low potencies of most homeopathic remedies can also cause allergic reactions. Several such 
adverse effects have been reported (
11
, 
12
). In cases of concomitant drug treatments, interactions with homeopathic remedies are conceivable, although as yet there 
seems to be no documented evidence of this. Finally, homeopathic remedies have on occasion been adulterated with potent allopathic medicaments (
13
, 
14
).
Adverse effects from homeopathic medications are assumed to be rare. In a survey of 386 users of complementary medicine, 10% of those who employed 
homeopathy had experienced adverse effects at some time (
15
). Reilly, in a systematic study of high dilutions of allergens, showed that approximately 24% of patients 
reported adverse aggravations. This resulted in about a 6% dropout rate from the study due to such adverse effects (
16
). Until reliable prevalence figures exist, we 
cannot conclude that homeopathic remedies are entirely risk-free.
H
OMEOPATHIC
 A
GGRAVATION
The term homeopathic “aggravation” describes the concept that symptoms may get initially worse if the optimal remedy has been administered. Homeopathic 
physicians see aggravations as a positive sign on the correct route to recovery. Patients, however, may view aggravation as an adverse effect. In our survey (
15
), 
about half of the adverse effects reported by users of homeopathy were associated with aggravation of symptoms. Sensitive patients are said to often produce severe 
aggravations. One of the authors (WJ) has seen a case of sudden severe aggravation of asthma necessitating hospitalization from homeopathic treatment. A recent 
placebo-controlled trial of homeopathy showed that homeopathic aggravation happened in the placebo group roughly as frequently as in the verum group (
17
), 
casting doubts on the existence of the phenomenon.
Homeopathic aggravation is a complex issue. Until more evidence is available to demonstrate the existence of homeopathic aggravation, it may be best to view it as a 
potential for adverse events.
The fact that homeopaths believe it does may significantly increase the risk of delaying mainstream treatments. Thus, the issue of homeopathic aggravation could be 
relevant in terms of both direct and indirect risks of homeopathy.
D
OSE
-D
EPENDENT
 R
EVERSE
 E
FFECTS
Dose-dependent reverse effects may also be an example of indirect effects from homeopathic preparations. Doutremepuich has demonstrated in several models that 
serially agitated preparations of aspirin can induce rather than inhibit thrombosis (
18
). Animal studies have shown that high potencies of arsenic can accelerate the 
elimination of arsenic when subsequently given in toxic dose (
19
, 
20
). One must consider whether homeopathic preparations of essential minerals and nutrients might 
also stimulate the elimination of such nutrients. If dose-dependent reverse effects occur for other substances, one wonders whether cytotoxic drugs, when given in 

homeopathic preparations, might stimulate cellular proliferation and carcinogenesis rather than inhibit it.
Indirect Risks
A second category of adverse effects in homeopathy relates to indirect effects. If an effective therapy exists, then treatment with ineffective therapy, be it homeopathy 
or conventional treatment, may result in unnecessary progression of the disease, which is an indirect adverse effect. These are “neglect” effects. An extreme example 
would be an attempt to use homeopathy in place of setting a fracture or getting emergency care. More subtle “neglect” effects can also occur, however. Comparing a 
drug to placebo is insufficient to assess the difference in adverse effects between two active drug treatments. A drug response over one placebo may not be the same 
as another drug over another placebo. Only direct comparison of two interventions gives information about optimal therapies, and thus the degree of “neglect” effects 
that occur through the administration of a suboptimal therapy.
Other indirect effects in homeopathy can occur because of its philosophy of healing. Homeopathy stimulates the healing response, which may be slow in some 
patients.
For example, some homeopaths claim there is a duration of action from particular potencies, even up to a year after a single dose. One of the authors (WJ) has seen 
cases where individuals with chronic illness, such as gingivitis and gall bladder disease, have been told to wait for the full duration of action of the remedy, often for 
very long periods, resulting in continued suffering. Homeopaths also claim that the return of old symptoms is a good sign during the action of the remedy, but for most 
individuals this is not desirable. These symptoms would also be classified as indirect adverse effects.
