The healthy hand in the CRPS brain –  digging deeper

‘Cortical reorganisation’ is a commonly used term in pain. In CRPS there has long been evidence of cortical changes; specifically that representation of the CRPS-affected hand in the primary somatosensory cortex (S1) is smaller than that of the S1 representation of the other, healthy, hand [1-5]. In 2015 we performed our own functional MRI investigation on this question and our findings surprised us. We did not find a smaller S1 representation of the affected hand, rather we found an enlarged representation of the healthy hand [6]. We wanted to explore this further. And that’s what led us to this current paper, An exploration into cortical reorganisation of the healthy hand in complex regional pain syndrome [7].

The obvious explanation for an apparently enlarged S1 representation of the healthy hand is use. It makes sense that if one hand is affected by CRPS then the other hand must compensate in daily life, and maybe the cortical changes – the enlargement – might reflect this. So we embarked on this current exploration wanting to know whether the S1 representation of the healthy hand is associated with altered hand use in response to CRPS. Specifically we were interested in whether the size of the S1 representation of the healthy hand is associated with the severity of the functional impairment of the CRPS-affected hand. We also thought that CRPS of longer duration might result in more extended use of the healthy hand, so we investigated whether we could find a positive relationship between the S1 size of the healthy hand and CRPS duration.

This study used neuroimaging data on 12 CRPS patients and 10 pain-free controls. Unfortunately (and I think we can be forgiven for this) we didn’t collect a lot of data on the healthy hand in our original MRI investigation. We did not obtain any objective measure of movement or use – and that, in hindsight, would have been very sensible to collect. However, we did collect subjective data on the use of the CRPS-affected hand, on overall daily function, and on self-efficacy – which is the belief in one’s own ability to achieve a desired outcome, in this case despite pain. So we used these measures to infer compensatory use. And we investigated the relationship between this inferred compensatory and the S1 representation of the healthy hand with multiple regression models. We also investigated the correlation between S1 representation of the healthy hand and CRPS duration.

We were surprised by our findings (again). We found that there was no relationship between the size of the healthy hand in S1 and the severity of functional impairment of the CRPS-affected hand relative to overall daily function or self-efficacy. We did not even find a trend. So the apparent enlargement of the S1 representation of the healthy hand doesn’t seem to reflect the extent to which the affected hand is incapacitated by CRPS. We also did not find any relationship between the S1 representation of the healthy hand and CRPS duration, which, again, did not support the notions we went in with.

There are a few things to note in the interpretation of these findings. We were careful to account for hand dominance in our analyses, because we had evidence from our previous neuroimaging investigation of a difference in dominant and non-dominant hand representation size in healthy controls. It is possible that a different pattern of S1 representation might arise depending on whether the CRPS affects the dominant or nondominant hand – unfortunately numbers would have been too small in each group to validly investigate this here.

What’s interesting is that as well as being nonsignificant, the relationships we found were also very small, implying that even if we were underpowered here and instead investigated these relationships with a larger sample, the results are unlikely to be meaningful in trying to explain the effect of use on S1 cortical representation. That we found no relationship between S1 representation and the duration of CRPS signs and symptoms is intriguing and raises some novel possibilities: is the difference in S1 representation between hemispheres pre-morbid and does it reflect a vulnerability to CRPS onset? Or might the difference between hemispheres arise early on in the disease, for instance soon after injury or during immobilisation?

While these questions are highly speculative and much work remains to be done before we have any definitive answers, it seems that the S1 cortical enlargement of the healthy hand is not explained by compensatory use in response to CRPS.

About Flavia Di Pietro

Flavia Di Pietro BodyInMindFlavia Di Pietro completed her PhD with the Body in Mind Sydney group early in 2014. Her project used functional MRI to investigate the brain’s patterns of activity in people with complex regional pain syndrome (CRPS) of the upper limb. She now works as a post-doctoral researcher at the University of Sydney with Associate Professor Luke Henderson. They are researching the electrical and chemical function of the brain in people with chronic orofacial pain.  She is also one of the Commissioning Editors on BiM


[1] Di Pietro, et al. Primary somatosensory cortex function in complex regional pain syndrome: a systematic review and meta-analysis. J Pain 2013; 14: 1001-18.

