Distorted Gait—A Problem of Shaping

FIGURE 1 Increasing walking speed to inhibit knee buckling.

Commentary: Distorted Gait: Biomechanical Implications, Historical Reflections, and Future Directions

Steven Z. George and Mark D. Bishop

It is a great privilege to be invited to help with an introduction to this chapter of this book. Indeed, for people inspired by Wilbert Fordyce’s writings, there can be no higher honor than to have a chance to comment on the man who “started it all.” After re-reading Chapter 11 we had a similar reaction to Lorimer Moseley’s commentary. It had been some time since we revisited this foundational material and we were amazed at how clear, compelling, and valid Fordyce’s clinical observations were. As might be expected based on our shared profession, we held common viewpoints on Fordyce’s work and impact. Lorimer Moseley has very eloquently described the state of neuroscience behind many disordered movement patterns, including potentially distorted gait.

In this commentary on Chapter 11, we add to an area intentionally left underdeveloped by Lorimer Moseley, namely Fordyce’s observation that ‘It may (also) be that the distorted gait has persisted so long that a substantial skill deficit for normal gait has developed and the patient may no longer know how to walk properly.’ We will do this by providing historical reflection, expanding on biomechanical details, and identifying some areas that may be ripe for future research. It is our hope that the original chapter and observations put forth in this accompanying commentary spur future clinicians and researchers to make progress on this universally perplexing condition known as chronic pain.


What was most striking about this chapter is how elegantly Fordyce integrated sensory, perceptual, and motor elements of pain. Interest in this integration is something that has become a research priority in the past 10–15 years but was not on the radar at the time of Fordyce’s writings. It is one thing to note that an author is “ahead of his time”, but in this particular case some dissection of exactly how far ahead of his time is warranted.

Philosophers, anatomists, and researchers have spent great time and energy deconstructing the pain experience into separate components, and then realizing how these insufficiently account for the entire experience. When one takes a historical perspective on Fordyce’s work, there is a realization that his writings were occurring at an extremely influential period in our understanding of pain. The Gate Control Theory of Pain, a ground-breaking model based on pattern recognition, had already been published and one of its many influences was to provide the neurophysiologic groundwork for the separation of the perception of pain into fundamental sensory and emotional elements. In chapter 11, Fordyce clearly acknowledges this separation, but unlike others who observed it, he did not dwell on measuring or deciphering the individual elements of pain. Instead, the solution he advocated was to focus on the consequences for motor function (i.e. limping, falling, or knee buckling) and the subsequent behavior (i.e. distorted gait).

This approach of synthesis, rather than reduction, was a noticeable contrast to what had historically occurred in the field of pain research. The culmination and genius of the chapter is Fordyce’s clear demonstration of the practical advantage of such a perspective on managing pain. Fordyce was able to offer his patients a viable and effective clinical solution that was not based on one element of the pain experience, but on all of them. Fordyce’s approach suggests that instead of targeted and focused approaches that rely on identification of primary pathways for providing pain relief, we could “start at the end” by altering resultant behaviors that no longer serve their original protective function.

The message that treatment options exist for the output of pain, rather than only having the option to modulate nociceptive input from individual causal elements was a ground breaking one. Interestingly this important message is one about which the pain research field needs constant reminding. Fordyce’s work is particularly pertinent when highly touted chronic pain “cures” focused on a single receptor or pathway inevitably fail to meet their original promise when tested in clinical settings.


In this chapter, Fordyce describes the management of a patient who has a problem with walking because of back pain – in this case the leg “gives out”. The reader may observe the following: Next time you go for a walk, pay attention to the way your body moves and you will notice that your thorax (chest) moves in the opposite direction to your pelvis. Quite a few studies have indicated that this ‘counter-rotation’ is reduced or missing in people with back pain and there are associated changes in how the back muscles are working [6]. If there is counter-rotation, the speed at which this happens is much lower in people with back pain than in those without, particularly as that person needs to increase how fast they are walking [3,6]. In essence, these people are walking more stiffly. Also, the motion of the trunk in people with back pain is less variable or has lower degrees of freedom. Think of variability in this context as available options for movement. It is better to have more available options for movement so that when you are walking you are able to keep walking if you need to scratch a leg, reach for a phone, or anything else you can think of doing. Those with back pain seem to lose this freedom. So it seems that the person with back pain can’t adjust how they are walking when and if they need to.

Changes in movement in people with chronic low back pain not only happen during walking but are also present during other activities. For example, people with low back pain don’t bend the back as much when they are reaching for something as people without pain, even when the potential for movement is there (i.e. there is no restriction), and the change in the way they reach because of this pain still persists even 4 weeks after the pain has gone [10]. Hodges and Tucker [2] propose a model that suggests that motor adaption to pain includes changes in muscle activity and modified movement that has short-term benefits but long-term consequences (such as decreased movement and decreased variability). The short term benefit of modifying movement is to decrease stresses placed on the currently injured and inflamed tissues or structures, thus protecting the injured tissue or structure from further damage. These adaptions to movement do not take very long to occur. Give someone who is pain-free back pain and they very quickly adopt protective strategies to decrease movement of the back. This is shown nicely by Trost et al [12] who induced back pain experimentally in otherwise healthy people using high intensity exercise. These authors observed that participants reduced how much forward bending in their backs they used when they were asked to reach for a target.

However, as mentioned above, the changes in movement observed that initially served to protect against further injury can persist in people with low back pain weeks/months after the pain has resolved [10]. Now there is a long-term consequence. In this case, the person has no back pain but continues to move as if they do. This reduces options for moving his or her back and limits the available ways in which he or she is able move to perform any tasks that might include bending the back. If a person can only perform a task using one movement strategy, he or she becomes more likely to avoid activities that require a different movement to complete the same task. For example, having to change position rapidly to move out of the way of an unseen object or change his or her gait to avoid stepping in a hole or up a curb.

