Hanspeter Seiler
Publikation Vithoulkas-Seiler

New Facts about the Munich Headache Study

by G. Vithoulkas and Hp. Seiler

Since its publication, the Munich headache study by Walach  has been the subject of controversial discussion.  In spite of this, even in homeopathic circles, Walach’s study is still regarded as a seri-ous scientific trial with negative implications for classical home-opathy and has influenced all meta-analyses since published. As a result, it ”damaged homeopathy more than anything else that had so far surfaced in medical journals“  and has become a main pillar of Walach’s interpretation of homeopathy as “non-causal” respec-tively “magical”,  contradicting Hahnemann’s original principles.
This has recently prompted Seiler to carry out a detailed review of Walach’s study.  This shows that Vithoulkas’ original criticism that the verum group was suffering from homeopathic aggravations can be proven to be correct. Walach’s data concerning the thera-peutic reactions of verum and placebo have been interchanged for the most part and are interpreted in a clinically inadequate man-ner; moreover, an essential error in randomization has been over-looked and the clinical parameters for migraine have been used inappropriately. The following text includes a review of the history and the most important critical aspects of the Munich study.

Homeopathic headache-studies, double-blinded studies, interpre-tation of homeopathy.

1. The legendary migraine study of Brigo and Serpelloni
This trial was published in 1991 by the Italian homeopathic re-searcher, Dr. med. B. Brigo, in cooperation with Dr. med. G. Ser-pelloni.  Its aim was to prove the efficiency of high-potencies in severe chronic cases in a strict double-blinded setting. For that purpose, Brigo selected chronic migraine patients with a mean age of 37.5 years. Only patients with a first medication of Lache-sis, Natrium muriaticum, Silicea and Sulphur were admitted to the study. Randomization was perfect and the patients were allowed to continue their usual allopathic medication and dietary habits. Homeopathic medication was administered in C 30 at intervals of 14 days. Changing the remedy was allowed and this possibility was used in most cases. The duration of the study was 4 months.
Its results were spectacular: headache frequency, the most im-portant clinical parameter in migraine studies, displayed about 70% reduction in the homeopathic verum group. The comparison group taking placebo showed only 20% reduction.
This almost too brilliant Italian result clearly surpasses an average homeopathic practitioner’s daily experience with severe chronic migraine cases.  Possibly it just reflects the extraordinary homeo-pathic skill of Dr. Brigo with the assistance of some statistical luck.  Nevertheless, the data were double-blinded and statisti-cally highly significant. Therefore, they provided further proof that homeopathic remedies potentized far beyond the number of Avogadro are also effective in chronic cases.

2. The Munich headache study – a differently conceived re-construction with paradoxical results
Some years later Prof. Dr. phil. H. Walach et al. decided to re-examine Brigo’s results.  However, they inexplicably chose a very different setting which will be discussed below in some detail.
The results of the Munich study were a shock for the homeopathic world and a triumph for its adversaries: both placebo and verum showed virtually the same, only very slight reduction in headache frequency amounting to about 6%, much less even than Brigo’s placebo rate of 20%. What was even worse, Walach’s placebo group showed a clear tendency to achieve better results than ho-meopathy.
A further strange finding was that all the placebo results displayed far more fluctuations than verum. Walach was unable to interpret this phenomenon.  
In this article, we will try to show that these paradoxical results of the Munich study can be explained by its inadequate conception, homeopathic long-term aggravations and an essential error in randomization.

