Manual Therapy Journal - Volume 8, Issue 3, Pages 129-192 (August 2003)

Manual Therapy (2003) 8(3), 129 r 2003 Elsevier Science Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:...

Author: Editors: Ann Moore and Gwen Jull

23 downloads 620 Views 903KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Manual Therapy (2003) 8(3), 129 r 2003 Elsevier Science Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00039-0

Editorial

Masterclasses Reconsidered and Revisited A consistent feature of Manual Therapy Journal from its first issue in 1995 has been the Masterclass section. Each issue has contained a commissioned Masterclass paper enabling an expert clinician to discuss and explore a particular system of treatment and or examination, which they have developed or they have used frequently in their own practice setting. From the outset Manual Therapy Journal has wished to maintain a strong clinical relevance for manual therapists. The Masterclass adds to this dimension complementing original research work, technical notes, professional issues and case reports which can all contribute to clinical practice in various ways. The Masterclass section aims to reflect current clinical practice within the scheme of evidence-based/ evidence-informed practice. Some Masterclasses are entirely research based whilst others are partially unsupported by current hard evidence but they are strongly supported by expert opinion. Ideally, all the Masterclass manuscripts published in the journal should be entirely based on evidence. This is not always possible at the present stage of development in musculoskeletal research and its knowledge base. While expert opinion or expert consensus is given low weighting on scales of evidence, an important function that these mauscripts serve is in raising relevant research questions and possible directions for future research. The response to the call for evidence-based practices by musculoskeletal researchers and therapists has been tremendous in the last 5– 10 years (see Pedro data base). The Editors expect to see an exponential increase in the number of Masterclass manuscripts, which are entirely research based in the next 5 years. This year sees the publication by Elsevier Science Ltd of two new textbooks, which include a selection

of Masterclass manuscripts published in Manual Therapy Journal over the last 8 years. These texts are edited by Karen Beeton, the editor of the Masterclass section of Manual Therapy Journal. There are two volumes a peripheral and a spinal volume and these publications have given the original authors the opportunity to update their previously published articles in particular in relation to any new research findings that may support or refute them. We hope that readers of Manual Therapy Journal will find these new volumes valuable and would like to congratulate Karen Beeton on putting these volumes together. As editors we are always interested in any comments that you have about any aspect of the Journal. Karen Beeton is especially keen to receive comments on the Masterclass section. Please let us know if you have any comments, we will as a matter of course pass your comments onto the authors of the work subject to comment who then will be able to respond to your comments publicly extending the academic debate. On reflecting on the 2002 publication of Manual Therapy Journal the editors would like to thank readers for their patience at a time when the publishers of the Journal, Elsevier were going through a number of internal re-organizations, which delayed publication of 2 issues of Manual Therapy Journal in 2002 by some weeks. We value the support and understanding you showed to the editorial board and also the Journal during this difficult time and we are pleased that Manual Therapy Journal along with many other Elsevier journals are now back on their planned production schedules. Ann Moore and Gwen Jull Editors

129

Manual Therapy (2003) 8(3), 130–140 r 2003 Elsevier Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00051-1

Masterclass

A pain neuromatrix approach to patients with chronic pain G. L. Moseley*,w Department of Physiotherapy, University of Queensland, Australia, wDepartment of Physiotherapy, Royal Brisbane Hospital, Brisbane, Australia

n

SUMMARY. This paper presents an approach to rehabilitation of pain patients. The fundamental principles of the approach are (i) pain is an output of the brain that is produced whenever the brain concludes that body tissue is in danger and action is required, and (ii) pain is a multisystem output that is produced when an individual-specific cortical pain neuromatrix is activated. When pain becomes chronic, the efficacy of the pain neuromatrix is strengthened via nociceptive and non-nociceptive mechanisms, which means that less input, both nociceptive and non-nociceptive, is required to produce pain. The clinical approach focuses on decreasing all inputs that imply that body tissue is in danger and then on activating components of the pain neuromatrix without activating its output. Rehabilitation progresses to increase exposure to threatening input across sensory and non-sensory domains. r 2003 Elsevier Ltd. All rights reserved.

It is important to note that there are assumptions underlying the present approach that have not yet been validated and this paper is occasionally speculative for the sake of clinical relevance. There are no illusions that one approach provides the panacea for persistent pain. Rather, it is hoped that the astute reader will take from this work those aspects that can be integrated with their clinical experience and approach in order to promote better outcomes in a population for which success is elusive.

INTRODUCTION Pain and movement are the primary currencies in manual therapy. Most patients present for therapy because they are in pain and most therapies incorporate movement into assessment, diagnosis, aetiology and management. Indeed, many therapies attempt to restore movement in the hope that pain will automatically get better as movement improves. However, from aetiologic and therapeutic perspectives, it is difficult to determine the chicken and the egg: is pain caused by abnormal movement or is abnormal movement caused by pain? The current paper presents a model for management in which pain and changes in motor control are considered two dimensions of a multidimensional output of the ‘pain neuromatrix’. The model is based on the author’s interpretation of the current thought across the pain sciences. The theoretical background for this model is presented and the implications for assessment and intervention are discussed.

BACKGROUND A fundamental principle of this approach is that pain is produced by the brain when it perceives that danger to body tissue exists and that action is required. All dimensions of pain serve to promote this objective. Thus pain is a multiple system output, not just ‘an unpleasant sensory and emotional experience associated with actual or potential damagey’ (Merskey & Bogduk 1994). The principle that pain is primarily aimed at action is not novel — it was proposed by Patrick Wall a decade ago (Wall 1994) — however, it is yet to gain widespread acceptance. This is somewhat surprising because the notion that pain is a reliable informant of what is actually happening in the tissues is no longer tenable. There

Received: 16 March 2003 Revised: 8 April 2003 Accepted: 12 April 2003 Lorimer Moseley, PhD, B.App.Sc. (Phty) (Hons), Department of Physiotherapy, Royal Brisbane Hospital, Herston, 4029, Australia. Tel.: +61-7-3636-2590; Fax: +61-7-3636-2595; E-mail: [email protected] 130

A pain neuromatrix approach 131

are many factors that are important in determining what hurts for whom and when, for example context, company, competitive stimuli, meaning. Discussion of those factors is not appropriate here. However, their undeniable impact points clearly to (a) the complexity of pain as a multiple system mechanism, and (b) the simplicity of its modus operandi: the body is in danger and action is required. Each dimension of pain is consistent with an immediate relevance to survival. Experientially, pain is unpleasant (Merskey & Bogduk 1994) and demands attention (Eccleston & Crombez 1999). Pain reduces cortical processing capacity (Derbyshire et al. 1998), slows decision making (Crombez et al. 1996), and increases cognitive error rate (Buckelew et al. 1986). Not surprisingly, chronic pain patients often report being forgetful and easily distracted (Jamison et al. 1988; Parmelee et al. 1993; Schnurr & MacDonald 1995). During pain, immune activity is modified (Watkins & Maier 2000), hypothalamus– pituitary–adrenal axes and sympathetic nervous system activity is altered (Melzack 1999), reproductive system function is reduced (Negro-Vilar 1993; Rivier 1995) and visuomotor systems are activated (Price 2000). Thus, pain is the cortical output of highest priority. Motor output as a dimension of pain varies according to the task at hand, but generally serves to both promote escape and limit provocation of the painful part. Limb muscle studies have shown that to this end, muscle pain is associated with increased and decreased activity of the agonist muscle and antagonist muscles, respectively (Lund et al. 1991; GravenNielsen et al. 1997). Trunk muscle studies have shown that pain is associated with reduced modulation of muscle activity during dynamic movements (ArendtNielsen et al. 1996; Zedka et al. 1999) and reduced and increased activity of deep and superficial trunk muscles, respectively, during single limb movements (Hodges et al. 2003). This type of response is consistent with splinting the trunk and/or preparing the body for torque production in the limbs. Taken together, the data suggest that motor output changes associated with pain are predominantly (although not exclusively) driven by higher centres.

What is the pain neuromatrix? In the current therapeutic approach, the pain neuromatrix is that combination of cortical mechanisms that when activated produce pain. The term is taken from Melzack’s ‘Neuromatrix theory’ (1996), but also acknowledges progress in imaging studies that identify neuroanatomical correlates of pain (the ‘Pain matrix’ — including anterior cingulate cortex (ACC), insular cortex, thalamus and sensorimotor cortex). Manual Therapy (2003) 8(3), 130–140

The conceptual view of the pain neuromatrix Although the brain is a remarkable information processor, it is probably unable to create an experience instantaneously based on incoming sensory information, even though this is commonly assumed to be the case. Rather, the brain probably produces a common output that extends to awareness and motor tasks, much in the manner proposed by Hebb (1949) and Melzack (1996). Those theories conceptualize a network of cells (termed ‘the neuromatrix’ by Melzack), determined genetically and/or on the basis of sensory input, that produces a constant perceptual and motor output, such that both outputs can be activated by a single input, or in some cases no inputs at all. To illustrate this, consider sitting in a train while the adjacent train begins to move. In this situation, visual cues alone are often sufficient to produce the experience of moving and a postural response appropriate to that experience, even though there is no corroborative vestibular or proprioceptive input. The neuroanatomical pain matrix Imaging studies demonstrate that there is no single ‘pain centre’. Many cortical areas can be activated during pain and wide variability exists within and between individuals (see Ingvar and Hsieh 1999 for review). However, some cortical areas are involved more often than others. These structures are known as the ‘pain matrix’ and provide a neuroanatomical reference for the current therapeutic model. Most studies report activity in the ACC during pain (e.g. Ingvar & Hsieh 1999; Creac’h et al. 2000; Apkarian et al. 2001; Bantick et al. 2002), although most imaging studies not investigating pain also report ACC activity. Across studies, the middle portion of the ACC is thought to be important for deciding ‘what should I do?’, such that it can be considered an action centre (the ACC has been termed the limbic-motor cortex, Craig 2002). In pain studies, the ACC is considered to serve to (i) establish an emotional valence of pain, and (ii) coordinate the selection and planning of an appropriate behavioural/motor response strategy (Price 2000). Similar ACC activity has been reported during non-nociceptive but biologically threatening events such as anticipated pain (Sawamoto et al. 2000) and anxiety (Kimbrell et al. 1999; Osuch et al. 2000) and the ACC is chronically active in chronic pain patients (Hsieh et al. 1995). Other key areas include the thalamus (Bushnell & Duncan 1989), anterior insula, prefrontal and posterior parietal cortices (Ingvar & Hsieh 1999). These areas are implicated in the affective-emotional and motoric dimensions of pain and together can be considered primary substrates of the pain neuromatrix (Fig. 1). In terms of the therapeutic approach presented here, the exact neuroanatomy of the pain neuror 2003 Elsevier Ltd. All rights reserved.

132 Manual Therapy

Frontal cortex

ACC & insular Thalamus

Premotor cortex Sensorimotor cortex

L

P Fig. 1—Functional magnetic resonance image (axial view, multiple slices) of brain activity during painful thermal stimulation (571C) of the left hand. Components of the pain matrix are circled — thalamus, anterior cingulate cortex (ACC), insular, frontal, premotor and primary sensory and motor cortices. Note that the image is left-right inverted such that L marks the left cortical hemisphere. P denotes posterior.

matrix is not important. Rather, the above studies suggest cortical regions that are likely to be involved in what is ultimately an individual-specific pain neuromatrix.

The virtual body A critical component of the current approach is that pain is experienced in the body image that is held by the brain, labelled here the ‘virtual body’. Although this notion seems at first glance to be counterintuitive, this is only because the virtual body is so effective — the phantom experiences of an amputated limb is a profound case in point. Phantom experience has been discussed at length by Melzack (Melzack 1989; 1990; 1996; Melzack et al. 1997). Neuroanatomically, the primary somatosensory homunculus is the most well-known spatial representation of the internal and external physical environment (Deiber et al. 1991; Grafton et al. 1992; Grafton et al. 1996), but there are other representations as well. The dorsal insular cortex is proposed to contain representation of the physiological condition of the entire body (Craig 2002). This proposal is based firstly on studies that show activity of the dorsal insular during homeostatic mechanisms (including pain), and secondly on studies that show its connection almost exclusively with small-diameter afferents. It is noteworthy in this regard that small diameter afferents, although conventionally called nociceptors, are more accurately considered interoceptors — they detect changes in the body tissue1 (MacIver & Tanelian 1992; Carlton et al. 2001; Cook & McCleskey 2002). Different virtual bodies may dominate experience at different times. For example, Andre et al. (2001) 1

An excellent example of the distinction between nociceptors and interoceptors was described by Vallbo et al. 1999 when they observed that C fibres are exquisitely sensitive to slow, weak mechanical stimuli that evoke sensual touch. r 2003 Elsevier Ltd. All rights reserved.

demonstrated temporary non-painful and movable phantom limbs in amputees after vestibular caloric stimulation (inserting 201C water into one ear), regardless of whether, prior to stimulation, they had no phantom, a painful phantom or a dysmorphic phantom. Those authors concluded that sudden vestibular stimulation activated a stable and intact virtual limb in order to provide the postural frame of reference on which to base a postural response to the perturbation. That is, caloric stimulation caused the virtual body to be overridden by a relatively permanent counterpart. However, in the current discussion, the incumbent virtual body is important for the very reason that it is continually updated by sensory input. Therefore, it may be an important part of the pain neuromatrix because it provides a neural substrate for allocating pain an anatomical reference. In light of this, treatment aimed at reducing pain is aimed at the virtual body, albeit commonly (but not exclusively) accessed through the corresponding body tissues. The virtual body is also indispensable for the maintenance of coherent motor commands, including the coordination of postural and movement responses. The central nervous system (CNS) produces motor commands on the basis of (i) the predicted requirements of the movement including the predicted disturbance to stability, and (ii) the perceived current position, movement and stability of the body, such that postural control commands are effected prior to movement (Belen’kii et al. 1967; Bouisset & Zattara 1987). Thus, the virtual body provides a common platform from which experiential and motoric dimensions of pain can be launched, which make it an important consideration for clinicians interested in pain and movement. Added complexity of chronic pain — effects on activity of the pain neuromatrix Two inter-dependent mechanisms can contribute to chronicity — nociceptive (including humoral or immune-related dysfunction that stimulates nociceptive structures and body tissues) and non-nociceptive (cognitive–evaluative) mechanisms. In either case, there is an increase in the conviction of the CNS that body tissue is in danger and, therefore, there is an increase in activity of the pain neuromatrix. When pain persists, both the nociceptive system and the virtual body undergo profound changes, which increase sensitivity to noxious as well as nonnoxious input and corrupt the integrity of motor output. Review of the changes that occur is beyond the scope of this paper; however, potent changes occur peripherally and centrally. Alterations of wide dynamic range second-order nociceptors are particularly relevant (Doubell et al. 1999; Mannion & Woolf 2000) because these second-order nociceptors Manual Therapy (2003) 8(3), 130–140


dominate ascending connections to the brain areas identified as key components of the pain neuromatrix (see Price, 2000 for review). Further, reorganization of primary sensory and motor cortices, which also occurs with chronic pain (Flor et al. 1998), probably has even more profound effects across pain dimensions. Discussion of cognitive–evaluative mechanisms associated with chronicity is also beyond the scope of this paper. It is sufficient to highlight the importance of beliefs and attitudes that emphasise the threat value of pain (see Gatchel & Turk, 1999 for review). The threat value of pain is an important predictor of its unpleasantness, but more importantly it is of obvious relevance to the underlying principle of the current approach — that pain is produced by the brain when it perceives that danger to body tissue exists and that action is required. Finally, it is thought that alterations in both nociceptive and cognitive–evaluative input can lead to modification of the shape and properties of the virtual body, such that motor and postural responses become variable and inaccurate. The effect of chronicity — implications for the pain neuromatrix approach to chronic pain The current approach proposes that the primary effect of chronicity is enhanced synaptic efficacy (broadly speaking, the sensitivity) of the pain neuromatrix, such that less input is required for activation. This proposal is able to explain many clinical phenomena in the chronic pain patient group, e.g. dynamic functional allodynia (pain during movements that would not normally be painful), or pain elicited by seeing another person perform a painful movement. The main point is that smaller and seemingly less relevant inputs are sufficient to activate the neuromatrix and thus produce pain. Herein lies the primary challenges of this patient group: an unclear relationship between pain and tissue input; difficult-to-predict flare-ups; poor tolerance of normal therapeutic approaches; problems with physical and functional upgrading; difficulty generalizing gains to other activities.

THE CLINICAL APPROACH — DESENSITISATION AND GRADED ACTIVATION OF COMPONENTS OF THE PAIN NEUROMATRIX The aim of the current approach is to utilize functional components of the individual-specific pain neuromatrix that are appropriate for movement and that are consistent with the normal aims of treatment, without reinforcing the synaptic efficacy of the neuromatrix. There are three aspects of the approach: (i) reduction of threatening input so as to reduce Manual Therapy (2003) 8(3), 130–140

activity of the pain neuromatrix and thereby reduce its efficacy, (ii) targeted activation of specific components of the pain neuromatrix without activating the neuromatrix, and (iii) upgrading physical and functional tolerance by exposure to threatening inputs across sensory and non-sensory domains. 1. Reduction of threatening input Reduction of threatening input — nociceptive mechanisms Where possible, nociceptive mechanisms that contribute to threatening information should be treated. Obviously, treatment will depend on sound tissue and neural examination techniques and selection of appropriate therapeutic strategies. Broadly speaking, any strategy that has an inhibitory effect on nociceptive input is probably appropriate in the short term unless it simultaneously activates non-nociceptive threatening inputs. For example, although manual therapy strategies may activate endogenous inhibitory mechanisms, (e.g. Vicenzino et al. 1998), treatment may be delivered in such a context as to reinforce to the patient that there is something wrong in their tissues (i.e. the conviction that the body is in danger). Comprehensive appraisal of the mechanisms of various therapeutic strategies lies beyond the expertise of the current author and is not appropriate here. Reduction of threatening input — non-nociceptive mechanisms — education The main objective of education is to decrease the threat value associated with pain by increasing the patient’s understanding of human physiology. Our group has conducted several studies to evaluate the effect of such education as a way of altering beliefs and attitudes about the meaning of pain (e.g. Moseley 2002; 2003b). The education material is outlined in Table 1 and presented in detail in Butler and Moseley (2003). It includes high-level pain physiology information. Contrary to popular opinion among health professionals, patients are able to understand complex physiology if the information is presented appropriately. In fact, after education, patients appear to have a better understanding of pain physiology than most health professionals, excepting those who have participated in education themselves, or who have specialist training in pain sciences (Moseley 2003a). Pain physiology education differs from popular education strategies (for low back pain), which have focussed on anatomy and physiology of the lumbar spine. Such a focus is thought to have limited clinical effect (Cohen et al. 1994; Triano et al. 1995) although for a contrary view, see Waddell and Burton (2000). More recently, education about cognitive and behavioural responses, given prior to injury or after an r 2003 Elsevier Ltd. All rights reserved.

134 Manual Therapy Table 1. Material presented in education about pain physiology The neuron The action potential (message) The synapse Primary nociceptors (danger receptors) Second order nociceptors (danger messenger nerves) Brain output dependent on total perception of danger Descending input Primary nociceptor state-dependent functioning Second order nociceptor state-dependent functioning

Modality-specific receptors, axon, terminal bouton All or nothing, post-synaptic membrane potential, propogation, dromic and antidromic flow Neurotransmitters, inhibitory and facilitatory input, chemically driven ion channels, ion channel synthesis and absorption Respond to danger Sum of inhibitory and facilitatory input, interneurons from nonnociceptive fibres (normally inhibitory), project to many parts of the brain All information that is relevant to that decision is considered — thoughts, memories, beliefs, explanatory model, consequences Inhibitory and facilitatory — also dependent on above Potentiation and summation, ectopic pacemakers, dorsal root ganglion, neurogenic inflammation, allodynia and hyperalgesia Potentiation, active blocking of ion channels, increased receptor synthesis, sprouting, activation by endocrine mediators

initial episode, reduces chronic disability, although the effect on chronic pain is not known (Symonds et al. 1995; Burton et al. 1999). There are limited data that support the efficacy of such education with patients who already have chronic pain. Anecdotally, this may be because the response that is encouraged in such programs is counterintuitive for patients because their understanding of pain is based on a structural-pathology model (i.e. that pain is analogous to injury). Pain physiology education targets this limitation by aiming to reconceptualize the underlying physiological problem of a patient’s pain, on the assumption that an appropriate cognitive and behavioural/motor response will follow. This is sensible according to principles of ‘deep learning’, in which information is retained and understood and applied to problems at hand (Sandberg & Barnard 1997). In contrast, ‘superficial’ or ‘surface’ learning is that in which information is remembered but not understood or integrated with attitudes and beliefs (Evans & Honour 1997). To this effect, it is important to note that deep learning is facilitated when the learner is motivated (Sankaran 2001) and when the information presented is made personally relevant (Moreno & Mayer 2000), both of which are promoted by the method of education used here. It is critical that patients understand the material that is presented — care should be taken to use various graphics (including whiteboards, hand-drawn pictures, personalized manuals), and accessible metaphors in order to achieve this objective. Education can be time consuming but we have demonstrated an effect, albeit smaller, with group programs (Moseley 2003b).

direct questioning — ‘What in your body do you think is causing your pain?’ Most commonly, the explanatory model held by a patient is based on a structural-pathology model because that is the dominant model in the medical and lay arenas. However, the structural-pathology model is often not appropriate. Focus on a structural label for pain may actually heighten attention on the pain, emphasize the vulnerability of the body to damage and increase patients’ health care consumption (Jones et al. 1988; Nachemson 1992; Hirsch & Liebert, 1998). There are three key points to consider in modification of the explanatory model. First, care should be taken to present currently accurate information rather than an explanation that has been diluted for the sake of simplicity, to save time or to avoid a confrontation with the patient. This means that on the basis of the information that they have been presented, the patient should be unable to support an inappropriate or indefensible explanatory model. Second, the information presented should offer the patient an alternative explanatory model that is supported by the currently accurate information about human physiology. Third, the information must be presented in a manner that is respectful of the patient and acknowledges their suffering. At first glance this seems a moot point. However, the negative stigma associated with chronic pain is remarkably pervasive despite the fact that the basis of the stigma is na.ıve. In summary, to be effective with education, the therapist must (i) be an expert and be perceived as such by the patient, and (ii) be respectful and compassionate and be perceived as such by the patient.

Targeting the explanatory model Based on the underlying principle of the current work, the way that a patient explains their pain (‘explanatory model’) is an important consideration in targeting education and planning therapy. What story does the patient tell about their pain? Evaluation of the explanatory model is possible through

2. Activation of components of the individual’s pain neuromatrix

r 2003 Elsevier Ltd. All rights reserved.

Management framework The following framework for management is presented to patients (Fig. 2). It represents a practical response to the pain physiology education and Manual Therapy (2003) 8(3), 130–140

Demand on human system (e.g. tissue)


Pre-injury tissue tolerance

Pre-injury buffer

Pre-injury pain threshold

Current tissue tolerance

Flare-up zone

Maximum training load and progression

Flare-up line

Current buffer Current pain threshold

Fig. 2—A management framework that is presented to patients. The y-axis represents the demand on a particular human system (usually the musculoskeletal system) that is imparted by a certain activity or combination of activities. The pre-injury and current buffer conceptualise the protective gap between onset of pain and tissue damage. Note that in chronic pain, by virtue of enhanced synaptic efficacy of the pain matrix, the buffer has been expanded. This is an effective protective strategy. The training load (dashed line) begins below the flare-up line and progresses conservatively. The flare-up line and tissue tolerance gradually increase in line with progression of the training load.

incorporates an understanding of pain as the multiple system output of a sensitized pain neuromatrix. Prior to the onset of pain, there was a maximum amount of an activity that could be performed before tissue would in some way yield. This is termed the ‘previous tissue tolerance’. There was a reliable protective mechanism, mediated primarily by the sensory nerves within body tissue, that would usually activate the pain neuromatrix. However, three main effects occur after injury and with chronic pain: (i) reduction in tissue tolerance, by virtue of altered tissue properties, deconditioning and disuse, (ii) reduction in the activation threshold of the peripheral nociceptors such that the integrity of the tissue-based protective mechanisms is maintained, (iii) increase in and diversification of threatening input, mediated by increased sensitivity of the higher order nociceptive structures (spinal cord and brain) and/or cognitive– evaluative factors associated with threat, such that the pain neuromatrix is activated at low levels of threatening input.

Assessment — determining the baseline In the current context, ‘baseline’ is the extent to which the components of the pain neuromatrix can be utilized without activating the whole neuromatrix. That is, what level of nociceptive input (e.g. how many, how often, how demanding an exercise or functional task?) or other input (e.g. how threatening from a non-nociceptive perspective?) can occur without activating the pain neuromatrix? Determination of the baseline is more difficult with chronic pain than with acute or sub-acute problems, by virtue of the increase and diversification of threatening inputs. Manual Therapy (2003) 8(3), 130–140

Assessment consists of determining motor and functional baselines. Motor baselines Clinical assessment of motor strategies can involve those tests with which the therapist is most comfortable and competent (e.g. assessment of deep trunk muscle function or one-legged standing balance), however the tests may require more conservative baselines than is required in acute or sub-acute problems. Motor strategies can provide important insight into what inputs are likely to activate the pain neuromatrix. According to the theoretical basis outlined earlier, particular note should be made of the relative activation of torque producing and nontorque producing (so-called ‘stabilizing’) muscles. Broadly speaking, this is a consideration common to numerous body areas, for example, deep versus superficial trunk muscle activation in back pain (O’Sullivan et al. 1997; Richardson et al. 1999), vastus medialis obliquus versus other quadriceps muscles in knee pain (Voight and Wieder, 1991; Cowan et al. 2002), deep cervical flexors versus superficial neck muscles in neck pain (Jull et al. 1999). In a portion of cases the baseline will be relatively normal. In many cases, assessment will involve excursion from this normal baseline in a multidimensional manner. Fig. 3 outlines the options. If the patient can easily obtain a normal baseline, the new baseline can be determined by progressively making the task more threatening. This may involve increasing the physical demand of the task — increase the speed, amplitude and duration of the movement. Alternatively or in addition, it may involve modifying the context or environment, increase the implications r 2003 Elsevier Ltd. All rights reserved.

136 Manual Therapy

Normal baseline Not painful/appropriate motor strategy

Painful/inappropriate motor strategy

A

B

Increase threat

Physical demand:increased speed, torque, amplitude, duration.

Change context: work or injuryspecific, traumaspecific,

Decrease threat

Change emotional or cognitive load

Physical demand: break into components reduce speed, torque, amplitude, duration.

Change context: imagery, social environment context, humour.

Change posture, alter other inputs e.g. visual.

Imagined movements (non-painful), hypnosis.

Fig. 3—Options for determining the baseline and planning progression. (A) If the normal baseline level is performed pain-free (functional task) or using an appropriate motor strategy (motor task), then the aim is to increase threat via physical, contextual changes or emotional/ cognitive load. (B) If the normal baseline level elicits pain (functional task) or an inappropriate motor strategy (motor task), then the aim is to decrease threat, via similar means.

of poor performance (e.g. ‘someone of your ability should have no problem performing this’), or perform the same movement from a different postural reference or within a different emotional context. For example, a patient with chronic back pain is able to perform a near perfect voluntary abdominal muscle task in standing. Assessment of muscle activity while this patient imagines bending forward to pick up a box might reveal a loss of quality of contraction, for example marked activation of the superficial abdominal muscles/long back extensors. This finding implies that forward bending is a threatening movement and identifies success at this task as a suitable short-term goal. Threatening stimuli can be work-specific (e.g. work environment), context-specific (e.g. sitting in car) and posturespecific (that is, the postural frame of reference from which a movement is occurring). For many patients, the changes in nociceptive and cognitive–evaluative inputs are such that the baseline is difficult to find because conventional baselines are painful (i.e. they activate the pain neuromatrix). In this situation, the aim is to utilize as much of the pain neuromatrix as possible without activating it. That is, to reduce the threat associated with the task (Fig. 3b). This may involve breaking the movement down, performing imaginary movements (that are imagined pain-free), or changing the postural reference or context of the movement. Can the task be modified in order to make it non-threatening? How can components of the movement/task be performed in a manner that does not activate the pain neuromatrix? For example, if minimal cervical rotation in sitting is painful, try cervical rotation in standing, imagined r 2003 Elsevier Ltd. All rights reserved.

cervical rotation in sitting, performing the movement in the dark (i.e. removing visual stimuli that may add to the threat), or try imagery in which the patient is turning their head toward a cool breeze on a hot day (see Table 2). These strategies activate motor mechanisms that would normally activate the pain neuromatrix but do so in a manner that is explicitly non-threatening and therefore less likely to activate the pain neuromatrix.

Functional baselines Functional baselines are effectively determined by the patient, however persistent assistance from the therapist is required. The functional baseline is determined by the flare-up line presented in Fig. 2, which is the point at which the patient notices a marked increase in pain that persists for more than a few minutes. Patients invariably relate to the term flare-up and recognize it as a period of severe pain which is often accompanied by incapacity, inability to sleep, nausea and vomiting, and ‘drastic’ treatment options. Flare-ups need not be this severe and, anecdotally at least, are less severe if the patient has gained an accurate understanding of human physiology as it relates to his/her pain. However, the prevalence of this idea of a flare-up probably reflects in part that patients find it difficult to determine their own flare-up line. Physiologically, the flare-up line may coincide with activation of tissue-initiated nociception, although there are no data that evaluate this possibility. Persistent and skilled questioning should suffice to identify the baseline of any functional task. For Manual Therapy (2003) 8(3), 130–140

A pain neuromatrix approach 137 Table 2. Examples of multidimensional decrease and increase in threat Activity

Flare-up line

Mon

Tue

Wed

Thu

Fri

Sat

Sun

Deep neck muscles/head tilts Hanging washing Typing at desk

10 10 10 items 8 min

55 4 items 4 min

75 5 4.30

95 6 5

86 7 5.30

96 8 6

10 6 9 6.30

97 10 7

example, an interaction between therapist and patient; Therapist: ‘How long can you walk before you flare-up?’ Patient: ‘I can walk for 30 min but I pay for it the next day’ Therapist: ‘Can you walk for 20 min without flaring up?’ Patient: ‘No, but I have’ Therapist: ‘Can you walk for 10 min without flaring up?’ Patient: ‘Probably not — definitely not up hills’ Therapist: ‘5 min on a flat surface?’ Patient: ‘Probably’ Therapist: ‘3 min on a flat surface?’ Patient: ‘Definitely’ Thus, for this patient the baseline for walking is 3 min on a flat surface. This process of questioning, although laborious, is usually critical in order to get a clearly defined baseline and to impart an understanding to the patient of what is meant by ‘baseline’. The patient is usually able to undertake this process on their own once several activities have been discussed and baselines have been defined. It is critical that the agreed baseline is recorded, preferably in the patient’s diary. Progression — a little a lot Fig. 3a also provides a framework within which motor and functional tasks can be progressed. Once a baseline is identified, a goal is established. When that goal is achieved, it becomes the new baseline. Different issues are pertinent to progression of motor and functional tasks. In either case, progression must be conservative but continual — small, frequent increments in the training load. This can be frustrating for patients and therapists alike, however, it is important to both increase tolerance and avoid flareups: ‘persistence and patience’ is a good theme. Generally speaking, patients respond to this if they can understand the rationale, which is presented both as part of the education and in an ongoing way.