The implications of not bending the back are not only of biomechanical significance. The changes identified by Trost et al [12] were related to fear of the pain and other beliefs that the person might have, not necessarily the pain itself; that is, the more fearful someone was of the pain, the less they moved. This was the case even when the person had enough range of motion in his or her back to be able to do the full movement required to reach the target! Trost and colleagues also noted that fear of potential pain also predicted how much effort some would use during a test of strength [11]. The authors of these studies suggested that the reduced effort and movement is an example of a rapidly adopted avoidance strategy. This phenomenon is also clearly shown in that anticipated pain (pain that might occur during walking) is a better predictor of how fast someone would walk than actual pain (pain that did occur while walking) in those with chronic back pain [1]. Avoidance of activity, reduced physical performance, and decreased variability in movement likely lead to the restricted movements displayed by patients with chronic low back pain.

Fordyce’s observations in Chapter 11 were based on experiences with patients suffering persistent low back pain but, as pointed out by Lorimer Mosely, has broader implications for other pain conditions like Complex Regional Pain Syndrome (CRPS). Interestingly, other patients that are not typically thought to experience “chronic” pain also have shown signs consistent with the original observations. A recent systematic review of the results from studies of patients recovering from reconstructive surgery of the knee, for example, indicated that fear of re-injury altered movement of the leg, reduced the person’s function and slowed the return to regular sporting activities [9].

So, people with pain definitely move differently than people without. These changes in movement occur very quickly after the onset of pain and though initially may be adaptive in nature, they continue to persist once that pain has resolved even when no longer adaptive. While few of us may have the available time (or resources) to completely adopt Dr. Fordyce’s suggested intervention strategy verbatim to correct lingering pain behaviors that are no longer adaptive, the principles underpinning those strategies remain as true today as they were in 1976.


So, what are the implications for the field if one of the major conclusions in revisiting Chapter 11 is that Fordyce had a lot of this right? For those that have an interest in future directions for research and practice, there are some “devils in the details” that are worth identifying as we work towards improving the effectiveness of treating pain with behavioral approaches. In our opinion, a high priority should be placed on a better understanding of how motor adaptation occurs in response to pain. In particular, identification of malleable factors that influence persistence of strategies that were originally adaptive, but no longer serve that purpose are of how interest to the field. Such information is important because motor function and adaptation are key intermediaries between the perceptual and behavioral aspects of pain, yet are often understudied.

Pain Influences Motor Activity

Fordyce’s observation that heterogeneity in the pain experience (including behavior) is expected has been repeatedly supported in empirical investigations. To date, the research on heterogeneity in pain has expanded on Fordyce’s original observations by providing specific examples of the variability in the perceptual and sensory aspects of pain. For example, pain expression can be influenced by variability in aspects of personality and other psychosocial influences. Another example of how pain heterogeneity has been documented is through studies that use quantitative sensory testing (QST). QST is the application of standard nociceptive stimuli to individuals which enables a differentiation between characteristics of the individual and features of the stimulus within the overall pain experience. In this way, QST allows the practitioner or researcher to control the amount of stimulation, a situation that is not possible in most clinical settings. Wide variation in the sensory and perceptual experience of pain has been demonstrated in human human subjects using QST, even when receiving the same nociceptive stimulus.

In Chapter 11, Fordyce described an extreme in variation of motor responses to low back pain. In other words, it is not common to have limping, falling, and knee buckling, but it is within the spectrum of expected responses. Earlier in this commentary (page 352) some other, more subtle, motor responses to pain were described. It is likely that Fordyce used an extreme motor response because it directly challenged the existing paradigm for how pain should affect gait. The example he provided directly questioned the validity of peripheral explanations for the limping, falling, and knee buckling that in turn led to a distorted gait pattern. This example also allowed for a meaningful demonstration of how chronic pain could be effectively managed when the existing paradigm was shed and the patient was viewed through a different conceptual lens that operated independently of knowing more about the peripheral cause of pain. As it turns out, the extreme example that Fordyce selected was robust because we see notable behavior change even for subtle motor responses, and the variation in these responses is associated with patient beliefs about pain (see earlier examples on page 352).

Fordyce’s original observations on motor and behavioral responses have stood the test of time, but there are still many unanswered questions related to heterogeneity of motor expression in response to pain. The link between variability in motor adaptation and behavior in response to pain is an area that is ripe for further development and study because formative theories in this area have not adequately accounted for the variability in motor function following pain. For example, the “pain adaptation” theory suggests that motor function is uniformly inhibited in response to pain. However, in the literature there are examples of both muscle inhibition and activation occurring when painful stimuli are applied [2]. The previously mentioned model from Hodges and Tucker [2] provides a current perspective on motor adaptation for short and long term consequences of pain. This theoretical model serves as a great starting point for those interested in making progress in this area. In our opinion, future research priorities should include: 1) Studies that capture both short and long term responses to pain, as short term responses are more frequently reported; 2) Inclusion of patient populations, as induced pain in healthy subjects is commonly used as the experimental model; and 3) Making a more direct link to the clinical relevance of the findings by incorporating commonly used outcome measures in addition to the measures of motor activity. Further development and study that follows these priorities should ensure that future studies provide a better understanding of how variability in motor expression is integral in the development and maintenance of chronic pain.

