Acupuncture – the mysterious case of the missing razor

Acupuncture is all the rage in the treatment of pain. Recent clinical guidelines in the UK recommend it in the treatment of persistent back pain. This decision is somewhat controversial and has led to much discussion, because while the research in back pain suggests people feel somewhat better after acupuncture, it also demonstrates with clarity that they feel just as much improvement after receiving sham acupuncture. Nevertheless therapists seem to be queuing up for acupuncture training. There’s gold in them there needles.

So how about this acupuncture training? A paper just published in the Journal of Pain analysed the combined results from four big recent trials of acupuncture for chronic pain. They scrutinised this big data-set (almost 10,000 patients treated) to see what influence the characteristics of the treating clinician (primarily their level of training and clinical experience) had on the success or failure of treatment. Intriguingly no real influence of training or expertise was found. If you have just completed your training and started practising acupuncture for chronic pain then savour this moment. You may be at the peak of your (acupuncture specific) therapeutic powers.

These results add to a growing body of evidence that challenges the principles of acupuncture. A review across all conditions showed us that using the principles of Traditional Chinese Medicine (TCM – with all the trappings of Qi and meridians) to determine the location of the needles confers no added benefit to putting them in anywhere and across studies of pain real acupuncture does not perform convincingly better than sham acupuncture regardless of what type of sham you use. Even more interesting is that while clinical expertise and treatment characteristics make no difference to the clinical outcome, several studies show that the patients’ expectations of acupuncture (see here and here) and their beliefs about whether they are getting real or sham treatment do have an effect (see this great study by Barker Bausell -  the beliefs had more influence on success than the actual treatment received!).

So where does this leave us? Acupuncture is used because it has been around for a long time, is based on years of supposed wisdom and of course ”we know that it works”. But when we test the traditional principles underpinning acupuncture (the original reasons for doing it) they come up short. These days some clinicians follow the paradigm of Western Medical Acupuncture, ignoring the old TCM lore and based on the idea that this novel and specific sensory stimulus has specific physiological effects. But the evidence suggests that it doesn’t even matter even whether the patient can feel the stimulus, let alone where you put them and whether they penetrate the skin which implies that any effect might be rather non-specific .

Efforts continue to investigate the mechanisms of acupuncture. All of this reconceptualising and research effort is driven by one underlying construct that is accepted as truth: “We know that acupuncture works”. Of course this is a logical fallacy. You can’t know a treatment works until you’ve tested it in controlled trials and when we do that…..

What we can say with some confidence is that it does not seem to matter what you do with the needles, why you do it or how expert you are at it. If I were to be cheeky I might suggest that evidence-based acupuncture training could consist only of where you really mustn’t put the needles and the basics of antiseptic technique! We know that people feel a bit better after any convincing treatment (real or sham) and that adding a therapeutic ritual and empathetic clinician interaction enhances this effect (see here and here). Proving a negative is difficult but the weight of evidence strongly points in the direction that any clinical efficacy of acupuncture is due to the placebo effect.

Of course there may still be unknown reasons why this might not be so (e.g.“all of the sham treatments (even the non-penetrating ones and the ones that you can’t feel) have an active specific effect and are therefore also acupuncture” or “the trials are wrong” or “we just don’t know the mechanism yet”) but for me as the evidence base expands this becomes more of an exercise in creative thinking and starts to sound a little desperate. Worse, it pushes acupuncture into the realms of the unfalsifiable hypothesis. If we consider that old scientific principle of Occam’s Razor (what can be made with fewest assumptions is made in vain with more, or more simply, the simplest explanation that requires the fewest flights of fancy is the best one) then the placebo conclusion is the fairest.

