Acupuncture and a cracking review paper

Going over the archives, Neil wrote this after finding a  review paper written by Professor Donald Marcus “Is Acupuncture for Pain a Placebo Treatment? An examination of the evidence“.  First published in The Rheumatologist, it is open access – which means you can read the full paper for free.

Acupuncture, some dodgy maths and a cracking review paper

I have a challenge for you. Imagine you’re in ancient China and you’ve had this idea that health and disease hang on the flow of energy through invisible energy pathways called meridians that can be manipulated by applying needles in certain specific points. How do you go about systematically validating this theory? How do you know where the points are and which combinations work?

acupuncture research

Some colleagues* and I were thinking about what it would have taken to develop acupuncture systematically by gathering basic case study level evidence. We are not mathematicians though and have made some cheekily broad assumptions and our maths could be wrong (since, to a man, we consider ourselves borderline innumerate). Here is what we came up with, but if anyone has a flair for this kind of thing then wade in and do better. It amused us greatly.

Here goes:

There are around 400 different acupuncture points in the early acupuncture texts.
If you stick in just 4 needles the possible permutations of different needle patterns are 25,217,757,600
If you stick in 10 needles the possible permutations are 9.4 to the power of 25
If you stick in 20 needles the possible permutations are 6.8 to the power of 51
If you stick in 200 needles the possible permutations are 2.1 to the power of 254

Let’s be generous and stick with just 4 needles.

The population of China when the first known Chinese acupuncture tome was published (200AD) = 60,000,000

25,217,757,600/ 60,000,000 = 420 is the number of treatments each member of the population must have to endure to collect n of 1 study data for all treatment possibilities.

The Chinese life span at that time (generously estimated without data at 50 years) would mean that the number of separate n of 1 studies that every member of the Chinese population at the time would need to have had each year in order to systematically test acupuncture is: 420/50 = 8.4. So every member of the population would had to have the same condition throughout their lives and then receive 8 separate treatments per year for every year of their life in order to get all these combos tested (for just one condition –  and acupuncture is recommended for a lot more than one condition).

But then we realised we have been too generous. To systematically develop the system from scratch you would not know a priori where the points are. Thus you would need to needle the whole body to find your hotspots. So here goes:

The average body surface area is 17,000cm2 (this is based in Europeans as we don’t have data from the Chinese population 200AD). Based on giving each acupuncture point a specificity of 1cm2,  if you stick in just 4 needles the possible permutations of different needle patterns are 35,000,000,000,000,000 (3.5 to the power of 15)

Following the same calculations as above the number of different n of 1 studies that every member of the Chinese population at the time would have to have had each day in order to systematically test acupuncture: 3,176 per day. That’s a treatment every 27 seconds of their waking and sleeping life just to achieve case study level evidence (which does not sit high on our evidence based hierarchy).

Our model is ridiculously lazy (but fun), in large part due to the fact that we scribbled it down on the back of a paper napkin in the coffee shop. It doesn’t factor in time for a start, and there has been a lot of that. But nonetheless I guess it is unlikely that the theory was developed systematically. What we must be left with then is a case of “special wisdom”; that is someone a long time ago was privy to an understanding of the body that others were not and that this authority underpins acupuncture.  As with most arguments from authority it would more than surprise me if it were true.  As Benjamin Franklin famously declared “It is the first responsibility of every citizen to question authority.”

Mathematical tomfoolery aside in the latest edition of “The Rheumatologist” Professor Donald Marcus has written a comprehensive assessment of the evidence relating to acupuncture for the treatment of pain. I would strongly recommend that you read this review as it is one of the best I have seen. The evidence that acupuncture is essentially a placebo treatment, the real meaning of the current plethora of  “acupuncture lights up the brain” studies, the discussion regarding what this means for the provision of this therapy and how you might address the issues with patients strikes a great balance and is presented with genuine clarity. It is also open access with no pay-wall so get stuck in….

*A big hug for John Cossar, Physio lecturer extraordinaire who boldly got out his abacus for the good of science.