Easily remembered phrases are useful in this regard — e.g. ‘every day you do more than you did yesterday, but not much more’. Progression diary Compliance is a major barrier to chronic pain rehabilitation (see Nicholas 1995 for review). Most often, patients exceed the flare-up line and become entrenched on a boom-bust cycle of increasing disability and dysfunction. Providing patients with a sound rationale as to why a flare-up should be avoided counters this, at least in part. However, in order to balance progression with avoidance of flareups, it is often necessary to keep a progression diary, in which the motor and functional training load is well documented and planned in advance. The aim of a progression diary (see Table 3) is to map out the increments of motor and functional training and to ensure that patients do not exceed the training load. This is critical because patients will often, having completed the functional goal of a given day, be feeling relatively pain-free and satisfied with their progress, and decide to do just a little more. Anecdotally, exceeding the training load in this manner, which often leads to flare-up, which in turn is poorly managed by virtue of an inaccurate understanding of the physiology underlying the flare-up, is the most common reason for failure in management. Training alarm Much functional and motor training is defined by time. Because distraction remains the most effective means of endogenous analgesia, it is both an ally and enemy in management because patients are at risk of exceeding the training load, simply because they do not realize they have done so. For example, consider the patient who has as a pre-planned training load of sitting for 14 min. The patient enjoys reading and uses reading as an effective distraction. However, the patient is sufficiently distracted that they read for 20 min at which time they have quite rapid onset of pain that leads to a flare-up. Alarm clocks are

Table 3. Example of a progression diary, applied to functional and motor tasks Target

Change context

Change posture

Recruit premotor and Alter sensory input motor no movement

Change emotional state

To decrease threat

Rotate while talking with friends Rotate in car

Rotate in standing/ lying/side lying Rotate with arms elevated

Imagine pain-free rotation Imagine rotating against resistance

Rotate while listening to favourite music Rotate when angry

To increase threat

Manual Therapy (2003) 8(3), 130–140

Rotate with eyes closed Rotate while eating


138 Manual Therapy

invaluable in preventing this scenario and are easy to use. In the previous example, by setting the alarm to 14 min the patient is able to benefit from distraction by reading but not exceed the training load. Progression of motor tasks Each time a new baseline is established, progression should involve exposure to more threatening inputs in a multidimensional manner. If the increment of progression according to physical demand is too large, it is possible to progress in other dimensions, for example, perform the motor task under stressful conditions, in more threatening emotional states, or while performing a cognitive task. Table 2 provides examples of multidimensional increases in threat associated with rehabilitation of the patient with cervical spine pain. This strategy is consistent with the theoretical goal of activating components of the neural network without triggering it to produce pain. Progression of work Often the success of a return to work trial is prevented by the patient exceeding a suitable training load, usually because the increment in physical load is too large. We have had greater success in the chronic pain patient group by pursuing more conservative training loads that are increased more often and with smaller increments. For example, M was a patient who presented with 4 year history of debilitating arm pain initiated and aggravated by keyboard work. She had failed two previous attempts to return to work that began with three 4 hour(h) shifts (12 h) per week. On both occasions she did not complete 2 weeks of work. In our revised progression M began with 15 min at work and incremented initially daily in 5 min blocks. Importantly, she went to work everyday. She successfully graduated to five 3-h shifts (15 h) per week by the 12th week of the program and proceeded to full time work duties in B6 months. At first glance, this seems like a laborious return to work program. However, it was both quicker and cheaper (in terms of time off work) than previous attempts.

SUMMARY AND CONCLUSION The current paper has presented the background, guiding principles and clinical approach to treating chronic pain using a pain neuromatrix approach. The approach is founded on key principles — that pain is a multiple system output that is activated by an individual-specific pain neuromatrix; that the pain neuromatrix is activated whenever the brain concludes that body tissue is in danger and action is required; and that pain is allocated an anatomical reference in the virtual body, upon which coherent motor output is also dependent. When pain persists, r 2003 Elsevier Ltd. All rights reserved.

there is (i) an increase and diversification of threatening inputs such that the pain neuromatrix can be activated by all manner of threatening stimuli, nociceptive (including immune and endocrine-driven mechanisms) and cognitive–evaluative, and (ii) alterations in the morphology and behaviour of the virtual body, thus further altering motor output. The therapeutic aspects of the approach focus on reducing the sensitivity and activity of the pain neuromatrix, via reduction of threat. The key components are education about human physiology and a systematic approach to identification and progression of motoric and functional baselines across sensory and non-sensory domains.

Acknowledgements Lorimer Moseley is supported by grant number 210348 from the National Health and Medical Research Council of Australia.

References Andre JM, Martinet N, Paysant J, Beis JM, Le Chapelain L 2001 Temporary phantom limbs evoked by vestibular caloric stimulation in amputees. Neuropsychiatry Neuropsychology and Behavioral Neurology 14(3): 190–196 Apkarian AV, Krauss BR, Fredrickson BE, Szeverenyi NM 2001 Imaging the pain of low back pain: functional magnetic resonance imaging in combination with monitoring subjective pain perception allows the study of clinical pain states. Neuroscience Letters 299(1–2): 57–60 Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P 1996 The influence of low back pain on muscle activity and coordination during gait: A clinical and experimental study. Pain 64(2): 231–240 Bantick SJ, Wise RG, Ploghaus A, Clare S, Smith SM, Tracey I 2002 Imaging how attention modulates pain in humans using functional MRI. Brain 125(Pt 2): 310–319 Belen’kii V, Gurfinkel VS, Paltsev Y 1967 Elements of control of voluntary movements. Biofizika 12: 135–141 Bouisset S, Zattara M 1987 Biomechanical study of the programming of anticipatory postural adjustments associated with voluntary movement. Journal of Biomechanics 20: 735–742 Buckelew S, De Good D, Schwartz D, Kerler R 1986 Cognitive and somatic item response patterns of pain patients, psychiatric patients and hospital employees. Journal of Clinical Psychology 42: 852–860 Burton AK, Waddell G, Tillotson KM, Summerton N 1999 Information and advice to patients with back pain can have a positive effect. A randomized controlled trial of a novel educational booklet in primary care. Spine 24(23): 2484–2491 Bushnell MC, Duncan GH 1989 Sensory and affective aspects of pain perception: Is medial thalamus restricted to emotional issues? Experimental Brain Research 67: 415–418 Butler DG, Moseley GL 2003 Explain Pain. NOI Group Publishing, Adelaide. Carlton S, Du J, Zhou S, Coggeshall R 2001 Tonic control of peripheral cutaneous nociceptors by somatostatin receptors. Journal of Neuroscience 21: 4042–4049 Cohen JE, Goel V, Frank JW, Bombardier C, Peloso P, Guillemin F 1994 Group education interventions for people with low back pain. An overview of the literature. Spine 19(11): 1214–1222 Cook S, McCleskey E 2002 Cell damage excites nociceptors through release of cytosolic ATP. Pain 95: 41–47 Cowan SM, Hodges PW, Bennell KL, Crossley KM 2002 Altered vastii recruitment when people with patellofemoral pain Manual Therapy (2003) 8(3), 130–140


syndrome complete a postural task. Archives of Physical Medicine and Rehabilitation 83(7): 989–995 Craig A, 2002. How do you feel? Interoception: The sense of the physiological condition of the body. Nature Review Neuroscience 3: 655–666 Creac’h C, Henry P, Caille JM, Allard M 2000 Functional MR imaging analysis of pain-related brain activation after acute mechanical stimulation. American Journal of Neuroradiology 21(8): 1402–1406 Crombez G, Eccleston C, Baeyens F, Eelen P 1996 The disruptive nature of pain: An experimental investigation. Behaviour Research and Therapy 34(11-12): 911–918 Deiber MP, Passingham RE, Colebatch JG, Friston KJ, Nixon PD, Frackowiak RS 1991 Cortical areas and the selection of movement: A study with positron emission tomography. Experimental Brain Research 84(2): 393–402 Derbyshire SW, Vogt BA, Jones AK 1998 Pain and Stroop interference tasks activate separate processing modules in anterior cingulate cortex. Experimental Brain Research 118(1): 52–60 Doubell TP, Mannion RJ, Woolf CJ 1999 The dorsal horn: statedependent sensory processing, placticity and the generation of pain. In: Wall P, Melzack R (eds) The Textbook of Pain. Churchill Livingstone, Edinburgh, pp 165–181 Eccleston C, Crombez G 1999 Pain demands attention: A cognitive-affective model of the interruptive function of pain. Psychological Bulletin 125(3): 356–366 Evans B, Honour L 1997 Getting inside knowledge: The application of Entwistle’s model of surface/deep processing producing open learning materials. Educational Psychologists 17(1–2): 127–139 Flor H, Elbert T, Muhlnickel W, Pantev C, Wienbruch C, Taub E 1998 Cortical reorganization and phantom phenomena in congenital and traumatic upper-extremity amputees. Experimental Brain Research 119(2): 205–212 Gatchel RJ, Turk DC (eds), 1999 Psychosocial Factors in Pain. Guilford Press, New York Grafton ST, Fagg AH, Woods RP, Arbib MA 1996 Functional anatomy of pointing and grasping in humans. Cerebral Cortex 6(2): 226–237 Grafton ST, Mazziotta JC, Woods RP, Phelps ME 1992 Human functional anatomy of visually guided finger movements. Brain 115(Pt 2): 565–587 Graven-Nielsen T, Svensson P, Arendt-Nielsen L 1997 Effects of experimental muscle pain on muscle activity and co-ordination during static and dynamic motor function. Electroencephalography and Clinical Neurophysiology 105(2): 156–164 Hebb D 1949 The Organization of behavior. Wiley, New York. Hirsch MS, Liebert RM 1998 The physical and psychological experience of pain: The effects of labeling and cold pressor temperature on three pain measures in college women. Pain 77(1): 41–48 Hodges PW, Moseley GL, Gabrielsson A, Gandevia SC 2003 Evidence for a direct effect of pain on postural recruitment of the trunk muscles in pain free humans. Experimental Brain Research (in press). Hsieh JC, Belfrage M, Stone-Elander S, Hansson P, Ingvar M 1995 Central representation of chronic ongoing neuropathic pain studied by positron emission tomography. Pain 63(2): 225–236 Ingvar M, Hsieh J 1999 The image of pain. In: Wall P, Melzack R (eds), The textbook of Pain. Churchill Livingstone, Edinburgh, pp 215–234 Jamison RN, Sbrocco T, Parris WC 1988 The influence of problems with concentration and memory on emotional distress and daily activities in chronic pain patients. International Journal of Psychiatry in Medicine 18(2): 183–191 Jones SL, Jones PK, Katz J 1988 Compliance for low-back pain patients in the emergency department. A randomized trial. Spine 13(5): 553–556 Jull G, Barrett C, Magee R, Ho P 1999 Further clinical clarification of the muscle dysfunction in cervical headache. Cephalalgia 19(3): 179–185 Kimbrell TA, George MS, Parekh PI, Ketter TA, Podell DM, Danielson AL, Repella JD, Benson BE, Willis MW, Herscovitch P, Post RM 1999 Regional brain activity during transient self-induced anxiety and anger in healthy adults. Biological Psychiatry 46(4): 454–465 Manual Therapy (2003) 8(3), 130–140

Lund JP, Donga R, Widmer CG, Stohler CS 1991 The pain-adaptation model: A discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of Physiological Pharmacology 69(5): 683–694 MacIver M, Tanelian D 1992 Activation of C fibers by metabolic perturbations associated with tourniquet ischemia. Anesthesiology 76: 617–623 Mannion RJ, Woolf CJ 2000 Pain mechanisms and management: A central perspective. Clinical Journal of Pain 16(3 Suppl): S144–S156 Melzack R, 1989 Phantom limbs, the self and the brain. Canadian Psychology 30: 1–16 Melzack R, 1990 Phantom limbs and the concept of a neuromatrix. Trends in Neuroscience 13: 88–92 Melzack R 1996 Gate control theory. On the evolution of pain concepts. Pain Forum 5(1): 128–138 Melzack R, 1999. Pain and stress: A new perspective. In: Gatchel R, Turk DC (eds) Psychosocial Factors in Pain. Clinical perspectives. The Guilford Press, New York, pp 89–106 Melzack R, Israel R, Lacroix R, Schultz G 1997 Phantom limbs in people with congenital limb deficiency or amputation in early childhood. Brain N 120(Pt 9): 1603–1620 Merskey H, Bogduk N 1994 Classification of chronic pain. IASP Press, Seattle Moreno R, Mayer R 2000 Engaging students in active learning: The case for personalized multimedia messages. Journal of Educational Psychology 92(4): 724–733 Moseley GL 2002 Physiotherapy is effective for chronic low back pain. A randomised controlled trial. Australian Journal of Physiotherapy 48: 297–302 Moseley GL 2003a Unravelling the barriers to reconceptualisation of the problem in chronic pain: The actual and perceived ability of patients and health professionals to understand the neurophysiology. Journal of Pain 7(4): 184–189 Moseley GL 2003b Joining forces – combining cognition-targeted motor control training with group or individual pain physiology education: A successful treatment for chronic low back pain. Journal of Manual and Manipulative Therapeutics (in press) Nachemson AL 1992 Newest knowledge of low back pain. A critical look. Clinical Orthopaedics 279: 8–20 Negro-Vilar A 1993 Stress and other environmental factors affecting fertility in men and women: Overview. Environmental Health Perspective 101: 59 Nicholas MK 1995 Compliance: A barrier to occupational rehabilitation? Journal of Occupational Rehabilitation 5(4): 271–282 Osuch EA, Ketter TA, Kimbrell TA, George MS, Benson BE, Willis MW, Herscovitch P, Post RM, 2000. Regional cerebral metabolism associated with anxiety symptoms in affective disorder patients. Biological Psychiatry 48(10): 1020–1023 O’Sullivan P, Twomey L, Allison G, Sinclair J, Miller K, Knox J 1997 Altered patterns of abdominal muscle acitvation in patients with chronic low back pain. Australian Journal of Physiotherapy 43(2): 91–98 Parmelee PA, Smith B, Katz IR 1993 Pain complaints and cognitive status among elderly institution residents. Journal of American Geriatric Society 41(5): 517–522 Price DD 2000 Psychological Mechanisms of Pain and Analgesia. IASP Press, Seattle. Richardson C, Jull G, Hodges P, Hides J 1999 Therapeutic Exercise for Spinal Segmental Stabilization in Low Back Pain. Churchill Livingstone, London Rivier C 1995 Luteinizing-hormone-releasing hormone, gonadotropins, and gonadal steroids in stress. Annals of New York Academy of Science 771: 187 Sandberg J, Barnard Y 1997 Deep learning is difficult. Instructional Science 25(1): 15–36 Sankaran S 2001 Impact of learning strategies and motivation on performance: A study in web-based instruction. Journal of Instructional Psychology 28(3): 191–198 Sawamoto N, Honda M, Okada T, Hanakawa T, Kanda M, Fukuyama H, Konishi J, Shibasaki H 2000 Expectation of pain enhances responses to nonpainful somatosensory stimulation in r 2003 Elsevier Ltd. All rights reserved.

140 Manual Therapy

the anterior cingulate cortex and parietal operculum/posterior insula: An event-related functional magnetic resonance imaging study. Journal of Neuroscience 20(19): 7438–7445 Schnurr RF, MacDonald MR 1995 Memory complaints in chronic pain. Clinical Journal of Pain 11(2): 103–111 Symonds TL, Burton AK, Tillotson KM, Main CJ 1995 Absence resulting from low back trouble can be reduced by psychosocial intervention at the work place. Spine 20(24): 2738–2745 Triano JJ, McGregor M, Hondras MA, Brennan PC 1995 Manipulative therapy versus education programs in chronic low back pain. Spine 20(8): 948–955 Vallbo A, Olausson H, Wessberg J 1999 Unmyelinated afferents constitute a second system coding tactile stimuli of the human hairy skin. Journal of Neurophysiology 81: 2753–2763 Vicenzino B, Collins D, Benson H, Wright A 1998 An investigation of the interrelationship between manipulative therapy-induced hypoalgesia and sympathoexcitation. Journal of Manipulative and Physiology Therapeutics 21(7): 448–453


Voight ML, Wieder DL 1991 Comparative reflex response times of vastus medialis obliquus and vastus lateralis in normal subjects and subjects with extensor mechanism dysfunction. An electromyographic study. American Journal of Sports Medicine 19(2): 131–137 Waddell GA, Burton 2000 Evidence review. In: Carter J, Birrell L. (eds) Occupational Health Guidelines for the Management of Low Back Pain at Work — Principal Recommendations. Faculty of Occupational Medicine, London Wall P 1994. Introduction to the edition after this one. In: Wall P, Melzack R. (eds). The Textbook of Pain. Churchill Livingstone, Edinburgh Watkins L, Maier S 2000 The pain of being sick: Implications of immune-to-brain communication for understanding pain. Annual Reviews in Psychology 51: 29–57 Zedka M, Prochazka A, Knight B, Gillard D, Gauthier M 1999 Voluntary and reflex control of human back muscles during induced pain. Journal of Physiology (London) 520(Pt 2): 591–604

Manual Therapy (2003) 8(3), 130–140

Manual Therapy (2003) 8(3), 141–150 r 2003 Elsevier Science Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00010-9

Systematic review

The efficacy of stretching for prevention of exercise-related injury: a systematic review of the literature S.M. Weldon, R.H. Hill Osteopaths, Private Practice, London, UK

SUMMARY. The objective of this study was to conduct a systematic analysis of the literature to assess the efficacy of stretching for prevention of exercise-related injury. Randomized clinical trials (RTCs) and controlled clinical trials (CCTs) investigating stretching as an injury prevention measure were selected. A computer-aided search of the literature was conducted for relevant articles, followed by assessment of the methods of the studies. The main outcome measures were scores for methodological quality based on four main categories (study population, interventions, measurement of effect, and data presentation and analysis) and main conclusions of authors with regard to stretching. One RCT (25%) and three CCTs (100%) concluded that stretching reduced the incidence of exercise-related injury. Three RCTs (75%) concluded that stretching did not reduce the incidence of exercise-related injury. Only two studies scored more than 50 points (maximum score=100 points) indicating that most of the studies selected were of poor quality. Neither of the two highest scoring RCTs showed positive effects for stretching. Due to the paucity, heterogeneity and poor quality of the available studies no definitive conclusions can be drawn as to the value of stretching for reducing the risk of exercise-related injury. r 2003 Elsevier Science Ltd. All rights reserved.

rationale for stretching prior to exercise is to increase the compliance of muscle, reduce muscle stiffness and consequently less force will be generated in the muscle for a given stretch (Noonan & Garrett 1992), the suggestion being that there is a critical force at which a muscle will fail (Garrett et al. 1987; Mair et al. 1996). Recent in vivo (Gadjosik 1991; McHugh et al. 1992; Magnusson et al. 1996) and in vitro research (Taylor et al. 1990) has demonstrated transient viscoelastic stress relaxation in response to passive stretch. Some disagreement exists as to whether these transient changes in mechanical properties following stretching can reduce injury risk and what parameters are related to a tissue’s resistance to injury. Increases in length to failure (compliance) following cyclic stretching (Garrett 1996) were interpreted as a protective effect. Noonan et al. (1994) showed a non-significant increase in force to failure, deformation and energy absorbed in muscles passively stretched to 20% of failure force, but a nonsignificant decrease in the same parameters in muscle stretched to 30% of force to failure. Shrier (1999) has suggested that stretching prior to exercise may in fact

INTRODUCTION Despite the widespread use of stretching prior to exercise as an injury prevention measure, there is still no conclusive scientific evidence to support this practice. Muscle strain injury is cited as the most frequent sporting injury (Glick 1980; Salter 1983) and is characterized by a partial or complete tear of the muscle–tendon unit. Clinically, a muscle strain injury is an acute event, with focal pain and swelling (Noonan & Garrett 1992). Epidemiological research has cited reduced flexibility as an aetiological factor in acute muscle strain injury (Beaulieu 1981; Ekstrand & Gillquist 1982; Agre 1985; van Mechelen et al. 1992) concluding that improving flexibility by stretching may reduce injury risk. Hence the scientific Received: 11 February 2002 Revised: 17 October 2002 Accepted: 15 January 2003 Sarah M. Weldon BSc (Hons) Ost., Osteopath, Private Practice, Russell H. Hill BSc (Hons) Ost., Osteopath, Private Practice, London, UK. Correspondence to: SMW, 12 Titchwell Road, London SW18 3LW, UK. Tel.: +44 (0)20 8870 9631; Fax: +44 (0)20 7207 4868; E-mail: [email protected] 141

142 Manual Therapy

increase injury risk, citing an in vitro study by Noonan et al. (1993) in which the increase in compliance seen when muscles were warmed to 401C was associated with a reduction in their energy-absorbing capabilities, although the authors interpreted this as a protective effect. The majority of muscle strain injuries occur when muscles are active and functioning in an eccentric manner (Ciullo & Zarins 1983; Noonan & Garrett 1992). The ability of a muscle to absorb energy is dependent on both the active (contractile) components and its passive (connective tissue) components (Safran et al. 1989) and is significantly greater in active as opposed to passive muscle activity (Garrett 1987). A critical point is reached when the muscle is unable to prevent excessive sarcomere lengthening and the actin–myosin filaments are stretched beyond overlap (Morgan 1990). The ability of an active muscle to resist lengthening and hence injury, is therefore largely dependent on contractile strength and is substantially reduced when a muscle is fatigued (Mair et al. 1996; Safran et al. 1989). It is therefore the compliance of active muscle that is most relevant when looking at inury risk, which bears little relation to passive compliance, except at the extremes of stretch (Hawkins & Bey 1997). Whether passive stretching can influence the compliance of active muscle has been questioned (Shrier 1999). Various authors have investigated stretching as an injury prevention measure (Kerner & D’Amico 1983; Howell 1984; Jacobs & Berson 1986; Blair et al. 1987; Macera et al. 1989; Brunet et al. 1990). A number of reviews of the stretching literature exist (Shellock & Prentice 1985; Safran et al. 1989; Wilkinson 1992; Smith 1994), in which authors advocate stretching as an important part of an injury prevention programme, although these conclusions are not based on any clinical evidence. The poor scientific quality of such ‘narrative’ or ‘unsystematic’ literature reviews was highlighted by Mulrow (1987). In as much as primary research takes steps to avoid bias and random error, so too should the review article. In other words, a review should be subject to the same standards of scientific rigour as primary research. A recent systematic review by Shrier (1999) concluded that pre-exercise stretching did not reduce the incidence of local muscle injury. However, the cross-sectional design of five of the articles that concluded that stretching prior to exercise did not reduce injury meant it was impossible to determine whether subjects stretched before injury or because of injury (Kerner & D’Amico 1983; Howell 1984; Jacobs & Berson 1986; Blair et al. 1987; Brunet et al. 1990). In the cohort study that failed to show any evidence of effect (Macera et al. 1989) the authors failed to control for previous injury and weekly running distance in their analysis, both variables having been shown to be significant predictors of injury risk r 2003 Elsevier Science Ltd. All rights reserved.

(Ekstrand & Gillquist 1983; Blair et al. 1987; Brunet 1990). Although the randomized clinical trial (RCT) is considered to be the gold standard by which we judge the benefits of therapy (Riegelman & Hirsch 1996; Greenhalgh 1997), flaws in their design and conduct can result in overestimation or underestimation of effect, leading to false-positive or false-negative conclusions (van der Heijden et al. 1995). Two RCTs (Ekstrand et al. 1983; van Mechelen et al. 1993) and one controlled clinical trial (CCT) (Bixler & Jones 1992) were included in the review by Shrier (1999), but there was no qualitative analysis of study quality. Therefore a systematic review of published RCTs and CCTs on stretching for injury prevention is presented, with all trials scored for methodological quality and their results interpreted in light of the quality scores thereof.

METHOD Relevant studies were retrieved by means of a computer-aided literature search using MEDLINE, EMBASE, AMED, SPORT Discus, CINAHL and SIGLE databases, using the following Medical Subject Headings terms or text words: stretch, injury, clinical trial, controlled trial, muscles, sport, exercise. In addition, references given in the studies retrieved were further examined and key journals were handsearched for any relevant studies not recovered by other methods. To be included in the review, studies had to meet the following criteria: (1) randomized clinical trials (RCTs) or controlled clinical trials (CCTs) investigating stretching as an injury prevention measure (additional interventions were allowed); (2) study published from 1970 onwards; (3) abstracts and unpublished studies were excluded. Randomized clinical trials are considered the gold standard by which the benefits of therapy are judged (Riegelman & Hirsch 1996; Greenhalgh 1997); potential selection bias and confounding making it impossible to draw conclusive determinations of efficacy using non-randomized controlled clinical trials. However, where trials of this quality are lacking, it has been suggested that it would be foolish to ignore the potential for gaining information from other sources (de Bie 1996). Therefore, due to the paucity of relevant RCTs, relevant CCTs were included in this review in agreement with the method guidelines for systematic reviews (van Tulder et al. 1997). As such, all RCTs and CCTs meeting the criteria were retrieved. The quality of design and conduct of the selected studies were assessed using a modification of the method guidelines for systematic reviews by van Tulder et al. (1997) and the criteria used by Koes in Manual Therapy (2003) 8(3), 141–150

Efficacy of stretching 143

his study on the efficacy of spinal manipulation and mobilization (Koes et al. 1991). These criteria are based on generally accepted principles of intervention research as used by Ter Riet et al. (1990). These methodological principles are grouped into four categories; study population; interventions; effect; and data analysis. These four categories comprise 14 criteria (Table 1, A–N), which have been further divided to create a 41 item checklist (Appendix A). Each checklist item is weighted numerically, with a maximum score of 100 possible. The weights given to the criteria were arbitrarily defined but believed to reflect their relative importance for validity and precision. Higher weighting was given to those checklist items (C, I, K, L, M) considered as minimum criteria by the Editorial Board of the Cochrane Back Review Group (van Tulder et al. 1997), although empirical evidence of an association with bias (Schulz et al. 1995; Altman 1999) only exists for two of these criteria (C, I). All studies were subjected to the same methodological assessment, but the scores and results of the controlled clinical trials were tabulated separately in view of the caution with which their results should be interpreted. All trials were then scored by the authors independently, resulting in a hierarchical list according to methodological quality. Where disagreements in scoring occurred, these were solved by consensus. Trials were deemed positive if the authors concluded that stretching resulted in a reduction in injury risk. A study was deemed to be negative if the authors concluded that stretching failed to reduce injury risk, or increased injury risk.

Table 1. Criteria list for a methodological assessment of clinical trials of stretching for preventing injury Criterionn

Weighting

Study population A. Homogeneity B. Comparability of relevant baseline characteristics C. Randomization procedure D. >100 subjects in smallest group; >200 subjects in smallest group

30 2 12 6 10

Intervention E. Stretching procedure explicitly described F. Reference procedure explicitly described G. Co-interventions avoided or comparable H. Compliance reported and acceptable

20 5 5 5 5

Effect I. Assessor blinded J. Relevant outcome measures K. Drop outs described for each study group separately L. Attrition rate acceptable

35 10 12 5

Data Analysis M. Intention-to-treat analysis N. Frequencies presented for each group

15 10 5

Manual Therapy (2003) 8(3), 141–150

8

RESULTS Seven studies met the inclusion criteria, four randomized clinical trials and three controlled clinical trials. The RCTs ranged in quality from 12 to 68, of a possible 100 points. The CCTs achieved quality scores ranging from 16 to 30, of a possible 100 points. There were three negative RCTs, and only one positive RCT. All three CCTs were positive. Table 2 presents these studies arranged in hierarchical order based on their methodological scores. Only two studies (Pope et al. 1998, 1999) scored more than 50 points, indicating the poor overall methodological quality of most of the studies. To allow comparison between methodological quality of RCTs and CCTs, the same scoring system was used for both trial designs. However, it should be noted that by definition a CCT is unable to fulfil the criterion relating to random treatment allocation (criterion C). Common methodological flaws amongst RCTs concerned incomparability of subjects at baseline (criterion B), inadequate treatment allocation (criterion C), inadequate description of reference procedure (criterion F), failure to avoid co-interventions (criterion G) and failure to blind the assessor (criterion I). Common methodological flaws amongst CCTs concerned incomparability of subjects at baseline (criterion B), inadequate treatment allocation (criterion C), failure to describe dropouts (criterion K), high attrition rate (criterion L) and lack of intentionto-treat analysis (criterion M). Despite an attempt at randomization by two authors (Bixler & Jones 1992, Hartig & Henderson 1999), the pseudo-random procedure used for treatment allocation is not a reliable method of eliminating selection bias and therefore these studies are classified as CCTs. Despite some incomplete information, the studies were generally methodologically sound in the areas of homogeneity (criterion A), description of stretching procedure (criteria E), avoidance of co-interventions (criterion G) and adequate data presentation (criteria N). A sensitivity analysis of the checklist and the distribution of weights was undertaken, as utilized by van der Heijden et al. (1995), by recalculation of the weighted and unweighted method scores for the 14 criteria (A–N) and the 41 checklist items. The results of these recalculations, presented in Table 3, revealed no change in the hierarchical order of the studies. These recalculations show the robustness of the scoring system. Table 4 presents a summary of the four RCTs in hierarchical order. Two studies received method scores that exceeded 50 points (Pope et al. 1998, 1999). The three highest scoring RCTs showed no significant reduction in injury risk in subjects following a stretching programme. Protocols varied widely r 2003 Elsevier Science Ltd. All rights reserved.

144 Manual Therapy

Table 2. Clinical trials of the efficacy of stretching for prevention of injury in order of methods score Scores for methods criteria Authors

A 2

B 12

C 6

D 10

E 5

F 5

G 5

H 5

I 10

J 12

K 5

L 8

M 10

N 5

Total 100

Author’s conclusions

Randomized clinical trials Pope et al. (1999) Pope et al. (1998) Van Mechelen et al. (1993) Ekstrand et al. (1983)

1 1 1 2

4 4 6 2

3 3 0 0

10 10 10 0

5 4 4 1

4 4 0 0

5 5 0 0

0 0 2 0

10 0 0 0

6 4 4 4

5 5 2 0

0 0 8 0

1 10 0 0

5 5 5 3

68 55 42 12

Negative Negative Negative Positive

Controlled clinical trials Hartig and Henderson (1999) Cross and Worrell (1999) Bixler and Jones (1992)

1 1 1

2 2 2

0 0 0

5 5 0

5 4 4

1 2 0

5 5 0

0 0 2

0 0 0

6 4 4

0 0 0

0 0 0

0 0 0

5 3 3

30 26 16

Positive Positive Positive

Table 3. Sensitivity analysis of scoring system Weighted Scores (maximum score=100 points) Authors

Unweighted scores (maximum score=14 criteria)

Unweighted method (maximum score=41 items)

Rank

% Score

Rank

Score

% Score

Rank

Score

% Score

Randomized clinical trials Pope et al. (1999) Pope et al.(1998) van Mechelen et al.(1993) Ekstrand et al. (1983)

1 2 3 4

68 55 42 12

1 2 3 4

11 11 9 5

79 79 64 36

1 2 3 4

23 20 16 7

56 49 39 17

Controlled clinical trials Hartig and Henderson (1999) Cross and Worrell (1999) Bixler and Jones (1992)

1 2 3

30 26 16

1 2 3

8 8 6

57 57 43

1 2 3

14 13 10

34 32 24

across studies, both in duration of stretch and number of sessions. Pope et al. (1999) showed no significant effect of pre-exercise stretching on all-injuries risk (Hazard ratio=0.95, 95% CI 0.77–1.18) or soft tissue injuries (HR=0.83, 95% CI 0.63–1.09). Multivariate analysis showed fitness as a significant predictive indicator for injury (Po0.001). In the 1998 study by Pope et al., flexibility was shown to be a significant predictor of injury risk (LR=4.97; df=1; P=0.03) in agreement with other authors (Seto cited in Hartig & Henderson 1999). No significant difference in incidence of injury was found between groups, although a small but clinically significant reduction in injury risk could not be ruled out due to low statistical power. In the home-based stretching programme evaluated by van Mechelen et al. (1993), attrition rate was high (22.3%) and compliance with the prescribed programme was low (46.6%). Results were further complicated by differences in data collection methods between control and treatment group. In the only positive RCT (Ekstrand et al. 1983), the multi-faceted nature of the prophylactic programme and lack of similarity between control and intervention procedures made separate analysis of stretching effect impossible. Table 5 presents a summary of the three CCTs in hierarchical order. All CCTs showed a significant reduction in rates of injury in the intervention group. r 2003 Elsevier Science Ltd. All rights reserved.

