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

VOLUME 13 NUMBER 4 PAGES 277– 372 August 2008 Editors International Advisory Board Ann Moore PhD, GradDipPhys, FCSP, ...

Author: Editors: Ann Moore and Gwen Jull

155 downloads 752 Views 3MB 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

VOLUME 13 NUMBER 4 PAGES 277– 372 August 2008

Editors

International Advisory Board

Ann Moore PhD, GradDipPhys, FCSP, CertEd, FMACP Clinical Research Centre for Health Professions University of Brighton Aldro Building, 49 Darley Road Eastbourne BN20 7UR, UK

K. Bennell (Victoria, Australia) K. Burton (Hudders¢eld, UK) B. Carstensen (Frederiksberg, Denmark) M. Coppieters (Queensland, Australia) E. Cruz (Setubal Portugal) L. Danneels (Mar|¤ akerke, Belgium) S. Durrell (London, UK) S. Edmondston (Perth, Australia) J. Endresen (Flaktvei, Norway) L. Exelby (Biggleswade, UK) D. Falla (Aalborg, Denmark) T.W. Flynn (Denver, CO, USA) J. Greening (London, UK) C. J. Groen (Utrecht,The Netherlands) A. Gross (Hamilton, Canada) T. Hall (West Leederville, Australia) W. Hing (Auckland, New Zealand) M. Jones (Adelaide, Australia) S. King (Glamorgan, UK) B.W. Koes (Amsterdam,The Netherlands) J. Langendoen (Kempten, Germany) D. Lawrence (Davenport, IA, USA) D. Lee (Delta, Canada) R. Lee (Brighton, UK) C. Liebenson (Los Angeles, CA, USA) L. Ma¡ey-Ward (Calgary, Canada) E. Maheu (Quebec, Canada) C. McCarthy (Coventry, UK) J. McConnell (Northbridge, Australia) S. Mercer (Queensland, Australia) D. Newham (London, UK) J. Ng (Hung Hom, Hong Kong) S. O’Leary (Queensland, Australia) L. Ombregt (Kanegem-Tielt, Belgium) N. Osbourne (Bournemouth, UK) M. Paatelma (Jyvaskyla, Finland) N. Petty (Eastbourne, UK) A. Pool-Goudzwaard (The Netherlands) M. Pope (Aberdeen, UK) G. Rankin (London, UK) D. Reid (Auckland, New Zealand) A. Rushton (Birmingham, UK) C. Shacklady (Manchester, UK) M. Shacklock (Adelaide, Australia) D. Shirley (Lidcombe, Australia) V. Smedmark (Stenhamra, Sweden) W. Smeets (Tongeren, Belgium) C. Snijders (Rotterdam,The Netherlands) R. Soames (Dundee, UK) P. Spencer (Barnstaple, UK) M. Sterling (St Lucia, Australia) P. Tehan (Victoria, Australia) M. Testa (Alassio, Italy) M. Uys (Tygerberg, South Africa) P. van der Wur¡ (Doorn,The Netherlands) P. van Roy (Brussels, Belgium) B.Vicenzino (St Lucia, Australia) H.J.M.Von Piekartz (Wierden,The Netherlands) M.Wallin (Spanga, Sweden) M.Wessely(Paris, France) A.Wright (Perth, Australia) M. Zusman (Mount Lawley, Australia)

Gwendolen Jull PhD, MPhty, Grad Dip ManTher, FACP Department of Physiotherapy University of Queensland Brisbane QLD 4072, Australia Associate Editor’s Darren A. Rivett PhD, MAppSc, (ManipPhty) GradDipManTher, BAppSc (Phty) Discipline of Physiotherapy Faculty of Health The University of Newcastle Callaghan, NSW 2308, Australia Tim McClune D.O. Spinal Research Unit. University of Hudders¢eld 30 Queen Street Hudders¢eld HD12SP, UK Editorial Committee Masterclass Editor Karen Beeton PhD, MPhty, BSc(Hons), MCSP MACP ex o⁄cio member Associate Head of School (Professional Development) School of Health and Emergency Professions University of Hertfordshire College Lane Hat¢eld AL10 9AB, UK Case reports & Professional Issues Editor Je¡rey D. Boyling MSc, BPhty, GradDipAdvManTher, MCSP, MErgS Je¡rey Boyling Associates Broadway Chambers Hammersmith Broadway LondonW6 7AF, UK Book Review Editor Raymond Swinkels MSc, PT, MT Ulenpas 80 5655 JD Eindoven The Netherlands

Visit the journal website at http://www.elsevier.com/math doi:10.1016/S1356-689X(08)00091-X

ARTICLE IN PRESS

Manual Therapy 13 (2008) 277 www.elsevier.com/locate/math

Editorial

Collegiality in musculoskeletal therapy, a strength and a tool As we write this Editorial, IFOMT 2008 is only a few weeks away, a highlight in many musculoskeletal therapists’ calendar. We look forward to the presentations, the workshops, the meetings, the discussions and the opportunity to network, socialise and form new collaborations. We also welcome the chance to meet up with old friends and renew acquaintances. We would like to thank the NVMT team for creating this opportunity which we are sure will be full of successes and triumphs. Thank you on behalf of the Manual Therapy Editorial team who look forward to meeting up in Rotterdam during the conference and having the opportunity to meet face-to-face. We also welcome the opportunity that this kind of conference offers new researchers and cutting edge experienced researchers. IFOMT always promises to be the start of a number of publications within a range of journals. One of the great things about manual therapy conferences is the collegiality that exudes throughout the event. Somehow, musculoskeletal therapists appear to really enjoy being together, listening together, thinking together and IFOMT conferences are very well known for this collegiate atmosphere and to experience it is a very positive thing. Collegiality is important to professional groups and some are better at it than others. In some countries, for example, in the UK, the relevance of professions is being questioned and increasingly the message is ‘‘don’t try to be politically one professional group, link with other allied health professionals, so develop collegiality between professions, there is more mileage and greater

1356-689X/$ - see front matter r 2008 Published by Elsevier Ltd. doi:10.1016/j.math.2008.05.003

strength in numbers’’. Although this is probably sound advice in the current health service climate in the UK, it is a pity that professional identity is currently being frowned upon. Perhaps, for many physiotherapists, professional collegiality has been lost because of the growing number of specialisms and this has potentially diluted the physiotherapy profession’s research endeavours, service developments and educational foci. Critical mass is useful politically, but it is also persuasive in presenting a solid evidence base. The advantage that musculoskeletal therapy has is that in IFOMT we have a ready and able network of experienced practitioners, high quality academics and researchers, and importantly a collegiate atmosphere. Perhaps we should do more to collaborate, both nationally and internationally on a routine basis to solve research questions, problems and issues that are of general interest and of multinational importance taking into consideration differences in culture and healthcare standards, working practices and terminology. Let us share opportunities, values, beliefs and enterprise to create a strong and irrefutable evidence base for manual therapy. Let us use our international collegiality to best effect to support our speciality for the major beneﬁts it can offer to patient and community care.

Ann Moore, Gwen Jull University of Brighton, Aldro Building, 49 Darley Road, Eastbourne, East Sussex, BN20 7UR, UK E-mail address: [email protected] (A. Moore)

ARTICLE IN PRESS

Manual Therapy 13 (2008) 278–288 www.elsevier.com/locate/math

Review

Cervical arterial dysfunction and manual therapy: A critical literature review to inform professional practice Roger Kerrya,, Alan J. Taylorb, Jeanette Mitchellc,d,e, Chris McCarthyf a

Division of Physiotherapy Education, University of Nottingham, UK b Nottingham Nuffield Hospital, Nottingham, UK c Department of Zoology & Physiology, School of Biological Sciences, University of Wyoming, Laramie, USA d Department of Kinesiology & Health, Division of Health Sciences, University of Wyoming, Laramie, USA e Department of Physiotherapy, School of Allied Health Professions, University of the West of England, Bristol, UK f Warwick Emergency Care and Rehabilitation, The University of Warwick, UK Received 28 September 2006; received in revised form 29 October 2007; accepted 30 October 2007

Abstract An abundance of literature has attempted to provide insight into the association between cervical spine manual therapy and cervical artery dysfunction leading to cerebral ischaemic events. Additionally, speciﬁc guidelines have been developed to assist manual therapists in clinical decision-making. Despite this, there remains a lack of agreement within the profession on many issues. This paper presents a critical, re-examination of relevant literature with the aim of providing a contemporary, evidence-informed review of key areas regarding the neurovascular risks of cervical spine manual therapy. From a consideration of case reviews and surveys, haemodynamic principles, and blood ﬂow studies, the authors suggest that: (1) it is currently impossible to meaningfully estimate the size of the risk of post-treatment complications; (2) existing testing procedures have limited clinical utility; and (3) a consideration of the association between pre-existing vascular risk factors, combined with a system based approach to cervical arterial haemodynamics (inclusive of the carotid system), may assist manual therapists in identifying at-risk patients. r 2008 Elsevier Ltd. All rights reserved. Keywords: Manual therapy; Vertebrobasilar insufﬁciency; Internal carotid artery dissection; Cervical arterial dysfunction; Atherosclerosis

1. Introduction Manual therapists are aware that cervical spine manual therapy (MT) techniques, particularly those involving full-range, high-velocity, rotational movements, hold inherent risks of insult to the vertebral arteries (VAs) and internal carotid arteries (ICAs), which can result in cerebral ischaemia, stroke, or death. Extensive reviews of contemporary evidence, and reasoned debate, have been undertaken in an effort to Corresponding author. Division of Physiotherapy Education, University of Nottingham, Hucknall Road, Nottingham, NG5 1PB, UK Tel.: +44 (0) 115 8231790. E-mail address: [email protected] (R. Kerry).

1356-689X/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.10.006

provide clinical guidance for the safest implementation of such techniques and for pre-manipulative screening for vertebrobasilar insufﬁciency (VBI) (Barker et al., 2001; Magarey et al., 2004; APA, 2006). However, there is still widespread uncertainty and controversy regarding the association between cervical spine MT and cervical arterial dysfunction (CAD), the reliability and validity of functional screening tests, the speciﬁcity and sensitivity of these tests in identifying at-risk patients, and the medico-legal position of therapists and patients (Australian Journal of Physiotherapy (AJP), 2001; Jull et al., 2002; Kerry, 2002; Refshauge et al., 2002; Brew, 2004; Kerry, 2004). The purpose of this paper is to present a summarised, critical re-examination of current available published

ARTICLE IN PRESS R. Kerry et al. / Manual Therapy 13 (2008) 278–288

literature concerning cerebrovascular accidents (CVAs) in the context of MT. The review is non-systematic and aims to provide a broad picture of evidence and contemporary thought with regards to risks of cervical spine management. 1.1. Data sources and method Literature searches were conducted for the years 1990–2007, using the data-bases Medline (PubMed); CINAHL; Embase; AMED; CISCOM; and the Cochrane Library. Search terms were ‘vertebral artery’, ‘basilar artery’, ‘vertebrobasilar artery’, ‘internal carotid artery’; ‘anatomy and histology’, ‘cytology’, ‘embryology’, ‘injuries’, ‘innervation’, ‘pathology’, ‘physiology’, ‘radiology’, ‘radionuclide imaging’, ‘ultrasonography’, ‘ultrastructure’, and ‘adverse effects’, ‘adverse events’, ‘chiropractic’, ‘complications’, ‘manual therapy’, ‘osteopathy’, ‘physiotherapy’, ‘risk’, ‘safety’, ‘spinal manipulation’, ‘strokes’, ‘vascular accidents’. Related links, reference lists, personal ﬁles, www search engines and some publications prior to 1990, considered relevant according to the key terms used, were also searched. This search strategy was designed to widen the search beyond the conﬁnes of MT literature. The search process resulted in 833 articles, of which 224 were considered relevant to MT (referring to manipulative and non-manipulative techniques) via the themes selected for the discussion. Key clinical and theoretical themes consistent within this literature base are supported by a representation of the search results in this paper. The discussion and conclusions of this review focus on the risk of cervical spine MT, the usefulness of screening tests in assessing at-risk patients, haemodynamic factors, and the nature and signs and symptoms of arterial changes and insufﬁciency.

2. Reviews and surveys Reviews and surveys attempt to establish risk factors for, and incidence of, CVAs related to MT. 2.1. Questionnaire surveys Retrospective and prospective surveys reviewed include reports by both manual therapists and nonmanual therapists (neurosurgeons, neurologists, vascular surgeons). Non-manual therapists’ surveys report a higher number of patients suffering post-MT complications or associated symptoms (Dvorak et al., 1993; Lee et al., 1995; Rivett and Milburn, 1997). Retrospective surveys report estimates of CVA-risk incidence ranging from one in 9122 (Michaeli, 1993) to ﬁve in one million (Haynes, 1994). Some estimates have been based on inadequate sample sizes (Haynes, 1994;

279

Rivett and Milburn, 1996), and a probable ﬂaw with retrospective studies is the questionable reliability of recall of past events. Therefore inferences regarding size of risk should be made with caution. Most prospective surveys do not calculate estimates of risk incidence, some stating explicitly that it would be impossible and misleading to do so (Stevinson, et al., 2001). Prospective studies are also likely to report a higher number of cases of post-MT vascular complications, for example, (Leboeuf-Yde et al., 1997; Nadareishvili and Norris, 1999; Stevinson et al., 2001; Magarey et al., 2004), than retrospective estimates (because of under-reporting, reporting restrictions, non-association of event to treatment, etc) (Stevinson et al., 2001; Dupeyron et al., 2003). Dupeyron et al. (2003) suggest that post-MT VBI could be 30 times higher than published. A recent prospective study on— chiropractors reported no incidents of serious adverse events (Thiel et al., 2007). Although a well-structured, rigorous survey, it must be considered that, as stated above, any study targeted at manual therapists is likely to provide unreliable data due to reporting bias and lack of awareness of what constitutes an adverse event. 2.2. Narrative reviews These ranged from discursive (Rivett, 1995) to structured reviews (Haldeman et al., 1999), encompassing several themes. Firstly, in considering estimates of incidence, a major confounding variable is that the actual incidence is unknown (Rivett, 1995; Assendelft et al., 1996; Ernst, 2001a, b, 2004), partly because the number of persons in the population receiving MT is unknown or unreported. This apparent under-reporting (Assendelft et al., 1996; Ernst, 2004) may be inﬂuenced by MT practitioners’ awareness of post-treatment CVAs, physicians not making the link between MT and CVAs, or a low probability of cases being published (Ernst 2001a, 2002). As another theme illustrating potential vascular postMT trauma, Hurwitz et al. (1996) reviewed 145 reports of post-MT complications: of 118 documented cases, 21 patients died and 52 were left with serious neurological deﬁcits. Ernst (2004) reported that between 1995 and 2003 in Britain, 300 patients were adversely affected by cervical spine MT, the most frequently reported complication being stroke following arterial dissection. These reports highlight the necessity of manual therapists to be aware of mechanisms and presentations of vascular trauma, and possible underlying pathology. Lastly, the validity of movement-based pre-MT screening tests is questioned (Kunasmaa and Thiel, 1994; Rivett, 1997), some authors suggesting that testing should be stopped (Thiel and Rix, 2005). Recent reports have undertaken probability analysis to calculate the sensitivity and speciﬁcity of functional tests, using data

ARTICLE IN PRESS 280

R. Kerry et al. / Manual Therapy 13 (2008) 278–288

(ﬂow rates and reproduced symptoms) from blood ﬂow studies. Current reports present sensitivity of 0%–21% and speciﬁcity of 23%–90% (Kerry and Rushton, 2003; Gross et al., 2005; Ritcher and Reinking, 2005). 2.3. Systematic reviews Three major systematic reviews were identiﬁed (Ernst, 2001b, 2002; Rubinstein et al., 2005). Ernst (2001b) investigated the safety of cervical spine MT by summarising data of all prospective investigations. From ﬁve prospective studies, it was concluded that half of all patients experienced mild, transient adverse events following manipulation, that no reliable data of the incidence of adverse events exist, and that there is a need for further prospective investigations. In a systematic review of all original case reports between 1995 and 2001, Ernst (2002) retrieved 42 individual case reports of complications post-manipulation, 18 of which reported arterial dissection causing stroke. Again, it was concluded that, although serious complications do occur, it is impossible to rate their incidence based on existing reported data, and large rigorous prospective trials are needed. Rubinstein et al. (2005) undertook a systematic literature review, between 1966 and 2005, focussing on risk factors for VA or carotid artery (CA) dissection. Thirty-one case-control studies (from 1023 abstracts) were assessed for their methodological quality. After controlling for selection bias, lack of controls for confounding variables, and inadequate data analysis, there were still strong associations between VA or CA dissection and aortic root diameter (434 mm), migraine, relative common CA diameter change during the cardiac cycle, and trivial trauma (ie associated with cervical spine MT), and weak associations were found for plasma homocysteine and recent infections. The two MT papers (Rothwell et al., 2001; Smith et al., 2003) that Rubinstein et al. (2005) reviewed reported that few cases supported a clear temporal association between treatment and arterial dissection (n ¼ 4). However, Rubinstein et al. (2005) used a 24-h post-treatment cut-off point, thereby missing potential latent cases (a phenomenon previously reported by other authors). This calls into question the parameters set by these authors. 2.4. Case reviews Terrett (1987) reported on 107 cases of post-MT vascular accidents, between 1934 and 1986, equally distributed across the genders, with average ages of 39.8 yrs (female) and 36.2 yrs (male). DiFabio (1999) reported that death occurred in 18% of 177 cases and concluded that the literature at that time did not demonstrate that beneﬁts of cervical spine MT out-

weighed the risks. Frisoni and Anzola (1991) concluded from 39 cases that a priori risk identiﬁcation cannot be made in the majority of cases; symptoms may develop some time after uneventful treatment; mortality/longterm impairment occurs in 28% of cases, and a history of previous, transient neurological symptoms or upper cervical spine laxity should contra-indicate any neck movement-based treatment. Haldeman et al. (2002a) looked speciﬁcally at the effect of referral bias on clinical perceptions of VA dissection risk. Based on analyses of data from an insurance provider and a chiropractic survey, they reported that 1 in 48 manual therapists were likely to be aware of post-MT CVA, compared to 1 in 2 neurologists. This discrepancy could explain the apparent differences in perception and experiences between these two professions. Rothwell et al. (2001) conducted a population-based nested case-control review to test the association between MT and CVAs. Of 582 cases reviewed, they suggested that vertebrobasilar accidents were ﬁve times more likely to occur if a patient had MT one week prior to the stroke, and VA-stroke cases were ﬁve times more likely to have received MT than the control group. This review concurs with the report of Smith et al. (2003) who, after comparing 151 arterial dissection cases with a control group, concluded that arterial dissection was ﬁve times more likely to occur if MT had been administered within 30 days prior to, and twice as likely to have had increased neck/head pain preceding the stroke. They concluded that VA dissection was independently associated with stroke and increased neck/head pain. Haldeman et al. (2002a) reported that 92% of VA cases presented with neck/head pain, and 25% reported a sudden onset of pain (suggestive of a dissection in progress). Haldeman et al. (2002b) concluded that they were unable to identify risk factors from the 64 medicolegal cases and propose that vertebrobasilar dissection be considered an unpredictable, inherent, idiosyncratic, and rare complication of MT. These surveys and reviews show that there is a need for well-structured, large scale prospective surveys and trials to be undertaken (Rivett, 1995; Hurwitz et al., 1996; Shekelle and Coulter, 1997; Ernst, 2001b, 2004) before a more realistic impression of the neurovascular risk of MT can be established (Dvorak et al., 1993; Haynes, 1994; Senstad et al., 1996; Rivett and Milburn, 1997; Stevinson et al., 2001).

3. Haemodynamics, arterial trauma and vascular pathology MT has been burdened by its ongoing focus on the vertebral artery in isolation. Haemodynamic theory is a relatively new but important area of thinking in MT

ARTICLE IN PRESS R. Kerry et al. / Manual Therapy 13 (2008) 278–288

(Taylor, 2001). It is well documented and understood, that blood ﬂow in the VA and ICA systems is intricately linked via the circle of Willis (Cardon et al., 1998). It follows, therefore, that both VA and ICA blood ﬂow and pathology should be considered in pre-treatment risk assessment. ICA blood ﬂow velocity and volume are signiﬁcantly greater than in the VAs (Paivansalo et al., 1998; Schoning and Hartig, 1998; Sidhu, 2000; Scheel et al., 2000a, b). Therefore, the ICAs may have a signiﬁcant compensatory role during cervical spine movement and positioning. A negative pre-manipulative test relies critically on the patency of the ICAs and unaffected VAs (see blood ﬂow studies section) for adequate perfusion and, as such, tests the system as opposed to the concept that this somehow tests the VA in isolation. This concept is illustrated by Weintraub and Khoury (1998), who examined ICA and VA blood ﬂow changes in patients with known underlying pathology. They reported an absence of overt VBI symptoms on sustained end-range extension for 3–4 min in 160 cases (mean age: 66 years), 25% of whom had VA hypoplasia associated with marked basilar artery ﬂow alteration, whilst unsuspected ICA occlusion occurred in six cases and VA occlusion in two. Welch et al. (2000) also studied the effect of occlusion of the ICAs and VAs on the opposing vessel ﬂow, and reported that ﬂow in one part of the system was signiﬁcantly inﬂuenced by occlusion in the other part. These studies draw attention to the need to be aware of the effects cervical spine positioning and to identify pre-existing vascular disease as a reason for caution in both pre-treatment testing and practice of MT. Trauma to cervical blood vessels is generally classiﬁed as either dissection resulting from direct trauma to the vessel, or localised thrombogenesis and embolus formation in response to endothelial damage (Caplan and Biousse, 2004). Either pathological state may lead to stroke (Pollanen et al., 1992; Ross, 1999). Arterial dissection may occur after trivial trauma to the vessel, or spontaneously. This may be related to pre-existing, congenital weakness of the vessel wall or acquired vascular pathology (atherosclerosis). The mechanisms are thought to be arterial dissection with an intra-mural haematoma, resulting in vessel lumen narrowing with ensuing reduction in blood ﬂow and ischaemia; an extending dissection, leading to subarachnoid haemorrhage; and/or dissection leading to thrombus formation with secondary emboli, and stroke (Mann and Refshauge, 2001; Schievink et al., 2001). ICA dissection has been linked to multiple types of non-penetrating or blunt trauma, including motor vehicle accidents (Miller et al., 2002; Beaudry and Spence, 2003), manipulation (Haneline et al., 2003; Ernst, 2004), hand-held mechanical massagers (Grant and Wang, 2004), soft-ball injuries (Schievink et al.,

281

1998), strangulation (Clarot et al., 2005), effort (running) (Shimada et al., 2005), horse-hoof impact (Fletcher et al., 1995), sneezing (Taylor and Kerry, 2005). These examples, although unremarkable in isolation, should alert clinicians to the possibility that traumatic vascular injury may occur in the ICAs as well as the VAs. Furthermore, these arterial pathologies (VA and ICA) are known to present as neuromusculoskeletal symptoms, often isolated neck and head pain, without the commonly described VBI symptoms (Munari et al., 1994; Taylor and Kerry, 2005, Arnold et al., 2006). This concurs with other reports of VA dissection presenting as isolated neck pain (Krespi et al., 2002) and/or headache (Gates et al., 1997), without classical VBI signs. Although the literature suggests that the pathogenesis of arterial dissection remains uncertain, proposed risk factors include high plasma homocysteine concentrations (Giroud et al., 1994; Pezzini et al., 2002); speciﬁc genotypes (Pezzini et al., 2002); hereditary connective tissue disorders (i.e. Ehlers – Danlos syndrome) (Schwarze et al., 2000); connective tissue disorders (Brandt et al., 2001); ﬁbromuscular dysplasia (Van Damme et al., 1999), and recent infections (Grau and Buggle, 1999). There is much debate in the current literature as to the relevance and relative weight of these proposed risk factors and, in particular, why some individuals seem more prone to vascular injury and pathology than others (Rubinstein et al., 2005). As stroke or death is the ultimate adverse outcome of cervical spine MT (DiFabio, 1999), it is of note that ICA dissection is reported to be the cause of stroke in up to 20% of patients aged 18–44 years (Lisovoski and Rousseaux, 1991). Across all age groups, stroke is considered to be the leading cause of morbidity and the third-leading cause of mortality in the Western world (Flossmann et al., 2004). Li et al. (2005) reported that uncontrolled blood pressure is a key factor in the aetiology of stroke. Although the actual pathophysiogical mechanisms of post-cervical spine MT stroke are largely unknown, the occurrence of VA and/or ICA dissection, with or without pre-existing vascular pathology, is the most commonly accepted theory. Manual therapists should be acutely aware that patients may present with headache and neck pain as the early warning signs of impending stroke (Arnold et al., 2006). There are reports which suggest an association between ICA insufﬁciency/ischaemia and upper cervical spine instability (Volle and Montazem, 2001; Tominaga et al., 2002; Garg et al., 2004; Yamazaki et al., 2004). There are also reports that demonstrate anterior and posterior blood ﬂow dysfunction following road trafﬁc accidents (Chung and Han, 2002; Beaudry and Spence, 2003). These two points are of relevance to MT and should be given consideration, as pre-disposing factors, in the assessment of VA and ICA insufﬁciency.

ARTICLE IN PRESS 282

R. Kerry et al. / Manual Therapy 13 (2008) 278–288

Atherosclerosis is believed to be more prevalent in the ICAs than VAs in modern Western societies (Ersoy et al., 2003; Giannoukas et al., 2005). However, VA disease has been reported (Mitchell, 2002) and proposed as a risk factor for MT (Cagnie et al., 2006). VA and ICA disease has been shown to co-exist in the same patients and may be accompanied by coronary and peripheral occlusive disease and hypercholesterolaemia (Caplan et al., 1992). A large body of literature in this area considers potential risk factors such as hypertension, smoking, diabetes, hypercholesterolaemia, hyperhomocysteineamia, factors affecting blood coagulation, vascular trauma, infection, and migraine (Ross, 1999; Pezzini et al., 2002; Westaway et al., 2003; Migdalski et al., 2005; Arnold et al., 2006; Cagnie et al., 2006). This emphasises the need to incorporate both localised and systemic factors in pre-treatment clinical reasoning and risk assessment for all age groups. Contrary to this line of thinking are the conclusions from a sole MT-speciﬁc study which found no association between such risk factors and reported cases of post-MT stroke (Haldeman et al., 2002c). This report of 64 cases over a 16 year period concluded that cerebrovascular accident following manipulation was an unpredictable phenomenon. However, pre-incident data (the most important when considering proﬁling for chance of adverse event) was only available in 41% of the 64 cases reported, the range of events was restricted (i.e. only those that occurred within 48 h of a manipulation treatment and those which happened to get to the attention of a single physician), and the range of atherosclerotic risk factors was narrow and not representative of contemporary evidence-based risk factors. Considering these points, and the overwhelming amount of literature relating atherosclerosis to stroke, it is difﬁcult to refute the hypothesis that postMT strokes are associated with vascular pathology. It is clear that the interaction between mechanisms of vascular injury, underlying hereditary disorders, and proposed risk factors adds a body of new information for consideration. This knowledge may direct both the clinical reasoning process and future research into pretreatment cervical risk assessment. This may direct research toward the model used in suspected thromboembolism (Fancher et al., 2004), a condition which mirrors VA and ICA dissection in its complex pathogenesis and level of difﬁculty to diagnose or predict accurately.

4. Blood ﬂow studies The effects of cervical spine movements upon VA blood ﬂow have long been reported, with little consideration of ICA blood ﬂow and haemodynamic compensatory principles. Initial cadaver experiments showed decreased VA blood ﬂow, particularly on

contralateral cervical spine rotation (Tissington-Tatlow and Brammer, 1957; Toole and Tucker, 1960; Brown and Tissington-Tatlow, 1963; Selecki, 1969). Later in vivo studies, using ﬂowmetry, angiography, Doppler ultrasonography or magnetic resonance angiography, and documenting changes in VA blood ﬂow before and after cervical spine rotation, have produced contradictory results. Some researchers found no changes (Weingart and Bischoff, 1992; Thiel et al., 1994; Haynes and Milne, 2001; Zaina et al., 2003) while others reported a signiﬁcant reduction in contralateral blood ﬂow (Refshauge, 1994; Rossitti and Volkmann, 1995; Licht et al., 1998a; Li et al., 1999; Rivett et al., 1998, 1999; Mitchell, 2003; Arnold et al., 2004; Mitchell et al., 2004). This controversy may be because there is no standardisation of methods used. For example, healthy subjects and patients, who may or may not have had signs and symptoms of VBI at the time of measurement, were used. Both young and older subjects were included in some samples, and few authors compared males and females (Sturzenneger et al., 1994; Li et al., 1999; Rivett et al., 1999; Haynes et al., 2000; Scheel et al., 2000a, b; Mitchell, 2003, 2007). Blood ﬂow measurements at pre-transverse, cervical and intracranial parts of the VA have been reported. The cervical VA, as it traverses the bony transverse foramina, is difﬁcult to insonate accurately which may have confounded the ﬁndings (Johnson et al., 2000; Khaw et al., 2004). There is a logical argument that it is the change in blood ﬂow in the intracranial part of the VA (distal to the point of constriction) and not in the more proximal extracranial VA, that may correlate with potential hindbrain ischaemia most accurately (Zaina et al., 2003; Mitchell, 2005, 2008). However, few studies report blood ﬂow measurements at this site (Rossitti and Volkmann, 1995; Li et al., 1999; Mitchell, 2003; Mitchell et al., 2004). Various degrees of cervical spine rotation and extension have been implicated in compromised VA blood ﬂow and associated VBI (Refshauge, 1994; Sturzenneger et al., 1994; Licht et al., 1998b; Rivett et al., 1998, 1999; Li et al., 1999; Scheel et al., 2000a; Haynes and Milne, 2001). Reports of the effect of cervical spine rotation on both ICA and VA blood ﬂow found either no signiﬁcant change in ICA blood ﬂow or a decrease in both VA and ICA blood ﬂow on rotation and extension. (Schoning et al., 1994; Rivett et al., 1999; Scheel et al., 2000a, b; Licht et al., 2002). An increase followed by a decrease in VA and ICA blood ﬂow at 451 and full rotation, respectively, is also reported (Refshauge, 1994). Therefore, it is apparent from these studies that blood ﬂow changes in both the VA and ICA, associated with cervical spine movements, should be taken into account in patient pre-treatment risk assessment.

ARTICLE IN PRESS R. Kerry et al. / Manual Therapy 13 (2008) 278–288

Some researchers, using patients experiencing mild signs of VBI, found a decrease in blood ﬂow (at midcervical levels) on cervical spine rotation (Rivett et al., 1999). Others found no signiﬁcant decrease in VA blood ﬂow (at the pre-transverse level), despite positive VBI signs reported by their subjects (Licht et al., 2000, 2002). It is not clear from these ﬁndings if the VBI symptoms altered with cervical spine rotation. It is notable that most of the studies of decreased blood ﬂow following rotation, using normal subjects aged 20–40 years, report no accompanying signs and symptoms of VBI. Therefore, it is apparent that no correlation between cervical spine rotation and VBI symptoms can be clearly established from this research. The contradictory ﬁndings of these studies preclude deﬁnitive conclusions being made and emphasise the need for more rigorous research into the complex relationship between cervical spine movement and blood ﬂow changes which must, at this stage, be regarded as a guide only in pre-treatment assessment.

5. Discussion and conclusions This review of contemporary literature regarding cervical spine MT and arterial insufﬁciency highlights several areas of clinical importance and many of the outcomes correspond with previously published reviews in this subject area. There are some ﬁndings, however, which add to existing MT knowledge and hence may inﬂuence the development of clinical reasoning and research. This review, therefore, serves to inform professional practice. Most blood ﬂow studies reported refer to VA blood ﬂow in relation to cervical spine movement. Although some studies report negative ﬁndings, there is an overall trend suggesting that both VA and ICA blood ﬂow are inﬂuenced by full-range cervical spine movements, although there are not always corresponding cerebroischaemic symptoms. There is evidence that cervical spine movement, principally rotation, inﬂuences VA blood ﬂow. ICA blood ﬂow has been shown to be inﬂuenced more by extension with rotation (Scheel et al., 2000a, b), with some other reports implicating extension as the most inﬂuential movement for ICA ﬂow (Haneline et al., 2003). It is of interest to note here that the extension component of functional testing has been removed from MT pre-treatment guidelines (Magarey et al., 2004; APA, 2006). Recent guidelines (Magarey et al., 2004; APA, 2006) aim to assist in identifying patients who are more likely to suffer adverse neurovascular events should a particular treatment technique be used. One of the central components to the guidelines for pre-treatment screening of patients is the performance of a functional positioning test (the VBI test involving cervical spine

283

rotation and observing the reproduction of cardinal signs and symptoms). Many blood ﬂow studies that demonstrate changes in blood ﬂow during cervical spine rotation, for example, conclude that their ﬁndings support the use of the functional test (the test uses the position to alter blood ﬂow, altered blood ﬂow has been demonstrated in this study, therefore the test is valid). However, there is little evidence to indicate a correlation between blood ﬂow changes and symptoms of VBI (Licht et al., 1998a; Mitchell, 2003, 2007; Mitchell et al., 2004). Thus, a subject can have reduced blood ﬂow to the brain but display no signs of cerebral ischaemia. It has been reported earlier that the ICA and VA arterial supplies are parts of an interdependent and compensatory system. Therefore, it is unlikely that ischemic signs would be displayed if there is adequate compensatory/ collateral blood ﬂow. The disparity between symptoms and blood ﬂow has been highlighted further by probability analysis regarding utility of the functional test, which has reported its poor sensitivity and variable speciﬁcity (Kerry and Rushton 2003; Gross et al., 2005; Ritcher and Reinking, 2005). At the current level of evidence, pre-manipulative screening tests should be considered no more clinically valuable than Homan’s test, for example, which is still used as a small part of wider examination procedures–in lower limb deep vein thrombosis risk assessment protocols despite its known poor reliability (Levi et al., 1999). The cardinal signs and symptoms of VBI also lack consistent support from the literature reviewed. There is a diversity of presentations demonstrated in the cases reviewed. The most consistent recurring sign of VA and ICA dysfunction is an increase in neck/head pain. Localised, somatic pain referral, related to the ICA (Munari et al., 1994) and the VA (Krespi et al., 2002), has been reported. This presents an obvious diagnostic challenge for the manual therapist. Other symptoms can be attributed to dysfunction of the ICA (e.g. Horner’s syndrome, ﬂashing sensations) and may be erroneously interpreted as VBI, or of more concern, because these symptoms do not form part of the classic cardinal signs, not attributed to a neurovascular cause at all. An interesting trend in the literature is the distinction between younger patients (o45 years) who have been reported to suffer spontaneous or traumatic dissection but may not demonstrate any of the classical vascular risk factors (Rubinstein et al., 2005), and older patients (445 years), who show signs of vascular pathology, including atherosclerosis of the cervical vessels (Caplan et al., 1992; Cagnie et al., 2006), who may demonstrate classical (general or systemic) risk factors for vascular disease. However, other reports have suggested that, based on a cadaver study, the presence of VA atherosclerosis in young (o45 years) subjects cannot be excluded (Mitchell, 2002). This illustrates the

ARTICLE IN PRESS 284

R. Kerry et al. / Manual Therapy 13 (2008) 278–288

importance of understanding the complexity of CAD, particularly that cerebral ischaemia may occur due to a variety of underlying factors and pathologies. Considering that there is a risk, it is important to consider the size of the risk. Many reviews attempt to establish a risk ratio, but these are misleading ﬁgures based on under-reporting, methodological ﬂaws, and an erroneous perception of probability. The actual risk of post-treatment complications cannot be determined from the available evidence, and neither clinicians nor patients should be misled into believing that the risk is of a particular, established magnitude. The risk should be calculated on all current factors surrounding the particular situation at the particular time, for each patient. There have been few reports explicitly categorising deﬁnitive risk factors for VA and ICA dysfunction. Some have suggested that post-treatment accidents are unpredictable and that there are no deﬁnitive risk factors (Haldeman et al., 2002b). There are too few reports, however, which have focussed speciﬁcally on risk factors to make a conclusive judgment on this issue. Only well-structured, large prospective investigations could provide this information. There is a consistent suggestion in the literature reviewed that compromised blood ﬂow to the brain and thrombo-embolic events are related to various established vascular disease risk factors (hypertension, hypercholesterolaemia, atherosclerosis, trauma, infection, migraine etc). In the absence of strong evidence against this link, it might be prudent for the manual therapist not to dismiss such risk factors. Considering the inconsistencies in supporting evidence for: (1) cardinal signs of arterial insufﬁciency; (2) validity of functional testing; and (3) identiﬁcation of deﬁnite risk factors (for post-manipulation vascular injury), the acceptance of and adherence to clinical guidelines that incorporate inclusion of these areas presents an interesting professional challenge. The present medico-legal climate makes unreasoned compliance to guidelines that are not based on consistent evidence legally indefensible, and the opinions of expert witnesses may be challenged in court (Bolitho v City and Hackney, 1997; Foster, 2002, p. 116). 5.1. Risk versus benefit This paper has purposefully focussed on potentially serious adverse cervico-cranial events associated with MT. It was not the intention of the authors to report speciﬁcally on the suspected beneﬁts of cervical MT. However, in the clinical decision-making process, it is unrealistic to consider risk without beneﬁts, and vice versa. There have been several randomized controlled trials investigating the effectiveness of cervical MT (e.g. Koes et al., 1992a, b; Boline et al., 1995; Nilson et al., 1996;

Bronfort et al., 2001; Hoving et al., 2002; Jull et al., 2002a) with increasing evidence that cervical MT alone, or as part of multi-modal management strategy, is beneﬁcial for the relief of neck pain (Hurwitz et al., 1996; Bronfort et al., 2004; Gross et al., 2004). In order to make a statistically driven decision on the risk:beneﬁt ratio, valid data are needed on both sides of this analysis. As stated earlier, valid data relating to size of risk is incomplete. Intervention choices must also be made in the context of alternative interventions and their known risks. Commonly, authors compare cervical MT with common medical interventions, such as nonsteroidal anti-inﬂammatories, and state that the risk of MT is less than the known risks associated with medication (Dabbs and Lauretti, 1995; Hurwitz et al., 1996; Jull et al., 2002). Although this is likely, without supporting data such claims should be noted with a degree of caution. Despite the lack of data, a small number of authors have commented speciﬁcally on the risk:beneﬁt judgement and have come to ﬁrm conclusions. DiFabio (1999) explicitly states that the risks of cervical manipulation outweigh the beneﬁts. This follows a review of 177 cases of adverse events in the context of selected literature exploring the proposed beneﬁts of cervical manipulation. Conversely, a more recent study states that the potential beneﬁts of cervical therapy do outweigh the risks (Rubinstein et al., 2007). This statement is based on the results of a prospective follow-up survey of clinicians’ reports of responses to treatment (as discussed above). Similarly, whilst this is likely to be the case, such judgements need to be considered in the context of the limitations of the data and methodologies. Childs et al. (2005) embrace the inherent uncertainties in this clinical area and provide discussion on making reasoned judgements in the absence of good data – not only for risks and beneﬁts, but also in respect of the validity of screening procedures for potential risk. The outcomes of this review reﬂect this thinking: incorporating knowledge of haemodynamic and vascular pathophysiology into patient assessment may lead the clinician to an informed judgement as to the likelihood of serious adverse events. If the limitations of traditional screening processes are understood and the evidence regarding beneﬁt considered, a sound and reasonable decision may be reached regarding choice of intervention. 5.2. Recommendations There is a clear need for further research in a number of areas in this ﬁeld. The indication for large-scale, robust, prospective trials has been consistently reported as the best way to establish a reliable indicator of risk of arterial insufﬁciency related to cervical spine MT.

ARTICLE IN PRESS R. Kerry et al. / Manual Therapy 13 (2008) 278–288

However, considering the size of the task, this would be logistically difﬁcult. Until a stronger evidence-base can be established, it is advisable that manual therapists take note of the available evidence supporting all of the possible generic vascular risk factors, prior to cervical spine MT. Based on the ﬁndings of this review, the authors make the following recommendations: 1. Considering the limited support for the use of cervical spine functional pre-manipulative screening tests (e.g. the VBI test), as an indicator of either vascular patency or of the propensity of blood vessels to be injured as a result of MT, these tests should be used with caution and the results should be interpreted in the knowledge that the tests have little backing from the evidence base. 2. There is evidence supporting the relationship between vascular disease risk factors and CAD. As such the authors recommend a subjective assessment of vascular risk factors incorporating a ‘system’-based approach (i.e. incorporating ICA and VA knowledge). 3. Headache and neck pain are common presenting signs of vascular dissection. Therefore, careful consideration of the differential diagnosis of these symptoms by the manual therapist is recommended. 4. As cervical artery haemodynamics (inclusive of the ICA and therefore the system) are inﬂuenced by movement as a whole, and not exclusively by single thrust manipulative procedures, vascular risk assessment should be considered prior to all manual therapy procedures that induce movement in the cervical spine region (APA, 2006). 5. The evidence suggests that conﬁning pre-treatment risk assessment to the VA in isolation may represent a limitation in clinical reasoning. Consideration of the cervical arterial system, together with the range of vascular pathologies apparent within this system, may enhance the clinician’s reasoning process.

Acknowledgements The authors would like to acknowledge the support of and thank: the Manipulation Association of Chartered Physiotherapists (MACP) for its support in this project; Steven Barrett and the Research Support Unit, University of Nottingham, for assistance in the literature searches involved in this review; and Lindsay Skipper, University of Nottingham for proof-reading this paper. References Arnold C, Bourassa T, Langer T, Stoneham G. Doppler studies evaluating the effect of a physical therapy screening protocol on vertebral artery blood ﬂow. Manual Therapy 2004;9:13–21.

285

Arnold M, Cumurciuc R, Stapf C, Favrole P, Berthet K, Bousser MG. Pain as the only symptom of cervical artery dissection. Journal of Neurology Neurosurgery and Psychiatry 2006;77(9):1021–4. Assendelft WJJ, Bouter LM, Knipschild PG. Complications of spinal manipulation—a comprehensive review of the literature. The Journal of Family Practice 1996;42(5):475–80. Australian Physiotherapy Association (APA). Clinical guidelines for assessing vertebrobasilar insufﬁciency in the management of cervical spine disorders. /www.Physiotherapy.asn.auS. 2006. Australian Journal of Physiotherapy (AJP). Forum: pre-manipulative testing of the cervical spine. Australian Journal of Physiotherapy 2001;47(3):163–8. Barker S, Kesson M, Asmore J, Turner G, Conway J, Stevens D. Guidance for pre-manipulative testing of the cervical spine. Physiotherapy 2001;87(6):318–22. Beaudry M, Spence JD. Motor vehicle accidents: the most common cause of traumatic vertebrobasilar ischemia. Canadian Journal of Neurological Sciences 2003;30(4):320–5. Boline P, Kassak K, Bronfort G, Nelson C, Anderson AV. Spinal manipulation vs amitriptyline for the treatment of chronic tension type headaches—a randomized controlled trial. Journal of Manipulative and Physiological Therapeutics 1995;18:148–54. Bolitho v City and Hackney HA [1997] 4 All ER 771. Brandt T, Orberk E, Weber R, Werner I, Busse O, Muller BT, et al. Pathogenesis of cervical artery dissections: association with connective tissue abnormalities. Neurology 2001;57(1):24–30. Brew J. Consent to touch, consent to treat. Invited address. CSP Congress 2004, Birmingham, UK. Bronfort GR, Evans R, Nelson B, Aker PD, Goldsmith CH, Vernon H. A randomized clinical trial for patients with chronic neck pain. Spine 2001;26:788–99. Bronfort G, Haas M, Evans RL, Bouter LM. Efﬁcacy of spinal manipulation and mobilisation for low back pain and neck pain: a systematic review and best evidence synthesis. The Spine Journal 2004;4:335–56. Brown BStJ, Tissington-Tatlow WF. Radiographic studies of the vertebral arteries in cadavers. Radiology 1963;81:80–8. Cagnie B, Barbaix E, Vinck E, D’Herde K, Cambier D. Atherosclerosis in the vertebral artery: an intrinsic risk factor in the use of spinal manipulation? Surgical and Radiological Anatomy 2006; 28(2):129–34. Caplan LR, Biousse V. Cervicocranial arterial dissections. Journal of Neuroophthalmology 2004;24(4):299–305. Caplan LR, Amarenco P, Rosengart A, Lafranchise EF, Teal PA, Belkin M, et al. Embolism from vertebral artery origin occlusive disease. Neurology 1992;42(8):1505–12. Cardon A, Kerdiles Y, Lucas A, Podeur L, Ferte L, Le Du J, et al. Results of isolated carotid surgery in patients with vertebrobasilar insufﬁciency. Annals of Vascular Surgery 1998; 12(6):579–82. Childs JD, Flynn TW, Fritz JM, Piva SR, Whitman JM, Wainner RS, Greenman PE. Screening for vertebrobasilar insufﬁciency in patients with neck pain: manual therapy decision-making in the presence of uncertainty. Journal of Othopaedic and Sports Physiotherapy 2005;35(5):300–6. Chung YS, Han DH. Vertebrobasilar dissection: a possible role of whiplash injury in its pathogenesis. Neurological Research 2002;24(2):129–38. Clarot F, Vaz E, Papin F, Proust B. Fatal and non-fatal bilateral delayed carotid artery dissection after manual strangulation. Forensic Science International 2005;149(2–3):143–50. Dabbs V, Lauretti WJ. A risk assessment of cervical manipulation vs NSAIDs for the treatment of neck pain. Journal of Manipulative and Physiological Therapeutics 1995;18:530–6. DiFabio RP. Manipulation of the cervical spine: risks and beneﬁts. Physical Therapy 1999;79(1):50–65.

