Nature Scientific Reports publishes peer-reviewed paper showing efficacy of homeopathy

October 24, 2018 • 10:30 am

Reader Ryan (aka “vampyricon”) recently called my attention to this paper in Nature’s Scientific Reports (reference at bottom, click on screenshot below to read it online, and get pdf here). It’s important because it claims to demonstrate that homeopathy works. In this case, a homeopathic dilution of up to 10-30 of a solution of poison oak (Toxicodendron pubescens) was observed to reduce oxidative stress and other processes associated with pain in cultured cells, as well as reducing the pain response in rats.

 

You can read the paper for yourself (I’ve only scanned it), but also read the Nature News and Views article below that points out potential weaknesses of the study and criticisms by other scientists.

I’m loath to dismiss this result from the outset as I haven’t read the paper and at any rate am not professionally equipped to judge the metrics; we should always be careful not to immediately dismiss results that go against our biases. But of course there’s a huge reason to distrust studies like this, because dilutions so extreme can contain no active molecules of whatever substance is supposed to be working. That is, there’s no physical reason one can imagine that would cause such dilutions to work, apart from the unsubstantiated claim that somehow the dilutions leave their imprint on water, or in this case on the ethanol used to dilute the plant powder.

Among the problems cited in the Nature report are these:
  • The pain assessment study (on only 8 rats; a small sample) was not done blind, so that those who assessed the pain response could have known which treatment they were giving. I consider this a serious problem.
  • The results may not be generalizable to humans since pain was measured by observing how rats withdrew their paws from hot or cold stimuli. This is not a serious issue in my view.
  • There were discrepancies between the figures, including identical pictures that supposedly describe different experiments, as well as verbal descriptions of significant effects of dilutions that were more dilute than concoctions shown in the figures themselves.
  • There are reports that the same data were used in two different experiments.
  • One critic says that some authors of the paper, including Patil (see below) published another paper two years ago in the same journal that also had “inappropriately duplicated images”. That’s worrying.

In response, author Chandragouda Patil says that these mistakes were “unintentional” and the results of typos. The report has, according to the website, been updated, and Patil says that the experiments were done “with the utmost integrity”, with no ulterior motive to find pro- or anti-homeopathic results.

At any rate, I’ll leave it to readers with appropriate expertise to evaluate the paper itself. It was peer-reviewed, however, which means that homeopaths can now claim vindication for their whole practice, though other studies show no effects. I myself think that it can’t be right on first principles, since dilutions containing no efficacious molecules cannot be efficacious. Unless there’s some principle of chemistry of physics that has completely eluded us, something’s wrong. I suspect that, like the faster-than-light neutrinos that turned out to be the result of a loose measuring wire, this too shall pass. But the field will sort it out.

________

Magar, S., D. Nayak, U. B. Mahajan, K. R. Patil, S. D. Shinde, S. N. Goyal, S. Swaminarayan, C. R. Patil, S. Ojha, and C. N. Kundu. 2018. Ultra-diluted Toxicodendron pubescens attenuates pro-inflammatory cytokines and ROS- mediated neuropathic pain in rats. Scientific Reports 8:13562.

78 thoughts on “Nature Scientific Reports publishes peer-reviewed paper showing efficacy of homeopathy”

  1. Whether the result can be replicated is key. The small N and the absence of a blind process in the assessment of the pain response suggest caution in judgment about these results, caution that most homeopathy fans will throw to the wind.

    IF this results is replicated by many other experiments with a large number of subject and a blind process to assess pain, THEN it’s time to take it seriously as evidence for some unknown effect. For now it’s an interesting curiosity.

    1. I agree, those two factors leap out. I’m amazed that this passed peer review. The small sample size and methodologic issues would seem to preclude acceptance (our host believes only about 15% of submitted papers warrant acceptance, I doubt this would make the cut)

      The use of the same image to describe different experiments is also a red flag but perhaps less obvious without the post hoc digital analysis that’s been done.

      I’ll reserve final judgment, but I’m not heading off to buy homeopathic remedies based on this.

  4. Preparation of RT ethanolic extract and dilutions
    The procedure prescribed in the monograph of Indian Homeopathic pharmacopoeia was followed for the preparation of RT extract and its ultra-dilutions except the characteristic successions used in preparation of homeopathic dilutions. Dried and coarse powder of RT leaves was pulverized. Exactly weighed (10 gm) powder was mixed with 100 mL of ethanol (70%) and kept in the glass jar for cold maceration up to 7 days with occasional shaking during each day36,54. On 8th day, mixture was filtered through Whatman filter paper and the filtrate was used as an alcoholic extract of RT. Various dilutions of extract in ethanol were prepared to obtain the final RT concentrations of 1 × 10−2, 1 × 10−4, 1 × 10−6, 1 × 10−8, 1 × 10−10, 1 × 10−12, 1 × 10−14, 1 × 10−16, 1 × 10−18, 1 × 10−20, 1 × 10−22, 1 × 10−24, 1 × 10−26, 1 × 10−28, 1 × 10−30, 1 × 10−32, 1 × 10−34, 1 × 10−36 (Fig. 7).