3. Problematic use of the parameter “headache-frequency” in the Munich study
One of the main differences between Brigo’s and Walach’s trial is the fact that the Munich study was not confined to migraine pa-tients but included all kinds of severe chronic headache, particu-larly also permanent tension headache. Indeed, more than one third of Walach’s patients (37%) suffered from permanent head-ache.  
In spite of this fact, Walach uses the term “headache-frequency”, which is normally restricted to pure migraine studies, as the pri-mary parameter in his mixed study.  This is already quite prob-lematic. As every practitioner knows, patients with permanent tension headache per definitionem do not suffer from the periodic pain attacks with pain-free intervals characteristic of migraine. As a result, permanent sufferers who are successfully treated first feel a reduction in their pain intensity, which may later result in their headache ceasing to be permanent and becoming periodic. Per contra, migraine patients who are being successfully treated experience a reduction in their headache frequency right from the outset. This reduction may even precede the lessening of pain intensity.
Walach tries to avoid this problem by using a statistical manoeu-vre. Instead of asking his patients about the frequency of their pain attacks, as is usual in migraine studies, he defines his “head-ache frequency” as the “number of positive answers to the ques-tion: Did you suffer from headache today?”  over a certain period of time. This naturally means that the “headache frequency” of all his permanent sufferers is registered as “daily attacks”. In this way, all permanent headache patients in the Munich study were transformed into “migraine patients” – merely with seven “at-tacks” a week!
This confusion of chronic migraine and permanent tension head-ache is not just a statistical problem but essentially affects the clinical significance of the Munich study. Since permanent head-ache that is getting better will only show a reduction in frequency at a much later date than migraine, it becomes evident that the change in Walach’s parameter “headache frequency” is slower and less distinctive than it is the case in pure migraine trials.
4. Advanced age of the patients further slowed down the reactivity of the Munich study
Age was an additional factor slowing down the reactivity of the Munich-study. The mean age of Walach’s patients was substan-tially higher than Brigo’s (48.5 vs. 37.5 years) and the duration of their suffering longer (23 vs. 16.3 years). This difference was even greater in the verum group. The mean age of Walach’s verum patients was 51 years (Brigo’s 37) and they had been suffering from headaches for a mean of 23 years (compared to 14 years in Brigo’s study).
In his first critique, Vithoulkas already mentioned the fact that it would be very difficult to obtain useful results with such a problematic collective in a trial lasting merely a few months.  This was also the final opinion of the physicians conducting the Munich study.  
How slow the reactivity of the Munich study really was, particu-larly in regard to frequency reduction, can be shown by compar-ing its placebo results with those of pure migraine studies. In comparable studies, the placebo results should be roughly the same irrespective of the verum medication administered because the placebo patients all get the same medication viz. nothing. As mentioned above, the reduction in migraine frequency in Brigo’s Italian placebo group was about 20%, other pure migraine studies show about 15%.  In Walach’s study, however, as mentioned above, the placebo reduction in “headache frequency” was only about 6%.

5. Unrealistically high “headache frequencies” at the be-ginning of the trial
Furthermore, in Walach’s study, because of the many patients suffering from “daily attacks”, there was also an exceedingly high initial “headache-frequency” of more than 16 per month.  What is even more remarkable is that during the six-week observation period without medication that preceded the medication phase of the Munich study, the “headache frequency” of the placebo group showed an isolated, inexplicable rise to about 18 attacks per month.
These values are, of course, clinically absolutely unrealistic. Brigo recorded 10 attacks per month among his severely afflicted Italian patients, which was already a very high value, whereas other homeopathic migraine studies show just 4 - 5.  
However, this passes almost unnoticed because Walach never ex-plicitly mentions his artificially inflated “headache frequencies”. He restricts the use of this questionable term to the summary of his main publication and other general conclusions where absolute numbers do not have to be given. In his circumstantial analysis and in the only detailed graph published in his main article,  however, he carefully avoids the term “headache frequency” and more correctly speaks of the “percentage of patients with head-ache per day”. The rather complicated method of calculating the corresponding “headache frequency” from this extremely unclear expression  is not explained.

6. Walach’s more appropriate parameter “pain intensity” showed the most paradoxical results
Apart from his problematic “headache frequency”, Walach of course also registered the parameter “pain intensity” which is bet-ter adapted to permanent pain. As expected, this indicator re-acted more distinctly.
However, in contrast to the exceedingly high initial values of Walach’s “headache frequency”, the initial average pain intensity of his patients was much lower than in pure migraine studies. Walach reports only 15 mm on the VAS scale whereas Brigo indi-cates 87 mm and Straumsheim 53.7.  
This divergence can again be explained by Walach’s mix of mi-graine and permanent tension headache patients. Chronic tension headache never usually reaches the very high pain score of se-vere migraine attacks. What is more, the migraine patients’ days free of symptoms may even have been registered as zero pain-values. If this were the case, the sensitivity of the Munich study would have decreased even further.
Be that as it may, in spite of its very low basic value, Walach’s more reliable parameter “pain intensity” yielded a clearer result than his “headache frequency”. However, this made the para-doxes in his study more apparent in equal measure: During the observation period, there was an even more pronounced increase in the pain score of the placebo group from 15 to 20 mm VAS whereas verum once more remained practically stable.  Then, in the treatment phase, the comparison group showed a consider-able reduction of 23.5% from this elevated basic value. This is decidedly more than the 6% displayed in “headache frequency”. Verum, on the other hand, only showed 7.5% reduction in pain intensity which is also somewhat more than in “headache fre-quency” (6%) but at this point clearly less than the placebo group.  
If placebo shows better results than verum in a study, this must be due to one of the following two causes:
1. A hidden artefact of randomization makes the placebo group decidedly more sensitive to positive environmental factors of the study than verum. In this case the trial concerned evidently be-comes useless.
2. The verum medication has caused a real aggravation. Natu-rally, this can happen, particularly in a homeopathic trial. But in this case correct application of the well-selected remedy and ap-propriate length of the study make it possible to avoid this error.
These two factors can have a combined effect of course. In the following, we will show that this was most probably the case in Walach’s study.