Motor Activity Influences Pain

In general, we know that exercise and/or resumption of normal activities when still in pain is more beneficial than waiting until pain free before initiating movement. Fordyce described quota based graded activity as one specific approach to prescribing movement while still in pain and its effectiveness has been explored in clinical trials. In a systematic review and meta-analysis, the pooled effects of graded activity showed a benefit over minimal interventions but not when compared to other exercise approaches [4]. Therefore, there is an indication that not all patients with chronic pain receive an added benefit from graded activity approaches when compared to general exercise. There are a number of possible reasons for the somewhat equivocal findings in the systematic review. One reason of relevance to this commentary is the need for improved knowledge about the processes involved with how movement influences pain perception, especially when movement is applied with therapeutic goals in mind. Increased understanding in this area will help us come “full circle” on Fordyce’s original observations as we will be able to discriminate when graded activity is appropriate for a patient, and when another approach may be a better choice.

In particular we need to know more about when motor activity increases the perception of pain. This is a challenging dilemma for practitioners and scientists because evidence consistently suggests that movement and activity should be beneficial, but patients often report increased pain with activity and exercise. It is even more perplexing when the patient modifies behavior appropriately but the pain continues to worsen. Observations like these have led to the investigation of whether there are subgroups of individuals in which movement can increase the sensory aspect of pain. For example, Michael Sullivan and his group have developed a repetition induced summation of activity-related pain (RISP) protocol to determine how standardized movement impacts reports of pain [7,8]. These interesting studies show clear proof of principle that activity increases the sensory perception of pain, and that non-behavioral treatment, such as Transcutaneous Electrical Nerve Stimulation (TENS) reduces the summative aspect related to movement [5]. Behavioral analysis of individuals who show a summation of pain with movement might provide additional insight on how motor activity influences pain perception. This approach would offer another avenue of exploration for Fordyce’s observations so that we can better tailor movement based treatments for those with chronic pain.


In summary, Fordyce’s Chapter 11 synthesized the sensory and perceptual aspects of pain and identified appropriate direction for future investigation of the motor and behavioral consequences of pain. At the time this was a much needed prompt for the field that has inspired many important discoveries and advancements. However, there are still many nuances related to his general observations that need to be teased out so that in the future we will better serve patients with chronic pain by knowing which motor activity or behavioral treatment paradigm is the best fit for their particular presentation of pain.


  1.Al-Obaidi SM, Al-Zoabi B, Al-Shuwaie N, Al-Zoabi N, Nelson RM. The influence of pain and pain-related fear and disability beliefs on walking velocity in chronic low back pain. Int J Rehabil Res 2003;26:101–8.

  2.Hodges PW, Tucker K. Moving differently in pain: a new theory to explain the adaptation to pain. Pain 2011;152(Suppl):S90–8.

  3.Lamoth CJC, Daffertshofer A, Meijer OG, Beek PJ. How do persons with chronic low back pain speed up and slow down? Trunk–pelvis coordination and lumbar erector spinae activity during gait. Gait Posture 2006;23:230–9.

  4.Macedo LG, Smeets RJ, Maher CG, Latimer J, McAuley JH. Graded activity and graded exposure for persistent nonspecific low back pain: a systematic review. Phys Ther. 2010 Jun;90:860–79.

  5.Mankovsky-Arnold T, Wideman TH, Larivière C, Sullivan MJ. TENS attenuates repetition-induced summation of activity-related pain following experimentally induced muscle soreness. J Pain 2013;14:1416–24.

  6.Seay JF, Van Emmerik RE, Hamill J. Low back pain status affects pelvis–trunk coordination and variability during walking and running. Clin Biomech 2011;26:572–8.

  7.Sullivan MJ, Thibault P, Andrikonyte J, Butler H, Catchlove R, Larivière C. Psychological influences on repetition-induced summation of activity-related pain in patients with chronic low back pain. Pain 2009;141:70–8.

  8.Sullivan MJ, Larivière C, Simmonds M. Activity-related summation of pain and functional disability in patients with whiplash injuries. Pain 2010;151:440–6.

  9.te Wierike SC, van der Sluis A, van den Akker-Scheek I, Elferink-Gemser MT, Visscher C. Psychosocial factors influencing the recovery of athletes with anterior cruciate ligament injury: a systematic review. Scand J Med Sci Sports 2013;23:527–40.

10.Thomas JS, France CR, Lavender SA, Johnson MR. Effects of fear of movement on spine velocity and acceleration after recovery from low back pain. Spine 2008;33:564–70.

11.Trost Z, France CR, Thomas JS. Pain-related fear and avoidance of physical exertion following delayed-onset muscle soreness. Pain 2011;152:1540–7.

12.Trost Z, France CR, Sullivan MJ, Thomas JS. Pain-related fear predicts reduced spinal motion following experimental back injury. Pain 2012;153:1015–21.

Commentary: From Gait Disturbance to Cortical Reorganization and Back Again

G. Lorimer Moseley

I am one of those physical therapists who have taken the road less travelled, from clinical ‘musculoskeletal’ physical therapy, to chronic pain rehabilitation and thence to researching chronic pain and its treatment. I suspect that the majority of us who have taken this road have done so because our fascination, and indeed frustration, with persistent pain and movement problems, compelled us to do so. It was from this background that I re-visited the chapter on gait disturbance, and indeed much of this very important book. My current vantage point is one of a clinical neuroscientist with an intense interest in persistent pathological pain disorders and the role of the brain and mind in their development, maintenance, and treatment. I confess to having been slightly daunted on being asked to provide a commentary on a chapter of such an important book, written by such an important clinical scientist. I have, as it turns out, constructed my commentary around the wry smiles and ‘aha!’ moments I had on reading it again. Wry smiles when I read a passage that clearly shows Fordyce was beyond his time – I can hear the self-conscious whispering in my mind telling me that some of our recent ‘discoveries’ may have conjured less self-congratulatory back slapping if we had re-read this chapter before we did the experiment. ‘Aha!’ moments when I realize now what I did not realize on reading this work many years ago the first time around—before I had the clinical experience and maturity to really understand what Fordyce was saying.