[2] Juottonen, et al. Altered central sensorimotor processing in patients with complex regional pain syndrome. Pain 2002; 98: 315-23.

[3] Maihofner, et al. Patterns of cortical reorganization in complex regional pain syndrome. Neurology 2003; 61: 1707-15.

[4] Pleger, et al. Mean sustained pain levels are linked to hemispherical side-to-side differences of primary somatosensory cortex in the complex regional pain syndrome I. Exp Brain Res 2004; 155: 115-9.

[5] Vartiainen, et al. Central processing of tactile and nociceptive stimuli in complex regional pain syndrome. Clin Neurophysiol 2008; 119: 2380-8.

[6] Di Pietro, et al. Interhemispheric somatosensory differences in chronic pain reflect abnormality of the healthy Hum Brain Mapp 2015; 36: 508-18.

[7] Di Pietro, et al. An exploration into the cortical reorganisation of the healthy hand in upper-limb complex regional pain syndrome. Scan J Pain 2016; 13: 18-24.


  1. Flavia, thanks for this post and your work – continuous curiosity.
    I agree with you that the underlying principles of getting to mirror therapy is that movement of the affected limb can be stimulated by visual cues originating from the opposite side of the body – level A evidence in stroke rehab – makes sense – getting there via a sequential process that minimizes fear of movement is important. Our core-self is laid out in motor coordinates.

    What I continuously observe is that people with CRPS that I see come back to a position of safety (protective pattern) after doing a movement sequence / pattern. From my perspective, this would result in periods of quiescence in M1 (especially newer M1) and correspondingly in S1 – just my thoughts.

    Finally, I would like someone to comment on the studies that have shown that signs and symptoms of CRPS can be induced in ‘healthy’ individuals with prolonged immobilization and inflammation. Please provide perspective.

  2. EG – I apologize for the miscommunication – I am not saying to stifle discussion about pain. I am saying pain is difficult to express for specific reasons – I believe that babies experience the same messages of pain as adults and yes, there are many more ways to modulate the message as adults – I don’t think pain is learned. It is a message of surprise for a reason.

    I believe that Michael Graziano identified that awareness is a model of attention (a sketch) and in the same way, language is a sketch of the communication of pain – we become trapped in language as is pre-language.
    Is it like trying to explain machine code with Fortran or Pascal (dating myself)?

    However, creating a safe space for communication is critical.*a safe space*
    On reflection, if I had thought it through, the key is trying to understand the person’s sketch and and it happens through communication. Drawing a useful sketch that has meaning is an art that I am slowly learning.

    In terms of Sadhu Amar Bharati, what is helpful to me is that immobility has an effect (there are significant effects to tissues with disuse) and I think that if we are to come back to the point of this blog, it was highlighting that the perspective was that use of the healthy hand is not important in shaping the reorganization of the cortex. What I was trying to highlight, clumsily, was that disuse is what creates the reorganization and implied use was not the same as actual use. Those are my thoughts. I will shut up now and wait for clarity.

  3. EG, I am humbled by my ignorance and with the depth of suffering that I see.
    I have been in many dark places. I look for connections. Appreciate your help.

    Many of the people I see speak multiple languages and have experienced events that are much more than I think I could handle. I still try to help.
    Work in progress.

    I believe that the communication of pain is pre-language or sub-language.
    My apologies to Melzack and McGill. The social interaction is about creating a safe place. I work with Psychology. I don’t fully understand their mind-set.
    I realize we are all different and not categories. I look for teachable moments.
    I have many blind spots. You have provided insight and I appreciate it.

    I still think that disuse reorganizes – finding the meaning in movement is hard work especially with persistent pain. The converse is also true I think.

    I have read about Sadhu Amar Bharati who has raised his right hand in the air for > 40 years in the hopes of peace and I think that there is strength in that. Professionally speaking, I am amazed by the strength of mind of others.

    EG Reply:

    “I believe that the communication of pain is pre-language or sub-language”.