    So, they diluted the extract in ethanol… and then reported reduced pain responses in obviously drunk rats? They got the rats drunk. Scrape away all the sciencey crap and they basically admitted to making herbal gin and then giving it to rats and recording a reduction in reactions to hot/cold stimulus. Forget all the problems with the sample size and blind trials, this one just gets a plain ‘ol, “No Shit Sherlock”.

      2. I personally know some who use alcohol — even for children. (I once worked in a Waldorf/Steiner School that had such a concoction in their first aid kit.)

        1. As far as I am aware, the first extraction could well have been a strong solution in alcohol (Tinctura madre or mother liquor). That is the starting point also for making an extract for “normal” use as a herbal preparation.
          This is how laudanum is and was prepared (tincture of opium, sometimes sold as “black drop”).

          I’m assuming that further serial dilution of the mother liquor will have been made using water and not alcohol.

          It is not clear to me what they were using as control doses (they ought perhaps to have two controls:
          a) similar serial dilutions of alcohol but not containing RT and
          b)serial dilutions of a volume of water only but equivalent to the original alcohol sample used for the mother liquor extraction).

          This would allow three ranges of doses allowing distinction in the results between:
          a) the effect of the drug itself,
          b) the serially diluted alcohol alone and
          c) of serially diluted water alone.

          The absence of double-blind testing of randomised samples and dose is rather damning.

          The key would be the statistically significant difference between sufficiently large groups with high reproducibility.

          I would be curious to know who funded the project and what their expectations might be in relation to outcome.

    3. I dunno, but if they get the poison oak the hell outta there completely, they might have the makings of a decent drink. 🙂

  5. I’ll just put it out there – a dilution of 1 in 10 -30 of anything is a joke.

    I’ll now go eat my filet mignon that is diluted to that level.

    1. First thing that came to mind was Avogradro’s Number, which only goes to ~ 10**23.

      Now they’re talking 10**-30? Yeah, right. We’re talking minor pieces of atoms here. REALLY minor pieces of atoms.

      1. Hence the expression, “You have an Avogadro’s chance in hell of finding one in that sample”

    1. The 10^-36 part of 1 [1 representing the whole of something] isn’t any particular size because it’s a ratio. It all depends on the size of the starter pie you are dividing into 10^36 parts!

      If the pie was all the sand grains in Earth you would need all the sand from 1,000,000,000,000,000,000 Earths to do the dilution [division] – you would end up with 10^36 piles of sand each pile consisting of one grain of sand

    2. I’m not a physicist either, but

      1.0 x 10-30 = 0.0000000000000000000000000000010

      Relatively speaking, you might as well be drinking pure alcohol. 30 should be superscripted.

  7. Let the pros hash this one out. In the meantime, I’ll take it with a tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny, tiny grain of salt.

  8. It should be noted that the Indian Government set up Central Council of Homeopathy & it’s a part of life. It also fits well with certain Hindu beliefs such as ‘like cures like.’

    The five institutions that vomited up these ‘researchers’ – note the name of the 4th one:

    [1] Department of Pharmacology, R. C. Patel Institute of Pharmaceutical Education and Research, Shirpur- 425405,
    Dist. Dhule, Maharashtra, India.

    [2] School of Biotechnology, Kalinga Institute of Industrial technology (a deemed
    to be University), Campus-11, Patia, Bhubaneswar, Odisha, Pin-751024, India.

    [3] SVKM’s Institute of Pharmacy,
    Dhule-424001, Dist-Dhule, Maharashtra, India.

    [4] Janmangal Homeopathy and Wellness Centre, Bopal, Ahmedabad,
    Gujarat, 380058, India.

    [5] Department of Pharmacology & Therapeutics, College of Medicine & Health Sciences, UAE
    University, Al Ain, UAE.

    I did a search for homeopathy-style papers from the above addresses, but not much. However when I searched for “ultra diluted” instead of the variations of “homeopathy” I got plenty of hits [many overlaps though]. I couldn’t be bothered to tabulate my results as it’s a lot of work for nil return.

    A very dilute effort from me!