Symptom control without resolution of the underlying problem is another indirect adverse effect. The only evidence for this in homeopathy comes from case reports. 
For example:
A woman with staples inadvertently left in the skin of her scalp postoperatively was taking the homeopathic remedy  Hypericum to treat the pain. This resulted in 
the staples being undiscovered for longer than would have without such treatment.
A woman was treating what she thought were menstrual cramps with a homeopathic remedy. Later it was discovered that she had a ruptured ovarian cyst and 
was bleeding internally.
These adverse effects, of course, can occur with any kind of misdiagnosis and symptom control. There are certain conditions, such as acute myocardial infarction, 
fractures, bacterial meningitis, and so on, which require immediate conventional attention for managing. The attempt to use homeopathic medications in these 
conditions can result in adverse outcomes.
Even if homeopathic remedies were totally safe, the homeopath might not always be. In all professions, there should be proper training and competence for scope of 
practice (
21
). If a homeopathic practitioner takes over full medical responsibility for a patient, he or she should be medically competent to dose, including full 
diagnostic and management skills. If competence is insufficient, disasters may occur. A review of the medical literature indicates that these indirect risks are not 
merely an academic issue. There are distressing case reports, some with fatal outcomes (
22
, 
23
 and 
24
).
In this context, vaccination is a particularly relevant and clear example (
25
, 
26
). Some homeopaths (as well as some chiropractors and naturopaths) hinder access to 
immunization by advocating a homeopathic vaccination (
27
), which is not documented to be effective (
28
), or by advising their clients not to comply with immunization 
programs. Our own survey shows that none of the nonmedically qualified homeopaths recommended orthodox vaccination, whereas the majority of 
physician–homeopaths do (
29
). Certainly, there can be adverse effects from immunizations and these vary depending on the type of immunization. In view of their 
potential benefit, however, a general rejection of immunization puts the individual patient and the population at large at risk.
Illness Fabrication and Misclassification
Adverse effects in homeopathy also can arise from misclassification through the creation of false-positive syndromes. The classical homeopathic approach requires 
assessing the “totality” of symptoms in the assessment and treatment of a case. A homeopath may spend 60 to 90 minutes getting all the symptoms of a patient and 
prescribe a remedy for this “totality.” Six weeks later, the vast majority of these systems are likely to have regressed due to spontaneous remission or statistical 
regression to the mean or other factors. In this situation, the remedy will appear to have been successful, thus giving a false sense of efficacy. Repeated experiences 
such as this can result in a kind of treatment addiction to these false syndromes. The patient returns to have further symptoms assessed and apparently resolved 
through such nonspecific means. This also risks turning the therapeutic interaction into worry management. The patient attempts to resolve all symptoms or reduce all 
risk, resulting in repeated visits to the physician. This then is an adverse effect.
A similar phenomenon may occur when practitioners use electrodermal diagnostic techniques that apply homeopathic remedies for treatments. In this technique, 
electrical potential changes on acupuncture points of the fingers and toes are said to be related to premorbid health problems. These electrodermal changes then 
become diagnostic categories in themselves, requiring treatment with repeated visits and unnecessary medication. This results in the fabrication of illness. The 
assumption is that “treatment” of these electrical potentials will prevent health problems in the future. If patients then neglect well-known lifestyle and other prevention 
activities, such as exercise, appropriate diet, and smoking cessation, they may assume they are protected, when they are not. This may result in adverse effects from 
failure to take concrete steps to prevent illness. A number of other safety issues related to homeopathy are beyond the scope of this chapter but have been reviewed 
in detail elsewhere (
30
).
CONCLUSIONS
The risks of any type of medicine should not be evaluated in isolation. Whenever a practitioner prescribes a treatment, he or she should weigh its risks against its 
benefits. If the former is small, the latter should be minute so that, on balance, the benefits clearly outweigh the risks. Both homeopathy's benefits and its potential 
risks need to be established with a much higher degree of certainty before definitive answers to the complex question of safety can be given. Safety in homeopathy 
cannot be assumed without empiricalal testing and verification. Systematic evaluation of the direct, indirect, misclassification, and paradigmatic issues of homeopathic 
treatment is necessary.
