Few people would argue that Cochrane reviews tend to be the most rigorous, independent and objective assessments of therapeutic interventions we currently have. Therefore, it is relevant to see what they tell us about the value of acupuncture.

Here is a fascinating overview of all Cochrane reviews of acupuncture. It was compiled by the formidable guys at ‘FRIENDS OF SCIENCE-BEASED MEDICINE‘ in Australia. They gave me the permission to publish it here (thanks Loretta!).


Considering this collective evidence, it would be hard to dispute the conclusion that there is no convincing evidence that acupuncture is an effective therapy, I believe.

What do you think?

40 Responses to Acupuncture: there is no convincing evidence that it is an effective therapy

  • Unfortunately anecdotal evidence and wishful thinking are more convincing than any amount of scientific evidence for many people.

    • Just like spinal surgery and arthroscopic surgery Leigh. The lack of clinical evidence yet it still goes on. What do you suggest? Wait for the research? These surgeons are criminals in my opinion. All surgery must undergo testing before they start cutting. At least the drug companies run paid trials till they get a positive result.

      • Don’t worry, the surgery companies will run some paid trials too and we will have those positive results soon as well. There is always room for improvement these days.

      • Yes Roger.
        Like you, we do also think acupuncture should be forbidden just like cupping, blood-letting or any kind of injury that has been shown to be useless, even including certain types of surgical interventions that have been shown to be futile, as you point out quite correctly even if you seem to have missed the fact that surgical interventions are indeed studied and evaluated.
        It is a bit harsh to consider all acupuncturists and other purveyors of useless injury to be criminals. Most of them are well meaning but misguided and stuck in a combination of romantic infatuation, cognitive dissonance and financial dependency, with a dash of the Dunning-Kruger effect as they do not have the biomedical knowledge and training that helps understanding why injury is unlikely to confer any medical benefit.
        The problem arises when these people, be it acupuncturists or sugeons or whatever, persevere despite being presented with facts and evidence that their actions are doing more harm than good. Then it may become necessary to resort to judicial measures. This applies also to medical professionals in most civilised parts of the world where health care is properly supervised.

        • Oh I’m not worried about acupuncture. Its harmless and either useless or has some effect. This is not a dangerous medicine compared to the abusive and invasive methods of modern medicine. Western medicine has a lot of potential but is still like a small child running through a toy shop. It is young. We won’t see an effective medical model for another 100 years. The narrow thinking on this blog is certainly more of a block to healthy growth and development than helpful.

      • I know nothing about surgery. All clinical trials by vested interests should be carefully vetted. If what you claim is true then start a blog to inform the world.

        When I first started looking at the evidence base for acupunture I was somewhat sceptical but thought sticking needles into flesh could conceivably have some benefit even if it wasn’t obvious how. The Chinese mystical mumbo-jumbo was obvious rot. I was motivated to investigate for myself when a friend embarked on an acupuncture degree. She had a high regard for a doctor who had personally endorsed it. I was curiously doubtful but open to solid evidence if it existed.

        It became obvious that there was an ocean of trials without a drop of anything worth drinking. What’s more, claims being made by acupuncture organisations were clearly untrue. The most egregious example being that the WHO recommends acupuncture.

        Big pharma has many hawks keeping a watchful and critical eye on them. Rightly so. Caveat emptor rules, OK.

  • The current comical flap about ‘Kinesiology tape’ and the sport gullibles who use it deserves a mention here.

  • I wasn’t able to look at the list properly on this blog due to formatting issues, but tracked down what I assume is the same thing on the FSM website:

    Why was migraine omitted, for example? And surely it’s clear to anyone that the level of detail given in this list is inadequate to tell the whole story? Much more positive one-liners for most of those reviews could be cherrypicked, in the same way that FSM has been done – to create a false impression. The evidence is much more nuanced than this.