Hartig and Henderson (1999), who utilized the greatest total stretching stimulus of all the studies, showed a significant (P=0.02) reduction in lower extremity injuries in the intervention group (RR= 0.63, 95% CI 0.41–0.99). Relative risk could not be calculated for either of the two lowest scoring CCTs (Bixler & Jones 1992; Cross & Worrell 1999) due to lack of information regarding changes in exposure. To assess whether the quality of published trials has increased over the past decades, a graph of methodological quality scores (after Koes et al. 1995) against the year of publication was plotted (Fig. 1). A linear trend line shows that there has been a gradual increase in the quality of RCTs and CCTs over the past two decades. However, overall the quality of the studies is still poor, there being only two studies in the past decade that attained quality scores of more than 50 points.

DISCUSSION Due to the heterogeneity of the studies reviewed no meta-analysis was undertaken, however a vote count was performed in accordance with the recommendations in the method guidelines for systematic reviews (van Tulder et al. 1997). A vote count of positive and negative RCTs would suggest that stretching does not reduce the incidence of injury, there being three Manual Therapy (2003) 8(3), 141–150


Table 4. Randomised clinical trials of the efficacy of stretching for preventing injury Author

Intervention

Control

Pope et al. (1999)

One 20 s static stretch for each of 6 lower extremity muscle groups during warm-up. 40 sessions over 12 weeks.

Warm-up only.

Pope et al. (1998)

Two 20 s static stretches for each of their soleus and gastrocnemius muscles during warm-up.

Two 20 s static stretches for wrist flexors and triceps muscles during warm-up.

Van Mechelen et al. (1993)

Ekstrand et al. (1983)

11 week programme of stretching prior to intense physical activity Three 10 s static stretches for each of five lower extremity muscle groups following warm-up. Cooldown also performed. Twice daily for 16 weeks. 10 min of contract-relax stretching for five lower extremity muscle groups performed as part of warm-up. Stretching performed as part of a seven-part prophylactic programme. Stretches performed prior to every practice session or game over 6 months.

Subjects continued with normal programme of stretching, warm-up and cool-down. Subjects continued with usual activities.

Results

Quality score

No significant effect of pre-exercise stretching on all-injuries risk (HR=0.95, 95% CI 0.77–1.18), soft-tissue injury risk (HR=0.83, 95% CI 0.63–1.09) or bone injury risk (HR=1.22, 95% CI 0.86–1.76). No significant effect of pre-exercise stretching on all-injuries risk (HR=0.92, 95% CI 0.52–0.61).

68

No significant difference (P>0.05) in injury incidence between groups. Relative risk for injury was 1.12 (95% CI 0.56–2.72). A significant difference in injury incidence between groups was found (Po0.001). Incidence of 0.6 injuries per month in the stretch group compared with 2.6 injuries per month in the control group.

42

55

12

Table 5. Controlled clinical trials of the efficacy of stretching for preventing injury Author Hartig and Henderson (1999)

Cross and Worrell (1999)

Bixler and Jones (1992)

Intervention Five 30 s static stretches for the hamstrings performed three times daily in addition to normal pre-exercise stretching. 13 week programme. Three 15 s stretches for each of four lower extremity muscle groups as part of pre-practice schedule. Stretches performed as part of practice schedule over course of a year. 90 s static stretching routine performed following 90 s warm-up at the end of half-time.

Control

Results

Quality score

Normal pre-exercise stretching.

Subjects in the intervention group had significantly fewer lower extremity overuse injuries (P=0.02).

30

General pre-practice stretching for upper and lower extremities.

A significantly lower incidence of injury was found in the intervention group (Po0.05).

26

Normal half-time activities

Subjects in the intervention group had significantly lower thirdquarter sprains and strains per game (Po0.05).

16

Programme lasted one football season.

negative RCTs (75%) and only one positive RCT (25%). All three CCTs (100%) concluded that stretching did reduce injury risk, but due to the weaker trial design, less emphasis is placed on these results. It is of note that the four positive trials also received the lowest methodological quality scores. Manual Therapy (2003) 8(3), 141–150

However, de Bie (1996) emphasises that vote counting may result in small but clinically important effects being overlooked, particularly if studies with statistically non-significant results are counted as negative. The two highest scoring RCTs (Pope et al. 1998, 1999) both showed a small but non-significant r 2003 Elsevier Science Ltd. All rights reserved.

146 Manual Therapy

80

Quality Scores

60

40

20

0 1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Year of Publication

Fig. 1—Relationship between quality scores and year of publicaton: m, RCTs; ’, CCTs; —, Linear (RCTs); —, Linear (CCTs).

reduction in all-injuries risk. In the 1999 study by Pope et al., statistical power was sufficient to rule out a clinically useful reduction in injury risk, although the authors reduced the stretching duration to one 20 s stretch per muscle group, despite the previous study having shown no significant reduction in injury risk with greater total stretch stimulus. The best estimate of the effect of pre-exercise stretching was that it reduces all-injuries risk by 5%, with a 23% or greater reduction in all-injuries risk being ruled out with 95% confidence (HR=0.95, 95% CI 0.77–1.18). When this is expressed in absolute terms, the authors calculated that on the basis of the protocol used subjects would need to stretch for an average of 260 h to prevent one injury. It should be noted however, that in both the 1998 study by Pope et al. (RR=0.76, 95% CI 0.37–1.54) and the 1999 study (HR=0.83, 95% CI 0.63–1.09) there was a greater, but still statistically insignificant, reduction in risk of softtissue injuries. This translates to a best estimate of a 17% reduction in risk of soft-tissue injuries. When expressed in absolute terms, based on the protocol utilized (Pope et al. 1999), subjects would have to stretch for an average of 212 h to prevent one softtissue injury. On this basis it would seem that the practice of pre-exercise stretching for reducing injury is of limited clinical significance. Although the protocol used was based on established methods of stretching (St George 1989) that result in significant increases in flexibility, it has been shown (Halbertsma et al. 1996; Magnusson et al. 1996) that typically advocated protocols only produce a transient change in viscoelastic properties, maintained increases in flexibility being attributed to an increased stretch tolerance. Nevertheless, even transient increases in biomechanical properties of muscle should theoretically result in a reduction in injury for the duration of that effect (Glick 1980; Ciullo & Zarins 1983; Noonan & Garrett 1992; Gleim r 2003 Elsevier Science Ltd. All rights reserved.

& McHugh 1997). As ethical reasons prevent the testing of failure properties of human skeletal muscle, research findings are drawn from the animal model in which controlled strain injury has been extensively studied (Gleim & McHugh 1997). However, although some research on the effect of stretching on passive failure properties exists (Best et al. 1989), as yet the effect of passive stretching on active failure properties remains undocumented. As the majority of injuries occur when muscle is active and eccentrically contracting (Noonan & Garrett 1992), it is in this area that further research is required. It is however feasible that passive stretching of sufficient duration could reduce injury risk. In vitro research (Goldspink et al. 1995; Yang et al. 1997) has demonstrated hypertrophy of muscle fibres and an increase in the expression of Insulin-like Growth Factor (IGF-1), following prolonged immobilization in a lengthened position. Although no evidence exists as to whether similar effects can be achieved in vivo, if similar hypertrophy can be achieved following passive stretching, it is reasonable to assume that this would result in an increase in energy absorption and reduction in injury risk. Indeed, the CCT by Hartig and Henderson (1999) in which subjects stretched four times daily, amounting to a total stretching duration of greater than 40,950 s per muscle group (11.375 h), showed a significant reduction (P=0.02) in incidence of lower extremity overuse injuries. Although the protocol utilized by van Mechelen et al. (1993) was also of significant duration (6720 s per muscle group, 1.867 h), compliance with the prescribed procedure was extremely low (47%), making evaluation of efficacy difficult. In one of the negative RCTs (van Mechelen et al. 1993), subjects who stretched were actually found to be at higher risk of injury, in agreement with previous authors (Kerner & D’Amico 1983; Howell 1984; Jacobs & Berson 1986). Basic science evidence would Manual Therapy (2003) 8(3), 141–150


indicate that injudicious stretching might indeed increase risk of injury. Noonan et al. (1994) showed a significant decrease in maximal contractile force in muscles passively stretched to 30% of force to failure, accompanied by histologic evidence of focal areas of muscle fibre rupture and haemorrhage near the distal musculo-tendinous junction. Hasselman et al. (1995) also found a significant reduction in maximal contractile force in muscles actively stretched to 70% of passive force to failure. Similar reductions in maximal contractile force, seen following repeated muscular contractions, have been attributed to fatigue (Taylor et al. 1993). Fatigue has been shown to significantly reduce the ability of muscle to absorb energy (Mair et al. 1996) and is widely believed to be a predisposing factor in muscle injury (van Mechelen et al. 1992). In two of the positive trials (Ekstrand et al. 1983; Bixler & Jones 1992) the stretching procedure was performed as part of a routine which included warmup and therefore efficacy of the stretching procedure itself could not be evaluated. In vitro studies by Strickler et al. (1990) demonstrated significant increases in force and length to failure following passive warming of muscle, as well as a non-significant increase in energy absorbed prior to failure. Safran et al. (1988) showed significant (Po0.01) increases in length and force to failure following submaximal contraction, implying that a protective effect may be gained from the warmup procedure. However, it is important to ensure warmup intensity is not too vigorous as a 37% reduction in peak contractile force has been shown in vitro (Taylor et al. 1997) following rapid isometric contractions. Alternatively, PNF stretching incorporating a submaximal contraction (Ekstrand et al. 1983) may be more appropriate as an injury prevention measure. It has been shown that individuals with reduced flexibility are at significantly higher risk of injury (Ekstrand & Gillquist 1983; Worrell et al. 1991; Pope et al. 1998; Seto cited in Hartig & Henderson 1999). A significant correlation has been shown between static flexibility and stiffness of the musculature (Wilson et al. 1991). In a compliant system forces will be absorbed and dampened by the musculotendinous junction, thereby reducing trauma to muscle fibres. However, where there is high stiffness, forces will be transferred to the contractile apparatus with little reduction in force. This provides a mechanism to explain the noted association between reduced flexibility and occurrence of injury. Evidence for this is found in the in vivo study by McHugh et al. (1999), which found increased evidence of muscle damage following eccentric exercise in subjects with greater passive stiffness. It is suggested (Armstrong et al. 1991) that such damage occurs on the descending limb of the length-tension curve, in which Manual Therapy (2003) 8(3), 141–150

a decrease in active force and an increase in passive force is seen during eccentric exercise. This is consistent with the work by Hawkins and Bey (1997) which showed that in the outer ranges of movement as tendon stiffness increases, greater passive forces are generated within the muscle. In people with stiffer musculature even greater passive muscle forces would be expected to develop which would therefore increase their risk of muscle injury. A decrease in passive stiffness, as well as an increase in energy absorbing capabilities, would therefore seem necessary to reduce injury risk. Although Magnusson et al. (1997) found that endurance athletes with reduced flexibility had significantly (Po0.01) stiffer hamstring muscles, viscoelastic stress relaxation was similar in both tight and normal subjects. This is consistent with the findings by Pope et al. (1998) that there was no difference in the flexibility gains between tight and normal subjects. It would therefore appear that passive muscle stiffness is a function of the number of cross-links or collagen content, which are unlikely to be affected by non-disruptive stretching. It is proposed that in order to effect a change in normal, but short, healthy tissue it is therefore necessary to initiate the plastic change only achieved by creating minor muscle damage (Lederman 1997). From the basic science literature reviewed it is obvious that such stretching performed prior to exercise has the potential to increase the risk of injury. Evidence would therefore suggest that if stretching is performed as part of a warmup procedure it is necessary to avoid the potential damage or fatigue that can be caused by overstretching. The problem exists as to how to quantify the point at which a muscle enters the plastic region. Typically advocated stretching protocols define the end point as a point creating a sensation of stretch, not pain (Hartig & Henderson 1999, Pope et al. 1999). However, stretching has been shown to increase the pain threshold (Halbertsma et al. 1996; Magnusson et al. 1996) and may therefore adversely affect the ability to perceive the point at which damage occurs. PNF stretching, although potentially offering greater injury preventing effects, should also be used with caution due to the increased possibility for overstretching when enlisting the help of a third party. Determination of a reliable and pragmatic method of achieving the desired end point requires further research. There is some clinical evidence to suggest that prolonged stretching performed outside of the preexercise period can reduce incidence of injury (Hartig & Henderson 1999). Prolonged low load stretching in animals has been shown to increase muscle length and hypertrophy (Goldspink et al. 1995; Lederman 1997; Yang et al. 1997) as well as permanently lengthening connective tissue (Sapega et al. 1981). Similar effects have been achieved in human subjects r 2003 Elsevier Science Ltd. All rights reserved.

148 Manual Therapy

with osteoarthritic hips (Leivseth et al. 1989) and joint contractures (Wessling et al. 1987). Evidence of similar effects in healthy subjects with foreshortened muscles has yet to be documented. It is apparent that a need exists for carefully designed and conducted RCTs in this field if we are to make informed and unbiased decisions as to whether stretching can reduce injury risk and further our understanding of the mechanisms involved. The development by the Cochrane Collaboration of a database of RCTs in the area of physical therapy (Newham 1995) will hopefully encourage researchers in the field of complementary medicine to recognize the need for quality trials if our sphere of medicine is to continue gaining credibility amongst orthodox practitioners. With further evidence from highquality RCTs it will be possible to implement an effective injury prevention programme founded on valid and reliable evidence-based research.

CONCLUSION No definitive conclusions could be drawn as to whether stretching reduces the incidence of exerciserelated injury due to the heterogeneity and poor quality of the selected studies. A need exists for carefully controlled clinical trials of sufficient power to identify a clinically significant effect, and with much more attention paid to the proper design and conduct of such studies. Available evidence would suggest that pre-exercise stretching may increase the risk of injury. However, basic science and preliminary clinical evidence would indicate that prolonged stretching in the post-exercise period may increase the energy absorbing capabilities of muscle thereby reducing the risk of injury. Further research is required to clarify these findings, although a rethinking of current practices are indicated. References Agre JC 1985 Hamstring injuries: proposed aetiological factors, prevention and treatment. Sports Medicine 2: 21–33 Altman DG 1999 What randomised trials and systematic reviews can offer decision makers. Hormone Research 51 (Suppl 1): 36–43 Armstrong RB, Warren GL, Warren JA 1991 Mechanisms of exercise-induced muscle fibre injury. Sports Medicine 12: 184–207 Beaulieu JE 1981 Developing a stretching program. Physician & Sportsmedicine 9: 59–69 Best TM, Glisson RR, Seaber AV, Garrett Jr, WE 1989 The response of muscle–tendon units of varying architecture to cyclic passive stretching. Transactions of the Orthopaedic Research Society 14: 294 Bixler B, Jones RL 1992 High-school football injuries: effects of a post-halftime warm-up and stretching routine. Family Practice Research Journal 12(2): 131–139 Blair SN, Kohl HW III, Goodyear NN 1987 Relative risks for running and exercise injuries: Studies in three populations. Research Quarterly 58: 221–228 r 2003 Elsevier Science Ltd. All rights reserved.

Brunet ME, Cook SD, Brinker MR, Dickinson JA 1990 A survey of running injuries in 1505 competitive and recreational runners. Journal of Sports Medicine & Physical Fitness 30: 307–315 Ciullo JV, Zarins B 1983 Biomechanics of the musculotendinous unit: Relation to athletic performance and injury. Clinics in Sports Medicine 2(1): 71–86 Cross KM, Worrell TW 1999 Effects of a static stretching program on the incidence of lower extremity musculotendinous strains. Journal of Athletic Training 34(1): 11–14 de Bie RA 1996 Methodology of systematic reviews: An introduction. Physical Therapy Review 1: 47–51 Ekstrand J, Gillquist J 1982 The frequency of muscle tightness and injury in soccer players. American Journal of Sports Medicine 10(2): 75–78 Ekstrand J, Gillquist J 1983 The avoidability of soccer injuries. International Journal of Sports Medicine 4: 124–128 Ekstrand J, Gillquist J, Liljedahl S-O 1983 Prevention of soccer injuries: Supervision by doctor and physiotherapist. American Journal of Sports Medicine 11(3): 116–120 Gajdosik RL 1991 Effects of static stretching on the maximal length and resistance to passive stretch of short hamstring muscles. Journal of Orthopaedic and Sports Physical Therapy 14(6): 250–255 Garrett Jr, WE, Safran MR, Seaber AV, Glisson RR, Ribbeck BM 1987 Biomechanical comparison of stimulated and nonstimulated skeletal muscle pulled to failure. American Journal of Sports Medicine 15: 448–454 Garrett Jr, WE 1996 Muscle strain injuries. American Journal of Sports Medicine 24: S2–S8 Gleim GW, McHugh MP 1997 Flexibility and its effect on sports injury and performance. Sports Medicine 24(5): 289–299 Glick JM 1980 Muscle strains: Prevention and treatment. Physician and Sportsmedicine 8(11): 73–77 Goldspink DF, Cox VM, Smith SK, Eaves LA, Osbaldeston NJ, Lee DM, Mantle D 1995 Muscle growth in response to mechanical stimuli. American Journal of Physiology 268: E288–E297 Greenhalgh T 1997 How to read a paper: The Basics of Evidence Based Medicine, 2nd edn. BMJ Publishing Group, London . Halbertsma JP, van Bolhuis AI, Goeken LNH 1996 Sport stretching: Effect on passive muscle stiffness of short hamstrings. Archives of Physical Medicine and Rehabilitation 77: 688–692 Hartig DE, Henderson JM 1999 Increasing hamstring flexibility decreases lower extremity overuse injuries in military basic trainees. American Journal of Sports Medicine 27(2): 173–176 Hasselman CT, Best TM, Seaber AV, Garrett WE 1995 A threshold and continuum of injury during active stretch of rabbit skeletal muscle. American Journal of Sports Medicine 23(1): 65–73 Hawkins D, Bey M 1997 Muscle and tendon force-length properties and their interactions in vivo. Journal of Biomechanics 30(1): 63–70 Howell DW 1984 Musculoskeletal profile and incidence of musculoskeletal injuries in lightweight women rowers. American Journal of Sports Medicine 12: 278–282 Jacobs SJ, Berson BL 1986 Injuries to runners: a study of entrants to a 10 000 meter race. American Journal of Sports Medicine 14: 151–155 Kerner JA, D’Amico JC 1983 A statistical analysis of a group of runners. Journal of the American Podiatry Association 73: 160–164 Koes BW, Assendelft WJJ, van der Heijden GJMG, Bouter LM, Knipschild PG 1991 Spinal manipulation and mobilisation for back and neck pain: a blinded review. British Medical Journal 303: 1298–1303 Koes BW, Bouter LM, van der Heijden GJMG 1995 Methodological quality of randomized clinical trials on treatment efficacy in low back pain. Spine 20(2): 228–235 Lederman E. 1997 Fundamentals of Manual Therapy: Physiology, Neurology and Psychology. Churchill Livingstone, London Leivseth G, Torstensson J, Reikeras O 1989 The effect of passive muscle stretching in osteoarthritis of the hip. Clinical Science 76: 113–117 Manual Therapy (2003) 8(3), 141–150


McHugh MP, Magnusson SP, Gleim GW, Nicholas JA 1992 Viscoelastic stress relaxation in human skeletal muscle. Medicine & Science in Sports & Exercise 24(12): 1375–1382 McHugh MP, Connolly DAJ, Eston RG, Kremenic IJ, Nicholas SJ, Gleim GW 1999 The role of passive muscle stiffness in symptoms of exercise-induced muscle damage. American Journal of Sports Medicine 27(5): 594–599 Macera CA, Pate RP, Powell KE, Jackson KL, Kendrick JS, Craven TE 1989 Predicting lower extremity injuries among habitual runners. Archives of Internal Medicine 149: 2565–2568 Magnusson SP, Simonsen EB, Aagaard P, Dyhre-Poulson P, Malachy P, McHugh MA, Kjær M 1996 Mechanical and physiological responses to stretching with and without preisometric contraction in human skeletal muscle. Archives of Physical Medicine and Rehabilitation 77: 373–378 Magnusson SP, Simonsen EB, Aagaard P, Boesen J, Johannsen F, Kjaer M 1997 Determinants of musculoskeletal flexibility: Viscoelastic properties, cross-sectional area, EMG and stretch tolerance. Scandinavian Journal of Medicine and Science in Sports 7: 195–202 Mair SD, Seaber AV, Glisson RR, Garrett WE 1996 The role of fatigue in susceptibility to muscle strain injury. American Journal of Sports Medicine 24(2): 137–143 Morgan DL 1990 New insights into the behavior of muscle during active lengthening. Journal of Biophysiology 57(2): 209–221 Mulrow CD 1987 The medical review article: state of the science. Annals of Internal Medicine 106(3): 485–488 Newham D 1995 The Cochrane Collaboration. What is it, how does it work, and what has it to do with physiotherapy? Physiotherapy 81(7): 405–407 Noonan TJ, Garrett Jr, WE 1992 Injuries at the myotendinous Junction. Clinics in Sports Medicine 11(4): 783–806 Noonan TJ, Best TM, Seaber AV, Garrett Jr, WE 1993 Thermal effects on skeletal muscle tensile behaviour. American Journal of Sports Medicine 21(4): 517–522 Noonan TJ, Best TM, Seaber AV, Garrett WE 1994 Identification of a threshold for skeletal muscle injury. American Journal of Sports Medicine 22(2): 257–261 Pope RP, Herbert RD, Kirwan JD 1998 Effects of ankle dorsiflexion range and pre-exercise calf muscle stretching on injury risk in army recruits. Australian Physiotherapy 44(3): 165–177 Pope RP, Herbert RD, Kirwan JD, Graham BJ 1999 A randomized trial of pre-exercise stretching for prevention of lower-limb injury. Medicine & Science in Sports & Exercise 32(2): 271–277 Riegelman RK, Hirsch RP 1996 Studying a Study and Testing a Test: How to Read the Health Science Literature. Little, Brown & Co., Boston Safran MR, Garrett WE, Seaber AV, Glisson RR, Ribbeck BM 1988 The role of warm-up in muscular injury prevention. American Journal of Sports Medicine 16(2): 123–129 Safran MR, Seaber AV, Garrett Jr WE 1989. Warm-up and muscular injury prevention: an update. Sports Medicine 8(4): 239–249 Salter RB 1983. Textbook of disorders and injuries of the musculoskeletal system. Williams & Wilkins, Baltimore Sapega A, Quedenfeld T, Moyer R, Butler R 1981 Biophysical factors in range-of-motion exercise. Physician and Sportsmedicine 9: 57–65 Schulz KF, Chalmers I, Hayes RJ, Altman DG 1995 Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. Journal of the American Medical Association 273(5): 408–412 Shellock FG, Prentice WE 1985 Warming-up and stretching for improved physical performance and prevention of sportsrelated injuries. Sports Medicine 2: 267–278 Shrier I 1999 Stretching before exercise does not reduce the risk of local muscle injury: A critical review of the clinical and basic science literature. Clinical Journal of Sports Medicine 9: 221–227 Smith CA 1994 The warm-up procedure: to stretch or not to stretch. a brief review. Journal of Orthopaedic and Sports Physical Therapy 19(1): 12–17 St. George F 1989 The muscle fitness book. Simon & Schuster, Sydney Manual Therapy (2003) 8(3), 141–150

Strickler T, Malone T, Garrett WE 1990 The effects of passive warming on muscle injury. American Journal of Sports Medicine 18(2): 141–145 Taylor DC, Dalton JD, Seaber AV, Garrett Jr, WE 1990 Viscoelastic properties of muscle-tendon units: The biomechanical effects of stretching. American Journal of Sports Medicine 18(3): 300–309 Taylor DC, Dalton JD, Seaber AV, Garrett WE 1993 Experimental muscle strain injury. Early functional and structural deficits and the increased risk for reinjury. American Journal of Sports Medicine 21(2): 190–194. Taylor DC, Brooks DE, Ryan JB 1997 Viscoelastic characteristics of muscle: Passive stretching versus muscular contractions. Medicine & Science in Sports & Exercise 29(12): 1619–1624 Ter Riet G, Kleijnen J, Knipschild P 1990 Acupuncture and chronic pain: A criteria-based meta-analysis. Journal of Clinical Epidemiology 43: 1191–1199 van der Heijden GJMG, Beurskens AJHM, Koes BW, Assendelft WJJ, de Vet HCW, Bouter LM 1995 The efficacy of traction for back and neck pain: A systematic, blinded review of randomized clinical trial methods. Physical Therapy 75(2): 93–104 van Mechelen W, Hlobil H, Kemper HCG 1992 Incidence, severity, aetiology and prevention of sports injuries. Sports Medicine 14(2): 82–99 van Mechelen W, Hlobil H, Kemper HCG, Voorn WJ, de Jongh R 1993 Prevention of running injuries by warm-up, cool-down, and stretching exercises. American Journal of Sports Medicine 21(5): 711–719 van Tulder MW, Assendelft WJJ, Koes BW, Bouter LM, the Editorial Board of the Cochrane Collaboration Back Review Group 1997 Method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group for spinal disorders. Spine 22(20): 2323–2330 Wessling KC, DeVane DA, Hylton CR 1987 Effects of static stretch versus static stretch and ultrasound combined on triceps surae muscle extensibility in healthy women. Physical Therapy 67(5): 674–679 Wilkinson A 1992 Stretching the truth. A review of the literature on muscle stretching. Australian Physiotherapy 38(4): 283–287 Wilson GJ, Wood GA, Elliott BC 1991 The relationship between stiffness of the musculature and static flexibility: An alternative explanation for the occurrence of muscular injury. International Journal of Sports Medicine 12: 403–407 Worrell TW, Perrin DH, Gansneder BM, Gieck JH 1991 Comparison of isokinetic strength and flexibility measures between hamstring injured and non-injured athletes. Journal of Orthopaedic and Sports Physical Therapy 13(3): 118–125 Yang S, Alnaqeeb M, Simpson H, Goldspink G 1997 Changes in muscle fibre type, muscle mass and IGF-1 gene expression in rabbit skeletal muscle subjected to stretch. Journal of Anatomy 190: 613–622

Appendix A. Checklist for Assignment of Methodological Scores Details of criteria listed in Table 1. Each criterion must be applied independently of the other criteria. A. Description of inclusion and exclusion criteria (1 point). Restriction to a homogenous study population, ie same athletic population, previous injuries (1 point). B. Comparability for age, sex, previous injuries, hours of training, intensity of training and value of outcome measures (2 points each). C. Randomization procedure explicitly described (3 points); randomization procedure which excludes bias—for example, sealed envelopes (3 points). r 2003 Elsevier Science Ltd. All rights reserved.

150 Manual Therapy

D. Smallest group immediately after randomization. E. Stretching procedure explicitly described, ie type of stretch used, description of procedure, duration of stretch, number of repetitions, number of times daily/weekly during experimental period (1 point each). F. Reference procedure explicitly described, ie type of reference activity, description of procedure, duration of procedure, number of repetitions, number of times daily/weekly during experimental period (1 point each). G. Other mobilizing exercises avoided or comparable between groups. H. Compliance with allocated procedure reported (2 points); compliance greater than 90% in each group after randomization (3 points) I. Assessor blind to subjects experimental grouping J. Relevance of outcome measures—occurrence of injury, type of injury, flexibility, passive muscle stiffness, strength, fitness/ gait economy (2 points each).

r 2003 Elsevier Science Ltd. All rights reserved.

K. Number of subjects who withdraw given for each group without reasons for withdrawal (2 point); no dropouts or number of patients for each group with reasons for withdrawal (5 points). L. All randomized subjects minus the number of subjects at the main point of measurement for the most important outcome measure, as a proportion of all randomized subjects. Less than 20% attrition rate in each group (4 points), less than 10% attrition rate in each group (8 points). M. When less than 10% attrition rate: analysis on all randomized subjects for main outcome measure and on the most important points of measurement minus missing values, regardless of noncompliance (10 points). When greater than 10% attrition rate: alternative analysis such as baseline comparison for differences between groups at baseline measurement (10 points). N. Means and standard deviations presented for each group (5 points); means only presented (3 points).

Manual Therapy (2003) 8(3), 141–150


Original article

Psychometric properties of a generic health measure in Chinese patients with low back pain in Hong Kong A. S. L. Leung*, T.-H. Lamw, A. J. Hedleyw, L. T. Twomeyz n

Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hunghom, Kowloon, Hong Kong, Department of Community Medicine and Unit for Behavioural Sciences, The University of Hong Kong, Hong Kong, z The Curtin University of Technology, Western Australia w

SUMMARY. In Hong Kong, the measurement of perceived health status in patients with low back pain (LBP) can be facilitated by the availability of a health profile specifically designed for the Chinese culture. This prospective observational study investigated the psychometric properties of the generic Current Perceived Health 42 (CPH42) Profile in four separate samples (totalling 473) of Chinese patients with LBP in Hong Kong. The patients completed the CPH42 Profile and the Roland LBP Disability Scale at various points in the course of physiotherapy. Their pain intensity was measured using the 11-point pain numerical rating scale (NRS). The test– retest reliability and internal consistency of the CPH42 Profile demonstrated high intra-class correlation coefficient of 0.92 and Cronbach’s alpha of 0.90. Validity was confirmed by a moderate correlation with the Chinese adaptations of the Roland LBP Disability Scale and the NRS at the commencement of physiotherapy (Spearman’s correlation coefficients were 0.48 and 0.42, respectively). The responsiveness, measured from the commencement of physiotherapy to weeks 3 and 6 (standard response means of 0.33 and 0.58, respectively), were commensurate with the respective changes in pain intensity. The psychometric properties of the CPH42 Profile suggest its suitability for use as an outcome instrument in future efficacy studies on LBP intervention. r 2003 Elsevier Science Ltd. All rights reserved.

month prevalence of 39% and 21%, respectively, for adults sampled from two typical housing blocks (Lau et al. 1995). With such a high prevalence in the population, it may not be surprising to find that LBP is one of the most common conditions confronting physiotherapists in out-patient care (Battie et al. 1994; Jette et al. 1994). Typically, over 30% of monthly adult referrals in frontline physiotherapy out-patient departments (POPDs) in Hong Kong have been attributed to LBP (Physiotherapy Management Protocol for Low Back Pain Patients 1996). This figure surpasses those reported in the United States where a quarter of out-patient referrals for physiotherapy were for the treatment of LBP (Battie et al. 1994; Jette et al. 1994; Swanson, 1995). The availability of a valid and reliable outcome measure is essential to facilitate the study of the efficacy on LBP interventions. A number of LBPspecific disability scales have since been developed (Fairbank et al. 1980; Roland & Morris 1983;

INTRODUCTION Low back pain (LBP) is a major health problem in industralized countries (Dodd 1997; Cassidy et al. 1998). It has been estimated that between 70% and 80% of the adult population in the United States and Britain have suffered from LBP at some time in their lives (Frymoyer & Cats-Baril 1991; Walsh et al. 1992). A Hong Kong survey reported a life and 12 Received: 26 June 2002 Revised: 5 November 2002 Accepted: 16 January 2003 Arran S. L. Leung, PhD, Assistant Professor, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Tai-Hing Lam, MD, Professor, Anthony J. Hedley, Professor and Head, Department of Community Medicine and Unit for Behavioural Sciences, The University of Hong Kong, Lance T. Twomey, PhD, Vice-Chancellor, The Curtin University of Technology. Correspondence to: ASL Leung, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hunghom, Kowloon, Hong Kong. Tel.: +852 2766 6722; Fax.: +852 2764 1435. 151

152 Manual Therapy

Greenough & Fraser 1992; Ruta et al. 1994; Kopec et al. 1995). However, the use of generic non-LBPspecific profiles also has a place in measuring perceived health status of people with back pain. These profiles, such as the Sickness Impact Profile (SIP) (Bergner et al. 1981) and the SF36 Health Profile (Ware & Sherbourne 1992; Ware 2000), assess health status across a number of dimensions such as pain, mental health, physical and social functions. Also, the use of generic profiles allows comparison of the impact of perceived health status arising from LBP with other disorders. Their value as LBP outcome measures has also been demonstrated in previous studies (Follick et al. 1985; Lansky et al. 1992). Hence, back pain researchers have advocated the use of a LBP-specific scale, together with a generic profile, as being a more valid package for the assessment of health status in patients with LBP (Bergner et al. 1981; Garratt et al. 1993; Ruta et al. 1994). In Hong Kong, measurement of perceived health status in patients with LBP is hampered by the lack of an instrument rigorously tested for reliability, validity, responsiveness and feasibility, as well as suitable for use in busy clinical environments commonly found in frontline POPDs. Physiotherapists often use Chinese adaptations of existing scales, including the SF-36 Health Profile, in which measurement properties remain to be fully investigated, particularly in the complex cultural context of Hong Kong (Lam et al. 1998). This dilemma may be redressed, to a certain degree, with the development of a generic health profile, specifically designed, developed and validated for Chinese use in Hong Kong and Southern China. The Current Perceived Health 42 Profile (CPH42) (Li & Fielding 1995; Fielding & Li 1997) was developed for this purpose. As this health profile was developed in the same language and culture as the target group of patients, no adaptation is therefore necessary (Beaton et al. 2000). The CPH42 Profile is shown in the Appendix. The reliability and validity of the CPH42 Profile in well and in patient populations had been confirmed in its original report (Li & Fielding 1995). Its potential, however, as a measure of perceived health status in patients with LBP remains to be investigated. The objective of this study was to evaluate the reliability, validity and responsiveness of the CPH42 Profile in samples of Chinese patients with LBP in Hong Kong.

the subject’s status in the previous week. Questions are forced choice, requiring the respondent to choose one out of four possible responses on a Likert scale. Each response is assigned a score (from 0 to 3), which is blind to the respondents. The score of each itemquestion is then summed and converted to a percentage to produce an index score between 0 and 100. Patients who report the lowest possible levels of health status in all item-questions would score 100, and those who report the highest possible level for each item-question would score 0.

