What to call the amplification of nociceptive signals in the CNS that contribute to widespread pain?

Clifford Woolf, who some may know as the ‘father of central sensitisation’, recently wrote a commentary to PAIN.[1] It piqued our interest because it was about the use of the term ‘central sensitisation’, and we suspect we know people who’d insist that the term ‘central sensitisation’ should only be used to describe changes at the dorsal horn, and who’d think of that insistence as being true to the beliefs of the person who first described the phenomenon.

Per Hansson had written a topical review objecting to the frequent, inappropriate use of the term ‘central sensitisation’ (which I shall now call ‘CS’) as though its definition extended beyond that provided by the IASP.[2] Hansson had said that the term is used loosely, and often to name a reason for clinical phenomena that have not been demonstrably linked to CS. His solution was to remove the term from the clinical context entirely, and to restrict its use to the preclinical context, where neural activity can be controlled and measured.

Woolf wrote a concise, pointed objection to Hansson’s essay, nailing his own colours to the mast in a few succinct points.

First, Woolf tackled the IASP’s definition of CS, saying that ‘CS’ should refer to “all forms pain sensitisation that arise within the central nervous system”. This was fundamentally different to Hansson’s ideas, which seemed to restrict CS to describing the phenomenon as it was first discovered: dorsal horn changes induced by peripheral input.

Woolf argued that there is little basis for such a restriction of use, because processes similar to the ‘original CS’ have been identified in other central nervous system locations and with other drivers. The concept of CS is useful principally because it distinguishes between pain problems that are driven by peripheral dysfunction and those driven by central dysfunction.

Hansson had also suggested that ‘CS’ ought to be used with subcategories denoting possible causative mechanisms. Woolf’s rebuttal was that the mechanistic drivers of the sensitisation are often difficult to identify or isolate, and may change with time, and so to limit the use of CS to sensitisation due to a particular cause would not be helpful. However, he said, further qualifying terms could be used in conjunction with the term ‘CS’ where they are known and considered helpful. Woolf’s priority seems to be to reserve the clinical utility of the term. His definition allows it to serve as a useful descriptor which hints at possible causative mechanisms, but does not require recordings of neuronal activity.

At this point we must ask for clarity on the distinction between central sensitisation and a lowered pain threshold to a given stimulus: what is the difference? To me, it seems clear that a lowered pain threshold is a clinical finding, whereas (in Woolf’s view) central sensitisation is one of two mechanisms that could underlie that finding. Peripheral sensitisation is the other option; if that can be ruled out, then the patient’s lowered pain threshold is probably due to central sensitisation.

I find it interesting that Woolf does not distinguish, in his definition of CS, between pain and nociception. Of course, to make it about sensitisation of nociception would require measurement of said nociception, which would drastically limit the clinical utility of the term ‘CS’, and indeed, perhaps the idea. However, sensitisation of pain is tricky, too, because it implies that the usual amount of pain to a given stimulus is both known and reliable. Perhaps, though, this is exactly what we do when we identify hyperalgesia: we assume that we know how much pain the person should be feeling – a questionable assumption in itself.

Hansson had highlighted the paucity of research into what he describes as a patient “setting new criteria for pain reporting as a result of a concomitant pain experience”, and how such shifting criteria might influence clinical findings. He suggests a new term for such a situation: “cognitive-emotional sensitisation”. Woolf objects to this idea by stating the obvious: that teasing apart psychological from neurological drivers is not relevant if psychological drivers influence neuronal activity – these psychological factors would merely represent another driver of CS. And then he tackles an ‘elephant in the room’: the idea that it is acceptable to think that patients are not being truthful when they report experiences that we do not understand. Although I am not convinced that Hansson himself suggested such an idea, it is a helpful and important reminder to us all to listen to the patient.

As a spectator to this particular show of words and arguments, it occurs to me that this conversation seems rather overdue. I have heard the term ‘central sensitisation’ bandied about with all sorts of meanings, and few people seem to be entirely clear about what is meant, or whether everyone is really talking about the same thing. Perhaps this Hansson-Woolf-IASP show will lend greater clarity to those conversations in the future.

Tory Madden

Tory MaddenTory arrived from South Africa to start her PhD at BiM.  She is a physiotherapist who worked clinically before turning her focus toward research.  She is interested in pretty much anything related to pain and neuroscience, thanks to some particularly inspirational teaching by Romy Parker during her undergraduate training at the University of Cape Town.
Tory’s research looks at classical conditioning and pain.  She is also an associate editor for BiM.  She tries to spend much of her spare time exercising to compensate for the vast quantity of chocolate that lives in her bottom desk drawer.  Luckily, she loves trail running as much as she does food.