I have another confession—that although I have an honors degree in three dimensional kinematics of the rear foot during walking [27], and I spent the first years of my clinical life carefully observing the way athletes—from Olympians to Weekend Warriors—walked, ran, paddled, and threw, I have not developed any real expertise in gait and I am relieved that others commentating on this chapter are superbly placed to compensate for this substantial shortcoming. I will draw instead on both my own work, and that of others, that investigates other aspects of movement dysfunction that is associated with chronic pain.

Early on in this chapter, Fordyce states

‘It may (also) be that the distorted gait has persisted so long that a substantial skill deficit for normal gait has developed. The patient may no longer know how to walk properly.’

Whether the clinical insights of greats such as Fordyce triggered it or not, this kind of observation has been extensively explored over the last 10–15 years, ultimately leading to the now accepted notion that the motor and sensory cortices are organized differently in people with chronic pain than they are in people without chronic pain [43]. Intriguingly, that ‘final’ picture is perhaps not as clear as we thought only months ago, but along the way our industry has been driven by the common clinical observation that people with chronic pain really do seem to have a substantial skill deficit—they no longer know how to do some things properly.

To unpack this significant issue, it is necessary to first, establish what it is that we know about this movement disturbance, and second, describe how we think movement itself is generated by the brain and how this might be open to the sort of effects that Fordyce observed with gait.


The idea that ‘neurons that fire together, wire together’, still courses the veins of any self-respecting brain scientist. It has clearly also influenced the way we interpret some of the exciting developments in the study of the brain in pain. The earliest work in this area focused on primary sensory cortex (S1) and it is useful to briefly review that work because it is an important background to work on the motor cortex (M1), which is clearly critical here. The key finding regarding S1 was that the response profile of primary sensory cortex (S1) changes when monkeys have a finger amputated—the area of S1 that responds to the non-amputated fingers expands to ‘invade’ the area formerly dedicated to the now-amputated finger [32]. That work was extended to humans who had suffered an amputation and, on the whole, was corroborated (see [9] for review). The significant step in this line of research, from a clinical pain science perspective, involved comparing amputees with phantom limb pain to amputees without phantom limb pain. The results were striking—those with phantom limb pain showed the same ‘invasion’ of S1 representation of the intact areas (in this case of the lip in people who had an amputated arm) into the area formerly dedicated to the now missing limb. This abnormality was not present in amputees without phantom limb pain [8]. A substantial body of research, undertaken by several different research teams, led to the development of the ‘maladaptive neuroplasticity’ theory of phantom limb pain [9].

The next step was to investigate this cortical reorganization in other types of pain and the most obvious candidates were complex regional pain syndrome (CRPS), which shares several clinical characteristics with phantom limb pain [1], and low back pain, which is the most common chronic pain complaint and indeed, the world’s most burdensome healthy issue [40]. The CRPS research paints a very interesting picture. A systematic and meta-analytical review revealed compelling evidence that the hand representation in S1 is larger for the healthy hand than it is for the painful CRPS hand [7]. The assumption then was according to a ‘use it or lose it’ position—that the side-to-side imbalance reflected shrinkage of the affected S1. There was, however, poor evidence to support this assumption and the most recent, and rigorous, work suggests the opposite may be true—that the imbalance reflects a larger representation on the healthy side (Di Pietro et al., under review).

Let’s then extend this to investigations of the motor cortex. Here our understanding relies almost completely on an experimental approach in which we experimentally stimulate M1, using electrodes implanted in the cortex or magnetic stimulators fired just over the skull, and record the muscle twitch evoked in response. The most common approach in humans is to use transcranial magnetic stimulation (TMS), and we develop a ‘motor map’ of the muscles that move, for example, the hand. This technique is blunt and the leap from looking at single muscle twitches to the function-based organization of the motor cortices (see below), leaves a cloud over how we interpret the results. Nonetheless, it provides a useful window and it is very widely accepted as the best we can do with current techniques.

Systematic and meta-analytical review of the TMS literature shows clearly that M1 is disinhibited in people with chronic pain but that this disinhibition is not confined to the M1 representing the painful side of the body [6]. For studies in low back pain, the general pattern is that motor maps shift and become less precise [43]. This finding needs to be considered together with studies of S1 because: (i) there are extensive intra- and inter-hemispheric connections between primary motor and sensory cortex, (ii) on-line feedback from sensory systems is considered critical for motor control and motor learning (see early writings on this idea [35,39]), and (iii) there is compelling evidence that both motor and sensory maps are less precise in people with chronic pain.

The critical limitation of our knowledge here that is a kind of ‘inconvenient truth’—is that we do not know what is the chicken and what is the egg. Does chronic pain change the brain, as has been suggested in several loosely interpreted studies, or does cortical reorganization underpin chronicity (as has been suggested in other loosely interpreted studies)—see [23] for review. A final possibility is that the two phenomena are causally unrelated. Although the wider body of literature comfortingly tends to disconfirm this possibility, I would suggest that we cannot rule it out just yet. We can, however, conclude that the sensory and motor cortices of people in chronic pain are different from those of people without chronic pain. The common term for this difference is ‘cortical reorganization’ but, as stated, there is no clear evidence that the brain is reorganized so much as organized differently.


The investigation of movement has been integral to the investigation of the brain itself, for hundreds of years. As a very brief overview—the field seemed to have been kicked off by Swedenborg’s 18th century treatise on the brain [11]. Jackson’s somewhat ambiguous but nonetheless brilliant investigations of seizures [13] built on earlier work and laid the platform for Fritsch and Hitzig’s to describe a motor cortex organized according to movements rather than muscles [10]. Sherrington and others described several motor cortices—including the pre motor and supplementary motor areas [34] and Penfield’s laborious descriptions of over 150 brain surgery patients remain our most comprehensive account of human motor cortex organization [29,30,31].

