The Graded Motor Imagery Handbook – a review

I love getting presents, and I love books, so what could be better than getting a book to review as a present!

Graded motor imagery (GMI) has become incredibly popular in pain management, especially for people with unilateral pain.  It’s a treatment that is intensive for patients/participants, but is non-invasive, means the person with pain develops self management skills, and has level B1 evidence.  For those who don’t know – level B1 evidence means there are several RCT’s, and at least one meta-analysis showing support for this approach.

Back to the book.  Like all the NOI books it’s an unusual size, has groovy graphics and an easy-to-use layout. It’s a spiral bound book of over 140 pages with a great index (yay!), logical layout and has room for notes.  The illustrations and photographs are clear and provide excellent guidance for clinicians.  Chapters divide the book into sections of background info including theory and evidence; how to conduct treatment with GMI (clinical reasoning); metaphors (David Butler’s favourite teaching tools); and a whole chapter on how to use the materials available from NOI to support GMI treatment.   Each chapter can stand alone, and it’s not necessary to read from beginning to end – but of course, it does help!

This book isn’t for beginner therapists working in pain management. There are some assumptions about the level of clinical reasoning required and patient selection that are not fully explored, and true psychosocial aspects of managing pain – and the translation into the “real world” – are omitted. This is fine as long as clinicians are aware of the need to identify people who will benefit from the approach and as long as clinicians work within an interdisciplinary team environment.

Patients/participants need to be motivated, committed, and relatively psychologically well, without complex psychosocial contexts such as family/relationship issues, litigation, personality disorders, drug/alcohol problems or other cognitive impairment.  GMI has best application in people with unilateral pain such as CRPS and phantom limb pain – although it has been extended to other pains.  These factors may influence the degree of engagement and time required to carry out GMI, and may influence the outcome.

Back to the book again!   I love the chapter written by Lorimer Moseley on the neuroscience underpinning GMI.  His writing is clear and provides an excellent scientific basis for the approach.  He doesn’t extend his writing into psychological aspects of pain beyond the concepts of what he calls “neurotags”, or “interconnected neurones … that produce an output”.  Neurotags involve areas across the whole brain and, when activated, produce, for example, the experience of a whiff of fresh bread (along with the scent, the associated emotions and cognitions from past learning and the anticipation of future action).  I have learned these associations as just that – associations between various aspects of learning and anticipation, and have called them the neuromatrix – but NOI have used the term neurotag, and I guess it’s as good a name as any.

The chapter on conducting GMI treatment written by Tim Beames is extremely clear and well written.  While it’s possible to use this as a sort of cook-book to treatment, with the information from other chapters such as Lorimer’s neuroscience, and Butler’s metaphors, it becomes far more flexible.  I like this.  It is a chapter that I think many clinicians will turn to regularly – but as is emphasised throughout this handbook, patients/participants should read this stuff too.

Worth getting? Yes, I think so. 

My caution lies in over-interpreting the application of GMI beyond the evidence-base.  If you intend to try it with a patient/person with pain, please explain that this is an experiment that you and the person are conducting to see how this treatment works for him or her.  Select patients appropriately, checking for motivation, factors that could distract from engagement in treatment, and type of pain.  Record a baseline. Monitor progress. And involve the other members of your treatment team (particularly occupational therapists) to help transfer what is practiced out into the wide, wide world.  After all, the most complex context of all is being engaged in occupations like grocery shopping, driving, cooking a meal, playing a sport – where the environment is always changing, contains all those triggers, and where the brain is involved in multiple decisions moment-by-moment.

Where to get it? Go here – and let ’em know I sent you.


  1. It is a good book! Covers all the bases in an easily understood (for clinicians) manner. Not sure whether some of my patients would find it as easy to read, but it certainly gives the therapists the tools to educate the patients on why we would use GMI. I recommend reading explain pain or something similar prior to this book to give you a good grounding in general pain knowledge.
    I give it 4 out of 5 neurotags.

    1. Liked your review Bronnie. I have just read both the GMI book and a paper from the european Journal of Pain 16 2012 looking at an audit into the use of GMI in 2 pain management units in the UK. They did not find an improvement in pain but reading between the lines it is likely that the patient population had myriad psychosocial issues as do most patients who are seen in interdisciplinary pain units. In my experience its hard to focus on GMI when someones relationship is breaking down, they have no money etc.! I agree that it needs to be used thoughtfully. You dont want to try it with someone at the wrong time and be unsuccessful, because it may work at another time for them and you dont want to create negative expectations. I also agree that it is not for inexperienced therapists. Mick Thacker said something at the NOI conference that I felt was very true, namely that we need to give patients ‘the tip of the iceberg’ pain education but that needs to be supported by our own pain knowledg which is the rest of the iceberg. Books like Explain Pain or manage your Pain are the iceberg tip. If we are serious about pain our knowledge needs to go much deeper than this. Pain is complicated and books written to share with patients must necessarily simplify things.
      Ruth Hardman

      1. Nice summary of my own feelings, Ruth! I guess that’s what I mean about not “over-interpreting” the evidence base. For every person we work with, we need to consider the relevant factors to their presentation – why has this person presented in this way at this time? Develop a formulation, and work from that, rather than applying anything carte blanche.

  2. I’m debating whether to buy this book or not.

    It’s not clear to me how good this approach is, i don’t believe enough studies have ruled out that it maybe no better than other placebos.