    2. Still a laudible effort.
      ‘Successions’ are here performed by banging your head against a wall. This is believed to imprint and amplify the probable delusions.

      1. I thought it was ‘succussions’? Autocorrect strikes again?

        Actually that word always makes me think ‘concussion’, which, in the case of heads and walls, is probably apt.

        cr

    3. certain Hindu beliefs such as ‘like cures like.’

      As someone who has lived in India all my life, I have never heard of any such “Hindu belief”. Of course, many Indian languages have their own idioms with roughly the same semantic space as “it takes a thief to cure a thief” or “like cures like”. But absent compelling evidence, the existence of such idioms are not very useful for explaining government interest in homeopathy in India.

      A much more parsimonius explanation is legacy and lobbying.

      1. According to various sources the Bhagwat Purana refers to the “law of similars” in its Sanskrit text. But, it would not surprise me to find out that this connection is due to homeopaths scouring ancient literature looking for phrases that could be interpreted as being in support of homeopathy. So thank you for the clarification. There are hundreds of examples of the above claim on the net, but they do fit your description – the few I’ve checked.

        1. As far as I am aware the “doctrine of signatures” is an old medical idea which is found in Indian (Ayurvedic herbal medicine), Chines and even European (Galen).
          The idea was that plants are “signed” according to their nature and their purpose for man (possibly as designated by the gods).

          I believe that “like cures like” is connected with this.

          Ayurvedic medicine is mostly based on actual herbal extracts used in a dose related way (herbal medicines at actual therapeutic dose can also be dangerous since the plants are often poisons).
          Some modern drugs have been derived from Ayurvedic medicine.

    1. Think that caveat’s been an key part of the scientific method since our biases told us the earth was flat and the big orange orb traversed a quotidian path across its sky.

      Personally, I’m still tryin’ to wrap my biases around the double-slit experiment. 🙂

    2. “we should always be careful not to immediately dismiss results that go against our biases.”

      While this is true, it takes no special education (or intelligence, for that matter) to conclude that a “homeopathic dilution of up to 10-30 of a solution of poison oak” + n=8 sample size + study isn’t blind = rubbish, squared.

      It should not be required, in a roomful of science leaning intellectuals, to point out that an experiment conducted improperly invalidates the result.

      Let’s be open minded, yes, but not so open minded that our brains fall out.

    3. I agree heartily with saying the intended, except “biases” should surely not be used. I’d prefer something like ‘strong theoretical expectations based on a very well established fact of the existence of atoms’!!

      After all, various dictionaries in defining ‘bias’ use phrases like ‘prejudice’, and ‘unfair’, and ‘personal opinion’, and ‘disproportionate weight in favour of’, etc.

    4. That might be my fault. In the email I said

      I don’t think this will change my opinion of it significantly. Although it is evidence for homeopathy, it still contradicts everything we know about physics and chemistry so it is more likely that this is a fluke rather than a result that will hold up.

      -Ryan

  10. “The more you dilute it, the stronger it gets.”

    Normally, when you’re dealing in pseudoscience or spiritual woo there’s something intuitive behind the misunderstanding. With vaccine denial there’s the folk wisdom involving poison being poison regardless of the dose. With Creationism there’s the underlying familiarity with the experience where parts don’t work together till we put them together for a purpose. Most bullshit is a matter of common sense applied indiscriminately.

    Till we get to homeopathy. Where is the tiny little analogy with something we “know” — if we’re talking about “the more you dilute it, the stronger it gets?” Nobody instinctively cuts down on the lemons if they’re making lemonade for more people than originally expected. It’s counterintuitive.

    The only examples I can think of involve feelings and essences. “Absence makes the heart grow fonder.” “The less variation from a standard X, the purer and stronger the person/object is.” “If you believe on little evidence, that faith is a more powerful conviction than if there were reasons.”

    From what I can tell Homeopathy seems to be pulling its plausibility from that sort of thing. These “similar” bits of wisdom of common experience grant it credibility. Which, if so, is another problem.

    1. The ‘ancient wisdom’ of ‘Like cures like’ (supposedly from Hippocrates), seems to be the great unifying principle here. Hanneman just perfected the delivery system for it.

      Unfortunately the ancient wisdom of Aristotle’s list of logical fallacies still hasn’t filtered through to these people.

    2. My theory (which is mine, but probably not mine alone) is that homeopathy came from a time when many accepted medical treatments were often worse than doing nothing, so those patients who received water with a minute amount of poison were still better off than their fellows. And of course, as the poison was diluted further, the deleterious effects became smaller and smaller.