    One only needs to understand that one will be safe to make such an expression of pain without being shamed or stifled. Many cultures allow some degree of expression of acute physical pain, but chronic pain is often shamed, and chronic/deep emotional pain is often strongly shamed and stifled. Shaming only serves to embed and deepen the pain.

    “I have read about Sadhu Amar Bharati who has raised his right hand in the air for > 40 years in the hopes of peace and I think that there is strength in that”.

    I had to Google that. A silent protest ; a self-punishment disguised as religious asceticism. His inner voice might be like so – “If I damage my body in a spectacular way, will you please, please attend to me and notice me?”. He’s using religion as a prop in a misguided attempt to repair his damaged self-image. [It’s well know that extreme asceticism is counterproductive in spiritual growth. Buddha proved that through extensive experimentation]. He is going about things in entirely the wrong way, unfortunately. I’m stunned no one has attempted to help him.

  4. I will answer myself as best I can. Nothing works in isolation. I am naive.
    Does the ‘surface’ in mammals lie in the salience matrix for the internal monologue of danger / threat? Teaching grandma to suck eggs?

    There are reasons why birds and reptiles look at a danger or threat through their left eye and we don’t unless severely threatened. Layers of defence.
    Does the regulation of autonomic control have left right discrimination with the left hand being activated first with cooling when we are not safe? Right dorsal insular cortex for initial activation). Crossing the arms alters the effect. You see this with infrared thermometer readings.
    Once the person feels safe, more similar left / right readings from my limited understanding – I need to check more and back up my speculations. Balance of left / right discrimination indicates a level of safety or understanding (leftwards bias in line bisection with patients with CRPS is not ‘pseudo neglect’ but evidence of ‘hyper vigilance’ from my understanding). Visual imagery and mirror therapy which now have strong evidence in stroke still rely on a level of safety for movement, whether visualized or mirrored.
    I will answer EG and wait for help. I would appreciate dialogue. Thanks.

    EG Reply:

    Hi Stu,

    Having high expectations without the requisite knowledge and skill will cause failed interventions and burnout. I’ve been to Burnoutsville, and it’s a shit of a place. But realistically high expectations are fine. 5 treatments without change will usually make suggest ceasing treatment. Sometimes I persist… depends.

    I know that if I’m fully present and congruent, then I’m providing ‘therapeutic space’. The client recognizes the ‘space’ and either: 1) begins to tell me about the *real* issue underlying the physical pain, or 2) begins to repeat certain words/phrases laden with hidden meanings (Freudian slips). In the first scenario, just stay present, don’t fix or analyze anything. In the second, then the words/phrases can be used to construct a new ‘frame’ which is non-threatening.

    I rarely look too deeply into any theory where I can’t imagine a therapeutic application. There’s just wayyy too much to read without having a “and how am I supposed to use this info..??” filter. Sometimes when I read your stuff I want you to expand into your professional experiences rather than just stay on the theory level.

  5. Hi Flavia,
    Again I waited for dialogue from you or others. I have too much an urgency.
    In terms of motor systems, I think I would agree with you and others that there are classifications and a continuum of movement from basic reflexes of protection to synergies to fine motor control. I will carry on the conversation.

    The communication of pain is an internal monologue even though it is shared with others in mammals that are socially interacting. Mogil’s work is key.
    The awareness piece is helpful with Graziano’s work and the analogy of bubbles floating to the surface with the brain as a filter for information is helpful to me. The sketch or model of attention that awareness may be is important as it minimizes the debate (input / output). Where is the surface?

    That is my understanding. The synergies are housed in motor synergy encoder neurones possibly in the dorsal horn (Levine’s work in 2014 was enlightening). The work of Butler and others on neurodynamics is helpful to gain information and Sean Gibbon’s work in teaching was helpful in looking for what is ‘uncovered’ when pain persists. John Pepper’s understanding when treating his shadow was that there are processes going on under the radar that he had conscious control of and he was able to walk away his Parkinson’s. I love the analogies that have been provided of the ‘pie’ in which medications are a piece of the pie in many cases but need to be a part only.
    Just as adding more soap to the washing machine does not get the clothes any cleaner and can have unwanted side effects (the bubbles overflowing).