The very human error common in the therapies is that when research does not confirm our opinion we are quick to find reasons for this failure that do not include “our theory was wrong”. Acupuncture research will no doubt continue to test all of these possible reasons but I would suggest that it’s now about time we had good look for that old lost razor (it probably fell down the awkward gap behind the sink).

rb2 large gray Acupuncture – the mysterious case of the missing razor
Witt CM, Lüdtke R, Wegscheider K, & Willich SN (2010). Physician characteristics and variation in treatment outcomes: are better qualified and experienced physicians more successful in treating patients with chronic pain with acupuncture? The journal of pain:official journal of the American Pain Society, 11 (5), 431-5 PMID: 20439056

MOFFET, H. (2008). Traditional acupuncture theories yield null outcomes: a systematic review of clinical trials Journal of Clinical Epidemiology, 61 (8), 741-747 DOI: 10.1016/j.jclinepi.2008.02.013

Madsen, M., Gotzsche, P., & Hrobjartsson, A. (2009). Acupuncture treatment for pain: systematic review of randomised clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups BMJ, 338 (jan27 2) DOI: 10.1136/bmj.a3115

Kalauokalani D, Cherkin DC, Sherman KJ, Koepsell TD, & Deyo RA (2001). Lessons from a trial of acupuncture and massage for low back pain: patient expectations and treatment effects. Spine, 26 (13), 1418-24 PMID: 11458142

Linde K, Witt CM, Streng A, Weidenhammer W, Wagenpfeil S, Brinkhaus B, Willich SN, & Melchart D (2007). The impact of patient expectations on outcomes in four randomized controlled trials of acupuncture in patients with chronic pain. Pain, 128 (3), 264-71 PMID: 17257756

Bausell RB, Lao L, Bergman S, Lee WL, & Berman BM (2005). Is acupuncture analgesia an expectancy effect? Preliminary evidence based on participants’ perceived assignments in two placebo-controlled trials. Evaluation & the health professions, 28 (1), 9-26 PMID: 15677384

Madsen MV, Gøtzsche PC, & Hróbjartsson A (2009). Acupuncture treatment for pain: systematic review of randomised clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups. BMJ (Clinical research ed.), 338 PMID: 19174438

Kaptchuk TJ, Stason WB, Davis RB, Legedza AR, Schnyer RN, Kerr CE, Stone DA, Nam BH, Kirsch I, & Goldman RH (2006). Sham device v inert pill: randomised controlled trial of two placebo treatments. BMJ (Clinical research ed.), 332 (7538), 391-7 PMID: 16452103

Kaptchuk TJ, Kelley JM, Conboy LA, Davis RB, Kerr CE, Jacobson EE, Kirsch I, Schyner RN, Nam BH, Nguyen LT, Park M, Rivers AL, McManus C, Kokkotou E, Drossman DA, Goldman P, & Lembo AJ (2008). Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome. BMJ (Clinical research ed.), 336 (7651), 999-1003 PMID: 18390493

All blog posts should be attributed to their author, not to BodyInMind. That is, BodyInMind wants authors to say what they really think, not what they think BodyInMind thinks they should think. Think about that!


  1. Nice work. Keep the Skeptic flag flying. Too many people are suckered by this – too many governments too, for that matter.


  2. Diane Jacobs says:

    Wonderful blogpost, couldn’t agree more – THANK YOU.


  3. Maybe acupuncture is overrated as a teaching – and all that is required is good old cutting or stabbing such as self-inflicted injuries make borderline personality disorder patients feel temporarily better. Yet, body areas are mapped at the thalamic level and I have entirely repeatable and strange cross links between body functions and phantom pain that are just as easily influenced by regional pain, injury or the like. For example, a full bladder is of no concern to me in that respect – but the VERY MOMENT OF OPENING THE URINARY SPHINCTER sets off a very strong phantom pain. Always, each and every time. As long as a neurological model does not explain that it may well fail acupuncture.

    - – -

    Adenosine A1 receptors mediate local anti-nociceptive effects of acupuncture

    Nanna Goldman, Michael Chen, Takumi Fujita, Qiwu Xu, Weiguo Peng, Wei Liu, Tina K Jensen, Yong Pei, Fushun Wang, Xiaoning Han, Jiang-Fan Chen, Jurgen Schnermann, Takahiro Takano, Lane Bekar, Kim Tieu & Maiken Nedergaard