ARTICLE IN PRESS 286

R. Kerry et al. / Manual Therapy 13 (2008) 278–288

Dupeyron A, Vautravers JL, Isner-Horobeti ME. Complications following vertebral manipulation – a survey of a French region physicians. Annales de Readaptation et de Medicine Physique 2003;46(1):33–40. Dvorak J, Loustlot D, Baumgarner H, Antinnes. Frequency of complications of manipulation of the spine – a survey among the members of the Swiss Medical Society of Manual Therapy. European Spine Journal 1993;2:136–9. Ernst E. Life-threatening complications of spinal manipulation. Stroke 2001a;23:809–10. Ernst E. Prospective investigations into the safety of spinal manipulation. Journal of Pain and Symptom Management 2001b;21(3): 238–42. Ernst E. Manipulation of the cervical spine: a systematic review of case reports of serious adverse events 1995–2001. Journal of the Royal Society of Medicine 2002;176(4):376–80. Ernst E. Cerebrovascular complications associated with spinal manipulation. Physical Therapy Reviews 2004;9:5–15. Ersoy H, Watts R, Sanelli P, Zimmerman RD, Kent KC, Bush HL, et al. Atherosclerotic disease distribution in carotid and vertebrobasilar arteries: clinical experience in 100 patients undergoing ﬂuoro-triggered 3D Gd-MRA. Journal of Magnetic Resonance Imaging 2003;17(5):545–58. Fancher TL, White RH, Kravitz RL. Combined use of rapid D-dimer testing and estimation of clinical probability in the diagnosis of deep vein thrombosis: systematic review. British Medical Journal 2004;329(7470):821. Fletcher J, Davies PT, Lewis T, Campbell MJ. Traumatic carotid and vertebral artery dissection in a professional jockey: a cautionary tale. British Journal of Sports Medicine 1995;29(2):143–4. Flossmann E, Schulz UG, Rothwell PM. Systematic review of methods and results of studies of the genetic epidemiology of ischemic stroke. Stroke 2004;35(1):212–27. Foster C. Civil Procedure, Trail Issues and Clinical Guidelines. In: Tingle J, Foster C, editors. Clinical Guidelines: law, policy and practice. Cavendish: London; 2002. Frisoni B, Anzola GP. Vertebrobasilar ischemia after neck motion. Stroke 1991;22:1452–60. Garg A, Gaikwad SB, Kanodia A, Suri A, Gupta V, Mishralf NK. Positional occlusion/stasis of vertebral arteries in a case of cervical rheumatoid arthritis presenting with multiple posterior circulation infarcts: a case report with angiographic demonstration. Spine 2004;29(15):E321–5. Gates J, Hartnell GG, Quintas J. Young adult with ‘‘worst ever headache’’ following mild trauma. Acadademic Radiology 1997;4(4):312–3. Giannoukas AD, Labropoulos N, Smith FC, Venables GS, Beard JD. Management of the near total internal carotid artery occlusion. European Journal of Vascular and Endovascular Surgery 2005;29(3):250–5. Giroud M, Fayolle H, Andre N, Dumas R, Becker F, Martin D, et al. Incidence of internal carotid artery dissection in the community of Dijon. Journal of Neurology Neurosurgery and Psychiatry 1994;57(11):1443. Grant AC, Wang N. Carotid dissection associated with a handheld electric massager. Southern Medical Journal 2004;97(12):1262–3. Grau A, Buggle F. Infection atherosclerosis and acute ischemic cerebrovascular disease. Review of Neurology 1999;29(9):847–51. Gross AR, Hoving JL, Haines TA, Goldsmith CH, Kay T, Aker P. A Cochrane review of manipulation and mobilization for mechanical neck disorders. Spine 2004;29:1541–8. Gross AR, Chesworth B, Binkley J. A case for evidence based practice in manual therapy. In: Boyling J, Jull GA, editors. Grieve’s modern manual therapy—The Vertebral Column. 3rd ed. Edinburgh: Churchill Livingstone; 2005. p. 569–80 chapter 39. Haldeman S, Kohlbeck FJ, McGregor M. Risk factors and precipitating neck movements causing vertebrobasilar artery

dissection after cervical trauma and spinal manipulation. Spine 1999;24(8):785–94. Haldeman S, Carey P, Townsend M, Papadopoulos C. Clinical perceptions of the risk of vertebral artery dissection after cervical manipulation: the effects of referral bias. The Spine Journal 2002a;2:334–42. Haldeman S, Kohlbeck FJ, McGregor M. Stroke, cerebral srtary dissection, and cervical spine manipulation. Journal of Neurology 2002b;249:1098–104. Haldeman S, Kohlbeck FJ, McGregor M. Unpredicatbility of cerebrovascular ischemia associated with cervical spine manipulation therapy—a review of sixty-four cases after cervical spine manipulation. Spine 2002c;27(1):49–55. Haneline MT, Croft AC, Frishberg BM. Association of internal carotid artery dissection and chiropractic manipulation. Neurologist 2003;9(1):35–44. Haynes MJ. Stroke following cervical manipulation in Perth. Chiropractic Journal of Australia 1994;24(2):42–6. Haynes MJ, Milne N. Color Duplex sonographic ﬁndings in human vertebral arteries during cervical rotation. Journal of Clinical Ultrasound 2001;29:14–24. Haynes M, Hart R, McGeachie J. Vertebral arteries and neck rotation: Doppler velocimeter interexaminer reliability. Ultrasound in Medicine and Biology 2000;26:1363–7. Hoving JL, Koes BW, deVet HC, van der Windt DA, Assendelft WJ, van Mameren H, et al. Manual therapy, physical therapy or continued care by a general practitioner for patients with neck pain. Annals of Internal Medicine 2002;136:713–22. Hurwitz EL, Aker PD, Adams AH, Meeker WC, Shekelle PG. Manipulation and mobilization of the cervical spine: a systematic review of the literature. Spine 1996;21(15):1746–59. Johnson C, Grant R, Dansie B, Taylor J, Spyropolous P. Measurement of blood ﬂow in the vertebral artery using colour duplex Doppler ultrasound: establishment of the reliability of selected parameters. Manual Therapy 2000;5:21–9. Jull G, Magarey M, Niere K, Elvey R. Physiotherapy, a responsible profession to use cervical manipulation. Response to Refshauge et al. Australian Journal of Physiotherapy 2002;48(2):180–5. Kerry R. Pre-manipulative procedures for the cervical spine—new guidelines and a time for dialectics: knowledge, risks, evidence and consent. Physiotherapy 2002;88(7):417–20. Kerry R. Vertebrobasilar Insufﬁciency—the debate and the evidence? Keynote lecture. CSP Congress 2004, Birmingham, UK. Kerry R, Rushton A. Decision theory in physical therapy. World Confederation for Physical Therapy 14th International Congress 2003. Barcelona, Spain. Khaw AV, Schminke U, Bleiss A, Kessler C. Echo-enhanced duplex sonography of extracranial vertebral arteries. Ultrasound in Medicine and Biology 2004;30:139–46. Koes BW, Bouter LM, van Mameren H. The effectiveness of manual therapy, physiotherapy, and treatment by the general practitioner for non-speciﬁc back and neck complaints. A randomized clinical trial. Spine 1992a;17:28–35. Koes BW, Bouter LM, van Mameren H. Randomised clinical trial of manipulative therapy and physiotherapy for persistent back and neck complaints: results of one year follow-up. BMJ 1992b;304:601–5. Krespi Y, Gurol ME, Coban O, Tuncay R, Bahar S. Vertebral artery dissection presenting with isolated neck pain. Journal of Neuroimaging 2002;12(2):179–82. Kunasmaa KTTT, Thiel HW. Vertebral artery syndrome: a review of the literature. The Journal of Orthopaedic Medicine 1994;16(1): 17–20. Leboeuf-Yde C, Hennius B, Rudberg E, Leufvenmark P, Thunmann M. Side effects of chiropractic reatment: a prospective study. The Journal of Manual and Manipulative Therapy 1997;20(8):511–5.

ARTICLE IN PRESS R. Kerry et al. / Manual Therapy 13 (2008) 278–288 Lee PK, Carlini WG, McCormick GF, Albers GW. Neurologic complicatios following chiropractic manipulation: a survey of California neurologists. Neurology 1995;45(6):1213–5. Levi M, Hart W, Buller HR. Physical examination; the signiﬁcance of Homan’s sign. Ned Tijdschr Geneeskd 1999;143(37):1861–3. Li C, Engstrom G, Hedblad B, Berglund G, Janzon L. Blood pressure control and risk of stroke: a population-based prospective cohort study. Stroke 2005;36(4):725–30. Li YK, Zhang YK, Lu CM, Zhong SZ. Changes and implications of the blood ﬂow velocity of the vertebral artery during rotation and extension of the head. Journal of Manipulative and Physiological Therapeutics 1999;22:91–5. Licht PB, Christensen HW, Hojgaard P, Hoilund-Carlsen PF. Triplex ultrasound of vertebral artery ﬂow during cervical rotation. Journal of Manipulative and Physiological Therapeutics 1998a;21: 27–31. Licht PB, Christensen HW, Hojgaard P, Marving J. Vertebral artery ﬂow and spinal manipulation: a randomized, controlled and observer-blinded study. Journal of Manipulative and Physiological Therapeutics 1998b;21:141–4. Licht PB, Christensen HW, Hoilund-Carlsen PF. Is there a role for premanipulative testing before cervical manipulation? Journal of Manipulative & Physiological Therapeutics 2000;23:175–9. Licht PB, Christensen HW, Hoilund-Carlsen PF. Carotid artery blood ﬂow during premanipulative testing. Journal of Manipulative and Physiological Therapeutics 2002;25(9):568–72. Lisovoski F, Rousseaux P. Cerebral infarction in young people. A study of 148 patients with early cerebral angiography. Journal of Neurology Neurosurgery and Psychiatry 1991;54(7):576–9. Magarey ME, Rebbeck T, Coughlan B, Gimmer K, Rivett DA, Refshauge K. Pre-manipulative testing of the cervical spine review, revision and new clinical guidelines. Manual Therapy 2004;9(2): 95–108. Mann T, Refshauge KM. Causes of complications from cervical spine manipulation. Australian Journal of Physiotherapy 2001;47(4): 255–66. Michaeli A. Reported occurrence and nature of complications following manipulative physiotherapy in South Africa. Australian Journal of Physiotherapy 1993;39(4):309–15. Migdalski A, Kotschy M, Jawien A. Tissue factor, tissue factor pathway inhibitor and vascular endothelial growth factor-A in carotid atherosclerotic plaques. European Journal of Vascular and Endovascular Surgery 2005;30(1):41–7. Miller PR, Fabian TC, Croce MA, Cagiannos C, Williams JS, Vang M, et al. Prospective screening for blunt cerebrovascular injuries: analysis of diagnostic modalities and outcomes. Annals of Surgery 2002;236(3):386–93. Mitchell J. Vertebral artery atherosclerosis: a risk factor in the use of manipulative therapy? Physiotherapy Research International 2002;7(3):122–35. Mitchell J. Changes in vertebral artery blood ﬂow following normal rotation of the cervical spine. Journal of Manipulative and Physiological Therapeutics 2003;26:347–51. Mitchell J. The vertebral artery: a review of anatomical, histopathological and functional factors inﬂuencing blood ﬂow to the hindbrain. Physiotherapy Theory and Practice 2005;21:23–36. Mitchell J. Is mechanical deformation of the suboccipital vertebral artery during cervical spine rotation responsible for vertebrobasilar insufﬁcienct? Physiotherapy Research International; 2007a [ 15 May 2007] /http://www.interscience.wiley.comS. Mitchell J. Doppler insonation of vertebral artery blood ﬂow changes associated with cervical spine rotation—implications for manual therapists. Physiotherapy Theory and Practice 2008;26:303–13. Mitchell J, Keene D, Dyson C, Harvey L, Pruvey C, Phillips R. Is cervical spine rotation, as used in the standard vertebrobasilar insufﬁciency test, associated with a measurable change in

287

intracranial vertebral artery blood ﬂow? Manual Therapy 2004;9:220–7. Munari LM, Belloni G, Moschini L, Mauro A, Pezzuoli G, Porta M. Carotid pain during percutaneous angioplasty (PTA). Pathophysiology and clinical features. Cephalalgia 1994;14(2): 127–31. Nadareishvili Z, Norris JW. Stroke from traumatic arterial dissection. The Lancet 1999;354:159–60. Nilson N, Christenson HW, Hartvigsen J. Lasting changes in passive range of motion after spinal manipulation: a randomized, blind, controlled trial. Journal of Manipulative and Physiological Therapeutics 1996;19:165–8. Paivansalo M, Heikkila O, Tikkakoski T, Leinonen S, Merikanto J, Suramo I. Duplex ultrasound in the subclavian steal syndrome. Acta Radiology 1998;39(2):183–8. Pezzini A, Del Zotto E, Padovani A. Hyperhomocysteinemia: a potential risk factor for cervical artery dissection following chiropractic manipulation of the cervical spine. Journal of Neurology 2002;249(10):1401–3. Pollanen MS, Deck JH, Boutilier L, Davidson G. Lesions of the tunica media in traumatic rupture of vertebral arteries: histologic and biochemical studies. Canadian Journal of Neurological Science 1992;19(1):53–6. Refshauge KM. Rotation: a valid premanipulative dizziness test? Does it predict safe manipulation? Journal of Manipulative and Physiological Therapeutics 1994;17:15–9. Refshauge KM, Parry S, Shirlet D, Larsen D, Rivett DA, Boland R. Professional responsibility in relation to cervical spine manipulation. Australian Journal of Physiotherapy 2002;48(2):171–9. Ritcher RR, Reinking MF. Clinical Question: how does evidence on the diagnostic accuracy of the vertebral artery test inﬂuence teaching of the test in a professional physical therapist education program? Physical Therapy 2005 /http://www.ptjournal.org/ PTJournal/Jun2005/Jun05_EiP.cfmS. Rivett DA. Neurovascular compromise complicating cervical spine manipulation: what is the risk? The Journal of Manual and Manipulative Therapy 1995;3(4):144–51. Rivett DA. Preventing neurovascular complications of cervical spine manipulation. Physical Therapy Reviews 1997;2(1):29–37. Rivett DA, Milburn P. A prospective study of complications of cervical spine manipulation. The Journal of Manual and Manipulative Therapy 1996;4(4):166–70. Rivett DA, Milburn P. Complications arising from spinal manipulation therapy in New Zealand. Physiotherapy 1997;83(12):626–32. Rivett DA, Milburn PD, Chapple C. Negative pre-manipulative vertebral artery testing despite occlusion: a case of false negativity? Manual Therapy 1998;3:102–7. Rivett DA, Sharples KJ, Milburn PD. Effect of premanipulative tests on vertebral artery and internal carotid artery blood ﬂow: a pilot study. Journal of Manipulative and Physiological Therapeutics 1999;22:368–75. Ross R. Atherosclerosis is an inﬂammatory disease. American Heart Journal 1999;138:419–20. Rossitti S, Volkmann R. Changes of blood ﬂow velocity indicating mechanical compression of the vertebral arteries during rotation of the head in the normal human measured with transcranial Doppler sonography. Arquivos de Neuro-Psiquiatria 1995;53:26–33. Rothwell DM, Bondy SJ, Williams JL. Chiropractic manipulation and stroke: a population-based case-control study. Stroke 2001;32: 1054–60. Rubinstein SM, Peerdeman SM, van Tulder MW, Riphagen I, Halderman S. A systematic review of the risk factors for cervical artery dissection. Stroke 2005;36:1575–80. Rubinstein SM, Leboeuf-Yde C, Knol DL, deKoekkoek TE, Pﬁeﬂe CE, van Tulder MW. The beneﬁts outweigh the risks for patients undergoing chiropractic care for neck pain: a prospective, multi-

ARTICLE IN PRESS 288

R. Kerry et al. / Manual Therapy 13 (2008) 278–288

center, cohort study. Journal of Manipultive and Physiological Therapeutics 2007;30(6):408418. Scheel P, Ruge C, Schoning M. Flow velocity and ﬂow volume measurements in the extracranial carotid and vertebral arteries in healthy adults: reference data and the effects of age. Ultrasound in Medicine and Biology 2000a;26:1261–6. Scheel P, Ruge C, Petruch UR, Schoning M. Color duplex measurement of cerebral blood ﬂow volume in healthy adults. Stroke 2000b;31:147–50. Schievink WI, Atkinson JL, Bartleson JD, Whisnant JP. Traumatic internal carotid artery dissections caused by blunt softball injuries. American Journal of Emergency Medicine 1998;16(2):179–82. Schievink WI, Thompson RC, Loh CT, Maya MM. Spontaneous retroclival haematoma presenting as a thunderclap headache. Case report. Journal of Neurosurgery 2001;95(3):522–4. Schoning M, Walter J, Scheel P. Estimation of cerebral blood ﬂow through color duplex sonography of the carotid and vertebral arteries in healthy adults. Stroke 1994;25:17–22. Schoning M, Hartig B. The development of hemodynamics in the extracranial carotid and vertebral arteries. Ultrasound Medical Biology 1998;24(5):655–62. Schwarze U, Atkinson M, Hoffman GG, Greenspan DS, Byers PH. Null alleles of the COL5A1 gene of type V collagen are a cause of the classical forms of Ehlers – Danlos syndrome (types I and II). American Journal of Human Genetics 2000;66(6):1757–65. Selecki BR. The effects of rotation of the atlas on the axis: experimental work. Medical Journal of Australia 1969;1:1012–5. Senstad O, Leboeuf-Yde C, Borchgrevink CF. Side-effects of chiropractic spinal manipulation: types, frequency, discomfort and course. Scandinavian Journal of Primary Health Care 1996;14: 50–3. Shekelle PG, Coulter I. Cervcal spine manipulation: summary report of a systematic review of the literature and a multidisciplinary expert panel. Journal of Spinal Disorders 1997;10(3):223–8. Shimada T, Toyoda K, Hagiwara N, Sayama T, Inoue T, Yasumori K, et al. Recurrent embolic stroke originating from an internal carotid aneurysm in a young adult. Journal of Neurological Science 2005;232(1–2):115–7. Sidhu PS. Ultrasound of the carotid and vertebral arteries. British Medical Bulletin 2000;56(2):346–66. Smith WS, Johnston SC, Skalabrin EJ, Weaver M, Azari P, Albers GW, et al. Spinal manipulation therapy is an independent risk factor for vertebral artery dissection. Neurology 2003;60(5):1424–8. Stevinson C, Honan W, Cooke B, Ernst E. Neurological complications of cervical spine manipulation. Journal of the Royal Society of Medicine 2001;94(3):107–10. Sturzenneger M, Newell DW, Douville C, Byrd S, Schoonover K. Dynamic transcranial Doppler assessment of positional vertebrobasilar ischemia. Stroke 1994;25:1776–83.

Taylor AJ. Vascular assessment in musculoskeletal examination. Physiotherapy 2001;87(6):282–4. Taylor AJ, Kerry R. Neck pain and headache as a result of internal carotid artery dissection: implications for manual therapists. Manual Therapy 2005;10(1):73–7. Terrett AGJ. Vascular accidents from cervical spine manipulation: report on 107 cases. Journal of the Australian Chiropractors’ Association 1987;17(3):15–24. Thiel H, Rix G. Is it time to stop functional pr-manipulation testing of the cervical spine? Manual Therapy 2005;10(2):148–54. Thiel H, Wallace K, Donat J, Yong-Hing K. Effects of various head and neck positions on vertebral artery blood ﬂow. Clinical Biomechanics 1994;9:105–10. Thiel HW, Bolton JE, Docherty S, Portlock JC. Safety of chiropractic manipulation of the cervical spine: a prospective national survey. Spine 2007;32(21):2375–8. Tissington-Tatlow WF, Brammer HG. Syndrome of vertebral artery compression. Neurology 1957;7:331–40. Tominaga T, Takahashi T, Shimizu H, Yoshimoto T. Rotational vertebral artery occlusion from occipital bone anomaly: a rare cause of embolic stroke. Case report. Journal of Neurosurgery 2002;97(6):1456–9. Toole JF, Tucker SH. Inﬂuence of head position upon cerebral circulation. Archives of Neurology 1960;2:616–23. Van Damme H, Quaniers J, Limet R. Fibromuscular dysplasia. Revue Medicale de Leige 1999;54(12):935–42. Volle E, Montazem A. MRI video diagnosis and surgical therapy of soft tissue trauma to the craniocervical junction. Ear Nose Throat Journal 2001;80(1):41–8. Weingart JR, Bischoff H-P. Doppler-sonographische Untersuchung der A. vertebralis unter Berucksichtigung chirotherapeutisch relevanter Kopfpositionen. Manuelle Medizin 1992;30:62–5. Weintraub MI, Khoury A. Cerebral hemodynamic changes induced by simulated tracheal intubation: a possible role in perioperative stroke? Magnetic resonance angiography and ﬂow analysis in 160 cases. Stroke 1998;29(8):1644–9. Welch HJ, Murphy MC, Raftery KB, Jewell ER. Carotid duplex with contralateral disease: the inﬂuence of vertebral artery blood ﬂow. Annals Vascular Surgery 2000;14(1):82–8. Westaway MD, Stratford P, Symons B. False-negative extension/ rotation pre-manipulative screening test on a patient with an atretic and hypoplastic vertebral artery. Manual Therapy 2003;8:120–7. Yamazaki M, Okawa A, Aramomi MA, Hashimoto M, Masaki Y, Koda M. Fenestration of vertebral artery at the craniovertebral junction in Down syndrome: a case report. Spine 2004;29(23): 551–4. Zaina C, Grant R, Johnson C, Dansie B, Taylor J, Spyropolous P. The effect of cervical rotation on blood ﬂow in the contralateral vertebral artery. Manual Therapy 2003;8:103–9.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 289–294 www.elsevier.com/locate/math

Original article

What is the normal response to structural differentiation within the slump and straight leg raise tests? Lee Herringtona,b,, Katie Bendixa, Catherine Cornwella, Nicola Fieldena, Karen Hankeya a

Directorate of Sport, University of Salford, UK Centre for Rehabilitation and Human Performance Research, University of Salford, UK

b

Received 28 September 2005; received in revised form 20 December 2006; accepted 15 January 2007

Abstract The purpose of the study was to assess the effect of structural differentiation or sensitising manoeuvres on responses of normal subjects to standard neurodynamic tests of straight leg raise (SLR) and slump test. Eighty-eight (39 males and 49 females) asymptomatic subjects were examined (aged 18–39 mean age 21.974.1 years). Knee ﬂexion angle was measured using a goniometer during the slump test in two conditions cervical ﬂexion and extension. Hip ﬂexion angle was measured using a goniometer during SLR test in two conditions; ankle dorsi-ﬂexion and neutral. The change in knee ﬂexion, following addition of the structural differentiating manoeuvre to the slump test, was a signiﬁcant increase in knee ﬂexion angle for both males (change in knee angle; 6.674.71/18.7717.5%, po0.01) and females (change in knee angle5.475.81/17.6723.7%, po0.01), though showed no difference between sides (p40.05). During the SLR test, a signiﬁcant reduction in hip ﬂexion occurred following structural differentiation for both groups (change in hip angle; males ¼ 9.578.31/21.5718.8%, po0.01; females ¼ 15.279.51/25.9713.9%, po0.01), though showed no difference between sides (p40.05). Structural differentiating manoeuvres have a signiﬁcant effect on test response in terms of range of movement even in normal asymptomatic individuals. These responses should be taken into account during the assessment clinical reasoning process. r 2007 Elsevier Ltd. All rights reserved. Keywords: Neurodynamic testing; Structural differentiation; Lower limb

1. Introduction Neurodynamic assessment has now emerged as a popular adjunct in the investigation of musculoskeletal injuries. These tests are used in assessment to gain an impression of neural tissue mobility and sensitivity to mechanical stress. The key to the successful interpretation of these tests is the use of structural differentiation during the tests (Shacklock, 2005a). Structural differentiation is the use of manoeuvres applied during neurodynamic testing in order emphasise the role of Corresponding author. Directorate of Sport, University of Salford, Allerton Annexe, Frederick Road, Salford, Manchester M6 6PU, UK. Tel.: +44 161 295 2326; fax: +44 161 295 2395. E-mail address: [email protected] (L. Herrington).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.01.013

neural tissue as opposed to musculoskeletal tissues in creating a change in the test outcome (Butler, 2000). To accurately interpret these tests, the clinician must distinguish between motion restrictions due to pain and dysfunction and those which may be considered a normal or average response found in an asymptomatic population. Within the literature there have only been a limited number of studies, which have reported the response of asymptomatic subjects to the neural structural differentiation during the lower quadrant neurodynamic tests of straight leg raise (SLR) and slump tests. The addition of cervical ﬂexion during the slump test or ankle dorsiﬂexion during the SLR test would be examples of structural differentiating manoeuvres. The addition of these manoeuvres implies increased sensitivity of neural

ARTICLE IN PRESS 290

L. Herrington et al. / Manual Therapy 13 (2008) 289–294

tissue if they result in a decreased range of movement of an element of the test movements, for example with the slump test knee ﬂexion angle, or with the SLR test hip ﬂexion angle or a reproduction of symptoms (Coppieters et al., 2005; Cleland et al., 2006). The diagnostic value of the neurodynamic tests in differentiating between neural and non-neural structures has been regarded by certain authors as limited (Di Fabio, 2001), principally because of the alternate explanation for ﬁndings which can be offered due to the continuity of the fascial system (Gajdosik et al., 1985; Lew and Briggs, 1997). The thoraco-lumbar fascia has been shown to have direct connection through into the lower limb (Vleeming et al., 1995) and to the tendons of splenius capitis and cervivis muscles in the neck (Barker and Briggs, 1999). This has led to the argument that positive ﬁndings particularly with the slump test and its differentiating manoeuvre of cervical ﬂexion could be related equally to increased tension in the fascial system (Barker and Briggs, 1999). The hypothetical conclusions from these studies investigating fascial attachment in cadavers are difﬁcult to dispute. One study has attempted to show the ability of the structural differentiating manoeuvres to discretely distinguish neural tissue involvement (Coppieters et al., 2005). Coppieters et al. (2005) using experimentally induced calf muscle pain demonstrated that the addition of structural differentiating manoeuvres to both the slump test and SLR test that had no signiﬁcant additive effect on pain perception. In fact, as progressively increased tension was applied with these manoeuvres, there was a trend towards decreased pain perception (Coppieters et al., 2005). In a similar vein, Lew and Briggs (1997) found that the addition of cervical ﬂexion in slump, with the thigh held in a position of maximal knee extension increased posterior thigh pain, but did not increase muscle tension, measured indirectly using EMG. These ﬁndings may to a degree dispute the role of fascial tensioning causing increased pain perception with the application of structural differentiating manoeuvres during lower limb neurodynamic testing (Cleland et al., 2006) and increase the evidence base in support of their ability to discretely distinguish neural tissue involvement. The effect of the addition of the manoeuvre of neck ﬂexion during the slump test has been investigated by a number of authors (Maitland, 1985; Johnson and Chiarello, 1997; Lew and Briggs, 1997; Yeung et al., 1997; Coppieters et al., 2005). In the study of Yeung et al. (1997), the addition of neck ﬂexion brought about a 472.81 increase in knee ﬂexion angle in a group of 40 female subjects. Johnson and Chiarello (1997) report the effect on 34 male subjects, showing neck ﬂexion to increase knee ﬂexion by 679.11. Coppieters et al. (2005) in a mixed group of 22 males and 3 females showed an increase of 2.371.71. Maitland (1985) reported changes

in knee ﬂexion angle with the addition of neck ﬂexion, but failed to measure or quantify these changes. Lew and Briggs (1997) in their study did not measure changes in knee ﬂexion angle brought about by the addition of neck ﬂexion, but rather changes in muscle tension, the ﬁndings of which have been discussed above. The wide variety of results reﬂects the different methods used both to measure knee ﬂexion angle and the maintenance of the slump position, both of which will signiﬁcantly inﬂuence the range of knee ﬂexion obtained (Herrington, 2004). The consequence of the addition of ankle dorsi-ﬂexion to the range of hip ﬂexion during the SLR test in asymptomatic subjects has received less attention than investigations into the slump test (Gajdosik et al., 1985; Boland and Adams, 2000). In the study of Boland and Adams (2000), they reported that addition of ankle dorsi-ﬂexion brought about an 8.87181 decrease in hip ﬂexion angle in a group of 10 male and 10 female asymptomatic subjects. Gajdosik et al. (1985) examining a group of 10 male and 12 female normal subjects reported a 101 decrease in hip ﬂexion angle with the addition of ankle dorsi-ﬂexion to the SLR test. With the increased use of neurodynamic testing and treatment within musculoskeletal manual therapy practice, the need to establish normal neurogenic responses to the tests becomes of increasing importance, in order to allow appropriate clinically reasoned responses to the ﬁndings of the tests (Shacklock, 2005a). The purpose of this study was to establish what effect the addition of structural differentiating manoeuvres has on range of movement of the base tests of SLR and slump. Additionally, the study aimed to establish if the level of response differed between male and female subjects or between limbs. 2. Method Subjects: Eighty-eight (39 males and 49 females) asymptomatic subjects were examined (aged 18–39 mean age 21.974.1 years). Subjects were excluded if they had any history of low back pain, lower limb muscular or joint injury, neurological or vascular impairment. All subjects gave informed written consent to participate prior to commencing study and the study was approved by the Institutional Research Ethics Committee. 2.1. Procedures 2.1.1. Slump test The subject sat on the table in the stabilisation frame with the pelvis ﬁxed against the upright support, to maintain the sacrum in a vertical position (Fig. 1). The thighs were fully supported with the popliteal fossa against the edge of the table, and the knees together. A strap was fastened across the thighs to ensure that no

ARTICLE IN PRESS L. Herrington et al. / Manual Therapy 13 (2008) 289–294

291

presence of pain in excess of a mild stretching sensation, if this occurred during testing, the test was terminated. All subjects were able to fully participate using these criteria, without the generation of symptoms and with none requesting withdrawal.

Fig. 1. Subject ﬁxed in slump position with overpressure to cervical ﬂexion applied.

hip ﬂexion took place. The subject was asked to ‘slump’ or ‘sag’ the trunk while the cervical spine was kept in a neutral position. To maintain a consistent overpressure to thoraco-lumbar ﬂexion, a strap was positioned across the shoulders just below C7 vertebra. A universal 1801 goniometer (Physiomed, UK) was positioned with the stationary arm aligned between the lateral condyle of the knee and the greater trochanter, and the moving arm aligned between the lateral condyle of the knee and the lateral malleolus of the ankle (Palmar and Epler, 1998). Zero was taken as the point of full knee extension. The cervical spine was moved into either ﬂexion or extension as speciﬁed by the randomised test order. For the position of cervical extension the subject was told to look up towards the ceiling as far as they could, and to maintain that position. For cervical ﬂexion, the subject was asked to put their chin to their chest. In this position, an assistant applied overpressure. Once the correct neck position had been achieved, the examiner applied ankle dorsi-ﬂexion and passively extended the knee, until the onset of resistance. The range of knee ﬂexion at this point was measured relative to zero. This was repeated three times with the average range recorded. The test was repeated for the other neck position and limb. At all times, the subject was monitored for the presence of adverse symptoms (pins and needles and sensation changes) during testing and

2.1.2. SLR test The subject lay supine on the treatment couch with the trunk, shoulders and hips in a neutral position, with the head kept ﬂat. The goniometer was positioned with the stationary arm aligned between the greater trochanter and the mid-line of the trunk, and the moving arm aligned between the greater trochanter and the lateral condyle of the femur (Palmar and Epler, 1998). This position was taken as zero. The leg to be examined was ﬁxed into knee extension by the examiner’s hand placed on the thigh just proximal to the knee joint. The opposite hand either supported the subject’s foot and ankle under the heel allowing it to assume a neutral position, or held the ankle in a position of maximal dorsi-ﬂexion, using the hand hold recommended by Butler (1991). A neutral foot position was assumed to be approximately 101 plantar-ﬂexion with no inversion or eversion. The leg was then passively elevated in a sagittal plane, moving at a ﬁxed point about the hip joint, until the onset of resistance hip ﬂexion angle was then measured. This was repeated three times with the average range recorded. The test was repeated for the other foot position and limb. At all times, the subject was monitored for the presence of adverse symptoms (pins and needles and sensation changes) during testing and presence of pain in excess of a mild stretching sensation, if this occurred during testing the test was terminated. All subjects were able to fully participate using these criteria, without the generation of symptoms and with none requesting withdrawal. The order of testing either slump or SLR test ﬁrst was block allocated, with alternated subjects starting with slump test followed by SLR. Similarly, the limb examined was block allocated, here left leg with slump ﬁrst and then left with SLR ﬁrst and then right leg tested ﬁrst on the next pair of tests. 2.1.3. Intra-tester reliability Ten subjects were measured again on a separate occasion at the same time the following day, for both the slump and SLR tests. The previous block allocations i.e. test and limb order was reversed for each subject. 2.1.4. Analysis All data were analysed on the statistical computer package SPSS (version 11). The measurements from the data collected were converted to values representative of a change in the range of movement (degrees), due to the addition of the structural differentiation manoeuvre.

ARTICLE IN PRESS L. Herrington et al. / Manual Therapy 13 (2008) 289–294

The effect of the addition of the structural differentiation manoeuvre on knee ﬂexion angle (slump test) and hip ﬂexion angle (SLR test) was analysed using repeated measures t-tests. The a level was set at a ¼ 0.05. Intratester reliability was assessed using intra-class correlation co-efﬁcient calculated from the repeated measurements of 10 subjects for both the slump test and SLR. 3. Results 3.1. Intra-tester reliability SLR test intra-tester reliability was r ¼ 0.93 (po0.01), standard error of measurement (SEM) of 2.51. Slump test intra-tester reliability r ¼ 0.88 (po0.01) with a SEM 1.81.

SLR Ankle Neutral SLR Ankle dorsiflexed

60 50 40 30 20 10 0 Male

Female

Fig. 3. Average hip ﬂexion angles during SLR test.

3.2. Slump test In both male (p40.05) and female (p40.05) subjects no difference in the effect of structural differentiation manoeuvre was found between limbs. Both male and female subjects revealed an increase in range of knee ﬂexion (01 ¼ full knee extension) following the addition of structural differentiation (cervical ﬂexion) in the slump test (Fig. 2). There was a signiﬁcant difference in the amount of change in knee ﬂexion angle between male and female subjects (po0.05), average increase in knee ﬂexion angles for male subjects was 6.674.71 and for female subjects 5.475.81. For both male and female subjects, the addition of cervical ﬂexion brought about a statistically signiﬁcant increase in knee ﬂexion angle (po0.01). 3.3. Straight leg raise test In both male (p40.05) and female (p40.05) subjects, no difference in the effect of structural differentiation manoeuvre (ankle dorsi-ﬂexion) was found between 60 SLUMP Neck Flexed SLUMP Neck Extended

Knee Flexion angle (Degrees)

70

Hip Flexion Angle (Degrees)

292

50 40 30 20 10 0 Male

Female

Fig. 2. Average knee ﬂexion angles during the slump test.

limbs. Both male and female subjects revealed a diminished range of hip ﬂexion following the addition of structural differentiation (ankle dorsi-ﬂexion) in the SLR test (Fig. 3). There was a signiﬁcant difference in the amount of change in hip ﬂexion angle between male and female subjects (po0.01), average decrease in hip ﬂexion angles for male subjects was 9.578.31 and for female subjects 15.279.51. For both male and female subjects, the addition of ankle dorsi-ﬂexion brought about a statistically signiﬁcant decrease in hip ﬂexion angle (po0.01). 4. Discussion The principal aim of the study was to establish in asymptomatic subjects the response to the neurodynamic tests of SLR and slump of the addition of structural differentiating manoeuvres. Secondary to this, was to investigate if male and female subjects respond differently to the tests. In the group of asymptomatic subjects examined, all could be regarded as having a positive neurogenic response to the addition of the structural differentiating manoeuvre (Butler, 2000; Shacklock, 2005b). The level of normal response to neurodynamic testing has not previously been reported in the literature, though it has been implied that the normal response to these tests is that some subjects will not have a response indicative of neural tissue sensitivity (Butler, 2000). The ﬁndings of this study would imply that all subjects will show a response to neurodynamic testing indicative of a degree of sensitivity of neural tissue; this though could be regarded as a normal neurogenic response to the test (Shacklock, 2005b). This was deﬁned by Shacklock (2005a) as a symmetrical response which produces neither overt nor covert symptom reproduction, as was the case here with no reproduction of symptoms or asymmetry. It is important to understand the level of normal neurodynamic

ARTICLE IN PRESS L. Herrington et al. / Manual Therapy 13 (2008) 289–294

response to testing in order to avoid the generation of false positive test results. The use of this deﬁnition for the ﬁndings is further supported because of the symmetry between limbs found during both tests. The difference between limbs within subjects on application of structural differentiating manoeuvres for both tests and genders was not a statistically signiﬁcant one (p40.05). The question of symmetry of response has not previously been investigated; this makes comparison of ﬁndings impossible, though it has been implied that symmetry should occur and asymmetry should be regarded as indicative of a problem (Maitland, 1985; Butler, 2000). It should be noted though that there was not a statistically signiﬁcant difference between limbs, this is not the same as the response between limbs being identical. In fact, out of the 88 subjects only 10 subjects had a truly identical response to either test, the rest showed differences in response between limbs which though present were not of signiﬁcant enough magnitude to be regarded as statistically different, that is they did not vary signiﬁcantly from the group mean difference between limbs. The actual mean values for difference between limbs were 2.71 and 21 for the slump and SLR tests, respectively. These values as they are close to the SEM for the measurement are in all probability just as likely to be related to measurement error as actual differences between limbs. This again may prove clinically important with non-symmetrical responses to tests only being relevant if they exceed the reported ranges of test response. In the group that studied the normal neurogenic response to the addition of the structural differentiating manoeuvre of cervical ﬂexion to the slump test is to result in an 6.674.71 and 5.475.81 increase in knee ﬂexion angle of male and female subjects, respectively. In the SLR test, the addition of ankle dorsi-ﬂexion brought about a 9.578.31 and 15.279.51 decrease in hip ﬂexion angle in males and females, respectively. Previous literature has reported between 21 and 61 changes in knee ﬂexion angle during the slump test with the addition of cervical ﬂexion (Johnson and Chiarello, 1997; Yeung et al., 1997; Coppieters et al., 2005), despite the methodological differences this is within the range reported in the present study 5.4–6.61. Indicating this level of normal neurogenic response is consistent across a number of studies for the slump test. From the two studies, the range of change in hip ﬂexion angle during SLR test with the addition of ankle dorsi-ﬂexion was 8.8–101 (Gajdosik et al., 1985; Boland and Adams, 2000), the present study found a range of 9.5–15.21 which is outside that of the other studies. When the 181 standard deviation of the Boland and Adams (2000) is taken into account, the ranges of the present studies results are very similar to those previously reported. These ﬁndings have some signiﬁcance to the clinician wishing to avoid false positive ﬁndings from these tests.

293

Firstly, it would appear that the majority (if not all) normal asymptomatic subjects have a positive neurogenic response to structural differential of these tests. Therefore, the ﬁnding of positive structural differentiation does not necessarily imply the present of neural pathology. Secondly, for a test to be regarded as abnormal the change brought about by addition of structural differentiating manoeuvres should be greater than those reported by the present and previous studies and also asymmetrical, if this is relevant to the underlying problem. There were signiﬁcant differences in the nature of the test response between male and female subjects for both SLR (po0.01) and slump (po0.05) tests. This difference in test response between the sexes has not been previously reported; previous studies have either reported on a single sex or not reported on any gender differences, merging the data from both sexes. 5. Conclusion Structural differentiating manoeuvres have a signiﬁcant effect on neurodynamic test response in terms of range of movement even in normal asymptomatic individuals. These normal neurogenic responses to lower quadrant neurodynamic testing should be taken into account during the assessment clinical reasoning process, to avoid the development of management plans based on a false positive test result. References Barker P, Briggs C. Attachment of the posterior layer of lumbar fascia. Spine 1999;24:1757–64. Boland R, Adams R. Effects of ankle dorsiﬂexion on range and reliability of straight leg raising. Australian Journal of Physiotherapy 2000;46:191–200. Butler D. Mobilisation of the nervous system. Oxford: Churchill Livingstone; 1991. Butler D. The sensitive nervous system. Australia: Noigroup Publications; 2000. Cleland J, Childs J, Palmer J, Eberhart S. Slump stretching in the management of non-radicular low back pain: a pilot clinical trial. Manual Therapy 2006, in press. Coppieters P, Kurz K, Mortensen T, Richards N, Skaret I, McLaughlin L, et al. The impact of neurodynamic testing on the perception of experimentally induced muscle pain. Manual Therapy 2005;10:52–60. Di Fabio R. Neural mobilisation: the impossible. Journal of Orthopaedic and Sports Physical Therapy 2001;3:224–5. Gajdosik R, LeVeau F, Bohannon R. Effects of ankle dorsiﬂexion on active and passive unilateral straight leg raising. Physical Therapy 1985;65:1478–82. Herrington L. What is a normal neurogenic response to the slump test? In: Eighth international scientiﬁc conference of the international federation of orthopaedic manipulative therapists. Cape Town, 2004. Johnson E, Chiarello C. The slump test: the effects of head and lower extremity position on knee extension. Journal of Orthopaedic and Sports Physical Therapy 1997;26:310–7.

ARTICLE IN PRESS 294

L. Herrington et al. / Manual Therapy 13 (2008) 289–294

Lew P, Briggs C. Relationship between the cervical component of the slump test and change in hamstring muscle tension. Manual Therapy 1997;2:98–105. Maitland G. The slump test: examination and treatment. Australian Journal of Physiotherapy 1985;31:215–9. Palmar M, Epler M. Fundamentals of musculoskeletal assesment techniques. 2nd ed. Philadelphia: Lippincott Williams and Wilkins; 1998. Shacklock M. Improving application of neurodynamic (neural tension) testing and treatments: a message to researchers and clinicians. Manual Therapy 2005;10:175–9.

Shacklock M. Clinical neurodynamics: a new system of musculoskeletal treatment. Oxford: Elsevier; 2005. Vleeming A, Pool-Goudzwaard A, Stoeckart R, va Wingerden J, Snijders C. The posterior layer of the thoraco-lumbar fascia. Its function in load transfer from spine to legs. Spine 1995;20: 753–8. Yeung E, Jones M, Hall B. The response to the slump test in a group of female whiplash patients. Australian Physiotherapy 1997;43: 245–52.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 295–299 www.elsevier.com/locate/math

Original article

Prevalence of pain and dysfunction in the cervical and thoracic spine in persons with and without lateral elbow pain K.M. Berglund1, B.H. Persson, E. Denison Department of Public Health and Caring Sciences, Section of Caring Sciences, Uppsala University, Uppsala Science Park, S-751 83 Uppsala, Sweden Received 4 July 2006; received in revised form 28 December 2006; accepted 15 January 2007

Abstract The purpose of this study was to survey the prevalence of pain in the cervical and thoracic spine (C2–T7) in persons with and without lateral elbow pain. Thirty-one subjects with lateral elbow pain and 31 healthy controls participated in the study. The assessment comprised a pain drawing, provocation tests of the cervical and thoracic spine, a neurodynamic test of the radial nerve, and active cervical range of motion. Seventy percent of the subjects with lateral elbow pain indicated pain in the cervical or thoracic spine, as compared to 16% in the control group (po0.001). The frequency of pain responses to the provocation tests of the cervical and thoracic spine was signiﬁcantly higher (po0.05) in the lateral elbow pain (LEP) group, as was the frequency of pain responses to the neurodynamic test of the radial nerve (po0.001). Cervical ﬂexion and extension range of motion was signiﬁcantly lower (po0.01) in the LEP group. The results indicate a relation between lateral elbow pain and pain in the vertebral spine (C2–T7). The cervical and thoracic spine should be included in the assessment of patients with lateral elbow pain. r 2007 Elsevier Ltd. All rights reserved. Keywords: Lateral epicondylalgia; Manual therapy; Pain; Tennis elbow; Cervical and thoracic spine

1. Introduction Lateral elbow pain is a common disorder, which has been described in terms such as tennis elbow, epicondylitis, and lateral epicondylalgia. In this paper, the term lateral elbow pain is used. According to Vicenzino and Wright (1996), the prevalence of lateral elbow pain was found to be 3% in a random population sample, and most frequent among persons in their mid-40s. There were no gender differences, although symptoms seemed more pronounced and chronic in females. Relapse rates ranged between 33% and 50% in an 18-month follow-up period. The relative simplicity of the clinical presentaCorresponding author. Tel.: +46 21 103 138; fax: +46 21 101 633.

E-mail address: [email protected] (E. Denison). This study was completed in the fulﬁlment of a Master of Science degree at the Department of Physiotherapy, the Karolinska Institute, Stockholm, Sweden. 1

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.01.015

tion belies the complexity of the underlying pathophysiological and aetiological processes, and the mechanism of pain production in chronic lateral elbow pain is probably multifactorial, involving local pathophysiological mechanisms as well as nociceptive system mechanisms (Vicenzino and Wright, 1996). Among the latter, somatic pain referral from the cervical and thoracic spine, and radial nerve involvement have been proposed as possible factors (Vicenzino and Wright, 1996). Painful disorders in the cervical and thoracic spine have been suggested by several authors as common causes of referred pain in the lateral elbow area (Iselin, 1977; Lee, 1986; McGuckin, 1986; Widenfalk et al., 1988; Butler, 1991; Haker, 1993; Noteboom et al., 1994; DeFranca and Levine, 1995). In a similar vein, Wright et al. (1994) suggested that clinical signs of lateral elbow pain such as tenderness, pain spreading beyond its primary location, guarding of the pain area, and changes in skin temperature may be explained by central sensitisation processes that could have arisen from structures in the

ARTICLE IN PRESS 296

K.M. Berglund et al. / Manual Therapy 13 (2008) 295–299

lower cervical spine. Some uncontrolled studies have shown decreased pain intensity in the elbow when manual treatment was given to the cervical spine for patients with lateral elbow pain (Gunn and Mildbrandt, 1976; Maigne, 1988). In the studies by both Gunn and Mildbrandt (1976) and Maigne (1988), the participants had all obtained earlier treatment directed to the elbow without effect. The implication of the results in these two studies is that there may be a cervical factor in lateral elbow pain. However, the absence of a control group limits the conclusions that can be drawn from the results. Vicenzino et al. (1998) showed in a randomised and controlled study that manipulative therapy of the cervical spine produced hypoalgesic and sympathoexcitatory changes that were signiﬁcantly greater than placebo and control treatments. Vicenzino et al. (1996) demonstrated increased pain-free strength and increased shoulder abduction during responses to a neurodynamic test of the radial nerve after manual treatment at the C5–C6 level. Clinical evidence of radial nerve involvement in lateral elbow pain was presented by Yaxley and Jull (1993), who found signs of less extensible neural tissue in the arm with unilateral lateral elbow pain, as compared to the unaffected arm during a neurodynamic test of the radial nerve in subjects suffering from unilateral elbow pain. However, a possible confounding factor in referred pain is that degenerative changes in the cervical spine is very common in asymptomatic subjects, the C5–C6 segment showing the highest frequency of all MRI ﬁndings (Matsumoto et al., 1998). Because none of the studies referred to above used pain-free subjects as controls, it is not known whether signs of disorders in the cervical spine are present to a greater extent in persons with lateral elbow pain than in persons without lateral elbow pain. In sum, there is some support for both painful disorders in the cervical and thoracic spine and radial nerve involvement as aetiological factors in lateral elbow pain. Because comparison of symptoms and signs of spinal disorders such as pain intensity, restricted cervical range of motion, positive responses to pain provocation tests of the cervical and upper thoracic spine, or radial nerve involvement with subjects without lateral elbow pain is lacking in the literature, no ﬁrm conclusions can be drawn regarding the role of these disorders in the aetiology of lateral elbow pain. Thus, the purpose of the present study was to (1) survey the prevalence of self-reported pain in the cervical and thoracic spine (C2–T7) in persons with and without lateral elbow pain, and (2) compare persons with and without lateral elbow pain regarding (a) responses to several pain provocation tests in the cervical and thoracic spine, (b) responses to a neurodynamic test of the radial nerve, and (c) cervical active range of motion (ROM).

2. Methods 2.1. Design The study was descriptive and comparative, involving a group of persons with ongoing lateral elbow pain (LEP) and a pain-free control (C) group. 2.2. Subjects and setting All subjects were recruited from a large paper mill in the south of Sweden. The criterion for being included in the LEP group was lateral elbow pain for at least 6 weeks prior to the study, in order to avoid subjects with acute and possibly selflimiting pain. Exclusion criteria were ongoing treatment due to lateral elbow pain or treatment within the last month, any known systemic disease affecting joints and/or muscles, or lateral elbow pain caused by external trauma. Subjects in the C group were required to have had no lateral elbow pain for at least 6 months prior to the study, and no known systemic disease affecting joints or muscles. The subjects in the LEP group were recruited from the company’s occupational healthcare centre. The subjects in the C group were recruited by advertising within the company and those who volunteered were selected consecutively so that eventually there were equally many males and females in each group, making it less likely that gender would be a confounding variable. There were 31 subjects in each group, 23 males and 8 females. The mean age was 49 years in the LEP group and 47 years in the C group. The mean duration of elbow pain in the LEP group was 36 months. Twentythree of the subjects in the LEP group had received treatment directed to the elbow. All subjects completed the study. 2.3. Measures The tests used in his study were selected because of their frequent use in clinical assessment of patients with neck disorders. They are well described in the literature and are simple to perform. The purpose of the tests was to record painful responses only (Sandmark and Nisell, 1995). The sensitivity (the proportion of subjects reporting neck/thoracic pain who had a positive outcome of the test) and speciﬁcity (the proportion of subjects without neck/thoracic pain who had a negative outcome of the test) of the tests for cervical and thoracic pain of the radial nerve were assessed as described by Sandmark and Nisell (1995) by setting up 2 2 contingency tables and calculating the proportions of true positives and true negatives. Pain location was recorded with a pain drawing where subjects were asked to indicate the area/s where they felt

ARTICLE IN PRESS K.M. Berglund et al. / Manual Therapy 13 (2008) 295–299

pain. Pain drawings have been reported to be reliable and valid measures of pain location (Jensen and Karoly, 1992). Cervical active ROM was measured with a Myrin goniometer (Hagen et al., 1997), and comprised ﬂexion, extension, lateral ﬂexion, and rotation, using the procedure described by these authors. Hagen et al. (1997) reported test–retest reliabilities ranging between r ¼ 0.75 and 0.93. A neurodynamic test of the radial nerve was conducted in conformity with Butler (1991). The subjects were examined in the supine position. A positive response was recorded if a subject reported pain in the forearm at less than 401 of shoulder abduction. Yaxley and Jull (1991) reported that no signiﬁcant variation between two raters was found, and that there was no signiﬁcant difference between trials in a repeatability test. Spinal cervical pain was assessed by two tests for the cervical spine (C2–C7). Firstly, palpation of the nerve trunk just beyond its exit from the vertebral foramen was performed with subjects in the supine position, according to Cyriax (1982). Secondly, compression of the vertebral foramina was performed with subjects in the sitting position. The subject’s head was placed passively in dorsal ﬂexion, rotated, and laterally ﬂexed to the same side, according to Kaltenborn (1989). Responses to the tests were recorded as no pain, local pain, or pain referred to the elbow. Strender et al. (1997) reported an inter-rater agreement of 76% and k ¼ 0.43 for the foramen compression test. Sandmark and Nisell (1995) reported a sensitivity of 77% and a speciﬁcity of 92% for the foramen compression test. Spinal thoracic pain (T1–T7) was assessed by the Springing test, which was performed according to Kaltenborn (1989). The subjects were examined in the prone position. Responses to the test were recorded as no pain, local pain, or pain referred to the elbow. All tests were performed bilaterally.

2.4. Procedure Two physiotherapists with long clinical experience and formal education in manual therapy made the assessments. Before the assessments took place, the ﬁrst physiotherapist (KB) gave code numbers to each subject. The subjects chose appointment times from a list provided by the ﬁrst physiotherapist. In this way, the sequence of subjects was not inﬂuenced by any of the physiotherapists. The sequence of subjects was not available to the second physiotherapist (HP), who was thus blinded to group afﬁliation. Pain drawings were collected by the ﬁrst physiotherapist. Then, in a separate room, the second physiotherapist ﬁrst measured active cervical ROM. Then, the neurodynamic test of the radial nerve was performed.

297

Finally, the tests for spinal pain were performed. The assessments were completed in about 50 min. The local ethics committee approved the project. 2.5. Data analysis Positive responses to the tests (local pain or pain referred to the elbow) were collapsed into one category. Statistical analyses were performed in StatisticaTM for Macintosh. Differences between groups based on frequencies were analysed with w2 tests. Differences between groups based on means and standard deviations of cervical ROM were analysed with unpaired t-tests. The level of signiﬁcance was set at po0.05, two-tailed.