    Finding the meaning and the movement that allows the meaning is key. This may not be possible if we want or ‘need’ to do what we did before and I for one have been struggling and struggling when people can’t get back to what they ‘need’ to do. My compensation is to persist in trying to constrain patterns and provide meaningful input to assist with neuroplasticity.
    I rarely figure out whether I should do a PCS score or an IEQ score or some other score to determine the reasons for persistence. I need assistance.

    I do not think that CRPS and PLP are that different. I see both and see residual leg or arm pain as potentially separate as to where the perceptions lie. It is hard to separate when you are immersed in the ‘monologue’ and I do not in any way mean to think that my words are helpful – empathy is deeper.
    I personally struggle with relationships and psychological flexibility – I struggle to explain my understanding to the person in front of me and the support networks that they have. Challenging beliefs is hard and yet I present my hypotheses in an open forum. Is this wise? Towards meaningful dialogue.
    Towards understanding and respect. Perhaps I state what is already known.
    Compassion is the recognition that we are each doing the best we can based on our beliefs and capacities. Perhaps I should remain silent. Curiosity?
    Any help?

  6. Thanks Flavia – appreciate the response and insight. I will wait for further research and look for connections. In hopes of further understanding.

    I think of sensorimotor integration and have difficulty separating. Thoughts?
    In my mind, I don’t think that GMI and investigations of repetition and dose are directed at the affected hand only but at safety first and creating an environment to practice on meaningful goals with shared decision making.
    Sometimes, talking gets in the way of learning.

    In the moment, I struggle as do the people I work with – my own self regulation is important to create a space for connection and understanding.

    EG Reply:

    “In the moment, I struggle as do the people I work with – my own self regulation is important to create a space for connection and understanding”.

    Hey stu. Same here. You know we’re in a very unusual job! Most jobs provide reward for effort/outcome, and this allows people to *use* the job to boost their self esteem. However in the healing professions, the therapist has to show up already “rewarded” – that’s a unique situation. If he shows up with such an attitude, then he won’t need the patient to change for the better in order to feel satisfied with his work. A good outcome should ideally feel the same as a poor outcome (for the therapist). If the “heart sink” feeling comes into play when someone doesn’t improve, then the dynamics are all wrong.

    Effort is always counterproductive. The reason for this is that when the therapist displays effort, he belies a need for the patient to change. That’s not to say you discard goals and intentions, but attachment to outcome is harmful to the process.

    Physio can be very rewarding, but you cannot turn up wanting ‘reward for effort’, or ‘reward for outcome’. Very different to other jobs.

    EG Reply:

    So it’s ok to have an intention and desire for rapid and complete recovery for the patient. And it’s ok to enjoy when it happens. But if you need it too much and your self esteem rests on the outcome, then desperation (clinging/aversion) has entered the picture.

    If anyone here works in a generalist Physio practice or hospital setting, ask your colleagues (in a conversational tone) about chronic pain patients. It’s very, VERY common to hear the words “oh I HATE those sort of patients! I like sports injuries!”. And among those who check themselves and maintain a politically correct front, you’ll see the heart sink, the deep breath and the sigh.

    The requirements for a therapist working in chronic pain are extremely high. It’s not good enough to distance yourself and run an abdominal strengthening bootcamp, because no one will improve. Nor is it good enough to open yourself to the patient’s massive negativity and resistance and as a result, burn out.

    stu Reply:

    EG, thanks for your insight – always valuable.
    When do you decide that you cannot help? You have tried everything that is reasonable? I agree that the potential for burnout is high with expectations.
    Acceptance of own limitations and acceptance of the *present* is key.

    Your cost / benefit ratios are interesting however what happens when you are perceived to be the ‘end of the line’ for assistance? What’s reasonable?
    I have seen amazing results with colleagues and people who have persisted and I have seen amazing results with acceptance of limitations and regrowth. I enter with uncertainty.
    Towards further understanding.

  7. Hi Stu,

    Thanks for your comments. We’re always open for dialogue on a site like this.