C
HAPTER
 R
EFERENCES
1.
Linde K, Jonas WB, Melchart D, et al. Critical review and meta-analysis of serial agitated dilutions in experimental toxicology. Hum Exper Toxicol 1994;13:481–492.
2.
Linde K, Clausius N, Ramirez G, et al. Are the clinical effects of homeopathy all placebo effects? A meta-analysis of randomized, placebo controlled trials. Lancet 1997;350:934–843.
3.
Stearns G. Completion of the experiment with high dilutions of natrium muriaticum given to guinea pigs. J Am Inst Hom 1925;18(790):790–792.
4.
Dwarkanath SK, Stanley MM. Short term observations on the effect of China in relation to the rectal temperature of albino rats. Hahn Glean 1979;46(37):37–41.
5.
Chandrasekhar K, Sarma GHR. Effect of caulophyllum 200 and 10,000 potencies on the ovaries, the uteri and the thyroids of rats. Ind J Zootomy 1975;16(199):199–204.
6.
Kumar S, Srivastava AK, Chandrasekhar K. Effects of caulophyllum on the uteri and ovaries of adult rats. Br Homoeop J 1981;70(135):135–138.
7.
Kerr HD, Saryan LA. Arsenic content of homeopathic medicines. Clin Toxicol 1986;24:451–459.
8.
De Smet PAGM. Giftige metalen in homeopathische preparaten. Pharm Weekbl 1992;127:26,125–126.
9.
Montoya-Cabrera MA, Rubio-Rodriguez S, Valazquez-Gonzalez E, Avila Montoya S. Intoxicacion mercurial causada por un medicamento homeopatico. Gaceta Med Mex 1991;127:267–270.
10.
Oepen I. Kritische Argumente zur Homöopathie. Dtsch Apoth Ztg 1983;123:1105.
11.
Van Ulsen J, Stolz E, Joost T. Chromate dermatitis from a homoeopathic drug. Contact Derm 1988;18:56–57.
12.
Forsman S. Homeopati kan vara farling vid hudsjukdomar och allergier. Läkartidningen 1991;88:1672.
13.
Morice A. Adultered “homoeopathic” cure for asthma. Lancet 1986;1:862–863.
14.
Goulon M, Combes A. Crise thyrotoxique médicamenteuse. Des dangers d'une préparation prétendue homéopathique. Nouv Press Med 1977;6:3729–3731.
15.
Abbot NC, White AR, Ernst E. Complementary medicine. Nature 1996;381:361.
16.
Reilly DT, Taylor MA, McSharry C, Aitchinson T. Is homeopathy a placebo response? Controlled trial of homeopathic potency, with pollen in hayfever as model. Lancet 1986;2(8512):881–885.
17.
Walach H, Haeusler W, Lowes T, et al. Classical homeopathic treatment of chronic headaches. Cephalalgia 1997;17:119–126.
18.
Doutremepuich C, DeSéze O, LeRoy D, et al. Aspirin at very ultra low dosage in healthy volunteers: effects on bleeding time, platelet aggregation and coagulation. Haemostasis 
1990;20:99–105.
19.
Cazin JC, Cazin M, Boiron J. A study of the effect of decimal and centesimal dilutions of arsenic on the retention and mobilization of arsenic in rats. Hum Toxicol 1987;135(6):315–320.
20.
Boiron J. Comparaison de l'action de Arsenicum album 7 CH normal et chauffe a 120_C sur l'intoxication arsenicale provoquee. Homeopathie 1985;2(5):49–53.

21.
Ernst E. Competence in complementary medicine. Comp Ther Med 1995;3:6–8.
22.
Goodyear HM, Harper JI. Atopic eczema, hyponatraemia, and hypoalbuminaemia. Arch Dis Child 1990;65:231–232.
23.
Tsur M. Inadvertent child health neglect by preference of homeopathy to conventional medicine. Harefuah 1992;122:137–142.
24.
De Smet PAGM, Stricker BHCh, Nijman JJ. Indirecte risico's van alternatieve middelen - een oproep tot het rapporteren van concrete gevallen. Med Contact 1994;49:1593–1594.