[1] Woolf CJ (2014). What to call the amplification of nociceptive signals in the central nervous system that contribute to widespread pain? Pain, 155 (10), 1911-2 PMID: 25083929

[2] Hansson P (2014). Translational aspects of central sensitization induced by primary afferent activity: what it is and what it is not. Pain, 155 (10), 1932-4 PMID: 25067835


  1. Phil Austin says

    I find this central sensitisation (CS) argument fascinating in that it stimulates a fascinating spectrum of debate. I myself wonder if Per Hansson has a valid point. The blanket term of CS used by Woolf may reflect all forms of pain sensitisation in the central nervous system, but it does not reflect differing neural neurophysiological processes involving differing or sub-populations of neurotransmitters and receptor-types.

    There are some interesting recent studies of which I give two example (1-2) where researchers have used the term “CS” to reflect increased synaptic activity in the dorsal horn and the equally important term, that of “pain modulation” to reflect predominantly descending pain pathways as raised by Blaise and soon to be investigated by Andy. The two cited examples demonstrate distinct neurophysiological differences of pain sensitisation mechanisms within the CNS. Here, Bouwense and co-workers show pregabalin as an antihyperalgaesic reflected by reductions in QST ratings but where conditioned pain modulation (CPM) in these subjects remained unaltered. Oppositely, Yarnitsky and colleagues interestingly show the efficacious use of Duloxetine in subjects with reduced CPM, but where no changes were observed in temporal summation, commonly associated with increased synaptic activity in the dorsal horn.

    With several other studies showing similar results using differing pharmaceuticals, there appears to be differing processes involving different neurotransmitters and receptor types. From a clinical perspective, these findings are important as they have potential to help with clinical decision making where simple tests such as CPM and QST may help to reduce the trial and error process of finding the right medication.

    As a non-philosopher, reading this blog, has introduced me to the intrigue of neurophenomenology, so many thanks for that

  2. How does one stop the notifications of thread updates?

    Heidi Reply:

    Hi Dave
    You should be able to unclick the box at the bottom of the comments which says ‘notify me of follow up comments via email’

    Dave Reply:

    that box is not clicked when I open these 10/day emails; therefore I don’t know how to stop them


    Heidi Reply:

    The only other thing I can advise is that comments are automatically closed after a period of time so the volume of notifications will stop. I suspect as newer posts are published the number of comments will also go down. Hope that helps

  3. Adam Bjerre says

    For reference of the above see
    Deacon T, Koutroufinis S. Complexity and Dynamical Depth. Information 2014, 5, 404-423.

  4. Adam Bjerre says

    I absolutely agree with you, Mick. I’m familiar with both Friston and Clark. This way of thinking might be really useful to look at the questions that we have regarding pain, how it comes about and how it persists. Thank you so much for sharing your insights and way of thinking all these years. It has helped tremendously along with John’s contributions.

    I’m in over my head on these topics. Deacon is challenging the theories of information from Shannon and Boltzmann and it is an important contribution with respect to understanding information in biological systems IMO, but I’m groping around trying to find out how it fits somehow. But I don’t think that we can reliably say anymore that the brain passively “analyses and decides” on what to do about peripheral input. Precisely as you, John and Blaise have alluded to.

    Thank you all for this discussion. Very helpful as always.

    John Quintner Reply:

    Adam, for what it is worth, my wise neighbour keeps reminding me that our brains do not really “know” anything. Ultimately it is the person who knows and decides.

  5. Blaise Doran says

    This is a fabulous discussion, thanks for engaging.
    Unfortunately for me, I have an MSc assignment due and I’m all out of procrastination credits, so I’m out of the loop until at least the beginning of March.

    I do hope it continues, and if not… perhaps we can continue to keep in touch in a different format?


    Andrew McMullan Reply:

    I would regret a move to another format if this discussion was to be continued further unless it was open enough for some of us chumps slow on the uptake to read.

    While I can hold no candle in this discussion it is fascinating and echoes in my mind some of the darker discussions from within the motor control community. I would like to know more but am playing catch up.

    If it does move elsewhere can it be public please?


  6. Mick Thacker says

    Hi Adam – thanks for the interesting additions to the discussion – the predictive coding model outlined above is consistent with Karl Friston’s free energy principle (see Nature Reviews Neuroscience 11, 127-138 February 2010) – this theory brings together metabolic and thermodynamical considerations in relation to the functioning of the nervous system and whilst not directly focused on pain or emotion is highly relevant in my opinion to a future understanding of pain and helps explain some of the elements that we so far struggle with such as how does peripheral inputs lead to the experience of pain and how this may become “sensitised” over time.