7. An obvious error in randomization
The complicated process of randomization is carefully documented in Walach’s study. Unfortunately, it resulted in a rather considerable difference in the size of the groups. Of the total 98 patients, 61 were allocated to the verum group, compared to only 37 in the placebo collective. This disequilibrium made the study more susceptible to artefacts from the outset.
Walach had already noted an apparent lack of clinical homogene-ity in his two groups. Before the trial, 75% of the verum group were taking allopathic remedies whereas this was only the case with 58% of the placebo group.  The authors of the Munich study were unable to state a reason for this discrepancy.
Another inconsistency which corresponds to that mentioned above is not noted however. Before the trial, 27% of the placebo pa-tients underwent psychotherapy or similar psychosomatic treat-ment compared to just half of that amount (14%) in the verum group.
So in spite of the careful randomization, in the small comparison group we find a clear tendency to prefer psychotherapy to drugs. How can this be explained? A look at the distribution of the occupational groups among verum and placebo gives us the clue to this strange phenomenon. Through a very rare coincidence, with the almost incredible low probability of only about 1:350, all the patients belonging to creative-artistic professions (art, media etc.) happened to be allocated to Walach’s small placebo group.  As a result, these six people represented the second largest occupational group  in the comparison group of only 37 patients.
Consequently, the verum group was dominated by more prosaic people tending to practise rather down-to-earth professions.  As artists professionally have to be more sensitive than other people, this group difference most probably explains the increased psy-chosomatic susceptibility of the placebo collective which - amongst other things - made it prefer psychotherapy to drugs.
This artefact of randomization was increased by another acciden-tal imbalance. Of the six patients belonging to paramedical pro-fessions, no less than five were allocated to Walach’s small pla-cebo group. This disproportion is, admittedly, somewhat less im-probable than that of the artists but still very noteworthy. Nurses and medical assistants tend on average to be more sensitive to medical influences than other people. So finally the number of psychosomatically potentially more sensitive patients in Walach’s placebo group grows to almost one third (11 of 37).  

8. Walach’s more sensitive placebo group noted higher di-ary values
During the purely observational six-week period without new medication, all patients had to keep a detailed diary of their com-plaints. The end result of this diary run-in was taken as the basis value for the subsequent three-month medication phase. As al-ready mentioned, during this period the placebo values showed an isolated increase of about 10% in “headache frequency” (from about 16.5 to 18) and about 30% in pain intensity (from about 15 to 20) while the verum values remained practically stable.
This strange phenomenon can now be explained. As we have seen above, artists and nurses, who were clearly overrepresented in the placebo collective, usually exhibit more psychosomatic sensi-tivity than other people. So when starting a detailed diary without receiving any new medication, this group most probably tended to display more introspection and self-observation than the verum group  causing them to note down higher scores for their symp-toms (= diary effect).
9. Important changes in allopathic medication and with-drawal from coffee affected verum somewhat more than placebo
After the observation period, the three-month treatment phase was started. In contrast to Brigo, Walach’s patients were now forced to change their allopathic medication in significant as-pects.  These changes naturally applied to both groups, but the verum patients, who were more liable to take drugs, were proba-bly affected somewhat more. This might also have contributed to the lack of positive reaction of the verum group during the treat-ment period.
Furthermore, in contrast to Brigo, Walach’s patients now had to cut out coffee completely.  Coffee restriction is unfortunately documented somewhat inconsistently by Walach. However, it seems clear that although it had partially begun in the observa-tion period, it was only strictly controlled in the treatment pe-riod.
But at least theoretically it could be that an increased susceptibil-ity of the placebo people to coffee withdrawal might have caused the isolated rise in their parameters during the observation phase. This, however, is very improbable for two reasons:
1. All registered complaints concerning coffee withdrawal were noted by members of the verum and not by the placebo group.
2. Another homeopathic migraine trial, the Norwegian Straum-sheim study, also included a purely observational phase, but cof-fee withdrawal only started during the treatment phase. Nevertheless, this study also shows an augmentation in the num-ber of attacks from about 4 to more than 5 per month recorded in the diaries.  So surely the diary effect occurs independently from coffee withdrawal and is mainly due to psychological factors.
The increased sensitivity of the Munich verum group to coffee withdrawal can be explained by the fact that coffee is the best phytotherapeutic remedy for headache and that this collective generally showed more drug dependency. Accordingly, persistent symptoms stemming from coffee withdrawal might also have con-tributed to the negative result of the homeopathic collective in the treatment phase.  However, as we shall see, there are other more important explanations for this.