3. Results Twenty-two subjects (70%) in the LEP group and 5 subjects (16%) in the C group indicated pain in the cervical and/or thoracic areas as measured by the pain drawing (w2 ¼ 18.96, po0.001). There were signiﬁcant differences between the groups in the frequency of positive responses to both tests of cervical pain, and the Springing test (Table 1). The sensitivity and speciﬁcity for the nerve trunk palpation test were 40% and 68%, respectively, and for the foramen compression test the corresponding ﬁgures were 67% and 68%. The sensitivity and speciﬁcity for the Springing test were 62% and 85%, respectively. Eighteen subjects (58%) in the LEP group and 4 subjects (13%) in the C group reported pain in the forearm during the neurodynamic test of the radial nerve (w2 ¼ 14.84, po0.001). There were signiﬁcant differences between the groups in cervical ﬂexion and cervical extension, but not in any of the other measured directions (Table 2).

4. Discussion The results of this study show that 70% of the subjects in the LEP group indicated pain in the cervical Table 1 The frequency of painful responses to the tests for cervical and thoracic pain

Nerve trunk palpation test Compression of cervical vertebral foramina Springing-test of T1–T7

Elbow pain group (n ¼ 31)

Comparison group (n ¼ 31)

p

15 17

6 8

o0.05 o0.05

14

5

o0.05

ARTICLE IN PRESS K.M. Berglund et al. / Manual Therapy 13 (2008) 295–299

298

Table 2 Means (M) and standard deviations (SD) for active cervical ROM in the two groups

Cervical ﬂexion Cervical extension Right lateral ﬂexion Left lateral ﬂexion Right rotation Left rotation

Elbow pain group (n ¼ 31)

Comparison group (n ¼ 31)

M

SD

M

SD

60.2 62.4 38.1 39.7 68.6 69.7

11.1 11.9 8.9 8.3 13.6 12.0

66.9 70.6 41.0 40.5 73.7 74.4

8.6 9.3 7.6 7.6 9.7 11.0

p

o0.01 o0.01 NS NS NS NS

and/or thoracic spine, as compared to 16% in the C group. The frequency of positive responses to provocation tests of the cervical and thoracic spine was signiﬁcantly higher in the LEP group, as well as the frequency of positive responses to a neurodynamic test of the radial nerve. Finally, active cervical ﬂexion and extension ROM were signiﬁcantly lower. Since the two groups were comparable in age, degenerative and agerelated changes are not likely explanations of these differences. The 1-year prevalence of neck pain in a normal population has been estimated to about 43% (Bovim et al., 1994), which is clearly lower than the 70% in the LEP group of this study. The main difference between the groups was noted for the neurodynamic test of the radial nerve. Positive responses to this test have been shown in previous studies regarding lateral elbow pain that did not include a pain-free control group (Yaxley and Jull, 1993; Wright et al., 1994; Vicenzino and Wright, 1996). Yaxley and Jull (1993) recommended inclusion of tests of neural structures as routine for patients with lateral elbow pain. Although the neurodynamic test does not reveal the origin of symptoms (Moses and Carman, 1996), the results of the present, controlled study support the ﬁndings reported in previous studies (Yaxley and Jull, 1993; Wright et al., 1994; Vicenzino and Wright, 1996). The frequency of positive responses to the provocation tests of the spine (C2–T7) differed signiﬁcantly as well. Again, these results support ﬁndings of cervical and/or thoracic disorders in patients with lateral elbow pain reported in previous uncontrolled studies (Gunn and Mildbrandt, 1976; Maigne, 1988; Haker, 1993; Vicenzino et al., 1998). Wright et al. (1994) proposed that the nociceptive trigger activating the process of central sensitisation in patients with lateral elbow pain could have arisen from structures within the lower cervical spine. The ﬁndings in the present study support the presence of nociceptive processes in the cervical spine in patients with lateral elbow pain. However, because data regarding signs of central sensitisation

were not obtained, no conclusions can be drawn concerning relations between cervical pain and signs of central sensitisation. The active cervical ﬂexion and extension were signiﬁcantly lower in the LEP group, which is in line with the ﬁndings of Vicenzino and Wright (1996), who noted that 90% of subjects participating in studies of lateral elbow pain had segmental hypomobility in the lower cervical spine. Relapse is common in patients with lateral elbow pain, and Vicenzino and Wright (1996) reported relapse rates of 33–50%. In the present study, 23 of the 31 subjects (74%) in the LEP group had been given earlier treatment directed to the elbow. As for the interpretation of the results in this study, some important limitations should be noticed. Firstly, the sample was self-selected, which means that the population of patients with lateral elbow pain may not be well represented. However, the results reported here are in good accordance with previously reported results. Secondly, there is a lack of published data regarding the reliability and validity of the nerve trunk palpation test and the provocation test of the thoracic spine (the Springing test) that were used in the present study. However, the tests were used only to record painful responses, not hypo- or hypermobility (the Springing test). Sandmark and Nisell (1995) argued that pain is the most relevant parameter in clinical studies of musculoskeletal problems, not hypo- or hypermobility, which is not so uniquely tied to a syndrome that it could serve as a standard by which a condition is judged to be present or absent. In the present study, sensitivity and speciﬁcity were used to indicate the validity of the three provocation tests for the cervical and thoracic spine. While speciﬁcity was fair in all three tests, the nerve trunk palpation test showed low sensitivity, which means that the test may yield false negatives. Thus, some caution is warranted concerning the nerve trunk palpation test.

5. Conclusion The results of this controlled study show that pain in the cervical and/or thoracic spine was more common in subjects with lateral elbow pain than in healthy controls, and that the frequency of positive responses to provocation tests of the cervical and thoracic spine was signiﬁcantly higher. The frequency of positive responses to a neurodynamic test of the radial nerve was signiﬁcantly higher and active cervical ﬂexion and extension was signiﬁcantly lower. The results support ﬁndings in previous studies that did not include a painfree control group and suggest that the cervical and thoracic spine should be included in the assessment of patients with lateral elbow pain.

ARTICLE IN PRESS K.M. Berglund et al. / Manual Therapy 13 (2008) 295–299

Acknowledgements The realisation of this study was possible thanks to the kind co-operation of Stora Billerud AB, Division Paper, Sweden.

References Bovim G, Schrader H, Sand T. Neck pain in the general population. Spine 1994;19(12):1307–9. Butler DS. Mobilisation of the nervous system. London: Churchill Livingstone; 1991. p. 242–3 [Chapter 13]. Cyriax JH. Textbook of orthopaedic medicine, 8th ed. Scotland: Tomphson Litho Ltd; 1982. p. 92–6 [Chapter 7]. DeFranca DC, Levine LJ. The T4 syndrome. Journal of Manipulative and Physiologic Therapeutics 1995;18(1):34–7. Gunn CC, Mildbrandt WE. Tennis elbow and the cervical spine. Canadian Medical Association Journal 1976;114(8):803–9. Hagen KB, Harms-Ringdahl K, Enger NO, Hedestad R, Morten H. Relationship between subjective neck disorders and cervical spine mobility and motion-related pain in male machine operators. Spine 1997;22(13):1501–7. Haker E. Lateral epicondylalgia: diagnosis, treatment, and evaluation. Critical Reviews in Physical and Rehabilitation Medicine 1993; 5(2):129–54. Iselin DG. Inﬂuence de la colonne vertebrale sur l’epicondylite. Therapeutische Umschau 1977;34(2):88–91. Jensen MP, Karoly P. Self-report scales and procedures for assessing pain in adults. In: Turk DC, Melzack R, editors. Handbook of pain assessment. New York: Guilford; 1992 p. 135–51 [Chapter 9]. Kaltenborn F. Manuell mobilisering av ryggraden (Manual mobilisation of the spine). Oslo: Olaf Norli Bokhandel; 1989. p. 184–260 [Chapters 3–4]. Lee DG. Tennis elbow: a manual therapist’s perspective. Journal of Orthopaedic and Sports Physiotherapy 1986;8(3):134–42. Maigne R. Behandlung der epicondylitis durch manipulation: zervicaler und articula¨rer faktoren. Manuelle Medizin 1988;26: 69–72.

299

Matsumoto M, Fujimura Y, Nobumasa S, Nishi Y, Nakamura M, Yabe Y, et al. MRI of cervical intervertebral discs in asymptomatic subjects. Journal of Bone and Joint Surgery 1998;80-B(1):19–24. McGuckin N. The Th4 syndrome. In: Grieve GP, editor. Modern manual therapy of the vertebral column. Edinburgh: ChurchillLivingstone; 1986 pp. 370–6 [Chapter 36]. Moses A, Carman J. Anatomy of the cervical spine: implications for the upper limb tension test. Australian Journal of Physiotherapy 1996;42(1):31–5. Noteboom T, Curver R, Keller J, Kellog B, Nitz AJ. Tennis elbow: a review. Journal of Orthopaedic and Sports Physiotherapy 1994; 19(6):357–66. Sandmark H, Nisell J. Validity of ﬁve common manual neck pain provoking tests. Scandinavian Journal of Rehabilitation Medicine 1995;27:131–6. Strender L-E, Lundin M, Nell K. Interexaminer reliability in physical examination of the neck. Journal of Manipulative and Physiologic Therapeutics 1997;20(8):516–20. Vicenzino B, Wright A. Lateral epicondylalgia I: epidemiology, pathophysiology, aetiology, and natural history. Physiotherapy Review 1996;1:23–34. Vicenzino B, Collins D, Wright A. The initial effects of a cervical spine manipulative physiotherapy treatment on pain and dysfunction of the lateral epicondylalgia. Pain 1996;68:69–74. Vicenzino B, Collins D, Benson H, Wright A. An investigation of the interrelationship between manipulative therapy-induced hyperalgesia and sympathoecitation. Journal of Manipulative and Physiological Therapeutics 1998;21(7):448–53. Widenfalk B, Elfvin L, Wiberg M. Origin of the sympathetic and sensory innervation of the elbow in a rat. A retrograde axonal actin study with wet germ agglutinin conjugated horseradish preoxidase. Journal of Comparative Neurology 1988;217:313–8. Wright A, Thurnwald P, O’Callaghan J, Smith J, Vicenzino B. Hyperalgesia in tennis elbow patients. Journal of Musculoskeletal Pain 1994;2(4):83–97. Yaxley G, Jull G. The modiﬁed upper limb tension test: an investigation of responses in normal subjects. Australian Journal of Physiotherapy 1991;37(3):143–52. Yaxley G, Jull G. Adverse tension in the neural system. A preliminary study of the tennis elbow. Australian Journal of Physiotherapy 1993;39(1):15–22.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 300–306 www.elsevier.com/locate/math

Original article

Lower lumbar spine axial rotation is reduced in end-range sagittal postures when compared to a neutral spine posture Angus Burnett, Peter O’Sullivan, Lars Ankarberg, Megan Gooding, Rogier Nelis, Frank Offermann, Jannike Persson School of Physiotherapy, Curtin University of Technology, GPO Box U1987, Perth, Western Australia 6845, Australia Received 18 August 2006; received in revised form 28 December 2006; accepted 23 January 2007

Abstract Sports such as rowing, gymnastics, cycling and fast bowling in cricket that combine rotation with spine ﬂexion and extension are known to carry greater risk of low back pain (LBP). Few studies have investigated the capacity of the lumbar spine to rotate in various sagittal positions, and further, these studies have generated disparate conclusions. The purpose of this study was to determine whether the range of lower lumbar axial rotation (L3–S2) is decreased in end-range ﬂexion and extension postures when compared to the neutral spine posture. Eighteen adolescent female rowers (mean age ¼ 14.9 years) with no history of LBP were recruited for this study. Lower lumbar axial rotation was measured by an electromagnetic tracking system (3-Space FastrakTM) in end-range ﬂexion, extension and neutral postures, in sitting and standing positions. There was a reduction in the range of lower lumbar axial rotation in both end-range extension and ﬂexion (po0.001) postures when compared to neutral. Further, the range of lower lumbar axial rotation measurements in ﬂexion when sitting was reduced when compared to standing (p ¼ 0.013). These ﬁndings are likely due to the anatomical limitations of the passive structures in end-range sagittal postures. r 2007 Elsevier Ltd. All rights reserved. Keywords: In vivo; Lumbar spine; Axial rotation; Sagittal posture

1. Introduction A number of studies have shown an increased prevalence of low back pain (LBP) (predominantly at the lower lumbar spine) amongst adolescent athletes in sports such as rowing, gymnastics, cycling and fast bowling in cricket. These sports involve high mechanical spinal loading in association with coupled ﬂexion/ extension and axial rotation of the lumbar spine and medium to high volumes of training and competition (Tertti et al., 1990; Burnett et al., 1996, 2004; Balague´ et al., 1999). Recently, Perich et al. (2006) examined a large group of adolescent female rowers (N ¼ 356) and found the Corresponding author. Tel.: +61 8 9266 3662; fax: +61 8 9266 3699. E-mail address: [email protected] (A. Burnett).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.01.016

LBP point-prevalence to be 47.5%. In comparison an age, height, weight, socio-economic and physical activity matched non-rowing control group had LBP pointprevalence of 15.5%. In the rowers group, 64% of subjects reported that rowing in a sweep eight (rowing on one side of the body including spinal rotation) brought on or exacerbated their pain, whilst rowing in a single scull (14%) or quadruple scull (37%) (rowing on both sides of the body) pain was less common. Similarly, Burnett et al. (2004) reported that cyclists with chronic low back pain (chronic LBP) had a trend towards greater ﬂexion/rotation of the lower lumbar spine during the crank cycle whilst riding seated on a wind trainer when compared to cyclists without LBP. At the other extreme, fast bowlers in cricket whose spines are exposed to high volumes of extension/side bending/ rotational stress experience greater low back pain (LBP) and a higher risk of pathological spinal changes than

ARTICLE IN PRESS A. Burnett et al. / Manual Therapy 13 (2008) 300–306

non-bowlers (Burnett et al., 1996, 1998; Ranson et al., 2005). These data are strongly suggestive of ﬂexion or extension loading coupled with rotation and or side bending as a dominant factor in the aetiology and exacerbation of LBP in these speciﬁc populations. It has been proposed that the risk of tissue strain is increased at end-range of spinal motion where the passive spinal structures are maximally loaded (Panjabi, 1992a, b). The addition of rotation to a spinal segment that is already fully ﬂexed or extended may result in increased tissue loading as passive spinal structures (bone, ligament and disc) may be limiting movement. In contrast, when the spine is loaded or rotated within a more neutral position of the motion segment, there may be more compliance within the passive spinal structures (Panjabi, 1992a, b) therefore, reducing the risk of tissue strain. These considerations are consistent with the concept of neutral spine control, which is considered to be important in minimising spinal tissue strain. Although spinal rotation represents a known risk factor for spinal injury, few studies have investigated the biomechanics of lumbar axial rotation in sagittal postures. Previous in vivo studies have reported reduced lumbar axial rotation in forward ﬂexion when compared to upright sitting and standing (Gunzburg et al., 1991). Previous in vitro studies have reported reduced lumbar axial rotation in extension compared to neutral (Haberl et al., 2004) and a trend towards reduced axial rotation in ﬂexion (Gunzburg et al., 1991; Haberl et al., 2004). In these studies it was not reported where in the sagittal plane (relative to end range), lumbar axial rotation was measured. To our knowledge, no study has yet investigated the magnitude of lumbar axial rotation available in vivo, in neutral when compared to end-range ﬂexed and extended postures. Mid-range and end-range ﬂexed and extended postures are typical in sporting activities where LBP is common (Caldwell et al., 2003; Burnett et al., 2004; Perich et al., 2006; Ranson et al., 2005, 2007). Therefore, the purpose of this study was to determine whether the range of lower lumbar axial rotation differed in end-range ﬂexed and extended postures when compared to a neutral spine posture in sitting and standing positions. This research question was examined in a group of adolescent female rowers.

dent Girls Schools’ Sports Association in Western Australia. Potential subjects were initially identiﬁed by the Head of the Physical Education department in each participating school and those with no LBP were randomly selected and invited to undertake further testing. Subjects were excluded from the study if they had ever experienced LBP whilst rowing or if they experienced pain or discomfort whilst undertaking the experimental protocol. Approval to conduct the study was obtained from the Institutional Human Research Ethics Committee and parents/guardians were required to provide their informed consent. 2.2. Experimental protocol The experimental protocol involved subjects moving into maximal active lumbar axial rotation and holding the position for three seconds, in different combinations of sagittal spinal posture (neutral, end-range ﬂexion and end-range extension) in both standing and sitting. Two trials were completed in left and right rotation, with a total of 24 trials (6 positions/postures 2 sides 2 trials per side) being completed by each subject. The order of the different testing positions was randomised and the total protocol took 30 min. Prior to the testing protocol, each subject’s sagittal end-range ﬂexion and extension position in both sitting and standing was measured. This was done for two reasons; ﬁrstly, to ensure that sagittal plane end-range was actually achieved before and during the axial rotation measurements and secondly, to assist in setting the neutral postures in sitting and standing which were considered to be the mid-point between end-range ﬂexion and extension. Two experienced physiotherapists guided each movement in the protocol. In all trials, the lower lumbar spine position was measured by an electromagnetic tracking system as detailed below. During data collection, each position (standing and sitting) and sagittal posture (neutral, end-range extension and end-range ﬂexion) were standardised as follows:

2. Methods 2.1. Subjects This study utilised a cross-sectional repeated measures design, with 18 asymptomatic adolescent female rowers (mean7SD, age 14.970.9 years, mass 58.579.7 kg, height 1.6970.08 m) recruited from a total population of approximately 400 female rowers from the Indepen-

301

Position—Standing: Subjects stood with their feet parallel, shoulder width apart, with their arms crossed and hands resting on shoulders. Position—Sitting: Subjects sat on a stool with a ﬂat, horizontal surface and no back support. The height of the stool was adjusted to 1001 of hip and knee ﬂexion. Feet were positioned parallel, shoulder width apart, ﬂat on the ﬂoor. Arms were crossed and hands resting on shoulders. Sagittal posture—Neutral: The lumbar neutral spine posture was estimated by two physiotherapists in both standing and sitting positions as ‘‘neutral lordosis’’. The neutral lordosis was achieved by adjusting the pelvic tilt and thoracic alignment to

ARTICLE IN PRESS A. Burnett et al. / Manual Therapy 13 (2008) 300–306

302

being approximately upright to neutralise the curvature of the lumbar spine. Further, for all neutral position testing, subjects were positioned into the correct position with the assistance of feedback from a customised software program interfacing with the electromagnetic device. This provided the quantiﬁed mid-point between the pre-determined end-range ﬂexion and extension values. Sagittal plane angles were monitored closely during the axial rotation trials as a matter of quality control, and trials were considered void if the difference between the starting neutral posture and the neutral posture during axial rotation was greater than 51 during the trial. Sagittal posture—Flexion: Lumbar end-range ﬂexion in standing was deﬁned as maximal lumbar ﬂexion with knees extended and pelvis neutral (midway between anterior and posterior tilt). The ﬂexed position in sitting was deﬁned as posterior pelvic rotation and maximal thoraco-lumbar ﬂexion in a slumped posture with alignment of shoulders over the hips. Sagittal posture—Extension: Lumbar end-range extension was deﬁned as anterior pelvic rotation with maximal lumbar lordosis in both standing and sitting. Axial rotation in neutral: The therapist manually guided subjects into a neutral lordosis. Subjects were then asked to maintain this neutral posture and to actively axially rotate as far as possible to one side (randomly chosen left or right). This movement was guided by a second therapist via the shoulders to ensure minimal sagittal plane deviation during the rotation. Axial rotation in ﬂexion: Once ﬂexed the subject was then asked to maintain the ﬂexed posture and to actively axially rotate to end-range. This movement was guided by a second therapist via the shoulders to ensure minimal sagittal plane deviation during the rotation. Axial rotation in extension: The subject was asked to maintain the extended posture and to actively axially rotate to the end-range. This movement was guided by a second therapist via the shoulders to ensure minimal sagittal plane deviation during the rotation.

All data were collected using the device’s standard data collection software. Prior to testing, an experienced physiotherapist identiﬁed and marked the position of the spinous processes of L3 and S2, which was conﬁrmed by a second experienced physiotherapist. Pilot testing revealed that extreme care had to be taken with the sensor attachment, as motion artefact was substantial with direct attachment to the skin overlying the spinous processes with this combined movement protocol. Speciﬁcally, the sensor moved substantially with the skin over the erector spinae to the side contralateral to the direction of trunk rotation. Therefore, to minimise this artefact each FastrakTM sensor was attached to a 15 cm clear perspex ruler and placed horizontally on the skin, with each sensor aligned over the previously marked spinous processes (S2 and L3). During sensor attachment subjects were positioned in slight spinal ﬂexion to minimise displacement caused by skin movement. Each ruler was attached to the skin by two strips of double-sided tape along its upper and lower boarder (Fig. 1). Sensors were then ﬁrmly secured around the pelvis and trunk by a Nylatex strap. This method of attachment was derived from previous work examining three-dimensional movement of the trunk in fast bowling in cricket (Burnett et al., 1998). From repeated observation by the investigators prior to the study, this

2.3. Data collection During testing in each of the positions and postures, three-dimensional (3D) spinal kinematics data were recorded using the 3-Space FastrakTM (Polhemus Navigation Science Division, Kaiser Aerospace, Vermont). The Fastrak is a non-invasive electromagnetic device, which measures the position and orientation of points in space in 3D at a sampling frequency of 25 Hz. This apparatus has been shown to be both reliable and valid for measurement of lumbar spine movement with an accuracy of 0.21 (Pearcy and Hindle, 1989).

Fig. 1. The 3-Space Fastrak in use, prior to Nylatex strap ﬁxation. One sensor was mounted over the spinous process S2 and the second sensor was mounted over the spinous process of L3 via rulers, which were in turn attached to the skin with double-sided tape.

ARTICLE IN PRESS A. Burnett et al. / Manual Therapy 13 (2008) 300–306

Lower lumbar axial rotation: deﬁned as the relative axial rotation in the lower lumbar spine (between L3 and S2). Lower lumbar sagittal angles: deﬁned as the angle between two intersecting lines, one the tangent to sensor on the skin surface of L3 and one the tangent to the sensor on S2, with positive lumbar curvature values indicating extension.

2.4. Statistical analysis A two-way ANOVA with two within-subject variables was used to compare whether the range of axial rotation differed in posture (neutral, end-range ﬂexion and extension) and position (sitting and standing). Posthoc analysis was conducted using Least Squares Differences. Reliability of axial rotation measurements was conducted using the Standard Error of Measurement (SEM) (Norton et al., 2000). All statistical procedures were conducted using SPSS v12.0 for Windowss with the level of signiﬁcance set at po0.05. 3. Results The single measure SEM values ranged between 0.71 and 2.41 and for each subject, the two trials for each condition were averaged. This approach was justiﬁed as the reliability for all variables describing the magnitude of axial rotation in all positions and postures was acceptable when considering the magnitude of the differences measured in the study (as detailed below). Further, there were no signiﬁcant differences between the magnitude of left and right axial rotation in any of the experimental conditions. The differences between the amount of left and right axial rotation ranged between 0.21 and 1.11. Therefore, these values were summed to give the range of axial rotation.

The range of lower lumbar axial rotation decreased in both end-range ﬂexion and extension when compared to the therapist-deﬁned neutral posture (Fig. 2). Speciﬁcally, in extension the range of lower lumbar axial rotation decreased (po0.001) and in ﬂexion lower lumbar axial rotation also was signiﬁcantly less (po0.001). Further, there was a signiﬁcant difference evident for the range of lower lumbar axial rotation measurements in ﬂexion when sitting was compared to standing (p ¼ 0.013). To demonstrate that subjects closely replicated the end-range postures that were measured prior to testing during the experimental protocol, a comparison was made between the sagittal angles measured prior to testing and when being positioned at end-range just prior axial rotation commencing. This comparison revealed subjects were indeed placed at, or near endrange, during the experimental protocol with the average difference being no greater than 51 for any position and posture that was tested (Table 1). 35 Range of Axial Rotation (deg)

attachment method clearly replicated movement of the trunk. Prior to processing the raw data, a customised quality control programme written in LabVIEW V7.0 (National Instruments, TX, USA) was used to check for data collection errors. Then from each trial two dependent variables were calculated to evaluate the lumbar spine posture (Dankaerts et al., 2006) using the matrix algebra procedures similar to those outlined by Burnett et al. (1998). The only difference was that Cardan angles (a series of three rotations describing 3-D position in space with rotations occurring about all three axes) with an order of rotation of ﬂexion/extension (Y), axial rotation (X) and lateral bending (Z) with a moving frame of reference were used in this study. The neutral spine posture in sitting and standing was deﬁned as zero and the range of sagittal ﬂexion and extension was measured relative to this point. The two angles were as follows:

303

30

*

*

Neutral Extension Flexion

25

*

20

* 15 10 5 0 Sitting

Standing

Fig. 2. Range of lower lumbar axial rotation in end-range ﬂexed and extended postures when compared to the neutral spine posture in sitting and standing positions. * Indicates statistically signiﬁcant difference (po0.001) when compared to the neutral spine posture.

Table 1 Mean lower lumbar end-range sagittal angles (1) in sitting and standing positions at baseline and at the start of axial rotation range measurements (SD reported as mean of four trials)

Sitting Sagittal angle—baseline Sagittal angle—axial rotation measurement Difference Standing Sagittal angle—baseline Sagittal angle—axial rotation measurement Difference

Flexion

Extension

23.0 26.0 (1.1)

28.5 32.3 (3.2)

3.0

3.8

41.3 43.3 (1.7)

25.1 30.1 (2.3)

2.0

5.0

Flexion is represented by negative values and extension is represented by positive values.

ARTICLE IN PRESS 304

A. Burnett et al. / Manual Therapy 13 (2008) 300–306

4. Discussion Panjabi (1992a, b) proposed the concept of the ‘‘neutral zone’’ where a region of high ﬂexibility or relative laxity exists in neutral spine postures. In contrast, the ‘‘elastic zone’’ is deﬁned as a region of high stiffness, where signiﬁcant internal resistance to motion is provided. In ﬂexion, this resistance is thought to be provided by the posterior ﬁbres of the annulus and the posterior ligaments (Gunzburg et al., 1991, 1992; Pearcy, 1993) whilst in extension the facet joints create bony opposition (Schendel et al., 1993; Haberl et al., 2004). This increased internal resistance thereby places more strain on the associated spinal structures. This concept has been examined in vitro (Haberl et al., 2004) but not in vivo. The hypothesis of the current study was that there would be a decrease in the range of lower lumbar spine axial rotation in end-range ﬂexion and extension postures in both seated and standing positions when compared to a neutral spine posture due to reduced tissue compliance within these positions. From the results of this study the hypothesis was conﬁrmed. Haberl et al. (2004) investigated the magnitude of lumbar spine axial rotation in mid-range, ﬂexion and extension postures in vitro. They reported reduced axial rotation in extension when compared to a mid-range position, which is in agreement with the current study. They also showed a trend of reduced axial rotation in a ﬂexed spinal position. In contrast, the current study demonstrated that the reduction of rotation was greatest in ﬂexion. The study of Haberl and associates was the ﬁrst to compare the range of lumbar axial rotation in both ﬂexion and extension to a pre-determined midrange posture and their results in vitro are largely consistent with our in vivo ﬁndings. The fact that we found greater restriction of rotation in ﬂexion compared to extension may reﬂect the potentially greater compressive loading on the spine in ﬂexed postures limiting rotation during in vivo testing. The results of the current study are also in agreement with the ﬁndings of Gunzburg et al. (1991) who demonstrated that axial rotation was decreased in lumbar ﬂexion in vivo compared to upright postures in sitting and standing. Their testing was conducted on eight male subjects with Steinmann pins inserted into the spinous processes of the lower lumbar spine. The in vitro component of their study also showed a slight decrease in rotation of the entire lumbar spine (T12 to sacrum) in ﬂexion when compared to neutral, however this was not statistically signiﬁcant. It should be noted that the speciﬁc neutral posture and the magnitude of ﬂexion and extension with reference to end-range was not described in either of the in vivo or in vitro studies. Pearcy (1993) in contrast demonstrated there was a trend towards increased lumbar rotation in ﬂexed

postures, although this was only statistically signiﬁcant when comparing sitting to upright standing. It has been hypothesised by McGill (2001) and McGill and Cholewicki (2001) that there is a risk of low back injury when loading the spine in end-range ﬂexion due to increased loading of the passive spinal structures. Our ﬁndings demonstrate that axial rotation is reduced at both end-range ﬂexion and extension, supporting that passive spinal structures are less compliant to movement at either end of the sagittal plane range of motion. This is likely to be due to increased stiffness of the passive spinal system from increased connective tissue strain and facet joint loading as has been described previously in the case of lumbar spine ﬂexion (Gunzburg et al., 1991). These ﬁndings also indicate that axial loading the spine in end-range sagittal postures may provide a greater potential for risk of injury if tissue is loaded beyond its tolerance level, as this is the position where the passive structures appear to be at their maximal stiffness. Rotating beyond this point of tissue tolerance may result in increased shear through the disc (Bogduk, 1997) and other supportive structures. This concept is supported by the literature that demonstrates an increased prevalence of LBP in individuals who perform tasks that demand high volumes of spinal loading in mid-range and/or endrange positions associated with coupled rotation such as rowing (McGregor et al., 2002; Caldwell et al., 2003; Holt et al., 2003; Perich et al., 2006), gymnastics (Tertti et al., 1990) cycling (Burnett et al., 2004), fast bowling in cricket (Burnett et al., 1996, 1998; Ranson et al., 2005, 2007) and manual work (O’Sullivan et al., 2006). To our knowledge, this study is the ﬁrst in vivo attempt to measure the range of lower lumbar spine axial rotation in a deﬁned neutral posture in comparison with end-range ﬂexion and extension postures. When measuring the lower lumbar axial rotation range, there was minimal difference from the baseline end-range measurements in the sagittal plane (Table 1), indicating axial rotation was in fact measured very close to endrange ﬂexion and extension and consequently, within the respective regions of high stiffness. Further, there is no widely accepted deﬁnition of neutral posture and judgements are based on clinical opinion (Scannell and McGill, 2003). The actual mid-position of the sagittal plane, demonstrating the neutral posture as proposed by Panjabi (1992a, b) in vitro is calculated as the half way point of the end-range ﬂexion and extension values (from Table 1). In the current study the difference between the therapist-deﬁned neutral and calculated neutral in standing was 6.61 and the difference in sitting was 3.21, indicating that the therapist deﬁned neutral posture was within a neutral spinal posture (Panjabi, 1992a, b) The similarity of pre-rotational measurements with baseline measurements in end-range and neutral postures demonstrates the method of the current study

ARTICLE IN PRESS A. Burnett et al. / Manual Therapy 13 (2008) 300–306

produced measurements that reﬂect Panjabi’s concept of the neutral and elastic zones. A potential limitation of this study is that surface landmarks to quantify trunk movement were used and therefore are likely to overestimate ‘‘true’’ lumbar spine axial rotation due to skin distraction (Gregersen and Lucas, 1967; Pearcy and Tibrewal, 1984; Gunzburg et al., 1991; Russell et al., 1993). However, the study utilised a within-subjects design to compare the range of axial rotation in sagittal plane postures in sitting and standing which consequently strengthens the study. Further, as our measurements are comparisons of a young female sporting population, care has to be taken with extrapolation of the results to the general adult population. Further research is required examining different age groups and across genders.

5. Conclusion The results of this study demonstrate that reduced range of lower lumbar axial rotation exists in end-range ﬂexion and extension postures when compared to a neutral spine posture, in both sitting and standing positions in a group of adolescent females. Further, there was a reduction of axial rotation in ﬂexion when sitting was compared with standing. The reduction in axial rotation in end-range postures is likely to be due to the increased stiffness of the passive spinal structures in the elastic zone of motion as termed by Panjabi.

Acknowledgements The authors thank, for their kind assistance throughout this study, Dr. Ritu Gupta for statistical advice and Ms. Debra Perich for liaising with the schoolgirl rowing program. The authors declare the experiments of the study comply with the current laws of the country in which they were performed, with ethics approval obtained from the Institutional Human Research Ethics Committee and parents/guardians were required to provide their informed consent.

References Balague´ F, Troussier T, Salminen JJ. Non-speciﬁc low back pain in children and adolescents: risk factors. European Spine Journal 1999;8:429–38. Bogduk N. Clinical anatomy of the lumbar spine and sacrum. New York: Churchill Livingston; 1997. Burnett AF, Khangure MS, Elliott BC, Foster DH, Marshall RN, Hardcastle PH. Thoracolumbar disc degeneration in young fast bowlers in cricket: a follow-up study. Clinical Biomechanics 1996; 11:305–10.

305

Burnett AF, Barrett CJ, Marshall RN, Elliott BC, Day RE. Threedimensional measurement of lumbar spine kinematics for fast bowlers in cricket. Clinical Biomechanics 1998;13:574–83. Burnett A, Cornelius M, Dankaerts W, O’Sullivan P. Spinal kinematics and back muscle activity in cyclists: a comparison between healthy controls and non-speciﬁc chronic low back pain— a pilot investigation. Manual Therapy 2004;9:211–9. Caldwell JS, McNair PJ, Williams M. The effects of repetitive motion on lumbar ﬂexion and erector spinae muscle activity in rowers. Clinical Biomechanics 2003;18:704–11. Dankaerts W, O’Sullivan P, Burnett A, Straker L. Differences in sitting postures are associated with non-speciﬁc chronic low back pain disorders when patients are sub-classiﬁed. Spine 2006;31: 698–704. Gregersen GG, Lucas DB. An in vivo study of the axial rotation of the human thoracolumbar spine. Journal of Bone and Joint Surgery 1967;49A:247–62. Gunzburg R, Hutton W, Fraser R. Axial rotation of the lumbar spine and the effect of ﬂexion; an in vitro and in vivo biomechanical study. Spine 1991;16:22–8. Gunzburg R, Hutton WC, Crane G, Fraser RD. Role of the capsuloligamentous structures in rotation and combined ﬂexion-rotation of the lumbar spine. Journal of Spinal Disorders 1992;5:1–7. Haberl H, Cripton PA, Orr TE, Beutler T, Frei H, Lanksch WR, et al. Kinematic response of lumbar functional spinal units to axial torsion with and without superimposed compression and ﬂexion/ extension. European Spine Journal 2004;13:560–6. Holt PJE, Bull AMJ, Cashman PMM, McGregor AH. Kinematics of spinal motion during prolonged rowing. International Journal of Sports Medicine 2003;24:597–602. McGill SM. Low back stability: from formal description to issues for performance and rehabilitation. Exercise and Sports Science Reviews 2001;29:26–31. McGill SM, Cholewicki J. Biomechanical basis of stability: an explanation to enhance clinical utility. Journal of Orthopaedic and Sports Physical Therapy 2001;31:96–100. McGregor A, Anderton L, Gedroyc W. The assessment of intersegmental motion and pelvic tilt in elite oarsmen. Medicine and Science in Sports and Exercise 2002;34:1143–9. Norton K, Marfell-Jones M, Whittingham N, Kerr D, Carter L, Saddington K, Gore C. Anthropometric assessment protocols. In: Physiological Tests for Elite Athletes. South Australia, Human Kinetics Lower Mitcham, 2000. p. 66–85. O’Sullivan PB, Mitchell T, Bulich P, Waller R, Holte J. The relationship between posture and back muscle endurance in industrial workers with ﬂexion related low back pain. Manual Therapy 2006;11:264–71. Panjabi M. The stabilizing system of the spine. Part 1: function, dysfunction, application and enhancement. Journal of Spinal Disorders 1992a;5:383–9. Panjabi M. The stabilization system of the spine. Part 2: neutral zone and instability hypothesis. Journal of Spinal Disorders 1992b;5: 390–7. Pearcy MJ. Twisting mobility of the human back in ﬂexed postures. Spine 1993;18:114–9. Pearcy MJ, Hindle RJ. New method for the non-invasive threedimensional measurement of human back movement. Clinical Biomechanics 1989;4:73–9. Pearcy MJ, Tibrewal S. Axial rotation and lateral bending in the normal lumbar spine measures by three dimensional radiography. Spine 1984;9:582–7. Perich D, Burnett A, O’Sullivan P. Low back pain and the factors associated with it: Examination of adolescent female rowers. In: Schwameder H, Srtutzenberger G, Fastenbauer V, Lidinger S, Muller E, editors. 24th symposium of the international society of biomechanics in sports. The University of Salzburg, 2006. p. 355–8.

ARTICLE IN PRESS 306

A. Burnett et al. / Manual Therapy 13 (2008) 300–306

Ranson CA, Kerslake RW, Burnett AF, Batt ME, Abdi S. Magnetic resonance imaging ﬁndings of the lumbar spine in asymptomatic professional fast bowlers in cricket. Journal of Bone and Joint Surgery 2005;87Br:1111–6. Ranson C, Burnett A, King M, Patel N, O’Sullivan P. The relationship between bowling action classiﬁcation and three-dimensional lumbar spine motion in fast bowlers in cricket, 2007, in review. Russell P, Pearcy MJ, Unsworth A. Measurement of the range and coupled movements observed in the lumbar spine. British Journal of Rheumatology 1993;32:490–7.

Scannell J, McGill SM. Lumbar posture—should it and can it be modiﬁed? A study of passive tissue stiffness and lumbar position during activities of daily living. Physical Therapy 2003;83:907–16. Schendel MJ, Wood KB, Buttermann GR, Lewis JL, Ogilve JW. Experimental measurement of ligament force, facet force, and segment motion in the human lumbar spine. Journal of Biomechanics 1993;26:427–38. Tertti M, Paajanen H, Kujala UM, Alanen A, Salmi TT, Kormano M. Disc degeneration in young gymnast: a magnetic resonance imaging study. American Journal of Sports Medicine 1990;18:206–8.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 307–316 www.elsevier.com/locate/math

Original article

Reliability of stiffness measured in glenohumeral joint and its application to assess the effect of end-range mobilization in subjects with adhesive capsulitis Hui-Ting Lina, Ar-Tyan Hsub, Kai-Nan Anc, Jia-rea Chang Chiend, Ta-Shen Kuane, Guan-Liang Changa, a

Institute of Biomedical Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan b Department of Physical Therapy, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan c Biomechanics Laboratory, Division of Orthopedic Research, Mayo Clinic, College of Medicine, Rochester, MN 55905, USA d Department of Electrical Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan e Department of Orthopedic, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan Received 29 March 2006; received in revised form 19 January 2007; accepted 28 February 2007

Abstract End-range mobilization techniques are recommended for the treatment of patients with hypomobile joints. The purposes of this study were (1) to assess the reliability of a glenohumeral (GH) stiffness measurement technique and (2) apply the measurement technique on subjects with adhesive capsulitis to compare the GH end-range stiffness and rotational range of motions (ROMs) before and immediately after the application of end-range translational mobilization techniques. Fifteen normal subjects were recruited for assessment of test–retest reliability. Four men and two women with adhesive capsulitis in the glenohumeral joint (mean disease duration ¼ 6.5 months, SD ¼ 2.7) were treated with end-range mobilization by an experienced physical therapist. The passive abduction angles, rotational ROM and GH joint stiffness were measured by the same observer before and immediately after end-range mobilization treatment. The test–retest reliability was assessed and revealed good to excellent reliability in anterior– posterior glenohumeral joint stiffness and fair to excellent reliability of GH stiffness in posterior–anterior direction. The GH joint stiffness decreased and passive abduction range of motion increased immediately after end-range mobilization of the shoulder joint. The use of intensive mobilization techniques may help to decrease the risk of further stiffness or joint contracture progression in patients with adhesive capsulitis. r 2007 Elsevier Ltd. All rights reserved. Keywords: End-range mobilization; Glenohumeral joint; Range of motion; Stiffness

1. Introduction Examination of the shoulder routinely involves testing joint range of motion (ROM) and glenohumeral (GH) joint laxity. In the clinical setting, humeral translation tests, such as the anterior drawer test, posterior drawer test and sulus test, are the most common and important Corresponding author. Tel.: +886 6 2757575x63421; fax: +886 6 2343270. E-mail address: [email protected] (G.-L. Chang).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.02.003

tests to identify shoulder joint instability. However, since the measured joint laxity depends on the force applied by the examiner, alternatively measuring joint stiffness can avoid the force application dependence. GH joint stiffness is deﬁned as the ratio of the incremental force required to stretch the GH structures over a range of humeral head translation to the translational displacement, which provides information regarding the mechanical properties of the joint (Makhsous et al., 2004). GH stiffness may be altered in the case of shoulder injuries such as unstable shoulder

ARTICLE IN PRESS 308

H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

joint or adhesive capsulitis, and it is therefore generally considered to be one of the crucial aspects of the clinical shoulder joint examination. Several studies have quantiﬁed the passive GH force– displacement relationship (Hsu et al., 2000; Sauers et al., 2001; McQuade and Murthi, 2004) and have described rotation–moment behaviour (Novotny et al., 2000; Hsu et al., 2002b). However, data describing the passive joint stiffness and its reliability are still lacking (McQuade and Murthi, 2004; Borsa et al., 2006). McQuade and Murthi (2004) demonstrated a technique for quantitative clinical examination of the GH joint stiffness. The use of a computed internal humeral head centre for translation tracking, a true scapula reference frame to deﬁne the GH translation, and GH joint rotation relative to the scapula orientation enabled precise measurement of GH joint movement, but the reliability of joint stiffness was not reported in their study. In Borsa’s study, the GH stiffness is objectively characterized in baseball pitchers using an instrumented stress device (LigMasterTM, Sport Tech, Inc., Charlottesville, VA) while their shoulders were held at only one position (901 of abduction in the scapula plane and 601 of external rotation). The between-session ICC values were reported ranging from 0.29 to 0.89, depending on the direction of stiffness testing (Borsa et al., 2006). The term ‘‘adhesive capsulitis’’ was ﬁrst coined by Codman in 1934 and was subsequently deﬁned as an idiopathic condition of the shoulder characterized by the spontaneous onset of pain in the shoulder with restriction of mobility at the GH joint in every direction (Codman, 1934; Rundquist et al., 2003). Adhesive capsulitis has been divided into two types: (i) primary adhesive capsulitis, which refers to the idiopathic form of a painful and stiff shoulder, and (ii) secondary adhesive capsulitis, indicated as a loss of motion resulting from many predisposing factors such as trauma, stroke, upper extremity fracture, or surgery with immobilization (Mao et al., 1997; Beyers and Bonutti, 2004). Orthopaedic physical therapists frequently employ translational mobilization techniques to treat patients with adhesive capsulitis. Several authors have suggested the use of end-range mobilization techniques (EMTs) for treating these patients (Wadsworth, 1986; Edmond, 1993). Joint laxity and ROM are frequently used to assess the clinical effect of EMTs in the treatment of patients with adhesive capsulitis. In addition, clinical assessment of end-feel stiffness in the GH joint is universally employed to characterize the resistance of tissue to a manually applied force. However, experimental data from clinical efﬁcacy studies on the effect of EMTs on patients with adhesive capsulitis are lacking. Several in vitro studies have indicated that translational GH joint mobilization at a greater abduction range can stretch capsular ligaments and increase abduction ROM

to improve abduction hypomobility (Hsu et al., 2002a, b). However, the cadaver specimens used in these studies had no active muscle tension, the core temperature was not the same as in living tissue, and the material properties of living tissue may differ from those of cadaver tissue. A few studies have investigated the effect of mobilization techniques on the GH joint with capsular adhesions in vivo (Placzek et al., 1998; Gamage and Lasenby, 2002; Gokeler et al., 2003; Boyles et al., 2005). However, the performance of the technique (end-range or mid-range mobilizations combined with interscalene brachial plexus block or stretching exercise), duration of treatment (several weeks or several months) and the utilization of other modalities (hot packs, cold packs and home exercises) differ from the approach used in the present study. In order to exclude other effects from physical modalities, such as hot or cold pack, neither was used before the application of EMTs in the current study. The beneﬁcial effects on shoulder ROM of this manual technique were reported by Vermeulen et al. (2000). Nevertheless, in their study the shoulder ROM relative to trunk did not represent pure GH ROM. Although three-dimensional (3-D) motion analysis techniques can separate GH and scapulothoracic motion, capture true GH movement, and make in vivo estimation of the centre of rotation (COR) in the GH joint possible with reasonable precision and reliability, reliable data of GH joint stiffness is still lacking and no quantitative 3-D kinematics data related to detecting changes in GH stiffness or ROM after end-range mobilization treatment in the shoulder joint are available either. Therefore, the purpose of this study was to assess the reliability of this study’s GH stiffness measurement technique. This measurement technique was also applied on the subjects with adhesive capsulitis to compare the GH end-range stiffness and ROM before (T0) and immediately after (T1) the application of endrange translational mobilization technique.

2. Method 2.1. Subjects Fifteen healthy volunteers with no shoulder symptoms and six patients with a clinical diagnosis of unilateral adhesive capsulitis having a painful stiff shoulder for at least 3 months were recruited. All patients were referred according to clinical criteria of having a shoulder disorder with a pattern of restricted joint mobility of more than 50% in passive shoulder abduction, forward ﬂexion and external rotation compared with the contralateral site (Kelly and Clark, 1995; Vermeulen et al., 2000). Patients with a history of neurological disorders, shoulder impingement syndrome,

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

309

Table 1 Demographics for normal subjects (eight males, seven females) Variable Age (years) Height (cm) Weight (kg) Joint play Shoulder ROM (1)a Flexion

Abduction

External rotation

Internal rotation

Active

Active

Passive

Active

Passive

173.0 179.6 164.3 173.3 101.3 7.0 6.3 7.9 7.2 9.1 165–190 170–200 145–170 155–180 90–120

112.6 11.3 95–130

55.3 12.0 40–80

69.6 14.6 45–90

Active Mean SD Range a

23.2 2.6 20–29

167.6 5.9 155–179

57.0 8.4 44–71

3 0 3

Passive

Passive

Shoulder ROM relative to the trunk.

Table 2 Demographics and clinical characteristics for six patients with adhesive capsulitis Subjects

Sex

Age

Side

Occupation

Duration of complaints (months)

No. of previous injections

Sleep cycle disturbances

1 2 3 4 5 6

F M M M M F

50 70 74 64 49 50

Non-D D D D D D

Employee Metal worker None Teacher Teacher Teacher

3.5 11 4 6 8 7

1 0 1 0 0 0

Y N Y N Y Y

rotator cuff tear, severe osteoarthritis, or osteoporosis were excluded. All participants read and signed informed consent forms before participating in this study. The research protocol was approved by the Institutional Review Board of the National Cheng Kung University Hospital, Tainan, Taiwan. Normal subject demographics, shoulder ROM and joint play were recorded and are presented in Table 1. The maximal active and passive ROM of the shoulder in normal subjects was assessed by a goniometer. The aetiology of the adhesive capsulitis (primary or idiopathic) was unknown for the six patient participants of this study, and the dominant arm was involved in ﬁve of the six patients. The patients received physical therapy treatment sessions including physical modalities (ultrasound, short-wave therapy, or electrotherapy), mid-range mobilization techniques, passive stretching and active exercise before enrolment in this study. Two patients had received a corticosteroid injection in their symptomatic shoulder from their physicians, and four subjects had sleep-cycle disturbances. The demographic data for the patients are presented in Table 2. 2.2. Instrumentation A 3-D electromagnetic tracking device with a transmitter and four receivers (Polhemus Inc., 40 Hercules Dr, Colchester, VT) was used to collect the kinematic information from both groups. Three electromagnetic

sensors were used to track the position and orientation of the thorax, scapula and humerus. A stylus was used to digitize anatomic landmarks for deﬁning the anatomic coordinate of each segment. The residual errors of angles and linear translations after calibration were less than 0.121 and 0.7 mm, respectively. A strain gauge and a torque transducer (MLP-50 and SWS-100, Transducer Techniques, Temecula, CA, USA) were used to measure the force and torque applied by the examiner so as to ensure that they were repeatable and reliable. 2.3. Procedure Before testing, the tracking system, force transducer and torque transducer were calibrated. All subjects were seated on a specially designed sturdy chair with clamps grasping the spine of the scapula superiorly and posteriorly and the clavicle anteriorly. The lateral borders of the scapula were also blocked with another clamp. The subjects were instructed to sit relaxed with their trunk stabilized with a pelvic and chest belt. An adjustable elbow brace was applied on the elbow to immobilize the elbow joint at 901 of ﬂexion. Three sensors were attached with one to the sternum notch, another was attached to the ﬂat surface of the acromion, and the last to a cuff around the upper arm with adhesive tape. While subjects sat with their symptomatic arms relaxed in neutral position, nine bony landmarks on the thorax, humerus and scapula were palpated and digitized.

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

310

Yt GH

ys

Zt

MR

LR

C7

AT

SN

T8

xs

PX Xt

RS

UR zs

PT

AA

DR

Yh

ER IA

AB

EM

Zh

IR

EL

Ext Flex

Xh

A

Fig. 1. Bony landmarks used to deﬁne local coordinate system of the thorax, humerus and scapula according to the suggestion of the International Shoulder Group. SN: suprasternal notch, C7: spinous process of the seventh cervical vertebra, T8: spinous process of the eighth thoracic vertebra, PX: xiphoid process, EM: medial epicondyle, EL: lateral epicondyle, GH: glenohumeral rotation centre, AA: acromial angle, RS: root of the spine of the scapula, IA: inferior angle of the scapula. IR/ER: internal rotation/external rotation, Flex/Ext: ﬂexion/extension, AB/AD: abduction/adduction, MR/LR: medial rotation/lateral rotation, PT/AT: posterior tipping/anterior tipping, UR/DR: upward rotation/downward rotation.

