The Brain in CRPS-More Barriers or New Opportunities

From a biological perspective, I find CRPS very perplexing. So do many other people—the community of scientists that are pursuing better treatments, through a better understanding of the basic mechanisms, or by developing better animal models of CRPS, or by undertaking clinical investigations and clinical trials, stretches across many countries and many scientific and clinical disciplines. With all of those supposedly clever people working on it, one might expect that by now there would be better ways to prevent and treat this debilitating disorder. For many people with CRPS, and for many clinicians trying their darndest to help them, the situation can feel a bit hopeless. That is one place where organizations like Reflex Sympathetic Dystrophy Syndrome Association (RSDSA) play a crucial role—reminding patients and clinicians that significant gains are being made, and that there is hope. In fact, all of those supposedly-clever people are, in fact, developing a better understanding of the mechanisms and possible treatments of CRPS.

When Jim Broatch asked if I would write an article for this newsletter, he suggested that I provide an update of my own research and what it might mean for people with CRPS and the clinicians who treat them. From where I sit, my own research looks rather simple, compared at least to the very sophisticated work going on elsewhere.

I get a little bamboozled with the molecular biology—just remembering what the acronyms stand for seems an arduous task. The thing that interests me is what it is like to have CRPS, what goes wrong in the way the limb feels, how the brain might contribute to this, how these changes might contribute to the problem, and what we can possibly do about it. In this article, I will concentrate on the things that, at first glance, seem a little daunting because they involve changes in the way the brain works, but, at second glance, are actually opportunities because the brain, more than any other organ we have, is very responsive to training. These are the things I have been a part of that I think are important for people with CRPS and their team.

People with chronic CRPS tend to think that their affected limb is bigger than it really is [12]

I first wondered about this when I noticed that patients would describe their limb as feeling swollen, and describing the size of the limb in a way that made me think, “Hang on! It’s not that big!” Now, if someone is saying that their arm feels big when it is not, they are either lying or it does actually feel bigger than it really is. I don’t think many people, patients or otherwise, are liars, so we set up an experiment. In short, we took a photograph of both limbs and then manipulated the photograph so that the affected limb looked bigger or smaller than it really was. We then asked patients to select the photograph that they thought was the real photograph. People without CRPS tend to pick the correct one. People with chronic pain that is not associated with CRPS tend to make more mistakes but still get it about right. People with chronic CRPS tend to pick the photograph that shows the affected limb to be slightly bigger than it really is. We have done subsequent experiments to confirm this phenomenon.

Movement causes a bigger increase in pain and swelling when the visual image of the painful body part is magnified [3]

On the basis of the above, we decided to see if this distortion of perceived size of the painful limb might contribute to pain and swelling. We did an experiment with 10 people with CRPS of one hand. They did a set of movements while they looked at their hand under three conditions: normal, through a set of glasses, through a set of magnifying glasses. Remarkably, when they watched through the magnifying glasses, the increase in pain and swelling was significantly worse. So, when a painful limb looks bigger, it actually gets more painful and more swollen. The groovy and unexpected finding was that when the magnifying glasses were inverted, the increase in pain and swelling was reduced! We don’t know why that happened, but we have started to think about how to use this clinically.

The brain’s map of the affected body part shrinks [4-5]

This is a little tricky to understand, but give it a go because if you can understand this bit, you will understand how we go about treating it. The brain holds maps of the body so that it can tell you where exactly you are, where you end and the space around your body begins. These maps are really important for the feelings that we have of our own body—that it is ours and that we own it. The group of brain cells that holds the map of a specific body part—let’s say your arm—is reasonably consistent between people and it is kept accurate and precise by a very complex system of brain cells influencing brain cells. For some reason, in CRPS something goes wrong with this system so that fewer brain cells are responsible for the map and the map becomes ‘smudged’—the outlines are not as clearly delineated, which really means some brain cells become part of two or three maps instead of one). It is not only CRPS in which this happens. It also happens in people with phantom limb pain after amputation. The opposite thing happens in people who use a body part a lot. For example, more brain cells are involved in the map of the left hand of a violinist, or—get this—Braille readers have more brain cells involved in the map of their index finger during the week than they do on weekends! Those discoveries are very important because they show that you can train the brain to change the maps.