About Neil O’Connell

Neil O'Connell 2As well as writing for Body in Mind, Dr Neil O’Connell, (PhD, not MD) is a researcher in the Centre for Research in Rehabilitation, Brunel University, West London, UK. He divides his time between research and training new physiotherapists and previously worked extensively as a musculoskeletal physiotherapist.
He also tweets! @NeilOConnell
Neil’s main research interests are chronic low back pain and chronic pain more broadly with a focus on evidence based practice. He has conducted numerous systematic reviews including some for the Cochrane Collaboration. He also makes a mean Yorkshire pudding despite being a child of Essex.
Link to Neil’s published research here. Downloadable PDFs here.


Donald M. Marcus (2010). Is Acupuncture for Pain a Placebo Treatment? An examination of the evidence. The Rheumatologist.  Free full text online


  1. John Quintner says:

    For those interested in the topic of “dry needling” of myofascial trigger points, please note that I have recently posted a provocative commentary on the National Pain Report (USA).

    In my opinion, the American Physical Therapy Association has a case to answer. Of course others may not agree with me.

  2. John Ware, PT says:

    @ June: “As for placebo effect; yes acupuncture may well stimulate a placebo response, bringing about a calming effect in the subconscious brain due to altered attention. Good.”

    Maybe not so good if you end up puncturing a lung or causing some nasty infection. There are many documented cases of both of these complications from inserting acupuncture needles.

  3. John Quintner says:

    Jill, needling parts of the body will certainly attract the interest of the brain, as will other noxious stimuli. I for one do not make fun of acupuncture as it has a definite place in other cultures. The “dry needlers” amongst the PTs have deliberately distanced themselves from traditional Chinese medicine and have thereby exposed themselves to criticism on the basis of the modern scientific knowledge they claim to possess. They have placed themselves in a parlous position where their “house of cards” is in danger of collapse.

  4. June Trenholm says:

    Well there is The Journal of Pain vol 14 no. 3 March 2013 to consider; Inserting Needles into the body: a meta-analysis of brain activity associated with acupuncture needle stimulation, Chae, Chang, Lee, Jung et al. Not all is dismal in the Acupuncture world. I find that when people make fun of acupuncture because of 3000 year old language and world views, it makes my skin crawl. Needles are a way of stimulating the nervous system; one way in many including the applied sham treatments. The meridian acupuncture points are often where there is a space between muscles that give easier access to the nerves. It is my understanding that initially these points were considered free standing points and that the meridian explanation came along later to help describe the relationships between the points and the story was in context for the people in that culture at the time. We could explain it today using pain science. This paper seems like a good study to me. As for point specificity, I think we need to wait until we have better methods for investigating brain function. As for placebo effect; yes acupuncture may well stimulate a placebo response, bringing about a calming effect in the subconscious brain due to altered attention. Good. Sometimes you just can’t reason with the conscious brain. I think placebo got a bad rap when it became synonymous with nothing being done. I’m not trying to turn people who don’t use acupuncture into raving advocates. I would like the investigations to continue.

  5. John Quintner says:

    @ Jill. I think you are referring to a meaningful engagement taking place between clinician and patient in the “third space”. I can see no rational scientific reason for one of the partners to insert needles into the other. What makes it worse is that money changes hands as a result of this pseudo-treatment.

  6. Whether its others experience of pain or effect of acupuncture we stand on the outside looking in. We judge another’s truth and lived experience from our own belief point. As a case manager I work with people with highly complex injuries resulting from accidents that have had devastating impact on their life. That many make incredible changes and move on, often from a stuck point, several years post accident is, I believe, a testimony to their own ability to mend and recover. What fascinates me is how they do that. How they gather their resources and at this time, on that day, release from stuck state and tension.
    Fear is there in all of them. Deep inside. Maybe for some a fear of loss of their identity, others a fear of the future, fear of feeling unlike their usual self. We all have fear, but my guys have a LOT of fear.
    I’m sure great warriors in Chinese warring times had fear. The elite Judoki and Rowers I physiod had great fears and needed a lot of psychological support. Maybe the Chinese warriors too felt some sense of support from the wise worker who massaged their limbs and pricked their skin with needles to ‘balance their system’. Maybe the whole experience did help them re-balance and reach a best point of homeostasis for healing to happen. Maybe that’s what we are doing as therapists but its not just as simple as that.
    Give the same person to three or four or five practitioners and they will respond to probably one. There is often a clear tipping point in the process of case management when the ‘right’ therapist gets involved on the ground. There is an almost palpable shift in the state of the person and they start to get better. All their senses shift, they start to relax, laugh, they move back to an adult place, and they report they feel better