The location of the humeral head centre [centre of the glenoid (CGL) and centre of the humeral head (CHH)] was determined using Gamage’s method with a smallarc movement of the humerus (Gamage and Lasenby, 2002). Using these digitized points and estimated humeral head centre, the local coordinate systems (LCS) of trunk, humerus and scapula were determined according to the suggestion of the International Shoulder Group (Fig. 1). The transformation matrix between receiver data and the local anatomically based coordinate systems was then constructed. An 80-N anterior–posterior (A–P) or posterior– anterior (P–A) force, monitored by a force sensor, was applied during A–P and P–A glides, and the markers were tracked to assess the mobility of the GH joint at 101 increments of GH abduction in neutral rotation (Fig. 2A, B). A diagram demonstrating the procedure to compute the amount of displacement is shown in Fig. 3. As was stated above, the CGL and CHH coincide at the COR of the GH joint during small-arc movements of the GH joint. Once the COR of the GH joint was computed, the relationship between the distal humeral receiver position and the COR (and thus the CHH) within the humeral LCS could be determined. Likewise, the relationship between the acromial receiver position and the COR (and thus the CGL) within the scapular LCS could also be established. Once these relationships were established, the coordinates of CHH and CGL could be derived separately from the humeral and acromial sensor data, irrespective of the initial COR position of the GH joint. During this displacement testing the amount of translation was computed from

the vector difference with respect to the anterior axis of the plane of the scapula between the CGL and CHH at the initial position (neutral rotation at every abduction angle in the plane of the scapula) and that of the end position of the A–P and P–A procedure. The end-range stiffness in each abduction position was calculated as the slope of the best-ﬁt line on the linear region of the force– displacement curve with Matlab software (Su et al., 2005) (Fig. 4). The torque transducer was monitored and used to ensure a consistent torque of 4 N m for the internal rotation (IR) and external rotation (ER) moments. The rotational ROM (IR, ER) of the GH joint was determined in response to 4 N m of IR and ER moment. The coordinates and orientations of the receivers were tracked at every 101 increment from the neutral position to the end-range of GH abduction during rotation ROM testing in the plane of the scapula. For GH joint rotation, the anatomic landmarks of the scapula and humerus are needed to be recorded to specify the relative orientation of each segment (scapula and humerus) based on anatomically LCS. The receiver orientation data during motion was transformed to describe relative positions of the anatomical LCS and align with the anatomically based clinically meaningful coordinate system according to the constructed transformation matrix between receiver data and the anatomically based LCS. The change in the humeral orientation relative to the scapula can be described as the Euler angle. The Y–X0 –Z00 rotation sequence of Euler angles was used to describe the three-axis rotation angles of the GH joint. Three trials were performed to measure end-range

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

311

Fig. 2. An 80-N posterior–anterior force was applied to the head of the humerus (A); a force sensor was used to monitor applied force (B); a torque transducer was used during rotational ROM testing to ensure a consistent torque of 4 N m for internal rotation (C) and external rotation (D) procedures.

* Small-arc movement Receiver at humerus Gamage method # Static position Receiver at scapula CGL

=

CHH

Constant vector relationship within scapula LCS

Dynamic motion Receiver at scapula

# Static position Receiver at humerus

GH center

Constant vector relationship within humerus LCS

CGL at motion

CHH at motion

Dynamic motion Receiver at humerus

Displacement of head relative to glenoid Fig. 3. A diagram that demonstrates the procedure to compute the amount of displacement. CHH: centre of the humeral head, CGL: centre of the glenoid, GH: glenohumeral, LCS: local coordinate system.

ARTICLE IN PRESS 312

H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

force is used but the stretching of soft tissue around the joint lasts for a shorter period of time. If the patients have constant pain in the shoulder joint through the ROM, the joint should be treated with a large amplitude grade 3 mobilization. If the ROM of the patient is not restricted by increased pain while moving to the end-of-range placement, the grade 4 technique is suggested. 2.5. Data analysis

Fig. 4. The force–displacement curve for one subject showing the linear region of anteroposterior glide (AP) and posteroanterior glide (PA). Stiffness was deﬁned as the slope of the linear region.

stiffness measures and quantiﬁed rotational ROM at each angle of GH abduction (01, 101, 201, 301, 401, 501 and end-range of GH abduction range) in the plane of the scapula for normal subjects as well as before and after 30 min of end-range joint mobilization treatment for the subjects with adhesive capsulitis. 2.4. Intervention A physical therapist with more than 10 years of experience performed the EMT originally described by Maitland (1991). The physical therapist examined the shoulder ROM of patients to obtain the end-range position and end-feel of the GH joint prior to each GH EMT treatment. The therapist began the intervention with a few minutes of mid-range mobilization for warmup purposes. The patient was then employed with various mobilization techniques of grade 3 or 4 in various end-range positions of the GH joint while the patient was positioned supine or side lying. The therapist’s hands were placed close to the GH joint, and the patient’s humerus was placed into a position of maximal ﬂexion, maintained in the sagittal plane, or into a position of maximal abduction in the coronal plane. Mobilizations were performed using the therapist’s hand to improve the gliding of the humeral head. After 10–15 repetitions of intensive mobilization techniques in each end-range position, the direction of mobilization was altered by varying the degree of rotation or by varying the plane of elevation. In addition, the movement included inferior, A–P, and P–A glides, and various distraction techniques in several end-range GH joint positions. The mobilization grade (grade 3 or 4) and duration of applied stress was varied to the tolerance of the patient for each direction of the EMT treatment. In general, the grade 3 mobilization was considered more comfortable than a grade 4 because the same degree of

The repeatability of the measurement during three trials on the same day was determined. The intraclass correlation coefﬁcient (ICC2,1) was used to test the reliability of the respective stiffness at 01, 101, 201, 301, 401, 501 and end-range of GH abduction for both groups. The reliability of rotational ROM was also assessed. The mean absolute difference between the second and the third measures of stiffness in normal subjects at each GH abduction was then determined as the measurement error to test whether it is precise to conclude that there is a real change after EMTs intervention. The Wilcoxon Sign Rank Test was used to identify statistically signiﬁcant differences in GH joint stiffness before and immediately after EMTs. The level was set as 0.05. All data analyses were performed using SPSS for window 11.0 (SPSS, Inc., Chicago, IL).

3. Results In normal subjects, the ICC values of measurements of stiffness (N/mm) were 0.82–0.96 for A–P and 0.68– 0.94 for P–A glide. The within-day mean absolute differences (measurement error) for normal GH stiffness are shown in Table 3. The intrasession test–retest reliability of the torque ROMs (TROM, i.e. IR and ER) ICC values were 0.98–0.99 for IR ROM and 0.97–0.99 for ER ROM in normal subjects (Table 3). A–P and P–A stiffness in patients with adhesive capsulitis ranged from 0.69 to 0.92 and 0.45 to 0.98, respectively. The patients’ values of ICC for the intrasession test–retest reliability of the TROM (IR and ER) were 0.97–0.99 (IR) and 0.89–0.99 (ER) (Table 3). A representative force–displacement curve before and immediately after the EMT treatment is displayed in Fig. 5. The A–P and P–A stiffness of the affected GH joint at multiple abduction angles before and immediately after the EMT treatments are shown in Fig. 6. The A–P stiffness at 501 of GH abduction and the P–A stiffness of patients at 101 of GH abduction position were decreased after the treatment in the plane of the scapula. We compared the change in passive mobility of abduction, ER and IR in the GH joint immediately

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

313

Table 3 Repeatability of glenohumeral joint A–P and P–A stiffness and torque ROM (IR, ER) among normal subjects and patients Position

0 10 20 30 40 50 End

Normal subjects

Patients

ICC Intra (AP/PA)

ICC Intra (IR/ER)

Diff23 (AP/PA)

ICC Intra (AP/PA)

ICC Intra (IR/ER)

0.91/0.90 0.89/0.68 0.82/0.82 0.90/0.86 0.96/0.89 0.95/0.87 0.89/0.94

0.99/0.99 0.99/0.99 0.98/0.99 0.99/0.97 0.98/0.97 0.99/0.99 0.98/0.98

0.31/0.65 0.53/0.79 0.41/0.66 0.67/0.80 0.66/1.01 0.70/0.60 1.32/1.15

0.92/0.82 0.87/0.45 0.78/0.87 0.69/0.87 0.71/0.98 0.71/0.85 0.85/0.90

0.99/0.98 0.99/0.99 0.99/0.98 0.99/0.99 0.98/0.97 0.97/0.89 0.98/0.98

0, 10, 20, 30, 40, 50 and end represent 01, 101, 201, 301, 401, 501 and end-range of glenohumeral abduction. AP: anterior–posterior stiffness; PA: posterior–anterior stiffness. Diff23: The mean difference between the second and the third measures of glenohumeral joint stiffness (unit: N/mm).

16 Before EMTs After EMTs

A-P stiffness (N/mm)

14

*

12 10 8 6 4 2 0 0

Fig. 5. Representative force–displacement curve for one subject before and immediately after end-range mobilization techniques.

10

20 30 40 50 Abduction position (degrees)

end

18 16 P-A stiffness (N/mm)

after the EMT treatments in the six patients (Table 4). The mean GH abduction improved from 45.651 (SD ¼ 15.621, range ¼ 24.47–59.291) to 54.951 (SD ¼ 5.241, range ¼ 48.44–60.761), and the mean ER ROM at the end-range of abduction increased from 27.191 (SD ¼ 19.541) to 55.941 (SD ¼ 30.531). However, a small difference in the mean IR ROM at the neutral position (01 of abduction) was observed between T0 and T1, and the mean IR angle at the end-range and ER at the neutral position decreased immediately after the end-range manipulative intervention. The mean total rotational ROM in subject 1 and subject 2 increased after EMTs at the end-range of abduction. Subject 3 and subject 4 exhibited increased rotational ROM at the lower abduction position. Rotational ROM in subject 5 increased from 69.681, 65.641, 57.701 and 45.761 to 80.701, 71.061, 66.901 and 56.071 at 301, 401, 501 and the end-range of abduction, respectively. The improvement in rotational ROM in subject 6 was evident at abduction angles from 201 to the end-range (Table 5).

*

14

Before EMTs After EMTs

12 10 8 6 4 2 0 0

10

20 30 40 Abduction position (degrees)

50

end

Fig. 6. Glenohumeral joint stiffness before and immediately after endrange mobilization techniques during the glide tests: (A) anteroposterior glide and (B) posteroanterior glide.

4. Discussion The test–retest reliability of the GH joint stiffness in our measurement technique was sufﬁcient for normal

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

314

inﬂammation condition is often relieved after at least 3 months; therefore our outcome measures were not affected by the presence of severe pain. The patients group did not receive uniform intervention (i.e. modalities or other treatment) before enrolment into our study; however, since the current study involved measurements of GH joint stiffness and rotational ROM before and immediately after mobilization, the instant differences after EMTs are not likely to be relevant to whether these patients received different treatment prior to EMT intervention. The end ROM values for the mobilization techniques used were deﬁned at the location that each subject could comfortably approach. In the current study, the patients with adhesive capsulitis were treated with intensive (sometimes painful) mobilizations. It is occasionally necessary to change the intensity or the direction in order to decrease the reﬂex muscle activity and muscle guarding during the mobilization manoeuvre. Sometimes massage and rhythmic low-amplitude oscillation can be used to relax the guarding condition of the surrounding muscle (Vermeulen et al., 2000). During treatment, the guarding condition which would cause resistance to the mobilization technique can be assessed with the GH joint play performed by the physical therapist. It is imperative to ensure that the surrounding muscle is relaxed during EMT treatment and during the measurement of GH joint ROM and stiffness. In our study, the patients were instructed to inform the physical therapist of any pain during and after intervention. Although some patients stated that they had pain around the shoulder joint, the therapist believed that there was no reﬂex muscle activity or any muscle guarding during EMT intervention. Thus, the soft tissue around the GH joint of the patients may be stretched effectively during EMT. In this study, the mean A–P and P–A stiffness decreased immediately after mobilization at abduction angles of 501 and 101, respectively. According to the previous GH specimen study (Hsu et al., 2002a), the mobilization force (grade 3) performed by experienced physical therapists falls safely in the linear elastic region

subjects and patients with adhesive capsulitis. Since the within-day efﬁcacy of EMTs was evaluated for the GH joint of patients in this study design, the test–retest reliability was assessed for the within-day condition instead of between-days. The reliability of GH stiffness can be inﬂuenced by multiple factors, such as subject positioning, muscle tension, palpation of humeral head position, or ﬁxation of the scapula (McQuade and Murthi, 2004; Bryde et al., 2005; Borsa et al., 2006). Reliability of GH stiffness is even further inﬂuenced by the above factors for between-day sessions. Low between-session ICC values have been reported (Borsa et al., 2006). Thus, between-day reliability for GH joint stiffness was not investigated in the present study. The stiffness measurement error was less than the average value of the absolute difference of stiffness between before and immediately after EMTs at the GH abduction angles of 01, 201, 301, 501 for A–P glide and those of 01, 101, 201, 301, 501 for P–A glide. It indicated that the difference after EMT intervention may not be the result of measurement error. Therefore, it appears to show that the changes in stiffness after intervention vary in response to changes with the applied load at most GH abduction positions. The present study included patients with symptoms of adhesive capsulitis that had been present for at least the previous 3 months. Application of the EMT does not cause too much pain in the chronic stage (in contrast to the acute stage). We believe that the adhesive capsulitis Table 4 Passive mobility measurement (in degrees) before (T0), immediately after (T1) end-range mobilization techniques (EMTs) T1

T0

Abduction Internal rotation 0 External rotation 0 Internal rotation end External rotation end

X

SD

X

SD

45.65 23.84 46.05 31.86 27.19

15.62 22.73 23.45 12.65 19.54

54.95 22.80 37.96 18.16 55.94

5.24 19.87 24.15 13.42 30.53

Table 5 Passive total rotational range of motion measurements (in degrees) before (T0), immediately after (T1) end-range mobilization techniques (EMTs) Patient no.

1 2 3 4 5 6

TR at 0 (1)

TR at 10 (1)

TR at 20 (1)

TR at 30 (1)

TR at 40 (1)

TR at 50 (1)

TR at end-range

T0

T1

T0

T1

T0

T1

T0

T1

T0

T1

T0

T1

T0

T1

46.68 35.01 38.27 97.14 104.47 105.62

29.77 15.09 30.49 102.02 103.33 83.89

51.56 57.60 40.88 106.59 106.06 103.49

35.87 37.45 44.77 105.77 100.71 104.58

54.79 66.33 40.83 107.13 110.84 89.83

50.78 52.74 49.75 111.17 90.33 100.68

– 60.58 – 124.01 69.68 83.69

– 40.03 – 94.97 80.70 98.67

– 55.95 – 113.44 65.64 66.34

– 51.01 – 99.04 71.06 98.65

– 61.64 – 104.62 57.70 66.54

–

67.33 41.22 55.61 94.09 45.76 53.20

67.59 51.79 55.41 75.30 56.07 98.52

TR: total rotation ROM; –: measurement that the subject cannot reach.

33.75 – 79.18 66.90 105.91

ARTICLE IN PRESS H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

of the load–displacement relationship. In the terminology of manual therapy, grade 3 and 4 mobilizations refer to a large-amplitude joint movement that reaches the end ROM and a small-amplitude movement within a joint performed at the end of the ROM, respectively (Maitland, 1991). Both grades reach the same point end ROM and therefore, the same force is required to create them. The only difference is in the amplitude of the movement. Therefore, grade 3 or 4 mobilizations should theoretically be located at the linear elastic region of the load–displacement curve in the loose-pack or close-pack position. In our opinion, at the end-range of abduction, the GH ligament may be stretched decreasing the slack on the ligament. Therefore, at the limit of abduction during EMTs the load can be applied directly on the stretched tissues and is more effective at decreasing joint stiffness, improving joint mobility and producing elastic deformation in the stiff joint when compared to loading performed in neutral position. The present study has found that A–P, P–A and inferior glides performed close to the end-range of abduction will improve abduction ROM. Hsu et al. reported a signiﬁcant increase in GH abduction immediately following an A–P glide and an improvement in GH abduction immediately after a caudal glide in cadavers (Hsu et al., 2000, 2002b). Increases in active and passive abduction ROM after 3 months of EMT treatment have also been reported in a multiple-subject case report (Vermeulen et al., 2000). An increase in abduction ROM can be explained as follows: the posterior and anterior bands of the inferior GH ligament and the axillary pouch become the primary stabilizers against inferior gliding of the humeral head on the glenoid fossa as the GH joint approaches the endrange of abduction (O’Brien et al., 1995; Hsu et al., 2000, 2002b). In this position, A–P, P–A and inferior glides will most effectively stretch the posterior band, anterior band of the inferior GH ligament and axillary pouch. Therefore, the entire tightened inferior GH ligament could be released, increasing the abduction ROM after EMT treatments. It is also important to know whether this treatment is effective on other ROM measures, such as rotation. In the present study, the mean passive ER ROM appeared to increase at the end of abduction, with the total rotational ROM of the subjects increasing with differing patterns. In the study by Roubal et al. (1996), eight patients received GH gliding under anaesthesia. Immediately following manipulation, the mean increases in passive ROM for ﬂexion, abduction, ER and IR were 681, 771, 491 and 451, respectively. Vermeulen et al. (2000) reported increases in passive ROM including mean abduction, mean ﬂexion and mean lateral rotation after 3 months of treatment. The results of the current study are consistent with previous studies demonstrating that end-range mobilization treatments appear to

315

produce positive outcomes in increasing abduction and rotational ROM (Wadsworth, 1986; Maitland, 1991; Edmond, 1993). To our knowledge, viscoelastic tissue will deform progressively until a newly stable length is reached when a load is applied over a long period, whereas the tissue will gradually return to the original length when a load is applied within the linear elastic region over a short period (Threlkeld, 1992). In the present study, we evaluated the immediate rather than the long-term efﬁcacy of EMTs. Therefore, although beneﬁcial effects on immediate decrease of stiffness and increase of abduction ROM were observed, we believe that the long-term application of EMT treatment to the GH joint is needed in patients with adhesive capsulitis to obtain the permanent deformation of stiff tissues. 5. Conclusion This investigation has established that our measurement technique for glenohumeral joint stiffness is reliable within the same day for normal subjects. We applied the measurement technique on patients with adhesive capsulitis to assess the effect of end-range mobilization. Patients with adhesive capsulitis treated with end-range mobilization appear to have the tendency to immediately improve in abduction range of motion and decreases of the glenohumeral joint stiffness. References Beyers M, Bonutti P. Frozen shoulder. In: Donatelli R, editor. Physical therapy of the shoulder. New York: Churchill Livingstone; 2004. p. 319–36 [chapter 11]. Borsa PA, Dover GC, Wilk KE, Reinold MM. Glenohumeral range of motion and stiffness in professional baseball pitchers. Medicine & Science in Sports and Exercise 2006;38:21–6. Boyles RE, Flynn TW, Whitman JM. Manipulation following regional interscalene anesthetic block for shoulder adhesive capsulitis: a case series. Manual Therapy 2005;10:80–7. Bryde D, Jane Freure B, Jones L, Werstine M, Kathryn Briffa N. Reliability of palpation of humeral head position in asymptomatic shoulders. Manual Therapy 2005;10:191–7. Codman EA. The shoulder. Rupture of the supraspinatus tendon and other lesions in or about the subacromial bursa. Boston: Thomas Todd; 1934. Edmond SL. Manipulation and mobilizaton—extermities and spinal techniques. St. Louis: Mosby; 1993. Gamage S, Lasenby J. New least square solutions for estimating the average center of rotation and the axis of rotation. Journal of Biomechanics 2002;35:87–93. Gokeler A, van Paridon-Edauw GH, DeClercq S, Matthijs O, Dijkstra PU. Quantitative analysis of traction in the glenohumeral joint. In vivo radiographic measurements. Manual Therapy 2003;8: 97–102. Hsu AT, Ho L, Ho S, Hedman T. Joint position during anteriorposterior glide mobilization: its effect on glenohumeral abduction range of motion. Archives Physical Medicine and Rehabilitation 2000;81:210–4.

ARTICLE IN PRESS 316

H.-T. Lin et al. / Manual Therapy 13 (2008) 307–316

Hsu AT, Ho L, Chang JH, Chang GL, Hedman T. Characterization of tissue resistance during a dorsally directed translational mobilization of the glenohumeral joint. Archives Physical Medicine and Rehabilitation 2002a;83:360–6. Hsu AT, Hedman T, Chang JH, Vo C, Ho L, Ho S, et al. Changes in abduction and rotation range of motion in response to simulated dorsal and ventral translational mobilization of the glenohumeral joint. Physical Therapy 2002b;82:544–56. Kelly MJ, Clark WA. Orthopedic therapy of the shoulder. Philadelphia: J.B. Lippincott Company; 1995. p. 139–142. Maitland GD. Peripheral manipulation. Boston: ButterworthsHeinemann; 1991. Makhsous M, Lin F, Zhang LQ. Multi-axis passive and active stiffnesses of the glenohumeral joint. Clinical Biomechanics (Bristol, Avon) 2004;19:107–15. Mao CY, Jaw WC, Cheng HC. Frozen shoulder: correlation between the response to physical therapy and follow-up shoulder arthrography. Archives Physical Medicine and Rehabilitation 1997;78:857–9. McQuade KJ, Murthi AM. Anterior glenohumeral force/translation behavior with and without rotator cuff contraction during clinical stability testing. Clinical Biomechanics 2004;19:10–5. Novotny JE, Woolley CT, Nichols CE, Beynnon BD. In vivo technique to quantify the internal-external rotation kinematics of the human glenohumeral joint. Journal of Orthopaedic Research 2000;18:190–4. O’Brien SJ, Schwartz RS, Warren RF, Torzilli PA. Capsular restraints to anterior-posterior motion of the abducted shoulder:

a biomechanical study. Journal of Shoulder and Elbow Surgery 1995;4:298–308. Placzek JD, Roubal PJ, Freeman DC, Kulig K, Nasser S, Pagett BT. Long-term effectiveness of translational manipulation for adhesive capsulitis. Clinical Orthopaedics and Related Research 1998: 181–91. Roubal PJ, Dobritt D, Placzek JD. Glenohumeral gliding manipulation following interscalene brachial plexus block in patients with adhesive capsulitis. Journal of Orthopaedics Sports Physical Therapy 1996;24:66–77. Rundquist PJ, Anderson DD, Guanche CA, Ludewig PM. Shoulder kinematics in subjects with frozen shoulder. Archives of Physical Medicine and Rehabilitation 2003;84:1473–9. Sauers EL, Borsa PA, Herling DE, Stanley RD. Instrumented measurement of glenohumeral joint laxity and its relationship to passive range of motion and generalized joint laxity. American Journal of Sports Medicine 2001;29:143–50. Su BW, Protopsaltis TS, Koff MF, Chang KP, Strauch RJ, Crow SA, et al. The biomechanical analysis of a tendon ﬁxation device for ﬂexor tendon repair. Journal of Hand Surgery 2005; 30:237–45. Threlkeld AJ. The effects of manual therapy on connective tissue. Physical Therapy 1992;72:893–902. Vermeulen HM, Obermann WR, Burger BJ, Kok GJ, Rozing PM, van Den Ende CH. End-range mobilization techniques in adhesive capsulitis of the shoulder joint: a multiple-subject case report. Physical Therapy 2000;80:1204–13. Wadsworth CT. Frozen shoulder. Physical Therapy 1986;66:1878–83.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 317–324 www.elsevier.com/locate/math

Original article

The inﬂuence of neck pain on balance and gait parameters in community-dwelling elders Eliza Poole, Julia Treleaven, Gwendolen Jull Division of Physiotherapy, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane 4072, Qld., Australia Received 8 June 2006; received in revised form 20 November 2006; accepted 28 February 2007

Abstract Neck pain has been shown to be associated with balance disturbances. Balance and gait speed are also known to decline with ageing. The aim of this study was to determine whether the presence of neck pain was associated with a decline in postural stability and gait speed over and above what is expected with normal ageing. Twenty female subjects with idiopathic neck pain and 20 healthy female controls aged between 65 and 82 years were studied. Subjects performed balance tests on a computerised force plate under conditions of eyes open, eyes closed on ﬁrm and soft surfaces in comfortable and narrow stance. Sway energy and root mean square (RMS) amplitude of sway were measured. Subjects also undertook a Timed Ten Metre Walk Test, with and without head turning. There were trends for the elderly group with neck pain to have poorer balance than the healthy controls across most balance conditions, although differences were signiﬁcant only in the following tests; comfortable stance—eyes closed on a ﬁrm surface (p ¼ 0.02), eyes open on a soft surface (p ¼ 0.01); narrow stance—eyes open on a ﬁrm surface (p ¼ 0.02). In the Timed Ten meter Walk Test, elderly subjects with neck pain had a slower self-selected gait speed (p ¼ 0.02) and cadence (p ¼ 0.04) in the head turn condition, as well as a longer gait cycle duration both with (p ¼ 0.00) and without head turns (p ¼ 0.04). The results of this study suggest that neck pain in the elderly may contribute to some disturbance in balance and gait parameters over and above that which occurs with normal ageing. r 2007 Elsevier Ltd. All rights reserved. Keywords: Balance; Neck pain; Proprioception; Elderly

1. Introduction Age-related functional decline in the motor and sensory systems may affect balance function (Kerr et al., 1985; Woollacott et al., 1986; Toupet et al., 1988; Maylor & Wing, 1996). Balance and postural control have been shown to decline with age (Speers et al., 1998; Woollacott, 2000; Isles et al., 2004) and the role of factors such as vestibular function, motor control of back and pelvic musculature and muscle strength has been investigated in elderly people (Alhanti et al., 1997; Speers et al., 1998; Mientjes and Frank, 1999). Corresponding author. Tel.: +61 7 3365 2275; fax: +61 7 3365 2775. E-mail address: [email protected] (J. Treleaven).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.02.002

Musculoskeletal conditions, and speciﬁcally neck pain in this instance, might also contribute to balance deﬁcits in the elderly. Neck pain is not uncommon in this age group. March et al. (1998) documented a neck pain prevalence of 40.5% in elderly women and 36.1% in elderly men living independently in the community. Cervical afferent input is an important contributor to balance (Karlberg et al., 1996a; Lekhel et al., 1998; Bove et al., 2002) and balance disturbances have been documented in young and middle aged individuals with neck pain of both insidious and traumatic onset (Michaelson et al., 2003; Treleaven et al., 2005a,b). Subsequent improvements in balance have been demonstrated following localised treatment to the cervical spine (Fattori et al., 1996; Karlberg et al., 1996a). In the absence of vestibular pathology, such disturbances are

ARTICLE IN PRESS 318

E. Poole et al. / Manual Therapy 13 (2008) 317–324

considered to result from altered cervical somatosensory input and integration in the postural control system (DeJong and DeJong, 1977; Pyykko et al., 1989; Karlberg et al., 1996b; Ishikawa et al., 1998; Kavounoudias et al., 1999; Bove et al., 2002; Courtine et al., 2003; Schieppati et al., 2003; Gosselin et al., 2004). McPartland et al. (1997) determined a signiﬁcant correlation between poor balance control and fatty inﬁltration of the cervical extensor musculature in those with neck pain and others have demonstrated the adverse effects of neck extensor muscle fatigue on postural sway (Schieppati et al., 2003; Gosselin et al., 2004). Stimulation of the muscle spindle afferents via neck muscle vibration has also been shown to not only increase postural sway (Kavounoudias et al., 1999) but also inﬂuence the velocity and direction of gait and running in asymptomatic healthy individuals (Bove et al., 2002; Courtine et al., 2003). As a signiﬁcant number of elderly persons in the community experience neck pain, we questioned whether or not there was a relationship between neck pain and a further decline in balance and/or gait parameters in the aged. Although balance and gait speed declines with age, both poor balance and slowness of gait have been recognised as risk factors for falls (Tinetti et al., 1988; Cress et al., 1995). If a relationship exists, it is possible that speciﬁc treatment directed to the neck pain may help to reduce the contributors to balance and gait disorders and risk of falls in this group. Thus, the aim of the study was to determine if any differences existed in selected standing balance tests and gait speed parameters between elderly subjects with neck pain when compared to elderly subjects without neck pain.

2. Method 2.1. Participants Forty community-dwelling women, 65 years or older, with neck pain (n ¼ 20) and without neck pain (n ¼ 20) participated in the study. Participants were recruited from the Brisbane metropolitan area using a method of convenience sampling. Volunteers were sought from the Australasian Centre for Ageing’s 50+ Registry, and advertisements were placed in local newspapers and community settings where a high proportion of older people met. Physiotherapy practices in the area were approached to assist in the recruitment of participants with neck pain. To be included in the neck pain group, participants were to report neck pain of greater than three months’ duration with a score of at least 10 out of 100 on the Neck Disability Index (Vernon, 1996). In order to control for other factors which may inﬂuence balance,

potential participants were excluded on the following criteria: taking more than four medications, a history of falls, recent orthopaedic surgery (hip/knee/ankle problems), inner ear pathology, stroke, head injury, diabetes, neurological or vestibular pathology, arthritis that required active management, those on pain management, or acute injuries (such as ankle/knee sprains). In an attempt to access a sample representative of the community-dwelling population aged over 65 years, those with minor ailments and common medical conditions were not excluded from the study. Detailed information concerning co-morbidities was collected from the participants at the testing session and from answers to screening questionnaires. The information was coded into four categories deﬁned as: (1) musculoskeletal conditions affecting the lumbar spine or lower limb, encompassing arthritis and non-speciﬁc low back pain; (2) common medical conditions including hypertension, heart problems, osteoporosis and depression; (3) dizziness and (4) previous traumatic neck injury. Each participant’s co-morbidities were placed in a code and the number of ‘positive’ codes was summed, with the total possible equal to four. For example, if a participant had arthritis in the knees and low back pain then this would be a positive code 1. If the participant also had hypertension (positive code 2), her total co-morbidities would be two. Ethical Clearance for the study was obtained from the Medical Ethics Committee of The University of Queensland and all procedures were conducted according to the Declaration of Helsinki. All the subjects gave their informed consent to participate. 2.2. Measures Questionnaires were administered to collect demographic data, level of current resting neck pain (VAS) and selfreported neck disability (Neck Disability Index (NDI)) (Vernon, 1996). Two sets of measures were taken from each participant: standing balance and timed walking tasks. 2.3. Standing balance A modiﬁed Clinical Test of Sensory Integration and Balance (CTSIB) was used to assess standing balance (Shumway-Cook and Horak, 1986; Cohen et al., 1993; Alhanti et al., 1997). The CTSIB is a timed test that was developed for systematically testing the inﬂuence of visual, vestibular and somatosensory input on standing balance (Shumway-Cook and Horak, 1986). Eight tests were used as in our previous research (Treleaven et al., 2005b). The conditions of the CTSIB tested were: eyes open and closed on both ﬁrm and soft surfaces, in both comfortable and narrow stance. The soft surface was a piece of high-density (10 cm thick) foam rubber placed on the force platform. Tandem stance was not included

ARTICLE IN PRESS E. Poole et al. / Manual Therapy 13 (2008) 317–324

as it has been previously determined that it is too challenging for healthy elders (Speers et al., 1998). Computerised dynamic posturography (CDP) was used to measure the amount of body sway under each condition. The subjects stood on a computerised, stable force platform of 40 cm 60 cm with strain gauges located at each corner of the force platform. Changes in standing balance were recorded in both the mediolateral (ML) and anteroposterior (AP) directions under altered visual and support conditions. Each individual was required to stand in the comfortable stance (with the feet approximately shoulder width apart) and in the narrow stance (with the feet together). A piece of paper, with the pre-determined comfortable foot position traced on it, was used to ensure that the subjects’ feet were placed in the exact same position for each comfortable stance test (McIlroy and Maki, 1997). For narrow stance, the subjects were asked to place their feet as close together as possible, centred over the marked midpoint of the force plate. 2.4. Timed walking tasks 2.4.1. 10 Metre Walk test with and without head turning, measured with a stride analyser The Ten Metre Walk (TMW) test is a test of physical mobility and balance that evaluates a person’s ability to adjust their centre of gravity continuously over a moving base of support (Arnadottir and Mercer, 2000). It is a measure of self-selected walking speed, which has been said to be a primary predictor of selfperceived function (Cress et al., 1995). As gait speed is easily assessed, this measure offers an efﬁcient clinical tool for assessing physical function and was selected to compliment the information gained from the static standing measures on the force platform. The walk was completed with head facing straightforward and repeated with the subject rotating their head from side to side. A stride analyser (model SA-III) was used to measure a number of variables during each subject’s walk including: time taken to complete the 10 m (seconds), number of strides taken, cadence (steps/seconds), stride length (centimeters) and gait cycle duration (seconds). The stride analyser consists of two foot pads inserted into the subject’s left and right shoes. These pads are connected to a small device, worn around the subject’s waist, which records the information collected with each foot strike on activation of a switch. The data from stride analyser variables were downloaded onto an IBM computer.

3. Procedure Prospective participants were initially screened via telephone interview by one of the investigators. Subjects

319

who met the inclusion criteria and entered the study completed both standing balance and timed walking tasks. The order of the standing balance tasks was set to ensure the least challenging tasks were tested ﬁrst, to prevent subject fatigue. Standing balance was tested before the walking task. For the CTSIB tests, the subjects stood on the computerised force plates. They were given clear instructions to stand as still as possible, and look straight ahead at a wall 1.5 m in front of them with their arms by their sides. Each subject performed one 30 s trial for each of the eight balance conditions. For safety, two researchers were present and stood close to the subjects during each test. Furthermore, the subjects were permitted adequate rest time during the test procedures to minimise any effect of fatigue. For the TMW test, the subjects stood at the end of a 14 m walkway with markers at 2 m from each end of the walkway to indicate the start and ﬁnish of the measurement area. The investigator stood beside or behind the subject. The standardised instruction was given, ‘‘I want you to walk to the far line at your comfortable speed. Do not stop until you reach the far line.’’ The investigator walked beside or behind the subject without providing hands on support unless the subject appeared to be overbalancing. If overbalancing occurred the test was repeated. (No overbalancing occurred in our study.) The tester pressed the switch of the Stride Analyser to commence and end recording as the subject crossed the 2-m and 12-m marks, respectively. The test was then repeated with head rotation. Subjects were instructed to, ‘‘Walk to the far line at your comfortable speed, whilst continually turning your head from side to side. Do not stop until you reach the far line.’’

4. Statistical analysis Two characteristics of sway were analysed, total energy (Wavelet analysis using Daubechies ﬁlter 6) (Treleaven et al., 2005b) and root mean square (RMS) amplitude in millimetres (Labview, National Instruments). Measures of both amplitude and frequency/ velocity were used as they are both considered important components to reﬂect the strategy of postural control (Dault et al., 2001). Wavelet analysis was chosen to measure the energy of the sway signal (Treleaven et al., 2005b) and RMS was chosen to demonstrate the average amplitude travelled by the centre of pressure. The latter is a robust measure of amplitude (Prieto et al., 1996; Rocchi et al., 2004) and has demonstrated differences in postural control between clinical populations (Rocchi et al., 2002). Data of the AP and ML traces in comfortable and narrow stance conditions were analysed. Normality of the data set was assessed using

ARTICLE IN PRESS E. Poole et al. / Manual Therapy 13 (2008) 317–324

320

Q–Q plots to verify parametric test use and data were logged. Differences between the neck pain and control groups for age, co-morbidities, number of medications and NDI scores were assessed initially using a series of one-way analysis of variances (ANOVAs). Group differences due to signal energies and RMS were examined using a generalized linear model, multivariate analysis of variance (MANOVA). The a level was set at 0.05.

Table 2 presents the group characteristics for the Timed Ten Metre Walk Test. Missing data were present for both groups due to a malfunction of the foot switch and thus the data were analysed with n ¼ 16 for both the control and neck pain groups. Signiﬁcant differences were found between groups for both the time it took to complete the test (p ¼ 0.02) and the cadence (steps per second) (p ¼ 0.04) when walking with head turns. The neck pain group took a longer time to complete the walk and used slower steps. The neck pain group also had a signiﬁcantly longer gait cycle duration when walking both with (p ¼ 0.00) and without head turns (p ¼ 0.04).

5. Results Group characteristics for age, NDI, co-morbidities and number of medications are presented in Table 1. There were no signiﬁcant between group differences for age (p ¼ 0.277) or co-morbidities (p ¼ 0.19). As expected, the neck pain group had a signiﬁcantly higher NDI (p ¼ 0.00) and were taking more medications than the control group (p ¼ 0.015). To account for between-subjects effects, the number of medications taken and age were included as factors in the MANOVA. Whilst there were no mean age differences between the groups, the subjects were not precisely age matched. Figs. 1 and 2 present the mean and standard errors for the logged energy values for each test in the AP and ML directions for the neck pain and control groups and Figs. 3 and 4 present those for the amplitude of sway (RMS values). Overall there was a tendency for higher levels of both total energy and RMS amplitude in sway in the neck pain group. Differences reached signiﬁcance between the neck pain and control groups in the AP direction in comfortable stance in the condition of eyes closed on a ﬁrm surface for both total energy (p ¼ 0.02) and RMS (p ¼ 0.02) (Figs. 1 and 3). The RMS amplitude was signiﬁcantly greater in the neck pain group in the test of eyes open on the soft surface in the AP direction (p ¼ 0.01). Signiﬁcant differences were also seen in both total energy (p ¼ 0.02) and RMS (p ¼ 0.02) in the ML direction in narrow stance with eyes open on the ﬁrm surface (Figs. 2 and 4). Table 1 Characteristics of the neck pain and control groups

NDI Age Co-morbidities Number of medications po0.02.

Controls N ¼ 20 Mean (standard error)

Neck pain N ¼ 20 Mean (standard error)

3.00 (0.78) 71.4 (1.1) 2.00 (0.23) 1.50 (0.29)

23.95 (2.3) 70.3 (0.88) 2.45 (0.15) 2.70 (0.36)

6. Discussion The results of this study demonstrate that elderly subjects with neck pain demonstrate some deﬁcits in standing balance and gait parameters when compared to asymptomatic elderly subjects, which may alter their functional balance and gait ability and possibly increase their risk of falling. These ﬁndings point to the need to consider the contribution of a cervical disorder to balance and gait speed deﬁciencies in the older person. They also point to the need for routine assessment of these parameters in all elderly patients presenting for either treatment of neck pain or conversely for falls prevention programs if treatment programs are to be holistic. There were trends for increased postural activity on all static standing tests in the neck pain group, although a signiﬁcant increase in both amplitude (RMS) and energy of sway was present in only two of the eight test conditions (eyes closed on a ﬁrm surface in comfortable stance and eyes open on a ﬁrm surface in narrow stance). An increase in amplitude but not energy was seen in one other test (eyes open on a soft surface in comfortable stance). There were no test conditions where energy, but not amplitude, was signiﬁcantly higher in the neck pain group. Increases in amplitude of sway in the neck pain group suggest that these subjects may have a decreased awareness of altered stability. The result is an increase in amplitude of sway rather than the subject choosing a stiffening strategy to maintain stability. A stiffening strategy would present as a high energy (frequency) but a reduced amplitude sway path (Dault et al., 2001). It has been suggested that changes measured on the force platform tests may be associated with an increase risk of falls (Piirtola and Era, 2006). Speciﬁcally, Melzer et al. (2004) demonstrated an increase in ML sway in narrow stance in elders with recurrent falls. Others have also noted an association between an increase in ML RMS sway amplitude during normal standing and future falls in elderly subjects (Topper et al., 1993; Maki et al., 1994). In the current study, there was an

ARTICLE IN PRESS E. Poole et al. / Manual Therapy 13 (2008) 317–324

321

2 1.8

Controls Neck pain

Loggeg total energy

1.6 *

1.4 1.2 1 0.8 0.6 0.4 0.2 0 ceof

cecf

ceos

cecs

neof

necf

neos

necs

C = Comfortable foot position, N = narrow foot position, EO = eyes open, EC = eyes closed, F = firm surface, S = Soft surface Fig. 1. Comparison of antero-posterior total energy between control subjects and subjects with neck pain.

2.5 Controls Neck pain

Logged total energy

2

* 1.5

1

0.5

0 ceof

cecf

ceos

cecs

neof

necf

neos

necs

C = comfortable foot position, N = narrow foot position, EO = eyes open, EC = eyes closed, F = firm surface, S= soft surface Fig. 2. Comparison of medical-lateral total energy values between control subjects and subjects with neck pain.

increase in ML RMS amplitude and energy of sway in narrow stance with eyes open in the neck pain group. However, all other signiﬁcant differences were related to changes in the AP direction. Further research is warranted to determine the signiﬁcance of these ﬁndings from the force platform in relation to risk of falls and altered functional balance ability in elderly with neck pain. The Timed Ten Metre Walk Test revealed that elderly subjects with neck pain had a slower self-selected gait speed and cadence when walking whilst turning their head from side to side and a signiﬁcantly longer gait cycle duration when walking both with and without head turns. These results suggest that those with neck pain were more cautious or apprehensive with their

walking. Additionally turning the neck may cause pain or elicit fear of pain, which may further alter cervical somatosensory input to the postural control system, which could subsequently exacerbate balance and gait difﬁculties. The differences in gait parameters could also be a consequence of a dual tasking. Toulotte et al. (2006) recently found no change in gait parameters between fallers and non-fallers in a single task condition but signiﬁcant changes when subjects were asked to perform another task while walking. Such changes in gait quality measured in the current study may place those with neck pain at a greater risk of falls than non- neck pain elderly subjects. Wolfson et al. (1990) demonstrated that stride length, walking velocity and gait quality were reduced in elderly

ARTICLE IN PRESS E. Poole et al. / Manual Therapy 13 (2008) 317–324

322

10 9 Controls Neck pain

8

RMS (mm)

7

* *

6 5 4 3 2 1 0 ceof

cecf

ceos

cecs

neof

necf

neos

necs

C = Comfortable foot position, N = narrow foot position, EO = eyes open, EC = eyes closed, F = firm surface, S = Soft surface Fig. 3. Comparison of antero-posterior root mean square (RMS) amplitude between control subjects and subjects with neck pain.

12

10 Controls Neck pain

RMS (mm)

8 * 6

4

2

0 ceof cecf ceos cecs neof necf neos necs C = comfortable foot position, N = narrow footposition, EO = eyes open, EC= eyes closed, F = firm surface, S = soft surface Fig. 4. Comparison of medical-lateral root mean square (RMS) amplitude between control subjects and subjects with neck pain.

nursing home residents with a history of falls compared to control subjects. Imms and Edholm (1979, 1981) and Brauer et al. (2000) have also reported differences in stride length and walking speed between fallers and control subjects in community dwelling elderly. Self-selected gait speed is a predictor of self-perceived function (Cress et al., 1995). Nevertheless the average gait velocity demonstrated by the neck pain subjects in this study was still relatively high (greater than 1.1 ms) and thus may not necessarily relate to functional deﬁcits (Montero-Odasso et al., 2005). Further research is required to determine any clinical signiﬁcance of the changes to gait in the neck pain population.

The differences between the groups in this study could not be attributed to co-morbidities or medication intake. It is known that both use of more than four medications and the greater the number of co-morbidities a person has, the greater their risk of falls (Tinetti et al., 1988). Therefore, we speciﬁcally excluded any person with more than four co-morbidities and/or taking more than four medications, which challenged recruitment in the elderly population in this study. There were no differences between groups with respect to the number of co-morbidities, but the neck pain group, on average, took a greater number of medications. The number of medications taken was included as a covariate in the analysis but did not inﬂuence the differences in balance

ARTICLE IN PRESS E. Poole et al. / Manual Therapy 13 (2008) 317–324 Table 2 The results of the features measured in the Timed Ten Metre Walk Test for the neck pain and control groups Test

Controls

Neck pain

pValue

N ¼ 16 N ¼ 16 Mean (SE) Mean (SE) Time no head turn 7.43 (.25) 8.13 (.25) 0.08 Time with head turn 7.81 (.31) 8.93 (0.30) 0.02 Strides no head turn 5.85 (.15) 6.11 (.15) 0.24 Strides with head turn 6.26 (.22) 6.45 (.22) 0.58 Cadence no head turn 117.0 (3.3) 112.49 (3.3) 0.37 Cadence with head turn 121.64 (3.9) 108.97 (3.9) 0.04 Stride length no head turn 1.37 (.03) 1.32 (.03) 0.25 Stride length with head turn 1.32 (.03) 1.29 (.03) 0.68 Gait cycle duration no head turn 1.00 (.02) 1.07 (.02) 0.04 Gait cycle duration with head turn 1.01 (.02) 1.11 (.02) 0.00 SE ¼ Standard error. Cadence ¼ steps/seconds, gait cycle duration (seconds), stride length (centimeters).

seen between the groups, a feature previously determined in a younger population with neck pain (Treleaven et al., 2005b). Despite relatively small subject numbers, the results of this study suggest that older persons with neck pain demonstrate some disturbances to their balance and gait over and above those, which occur with normal ageing. Such disturbances are likely due to altered cervical somotosensory input and integration to the postural control system. Functional impairment of cervical muscle and joint receptors such as muscle fatigue (Falla et al., 2004) altered muscle spindle activity due to the presence of inﬂammatory mediators (Wenngren et al., 1998; Thunberg et al., 2001) and the effects of pain itself on both nociceptor and mechanoreceptor activity locally, at the spinal cord and within the central nervous system (Le Pera et al., 2001; Ageborg, 2002), are possible mechanisms of altered cervical input. The results of this study suggest that clinicians might consider management of neck pain and speciﬁc exercises to improve cervical afferent input in programmes directed at prevention of falls in older age and conversely, consider balance and gait assessment in those elders presenting with neck pain.

References Ageborg E. Consequences of a ligament injury on neuromuscular function and relevance to rehabilitation—using the anterior cruciate ligament-injured knee as a model. Journal of Electromyography and Kinesiology 2002;12:205–12. Alhanti B, Bruder L, Creese W, Golden R, Gregory C, Newton R. Balance abilities of community dwelling older adults under altered visual and support surface conditions. Physical and Occupational Therapy in Geriatrics 1997;15(1):37–53.

323

Arnadottir S, Mercer V. Effects of footwear on measurements of balance and gait in women between the ages of 95 and 93 years. Physical Therapy 2000;80(1):17–27. Bove M, Courtine G, et al. Neck muscle vibration and spatial orientation during stepping in place in humans. Journal of Neurophysiology 2002;88(5):2232–41. Brauer S, Burns Y, Galley P. A prospective study of laboratory and clinical measures of postural stability to predict community-dwelling fallers. Journal of Gerontology 2000;55A(8): 469–76. Cohen H, Blatchly C, Gombash L. A study of the clinical test of sensory integration and balance. Physical Therapy 1993;73:346–51. Courtine G, Papaxanthis C, et al. Gait-dependent integration of neck muscle afferent input. Neuroreport 2003;14(18):2365–8. Cress M, Schechtman K, Mulrow C, Fiatarone M, Gerety M, Buchner D. Relationship between physical performance and self-perceived physical function. Journal of the American Geriatric Society 1995;43:93–101. Dault MC, Geurts ACH, et al. Postural control and cognitive task performance in healthy participants while balancing on different support-surface conﬁgurations. Gait and Posture 2001;14(3): 248–55. DeJong PI, DeJong JM. Ataxia and nystagmus induced by injection of local anaesthetics in the neck. Annals of Neurology 1977;1977(1): 240–6. Falla D, Jull G, et al. Neck ﬂexor muscle fatigue is side speciﬁc in patients with unilateral neck pain. European Journal of Pain 2004;8(1):71–7. Fattori B, Borsari C, Vannucci G, Casani A, Cristofani R, Bonuccelli L, et al. Acupuncture treatment for balance disorders following whiplash injury. Acupuncture & Electro-Therapeutics Research 1996;21(3–4):207–17. Gosselin G, Rassoulian H, et al. Effects of neck extensor muscles fatigue on balance. Clinical Biomechanics 2004;19(5):473–9. Imms F, Edholm O. The assessment of gait and mobility in the elderly. Age and Ageing 1979;8:261–7. Imms F, Edholm O. Studies of gait and mobility in the elderly. Age and Ageing 1981;10:147–56. Ishikawa K, Matsuzaki Z, et al. Effect of unilateral section of cervical afferent nerve upon optokinetic response and vestibular nystagmus induced by sinusoidal rotation in guinea pigs. Acta Otolaryngology Supplement 1998;537:6–10. Isles R, Low Choy N, Steer M, Nitz JC. Normal values of balance tests in women aged 20–80. Journal of the American Geriatric Society 2004;52(8):1367–72. Karlberg M, Magnusson M, Malmstrom E, Melander A, Moritz U. Postural and symptomatic improvement after physiotherapy in patients with dizziness of suspected cervical origin. Archives of Physical Medicine and Rehabilitation 1996;77(9):874–82. Karlberg M, Johansson R, et al. Dizziness of suspected cervical origin distinguished by posturographic assessment of human postural dynamics. Journal of Vestibular Research-Equilibrium and Orientation 1996;6(1):37–47. Kavounoudias A, Gilhodes JC, et al. From balance regulation to body orientation: two goals for muscle proprioceptive information processing? Experimental Brain Research 1999;124(1):80–8. Kerr B, Condon S, McDonald L. Cognitive spatial processing and the regulation of posture. Journal of Experimental Psychology— Human Perception and Performance 1985;11(5):617–22. Lekhel H, Popov K, Bronstein A, Gresty M. Postural responses to vibration of neck muscles in patients with uni- and bilateral vestibular loss. Gait and Posture 1998;7(3):228–36. Le Pera D, Graven-Nielsen T, et al. Inhibition of motor system excitability at cortical and spinal level by tonic muscle pain. Clinical Neurophysiology 2001;112(9):1633–41.