METHODS

Part I: Test–retest reliability and internal consistency Consecutive patients, with LBP from three study centres (Centres 1, 2 and 3), were used to examine the test–retest reliability and internal consistency of the CPH42 Profile. When patients first visited the POPDs to make their initial appointment for physiotherapy

The CPH42 Profile is a self-administered questionnaire, which measures health status across pain, physical and social functions, with an emphasis on mental health. The 42 item-questions enquire about r 2003 Elsevier Science Ltd. All rights reserved.

Subjects The patients were recruited from six frontline POPDs in The Hospital Authority between October, 1995 to May, 1996. The Hospital Authority is a publicfunded organization which provides a comprehensive range of medical and health services to the public in Hong Kong. The POPDs accept patients who are referred primarily from medical clinics or units in The Hospital Authority which provide the majority of ambulatory physiotherapy care in Hong Kong. Consecutive Chinese patients with LBP who sought treatment from the POPDs were invited by the evaluators, who were senior physiotherapists, to participate in this study. The outline of the study and its objectives were then explained to each patient by the evaluator. Potential subjects were required to be 18 years of age or above and referred to physiotherapy for mechanical LBP, with or without radiation of symptoms. Written informed consent was then obtained from them. They were also assured that their human rights would be protected to the best of the investigators’ abilities. Patients were excluded from the study on the following criteria: (1) specific LBP caused by metastasis, infection, inflammation or a recent fracture; (2) LBP exhibiting symptoms and signs of cord or cauda equina compressions or structural instability in the spine; (3) concurrent symptoms in the upper back and neck; (4) pregnancy; (5) pending litigation; and (6) overt and excessive sickness behaviour as perceived by the evaluator as restlessness, inconsistency in response and extreme anxiety. To investigate the psychometric properties of the CHP42 Profile, there were four parts in this study: Part I (test–retest reliability and internal consistency); Part II (cross-sectional construct validity); Part III (longitudinal construct validity); and Part IV (responsiveness).

Manual Therapy (2003) 8(3), 151–160

Generic health measure for low back pain 153

for their back pain, on their written consent, they were then invited by the evaluator to complete the first CPH42 Profile. On returning for their first physiotherapy session, usually around 10 days (mean=9.6 days, SD=3.9 days), the patients who reported no change in back pain since their initial visit, were requested to complete the second CPH42 Profile. The interval of time between the first and second questionnaires conforms to the observation of Streiner & Norman (1995). They suggested that for scales measuring LBP, an interval of 7–14 days is recommended so that the two occasions should be close enough to avoid genuine changes, yet should be far enough apart to avoid memory effect. Here, those patients who reported a change in their back pain condition did not proceed to complete the second questionnaires and were screened out of the study. Altogether 204 patients with LBP completed the first and second sets of questionnaires (Sample A). Part II: Cross-sectional construct validity In studying the construct validity of the CPH42 Profile in patients with LBP, hypotheses were tested on the following assumptions in patients in Sample A using the unpaired t-test: (1) Patients who had intended to seek or had sought medical consultation for their back pain in the past week would be less healthy than those who had not. (2) Patients who had taken medication for their back pain in the past week would be less healthy than those who had not. Testing the construct validity of the CPH42 Profile using the criterion approach is not feasible as there is no ‘gold standard’ in the measurement of quality of life from LBP. Instead, the approach of testing the external construct validity is used. In studying the cross-sectional construct validity of a measure, the correlation between it and an external measure at different points in time is established (Carmines & Zeller 1979; Kirshner & Guyatt 1985). In testing the cross-sectional construct validity of the CPH42 Profile in patients with LBP, the association between it and two external scales were examined: (1) the Roland LBP disability Scale (Roland & Morris 1983) and (2) pain intensity measured with a 11-point numerical rating scale (NRS). The Roland LBP disability Scale was used as it is widely used in POPDs in Hong Kong and it has acceptable psychometric properties (Beurskens et al. 1995). The Chinese version of the Roland Scale was investigated using 50 Hong Kong Chinese patients with LBP (Tsang 1999). The test–retest reliability on two administrations, with an intervening period of around 10 days when patients’ condition remained Manual Therapy (2003) 8(3), 151–160

stable, was found to be good (intra-class correlation coefficient=0.94, CI=0.86–0.95). The NRS is easy to administer and score. Therefore, it is the most common means of measuring pain intensity in Hong Kong. Its validity has been well studied (Jensen et al. 1989; Wilkie et al. 1990; Von Korff et al. 2000). Each patient was asked to rate the present intensity of pain from 0 to 10, after explaining that 0 represented no pain, and 10 represented an intolerable level of pain. Two separate samples of patients completed the CPH42 Profile, the Roland Scale and the NRS: Sample B (n=114) at their first physiotherapy session, and Sample C (n=94) at their last physiotherapy session before discharge. The inclusion criteria of Samples B and C were identical to those of Sample A. Part III: Longitudinal construct validity The longitudinal construct validity of the CPH42 Profile was studied in patients with LBP in Sample D (n=61). Using the study method of Kirscher and Guyatt (1985), longitudinal construct validity is established when the change in the investigated scale correlates with the corresponding change in an external measure over the same period of time. The association between the CPH42 Profile score and the pain intensity score using the NRS was studied at the commencement of physiotherapy, after 3 weeks of treatment and after 6 weeks of treatment. Part IV: Responsiveness Responsiveness of the CPH42 Profile was measured in patients of Sample D (n=61 at entry) in term of its effect size (standard response mean, SRM) at weeks 3 and 6 after the commencement of physiotherapy. This duration was based on the observation that many clinical investigations have evaluated outcome measures within this time frame (Koes et al. 1992; Erhard et al. 1994; Stratford et al. 1996a; 1998). The SRM is defined as the mean change in the CPH42 Profile scores at weeks 3 and 6 divided by the standard deviation of the change (Katz et al. 1987; Lurie 2000). The corresponding SRM for the pain NRS was also calculated for comparison with that of the CPH42 Profile. The sensitivity of the NRS to treatment has been established in previous studies (Jensen et al. 1989; Wilkie et al. 1990; Von Korff et al. 2000). Another element in studying a health measure’s responsiveness is the minimal level of detectable change in individual patients for that measure. The method to calculate the minimal detectable change (MDC) as suggested by Stratford et al. (1996b) and Beaton (2000) was used. The MDC was calculated using the formula: standard error of measurement (SEM) z-score at the two-sided confidence intervals of 95% (z=1.96) O2. The SEM was determined using the formula: SEM = SD1 (1–R), where SD1 is r 2003 Elsevier Science Ltd. All rights reserved.

154 Manual Therapy

the standard deviation of the measure at baseline and R is the test–retest reliability coefficient. The product of the SEM and the z-score was multipled by 2 as there were two-sample measurements (first and repeated measurements) compared, rather than a sample value or a true score. The MDC in this study was calculated on data from Sample A. After the MDC for individual patients was calculated, the MDC for the group of patients was calculated by reducing the MDC by a factor of one over the square root of the sample size using the method suggested by Ravaud et al. (1999).

RESULTS Altogether, there were four independent samples, totalling 473 patients with LBP observed and measured in this study (Samples A–D). These independent samples of patients with LBP were recruited from six different POPDs (Centres 1–6) so that results would be generalizable to the patients seeking public physiotherapy service in Hong Kong. The patient demographics were presented in Table 1. Most of the patients in all samples (49.3%) have had more than four episodes of LBP, while some (18.0%) sought physiotherapy for their first episode. For the present episode of LBP, 63.2% of the patients had LBP for over 14 days before seeking physiotherapy at the POPDs, while the rest sought treatment earlier.

For those patients who had received physiotherapy (patients in Samples C and D), the modalities included electrotherapy, traction, passive mobilization, therapeutic exercises and back education.

Part I: Reliability and internal consistency In this part of the study, there were, however, 171 patients who did not complete the second questionnaire. Of these patients, 35.7% (n=61) of them reported that their LBP had improved, while 19.3% (n=33) reported a worsened condition since they had completed their first questionnaires. For those who did not complete the two questionnaires their reasons were attributable to refusal (n=9), inability to read or understand the item-questions (n=11), or to other various reasons (n=8). The overall response rate for the 375 patients (204 + 171) who completed the first questionnaires was 92.5%, which is over the acceptable level of 90% (Waddell et al. 1982). Results for the test–retest reliability study of the CPH42 Profile on 204 patients in Sample A are given in Table 2. Here it can be seen that the high ICC coefficient of 0.92 (CI=0.90–0.94) remained stable in subsamples stratified by age and gender notwithstanding the fact that they could be confounding variables (Portney & Watkins 2000). The mean difference in the initial and repeat scores was 0.15% (SD=7.03%, 95% CI=13.9% and 14.2%). These figures support that

Table 1. Patient characteristics Study part

I, Test–retest reliability and internal consistency

II, cross-sectional construct validity

III, longitudinal construct validity IV, responsiveness

Sample (Centre)

A* (1,2,3)

B (1)

C (1)

D (1,4,5,6)

n Age Mean (years) SD (years) Range (years)

204

114

94

61

41.70 11.19 20–77

39.18 12.64 15–76

40.50 11.04 18–75

38.28 11.51 19–64

Male Female History 1st Episode 2nd–4th episode Over four episodes

n 91 113

% 44.6 55.4

n 50 64

% 43.9 56.1

n 45 49

% 47.9 52.1

n 37 24

% 60.7 39.3

36 43 125

17.6 21.1 61.3

17 63 34

14.9 55.3 29.8

19 29 46

20.2 30.9 48.9

13 20 28

21.3 32.8 45.9

89 19 96

43.6 9.3 47.1

34 34 46

29.8 29.8 40.4

26 10 58

27.7 10.6 61.7

25 9 27

41.0 14.8 44.2

Duration before physiotherapy 0–14 days 15–28 days Over 28 days

*Patients in Sample A also participated in the study on the construct validity of the CPH42 Profile. Centre 1: Prince of Wales Hospital, New Territories. Centre 2: Kowloon Rehabilitation Centre, Kowloon. Centre 3: Kwong Wah Hospital, Kowloon. Centre 4: Queen Elizabeth Hospital, Kowloon. Centre 5: David Trench Rehabilitation Centre, Hong Kong. Centre 6: Yan Chai Hospital, New Territories. r 2003 Elsevier Science Ltd. All rights reserved.

Manual Therapy (2003) 8(3), 151–160

Generic health measure for low back pain 155 Table 2. Test–retest reliability and internal consistency coefficients for the CPH 42 Profile in patients of Sample A Part I: Reliability Test–retest reliability

All Ager40 years >40 years Male Female

Internal consistency

ICC (95% CI)

n

Cronbach’s alpha (95% CI)

Equal–length Spearman–Brown coefficient (95% CI)

0.92 (0.90–0.94) 0.94 (0.91–0.96) 0.95 (0.93–0.97) 0.95 (0.93–0.97) 0.94 (0.92–0.96)

204

0.90 (0.86–0.93) 0.89 (0.84–0.93) 0.90 (0.86–0.93) 0.90 (0.86–0.93) 0.89 (0.85–0.92)

0.87 (0.84–0.90) 0.87 (0.81–0.91) 0.88 (0.82–0.92) 0.90 (0.85–0.93) 0.85 (0.79–0.89)

100 104 91 113

ICC=Intra-class correlation coefficients (repeated measures). CI=Confidence intervals. As the results of the second test were almost identical, only results of the first test were reported here.

test–retest reliability of the CPH42 Profile was highly acceptable for clinical measurement. Results of the internal consistency study for the CPH42 Profile are also summarized in Table 2. Since the results of the first CPH42 Profile scores were almost identical to those of the repeat scores, only the former are presented here. The Cronbach’s alpha and the equal-length Spearman–Brown (split-half) coefficients were 0.90 (CI=0.86–0.93) and 0.87 (CI=0.84–0.90), respectively, on subjects in Sample A and demonstrated good internal consistency. Examination of the results showed that these two measures of the internal consistency of the CPH42 Profile also remained stable for the subsamples stratified on age and gender. On the third measure of internal consistency, the item-total correlation coefficients on all 42 item-questions, all but three item-questions scored above 0.2 in their item-total correlation coefficients. These three included Question #8 (Have you been staying up till very late in the evenings?), #12 (Have you been able to stay clearminded?), and #31 (Have you been wakened up very early in the morning?).

that (1) patients who had intended to seek or had sought medical consultation for their back pain in the past week would be less healthy than those who had not and (2) patients who had been taking medication for their back pain in the past week would be less healthy than those who had not. It can be seen that those who had intended to seek or had sought medical care were significantly, though marginally, less healthy than those who had not (t=2.00, P=0.05). Also those patients who had been taking medication for their back pain were significantly less healthy than those who had not (t=2.72, P=0.01). Table 4 presents the results of the cross-sectional construct validity of the CPH42 Profile with the Roland LBP Scale and pain intensity NRS in patients in Samples B and C. The correlation between the CPH42 Profile and the Roland Scale was shown to be 0.48 at entry to study and 0.53 at discharge. The correlation with the pain intensity NRS were 0.42 at entry and 0.22 at discharge. Part III: Longitudinal construct validity

Part II: Cross-sectional construct validity Table 3 presents the results of the unpaired t-tests, for patients in Sample A, which tested the hypotheses

Results of the longitudinal construct validity of the CPH42 Profile are given in Table 4 for patients in Sample D. The correlation between the change in CPH42 Profile score with the change in pain intensity

Table 3. Mean scores in the CPH 42 Profile in patients in Sample A. Results of the unpaired t-test on their mean scores are also given Construct validity (1) No

(2) Yes

(1)–(2)

t

P-value

No–Yes

Mean score (n)

Mean score (n)

Mean score difference

Seeking medical care

33.29 (45)

38.15 (159)

4.86

2.004

0.05

Taking medication

33.40 (60)

38.61 (144)

5.21

2.718

0.01

Manual Therapy (2003) 8(3), 151–160


156 Manual Therapy Table 4. Correlations between the CPH 42 Profile (change in the longitudinal validity study) and other measures of disability and pain Construct validity

Part II Cross- sectional Validity

Part III Longitudinal Validity

rs

95% CI

P-value

n

Sample

Roland Scale

At entry At discharge

0.48 0.53

0.32–0.61 0.37–0.67

0.01 0.01

113 91

B C

Pain NRS

At entry At discharge

0.42 0.22

0.28–0.72 0.16–0.40

0.01 0.05

113 94

B C

0.09 0.42

0.32–0.25 0.08–0.67

0.57 0.05

45 29

D

Change in pain NRS At week 3 At week 6

Pain NRS—11-point pain numerical rating scale. rs =Spearman’s coefficient. 0.81–1.00 Very strong 0.49–0.80 Strong 0.25–0.48 Moderate 0.09–0.24 Low 0.00–0.08 Weak Table 5. The SRMs, mean change in the CPH42 Profile Part IV: Responsiveness

Standard response means* From entry to week 3 From entry to week 6

Mean change (and SD) From entry to week 3 From entry to week 6

CPH42 (n)

NRS (n)

Sample

0.33 (35) 0.58 (21)

0.32 (40) 0.49 (27)

D

CPH42

NRS

2.83 (8.66) 6.67 (11.51)

0.61 (1.89) 1.15 (2.34)

*Standard response mean=mean change in scores/standard deviation of change in scores. NRS=Pain numerical rating scale.

NRS at week 3 was 0.09. The corresponding correlation at week 6, however, was 0.42.

Part IV: Responsiveness The results for the responsiveness of the CPH42 Profile are provided in Table 5. It can be seen here that the responsiveness of the CPH42 Profile measured from the commencement of physiotherapy to week 3 (SRM=0.33) and from the commencement to week 6 (SRM=0.58) were approximately the same as those of the pain intensity NRS (SRMs=0.32 and 0.49, respectively). In studying the MDC of the CPH42 Profile in individual patients, the SEM was first calculated. As the standard deviation of the first CPH42 score in Sample A was 12.76%, and the test–retest reliability coefficient was 0.92, the SEM was therefore worked out to be: 3.61% (12.76% O (1–0.92)). The MDC at 95% CI was then calculated to be 10.01% (3.61% 1.96 O 2). The MDC for the group of patients was 0.70% (10.01% 1/ O 204). r 2003 Elsevier Science Ltd. All rights reserved.

DISCUSSION The high response rate (92.5%) in the completion of the CPH42 Profile demonstrated the acceptability of its use in Chinese patients with LBP in busy frontline POPDs in Hong Kong. Results of the test–retest reliability in 204 patients in Sample A (Table 2) indicate that the CPH42 Profile is a highly reliable instrument. The ICC coefficient of 0.92 exceeded that reported for the SIP (r=0.85) for patients with LBP (Deyo & Diehl 1983). Confidence interval analysis was used to illustrate the size of the difference between the first and repeat administrations of the CPH42 Profile (Bland & Altman 1986). The 95% confidence interval for the difference in scores between the two administrations ranged from 13.9% to 14.2%. As the interval was non-systematic and comparable to that of the Aberdeen LBP Disability Scale, a condition-specific measure (comparable confidence interval of 14.25% and 12.4%), it is perhaps adequate for clinical measurement. Confident interval analysis has not been reported for other generic health profiles. Manual Therapy (2003) 8(3), 151–160


The internal consistency of the CPH42 Profile in 204 patients, in Sample A (Table 2), was satisfactory (Cronbach’s alpha=0.90; Spearman–Brown splithalf coefficient=0.87). All item-questions exceeded the lowest acceptable item-total correlation coefficients of above 0.2 (as stipulated by Abramson 1990; Streiner & Norman 1995) except Questions #8, 12 and 31. These results are not unexpected as the CPH42 Profile is a generic questionnaire which comprises item-questions which may not be sufficiently sensitive and relevant in measuring current perceived health in LBP. The results here may also suggest that the CPH42 Profile can be further refined by deleting Questions #8, 12 and 31 from the Profile when used specifically for use in LBP measurement. In evaluating the construct validity of the CPH42 Profile in patients in Sample A (Table 3), results of the unpaired t-tests showed that patients who had intended to seek or had sought medical care and took medication for their back pain in the past week were significantly less healthy than those who did not (t=2.00, P=0.05 for medical care; t=2.72, P=0.01 for medication). These results suggest that the construct of the CPH42 Profile in assessing perceived health in patients with LBP is basically robust. Cross-sectional construct validity was observed on the correlation between the CPH42 Profile and the two other measures of disease severity: the Roland LBP Scale and the pain intensity NRS at the start of physiotherapy (Sample B, n=114) and at discharge (Sample C, n=94) (Table 4). It should be pointed out here that the results for the Roland LBP Scale should be interpreted with some caution as the construct validity of the Chinese version has not been fully investigated (Tsang 1999). The Spearman’s correlation coefficients with the Roland Scale ranged between 0.48 and 0.53, indicating a partial association between the two measures. These results show that the CPH42 Profile had a weaker association than the SIP when the SIP was correlated with the Roland Scale (r=0.85) (Deyo 1986). Closer examination on the SIP study revealed that the Roland Scale correlated less strongly with the psychosocial dimension (r=0.59) than with the physical dimension. As the CPH42 Profile has a strong emphasis on psychosocial health, it does not lessen its value as an adjunct health measure for LBP. Indeed, the partial association between the CPH42 Profile and the Roland Scale may indicate its supplementary value when it is used side by side with a LBP-specific scale in which the physical dimension predominates. The correlation of the CPH42 scores with pain intensity NRS was moderate to low, yet significant ( P=0.05, rs ranged between 0.42 at entry and 0.22 at discharge) (Pett 1997). It may be due to the fact that the CPH42 Profile is a multi-dimensional measure in which pain is only one of the many attributes. This Manual Therapy (2003) 8(3), 151–160

may agree to the observation of Ruta et al. (1994) that pain intensity provides only a narrow definition of LBP outcome. However, it should be pointed out here that the NRS, though widely used in Hong Kong, was not validated in Chinese patients with LBP. The association may also indicate that LBP disability may have a general health implication, for example, psychological and social functions. The use of the CPH42 Profile would open up the perspective of the health-care providers to wider health implications of LBP. Ideally, the CPH42 Profile should be validated with a ‘gold standard’ measure of quality of life or psychosocial status. However, such a ‘standard’ is not yet available in the Chinese language. This is a limitation of this study, thus providing an area for future study. Longitudinal construct validity of the CPH42 Profile was established when the change in score was correlated with change in pain intensity NRS at weeks 3 and 6 after the commencement of physiotherapy (Sample D, Table 4). Association with the pain intensity NRS was moderate at week 6 (rs=0.42) though it was low at week 3 (rs=0.09) (Pett 1997). This may indicate the value of the CPH42 Profile in measuring LBP disability in the longer term. The responsiveness of the CPH42 Profile measured from entry to week 3 and then to week 6 in patients in Sample D (n=61, Table 5) were intermediate (SRMs=0.33 and 0.58, respectively). The higher SRMs of the CPH42 Profile suggest its slight advantage over the pain intensity NRS. This agrees with the observation of Ruta et al. (1994) that pain intensity is a narrow definition of health outcome. It may need another study to support the implication that physiotherapy may be more effective in improving general health than pain. The CPH42 performed less well in comparison with the SF36 Health Profile (SRM for the overall measure was 1.09) (Beaton et al. 1997; Lurie 2000). The failure to measure the Roland scores at weeks 3 and 6 is one of the limitations of this study. However, compared with the Aberdeen LBP Scale (the highest SRM was 0.62) (Ruta et al. 1994), the responsiveness of the CPH42 Profile appears adequate. A detectable change in the CPH42 Profile in a future efficacy study may reflect a change in health status primarily across the psychological and social functions which are presumably less susceptible to change. The MDC of the CPH42 score in individual patients was found to be 10.01%. This represents the smallest change score in an individual patient that a clinician can confidently consider to be indicative of a change and not just measurement error. For example, in a patient with an initial CPH42 score of 30%, a follow-up score of either 20% or 40% would be a true improvement or worsening that has occurred in that patient. r 2003 Elsevier Science Ltd. All rights reserved.

158 Manual Therapy

Studying responsiveness in subjective health measures using the MDC method is relatively uncommon. Not many studies reported MDC. In examining the LBP-specific Roland Scale, the MDC in individual patients as reported by Stratford et al. (1996b; 1998), Beaton (2000) was 5.0 points, turning into a percentage of 20.83% (5/24 100%, where 24 is the highest score in the Roland Scale). In this respect, with a MDC of 10.01%, the CPH42 Profile is likely to be more responsive to change than the Roland Scale. On investigating the reliability, validity and responsiveness of the CPH42 Profile on four samples of Chinese patients with LBP under clinical conditions in frontline POPDs in Hong Kong, results suggest that it may be considered as an useful outcome measure for LBP in efficacy studies. As the CPH42 Profile purports to measure, in particular, psychosocial health, it may be used as a broad-spectrum and multi-dimensional supplementary measure to LBP-specific functional scales. The use of a generic health profile as part of a more comprehensive portfolio of measures is recommended to assess LBP (Garratt et al. 1993; Ruta et al. 1994; Bombardier 2000). The provision of a generic perceived health profile also allows an opportunity for researchers and clinicians to compare health status between patients afflicted with different conditions through the use of a common measure. The psychometric properties of the CPH42 Profile, however, remain to be tested in patients who have chronic LBP for an extended period of time, or in situations where they need to be observed and measured in follow-up to conservative or operative interventions. Acknowledgements The authors thank patients and staffs of Physiotherapy Departments, particularly Mr Joseph Chiu of David Trench Rehabilitation Centre, Mr Gordon Mak of Jockey Club Kowloon Rehabilitation Centre, Mr Lawrence Fung of Kwong Wah Hospital, Mr Herman Lau of Prince of Wales Hospital, Ms Polly Lau, Mr Gilbert Ho and Ms Jocelyn Cho of Queen Elizabeth Hospital, and Mr Calvin Luk of Yan Chai Hospital, The Hospital Authority, Hong Kong. Thanks must also go to Dr Rhonda J. Scudds for her editorial work on this paper. We are indebted to The Health Services Research Committee, Hong Kong, and the Staff Development Programme, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, which provided the opportunity and funding to perform this study. This work is attributed to The Department of Community Medicine and Unit for Behavioural Sciences, The University of Hong Kong. This work was supported by The Health Services Research Committee, Hong Kong and Staff Development Programme, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University.

References Abramson JH 1990 Survey Methods in Community Medicine, Churchill Livingstone, Edinburgh, ch 13, p 127 r 2003 Elsevier Science Ltd. All rights reserved.

Battie MC, Cherkin DC, Dunn R, Ciol MA, Wheeler KJ 1994 Managing low back pain: Attitudes and treatment preferences of physical therapists. Physical Therapy 74: 219–226 Beaton DE 2000 Understanding the relevance of measured change through studies of responsiveness. Spine 25(24): 3192–3199 Beaton DE, Bombarider C, Guillemin F, Ferraz MB 2000 Guidelines for the process of cross-cultural adaptation of self-report measures. Spine 25(24): 3186–3191 Beaton DE, Hogg-Johnson S, Bombarider C 1997 Evaluating changes in health status: Reliability and responsiveness of five generic health status measures in workers with musculoskeletal disorders. Journal of Clinical Epidemiology 50: 79–93 Bergner M, Bobbitt RA, Carter WB, Gibson BS 1981 The Sickness Impact Profile: Development and final revision of a health status measure. Medical Care 19: 787–805 Beurskens AJ, de Vet HC, Koke AJ, Heijden GJ, Knipschild PG 1995 Measuring the functional status of patients with low back pain. Spine 20(9): 1017–1028 Bland JM, Altman DG 1986 Statistical methods for assessing agreement between two methods of clinical measurement. Lancet i: 307–310 Bombardier C 2000 Outcome assessments in the evaluation of treatment of spinal disorders: Summary and general recommendations. Spine 25(24): 3100–3103 Carmines EG, Zeller RA 1979 Reliability and Validity Assessment, Sage Publications, Beverley Hills Cassidy JD, Carroll LT, Cote P 1998 The Saskatchewan health and back pain survey: The prevalence of low back pain and related disability in Saskatchewan adults. Spine 23(17): 1860–1867 Deyo RA 1986 Comparative validity of the Sickness Impact Profile and shorter scales for functional assessment in low back pain. Spine 11(9): 951–954 Deyo RA, Diehl AK 1983 Measuring physical and psychosocial function in patients with low back pain. Spine 8(6): 636–642 Dodd T 1997 The Prevalence of Back Pain in Great Britain in 1996: A Report On Research for The Department of Health Using the ONS Omnibus Survey. The Stationery Office, London Erhard RE, Delitto, Cibulka MT, 1994 Relative effectiveness of an extension programme and a combined programme of manipulation and flexion and extension in patients with acute low back syndrome. Physical Therapy 74: 1093–1100 Fairbank JCT, Cooper J, Davies JB, O’ Brien JP 1980 The Oswestry low back pain disability questionnaire. Physiotherapy 66: 71–73 Fielding R, Li J 1997 A validation of the concept of current perceived health and the current perceived health-42 (CPH42) questionnaire. Quality of Life Research 6: 35–42 Follick MJ, Smith TW, Ahern DK 1985 The Sickness Impact Profile: A global measure of disability in chronic low back pain. Pain 21: 67–76 Frymoyer JW, Cats-Baril WL 1991 An overview of the incidences and costs of low back pain. Orthopaedic Clinics North America 22: 263–271 Garratt AM, Ruta DA, Abdalla MI Buckingham JK, Russell IT 1993 The SF36 health survey questionnaire: An outcome measure suitable for routine use within the NHS? British Medical Journal 306: 1440–1444 Greenough CG, Fraser RD 1992 Assessment of outcome in patients with low back pain. Spine 17: 36–41 Jensen MP, Karoly P, O’Riordan EF, Bland F, Burns RS 1989 The subjective experience of acute pain: An assessment of the study of 10 indices. Clinical Journal of Pain 5: 153–159 Jette AM, Smith K, Haley SM, Davies KD 1994 Physical therapy episodes of care for patients with low back pain. Physical Therapy 74: 101–105 Katz JN, Larson MG, Philips CB, Fossal AM, Liang MH 1987 Comparative measurement sensitivity of short and longer health status instruments. Medical Care 27: S178–S189 Kirshner B, Guyatt G 1985 A methodological framework for assessing health indices. Journal of Chronic Diseases 35: 27–36 Koes BW, Bouter LM, van Mameren H, Essers AH, Verstegen GMJR, Hofhuizen DM, Houben JP, Knipschild PG 1992 The effectiveness of manual therapy, physiotherapy, and treatment by the general practitioner for non-specific back and neck complaints: A randomised clinical trial. Spine 17: 28–35 Manual Therapy (2003) 8(3), 151–160


Kopec JA, Esdaile JM, Abrahamowicz M Abenhaim L, WoodDauphinee S, Lamping DL 1995 The Quebec back pain disability scale. Spine 20(3): 341–352 Lam CLK, Gandek B, Ren XS, Chan MS 1998 Tests of scaling assumptions and construct validity of the Chinese (Hong Kong) version of the SF36 Health Survey. Journal of Clinical Epidemiology 51(11): 1139–1147 Lansky D, Butler JBV, Waller FT 1992 Using health status measures in the hospital setting: from acute care to ‘‘outcome management’’. Medical Care 30: MS57–MS73 Lau EMC, Egger P, Coggon D, Cooper C, Valenti L, O’Connell D 1995 Low back pain in Hong Kong: Prevalence and characteristics compared with Britain. Journal of Epidemiology and Community Health 49: 492–449 Li J and Fielding R 1995 The measurement of current perceived health among Chinese people in Guangzhou and Hong Kong, Southern China. Quality of Life Research 4: 271–278 Lurie J 2000 A review of generic health status measures in patients with low back pain. Spine 25(24): 3125–3129 Pett MA 1997 Non-parametric Statistics for Health Care Research. Sage Publications, California, ch 9, p 251 Physiotherapy Management Protocol for Low Back Pain Patients 1996 Co-ordinating Committee for Physiotherapists, The Hospital Authority, Hong Kong Portney LG, Watkins MP 2000 Foundations of clinical research: applications to practice, 2nd edn. Appleton and Lange, Norwalk, ch 26 pp 560–566, 575–577 Ravaud PH, Giraudeau B, Auleley GR, Edouard-Noel R, Dougados M, Chastang CL 1999 Assessing smallest detectable change over time in continuous structural outcome measures: Application to radiological change in knee ortharthritis. Journal of Clinical Epidemiology 52(2): 1225–1230 Roland M, Morris R 1983 A study of the natural history of back pain, Part 2: Development of guidelines for trials of treatment in primary care. Spine 8: 145–150 Ruta DA, Garratt AM, Wardlaw D, Russell IT 1994 Developing a valid and reliable measure of health outcome for patients with low back pain. Spine 19(17): 1887–1896 Stratford PW, Binkley JM, Solomon P, Finch E, Gill C, Moreland J 1996a Defining the minimum level of detectable change

for the Roland Morris Questionnaire. Physical Therapy 76(4): 359–365 Stratford PW, Finch E, Solomon P, Binkley J, Gill C, Moreland J 1996b Using the Roland Morris Questionnaire to make decisions about individual patients. Physiotherapy Canada 48(2): 107–110 Stratford PW, Binkley JM, Riddle DL, Guyatt GH 1998 Sensitivity to change of the Roland Morris Back Pain Questionnaire: Part I . Physical Therapy 78(11): 1186–1196 Streiner GL, Norman DR 1995 Health Measurement Scales: A Practical Guide to their Development and Use 2nd edn. Oxford University Press, Oxford ch 5 pp 65–66, ch 8 pp 104–127 Swanson G 1995 Use of outcome reports justifying the need for physical therapy services. Orthopaedic Clinics North America 4: 253–268 Tsang R 1999 Examining reliability, validity and responsiveness of the Hong Kong version of the Roland Morris Disability Questionnaire as an outcome measure of patients with low back pain: A multi-centre study. Working Group on Low Back Pain Outcome Assessment, Physiotherapy Specialty Forum in Orthopaedics and Traumatology, The Hospital Authority, Hong Kong Von Korff M, Jensen MP, Karoly P 2000 Assessing global pain severity by self-report in clinical and health services research. Spine 25(24): 3140–3151 Waddell G, Main CJ, Morris EW, Venner RM, Rae PS, Sharmy SH, Galloway H 1982 Normality and reliability in the clinical assessment of backache. British Medical Journal 284: 1519–1523 Walsh K, Cruddas M, Coggon D 1992 Low back pain in eight areas of Britain. Journal of Epidemiology and Community Health 46: 227–230 Ware JE 2000 SF36 Health survey update. Spine 25(4): 3130–3139 Ware JE, Sherbourne CD 1992 The SF36 health status survey: I. Conceptual framework and item selection. Medical Care 30: 473–483 Wilkie D, Lovejoy N, Dodd M, Tesler M 1990 Cancer pain intensity measurement: concurrent validity of three tools— finger dynamometer, pain intensity number scale, visual analogue scale. Hospice Journal 6(1): 1–13

Appendix. English version of the CPH 42 Profile This questionnaire helps us to understand how much people experience the feelings listed. This is helpful for us to plan better health services. You do not need to give your name and all information will be kept confidential. Please answer as honestly as possible, how you have been feeling generally over the past week. Please circle the answer that most applies to you. 0=never,

1=sometimes,

2=often,

3=most of time

During past seven days, have you been 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

feeling tired all the time? feeling satisfied? feeling worried about your family or close friends? feeling disordered and in confusion? only able to walk about indoors? losing confidence in yourself? feeling relaxed? going to bed at late hours? finding it hard to dress yourself? feeling safe? feeling that you are short of friends, and want to have some more? feeling clear-headed?

