The Scientific Evidence for the Efficacy of Homeopathy


by Peter Fraser


In an editorial in The Townsend Letter1 Dr Gaby looked at the question of the effectiveness of homoeopathy from a basically sceptical point of view. The arguments against homoeopathy are ones of scientific common sense. Why should a substance that causes a symptom not make it worse rather than better? How can a dilution that no longer contains the original substance have a physiological effect? Even more questionable is the proposition that diluting even further increases that effect. On the other hand, Dr Gaby gives four reasons: anecdotal evidence, the test of time, basic Ultra High Dilutions (UHD) research and clinical evidence, which lead him not to discount homoeopathy, however much scientific instinct tells him he should. These make an excellent framework to analyse the evidence.


About the author

     Peter Fraser practices homoeopathy in Bristol and London. He is involved in homoeopathic research and the proving of new remedies and administers the Homoeopathic Information Service website ( He can be contacted at 23 Berkeley Road, Bishopston, Bristol, BS7 8HF. Tel: 0117 944 51447,




A GP Homoeopath’s View Regarding Research


by Dr Angela Jones, MA, BM, BCh, DCH, DRCOG, DFFP, MRCGP, MFHom


   Doctors who use homoeopathy face a doubly difficult task. Not only do they have to convince their patients that the therapy works – they also have to convince their colleagues. There was a time when general practitioners could practise more or less however they liked, as long as the patients came to no harm. Thus, homoeopathy was used in NHS primary care by a staunch group of medical homoeopaths for the benefit of their patients. Likewise, the four NHS homoeopathic hospitals were never short of patients who wished to be referred.

   However, the climate has changed fundamentally in the last five to ten years with the introduction of the internal market in the NHS. This has meant that health authorities have to make active decisions as to which treatments they are going to purchase. In the ever- cashstrapped NHS, homoeopathy has been a popular target for cuts. These cuts were made on the basis that “there is no evidence for homoeopathy”.

   Yet there is evidence that homoeopathy works. Major analyses of the available research have been performed by sceptical scientists and have concluded that homoeopathy is having an effect which is over and above any placebo response.1,2 A good proportion of the research which has been carried out uses standard modern methods such as the double-blind, randomised, controlled trial and therefore should be accepted by conventional practitioners, even if the way in which homoeopathy works is, as yet, unknown.

   The medical homoeopathic community is now concentrating a great deal of effort on large-scale studies to evaluate the outcome of homoeopathic treatment in the field. Several studies are measuring outcomes not only in terms of the specific medical outcome relating to the original complaint, but also with quality of life measurements. These should show significant improvements if our clinical impression of the overall benefits to the patient of homoeopathic treatment is correct.

   Homoeopathic practitioners cannot afford to turn their backs on research in the present climate. In the UK, where all medical interventions have been scrutinised for their “evidence base”, the very survival of homoeopathy as a therapy is at stake.


1.    Kleijnen J, Knipschild P and Ter Riet, G. Clinical Trials of Homeopathy. British Medical Journal. 302: 316-323. 1991.

2.    Linde K, Clausius N, Ramirez G, Melchart D, Eitel F, Hedges LV and Jonas WB. Are the clinical effects of homeopathy placebo effects? A meta-analysis of placebo-controlled trials. The Lancet. 350: 834-843. 1997.




 Anecdotal Evidence


     Anecdotal evidence is not accepted as “scientific”; yet when it comes from a reliable and trustworthy source it demands attention. For most homoeopaths, their patients and their prospective patients, this is what is most important. Homoeopaths see their patients getting better, sometimes quite dramatically, while their patients experience noticeable improvements and tell other people. Homoeopathy has always thrived on people’s experience of its effectiveness. Its first major expansion occurred in the middle of the nineteenth century when Hahnemann’s students proved to be so effective in the cholera epidemics that swept Europe. The expansion of homoeopathy in America in the second half of the nineteenth century and the growth seen in Britain over the last twenty years have been fuelled almost entirely by people’s experience of its effectiveness. The death rates in homoeopathic hospitals and hospitals and institutions supervised by homoeopaths in nineteenth century America were a half to an eighth of what they were in conventional institutions.2 Declines, especially the one that occurred in America at the turn of the century, have similarly come about because too many poorly trained practitioners have proved ineffective.