ARTICLE IN PRESS 324

E. Poole et al. / Manual Therapy 13 (2008) 317–324

March LM, Brnabic AJ, Skinner JC, Schwarz JM, Finnegan T, Druce J, et al. Musculoskeletal disability among elderly people in the community. Medical Journal of Australia 1998;168:439–42. Maki BE, Holliday PJ, Topper AK. A prospective study of postural balance and risk of falling in an ambulatory and independent elderly population. Journal of Gerontology 1994;49(2):72–84. Maylor E, Wing A. Age differences in postural stability are increased by additional cognitive demands. Journal of Gerontology 1996; 51(3):143–54. McIlroy WE, Maki BE. Preferred placement of the feet during quiet stance: development of a standardized foot placement for balance testing. Clinical Biomechanics 1997;12(1):66–70. McPartland JM, Brodeur RR, et al. Chronic neck pain, standing balance, and suboccipital muscle atrophy—A Pilot Study. Journal of Manipulative and Physiological Therapeutics 1997;20 (Number 1):24–9. Melzer I, Benjuya N, Kaplanski J. Postural stability in the elderly: a comparison between fallers and non-fallers. Age and Ageing 2004;33(6):602–7. Michaelson P, Michaelson M, Jaric S, Latash ML, Sjolander P, Djupsjobacka M. Vertical posture and head stability in patients with chronic neck. Journal of Rehabilitation Medicine 2003; 35(5):229–35. Mientjes M, Frank J. Balance in low back pain patients compared to healthy people under various conditions in upright standing. Clinical Biomechanics 1999;14:710–6. Montero-Odasso M, Schapira M, Soriano ER, Varela M, Kaplan R, Camera LA, et al. Gait velocity as a single predictor of adverse events in healthy seniors aged 75 years and older. Journals of Gerontology Series A—Biological Sciences and Medical Sciences 2005;60(10):1304–9. Piirtola M, Era P. Force platform measurements as predictors of falls among older people—a review. Gerontology 2006;52(1):1–16. Prieto TE, Myklebust JB, Hoffmann RG, Lovett EG, Myklebust BM. Measures of postural steadiness: differences between healthy young and elderly adults. IEEE Transactions on Biomedical Engineering 1996;43(9):956–66. Pyykko I, Aalto H, et al. Hierarchy of different muscles in postural control. Acta Otolaryngology Supplement 1989;468: 175–80. Rocchi L, Chiari L, Cappello A. Feature selection of stabilometric parameters based on principal component analysis. Medical & Biological Engineering & Computing 2004;42(1):71–9. Rocchi L, Chiari L, Horak FB. Effects of deep brain stimulation and levodopa on postural sway in Parkinson’s disease. Journal of Neurology Neurosurgery and Psychiatry 2002;73(3):267–74.

Schieppati M, Nardone A, et al. Neck muscle fatigue affects postural control in man. Neuroscience 2003;121(2):277–85. Shumway-Cook A, Horak F. Assessing the inﬂuence of sensory integration on balance. Physical Therapy 1986;66:1548–50. Speers RA, Ashton-Miller JA, Schultz AB, Alexander NB. Age differences in abilities to perform tandem stand and walk tasks of graded difﬁculty. Gait and Posture 1998;7(3):207–13. Tinetti M, Speechley M, Ginter S. Risk factors for falls among elderly persons living in the community. New England Journal of Medicine 1988;319:1701–7. Thunberg J, Hellstrom F, et al. Inﬂuences on the fusimotor–muscle spindle system from chemosensitive nerve endings in the cervical facet joints in the cat; possible implications for whiplash induced disorders. Pain 2001;91:15–22. Topper AK, Maki BE, Holliday PJ. Are activity-based assessments of balance and gait in the elderly predictive of risk of falling and/or type of fall? Journal of the American Geriatric Society 1993;41(5):479–87. Toulotte C, Thevenon A, Watelain E, Fabre C. Identiﬁcation of healthy elderly fallers and non-fallers by gait analysis under dualtask conditions. Clinical Rehabilitation 2006;20(3):269–76. Toupet M, Gage P, Heuschen S. Vestibular patients and aging subjects lose use of visual input and expend more energy in static postural control. In: Vellas B, Toupet M, Rubenstein L, et al., editors. Balance and gait disorders in the elderly. Paris: Elsevier; 1988. p. 183–98. Treleaven J, Jull G, LowChoy N. Standing balance in chronic whiplash—Comparison between subjects with and without dizziness. Journal of Rehabilitation Medicine 2005;37(4): 224–9. Treleaven J, Murison R, Jull G, Low Choy N. Is signal analysis important for measuring standing balance in chronic whiplash? Gait and Posture 2005;21(4):395–402. Vernon H. The Neck Disability Index: patient assessment and outcome monitoring in whiplash. Journal of Musculoskeletal Pain 1996; 4(4):95–104. Wenngren B, Pedersen J, et al. Bradykinin and muscle stretch alter contralateral cat neck muscle spindle output. Neuroscience Research 1998;32:119–29. Wolfson L, Whipple R, Amerman P, Tobin J. Gait Assessment in the elderly: a gait abnormality rating scale and its relation to falls. Journal of Gerontolgy 1990;45(1):12–9. Woollacott M. Systems contributing to balance disorders in older adults. Journal of Gerontology 2000;55A(8):424–8. Woollacott M, Shumway-Cook A, Nashner LM. Ageing and posture control—changes in sensory organisation and muscular coordination. International Journal of Ageing and Human Development 1986;23(2):97–114.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 325–333 www.elsevier.com/locate/math

Original article

Effects of slouching and muscle contraction on the strain of the iliolumbar ligament Chris J. Snijdersa,, Paul F.G. Hermansa, Ruud Niesinga, Gert Jan Kleinrensinkb, Annelies Pool-Goudzwaarda a

Department of Biomedical Physics and Technology, Erasmus MC, University Medical Center Rotterdam, The Netherlands b Department of Neurosciences, Erasmus MC, University Medical Center Rotterdam, The Netherlands Received 27 October 2004; received in revised form 31 August 2005; accepted 5 March 2007

Abstract The study consisted of biomechanical modelling and in vitro experiments. The objective of the study was to ﬁnd a mechanical cause of acute low back pain (LBP) in everyday situations. The precise mechanism producing LBP is still under discussion. Most biomechanical studies link the concepts of stooped postures and buckling instability of the spine under high compressive load. No biomechanical model addresses situations with small or neglectable compressive spinal load. The proposed conceptual model describes strain on the iliolumbar ligaments (ILs) when slouching from standing upright. Delayed or absent recruitment of back muscles that protect against hyperkyphosis of the lumbar spine is a conditional factor. Erector spinae and multiﬁdus muscle forces are included, representing a bifurcation in back muscle force: one part acting on the iliac bones and one part acting on the sacrum. The multiﬁdus muscle action on the sacrum may produce nutation which can be counteracted by pelvic ﬂoor muscles, which would link back problems and pelvic ﬂoor problems. The effect of simulated muscle tension on the ILs and the L5-S1 intervertebral disc angle was measured using embalmed specimens. Forces were applied to simulate erector spinae and sacral part of multiﬁdus tension, bilateral up to 100 N each. Strain gauge sensors registered elongation of the ILs. Explorative biomechanical model calculations show that dynamic slouching, driven by upper body weight and (as an example) rectus abdominis muscle force may produce failure load of the spinal column and the ILs. The quasistatic test on embalmed specimens showed a signiﬁcant increase of IL elongation with simulated rectus abdominis muscle force. Adding erector spinae or multiﬁdus muscle tension eased the ILs. Sudden slouching of the upright trunk may create failure risk for the spine and ILs. This loading mode may be prevented by controlling loss of lumbar lordosis with erector spinae and multiﬁdus muscle force. r 2007 Elsevier Ltd. All rights reserved. Keywords: Low back pain; Multiﬁdus muscle; Pelvic ﬂoor; Iliolumbar ligament

1. Introduction The cause of low back pain (LBP) is often attributed to intolerable high intradiscal pressure. Use of the spinal compression model is often referred to for workload standards (Kumar, 1994) and is the starting point for spinal buckling instability models (Howarth et al., 2004) for lifting in stooped postures. The precise mechanism Corresponding author. Tel.: +31 10 408 73 68; fax: +31 10 408 94 63. E-mail address: [email protected] (C.J. Snijders).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.03.001

producing back sprain is, however, still under discussion. Therefore, we decided to explore a novel approach. In contrast to established biomechanical research we do not relate injury risk to forward trunk inclination, but take the unconstrained erect posture as a starting point. In a previous study (Snijders et al., 2004) we developed a biomechanical model on sitting with hyperkyphosis while leaning against a high backrest. For veriﬁcation of the model we measured in vitro stepwise backward tilt of the pelvis combined with forward ﬂexion of the spine. We found that during forward ﬂexion of the L5 vertebra the sacrum moved in the

ARTICLE IN PRESS 326

C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

opposite direction (counternutation). During the same test we measured (indirectly) elongation of the iliolumbar ligaments (ILs). The increase of strain on the IL by forward ﬂexion of L5 was similar to that reported earlier (Mu¨ller-Gerbl et al., 1988, Paul, 1989). Because patients suffering from acute LBP often present with pain at the site of the IL we decided to develop a model on sudden slouching. Starting point was the absence or delay of protective muscle force. A higher incidence of LBP was found in athletes showing delayed muscle reﬂex response on a quick force release in trunk ﬂexion, extension and lateral bending (Cholewicki et al., 2005). In continuation of our earlier biomechanical model on sitting we decided to model dynamic slouching of the upright trunk. The aim of the present study was to assess failure risk of the IL by means of explorative calculations (see Appendix A) and to measure in vitro if such risk could be prevented by back muscles. The following hypothesis was postulated: tension in the IL increases with forward ﬂexion of L5 and decreases by multiﬁdus and erector spinae muscle contraction.

2. Materials and methods 2.1. Materials Four embalmed specimens (age range 63–86 years) consisting of L4, L5 and pelvis with intact ligaments and intervertebral discs were loaded on a specially designed apparatus (see Fig. 1). 2.2. Methods An embalmed pelvis was placed upright with 12 degrees backward inclination of the tangent plane to the symphysis and the left and right spina iliaca anterior superior (Fig. 1). The upper part of L4 was attached to a vertical bar by means of a clamp with screws. Screws were inserted in the specimen and ropes were adjusted to simulate the rectus abdominis, the sacral part of the multiﬁdus and the lateral part of the erector spinae muscles. The site of muscle attachments was taken as an average of 5 cadavers and measured with respect to the axis of the spine. The distances were at the L4 level 11, 5.5 and 3 cm, respectively. The tension in the cords was increased with steps of 20 N. We used a ﬂexion moment of 22 N m. A smaller magnitude (10 Nm) was used in earlier studies (Mu¨llerGerbl et al., 1988; Paul, 1989; Yamamoto et al., 1990), resulting in an average strain (n ¼ 6) of 6.8% (SD 12%) in the ventral band of the IL (Paul, 1989). Strain of the IL was measured with sensors attached to the ventral band of each IL halfway between the transverse process of L5 and the ilium.

Fig. 1. Load test on embalmed specimen in the posture shown in Fig. 3B (Appendix A). The IL elongation is measured while ﬂexing the spine. The pelvis is rotated backward by means of cords representing the rectus abdominis muscles (RA). Additional cords are tightened simulating erector spinae (ES) or multiﬁdus (MM) muscles.

The movements of the ventral side of L5, sacrum and ilium were recorded with a 3-D videorecording system. 2.3. Instrumentation Partial upper body weight was simulated by means of a construction (75 newton ¼ 75 N ¼ ca. 7.5 kg) ﬁxed on top of L4. Pelvic inclination was adjusted with a spindle via a horizontal bar hinging with a vertical bar. Linear actuators (motor LA12-100-24-001, 0–500 N, Linak, Breda, the Netherlands) tightened cords (loading rate 2 mm/s), guided by pulleys, to simulate muscle forces with ﬁxed lever arms of these muscles with respect to the axis of the vertical bar. Tension in the cords was measured using custom- made strain-gauged force transducers (linearity 1.2%, 0–100 N). For the assessment of change in ligament elongation both ends of a custom-made sensor (U-shape bend strip) with straingauges (Mu¨ller-Gerbl et al., 1988; Paul, 1989) were attached to the IL by means of pins at both ends of the strip (linearity 0.7%, 0–5%). For 3-D videorecording with 2 CCD cameras of relative movements of bones, retroreﬂective markers were attached to the bones; accuracy was 0.1 degree (Keemink et al., 1991). 2.4. Data analysis The t-test for zero slope was applied with signiﬁcance level Po0.05. We were interested in the trend of measured quantities (increase or decrease), not in the absolute values.

3. Results Simulation of rectus abdominis force in the erect position of the embalmed specimens resulted in elongation of the IL, left and right (results of the right IL of

ARTICLE IN PRESS C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

every tested pelvis are in Fig. 2A). The IL was released by raising erector spinae or multiﬁdus force separately to 200 N (each side 100 N), with rectus abdominis force kept constant at 100 N (each side 50 N). Forward rotation of L5 with respect to the sacrum occurred when applying rectus abdominis force (Fig. 2B). The opposite occurred with the application of erector spinae and multiﬁdus force while keeping the rectus abdominis force at a constant 100 N. Rotation between the sacrum and ilium remained smaller than the error of measurement; this is attributed to stiffness of the embalmed specimens.

327

The slopes of the (non-linear) regression lines differed signiﬁcantly from zero slope over the whole loading range.

4. Discussion 4.1. Slouching The unconstrained upright standing posture was the starting point of this study on slouching. In this position back muscle activity is minimal or absent because the

Fig. 2. (A) The elongation (in %) of the anterior part of the right IL increases with simulated force (in N) in rectus abdominis muscles (RA). Additional force in erector spinae (ES) or multiﬁdus (MM) muscle with RA kept constant on 100 N results in ease of the IL. (B) Relation between average rotation (in degrees,+is forward) of L5 with respect to the sacrum and simulated muscle forces (in N). All four specimens show the same trend. Negative values of deformation are attributed to preload by the weight on top of L4.

ARTICLE IN PRESS 328

C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

Fig. 3. (A) Figure taken from an X-ray picture (male, age 25 years) in unconstrained standing. Low or almost absent dorsal back muscle activity. (B) Risk situation: sudden slouching from standing erect by backward tilt of the pelvis combined with forward ﬂexion of the spine (click-clack movement) and ﬂexion of hip and knee joints. Upper body weight (Fg) and (as an example) rectus abdominis muscle force (Fa) have lever arms g and a with respect to the middle of the disc L5-S1. Flexion of L5 puts strain on the IL (small circle). The downward displacement (h) of the upper body mass produces kinetic energy.

upper body weight is centred above the spine and above or just behind the hip joints. Therefore, during sudden and fast slouching, the response of back muscles to protect the change from lumbar lordosis into lumbar hyperkyphosis may be delayed. Muscle delay can exist after a period of prolonged stretch of dorsal ligaments and muscles (McGill and Brown, 1992; Solomonow et al., 1999; Dolan and Adams, 2001) and muscle fatigue (Adams et al., 2002). The explorative calculations in Appendix A show that failure load of the spine and the IL could be reached during sudden slouching. This change of form in the upright position is called the clickclack movement (Fig. 3, Appendix A) and is characterized by backward tilt of the pelvis combined with forward ﬂexion of the spine (Snijders, 1972; Snijders et al., 2004).

4.2. Strain on the IL The present study on the strain of the IL by spinal ﬂexion may provide an explanation for acute LBP in situations with neglectable axial spinal load. Newman (1952) states that slouching resembles the primitive reﬂex mechanism including limb movement to the centre of the body; he refers to this type of movement in patients who have had an acute attack of lumbago while shaving and he attributes the acute attack of pain to bending forward in a position of ﬂexion of the spine with the erector spinae muscles relaxed. This ﬂexionrelaxation mechanism was, however, revised by Andersson et al. (1996), who used measurements with ﬁne-wire electrodes to demonstrate that quadratus lumborum and deep lateral erector spinae were activated when the

ARTICLE IN PRESS C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

ﬂexion-relaxation phenomenon was present for the superﬁcial medial erector spinae muscles. Therefore, we do not refer to forward trunk inclination but rather to the absence of protective dorsal muscle activity while slouching in the upright posture. We suppose that this only occurs in humans. A human, when in a static standing posture, has his centre of gravity positioned just behind the hip joints, requiring low or absent dorsal muscle activity (Campbell and Loy, 2000). This enables the erect slouched posture to be adopted by humans, but not by apes. The bent-legged bipedal posture of the ape is the consequence of limited hip extension. Therefore, the centre of gravity is anterior to the hip joints requiring stabilizing dorsal muscle activity (Campbell and Loy, 2000). Strain of the ILs may occur in dynamic (see Appendix A) and in static situations. Deursen van et al. (2002) reported that LBP patients experience signiﬁcantly more pain provocation in slouched standing postures (e.g. vacuuming, brushing teeth and washing dishes) than in stooped postures. Sims and Moorman (1996) suggest that stress at the ligamento-osseous junction of the IL at the ilium, and stress in the innervated IL, can trigger pain from both tissues, which is the premise behind local injections of anaesthetics. Although controversial, evidence exists for the signiﬁcant role of the IL in LBP (Boebel von, 1961; Gutmann et al., 1981; Paul, 1989; Sims and Moorman,

329

1996; Deursen van et al., 2002). Of interest to our study is that, of the 19 clinical parameters addressed during medical examination of a group of children reporting LBP, only one was signiﬁcantly more prevalent: i.e. pain on palpation at the insertion site of the IL on the iliac crest (Gunzburg et al., 1999). Further support for our model on slouching can be found in the loss of thoracolumbar curvature in pregnant women (Snijders et al., 1976; Dumas et al., 1995), in astronauts during ﬂight (Thornton et al., 1977; Snijders et al., 2005) and in individuals suffering from psychosocial distress. Our study on the slouched sitting was reported in a previous article (Snijders et al., 2004), which included measurements on fresh human specimen. For the present study we used a small sample size of embalmed specimen which, however, resulted in similar load-deformation patterns. This is in agreement with expectation because the orientation of the IL is, although individual differences are reported in literature (Fujiwara et al., 2000), in all cases in the direction of loading. 4.3. Protective muscle contractions The present study provides preliminary information on the effects of muscle contraction on the strain of the IL. It shows that the lateral erector spinae muscles and the sacral part of the multiﬁdus muscles are capable of

Fig. 4. (A) Bifurcation in back muscle force: one part acting on the ilium (Fe) and one part acting on the sacrum (Fm) with their reaction force in the spine (Fre+Frm). (B) Back muscle force (Fe) tends to tilt the innominate forward. This can be counteracted by hip muscle force (Fh). (C) The sacrum tends to rotate forward (nutation) by multiﬁdus muscle force (Fm) which can be counteracted by pelvic ﬂoor muscle force (Fp) according to the agonist-antagonist action. (The weight force from the upper body is not drawn.)

ARTICLE IN PRESS 330

C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

Fig. 5. The relation between forward rotation of L5 and backward rotation of the sacrum (counternutation). Fig. 4 represents movement in the opposite direction (Snijders et al., 2004). Waste of multiﬁdus force or hyperactivity may result in dysfunction of the pelvic ﬂoor muscles resulting in e.g. urine incontinence, frequency, constipation or sexual complaints (Pool-Goudzwaard et al., 2005). Iliolumbar ligament (1), axis of rotation of sacroiliac joint (2), sacroiliac joint surface (3), pelvic ﬂoor muscles (4), anus (5), vagina (6), urethra (7), pubic symphysis (8).

protecting the L5-S1 disc against hyperﬂexion and the IL against excessive tension, i.e. by hollowing of the lower back. This followed from load tests with the use of embalmed specimen, which implies stiffer material than in vivo. We expect, however, that the relation between the applied load on L5 and the IL elongation will show the same trend in the in vivo situation. The rotation between the sacrum and ilium remained smaller than the error of measurement, which is attributed to stiffness of the embalmed specimens. For the effect of back muscle force on the sacroiliac joints we made a distinction between the effect of erector spinae and multiﬁdus muscle (Fig. 4, Appendix A). This is of interest because the sacral part of the multiﬁdus muscle has the unique ability to produce the isolated action of extension of the L5-S1 intervertebral disc together with forward rotation of the sacrum with respect to the ilium (nutation). This suggests coactivation with pelvic ﬂoor muscles according to the agonist-antagonist action about the sacroiliac joints (Fig. 5, Appendix A). Therefore, it may be expected that LBP due to strain on the IL may be related to pelvic

Fig. 6. Comprehensive biomechanical model on sacroiliac joint stability by self-bracing with transversely oriented abdominal muscles (transversus abdominis), back muscles (sacral part of multiﬁdus) and pelvic ﬂoor muscles (coccygeus).

ﬂoor problems. Pool-Goudzwaard et al. (2005) demonstrated a signiﬁcant increase of pelvic ﬂoor activity in a population of postpartum low back and pelvic pain patients accompanied by frequency, urgency and stress incontinence. The present study has led to further understanding of the biomechanical relationship incorporating the ‘‘low back and pelvic pain’’ along with the concept of selfbracing with abdominal muscles (transversus abdominis), back muscles (sacral part of the multiﬁdus) and pelvic ﬂoor muscles in one comprehensive biomechanical model (Snijders et al., 1998, 2005; Richardson et al., 2002) (Fig. 6, Appendix A). This model supports conservative treatments which promote lumbar lordosis (Gard et al., 2000) and train the recruitment of multiﬁdus muscles in co-contraction with the transversus abdominis muscles (Hides et al., 1994, 2001; Richardson et al., 2002, 2004). Future in vivo research is required to further substantiate the biomechanical model on slouching. It is, however, an ethical problem to provoke back sprain in healthy subjects.

5. Conclusions This study lends credibility to the idea that back sprain can be the result of slouching in the upright

ARTICLE IN PRESS C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

posture. The IL is designated as a possible source of pain, because this ligament was strained by ﬂexion of L5. In line with these measurements were the explorative calculations pointing to the possibility of injury to the IL and spine by slouching. Ease of the IL was obtained by simulated back muscle force. Conditional for the conceptual model on back sprain is the delay or absence of muscle recruitment which has to prevent the ‘‘clickclack movement’’, i.e. backward tilt of the pelvis combined with forward ﬂexion of the spine. This suggests that the attention for lifting in a stooped posture may be diverted to dynamic or static slouching with the trunk in the upright position as the main cause of LBP in daily life activities.

Acknowledgement The authors would like to thank D. McCook, M. de Groot, C.W. Spoor, F.C. Velkers and E. Vlaanderen for their valuable contributions. Financial support was provided by the Anna Foundation.

Appendix A A.1. Conceptual biomechanical model The starting point for the following model is that the upper body weight is centred above or just behind the hip joints in standing. This allows for equilibrium in the relaxed posture with small or without muscle action at the dorsal side of the spine necessary prior to sudden movement into a slouch from standing erect. Sudden slouching can be seen as a click-clack movement (Snijders, 1972; Snijders et al., 2004) from standing (Fig. 3). This click-clack movement of the pelvis into a sudden postural slouch involves a combination of backward tilt of the pelvis and forward ﬂexion of the spine into a lumbar kyphosis. In the following, rough calculations are made based on the conceptual model that injury risk can occur when certain conditions are fulﬁlled, such as (a) standing or squatting with the trunk upright and then (b) sudden slouch/ﬂexion, involving a click-clack movement with bending of hips and knees. Protection against hyperﬂexion is normally provided by dorsal muscles (Adams and Hutton, 1986), but in the case of wasted or delayed muscle response or in muscle inhibition due to pain/ dysfunction post injury, it is assumed that the strain is allowed to increase until the disc and/or the IL tears. The spine may also yield at a level above L5. The possibility of injury is based on the following assessment. Abrupt head-down tilt and curving of the trunk driven by upper body mass and, as an example,

331

rectus abdominis muscle force, produces a ﬂexion moment of force with respect to the spine. The rectus abdominis muscle can produce a considerable moment of force about the lumbar level of the spine. Isometric rectus abdominis force is estimated to be up to 1000 N (total of both sides for a male). Here we used a physiological cross-section of 18 cm2 (SD 4.6) (Marras et al., 2001) and following McGill et al. (1988) we used about 50 N/cm2. A lever arm of 10.3 cm (SD 2.07) (McGill et al., 1988) results in a bending moment of about 100 Nm. This is well above 60 Nm, which is considered injurious to the spine (Dolan and Adams, 1993; Adams et al., 2002). Injury risk increases after fatigue loading of the posterior structures (Adams et al., 2002) and after sustained ﬂexion which causes creep (McGill and Brown, 1992) and reduces motion segment resistance to bending by 42% in just 5 min, and by 67% in 1 h (Adams et al., 2002). Applying 100 N m to our model would result in possible ruptures in the intervertebral regions of the lower lumbar levels (Dolan and Adams, 1993; Adams et al., 2002). To assess strain on the IL, we consider as a ﬁrst approach the extreme situation without resisting bending moment from the L5-S1 disc against 100 N m in the above situation. Then, with reference to our earlier study (Snijders et al., 2004) a force is applied in dorsal direction at the L5-S1 level and in ventral direction on the IL of 4000 N. If this force is divided over two IL, and taking into account an angle of 201 between IL and coronal plane (Paul, 1989; Fujiwara et al., 2000), this would result in 2000:0.3420 ¼ 5850 N tensile force. A cross-section of about 50 mm2 of the ventral and dorsal part of the ligament (Hanson and Sonesson, 1994) results in 117 megapascal ¼ 117 MPa ¼ 117 MN/m2 tensile stress. This is well above the tensile strength of the anterior longitudinal ligament at the L4-L5 level (36.9 MPa), the anterior cruciate ligament (45.7 (SD 19.5) MPa) and the coracoacromial ligament (46.9 (SD 30.7) MPa) (Abe´ et al., 1996). Moreover, the contribution of 500 N upper body weight (Snijders, 2001) is 22 N m bending moment (see proportion of lever arms in Fig. 3B) resulting in 26 MPa stress in the IL. Sudden slouching with ﬂexion of hip and knee joints imply acceleration of body mass and conversion of potential energy into kinetic energy. As an estimation of the order of magnitude, this energy can be about 500 N 0.5 (part of upper body weight above L5) (Snijders, 2001) 0.1 m (h in Fig. 3 measured from a subject with body height 178 cm) ¼ 25 J. The dorsal part of the IL may yield by 1.5 J (ligament volume 0.46 cm3 3.3 Nm/cm3 strain energy density taken from the anterior cruciate ligament) (Hanson and Sonesson, 1994; Abe´ et al., 1996). At the same time, work may be generated by longitudinal and oblique abdominal muscles adding kinetic energy. We assume that back

ARTICLE IN PRESS 332

C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333

muscles do not oppose this build up of kinetic energy in cases of delayed activation. At the end of the click-clack movement the kinetic energy must be absorbed by deformation of soft tissue, including the ligaments around the SIJ. Therefore, dynamics increase the loading mode illustrated in Fig. 3. Additional force and energy may be produced by e.g. pulling an unexpected load or squatting. Although not veriﬁed by experiments, these rough calculations provide a plausible explanation for the occurrence of microtrauma in the IL, particularly in sudden slouching. Ruptures in the IL (Gutmann et al., 1981) and calciﬁcation of the IL at the bony insertion (Boebel von, 1961) have been reported. A.2. Bifurcation in back muscle force and the relation between multifidus and pelvic floor muscles At the dorsal side of the lumbopelvic region the resultant back muscle force can be divided into two components (Fig. 4A), the ﬁrst acting on the ilia (Fe) and the second acting on the sacrum (Fm). In Fig. 4B the erector spinae muscle force (Fe) acts on the left ilium with the reaction force (Fre) acting on the sacroiliac joint. These forces produce the moment of force Fe e which results in forward ﬂexion of the innominate. Muscles attached to the sacrum produce a completely different effect. In Fig. 4C the multiﬁdus muscle force (Fm) with the reaction force Frm produce the moment of force Fm m which tends to forward ﬂexion of the sacrum (nutation). Because Fm is equal to Frm, the sacroiliac joint does not experience shear in the longitudinal direction of the spine. In the case of erector spinae muscle force, however, the reaction force in the spine (Fre) has to cross the sacroiliac joint (shear) to produce equilibrium for the innominate. The rotary effect of the multiﬁdus force (Fm) about the sacroiliac joint can be counteracted by pelvic ﬂoor muscle force (Fp) which produces a moment of force Fp p. Equilibrium of moments of force about the sacroiliac joints follows from Fm m ¼ Fp p which represents an agonist-antagonist action. This relation between muscle forces can be translated to movements. Fig. 5 illustrates forward ﬂexion of L5, resulting in strain on the ILs, and backward rotation of the sacrum (counternutation) according to our earlier study with the use of human bodies prior to embalming (Snijders et al., 2004). The opposite is that backward rotation of L5 goes together with forward rotation of the sacrum (nutation), which according to Fig. 2, would ease the IL. Fig. 6 shows the deep muscle corset for the control of sacroiliac joint stability. We suggest that each of the respective muscles may be considered as primary mover of the sacroiliac joints, depending on speciﬁc demands, albeit that the movements are minute.

References Abe´ H, Hayashi K, Sato M. Data book on mechanical properties of living cells, tissues, and organs. Tokyo: Springer; 1996. Adams MA, Hutton WC. Has the lumbar spine a margin of safety in forward bending? Clinical Biomechanics 1986;1:3–6. Adams MA, Bogduk N, Burton K, Dolan P. The biomechanics of back pain. London: Churchill Livingstone; 2002. Andersson EA, Oddsson LIE, Grundstro¨m H, et al. EMG activities of the quadratus lumborum and erector spinae muscles during ﬂexion-relaxation and other motor tasks. Clinical Biomechanics 1996;11(7):392–400. von Boebel R. Anatomische Untersuchungen am Ligamentum iliolumbale; ein Beitrag zur Kenntnis des lumbosakralen U¨berganges der Wirbelsa¨ule beim Menschen. Zeitschrift fuer Orthopaedie 1961;95(2):131–9. Campbell BG, Loy JD. Humankind emerging, 8th ed. Boston: Allyn and Bacon; 2000. p. 100–02; 192–9; 251–3. Cholewicki J, Silﬁes S, Shah R et al. Delayed trunk muscle reﬂex responses increase the risk of low back injuries. 32nd Annual Meeting of the International Society for the Study of the Lumbar Spine, New York, 2005. p. 2–3. van Deursen LLJM, Snijders CJ, Patijn J. Inﬂuence of daily life activities on pain in patients with low back pain. Journal of Orthopaedic Medicine 2002;24(3):74–6. Dolan P, Adams MA. Inﬂuence of lumbar and hip mobility on the bending stresses acting on the lumbar spine. Clinical Biomechanics 1993;8:185–92. Dolan P, Adams MA. Recent advantages in lumbar spinal mechanics and their signiﬁcance for modelling. Clinical Biomechanics 2001; 16(1):S8–S16. Dumas GA, Reid JG, Wolfe LA, Grifﬁn MP, McGrath MJ. Exercise, posture, and back pain during pregnancy. Clinical Biomechanics 1995;10(2):98–103. Fujiwara A, Tamai K, Yoshida H, et al. Anatomy of the iliolumbar ligament. Clinical Orthopaedics and Related Research 2000;380: 167–72. Gard G, Gille KA, Degerfeldt L. McKenzie method and functional training in back pain rehabilitation; a brief review including results from a four-week rehabilitation programme. Physical Therapy Review 2000;5:107–15. Gunzburg R, Balague F, Nordin M, et al. Low back pain in a population of school children. European Spine Journal 1999;8(6): 439–43. Gutmann G, Sassendorf B, Biedermann H. Sklerosierungstherapie im Bereich der Wirbelsa¨ule. In: Die Wirbelsa¨ule in Forschung and Praxis, 1981. p. 167–70. Hanson P, Sonesson B. The anatomy of the iliolumbar ligament. Archives of Physical Medicine and Rehabilitation 1994;75(11): 1245–6. Hides JA, Jull GA, Richardson CA. Long-term effects of speciﬁc stabilizing exercises for ﬁrst-episode low back pain. Spine 2001; 26(11):E243–8. Hides JA, Stokes MJ, Saide M, et al. Evidence of lumbar multiﬁdus muscle wasting ipsilateral to symptoms in patients with acute/ subacute low back pain. Spine 1994;19:165–72. Howarth SJ, Allison AE, Grenier SG, Cholewicki J, McGill SM. On the implications of interpreting the stability index: a spine example. Journal of Biomechanics 2004;37:1147–54. Keemink CJ, Hoek van Dijke GA, Snijders CJ. Upgrading of efﬁciency in the tracking of body markers with video techniques. Medical and Biological Engineering and Computing 1991;29: 70–4. Kumar S. Lumbosacral compression in maximal lifting efforts in sagittal plane with varying mechanical disadvantage in isometric and isokinetic modes. Ergonomics 1994;37(12):1975–83.

ARTICLE IN PRESS C.J. Snijders et al. / Manual Therapy 13 (2008) 325–333 Marras WS, Jorgensen MJ, Granata KP, et al. Female and male trunk geometry: size and prediction of the spine loading trunk muscles derived from MRI. Clinical Biomechanics 2001;16(1): 38–46. McGill SM, Brown S. Creep response of the lumbar spine to prolonged full ﬂexion. Clinical Biomechanics 1992;7:43–6. McGill SM, Patt N, Norman RW. Measurement of the trunk musculature of active males using CT scan radiography: implications for force and moment generating capacity about the L4/L5 joint. Journal of Biomechanics 1988;21(4):329–41. Mu¨ller-Gerbl M, Paul HJ, Putz R. Analysis of function of iliolumbar ligament by means of strain gauges. Journal of Biomechanics 1988; 21(10):868. Newman PH. Sprung back. Journal of Bone and Joint Surgery 1952; 34B(1):30–7. Paul HJ. Experimentelle Untersuchung zur Funktion des Ligamentus iliolumbale. Thesis, Albert Ludwigs-Universita¨t Freiburg im Breisgau; 1989. Pool-Goudzwaard AL, Slieker ten Hove MCPH, Vierhout ME, Mulder PGH, Pool JJM, Snijders CJ, et al. Relations between pregnancy related low back, pelvic ﬂoor activity and pelvic ﬂoor dysfunction. International Urogynaecology Journal 2005;16(6): 468–74. Richardson CA, Hodges PW, Hides JA. Therapeutic exercise for lumbopelvic stabilization: a motor control approach for the treatment and prevention of low back pain. 2nd ed. London: Churchill Livingstone; 2004. Richardson CA, Snijders CJ, Hides JA, et al. The relationship between the transversely oriented abdominal muscles, sacroiliac joint mechanics and low back pain. Spine 2002;27(4):399–405.

333

Sims JA, Moorman SJ. The role of the iliolumbar ligament in low back pain. Medical Hypotheses 1996;46(6):511–5. Snijders CJ. The form of the spine related to the human posture. Agressologie 1972;13B:5–14. Snijders CJ. Standing, sitting, and lying. In: Nordin M, Frankel VH, editors. Basic Biomechanics of the musculoskeletal system. New York: Lippincott Williams & Wilkins; 2001. p. 420–36. Snijders CJ, Richardson CA, Pool-Goudzwaard AL, Hides JA. Low back pain in microgravity; causality and countermeasures. Graz, Austria: Abstract, 15th IAA humans in space symposium, 2005. Snijders CJ, Ribbers MTLM, de Bakker JV, et al. EMG recordings of abdominal and back muscles in various standing postures: validation of a biomechanical model on sacroiliac joint stability. Journal of Electromyography and Kinesiology 1998;8:205–14. Snijders CJ, Seroo JM, Snijder JGN, et al. Change in form of the spine as a consequence of pregnancy. Ottawa: Digest 11th ICMBE, 1976. p. 670–1. Snijders CJ, Hermans PFG, Niesing R, Spoor CW, Stoeckart R. The inﬂuence of slouching and lumbar support on iliolumbar ligaments, intervertebral discs and sacroiliac joints. Clinical Biomechanics 2004;19(4):323–9. Solomonow M, Zhou BH, Baratta RV, et al. Biomechanics of increased exposure to lumbar injury caused by cyclic loading: part 1. Loss of reﬂexive muscular stabilization. Spine 1999;24(23): 2426–34. Thornton WE, Hofﬂer GW, Rummel JA. Anthropometric changes and ﬂuid shifts. In: Biomedical results from Skylab, NASA; 1977. Yamamoto I, Panjabi MM, Oxland TR, et al. The role of the iliolumbar ligament in the lumbosacral junction. Spine 1990;15(11): 1138–41.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 334–340 www.elsevier.com/locate/math

Original article

Real-time morphologic changes of the iliotibial band during therapeutic stretching; an ultrasonographic study$ Hsing-Kuo Wanga, Tiffany Ting-Fang Shihb,c, Kwan-Hwa Lind, Tyng-Guey Wange, a

Sports Physiotherapy Laboratory, School and Graduate Institute of Physical Therapy, College of Medicine, National Taiwan University, Taiwan, ROC b Department of Medical Image, National Taiwan University Hospital, Taiwan, ROC c Department of Surgery, College of Medicine, National Taiwan University, Taiwan, ROC d School and Graduate Institute of Physical Therapy, College of Medicine, National Taiwan University, Taiwan, ROC e Department of Physical Medicine & Rehabilitation, National Taiwan University Hospital, School of Medicine, National Taiwan University, 7 Chung Shan South Road, Zhongzheng District, Taipei City 100, Taiwan, ROC Received 28 August 2006; received in revised form 12 January 2007; accepted 2 March 2007

Abstract The purpose of this study was to evaluate the utility of ultrasonography (US) in determining the morphological changes of the iliotibial band (ITB) with the modiﬁed Ober maneuver. Forty-four subjects (23 men and 21 women, mean age (7 SD), 24.774.7 years) who had no previous history of lower back, gluteus, hip or knee pain and satisﬁed additional inclusion criteria were recruited. Twenty out of the 44 subjects were initially examined by both MRI and US for measurement conﬁrmation. Band width of the left ITB (the measures of which were highly correlated between techniques) was then assessed for these 44 subjects by US with the modiﬁed Ober maneuver in three gradually increased hip adduction positions; neutral, adducted and adducted with weight in these 44 subjects. In addition, examiner reliability was assessed by conducting duplicate measurements in 20 randomly chosen subjects. Results demonstrated that measures of band width, but not thickness, were highly correlated between MRI and US (po0:001, r ¼ 0:850). Signiﬁcant reductions in band width were observed between the three positions with the modiﬁed Ober maneuver (po0:001). Intratester reliability was high (intraclass correlation coefﬁcient (ICC) ¼ 0.86–0.94). Band width changes indicated that the ITB was subjected to a signiﬁcant stretching force during hip adduction. We conclude that US is a reliable means to directly assess the real-time effects of stretching exercises. r 2007 Elsevier Ltd. All rights reserved. Keywords: Iliotibial band; Stretch; Ultrasonography

1. Introduction Therapeutic stretching is commonly prescribed for patients with soft tissue shortness or joint tightness. Physical therapists often employ a static stretch while gradually moving the joint in an antagonistic direction $ This research is completed in School and Graduate Institute of Physical Therapy, College of Medicine, National Taiwan University, Floor 3, No.17, Xuzhou Rd., Zhongzheng District, Taipei City 100, Taiwan, ROC. Corresponding author. Tel.: +886 2 23123456x7588; fax: +886 2 23832834. E-mail address: [email protected] (T.-G. Wang).

1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.03.002

to increase tissue ﬂexibility and joint range of motion (ROM) (Bandy and Irion, 1994). In addition to ROM and ﬂexibility, the acute effects of static stretching on neural and muscular properties can be measured immediately after the exercise (Bandy and Irion, 1994; Fowles et al., 2000; Behm et al., 2004). These outcomes have usually been described after the stretch has been performed, but not during the exercise. To data, no study has used ultrasonography (US) to objectively report real-time responses of target tissues, such as the iliotibial band (ITB), to therapeutic stretching. The effects of real-time ITB stretching have been reported from a subjective standpoint (such as by

ARTICLE IN PRESS H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

palpation) (Gajdosik et al., 2003), however, qualitative measurements were not made. Other studies have reported real-time measurements by way of cinemography or goniometry, techniques which lack speciﬁcity (Fredericson et al., 2002; Reese and Bandy, 2003). Therefore, a valid and quantiﬁable method is needed to determine real-time target tissue responsiveness under stretch. Such a method could enhance the validity of therapeutic stretching in evidence-based medicine, and further our understanding of how real-time soft tissue responses affect consequent acute muscular responses. Recent real-time, noninvasive laboratory studies using US have demonstrated morphological changes in soft tissues through different joint positions. In one study, the gastrocnemius tendon increased in length when the leg was moved from an ankle-neutral position to dorsiﬂexion (Maganaris and Paul, 2002). In other studies, the patellar tendon and posterior cruciate ligaments increased in length when the knee was repositioned from full extension to ﬂexion (Sheehan and Drace, 2000; Li et al., 2004). In addition to elongation, the main morphologic changes observed in tendon or ligaments during a longitudinal stretch are simultaneously decreased cross-sectional area, width, or thickness, as conﬁrmed by cadaveric studies of ankle and shoulder (Lewis and Shaw, 1997; Costic et al., 2003). Few studies have analyzed real-time morphological changes of soft tissue in clinical practice, or used these values to assess the real-time effects of therapeutic stretching. Both magnetic resonance imaging (MRI) and US are high resolution, noninvasive techniques that can be used to monitor real-time morphological changes in soft tissues during stretch. In comparison to US, MRI provides higher contrast for soft tissue but has real-time imaging limitations. US now is widely used by physical therapists for both education and research (Ferreira et al., 2004; Wang et al., 2005). Ultrasound has also been used to measures features of the ITB (Bonaldi et al., 1998; Goh et al., 2003) which is often stretched using the modiﬁed Ober maneuver (Fredericson et al., 2000). In this maneuver, the ITB is initially stretched and tightened when the ipsilateral hip is adducted 4101. A further hip adduction angle is created by applying pressure on the lateral side of the knee while in this hip position if a greater ITB stretch is required (Kendall et al., 1970). The purpose of this study was to determine utility of US in assessing real-time morphological changes in the ITB following stretch as initiated using the modiﬁed Ober maneuver. Toward this end, we performed two experiments. In the ﬁrst experiment, morphological characteristics were measured using two imaging methods: MRI and dynamic ultrasound. The two methods were subsequently compared to level of correlation and agreement, using MRI as the reference method and

335

dynamic US as the comparison method. The aim of the ﬁrst experiment was to determine whether the two imaging methods correlate sufﬁciently for the US imaging to be used as a valid measurement in clinical setting. In the second experiment, our aims were to determine morphological changes during therapeutic stretching by US imaging and to assess intratester measurement reliability of the method.

2. Materials and methods 2.1. Subjects Our institutional review board approved this study, and all subjects provided informed consent. From August 2005 to April 2006, we recruited 50 healthy volunteers through the student center and the department of Rehabilitation Medicine at a university hospital. Inclusion criteria were as follows: subjects (1) were younger than 35 years of age; (2) had no prior history of lower back, gluteus, hip or knee pain that caused them to seek medical help one year prior to the study; (3) had a leg-length discrepancy (distance from the anterior superior iliac spine to the superior surface of the most prominent aspect of the medial malleolus) equal to or less than 1.5 cm; (4) had no genu varum with tibiofemoral angle of less than 41; (5) had no functional overpronation of the foot arch (the angle formed between the distal medial malleolus, the navicular tuberosity, and the ﬁrst metatarsal head less than or equal to 901). With respect to criteria (2)–(4), all of these criteria are considered risk factors predisposing an asymptomatic individual to develop irritation or tightness of the ITB (Nishimura et al., 1997; Bonaldi et al., 1998; Muhle et al., 1999). Subjects with these risk factors were excluded in order to validate the results of this study in healthy ITB. Note that these assessments regarding leg-length discrepancy, genu varum and foot arch were conducted on subjects in a standing position using a series of bilateral anthropometric and goniometric measurements and a standardized setting (Yamakoshi Seisakusho Co., Tokyo, Japan) by authors of this study (two board certiﬁed physiotherapists). During the MRI and US measurements, subjects were excluded from the study if: (1) the MRI indicated tendonitis or peritendinous edema, i.e. there was abnormal signal changes existed in the surrounding (ﬂuid collection) or within (swelling) the tendinous part of the ITB (Nishimura et al., 1997; Muhle et al., 1999); (2) US revealed a hypoechoic area of the ITB or joint ﬂuid located deep with respect to the ITB and superﬁcial to the lateral femoral condyle (Bonaldi et al., 1998); (3) the US measurements obtained with the modiﬁed Ober maneuver demonstrated a positive response consistent with a restricted ITB (i.e. a hip adduction angle of less

ARTICLE IN PRESS 336

H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

than 101). In total, six of the original 50 subjects were excluded (three each for criteria 2 and 3). Data were analyzed from 44 subjects (23 men and 21 women, mean age (7 SD), 24.7 7 4.7 years; height, 173.2 7 7.4 cm; weight, 65.1 7 8.5 kg). 2.2. Procedures and outcome measures Both MRI and US were used to characterize the morphology and measure the thickness and width of the tendinous part of the ITB. We assumed that the ITB is the target tissue for a longitudinal stretch by modiﬁed Ober maneuver, and that morphological changes are mostly due to associated structural strain. MRI and follow up US were conducted on the left knee, with subjects in a supine position, and the knees in a relaxed and extended position. Measurements were performed on the left knee due to the conﬁguration of the US theatre. MRI measurements were made using a 1.5-T Magnetom Sonata unit (Siemens, Malvern, PA, USA) with an extremity surface coil. The tendinous part of the ITB was identiﬁed by palpation (Fairclough et al., 2006) and a cod liver oil capsule was placed over it at an axial level corresponding to the superior border of patella (to act as a reference mark visible in the MR scan, see Fig. 1A). The routine MRI protocol at our hospital is as follows: fast spin-echo T1-weighted imaging (repetition time/echo time ¼ /19 ms, 800 ms/ 95 ms, echo train length ¼ 3) and fat-saturated fast spinecho proton density-weighted imaging (repetition time/ echo time ¼ 4000 ms/19 ms, echo train length ¼ 10) in the sagittal plane, and gradient-echo T2-weighted

imaging (repetition time/echo time ¼ 600 ms/15 ms, ﬂip angle ¼ 201) in the coronal plane. These sequences were performed to facilitate detection of peripheral lesions around the ITB or below the subcutis (exclusion criteria 1). Imaging parameters also included a 256 192 matrix, 3-mm section thickness with a 0.5 or 1 mm intersection gap, and a 16-cm ﬁeld of view for all sequences. Subjects were transferred by wheelchair, without moving their knees, to the US theatre after the MRI scan. The US measurements were made using a HDI 5000 imaging unit (Philips Medical Systems, Bothell, WA, USA), with a 12–5 MHz linear-array transducer. Subjects were in the same supine position as for MRI scans, and the cod liver oil capsule was removed. US measurements on the ITB were then taken in a transverse scan at the same location where the cod liver oil capsule had been placed. B mode and a view of near ﬁeld (depth of view: 2 cm) with focus depth between 0.5 and 1.0 cm were set when measuring the ITB (Fig. 1B). The width and thickness of the tendinous part of the ITB were ascertained by caliper, using computerized distance measurements from both scans. The measurements were made by double blinded examiners, one of whom was a board certiﬁed musculoskeletal ultrasonographer and the other a radiologist. The musculoskeletal ultrasonographer measured the ITB width and thickness with HDI 5000 and the radiologist measured the ITB with a Sonata unit, without knowing the results of the other measurement. Subjects were then turned into a left up side lying position for following real-time stretch effects and reliability tests. These measurements were conducted

Fig. 1. A 22-year-old man with asymptomatic left knee: (A) axial fat-saturated spin echo proton density MR image shows the location of the tendinous part iliotibial band (arrowhead). * Indicate the cod liver oil capsule. (B) Ultrasound image of the same knee. The tendinous part iliotibial band is identiﬁed by calipers.