Training the map of the body decreases pain and swelling [6,7,8]

The first three discoveries are all related and all contribute to the idea that we might be able to reduce symptoms by training the brain. A neuroscience superstar in Germany, Herta Flor, PhD, did exactly that in amputees with phantom limb pain. She showed that with just two weeks of training, amputees could discriminate between different types and locations of electrical stimuli delivered to the stump, which significantly reduced pain AND returned the map in the brain back to normal. This was a profound finding. We did a dumbed-down version of this in people with CRPS. We used a wine cork and a pen lid to touch the affected limb in one of five places. The patients had to discriminate between the two stimuli and between the five places. Again, in two weeks, there was a significant reduction in pain and swelling, but only if patients had to discriminate between the stimuli—not if they were just touched and didn’t concentrate on it. We did a subsequent study that showed that this effect increases if you use a mirror, so that you can look at the mirror image of your opposite limb while doing the training.

The brain gets better at producing pain and swelling as CRPS progresses [910]

CRPS affects many brain functions. A key change occurs with the sensitivity of the mechanisms that cause the symptoms. We have shown that people with CRPS get worse when they just think about moving the body part, even if they don’t even move it a whisker. We have also shown that if we trick patients into thinking that we have touched their painful arm by getting them to watch us touching their other arm, in a mirror placed between their arms, it hurts just like it would if we touch their painful arm.

People with CRPS find it difficult to recognize whether a pictured hand is a left hand or a right arm, if the pictured hand is the same side as their own affected hand [1112]

The same is true for feet. We have done a range of studies exploring this intriguing finding and have found that the effect is not due to pain and is not confined to people with CRPS. However, in testing it, we noticed that people with CRPS could improve on this task with practice. We decided then to see if we could use this task as a stepping stone to other treatments.

Graded motor imagery reduces pain and swelling in chronic CRPS [10,1314]

Graded motor imagery involves three stages:

  1. recognizing pictured hands (or feet, depending on the affected limb) as being left or right,
  2. imagined movements, and
  3. mirror movements.

Each stage lasts about two weeks and patients practice hourly. Noigroup Australasia produces a Motor imagery pack, which includes access to resources, an online recognition program (Recognise), and a mirror box. We have undertaken subsequent investigations that show that frequent practice is essential—we saw nearly no effect when patients practiced only once or twice a day. Graded motor imagery is the only treatment for chronic CRPS for which there is both strong evidence of efficacy and no known risks and minimal side effects.


An important aspect of all of this research is that clinical interventions are always undertaken within the context of resourcing patients with as much accurate information about pain, about chronic pain and about CRPS, as we can give them. I don’t mean that we just bombard them with information—I mean that we take seriously the role of educating patients about these things so that they understand it, so that they ‘get it’ in the marrow of their bones. There is a whole body of literature that clearly shows that empowering patients in pain, with an understanding of how pain really works—the biological processes that underpin it and what changes when pain persists—improves outcomes [15]. This is not a trivial task, but I think it is a terrifically important one.

It is critical to make it really clear in any article like this, that although the results of clinical trials are promising, the treatments don’t work for everyone. It is also very important to emphasize that in the treatments I have described, the role of the physical therapist, doctor or other medical or health practitioner is really important. However, this person is more like a coach or supervisor, since all the real work is done by the patient. As I say to patients—patience, persistence and courage. If you can muster each of these in good measure, I believe you will more than likely make it through.

Finally, my research is only a small proportion of all the work that is being done to better understand, manage, and treat CRPS. To provide an overview of all that is being done is beyond the scope of Jim’s request. It is also beyond me and would constitute a very large book with very small writing. I have cited the key papers relating to what I have written about in the reference list. In addition, I have provided some relevant links or sources for further information.