  7. Andrew Cook says:

    Good old Confucius

    Placebo effect – what are the a priori assumptions for this? There is first of all a confusion in that there are two major types of placebo. First, the patient unconsciously wants to please the doctor or unconsciously has some agenda that gives the illness a higher priority than health. This occurs more frequently than you might think. So the medication merely has a temporary effect or the illness is displaced and arises in some other form. Second, the patient believes in the treatment and that belief brings about full (or at least substantial) and lasting remission or some substantial and otherwise inexplicable physiological change. Since all this goes on in the conscious and non-conscious mind and is related to both generic and specific belief systems, how do you measure how placebo-able each person is? If belief in the medication and the doctor prescribing it are so important, how do you assess each doctor (and group of doctors) for how capable they are at delivering placebo, and each interaction for congruity? The answer to both is – it is not done because it can’t be done, because there are already too many interdependent variables. So displacement is unquantified, the placebo-ability of the test population is unquantified, the ability of each prescribing doctor to be congruent enough or to give out appropriate body language to deliver an effective placebo is unquantified and the specific interactions are unquantified. All these are assumed to be “average”. The placebo effect itself is so ill defined that the various literature gives it an occurrence spanning almost an order of magnitude of % points. And in the case of displacement, how do you with the current medical (and research) model recognise when displacement has occurred?

    To add more confusion, if belief in medication can cause physiological change, then what about belief in something else? So how many people have a belief in something else (other than the medication or the medical help they are receiving) that affects their state of health, but that effect is included in placebo (i.e. belief in doctor or medication) statistics or even in effectiveness of medication? You can’t have it both ways – either placebo doesn’t exist at all, or it is belief in {a highly diverse set of factors, from medical to physical to personal to spiritual that are not considered worthy of data collection, and probably would be impossible to adequately define anyway} that brings about healing. Another unquantified variable.

    Take a look at medical trials, and if you change the placebo effect by just a few % points the results in many cases will look wildly different. On this basis, hundreds of billions of dollars of medications are bought and tens of millions of people are prescribed medication. And the existence or non-existence of health effects are measured. And furthermore, medical science, unlike any other branch of science, has decided that (this) one ill-defined arbiter of proof is a golden standard that has no peer and accepts no alternative.

    So I am amazed when a placebo controlled trial comes up with anything useful at all, because with all the a priori averaged and unquantifiable assumptions that exist even before the data begins to flow in, it’s not a lot better than random. And it’s called “science”. In effect, placebo controlled trials themselves are an expression of a pseudoscientific belief system based on an inadequately formulated and deeply conflicted model of the human body-mind.

    Medicine as a whole I have no problem with. Neither with medical professionals, who are remarkably dedicated to their patients. But “Placebo effect” as a purportedly scientific standard of measurement? No.

    “The study of fallacies … should attune the student to the omnipresent dangers to which we are exposed as a consequence of imprecise expressions – vague, ambiguous, or misdefined terms – or of unarticulated assumptions and presuppositions. [Toulmin 1979]

    John Quintner Reply:

    Andrew, now THAT is a real challenge that I doubt can be met by the thinking that characterizes our Western scientific community.

    Neil refers above to “special wisdom” of times long past. Perhaps it has always been with us but completely overlooked in our learned discussions. Let us call it “imaginability”.

    My friend the philosopher Horst Ruthrof sent me a copy of his latest book – Language and Imaginability (2014) – which I found difficult to penetrate. However, in the Introduction he suggests “If you are able to imagine what I am talking about and the way I am saying it, then there is meaning; if not, there is not. And vice versa, if I am able to imagine what others are talking about and the way they do so, an event of linguistic meaning has occurred.” He puts this commitment under the umbrella of the “imaginability thesis”.

    This is understood as both the human faculty of imagining and as a feature of things that can be imagined.

    When we apply these concepts to the placebo response, Andrew’s argument is more easily understood. Do we have any tools by which we can measure “imaginability” in ourselves and in others?

    Please accept my apology if the above comment is too abstruse.