ARTICLE IN PRESS H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

Fig. 2. Ultrasonographic measurements on the iliotibial band conducted with Modiﬁed Ober’s test.

with the modiﬁed Ober maneuver; i.e. the subjects were lying on their right side with their shoulders and pelvis perpendicular to the examining table. In addition, their right hip and knee were ﬂexed to ﬂatten the lower back. The left knee was maintained in a fully extended position (Fig. 2). Under these conditions, three positions (neutral, adducted, and adducted with weight) were employed to gradually increase the left hip angle with the modiﬁed Ober maneuver while measuring ITB morphology. Operational parameters (scan location, view depth and focus range) were identical to those used in the initial scan. Separate examiners assessed subject stabilization, hip angle, and US measurements. For the neutral position, the examiner (a physiotherapist) pushed just below the left anterior aspect of the superior ilium cephalady of the subject to stabilize the pelvis, preventing pelvic lateral and anterior tilt. Furthermore, this examiner also grasped just below the subject’s left knee, ﬁrst abducting, then hyperextending the left hip to keep it from internal rotation and ﬂexion during measurements (Fig. 2). A second examiner (also a physiotherapist) placed an inclinometer (AcuAngle, Middletown, USA), which had markings in 11 increments, over the lateral condyle of the femur to conﬁrm that the hip adduction angle was neutral (i.e. 01). The third examiner (an ultrasonographer) positioned the transducer perpendicular to the skin surface in such a manner as to avoid compressing the skin and underlying tissues. Ultrasonographic measurements of the width were always measured at the suprapatellar level of the ITB ﬁrst in the neutral and then in the adducted hip position. For the adducted hip position, the support that had been applied to the subject’s left knee in the neutral position was removed by grasping the left ankle lightly but with enough tension to keep the subject’s hip from undergoing ﬂexion and internal rotation. The test knee

337

was allowed to drop (adduct) by gravity into the available range of adduction until hip movement ceased. The examiner then stabilized the subject’s pelvis and the leg was lowered in a controlled manner so that the ultrasonographer could retain a clear image of the ITB. The end angle of hip adduction was measured by placing the inclinometer on the lateral condyle of the femur after the ultrasonographic measurements were recorded. Each subject was then instructed to relax and return the hip to its neutral position with stabilization by the examiner. To apply pressure and generate a greater hip adduction angle than that of the adducted position, a 3 kg sand bag was suspended on the distal end of the subject’s left knee to further stretch the ITB (position of adducted with weight). The examiner stabilized the subject’s pelvis and controlled the dropping speed of the knee joint using the same technique as for the adducted hip position. The left knee was allowed to stretch and drop by gravity as aided by the weight in the available range of adduction until hip movement stopped. The ITB width and the end angle of hip adduction were measured using US and the inclinometer. Each position was repeated three times with 5 min intervals, i.e. three measurements were taken in neutral, then three in adduction, then three in adduction plus weight. Average width and hip angle were recorded. To rule out confounding effects during the US measurements, we conﬁrmed that there was no signiﬁcant myoelectric activity of the upper and lower ﬁbers of gluteus maximus, gluteus medius and tensor fascia latae muscles using an eight-channel surface electromyograph (TSD150, Biopac Systems Inc., CA, USA). To assess intratester test-retest reliability, 20 of the 44 subjects were randomly chosen to undergo a repeat of the same protocol with the same investigators 24 h after the ﬁrst neutral position measurements. 2.3. Statistical analysis Variables from the US and MRI measurements included width and thickness of the tendinous part of the ITB. A Kolmogorov–Smirov test was used to evaluate if all data for all variables were normally distributed. The mean differences between these two techniques were compared using a paired t test (with 95% conﬁdence limits). Pearson bivariate correlation test was used to assess correlation of measurement results between the two techniques. One-way analysis of variance (repeated measures) with Duncan’s multiple range post hoc test was used to analyze differences in the width of the ITB in the three hip positions for the modiﬁed Ober maneuver. Mean differences in hip angle between the hip adducted and stretching positions were compared by paired t test. The intratester ICC and standard error of measurement were calculated to

ARTICLE IN PRESS H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

338

estimate the reliability of the US measurements. All data were analyzed using SPSS statistical software (SPSS 11.0 version; SPSS, Inc., Chicago, IL). Differences were considered signiﬁcant when po0:05.

3. Results On the Kolmogorov–Smirov test, the data for all variables were shown to be normally distributed. Morphologic parameters of US and MRI, including width and thickness, for the tendinous part of the ITB are presented in Table 1. No signiﬁcant difference existed between width determinations measured by MRI and US (p ¼ 0:375, 95% conﬁdence interval (CIs) lower: 0.0452; upper: 0.62022). A strong and signiﬁcant correlation was observed between the two techniques for this measure (r ¼ 0:850, po0:001). Results of US and MRI for ITB thickness demonstrated a signiﬁcant difference (po0:001, 95% CIs lower: 0.28815; upper: 0.57452). No correlation existed between these two imaging methods (p40:05). Therefore, only width measurements obtained by US agree and correlate sufﬁciently to those obtained by MRI. Width measurements were examined in the hip adduction studies.

Signiﬁcant reductions in the band width were observed for the modiﬁed Ober maneuver from neutral position to either adducted or adducted with weight position (Table 2, po0:001). Furthermore, differences in ITB width and hip angle between adducted and stretching position were signiﬁcant (p ¼ 0:004 and po0:001, respectively). The intratester ICC for measurement of width in the modiﬁed Ober maneuver ranged from 0.94 to 0.86 with a standard error of measurement of 0.2 or 0.3 mm (Table 2).

4. Discussion This study aimed to assess the utility of US for evaluating the real-time effects of therapeutic stretch. To do so, we ﬁrstly compared MRI and US measurements of the width of the tendinous part of the ITB, and found strong agreement between the techniques. We then examined the effects of hip adduction on real-time changes in ITB width, and detected signiﬁcant stretch. The average ITB thickness as assessed by US in our cohort (1.970.2 mm) is similar to that previously reported in a group of young and asymptomatic volunteers (1.970.27 mm) (Goh et al., 2003). However, the present study demonstrated a lack of correlation

Table 1 Correlation and difference of morphology between ultrasonography and MRI on the tendinous part of the iliotibial band in mean values and standard error (N ¼ 20) Ultrasonographic measurements

MRI measurements

P value and 95% CIsa of pair t test

P value of correlation test

Width (mm)

5.370.5

5.670.7

0.375 Lower: 0.0452 Upper: 0.62022

o0.001

Thickness (mm)

1.970.2

1.570.2

o0.001 Lower: 0.28815 Upper: 0.57452

40.05

Mean r ¼ 0:850. a

Mean conﬁdent interval.

Table 2 Reliability analysis of the iliotibial band width and hip angle with the modiﬁed Ober maneuver (N ¼ 44) in mean values and standard errors Neutral position

Adducted position

Adducted with weight position

Standard error of measurement

5.270.8 Post hoc: Neutral to adducted position p ¼ 0:027 0.94 Lower: 0.471 Upper: 0.997 0.2 mm

4.670.6 Post hoc: Neutral to adducted with weight position po0:001 0.89 Lower: 0.164 Upper: 0.994 0.3 mm

3.970.7 Post hoc: Adducted to adducted with weight position p ¼ 0:004 0.80 Lower: 0.265 Upper: 0.986 0.3 mm

Hip adduction angle (1)

0

21.175.5

28.775.8

Width (mm) One way ANOVA po0:001 ICC (N ¼ 20)a 95% CIS

A signiﬁcantly difference of hip angle between hip adducted position and adducted with weight position, po0:001. a

Subject number for reliability test was 20.

ARTICLE IN PRESS H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

between values obtained by MRI and US. This may be due to the fact that the border/contrast was less clear with US, making accurate measurement difﬁcult. Any real-time measurements made by US would likely estimate changes in ITB thickness without sufﬁcient accuracy. For this reason, we did not use band thickness as an indicator of real-time stretch effects in the study. In contrast to previous studies examining real-time effects of stretch on the ITB by palpation, cinemographic or goniometric analyses (Fredericson et al., 2002; Gajdosik et al., 2003; Reese and Bandy, 2003), this ultrasonographic study used morphological changes of the tendinous part of the ITB as objective and direct indications of stretching effects. In addition, the confounding effects of hip muscles, not addressed or controlled in the previous studies, were minimal as determined by surface electromyography in the present study. This is the ﬁrst report of in vivo, real-time morphological changes of soft tissues under stretching. Using the modiﬁed Ober maneuver, we demonstrated a signiﬁcant reduction in the band width (as detected by US) when a subject’s hip was dropped from the neutral to adducted position. With the addition of the 3 kg weight, a signiﬁcant reduction in the band width and a signiﬁcant increase of hip adduction angle (from 21.1175.51 to 28.7175.81) were observed in the position of adducted with weight. These reliable ﬁndings (ICC: 0.94–0.86, standard error of measurement of 0.2–0.3 mm) illustrate a signiﬁcant and gradual narrowing of band width in the modiﬁed Ober maneuver. Additionally, these results implied that the initial adduction caused the ITB to stretch, and that further stretch was elicited upon adduction with weight or greater hip adduction position. These ﬁndings suggest that under different levels of stretch, real-time morphological proﬁles and changes in the tendinous part of the ITB can be reliably detected by dynamic US. Most studies regarding the real-time morphological changes of tendons or ligaments under stretch have focused on mechanical properties (Lewis and Shaw, 1997; Maganaris and Paul, 2002; Costic et al., 2003) or function analysis (Sheehan and Drace, 2000; Li et al., 2004). This study is the ﬁrst to report testing of the realtime effects and validity of a clinical application. We have found that signiﬁcantly increasing the joint angle during a clinical stretching maneuver results in signiﬁcant real-time morphological changes of the target tissue (in this case the ITB). These real-time morphological changes are thought to underlie acute responses to stretch, such as impairments in neural output via Golgi tendon organ (Fowles et al., 2000; Behm et al., 2004). Whether or not there is a dose response relationship between the real-time morphological changes and the acute responses following stretch is unclear. Further studies are needed to assess the correlation between real-

339

time morphological soft tissue changes and consequent acute muscular responses to therapeutic stretching. In addition, we suggest that the real-time effectiveness of therapeutic stretching protocols can be evaluated by US. Physiotherapists have used US to analyze the neuromuscular mechanisms of lower back pain, and to screen sporting shoulder injuries (Ferreira et al., 2004; Wang et al., 2005). Dynamic US is also often conducted in conjunction with physical examination (e.g. the shoulder impingement test) in order to conﬁrm diagnosis (Desmeules et al., 2004). We hope the results of our study will encourage more physiotherapists to use US for research regarding measurement of the real-time effects of manipulation therapy. US may help to provide more direct real-time evidence, enhancing the validity of manipulation therapy. There are several limitations to our study. It is commonly assumed in clinical practice that the ITB or band is the main tissue targeted under loading in the modiﬁed Ober maneuver. However, without inserting a strain gauge parallel to the ITB and without measuring the length of the entire ITB directly, we could not exclude the possibility that the changes in width were a consequence of changes in subject hip position. In addition, although one researcher (a physiotherapist with more than ten years experience) was responsible for the stabilization of the pelvic in this study, measures ensuring a standardized stabilization were not performed. Further studies could incorporate the pressure biofeedback unit (Stabiliser Pressure Biofeedback, Chattanooga, Australia) to monitor the onset movement of lateral pelvic title and establish a protocol of standardized stabilization (Herrington et al., 2006). This study demonstrated real-time morphological changes of the ITB in the modiﬁed Ober maneuver. We conclude that US is an objective and reliable technique that can be used to verify topics in clinical maneuvers such as real-time stretch effects. References Bandy WD, Irion JM. The effect of time on static stretch on the ﬂexibility of the hamstring muscles. Physical Therapy 1994;74: 845–50. Behm DG, Bambury A, Cahill F, Power K. Effect of acute static stretching on force, balance, reaction time, and movement time. Medicine and Science in Sports and Exercise 2004;36:1397–402. Bonaldi VM, Chhem RK, Drolet R, Garcia P, Gallix B, Sarazin L. Iliotibial band friction syndrome: sonographic ﬁndings. Journal of Ultrasound in Medicine 1998;17:257–60. Costic RS, Vangura Jr A, Fenwick JA, Rodosky MW, Debski RE. Viscoelastic behavior and structural properties of the coracoclavicular ligament. Scandinavian Journal of Medicine & Science in Medicine 2003;13:305–10. Desmeules F, Minville L, Riederer B, Cote CH, Fremont P. Acromiohumeral distance variation measured by ultrasonography and its association with the outcome of rehabilitation for shoulder

ARTICLE IN PRESS 340

H.-K Wang et al. / Manual Therapy 13 (2008) 334–340

impingement syndrome. Clinical Journal of Sport Medicine 2004;14:197–205. Fairclough J, Hayashi K, Toumi H, Lyons K, Bydder G, Phillips N, et al. The functional anatomy of the iliotibial band during ﬂexion and extension of the knee: implications for understanding iliotibial band syndrome. Journal of Anatomy 2006;208:309–16. Ferreira PH, Ferreira ML, Hodges PW. Changes in recruitment of the abdominal muscles in people with low back pain: ultrasound measurement of muscle activity. Spine 2004;29:2560–6. Fowles JR, Sale DG, MacDougall JD. Reduced strength after passive stretch of the human plantarﬂexors. Journal of Applied Physiology 2000;89:1179–88. Fredericson M, Guillet M, DeBenedictis L. Quick solutions for iliotibial band syndrome. The Physician and Sportsmedicine 2000;28:53–68. Fredericson M, White JJ, MacMahon JM, Andriacchi TP. Quantitative analysis of the relative effectiveness of 3 iliotibial band stretch. Archives of Physical Medicine and Rehabilitation 2002;83:589–92. Gajdosik RL, Sandler MM, Marr HL. Inﬂuence of knee positions and gender on the Ober’s test for length of the iliotibial band. Clinical Biomechanics 2003;18:77–9. Goh LA, Chhem RK, Wang SC, Chee T. Iliotibial band thickness: sonographic measurement in asymptomatic volunteers. Journal of Clinical Ultrasound 2003;31:239–44. Herrington L, Rivett N, Munro S. The relationship between patella position and length of the iliotibial band as assessed using Ober’s test. Manual Therapy 2006;11:182–6. Kendall HO, Kendall FP, Boynton DA. Posture and pain. Baltimore: Williams and Wilkins; 1970. p. 135–8.

Lewis G, Shaw KM. Modeling the tensile behavior of human Achilles tendon. Biomedical Materials and Engineering 1997;7:231–44. Li G, Defrate LE, Sun H, Gill TJ. In vivo elongation of the anterior cruciate ligament and posterior cruciate ligament during knee ﬂexion. American Journal of Sports Medicine 2004;32: 1415–20. Maganaris CN, Paul JP. Tensile properties of the in vivo human gastrocnemius tendon. Journal of Biomechanics 2002;35:1639–46. Muhle C, Ahn JM, Yeh L, Bergman GA, Boutin RD, Schweitzer M, et al. Iliotibial band friction syndrome: MR imaging ﬁndings in 16 patients and MR arthrographic study of six cadaveric knees. Radiology 1999;212:103–10. Nishimura G, Yamato M, Tamai K, Takahashi J, Uetani M. MR ﬁndings in iliotibial band syndrome. Skeletal Radiology 1997;26: 533–7. Reese NB, Bandy WD. Use of an inclinometer to measure ﬂexibility of the iliotibial band using the Ober’s test and the modiﬁed Ober’s test: difference in magnitude and reliability of measurements. Journal of Orthopaedic and Sports Physical Therapy 2003;33: 326–30. Sheehan FT, Drace JE. Human patellar tendon strain: a noninvasive, in vivo study. Clinical Orthopaedics and Related Research 2000;370:201–7. Wang HK, Lin JJ, Pan SL, Wang TG. Sonographic evaluations in elite college baseball athletes. Scandinavian Journal of Medicine & Science in Medicine 2005;15:29–35.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 341–348 www.elsevier.com/locate/math

Original article

Postero-anterior movements of the cervical spine: Repeatability of force displacement curves Neil Tuttle, Rod Barrett, Liisa Laakso School of Physiotherapy and Exercise Science, Griffith University, Australia Received 22 June 2006; received in revised form 23 January 2007; accepted 7 March 2007

Abstract The repeatability of instrumented assessments of postero-anterior (PA) movements has been reported previously for lumbar and thoracic spines, but only in relation to limited parameters of the movement. This study describes a device for measuring PA movements of the cervical spine and reports on repeatability of the entire force/displacement (FD) curves. Repeatability was assessed using coefﬁcients of multiple determination (CMDs) and adjusted CMDs (where the mean offset between the two curves are removed and the shape of the curve can be more directly assessed) for inter-rater intra-day (inter-rater), intra-rater inter-day (inter-day) and intra-rater intra-day (intra-rater) repeated measurements. The mean CMD and mean adjusted CMD for intra-rater measurements (0.90 and 0.99, respectively) were signiﬁcantly higher than for the other measurement intervals. Inter-rater and interday mean CMDs were 0.76 and 0.73 and mean adjusted CMDs were 0.96 and 0.97. It is concluded that the maximum repeatability is achieved if the same operator reassesses the patient on the same day. It is hoped that the methodology described will form the basis for further research that will enable greater understanding of what characteristics of PA movements inform manual palpation and thereby enable improvement in both manual therapy treatment teaching efﬁciency of manual therapy skills. r 2007 Elsevier Ltd. All rights reserved. Keywords: Range of motion; Articular; Physical therapy techniques; Manipulation; Spinal

1. Introduction Musculoskeletal disorders of the spine are amongst the most common health problems in Australia (Bogduk et al., 2003). Spinal mobilization is a common form of treatment for spinal musculoskeletal disorders and is based on a presumed relationship between symptoms and intervertebral mobility. Postero-anterior (PA) movements typically produced by the manual application of force to an individual vertebra either on or lateral to the midline are commonly used for assessment and treatment of spinal symptoms (Maitland et al., 2005). During a PA movement, the clinician perceives the relationship between the force applied and the resulting Corresponding author. Tel.: +61 07 55528930; fax: +61 55528674.

E-mail address: n.tuttle@grifﬁth.edu.au (N. Tuttle). 1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.03.005

displacement. The PA movement is intended to provide the clinician with information about the source of the patient’s symptoms and is often represented by a force– displacement (FD) curve. Although PA movements were once thought to produce localized translational intervertebral movement, it is now clear from both in vivo (Lee and Evans, 1997; Caling and Lee, 2001; McGregor et al., 2001; Kulig et al., 2004; Lee et al., 2005; McGregor et al., 2005) and in vitro studies (Gal et al., 1997a, b; Sran et al., 2005) that in addition to segmental movement around a ﬂexion/extension axis, other movement also occurs including regional spinal movement, soft tissue compression, and movement of muscle and connective tissue. There are two challenges to assessing the source of a patient’s symptoms by PA movements. Firstly, it is not known how the characteristics of PA movements are altered in the presence of symptoms. Secondly, the displacement produced by a

ARTICLE IN PRESS 342

N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

PA movement is complex including deformation of soft tissues as well as the vertical movement of the vertebra. The vertical movement of the vertebra in turn not only involves the entire cervical spine, but is inﬂuenced by other factors including rocking of the head and compression of the padding of the plinth. The difﬁculty in interpreting the clinically relevant characteristics of PA movements is therefore in being able to extract a symptomatic structure’s as yet unknown inﬂuence from a PA movement’s already complex signal. In spite of the apparent difﬁculties, clinicians are able to detect clinically useful information from manual assessments of PA movements. For example, the location of congenital fusion can be reliably detected by manual motion palpation (Humphreys et al., 2004). In a more clinically relevant study, patients who received treatment corresponding to ﬁndings on manual palpation had better outcomes than those receiving randomly allocated treatment (Fritz et al., 2005). Several researchers have attempted to objectively assess PA movements using an indentor to apply a force over part of a vertebra and sensors to measure displacement and force (Lee and Svensson, 1990; Lee and Evans, 1992; Latimer et al., 1996a; Edmondston et al., 1998; Kawchuk and Fauvel, 2001). Attempts at characterization of PA movements from instrumented assessments relied primarily on single values of displacement or stiffness at speciﬁed force levels. Lee and Svensson, (1990), and Latimer et al. (1996a) considered that the FD curve of a lumbar spine could be considered as consisting of a ‘toe region’ where the slope was nonlinear followed by a linear region of the curve over 20 or 30 N. The slope of the linear portion of the FD curves, the overall displacement and the length of the toe region have all been used to characterize PA movements. Although the rationale for selection of these parameters is not clear, they have been widely used to describe PA movements (e.g. (Lee and Evans, 1992; Latimer et al., 1996a; Kaigle et al., 1998; Allison et al., 2001; Shirley et al., 2002; Chiradejnant et al., 2003; Sran et al., 2005). Differences in single measures of stiffness or displacement of PA movements have been found to be related to factors affecting the target intervertebral structures (Latimer et al., 1996b, c; Kawchuk et al., 2001b; Sran et al., 2005) as well as factors affecting regional or extraspinal factors (Edmondston et al., 1998; Kawchuk and Fauvel, 2001; Chansirinukor et al., 2003; Shirley et al., 2003; Colloca and Keller, 2004). The single measures used in these studies however do not appear to have overcome the challenges to speciﬁcally assessing the clinically relevant aspects of PA movements. Furthermore instrumented assessments have not agreed with clinician’s interpretations from manual assessments (Latimer et al., 1996b, c) and the single measures used to characterize PA movements do not appear to be the

same as the parameters that inform manual motion palpation (Maher et al., 1998). As a result, the characteristics of PA movements that clinicians consider during motion palpation remain elusive (Petty et al., 2002). Findings are emerging suggesting how changes in intersegmental stiffness may be related to pathology and how such changes might impact on PA movements. Gay et al. (2006) found differences in segmental stiffness occurring with lumbar disc degeneration were more pronounced near the neutral position rather than near the end of range. We performed computer-based modelling of PA movements, where alterations of the ‘neutral zone’ suggested that effects on PA movements would be complex with the greatest differences likely to occur in the early to middle portion of the PA movement (Tuttle et al., 2006). The single values of displacement or stiffness used previously may not be the clinically relevant characteristics of PA movements. Other, as yet unknown parameters of the FD relationship may be necessary to adequately characterize PA movements. The repeatability of measurements of the FD relationship of PA movements must therefore be established if measurements of PA movements are to be used to detect the clinically relevant characteristics of PA movements. In order to ensure appropriate experimental design for future studies, it is also important to know whether repeated measures can be reliably performed by the same or different practitioners and on the same or different days. The purpose of this study was to determine the interrater, intra-rater and inter-day repeatability of the passive movement assessment device (PMAD), which was developed to assess PA movements of the cervical spine. It was hypothesized that the repeatability would be best for the same operator on the same day and that the variation from the tests being performed on different days would be greater than the variation resulting from different operators. 2. Methods 2.1. Subjects and experimental design Subjects were recruited from university staff and students. Inclusion criteria were asymptomatic subjects deﬁned as having no neck symptoms within the past six months that required treatment and no contraindications or precautions to manual therapy assessment or treatment (Lee et al., 2004). Ten participants (six females and four males; mean age 37.2 years, range 21–50 years; mean weight 72.7 kg, range 52–92 kg; mean height 169.9 cm, range 155–179 cm) were recruited for the study. The experimental protocol was approved by the Grifﬁth University Human Research Ethics

ARTICLE IN PRESS N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

Committee and all individuals provided written conﬁrmation of their informed consent prior to participation. The procedures were explained and the subjects familiarized with the equipment prior to the ﬁrst trial. Subjects were assessed by two trials on each of two consecutive days with all trials performed by the same operator except the second trial on day two, which was performed by a second operator. Both operators were qualiﬁed musculoskeletal physiotherapists with over 10 years experience. Each trial consisted of PA movements to a total of six locations: both sides at each of three levels separated by 12 mm along the long axis of the treatment bed. The intention of this study was to assess the repeatability of the PMAD not to assess the ability to locate anatomical locations. It was therefore not necessary to undertake the imaging that would be required to locate positions anatomically. Rather the positions of the PMAD selected in the ﬁrst trial were repeated for subsequent trials as it was necessary only to ensure that the indentor and patient were located in the same position for repeated measurements. After the ﬁrst trial on each day, the subject stood and walked a few steps before adopting the same position for a second trial. 2.2. Instrumentation and data collection protocol ‘Unilateral’ PA movements were assessed with the force applied over the articular pillar. In order to maximize the relevance of our instrumented assessment to the clinical setting, our intention was to develop a methodology capable of assessment of PA movements of the cervical spine in a manner as similar as possible to that which occurs during manual palpation of PA movements. In principle, the device was similar to methods reported previously for the lumbar and thoracic regions, but was adapted in two ways. Firstly, the indentor was constructed to be similar in size to the human thumb and secondly, the force was applied manually rather than mechanically. The force being applied manually ensures that the movement is as similar as possible to PA movements in the clinical setting and also enables the device to be used in future studies to establish the physical factors corresponding to clinicians’ perceptions. The principal requirements of the PMAD (Fig. 1) developed for this purpose were identiﬁed as being able to: 1. Apply a PA force to repeatable locations. 2. Produce repeatable movements as similar as possible to manual assessment of PA movements. 3. Be perceived by the subject and operator to be as similar as possible to those occurring with manual assessment. 4. Accurately measure the force and displacement characteristics of the PA movements.

343

Fig. 1. PA measurement device (PMAD). The device measures force and displacement during movement produced by a force applied manually to the thumb-hold.

The indentor was intended to resemble the human thumb as used in manual assessment of PA movements to ensure maximum subject comfort without the indentor being larger than the contact used by a practitioner or wider than the height of a cervical vertebra. An indentor was constructed of a 25 mm length of 12 mm square aluminium section with edges rounded to a radius of approximately 1 mm. The assessment of each location consisted of ﬁve gradual applications of force up to 25 N performed at a frequency of approximately 1 Hz, which is within the range used in manual assessment (Snodgrass et al., 2006). The force and displacement data were simultaneously recorded while the operator gradually applied force to the thumb-hold until hearing an audible sound produced by the data acquisition programme when the force reached 25 N. The ﬁve applications of the force were angled medially by 101 and positioned such that the shaft was 15–25 mm from the midline when the indentor contacted the subject. The shaft was secured to a linear bearing to ensure smooth repeatable movement and the entire assembly could be repositioned to a corresponding position on the contralateral side. Both the operator and subjects found this combination of indentor size, position and direction most closely approximated the sensations felt during manual assessment of unilateral PA movements. The applied force was measured with a load cell (Transducer Technologies MLP-25) between the thumbhold and the indentor. The corresponding displacements were measured with a linear potentiometer (Hollywell LTS04N04KB5C) attached to the shaft of the PMAD. The non-repeatability of the load cell is reported by the manufacturer to be 0.05% and the linearity of the potentiometer to be 70.1%. A two-point linear calibration was performed for both the potentiometer

ARTICLE IN PRESS 344

N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

and load cell prior to each day’s testing. Three readings were taken at each of two displacements (separated by 69 mm) and two loads (deadweights of mass 27 g and 2362 g). The sensors were connected to a PC through a USB DAQ card (USB-6008, National Instruments) and sampled at 100 Hz. Data collection and storage was performed with custom software written in LabView Version 7i. A repeatable subject position was achieved by replacing the head section of a standard heightadjustable treatment bed with a speciﬁcally designed head-cradle lined with 2 mm of high density EVA foam. To determine how much movement was likely to occur as a result of compression of the head-cradle and treatment bed an anatomical skull model was placed in the head cradle to represent the subject’s head and a ﬂat plate 20 cm 20 cm weighing 5 kg was placed on the treatment table as a conservative representation of the subject’s chest. Forces of 12.5 N were applied to each and displacements measured using the PMAD. The measured displacement of the skull model of 0.2 mm and of the plate of 0.4 mm suggested that less than 0.4 mm of the PA movement on subjects would result from compression of the supporting head-cradle and treatment table. The reproducibility of positioning of the subject was tested by measuring the angle of a pointer held between the subject’s teeth and the position of the vertex. The angle and position were then remeasured after the subject stood up and repositioned themselves on the treatment bed. The ICC(3,1) were 0.94 and 0.99 for head angle and position, respectively, and the corresponding 95% limits of agreement (Bland and Altman, 1999) were 1.91 to 1.91 and 1.7 to 1.7 mm, respectively. A sliding frame able to be ﬁxed at 12 mm intervals along the long axis of the bed was ﬁxed to the head cradle to enable repeatable positioning of the device for the three levels tested. 2.3. Data analysis Following data collection, the data was processed with custom software using Matlab Version 7.04. Data from the second, third and fourth applications from 0.5 to 25 N of force were ﬁltered using a second order lowpass Butterworth ﬁlter with a cut-off frequency of 2.5 Hz. The displacement at 0.5 N was assigned a value of zero to create a common starting point for further comparisons and the three resulting curves were averaged. In preliminary trials soft tissue occasionally appeared to move under the indentor and the force in the resulting measure did not increase continuously. Therefore, when processing the data, if the force data did not continuously increase through the range being assessed for any of the middle three applications, the ﬁfth and, when necessary, the ﬁrst applications were used in their place for further analysis. Although data

from the ﬁrst application was only used on three occasions, it is important to note that our preliminary testing did not suggest the need for ‘preconditioning’ the movement as we did not detect any differences between the ﬁrst and subsequent force applications. Preconditioning may not have been required due to aligning all curves at a force level of 0.5 N rather than at a common location in space. Ninety-nine displacement values corresponding to forces from 0.5 to 25 N at 0.25 N intervals were then determined using a cubic spline interpolation. 2.4. Statistical analysis The coefﬁcient of multiple correlation (CMC) was advocated by Kadaba et al. (1989) and has since been used extensively to assess the repeatability of curve data related to gait. The adjusted coefﬁcient of multiple determination (CMD) is deﬁned as the square of the CMC and was used in the current study as it indicates the proportion of the variance accounted for within the data. CMDs have been used to assess the repeatability of measures of both gait (Kavanagh et al., 2006) and active spinal movement (Lee et al., 2003), but to our knowledge have not been used for the assessment of the repeatability of passive movements. CMD is deﬁned as CMD ¼ 1

s2e , s2g

where s2e represents the variance from the ensemble average FD curve and s2g represents the variance from the grand mean of the FD curves. The more similar the curves being compared, the more s2e approaches zero and the CMD approaches one. Conversely, the more dissimilar the curves, the more s2e approaches s2g and the CMD approaches zero. The offset (systematic bias) in pairs of FD curves was assessed by calculating the difference (in mm) between the grand means of the two curves. In order to facilitate comparison of the overall shape of the curves, adjusted CMDs were obtained by recalculating the CMDs for each pair of FD curves with the offset removed. In other words the repeatability of the original curves considers the curves with a common starting point, the offsets indicate any systematic bias as might occur with altered tissue compliance or muscular contraction while the adjusted curves compare the overall shape of the curves. The repeatability of the FD curves was assessed by calculating the CMDs for each of the three repeated measures that were assessed: intra-day, inter-rater (inter-rater); intra-day, intra-rater (intra-rater); and inter-day, intra-rater (inter-day). The CMDs were also calculated to compare the repeatability of the shape of the curves with the offsets removed. ANOVAs and post

ARTICLE IN PRESS N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

hoc Scheffe tests were used to assess main effects of repeated measure interval, side, and level on CMDs, offsets and adjusted CMDs. Statistical signiﬁcance was set at Po0.05.

1

345

Inter-rater Inter-day Intra-rater

CMD

0.9 0.8 0.7 0.6 0.5 4 2 Offset (mm)

Representative data for two pairs of repeated measures with the corresponding CMDs, offsets and adjusted CMDs are shown in Fig. 2. The graphs illustrate how the overall displacement and the shape of the FD curve can differ and give an indication of the extent of agreement between pairs of curves that corresponds with CMD, offset and adjusted CMD values. The CMDs for each repeated measure interval ranged from 0.72 to 0.90 with the intra-rater mean CMD being 0.90 (Fig. 3A). There were signiﬁcant differences between CMDs for the three repeated measure intervals (F ¼ 6.57, dF ¼ 2), with intra-rater repeatability being greater than inter-rater by 0.16 (CI ¼ 0.03–0.28) and greater than inter-day by 0.18 (CI ¼ 0.04–0.31). There were no differences in CMDs between sides or levels. The offsets for the repeated measure comparisons are shown in Fig. 3B. The mean offsets for the three repeated measure intervals ranged from 0.1 to 2.0 mm. There were signiﬁcant differences in the magnitude of the offsets (F ¼ 8.47, dF ¼ 2) with the inter-day offsets

0 -2 -4 1 0.9

Adjusted CMD

3. Results

0.8 0.7 0.6

Inter-day Subject A, Right-Low CMD = 0.79 Offset = 2.8 mm Adjusted CMD = 0.96

25

Low

Mid

Up

Left

Right

Total

Fig. 3. Repeatability for each level and side for each repeatability interval. CMDs are shown in A, Offsets (the difference between the ensemble mean values of each FD curve) are shown in B and the adjusted CMDs (comparing the shape of the FD curves with the offsets removed) are shown in C.

Intra-day Subject D,Left-Mid CMD = 0.97 Offset = 0.4 mm Adjusted CMD = 0.99

20 Force (N)

0.5

15

10

5

0 0

5

10

15

20

25

Displacement (mm) Fig. 2. Representative data from two pairs of repeated measures with their corresponding CMDs. The similarity of the intra-day curves is conﬁrmed by the high CMD. The adjusted CMD indicates the extent of agreement between the curves once the offset is removed. The high adjusted CMDs of both repeated measures indicate close agreement in the shape of the curves.

being larger than inter-rater by 0.16 mm (CI 0.34 to 2.77 mm) and larger than intra-rater by 2.22 mm (CI 0.80 to 3.63 mm). There were no signiﬁcant differences in the offsets between sides or levels. As shown in Fig. 3C, the adjusted CMDs for the three retest intervals were higher than the standard CMDs and ranged from 0.96 to 0.99. Again there were signiﬁcant differences between the three intervals (F ¼ 2.46, dF ¼ 2) with the intra-rater repeatability being greater than inter-rater by 0.03 (CI 0.01–0.05) and inter-day by 0.02 (CI 0.01–0.04). There was a small, but statistically signiﬁcant difference in the adjusted CMDs between sides (F ¼ 4.11, dF ¼ 1) with the left side being greater than the right by 0.01 (CI 0.00–0.03), but no differences between levels.

ARTICLE IN PRESS 346

N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

4. Discussion The purpose of this paper was to report on the repeatability of assessment of PA movements of the cervical spine using the PMAD. To our knowledge this is the ﬁrst report of the repeatability of assessment of PA movements of the cervical spine and the ﬁrst study to assess the repeatability of entire FD curves rather than single values for any region of the spine.

make direct comparisons between the current study and previous studies as different measures of repeatability were employed in each study. It could be expected that assessment of the cervical spine may be less repeatable than other regions of the spine as the subject’s position when assessing the cervical spine is less stable and the bony landmarks being assessed more difﬁcult to locate accurately. 4.2. Relevance to clinical practice

4.1. Repeatability of PA movements As expected the best repeatability for all measures occurred for intra-rater comparisons, where the repeated measures were taken by the same operator on the same day. The expected difference between inter-rater and inter-day repeatability was only found in relation to the adjusted CMDs and was most likely too small to be of practical signiﬁcance. Therefore, differences in PA movements would be expected to be easiest to detect if repeated measurements are taken by the same operator on the same day. The instrumentation can be considered to perform equally well regardless of side or level as indicated by the repeatability being comparable across sides and levels. The adjusted CMDs were consistently larger than the non-adjusted CMDs demonstrating greater consistency in the shape of the curve than when the curves had a common starting point. If only the shape of the curve is being considered, the repeatability of the current instrumentation is excellent with CMDs over 0.96 for all intervals of repeated measures. The reason for or signiﬁcance of the offsets being larger for intra-day than for the other comparisons is not clear. Repeatability of FD curves from assessments of PA movements has not been reported previously. Previous studies on in vivo assessment of PA movements have been performed on lumbar, thoracic or animal spines. Lee and Evans (1992) assessed PA movements of the lumbar spine and reported ICCs for inter-day testing of 0.99 and 0.95 for displacement with a maximum error of less than 1 mm. Lee and Svensson, (1990) also found good repeatability of the stiffness of the linear portion of the FD curve with ICCs of 0.88. A later device from the same group reported ICCs of 0.89–0.96, for intra-day, intra-rater repeated measures performed without repositioning the subject (Latimer et al., 1996a). Kawchuk et al. (2001a, b) added an ultrasound transducer to the tip of the indentor of a similar instrument to reduce possible errors in positioning as well as providing a more direct indication of the depth and movement of the target vertebra. The resulting accuracy for the linear region of the FD curve of porcine vertebrae was 1–2% for displacement and 6% for stiffness while for the nonlinear region of the curve the corresponding accuracies were 3–4% and 14%, respectively. It is not possible to

The PMAD described in this paper was intended to assess PA mobility in a way comparable to that used in manual palpation. Even though PA movements involve the entire cervical spine, their intention is to assess symptomatic structures and it appears that clinicians are able to extract clinically useful information. It is not understood what parameters of PA movement inform the practitioner’s interpretations, but it is clear that single values of displacement or stiffness do not adequately characterize the movements being assessed. At this time, we can only speculate on ways symptoms or pathology might affect FD curves measured using the PMAD. For example, increased muscular contraction might produce a linear change in the entire curve, a difference in the size of an intervertebral ‘neutral zone’ might produce a greater difference in the early part of the PA movement and alterations in viscosity as suggested Nicholson et al. (2001) might result in ratedependent differences that are uniformly distributed throughout the PA movement. The potential complexity of aspects of PA movements are further illustrated by Maher et al. (1998) who documented over 40 terms used by practitioners to describe their perceptions of mobility on motion palpation. Ramsay (1996) suggests that the signiﬁcant features of continuous data are often more apparent when one considers derivatives of the raw data. For example, the pattern of stiffness (the ﬁrst derivative of the FD curve) or the change in stiffness (the second derivative of the FD curve) may be factors considered by practitioners. As the meaning of differences in offsets in the current study is unclear, it is interesting to note that the offsets are eliminated when derivatives of the data are considered. Changes in stiffness for example may be related to descriptors such as R1 (the point where the ﬁrst ‘resistance’ to the movement is felt) and end-feel (the change in stiffness at the end of the movement). Some authors suggest that the difﬁculty in assessment of PA movements results from practitioners being informed by non-repeatable or idiosyncratic factors (Maher and Adams 1995). Others have suggested using clinicians who are ‘gold standard palpators, against which others can be calibrated’ (Hansen et al., 2006). It is hoped that methodologies such as described in this paper will help reduce the subjectivity of motion

ARTICLE IN PRESS N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

palpation and enable more objective characterization of the clinically relevant aspects of PA movements. A clearer understanding of the relevant characteristics of PA movements not only has the potential to improve the effectiveness of manual therapy diagnosis and treatment, but to improve the efﬁciency of teaching manual therapy skills to students. It is expected that the repeatability of the PMAD found in the current study would be sufﬁcient to enable studies using this methodology to detect differences relevant to clinical practice regardless of whether the differences are constant, rate dependent or unevenly distributed throughout the PA movements.

References Allison G, Edmonston S, Kiviniemi K, Lanigan H, Simonsen AV, Walcher S. Inﬂuence of standardized mobilization on the posteroanterior stiffness of the lumbar spine in asymptomatic subjects. Physiotherapy Research International 2001;6(3):145–56. Bland JM, Altman DG. Measuring agreement in method comparison studies. Statistical Methods in Medical Research 1999;8(2):135–60. Bogduk N, Bolton P, Jull G, Bellamy N. Acute neck pain. In: Brooks P, March L, Bogduk N, Bellamy N, editors. Evidence-based management of acute musculoskeletal pain. Bowen Hills, QLD: Australian Academic Press; 2003. p. 83–110. Caling B, Lee M. Effect of direction of applied mobilization force on the posteroanterior response in the lumbar spine. Journal of Manipulative and Physiological Therapeutics 2001;24(2):71–8. Chansirinukor W, Lee M, Latimer J. Contribution of ribcage movement to thoracolumbar posteroanterior stiffness. Journal of Manipulative and Physiological Therapeutics 2003;26(3):176–83. Chiradejnant A, Maher CG, Latimer J. Objective manual assessment of lumbar posteroanterior stiffness is now possible. Journal of Manipulative and Physiological Therapeutics 2003;26(1):34–9. Colloca CJ, Keller TS. Active trunk extensor contributions to dynamic posteroanterior lumbar spinal stiffness. Journal of Manipulative and Physiological Therapeutics 2004;27(4):229–37. Edmondston SJ, Allison GT, Gregg CD, Purden SM, Svensson GR, Watson AE. Effect of position on the posteroanterior stiffness of the lumbar spine. Manual Therapy 1998;3(1):21–6. Fritz JM, Whitman JM, Childs JD. Lumbar spine segmental mobility assessment: an examination of validity for determining intervention strategies in patients with low back pain. Archives of Physical Medicine and Rehabilitation 2005;86(9):1745–52. Gal J, Herzog W, Kawchuk G, Conway P, Zhang YT. Measurements of vertebral translations using bone pins, surface markers and accelerometers. Clinical Biomechanics 1997a;12(5):337–40. Gal J, Herzog W, Kawchuk G, Conway PJ, Zhang YT. Movements of vertebrae during manipulative thrusts to unembalmed human cadavers. Journal of Manipulative and Physiological Therapeutics 1997b;20(1):30–40. Gay RE, Ilharreborde B, Zhao K, Zhao C, An KN. Sagittal plane motion in the human lumbar spine: comparison of the in vitro quasistatic neutral zone and dynamic motion parameters. Clinical Biomechanics 2006;21(9):914–9. Hansen BE, Simonsen T, Leboeuf-Yde C. Motion palpation of the lumbar spine—a problem with the test or the tester? Journal of Manipulative and Physiological Therapeutics 2006;29(3):208–12. Humphreys BK, Delahaye M, Peterson CK. An investigation into the validity of cervical spine motion palpation using subjects with congenital block vertebrae as a ‘gold standard’. BMC Musculoskeletal Disorder 2004;5:19.

347

Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. Journal of Orthopaedic Research 1989;7(6):849–60. Kaigle A, Ekstrom L, Holm S, Rostedt M, Hansson T. In vivo dynamic stiffness of the porcine lumbar spine exposed to cyclic loading: inﬂuence of load and degeneration. Journal of Spinal Disorders 1998;11(1):65–70. Kavanagh JJ, Morrison S, James DA, Barrett R. Reliability of segmental accelerations measured using a new wireless gait analysis system. Journal of Biomechanics 2006;39(15):2863–72. Kawchuk GN, Fauvel OR. Sources of variation in spinal indentation testing: indentation site relocation, intraabdominal pressure, subject movement, muscular response, and stiffness estimation. Journal of Manipulative and Physiological Therapeutics 2001; 24(2):84–91. Kawchuk GN, Fauvel OR, Dmowski J. Ultrasonic indentation: a procedure for the noninvasive quantiﬁcation of force–displacement properties of the lumbar spine. Journal of Manipulative and Physiological Therapeutics 2001a;24(3):149–56. Kawchuk GN, Kaigle AM, Holm SH, Fauvel OR, Ekstrom L, Hansson T. The diagnostic performance of vertebral displacement measurements derived from ultrasonic indentation in an in vivo model of degenerative disc disease. Spine 2001b;26(12): 1348–55. Kulig K, Landel R, Powers CM. Assessment of lumbar spine kinematics using dynamic MRI: a proposed mechanism of sagittal plane motion induced by manual posterior-to-anterior mobilization. Journal of Orthopaedic and Sports Physical Therapy 2004; 34(2):57–64. Latimer J, Goodsel MM, Lee M, Maher CG, Wilkinson BN, Moran CC. Evaluation of a new device for measuring responses to posteroanterior forces in a patient population, part 1: reliability testing. Physical Therapy 1996a;76(2):158–65. Latimer J, Lee M, Adams R. The effect of training with feedback on physiotherapy students’ ability to judge lumbar stiffness. Manual Therapy 1996b;1(5):266–70. Latimer J, Lee M, Adams R, Moran CM. An investigation of the relationship between low back pain and lumbar posteroanterior stiffness. Journal of Manipulative and Physiological Therapeutics 1996c;19(9):587–91. Lee H, Nicholson LL, Adams RD. Cervical range of motion associations with subclinical neck pain. Spine 2004;29(1):33–40. Lee M, Svensson NL. Measurement of stiffness during simulated spinal physiotherapy. Clinical Physics and Physiological Measurement 1990;11(3):201–7. Lee R, Evans J. Load–displacement–time characteristics of the spine under posteroanterior mobilization. Australian Journal of Physiotherapy 1992;38:115–23. Lee R, Evans J. An in vivo study of the intervertebral movements produced by posteroanterior mobilization. Clinical Biomechanics 1997;12(6):400–8. Lee RY, Laprade J, Fung EH. A real-time gyroscopic system for threedimensional measurement of lumbar spine motion. Medical Engineering and Physics 2003;25(10):817–24. Lee RY, McGregor AH, Bull AM, Wragg P. Dynamic response of the cervical spine to posteroanterior mobilisation. Clinical Biomechanics 2005;20(2):228–31. Maher C, Adams R. Is the clinical concept of spinal stiffness multidimensional? Physical Therapy 1995;75(10):854–60. Maher CG, Simmonds M, Adams R. Therapists’ conceptualization and characterization of the clinical concept of spinal stiffness. Physical Therapy 1998;78(3):289–300. Maitland G, Hengeveld E, Banks K, English K. Maitland’s peripheral manipulation. Edinburgh; New York: Elsevier/Butterworth Heinemann; 2005 [xiii, 636pp.].

ARTICLE IN PRESS 348

N. Tuttle et al. / Manual Therapy 13 (2008) 341–348

McGregor AH, Wragg P, Gedroyc WM. Can interventional MRI provide an insight into the mechanics of a posterior– anterior mobilisation? Clinical Biomechanics 2001;16(10): 926–9. McGregor AH, Wragg P, Bull AMJ, Gedroyc WMW. Cervical spine mobilizations in subjects with chronic neck problems: an interventional MRI study. Journal of Back and Musculoskeletal Rehabilitation 2005;18(1–2):21–8. Nicholson L, Maher C, Adams R, Phan-Thien N. Stiffness properties of the human lumbar spine: a lumped parameter model. Clinical Biomechanics 2001;16(4):285–92. Petty NJ, Maher C, Latimer J, Lee M. Manual examination of accessory movements—seeking R1. Manual Therapy 2002;7(1): 39–43. Ramsay JO. Principal differential analysis: data reduction by differential operators. Journal of the Royal Statistical Society Series B (Methodological) 1996;58(3):495–508.

Shirley D, Ellis E, Lee M. The response of posteroanterior lumbar stiffness to repeated loading. Manual Therapy 2002;7(1):19–25. Shirley D, Hodges PW, Eriksson AE, Gandevia SC. Spinal stiffness changes throughout the respiratory cycle. Journal of Applied Physiology 2003;95(4):1467–75. Snodgrass SJ, Rivett DA, Robertson VJ. Manual forces applied during posterior-to-anterior spinal mobilization: a review of the evidence. Journal of Manipulative and Physiological Therapeutics 2006; 29(4):316–29. Sran MM, Khan KM, Zhu Q, Oxland TR. Posteroanterior stiffness predicts sagittal plane midthoracic range of motion and threedimensional ﬂexibility in cadaveric spine segments. Clinical Biomechanics 2005;20(8):806–12. Tuttle N, Laakso L, Barrett R. Passive movements of the cervical spine: does what we think and teach match what we feel? Australian Physiotherapy Association, Queensland State Conference, Marcoola Beach, Queensland, Australia, 2006.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 349–356 www.elsevier.com/locate/math

Original article

The reliability of paraspinal muscles composition measurements using routine spine MRI and their association with back function Annina Ropponena,, Tapio Videmanb, Michele C. Battie´b a

Institute of Biomedicine, Physiology/Ergonomics, University of Kuopio, P.O. Box 1627, 70211 Kuopio, Finland b Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, Canada Received 20 April 2006; received in revised form 24 January 2007; accepted 9 March 2007

Abstract This study examines the reliability of quantitative and qualitative muscle composition measurements of paraspinal muscle crosssectional areas (CSAs) from routine lumbar spine magnetic resonance images and their association with maximal isokinetic lifting performance. The extent of paraspinal muscle composition reﬂects back function is currently not known. Measurements were repeated 4–8 weeks apart and different measurements of related constructs were compared. Participants were a population-based sample of 169 males, 35–67 years old, without considering the presence or absence of a history of low back pain or related problems in the selection of subjects. The quantitative and qualitative muscle composition measurements for axial magnetic resonance (MR) images of paraspinal muscles at the L3–L4 lumbar spine level, isokinetic lifting force and work, and body fat percentage were the main outcome measures. Results showed that the reproducibility of different paraspinal muscle composition measurements at the L3–L4 level was excellent for CSAs (ICC ¼ 0.95–0.99) and quantitative muscle composition measurements using cerebrospinal ﬂuid adjusted signal intensity (ICC ¼ 0.96–0.99), and moderate for qualitative muscle composition ratings (Kappa ¼ 0.54–0.76). The correlations of the quantitative and qualitative muscle composition measurements with isokinetic lifting force and work were generally low (r ¼ 0.02–0.41), and favoured the qualitative assessments. In conclusion, quantitative and qualitative muscle composition measurements of paraspinal muscles are highly reproducible tissue measures, have low associations with body fat and isokinetic lifting performance, and show that paraspinal muscle morphology using routine spine magnetic resonance imaging (MRI) is poorly related to back function. r 2007 Elsevier Ltd. All rights reserved. Keywords: Reliability; Paraspinal muscles; Magnetic resonance imaging; Muscle composition measurements; Lumbar spine

1. Introduction The wide availability of magnetic resonance imaging (MRI) devices and lumbar axial views of patients with severe or chronic spinal disorders provides as opportunity to examine morphological aspects of the paraspinal muscles in relation to back pain problems. Also the available magnetic resonance (MR) images adds to the epidemiologic research of the lower back, where the focus would be on methods suitable for large subject samples, such as low back pain (LBP) patients. Easily Corresponding author. Tel.: +358 17 163111; fax: +358 17 163112.