  1. Lotze M, Moseley GL. Role of distorted body image in pain. Curr Rheumatol Rep. 2007;9(6):488-496.
  2. Moseley GL. Distorted body image in complex regional pain syndrome. Neurology. 2005;65(5):773.
  3. Moseley GL, Parsons TJ, Spence C. Visual distortion of a limb modulates the pain and swelling evoked by movement. Curr Biol. 2008;18(22):R1047-R1048.
  4. Flor H, Elbert T, Knecht S, et al. Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation. Nature. 1995;375(6531):482-484.
  5. Maihofner C, Handwerker HO, Neundorfer B, Birklein F. Patterns of cortical reorganization in complex regional pain syndrome. Neurology. 2003;61(12):1707-1715.
  6. Flor H, Denke C, Schaefer M, Grusser S. Effect of sensory discrimination training on cortical reorganisation and phantom limb pain. Lancet. 2001;357(9270):1763-1764.
  7. Moseley GL, Wiech K. The effect of tactile discrimination training is enhanced when patients watch the reflected image of their unaffected limb during training. Pain. 2009; 144(3):314-9.
  8. Moseley GL, Zalucki NM, Wiech K. Tactile discrimination, but not tactile stimulation alone, reduces chronic limb pain. Pain. 2008;137(3):600-608.
  9. Acerra NE, Moseley GL. Dysynchiria: Watching the mirror image of the unaffected limb elicits pain on the affected side. Neurology. 2005;65(5):751-753.
  10. Moseley GL, Zalucki N, Birklein F, Marinus J, Hilten JJv, Luomajoki H. Thinking about movement hurts: The effect of motor imagery on pain and swelling in people with chronic arm pain. Arthritis Care Res. 2008;59(5):623-631.
  11. Moseley GL. Why do people with complex regional pain syndrome take longer to recognize their affected hand? Neurology. 2004;62(12):2182-2186.
  12. Schwoebel J, Friedman R, Duda N, Coslett HB. Pain and the body schema: evidence for peripheral effects on mental representations of movement. Brain. 2001;124(Pt 10):2098-2104.
  13. Moseley GL. Graded motor imagery is effective for long-standing complex regional pain syndrome: a randomised controlled trial. Pain. 2004;108(1-2):192-198.
  14. Moseley GL. Graded motor imagery for pathologic pain – A randomized controlled trial. Neurology. 2006;67(12):2129-2134.
  15. Moseley GL. A pain neuromatrix approach to patients with chronic pain. Man Ther. 2003;8(3):130-140.

Additional Resources

Motor Imagery Program Products available from Noigroup


Butler DS, Moseley GL. Explain Pain. Noigroup publications; Adelaide: 2003 available from Noigroup

Moseley GL. Painful yarns. Metaphors and stories to help understand the biology of pain. Dancing Giraffe Press: 2007. Available from UK, US, everywhere else

Click on our Books tab to see a synopsis of what the books are about and a video on Explain Pain

RSDA Reproduced with permission from the Reflex Sympathetic Dystrophy Syndrome Assocation

Lorimer Moseley (2009). The Brain in CRPS-More Barriers or New Opportunities


  1. Hi there
    I just had to respond to the ‘Information is to behaviour change…’ – it was Nick Kendall who introduced me to it and repeated it over & again in the 10 years we worked together! Something must have made the Kiwi’s adopt it methinks!
    Anyway, I took a very quick look at the cigarette cessation literature, and one of the main reasons Motivational interviewing was introduced in addiction services was because of the spectacular failure of ‘information’ and ‘education’ – even kindly, detailed and personalised education/information – to change quit rates. (A recent paper to illustrate is: Etter, J.-F. (2007). Informing smokers on additives in cigarettes: A randomized trial. Patient Education and Counseling, 66(2), 188-191.)
    The thing is, you and I probably have loads of things we ‘never get around to doing’ – and it’s not because of lack of understanding, it’s simply that other things are more important. Something that we do know increases reports of pain intensity and distress is lack of a sense of control and an expectation that to manage chronic pain requires the same set of behaviours as is applied to acute pain. So… I’m not at all surprised that people report reduced distress, increased self efficacy and even reduced disability after they’ve received ‘information’ or ‘education’. It’s certainly a component of pain management. BUT the words ‘education’ or ‘information’ can trivialise the process that is used by a sensitive and well-informed clinician. While you and I might go about helping someone understand their pain using things like guided exploration, mini-experiments and Socratic questioning (probably because the psychologists got to us early on!), to many clinicians, ‘education’ and ‘information’ are simply standardised bits of information-dumping. And this is what distresses me when those words are used – because unless the person reconceptualises their own role in managing their pain, this ‘information’ floats over the top and fails to make changes to behaviour. And I think you and I are coming from the same place when we say we want to see people use this approach as a platform from which to launch the comprehensive pain management programme.