  8. John Ware, PT says:

    In keeping with John Q’s important and appreciated references to clarity of language and avoidance of tautology in our use thereof, while at the same time giving a nod to Andrew’s predilections for all things Eastern, I provide this:
    “If names be not correct, language is not in accordance with the truth of things. If language be not in accordance with the truth of things, affairs cannot be carried on to success.” -The Analects of Confucius, Book 13, Verse 3

    The irrationality surrounding the perpetuation of needling in all of its forms represents, I think, the epitome of confounded language, fallacious reasoning, and just general obfuscation. Meanwhile, the lack of “success” of this intervention based on the mounds of high (and even low) quality empirical evidence available is startlingly clear. I challenge someone to come up with an intervention that has been more thoroughly disproven to work for persistent pain problems.

    As a profession, PTs in particular need to be rid of this nonsense.

    Paul Major PT Reply:

    Excellent comment John. I’m afraid I’m a little more cynical about these issues. Profit and self interest certainly plays a part in the perpetuation of needling Why else do we work so hard to work around the overwhelming evidence and market these treatments. There is not a private clinic here in Canada that does not have acupuncture on the menu.

    John Quintner Reply:

    You can fool some of the people some of the time, but not all of the people all of the time! The word is now out.

  9. john Quintner says:

    “The study of fallacies … should attune the student to the omnipresent dangers to which we are exposed as a consequence of imprecise expressions – vague, ambiguous, or misdefined terms – or of unarticulated assumptions and presuppositions. [Toulmin 1979]

    John, I see such a lack of knowledge as being part of a wider problem in the Pain Medicine literature.

    Consider these two statements that appear in the same paper [Littlejohn, 2007]:

    “… it is thought that trauma or injury to a component of the muscle-tendon unit (MTU) precedes regional pain syndrome in most cases.”

    “Inputs generated from deeply placed MTU spinal structures, such as ligaments or muscles, initiate the process of referred pain.”

    In the absence of evidence to support the existence of “muscle-tendon unit trauma or injury”, these beliefs are conjectural. Nevertheless the author asserts, as if true, that nociceptive input therefrom is involved in the genesis of (regional) pain.

    Another author [Staud, 2009] has engaged in circular reasoning leading him to an unsupportable conclusion:

    “Because nociceptive input from muscles is powerful in inducing and maintaining central sensitization, FM muscle abnormalities may strongly contribute to pain through the important mechanism of pain amplification.”

    These “FM muscle abnormalities” turn out to include “trigger points” – a construct that has been shown to lack scientific validity!

    Finally, an obvious example of speculation being passed off as scientific fact is the invention of the wonderfully adaptable “latent trigger point” – a veritable “will-o’-the-wisp”.

    We need to keep a lookout for these and other fallacies! They are not hard to find.


    Littlejohn G. Regional pain syndrome: clinical characteristics, mechanisms and management. Nat Clin Pract Rheumatol 2007; 3: 504-511.

    Toulmin S, Rieke R, Janik A. An Introduction to Reasoning, 2nd ed. New York: Macmillan Publishing Co., 1979.

    Staud R. Abnormal pain modulation in patients with spatially-distributed chronic pain: fibromyalgia. Rheum Dis Clin N Am 2009; 35: 263–274

  10. If acupuncture is purely placebo, how is it’s effectiveness in the treatment of pain in animals explained?

    John Quintner Reply:

    Marie, the Royal College of Veterinary Surgeons provides detailed information about the protocol necessary to obtain informed consent before a member administers any treatment to an animal. Is there any scientific evidence to support the claimed analgesic properties of needle acupuncture when used in this context?

  11. John Ware says:

    Yes, I agree, Andrew should avail himself of the burgeoning research in the area of placebo. Benedetti has written a fairly recent and excellent review. I don’t recall the “five elements” being mentioned, however.

  12. John Ware says:

    Hi John,
    Allow me to educate you about the business of rehab in the US. PTs must append “therapeutic” to everything we do in order to get reimbursed. It’s the language game that we play here. You know, where words stop having their original meaning, and are merely added for show.

    Sorry for any confusion.

    john Quintner Reply:

    John, thanks for the education in the business of rehabilitation in the US of A.

    Taking this language game to its logical conclusion suggests to me that PTs could also be reimbursed for “therapeutic nihilism” – reductio ad absurdum?

  13. Thanks John. Yes it is.

  14. john Quintner says:

    @ EG. The extract below is from a monograph by Max Charlesworth (1982) entitled “Science, non-science and pseudo-science.” I believe it is relevant to your comment.