E-mail address: annina.ropponen@uku.ﬁ (A. Ropponen). 1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.03.004

obtainable measurements such as measuring muscle quality from routinely available low back MRIs would provide new possibilities to assess low back, whereas diagnostic purposes require advantaged methods, such as tissue speciﬁc imaging sequences in MRI (see e.g. Schick et al., 2002). Paraspinal muscle cross-sectional area (CSA) measurements have been used in both crosssectional and longitudinal follow-up studies of prognosis or treatment outcomes (Gibbons et al., 1997a; Peltonen et al., 1998; Marras et al., 2001; Kim et al., 2005). In addition, measurements of muscle morphology offer objective assessment of muscularity (Parkkola and Kormano, 1992; Ranson et al., 2005; Mengiardi et al., 2006), as compared to muscle function tests that may be

ARTICLE IN PRESS 350

A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

inﬂuenced by such factors as motivation and pain (Lackner and Carosella, 1999; Keller et al., 1999). Paraspinal muscle CSA has been associated to some degree with the muscle’s capacity to generate force and has been claimed to be an objective measurement for back function (Bruce et al., 1997; Keller et al., 1999; Guyton and Hall, 2000). However, the association between paraspinal muscle CSA and back function related factors, such as disability and back pain, have been controversial (Ka¨ser et al., 2001; Mannion et al., 2000). This lack of consistency in the relationship between CSA and back function may be because of failure to take into account paraspinal muscle composition, such as the degree of fatty inﬁltration, in addition to CSA. Isokinetic trunk extension performed with maximal force at constant speed has been found to be moderately associated with paraspinal muscle composition in one study (r ¼ 0.61), and poorly in another (R2 ¼ 4%) (Hultman et al., 1993; Keller et al., 1999). Back function tests of submaximal force levels and endurance had no clear association with paraspinal muscle composition (Hultman et al., 1993; Gibbons et al., 1997). Although the association between the paraspinal muscle CSA and the body fat percent has been shown to be moderate (Gibbons et al., 1998), we are not aware of studies based on the connection between paraspinal muscle composition and body fat percent. The intra- and inter-rater reproducibility of paraspinal muscle CSA measurements were good to excellent, with intra-class correlation coefﬁcients (ICC) ranging from 0.91 to 0.99 (Gibbons et al., 1997; Peltonen et al., 1998; Ra¨ty et al., 1999; Marras et al., 2001). However, the MR images of back muscles from healthy individuals and those with muscle atrophy look very different (Kader et al., 2000). Decreased general activity levels may inﬂuence the ratio of muscle, connective and fat tissues, reorganization of collagen ﬁber directions, atrophy in some sites, and ﬁbrosis in others, without affecting the CSA of the whole muscle (Reid and Costigan, 1987; Mayer et al., 1989). Thus, the quantitative muscle composition measurements have been used in only a few studies investigating lumbar axial MR images (Parkkola and Kormano, 1992; Flicker et al., 1993; Gibbons et al., 1997; Ranson et al., 2005; Mengiardi et al., 2006). Of these, only Gibbons et al. (1997) used 20 healthy subjects and Ranson et al. (2005) used six healthy subjects, where they reported the reproducibility of the paraspinal muscle composition measurements, using muscle signal intensity as an indicator of composition (ICC40.99). In comparison, paraspinal muscle density for 31 chronic LBP patients assessed from computed tomography images had a 2–4% measurement error due to the observer (Keller et al., 2003). Intra-rater agreement for the amount of intermuscular fat in upper cervical muscles from signal

intensities of MR images was also high, ICC ¼ 0.94– 0.98 (Elliott et al., 2005). The single study we found based on the reliability of the qualitative ratings for paraspinal muscle fat content in LBP patients showed good inter-rater reliability (Kappa ¼ 0.85) (Kader et al., 2000). We are aware of only three other studies that used qualitative ratings of muscle composition; Parkkola and Kormano (1992) studied 74 healthy subjects; Mooney et al. (1997) studied 8 patients with LBP, and Mengiardi et al. (2006) compared 25 chronic LBP patients with 25 matched asymptomatic volunteers but none reported the reproducibility of their measurements. A study combining both quantitative and qualitative muscle composition measurements (Parkkola and Kormano, 1992) has shown that the quantitatively analyzed non-muscle tissue categorized by a qualitative measurement of muscle composition corresponded well the quantitative measure of the intramuscular fat. Parkkola and Kormano (1992) concluded that the qualitative measurement would offer a reliable tool for muscle composition assessment whereas a recent study of Mengiardi et al. (2006) could not use the qualitative muscle composition measurement for differing LBP patients from volunteers. The overall objectives of the present study were to investigate the intra-rater reliability of quantitative and qualitative measurements of total muscle CSA and composition, using the digital data from routine lumbar spine MRIs of subjects without current back disorders. In addition to repeating the measurements, we investigated the correlations of the quantitative and qualitative ratings of the muscle composition (e.g., fat inﬁltration) with body fat percentage and isokinetic lifting strength. The correlations of isokinetic lifting strength with total paraspinal muscle CSA and paraspinal muscle CSA after adjusting for measurements of muscle composition were compared. It is postulated that the latter correlation with adjusted muscle CSA will be higher. It is also expected that the reproducibility for quantitative muscle composition measurement will be higher in the well deﬁned psoas major muscle than in the erector spinae muscle mass, which is a combination of three muscles (multiﬁdus, longissimus and iliocostalis muscles) with overlapping muscle fascias, and deep or superﬁcial cavitas between the muscles and fat deposits.

2. Methods 2.1. Subjects Study subjects were a population-based sample of 169 men, ranging from 35 to 67 years old with a mean age of 49 years (standard deviation [SD] 7.8), a mean height of 176 cm (SD 7.3), and a mean weight 80 kg (SD 11.5). The subjects consisted of a random sample from a larger

ARTICLE IN PRESS A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

survey on the effects of common exposures on back pain and spinal degeneration (Battie´ et al., 1995). However, the presence or absence of a history of LBP or related problems was not considered in subject selection and subjects were found to be representative of the larger general population of adult men with respect to symptom history (Simonen et al., 2003). Informed consent was obtained prior to data collection. Data acquisition included axial MR images at the L3–L4 spinal level, isokinetic lifting strength testing and body fat percentage (body fat) measured by bioelectrical impedance. The study received approval from the Ethical Committees of the Department of Public Health at the University of Helsinki and the Faculty of Rehabilitation Medicine at the University of Alberta. MR images were used to analyze paraspinal muscle size and morphology. Subjects were positioned supine on a MR imager and the imaging protocol took about 25–30 min per subject. Transverse sections with the inclination positioned after each intervertebral space were obtained for lumbar levels L2 to S1 using a 1.5-T Siemens Magnetom MR imager (Siemens AG, Erlangen, Germany), with a surface coil. Spin-echo sequence was used with a repetition time of 2450 ms and an echo time of 90 ms. Slice thickness was 3 mm and the spaces between the slices were 0.3 mm. The matrix was 192 256 and the ﬁeld of view was 260 mm. The sequence 2450/90 indicates that the low intensity pixels are likely to come from muscle tissue and high intensity pixels from tissues with low water content such as connective tissue, including fat, in particular (‘‘nonmuscle tissue’’). Variations in MRI background intensity or ﬁeld strength between images creates problems in comparing signal intensities of structures, such as paraspinal muscles, between scans or subjects. Thus, the signal intensity of cerebrospinal ﬂuid (CSF) from the same axial level of each MRI scan was used as a reference to adjust muscle CSA signal intensity, which allowed comparisons between subjects. CSF has been used previously for such adjustments to allow comparisons of signal intensities of other structures between individuals (Battie´ et al., 1995). 2.1.1. Muscle CSA measurements The perimeters of the right erector spinae (including the multiﬁdus, longissimus and iliocostalis, later referred to ‘erector muscle’), psoas major and quadratus lumborum muscles were traced visually with a cursor at the L3–L4 level (Fig. 1) by one of the authors (AR) and analyzed digitally. Muscle perimeters were traced excluding clear cavities of fat at the periphery of the muscle area visually identifying the edge of the muscle. The erector muscle was traced as the whole area occupied by multiﬁdus, longissimus and iliocostalis, which lie adjacent to each other, including the nonmuscular tissue between them. The sequence 2450/90

351

Fig. 1. Axial MR image at the level of the L3–L4 disc with tracing around the paraspinal muscles.

indicate that the low intensity pixels based on signal threshold are likely to originate from muscle tissue (with a high water content) and high intensity pixels based on signal intensity are from tissues with low water content, such as fat in particular (non-muscle tissue). The water content of a muscle reﬂects the quality of the muscle: the degeneration of the muscle leads to replacement of muscle tissue (with high water content) to fat and other tissues with lower amounts of water and atrophy of muscle tissue decreases the relative amount of water. For the in vivo measurements, the sequence 2450/90 provided an excellent opportunity to evaluate the muscle quality based on the sequences ability to visualize water signals. The threshold signal intensity value separating muscle (lower intensity pixels) from fat and connective tissue (high intensity pixels) was selected based on the bi-modal distribution of signal intensities and conﬁrmed from sampling of intensities from the image data using a custom-made spine image analysis program. Images from L5-S1 were omitted from the analysis due to the effect of lumbar curvature (lordosis) on the angle of muscles and their CSA measurements. Muscle CSA measures at the L3–L4 disc level on the right side were selected for the analysis because the paraspinal muscle CSAs have previously been found to be largest overall at the L3–L4 level (Ka¨ser et al., 2001; Marras et al., 2001) and on the right side (Marras et al., 2001). The paraspinal muscle CSAs were scanned perpendicular to the direction of the muscle ﬁbers (901 angle from the front and back walls of spinal canal, ligamentum ﬂavum). Paraspinal muscles are parallel to ligamentum ﬂavum, so the CSA scans best represent the anatomic scans of paraspinal muscles. The position of the subjects was also controlled to avoid errors and all subjects were supine with pillow under their knees. The neutral position of the subjects has been shown to provide most accurate measure of the anatomical CSA of paraspinal muscles (Jorgensen et al., 2003).

ARTICLE IN PRESS A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

352

Theoretically, the CSA at the thickest site of a muscle should provide the most reproducible CSA measurement and the most accurate correlation with force. No correction for muscle orientation was made because the differences between digitally obtained areas and true anatomical areas were found to be small (o3%) (McGill et al., 1993). The intra-rater repeatability of paraspinal muscle CSA at the L3–L4 disc level was determined from repeat measurements by the same operator of the 169 subjects performed 4 to 8 weeks apart. The order of the images was randomized at the second session and the measurer was blinded to the results of the ﬁrst session. 2.1.2. Quantitative muscle composition measurement After the CSA of the muscle was traced, the mean signal intensity was obtained as an indicator of muscle composition. The mean muscle signal intensity was adjusted by the signal intensity of the adjacent CSF, which was used as an intra-body reference, and provided a ratio that could be compared between individuals (Battie´ et al., 1995). Higher values represented higher non-muscle or fat composition. Muscle CSA was then adjusted by this value for analyses of the association of paraspinal muscle measurements with isokinetic lifting performance. 2.1.3. Qualitative muscle composition measurement In addition to the quantitative measures, non-muscle tissue area within the total muscle CSA were rated qualitatively for the erector spinae, psoas major and quadratus lumborum muscles at the L3–L4 level for the 32 randomly selected subjects based on visual evaluation using a 4-point visual scale (0 ¼ no apparent non-muscle tissue (e.g. fat), 1 ¼ minor deposits of non-muscle tissue, 2 ¼ moderate deposits of non-muscle tissue, and Table 1 Frequencies (%) of qualitative muscle composition ratingsa of nonmuscle tissue within erector, psoas major and quadratus lumborum muscles at the L3–L4 level Muscle

Erector muscle Psoas major Quadratus lumborum a

No. nonmuscle tissueb

Minor deposits of nonmuscle tissue

Moderate amount of nonmuscle tissue

High amount of nonmuscle tissue

7 (22) 18 (56) 20 (63)

6 (19) 14 (44) 12 (38)

14 (44) 0 0

5 (16) 0 0

The qualitative muscle composition measurement was a 4-point visual scale (0 ¼ no apparent non-muscle tissue (e.g. fat), 1 ¼ minor deposits of non-muscle tissue, 2 ¼ moderate deposits of non-muscle tissue, and 3 ¼ large areas of non-muscle tissue). b Non-muscle tissue was deﬁned as the amount of white or light gray tissue (pixels) within the CSA of a muscle.

3 ¼ large areas of non-muscle tissue) (Table 1). The non-muscle tissue was identiﬁed as higher intensity pixels within the muscle CSA, while the lower intensity pixels represented muscle tissue. Muscle CSA was also adjusted by the qualitative rating for analyses of the association with isokinetic lifting performance. Intra-rater reliability estimates of quantitative and qualitative muscle composition measurements were based on MR images of 32 randomly selected subjects evaluated twice, 4–8 weeks apart. The subjects MR images were placed in a random order for the second evaluation and the measurer was blinded to the results of the ﬁrst session. Isokinetic lifting testing was performed from a forward-bent, knees straight position, where the forward-bent position was measured by the MIE goniometer (Medical Research Limited, Leeds, UK) on the L3 vertebra at 601 ﬂexion. The subjects received standardized instructions and a demonstration and were asked to lift their body as rapidly and forcefully as possible. The isokinetic lifting is based on a lift movement performed at a concentric, constant speed. The lifting speed was pre-determined to be 0.5 m/s, but applied force was determined solely by the subjects that were given standardized instructions to lift with maximal force. Additional instructions other than to keep the contribution of extremities in force production to a minimum, and maintain their knees, wrists and elbows straight were not given. The reproducibility of the isokinetic lifting test has been previously found to be acceptable (r ¼ 0.87) (Latikka et al., 1995). Lifted forces were recorded in Newton’s, and lift speed and height were entered to calculate the work done in Joules. 2.1.4. Body fat measurement Bioelectrical impedance was used to measure body fat following the manufacturer’s equations (Spectrum II, RJL Systems, Detroit, MI, USA) and electrode settings as described by Van Loan (1990). The coefﬁcient of variation of reproducibility of the measurements has been in the order of 0.2–4.1% (Van Loan, 1990; Vehrs et al., 1998). 2.1.5. Statistical analysis Intra-class correlation coefﬁcients (ICCs) were calculated to evaluate intra-rater reliability of paraspinal muscle CSAs and the mean signal intensity values using a formula based on a two-way analysis of variance with a separate error within the factor (Strout and Fleiss, 1979). In addition, the absolute error (the mean of the difference between ﬁrst and second measurement) and the percentage error (the mean percentage of difference between ﬁrst and second measurement, later referred to as error %) were calculated for the paraspinal muscle CSAs and the mean signal intensity values. Weighted Kappa coefﬁcients (Dunn, 1989) were estimated for

ARTICLE IN PRESS A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

intra-rater reliability of qualitative ratings of nonmuscle tissue within a muscle. Relationships of quantitative measures of muscle CSA, composition and isokinetic lifting strength were evaluated using Pearson correlation coefﬁcients. Kendall’s rank correlation coefﬁcient tau-b was calculated for correlations with qualitative muscle composition ratings. The scale of Landis and Koch (1977) was used in describing correlations, which indicates that the strength of agreement through correlation coefﬁcients should be interpreted as 0.00–0.20 ¼ poor, 0.21– 0.40 ¼ fair, 0.41–0.60 ¼ moderate, 0.61–0.80 ¼ good, and 0.81–1.00 ¼ excellent. The qualitative muscle composition rating was reversed for partial correlations to have the same direction of results with the quantitative muscle composition measurement. Partial correlations were estimated for total muscle CSAs and isokinetic lifting strength measures, adjusting for the qualitative muscle composition measurement or CSF. All analyses were conducted using Stata (StataCorp. 2004).

3. Results Frequencies of qualitative muscle composition ratings of non-muscle tissue in paraspinal muscles are in Table 1. Means of right side paraspinal muscle CSAs, isokinetic lifting force and work and body fat are shown in Table 2.

Table 2 CSA of right side paraspinal muscles, isokinetic lifting force and work and body fat (mean, 95% conﬁdence intervals [95%CI] and range) of subjects Variable

Mean (95%CI)

Range

Total CSA (cm2) Erector muscle, n ¼ 167 Psoas major, n ¼ 167 Quadratus lumborum, n ¼ 167 Body fat (%), n ¼ 169 Isokinetic lifting force (N), n ¼ 169 Isokinetic lifting work (J), n ¼ 169

24.5 (23.9–25.0) 13.3 (12.8–13.7) 6.6 (6.4–6.9) 22.9 (21.9–23.8) 1039 (1000–1077) 569 (540–599)

14.2–34.1 6.5–20.3 2.6–12.6 12.8–40.4 542–1568 159–939

353

3.1. Reproducibility The intra-rater reproducibility of total muscle CSAs at the L3–L4 level was excellent for erector (ICC ¼ 0.98), psoas major (ICC ¼ 0.99) and quadratus lumborum muscles (ICC ¼ 0.95) (Table 3). Reproducibility of the mean signal intensity of the traced areas was also excellent in erector, psoas major and quadratus lumborum muscles (ICC ¼ 0.96–0.99). Reproducibility of qualitative muscle composition ratings of non-muscle tissue were good in erector (Kappa ¼ 0.63) and quadratus lumborum muscles (Kappa 0.66), and moderate in psoas major muscle (Kappa ¼ 0.51) (Table 3). Association of muscle composition measurements and back function. There was an association evident between qualitative muscle composition ratings of non-muscle tissue in the erector muscle CSA and overall body fat (r ¼ 0.41, po0.01). Higher signal intensity suggested more fat inﬁltration of the muscle and being associated with higher body fat. Moderate or weak associations were also detected between different muscle CSA and composition measurements with isokinetic lifting force and work. The relative amounts of non-muscle tissue in CSA reﬂected by qualitative and quantitative muscle composition ratings of erector muscle (r ¼ 0.03), quadratus lumborum (r ¼ 0.23) and psoas major (r ¼ 0.08) were not associated (Table 4). Adjusting CSA measurements by muscle composition measurements did not yield clearly higher associations with isokinetic lifting performance. The amount of fat within paraspinal muscles as indicated by qualitative muscle composition ratings had weak to moderate associations with isokinetic lifting performance (r ¼ 0.19–0.41), similar to those of muscle CSA (r ¼ 0.17–0.32).

4. Discussion In this study, the reliability of quantitative and qualitative muscle composition measurements of paraspinal muscle CSAs from routine lumbar spine magnetic resonance images and association of these composition measurements with back function was examined. In

Table 3 Intra-rater correlation coefﬁcients for quantitative and qualitative muscle composition measurements of right side erector, psoas major and quadratus lumborum muscles at L3–L4 level CSA (cm2)

Mean signal intensity

Qualitative ratings#

ICC

95%CI

Absolute error

Error %

ICC

95% CI

Absolute error

Error %

Kappa

% of

0.98 0.99 0.95

0.96–0.99 0.98–1.00 0.89–0.98

102 34 95

1.01 0.99 3.93

0.99 0.99 0.96

0.99–1.00 0.98–1.00 0.91–0.98

0.24 0.15 0.19

0.69 0.80 20.11

0.63 0.51 0.66

91 84 85

agreement Erector muscle Psoas major Quadratus lumborum

ARTICLE IN PRESS 354

A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

Table 4 Correlations of quantitative and qualitative muscle composition measurements of erector, psoas major and quadratus lumborum muscles with isokinetic lifting force and work Muscle composition measurements

Erector muscle Total CSA (cm2) Mean signal intensity Total CSA adjusted for signal intensitya Qualitative rating of muscle compositionb Total CSA adjusted for qualitative ratingc Psoas major Total CSA (cm2) Mean signal intensity Total CSA adjusted for signal intensitya Qualitative rating of muscle compositionb Total CSA adjusted for qualitative ratingc Quadratus lumborum Total CSA (cm2) Mean signal intensity Total CSA adjusted for signal intensitya Qualitative ratings of muscle compositionb Total CSA adjusted for qualitative ratingc

Isokinetic lifting Force

Work

0.21* 0.14 0.14 0.25 0.23

0.17* 0.08 0.28 0.41** 0.02

0.31*** 0.15 0.33*** 0.19 0.45*

0.32*** 0.11 0.34*** 0.23 0.31

0.26** 0.15 0.24 0.34* 0.08

0.22** 0.10 0.20 0.30 0.08

*po0.05, **po0.01, ***po0.001. a Total CSA adjusted for the signal intensity of that area divided by the signal intensity of adjacent CSF. b Kendall’s rank correlation coefﬁcient tau-b. c Total CSA adjusted for the qualitative rating.

general, the reliability of CSA measurements of erector, psoas major and quadratus lumborum muscles was excellent, which is well in line with previous studies (Gibbons et al., 1997; Peltonen et al., 1998; Marras et al., 2001). The reliability of the mean signal intensity measurements of erector, psoas major and quadratus lumborum muscles was also excellent, as in previous reports (Gibbons et al., 1997; Elliott et al., 2005; Ranson et al., 2005). Qualitative muscle composition ratings were found to be somewhat less reproducible, yielding similar results as those by Kader et al. (2000) of qualitative ratings of non-muscle tissue. In general, the low associations of quantitative and qualitative muscle composition measurements with isokinetic lifting force and work demonstrate that these muscle measurements are not good indicators of functional capacity. Contrary to our original hypothesis, adjusting CSA by muscle composition measurements did not yield consistently higher associations with isokinetic lifting performance measurements than CSA alone. This ﬁnding does not support the construct validity of the muscle composition measurements. On the other hand, the correlations of the qualitative muscle composition measurements and isokinetic lifting performance, while modest, were similar in magnitude as those for muscle CSA and lifting performance, providing

some evidence of muscle composition measurements’ ability to reﬂect the back function. Associations between paraspinal muscle signal intensity measurements and isokinetic and isometric back extension force has been weak in earlier studies as well (Gibbons et al., 1997; Mooney et al., 1997; Keller et al., 2003). In an effort to control the effect of lifting performance on associations with paraspinal muscle composition measures, all subjects received identical instructions to lift as forcefully as possible. In addition, the isokinetic lifting tests were the ﬁrst back function test performed following three warming-up lifts. Of the three test lifts, the best lift with the highest result was selected for the analysis. During the lifting, the tester visually conﬁrmed that the proper test position was maintained during lifting. Recognizing that motivation and other factors could inﬂuence performance were not completely controlled, following the test each subject was asked an open question ‘‘Were there any factors limiting your maximum performance during the lifting?’’ None of the subjects of this sample reported limiting factors. For lower extremities, it has been shown that muscle fatigue was not related to muscle size, quality or strength (Katsiaras et al., 2005). Collectively, our results and those of others suggest that both quantitative and qualitative muscle composition measurements using routine sequences in lumbar spine MR imaging are reliable, but evidence of their construct validity is modest. A factor related to the used muscle composition measurements affecting the magnitude of associations found may have been that clear superﬁcial veins or cavitas of non-muscle tissue at the periphery of the muscle CSAs were excluded. This may have led to measurements indicating lower muscle fat than if fat on the periphery of the muscle had been included, which may have affected the associations with back function and particularly with body fat, since the body fat measurement used in this study takes into account total body fat. There was a relatively high amount of intraand inter-muscular non-muscle tissue for the erector muscle, which was comprised of a combination of three different muscles. The qualitative muscle composition ratings for non-muscle tissue for this muscle mass were more strongly (although only moderately) associated with body fat than muscles with less intramuscular nonmuscle tissue, such as psoas major and quadratus lumborum. Yet, consistently differentiating layers of superﬁcial fat from fascial tissue surrounding the muscle in this region was not possible. Others have also reported weak or absent associations between the amount of fatty tissue within the upper cervical muscles or paraspinal muscles and body composition (Parkkola and Kormano, 1992; Wood et al., 1996; Elliott et al., 2005) and paraspinal muscle atrophy and body weight (Parkkola and Kormano, 1992). In addition, the less

ARTICLE IN PRESS A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

than ideal reproducibility of isokinetic lifting strength (r ¼ 0.87) likely reduced the associations found between isokinetic lifting strength and paraspinal muscle composition measurements. The effect of isokinetic lifting strength measurement accuracy on the observed correlation with paraspinal muscle measurements can be assessed with the formula (Francis et al., 1999): sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 1 r ¼ r, ð1 þ ðV ex =V x ÞÞð1 þ ðV ey =V y ÞÞ 0

where r0 is the clinically measured correlation coefﬁcient, Vex/Vey the variance of error component due to reproducibility of the test, Vx/Vy the variance of the measurement, and r the underlying correlation coefﬁcient assumed to be 1. The standard deviation of isokinetic lifting strength is 239 N and the variance is then 57 121 N2 (Vex and Vx respectively) and the standard deviation of paraspinal muscle CSA is for the total erector spinae 3.67 cm2 giving the variance of 13.47 (cm2)2 (Vey and Vy respectively). Adding these values to the formula above we get a correlation value of 0.78. This suggests, assuming the perfect underlying correlation between the isokinetic lifting strength and the total erector spinae CSA, the expected observed correlation coefﬁcient (r0 ) can only be on the average 0.78. Correlations involving other measures with greater error components, such as qualitative ratings of muscle quality, can be expected to be affected to a great degree. The lower than expected association may also be explained by neuromuscular factors, motor learning or motivation, which play a vital role determining the muscle strength, or other factors. Performance on a ‘back lift’ test, such as the one used is also affected by the muscles in the extremities. A basic principle of muscle physiology is that muscle CSA correlates with muscle strength (Bruce et al., 1997; Keller et al., 1999). However, it must be acknowledged that there are many other factors that inﬂuence strength and its measurement. Maximal lifting effort on isokinetic lifting tests has been shown to be mainly inﬂuenced by heredity (Ropponen et al., 2004) and depends on biological factors, such as muscle ﬁber type and neuromotor function. Neuromotor function has been shown to have a genetic component (Simoneau and Bouchard, 1995). Training status may be reﬂected through more neural activation and decreases in intramuscular fat, however, very few among our sample did resistance training and the main ‘training’ was routine daily activities in work and leisure time. Body fat percentage has at least some theoretical correlation with overall physical activity level (Kyle et al., 2001; LahtiKoski et al., 2002), which may be associated with muscle quality as reﬂected by fatty inﬁltration. Also body fat has been shown to have a signiﬁcant genetic component (Bouchard, 1991; Schousboe et al., 2004).

355

5. Conclusions The CSA and quantitative and qualitative muscle composition measurements of the lumbar paraspinal muscles are highly reliable tissue measurements. The generally low correlations of the quantitative and qualitative muscle composition measurements, alone or in combination with CSA, to isokinetic lifting performance showed that paraspinal muscle morphology based on routine lumbar MRI is poorly related to back function.

Acknowledgements This study was ﬁnancially supported by the Ministry of Education, Urheiluopistosa¨a¨tio¨, Finnish Cultural Foundation, the Finnish Work Environment Fund, the Finnish Academy (grant no. 42044), the National Institutes of Health (US grant no. AR 40857), the Alberta Heritage Foundation for Medical Research, Canada, The Canada Research Chairs Program and EURODISC – an EU-funded project (Contract no. QLK6-CT-2002-02582). We thank Esko Leva¨lahti, MSc, for statistical help. References Battie´ MC, Videman T, Gibbons LE, Fisher LD, Manninen H, Gill K. Determinants of lumbar disc degeneration. A study relating lifetime exposures and magnetic resonance imaging ﬁndings in identical twins. Spine 1995;20:2601–12. Bouchard C. Heredity and the path to overweight and obesity. Medicine & Science in Sports & Exercise 1991;23:285–91. Bruce SA, Phillips SK, Woledge RC. Interpreting the relation between force and cross-sectional area in human muscle. Medicine & Science in Sports & Exercise 1997;29:677–83. Dunn G. Design and Analysis of Reliability Studies. The Statistical Evaluation of Measurement Errors. New York: Oxford University Press; 1989. pp. 30–5, 38. Elliott JM, Galloway GJ, Jull GA, Noteboom JT, Centeno CJ, Gibbon WW. Magnetic resonance imaging analysis of the upper cervical spine extensor musculature in an asymptomatic cohort: an index of fat within muscle. Clinical Radiology 2005;60:355–63. Flicker PL, Fleckenstein JL, Ferry K, Payne J, Ward C, Mayer T, Parkey RW, Peshock RM. Lumbar muscle usage in chronic low back pain. Spine 1993;18:582–6. Francis DP, Coats AJS, Gibson DG. How high can a correlation coefﬁcient be? Effects of limited reproducibility of common cardiological measures. International Journal of Cardiology 1999; 69:185–9. Gibbons LE, Latikka P, Videman T, Manninen H, Battie´ MC. The association of trunk muscle cross-sectional area and magnetic resonance imaging parameters with isokinetic and psychophysical lifting strength and static back muscle endurance. Journal of Spinal Disorders 1997;10:398–403. Gibbons LE, Videman T, Battie´ MC, Kaprio J. Determinants of paraspinal muscle cross-sectional area in male monozygotic twins. Physical Therapy 1998;78:602–10. Guyton AC, Hall JE. Textbook of Medical Physiology, 10th ed. W.B. Saunders Company; 2000. p. 968.

ARTICLE IN PRESS 356

A. Ropponen et al. / Manual Therapy 13 (2008) 349–356

Hultman G, Nordin M, Saraste H, Ohlse´n H. Body composition, endurance, strength, cross-sectional area and density of MM erector spinae in men with and without low back pain. Journal of Spinal Disorders 1993;6:114–23. Jorgensen MJ, Marras WS, Gupta P. Cross-sectional area of the lumbar back muscles as a function torso ﬂexion. Clinical Biomechanics 2003;18:280–6. Kader DF, Wardlaw D, Smith FW. Correlation between the MRI changes in the lumbar multiﬁdus muscles and leg pain. Clinical Radiology 2000;55:145–9. Katsiaras A, Newman AB, Kriska A, Brach J, Krishnaswami S, Feingold E, Kritchevsky SB, Li R, Harris TB, Schwartz A, Goodpaster BH. Skeletal muscle fatigue, strength, and quality in the elderly: the Health ABC study. Journal of Applied Physiology 2005;99:210–6. Keller A, Johansen JG, Hellesnes J, Brox JI. Predictors of isokinetic back muscle strength in patients with low back pain. Spine 1999; 24:275–80. Keller A, Brox JI, Gunderson R, Holm I, Friis A, Reikera˚s O. Trunk muscle strength, cross-sectional area, and density in patients with chronic low back pain randomized to lumbar fusion or cognitive intervention and exercises. Spine 2003;29:3–8. Kim DY, Lee SH, Chung SK, Lee HY. Comparison of multiﬁdus muscle atrophy and trunk extension muscle strength: percutaneous versus open pedicle screw ﬁxation. Spine 2005;30:123–9. Kyle UG, Gremion G, Genton L, Slosman DO, Golay A, Pichard C. Physical activity and fat-free and fat mass by bioelectrical impedance in 3853 adults. Medicine and Science in Sports and Exercise 2001;33:576–84. Ka¨ser L, Mannion AF, Rhyner A, Weber E, Dvorak J, Mu¨ntener M. Active therapy for chronic low back pain. Part 2. Effects on paraspinal muscle cross-sectional area, ﬁber type size, and distribution. Spine 2001;26:909–19. Lackner JM, Carosella AM. The relative inﬂuence of perceived pain control, anxiety, and functional self-efﬁcacy on spinal function among patients with chronic low back pain. Spine 1999;24: 2254–61. Lahti-Koski M, Pietinen P, Helio¨vaara M, Vartiainen E. Associations of body mass index and obesity with physical activity, food choices, alcohol intake, and smoking in the 1982–1997 FINRISK Studies. American Journal of Clinical Nutrition 2002;75:809–17. Landis JK, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74. Latikka P, Battie´ MC, Videman T, Gibbons LE. Correlations of isokinetic and psychophysical back lift and static back extensor endurance tests in men. Clinical Biomechanics 1995;10:325–30. Mannion AF, Ka¨ser L, Weber E, Rhyner A, Dvorak J, Mu¨ntener M. Inﬂuence of age and duration of symptoms on ﬁbre type distribution and size of the back muscles in chronic low back pain patients. European Spine Journal 2000;9:273–81. Marras WS, Jorgensen MJ, Granata KP, Wiand B. Female and male trunk geometry: size and prediction of the spine loading trunk muscles derived from MRI. Clinical Biomechanics 2001;16:38–46. Mayer TG, Vanharanta H, Gatchel RJ, Mooney V, Barnes D, Judge L, et al. Comparison of CT scan muscle measurements and isokinetic trunk strength in postoperative patients. Spine 1989; 14:33–6. McGill SM, Santaguida L, Stevens J. Measurement of the trunk musculature from T5 to L5 using MRI scans of 15 young males

corrected for muscle ﬁbre orientation. Clinical Biomechanics 1993; 8:171–8. Mengiardi B, Schmid MR, Boos N, Pﬁrrmann CWA, Brunner F, Elfering A, Hodler J. Fat content of lumbar paraspinal muscles in patients with chronic low back pain and in asymptomatic volunteersquantiﬁcation with MR spectroscopy. Radiology 2006; 240:786–92. Parkkola R, Kormano M. Lumbar disc and back muscle degeneration on MRI: correlation to age and body mass. Journal of Spinal Disorders 1992;5:86–92. Peltonen JE, Taimela S, Erkintalo M, Salminen JJ, Oksanen A, Kujala UM. Back extensor and psoas muscle cross-sectional area, prior physical training, and trunk muscle strength—a longitudinal study in adolescent girls. European Journal of Applied Physiology 1998; 77:66–71. Ranson CA, Burnett AF, Kerslake R, Batt ME, O’Sullivan PBO. An investigation into the use of MR imaging to determine the functional cross-sectional area of lumbar paraspinal muscles. European Spine Journal 2005; May 14 [Epub ahead of print]. Reid JG, Costigan PA. Trunk muscle balance and muscular force. Spine 1987;12:783–6. Ropponen A, Leva¨lahti E, Videman T, Kaprio J, Battie´ MC. The role of genetics and environment in isokinetic and psychophysical lifting capacity and isometric trunk extension endurance. Physical Therapy 2004;84:608–21. Ra¨ty HP, Kujala U, Videman T, Koskinen SK, Karppi SL, Sarna S. Associations of isometric and isoinertial trunk muscle strength measurements and lumbar paraspinal muscle cross-sectional areas. Journal of Spinal Disorders 1999;12:266–70. Mooney V, Gulick J, Perlman M, Levy D, Pozos R, Leggett S, Resnick D. Relationships between myoelectric activity, strength and MRI of lumbar extensor muscles in back pain patients and normal subjects. Journal of Spinal Disorders 1997;10:348–56. Schick F, Machann J, Brechtel K, Strempfer A, Klumpp B, Stein DT, Jacob S. MRI of muscular fat. Magnetic Resonance in Medicine 2002;47:720–7. Schousboe K, Visscher PM, Erbas B, Kyvik KO, Hopper JL, Henriksen JE, Heitmann BL, Sørensen TIA. Twin study of genetic and environmental inﬂuences on adult body size, shape, and composition. International Journal of Obesity 2004;28:39–48. Simoneau J, Bouchard C. Genetic determinism of ﬁber type proportion in human skeletal muscle. The Federation of American Societies for Experimental Biology Journal 1995;9:1091–5. Simonen R, Videman T, Kaprio J, Leva¨lahti E, Battie´ MC. Factors associated with exercise lifestyle—a study of monozygotic twins. International Journal of Sports Medicine 2003;24:499–505. StataCorp. Stata statistical software: Release 8.2. College Station, TX: Stata Corporation; 2004. Strout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychology Bulletin 1979;86:420–8. Van Loan MD. Bioelectrical impedance analysis to determine fat-free mass, total body water and body fat. Sports Medicine 1990;10: 205–17. Vehrs P, Morrow Jr JR, Butte N. Reliability and concurrent validity of Futrex and bioelectrical impedance. International Journal of Sports Medicine 1998;19:560–6. Wood S, Pearsall DJ, Ross S, Reid JG. Trunk muscle parameters determined from MRI for lean to obese males. Clinical Biomechanics 1996;11:139–44.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 357–366 www.elsevier.com/locate/math

Original article

Sustained natural apophyseal glides (SNAGs) are an effective treatment for cervicogenic dizziness Susan A. Reid, Darren A. Rivett, Michael G. Katekar, Robin Callister School of Biomedical Sciences, Faculty of Health, The University of Newcastle, Callaghan, NSW 2308, Australia Received 2 June 2006; received in revised form 2 February 2007; accepted 11 March 2007

Abstract Cervicogenic dizziness is dizziness described as imbalance occurring together with cervical pain or headache. This study aimed to determine the efﬁcacy of sustained natural apophyseal glides (SNAGs) in the treatment of this condition. A double-blind randomised controlled clinical trial was undertaken. Thirty-four participants with cervicogenic dizziness were randomised to receive four to six treatments of SNAGs (n ¼ 17) or a placebo of detuned laser (n ¼ 17). Participants were assessed by a blinded assistant before treatment, after the ﬁnal treatment and at 6- and 12-week follow-ups. The primary outcome measures were severity of dizziness, disability, frequency of dizziness, severity of cervical pain, and global perceived effect; balance and cervical range of motion were secondary measures. At post-treatment, 6- and 12-week follow-ups compared to pre-treatment, the SNAG group had less (Po0.05) dizziness, lower (Po0.05) scores on the Dizziness Handicap Inventory (DHI), decreased (Po0.05) frequency of dizziness, and less (Po0.05) cervical pain. The placebo group had signiﬁcant (Po0.05) changes only at the 12-week follow-up in three outcome measures: severity of dizziness, DHI, and severity of cervical pain. Compared to the placebo group at post-treatment and 6-week follow-up, the SNAG group had less (Po0.05) dizziness, lower (Pp0.05) scores on DHI, and less (Po0.05) cervical pain. Balance with the neck in extension improved (Pp0.05) and extension range of motion increased (Po0.05) in the SNAG group. No improvements in balance or range of motion were observed in the placebo group. The SNAG treatment had an immediate clinically and statistically signiﬁcant sustained effect in reducing dizziness, cervical pain and disability caused by cervical dysfunction. r 2007 Elsevier Ltd. All rights reserved. Keywords: Dizziness; SNAG; Manual therapy; Randomised controlled trial; Cervical spine

1. Introduction Dizziness is a common presenting problem in clinical practice (Luxon, 1984; Kroenke and Mangelsdorff, 1989; Shumway-Cook and Horak, 1989; Kroenke et al., 1992; Colledge et al., 1996; Furman and Whitney, 2000) and is particularly prevalent in the elderly (Luxon, 1984; Colledge et al., 1996). Dizziness has substantial physical, social and emotional effects as well as ﬁnancial consequences on individuals and the community (Yardley et al., 1992). The physical problems include postural instability, unsteadiCorresponding author. Tel.: +61 2 49813111; fax: +61 2 49812111.

E-mail address: [email protected] (S.A. Reid). 1356-689X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.03.006

ness and falls. Yardley et al. (1992) found dizziness often leads to increased dependence on family members, anxiety about public exposure, withdrawal from work, travel and social activities, and depression. Dizziness is a term used to describe a wide range of symptoms, which have been loosely grouped into four main types: faintness, imbalance, vertigo, and disorientation (Drachman and Hart, 1972; Froehling et al., 1994; Enloe and Shields, 1997). One speciﬁc type of dizziness is cervicogenic dizziness, which is thought to be caused by dysfunction in the upper cervical spine. It is characterised by symptoms of dizziness described as imbalance or disequilibrium, commonly associated with neck pain, stiffness or headache (Wrisley et al., 2000), and is

ARTICLE IN PRESS 358

S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

frequently related to a neck injury such as whiplash or to spinal degeneration. It is a diagnosis of exclusion; other causes of dizziness must be excluded based on history, physical examination and vestibular function tests (Wrisley et al., 2000; Heikkila, 2004). There is evidence that cervicogenic dizziness is a result of perturbation in the information from sensory afferents in the cervical spine (de Jong et al., 1977; Brandt and Bronstein, 2001). Balance problems can be the result of abnormal afferent input to the vestibular nuclei from altered cervical articular proprioceptors (Terrahe, 1979; Wyke, 1979). Cervicogenic dizziness should be distinguished from the most common type of dizziness, vertigo, which is a spinning sensation and usually due to a problem with the vestibular system (Brandt, 1991; Heikkila, 2004). To date there is limited evidence to support the use of manual therapy in treating cervicogenic dizziness. A systematic review of the literature on this topic conducted by Reid and Rivett (2005) found that randomised controlled trials of all manual therapy treatments for cervicogenic dizziness were scarce and generally of poor methodological quality. Despite this limited evidence, manual therapy is advocated to treat cervicogenic dizziness in the clinical setting (Karlberg et al., 1996b; Galm et al., 1998; Zhou et al., 1999; Bracher et al., 2000). Brian Mulligan, a New Zealand physiotherapist, described a manual therapy treatment for this condition in 1991 entitled sustained natural apophyseal glides (SNAGs) (Mulligan, 1991). A SNAG is a sustained passive accessory glide in the plane of the zygapophyseal joint performed by the therapist while the patient actively moves their neck physiologically in the symptomatic direction (Mulligan, 1991). The accessory glide provides immediate pain/symptom relief when moving in the provocative direction (Mulligan, 1991). Since 1991 its use clinically has spread among physiotherapists despite not having been evaluated in any clinical trials (Exelby, 1995; Wilson, 1996; Mulligan, 1999). Thus, the aim of the present study was to determine the efﬁcacy of SNAGs in the treatment of the signs and symptoms of patients with cervicogenic dizziness. If effective, SNAGs would provide an option for patients, manual therapists and medical practitioners who previously had no evidence-based treatment for this disabling problem. In addition, if manual therapy to the cervical spine is shown to be an effective treatment for cervicogenic dizziness, it would provide indirect evidence of the existence and origins of this disorder, a topic of some controversy in the literature (Karlberg et al., 1996a; Brandt and Bronstein, 2001). 2. Methods 2.1. Research design This study was a double-blind randomised controlled clinical trial, conducted at the University of Newcastle,

Australia. Participants were randomly assigned to either a SNAGs manual therapy treatment group (SNAG group) or a sham laser placebo group. A research assistant blinded to group allocation performed all the evaluation measurements. The study was approved by the University of Newcastle Human Research Ethics Committee. All participants provided written informed consent prior to participation. 2.2. Sample size Calculations to determine the required group sample sizes (n ¼ 17 for each group) were based on the differences in the DHI and VAS for dizziness scores that would be needed for a statistically signiﬁcant effect of a treatment based on a change in the DHI of 15 units and a change in the VAS for dizziness of 2 units, with power set at 80% and the signiﬁcance level at 0.05. The magnitude of these differences in scores was determined from previous studies (Koes et al., 1992a, b; Newman and Jacobsen, 1993; Enloe and Shields, 1997; Storper et al., 1998; O’Reilly et al., 2000) using these evaluation tools to indicate clinically meaningful differences. The calculations were performed by a statistician using the PS Power and Sample Size Calculation computer program available online from the Vanderbilt Medical Centre (www.mc.vanderbilt.edu). 2.3. Subjects The inclusion and exclusion criteria for the participants in the study are provided in Table 1. Participants had to meet all the inclusion criteria to enter the study, but were excluded if they met any of the exclusion criteria. 2.4. Recruitment and screening Participants with suspected cervicogenic dizziness were recruited from a press release by the University’s media unit (378 respondents), and respondents were screened by telephone interview by a manipulative physiotherapist. After questioning, 83 respondents with dizziness described as imbalance or unsteadiness combined with a stiff or painful neck or headache were considered potential participants. Nine other participants were directly referred to the study by neurologists in the local region. All potential participants were invited to undergo comprehensive assessment by a neurologist (Table 2) in which other causes of dizziness (Table 1) were excluded. Following neurological assessment of 59 people, 25 people were excluded because they were diagnosed with other causes of dizziness and 34 entered the trial (Fig. 1).

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366 Table 1 Inclusion and exclusion criteria Inclusion criteria Dizziness described as imbalance or unsteadiness (not rotatory vertigo) Dizziness related to either movements or positions of the cervical spine, or occurring with a stiff or painful neck Symptoms 43 months 18–90 years of age Exclusion criteria By neurologist Vestibular disorders (e.g. Benign Paroxysmal Positional Vertigo, Meniere’s disease, peripheral vestibulopathy) CNS disorder (e.g.cerebellar ataxia, stroke, demyelination) Mal de debarquement syndrome Migraine associated vertigo Psychogenic dizziness Cardiovascular disorders A history of Vestibular disorder Active inﬂammatory joint disease Spinal cord pathology Cervical spine cancer or infection Bony disease or marked osteoporosis Marked cervical spine disc protrusion Acute cervical nerve root symptoms (severe pain, weakness, pins and needles or numbness in the arm or hand for less than six weeks) Recent (in the previous three months) fracture or dislocation of the neck Previous surgery to the upper cervical spine Physiotherapy or any manual therapy treatment to the neck in the previous month Pregnancy Inability to read English

2.5. Randomisation Participants were randomly allocated to either the SNAG group (n ¼ 17) or the placebo group (n ¼ 17) by choosing an opaque envelope with a computer generated random number inside. This number was then checked against a list (prepared by a biostatistician) that allocated the participant to either the placebo group or the SNAG group. The participants were unaware that the laser procedure was a sham treatment, so were blind to whether they were in the treatment or placebo group. The success of the blinding was evaluated by the Global Perceived Effect (GPE) Questionnaire (Koes et al., 1992a, b). 2.6. Interventions Both groups of participants received their interventions 4–6 times over 4 weeks. Most participants (15 participants in each group) received it four times. At the discretion of the treating manipulative physiotherapist,

359

Table 2 Clinical examination process used by the neurologist for determining the presence of cervicogenic dizziness 1. Presence of symptoms and signs of musculoskeletal pain and stiffness in the neck. 2. Current history of positionally aggravated dizziness, unsteadiness or imbalance; or a persisting disequilibrium. Volunteers with true rotatory vertigo were considered to have other disorders. 3. Exclusion of alternative causes of dizziness. These included, but were not limited to Benign Paroxysmal Positional Vertigo, peripheral vestibulopathy, Meniere’s disease, stroke, demyelination and other CNS disorders, and migraine associated vertigo. This was based on the history, standard neurological and neuro otological examination including: Hallpike test Examination for positional, gaze evoked or head shaking nystagmus Head thrust vestibulo-ocular reﬂexes Ocular smooth pursuit and saccades testing Romberg and Unterberger tests Tandem gait and blind gait testing. 4. Following clinical examination, caloric testing was performed in the neurologist’s laboratory by a trained technician to exclude other vestibular causes of dizziness.

two extra sessions were given to two participants in the SNAG group for whom it was felt it would be beneﬁcial to receive six treatments in 4 weeks. This was matched in the placebo group with two participants also receiving six sessions. All participants were asked to avoid new co-interventions throughout their participation in the trial, and as far as can be ascertained, all participants complied with this request. 2.6.1. SNAGS intervention One group of participants received the SNAG treatment as described by Brian Mulligan (Mulligan, 1999). From the participant’s history, the offending active cervical movement (i.e. the movement that predominantly caused their dizziness) was ascertained and used to determine the treatment direction. Mulligan suggests the most common active movement to bring on dizziness is cervical extension, though it can be provoked by rotation or ﬂexion (Mulligan, 1999). Active movement did not elicit reproducible symptoms (dizziness or cervical pain) in all participants at the time of treatment. In these cases, the provocative active cervical movement for dizziness, as determined by the history, was used in the SNAG treatment. With the participant in an upright (weight-bearing) sitting position, a manipulative physiotherapist of 18 years clinical experience applied a sustained passive accessory movement (glide) while the participant moved actively through their available physiological range in the direction that produced their symptoms. This was repeated six times. The participant was asked to report

ARTICLE IN PRESS 360

S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

Assessed for eligibility (n = 59) n = 50 from press release n=9 from neurologists Enrollment

Excluded by exclusion criteria (n = 25)

Randomized (n = 34)

Allocated to placebo group (n = 17) Received allocated intervention (n = 17)

Allocated to SNAG group (n = 17) Allocation

Received allocated intervention (n = 16) One participant left study due to previous bad experience with manual therapy

6-week follow-up (n = 17) 12-week follow-up (n = 17) for self-report (n = 15) for physical assessments (2 unable to attend)

Follow-Up

Analyzed (n = 17) for self-report (n = 15) for physical assessments

6-week follow-up (n = 16) 12-week follow-up (n = 16)

Analysis

Analyzed (n= 16)

Fig. 1. Flow chart of progression of participants through the study.

any dizziness or other symptoms during the application of the procedure to ensure the treatment was symptomfree. If the participant reported some dizziness or cervical pain then either the range of active movement was decreased, the angle of the glide was slightly altered or the point of application was changed to ensure the treatment was symptom-free in all participants. Consistent with the recommendations of Mulligan (1999), if cervical spine extension or ﬂexion was symptomatic then the glide was applied ventrally to the C2 spinous process while the participant slowly, actively extended or ﬂexed their neck. If left rotation was symptomatic then the anterior glide was applied to the left C1 transverse process while the participant rotated their neck slowly to the left. If dizziness persisted, the glide was then applied to the contralateral right transverse process of C1, while the participant still

performed active left rotation. The comparable approach was used for symptomatic right rotation. To progress the treatment at subsequent sessions, the number of repetitions of the SNAG was increased from six to 10. If the participant reported in their history that a sustained cervical position such as prolonged extension (e.g. changing a light bulb) or ﬂexion (e.g. reading) produced their dizziness, the end-range physiological position was maintained for up to 10 s at the discretion of the treating manipulative physiotherapist. A second active movement was added if symptoms persisted in that direction, again at the discretion of the treating physiotherapist. 2.6.2. Placebo procedure The second group of participants received a placebo consisting of sham treatment with a detuned laser and

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

carried out by the same therapist. The laser device used in the study (Omni Laser Micro 301, Serial No. 1875, manufactured by Laserdyne P/L, 17 Production Ave, Ernest, Queensland, Australia) had been deactivated by the manufacturer to produce no effective emission. The laser appeared to operate normally, emitting a light signal and a beeping sound. The detuned laser, which has been shown to have a very strong placebo effect (Irnich et al., 2001), was used for six applications of 20 s to various sites on the upper cervical spine, at a distance of 0.5–1 cm from the skin. Sham laser was chosen as it does not activate somatosensory receptors (Irnich et al., 2001).

beneﬁt, 4 ¼ a lot of beneﬁt, 5 ¼ great beneﬁt, 6 ¼ maximal beneﬁt) as used in other studies (Koes, 1991; Koes et al., 1992a, b). Secondary outcome measures were assessed as follows at pre-treatment, post-treatment and at the 12-week follow-up.