The Test of Time


     That homoeopathy has stood the test of time makes it hard to ignore. Many cures and therapeutic modalities come and go with remarkable rapidity. Even in conventional medicine, most treatments and drugs therapies are out of date and replaced within a few years. Homoeopathy has grown and been refined, but the principles and the research that led to it are as important and valuable today as they were two hundred years ago. If these principles were not as effective as they are they would require continual “adjustment” to keep them fitting the facts of the time.


Basic Ultra High Dilutions (UHD) Research


     It is the fact that any dilution beyond a certain level, which corresponds approximately to the 12c homoeopathic potency, will result in most samples containing no material substance that makes homoeopathy so unreasonable to scientists. Yet there are numerous studies of various types which have shown that both in vitro and in vivo effects can be caused by UHDs. Evidence ranges from Beneviste’s work on dilute IgE antiserum;3,4,5 and Conte’s work with NMR;6,7 to other studies using various homoeopathic dilutions.8,9 In an overview of UHDs AK Vallance concludes that there is considerable empirical evidence of them.10

     There are also emerging plausible and experimentally supported possible mechanisms for the actions of UHDs. These include the ordering of water in microtubules;11 and IE crystals found in water.12


Clinical Evidence


     Even the evidence for homoeopathy that scientists regard as the “gold standard”, double blind clinical studies, is not lacking. Research on respiratory disease in piglets and mastitis in cows has shown homoeopathy to be effective with animals where the placebo effect can safely be ignored.13,14 In humans, a tiny selection of the published research shows the range of evidence available. Studies have shown homoeopathy to be effective in: vertigo;15 pollinosis;16,17 postoperative ileus;18 upper respiratory tract infection and otitis media;19,20 and depression and anxiety.21

     A meta-analysis of placebo-controlled trials published in the Lancet concluded that the results were “not compatible with the hypothesis that the clinical effects of homoeopathy are completely due to placebo”.22 An earlier study also published in the Lancet also indicated that the evidence for homoeopathy is reproducible.23

     As Dr Gaby says: “While statistically significant results will occasionally occur by chance, it is extremely improbable that a worthless treatment would repeatedly produce statistically significant benefits.”24


Nux vomica and Acute Ethanol Poisoning


     Nux vomica is an important homoeopathic remedy. It is one of the major remedies used by homoeopaths to cure some patients of a number of conditions. It is made from the nut of a tree and contains a very high proportion of strychnine. Undiluted, it is highly toxic. Homoeopaths use the remedy in a highly dilute form that is no longer dangerous. Poisonings with the nut and provings in which the effect of the remedy was tested on volunteers indicated that the remedy causes a state of vertigo and stupefaction that resembles alcohol intoxication. On the homoeopathic principle that “like cures like” it has been given as a treatment for alcohol poisoning for nearly 200 years. Homoeopaths have observed and recorded its efficacy in this situation time and time again. I myself had a patient, a teenage boy, who had consumed an inordinate amount of alcohol and was comatose with an extremely weak pulse and imperceptible respiration. Given Nux vomica 30 every few minutes he was breathing normally in less than fifteen minutes and able to stand within an hour.

     In a very carefully designed and controlled experiment it was shown that giving mice Nux vomica 30 after they had been intoxicated with alcohol reduced the time between becoming intoxicated and being able to right themselves, the righting reflex, by 44%. This experiment was double blind and fulfilled all criteria expected of such experiments.25

     Conventional medicine has no effective treatment for acute alcohol poisoning. Here we have a treatment with 200 years of powerful clinical evidence for its efficacy, sufficient usage to show it has no serious side effects and “gold standard” animal tests that show its physiological effect.