    Nature Neuroscience (2010) doi:10.1038/nn.2562

    Received 16 March 2010 Accepted 27 April 2010

    Published online 30 May 2010

    Acupuncture is an invasive procedure commonly used to relieve pain. Acupuncture is practiced worldwide, despite difficulties in reconciling its principles with evidence-based medicine. We found that adenosine, a neuromodulator with anti-nociceptive properties, was released during acupuncture in mice and that its anti-nociceptive actions required adenosine A1 receptor expression. Direct injection of an adenosine A1 receptor agonist replicated the analgesic effect of acupuncture. Inhibition of enzymes involved in adenosine degradation potentiated the acupuncture-elicited increase in adenosine, as well as its anti-nociceptive effect. These observations indicate that adenosine mediates the effects of acupuncture and that interfering with adenosine metabolism may prolong the clinical benefit of acupuncture.


  4. Frédéric Wellens, pht says:

    Good comment. I am inclined to the same rational. This type of thinking could also apply quite neatly to many aspect of manual therapies as well. Now, why is it that the type of rational thinking you demonstrated in this comment seems to be the exception rather than the norm in our field of pain therapies?


  5. Graeme Campbell says:

    Hi Neil

    Sounds like TCM acupuncture is dead & buried under the weight of evidence that you have presented & my own impression is you are correct. However, I’d like to take issue with one point before we administer the last rites & I will play the devil’s advocate from the TCM standpoint. BTW -I have trained in traditional acupuncture & my in-laws are Chinese, some of whom practice TCM, but at heart I believe it to be nonsense in terms of the theory, so I have no cause to pursue (well as far as I’m aware that is).

    Your comment that “A review across all conditions showed us that using the principles of Traditional Chinese Medicine (TCM – with all the trappings of Qi and meridians) (my emphasis) to determine the location of the needles confers no added benefit to putting them in anywhere” The article you are referencing is a systematic review by Moffet. I wonder how accurate your statement about “all the trappings of TCM” actually is. That is significant because you/we risk accusations of not assessing real acupuncture, and are really only reviewing a bastardised western version of TCM, & condemning the non identical twin (TCM) for the sins of the other twin (sham TCM, which should not be confused with sham acupuncture).

    Allow me to diverge for one moment. Traditional Chinese Medicine (TCM) from my perspective is a completely different paradigm to Western Medicine & I am unable to see how you can apply Western diagnoses or outcome measures. For example, the TCM “pathogens” might be deficient chi or excessive wind, & this is diagnosed by feeling the three pulses at both sides of the wrist, observing the tongue & taking a history. Outcome of treatment is determined by a change in the TCM pulse, appearance of the tongue (objective signs) or change in symptoms. When we have ten patients with a western medicine diagnosis of something like diabetes, the TCM diagnosis may well reveal one has weak kidney meridian, another has excessive cold etc etc. To each of the diabetes patients, ten different TCM treatments might be dispensed as their tongue, pulse & TCM history dictates. In other words there is no or little common understanding that enables us to make comparisons as to appropriateness of treatment of to their efficacy.

    Faced with how to compare apples with oranges what do we do? We can take a cohort with a common western diagnosis & apply a westernised TCM approach & see what happens & assess that based on western concepts, which is reasonable, but it isn’t strictly using TCM; rather it is an attempt to marry the two & and arguably the best we can do in terms of applying scientific rigour.

    The articles that Moffet has referenced seem to do this, although I admit I haven’t had the time to track them down & read them apart from the titles except for one. This was- Scharf HP, Mansmann U, Streitberger K, Witte S, Kramer J, Maier C,et al. Acupuncture and knee osteoarthritis: a three-armed randomized trial. Ann Intern Med 2006;145:12-20. From what I can determine the “real acupuncture” they use is in Scharf’s article is exactly as I have described. Scharf et al say “Traditional Chinese acupuncture was “real” acupuncture according to Chinese protocols that specify the location and depth of needle placement in the treatment of knee pain”. Such a statement possibly demonstrates ignorance of TCM because there is no prescription for the treatment of knee pain because “knee pain” is not a TCM diagnosis, although some points might be commonly used in knee pain & be viewed as a “Chinese protocol”. I think there is a lot lost in translation & I suspect what has been used is a westernised reductionist formula of needle placement & from a TCM practitioner’s standpoint such a study would not be a reflection of real TCM.