      1. I think your theory has merit. Homeopathy was created in 1796 by Samuel Hahnemann, based on his doctrine of like cures like. In that time medicine was in it’s dangerous infancy and probably as likely to kill as cure.

  11. “dilutions so extreme can contain no active molecules of whatever substance is supposed to be working”

    Furthermore, it’s easy to forget that the substances themselves used in homeopathy, even if undiluted, also won’t have any therapeutic effect. They base the entire treatment on twisted logic (like must cure like, because it simply must be so), and not on physiology.

    1. As in religious sects like Mormonism and Scientology, though, it is an excellent strategy for filtering your potential client base. The customers arrive pre-screened for gullibility.

        1. Gullibility is the quality of believing in things too quickly. Someone with a lot of gullibility can be easily tricked. … in response, then you are proving to be gullible: a little too ready to believe what you hear. The word for this quality or tendency is gullibility. Being superstitious shows gullibility.

          1. My question was more why Mormons would be more susceptible that any other religious organization. I see Mormons today as mainline, and not as a small sect.

          2. LDS and Scientology suffer from the same defect. They are very recently created systems, which leaves them open to rather easy investigation and refutation. Only the least skeptical can swallow them hook and line. The origins of traditional religions are cloaked in the fog of time, which gives them ersatz authority.

          3. I’ve forgotten who said the difference between a cult and a religion is that in a cult, there is someone at the top of the hierarchy who knows it’s all a giant scam. In a religion, this person is dead.

  12. Looking at Fig 4 and the text above it, the two do not seem to agree.

    The text says the “sham” group was 23.2 +/- 3.8, but in the picture it looks more like +/- 16. It says in the picture that the +/- is SEM, by which I assume they mean standard error of the mean (standard deviation over root sample size).

    Furthermore, taking the “sham” group of 23.2 +/- 3.8 and CCNI as 15.9 +/- 2.1, and again assuming these are SEM values, a 2-sample t test gives (I make it) t_obs = 1.68, so definitely p>0.1, not p<0.001 as they claim for the full ANOVA with Bonferroni. This is cherry-picking this pair, full ANOVA would give more groups with the higher mean, but Bonferroni corrections would make it more difficult to reject the null.

    It is possible that the result they quote is because they have used a paired analysis, but only reported the unpaired results.

  13. Many years ago, my laboratory performed a homeopathic test of science publicity, by cutting up pages of reputable scientific journals to detect the smallest bit that would create publicity in the popular media. We found that a scrap from Nature containing only a comma could elicit a returned telephone call from several magazines, while a tiny fragment of the “N” on the title page could induce a tremor in The New York Times. The only physical explanation we could offer for these results was that the bits of paper retained a molecular memory of the scientific journal they were cut from.

  14. The pain assessment study (on only 8 rats; a small sample) was not done blind, so that those who assessed the pain response could have known which treatment they were giving. I consider this a serious problem.

    For me that is game over for any study of this type.

    The only question to ask is how it got published at all.

  15. The normal and sham operated rats received 1 ml of saline daily, while the treatment group received 1 ml of distilled water and 0.1 ml of the extraction diluted in ethanol. For a valid comparison, the normal and sham should have also received 1 ml of distilled water with 0.1 ml of ethanol without the extract.

    1. Good point.
      If you assume 200 gm rats (their range was 170-220) and humans at around 80 Kg (175 lb more or less), so you scale up by 400, you’re looking at a human equivalent dose of 40 mL ethanol, or around 3.5 oz of 80-proof booze. I don’t know about rat responses to alcohol, but 3.5 oz of 80-proof seems likely to me to affect the average human’s perception of pain.

    2. Yes indeed. And the ethanol without the extract is just as likely to have more stuff dissolved or otherwise floating around in it than the ethanol with extract.

  16. Ethanol at 20C is .7893 g/cm^3 and has atomic weight about 46. So there are roughly .7893*6.02*10^23/46 ~ 10^22 molecules of ethanol per mL. 10^30 molecules of ethanol takes up 10^8 mL or 100,000 liters. A dilution 10^-30 of magic treated ethanol in regular ethanol means one molecule of magic treated ethanol per 10^30 molecules of regular ethanol, i.e., 100,000 liters. One can’t get a dilution of 10^-30 without having 10^30 molecules of the diluting substance unless the experimenters are magically using fractional molecules.

    1. Excellent point, though you are forgetting that the woo woo quantum spin “imprint” of the macerated extract on the original ethanol molecules is transferred via quantum tunneling through nano wormholes to the fresh diluting ethanol molecules, you see…

  17. N=8 is weak sauce for anything. N=8 for something that would result in nobel prizes, no frigging way.

    Incidentally, why do homeopaths never do the “basic science” of homeopathy *first*? (Rhetorical question, I know, but …)

  18. Two notes:
    1) I suppose it is remotely possible that something akin to homeopathy works in one out of, say, 50 cases. (A number I pulled out from under my armpit.) If the researchers have hit on one, it does not prove that homeopathy in general is sound.