10. Problematic application of Q-potencies with risk of long-term aggravations
In Walach’s study, instead of a single doctor deciding about the constitutional remedy to be given, the decision was made by a group of homeopathic physicians. Furthermore, in contrast to Brigo, these homeopathic physicians were allowed to select their remedies without any restriction.
After beginning with their medication, patients underwent a first routine control after 6 weeks which was followed by a second treatment period of the same length concluded with the final ex-amination. However, they were free to contact their doctors at any time and changes in medication could be initiated whenever prescribed. But this option was used seldom: during the entire trial only 10 new prescriptions were issued outside the routine controls.  Furthermore, even when changes in medication at the occasion of routine-controls were included, new prescriptions were clearly less frequent than with Brigo.
In addition, the Munich physicians were free to choose among C- and Q-potencies. This is an important fact as finally 65 viz. 30% of the overall 217 documented prescriptions for Walach’s 98 pla-cebo und verum patients were Q-potencies.  Q-potencies are usually administered daily and need a more stringent case man-agement than with the Kentian method to which the control fre-quencies of the Munich study were adapted. Hahnemann con-trolled even his chronic patients treated with Q-potencies regu-larly at much shorter intervals of 1 – 2 weeks and frequently changed his prescriptions.  In his first discussion with Walach, Vithoulkas also stresses the necessity of prescribing a sequence of different remedies when treating patients with severe chronic headache.
In the Munich study, however, in at least some cases, Q-potencies were administered unchanged once or even twice daily during a whole treatment interval of 6 weeks without regular con-trol in between.  In addition, as mentioned above, extra-consultations involving change of prescription were rare.
If Q-potencies are administered in such a way, even long-term aggravations can easily occur. In addition, Vithoulkas, in his first critique of Walach’s study, already mentioned that in very severe chronic cases long-term aggravations may occur even when using Kentian technique.  
We shall see that the pattern of therapeutic reactions reported by the homeopathy group makes it almost certain that this was the case with a relevant number of patients.

11. Interchange of the data concerning the therapeutic re-actions of verum and placebo groups
After each of the two six-week treatment periods, patients were asked to note any side effects resulting from the treatment in their diary. First, they had to reply to the general question of whether they felt side effects with “yes” or “no” and then were asked to give specifications if applicable. Twenty-two homeopathy patients (36.1%) and 17 participants in the control group (45.9%) replied “yes” once or twice and many of them (17 of verum = 27.9%, 12 of placebo = 32.4%) also noted details of their reactions.  These specifications are, of course, of great im-portance in attempting to answer the question whether verum suffered from specific homeopathic aggravations or not.
In view of the importance of these data, we compared Walach’s list of specific reactions  with the original published by Walach’s statistician Prof. Gauss.  To our great astonishment, we found that in most cases (80%) Walach’s data of placebo and verum were interchanged.  This very remarkable error has been explic-itly confirmed by Gauss.  
This mistake clearly confirms that Walach’s analysis of the speci-fied reactions of verum and placebo was rather careless. Without giving any concrete argument (not to mention statistics), he simply notes that exactly 50% of the specifications given by both groups could be interpreted as typical homeopathic reactions.  So according to Walach, there was absolutely “no difference be-tween the two groups”.  
Hence, one could conclude that his confusion regarding the pa-tients’ data was not of great significance. However, a careful analysis of the adjusted data will show the contrary.