    BTW so far I see that rough 33% ‘improvement’ rate being applied to many different treatments – including standard Interdisciplinary Pain Management using a CBT approach. It does depend a lot on what you count as ‘improvement’ – reduced pain and improved gait pattern in clinic, may not translate into improved engagement in return to work or better relationships with others. At least, that’s been our outcome except for people with CRPS of less than 6 months duration. And we do see probably the most complex patients who present with not only CRPS but numerous other psychosocial and functional problems that complicate their disability.

    I guess I’d love to see some guideline as to when to prioritise pain reduction as opposed to increased function – back to the ‘what works for whom and when’. And most importantly, how to integrate the motor imagery (etc) within other rehabilitation approaches such as vocational management and relationships etc.

    I don’t think there is much argument that CRPS is an incredibly complex disorder, managing it isn’t going to be a case of simple question: simple answer, and that’s what makes both the tragedy for people with the problem, and the challenge and intrigue for researchers and clinicians alike.

    I’m with you – let’s keep this discussion going, it’s great!

  2. These are stellar contributions – great GREAT work! Some genuine insight and wisdom. I have a few responses. First, re education. My PhD supervisor, Prof Michael Nicholas at Uni Sydney, absolutely loves that saying – Education to behaviour change….. I am less convinced that education should be bundled up as one thing here. I reckon that it is the type of education that matters. I have done several RCT’s that show that teaching people about the biology of their pain directly changes behaviour – as measured by pain ratings on a straight leg raise and by functional measures. So, maybe Fordyce was right about that as far as telling people that smoking kills, or that one with pain should move despite pain, but i am not sure if the same lack of effect would occur if we sat down with a smoker and EXPLAINED it all to them in a memorable, interesting and intelligible way. clearly of course, we can’t JUST teach people things – i see it as the platform on which to launch the full training program.
    Re what works for whom. This is critical and i share Bronnie’s reluctance to get carried away on a ‘new’ treatment. Those who have heard me speak on this stuff will know that i am very keen to emphasise that these data are early. I firmly believe we should test a treatment in an RCT before we conclude it works. We now have 3 RCTs that suggest graded motor imagery works for long term CRPS. What that really means is that for one person who would not respond to ongoing treatment to get 50% better with graded motor imagery, 3 patients need to do it. So, for Bronnie’s patients who are end of the line, it sort of works out to a 33% success rate. This, believe it or not, is excellent for chronic CRPS.
    One thing I have learnt about these approaches is that we should consider it like we would stroke rehab. I say to patients that they are embarking on a long rehab program that will take weeks, and then months, to get real gains. For them to participate, they need, i believe, to understand. This is where the education bit comes in. I think we need to convince them that, in chronic CRPS, the problem is in the brain and we need to train the brain. etc etc. Relevant to this is my observation that most patients do not do enough training to make changes happen. We have audited patients and those who don’t succeed usually don’t train more than 20-30 mins a day. If they were stroke patients, we would not expect this to work.
    Regardless of these observations, i am with Bronnie in saying that there is still so much we don’t know about CRPS and we are still a long way from knowing that we can fix it. Finally, i want to iterate how much i LOVE these contributions. Thanks.