Unconvinced by the Evidence


     Why is it that homoeopathy has proved itself so clearly yet it is not accepted? Why is it that Nux vomica is not the standard treatment for alcohol poisoning in every Accident & Emergency department in the country? One reason is that because of homoeopathy’s lack of a demonstrable mechanism, homoeopathy must produce much more evidence of a much higher quality than would be required of a conventional treatment. It has been shown that for doctors “the more one considers that empirical data should be assessed independent of theoretical considerations, the higher are the demands which are placed on the quality and quantity of the empirical documentation”.26

     Yet this is not really the most important factor. Homoeopathy and conventional medicine have very different underlying philosophies and the methodology of conventional research is designed from the framework of conventional philosophy. To give an analogy: if you look at a plan of the island of Manhattan the World Trade Center is fairly insignificant and the most important feature is clearly Central Park; however, a picture of the skyline of New York is dominated by the World Trade Center and Central Park is nowhere to be seen. If homoeopathy is Central Park, and conventional medicine the skyscrapers, then conventional research is very definitely the view and not the plan.


Individualisation of Remedies


     Conventional research does not, and perhaps cannot, take account of many of the most important features of homoeopathy. The first and most important of these is the matching of the remedy to the individual characteristics of the patient. In both the studies on pollinosis cited above, the comparison was between a single remedy and either placebo or conventional betahistamines. Yet a standard textbook of homoeopathic therapeutics offers 25 remedies for hayfever,27 and that list is by no means exhaustive – it doesn’t include either of the remedies shown to be effective in the aforementioned tests. Only a small number of people will be helped by any one remedy. That small number will often be below the number who get better naturally on placebo and therefore will not show up in the sort of test accepted by conventional science. A certain condition might improve naturally in 10% of cases. A conventional treatment might help 25% of patients and the effect can be clearly demonstrated. Yet, though homoeopathic treatment might help 75% of the patients, it would use a dozen or more different remedies and it is possible that no individual remedy would be effective in 10% of cases. In other words, although homoeopathy helped three times as many people, the effectiveness of any of the remedies used could not be demonstrated. Homoeopaths who have treated thousands of patients will have only one or two cases of some remedies. However dramatic may have been the individual effects, they will never have statistical significance.

     Even where there is individualisation of the remedies used in research, the quality of the individualisation is a major variable. There is a large number of mediocre homoeopaths and it is impossible to tell whether the individualisation in a trial was good or not.


Definitions of Disease and Cure


     Definitions of cure and disease used by homoeopaths are very different from those of conventional medicine. For a homoeopath the disease is made up of the patient’s symptoms. “The unprejudiced observer only perceives the deviations from the former healthy state of the now sick patient… These perceptible signs represent the disease in its entire extent.”28 For a homoeopath there can be no such thing as asymptomatic disease. In a study on the effects of homoeopathic treatment in HIV infection, homoeopathy was shown to have a statistically significant effect in the symptomatic phase of the disease and no statistical effect in the asymptomatic phase.29 To a homoeopath this is entirely consistent with what is to be expected; to a conventional researcher it appears that homoeopathy is only partially effective or only palliative of the symptoms.

     Conventional medicine tends to act through a continuum of amelioration from slight to complete; whereas homoeopathy cures completely when the right remedy is given, and does little or nothing when the wrong remedy is given. Thus, even studies that use an individualised homoeopathic remedy may use definitions that work against homoeopathy. A study of homoeopathy for prophylaxis of migraine found no significant benefit for the outcome measure which was attack frequency; however, the reduction in attacks was in mild attacks in placebo but in medium and severe attacks in verum. The improvement in placebo reversed in the last month of the trial whereas it did not in the verum group.30 The study lasted for only three months of active treatment and did not allow for the changing of remedy. The homoeopathic treatment of an adult with migraine would normally last longer than three months and might be expected to involve more than one remedy. Again there is no way of knowing the quality of the individualisation of the remedy choice.


The Whole and the Parts


     Homoeopathic treatment is always of the person and not of the disease, much less of individual symptoms of the disease; yet conventional research can only take account of a single variable and thus a single symptom. In homoeopathic thinking, the clearing of a symptom is curative only if it is part of a more general improvement and it is decidedly a poor outcome if it is replaced by another more serious symptom. Thus, the removal of eczema cannot be considered curative if it is replaced by asthma, and the appearance of eczema is a curative sign if it replaces a more serious disease such as asthma. A study that shows the inappropriateness of conventional research methods to homoeopathy was one conducted to compare the effect of homoeopathy on warts in children.31 Warts are a nuisance but they are one of the least life-threatening, and therefore least serious, manifestations of disease. In treatment they would therefore be one of the last symptoms to disappear, probably after some time and possibly after more than one remedy. Not only does the trial take no account of this but the measure of success was at least a 50% reduction in the area occupied by warts. The difference between placebo and verum by this measure was insignificant and this was the conclusion of the study; however, if the measure used had been total cure the conclusion would have been very different, as five times as many children were completely cured on verum as were on placebo.