    • “Why was migraine omitted, for example?”
      do you mean the Cochrane review that included the sentence in the conclusions “here is no evidence for an effect of ‘true’ acupuncture over sham interventions”?

      • No, I mean the updated one that includes the sentences ‘The available evidence suggests that adding acupuncture to symptomatic treatment of attacks reduces the frequency of headaches. Contrary to the previous findings, the updated evidence also suggests that there is an effect over sham, but this effect is small.‘

        Isn’t it good practice to present the latest version of the evidence? This is one of several examples where the evidence for acupuncture has increased over time.

          I still cannot call this convincing evidence, sorry.
          sorry also for clicking on the 1st review that came up on Medline without checking for an update [it’s still early in the morning]

          • You are forgiven! Human after all 😉

            When taken in the context of the much-debated and thorny issue of placebo acupuncture, I (and many others, not all of which are acupuncturists) would argue that these results are far more interesting than they may seem. The effect beyond (probably active) placebo may be small, but the overall effect is considerable, and I see it making huge differences to people’s lives on a regular basis. Wouldn’t you conclude, as the review authors have, that a course of acupuncture is a valid option for migraine sufferers?

          • they would say that, wouldn’t they?

        • I don’t agree that this review is an update.

          5 trials in the previous review were replaced by 5 new ones in the new review. Had the 5 new trials been added to those included in the previous review it would indeed be an update.

          What we have are two different reviews showing contradictory results.

          What would result if all the evidence was considered? I would be interested to know.

    • Thanks for the link, Tom, I had difficulty viewing the document in the original post.

      It’s important to look at the bottom line conclusions, particularly with a scientifically implausible treatment such as acupuncture. It is always possible to put a more positive gloss on the findings of most studies, acupuncture studies seem particularly prone to doing this, but we have to see past the qualifiers – more studies are needed, or the maybes and might-bes – and look at the unvarnished results. That’s what science is about. If acupuncture and placebo acupuncture both improve headaches that isn’t a victory for acupuncture.

  • “The evidence is much more nuanced than this.”

    Indeed it is – that’s why it is not convincing.
    Res ipsa loquitur, quod erat demonstrandum, caveat emptor.

    • Sorry, I didn’t have the benefit of a classical education so the Latin is lost on me. My point is that picking out a negative line from a bunch of reviews, omitting more positive ones, and not discussing the context is inadequate if a proper understanding of the current evidence is the goal.

      • Res ipsa loquitor: the matter speaks for itself – the principle that the mere occurrence of some types of accident is sufficient to imply negligence.
        Quod erat demonstrandum: that which was to be proven – Some may also use a less direct translation instead: “thus it has been demonstrated.”
        Caveat emptor: let the buyer beware – the principle that the buyer alone is responsible for checking the quality and suitability of goods before a purchase is made.

        • Credat Emptor; “buyer have faith”.
          Clearly there is a strong, if not imperative aspect to the general populous regarding health-care providers, especially those the society opts to allow the moniker Dr.
          The tug-of-war between “buyer beware” & “buyer have faith” in regards health-care is sempiternal.
          The fraud thrives on the latter….and uses it to its greatest fraudulent extent….seems to me.

    • Actually, I don’t think the evidence is nuanced. The evidence matches expectations. Why would acupuncture work? Good evidence shows acupuncture doesn’t work. No surprise.

  • @Edzard ‘They would say that, wouldn’t they?’

    So Cochrane reviews are ‘the most rigorous, independent and objective assessments of therapeutic interventions we currently have’, unless you don’t like the conclusion?! You seemed pretty happy with the NICE decision to remove acupuncture as a recommendation for low back pain, but I didn’t see any ‘they would say that, wouldn’t they?’ analysis of that. The chairman of the committee declared 6 conflicts if interest, and another voting member declared 10. Not surprisingly, pharmaceutical options with known side-effects and a lack of evidence of efficacy beyond sham made the cut ahead of acupuncture, which has been shown to considerably reduce the intake of pain killers. As I said, plenty of nuance.