  7. Adam Bjerre says

    I think you are correct, Blaise. Our terms and vocabulary very much shapes the way we think about these complex problems.

    The information metaphor in cognitive science is widely used, but some (Terrence Deacon, Jesper Hoffmeyer, Claus Emmeche among others) has pointed out that it is primarily used in the sense of Shannon information (bits and bytes) when the information we’re often interested in is semantic information. He argues that we have divorced information from thermodynamics (and this has been extremely useful in engineering and computers) too soon and that dynamical concepts are crucial to understand information in biology. That means that we might begin to reframe some basic concepts that are also relevant for emotion and pain as a subjective experience in terms of the extent and intensity of (metabolic and neural) work and energy.

    Thanks for reference, John. There are discussions going on at the moment regarding autopoiesis versus autogenesis (see Deacon & Cashman, 2013). Very interesting POV’s.

    Deacon T, Cashman T. Teleology versus mechanism in biology: Beyond self-organization. in Beyond Mechanism: Putting Life Back Into Biology. 2013

  8. Adam Bjerre says

    Thank you so much for this discussion, all of you.
    These are deep and intriguing questions and I think (or feel?) they are the right questions.

    One of the central questions might be in what kind of language we frame our explanations in. When we try to explain the workings of our biology we use computational language. We use words like “input”, “processing”, “error”, “coding”, “data”, “information”, “apparatus”, “feedback”, “modulation” and “mechanisms”. The question is what do we mean when we talk about information in biology (us and other living processes)?

    Is the concept of information regarded as an “entity” or “cause”. That is often framed liked that and it might lead to a (too) reductionistic view of living processes when the point of fact is that we as whole persons are constantly changing or you could say generating ourselves with respect to our Umwelt.
    “Somebody” in the world of biology has to *interpret* the information we are talking about, the question is “what is the kind of unit that interprets?” Is it the brain (or the “central” nervous system”)?. Does the body serve the brain or does the brain, in fact, serve the body?

    Some are suggesting the we might have to frame our biology in a more dynamical framework (based on “self”-organizing processes) and I think this might help us understand the basic question of how pain comes about and why it feels like something. The question still hangs in the air: Can we have a real subject or a self in our scientific theory or do we still regard it as a metaphor?

    Blaise Doran Reply:

    Hi Adam
    I think you have just illustrated what John and Mick allude to; we are bound (at the moment) by our vocabulary and constructs; our somewhat mechanistic results of the reductionist epistemology. Perhaps (and with deep respect, John) this is why pain is an aporia to healthcare? We are asking the correct questions using the wrong cipher.

    Still, If I’ve understood you well, you are also referring to are scientific principles… Just ones we rarely apply to health, or illness and wellness states… no?

    Can you expand on what you mean?

    John Quintner Reply:

    Blaise and Adam, for what it is worth, I found this paper contained an enormous number of worthwhile insights derived from the work of Varela. I agree that we need to embrace a dynamical framework and this is on offer here (and it is freely downloadable).

    Rudrauf et al. From autopoiesis to neurophenomenology: Franciso Varela’s exploration of the biophysics of being. Biol Res 2003; 36(1): 27-65.

    Blaise Doran Reply:

    Once again, thanks for the guidance, John.

  9. Mick Thacker says

    Hi John – yes I would like to propose –
    expectation – reality mismatch !
    But it isn’t going to catch on because it doesn’t sound mechanistic enough !

    John Quintner Reply:

    Mick, I will need time to reflect upon your proposition. However, it could well be that our main difficulty in answering Tory’s question lies with the limitations imposed upon us by our current scientific language. As the late Janos Szentagothai once remarked: “This is a job for professional philosophers.” I agree you are heading in the right direction.

    Blaise Reply:

    And, John, a quote from Dennett may also be apposite:
    “But there is no such thing as philosophy-free science; there is only science whose philosophical baggage is taken on board without examination.”

    Perhaps we are just examining the baggage.


    John Quintner Reply:

    I wholeheartedly agree.


  10. Mick Thacker says

    Hi John – the temptation is to say yes but I don’t want to limit this to the brain only – the whole of the nervous system sculpts and is sculpted . I think this also applies to the person and their life world, to put it in the larger context. I hope that makes sense? This inherently has features of autopoiesis included within it a la Varela, Thompson etc.
    I see our nervous systems as embodied within us as a person who is embedded in the world with which we interact. This has led to interactions that extend into, and with the environment which in turn dynamically shapes and influences us ( and therefore our nervous system) in turn we shape the world in which we live.