The upshot of all this work is that we now know that several brain areas work together to provide normal motor function, and that non-reflexive movements are generated by networks of brain cells. These networks are not fixed—there is not a hip flexion1 cell or a rear foot supination2 cell. Rather, the anatomical substrates that subserve movements are dynamic and are highly open to modulation. Primary motor cortex (M1) is functionally organized and this provides the direct link to Fordyce’s observation: if there is a functional advantage in learning a new way of doing things, then the chances are the brain will work it out. Moreover, the effects of this learning should be detectable in the response profile of M1. The implication is clear, and clearly foretold by Fordyce: The patient may no longer know how to walk properly, or throw properly, or write, wink, stand, or sit properly.


I contend that the broad picture of the body of research in this field is that movement disorders in chronic pain are consistent with functional advantage insofar as they function to protect the painful part. Gait disturbance in its most obvious and observable way, with the characteristics astutely documented by Fordyce, clearly serves this purpose. Indeed, such gait disturbance promotes the avoidance of pain in an indirect way—by promoting reinforcement and ‘time out’– identified by Fordyce as characteristics of “avoidance learning.” However, gait disturbance, and other movement disturbance, extends beyond this interpretation by not necessarily leading to an observable operant behavior. For example, Claudine Lamoth has shown that people with chronic back pain walk with an idiosyncratic abnormality in the way their trunk rotates, but this abnormality is not necessarily observable, nor is it necessarily associated with functional limitation—walking velocity and cadence might be normal [14,15]. Importantly, this idiosyncratic movement pattern can be induced in healthy volunteers by injecting (sterile) salty water into their back muscles and the movement abnormality resolves once the experimentally induced pain does [16]. There are in fact a large number of studies that have looked at muscle activity during walking in people with clinical and experimentally induced back pain. That research is not directly in my area of expertise, so it is best covered more eloquently in the previous commentary. I will instead focus on a related area of research in which the skill loss referred to by Fordyce, is untangled from the pain, such that persistent movement abnormality exists, without persistent behavioral change and without persistent pain.


A vast proportion of work in this area comes from the research group led by the neurophysiologist Paul Hodges. His work investigates the way in which the brain prepares the body for perturbations to postural stability, most often perturbations imparted by movement. When we move a limb, for example, the brain must first stabilize the body so as to eliminate the effect of the limb movement on our postural stability. By inserting electrodes to a gaggle of muscles around the low back and pelvis, Hodges has shown (see [33] for an overview of his early and influential work) that, in healthy volunteers with no history of back pain, the trunk muscles are recruited in a very predictable, movement-specific and, with regards to the biomechanics of movement, a very sensible way. However, people with recurrent back pain show a different pattern, a pattern that is variable between people but which invariably serves to limit movement of the back. To iterate, even though these people are pain-free at the time of testing, and even though they move their limbs in a manner that looks identical to the movements undertaken by healthy controls (the limb movements are in fact almost identical), their brain activates their trunk muscles differently in a manner that is consistent with protection of the back.

In my own doctoral work, under the watchful eye of Professor Hodges, we showed that one could induce these protective strategies by giving healthy pain-free people a painful injection of hypertonic saline [12], or by inducing the expectation that their back is about to hurt [25]. Moreover, this shift in postural adjustment strategy—from normal to protective—does not return to normal in particular people who are characterized by a heightened sense of vulnerability of their back [24]. This type of finding led to a shift in the way we make sense of movement disorders in people with chronic pain—from a purely behavioral effect as observed by Fordyce, to the result of an evaluative process whereby the motor command matches the functional demand and the perceived capacity to meet it. I have written extensively on this topic elsewhere [22], but suffice here to say that we can now expand on Fordyce’s proposal that the patient no longer knows how to walk normally, to say that even when they are pain-free and appear to be walking normally, they may be so doing in an overly protective way.

The broad picture of movement disorders in chronic pain bears a clear semblance to the broad picture of pain itself. That is, I contend that there is a vast literature that supports the idea that pain reflects an implicit conviction that body tissue is in danger and action is required. This idea is not a new one—see Patrick Wall’s writings on this more than 25 years ago [41]—but it is an idea that is by no means universally endorsed. Rather than outline why this understanding of pain is consistent with both our clinical observations and the available evidence (which is in fact vast)—this has been done elsewhere [3]—I humbly request the reader to accept the idea for the moment, and then take the relatively tiny step to see its correlate in the movement disorder literature. Insofar as anything that modulates the brain’s conviction that body tissue is in danger and action is required can modulate pain, so too, one might predict that anything that modulates the brains conviction that body tissue is in danger and action is required could modulate motor output [3]. From here then, a multitude of physiological mechanisms, broadly encapsulated by the concept of learning, will give rise to protective motor patterns that persist beyond the need for protection. Unlike the data on M1 and S1 organizational differences in people with chronic pain, there is compelling evidence that motor abnormalities occur in line with chronic pain. There is a casual argument that they do not result from the pain, although both pain and motor abnormality can be conceptualized as resulting from persistent perceived threat to body tissue. That is, pain and motor abnormality can be considered the results, and persistent perceived threat the cause. It is useful here to emphasize that the threat may not actually exist, but the evaluative mechanisms of the brain have erroneously decided that it does. This situation might be easily evoked experimentally by tricking participants into thinking that they will receive a noxious stimulus when indeed they won’t, or clinically by unintentionally giving the naïve patient an inaccurate understanding that their back is fragile, when in fact it is not.