Manual Therapy (2003) 8(3), 151–160

0 0 0 0 0 0 0 0 0 0 0 0

1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3


160 Manual Therapy

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

feeling often out of breath or having difficulty in breathing? feeling run out of energy and unwilling to move any more? finding it hard to get on with people? feeling that your family expect too much from you? feeling sleepy? having unbearable pain? feeling confident? finding it hard to keep your mind on things? feeling it hard to do what you want to do? feeling delightful, joyful? feeling it painful when doing sports? feeling that you are soon run out of energy? feeling there are too many unhappy things going on? feeling like you are losing control of the important things in your life? going out as often as usual? feeling discomfort or pressure in your chest? in pain when doing things? feeling lonely? waking up in the early hours of the morning? needing help to walk about outside? feeling that no one can really help you to solve current difficulties? feeling agitated? feeling unable to change your circumstances? in pain during sleep? finding it hard to bend or reach for things? feeling depressed? having enough time to sleep? feeling you had no appetite at all? having trouble getting up or down stairs? feeling light of mood?


0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Manual Therapy (2003) 8(3), 151–160


Original article

Lumbar multifidus muscle size does not differ whether ultrasound imaging is performed in prone or side lying Y. Coldron*, M. Stokesw, K. Cookz St George’s Hospital Medical School, London, UK, w Royal Hospital for Neuro-disability, London, UK, St. George’s Fetal Medical Unit, London, UK

n

z

SUMMARY. This methodological study examined the influence of body position when measuring multifidus muscle size using real-time ultrasound imaging. Previous studies examined multifidus with the subject prone but people with certain conditions cannot be studied in this position, so side lying was investigated as an alternative posture. In 20 normal females (aged 19–45 years), the cross-sectional area (CSA) of lumbar multifidus was measured at the level of the fifth lumbar vertebra (L5) on both the right and left sides, with the subject in prone and in side lying. Multifidus CSA was highly correlated between the two positions on both the right (r=0.90) and left (r=0.91) sides. Paired t-tests found no significant differences between the measurements made in the two scanning postures (right P=0.77; left P=0.16). Bland and Altman plots showed good agreement between the two positions, with no systematic difference. These findings demonstrate that measurements of lumbar multifidus at L5 can be made in either prone or side lying and a valid comparison of the results obtained in both positions can be made. r 2003 Elsevier Science Ltd. All rights reserved.

Values for CSA measurements of lumbar multifidus reported in the literature were obtained from ultrasound scans taken with the subjects in prone lying. However, prone lying may not be a comfortable position for some subjects, for example postnatal women who are breast feeding, and may not be achievable in others, such as patients with certain neurological conditions or some individuals with acute low back pain. A validation study of multifidus CSA measurements made using ultrasound imaging vs magnetic resonance imaging (MRI) found no difference between the two methods, the latter scans being taken with the subject supine (Hides et al. 1995). Since the lumbar spine is not accessible for ultrasound imaging in supine, the present study evaluated side lying as an alternative position.

INTRODUCTION Ultrasound imaging has been used to measure the size of various skeletal muscles (see Stokes et al. 1997 for review). The lumbar multifidus muscle is of particular interest in patients with low back pain and has been studied in normal subjects (Hides et al. 1992) and patients (Hides et al. 1994, 1996, 1998). These studies have demonstrated that multifidus cross-sectional area (CSA) can be measured adequately by ultrasound imaging and used to compare different populations.

Received: 24 May 2002 Revised: 4 November 2002 Accepted: 15 January 2003 Yvonne Coldron, MSc, MCSP, MMACP, Research Fellow, St George’s Hospital Medical School, London, UK, Professor Maria Stokes, PhD, MCSP, Director of Research & Development, Royal Hospital for Neuro-disability, London, UK, Katy Cook, DCRR, DMU, Superintendent Ultrasonographer, Fetal Medicine Unit, St George’s Hospital, London, UK. Correspondence to: YC, Department of Biochemistry & Immunology, St George’s Hospital Medical School, University of London, Cranmer Terrace, London SW17 ORE, UK. Tel.: +44 208 725 5790; E-mail: [email protected]

METHODS Subjects Twenty normal females (aged 19–45 years) were studied. Exclusion criteria were: pregnancy, history 161

162 Manual Therapy

of low back pain and skin conditions in the areas to be scanned. Written informed consent was obtained from all subjects. The project was approved by both the Riverside Research Ethics Committee and Wandsworth Local Research Ethics Committee. Ultrasound imaging technique Real-time ultrasound imaging was used to measure lumbar multifidus CSA following the protocol first described by Hides et al. (1992). One of two scanners was used depending on machine availability, either an Aloka SSD 1200 (Aloka Co. Ltd. Mitaka-shi, Tokyo, Japan) or an Acuson 128 XP/10 (Siemens Medical Solutions, Bracknell, Berks., England). A 5 MHz curvilinear transducer was used for optimum penetration and resolution. Two scans were taken at the level of L5 on both the right and left sides, and the mean of the two measurements at each site was recorded. One investigator scanned each subject during one testing session in two positions: prone lying and side lying, in random order. When prone, the subject lay with the head in the midline position, with one pillow under the lower legs and another under the hips to reduce the lumbar lordosis. The spinous process of L5 was located and checked by palpating the levels of the iliac crests. The skin over the spinous process was marked with a chinagraph pencil and the site for L5 was later confirmed internally on the image on the scanner’s screen (see below). For the side lying position, subjects lay on their left side and the ultrasound scanner was placed behind the subject. One pillow was placed under the head and another between the knees, with the hips and knees positioned in sufficient flexion (approximately 701 at the hips but not measured) to reduce the lumbar lordosis. A rolled towel was placed under the waist to obtain a neutral position for the spine in a lateral direction. The spinous process of L5 was repalpated and the location confirmed by palpating L4/ 5 interspinous space from the position of the iliac crests. The skin was re-marked, as the site differed from when in prone lying. The transducer was placed longitudinally over the skin marking for L5 spinous process to orientate the investigator and thus confirm skin marking. The transducer was then rotated by 901 to obtain a transaxial view, showing the L5 spinous process and bilateral echogenic laminae (Fig. 1), and then moved directly laterally to image either the right or left muscle individually. The transducer was angled to allow the beam to be perpendicular to the lamina (i.e. where the image of the lamina was brightest) as this provided a consistent landmark. The image was r 2003 Elsevier Science Ltd. All rights reserved.

Fig. 1—Real-time ultrasound image of a cross-sectional view of the left multifidus at the level of the fifth lumbar vertebrae (L5). The spinous process (SP) is on the right of the picture, with the echogenic lamina (L) providing a landmark for the inferior border of the multifidus (H) muscle on the left side.

taken at this point. The lateral border of multifidus was identified. The images were frozen and then measured using on-screen calipers to trace the perimeter of the muscle and provide a measure of area (cm2). Data analysis The purpose was to determine whether there was agreement between the scans taken in prone and side lying i.e. there was no effect of position. The correlation coefficient was calculated for the right and left sides and regression plots for side lying against prone lying were produced. Paired t-tests were also calculated to compare postures. Bland (2000) recommended that two methods of measurement could be assumed to agree if the expected difference between observations obtained on the same subject is small enough for the two methods to be used interchangeably. He recommended testing the mean difference of the observations to determine if the methods differed systematically, and also calculating the 95% limits within which most differences will lie, to indicate the extent the two methods are likely to differ due to observer error when used in practice. Consequently for the important measures, the differences between the mean of measurements in prone and side lying were obtained and the mean of these differences calculated and tested against zero.

RESULTS All data analyses demonstrated that there was no effect of position on the scans taken in prone and side lying (Table 1). Manual Therapy (2003) 8(3), 161–165

Lumbar multifidus muscle size 163 Table 1. Results for multifidus cross-sectional area (cm2) measured at the fifth lumbar vertebra using ultrasound imaging in prone and side lying: mean differences between the two positions; correlation coefficients; regression coefficients; and limits of agreement Side Right Left

Mean7SD (cm2) P SL P SL

5.4871.22 5.4471.22 5.5471.02 5.3971.02

Correlation coefficient

Regression coefficient

Mean differences 7SD (cm2)

Limits of agreement 72SD (cm2)

0.91

*0.83

0.0370.52

70.99

0.90

*0.80

0.1570.47

70.94

P=prone, SL=side lying; n=20. *Po0.001.

The correlation coefficients for multifidus CSA measured in the two positions were high on both the right (r=0.90) and left (r=0.91) sides. Regression analysis showed a significant relationship between the two scanning positions for both sides (Po0.001). This relationship is illustrated in Fig. 2 for the left side, and the graph for the right side was very similar. The within-subject differences between the prone and side lying positions plotted against the mean of measurements with 95% limits of agreement are seen in Fig. 3. The limits of variation (dotted lines) are 2SD from zero and indicate the range of differences likely to occur by chance. There was good agreement between measures. For both the right and left muscles, the mean differences were close to zero and the paired t-tests were not significant (right P=0.77; left P=0.16). The 95% limits of agreement are shown in Table 1.

Fig. 2—Regression plot showing a close relationship between measurements of lumbar multifidus cross-sectional area (CSA; cm2) at the fifth lumbar vertebra (L5) for the left side.

DISCUSSION This methodological study has demonstrated that CSA measurements of the lumbar multifidus muscle at L5 do not vary for the two lying postures examined. It would therefore be valid to make ultrasound measurements of the muscles in side lying. In prone lying the lumbar lordosis was reduced by placing a pillow under the abdomen with the hips in slight flexion. In side lying the lumbar lordosis was reduced by instructing the subject to flex the hips until the spine was in a neutral position. It is unlikely that the lumbar extensors were lengthened, as the CSA measurements were not altered significantly. This finding supports the view that lumbar multifidus undergoes minimal length change during trunk movements (Comerford & Mottram 1998). At the level of L5, the fascicles of multifidus arising from the vertebral levels above it are included in the CSA measurements at this level. The present findings may reflect the fact that in the two positions studied, the position of the lumbar spine, i.e. neutral, was more important to standardize than the degrees of hip flexion, which did not appear to influence the spine or Manual Therapy (2003) 8(3), 161–165

hence multifidus muscle length. No significant differences were seen between the means for prone and side lying, therefore there is good evidence that measurements from these locations do not differ systematically. The L5 level was selected for the present study because previous studies found that pain symptoms and atrophy of multifidus were most common at this level (Hides et al. 1996). If other vertebral levels are to be studied with subjects in side lying, the present findings would need to be confirmed for those levels, in order for values of CSA to be compared to those published for normal populations. Alternatively, a study could include a control group of normal subjects measured in side lying at the same vertebral level. Severe atrophy of multifidus has been shown to occur at the level of symptoms in patients with acute low back pain (Hides et al. 1994) and recovery of muscle size is not automatic following cessation of symptoms and only returns after specific rehabilitation of multifidus (Hides et al. 1996). It has yet to be shown whether a less specific exercise programme r 2003 Elsevier Science Ltd. All rights reserved.

164 Manual Therapy

Fig. 3—Bland and Altman (1986) plots for multifidus size measured at the fifth lumbar vertebra (L5) showing 95% limits of agreement and differences from zero for the left side.

might be as effective for restoring multifidus muscle size. Therefore the ability to make accurate measurements of multifidus is important to aid assessment and monitor the effects of rehabilitation. A limitation of the present study was that the investigator obtained the images and made measurements real time, so no blinding of measurement was undertaken. However, permanent recordings of the scans and cursor measurements were obtained, providing evidence that the actual data used were true. In order to rule out the possibility of any bias, studies should preferably involve blinding of measurement by, for example, making measurements on coded scans off-line. However, at the time of the present study, a facility for off-line analysis was not available. The present study has further refined the ultrasound imaging technique for assessing lumbar multifidus by introducing an alternative position in which to place the subject. Measurements of multifidus CSA have now been measured in prone and side lying using ultrasound imaging, as well as in supine using MRI in the study by Hides et al. (1995). Other more functional positions could also be studied, such as standing and sitting, to see whether measurements alter, being careful to ensure that the lumbar spine was standardized in neutral by reducing the lordosis. Ultrasound imaging of lumbar multifidus requires skill and training, since this muscle is not easy to examine using this technique. In some rehabilitation studies, scanning was performed by a qualified ultrasonographer (e.g. Hides et al. 1996). Prior to commencing the present study, the lead investigator (a physiotherapist) was trained in the imaging technique. This included learning about the physics and technical aspects of ultrasound imaging, and undertaking an intra-rater reliability study that demonstrated good agreement between repeated r 2003 Elsevier Science Ltd. All rights reserved.

measures on different days. It is recommended that potential investigators undertake appropriate training and reliability testing prior to using ultrasound to image muscles for clinical practice or research (Stokes et al. 1997). The clinical implications of the present findings are that subjects in whom it is difficult to achieve a prone position to obtain real-time ultrasound scans of multifidus, can be studied in side lying without influencing the measurement of CSA at L5.

Acknowledgements The authors thank the subjects who took part in the study, Dr Anthony Swan for statistical advice, Dr Basky Thilaganathan for use of facilities in the Fetal Medicine Unit at St George’s Hospital, London and the Neuro-disability Research Trust for financial support. Part of this work was undertaken at the Royal Hospital for Neuro-disability, which received a proportion of its funding from the NHS Executive; the views expressed in this publication are those of the authors and not necessarily those of the NHS Executive.

References Bland M 2000 An Introduction to Medical Statistics, 3rd edn. Oxford University Press Oxford Bland JM, Altman DG 1986 Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307–310 Comerford M, Mottram S 1998 Stability dysfunction and low back pain. The Journal of Orthopaedic Medicine 20(2): 13–18 Hides JA, Cooper RG, Stokes MJ 1992 Diagnostic ultrasound imaging for measurement of the lumbar multifidus muscle in normal young adults. Physiotherapy Theory and Practice 8: 19–26 Hides JA, Richardson CA, Jull GA 1995 Magnetic resonance imaging and ultrasonography of the lumbar multifidus muscle: Comparison of two different modalities. Spine 20: 54–58 Hides JA, Richardson CA, Jull GA 1996 Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine 21(23): 2763–2769 Manual Therapy (2003) 8(3), 161–165

Lumbar multifidus muscle size 165

Hides JA, Richardson CA, Jull GA 1998 Use of real time ultrasound imaging for feedback in rehabilitation. Manual Therapy 3(3): 125–131 Hides JA, Stokes MJ, Saide M, Jull GA Cooper DH 1994 Evidence of lumbar multifidus wasting ipsilateral to

Manual Therapy (2003) 8(3), 161–165

symptoms in patents with acute/subacute low back pain. Spine 19(2): 165–172 Stokes M, Hides J, Nassiri D 1997 Musculoskeletal ultrasound imaging: Diagnostic and treatment aid in rehabilitation. Physical Therapy Reviews 2: 73–92



Original article

Side-to-side weight-bearing asymmetry in subjects with low back pain J. D. Childs*, S. R. Pivan, R. E. Erhardw, G. Hicksz n

Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, USA, University of Pittsburgh Medical Center Health System’s Spine Specialty Center, USA, zClinical Research Center, Gerontology Research Center, National Institutes on Aging, National Institutes of Health, USA

w

SUMMARY. The purpose of this project was to determine if subjects with low back pain (LBP) exhibit greater side-to-side weight-bearing (WB) asymmetry compared to healthy control subjects without LBP. This study utilized an observational double cohort design and consisted of 35 subjects with LBP and 31 healthy control subjects. Sideto-side WB asymmetry was calculated as the average of the absolute value of the difference between the right and left lower extremity from three trials. The percentage of the average side-to-side WB asymmetry relative to the total body weight was calculated to normalize expected differences in magnitude of asymmetry based on a subject’s total body weight. An 11-point numeric pain rating scale was used to represent the subject’s current level of pain. Patients with LBP demonstrated significantly greater normalized side-to-side WB asymmetry than healthy control subjects (8.8% vs. 3.6%, respectively, Po0.001). In patients with LBP, higher magnitudes of side-to-side WB asymmetry were significantly associated with increased pain (r = 0.39, P = 0.021). In conclusion patients with LBP exhibited increased side-to-side WB asymmetry compared to healthy control subjects without LBP. This asymmetry was associated with increased levels of pain. This finding is relevant for planning future studies that will attempt to provide evidence for the construct validity of manipulation by determining if side-to-side WB asymmetry normalizes after a manipulation intervention and if this improvement is associated with improvements in pain and function. r 2003 Elsevier Science Ltd. All rights reserved.

manipulation to be effective in patients with low back pain (LBP). (Rasmussen 1979; Postacchini et al. 1988; Koes et al. 1992; Wreje et al. 1992; Delitto et al. 1993; Erhard et al. 1994; Triano et al. 1995) Because the evidence continues to support the use of manipulation in patients with LBP, it seems important to continue to explore the mechanism by which manipulation may work to improve pain and function in these patients. Based on the absence of evidence to explain the mechanism for how manipulation may work to improve pain and function, it has recently been suggested (Ebrall 2000) that clinicians and researchers pursue this line of research. Based on our experience and the suggestion of others (Tullberg et al. 1998) it seems reasonable to suspect that lumbopelvic asymmetries in patients with LBP that may be addressed with manipulation may be attributed to changes in soft tissues in this region; however, little research has been conducted to examine the plausibility of this hypothesis. It is possible that discrepancy in side-to-side weight-

INTRODUCTION There is little evidence to support the mechanism by which manipulation works to improve pain and function. However, many studies have demonstrated Received: 24 June 2002 Revised: 7 November 2002 Accepted: 16 January 2003 John D. Childs, PT,MS,MBA,OCS,CSCS,FAAOMPT, AFIT Doctoral Candidate, Sara R. Piva, PT, MS, OCS, FAAOMPT, Doctoral Candidate, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, USA, Richard E. Erhard, DC,PT,FAAOMPT, Assistant Professor, Head Chiropractor, University of Pittsburgh Medical Center Health System’s Spine Specialty Center, USA, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, USA, Gregory Hicks, PT,PhD, PostDoctoral Fellow, Clinical Research Branch, Gerontology Research Center, National Institute on Aging-National Institute of Health, USA. Correspondence to: JDC, 59MDW/Wilford Hall Medical Center, Physical Therapy Service/MMKPP, 2200 Bergquist Dr., Suite 1, Lackland AFB, TX 78236, USA Tel.: 210-292-7735; Fax: +1-210-292-7991; E-mail: [email protected] 166

Low back pain and weight bearing asymmetry 167

bearing (WB) between the lower extremities could be a manifestation of soft tissue or other biomechanical dysfunction in the lumbopelvic region. However, no previous studies were identified that explored whether patients with LBP exhibited discrepancies in side-toside WB through their lower extremities compared to healthy control subjects. If patients with LBP indeed exhibit increased side-to-side WB discrepancy compared to healthy control subjects, these data would assist in the design of future studies to investigate the relationship between changes in side-to-side WB asymmetry after treatment with manipulation and pain and function in patients with lumbopelvic dysfunction. Thus the purpose of this project was to determine if subjects with LBP exhibit greater side-toside WB asymmetry compared to healthy control subjects without LBP. METHODS This study utilized an observational double cohort design to compare side-to-side WB asymmetry in patients with LBP compared to healthy control subjects without LBP. Subjects Consecutive patients referred to the University of Pittsburgh Medical Center Health System’s Spine Specialty Center, with a primary complaint of LBP agreed to participate in this study. The presence of LBP was determined by the presence of pain in the low back area observed on a pain diagram (Uden et al. 1988; Chan et al. 1993; Mann et al. 1993) that patients completed before being seen by a clinician. Patients with LBP were excluded if they had a history of surgery to the lumbopelvic region of the spine within the last year. Control subjects consisted of subjects with no complaints of LBP for the past 6 months. This study was approved by the University of Pittsburgh Institutional Review Board, and all subjects provided informed consent prior to participation in this study. Thirty-five subjects with LBP and 31 healthy control subjects agreed to participate in this study. Descriptive statistics for each group can be seen in Table 1.

Measures of pain and function The patients with LBP completed the following selfreport measures of pain and function. The control subjects were not asked to complete these self-report measures because they had no complaints of LBP, and we were only interested in exploring the relationship between side-to-side WB asymmetry and pain and function in the patients with LBP. 1. An 11-point numeric pain rating scale (NPRS) ranging from 0 (no pain) to 10 (worst imaginable pain) was used to assess current pain intensity, the best, and the worst level of pain during the last 24 h (Downie et al. 1978; Makuch & Simon 1978; Jensen et al. 1994). The average of these three pain scores was used to represent the subject’s level of pain. 2. The Oswestry Disability Questionnaire (ODQ) is a region-specific measure of functional limitations and disability for patients with LBP (Fairbank et al. 1980; Stratford et al. 1994; Fairbank 1995; Beurskens et al. 1996; Fairbank & Pynsent 2000). The questionnaire consists of ten items addressing different aspects of function, each scored from 0 to 5 with higher values representing greater disability. The ODQ used in this study was modified to improve compliance. The section on sex life was replaced with one regarding employment/homemaking. Previous research has demonstrated the modified version to have high levels of reliability, validity and responsiveness (Fritz & Irrgang 2001). Measurement of side-to-side wb asymmetry Subjects were asked to stand with one foot each on two independent Thinnerr digital scales (Measurement Specialties, Inc., New Jersey) placed side-byside in such a manner that their feet were approximately shoulder width apart in a neutral standing position. The scales were calibrated at the factory as part of the manufacturing process. No instructions were provided as to which leg should be used to first step up on the scales. The weight distributed on each lower extremity was recorded to the nearest pound. The examiner was not blinded to the subject’s group assignment (i.e. LBP or control subject). However,

Table 1. Descriptive statistics for patients with LBP and controls Patients with LBP (n = 35) Gender Age (years)—mean (SD) Total body weight (kg)—mean (SD) Pain—mean (SD) ODQ—mean (SD)

60% female 38.9 (11.7) 79.9 (19) 5.4 (1.9) 48.7 (18.1)

Control subjects (n = 31)

P-value

48.4% female 31.0 (11.4) 81.7 (24.9)

0.344 0.009* 0.738

n/a n/a

n/a n/a

*Significant at an a-level=0.05 (two-tailed test). Manual Therapy (2003) 8(3), 166–169


168 Manual Therapy

after an initial fluctuation in body weight for 3–5 s as the scales normalizes, the value on the scale becomes fixed and does not oscillate thereafter. The amount of WB is then simply recorded, thus the measurement is not subject to rater bias. The measurement was repeated three times. Subjects were blinded to their side-to-side WB asymmetry during the three trials by having the examiner mask the display on the scale. Subjects also stood on one scale to obtain a total body weight. The reliability between the two scales for recording total body weight was 0.99 (95% CI: 0.99, 1.0). Side-to-side WB asymmetry was calculated as the average of the absolute value of the difference between the right and left lower extremity from the three trials. The percentage of the average side-to-side WB asymmetry relative to the total body weight was then calculated to normalize expected differences in magnitude of asymmetry based on a subject’s total body weight. We determined the reliability of measuring side-to-side WB asymmetry on the bilateral scales in this sample using the first 45 subjects (18 controls and 27 subjects with LBP). Three trials for each subject were recorded by one of two independent raters. Three additional trials were then recorded by the second rater. The amount of WB asymmetry for this calculation represented the mean ratings of the three trials for each rater. Using an intraclass correlation coefficient (formula(2,k)), (Shrout & Fleiss 1979) the reliability coefficient was 0.88 (95% CI: 0.79 – 0.93). The standard error of the measurement (SEM) was calculated to be a 2% change in WB asymmetry relative to total body weight. Sample size estimation The sample size was calculated apriori using SamplePowert statistical software (SPSS Inc., Chicago, Il) (SPSS 1998) to ensure sufficient statistical power to detect a 5% difference in side-to-side WB asymmetry between the groups with a common standard deviation of 7% using a two-tail test at an a-level set to 0.05. These parameters resulted in a sample size of 32 subjects in each group. Data analysis Descriptive statistics, including frequency counts for categorical variables and measures of central ten-

dency and dispersion for continuous variables were calculated to summarize the description of the groups (Table 1) and the WB data (Table 2). Independent ttests were performed between the groups at an a-level set to 0.05 for subjects’ age and total body weight. A Pearson chi-square test was performed between the groups based on gender at an a-level set to 0.05. (Portney & Watkins 1993). An independent t-test was performed to compare the magnitude of side-to-side WB asymmetry and normalized side-to-side differences in WB asymmetry in patients with LBP compared to the healthy control subjects. The a-level used for the comparison was apriori set to 0.05 for each comparison. The Pearson product moment correlation coefficient was calculated to determine the relationship between the magnitude of side-to-side WB asymmetry and pain and function for the patients with LBP (Portney & Watkins 1993).

RESULTS The patients with LBP were significantly older than the healthy control subjects (38.9711.7 vs. 31.0711.4, respectively), however there were no differences between the groups for total body weight (Table 1). There were no differences in the distribution between males and female subjects within each group. Patients with LBP demonstrated significantly greater magnitude of side-to-side WB asymmetry and greater normalized side-to-side WB asymmetry than healthy control subjects (Table 2). In patients with LBP, higher magnitudes of side-to-side WB asymmetry were significantly associated with increased levels of pain (r=0.39, P=0.021). There was no significant relationship between side-to-side WB asymmetry and the score on the ODQ (r = 0.26, P=0.098).

DISCUSSION The results of this study demonstrate that patients with LBP exhibit side-to-side WB asymmetry that is just more than double that found in the healthy control subjects without LBP (Table 2). Although the clinical relevance of this finding is not currently known, increases in WB asymmetry were associated

Table 2. Comparison of side-to-side WB asymmetry

Magnitude of side-to-side WB asymmetry (kg)—mean (SD) Normalized side-to-side WB asymmetry (% of total body weight)—mean (SD)

Patients with LBP (n = 35)

Control subjects (n = 31)

P-value

6.8 (5.2) 8.8 (8.0)

2.8 (1.9) 3.6 (2.6)

0.001* o0.000*

*Significant at an a-level=0.05 (two-tailed test). r 2003 Elsevier Science Ltd. All rights reserved.

Manual Therapy (2003) 8(3), 166–169

Low back pain and weight bearing asymmetry 169

with an increased level of pain in these patients. This finding suggests at least preliminary evidence that the assessment of side-to-side WB asymmetry may have some merit in these patients. Further investigation of the meaningfulness of this finding is necessary. It has recently been proposed that manipulation may act at the level of soft tissues in the lumbopelvic region to improve pain and function in patients with LBP (Tullberg et al. 1998). The exact mechanism as to how manipulation may work to improve pain and function in patients with LBP is not clearly understood. We speculate that manipulation affects the tension of soft tissues in this area, thus normalizing the transmission of forces through the lumbopelvic region and restoring more equal weight-bearing distribution in the lower extremities. We do not suggest that the results of this study lend evidence to support the validity of the role of soft tissues in their contribution to lumbopelvic dysfunction. However, the results of this study will be important in the design of future investigations to examine the plausibility of this theory. Future work in this area could investigate changes in side-to-side WB asymmetry after manipulation in addition to changes in electromyographic (EMG) activity of key muscles in the lumbopelvic region that could contribute to side-side WB asymmetry in these patients. Force platforms could also be used to explore the dynamic properties of weight shifting during the measurement. If side-to-side WB asymmetry in patients with LBP normalizes after a manipulative intervention is performed, and if this improvement is associated with improvements in EMG activity of these key muscles, this information would provide preliminary evidence to support the role of soft tissues changes such as muscles in contributing to LBP and how manipulation may act to ameliorate these alterations to improve pain and function in patients with LBP.