    I am happy to attend the funeral of one twin and call me a nit picker if you like, but I’m not ready for the funeral of both twins just yet. My more immediate dilemma remains facing my wife when she gets home & admitting to her to being on the internet with you guys & not taking my TCM medicine whilst I’m home sick with the flu’!! Pray for me!


  6. Neil O'C says:

    Thanks all for your comments – really appreciated.

    @Wolf: I think the counter-irritation, diffuse noxious inhibitory controls (cause pain to relieve pain) hypothesis doesn’t work with acupuncture since penetration and skin sensation, and feelings of “De Qi” do not seem to mediate the effect:

    @ Frederic: Similar reasoning might well be applied to other modalities, and some are easier to sham than others. We should be willing to consider the possibility that manual therapy may continue to survive only because it is not possible to devise a decent sham condition. This review of modalities makes for interesting reading:


    I think your response is longer than my post! Thanks for all of these thoughts and devils advocates are always welcome.

    With regards to whether trials represent a fair test of TCM, a variety of approaches seem to be used in trials, some that have a predetermined treatment paradigm and some that allow the acupuncturist to choose. On this point though I would call up the idea of prior plausibility (see here:

    Ideas such as “weak kidney meridian” or “deificent chi” are prescientific and have more in common with superstition that rationality and one might ask why we might expect them, a priori, to be useful given our modern understanding of anatomy, physiology and pathology. These interesting if small studies found that TCM diagnosis is, perhaps unsurprisingly, a variable and unreliable feast:

    My personal view is that science is not culturally specific and since it is unlikely that 2 separate physiological worlds simultaneously co-exist, one where Western medicine has it right and one where TCM holds the answers, my money is on the one that possesses both plausbility and genuine evidence of progress (guess which). In fairness I imagine that many of us are no strangers to “excessive wind”.

    In terms of dry needling a recent review and meta-analysis found little evidence in support of this approach:

    The placebo question is a broader one. I have a blog post (in my head but not yet on paper) on this as I think it is a genuinely interesting and important topic for discussion. Maximising the efficacy of the clinical interaction is a great idea, the burning question is, is it OK to do this if we know we are being deceitful? I hope to write more on this soon….


  7. Graeme Campbell says:

    While I’m here, a few somewhat tangential thoughts come to mind as well:

    The article that was referenced by Kaptchuk – Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome – concluded “Factors contributing to the placebo effect can be progressively combined in a manner resembling a graded dose escalation of component parts. Non-specific effects can produce statistically and clinically significant outcomes and the patient-practitioner relationship is the most robust component”. There is another study, which I can’t recall the details of but it showed the best effect from acupuncture was on Chinese women delivered by a Chinese male acupuncturist in China, and the least efficacious scenario was delivery by a Western female acupuncturist treating a Chinese man outside of China. These sorts of findings raise lots of issues for me:

    1. Whilst it might be seen as an annoying artefact for research, should we utilise the placebo effect for clinical efficacy? I can’t see why not in acute cases. For chronic conditions, placebo has a sunset clause, so our grey hair, wearing a tie and a white coat, or other known powerful placebo generators would appear to be not so relevant.

    2. I work in a pain management centre & we use EBM in our pain management programs & aside from refugees from traumatised background or such likes, some ethnic groups consistently do not respond to treatments,. (We are allowed to ask politically incorrect questions here aren’t we? It seems to be something that many colleagues agree on but won’t admit to in public. Fear avoidance has many aspects does it not!?). Anyway, what is happening here? The example of the Chinese man in China is about placebo but does this have any bearing on how we might deliver more efficacious treatment to a broader patient base? (This is significant in Australia as I understand we are the second most ethnically diverse country after Israel). It begs the question for me as to how much a treatment should be tweaked to accommodate the individual, & the general sense I get about where EBM & the pain world is heading is that it is moving towards individual rather that group treatment.

    3. Our beliefs are important things as we can see from the Kaptchuk and other studies, both those of the patient and those of the practitioner, as well as the academic. Can we apply the same to our beliefs in EBM? Is there a level of scientism (with the sacred cow of EBM championing the cause) in how TCM is studied in the west?