Other Factors and Issues


     Conventional medicine relies on a diagnosis that can be confirmed by designated tests, and then on treatment that has been shown to be effective in controlled research. Homoeopathy does not use a diagnosis, only a prescription, and that can only be tested by observing its results. Since each case is unique, it is impossible to apply statistical analysis to it in the way that you can in conventional medicine. Conventional medicine has a set of rules that are relatively easy to apply. A doctor therefore needs to have a certain degree of knowledge and then apply the prescribed treatment in the way that has been judged to be most effective. This is fairly easy to regulate. The homoeopath likewise needs a certain amount of knowledge but there is no prescription for the way in which his knowledge is applied. Individual circumstances demand an entirely individual response. This is why homoeopathy is so difficult to regulate and why the setting of standards, such as the recent National Occupational Standards, are so problematic.

     The motivations in homoeopathy and conventional medicine, especially the economic ones, travel in very different directions. A pharmaceutical company is interested in proving that its treatment is the best and the doctor puts that proof into practice. Homoeopathy has no such industrial base, and the entire motivation comes from the homoeopath who is looking for the right remedy for his patient and wants only research that will help find it.


Homoeopathic Research

     Homoeopathic research is fundamentally different from that of conventional medicine. There is no point in trying to plot the details of Central Park with views of the skyscrapers. The research that homoeopathy needs involves provings and clinical case reports. Conventional research offers very little to homoeopaths or their patients.

     Even as a means of convincing sceptics, conventional research will probably fail because it is looking at the wrong things. In my view, the only research that will, in the long run, convince doctors and politicians is that which gives evidence of the overall efficacy and cost effectiveness of homoeopathy – research like that of Dr Jenifer Jacobs which showed that physicians using homoeopathy spent more time with their patients but ordered fewer expensive tests and prescribed fewer expensive conventional medicines.32



1    Gaby AR. Editorial. Townsend Letter. 195:13-14. Oct 1999.

2    Ullman Dana. What Skeptics Say About Homeopathy and How to Respond to Them. Townsend Letter. 195:124-8. Oct 1999.

3    Davenas E et al. Human basophil degranulation triggered by very dilute antiserum against IgE. Nature. 333: 816-8. Jun 1988.

4    Poitevin B, Davenas e and Benveniste J. In vitro immunological degranulation of human basophils is modulated by lung histamine and Apis mellifica. Br J Clin Pharmacol. 25(4): 439-44. Apr 1988.

5    Schiff Michel. The Memory of Water. Thorsons. London. 1995.

6    Conte Rolland et al. Theory of High Dilutions and Experimental Aspects. Dynasol Ltd. 1997.

7    Milgrom Lionel. Ou Etes-vous, Dr Conte. The Homoeopath. 72: 32-6. 1999.

8    Dittmann J and Harisch G. Characterization of

differing effects caused by homeopathically prepared and conventional dilutions using cytochrome P450 and other enzymes as detection systems. J Altern Complement Med. 2(2): 279-90. 1996.

9    Datta S, Mallick P and Bukhsh AR. Efficacy of potentized homoeopathic drug (Arsenicum album-30) in reducing genotoxic effects produced by arsenic trioxide in mice: comparative pre- and post-oral administration and comparative efficacy of two microdoses. Complement Ther Med. 7(2): 62-75. Jun 1999.

10    Vallance AK. Can biological activity be maintained at ultra-high dilution? An overview of homeopathy, evidence and Bayesian philosophy. J Altern Complement Med. 4(1): 49-76. 1998.

11    Hameroff SR. Ch’i: A Neural Hologram? Microtubules, Bioholography and Acupuncture. Am J Chin Med. 2(2): 163-70. 1974.