25.
Fisher P. Enough nonsense on immunization. Br Homoeop J 1990;79:198–200.
26.
English J. The issue of immunization. Br Homoeop J 1992;81:161–163.
27.
Ernst E. The attitude against immunisation within some branches of complementary medicine. Eur J Pediatr 1997;156:513–515.
28.
Burgess M. Homoeopathy and vaccination. Lancet 1994;344:1168.
29.
White AR, Ernst E. Homoeopathy and immunization. Br J Gen Pract 1995;48:629–630.
30.
Jonas WB: Safety in homeopathy. In: Ernzt E, ed. Homeopathy: a critical appraisal. London: Butterworth, 1998.

CHAPTER 9. A
DVERSE
 E
FFECTS OF
 A
CUPUNCTURE
Essentials of Complementary and Alternative Medicine
CHAPTER 9. A
DVERSE
 E
FFECTS OF
 A
CUPUNCTURE
Edzard Ernst
Infections
Trauma
Other Adverse Effects of Acupuncture
Chapter References
Acupuncture is one of the most popular complementary therapies worldwide. Contrary to claims made by some authors, acupuncture is not free of adverse effects or 
complications. Most of these are mild and transient (e.g., aggravation of symptoms, pain during needling, fainting). However, serious complications, including 45 
deaths (
1
, 
2
, 
3
 and 
4
), are on record.
INFECTIONS
Inadequate use of acupuncture needles carries an obvious risk of infection. One overview identified 126 published cases of hepatitis associated with acupuncture (
5
). 
One outbreak of hepatitis B was traced back to an acupuncturist who infected 36 patients by reusing needles (
6
). Six patients were infected through acupuncture 
treatments at a chiropractic institution (
7
). In a further instance, 35 patients of one acupuncturist tested positive for hepatitis B (
8
). Five cases of hepatitis B were 
reported in Israel, and an additional six people were infected but remained asymptomatic (
9
). Four similar cases have been documented in Europe (
10
, 
11
).
Epidemiologic studies contribute further important evidence. Kiyosawa and associates investigated two areas of Japan, one of which was endemic for hepatitis C. 
Acupuncture was significantly more prevalent in the region endemic with hepatitis C (
12
). During a screening program for hepatitis B, 651 individuals tested positive, 4 
of whom had a history of acupuncture (
13
). Other Japanese researchers investigated 262 hepatitis C carriers. They found that in 20% of cases, the most likely route 
of infection was via acupuncture (
14
). Finally, a survey in Singapore (
15
) revealed that the prevalence of hepatitis B in acupuncture patients was 8.7%, which was 
slightly less than for tattooing (10.9%) or blood transfusions (12.8%).
Even HIV infections have been linked to acupuncture. A Frenchman became HIV positive eight weeks after having being treated with acupuncture for six weeks (
16
). 
No other plausible origin of the infection could be identified. In a study of 148 AIDS patients, two individuals with no other recognized risk factor had previously used 
acupuncture (
17
).
Other serious infections include subacute bacterial endocarditis due to  Propionibacterium acnes infection after ear-acupuncture (
18
). Similar cases had been reported 
previously in which the causal agents were  Pseudomonas aeruginosa (
19
) and Staphylococcus aureus (
20
). All patients thus infected recovered with antibiotic 
therapy.
Two fatalities have been reported in which acupuncture led to S taphylococcus septicemia (
21
). Two further cases of septicemia are on record (
22
, 
23
). A dramatic 
incident was described (
19
) in which a patient received acupuncture treatment in the lumbar paraspinal region for low back pain and subsequently contracted bilateral 
psoas abscesses infected with S. aureus. He fortunately survived a severe and protracted illness.
Obviously, these infections are preventable. A survey of 500 Japanese acupuncturists showed that, in 1993, 60% of these therapists used disposable needles (
24
). It 
seems that use of sterile needles and correct handling of these needles are preconditions for safe acupuncture (
25
). If this precondition would be universally adhered 
to, such complications would be a thing of the past.