    Another thing comes to mind her as well:
    4. Dry needling of trigger points – is this not the same as Acupuncture needles applied to Ahshi points. (add a ‘t’ to the Ahshi & you’ll get the idea of what they are from the patient’s perspective). Is this a case of where “acupuncture” is effective – an exception to the findings that Neil has reported?


  8. Graeme Campbell says:

    Nope – my word count puts your original posting as longer than mine. Start worrying if you’re more long winded than me!! (PS I did include all your references so maybe I cheated!!)

    Ideas such as “weak kidney meridian” or “deficient chi/qi” may be prescientific, implausible, or delusional – I’m not disagreeing, not at all. But I don’t think you should call it TCM & claim the studies had all the trappings of TCM if it isn’t including & assessing & basing treatment on such concepts such as qi and element pathogens (wind, cold etc). Just call it what it is – westernised acupuncture, or a westernised version of TCM. Just because most people call silver beet “spinach”, still doesn’t make it spinach (or English spinach as we call it in Australia)!! Whatever the case may be the evidence points to nothing more than a placebo effect from the studies that have been reported. Furthermore we can’t say the same about TCM because it hasn’t been tested although I would expect the results would be telling us the same thing. Can I put it another way – it is like saying you have tasted spinach & proven it to be a disgusting but your concept of spinach is really silver beet. You then tell everyone spinach is disgusting but those who call English spinach “spinach” won’t believe you because you only tasted silver beet. Does that make sense?

    I guess what I was really hoping to bring to the fore by mentioning dry needling is that perhaps there is a point of overlap between TCM & Western Medicine in the case of Ahshi point & Trigger Points, which appear to be equivalent & where you can truly apply a TCM, a westernised acupuncture or a strictly westernised approach; and honestly be able to assess the outcomes using either western outcome measures or TCM for that matter. In that case maybe we are able to compare apples with apples after all. Too many food analogies – Maybe I should go & have dinner – Oh no you won’t believe it! – Left over Spinach ricotta pie is all we have left – I tell the truth!! I will be mindful of all the excessive wind pathogens – after dinner & when doing future posts!


  9. This may not look too much like a discussion that is concerned so much with the way acupuncture may or may not work. It seems to be more a discussion aimed at supporting critique against having clinical UK guidelines recommend acupuncture for lower back pain. As that, acupuncture does not have to be discussed at any level of possible neurological cross-section or connection, projection or switch (which, here, may follow but seems to not be there at this moment). For that purpose though, a simple weighing of cost, risk and effectiveness are sufficient. And while acupuncture may not lead that type of statistics (which leads it, anyway, opiates?) let’s start from the very bottom. And it appears that, for example, I was treated with Gabapentin / Neurontin for severe chronic pain, a drug that was recommended by a specialist neurologist, that was paid for by insurance and that apparently has a proven effect in a staggering 12% (bwahaha) of applications in similar instances. For the amount of money that cost, for its side effects and the really useless time I spent with it I much preferred the acupuncturist: the whole setup of his attempt at faking Chinese medicine was far more entertaining, it cost about a fifth or less of the useless drug, it also had no effect for my pain, but at least the first time (I had complained about being tired) I afterwards felt extremely awake all afternoon, far better than after all cigarettes and coffees – and after I told the man second time around, he scoffed and due to acupuncture placebo magic I fell asleep then despite his ~ 40 needles stinging with every little move. The guy may really have been a acupuncture geek but he could not tame my pain nerves so he lost that race. So for weighing one not-understood useless costly method against another but similarly ill-understood but (12% is hard to beat in terms of uselessness, and if you now find some information that it is actually 14% or 18% you may *really* make a point here) somewhat less unsuccessful method, the type, severity and details of side effects and cost may end up being the decisive moments. – - And so later, when I had a slow-healing painful wound, I went back to acupuncture and it did entertain and distract me with a surprisingly good effect on the pain, with no side effects as to driving, sleeping, stomach or vision. Now inasmuch as *mechanisms* are concerned, I sure would like to know how pain is processed with regard to complex relays (thalamic regions, pain modulation) and when needling, placebo, joy, stress and others all activate similar endogenous opioid production, how are you going to discriminate these in a test? I figure that once you produce a bottom up brain map to sufficiently explain and treat phantom pain, you will be able to pin down acupuncture as well. Mirror therapy changes stuff in the brain for phantom pains – and not always for the good, mind you. Since I started doing it I had a massive drop in baseline phantom sensation AND a strong intensity increase in the phantom pain jabs – at once. Except when my bladder is full and I open the sphincter – some thalamic pathway then links hell 1 and hell 2 together and I get a fireworks of phantom pain. Probably if you then stabbed a virtual needle into my phantom pain you could reverse trigger the sphincter and get the waterfall to stop…. but until we see into these from a scientific viewpoint, I regard acupuncture as an un-understood attempt just as gabapentin for pain – except it’s cheaper, it can be definitely more fun and it really has less side effects if done correctly.