    In any case, what I would like to see is a detailed explanation on how awareness and mirror box therapy illusions or even imaginations of painful limb movement significantly relieve real chronically painful limbs. If this is all about activating pain inhibition pathways down to sensitized dorsal horn neurons without causing pain then this can probably be done in many ways. Also, as was mentioned, GMI seems like only a portion of what contributes to a patients pain by not involving the whole psychology of chronic pain.

    1. I think any treatment needs to be considered in light of the evidence base and the individual presentation of the person you’re working with. I agree, there are many ways to activate pain inhibition – this is one that not only does that, but also appears to contribute to cortical reorganisation. It’s difficult to clearly articulate the way this approach works – like so many treatments, we don’t know exactly how the mechanisms influence the experience of pain. At the very least, GMI seems to have few side effects, unlike most pain medication, and this suggests it could be worth a shot.

      1. Would you expand a bit more on why cortical re-organization is a desirable thing?

        The brain is very plastic and adaptive. Parts take over for other parts especially if spares parts are available. If the reorganization of the sensory cortex to pain occurs then why is that considered undesirable instead of a desirable adaptation? Is it really because “spare brain” isn’t available???

        If unavailable then I think it’s the case that the sensory and motor system reorganize together. So maybe if I inhibit pain and also activate the motor system in non-painful ways that where previously painful by whatever approach then I should get similar or even better results as GMI.

        What do you think?

      2. I think you probably achieve cortical reorganisation through the process of “inhibiting pain” and activating the motor system in non-painful ways. For some conditions (and phantom limb, also CRPS both strike me as such conditions), pain inhibition is either difficult to achieve, or can’t be achieved without multiple undesirable side effects. In both these situations motor imagery may be a first step towards developing a plastic response that approximates normal. In both of these situations, it is thought that cortical reorganisation has occurred in an undesirable and abnormal way, giving rise to the pain experience – GMI provides one way for a person to begin the process of returning to homeostasis, or at least to more normal cortical arrangements.
        Personally I think this is just another approach to help people with intractable pain problems regain function, and as I pointed out, isn’t the only one and isn’t carried out in isolation. The advantages of this approach is the relatively low cost of materials (compared with medications), and the low side effect profile.
        In people who can manage to activate the motor system while pain is inhibited – great! Go for it! One question I don’t have the answer to (OK, yes, there are many questions I don’t have answer to!) is whether pharmacological or anaesthetic blocking might also begin the process of normalised movement. My observations over years are that they’re often given in isolation from self management and away from an environment in which the hard yards needed to “do rehab” are supported. And these are often not given in the context of the very clear “education” or reconceptualisation that is advocated in GMI. Maybe it’s simply that allied health are less expensive than biomedical clinicians and therefore both have the time AND the inclination to listen and respond to individuals’ beliefs. Thoughts?

  3. OK, this is great that you are responding with such thoughtful answers! Thank you!

    My next question is something alluded to in your qualified endorsement of GMI and that is the overall psychology of the person in chronic pain. This site has a book review of GMI with pictures:

    However, I don’t see how motivation, intent or meanings the patient has fits in those box diagrams / flow charts of brain regions. It’s like a partial description of the passive flow of nerve impulses.

    I have one more question on the relationship between sensory information from the body and muscle activation patterns. Are they related in a one-to-one fashion or in a one-to-many way. In other words, a pattern of sensory information from the body can be related to multiple but similar patterns of muscle activation (i.e., intensity, duration, different muscle groups) for an intended movement or is it the other way around in many-to-one fashion?

    Thanks again.

  4. Some reflections from an author.
    Thanks for your thoughts here – it all helps. I think (hope) in a few years we will look back as see GMI as a crude process but at least it got a lot of therapists into the brain.
    We are wary with it and Moseley and I have only just been game to call it “a toddler” after previously considering it “embryonic”. Never the less, it is often bloody hard work, it obviously does not stand alone, it requires clinical reasoning rather than algorithmic thinking, it probably requires significant conceptual change in pain notions, and many of the targeted pain states such as CRPS have been traditionally harder to manage than many cancers. The therapist who realises that there may be something new around and recalls long discharged patients with severe neuropathic pain may be in for a shock. I must admit to being scared when I hear about how some people are using it

    But the authors are cautiously excited and the book has taken a lot of pressure off us answering questions from therapists and patients
    Perhaps this example below is a summary of where we are currently at.?
    I sat with a spinal cord injured patient T12 complete with intractable left leg pain for 6 years. He was repeatedly slow ( though not inaccurate) to identify his left side in left/right discrimination testing. Thoughts swirled in my mind such as “is it relevant, is it worth chasing, what does it really mean, what else should I do, nothing else has even touched it and he has been everywhere, let’s have a go, what is going on in his spinal cord and brain.” When he left he commented “now at least I have something to work on”. When I realised that this perception was also founded on a lot of novel brain/pain education that I offered and not just thrill of a computer possibly easing the pain, I judged that a trial of imagery based therapy was clinically worthwhile and not just false hope.
    But it sums up where we are perhaps – There is some basic science support, some clinical trial support in other intractable neuropathic pain states and this is worth further exploratory and clinical trials and encouragement to use it in the clinic. Currently the book funds much of the exploratory research – pharma is not so interested). Discussion sites like this certainly help.

    David Butler

    1. Thanks so much for adding to the discussion on Healthskills Blog, David. I think clinicians, when thinking about any intervention, really need to get their clinical reasoning in action so there is a definite clinical rationale for the approach they choose. Chronic pain management isn’t at the point where we have an exact science of what works for whom and how, so it is a process of setting up hypotheses and systematically testing them. The handbook is great – lots of clear information, practical steps and the usual NOI humour!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

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