CONCLUSIONS Patients with LBP exhibit increased side-to-side WB asymmetry compared to healthy control subjects without LBP. This asymmetry was associated with increased levels of pain. This finding is relevant to plan future studies that will attempt to provide evidence for the construct validity of manipulation by determining if side-to-side WB asymmetry normalizes after a manipulation intervention and if this improvement is associated with improvements in pain and function. The results of this study will also help to provide evidence for the mechanism by which manipulation may work to improve pain and function in patients with LBP. Manual Therapy (2003) 8(3), 166–169

References Beurskens AJ, de Vet HC, Koke AJ 1996 Responsiveness of functional status in low back pain: A comparison of different instruments. Pain 65(1): 71–76 Chan CW, Goldman S, Ilstrup DM, Kunselman AR, O’Neill PI 1993 The pain drawing and Waddell’s nonorganic physical signs in chronic low-back pain. Spine 18(13): 1717–1722 Delitto A, Cibulka MT, Erhard RE, Bowling RW, Tenhula JA 1993 Evidence for use of an extension-mobilization category in acute low back syndrome: A prescriptive validation pilot study. Physical Therapy 73(4): 216–222 Downie WW, Leatham PA, Rhind VM, Wright V, Branco JA, Anderson JA 1978 Studies with pain rating scales. Annals of Rheumatic Disease 37(4): 378–381 Ebrall P 2000 Motion palpation: It’s time to accept the evidence. Journal of Manipulative and Physiological Therapeatics 23(7): 513–514 Erhard RE, Delitto A, Cibulka MT 1994 Relative effectiveness of an extension program and a combined program of manipulation and flexion and extension exercises in patients with acute low back syndrome. Physical Therapy 74(12): 1093–1100 Fairbank J 1995 Use of Oswestry Disability Index (ODI). Spine 20(13): 1535–1537 Fairbank JC, Couper J, Davies JB, O’Brien JP 1980 The Oswestry low back pain disability questionnaire. Physiotherapy 66(8): 271–273 Fairbank JC, Pynsent PB 2000 The Oswestry Disability Index. Spine 25(22): 2940–2953 Fritz JM, Irrgang JJ 2001 A comparison of a modified oswestry low back pain disability questionnaire and the Quebec back pain disability scale. Physical Therapy 81(2): 776–788 Gerlach UJ, Lierse W 1992 Functional construction of the sacroiliac ligamentous apparatus. Acta Anatomica (Basel) 144(2): 97–102 Jensen MP, Turner JA, Romano JM 1994 What is the maximum number of levels needed in pain intensity measurement? Pain 58(3): 387–392 Koes BW, Bouter LM, van Mameren H, Essers AH, Verstegen GM, Hofhuizen DM, Houben JP, Knipschild PG 1992 The effectiveness of manual therapy, physiotherapy, and treatment by the general practitioner for nonspecific back and neck complaints. A randomized clinical trial. Spine 17(1): 28–35 Levangie PK 1999 Four clinical tests of sacroiliac joint dysfunction: The association of test results with innominate torsion among patients with and without low back pain. Physical Therapy 79(11): 1043–1057 Makuch R, Simon R 1978 Sample size requirements for evaluating a conservative therapy. Cancer Treatment Report 62(7): 1037–1040 Mann NH, Brown MD, Hertz DB, Enger I, Tompkins J 1993 Initial-impression diagnosis using low-back pain patient pain drawings. Spine 18(1): 41–53 Portney LG, Watkins MP 1993 Foundations of Clinical Research: Applications to Practice. Appleton and Lange, Norwalk, CT Postacchini F, Facchini M, Palieri P 1988 Efficacy of various forms of conservative treatments in low back pain: A comparative study. Neuro Orthopedics, pp 628–635 Rasmussen GG 1979 Manipulation in treatment of low-back pain: A randomized clinical trial. Manual Medicine 18–10 Shrout PE, Fleiss JL 1979 Intraclass correlations: Uses in assessing rater reliability. Psychology Bulletin 86420–86428 SPSS I 1998 SPSS for Windows 1.1. (8.0) Stratford PW, Binkley J, Solomon P, Gill C, Finch E 1994 Assessing change over time in patients with low back pain. Physical Therapy 74(6): 528–533 Triano JJ, McGregor M, Hondras MA, Brennan PC 1995 Manipulative therapy versus education programs in chronic low back pain. Spine 20(8): 948–955 Tullberg T, Blomberg S, Branth B, Johnsson R 1998 Manipulation does not alter the position of the sacroiliac joint. A roentgen stereophotogrammetric analysis. Spine 23(10): 1124–1128 Uden A, Astrom M, Bergenudd H 1988 Pain drawings in chronic back pain. Spine 13(4): 389–392 Wreje U, Nordgren B, Aberg H 1992 Treatment of pelvic joint dysfunction in primary care—a controlled study. Scandinavian Journal of Primary Health Care 10(4): 310–315 r 2003 Elsevier Science Ltd. All rights reserved.


Original article

Anatomical and possible clinical relationships between the calcaneofibular ligament and peroneus brevis – a pilot study A. Dowling*, B. Downeyw, R. Greenz, P. Reddyy, J. Wickhamz Yarra Valley and Mooroolbark District Physiotherapy, Victoria, Australia, wPhysioworks, Cranbourne, Victoria, Australia, zSchool of Human Biosciences, La Trobe University, Victoria, Australia, y Epping Physiotherapy, Victoria, Australia n

SUMMARY. The authors have observed in their clinical practice patients presenting with chronic retromalleolar pain following lateral ankle injuries. It has been hypothesized that persistent retromalleolar pain following a supination sprain may be due to peroneus brevis (PB) tendon tears (Boruta et al. 1990). The aims of this study were to investigate whether an anatomical relationship exists between the calcaneofibular ligament (CFL) and PB, and if so, the significance of this relationship in the positions of supination sprain and talar tilt test. Seven out of eight cadaveric ankles demonstrated fibrous connecting tissue between the tendon of PB and CFL. Four of the eight ankles demonstrated PB tendon abnormalities. The presence of connecting tissue between CFL and PB suggests an anatomical basis for concomitant damage to the PB tendon with a supination sprain, thus supporting the hypothesis that there may be an anatomical basis for persistent retromalleolar pain subsequent to injury to the lateral ankle complex. r 2003 Elsevier Science Ltd. All rights reserved.

Bulucu et al. 1991; Brukner & Khan 2000) and account for the most time lost by athletes due to injury (Safran et al. 1999; Hunter & Fortune 2000). Supination1 sprains constitute over 80% of all ankle ligament sprains (Roy & Irvin 1983) and are said to be the cause of chronic ankle pain and disability (Major et al. 2000). Supination injuries have been shown to damage the three components of the lateral ligament complex namely, the anterior talofibular ligament (ATFL), the calcaneofibular ligament (CFL) and the posterior talofibular ligament (PTFL). Ankle sprains are graded from I to III in increasing order of severity with grade I being disruption of some ligament fibres and no loss of stability to grade III which is a complete rupture of all ligament fibres with joint instability (Bassett & Speer 1993; Brukner & Khan 2000).

INTRODUCTION The authors (AD, BD) have observed in their clinical practice patients presenting with chronic retromalleolar pain following lateral ankle injuries. This study was initiated to further investigate this problem. Ligamentous injuries of the lateral ankle joint are among the most common musculoskeletal complaint seen by the health practitioner (Boruta et al. 1990; Received: 20 February 2002 Revised: 6 January 2003 Accepted: 7 February 2003 A. Dowling (Physio), M.Manip. Physio., Private Practitioner, Yarra Valley and Mooroolbark District Physiotherapy, Victoria, Australia, Brian Downey, B.Physio. (Hons), M.Manip. Physio., Private Practitioner, Physioworks, Cranbourne, Victoria, Australia, Rodney Green, B.Sc. (Hons), M.Sc., Dip.Ed., Ph.D. Senior Lecturer, School of Human Biosciences, La Trobe University, Victoria, Australia, Prakashni Reddy, B.Physio., M.Manip. Physio., Private Practitioner, Epping Physiotherapy, Victoria, Australia, James Wickham, B.Sp.Sc., Dip.Ed., B.Sc. (Hons), Ph.D. Lecturer, School of Human Biosciences, La Trobe University, Victoria, Australia Correspondence to: RG, School of human Biosciences, La Trobe University, Victoria 3086, Australia. Tel.: +613 9479 5751; Fax: +613 9479 5784; E-mail: [email protected].

1

The term supination, rather than inversion, has been used to be consistent with the recent trend in anatomical and biomechanical texts to describe the triplanar movement occurring at the tarsal joints.

170

Calcaneofibular ligament and peroneus brevis 171

Although some ankle sprains respond readily to appropriate conservative management (Brukner & Khan 1991; DiGiovanni et al. 2000), a significant number of patients report persistent pain and swelling (Brukner & Khan 2000), recurrent instability (Larsen 1987) and chronic disability (Major et al. 2000) long after ligament repair would have been expected to occur. Ankle pain following a supination sprain which is localized to the postero-inferior aspect of the lateral malleolus has been postulated to occur secondary to tearing of the peroneal tendon sheath and damage to the peroneal tendons (Boruta et al. 1990). Although there may be anatomical variation, the anatomy of this area is typically as follows. Peroneus longus (PL) and brevis run distally in the lateral compartment of the leg and become tendinous just proximal to the lateral malleolus where they are held down by the superior peroneal retinaculum. The tendon of PB, lying anterior to PL, runs in a groove posterior to the malleolus and immediately superficial to the CFL (Fig. 1). The CFL runs from the anterior tip of the lateral malleolus postero-inferiorly to insert into the calcaneus. The PL and PB tendons are held down by the inferior retinaculum where they pass on the lateral aspect of the calcaneus inferior and superior to the peroneal trochlea, respectively. The

Fig. 1—Lateral view of peroneal tendons (PL and PB) and the calcaneofibular ligament (CFL) at the ankle joint (Modified and reproduced by kind permission of American Orthopaedic Foot and Ankle Society from Sobel et al. 1992) Manual Therapy (2003) 8(3), 170–175

PB tendon continues to insert into the base of the fifth metatarsal, whereas the PL tendon passes into the sole of the foot under the cuboid to insert into the base of the first metatarsal and medial cuneiform (Williams et al. 1995). Both muscles are innervated by the superficial peroneal nerve (L5/S1) (Williams et al. 1995; Sinnatamby 2000). Longitudinal splits or tears of the tendon or tendon sheath of PB have been implicated as the source of retromalleolar pain and disability (Major et al. 2000). Peroneal tendon injury has been demonstrated to occur concurrently with a single or repeated sprain of the ankle (Sobel et al. 1991; Bassett & Speer 1993). PB tendon splits have been graded from I to IV dependent upon the magnitude of the split with grade I being a splayed tendon and IV being a full thickness split greater than 2 cm (Sobel et al. 1992). These tendon splits have been proposed to occur secondary to mechanical attrition of the PB tendon (Larsen 1987; Sobel et al. 1990, 1992; Major et al. 2000) against the posterior aspect of the lateral malleolus due to compression by the PL tendon (Munk & Davis 1976; Sobel & Geppert 1992; Bassett & Speer 1993) (Fig. 2). Some authors suggest that the presence of a split in the PB tendon is infrequently an anatomical variant (Sobel et al. 1992; Rosenberg et al. 1997; Major et al. 2000). Additionally it has been suggested that the CFL, being extra-articular and intimately related to the peroneal tendon sheath, when disrupted following a supination sprain may cause concomitant tearing of the tendon sheath or peroneal tendons (Boruta et al. 1990). The talar tilt test is commonly

Fig. 2—Transverse section at the level of the fibular groove showing the proposed mechanism of injury of peroneus brevis (PB) due to compression by peroneus longus (PL) (reproduced by kind permission of American Orthopaedic Foot and Ankle Society from Sobel et al. 1992) r 2003 Elsevier Science Ltd. All rights reserved.

172 Manual Therapy

used clinically to assess the integrity of the CFL. In this test the tibia is stabilized with one hand and the heel is moved medially with the other and movement of the talus and calcaneus is assessed in relation to the tibia and fibula (Brukner & Khan 2000). Upon reviewing the literature no studies were found to have tested Boruta et al.’s (1990) hypothesis. In order to evaluate the evidence supporting Boruta et al.’s theoretical proposal, that retromalleolar ankle pain following a supination sprain may be secondary to tearing of the peroneal tendon sheath and damage to the peroneal tendons (Boruta et al. 1990), the aims of this study were firstly:

CFL. Any connecting tissue was examined, and its attachments documented and photographed. Any split or variation in structure of the PB tendon was graded from I to IV as described by Sobel et al. (1992). To the extent that movement was possible in the cadaver specimens, the position of the supination sprain was simulated and the talar tilt test was applied to each ankle. Tension in the CFL, PB and connecting tissue in each of these positions was observed and documented.

1. to investigate the presence of an anatomical link between the CFL and PB.

RESULTS

If such an anatomical link exists, secondary aims were: 1. to determine how this relationship may cause disruption to the PB tendon during a supination sprain, and 2. to examine the relationship between the CFL and PB in the talar tilt testing position.

METHOD Eight lateral ankle complexes were dissected in six embalmed human adult cadavers. The specimens dissected were all of those available in an anatomy teaching school on which the lateral ankle had not been previously dissected in any way and had no superficial evidence of previous ankle injury or abnormality. The skin and subcutaneous tissue were resected to expose the tendons of PL and PB restrained by the superior and inferior peroneal retinacula. The superior and inferior peroneal retinacula were then incised to release the peroneal tendons. The tendon of PB was lifted to observe whether an anatomical relationship existed connecting it and the

Fibrous attachments between the tendon of PB and the lateral aspect of the ankle joint were demonstrated in all eight ankles (Table 1). In seven of the eight ankles the tissue was observed to attach to the CFL (Fig. 3), but in specimen 1 the connecting tissue attached only to the joint capsule with no fibrous attachment to the CFL. In all cases the connecting tissue appeared thin and membranous with little similarity to ligamentous tissue, suggesting that it may be part of the parietal layer of the synovial tendon sheath. A full thickness longitudinal split in the tendon of PB was viewed in two of the specimens (Fig. 4). Both of these specimens had strong tissue attachment between the CFL and the posterior portion of the tendon of PB. Partial degeneration of the PB tendon was observed in two other tendons (Table 1). Both left and right ankles in two of the eight cadavers were dissected. In one of the cadavers there was no attachment between the CFL and PB on the left ankle (specimen 1) and a split PB tendon with strong fibrous attachment between PB and CFL on the right ankle (specimen 2). In the other cadaver there was a strong attachment between CFL and tendon of PB on the left (specimen 3) and between CFL and the split tendon of PB on the right (specimen 4).

Table 1. Descriptions of connecting tissue and PB tendons Specimen

Cadaver

Limb

Description & attachment of fibrous tissuen

Description tendon of PB tendon

1 2

86-year-old female

L R

Grade 2 split Grade 3 split

3 4

83-year-old male

L R

5 6 7 8

78-year-old male 83-year-old male 76-year-old male 90-year -old male

R R L R

Weak attachmentw between joint capsule and PB Strong attachment between CFL and posterior portion of split PB Strong attachment between CFL and PB Strong attachment between CFL and to posterior portion of split PB tendon Moderate attachment between CFL and PB Weak attachment between CFL and PB Moderate attachment between CFL and PB Strong attachment CFL and PB

Normal Grade 4 split Normal Normal Normal Grade 1 split

n

The use of the terms strong, moderate and weak is used to broadly classify the extent of the tissue present in each specimen rather than being indicative of the mechanical strength of the tissue. No attachment to CFL.

w


Manual Therapy (2003) 8(3), 170–175


Fig. 3—Lateral aspect of the right ankle with the peroneus longus retracted and the tendon of peroneus brevis (held in forceps) raised to demonstrate the connecting tissue (small arrow) between calcaneofibular ligament (large arrow)/joint capsule and peroneus brevis.

In the positions of supination sprain and talar tilt test, tightening of the PL, PB, CFL and the connecting tissue was seen to occur in all specimens. Although the tension was not quantified, increases in soft tissue tension were readily apparent to the naked eye. DISCUSSION Lateral ligament injuries due to supination sprain of the ankle joint are a common complaint seen by the health practitioner (Bulucu et al. 1991; Brukner & Khan 2000) and while most respond to conservative management a significant number report persistent retromalleolar pain (Sammarco & DiRaimondo 1989; Bassett & Speer 1993; DiGiovanni et al. 2000). Boruta et al. (1990) proposed that CFL damage following a supination sprain causes disruption of the CFL with concurrent tearing of the peroneal tendons and sheath. This study demonstrated the presence of connecting tissue which passed from the CFL and lateral aspect of the ankle joint capsule to the tendon of PB and was shown to come under tension in the positions of supination sprain and talar tilt. Although histological examination was not possible (the teaching specimens available had not been preserved in a manner which would have allowed any conclusive histological examination), the connecting tissue appears to constitute the parietal layer of the synovial tendon sheath. Gray’s anatomy describes the structure of a synovial tendon sheath as follows. The internal (visceral) layer is attached by loose connective tissue to the underlying tendon and the external (parietal) layer is similarly tethered with adjacent periosteum or other connective tissue structures. It should be added that the precise shape of the enclosed tendon and surrounding tissues varies much, from simple cylindrical to highly Manual Therapy (2003) 8(3), 170–175

Fig. 4—Lateral aspect of the right ankle demonstrating split PB tendon (large arrow) raised to demonstrate the connecting tissue between calcaneofibular ligament (CFL) and peroneus brevis (PB) (small arrow).

complex, depending on the shape of the enclosed tendon and the surrounding tissues and on other mechanical factors. (Williams et al. 1995)

However, neither Gray’s anatomy nor other recent anatomical texts (Williams et al. 1995; Moore & Dalley 1999; Sinnatamby 2000) provide any details on the attachments of the parietal layer of the peroneal synovial tendon sheathes nor mention of any other tissue connecting the CFL to the PB tendon. It appears that this level of detail in regard to the attachments of the peroneal synovial tendon sheathes has not previously been reported. Last (Sinnatamby 2000) describes the tendon sheath as being common to both the PL and PB tendons proximal to the peroneal trochlear on the lateral surface of the calcaneus. The location of the PB and the sheath covering it were found to be in close proximity to the CFL consistent with the deeper location of the PB tendon at this point. The connecting tissue/tendon sheath was observed to bear attachment only to the PB tendon sheath thus endorsing Boruta et al.’s hypothesis and may partially explain the findings of PB tears being more prevalent than PL tears (Sobel et al.’s 1990). Increased tension in the connecting tissue was observed when the cadaver ankles were placed in the supination sprain and talar tilt positions. It is possible that tension in this connecting tissue/tendon sheath generated either during a supination sprain or during a talar tilt test may contribute to damage to the PB tendon sheath or possibly the tendon itself. The current study is the first to provide anatomical evidence to support Boruta et al.’s proposal that damage to the peroneal tendons may occur as a consequence of damage to the CFL (Boruta et al. 1990). Chronic retromalleolar pain following a supination sprain may be due to longitudinal splits or tears of the tendon or tendon sheath of PB (Major r 2003 Elsevier Science Ltd. All rights reserved.

174 Manual Therapy

et al. 2000). The reported proportion of split PB tendons in the literature ranged from 11% to 37% (Sammarco & DiRaimondo 1989; Sobel et al. 1990, 1991; DiGiovanni et al. 2000). According to Sobel’s classification of PB tears, only grade III and IV tears have a full thickness split (grade I and grade II being splayed and a partial thickness split, respectively) (Sobel et al. 1992). Thus, the incidence of split tendons in this study was 25% (two of eight ankles) which is in agreement with previous reports. It is possible that this could be an asymptomatic anomaly; however, the incidence of anomalies has not been reported in the literature, and the clinical history of our ankles is unknown. Longitudinal tears in the PB tendon have been proposed to result from repetitive compression of the PB against the posterior margin of the lateral malleolus by the PL (Sammarco & DiRaimondo 1989; Major et al. 2000). This proposal does not fully explain cases of retromalleolar pain that occur following a single episode of supination sprain. It is possible, although given the nature of the tissue observed, unlikely, that tension transmitted via the connecting tissue/tendon sheath during a supination sprain could contribute to damage to the PB tendon. The presence of stretch receptors in tendons is well documented; however, recent animal studies have found nerve endings in rat tendon that are immunoreactive to both substance P (SP) and calcitonin gene-related peptide (CGRP) (Ackermann et al. 1999). These neuropeptides are known to be involved in nociception (Vanrossum et al. 1997). SP and CGRP have also been identified in neurones associated with the knee joint cavity in dogs (Tamura et al. 1998) and the synovial lining of the navicular bursa in foals (Bowker et al. 1995). It is therefore likely that these neurones, associated with nociception, are present both in tendons and synovial tendon sheaths in humans. The tearing of the tendon sheath tissue or tendon, as proposed by Boruta et al. (1990) could lead to ankle pain well after healing of the lateral ligament of the ankle has taken place. Retromalleolar pain reported by patients when performing the talar tilt test may originate in a variety of tissues, including the lateral joint capsule, the CFL, the PB tendon or tendon sheath. The findings of Bassett & Speer (1993) who found that athletes who sustained lateral ligament sprains still complained of retromalleolar pain even after their ankles were clinically stable may be related to pain originating in the PB tendon or tendon sheath. Therefore, a report of pain on the talar tilt test in acute or chronic injury situations may be due to a damaged CFL; however, it may also implicate the tendon sheath/connecting tissue or the PB tendon itself. r 2003 Elsevier Science Ltd. All rights reserved.

As cadaveric material possesses more rigidity than in vivo, extrapolation to living tissue is difficult. A study using fresh, rather than preserved, cadaveric specimens may allow more definitive examination of this tissue. The connecting tissue appeared thin and membranous with little similarity to ligamentous tissue, suggesting that it may be part of the parietal layer of the synovial tendon sheath. The connecting tissue was observed to come under tension but not tear when manually stressed. However, the tension was not quantified, and whether the tissue can transmit sufficient tension to damage the peroneal tendon or tendon sheath is yet to be ascertained, as the structure and composition of this tissue has not been determined. A histological examination of this tissue would help to establish its composition, innervation and vascularity in order to confirm this theoretical proposal. The CFL was not examined in any detail in this study (as this would have necessitated the removal of more superficial tissues (including any connecting tissues to PB). It is possible that variations in the orientation or attachments of the fibres of the CFL may be significant in any potential relationship between the CFL, connecting tissues and the PB tendon. However given the more general attachment of the connecting tissue to the lateral joint capsule variations in CFL may not be significant in the relationship between PB tendon and the connecting tissue. A further limitation of this study is the unknown living history of the subjects with respect to sporting, occupational and social pursuits. Previous ankle injury could have been responsible for the longitudinal splits observed in PB in some specimens. However, the absence of any surgical scars or obviously abnormality (other than split PB tendons) on dissection suggests that there is no reason to suspect any previous damage to structures present in the lateral ankle in the specimens dissected. Other than the age of the specimens (specified in Table 1), there is no known bias in the selection of specimens available for dissection.

CONCLUSION Tissue observed connecting the CFL/lateral aspect of joint capsule and PB was seen to come under tension in the positions of supination and talar tilt, supporting Boruta et al.’s (1990) theory that tears of the CFL can cause concomitant tearing of the tendon or tendon sheath of PB. Patients who present with chronic or recurrent retromalleolar pain following a supination sprain may have sustained damage to the CFL, connecting tissue or PB tendon or sheath. Therefore a painful talar tilt test not only implicates the CFL but also the connecting tissue/PB tendon Manual Therapy (2003) 8(3), 170–175


sheath and possibly PB. Palpation may assist in differentiating between the connecting tissue, PB or CFL. A patient with chronic ankle pain who is not responding to conservative management may need further investigations. MRI has been demonstrated to be sensitive to PB tendon tears (Major et al. 2000). Although a split PB tendon may be an asymptomatic anomaly referral for a surgical review is warranted if you suspect it to be the cause of the retromalleolar pain. Recommendations for further research would include a histological study of this connecting tissue to determine its structure and composition, also in vivo studies, for example during ankle surgery, and subjects from a broader demographic base.

Acknowledgements The authors would like to thank Kelly Deed for her assistance in the preparation of the cadavers; also Robyn Anker for her photographic expertise.

References Ackermann PW, Finn A, Ahmed M 1999 Sensory neuropeptidergic pattern in tendon, ligament and joint capsule. A study in the rat. Neuroreport 10(10): 2055–2060 Bassett FH, Speer K P 1993 Longitudinal rupture of the peroneal tendons. The American Journal of Sports Medicine 21(3): 354–357 Boruta PM, Bishop JO, Grant Braly W, Tullos HS 1990 Foot fellows review. Acute lateral ankle ligament injuries: A literature review. Foot and Ankle 11(2): 107–113 Bowker RM, Linder K, Sonea IM, Holland RE 1995 Sensory innervation of the navicular bone and bursa in the foal. Equine Veterinary Journal 27(1): 60–65 Brukner P, Khan K 1991 The difficult ankle. Australian Family Physician 20(7): 919–921, 924–930 Brukner P, Khan K 2000 Clinical Sports Medicine, 2nd edn. McGraw-Hill, Roseville, NSW, Australia Bulucu C, Thomas KA, Halvorson TL, Cook SD 1991 Biomechanical evaluation of the anterior drawer test: The contribution of the lateral ankle ligaments. Foot and Ankle 11(6): 389–393

Manual Therapy (2003) 8(3), 170–175

DiGiovanni BF, Fraga CJ, Cohen BE, Shereff MJ 2000 Associated injuries found in chronic lateral ankle instability. Foot and Ankle 21(10): 809–815 Hunter LJ, Fortune J 2000 Foot and ankle biomechanics. South African Journal of Physiotherapy 56(1): 17–20 Larsen E 1987 Longitudinal rupture of the peroneus brevis tendon. The Journal of Bone and Joint Surgery 69-B(2): 340–341 Major NM, Helms CA, Fritz RC, Spper KP 2000 The MR imaging appearance of longitudinal split tears of the peroneus brevis tendon. Foot and Ankle 21(6): 514–519 Moore KL, Dalley AF (eds) 1999 Clinically Oriented Anatomy, 4th edn. Lippincott Williams & Wilkins, Baltimore Munk RL, Davis PH 1976 Longitudinal rupture of the peroneus brevis tendon. The Journal of Trauma 16(10): 803–806 Rosenberg ZS, Beltran J, Cheung YY, Colon E, Herraiz F 1997 MR features of longitudinal tears of the peroneus brevis tendon. American Journal of Roentgenology 168: 141–147 Roy S, Irvin R 1983 Sports Medicine. Prevention, Evaluation, Management, and Rehabilitation. Prentice-Hall, Englewood Cliffs, NJ Safran MR, Benedetti RS, Bartolozzi III AR, Mandelbaum BR 1999 Lateral ankle sprains: A comprehensive review. Part 1: Etiology, pathoanatomy, histopathogenesis, and diagnosis. Medicine & Science in Sports & Exercise 31(7)(Suppl): S429–S437 Sammarco GJ, DiRaimondo CV, 1989 Chronic peroneus brevis tendon lesions. Foot and Ankle 9(9): 163–170 Sinnatamby CS (ed) 2000 Last’s Anatomy: Regional and Applied, 10th edn. Churchill Livingstone, Edinburgh Sobel M, Bohne WHO, Levy ME 1990 Longitudinal attrition of the peroneus brevis tendon in the fibular groove: An anatomic study. Foot and Ankle 11(3): 124–128 Sobel M, DiCarlo EF, Bohne WHO, Collins L 1991 Longitudinal splitting of the peroneus brevis tendon: An anatomic and histologic study of cadaveric material. Foot and Ankle 12(3): 165–170 Sobel M, Geppert MJ 1992 Repair of concomitant lateral ankle ligament instability and peroneus brevis splits through a posteriorly modified brostrom gould. Foot and Ankle 13(4): 224–225 Sobel M, Geppert MJ, Olsen EJ, Bohne WHO, Arnoczky SP 1992 The dynamics of peroneus brevis tendon splits: a proposed mechanism, technique of diagnosis, and classification of injury. Foot and Ankle 13(7): 413–422 Tamura R, Hanesch U, Schmidt RF, Kumazawa T, Mizumura K 1998 Examination of colocalization of calcitonin gene-related peptide- and substance p-like immunoreactivity in the knee joint of the dog. Neuroscience Letters 254(1): 53–56 Vanrossum D, Hanisch UK, Quirion R 1997 Neuroanatomical localization, pharmacological characterization and functions of CGRP, related peptides and their receptors. Neuroscience & Biobehavioral Reviews 21(5): 649–678 Williams PL, Warwick R, Dyson M, Bannister LH (eds) 1995 Gray’s Anatomy, 38th edn. Churchill Livingstone, Edinburgh



Technical & measurement report

Measuring range of active cervical rotation in a position of full head flexion using the 3D Fastrak measurement system: an intra-tester reliability study M. Amiri, G. Jull, J. Bullock-Saxton Department of Physiotherapy, The University of Queensland, Australia

SUMMARY. Most external assessments of cervical range of motion assess the upper and lower cervical regions simultaneously. This study investigated the within and between days reliability of the clinical method used to bias this movement to the upper cervical region, namely measuring rotation of the head and neck in a position of full cervical flexion. Measurements were made using the Fastrak measurement system and were conducted by one operator. Results indicated high levels of within and between days repeatability (range of ICC2,1 values: 0.85–0.95). The ranges of axial rotation to right and left, measured with the neck positioned in full flexion, were approximately 56% and 50%, respectively of total cervical rotation, which relates well to the proportional division of rotation in the upper and lower cervical regions. These results suggest that this method of measuring rotation would be appropriate for use in subject studies where movement dysfunction is present in the upper cervical region, such as those with cervicogenic headache. r 2003 Elsevier Science Ltd. All rights reserved.

sensitivity of motion loss for cervicogenic headache. Most external measures of cervical motion consider movements of both the upper and lower cervical regions simultaneously. However as cervicogenic headache has a primary involvement in the upper cervical segments, measurement of motion biased to the upper cervical region could be a more relevant measure. Dhimitri et al. (1998) investigated the reliability of a quantitative method of measuring upper cervical flexion and extension in vivo, but to date no research has reported the reliability of measuring range of active rotation in upper cervical region. The C1-2 segment is commonly involved in cervicogenic headache (Bogduk 2001) and the segment accounts for 40–60% of total range of neck rotation (Panjabi et al. 1988). It seems reasonable that measurement of rotation purportedly biased to this upper cervical region could be a relevant clinical measure in the differential diagnosis of cervicogenic headache. Clinically, rotation in the upper cervical region is differentiated from that in the lower cervical region by pre-tensioning the structures of mid-lower cervical spine in a full flexion position and then rotating the head on the pre-flexed cervical spine (Dvorak et al. 1984). The objective of this study was to investigate the reliability of this clinical test by assessing

INTRODUCTION Cervicogenic headache arises from musculoskeletal dysfunction in the cervical region, predominantly from the upper three cervical segments (Bogduk 1997). There is some symptomatic overlap between cervicogenic and other chronic headache forms such as tension-type headache and migraine (Sjaastad & Bovim 1991). This realizes the importance of possessing reliable measures to identify the physical signs of cervical musculoskeletal dysfunction to assist differential diagnosis. Reduced range of cervical motion is one physical diagnostic criterion for cervicogenic headache (Sjaastad et al. 1998). Zwart (1997) measured cervical range of motion in a comparative study of subjects with cervicogenic headache, migraine and tension headache and asymptomatic controls and confirmed the diagnostic Received: 18 July 2002 Revised: 7 October 2002 Accepted: 15 January 2003 Mohsen Amiri, BPhty, MPhty, PhD student, Gwendolen Jull MPhty, PhD, FACP, Associate Professor, Joanne Bullock-Saxton BPhty, PhD, Senior Lecturer, Department of Physiotherapy, The University of Queensland, Brisbane, Queensland 4072, Australia. Corresponding author: Tel.: +61 7 33654691; Fax: +61 7 33652775; E-mail: [email protected] 176

Measuring range of active cervical rotation 177

within and between day repeatability using a threedimensional external measurement system.