  3. This article is absolutely great! I’ve often noticed “weird stuff” with my body and CRPS, I am especially thinking of how we noticed that when I try to tell you where my leg is in space, I think it’s farther forward than it really is. I’d also noticed that it feels bigger to me, but looked smaller to the sighted people around me. It would be interesting to figure out how this applies when one can’t see.

    It’s so exciting to see these studies and this information. I always learn a ton from your articles!

  4. It’s great to know that some people, like you, are helped with mirror therapy – and it’s always great to hear of people recovering. In my job and at the centre in which I work, sadly we don’t see that many probably because we are a tertiary treatment centre often thought of as ‘the last resort’. As with any treatment, it’s important to work out what works for whom and when. The problem with any treatment is that it can become that ‘one size fits all’ and for the people for whom it doesn’t work, it can feel like they’ve failed rather than a treatment simply not working in their situation.
    We have found that while some people regain good function in therapy with this type of treatment, resumption of function including working, thinking of oneself as a person first rather than the pain, and returning to a ‘normal’ lifestyle doesn’t occur. Whether this is because their particular combination of problems is not simply about retraining the brain to resolve the CRPS (and actually requires other components to be added), or because as treatment providers we haven’t been astute enough to tailor the treatment to fit them, I don’t know. As ever, it really is still an hypothesis testing situation – we hypothesise that this problem is arising because of X,Y, Z and go about systematically testing each hypothesis. Until we have a much clearer idea of how and why and which mechanisms are involved, I think we need to be cautious about endorsing any single treatment modality. Call me conservative, but in pain management so often I’ve seen a ‘new wave’ of treatment come in with a hiss and a roar (and even some good science!) only to find out later that we need to be careful to work out what works for whom and when. I see a theme happening?!

  5. Hope you don’t mind if I add abit here. Bronnie I had crps for over 10 years and was diagnosed with a whole body problem including dysautonomia involving cardiac symptoms etc. Over a long period I used mirror therapy to relieve one sided flare ups. This in time resulted in a gradual reduction in overall pain and symptoms. After many months I had a period of a few days with no major flare. This increased in time the more I used mirror therapy. Mirror therapy wasn’t all I did but it was the single thing that made a big difference and gave me some control over the management of symptoms. I’m in remission now. All crps symptoms and dysautonomia have gone. Dr Moseley’s description of distorted perception of size of affected body part is accurate in my case. I now perceive my body to be normal. Mirror therapy may not work for everyone. However it has worked for me with chronic, long term, previously considered intractable crps. The few times it hasn’t worked I realized in the end t was because just I wasn’t doing it the right.

  6. Great writing as ever! I guess I wonder about a couple of things.
    The first is your point about ‘educating’ patients. I think this is a misnomer – we’re not educating as much as helping people reconceptualise or reinterpret their situation. As Bill Fordyce is reputed to have said: ‘information is to behaviour change as spaghetti is to a brick’. Informing or educating people is insufficient – or we wouldn’t have people who were overweight, didn’t exercise or smoked! You hint at this when you describe it as ‘so that they ‘get it’ in the marrow of their bones’.
    I wonder too whether describing the brain changes that occur when people are treated with motor imagery actually changes our clinical practice: after all for years and years we have carried out graded exposure for phobia, with many people using imagery (with hypnosis), virtual reality and gradually increasing tolerance or coping with their experience. Are we really only finally uncovering the underlying neural mechanisms as to how these processes ‘work’?
    As a clinician working with people who have long-standing CRPS, who find themselves utterly disappointed when mirrorbox and brain training don’t work, I’m also worried about this approach being touted as a ‘one size fits all’. I have no doubt at all that it’s effective in sub-acute cases, and especially in cases without other psychosocial factors, but I do worry that it’s less effective in the more complex and long-standing cases – it will be great one day to know ‘what works for whom and when’.

    Regarding the possible evolutionary rationale for CRPS? I don’t think there is one. I think it’s just possible that, like in every other body system we have, there are times the nervous system fails to function. Perhaps the main learning is that we’re made to move, and with good acute analgesia we can keep moving, and maybe we can reduce the incidence of this nasty disorder.