12. Aggravations in the first half of the trial were much more frequent among the homeopathy group
After correcting Walach's error, we could analyze the specified notes about the therapeutic reactions in detail. They were listed in the following manner: if a patient noted several symptoms of dif-ferent organ systems (e.g. digestive and circulatory problems in the same patient), they were counted separately. Symptoms noted by the same patient both in phase I and II of the treatment period were also noted separately. This resulted in our  finding 27 specifications of therapeutic reactions with verum and 20 with pla-cebo.
There is already an astonishing difference between verum and placebo when analyzing the pattern of occurrence of these speci-fied symptoms, either in phase I or II. Of the 27 therapeutic reac-tions of the homeopathy group, no less than 21 were registered in the first period of treatment. This is an overwhelming majority of 78%.
In the control group, however, only 8 of the 20 detailed symp-toms (40%) were noted in phase I. So with homeopathy we find a relationship of early to late phase reactions of 21:6 or 3.5:1 whereas with placebo only 8:12 respectively 0.7:1.
If homeopathy and control group had been equal in their reaction pattern, these coefficients should, of course, be about the same and accordingly their ratio about 1:1. But in Walach’s trial we find a significant difference in this relationship of 3.5:0.7 respectively 5:1.
This clearly proves that the homeopathy group suffered more in phase I of the treatment. To a certain extent, this can be ex-plained by the forced changes in chemical and phytotherapeutic medication affecting verum somewhat more than placebo as dis-cussed above. But it will be shown below that most probably ho-meopathic aggravations were responsible for at least some of these early therapeutic reactions.
The slight increase in placebo or rather nocebo “side-effects” in phase II of the trial is typical. It is well known that the positive reaction to placebo medication is always strongest at the begin-ning of treatment. Subsequently, disappointment sets in more or less rapidly and consequently complaints increase. Brigo’s study already showed that with migraine patients, the placebo effect al-ready gets decidedly weaker after two months.

13. The homeopathy group suffered significantly more from typical homeopathic aggravations.
To judge a symptom as a typical homeopathic reaction, Seiler classified the specified therapeutic reactions in seven categories according to their homeopathic importance. This resulted in the following ranking list:

I. Initial aggravation followed by explicit amelioration
Verum: 1 case.
This patient noted with an exclamation mark that she or he was ”worse for three days, then well!“
Placebo: No such records.

II. Old symptoms returning after a longer period of ab-sence
Verum: 2 cases
a. Return of depression after a longer phase of stability. This was so severe that the patient’s family judged it to be unacceptable.
b. Unbearable headache on two days exactly noted with date. This had never happened to such a degree since the birth of the pa-tient’s son two years earlier.
Placebo: No such records.

III. Symptoms explicitly marked as new or unusual
Verum: 2 cases
a. Menstruation four days earlier than usual which had never happened before.
b. Unusual pattern of headache on four days exactly noted with date.
Placebo: No such records.
So all five symptoms ranking among the first three categories of typical homeopathic reactions were noted by the verum group. Even considering the difference in group size (61 verum, 37 pla-cebo), this ratio of 5:0 in favour of the homeopathy group still remains at about 3:0.

IV. Probably new, characteristic and striking symptoms
Verum: 8 cases
Placebo: 2 cases
If we add this category, which is still typical for a homeopathic re-action, the preponderance of the homeopathy group increases further to a ratio of 13:2 = 6.5:1. Considering the difference in group size we still find about 4:1.
It is only from the next lower category of symptoms onward that the ratio of homeopathy to placebo changes to an increasing preponderance of placebo.

V. Possibly new but poorly characterized and non-specific symptoms
Verum: 7 cases
Placebo: 7 cases
In this group of symptoms, which also includes typical negative placebo or nocebo reactions, we find a relationship of 1:1. Con-sidering the difference in group-size this coefficient even in-creases to 1.7:1 in favour of placebo.

VI. Aggravation of current symptoms (in most cases head-ache) without explicit amelioration
Verum: 2 cases
Placebo: 4 cases.
As expected in this category, which includes the most typical kind of nocebo reaction, we find a clear dominance in the control group.