METHODS Subjects A sample of convenience of 15 healthy subjects was studied. Subjects were volunteers from the students and staff of The University of Queensland and comprised six males and nine females, ranging in age from 20 to 50 years (27.6077.85 years). Subjects were not considered if they had a history of headache or neck surgery or had received treatment for neck or shoulders conditions within the past 3 months. They were requested to avoid strenuous exercise involving the neck and shoulder regions 3 h prior to measurements and also to attend data collection sessions of approximately 20 min duration on two different days. Ethical clearance for the study was gained from the Medical Ethics Committee of The University of Queensland and subjects gave informed consent. Measurements Cervical movement was measured with a 3-Space Fastrak (Polhemus, Navigation Science Division, Kaiser Aerospace Vermont). The Fastrak is a non-invasive electromagnetic device, which tracks the positions of sensors relative to a source in three dimensions. The system has been shown to be accurate to within 70.21 (Pearcy & Hindle 1991). In this study, one sensor was fixed to an adjustable semi-rigid plastic headband, which was placed around the subjects’ forehead so that the sensor was aligned with the bridge of the nose. A second sensor was placed over the C7 spinous process using doublesided tape to prevent movement of the sensor in relation to the skin (Fig. 1). The leads were also secured with tape to prevent traction on the sensor. The electromagnetic source was placed in a box attached to the back of a wooden chair. The Fastrak was connected to a compatible PC, which continually recorded the position of the sensors relative to the source during each test sequence. A software program was written to format and process the data for 3D analysis of the neck movement. Data was converted into files and graphs so that the process could be visualized in real time. Data consisted of a 3 3 matrix of direction cosines, for the orientation of the forehead sensor relative to the sensor at C7. This was then analysed to give a 3D measurement of the position of the head relative to the C7. Baseline measures of full cervical flexion and rotation were taken from a neutral starting position (zeroed on the Fastrak). For the measurement of head and neck rotation in the pre-flexed cervical Manual Therapy (2003) 8(3), 176–179

Fig. 1—The position of the subjects’ head and neck and the location of Fastrak sensors in measurement active rotation range in a position of full cervical flexion.

position, the subjects fully flexed their cervical spine and in that position, rotated the head to the left and then right sides. In this test, the flexion range was recorded on the X-axis of the Fastrak system as a negative value and the head and neck rotation to the left and right sides, as negative and positive values respectively on the Z-axis (Fig. 2). Procedure Subjects received full information of the measurement protocol and tasks required, prior to commencement of the study. They were then seated on a wooden chair with a backrest and the sensors were attached. The subjects performed each movement three times, which served as a familiarization and warm-up session. For the measurement session, subjects were instructed to sit tall with their thoracic spine resting against the backrest of chair and to look straight ahead. This position was marked as the zero reference starting point on the Fastrak and prior to the commencement of each test movement, the subjects’ head was relocated to this reference position using the real time viewing facility of the software program. Measurements were taken of cervical flexion and rotation to the left and to the right. For the measurement of rotation in a position of cervical pre-flexion, the subjects were requested to first fully flex their neck. A researcher gently maintained the neck in this flexed position and the subject was asked to then turn their head to the left or right (Fig. 1). Three repetitions of the movement sequence were performed within each trial and three trials were performed on each day with a 3-min rest between r 2003 Elsevier Science Ltd. All rights reserved.

178 Manual Therapy

RESULTS The mean (7SD) for the total range of cervical flexion was 66.2177.61 and for rotation to the right and left sides were 78.6179.51 and 81.11710.31, respectively. The means (7SD) of the ranges of rotation measured with the neck in a pre-flexed position as well as the degree of flexion in which the movements were performed, on each day (inter trial) and between days (test-retest) are presented in Table 1. This table also includes the intraclass correlation coefficient (ICC) and SEM. These results indicate that the measures have good to excellent repeatability between and within days, respectively (range of ICC2,1 values 0.85–0.95). The SEMs were similar in both rotation directions, ranging from 1.6 to 1.91 for within day trials and 1.4 to 2.01 for between days trials.

Fig. 2—A sample of a Fastrak trace of measuring active cervical rotation range to the right and left sides (Z-axis) in a position of full cervical flexion (X-axis).

each trial. The order of performance of the movements was randomized between subjects on each day. Subjects were retested on a second occasion 3 days later.

DISCUSSION The results of this study suggest that the clinical method purported to bias rotation to the upper cervical region (head and neck rotation in a position of full neck flexion) as described by Dvorak et al. (1984) can be measured repeatedly both within and between days, as measured by one operator in this study. The ICCs for the measures derived by the Fastrak measurement system revealed good to excellent repeatability (Table 1) and ICC values are comparable to those of other studies (Dhimitri et al. 1998, Jordan et al. 2000). The SEM indicated that there was minimal error on repeated measurement within and between days. The flexion position can be maintained quite well during the test and subjects performed the head rotation within approximately 4 and 11 (left rotation) of full flexion (Table 2). That flexion was not maintained as fully with head rotation to the right, might account for the slightly greater proportion of total cervical rotation in this direction found in this study. This could reflect that 2 or 31 of rotation in the lower cervical segments contributed to the range of head rotation to the right. Nevertheless, this was a

Data analysis The mean of three repetitions in each trial and also the mean of three trials on both days were calculated in preparation for statistical analysis. The data were inspected to ensure that statistical assumptions were met. Calculation of skewness for all measures were within those values accepted for reliability tests (o3). Intra-class correlation coefficients (ICC2,1) were calculated from a repeated measures analysis of variance (ANOVA) to examine the within and between day reliability of the measures of rotation in the pre-flexed cervical position (Hass 1991, Domholdt 1993). The standard errors of measurement (SEM) were computed as a measure of absolute reliability, by which the variability of the scores obtained from measurement to measurement was examined (Domholdt 1993). The SEM is calculated from the square root of the error variance from the ANOVA and expresses the magnitude of the measurement error in the same units (Bland et al. 1996a, b) All analyses were computed with the SPSS program.

Table 1. The means7standard deviations (SD) and inter-trial (within day 1 and day 2) and test–retest (between days) intraclass correlation coefficients (ICC) for active cervical rotation range in a position of full cervical flexion Range of active rotation in a position of full cervical flexion (deg) Right Day

Mean7SD

One Two One vs Two Allied flexion

44.376.3 44.276.8 44.376.4

Left

SEM

ICC

Mean7SD

1.7 1.9 1.4 62.178.1

0.93 0.92 0.95

39.775.6 39.975.3 39.875.1

SEM 1.6 1.7 2.0 65.179.6

ICC 0.92 0.90 0.85

The degree of cervical flexion at the point of measurement of cranio-cervical rotation is also recorded (n=15 subjects). r 2003 Elsevier Science Ltd. All rights reserved.

Manual Therapy (2003) 8(3), 176–179

Measuring range of active cervical rotation 179 Table 2. The mean7standard deviations (SD) for the measures of total cervical flexion and rotation (n=15) Total neck movement

Mean7SD (deg)

Flexion Rotation to the right Rotation to the left

66.1777.63 78.5779.48 81.14710.29

the assessment of patients with dysfunction in the upper cervical region such as those with cervicogenic headache, for purposes of differential diagnosis and assessment of treatment outcomes.

References systematic difference, which occurred both within and between days, as reflected by the consistency of the measure within and between days (Table 1) and may reflect an anatomical variation. This study could not address the issue of validity of the clinical test, that is, whether testing rotation in a position of pre-flexion of the cervical spine, limits the rotation motion to the upper cervical region. In this study, there was no attempt to externally fixate C2 or the lower cervical region, other than by positioning and gently assisting the subject to maintain the full flexion position while they turned their head. There is a high likelihood that some rotation occurred below the C1-2 segment in the measurement of head rotation. Thus the present results must be interpreted with caution, bearing in mind the clinical assumption that the rotation was performed predominantly in upper cervical region. The conclusion of the current study might be strengthened if such an assumption could be examined in future using a form of radiographic measurement to establish the validity of the measurement.

CONCLUSION This study has indicated that the measure of head and neck rotation in a position of full cervical flexion can be performed reliably. It would be reasonable to recommend that this clinical assessment be used in

Manual Therapy (2003) 8(3), 176–179

Bland JM, Altman DG 1996a Measurement error. British Medical Journal 313: 744–753 Bland JM, Altman JM 1996b Measurement error and correlation coefficients. British Medical Journal 313: 41–42 Bogduk N 1997 Headache and the neck. In: Goadsby P, Silberstein SD (eds) Headache, 17th edn. Butterworth-Heinemann, Melbourne, Ch 22, pp 369–381 Bogduk N 2001 Cervicogenic headache: Anatomical basis and pathophysiological mechanisms. Journal of Current Pain Headache Report 5: 382–386 Dhimitri K, Brodeur S, Croteus M, Richard S, Seymour C 1998 Reliability of the cervical range of motion device in measuring upper cervical motion. Journal of Manual and Manipulative Therapy 6: 31–36 Domholdt E 1993 Physical Therapy Research: Principles and Applications, 1st edn. W.B. Sanders, Philadelphia, pp 153–157 Dvorak J, Dvorak V, Schneider W 1984 Manual Medicine, 1st edn. Springer, Berlin, Heidelberg Ch 2, pp 70–79 Hass M 1991 Statistical methodology for reliability studies. Journal of Manipulative and Physiological Therapeutics 14: 119–128 Jordan K, Dziedzic K, Jones PW, Ong BN, Dawes PT 2000 The reliability of the three-dimensional FASTRAK measurement system in measuring cervical spine and shoulder range of motion in healthy subjects. Journal of Rheumatology 39: 382–388 Panjabi M, Dvorak J, Duranceau J, Yamamoto I, Gerber M, Rauschning WG 1988 Three-dimensional movements of the upper cervical spine. Spine 13: 726–731 Pearcy MJ, Hindle RJ 1991 Axial rotation of lumbar intervertebral joints in forward Flexion. Proceeding of the Institution of Mechanical Engineers Part H- Journal of Engineering in Medicine 205: 205–209 Sjaastad O, Bovim G 1991 Cervicogenic headache. The Differentiation from common migraine: An overview. Journal of Functional Neurology 6: 93–100 Sjaastad O, Fredriksen TA, Pfaffenrath V 1998 Cervicogenic headache: Diagnostic criteria. Journal of Headache 38: 442–445 Zwart JA 1997 Neck Mobility in different headache disorders. Journal of Headache 37: 6–11



Case report

Post-herpetic neuralgia: Possible mechanisms for pain relief with manual therapy M.I. Rabey Fulham Physiotherapy, Swan Mews, Parsons Green Lane, SW6 4QT, London, UK

light touch of the affected skin (dynamic mechanical allodynia) (Baron 1996). The pathophysiology underlying ongoing symptoms is unclear. Post-mortem histological study of sufferers of PHN reveals structural differences compared to subjects, who only had acute herpetic infection but no PHN. The subjects who had PHN demonstrated atrophy of the dorsal horn, due partly to myelin loss. Further demyelination, along with axonal loss and inflammatory cells have been noted in the DRG and peripheral nerves. Fibrotic lesions were also shown in the DRG. Dorsal horn atrophy, and myelin and axonal loss and fibrosis in the DRG may not occur in subjects without ongoing pain (Watson et al. 1991). It is postulated that the allodynia is due to abnormal nociceptor sensitization. In acute herpes zoster, the inflammatory response is suggested to lead to sensitization of peripheral nociceptors and damage to peripheral neurones. This leads to a reduction in activation thresholds, spontaneous activity and increased discharge to suprathreshold stimuli at the axon terminals, and spontaneous activity in the proximity of the DRG. Ongoing afferent input from these nociceptive fibres initiates and may then maintain central sensitization (Rowbotham et al. 2001). Central sensitization, leading to Ab-mediated dynamic allodynia may occur following an inflammatory stimulus that causes intraneural oedema and immune cell activation but minimal fascicular degeneration (Eliav et al. 1999). However, there is no evidence for infection of afferents from deep tissues, and no evidence of inflammation in skin biopsies of sufferers of PHN; hence, no histopathological reasons behind prolonged sensitization of nociceptors. Following the acute infectious period, without a continuing inflammatory presence in the skin, nociceptors and hence the dorsal horn cells should return to their pre-morbid level of sensitivity (Wall 2001). Axonal loss has been noted in PHN (Watson et al. 1991). C fibre degeneration may only involve the peripheral branch leaving ectopic activity in the DRG

INTRODUCTION This case report details the history, presentation, and treatment of a patient with a diagnosis of postherpetic neuralgia (PHN). The pathophysiology behind the development of PHN following infection with the varicella zoster virus (VZV) is discussed, although the exact mechanisms remain unclear. The subjective and objective findings in this patient with PHN are highlighted to alert other clinicians to the possibility of PHN as a differential diagnosis when patients attend the clinic for what they perceive to be pain of musculoskeletal origin. Following a description of the simple treatment approach that was adopted with this patient, possible mechanisms of action for this intervention are considered in relation to what is known of the underlying pathophysiology. However, it is not the author’s intention to propose manual therapy as a standard intervention for this condition.

PATHOPHYSIOLOGY ‘Shingles’, the vesicopapular rash associated with acute herpes zoster infection is caused by reactivation of the VZV that may remain dormant in the dorsal root ganglion (DRG) of people who have had ‘chicken pox’. The acute condition involves intense inflammation in the dorsal horn, DRG, dorsal root, peripheral nerve and skin (Haanpaa 2001), and may lead to the development of PHN. However, less than 4% of patients who have an acute herpetic rash have pain 1 year later (Volmink et al. 1996). These patients often complain of constant or spontaneous burning or shooting pain and sharp or burning pain upon Received: 10 December 2002 Accepted: 9 January 2003 Martin Ian Rabey, M. Manip. Th., BSc (Hons) Phty, Chartered Physiotherapist, Fulham Physiotherapy, Swan Mews, Parsons Green Lane SW6 4QT, London. Correspondence to: MIR. Tel.: +44-207-371-7666; Fax: +44207-371-7636; E-mail: [email protected] 180

Post-herpetic neuralgia 181

to continue to stimulate the dorsal horn (Rowbotham et al. 2001). Regenerating axons may not reach their target tissues forming small neuromas, possibly due to an inability to penetrate herpetic scars (Bennett 1994). It is likely that these neurones would then develop spontaneous activity and sensitivity to adrenaline (Wall 2001). Uninjured C fibres may develop spontaneous activity when neighbouring axons degenerate, causing further afferent input to the dorsal horn (Wall 2001). One consequence of the increased afferent barrage to the dorsal horn might be excitotoxicity, causing degeneration of inhibitory interneurones. This may explain the atrophy noted in this region (Watson et al. 1991). This disinhibition facilitates pain perception. It is also possible that with these dorsal horn changes and C fibre deafferentation novel anatomical connections due to sprouting may also contribute to allodynia (Rowbotham et al. 2001). Hence, despite reduced afferent input to the central nervous system any remaining input is amplified (Wall 2001). Overall weight of evidence suggests that even if the acute inflammatory response sensitizes peripheral nociceptors this acts merely to initiate central sensitization. The mechanical allodynia in PHN is from then on centrally mediated. This case report details the presentation of a patient with the diagnosis of PHN. It suggests mechanisms for the reduction in the patient’s pain due to the manual therapy that was administered, relating these mechanisms to the proposed pathophysiology of the disease. It is not the author’s intention to propose manual therapy as a standard intervention for this condition.

CASE PRESENTATION History A 71-year-old female presented with left heel pain of 2 months duration. She described her pain as intermittent bony tenderness, accompanied by stiffness in her left calf musculature. The sensation had come on the morning after spending approximately 15 min in the garden simultaneously reaching towards plants in one direction, and speaking to a neighbour in the opposite direction. On questioning she admitted that initially the pain had been down the whole of the posterolateral aspect of her thigh and calf, but not in the low back, and had been sharper in nature than it was when she presented to physiotherapy. On asking whether she had consulted anyone else regarding her pain she said that she had had five sessions of acupuncture that had relieved her of virtually all of her symptoms except for the heel pain. The heel pain had plateaued some time prior to her first physiotherapy consultation. When asked if she Manual Therapy (2003) 8(3), 180–184

Fig. 1—Body chart.

had consulted her general practitioner (GP) regarding the symptoms she said no, but she had spoken to the GP a couple of days after the onset of pain, because she had developed a rash on the posterolateral aspect of her left thigh (see Fig. 1). She telephoned the GP, who asked if she had recently changed her washing powder. Since this was the case, the GP stated that she should return to her normal washing powder, and that her rash would disappear within a few days, which it did. However, when asked whether the rash had been anywhere else on her body the patient’s response was negative. Her most aggravating factor for the heel pain was heel strike during gait when the pain was rated at 7/10. Her maximum walking distance was 200 m. This led to obvious activity limitation. Pain was markedly eased immediately when non-weightbearing. She also complained of pain when supine with pressure on the heel. The patient was otherwise well, took no medication, and described no symptoms of cord or cauda equina involvement. Objective assessment On examination, the patient had numerous small dry darkly pigmented scars on her posterolateral thigh and upper calf. Observation of gait revealed decreased weight bearing on the left, with an early heeloff, but decreased push off. Active lumbar flexion and right lateral flexion provoked a ‘tightness’ sensation in the posterior aspect of the left thigh. Other lumbar movements were unremarkable. Provocative tests of the hip and sacroiliac joints were negative. Peripheral neurological screening revealed notable weakness of plantarflexors, and some weakness of hamstrings, peroneii and toe flexors; and decreased appreciation r 2003 Elsevier Science Ltd. All rights reserved.

182 Manual Therapy

of light touch at the great toe. Neural tissue provocation tests revealed straight leg raising (SLR) to provoke ‘tightness’ in the left posterior thigh accompanied by palpable hamstring activity at 501. The right hand side was less ‘tight’ at 801. Palpation over the medial calcaneal branch of the tibial nerve was exquisitely tender on the left compared to the right. Passive physiological intervertebral motion palpation revealed grade I restriction in all directions at L5/S1. Passive accessory intervertebral motion palpation revealed allodynia when pressure was applied over the left L5/S1 zygapophyseal joint. At this point, the physiotherapist asked the patient’s GP to attend the consultation. The GP confirmed that the patient had had ‘chicken pox’, and that the scars on the posterolateral aspect of the leg were as a result of ‘shingles’. A diagnosis of PHN was made. The most affected nerve root was probably S1. Treatment At this stage it was suggested to the patient that there is no physiological reason why manual therapy should impact upon a virally caused neuropathy. Also with the affected nerve root appearing to be S1, it was stated that it was improbable that manual therapy would be able to have any physiological effect on this structure. Since manual therapy interventions aim to have beneficial physiological effects by improving motion of an affected segment, and that the closest motion segment was L5/S1, treatment at this segment could only affect the transiting S1 root. The virus remains dormant in the DRG, which relates anatomically to the intervertebral foramen (Taylor & Twomey 2000). It would be impossible to achieve a physiological effect near the S1 DRG, since it would be impossible to alter motion at the S1/2 region of the sacrum. The GP suggested that he should review the patient in 2 months time and if she were still experiencing symptoms he would prescribe appropriate medication. It was suggested to the patient that since they were already at the clinic that one session of manual therapy could be undertaken as a trial. Again, the highly unlikely possibility of treatment being beneficial was outlined. Three minutes of left unilateral posteroanterior oscillatory pressure was applied over the L5/S1 zygapophyseal joint. No change in SLR was achieved. The patient was contacted by telephone 3 days later. She stated that she had gained a reduction in pain intensity following treatment. She returned to the clinic 4 days later. She now rated her pain on heel strike as 4/10. Lumbar flexion still provoked a ‘tight’ sensation in the posterior left thigh. SLR was sensitive at 601. The second treatment involved further mobilisation over the left L5/S1 zygapophyseal joint. This improved SLR to 75 degrees before ‘tightness’, and reduced this sensation r 2003 Elsevier Science Ltd. All rights reserved.

on lumbar flexion. At the end of the session, the patient described a further reduction in heel pain when walking. Nine days later, the patient reported pain to be 2/ 10, and predominantly present in the evenings. Pain on ambulation was much reduced, and maximum walking distance was half a mile. SLR produced ‘tightness’ at 701, which was mimicked by lumbar flexion. Further mobilization as before improved SLR to 751, and reduced the ‘tightness’ on lumbar flexion. One week later, the same treatment was repeated. Plantarflexor muscle strengthening exercises were also implemented, as the weakness that was evident remained unchanged. A further week later and pain levels were reported as 0.5/10. Walking was no longer restricted, and the patient had resumed her activities in the garden. SLR was now equal left side compared to right, and there was no ‘tightness’ on lumbar flexion. Palpation over the left L5/S1 zygapophyseal joint still revealed slight hyperalgesia, as did palpation over the medial calcaneal branch of the tibial nerve. At this stage, weakness was only appreciable in the plantarflexors. Further mobilization was undertaken, and strengthening exercises progressed. A review took place nearly 4 weeks later. Pain was reported as 0.1/10, and only present intermittently when supine. Strength was improving. The exercises were progressed. One month later, the patient complained of no pain. Examination still highlighted slight weakness of the plantarflexors. Passive physiological intervertebral motion palpation revealed grade II restriction at L5/S1 on flexion and extension, and passive accessory intervertebral motion evoked no discomfort. There was still slight hyperalgesia on palpation over the medial calcaneal branch of the tibial nerve. Strengthening exercises were progressed once more. At followup 3 months later the patient had no pain and no functional weakness, despite some degree of weakness still being evident on formal testing of the plantarflexors.

DISCUSSION The VZV remains dormant in the DRG. PHN normally presents with only sensory symptoms. However, clinically detectable myotomal weakness can occur in 1–5% of patients (Haanpaa 2001), and abnormal EMG activity may occur in up to 80% (Haanpaa et al. 1997). One might therefore argue that the diagnosis of PHN may have been incorrect, since clinically detectable segmental muscle weakness is uncommon. It is possible, but unlikely, that the patient may have had an acute attack of ‘shingles’ producing no subsequent PHN, and a concurrent spinal disorder that resolved with manual therapy. Manual Therapy (2003) 8(3), 180–184

Post-herpetic neuralgia 183

Bearing in mind the pathology of PHN it was thought unlikely that manual therapy would have impacted upon the symptoms. It is possible that natural history was the cause of the patient’s improvement; however, the patient stated that there had been a plateau in the recovery for some time prior to attending for physiotherapy. At least 85% of patients who have an acute herpetic rash will have no pain 1 month later. Of those still having pain, only one-quarter of these will still be symptomatic 1 year later (Volmink et al. 1996). The placebo effect, and other indirect effects (clinician charisma and enthusiasm, rational explanation for symptoms, empathy, paternalism, etc.) must also be taken into consideration, but it was emphasized to the patient that there was no reason why manual therapy should have had an effect on a virally induced neuropathy. This however cannot rule out these effects. Pharmacological studies have shown that pain relief can be obtained with tricyclic antidepressants and topical capsaisin (Volmink et al. 1996). Antiviral, corticosteroid (Lancaster et al. 1995) and antiprostaglandin medication, as well as TNS and acupuncture have insufficient evidence to prove their effectiveness in PHN (Volmink et al. 1996). The author was unable to find any evidence of the use of manual therapy in the treatment of PHN. When considering how manual therapy may have benefited the patient explanations must be based upon physiological processes that may be affected by this intervention. A possible theoretical basis for the patient’s improvement will be discussed. An oscillatory posteroanterior pressure was applied over the left L5/S1 zygapophyseal joint. The application of this technique may have had effects on both the immediate environs of the neural tissue, and upon its physiological functioning. However, it is unlikely that it will have had any direct effect on the activity of the VZV that remains present within the DRG. In PHN, authors describing histological examination have noted inflammatory cells in both the peripheral nerves and the DRG (Watson et al. 1991). The L5/S1 zygapophyseal joint lies in immediate proximity to the DRG of L5 (Taylor & Twomey 2000). Manual therapy may have offered some beneficial physiological effects to this inflammatory reaction (Elvey 1986). However, the most affected DRG was postulated to be S1. Toe flexor weakness and allodynia/hyperalgesia over the medial calcaneal branch of the tibial nerve (S1, 2, Williams et al. 1995) suggest that the L5 nerve root cannot have been the main site of pathology. If this is the case manual therapy applied over the L5/S1 zygapophyseal joint would only have been able to have an indirect effect, if any, on the transiting S1 root. Spontaneous activity in neuromas may be suppressed by the repeated application of C fibre stimulation. The technique applied in this case was an Manual Therapy (2003) 8(3), 180–184

oscillatory posteroanterior pressure. The frequency at which the oscillation occurs during joint mobilization techniques is within that capable of suppressing this spontaneous activity (1–5 Hz) (Devor & Dubner 1988). However, it is necessary to query whether C fibres would actually have been stimulated with this technique. The intervention was certainly applied at an intensity that reached the patient’s pain threshold, but if the underlying pathophysiology is borne in mind the pain perceived may have been as the result of Ab activation. However, if stimulation were at greater than 10 Hz it may block conduction of impulses from Ab fibres to wide dynamic range neurones via their small collateral branches and hence reduce symptoms (Campbell et al. 1990). Noxious stimuli applied to the affected segment may stimulate Ad fibres, which via interneurones can inhibit further transmission along ‘pain pathways’. Ab activation at the affected segment may, via inhibitory interneurones, presynaptically inhibit onwards transmission. These possibilities however assume normal architecture of the dorsal horn that might not be present in PHN. It has been suggested that treatment directed towards the L5 DRG via the L5/S1 zygapophyseal joint was incorrect, since the S1 DRG was the level likely to be the site of greatest pathology. While the rash associated with reactivation of the VZV is usually restricted to one dermatome, the pathological changes underlying the disease are not. Widespread, multi-segmental inflammatory changes involving both the dorsal and ventral horn have been described (Haanpaa 2001). There are numerous possible explanations for pain reduction by treatment of a level that is not the main site of pathology. Vibration applied to the affected dermatome in experimentally induced pain had no effect on pain intensity. However, when applied to an adjacent dermatome ipsilaterally pain intensity was reduced (Yarnitsky et al. 1997). If afferent noxious input generated by treatment was to reach the spinal cord at a level different to the most sensitized level it is still possible that nociceptive stimuli from the affected level would be prevented from reaching consciousness. This noxious input at a distant level of the spinal cord may lead to inhibition of the dorsal horn at all other levels, including the level associated with the pathology (Diffuse noxious inhibitory control) (Le Bars & Villanueva 1998). Painful stimuli during the treatment may have elicited a response from the dorsal periaqueductal gray, which in turn causes descending noradrenergic analgesia at spinal level (Wright 1995). Motor symptoms in PHN may be related to disease processes within the central or peripheral nervous systems. Anterior horn and ventral root involvement may occur (Haanpaa et al. 1997). Inflammatory changes have been noted in ventral roots on r 2003 Elsevier Science Ltd. All rights reserved.