    • so, acupuncture does not work against paranoia?

    • “the most rigorous, independent and objective assessments of therapeutic interventions we currently have” does not mean they are always faultless. I have criticised many even on this blog.

    • …acupuncture, which has been shown to considerably reduce the intake of pain killers.

      This is a common trope these days in the wake of the opioid epidemic, which is of course not caused by the lack of acupuncture.
      I presume you are referring to chronic conditions. Any activity or pastime can be shown to have a positive effect on the need for pain killers in most situations.
      Take chronic back pain for example. Physical exercise, hiking, dancing, skiing are all examples of practically side effect free measures that have been shown to reduce the need for pain medication in such conditions, and they have positive effects on many other aspects of health, not the least psychological well-being. Acupuncture and its related ritual theatricals is a type of pastime of course and it is no surprise that it may have a transient effect as such. But I am not aware that it has been shown to considerably reduce the need for pain killers and it is certainly not superior to such simple, less costly and demonstrably effective measures. Even more specialised interventions such as cognitive behavioural therapy are demonstrably effective in reducing need for pain killers, without the risks involved in inflicting unnecessary injury. As we have repeatedly made clear here in the discussions on this blog, acupuncture is far from free from a risk of serious complications.
      Now please tell us Tom, have you tried taking your patients for a hike and to the movies instead of perforating them and chanting Chinese wisdom or whatever the ritual involves? How would that (with several heaped tablespoons of verbal reassurance and medicine-bashing of course) measure up to acupuncture in less need for pain killers?

      And also Tom… tell us what you think of the information in this article and its abundant linked material:
      (please spare us your opinion of Dr. Gorski, we’we already heard all those rants from others)

      • I’m not in the habit of taking my patients for a hike, but I certainly recommend all sorts of similar things. However, I do not ‘bash’ conventional medicine, as you imply. There are many situations when I refer people to their doctors.

        I haven’t had time to look into this in detail, but here’s one recent review that suggests reduced and delayed opioid consumption with acupuncture, but not with passive motion/preoperative exercise:

        • Hmm…. Tom.
          I read the abstract differently. It only says acupuncture “delayed” not “reduced and delayed”. The conclusion is misleading in this respect. And what is more there is no mention of wether they only compared acceptably controlled studies. From the text in the abstract they found that any intervention either delayed or with low certainty or very low certainty reduced opioid intake.
          All the modalities entered in this review seem to have had similar effect so one wonders if a Vaudeville show might have produced the same desired effect? Remember that we will want to separate the efficacy from the effect so we know whether the intervention has an independent activity from that of simply entertaining the patient. (as in theatrical placebo). The only way to separate efficacy from effect is by proper control. This is possible in acupuncture and has been done, with disappointing results for those who have vested interests in its utility.

          This reference is as far as the abstract goes, not at all convincing. I would have to scrutinise the whole opus, which is behind a paywall, but I am not hopeful that the review is indeed corroborating your ideas about acupuncture being useful for reducing opioid use.

  • An update of the landmark 2012 study by the international Acupuncture Trialists’ Collaboration reinforces the evidence that acupuncture is effective for the treatment of chronic musculoskeletal pain and that its benefits cannot be explained solely in terms of placebo effects. The new individual patient data meta-analysis analysed data from an additional 13 trials, giving a total data set of 20,827 patients from 39 trials. Acupuncture was superior to both
    sham and no-acupuncture control for all four chronic pain conditions examined (back and neck pain, osteoarthritis, chronic headache and shoulder pain). (Acupuncture for Chronic Pain: Update of an Individual Patient Data Meta-Analysis. J Pain. 2017 Nov 30. wii: S1526-5900(17)30780-0)

    National Institute for Health Research: An NIHR systematic review compared acupuncture, sham acupuncture and usual care and concluded that acupuncture was better than sham acupuncture and usual care for relieving pain from musculoskeletal conditions, knee osteoarthritis and chronic headache.