VII. Symptoms marked by patients explicitly as insecure or caused by other reasons than medication
Verum: 5 cases.
Placebo: 7 cases.

Of course this classification of symptoms has some arbitrary as-pects and could be handled in a somewhat different way too. Nevertheless, the outcome of an exact analysis would be about the same: typical homeopathic reactions were decidedly more frequent within the verum group.
How could such obvious differences between the therapeutic reac-tions of verum and placebo pass unnoticed in the analysis of the Munich study?  - Maybe Walach was just unable to see what he did not want to see. In his vehement dispute with Vithoulkas about the influence of homeopathic aggravations on the outcome of his study, he even went so far as to call homeopathic aggrava-tion “a myth”.  This statement could only come from someone lacking true experience in classical homeopathy. A good homeo-pathic practitioner sees aggravations among his patients on a day-to-day basis.  So it seems important to the authors of this article that future important homeopathic studies should only be monitored and analyzed by the most experienced homeopathic practitioners and not by mere theoreticians.  

14. The analysis of the Munich study must be revised
So the paradoxical findings of the Munich study can be exposed in the following way:

I. Placebo group
During the diary run-in without medication, the placebo group, which was psychosomatically more sensitive, noted higher scores than verum in all parameters and consequently reached more ele-vated starting values.
Once treatment began, this group for the same reason showed a decidedly positive placebo reaction with clear amelioration.
Only at the end of phase II, with decreasing placebo effect, did this amelioration turn into a final aggravation of all parameters. However, this third distinct fluctuation of the placebo values did not reach the elevated starting point seen at the end of the diary run-in period.
Consequently, in spite of the very low reactivity of Walach’s study, there remained a slightly positive result in favour of pla-cebo.

II. Homeopathy group
In contrast, the homeopathic verum group consisting of more down-to-earth people recorded almost stable scores during the diary run-in.
This surprisingly did not change greatly during the treatment phase. So we have to conclude that antithetical factors were can-celling each other out in the treatment phase:
- On the negative side, we surely have to take into consideration, as explained above, homeopathic aggravations that were poten-tially severe and long lasting. Furthermore, the verum group, be-ing more dependent on drugs, suffered at least somewhat more than the placebo group from the significant changes in allopathic medication and the total coffee withdrawal imposed by the Munich study.
- On the positive side, we surely have to take into consideration a certain placebo effect in the homeopathy group too. However, for the reasons mentioned above, this factor has to be weighted somewhat less than for the placebo group. Furthermore, in spite of the problematic setting of the Munich study and the difficulties already mentioned, we also have to take into consideration at least a small number of patients who, without prolonged initial aggravation, showed positive reactions to their well selected and correctly administered homeopathic remedy.  However, for the reasons given above, these positive therapeutic reactions were neither frequent nor striking.
In this way, we can explain why the positive factors were can-celled out almost completely by the negative and the verum group showed merely a very slight amelioration which remained even below that of the placebo group.


Finally, we can only confirm Vithoulkas’ earlier statement that se-vere chronic headaches are not the ideal subject for a homeo-pathic trial of a few months duration.  Brigo’s splendid result re-mains an isolated case, but it has to be strictly reemphasized that it has never been exactly replicated. Other double-blinded ho-meopathic migraine studies with a different, more Walach-like setting, had to struggle with less severe, but similar difficulties to those experienced in the Munich study.  Nevertheless, a careful analysis shows that both these trials, in spite of generally being qualified as negative by Walach and others, yielded significantly positive partial results in favour of homeopathy.  Therefore, after definitively qualifying Walach’s trial as defective, it can be said that all three remaining double-blinded homeopathic migraine studies (Brigo, Straumsheim and Whitmarsh) at least partially showed positive results.