    The findings are difficult to interpret aren’t they? We too were surprised by what our exploration threw up. But exploration is the key word. What (we hope) came across in the publication itself, if not here, was that we were forced to infer compensatory use because we did not have direct measurements related to the healthy hand from our previous study (Di Pietro et al., Human Brain Mapping 2015) – subjective or objective. Another thing to keep in mind is that this sample was small.

    So clearly, yes, we’ve made inferences. But there are no strong conclusions to be drawn from this exploration. This study is really about hypothesis generating rather than hypothesis testing, on the back of our recent interesting neuroimaging findings of an apparently enlarged healthy hand representation in CRPS – something new.

    What is interesting though is that, in spite of all of our study’s limitations, there was not even evidence of trends in the expected directions – that is, we have no evidence of a trend of a relationship between enlarged healthy hand representation in S1 and increased use of that healthy hand. The small values imply to us that even if we were underpowered, even recruiting a larger sample may not yield evidence of the expected relationships.

    You’re so right. Perhaps the pathophysiology of stroke, CRPS (and indeed phantom limb pain) are all too different. And I think that’s where we often have problems in neuroimaging research. A lot of the CRPS and functional ‘reorganisation’ literature, for instance, is on the back of the phantom limb literature – an entirely different condition!

    Bear in mind that studies of GMI etc, and their investigations of repetitions or dose, are all directed at the affected hand. This was an exploration of relationships with the unaffected, healthy hand. There could indeed be differences there based on pathology.

    I think future work lies more in the motor system on this one. Use and movement are surely going to impart a greater effect there. And future work would definitely benefit from more objective measures of use. In the meantime, we are interested in other issues, like interhemispheric interaction in the sensory system, as being possible factors at play here in the apparently enlarged healthy hand representation in the brain.

    Thanks for your interest,

  8. Flavia, thanks again for your research. From my reading and understanding, there were a lot of inferences and some strong conclusions. I guess there is no opportunity for dialogue. Your ideas are novel which is important.
    However, inferred use (if pain was not the issue) seems to lack relevance.

    At some point, for clinicians and people with complex pain, dialogue may not be helpful. In terms of a sanctuary for discussion, I will sit back and listen.

  9. Hi Flavia, I am not sure how to fully interpret your findings. The fact that you stated you found absolutely no correlation on functional use is baffling.
    I realize it is very complex and I am still learning however I wanted insight.

    How did you rate the use of affected hand? MAL scores? I know that post stroke, often in the first 3 months or so, it is still considered a critical period in which bilateral activities and intensity of meaningful functional use are key. From my understanding, this is why there is still a tendency to front-load funding for therapy early. After a certain point, constraint of the unaffected hand seems necessary to ‘reactivate’ the affected representation in somatosensory cortex due to disuse with LOTS of intensity (Taub’s research).

    I realize there is a vast difference between someone with stroke and CRPS (however I have seen people with both and who have ‘recovered’)

    A threshold of use? Beliefs? Expectations? Is it my right to even question? – perhaps the effects of PET has benefit if it targets the belief of the person with CRPS and support network. Otherwise, expectations dictate efficacy?

    You seem to be indicating that it has nothing to do with functional use whatsoever. When we discuss thresholds of use, the research points to thousands of repetitions. It could be virtual repetitions which speaks to the benefits of GMI (at least the way it was provided in Lorimer’s work in 2004 and 2006) but not the way it was provided in other studies. More questions.
    Thanks for the study however I am hoping for further insight – any help?

  10. Hi Jennifer, there is certainly reason to believe that brain areas other than S1 (and indeed the cortex) are involved in the pathophysiology of pain, yes.
    Glad you enjoyed the piece!

  11. Jennifer Gansen says

    Thanks for this thought-provoking summary. I’ve appreciated Mike Merzenich’s book Soft-Wired, in which he describes the concept of signal-to-noise ratio to explain many conditions of disordered sensorimotor processing, including chronic pain and focal dystonia. Any thoughts on whether CRPS is related to “noisy signal processing” involving areas other than S1?