    John Quintner Reply:

    Mick, it does make sense to me. But can we now answer Tory’s question?

  11. John Quintner says

    Mick, when you use the term “sculpt” are you embracing the concept of the self-organizing brain? In other words, is the brain both the sculptor and the sculpture?

  12. Mick Thacker says

    Hi Blaise

    Simply put yes to your last paragraph – in fact modulation is exactly what i plan to focus on with Andy, that’s the reason for the subtle little caveat
    “……..hierarchically separated although one could easily argue against this” in my post. I think what i wanted to get across was that we should be wary of saying that one variable always co exists with another i.e. are inseparable.
    I think the work closely together part of the quote from Andy

    “Perception, cognition, and action – if this unifying perspective
    proves correct – work closely together to minimize sensory prediction errors by selectively sampling, and actively sculpting, the stimulus array.

    supports this position also. There is no doubt that whilst there is some degree of hierarchical determination that this does not exclude the co-operation of the higher and lower centres to sculpt the input signal in this case nociceptive activity. This is what i thought i referred to above but realise that i did not make this link explicit to the reader as i suggested that the higher centres can model the lower level inputs , i should more correctly have said model and sculpt!
    I hope this makes sense?

    Blaise Reply:

    Thanks Mick,
    That does make sense to me. I wonder if it might be good to gauge Andy’s thoughts on modulation of, and by, any of the ‘responsive’ (effector) mechanisms one might care to broaden this out to: Endocrine/neuroendocrine; immunological/neuroimmunological, maybe?

  13. Mick Thacker says

    Hi John and Blaise – I think is is a fascinating discussion, my take on it is that we can not say with certainty that enhanced nociception and hyper vigilance are the same – although i wish in many ways we could! Perception and understanding are high level constructs albeit shaped by error signals from lower hierarchical levels – I think that nociception and hyper vigilance are hierarchically separated although one could easily argue against this. Importantly I think a predictive coding (reliant on Bayesian modelling) perspective argues against “absolutes” in this way as the context (constantly updated by Bayesian processing as detailed by Blaise) may allow separation between enhanced nociception and hyper vigilance – clinically not all patients demonstrating enhanced nociceptive processing show hyper vigilance in many cases they show a decrease in attention/ “neglect like” attitudes. (I do however acknowledge that we must also define what we mean by hyper vigilance).

    Blaise Doran Reply:

    Thanks, what you write succinctly expresses some of the things I am trying to sort out in my head. And one of the reasons why I say that that, while it’s beguiling to think that way, it may not necessarily explain the phenomena we see. The advantage, however, in crash-testing this hypothesis (as I see it, at least) is that it it does have the in-built fluidity that you allude to.

    The question of cascading probabilistic influences from top down, and the barrage of ‘real time’ information from bottom up (the top-down, bottom up tango), and its self-feeding updates from prior experience needs a different, perhaps more encompassing construct within which to nest. Possibly, something that more successfully contains the fluidity of the underlying constructs, whilst allowing for the myriad, and on occasion contradictory, signs (or maybe more correctly ‘representamina’.)

    Apologies for the somewhat vague response; it reflects the relatively tenuous grasp I have on these concepts.

    Blaise Doran Reply:

    Hi again Mick
    I’ve had a bit of a think about this:
    “Perception and understanding are high level constructs albeit shaped by error signals from lower hierarchical levels – I think that nociception and hyper vigilance are hierarchically separated although one could easily argue against this.”

    I follow your argument here, I think. Although, perhaps we might need to consider Andy Clark’s concept of “action-oriented predictive processing” (see section 1.5 in Clark, 2013) and extrude it slightly to include not only motor but other effector systems.

    While Andy concentrates on how predictive coding may be applied to the motor system, but what of other descending modulation, such as with brainstem modulation of dorsal horn sensitivity, as part of a prediction model?
    As Andy puts it:
    “Perception, cognition, and action – if this unifying perspective
    proves correct – work closely together to minimize sensory prediction errors by selectively sampling, and actively sculpting, the stimulus array.”

    Thus, could not the perception (higher up the hierarchy) influence the effector systems lower down; as with motor output, so with sensitisation of nociceptor apparatus through decending influences?

    Does any of that make sense?

