Neil OConnell on A Cup of weak Qi

Brain imaging studies are often both fascinating and frustrating in equal measure. They shine a light on activity patterns within the brain that occur during various aspects of behaviour, movement and perception. The problem comes when we try to negotiate the minefield of drawing conclusions.

Acupuncture is a widely used and advocated treatment for pain (as well as almost every other clinical condition you may have heard of) and currently enjoys great popularity. Many acupuncturists believe that when needles are inserted it is important that the patient feels a sensation of ‘deqi’ (pronounced ‘day chii’) to achieve therapeutic benefit.  Deqi is described as distinct from acute pain and characterised as ‘aching’ ‘dull’ ‘heavy’ ‘numb’ ‘radiating’ ‘spreading’ and ‘tingling’.

Newly published research from the UK has used functional MRI scanning to compare the brain activity associated with feeling deqi-like sensations to the activity associated with feeling acute pain elicited by acupuncture to a point on the hand (Ashgar et al. 2010). The results show that when people feel deqi-like sensations regions of the brain tend to show de-activation whereas when people feel acute pain there is mix of activation and deactivation of some brain regions. Specifically in areas of the limbic system (a big area for emotional processing) and the cerebellum (the big cheese in integrating sensory information) deqi was associated with deactivation but acute pain was associated with increases in activation.

ResearchBlogging.orgSo we can say with some confidence that when we prick people with a thin needle, the different sensations that people feel are associated with different brain activity. Now comes the tricky part – what does it mean? The authors of the study point out that the limbic system has been shown to be a player in the brain activity underpinning the experience of pain and suggest that these differences might help explain the pain-relieving effects of acupuncture needling when deqi is achieved. They rightly recognise that it isn’t possible to say whether this is caused by the physical effects of acupuncture needling or psychological factors. After all, conscious sensations are the end result of both sensory input and central evaluation and asking the participants to rate different sensations involves a degree of suggestion.

I find myself wondering whether acupuncture itself might be a bit of a red herring when we try to interpret these results so I’m going to do a bit of speculating (bear with me). The authors suggest that deqi and pain sensations have different ‘impacts’ on brain activity. The pedant in me wonders whether it would be more accurate to replace the word ‘impact’ with something like ‘association’. The activity in the brain may be the actual neural substrate of the sensations reported, or in simpler language the activity is the sensation. From that perspective what we know is not that acupuncture itself leads to specific brain activity but that feeling pain or the variety of sensations that comprise deqi are underpinned by specific brain activity. As a study of how the brain processes sensory perception these results add to a fascinating body of accumulating evidence. Drawing conclusions about the therapeutic mechanisms of acupuncture is a little more problematic.

A recent systematic review suggests that acupuncture has little or no effect on pain in comparison to sham or placebo acupuncture for any number of different painful conditions (Madsen et al. 2009).  They also showed with some fancy-but-fair statistics that it doesn’t matter what type of placebo acupuncture you use, the real thing still isn’t more effective. It doesn’t seem to matter where or how deep you put the needles, whether or not you ‘twiddle’ them, or whether or not the patient feels deqi.  In fact it doesn’t even seem to matter whether you insert the needle at all!

So maybe sensation is not as important to the clinical effects of acupuncture as the idea of having the treatment and its related expectations. This does not detract from some fascinating data on the different ways in which the brain processes different sensations, it just doesn’t tell us a great deal about how or if acupuncture works.


Asghar, A., Green, G., Lythgoe, M., Lewith, G., & MacPherson, H. (2010). Acupuncture needling sensation: The neural correlates of deqi using fMRI Brain Research, 1315, 111-118 DOI: 10.1016/j.brainres.2009.12.019
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

About Neil

Neil O’Connell is a researcher in the Centre for Research in Rehabilitation, Brunel University, West London, UK. He confesses to having a sceptical disposition to all things alternative (except music).


  1. Thanks Ian,

    That sounds a fascinating dissertation. I agree with you on all points. The placebo effect is all about the wider meaning on the therapeutic interaction. The technology definitely has a role in developing our understanding of this, the trick is not to overinterpret it. As for your later point I would agree and suggest that the implications are that to maximise your therapeutic effect you need to try to understand the patient and the context of their problem as fully as possible. Of course thats exactly what good pain management strives to do.

  2. ian stevens says:

    Neil have you a copy of your shameless self – citation ?!
    I did my disseration on meaning/placebo and find the subject fascinating. You may disagree , but to understand ‘meaning’ and clinical effects its interesting to look at the wider cultural beliefs as well as neuroarchitecture displayed in a brain scanner. David Morris writes well in this area . I think people can get seduced by technology myself, although i have to agree that it is fascinating…..Although this is a different subject I think that this discussion (in the link below) is interesting in relation to the fact that the brain is not a single ‘thing’ but is modulated and shaped by community . We deal with individuals but often forget they are embedded in social networks which probably shape beliefs and ultimately motor activity( if this is a focus of ‘treatment’)

  3. Thanks for your nice comments Bronnie. Moermans writings on placebo are fantastic – all clinicians, researchers and patients should read them. It appears to me sometimes that most acupuncture research is interpreted from the perspective that “we know that it works” and then uncritically reported in the media as such (shameless plug to a related paper I have written with the great Ben’s Wand and Goldacre below). The problem is that the quality evidence would suggest more strongly that “we know that it doesn’t work better than placebo”. This shouldn’t be so surprising given the pre-scientific ideas underpinning it and their inherent lack of biological plausbility.

    There is much talk now of “Western Medical Acupuncture” which dispenses with traditional chinese medicine theory and suggests that acupuncture works (mostly for pain) as a kind of novel form of sensory stimulus that alters pain processing. The problem with this is that, as I mentioned in my post, no variation in the sensory stimulus seems to elicit more or less therapeutic effects. Expectations do seem to play a part though (see this lovely paper by Barker Bausell: Bausell RB, Lao L, Bergman S, Lee WL, Berman BM Is acupuncture analgesia an expectancy effect? Preliminary evidence based on participants’ perceived assignments in two placebo-controlled trials. Eval Health Prof. 2005 Mar;28(1):9-26.).

    Some acupuncture researchers would suggest that even non-penetrating sham needles are acupuncture too, but I worry that they are changing their own goalposts and making acupuncture an unfalsifiable hypothesis. I guess the bottom line for brain imaging studies that look for things other than the placebo effect to explain how acupuncture works is that it is very difficult if a specific therapeutic effect beyond placebo can’t be found.

    Of course thats just my view…

    Shameless self-citations:

    O’Connell NE, Wand BM, Goldacre B Interpretive bias in acupuncture research?: A case study. Eval Health Prof. 2009 Dec;32(4):393-409.

  4. Thanks so much for this summary of the association between brain imaging findings and experience. The correlation or causation situation in brain imaging makes it so difficult for uncritical thinkers to unpack, and so often we read headlines like the ones debunked a couple of days ago here! I do like Dan Moerman’s description of the ‘meaning effect’ of health provider interactions, and it doesn’t seem so far-fetched for it to apply to acupuncture (even with effective sham acupuncture methodology).