    • Although I have scrutinised many, I do not pretend that systematic reviews are within my field of education or experience.
      But I had a look this morning at some relevant chapters in the study referenced by Peter Deadman. My first impression is that this is a peculiar conglomeration of projects, lacking focus on clinically comparable problems and the authors, despite repeated declarations of rigour, went out of their way to welcome non-rigorous trials into the analysis.

      I look forward to seeing expert reviews of this curious review.

  • I just found this in a post in Linkedin and was shocked. Could it be that the “opioid crisis” in the US opens the door for quackery to step in? The OMT practising osteopathic physicians D.O. are full of hope to get better conditions by the health insurances now.

  • Dear Prof. Ernst and everyone else who is reading it,
    I am writing to you to challenge your view on acupuncture, which probably is close to impossible. I will say right now, that this is a lot to take into account but it took me quite some time to compile this together and I would be very honoured to receive back an answer from you at some time later.
    I have been reading a lot of criticism on this topic on your website and I would consider myself a critic too, however, looking at recent research that has shown increasingly better study methodology comparing it to the studies before 2013 especially those conducted in China before 2000, I put together some of the studies that I have found showing some really interesting outcomes. I believe you have not reacted to those since most of them were published after 2013. I couldn´t check with all your articles on acupuncture if you have already frowned upon these studies listed below so I am hoping to expose you to research that is new to you. I know, that I am not a medical student and I am not in a position of being able to perfectly judge if a study is impeccable or not. That is why I am writing to you. I tried to limit myself to sham-controlled studies. They are not equally well-conducted, I know that. And I have also read your article “Acupuncture: there is no convincing evidence that it is an effective therapy” showing an overview of the meta-analysis.

    (I believe people might be too busy blaming “Qi” as an actual thing whereas it might have been just a metaphor for something the Chinese could not yet explain. Some of the studies you can find below do prove that there are often major nerves stimulated below these acupoints (contradicting one of your arguments in your essay claiming that nerve location could not be associated with acupoints). This is not me saying that there is such a thing as a real “system” connecting all these points (even though new research being conducted is very striking: + see the document attached that shows a very good correlation to the acupoints).
    From this standpoint, it is obvious that acupuncture fails to be efficient in some trials since it might just not be as perfect as Chinese physicians declared it or thought it to be when developing it. For example that maybe some acupoints are just made up or not perfectly placed. No medical intervention is flawless. But enough of these unproven hypotheses. Let´s cut to the interesting bit.)