    Andy Clark (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36, pp 181-204 doi:10.1017/S0140525X12000477
    (available here: http://users.monash.edu/~naotsugt/Tsuchiya_Labs_Homepage/Bryan_Paton_files/Commentary%20.pdf)

  14. Mick Thacker says

    Hi Blaise – thanks for you comment and great links to Andy’s work – he mentioned you had been in contact with him – perhaps you would like to collaborate with us in the future?

    Blaise Doran Reply:

    Hi Mick – I’d like that very much, and would be most honoured.

    Except, I should ‘fess up that I’m just a jobbing physio (albeit one that cares enough about the underlying processes to fry my brain with these concepts.)
    Perhaps we could have more of a ‘fireside chat’, and you can sus out if I’m smart enough to play with the big boys (…meant with the deepest respect to you both.)

    If you wish, Andy has my email address. Otherwise, would you object to me contacting me via your KCL email?

    Mick Thacker Reply:

    Would be a pleasure to here from you Blaise so yes please – as far as I am concerned I am equally out of my depth – wouldn’t dare bracket myself with Andy in this arena or any other for that matter – just lucky/happy to be taken seriously!
    Look forward to hearing from you.

  15. John Quintner says

    Thanks for the very informative links, Blaise. I was interested to hear Andy state that based upon the model of hierarchical predictive processing, perception and understanding are inseparable. May I take this to infer that enhanced nociception and hypervigilance are also inseparable?

    Blaise Doran Reply:

    John, thanks for your comments.
    I can see from your first post that perhaps you support this view. Or have I misunderstood you? I’d hedge a tentative “yes” to your question, but I am also tempted to think that the predictive coding model is an incomplete explanation, or at least part of a larger, more inclusive model that my tired brain is struggling to understand (hence, I won’t speak to it in this response.)

    I think, perhaps the best I can say is that this clinically makes sense to me from on a number of levels, but it’s hard for me to articulate clearly (but that won’t stop me trying for the sake of it):
    Say we start with the (most basic) ‘triad’ of Bayesian modelling – we have the posteriors (prior knowledge); the probabilistic modelling (best guess influenced by prior knowledge); and prediction error (the residual error between the data coming in and the prediction; in the case of pain, from sensory afferent information).

    If the brain does, indeed, work in this manner in terms of all levels of processing (and we don’t really know; it’s essentially a hypothesis) there are also mechanisms whereby the brain can have such a strong predictive model that it overrides the residual error. (This is perhaps how conjurers make a living; they set you up to expect one thing to happen, then it doesn’t. But your predictive model is so strong that you experience surprise, awe and wonderment at their trickery).

    You can apply this to most situations that involve the nociception to pain transduction/translation process. (Think Lorrimer’s snake bite anecdote, parts one and two.)

    So, perhaps because of this we can also then postulate that the posteriors (prior knowledge) re-configure the prediction to expect – in a form of “if-this-then-that” manner – that the next time one is in such a situation it is going to be harmful, then the residual error may be overridden by a strong prediction model. (Might help explain fear avoidance?)

    But, does this mean that the nociceptive apparatus can primed to expect harm, even if it doesn’t exist, because of such a robust prediction? One could argue that, potentially. Actually, I think it’s rather tempting to think in this manner… (and so that makes me a bit sceptical of my own thinking!)

    One thing that I don’t think this hypothesis does is fall the Cartesian theatre trap, not as far as I can tell, as it relies on a ‘bootstrapping’ process whereby what is experienced (+/- the residual error being accounted for) feeds into the posteriors which then feed into the prediction, rather than that little man in your head running the show ad infinitum.

    Well, after all that, it doesn’t quite answer your question, does it?

    Anyway, John, perhaps a clinician of your erudition should collaborate with someone like Jakob Hohwy. He’s at Monash University in Melbourne.

    John Quintner Reply:

    Thanks for the complement, which is quite undeserved. My colleague Milton Cohen first pointed out to me the impossibility of distinguishing between enhanced nociception and hypervigilance as he suggested they were one and the same thing.

    Yes, it is so easy to fall into the “Cartesian trap”. In fact it was this very problem that led us to explore what has been called The Santiago Theory of Cognition developed in the 1970s by Humberto Maturana and Franciso Varela. A good introduction to this theory can be found in “The Hidden Connections” by Fritjof Capra (2002). It allowed us to escape the dreaded dualism and resulted in our 2008 paper – Pain Medicine and its Models, helping or hindering.

    Blaise Doran Reply:

    John, thanks for these leads.
    As to the complement – I would hazard that there are a great many people who admire your writing and, by extension, thinking. I wasn’t (to use the American vernacular) “blowing smoke”; merely going by the work of yours I have read, with whom you have collaborated, and your consistently keeping us on our toes.