184 Manual Therapy

histologic examination (Watson et al. 1991). Spread of inflammation to mixed peripheral nerves may account for motor symptoms (Akiyama 2000). The spread of inflammation is not restricted to one segment of the spinal cord, but extends above and below the level of the rash, and contralaterally. EMG changes to support this spread have been documented (Cioni et al. 1994; Greenberg et al. 1992; Haanpaa et al. 1997). It is postulated that secondary degeneration of motor neurones or anterior horn cells may occur following the widespread inflammatory response (Haanpaa et al. 1997). The patient’s motor weakness did not initially improve when only the oscillatory passive movements were performed. This intervention may not have had a direct effect on the pathophysiology underlying the weakness. When strengthening exercises were commenced improvement was achieved. It is possible that as the gross inflammatory response reduced motor function resolved. Complete or near complete recovery of motor weakness is expected in over 75% of patients with PHN (Thomas & Howard 1972). CONCLUSION This case report details the history of a patient with a diagnosis of PHN who responded to manual therapy intervention. Although this is a virally mediated neuropathy there are possible mechanisms through which the intervention prescribed may have had an effect on the underlying pathophysiology. However, it is impossible to differentiate which if any of these possible effects had any impact upon this patient. Indeed, it is also impossible to categorically state that the improvement achieved was not due to natural history, placebo and other indirect effects. It is not the author’s intention to propose manual therapy as a standard intervention for this condition. References Akiyama N 2000 Herpes zoster infection complicated by motor paralysis. Journal of Dermatology 27: 252–257 Baron R 1996 Pathophysiological mechanisms in post-herpetic neuralgia. Pain Clinic 9(2): 217–233 Bennett G 1994 Hypotheses on the pathogenesis of herpes zosterassociated pain. Annals of Neurology 35: S38–S41


Campbell J, Davis K, Meyer R, North R 1990 The mechanism by which dorsal column stimulation affects pain: Evidence for a new hypothesis. Pain (Supplement 5, Abstract 443): S228 6th World Congress on Pain, Adelaide, Australia, 1–6 April Cioni R, Giannini F, Passero S, Paradiso C, Rossi S, Fimiani M, Battistini N 1994 An electromyographic evaluation of motor complications in thoracic herpes zoster. Electromyographical and Clinical Neurophysiology 34(2): 125–128 Devor M, Dubner R 1988 Centrifugal activity in afferent C-fibers influences the spontaneous afferent barrage generated in nerveend neuromas. Brain Research 446: 396–400 Elvey R 1986 Treatment of arm pain associated with abnormal brachial plexus tension. Australian Journal of Physiotherapy 32: 225–230 Eliav E, Herzbery U, Ruda M, Bennett G 1999 Neuropathic pain from an experimental neuritis of the rat sciatic nerve. Pain 83: 169–182 Greenberg M, McVey A, Hayes T 1992 Segmental motor involvement in herpes zoster: An EMG study. Neurology 42: 1122–1123 Haanpaa M 2001 Neurological complications of herpes zoster. In: Watson C, Gershon A (eds) Herpes Zoster and Postherpetic Neuralgia, 2nd edn. Elsevier, London, ch 8, pp 89–96 Haanpaa M, Hakkinen V, Nurmikko T 1997 Motor involvement in acute herpes zoster. Muscle and Nerve 20: 1433–1438 Lancaster T, Silagy C, Gray S 1995 Primary care management of acute herpes zoster: systematic review of evidence from randomised controlled trials. British Journal of General Practice 45: 39–45 Le Bars D, Villanueva L 1988 Electrophysiological evidence for the activation of descending inhibitory controls by nociceptive afferent pathways. Progress in Brain Research 27: 275–298 Rowbotham M, Baron R, Petersen K, Fields H 2001 Spectrum of pain mechanisms contributing to PHN. In: Watson C, Gershon A (eds) Herpes Zoster and Postherpetic Neuralgia, 2nd edn. Elsevier, London, ch 16, pp 183–195 Taylor R and Twomey L 2000 The natural history of the lumbar spine. In: Twomey L, Taylor R (eds) Physical Therapy of the Low Back, 3rd edn. Churchill Livingstone, London, ch 1, pp 1–58 Thomas J, Howard F 1972 Segmental zoster paresis: A disease profile. Neurology 22: 458–466 Volmink J, Lancaster T, Gray S, Silagy C 1996 Treatments for postherpetic neuralgia: A systematic review of randomised controlled trials. Family Practice 13: 84–91 Wall P 2001 An essay on the mechanisms which may contribute to the pain of postherpetic neuralgia. In: Watson C, Gershon A (eds) Herpes Zoster and Postherpetic Neuralgia, 2nd edn. Elsevier, London, ch 17, pp 197–208 Watson C, Deck J, Morshead C, Van der Kooy D, Evans R 1991 Post-herpetic neuralgia: Further post-mortem studies of cases with and without pain. Pain 44: 105–117 Williams P, Bannister L, Berry M, Collins P, Dyson M, Dussek J, Ferguson M (eds) 1995 Gray’s Anatomy, 38th Edn. Churchill Livingstone, London Wright A 1995 Hypoalgesia post-manipulative therapy: A review of a potential neurophysiological mechanism. Manual Therapy 1: 11–16 Yarnitsky D, Kunin M, Brik R, Sprecher E 1997 Vibration reduces thermal pain in adjacent dermatomes. Pain 69: 75–77

Manual Therapy (2003) 8(3), 180–184

Manual Therapy (2003) 8(3), 185–186 r 2003 Published by Elsevier Science Ltd. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00040-7

Abstracts—Manipulation Association of Chartered Physiotherapists (UK) Research Awards 2002

Manipulation Association of Chartered Physiotherapists (UK) Research Presentation Award: A pragmatic randomized controlled trial of stabilization exercises in the management of recurrent low back pain

Manipulation Association of Chartered Physiotherapists (UK) Student Research Awards: Criteria for the determination of a cervicogenic component of headache based on signs and symptoms: The expert clinician’s perspective

Introduction: Altered muscular function of the deep abdominal and back muscles has been implicated as a factor in the development and continuation of low back pain (LBP). Exercise techniques designed to address these dysfunctions have become increasingly popular. Although small-scale studies, on specific sub-groups have suggested that addressing this dysfunction may improve outcome, their effectiveness in a recurrent, non-specific LBP population is unknown. A multi-centred, pragmatic randomized clinical trial was designed to investigate the effectiveness of incorporating stabilization exercises within the physical therapy management of recurrent LBP patients. Methods: Following ethical approval, consenting patients with recurrent LBP, without significant levels of distress (as measured by the Distress Risk Assessment Method (DRAM), were randomized into two groups:

Randomization was stratified for laterality, duration of symptoms and initial functional level (Roland Morris Disability Questionnaire—RMDQ). Functional ability (RMDQ) was the main outcome, and generic, diseasespecific and psychological measures were also collected. To detect a 5-point difference between groups using 90% power, a total of 97 patients were recruited from three physical therapy departments in the UK. Results: All patients were between the ages of 19 and 60 years (mean 38.6, SD: 10.5). The mean change (reduction) and 95% CI in RMDQ score from baseline to 6-months was 5.19 (3.48– 6.90) for the CT (n=37) and 4.97 (3.30–6.64) for the SSSE group (n=36) with 76% follow-up. The mean difference between groups at 6 months was 0.22 (95% CI 2.14 to 2.57; P=0.75, df=71). Discussion: Both groups achieved a clinically important improvement in function, which was maintained in the short-term (6 months). The addition of stabilization exercises to conventional physical therapy did not provide any obvious additional benefit in the short term. Full-term follow-up (12 months) will be completed by November 2001 and presented. Presented by Dr M. Cairns at the International Society for the study of the Lumbar Spine Annual Conference, Cleveland, OH, USA 14–18 May 2002. Contact—[email protected]

Objectives: The treatment rationale for any condition should be based upon the identification of physical impairments in the musculo-skeletal system, which are linked, in this case, to signs and symptoms of cervicogenic headache (Jull et al., 1999). Current diagnostic guidelines for cervicogenic headache (IHS, 1988; IASP, 1994; Sjaastad et al., 1998) are strongly medically based and may lack the precision required for physiotherapists to use within clinical practice. The purpose of this study was to develop a clinically relevant set of criteria to determine a cervicogenic component of headache based on signs and symptoms from the expert clinician’s perspective. Methodology: Following an extensive review of the literature a preliminary set of diagnostic criteria was created. A qualitative approach was selected using focus groups (Kruger & Casey, 2000). The first explored the clinical reasoning of six expert clinicians from three manual therapy organizations (MACP, SOM and the McKenzie Institute), with respect to headache patients. The second utilized a group of 5 MACP members to examine the acceptability and usefulness of the criteria. Results: The final criteria included symptom characteristics with one essential feature and ten desirable features. Secondly, signs of neck involvement incorporated two essential signs and one desirable sign with operational definitions describing the examination findings that must be present for each criterion. The results of the groups highlighted the complexity of differential diagnosis. However, it was possible to identify a series of diagnostic pointers. Although it was recognized that each clinician may manage their patients differently, reflecting the professional autonomy of the individual, the group discussion demonstrated major differences in clinical perspectives across the manual therapy groups. The MACP members were broadly able to agree on a number of basic criteria providing there was flexibility to explore a pathway further from an individual perspective. Conclusions: Focus groups can serve a useful role as a first stage in the development of criteria to determine a cervicogenic component of headache. At present, there appears to be a gap between the clinician’s perspective and the limited research available at this time to support the examination tests used. This pilot study involved a small group of experts, thus it would be advisable to conduct a larger scale study, perhaps on an international level, to see whether consensus could be reached between different manual therapy organizations. Contact—[email protected]

M CAIRNS

J BARGE

* *

‘conventional’ physical therapy (CT). ‘conventional’ physical therapy with additional specific spinal stabilization exercises (SSSE).

185

186 Manual Therapy

REFERENCES Headache Classification Committee of the International Headache Society—IHS 1988 Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 8(Suppl.7): 1–96 International Association for the Study of Pain—IASP 1994 Cervicogenic Headache. In: Merskey H, Bogduk N (eds) Classification of Chronic Pain. Descriptions of Chronic Pain Syndromes and Definitions of Pain Terms, 2nd edn. IASP Press, Seattle pp 94–95 Jull GA, Barrett C, Ho P 1999 Further clinical clarification of the muscle dysfunction in cervical headache. Cephalalgia 19: 179–185 Krueger RA, Casey MA 2000 Focus Groups: A Practical Guide for Applied Research, 3rd edn. Sage, London Sjaastad O, Fredriksen T A, Pfaffenrath V 1998 Cervicogenic headache: Diagnostic criteria. Headache 38: 442–445

Manipulation Association of Chartered Physiotherapists (UK) Student Research Awards: The effect of nerve gliding exercises on upper limb tension test 1 in a group of office workers Objectives: To determine if a nerve gliding exercise programme has an effect on upper limb tension test 1 (ULTT1). The development of non-specific arm pain (NSAP) amongst keyboard workers has been widely acknowledged, however there is little written on the use of exercise programmes to prevent the onset of such symptoms. This research investigated the effects of a nerve gliding exercise programme on the onset of resistance (R1) in ULTT1. Methodology: Measurements were carried out on asymptomatic office workers, using a keyboard for >40% of their work time. Subjects were screened prior to participation to exclude those whose ULTT1 may have been influenced by external factors. They were assigned to either an exercise (n=10), who were instructed in postural correction and carried out a programme of nerve gliding exercises, or control (n=8) group who were instructed only in postural correction. Subjects were aged between 20 and 45 years. Results: Following a 4-week period of nerve gliding exercises, subjects in the exercise group demonstrated a significant improvement in ULTT1 (P=0.01 for left and right arm measurements), compared to the control group. Measurements of scapular positioning showed a trend

r 2003 Published by Elsevier Science Ltd.

towards improvement after the study, however following statistical analysis these results were not significantly different from pre-study measures. Conclusions: This research indicates that nerve gliding exercises may be beneficial as part of a programme aiming to prevent the onset of NSAP in keyboard users. Contact— [email protected] G KLEIN

The effects of unilaterally applied lumbar mobilization technique on peripheral sympathetic activity in the lower limbs Objectives: To determine the nature of the sympathetic nervous system effects of a physiotherapeutic accessory joint mobilizations technique clinically utilized in the management of the symptoms of low back and unilateral leg pain. Methodology: A double blind independent groups placebo controlled design with random assignment of 45 normal healthy males subjects (aged 18–25) into one of three intervention groups (15 per group) undergoing either control, validated placebo or treatment conditions. The treatment condition comprised of a left sided L4-5 facet joint mobilization performed at a rate of 2 Hz. All subjects underwent the experimental condition whilst having their Skin conductance (SC) values monitored in the pre-, periand post-experimental periods. Results: Results indicated that there was a significant (Po0.001) increase in SC values (114.5%) (compared to placebo and control conditions) that were specific to the side treated for the treatment group during the intervention period. The times taken to maximum and minimum effects were not side-specific. Conclusions: A unilaterally applied lumbar joint mobilization technique can influence both spinal and supra-spinal reflex pathways and provides a link between published upper quadrant research and the lumbar spine and lower limb. This provides preliminary evidence for the use of joint mobilisation to influence the nervous system. Contact— joperry [email protected] J PERRY

Manual Therapy (2003) 8(3), 185–186

Manual Therapy (2003) 8(3), 187–188 r 2003 Published by Elsevier Science Ltd. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00045-6

Book review Clinical application of neuromuscular techniques. Volume 2: the lower body. Leon Chaitow & Judith Walker DeLany. 1st edn. 2002. Churchill Livingstone. Price: d49.99. 598pp. ISBN: 0443-06284-6.

The malalignment syndrome: biomechanical and clinical implications for medicine and sports. Schamberger Wolf. 2002. 1st edn. Churchill Livingstone. Price: d39.99. 450pp. ISBN: 0443064717. The author has challenged the problem of writing a book on a specific syndrome, called the malalignment syndrome. All health professionals working with the locomotor system know and realise the problems that may result from an inbalanced posture caused by e.g. an exaggerated inversion of the knee, or the movement restrictions resulting after ankle trauma etcetera. This book focuses on the most common presentations of malalignment, the biomechanical changes associated with them, their diagnosis and treatment. It outlines the significance of these changes to clinicians working in general practice, sports medicine and other medical specialities, and those involved in specific sports. The importance of the active participation of the patient/athlete in the day-to-day treatment process in order to achieve long-term results is stressed. Every work dealing with a syndrome has to challenge the difficulty of delineating the concept. This is what the author does in his first chapter, by describing its common presentations and diagnostic techniques. In this chapter the author refers to many researchers who have focused on the sacroiliac joint and the lumbopelvic region. In the second chapter the different presentations of malalignment are discussed in detail while in the next chapter the associated pain phenomena are discussed with reference to the implications for several specialist medical domains like cardiology and neurology etcetera. The relevance of malalignment syndrome for specific sports is discussed in the next chapter, while a separate chapter is dedicated to horse riding. In the last two chapters, the possible treatment approaches of malalignment are extensively discussed with a special short chapter dedicated to manual therapy modalities. The author has to be congratulated on this work and the step-by-step approach to the subject. While reading the first chapter it becomes more and more obvious that malalignment syndrome, just like most other syndromes, refers to a more or less specific combination of a varying number of aspecific findings in the clinical examination. Although the author has used an extensive list of literature in his analysis of malalignment syndrome, the references are used somewhat unsystematically. References seem to be included where they fit the purpose of supporting the analysis made by the author, not necessarily providing underpinning scientific evidence. From this point of view it is somewhat disappointing that there is no discussion for example on the reliability of the diagnostic process of malaligment, which mainly relies on inspection and palpation. For those working in the field and having to deal with inbalances in the musculoskeletal system this book may be very interesting, and it can offer a base for a more critical approach to the problem of malalignment.

This book is the companion to Clinical application of neuromuscular techniques, Volume 1: the upper body. This expansive book consists of 14 chapters. The first six chapters propose a variety of biological models underpinning the assessment and treatment techniques described in the remaining chapters. The authors state that these background chapters complement and further develop the theoretical information presented in Volume 1. Although a fascinating read the striking feature of these chapters is the glaring omission of a scientific basis to the opinion presented. The reader is provided with plausible hypotheses for which little or no evidence of any depth is provided. In addition the use of jargon, for example, gaiting or acture has the effect of hiding the message rather than clarifying it. Chapters 7–9 present ideas of self help strategies, aspects to consider during examination and a summary of modalities. The book then focuses on the lower limb where chapters (10–14) each deal with the lumbar spine or region of the lower limb. The general plan of each chapter is to describe the regional anatomy, provide an overview of assessment techniques and then concentrate on the anatomy, assessment and treatment techniques of associated muscles. The summary boxes, dense in print, provide a wealth of opinion. The descriptions of suggested techniques are clear and well illustrated. Written in an authoritative manner these chapters however lack any acknowledgement of issues of reliability or validity. Patient self-help exercises in a format designed to be copied and used are presented in an appendix. This is an excellent idea although some of the text is overly detailed for general patient use. This book provides a wealth of expert opinion and a useful compendium of neuromuscular techniques. However, the reader should be aware that greater justification is required before accepting propositions based on opinion and anecdote. Susan R Merce, BPhty(Hons), PhD, FNZCP Department of Anatomy & Structural Biology University of Otago, Dunedin, New Zealand

A manual therapist’s guide to surface anatomy & palpation skills. David Byfield & Stuart Kinsinger. 2002. ButterworthHeinemann. Price: d25. 173pp. ISBN: 0750644842. This new book is written by two chiropractic teachers and dedicated to students in manual therapy studies. In their work they were assisted by Matthew Clancy and Vanessa Kelly. The authors consider a complex psychomotor skill like palpation as essential for all students and practitioners of manual therapy. From this point of view the first chapter of the book is very important. Besides several didactic issues it gives a critical consideration of the scientific value of

Erik Cattrysse, PT, Msc MT Vrije Universiteit Brussel, Belgium 187

188 Manual Therapy

palpation skills and discusses the reliability and validity of this technique for both static and motion palpation. Although the scientific value of palpation techniques seems to be low and debatable, the authors correctly conclude that there is not enough information to exclude palpation techniques from clinical practice. Indeed they consider palpation as only one of the instruments used in the clinical evaluation of the patient. The final decision on a patients situation should be made on the combined findings of anamnesis, palpation and functional examination techniques. The second chapter gives a basic introduction to general active movement analysis. In the next five chapters palpation techniques are described and demonstrated by many clear pictures for each anatomical region. The lumbo-pelvic region, the thoracic spine, the cervico-thoracic spine and occipital region, the lower extremities and upper extremities are separately discussed by chapter. The book differs from other manuals on palpation in that it also considers motion palpation from a manual therapy perspective. It presents an interesting link between two previously completely separated aspects of evaluation in physiotherapy and manual therapy. Though it will always be difficult to cover the complete range of manual palpatory evaluation techniques, the authors have made a well worth attempt from a didactic perspective.

Erik Cattrysse, PT, Msc MT Vrije Universiteit, Brussel Belgium

Journal of whiplash & related disorders. Eds. C.J. Centeno & M.D. Freeman. Volume 1, Number 1, 2002. The Haworth Medical Press. Price: $ 45/year. 123pp.

The first peer-reviewed multidisciplinary journal concerning the problem of whiplash from clinical, biomedical, engineering, societal and traffic safety viewpoints has been published. It delivers recent information on whiplash and helps care providers, researchers, and politicians to access high-quality whiplash papers.

r 2003 Published by Elsevier Science Ltd.

The first issue consists of eight original papers, an abstract section listing several abstracts that have been recently published in different journals and information about future international whiplash congresses. The first paper reports on the development and methodological quality of the Bournemouth Questionnaire for the neck. This outcome measure is considered to be a valid, reliable, responsive and practical instrument for use in clinical practice. The second paper describes the role of cervical muscles in whiplash injury, and the third paper presents a new method for computing changes in velocity. The next paper is a case report. It describes the post-mortem pathoanatomical changes and their possible relation with the chronic symptoms of a man who sustained in a car accident 12 years ago. The fifth paper retrospectively reports on the presence of inner ear damage in chronic whiplash cases, and the sixth paper emphasizes the diagnosis of an additional mild traumatic brain injury during whiplash, especially for patients with impairments in cognition and coping with pain. The seventh paper analyses cervical spine motion during simulated rear-end collisions with human volunteers. The authors concluded that abnormal segmental movements mainly occurred at the C5–C6 level and this probably resulted in facet impingement. The last paper shows the effect of active eye–head–neck coordination exercises on cervico-cephalo kinaesthesia and pain intensity for patients with chronic whiplash. Several papers are not aimed at physical therapists or manual therapists primarily. However, for persons who have special interest in whiplash, this journal is highly recommended. Overall, the Journal of Whiplash and Related Disorders includes several very useful and interesting papers on aspects of diagnosis and treatment for patients with whiplash injuries. But it also contains papers that are probably not the core business of our profession.

Gwendolijne G.M. Scholten-Peeters, MSc, PT, MT Department of Manual Therapy Vrije Universiteit Brussel Belgium Dutch National Institute of Allied Health Professions Amersfoort, The Netherlands

Manual Therapy (2003) 8(3), 187–188

Manual Therapy (2003) 8(3), 189 r 2003 Elsevier Science Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00049-3

Letter to the editor

Comments on ‘Achilles tendinopathy’ In my preparation of a paper for the Sports Physiotherapy Australia 2003 conference, I have read fairly extensively on the topic of Achilles tendinopathy. I found the article by Cook et al. (2002) (Achilles tendinopathy, Manual Therapy 7(3): 121– 130) most interesting and proceeded to utilize the VISA-A questionnaire to document change in the clinical severity of a patient undergoing treatment for Achilles tendinopathy. I experienced difficulty with the scoring and so sourced the original article on the development of the VISA-A questionnaire. I noted that the example of the VISA-A questionnaire in Manual Therapy (2002) 7(3) page 124 is incorrect and that for the first

item the 0 and 100 min have been transposed! Also the zero on the scale is missing from questions 1–6 and this would appear to bias the user and is inconsistent with the original questionnaire (Robinson et al. 2001 The VISA-A questionnaire: a valid and reliable index of the clinical severity of Achilles tendinopathy. British Journal of Sports Medicine 35: 335–341). Perhaps you could draw readers’ attention to theses errors and reprint a corrected version of the VISA-A questionnaire. Lynn Bardin, MSc Australia

189

Manual Therapy (2003) 8(3), 190–191 r 2003 Elsevier Science Ltd. All rights reserved. 1356-689X/03/$ - see front matter doi:10.1016/S1356-689X(03)00050-X

Letter to the editor

Authors response The VISA-A questionnaire published in the Masterclass was not a complete and correct version of the scale. There was an error in the first question. In this question you score points for how long you have morning stiffness, the longer the tendon is stiff the LESS points you should score. As it was published, the score gave you more points for being stiffer in the morning, and this is clearly not correct, as the VISA score for a normal tendon should be 100 points. Similarly, for all questions there should be an

opportunity to score 0 points, these were not on the questionnaire published in the Achilles masterclass. The VISA-A scale is a valid and reliable outcome measure for Achilles tendinopathy and I trust it will still be used by clinicians and researchers alike. A correct version is printed with this correction. The errors arose in the authors copy sent to the publishers and were not picked up by the author’s proofs.

The VISA-A questionnaire: An index of the severity of Achilles tendinopathy In this questionnaire, the term pain refers specifically to pain in the Achilles tendon region 1. For how many minutes do you have stiffness in the Achilles region on first getting up?

100 min

0 min 0

1

2

3

4

5

6

7

8

9

POINTS &

10

2. Once you are warmed up for the day, do you have pain when stretching the Achilles tendon fully over the edge of a step? (keeping knee straight)

strong severe pain

no pain 0

1

2

3

4

5

6

7

8

POINTS &

9 10

3. After walking on flat ground for 30 min, do you have pain within the next 2 hours? (If unable to walk on flat ground for 30 min because of pain, score 0 for this question).

strong severe pain 0

1

2

3

4

5

6

7

8

no pain

POINTS &

no pain

POINTS &

9 10

4. Do you have pain walking downstairs with a normal gait cycle?

strong severe pain 0

1

2

3

4

5

6

7

8

9 10

5. Do you have pain during or immediately after doing 10 (single leg) heel raises from a flat surface?

strong severe pain

no pain 0

1

2

3

4

5

6

7

8

9 10

190

POINTS &

Letter to the editor 191

6. How many single leg hops can you do without pain? POINTS & 0

1

2

3

4

5

6

7

8

9 10

7. Are you currently undertaking sport or other physical activity? 0 4 7 10

& & & &

Not at all POINTS Modified training7modified competition & Full training7competition but not at same level as when symptoms began Competing at the same or higher level as when symptoms began

8. Please complete EITHER A, B or C in this question. * *

*

A.

If you have no pain while undertaking Achilles tendon loading sports please complete Q8a only. If you have pain while undertaking Achilles tendon loading sports but it does not stop you from completing the activity, please complete Q8b only. If you have pain which stops you from completing Achilles tendon loading sports, please complete Q8c only. If you have no pain while undertaking Achilles tendon loading sports, for how long can you train/practise? NIL & 0

1–10 min & 7

11–20 min 21–30 min >30 min & 14

& 21

& 30

POINTS &

OR B. If you have some pain while undertaking Achilles tendon loading sport, but it does not stop you from completing your training/practice for how long can you train/practise? NIL & 0

1–10 min & 4

11–20 min 21–30 min >30 min & 10

& 14

& 20

POINTS &

OR C. If you have pain that stops you from completing your training/practice in Achilles tendon loading sport, for how long can you train/practise? NIL & 0

1–10 min & 2

11–20 min 21–30 min >30 min & 5

& 7

TOTAL SCORE (/100)

& 10

POINTS &

&%

Jill Cook

Manual Therapy (2003) 8(3), 190–191


Manual Therapy (2003) 8(3), 192

Diary of events

9–10 August 2003, Auckland, New Zealand

21–26 March 2004

The New Zealand Manipulative Physiotherapists Association biennial Scientific Conference, entitled ‘Back to Clinical Relevance’, will be held at the Sheraton Hotel in Auckland. For further information please contact Vicki Reid at [email protected], or view the NZMPA website for further details on www.nzmpa.org.nz

8th IFOMT Congress (International Federation of Manipulative Therapy), International Convention Centre, Cape Town, South Africa. This promises to be one of the most exciting events in the history of physiotherapy in South Africa and you are encouraged to start your planning now. The theme for the conference is ‘‘Balancing the Outcome of Manual Therapy.’’ The programme will range from research based to clinical outcomes papers. Several speakers of international standing have indicated their willingness to participate. The programme will include sessions on Pain, Outcome Based Research, Community and Industrial Considerations, Musculo-Skeletal Spinal and Peripheral Dysfunctions. Website: Our website will be continually updated as further information becomes available – please keep watching: www.uct.ac.za/depgc/pgc/ Enquiries: If you would like further information, please send an expression of interest to: Sally Elliott, Conference Management Centre, UCT Medical School, Anzio Road, Observatory 7925, Cape Town, South Africa Tel: +27 21 406-6381; Fax: +27 21 448-6263; E-mail: [email protected]

2–6 September 2003, Prague, Czech Republic Pain in Europe IV: 4th Congress of the European Federation of IASP Chapters (EFIC). Organizing Secretariat: Congress Business Travel Ltd (CBT), Sˇteˇpańska´ 6/535, 120 00 Prague 2, Czech Republic. Tel: +420-2-2494-2575; Fax: +420-2-2494-2550; E-mail: [email protected]; www.pain2003.cz 11 September 2003, The Netherlands Body at Work Symposium. Website: www.bodyatwork.nl 25–27 September, 2003, Toledo, USA American Society of Biomechanics. You will get further details from these two addresses: www.asb-biomech.org/conference03 www.bioe.eng.utoledo.edu/ASB

21–26 August 2005, Sydney, Australia 11th World Congress on Pain, Workshop and Plenary Proposals. Please send proposals to the Chair of the Scientific Program Committee: Herta Flor, PhD, Central Institute of Mental Health, Dept of Clinical and Cognitive Neuroscience, PF 12 21 20, 68072 Mannheim, Germany. Tel: 49-621-170-3922; Fax: 49-621-170-3932; E-mail: fl[email protected] Workshop and plenary suggestions should be submitted by 15 March 2003 at the latest so that they can be considered by the Scientific Program Committee. Note that announcements, deadlines, and other information relating to the 2005 Congress will be routinely updated on the IASP Web page: www.iasp-pain.org

21–22 November 2003, London, UK 10th Anniversary Symposium on Complementary Health Care at the Royal College of Physicians in London. It is an international symposium with a focus on original research and comprises two days of platform and poster presentations as well as pre-conference workshops and satellite meetings. Contact: Barbara Wider, Complementary Medicine, Peninsular Medical School, Universities of Exeter & Plymouth, 25 Victoria Park Road, Exeter EX2 4NT, UK. E-mail: [email protected] Website: www.exeter.ac.uk/FACT/sympo 10–13 November, 2004, Melbourne, Australia

Janet G. Travell, MD Seminar Series, Bethesda, USA

5th Interdisciplinary World Congress on Low Back & Pelvic Pain Effective Diagnosis and Treatment of Lumbopelvic Pain. Information and Call for Papers: www.worldcongresslbp.com http://www.worldcongresslbp.com Deadline submitting papers: December 15, 2003 For all further information: [email protected] mailto: [email protected]

For information, contact: Myopain Seminars, 7830 Old Georgetown Road, Suite C-15, Bethesda, MD 20814-2432, USA. Tel.: +1 301 656 0220; Fax: +1 301 654 0333; website: www.painpoints.com/seminars.htm; e-mail: [email protected]

27–30 November 2003, Sydney, Australia

Evidence-based manual therapy congress Further information: www.medicongress.com

13th Biennial Conference of Musculoskeletal Physiotherapy Australia—Best Evidence Better Care. Includes Multidisciplinary Whiplash Symposium. Sydney Australia, Contact: Linda Phillips 2003 Conference Secretariat, GPO Box 2609 Sydney, Australia. Tel.: 61-02-9241-1478; Fax: 61-02-9251-3552; E-mail: [email protected]; Website: www.mpa2003.com

Intensive courses in Manual Therapy Further information: http://allserv.rug.ac.be/bvthillo If you wish to advertise a course/conference, please contact: Karen Beeton, Department of Physiotherapy, University of Hertfordshire, College Lane, Hatfield, Herts AL10 9AB, UK. There is no charge for this service.

192

Manual Therapy Journal - Volume 8, Issue 3, Pages 129-192 (August 2003)

Manual Therapy Journal - Volume 8, Issue 1, Pages 1-62 (February 2003)

Manual Therapy Journal - Volume 8, Issue 4, Pages 193-270 (November 2003)

Manual Therapy Journal - Volume 8, Issue 2, Pages 63-128 (May 2003)

Manual Therapy Journal - Volume 12, Issue 3, Pages 199-296 (August 2007)

Manual Therapy Journal - Volume 10, Issue 3, Pages 175-238 (August 2005)

Manual Therapy Journal - Volume 9, Issue 3, Pages 123-182 (August 2004)

Manual Therapy Journal - Volume 11, Issue 3, Pages 167-240 (August 2006)

Manual Therapy Journal - Volume 13, Issue 4, Pages 277-372 (August 2008)

Manual Therapy Journal - Volume 14, Issue 4, Pages 353-460 (August 2009)

Manual Therapy Journal - Volume 13, Issue 3, Pages 181-276 (June 2008)

Manual Therapy Journal - Volume 14, Issue 3, Pages 241-352 (June 2009)

Manual Therapy Journal - Volume 14, Issue 2, Pages 117-240 (April 2009)

Manual Therapy Journal - Volume 13, Issue 2, Pages 91-180 (April 2008)

Manual Therapy Journal - Volume 14, Issue 5, Pages 461-584 (October 2009)

Manual Therapy Journal - Volume 12, Issue 4, Pages 297-396 (November 2007)

Manual Therapy Journal - Volume 10, Issue 4, Pages 239-304 (November 2005)

Manual Therapy Journal - Volume 13, Issue 1, Pages 1-90 (February 2008)

Manual Therapy Journal - Volume 10, Issue 1, Pages 1-92 (February 2005)

Manual Therapy Journal- Volume 13, Issue 6, Pages 475-564 (December 2008)

Manual Therapy Journal - Volume 12, Issue 1, Pages 1-84 (February 2007)

Manual Therapy Journal - Volume 12, Issue 2, Pages 85-198 (May 2007)

Manual Therapy Journal - Volume 9, Issue 1, Pages 1-58 (February 2004)

Manual Therapy Journal - Volume 10, Issue 2, Pages 93-174 (May 2005)

Manual Therapy Journal - Volume 14, Issue 1, Pages 1-116 (February 2009)

Manual Therapy Journal - Volume 9, Issue 2, Pages 59-120 (May 2004)

Manual Therapy Journal - Volume 14, Issue 6, Pages 585-716 (December 2009)

Manual Therapy Journal - Volume 11, Issue 1, Pages 1-90 (February 2006)

Manual Therapy Journal - Volume 9, Issue 4, Pages 183-246 (November 2004)

Manual Therapy Journal - Volume 11, Issue 4, Pages 241-358 (November 2006)

Manual Therapy Journal - Volume 11, Issue 2, Pages 91-166 (May 2006)

Manual Therapy Journal - Volume 8, Issue 3, Pages 129-192 (August 2003)

Manual Therapy Journal - Volume 8, Issue 1, Pages 1-62 (February 2003)

Manual Therapy Journal - Volume 8, Issue 4, Pages 193-270 (November 2003)

Manual Therapy Journal - Volume 8, Issue 2, Pages 63-128 (May 2003)

Manual Therapy Journal - Volume 12, Issue 3, Pages 199-296 (August 2007)

Manual Therapy Journal - Volume 10, Issue 3, Pages 175-238 (August 2005)

Manual Therapy Journal - Volume 9, Issue 3, Pages 123-182 (August 2004)

Manual Therapy Journal - Volume 11, Issue 3, Pages 167-240 (August 2006)

Manual Therapy Journal - Volume 13, Issue 4, Pages 277-372 (August 2008)

Manual Therapy Journal - Volume 14, Issue 4, Pages 353-460 (August 2009)

Manual Therapy Journal - Volume 13, Issue 3, Pages 181-276 (June 2008)

Manual Therapy Journal - Volume 14, Issue 3, Pages 241-352 (June 2009)

Manual Therapy Journal - Volume 14, Issue 2, Pages 117-240 (April 2009)

Manual Therapy Journal - Volume 13, Issue 2, Pages 91-180 (April 2008)

Manual Therapy Journal - Volume 14, Issue 5, Pages 461-584 (October 2009)

Manual Therapy Journal - Volume 12, Issue 4, Pages 297-396 (November 2007)

Manual Therapy Journal - Volume 10, Issue 4, Pages 239-304 (November 2005)

Manual Therapy Journal - Volume 13, Issue 1, Pages 1-90 (February 2008)

Manual Therapy Journal - Volume 10, Issue 1, Pages 1-92 (February 2005)

Manual Therapy Journal- Volume 13, Issue 6, Pages 475-564 (December 2008)

Manual Therapy Journal - Volume 12, Issue 1, Pages 1-84 (February 2007)

Manual Therapy Journal - Volume 12, Issue 2, Pages 85-198 (May 2007)

Manual Therapy Journal - Volume 9, Issue 1, Pages 1-58 (February 2004)

Manual Therapy Journal - Volume 10, Issue 2, Pages 93-174 (May 2005)

Manual Therapy Journal - Volume 14, Issue 1, Pages 1-116 (February 2009)

Manual Therapy Journal - Volume 9, Issue 2, Pages 59-120 (May 2004)

Manual Therapy Journal - Volume 14, Issue 6, Pages 585-716 (December 2009)

Manual Therapy Journal - Volume 11, Issue 1, Pages 1-90 (February 2006)

Manual Therapy Journal - Volume 9, Issue 4, Pages 183-246 (November 2004)

Manual Therapy Journal - Volume 11, Issue 4, Pages 241-358 (November 2006)

Manual Therapy Journal - Volume 11, Issue 2, Pages 91-166 (May 2006)

Recommend Documents