    The question I would like to raise is the following: When it comes to pain-related conditions acupuncture struggles to show consistent superiority over a placebo effect. However, as we have seen in the German acupuncture trials for example, placebo, as well as sham-acupuncture, were significantly more efficient in decreasing pain compared to normal pain-killer drugs. This is a result of a lot of studies I have stumbled upon when acupuncture was compared to normal drugs. So that both verum and sham was effective in bettering a patients symptoms significantly. Having in mind that both the sham-acupuncture effects as well as the verum-acupuncture effects were successfully measured in high-quality studies looking at different brain function, receptor-building, etc. of mice as well as humans (like this very astonishing one from Oxford or from the list I gave you on studies in mice), and knowing that there is “good(?)” evidence that real acupuncture is capable of having a different biological function than sham-acupuncture (as shown in the studies below), can we say that even though regarding pain-related issues there is no significant difference in pain reduction in patients (it´s still often better than/as good as drugs whatsoever) real acupuncture is still effective and would work on proven biological levels besides the placebo effect? So that if you take away the placebo-effect it would still work on the basis of the studies I mentioned, however, in some studies (especially in pain-related ones where the pain assessment is often lacking objective parameters like receptor-building that is used in the ones below showing that real-acupuncture works differently than sham) the placebo effect of acupuncture is felt just as strong by the patients as the real acupuncture in the other group (the placebo effect is very strong in acupuncture as brain scans have proven; probably stronger than in normal drugs) and therefore couldn´t be distinguished by the patients but was very strong in both groups and better/equal to normal drugs/significantly improved the symptoms. I hope I could make my point clear. If you have any questions feel free to ask me and I would be very delighted to hear from you soon!
    For me looking at all the newer research especially those studies showing significant, measurable differences in animal-models highlight, that there must be something superior to the placebo-effect and can be explained on a biological basis apart from the placebo-effect.
    I would love to receive an answer from you as experts and probably the strongest critics on acupuncture on the internet. This may provide new material for your articles, but I would also be very happy to receive a direct answer from you evaluating some of the studies below.
    I divided my findings into two sections. Starting with the biological/neurological function of acupuncture that researchers were able to determine/measure, I found this study the most compelling one:
    This was also promoted by Harvard ( It is interesting because patients were anesthetized during the treatment. Plus, this study took place outside China and in the US for a change. I was keen on getting to know your reaction to the outcome of it (I know it´s electro-acupuncture, however sham-controlled). And yes, it is electro-acupuncture and this, of course, is more likely to stimulate nerve function (classical acupuncture does too as presented in the studies below), but it was significantly stronger on acupuncture points compared to sham. The impact was also captured by MRI-control.
    These following studies present an overview of the potential/probable measurable effects of acupuncture that has been performed on both non-acupoints and the “real” acupoints in mice. They are all somehow connected to pain-relief:
    This overview ( stood out to me because the measured impact in mice was also found in humans afterward. Challenging the argument that acupoints in mice cannot be differentiated from non-acupoints:
    These are further studies attempting to prove the “pain-killing” function of acupuncture treatment:
    • How can it be that sham and verum acupuncture are different in their opioid-response?

    These studies, also mostly performed on animals, evaluated other functions of acupuncture than pain-relief that were measured during the treatment:,%20193.pdf

    Finally, I also put together a list of more “recent”, mostly double-blind RTCs showing that verum-acupuncture was superior to sham. I know, that meta-analysis is even better, nonetheless, these are studies you haven´t reacted to yet I believe and maybe this is something you are interested by: (not double-blind) (not double-blind) (not double-blind) (not double-blind)

    This article by the Guardian additionally summarizes some of my arguments very well I know this is all an overwhelming amount of research (I am sure there is even more novel material that I didn’t include) and it takes a lot of time to critically read through this. If you can manage to look at the works that I commented on myself like the study published in the Brain I would be very delighted. I am very interested in your view on the biological effects of acupuncture I talked about in the first half of this email. You by far can evaluate better on studies and it would be very helpful to get to know your professional opinion on these measurable impacts of acupuncture on real acupoints compared to sham and the changes seen in mice. How do you explain the measured changed? In case you have already tackled some of these studies in an article please let me know.
    I am looking forward to hearing from you soon. Thank you very much in advance for reading this message and I would like to excuse mistakes in the English language on my behalf. I am very fond of your work and would be honoured to contribute to it in this way.
    Yours sincerely,
    Caro Bröcker

    • thank you for this – unfortunately I have not enough time to go into the details.
      I grant you that the evidence on acupuncture is exceedingly complex, and there are no easy answers.
      I have tried to summarise my own take on them here [I know this is not a new paper, but my views are essentially unchanged]:
      as to modes of action, I feel we need proof of efficacy first – without it the mechanisms are irrelevant.
      sorry to have to make this so brief.

    • Caro!! Thank you for taking the time to do this 7 years ago.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Subscribe via email

Enter your email address to receive notifications of new blog posts by email.

Recent Comments

Note that comments can be edited for up to five minutes after they are first submitted but you must tick the box: “Save my name, email, and website in this browser for the next time I comment.”

The most recent comments from all posts can be seen here.