    I have much to learn, so rarely stick my head above the parapet. Thanks for indulging me.

  16. I’m growing increasingly frustrated with clinical trials that assume to have identified a clear demarcation between “nociceptive” pain and that arising through other supposedly distinct mechanisms, such as CS. A trial (Aasa et al, pp 77-85) was just published in JOSPT this month making such a claim in patients with low back pain of >3 months duration. I quote from the Methods section: “Negative pain beliefs and non-optimal coping strategies were allowed, but it was ensured that mechanical factors, and not beliefs, provoked the experience of pain.” They then went on to describe how they made this distinction between beliefs- and mechanical-provoked pain by having the subject perform the “corrected” provocative movement, during which “the participant’s pain was expected to decrease.” Therefore, I suppose only those patients whose pain was reduced with the “corrected” movement qualified for inclusion.

    Does anyone else see the error of assumptions here? I see this kind of thing all the time in the PT literature, and, what’s more, this is the way a lot of clinicians think when they assess patients in the clinic.

    John Quintner Reply:

    Well spotted, as we say in photographic parlance! How some of this stuff ever gets past the reviewers is a burning question. There is such a lot of muddled thinking out there!

  17. John,
    I conceptualize chronic pain of a benign nature as being a central phenomenon with over-attention to sensory input as manifested by a focus that locks into a pathway between brain (in chronic, per Apkarian (with Hashmi, Brain 2013 on chronification), it has shifted from sensory cortex to emotional centers) and area affected. This can be experienced clinically as a “stuck” cycle, if you will.
    MT’s comment on generative causation jives with Derbyshire’s report on Hypnotically Induced Pain in Neuroimage 2004.


    John Quintner Reply:

    Thanks David. Your response harks back to the “good” pain/”bad” pain (eudynia/maldynia) idea that we attacked in our fairly recent contributions to Pain Medicine.

    David Schechter MD Reply:

    would enjoy that reference; a good attack means that something is stimulating to someone in some way!



    John Quintner Reply:

    David, yes, it is a stimulating issue.

    Here is the reference:

    Quintner JL, Buchanan D, Cohen ML. Maldynia as a moral judgment [letter]. Pain Medicine 2011; 12: 1130.

    P.S. you might also like to read our 2013 paper that touches upon this issue.

    Cohen ML, Quintner JL, Buchanan DA. Is chronic pain a disease? Pain Med 2013; 14: 1284-1288.

  18. to Mick Thacker,
    My clinical experience and review of the literature totally agrees with your comment:
    “…many neuroscientists including myself are becoming interested in the concept of generative causation – that is the ability of the nervous system (particularly centrally) to be able to generate accurate “copies” of peripheral inputs therefore not requiring any ongoing activity from the periphery although these generative models are shaped and based on previous exsposure to afferent inputs. In addition centrally mediated retrograde signals such as NO and PGE2 at the dorsal horn once established do not require further ongoing inputs to maintain their release. ”
    I have seen this happen and I have seen patients be untrained of this acquired “skill” (a negative one to have). I do not believe that persistent peripheral stim is required. The neural loop develops and gets stuck…. there are ways (cognitive/psychological) to unstick it.

    David Schechter MD
    USA Los Angeles

    John Quintner Reply:

    David, can you please explain what you mean by “gets stuck” when you use this term in relation to (re-entrant?) neural loops? I might then be able to follow your argument.

  19. Mick Thacker says

    Hi Tory et al – I am not sure that I can agree with Geoff’s point that most of the pain world believe that a peripheral input is needed to maintain CS. I admire Geoff’s work considerably but as a self confessed nociceptor jockey he may want us to believe that but many neuroscientists including myself are becoming interested in the concept of generative causation – that is the ability of the nervous system (particularly centrally) to be able to generate accurate “copies” of peripheral inputs therefore not requiring any ongoing activity from the periphery although these generative models are shaped and based on previous exsposure to afferent inputs. In addition centrally mediated retrograde signals such as NO and PGE2 at the dorsal horn once established do not require further ongoing inputs to maintain their release. On the last pooint we are very much in our infancy at developing assessment strategies that can accurately determine peripheral and central mechanisms.

    John Quintner Reply:

    Mick, are you building upon the work on perception by Gerald Edelman? Are you arguing the case for a dynamic interaction between top-down imagery and bottom-up sensory signals?