12    Lo Shui-Yin. Anomolous States of Ice. Modern Physics Letters B. 10(19): 909-19. 1996.

13     Albecht H and Schutte A. Homeopathy versus antibiotics in metaphylaxis of infectious diseases: a clinical study in pig fattening and its significance to consumers. Altern Ther Health Med. 5(5): 64-8. Sep 1999.

14     Merck CC, Sonnenwald B and Rollwage H. The administration of homeopathic drugs for the treatment of acute mastitis in cattle. Berl Munch Tierarztl Wochensschr. 102(8): 266-72. Aug 1989.

15    Weiser M, Strosser W and Klein P. Homeopathic vs conventional treatment of vertigo: a randomized double-blind controlled clinical study. Arch Otolaryngol Head Neck Surg. 124(8): 879-85. Aug 1998.

16    Reilly DT, Taylor MA, McSharry C and Aitchson T. Is homeopathy a placebo response? Controlled trial of a homoeopathic potency, with pollen in hayfever as model. Lancet. 2(8512): 881-6. Oct 1986.

17    Ludtke R and Wiesenauer M. A meta-analysis of homoeopathic treatment of pollinosis with Galphimia glauca. Wien Med Wochenschr. 147(14): 323-7. 1997.

18    Barnes J, Resch KL and Ernst E. Homeopathy for postoperative ileus? A meta-analysis. J Clin Gastroenterol. 25(4): 628-33. Dec 1997.

19    Friese KH, Kruse S, Ludtke R and Moeller H. The homoeopathic treatment of otitis media in children – comparisons with conventional therapy. Int J Clin Pharmacol Ther. 35(7): 296-301. Jul 1997.

20     Lewith GT. Upper respiratory tract infections and otitis media. Complement Ther Med. 5: 99-100. 1997.

21    Davidson JR, Morrison RM, Shore J, Davidson RT and Bedayn G. Homeopathic treatment of depression and anxiety. Altern Ther Health Med. 3(1): 46-9. Jan 1997.

22    Linde K, Clausisus N, Ramirez G, Melchart D, Eitel F, Hedges LV and Jonas WB. Are the clinical effects of homeopathy placebo effect? A meta analysis of placebo-controlled trials. Lancet. 351(9097): 834-43, Jan 1998.

23    Reilly D, Taylor MA, Beattie NG, Campbell JH, McSharry C and Aitchson TC, Carter R and Stevenson RD. Is evidence for homeopathy reproducible? Lancet. 344(8512): 1601-6. Dec 1994.

24    Gaby AR. Editorial. Townsend Letter. 195:13-14. Oct 1999.

25    Sukul A, Sinhabau SP and Sukul NC. Reduction of alcohol induced sleep time in albino mice by potentized Nux vomica prepared with 90% ethanol. Br Homoeopath J. 88: 58-61. 1999.

26    Lynoe N and Svensson T. Doctor’s attitudes towards empirical data – a comparative study. Scand J Soc Med. 25(3): 210-6. Sep 1997.

27    Lilienthal Samuel. Homoeopathic Therapeutics. Sett Dey & Co. Calcutta. pp505-7. 1950.

28    Hahnemann Samuel. O’Reilly WB ed. Organon of the Medical Art. Birdcage Books. Redmond, Washington. §6. 1996.

29    Rastogi DP, Singh VP, Singh V, Dey SK and Rao K. Homeopathy in HIV infection: a trial report of double-blind placebo controlled study. Br Homeopath J. 88(2): 49-57. Apr 1999.

30    Whitmarsh TE, Coleston-Shields DM and Steiner TJ. Double-blind, randomized placebo-controlled study of homoeopathic prophylaxis of migraine. Cephalagia. 17(5): 600-4. Aug 1997.

31    Kainz JT, Kozel G. Haidvogel M and Smolle J. Homoeopathic versus placebo therapy of children with warts on the hands: a randomized double-blind clinical trial. Dermatology. 193(4): 318-20. 1996.

32    Jacobs J, Chapman EH and Crothers D. Patients’ characteristics and practice patterns of physicians using homeopathy. Archives of Family Medicine. 7: 537-40. 1998.