15. Walach’s non-causal or magical interpretation of ho-meopathy – a superfluous and inappropriate assumption
As mentioned above, the seemingly controversial results of ho-meopathic trials and particularly the negative outcome of his important Munich study led Walach to interpret homeopathy as being non-causal in the sense of the statistical Copenhagen inter-pretation of quantum mechanics.  This implies that homeopathy would function in a way that is similar to sorcery and could never be proven by classical scientific methods, such as double-blind trials.
However, a critical meta-analysis of homeopathic studies  shows the contrary. In spite of the fact that double-blind trials are not the ideal test for a holistic method like classical homeopathy,  the results of Frei,  Jacobs,  Reilly  and others together with the present definitive proof of the invalidity of the Munich headache study show that the efficiency of homeopathy can be satisfactorily proven providing the trial is well-conceived. This experimental evidence already makes Walach’s assumption superfluous.
Furthermore, Walach’s theory lacks sound foundations on the theoretical level, too. It is well known that Einstein, still consid-ered to be the major physicist in modern science, strictly refused to accept the purely statistical approach to physical reality postu-lated by Bohr’s Copenhagen interpretation of quantum mechanics right up to his death: “God doesn’t play dice!”. Well-founded al-ternatives have been elaborated by de Brogle and Bohm, and the real physical meaning of quantum theory still remains one of the most puzzling open questions in modern physics.  
It is unnecessary to say that Hahnemann, an undisputed repre-sentative of the Age of Enlightenment, in spite of his belief that the essence of potentiation is the activation of a universal life force slumbering in matter, strictly saw homeopathy as a rational science and not an occult mystery! Life-energy and science at Hahnemann's time were not yet as strictly separated as – unfor-tunately – they are today. Hahnemann’s renowned predecessor in bio-energetic medicine, F.A. Mesmer, whose healing magnetism is integrated in the Organon,  had already elaborated the very in-teresting physical theory that matter is nothing other than a dy-namic manifestation of life-energy.  
It is still too little known that this theory not only gives us a useful model of high-potencies but also a new possibility to elucidate the controversial hidden parameters postulated by Einstein, de Broglie and Bohm to reintegrate quantum mechanics into classical causal physics.  Therefore, it is a major mistake to connect the non-causal statistical interpretation of quantum mechanics, possibly one of the more important errors of modern physics, to homeo-pathic science, as Walach has done. - On the contrary, modern physics should learn from homeopathy!

1. The Munich headache-study by Walach et al. has to be consid-ered invalid by the following reasons:
- The data concerning the reactions of the patients have been in-terchanged for the most part and are interpreted in a clinically in-adequate manner.
- An important error of randomisation has been overlooked.
- The sensitivity of the study has been decreased substantially by intermixing migraine and permanent headache.
2. The effectiveness of homeopathy can be proven by double-blinded studies too, particularly also by migraine-studies, pro-vided the conception is appropriate. This makes Walachs non-causal respectively magical interpretation of homeopathy super-fluous.

Brigo B., Serpelloni G.: “Homoeopathic Treatment of Migraines: A Randomized Double-blind Controlled Study of Sixty Cases.”
 The Berlin Journal on Research in Homoeopathy Vol. 1 (1991), 2:

Frei H. et al.: „Homeopathic treatment of children with attention deficit hyperactivity disorder: a randomized, double blind, placebo controlled crossover trial.“
 European Journal of Pediatrics 27.7. 2005, Springer-Verlag,
Frei H. et al.: „Die homöopathische Behandlung von Kindern mit ADS/ADHS.“     
 Haug-Verlag, Heidelberg 2007.
Gaus W. et al.: „Die Wirksamkeit der klassischen Homöopathie bei
chronischen Kopfschmerzen.“
 Der Schmerz, 1992/6, S. 134 – 140.
Gaus W.: Biometrischer Auswertungsbericht der Studie „Die Wirk-samkeit der klassischen Homöopathie bei chronischen Kopfschmerzen.“
 Universitätsklinikum Ulm 1994.
Hahnemann S.: “Organon der Heilkunst.”
6., unveränderte Auflage.  
 Hippokrates, Stuttgart, 1955.
Jacobs J. et al.: „Homeopathy for childhood diarrrhea.“
  The Pediatric Infectious Disease Journal, Vol. 22 (2003), p. 229
 – 34.

Kivellos S., Vithoulkas G. et al.: “Observational prospective study of homeopathic treatment in patients with migraine, attending the headache clinic “G. Gennimatas” Athens General Hospital.“
 Cephalalgia, Journal of the Hellenic Headache Association,
 Volume 13, Number 3, July-September 2006.