    Mick Thacker Reply:

    Hi john – this perspective is similar to Gerald Edelman’s ideas but is based primarily on the work of Karl Friston, Andy Clark and Jakob Hohwy – predictive coding like Edelman’s view does depend on interactions between top down and bottom up neural processing. Implicit is the idea that the nervous system is hierarchical and that higher levels are able to generate “models” that predict lower level inputs – errors between the expected generative model and the actual inputs result in error signals which are passed to higher levels to update the model in a fashion consistent with Bayesian theory. Thus the higher centres are able to construct(generate) “peripheral inputs” including nociceptive inputs – therefore we believe that there is no need for ongoing peripheral inputs in order to maintain CS. Importantly though this perspective is not brain bound I.e. these models are influenced by previous and/or current exteroceptive and introspective experiences of the person in the environment they live. I am hoping to work directly with Andy Clark to investigate the generation and maintenance of pain as we believe it has the potential to progress our current understanding of pain and CS. Abby Tabor has done some tremendous work during her PhD in this field and led a workshop at The World Congress of Pain covering some of these topics.

    John Quintner Reply:

    Mick, yours is a tremendously exciting project that lands you fair and square in neurophenomenological territory. My reading abruptly stopped short when I could not cope with Varela’s preliminary work in this area. Anyway, here is the reference, if you have not already seen his Chapter: Varela FJ. The specious present: a neurophenomenology of time consciousness. In: Petitot J, Varlela FJ, Pachoud B, Roy J-M, eds. Naturalizing Phenomenology: Issues in Contemporary Phenomenology and Cognitive Science. Stanford: Stanford University Press, 2000: 266-399.

    Blaise Doran Reply:

    Glad to see these concepts filtering in.
    For those who are less familiar with his work, Andy Clark’s paper Whatever next? (http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8918803&fileId=S0140525X12000477) and the commentaries that accompany it are worth picking through.
    He did a great lecture at the 9th International Symposium of Cognition, Logic and Communication held at the University of Latvia in Riga in 2013 (available on YouTube – https://www.youtube.com/watch?v=05P41FQlgjI&feature=youtu.be&t=11m50s ).
    There are a few ‘coal-face’ clinicians around the world interested in this concept and how it may be applied to those with chronic pain.
    It may not be the full answer, but it is taking us in an interesting direction.

    I do hope we see the fruits of your collaboration with Andy Clark soon.

  20. Thanks Tory. That word “predominantly” makes a big difference alright in terms of removing any suggestion of a dichotomy.

  21. Tory Madden says

    Thanks, everyone, for your comments and discussions. Keep them going!
    Geoffrey and Kieran, I must apologise for that phrasing: I did not intend to imply that peripheral and central drivers are mutually exclusive! I actually intended that sentence to be a paraphrase of what I understood Clifford Woolf to be saying, but I see that it implied a dichotomy. Perhaps it is more helpful to provide the original phrasing, which was: “Central sensitization continues, in my opinion, to have utility because it helps to distinguish those forms of pain that arise predominantly peripherally (such as nociception, peripheral sensitization, and spontaneous pain associated with ectopic activity after peripheral nerve lesions) from those in which there is a combination of a definable central component and an increase in pain sensitivity” (p.1911 – see full reference above).
    Kieran regarding your question about distinguishing peripheral and central contributors… yes, I think we usually try to work out how the balance of influence lies, in the clinical context – at least, that is my approach. I think that is what you meant with your last suggestion. I’m not aware of any way to identify a purely peripheral or purely central problem, and in fact, I’m not aware of any evidence that would suggest that such a thing exists. Again, I think I am just repeating what you said – and agreeing!

  22. super post Tory. thought provoking. and hard to argue with Prof Woolf!

    Geoffrey’s point that “most of us in the pain research world are coming to the conclusion that it (CS) has to be maintained by input from the periphery as well, albeit less input” is worthy of reflection too as people sometimes still try to suggest pain is either “central” OR “peripheral”.
    Based on that, a quick question for you Tory (or anyone interested really). You stated that the “concept of CS is useful principally because it distinguishes between pain problems that are driven by peripheral dysfunction and those driven by central dysfunction” – any good data on how these can be differentiated with validity clinically? and are they really being distinguished (as distinct entities), or is it more that we are trying to see which is more dominant in this individuals experience of pain?

  23. Geoffrey Bove says

    Just looking through newest issue of PAIN:
    Peripheral and spinal circuits involved in mechanical allodynia
    Alice Arcourt, Stefan G. Lechner
    Free access, though the figure seems messed up a bit.