  10. Neil O'C says:

    OK I accept that one persons TCM may be different from anothers and so “all the trappings” may be over-egged, but it was probably “all the trappings as the acupuncturists involved in designing the trials saw it”. Often in these things it is very difficult to pin down what the “real” version of something is – the discussion elsewhere on BIM on trigger points is a good example. It’s hard enough in science let alone in tradition. On balance I think the data holds up that acupuncture using meridian theory/ TCM as practised in trials and the West is inneffective and I take your point that you are looking at possible alternative explanations. How often have we heard that the trials are wrtong because “they didn’t do it like I would do it”.

    Now I guess the question is this: is there a reason to think the other versions of TCM might be more effective or can we accept the existing analyses?

    Hope the pie was nice, good to have quality left-overs!


  11. LORIMER says:

    Great dialogue gentlemen – pleasure to read.


  12. Thought this (FItzgerald et al; Journal of Urology – August 2009 (Vol. 182, Issue 2, Pages 570-580, DOI: 10.1016/j.juro.2009.04.022) might be of interest to those who haven’t seen it, particularly due to the nature of the urethral pain described and the discussion on trigger points (TP) etc. Its a randomised multicentre feasibility study on manual therapy treatment of chronic pelvic pain(CPP) using TP, connective tissue massage.

    Clinically, in addition to explaining pain biology as we currently understand it, my collegues and I treat men and women who c/o urethral pain, CPP etc with neural and connective tissue techniques, TP release of the pelvic floor (rectally or vaginally). It seems particularly effective if we can reproduce the patients pain/symptoms. However using ultrasound, I can immediately “see” the changes that occur after TP release. Having just finished my PhD on the dynamic evaluation of the (female) pelvic floor muscles (PFM) using ultrasound, it does seem that we have an instrument that can reliably measure subtle differences in PFM function. In the future I reckon it would be kinda cool to measure the effect of manual therapy to the pelvic floor in this way. Although I recognise that I maybe reducing the threat level/listening/talking the patients better, IF (its a big if) I was able to keep my mouth shut, it may have the possibility to illustrate any physical changes that occur due to manual therapy intervention. Knowing how tricky it would be not to have some element of chat, we could have one arm of the trial explaining chronic pelvic pain as well as physical therapy, and the other arm physical therapy only, measuring PF with ultrasound pre and post intervention. Would be cool to combine that with brain imaging pre and post treatment :-) What do you reckon?


  13. Your story was really inomfratvie, thanks!



  1. [...] out this interesting article on the effects of acupuncture on patient satisfaction and health and decide for [...]

  2. [...] that it is very effective. I use placebo freely at home but is it right to do this in the clinic? In a recent post I suggested that we can be more confident that acupuncture is essentially a placebo treatment. [...]

  3. [...] O'Connell,Pain,Research,Treatments TweetHaving written a number of posts on acupuncture (see here, here, and here) I guess my particular biases are reasonably apparent. So imagine my surprise when [...]

  4. [...] active treatment since these networks are also implicated in the placebo response. Also, as I have oft repeated, given that our best evidence clearly indicates that it really doesn’t matter what you do with [...]

Speak Your Mind