Kösters C.: “Ueberlegungen zur Münchener Migräne-Studie.”
 Allgemeine homöopathische Zeitschrift (AHZ) 1/1998, S. 12 –
Mesmer F.A.: „Mesmerismus oder System der Wechselwirkun-gen.“
 Nikolaische Buchhandlung, Berlin 1814.
Reilly D.T. et al.: „Is Homoepathy a Placebo Response?“
 The Lancet, October 1986, S. 881 – 85.  
Reilly D.T. et al.: „Is evidence for homoeopathy reproducible?“
 The Lancet, Vol. 344, December 1994, S. 1601 – 6.
Rey L.: “Thermoluminescence of ultra-high dilutions of lithium-chloride and  sodium chloride.“
 Physica A 323 (2003) S. 67 – 74.
Seiler Hp.: „Der Kosmonenraum - Ansätze zu einer ganzheitlichen Betrachtung von Raum, Zeit, Leben und Materie aufgrund einer erweiterten Neuformulierung der Physik F.A. Mesmers.“
 Verlag für Ganzheitsmedizin, Essen, 1986.

Seiler Hp: „Die Entwicklung von S. Hahnemanns ärztlicher Praxis.“
 Haug-Verlag, Heidelberg 1988.
Seiler Hp.: „Raum, Zeit, Leben und Materie – Geschichte und neue  Perspektiven der Aetherwirbeltheorie.“
 In Emotion Nr. 12/13, Berlin 1997.

Seiler Hp.: „Offener Brief Seiler an Walach.“
http://www.dzvhae.com/portal/loader.php?navigation=43301&org =36300&seite=43308&PHPSESSID=95e2887ad5e05050b1106363bd2    1ad86     
(= Seiler 2006/1, easier to download at www.hanspeterseiler.ch)

Seiler Hp.: „Replik Seiler an Walach.“
http://www.dzvhae.com/portal/loader.php?navigation=43301&org=36300&seite=43308&PHPSESSID=95e2887ad5e05050b1106363bd2    1ad86
(= Seiler 2006/2, easier to download at www.hanspeterseiler.ch)

Sexl R. et al.: „Die Deutungen der Quantentheorie.“
 Vieweg, Braunschweig 1986.

Straumsheim P. et al.: „Homeopathic treatment of migraine.“
 British Homepathic Journal 89 (2000), S. 4 -7.

Vithoulkas G.: „Homeopathic treatment of chronic headache: a critique.“
 Homeopathy 2002/91 p. 32 – 34 (= Vithoulkas 2002/1).

Vithoulkas G.: Answer to Walachs critique of Vithoulkas.
 Homeopathy 2002/91 p. 186 – 188 (= Vithoulkas 2002/2).

Walach H. et al.: „Classical Homeopathic Treatment of chronic headaches.“
 Cephalalgia 17 (1997), S. 119 – 126.
Walach H.: „Magic of Signs: A Nonlocal Interpretation of Home-opathy.“
 Journal of Scientific Exploration, Vol. 13 (1999), Nr. 2, S. 291 –
Walach H.: “Wissenschaftliche Untersuchungen zur Homöopathie.”
 KVC-Verlag, Essen 2000.

Walach H.: “Response to Vithoulkas: homeopathic fantasies about science, a     meta-critique.“
 Homeopathy 2002/91 p. 35 – 39 (= Walach 2002/1)

Walach H.: “Reply to Vithoulkas and Oberbaum.“
 Homeopathy 2002/91 p. 188 – 189 (= Walach 2002/2)

Walach H.: „Antwort Walach an Seiler.“
http://www.dzvhae.com/portal/loader.php?navigation=43301&org=363 00&seite=43308&PHPSESSID=95e2887ad5e05050b1106363bd21ad86
(= Walach 2006).

Whitmarsh T.E. et al.: „Double-blind randomized placebo-controlled study of  homeopathic prophylaxis in migraine.“
 Cephalalgia 17 (1997), S. 600 - 604.
Prof. George Vithoulkas
International Academy of Classical Homeopathy
Alonissos, Nothern Sporades,
Zip code: 37005
Ms. Georgia Liakou, Secretary to Prof. George Vithoulkas
Tel: +30 (24240) 65142
Fax: +30 (24240) 65147
E-Mail: academy@vithoulkas.com

Dr. med. Hanspeter Seiler
FMH General medicine
Ancient medical director of the Bircher-Benner-Clinic Zurich
Im Dörfli
CH-8124 Maur  
Telephone: 044 980 47 80
Fax:  044 980 42 69
E-Mail: praxisseiler@bluewin.ch