Herein lies an important question–Is this persistent protective motor disturbance a problem, and does it in some way cause persistent pain? Those in the affirmative camp argue that although this protective strategy is clearly associated with gain—it limits movement of the endangered body part—it is also associated with a cost—potentially dangerous loads on tissues [33]. My view is that the evidence is not compelling, but it is certainly intuitively sensible–if overprotection was not disadvantageous in any way one would presume that we would do it all the time—extra protection with no strings attached. I suspect that our intuition will eventually prove to be correct, and physical therapists have led the charge in adopting the very same approach advocated by Fordyce – “proceed as if a skill deficit exists, regardless of whether in fact it does so.” The approach to this over-protective motor disturbance in back pain, is to proceed to correct the disturbance by teaching people precise contractions of particular hard-to-contract muscles [33], an approach that has conceptualized the disturbance as evidence of end-organ dysfunction and the effect of training as end-organ correction [33]. I find the theoretical argument behind this conceptualization implausible and find Fordyce himself offering a more likely explanation—that of learning the movement disturbance over time, and that of the clinician correcting it through graded exposure and conditioning – helping the patient to “unlearn” it if you will.


In this chapter on gait, Fordyce’s recommendations for correcting motor disturbance are still largely upheld by the vast amount of research undertaken in this area since the work was first penned. To illustrate this I will focus on my own work with people with complex regional pain syndrome (CRPS). CRPS is a very debilitating, very painful disorder that occurs in about 3% of people after minor injury to one limb. From an academic perspective, it is an attractive pain disorder because, more so than any other pathological pain, all that can go wrong in chronic pain, does go wrong and does so very quickly. Diagnosis depends on (i) severe pain, which is not confined to a particular structure, but often extends to the entire limb, and (ii) multiple system dysfunction—motor disturbance, endocrine and sympathetic disturbances. The pathophysiology of CRPS is not well understood, but inflammatory, autonomic and central nervous system contributors have all been identified [17]. There is compelling evidence that the acute stage of the disease is dominated by aberrant inflammation and that the majority of cases resolve inside three months. However, those that persist become dominated by CNS dysfunction and anti-inflammatory treatments are useless. It is this group—those with persistent CRPS–who impart the bulk of the economic and social load of CRPS.

People with persistent CRPS have well documented movement disturbances. In fact, a small number develop such severe dystonia that the limb is rendered completely useless and contractures are almost inevitable [37,38]. The nature of dystonia is consistent with “hyper-protection” [28], a situation that would be predicted on the basis of the arguments I have made thus far. Most persistent CRPS patients do not develop this severe dystonia, but are still characterized by motor disturbance that is analogous to Fordyce’s gait disturbance observed the more generic (i.e., non-CRPS) pain population.


Currently the most effective treatment approach to persistent CRPS is graded motor imagery [18], although a cautionary note is important here –the stand-out evidence for graded motor imagery reflects both the encouraging results of three small randomized controlled trials [2,19,21] and a large number of controlled case series, but also the disheartening lack of evidence for anything else. Graded motor imagery was devised on the back of the observation that even imagined movements can cause pain and swelling in people with persistent CRPS [20,26]. From a neurophysiological perspective, the most likely explanation for this observation is that activation of the motor command to perform a movement was enough to trigger the protective output of pain and, presumably, the disturbed motor command. This interpretation is consistent with the data on altered M1 organization outlined above and has some empirical support from imaging studies [42]. The functional response to this dire situation would most predictably be to avoid even thinking about movement and, perhaps not surprisingly, CRPS is characterized by a suite of neglect-like disturbances.

Graded motor imagery attempts to get around this situation by training implicit motor imagery. Implicit motor imagery involves tasks that require the planning and preparation of movement without evoking the command. The method of choice for graded motor imagery is to use left/right judgments of pictured body parts. There is a very well established neurophysiology of left/right judgments (see [18] for comprehensive review), and can be summarized thus: when we see a body part that we must judge as left or right, we make an initial ‘automatic’ decision, which we then confirm by preparing to move our own corresponding limb into the same position as that shown in the image. Such a task activates motor networks in the brain without activating M1, which you will remember is organized according to functional tasks. The theory upon which graded motor imagery rests is that repeated training of this implicit motor imagery serves to reinstate normal inhibitory control in M1 and thus remove the skill deficit. Behavioral data appear supportive—performance improves and imagined movements no longer increase pain and swelling. Graded motor imagery progresses from left/right judgments to imagined movements, to mirror therapy and then to functional exposure. There is good evidence that the order of the first two components is important although the importance of mirror therapy as part of graded motor imagery remains to be empirically demonstrated.

It would not be accurate to claim that graded motor imagery was based on Fordyce’s book, but my wry smile certainly emerged on re-reading his account of the essential elements of normalizing the skill deficit of disturbed gait. Here I paraphrase Fordyce’s list from Page 343 (above), to apply it to our own implementation of graded motor imagery for CRPS of the foot3:

   1.  Distorted gait is not allowed additional rehearsal and reinforcement. This is a fundamental principle of graded motor imagery. After comprehensive and targeted education about the rationale behind the approach, the need for patience and persistence and ‘training the brain’ (see [18] for review), functional exposure is wound right back and treatment begins with multiple daily sessions of implicit motor imagery.

   2.  Gait is broken down into its constituent parts. This principle is applied at least twice in graded motor imagery. First, we have conceptualized graded motor imagery as “biologically based graded exposure,” whereby the cortical processes involved in gait are broken down into the movement preparation and planning stage and this stage is practiced extensively. Second, the nature of the images used in graded motor imagery follow the same process of beginning with feet shown in trivial and non-threatening postures and tasks, and leading to progressively more threatening postures and tasks, for example running.

   3.  Each stage is practiced to mastery. Perhaps the most challenging aspect of graded motor imagery is the amount of practice and perseverance that is required on the part of the patient. The criterion for progression of graded motor imagery is indeed mastery of the current stage. We evaluate this with quantifiable parameters—accuracy and reaction time.