    “It is important not to confuse science with a particular philosophical view or ideology of science which has become attached to it and which we might call ‘scientism’.

    Scientism sees science as the one and only true form of knowledge, as the very model of rational enquiry, and as the index of civilisation … scientism gives a central place to scientific values such as objectivity, neutrality, rationality, so that other values – the subjective, the personal, the emotional, the intuitive, the imaginative – are correspondingly downgraded.

    However, there is no necessary connection between science and this ideological view of science and we can acknowledge the place and value of science without subscribing to scientism … it is possible to recognise that science is one of the most astonishing inventions of the human mind without claiming that it is the one and only, or even the most important, invention of homo sapiens.”

    Charlesworth M. Science, non-science and pseudo-science. ABC Science Show Lectures. Deakin University Press, 1982.

  15. john Quintner says:

    John, can you explain what you mean by “therapeutic dry needling”? Sounds like a post hoc argument to me: it worked therefore it was therapeutic. When it does not work, do you then call it “non-therapeutic dry needling?”

  16. The meridians and points could not have been arrived at through experimentation – this was well explained using some maths. After a while the permutations and combinations would make one’s head spin.

    The question that should follow is: “How else can knowledge be obtained, if not through scientific experimentation?” B&M is excellent at refuting different approaches to health care, and I applaud that, but it’s possible to miss a huge chunk of learning if science is the extent of your investigation.

    [disclosure – I don’t use acupuncture or dry needling].


  17. John Ware says:

    Andrew states: “The effectiveness or not – as you probably know by now, I have very little positive to say about the concept of placebo as a scientific principle. It is far less well defined than meridiens.”

    Is this a joke? Where’s the emoticon? Five elements? What is this the Middle Ages? Can I ask one more question in this comment so that I make it an even five?

    Edward Grigoryan Reply:

    No really, this has got to be a joke. Starting with the false analogy of accurately determining the validity of acupuncture points to finding a home in Brisbane without a GPS, and ending with the statement on placebo, along with all the good bits on “feeling” the 5 elements and acupuncture meridians in between (I knew the knee pain I’ve been experiencing is a result of too much fire in my knee. Or wait, was it black bile… I confuse all of these exquisitely elegant concepts sometimes). Andrew, please read the chapter on placebo in Pain: The Science of Suffering by Wall. You owe it to yourself if you are involved in providing care to people who are in pain.

  18. Hi Neil
    Based on the experience of having practiced soft martial arts and bodywork for about 20 years …
    I honestly think that acupuncture and TCM theory in general arose as a result of people experiencing real processes in their own body (i.e. 5 elements) and by literally feeling meridiens and acupoints becoming active in certain circumstances. In detail. So the kind of mathematical analysis you have done is a bit like asking how someone can find their way home when there are 200,000 houses in Brisbane, 5000 streets, and they are totally ignorant of the geographic coordinate system that you base all your navigation on and they have no GPS. In fact, the very idea of this innate cordinate-free homing instinct reportedly posessed by many inhabitants of Brisbane is so freaky that it probably counts as pseudoscience.
    Yes – there is some fairly garbled theory that has been added to TCM over its 2500 year history, and the jargon is unfamiliar to our system of navigating the body. But even in the west many people have embodied experiences that – even though they are totally unfamiliar with TCM – are a direct experience of meridiens and/or 5 element processes.
    The effectiveness or not – as you probably know by now, I have very little positive to say about the concept of placebo as a scientific principle. It is far less well defined than meridiens. All the best.

    Matt Reply:

    How certain are you that this ” innate cordinate-free homing instinct” is merely wandering around lost until they spot something they recognise?

    Despite to difficulties of getting an experiment where you randomly abandon people in unknown places to see how they get home past the ethics board, I sure you have plenty of documented & published evidence to support this assertion of the ” innate cordinate-free homing instinct”.

  19. John Ware says:

    You know what your problem is Neil? You don’t acknowledge the difference between “acupuncture” and “therapeutic dry needling (TDN)”. 😉

    Can you get back with us on how frequent the population of, let’s say, New York City would need to undergo TDN if treating two myofascial trigger points in the upper trapezius for neck pain to achieve an n=1 level of analysis?

    No rush.