    2) Re: No known sound mechanism i.e. “That is, there’s no physical reason one can imagine that would cause such dilutions to work,”. This is akin to something in the mid-1980s anti-astrology statement produced by “The Humanist” which Carl Sagan said was the worst possible reason to be against astrology.
    Carl Sagan and Paul Feyerabend were the two allies of that publication that declined to sign their anti-astrology manifesto. Sagan discussed his reasons both in a letter to “The Humanist” (which appeared in the following issue) and his book “The Demon-Haunted World”. He cited multiple theories that turned out to be true originally rejected because there was no conceivable causal mechanism by which they could work.

    1. I have deep admiration for Sagan, but I think he’s wrong here.

      It isn’t just that there is no known mechanism, but that basically all chemistry is wrong if homeopathy does anything whatever related to what proponents claim. There is a wall of anti-evidence to get rid of *first*.

      This is similar to perpetual motion machines. (Which many countries, the US and Canada included, will not even *consider* for patent applications because the background knowledge we have makes them so massively implausible.)

      If only to save money and time on the part of everyone, the “appeal to background ridiculuousness” should be acceptable.

  19. Never purchase a homeopathic remedy unless the product information sheet states the degree of succussion the remedy was subjected to during manufacture. It could be very much more powerful than you realize.
    https://en.wikipedia.org/wiki/Homeopathy#Preparation

    Hahnemann found that undiluted doses caused reactions, sometimes dangerous ones, so specified that preparations be given at the lowest possible dose. He found that this reduced potency as well as side-effects, but formed the view that vigorous shaking and striking on an elastic surface – a process he termed Schütteln, translated as succussion – nullified this.[87] A common explanation for his settling on this process is said to be that he found preparations subjected to agitation in transit, such as in saddle bags or in a carriage, were more “potent”.[58]:16 Hahnemann had a saddle-maker construct a special wooden striking board covered in leather on one side and stuffed with horsehair.[88]:31 Insoluble solids, such as granite, diamond, and platinum, are diluted by grinding them with lactose (“trituration”).[58]:23

    The process of dilution and succussion is termed “dynamization” or “potentization” by homeopaths.[9][89] In industrial manufacture this may be done by machine.

    So always shake the bottle in case the active ingredients settle to the bottom in transit. Or don’t shake the bottle. If the active ingredients end up on the bottom, then the solution is even more dilute, and therefore even more powerful.

    P.S. And always check the use-by date.

    1. According to Hahnemann, it should not only be shaken but also tapped against a Bible, preferably leather bound. Homeopaths have dropped this idea for some reason.

  20. This study stinks on so many levels. Total rubbish! Didn’t read acknowledgments. My bet for “Funded by …” Gywneth Woo Paltrow.

  23. FYI, alternative medicine researcher Edzard Ernst has weighed in on this paper:

    I am utterly under-whelmed by in-vitro experiments (which are prone to artefacts) and animal studies (especially those with a sample size of 8!) of homeopathy. I think they have very little relevance to the question whether homeopathy works.

    But there is more, much more!

    It has been pointed out that there are several oddities in this paper which are highly suspicious of scientific misconduct or fraud. It has been noted that the study used duplicated data figures that claimed to show different experimental results, inconsistently reported data and results for various treatment dilutions in the text and figures, contained suspiciously identical data points throughout a series of figures that were reported to represent different experimental results, and hinged on subjective, non-blinded data from a pain experiment involving just eight rats.

    Lastly, others pointed out that even if the data is somehow accurate, the experiment is unconvincing. The fast timing differences of paw withdraw is subjective. It’s also prone to bias because the researchers were not blinded to the rats’ treatments (meaning they could have known which animals were given the control drug or the homeopathic dilution). Moreover, eight animals in each group is not a large enough number from which to draw firm conclusions, they argue.

    As one consequence of these suspicions, the journal has recently added the following footnote to the publication:

    10/1/2018 Editors’ Note: Readers are alerted that the conclusions of this paper are subject to criticisms that are being considered by the editors. Appropriate editorial action will be taken once this matter is resolved.>/blockquote>

    https://web.archive.org/web/20181012123512/https://edzardernst.com/2018/10/is-the-latest-proof-of-homeopathy-fraudulent/

    (link from the Wayback Machine because his website seems to be down atm)

Comments are closed.