   4.  Development of new (or reinstated) skills is generalized by broadening demand and context. Graded motor imagery has generalization and progression of motor skills at its heart, both within the successive stages and as it is progressed to functional exposure.

   5.  Significant others are included in the training process so as to ensure appropriate social and activity-based reinforcements. This principle is integral throughout graded motor imagery, and utilized with performance-based reinforcers (inbuilt to software and via clinician and significant others).

Insofar as graded motor imagery is progressed on time-contingent and complexity-contingent grounds, not on pain-contingent grounds, it is very similar to Fordyce’s very prescriptive account presented on the subsequent pages. The relevance of Fordyce’s work to graded motor imagery has actually become clearer as the body of literature and web and social media traffic concerning graded motor imagery builds. It is clear that some clinicians are experiencing exhilarating results with formerly untreatable patients and others are experiencing nothing but perspiration, exasperation and frustration at not being able to make the same sort of gains that are evident in the clinical trials. My own audit data, of over 1000 patients treated at nine different clinical centers, clearly support this pattern—clinician-specific recovery rate at six months after presentation vary from less than 10% to over 60%. Considering that spontaneous recovery would nudge 5 % to 6% in this population, these are both exciting and sobering data. We have theories as to why such variability exists [18], but one possibility that has hitherto been only partially considered relates to Fordyce’s work—perhaps the key is actually in the behavioral therapy we are implementing using the vehicle of graded motor imagery. Perhaps some clinicians are excellent behavioral therapists and others are not. Perhaps some are implementing graded motor imagery according to an end-organ conceptualization such as that mentioned above. These questions remain to be answered but they certainly seem worthy of investigation.

The argument for considering the role of behavioral therapy in graded motor imagery is strengthened by preliminary results of a purely behavioral approach to CRPS. Two approaches are most pertinent here—fear exposure therapy [4,5] and pain exposure therapy [36]. In fear exposure therapy, behavioral therapy principles are used to facilitate patients in undertaking the very tasks of which they are most fearful [4,5]. In pain exposure therapy, the same principles are used to facilitate patients in gaining function regardless of pain. These approaches seem to me to be ‘more Fordyce-based’ than graded motor imagery, although they do not seem to involve removing the distorted movement, listed by Fordyce as the first essential ingredient of training. Aside from that, they seem to consider pathophysiology as unimportant in the treatment of the disorder (although not necessarily unimportant in the disorder itself). This is not surprising as they, unlike graded motor imagery, are born directly out of the behavioral therapy tradition. I suspect that their authors had fewer wry smiles, and more ‘“I told you so’s,” on re-reading Fordyce’s book than I did.


The more I read Fordyce’s writings, the more I find myself asking “What would Bill do?” The resonance between his observations and approach to treatment of pain-related movement disturbances on the one hand, and our increasing understanding of the “brain in pain” (S1 and M1 in particular), on the other, is quite remarkable. We now have a biologically plausible mechanism by which patients may “no longer know how to move (walk) properly” a mechanism that is at once consistent with offering a functional advantage, and open to correction by targeted training, precisely metered reinforcement, and progression that is not contingent on pain. Whether the treatment paradigm be focused solely on Fordyce’s principles, or incorporated into a wider biologically-based approach, is perhaps less important than the principles themselves. In this sense, the insightful and important work of Fordyce continues to have a profound influence on what we do and how we do it.


  1.Acerra NE, Souvlis T, Moseley GL. Stroke, complex regional pain syndrome and phantom limb pain: can commonalities direct future management? J Rehabil Med 2007;39:109–14.

  2.Bowering KJ, O’Connell NE, Tabor A, Catley MJ, Leake HB, Moseley GL, Stanton TR. The Effects of Graded Motor Imagery and Its Components on Chronic Pain: A Systematic Review and Meta-Analysis. J Pain 2012;14:3–13.

  3.Butler D, Moseley GL. Explain Pain. Adelaide: Noigroup publications, 2013.

  4.de Jong JR, Vlaeyen JW, de Gelder JM, Patijn J. Pain-related fear, perceived harmfulness of activities, and functional limitations in complex regional pain syndrome type I. J Pain 2011;12:1209–18.

  5.de Jong JR, Vlaeyen JWS, Onghena P, Cuypers C, Hollander Md, Ruijgrok J. Reduction of pain-related fear in complex regional pain syndrome type I: The application of graded exposure in vivo. Pain 2005;116:264–75.

  6.Di Pietro F, McAuley JH, Parkitny L, Lotze M, Wand BM, Moseley GL, Stanton TR. Primary motor cortex function in complex regional pain syndrome: a systematic review and meta-analysis. J Pain 2013;14:1270–88.

  7.Di Pietro F, McAuley JH, Parkitny L, Lotze M, Wand BM, Moseley GL, Stanton TR. Primary Somatosensory Cortex Function in Complex Regional Pain Syndrome: A Systematic Review and Meta-Analysis. J Pain 2013;14:1001–18.

  8.Flor H, Elbert T, Knecht S, Wienbruch C, Pantev C, Birbaumer N, Larbig W, Taub E. Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation. Nature 1995;375:482–4.

  9.Flor H, Nikolajsen L, Jensen TS. Phantom limb pain: a case of maladaptive CNS plasticity? Nat Rev Neurosci 2006;7:873–881.

10.Fritsch G, Hitzig E. Electric excitability of the cerebrum (Uber die elektrische Erregbarkeit des Grosshirns). Epilepsy Behav 2009;15:123–30.

11.Gross CG. Emanual Swedenborg: A neuroscientist before his time. Neuroscientist 1997;3:142–147.