    Please note Reference 7 in this paper, it was a clear cut finding, oft cited.

  24. Geoffrey Bove says

    This is a wonderful post, thank you Tory. I have three points.
    1. As a nociceptor jockey —- we do not know where nociception ends and pain begins. In the spinal cord, all input from nociceptors is modulated — does “pain” start there?
    2. Central sensitization is a quite new phrase, but a fairly old phenomenon. Dr. Woolf would be the best to describe the history, which we always must acknowledge. Research in this area was triggered by memory research, applied to the dorsal horn. We used to teach “long term potentiation” and “long term depression” as components of “spinal memory.” When pain research blossomed, the new term was devised.
    3. Most important. You stated “The concept of CS is useful principally because it distinguishes between pain problems that are driven by peripheral dysfunction and those driven by central dysfunction.” I’m not sure what Dr. Woolf wold say to this (perhaps write and ask him?) but CS almost only happens secondary to a peripheral dysfunction, and most of us in the pain research world are coming to the conclusion that it has to be maintained by input from the periphery as well, albeit less input. Actually the word “sensitization” implies this.

  25. David Schechter MD says

    People ask me if Tension Myoneural Syndrome (TMS) is central sensitization and how they overlap. The article above helps clarify the need for widespread pain in the CS definition. In the TMS definition, the pain pathways are central but the peripheral manifestations may be localized, migratory, or widespread. The improvement with a clinical program based upon cognitively directing the patient’s thoughts away from the pathway or pain “loop” has overlap with CS treatment. My book Think Away Your Pain reviews this in detail.

    John Quintner Reply:

    David, if I am correct, what you are describing as TMS is based upon Dr Sarno’s hypothesis of “downward causation” of spinal pain. This harks back to my question as raised above – how can anyone tease apart hypervigilance and enhanced nociception? My reading of Dr Sarno’s work suggests that he has fallen into a trap on this very issue.

    David Schechter MD Reply:

    Someone wrote me today and said– Sarno came up with a treatment, now we are all reverse engineering the mechanism.
    Clinically I help people, often in powerful and amazing ways with the updated version of this sarno model that I use. Explaining it continues to evolve as we learn more about the nervous system.
    I note that some enjoy tearing apart his theoretical work; but myself and the other doctors doing similar work these days (Dr. Sarno retired) get great results with slight variations on a theme.

  26. How about Chronic Sensory Feedback (CSF – though it’s unfortunately also an acronym for cerebrospinal fluid)? I work with a lot of these cases and (although there are possibly multiple possible causes) the main cause appears to be a lack of choice in what is given attention and how that attention is applied. Whatever sensation we allow to be the dominant one in our awareness will be the one our sensory system thinks is most important to us. And the emotion we have when we feel that (and the thought loop we then get into that reinforces the emotion) sets up strong good/bad associations in both the amygdala and hindbrain. Stephen Levine’s Somatic Experiencing or Pat Ogden’s SensoriMotor Psychotherapy provide effective tools for working with this, by retraining attention to seek out health instead of pathology. I have come across a handful of people who are so terrified of their bodies that Resourced Attention retraining is way too soon (and these people often need help for dissociative issues – which is complex work), but most people pick it up very quickly and the pain reduces, sometimes very quickly, sometimes with a little bodywork to address usually relatively small physical causal issues.

    John Quintner Reply:

    Andrew, although I am not a philosopher, it seems to me that your suggestion (Whatever sensation we allow to be the dominant one in our awareness will be the one our sensory system thinks is most important to us) is in line with that which Daniel Dennett called Cartesian materialism:

    ‘Let’s call the idea of such a centered locus in the brain Cartesian materialism, since it’s the view you arrive at when you discard Descartes’s dualism but fail to discard the imagery of a central (but material) Theater where “it all comes together.”‘ [p. 107]

    Are you postulating the existence of what Dennett calls “an authoritative traffic cop homunculus already in place” to make the choices as to “what is given attention and how that attention is applied”? [p. 267]

    Is that fair comment?

    Reference: Dennett DC. Consciousness Explained. London: Penguin Books, 1991.

  27. Tom Shelton says

    Agreed, it would appear to fit in with what I have learnt about other body systems actions to say that it’s probably a bit of both and subsequently CS would be justifiably a broad term. I also agree that the variable understanding of the word should ideally be clarified, however is that not the inherent nature of words and their meanings?

  28. John Quintner says

    Tory, thanks for raising this important issue. Surely it will prove impossible to tease apart enhanced nociception and hypervigilance. They appear to be one and the same thing.