12.Hodges PW, Moseley GL, Gabrielsson A, Gandevia SC. Experimental muscle pain changes feedforward postural responses of the trunk muscles. Exp Brain Res 2003;151:262–71.

13.Jackson JH. A study of convulsions. Trans St Andrews Med Grad Assoc 1870;3:1–45.

14.Lamoth CJ, Meijer OG, Daffertshofer A, Wuisman PI, Beek PJ. Effects of chronic low back pain on trunk coordination and back muscle activity during walking: changes in motor control. Eur Spine J 2005 15:23–40.

15.Lamoth C, Meijer O, Wuisman P, van Dieen J, Levin M, Beek P. Pelvis-thorax coordination in the transverse plane during walking in non-specific low back pain. Spine 2002;27:E92–99.

16.Lamoth CJC, Daffertshofer A, Meijer OG, Moseley GL, Wuisman P, Beek PJ. Effects of experimentally induced pain and fear of pain on trunk coordination and back muscle activity during walking. Clin Biomech 2004;19:551–63.

17.Marinus J, Moseley GL, Birklein F, Baron R, Maihofner C, Kingery WS, van Hilten JJ. Clinical features and pathophysiology of complex regional pain syndrome. Lancet Neurol 2011;10:637–48.

18.Moseley G, Butler D, Beames T, Giles T. The graded motor imagery handbook. Adelaide: Noigroup publishing; 2012.

19.Moseley GL. Graded motor imagery is effective for long-standing complex regional pain syndrome: a randomised controlled trial. Pain 2004;108:192–198.

20.Moseley GL. Imagined movements cause pain and swelling in a patient with complex regional pain syndrome. Neurology 2004;62:1644.

21.Moseley GL. Graded motor imagery for pathologic pain—A randomized controlled trial. Neurology 2006;67:2129–34.

22.Moseley GL. Trunk muscle control and back pain: chicken, egg, neither or both? In: Hodges PW, Cholewicki J, van Dieën JH, editors. Spinal control: the rehabilitation of back pain: state of the art and science. Edinburgh: Elsevier; 2013.

23.Moseley GL, Gallace A, Spence C. Bodily illusions in health and disease: physiological and clinical perspectives and the concept of a cortical ‘body matrix’. Neurosci Biobehav Rev 2012;36:34–46.

24.Moseley GL, Hodges PW. Reduced variability of postural strategy prevents normalization of motor changes induced by back pain: A risk factor for chronic trouble? Behav Neurosci 2006;120:474–6.

25.Moseley GL, Nicholas MK, Hodges PW. Does anticipation of back pain predispose to back trouble? Brain 2004;127:2339–47.

26.Moseley GL, Zalucki N, Birklein F, Marinus J, van Hilten JJ, 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;59:623–31.

27.Moseley L, Smith R, Gant R, Hunt A. Three dimensional kinematics of the rearfoot during stance phase of walking in yound adult males. Clin Biomech 1996;56:34–78.

28.Munts AG, Mugge W, Meurs TS, Schouten AC, Marinus J, Moseley GL, van der Helm FCT, van Hilten JJ. Fixed Dystonia in Complex Regional Pain Syndrome: a Descriptive and Computational Modeling Approach. BMC Neurol 2011;11(53). doi:10.1186/1471-2377-11-53

29.Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man studied by electrical stimulation. Brain 1937;60:389–443.

30.Penfield W, Jasper H. Epilepsy and the functional anatomy of the human brain. Boston: Little & Brown; 1954.

31.Penfield W, Rasmussen T. The cerebral cortex of man: a clinical study of localization of function. New York: Macmillan; 1950.

32.Pons TP, Garraghty PE, Ommaya AK, Kaas JH, Taub E, Mishkin M. Massive cortical reorganization after sensory deafferentation in adult macaques. Science 1991;252:1857–60.

33.Richardson C, Jull G, Hodges P, Hides J. Therapeutic exercise for spinal segmental stabilization in low back pain. London: Churchill Livingstone; 1999.

34.Sherrington C. The integrative action of the nervous system. New Haven: Yale University Press; 1906.

35.Sperry R. Neural basis of the spontaneous optokinetic responses produced by visual neural inversion. J Comp Physiol Psychol 1950;43:482–9.

36.van de Meent H, Oerlemans M, Bruggeman A, Klomp F, van Dongen R, Oostendorp R, Frolke JP. Safety of “pain exposure” physical therapy in patients with complex regional pain syndrome type 1. Pain 2011;152:1431–8.

37.van Hilten JJ, Beek WJ, Vein AA, van Dijk JG, Middelkoop HA. Clinical aspects of multifocal or generalized tonic dystonia in reflex sympathetic dystrophy. Neurology 2001;56:1762–5.

38.van Rijn MA, Marinus J, Putter H, van Hilten JJ. Onset and progression of dystonia in complex regional pain syndrome. Pain 2007;130:287–293.

39.Von Holst H. Relations between the central nervous system and the peripheral organs. Br J Anim Behav 1950;2:89–94.

40.Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, Shibuya K, Salomon JA, Abdalla S, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2163–96.

41.Wall PD. On the relation of injury to pain. The John J. Bonica lecture. Pain 1979;6:253–264.

42.Walz AD, Usichenko T, Moseley GL, Lotze M. Graded motor imagery and the impact on pain processing in a case of CRPS. Clin J Pain 2013;29:276–279.

43.Wand BM, Parkitny L, O’Connell NE, Luomajoki H, McAuley JH, Thacker M, Moseley GL. Cortical changes in chronic low back pain: current state of the art and implications for clinical practice. Man Ther 2011;16:15–20.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 9, 2018 | Posted by in Uncategorized | Comments Off on Distorted Gait—A Problem of Shaping

Full access? Get Clinical Tree

Get Clinical Tree app for offline access