2.7. Outcome measures Outcome measurements were obtained pre-treatment, following the ﬁnal treatment, and at 6 and 12 weeks after the ﬁnal treatment (16–18 weeks after randomisation). The primary outcomes were the following selfreport measures:

Severity of dizziness (an average level over the previous few days) was measured with a 10 cm visual analogue scale (VAS). The VAS has been used to measure dizziness in other studies (Gill-Body et al., 1994; Cohen, 1999; Heikkila et al., 2000; Boismier and Disher, 2001). Disability caused by dizziness was measured with the Dizziness Handicap Inventory (DHI). The DHI assesses the impact of dizziness on the functional, emotional and physical aspects of everyday life (Jacobsen and Newman, 1990). The highest possible score is 100, indicating maximum self-perceived handicap. The DHI has been shown to be a highly reliable and responsive tool (Newman and Jacobsen, 1993; Enloe and Shields, 1997). Signiﬁcant correlations between DHI scores and speciﬁc objective measures of balance and gait have been demonstrated (Whitney et al., 2004; Treleaven, 2006). Frequency of dizziness was measured on a six-point rating scale (0 ¼ no dizziness, 1 ¼ dizziness less than once per month, 2 ¼ 1–4 episodes of dizziness per month, 3 ¼ 1–4 episodes of dizziness per week, 4 ¼ dizziness once daily, 5 ¼ dizziness more than once a day or constant). This scale was used by Karlberg et al. (1996b) and Gill-Body et al. (1994) to measure change in frequency of dizziness after manual therapy. Severity of cervical pain and headache (an average level over the previous few days) was assessed with a 10 cm VAS. There is much evidence supporting the validity of the VAS for measuring pain intensity (Scott and Huskisson, 1976; Turk and Melzack, 1992; Murphy et al., 1998; Maher et al., 2000; Scudds, 2001). GPE was measured by self-assessment on a six-point scale (1 ¼ no beneﬁt, 2 ¼ minimal beneﬁt, 3 ¼ some

361

Balance was measured with the Chattecx Balance Dynamic System (Serial No. 1001, Chattecx Corporation, the Chattanooga Group, Tennessee) for 10 s during quiet standing with eyes open, with eyes closed, and in cervical extension. The sway index (SI) obtained is a computer-generated single number that expresses the extent of sway in each direction over four footplates. Cervical range of motion was measured with the cervical range of motion instrument (CROM; Performance Attainment Associates, St Paul, MN) by the research assistant, who had been trained in the use of the CROM. Single measurements of cervical ﬂexion, extension, left rotation, right rotation, left lateral ﬂexion and right lateral ﬂexion were obtained. Moderate to high reliability of CROM measurements has been demonstrated independent of the investigator performing the measurements (Dhimitri et al., 1998). Also, CROM values have been shown to be highly correlated (0.97) with the radiographic method for measuring cervical ﬂexion and extension, strongly supporting its validity (Tousignant et al., 2000).

2.8. Statistical analysis An intention-to-treat analysis was performed on the study cohort. The baseline characteristics of the two groups were compared using univariate analyses. T-tests were used for continuous measures and w2 tests for categorical data. An alpha level of 0.05 was chosen to indicate a signiﬁcant difference between groups (Altman, 1996). The effects of the interventions were determined by two-way repeated measures analyses of variance with group (treatment) as the between factor and time as the repeated factor. Subsequent analyses depended on the presence of interactions or main effects, and included repeated measures analysis of each group independently and group comparisons at each time point. Also, the difference between the assessment scores at each time point and the baseline scores was calculated for each outcome measure. These differences were also calculated as percentage change. 2.9. Clinical significance A change of 15 units in DHI scores was considered clinically relevant (Newman and Jacobsen, 1993; Enloe

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

362

and Shields, 1997; Storper et al., 1998; O’Reilly et al., 2000). A change of 2 points in VAS scores for the main complaint (in this case dizziness and pain) was considered clinically signiﬁcant (Koes et al., 1992a, b).

3. Results 3.1. Participants The characteristics of the participants prior to treatment are shown in Table 3. The groups were similar for age, gender and all outcome measures, except posturography. One participant in the SNAG group withdrew from the study after the second treatment as she had previously (prior to the study) had an adverse experience with manual therapy (Fig. 1). No participants experienced any adverse effects from the interventions received in the study. 3.2. Severity of dizziness responses to interventions In the SNAG group dizziness was signiﬁcantly less severe post-treatment (P ¼ 0.001), at 6 weeks (Po0.001) and at 12 weeks (Po0.001) compared to pre-treatment Table 3 Characteristics of participants with cervicogenic dizziness prior to treatment Characteristic

Placebo group (n ¼ 17)

SNAG group (n ¼ 17)

P

Gender no. (%)a Age (years) VAS dizziness DHI Frequency of dizziness VAS pain Posturography Quiet standing/eyes open Eyes closed Cervical extension/eyes open CROM (deg) Cervical ﬂexion Cervical extension Left rotation Right rotation Left lateral ﬂexion

F:10 63.6 5.6 43.9 3.4 4.7

F:11(65) 63.4 (13.1) 5.6 (2.2) 44.4 (18) 3.3 (1.4) 4.7 (2.5)

0.12b 0.97 1.00 0.94 0.82 1.00

3.6 (1.4)

5.2 (2.2)

0.02

6.6 (2.6) 7.7 (4.5)

10.7 (10.1) 10.3 (8.0)

0.12 0.26

(59) (13.7) (2.9) (18.4) (1.1) (2.2)

(Fig. 2). The placebo group had no change posttreatment or at 6 weeks, although there was a signiﬁcant decrease by 12 weeks (P ¼ 0.02) (Fig. 2). Dizziness was signiﬁcantly less severe in the SNAG group compared to the placebo group post-treatment (P ¼ 0.03) and at 6 weeks (P ¼ 0.03) but no longer different at the 12-week follow-up (P ¼ 0.09). Using the criteria for clinical signiﬁcance of a 2-unit change in scores, the SNAG group had a clinically signiﬁcant reduction in dizziness severity post-treatment (mean change 3.1 units) that was maintained at the 12-week follow-up (mean change 3.3 units). The changes in the placebo group were not clinically signiﬁcant (mean changes of 1.0 post-treatment and 1.9 units at 12-weeks follow-up. 3.3. DHI responses to interventions In both the SNAG and placebo groups pre-treatment, 30% of participants reported mild handicap from their dizziness (scoring 0–30 on DHI), 53% moderate (score 31–60) and 17% severe handicap (score 61–100). Posttreatment in the SNAG group 68% of participants reported mild handicap, 32% moderate and none severe. Post-treatment in the placebo group 22% reported mild, 66% moderate and 12% severe handicap. The SNAG group had signiﬁcantly lower scores on DHI at all reassessment times compared to pre-treatment (P ¼ 0.001 post-treatment, Po0.001 at 6 weeks, Po0.001 at 12 weeks), while the placebo group had a signiﬁcant change only at the 12-week follow-up (P ¼ 0.01) (Fig. 3). The SNAG group had lower scores on DHI compared to the placebo group post-treatment (P ¼ 0.02) and at the 6-week follow-up (P ¼ 0.05), but this difference was not statistically signiﬁcant at the 12-week follow-up (P ¼ 0.06).

10

46.2 42.7 28.9 29.1 47.2

(13.7) (12.5) (11.0) (10.4) (11.1)

40 43.4 30.6 29 44.9

(15.1) (13.2) (8.1) (10.1) (12.2)

0.23 0.88 0.61 0.98 0.57

Values are mean (SD) unless stated otherwise. Note: SNAG ¼ sustained natural apophyseal glide; M ¼ males; F ¼ females; DHI ¼ dizziness handicap inventory; VAS ¼ visual analogue scale; CROM ¼ cervical range of motion. Signiﬁcant at Pp0.05. a For gender, the number and percentage of participants are reported. b 2 X.

VAS Dizziness Scores

Placebo Gp

SNAGs Gp

8 6 ∗ 4 2

Pre Treatment

∗

∗

Post Treatment

6-Weeks Follow-up

∗

12-Weeks Follow-up

Fig. 2. Effect of interventions on Visual Analogue Scale for dizziness scores for Placebo group (n ¼ 17) and SNAG group (n ¼ 16) over time. Data are mean7SE. *Signiﬁcantly different (Po0.05) to pretreatment score.

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

5

100

Placebo Gp

80 60 ∗ 40 20

Pre Treatment

∗

∗

∗

Post Treatment

6-Weeks Follow-up

12-Weeks Follow-up

Fig. 3. Effect of interventions on Dizziness Handicap Inventory scores for Placebo group (n ¼ 17) and SNAG group (n ¼ 16) over time. Data are mean7SE. *Signiﬁcantly different (Po0.05) to pre-treatment score.

For the SNAG group the change in score (18.1 posttreatment and 19.7 at 12-week follow-up) exceeded the 15 points required for a clinically signiﬁcant true change that is attributable to the treatment. The small decrease observed in DHI for the placebo group (2.2 posttreatment and 6.5 at 12-week follow-up) did not reach this level and therefore, could be due to variability in the measuring tool (Newman and Jacobsen, 1993). 3.4. Frequency of dizziness responses to interventions At initial assessment 48% of participants reported dizziness at least once daily. The SNAG group had signiﬁcantly decreased frequency of dizziness posttreatment (P ¼ 0.02), at 6 weeks (P ¼ 0.02) and at 12 weeks (P ¼ 0.03) compared to pre-treatment (Fig. 4). For the placebo group there was no signiﬁcant change at any time point (Fig. 4). There was no signiﬁcant difference in the frequency of dizziness episodes between the SNAG and placebo groups at any of the assessment times.

Frequency of Dizziness

SNAGs Gp

SNAGs Gp

4 3 2

∗

∗

∗

6-Weeks Follow-up

12-Weeks Follow-up

1

Pre Treatment

Post Treatment

Fig. 4. Effect of interventions on frequency of dizziness scores for Placebo group (n ¼ 17) and SNAG group (n ¼ 16) over time. Data are mean 7 SE. *Signiﬁcantly different (Po0.05) to pre-treatment score.

10 Placebo Gp

SNAGs Gp

8 VAS Pain Scores

Placebo Gp

DHI Scores

363

6 ∗

4 2

Pre Treatment

∗

∗

∗

Post Treatment

6-Weeks Follow-up

12-Weeks Follow-up

Fig. 5. Effect of interventions on Visual Analogue Scale for pain scores for Placebo group (n ¼ 17) and SNAG group (n ¼ 16) over time. Data are mean7SE. *Signiﬁcantly different (Po0.05) to pretreatment score.

achieved clinical signiﬁcance at 12-weeks follow-up with a 32% decrease in pain.

3.5. Severity of pain responses to interventions 3.6. Global perceived effect responses to interventions The SNAG group had signiﬁcantly less pain posttreatment (Po0.001), at 6 weeks (P ¼ 0.001) and at 12 weeks (P ¼ 0.01) compared to pre-treatment (Fig. 5). There was no signiﬁcant change in the placebo group post-treatment and at 6 weeks, although there was a signiﬁcant change by 12 weeks (P ¼ 0.02) (Fig. 5). The SNAG group had less pain than the placebo group posttreatment (P ¼ 0.001) and at 6-week follow-up (P ¼ 0.048) but not at 12 weeks (P ¼ 0.32). In the SNAG group, pain decreased 69% post-treatment and 54% at 12-weeks follow-up, easily meeting the criteria for a clinically signiﬁcant change of 28%. The placebo group was only 4% lower post-treatment but just

The GPE ratings of the SNAG group were signiﬁcantly higher than those of the placebo group posttreatment, and at the 6- and 12-week follow-ups (all Po0.001) (Table 4). On average, participants in the SNAG group rated the treatment as being ‘a lot of beneﬁt’ post-treatment and at 6-week follow-up and of ‘great beneﬁt’ at the 12-week follow-up. Participants receiving the placebo intervention rated it as having ‘minimal beneﬁt’ to ‘some beneﬁt’ at all assessment points suggesting that, although it had much less effect than SNAGs, they did not know it was a placebo, and therefore blinding to group allocation was successful.

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

364 Table 4 Scores for global perceived effect

Posttreatment 6-week follow-up 12-week follow-up

Placebo group (n ¼ 17)

SNAG group (n ¼ 16)

Mean difference (95% CI of the difference)

P

2.1 (1.2)

4.1 (1.2)

1.95 (2.87, 1.13)

o0.001

2.4 (1.1)

4.3 (1.3)

1.90 (2.75, 1.04)

o0.001

2.4 (1.1)

4.5 (1.2)

2.15 (2.97, 1.32)

o0.001

Values are mean (SD) with mean differences (95% conﬁdence intervals [CI]).Note: SNAG ¼ sustained natural apophyseal glide. Signiﬁcant at 0.05.

3.7. Balance responses to interventions There was no signiﬁcant change (P ¼ 0.29) in SI during quiet standing in the SNAG group at any stage following treatment, whereas the SI increased signiﬁcantly (P ¼ 0.01) (i.e. poorer balance) at the 12-week follow-up for the placebo group. When the SI was measured with the participant’s eyes closed, there was a 24% decrease in SI from pre- to post-treatment in the SNAG group, although this was not statistically signiﬁcant. During balance assessment with the cervical spine in extension, the SI for the SNAG group was lower (i.e. better balance) at the 12-week follow-up (P ¼ 0.05) but there was no change in the placebo group (P ¼ 0.23). 3.8. Cervical range of motion responses to interventions There was a signiﬁcant increase (P ¼ 0.01) in cervical extension range of motion in the SNAG group posttreatment with an increase in mean range of movement from 441 to 501 with only minor regression at the 12-week follow-up (491). In the placebo group there was no signiﬁcant change in extension range of movement. In the SNAG group there was a trend for increased cervical left rotation range of motion after treatment (P ¼ 0.065) and an increase at the 12-week follow-up (P ¼ 0.05). There was no signiﬁcant change in the placebo group. There was no statistically signiﬁcant difference between the two groups at any time, although for cervical extension range of motion post-treatment and at the 12-week follow-up there was a trend (P ¼ 0.09) for the two groups to be different.

4. Discussion This is the ﬁrst time the SNAG manual therapy technique has been investigated in the treatment of cervicogenic dizziness or any other cervical spine

condition. The results provide evidence of the beneﬁts of this manual therapy technique for cervicogenic dizziness and pain, and support the cervicogenic origins of this form of dizziness. Although there is some evidence that multi-modal treatment may be effective for cervicogenic dizziness (Karlberg et al., 1996b), this is the ﬁrst time an isolated manual therapy technique, which is more speciﬁc and less time consuming than multi-modal treatment, has been shown to be effective (Gross et al., 2002; Reid and Rivett, 2005) . The key ﬁndings of this study were that the SNAG treatment of cervicogenic dizziness results in substantial and clinically meaningful decreases in the severity of the dizziness (measured with VAS), disability associated with the condition (measured by the DHI), frequency of dizziness (measured on a six-point scale) and the severity of the cervical pain (measured with VAS). The improvement in the SNAG treatment group was immediate and maintained over the 12-week follow-up period after the cessation of treatment. Improvements in DHI and severity of dizziness and pain in the placebo group were of lesser magnitude and followed a much slower time course, reaching statistical but not necessarily clinical signiﬁcance at the 12-week follow-up. The gradual improvement over time in the placebo group could be due to the placebo intervention, or more likely, to be a result of slow natural resolution over time. It is reasonable to assume that most patients would prefer greater and more immediate relief from these disabling symptoms. Based on GPE ratings, the SNAG treatment was perceived by the participants to be of greater beneﬁt than the placebo intervention. Most participants receiving SNAGs were very satisﬁed with their treatment and there were no adverse effects. One of the clear beneﬁts of the SNAG treatment was the shorter time frame over which it proved effective. For the four primary outcome measures (severity of dizziness, DHI, frequency of dizziness and severity of pain) the effectiveness of the treatment was observed at the end of the intervention period in the SNAG treatment group. Importantly, this improvement was maintained throughout the follow-up period. Using the DHI, this study was able to demonstrate that cervicogenic dizziness had a substantial negative impact on the functional, emotional, and physical aspects of everyday life. Prior to treatment, the participants had a mean score on the DHI of 44.4, and four participants had severe disability, scoring between 76 and 78 out of a possible 100. After receiving the SNAG treatment there was a signiﬁcant reduction in the DHI score (mean 24.7) for that group. This improvement indicates that the SNAG treatment had a marked, positive effect on their quality of life and level of disability associated with their dizziness. It is possible that the SNAG treatment may normalise input from Type IV nociceptors, as there was a decrease

ARTICLE IN PRESS S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

in pain parallel to the decrease in dizziness in participants treated with SNAGs. The rapid pain relieving effect produced by SNAGs is similar to the rapid pain relief reported in the treatment of lateral epicondylalgia using another Mulligan technique, mobilisation with movement. This pain relief was shown to be the result of activation of the non-opioid descending noradrenergic system of analgesia (Paungmali et al., 2003). Extension was the movement that most commonly reproduced the participants’ symptoms and was thus the direction of movement used in the SNAG treatment in 59% of cases. Mulligan (1999) reported that cervical extension is the most common movement linked to cervicogenic dizziness. A signiﬁcant increase in cervical extension range of motion was seen at post-treatment in the SNAG group, and this was maintained at the 12-week follow-up. No improvement in cervical extension was found in the placebo group. Since both dizziness and extension range of motion improved there is possibly a relationship between the two. Indeed, it has been shown experimentally that loss of normal input from Type I cervical articular mechanoreceptors leads to dizziness and poor balance (Wyke, 1979). Previous studies have shown the presence of mechanoreceptive and nociceptive nerve endings in the cervical zygapophyseal joint capsules which implies that neural input from these joints is important in proprioception and pain sensation (Wyke, 1979; Hulse, 1983; Hinoki, 1985). It has been proposed that joint hypomobility may lead to pain and further reduction in movement (Kaltenborn and Evjenth, 1989; Oostendorp et al., 1993), and consequently a decrease in the number of mechanoreceptors stimulated (Oostendorp et al., 1993). Several clinical authors suggest that once normal joint accessory movement is restored, symptom-free voluntary physiological movement is then encouraged, facilitating pain-free normal function (Maitland, 1986; Kaltenborn and Evjenth, 1989; Mulligan, 1994). These theories are supported by the present study that has shown parallel improvements in range of motion, pain and dizziness following manual therapy directed to the cervical spine. Because participants had a decrease in their dizziness after manual therapy treatment to the upper cervical spine, it can be inferred that dysfunction in the upper cervical spine was the likely source of the problem. This provides evidence in support of the existence of cervicogenic dizziness, a topic still of some controversy to date (Karlberg et al., 1996a; Oostendorp et al., 1999; Brandt and Bronstein, 2001; Tjell, 2001). One limitation of present the study is that the manipulative physiotherapist providing both interventions was aware that the laser procedure was a sham treatment. Even though the therapist attempted to treat all participants equally, she may have unknowingly been

365

biased towards the SNAG treatment. Also, larger sample sizes may be required when measures such as posturography and cervical range of motion are used as the magnitude of the changes observed are small but may be functionally valuable.

5. Conclusion The present study found that SNAGs are a safe and effective manual therapy technique for the treatment of cervicogenic dizziness and pain. SNAGs were shown to have a clinically and statistically signiﬁcant immediate and sustained effect in reducing dizziness, neck pain and disability caused by cervical spine dysfunction. The results of this study provide a new evidence-based treatment option for patients suffering this disabling condition. References Altman DG. Practical statistics for medical research. London: Chapman & Hall; 1996. Boismier T, Disher M. Spontaneous vertigo and headache: endolymphatic hydrops or migraine? Ear, Nose and Throat Journal 2001;80(12):881–4. Bracher E, Almeida CIR, Almeida RR, Bracher CBB. A combined approach for the treatment of cervical vertigo. Journal of Manipulative and Physiological Therapies 2000;23(2):96–100. Brandt T. Cervical vertigo. In Vertigo: its multisensory syndromes. London: Springer; 1991. p. 277–81. Brandt T, Bronstein AM. Cervical vertigo. Journal of Neurology, Neurosurgery and Psychiatry 2001;71(1):8–12. Cohen H. Efﬁcacy of treatments for posterior canal benign paroxysmal postitional vertigo. Laryngoscope 1999;109:584–90. Colledge NR, Barr-Hamilton RM, Lewis SJ, Sellar RJ, Wilson JA. Evaluation of investigations to diagnose the cause of dizziness in elderly people: a community based controlled study. British Medical Journal 1996;313(28 September):788–93. de Jong PTVM, de Jong JMBV, Bernard C, Jongkees LBW. Ataxia and nystagmus induced by injection of local anaesthetics in the neck. Annals of Neurology 1977;1:240–6. Dhimitri K, Brodeur S, Croteau M. Reliability of the cervical range of motion device in measuring upper cervical motion. The Journal of Manual and Manipulative Therapy 1998;6(1):31–6. Drachman D, Hart C. An approach to the dizzy patient. Neurology 1972;22:323–34. Enloe LJ, Shields RK. Evaluation of health-related quality of life in individuals with vestibular disease using disease-speciﬁc and general outcome measures. Physical Therapy 1997;77(9):890–903. Exelby L. Mobilisation with movement—a personal view. Physiotherapy 1995;81(12):724–9. Froehling D, Silverstein M, Mohr D, Beatty C. Does this dizzy patient have a serious form of vertigo? Journal of American Medical Association 1994;271(5):385–9. Furman J, Whitney S. Central causes of dizziness. Physical Therapy 2000;80(2):179–87. Galm R, Rittmeister M, Schmitt E. Vertigo in patients with cervical spine dysfunction. European Spine Journal 1998;7:55–8. Gill-Body KM, Krebs DE, Parker SW, Riley PO. Physical therapy management of peripheral vestibular dysfunction: two clinical case reports. Physical Therapy 1994;74(2):36–50.

ARTICLE IN PRESS 366

S.A. Reid et al. / Manual Therapy 13 (2008) 357–366

Gross A, Kay T, Hondras M. Manual therapy for mechanical neck disorders: a systematic review. Manual Therapy 2002;7(3):131–49. Heikkila H. Cervical vertigo. In: Jull GA, editor. Grieve’s modern manual therapy, the vertebral column. Edinburgh, UK: Elsevier Churchill Livingstone; 2004. Heikkila H, Johansson M, Wenngren BI. Effects of acupuncture, cervical manipulation and NSAID therapy on dizziness and impaired head repositioning of suspected cervical origin: a pilot study. Manual Therapy 2000;5(3):151–7. Hinoki M. Vertigo due to whiplash injury: a neurotological approach. Acta Otolaryngologica (Stockholm) Supplementum 1985;419:9–29. Hulse M. Disequelibrium caused by a functional disturbance of the upper cervical spine, clinical aspects and differential diagnosis. Manual Medicine 1983;1(1):18–23. Irnich D, Behrens N, Molzen H, Konig A, Glegitsch J, Krauss M, et al. Randomised trial of acupuncture compared with conventional massage and ‘‘sham’’ laser acupuncture for treatment of chronic neck pain. British Medical Journal 2001;322(7302):1574. Jacobsen GP, Newman CW. The development of the dizziness handicap inventory. Archives Otolaryngology Head Neck Surgery 1990;116:424–7. Kaltenborn FM, Evjenth O. Manual mobilization of the extremity joints. Oslo, Norway: Olaf Noris Bokhandel; 1989. Karlberg M, Johansson M, Magnussen M, Frannson P. Dizziness of suspected cervical origin distinguished by posturographic assessment of human postural dynamics. Journal of Vestibular Research 1996a;6(1):37–47. Karlberg M, Magnussen M, Malmstrom E-M, Melander A, Moritz U. Postural and symptomatic improvement after physiotherapy in patients with dizziness of suspected cervical origin. Archives Physical Medicine and Rehabilitation 1996b;77(September): 874–82. Koes B. Spinal manipulation and mobilisation for back and neck pain: a blinded review. British Medical Journal 1991;303: 1298–302. Koes B, Bouter L, van Mameran H, Essers AH. A blinded randomized clinical trial of manual therapy and physiotherapy for chronic back and neck complaints: physical outcome measures. Journal of Manipulative and Physiological Therapies 1992a;15(1):16–23. Koes B, Bouter L, van Mameran H, Essers AHM, Verstegen GMJR, Hofhuizen DM, et al. Randomised clinical trial of manipulative therapy and physiotherapy for persistent back and neck complaints: results of one year follow up. British Medical Journal 1992b;304. Kroenke K, Mangelsdorff A. Common symptoms in ambulatory care. American Journal of Medicine 1989;86:262–6. Kroenke K, Lucas C, Rosenburg M, Scherokman B, Herbers J, Wehrle P, et al. Causes of persistent dizziness. Annals of Internal Medicine 1992;117(11). Luxon LM, editor. Vertigo in old age. Vertigo. Chichester, UK: Wiley; 1984. Maher C, Latimer J, Refshauge K. Atlas of clinical tests and measures for low back pain. Sydney, Australia, 2000. Maitland G. Vertebral manipulation. London: Butterworth Heineman; 1986. Mulligan BR. Vertigoy Manual therapy may be needed. In: Manipulative Physiotherapists Association of Australia seventh Biennial Conference, Blue Mountains, Australia, 1991. Mulligan BR. SNAGS: mobilisations of the spine with active movement. In: Palastanga N, editor. Grieve’s modern manual therapy, the vertebral column. Edinburgh, UK: Churchill Livingstone; 1994. p. 733–43.

Mulligan BR. Manual therapy ‘‘NAGS,’’ ‘‘SNAGS,’’ ‘‘MWMS’’ etc. Wellington, New Zealand: Hutcheson Bowman & Stewart Ltd; 1999. Murphy D, McDonald A, Power C, Unwin A, Macsullivan R. Measurement of pain: a comparison of the visual analogue with nonvisual analogue scale. The Clinical Journal of Pain 1998;3: 197–9. Newman CW, Jacobsen GP. Application of self-report scales in balance function handicap assessment and management. Seminars in Hearing 1993;14(4):363–74. Oostendorp R, van Eupen AAJM, Elvers JWH, Bernards J. Effects of restrained cervical mobility on involuntary eye movements. The Journal of Manual and Manipulative Therapy 1993;1(4):148–53. Oostendorp RAB, van Eupen AAJM, Van Erp J, Elvers H. Dizziness following whiplash injury: a neuro-otological study in manual therapy practice and therapeutic implication. The Journal of Manual and Manipulative Therapy 1999;7(3):123–30. O’Reilly R, Elford B, Slater R. Effectiveness of the particle repositioning manoeuvre in subtypes of BPPV. Laryngoscope 2000;110(August):1385–8. Paungmali A, Vincenzio B, Smith M. Hypoalgesia induced by elbow manipulation in lateral epicondylalgia does not exhibit tolerance. Journal of Pain 2003;4(8):448–54. Reid S, Rivett DA. Manual therapy treatment of cervicogenic dizziness: a systematic review. Manual Therapy 2005;10:4–13. Scott J, Huskisson EC. Graphic representation of pain. Pain 1976;2:175–84. Scudds RA. Pain outcome measures. Journal of Hand Therapy 2001;14:86–90. Shumway-Cook A, Horak F. Vestibular rehabilitation: an exercise approach to managing symptoms of vestibular dysfunction. Seminars in Hearing 1989;10(2):196–207. Storper I, Spitzer J, Scanlan M. Use of glycopyrrolate in the treatment of Menieres’s disease. Laryngoscope 1998;108(October):1442–5. Terrahe K. Vertigo and disturbances of equilibrium in upper cervical syndrome. Therapiewoche 1979;29(9):1392–6. Tjell C. Cervicogenic vertigo: with special emphasis on whiplashassociated disorder. In: Vernon H, editor. The Cranio-cervical syndrome. Toronto, Canada: Butterworth-Heinemann; 2001. Tousignant M, de Bellefeuille L, O’Donoughue S. Criterion validity of the cervical range of motion (CROM) goniometer for cervical ﬂexion and extension. Spine 2000;25(3):324–30. Treleaven J. Dizziness Handicap Inventory (DHI). Australian Journal of Physiotherapy 2006;52(1):67. Turk DC, Melzack R. Handbook of pain assessment. New York: The Guilford Press; 1992. Whitney S, Wrisley D, Brown K, Furman J. Is perception of handicap related to functional performance in persons with vestibular dysfunction? Otology and Neurotology 2004;25(2):139–43. Wilson E. Headaches and vertigo: a simple complex. In Touch 1996;78:5–9. Wrisley D, Sparto P, Whitney S, Furman J. Cervicogenic dizziness: a review of diagnosis and treatment. Journal of Orthopaedic and Sports Physical Therapy 2000;30(12):755–66. Wyke B. Cervical articular contributions to posture and gait: their relation to senile disequilibrium. Age and Ageing 1979;8:251–8. Yardley L, Todd AM, Lacoudraye-Harter MM, Ingham R. Psychosocial consequences of recurrent vertigo. Psychology and Health 1992;6(85–96). Zhou W, Jiang W, Li X, Zhang Y, Wu Z. Clinical study on manipulative treatment of derangement of the atlantoaxial joint. Journal of Traditional Chinese Medicine 1999;19(4):273–8.

ARTICLE IN PRESS

Manual Therapy 13 (2008) 367–371 www.elsevier.com/locate/math

Case report

Positive patient outcome after manual cervical spine management despite a positive vertebral artery test Eric G. Johnsona,, Rob Landelb, Randall S. Kusunosec, Tetiana D. Appela a

Department of Physical Therapy, Loma Linda University, School of Allied Health Professions, Loma Linda CA, 92350, USA Department of Biokinesiology and Physical Therapy at the School of Dentistry, University of Southern California, Los Angeles, CA, USA c The Jones Institute, Carlsbad, CA, USA

b

Received 7 June 2007; received in revised form 21 November 2007; accepted 6 December 2007

1. Introduction Dizziness is one of the most common symptoms reported to physicians and approximately 40% of adults experience clinically signiﬁcant dizziness during their lifetime (Sloane, 1989; Koziol-McLain et al., 1991; Dallara et al., 1994; Furman and Cass, 1996; Schubert, 2007). The pathology responsible for the dizziness impairment is often difﬁcult to determine (Furman and Cass, 1996). Vertebrobasilar insufﬁciency (VBI) has been described in the literature as a pathology that frequently produces dizziness (Williams and Wilson, 1962; Grad and Baloh, 1989; Clendaniel and Landel, 2007; Schubert, 2007). VBI is deﬁned as an occlusion of blood ﬂow during cervical rotation or extension in the ‘‘area of junction for the vertebral and basilar arteries’’ (O’Sullivan and Schmitz, 2007). Physical therapists use the vertebral artery test (VAT) to assess patient tolerance to cervical extension and rotation as well as for screening for VBI. The VAT involves placing the patient into cervical extension, rotation, or combined extension and rotation (cervical quadrant), sustaining the position for 10 s or more, and assessing for signs or symptoms (Magee, 2002; Clendaniel and Landel, 2007). Signs and symptoms associated with VBI include dizziness, vertigo, syncope, headaches, imbalance, nausea, vomiting, visual disturbances, dysarthria, sensory changes, disorientation, and extremity weakness (Grad Corresponding author. Tel.: +1 909 558 4632x47471; fax: +1 909 558 4291. E-mail address: [email protected] (E.G. Johnson).

1356-689X/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2007.12.001

and Baloh 1989; Baloh and Honrubia, 1990). The VAT has become standard clinical practice even though there is considerable controversy as to whether or not clinical tests purported to screen for the presence of VBI identify those at risk for vertebral artery dissection (Cote et al., 1996; Licht et al., 2000; Haldeman et al., 2002; Childs et al., 2005). Although cases of true VBI are exceedingly rare, ignoring the suspicion that a patient has VBI risks causing a catastrophic injury (Haldeman et al., 1999; Furman and Whitney, 2000). There are several diagnostic tests for vertebral artery occlusion including magnetic resonance angiography (MRA) and computed tomography scan. However, the duplex doppler ultrasound appears to be the most widely used due to its accessibility and cost-effectiveness for patients and health providers (Licht et al., 1998, 2000; Rivett and Milburn, 1999; Barker et al., 2000; Haynes, 2002; Mitchell, 2003). Research suggests that coupled cervical spine rotation and extension decreases the blood ﬂow of the contralateral vertebral artery (Licht et al., 1998, 1999; Rivett and Milburn, 1999; Yi-Kai et al., 1999; Mitchell, 2003; Arnold et al., 2004). Licht et al. (1999), in their study on vertebral artery volume ﬂow, stated ‘‘During rotation, the ipsilateral atlantoaxial joint is ﬁxed, but the contralateral joint slides forward and down. This causes the vertebral artery to stretch, kink, and narrow because it is ﬁxed in the surrounding transverse foramina, paravertebral muscles, and ﬁbrous ligaments.’’ The authors further state that they have observed such narrowing in diagnostic tests of vertebral artery blood ﬂow. This ﬁnding, however, is not universally accepted

ARTICLE IN PRESS 368

E.G. Johnson et al. / Manual Therapy 13 (2008) 367–371

in the literature and there is evidence that suggests vertebral artery blood ﬂow is not decreased as the result of cervical spine position (Cote et al., 1996). Muscle imbalances leading to painful trigger points could potentially produce symptoms similar to those of VBI. Muscle balance is ‘‘a state of equilibrium that exists when there is a balance of strength of opposing muscles acting on a joint, providing ideal alignment for movement and optimal stabilization’’ (Kendall and McCreary, 1993). Simons et al. (1999) reported that trigger points located in the sternocleidomastoid muscle (SCM), and/or upper trapezius can elicit symptoms of dizziness, nausea, spatial disorientation and syncope, particularly if the trigger points are severe and placed on sudden stretch, such as head rotation. 2. Case description 2.1. Examination The patient was a 24-year-old female with a one-year history of dizziness provoked by left cervical rotation. The dizziness was described as a feeling of anxiety and difﬁculty communicating. She reported no mechanism of injury and was otherwise healthy. Because the primary complaint was dizziness with cervical rotation, the examiner performed the VAT as described by Schubert (2007). This version of the VAT combines cervical spine rotation and extension with the patient in the seated position to minimize disruption of the semicircular canals (Fig. 1). This was an important consideration in the event her dizziness was being caused by benign paroxysmal positional vertigo (BPPV). The Hallpike–Dix test is used to assess for the presence of BPPV and because of it’s similarities to the supine version of the VAT, the clinician would not be able to determine whether symptoms of dizziness were due to VBI or BPPV (Appendix 1). For a detailed description of BPPV and the Hallpike–Dix test the reader is referred to Herdman (2007). The VAT was negative when performed to the patient’s right but when performed to patient’s left, it provoked the patient’s dizziness and slowed verbal responses to questions were observed. The VAT was therefore judged to be positive. The patient was referred to her physician for further evaluation, and subsequently underwent a duplex doppler ultrasound. The ﬁndings from this test indicated that there was no evidence of any signiﬁcant stenosis in the bilateral common carotid, internal carotid, external carotid, or vertebral arteries. Given the negative radiology report, further investigation for a mechanical cause of the symptoms was deemed appropriate and of minimal risk. A palpatory examination of the muscles in the upper quarter region identiﬁed several tender points bilaterally including the upper trapezius, SCM, levator scapulae, and anterior scalene muscles.

Fig. 1. Seated VAT position 1 (A), position 2 (B), position 3 (C). (A model was used for photographs).

ARTICLE IN PRESS E.G. Johnson et al. / Manual Therapy 13 (2008) 367–371

369

and (3) returning the patient slowly back to neutral avoiding any sudden return (Jones et al., 1995). SCS is deﬁned as the ‘‘relief of rheumatic pain by placing a joint in its position of greatest comfort’’ (Jones et al., 1995; Lewis and Flynn, 2001). The dysfunction of proprioceptive reﬂexes is thought to result in false messages of strain causing a protective muscle spasm. Relief of these false messages of strain can potentially be achieved by applying strain in the opposite direction. This is accomplished by shortening the muscle that contains the false message of strain to the point that it stops reporting strain, for a period of ninety seconds (Kusunose, 1993; Jones et al., 1995). When a muscle is shortened, it places the muscle spindle on slack and decreases the afferent discharge of information to the central nervous system, thereby relieving the muscle spasm and improving joint range (Kusunose, 1993; Jones et al., 1995). For a detailed description of SCS, the reader is referred to Jones et al. (1995). There was no other intervention administered. The patient was tested with VAT again after several weeks and again one-year later and the VAT remained negative.

3. Discussion

Fig. 1. (Continued)

2.2. Interventions and outcomes Strain-counterstrain (SCS) techniques were performed to the bilateral upper trapezius, levator scapulae, anterior scalenes, and SCM muscles based on the palpatory examination. Three principles common to all SCS techniques include the following: (1) passively moving the affected joint into its position of greatest comfort, (2) maintaining the position for ninety seconds,

Despite the fact that dizziness is a common complaint of patients, differential diagnosis of dizziness can be extremely challenging for both the physician and the physical therapist (Furman and Cass, 1996). The purpose of this case study was to describe the diagnosis and management of a patient who had a positive VAT but a negative doppler ultrasound examination. Complaints of dizziness can be the result of more than one pathology, requiring the clinician to generate a list of hypothetical causes. Because of a history of dizziness provoked by sustained cervical rotation, VBI, cervical muscular disorder and BPPV were included as potential contributory pathologies. The examination strategy included performing the VAT ﬁrst because of the potential severity of complications due to compromised cerebral blood ﬂow and dissecting aneurysms. Based on a positive VAT, physical therapy was deemed to be inappropriate at that time, and the patient was referred to a radiologist for a diagnostic vascular evaluation. Despite a positive VAT, having a negative duplex doppler ultrasound is not a surprising ﬁnding. The ﬁndings in our patient case are similar to those reported by Licht et al. (2000) who found that none of the 15 patients they studied, referred for evaluation of vertebral artery ﬂow velocity because of symptoms produced during premanipulative testing, had signiﬁcant decrease in blood ﬂow in any of the head positions tested. There is evidence that the VAT has a sensitivity of 0% and a positive predictive value of 0% for detecting decreased vertebral artery blood ﬂow (Cote et al., 1996). Others

ARTICLE IN PRESS 370

E.G. Johnson et al. / Manual Therapy 13 (2008) 367–371

have reported similar concerns about the usefulness of the VAT as a screening tool (Haldeman et al., 2002; Childs et al., 2005). On the contrary, Asavaopon et al versus Asavasopon and Jankoski (2005) identiﬁed a positive VBI using the VAT and their test result was conﬁrmed via duplex doppler ultrasonography and MRA. Arnold et al. (2004) also reported that the doppler ultrasound was able to detect blood ﬂow reductions during the VAT in a prospective study of 22 men and women. Since the patient denied any spinning sensation in her history, BPPV was deemed unlikely. With VBI ruled out based on duplex doppler ultrasound, the physical therapist next investigated the possibility of a muscular disorder in the cervical region as the primary pathology. The palpatory examination of the upper quarter musculature revealed several tender points in the bilateral cervical spine region, which supported the hypothesis. The goal was for the patient to be symptom-free while turning her head to the left. The next step was to formulate a plan for determining testing criteria and for assessing the status of the problems and goals. The testing criterion used to examine the validity of the hypothesis was re-evaluating the results of the VAT. The therapist was then able to plan an intervention strategy based on the hypothesis and indicate how the intervention would be implemented. The treatment tactics implemented were SCS techniques to the bilateral upper trapezius, levator scapulae, anterior scalenes, and SCM muscles. Because her initial complaint was dizziness with active range of motion (AROM) of left cervical spine rotation, post-intervention re-assessment included AROM of left cervical spine rotation. The result of this AROM test was negative. Subsequently, the VAT was performed and it was negative as well. Licht et al. (2000) reported on patients who have had a positive VAT but normal vertebral artery sonography, and who were subsequently treated with spinal manipulation with resolution of their symptoms. To our knowledge there are no reports of improvement of dizziness in the extension-rotation position after treatment of the cervical soft tissues. The scientiﬁc theory underlying the SCS technique is that dysfunction of proprioceptive reﬂexes results in a false message of strain causing a protective muscle spasm (Jones et al., 1995). Restoration of normal muscle balance by resolving the reﬂexive dysfunction could theoretically reduce the mechanical compression on the arterial structures during the clinical VBI test. A case study by Lewis and Flynn (2001) reported that two to three sessions of SCS resolved ‘‘aberrant neuromuscular activity’’ in four patients with lower back pain. No evidence exists to support using these techniques in the cervical spine other than clinical anecdotes. The decision to use these techniques in this case was based on the therapist’s level of expertise and clinical experience.

Another possible etiology leading to symptoms similar to those of VBI is mechanical deformation of baroreceptors in the carotid artery. Baroreceptors detect mechanical deformations that occur in vascular walls and regulate them by opening or closing ion channels. These channels send signals to the control mechanisms that determine the appropriate changes needed for the change in pressure (Guyton, 1971). Wijetunga and Schatz (2005) reported that in patients with carotid sinus hypersensitivity (CSH), ‘‘mechanical deformation of the carotid sinus (located at the bifurcation of the common carotid artery) leads to an exaggerated response with bradycardia or vasodilatation, resulting in dizziness or syncope.’’ In the human body, the carotid artery runs parallel to the jugular vein. These vessels are fairly superﬁcial with the omohyoid and SCM, respectively, crossing just above the bifurcation of the common internal/external carotid arteries (Netter and Hansen, 2003). If the VBI test is performed on a patient with a muscle imbalance in the cervical region and CSH, it is plausible that the testing position could exaggerate the physical compression on the arterial structures, stimulating the baroreﬂex, thereby provoking symptoms including dizziness. This case demonstrates that signs and symptoms found upon positioning the patient in cervical extension and rotation, which would classically be attributed to VBI, were likely due to non-vascular causes. The speciﬁc etiology of the signs and symptoms elicited by the VAT remains unclear and warrants further investigation in controlled studies. Based on the response to manual intervention directed at the cervical soft tissues, it seems clear that non-vascular causes must be considered. The cause and effect relationship between the cervical soft tissues and the symptoms needs further study. Financial disclosure and conﬂict of interest All the authors conﬁrm that here are no conﬂicts of interest or ﬁnancial beneﬁts associated with the writing and publication of this paper. Acknowledgement Work attributed to: Loma Linda University, Department of Physical Therapy. Appendix 1. A brief description of the Hallpike–Dix test The Hallpike–Dix test is used to identify BPPV. BPPV is a condition where patients experience episodes of dizziness, or vertigo, particularly after position changes affecting the head and neck. The Hallpike– Dix test is performed by having the patient long-sit on a plinth with their head rotated approximately 30-45 degrees (A). The examiner stands behind the patient

ARTICLE IN PRESS E.G. Johnson et al. / Manual Therapy 13 (2008) 367–371

with one hand supporting their head/neck and other supporting their trunk (B). The patient is then assisted into a supine position with the patient’s head slightly below the horizontal plane and the position is maintained for 30–60 s (C). The Hallpike–Dix test is performed on both sides and provocation of dizziness and nystagmus (involuntary eye movement) is considered a positive test. (A model was used for photographs).

References Arnold C, Bourassa R, Langer T, Stoneham G. Doppler studies evaluating the effect of a physical therapy screening protocol on vertebral artery blood ﬂow. Manual Therapy 2004;9(1):13–21. Asavasopon S, Jankoski J. Godges JJ. Clinical diagnosis of vertebrobasilar insufﬁciency: residents case problem. Journal of Sports Physical Therapy 2005;35(10):645–50. Baloh RW, Honrubia V. Clinical Neurophysiology of the Vestibular System. 2nd ed. F.A. Davis Company; 1990. Barker S, Kesson M, Ashmore J, Turner G, Conway J, Stevens D. Guidance for pre-manipulative testing of the cervical spine. Manual Therapy 2000;5(1):37–40. Childs JD, Flynn TW, Fritz JM, Piva SR, Whitman JM, Wainner RS, et al. Screening for Vertebrobasilar Insufﬁciency in Patients with Neck Pain: Manual Therapy Decision-Making in the Presence of Uncertainty. Journal of Orthopedic Sports Physical Therapy 2005;35(5):300–6. Clendaniel RA, Landel R. Non-vestibular Diagnosis and Imbalance: Cervicogenic Dizziness. In: Herdman SJ, editor. Vestibular Rehabilitation. 3rd ed. Philadelphia, PA: F.A. Davis Company; 2007. p. 475–6 [chapter 29]. Cote PB, Kreitz G, Cassidy JD, Thiel H. The validity of the extensionrotation test as a clinical screening procedure before neck manipulation: a secondary analysis. Journal of Manipulative and Physiological Therapeutics 1996;19(3):159–64. Dallara J, Lee C, McIntosh L, Sloane PD, Morris D. Emergency department length-of-stay and illness severity in dizzy and chestpain patients. American Journal of Emergency Medicine 1994;12(4):421–4. Furman JM, Cass SP. Balance Disorders: A Case-Study Approach. Philadelphia, PA: F.A. Davis Company; 1996 preface, p. iii. Furman JM, Whitney SL. Central causes of dizziness. Physical Therapy 2000;80(2):179–87. Grad A, Baloh RW. Vertigo of vascular origin: Clinical and ENG features in 84 cases. Archives of Neurology 1989;46(3):281–4. Guyton AC. Textbook of Medical Physiology, 4th ed. W.B. Saunders Company; 1971. Haldeman S, Kohlbeck FJ, McGregor M. Stroke, cerebral artery dissection, and cervical spine manipulation therapy. Journal of Neurology 2002;249(8):1098–104. Haldeman S, Kohlbeck FJ, McGregor M. Risk factors and precipitating neck movements causing vertebrobasilar artery dissection after cervical trauma and spinal manipulation. Spine 1999;24(8):785–94.

371

Haynes MJ. Vertebral arteries and cervical movement: Doppler ultrasound velocimetry for screening before manipulation. Journal of Manipulative and Physiological Therapeutics 2002;25(9):556–67. Herdman SJ. Vestibular Rehabilitation. 3rd ed. Philadelphia PA: F.A. Davis; 2007. Jones LH, Kusunose R, Goering E. Jones Strain-Counterstrain. Boise, ID: Jones Strain-Counterstrain; 1995. Kendall FP, McCreary EK. Provance PG: Muscles Testing and Function, 4th ed. Baltimore: Williams & Wilkins; 1993 [chapter 1, p. 3]. Koziol-McLain J, Lowenstein SR, Fuller B. Orthostatic vital signs in emergency department patients. Annals of Emergency Medicine 1991;20(6):606–10. Kusunose R. Strain and Counterstrain. In: Basmajian JV, Nyberg R, editors. Rational Manual Therapies. Baltimore: Williams and Wilkins; 1993. p. 323–33. Lewis C, Flynn TW. The use of strain-counterstrain in the treatment of patients with low back pain. The Journal of Manual and Manipulative Therapy 2001;9(2):92–8. Licht PB, Christensen HW, Hoilund-Carlsen PF. Vertebral artery volume ﬂow in human beings. Journal of Manipulative and Physiological Therapeutics 1999;22(6):363–7. Licht PB, Christensen HW, Hoilund-Carlsen PF. Is there a role for premanipulative testing before cervical manipulation? Journal of Manipulative and Physiological Therapeutics 2000;23(3):175–9. Licht PB, Christensen HW, Hojgaard P, Hoilund-Carlsen PF. Triplex ultrasound of vertebral artery ﬂow during cervical rotation. Journal of Manipulative and Physiological Therapeutics 1998;21(1):27–31. Magee DJ. Orthopedic Physical Assessment, 4th ed. USA: Elsevier Sciences; 2002 [chapter 3, p. 152–156]. Mitchell JA. Changes in vertebral artery blood ﬂow following normal rotation of the cervical spine. Journal of Manipulative and Physiological Therapeutics 2003;26(6):347–51. Netter FH, Hansen JT. Atlas of Human Anatomy. 3rd ed. R.R. Donnelly; 2003. O’Sullivan SB, Schmitz TJ. Physical Rehabilitation: Assessment and Treatment, 5th ed. Philadelphia: F.A. Davis Company; 2007 glossary, p. 1348. Rivett DA, Milburn PD. Effect of premanipulative tests on vertebral artery and internal carotid artery blood ﬂow: a pilot study. Journal of Manipulative and Physiological Therapeutics 1999;22(6):368–76. Schubert MC. Vestibular disorders. In: O’Sullivan SB, Schmitz TJ, editors. Physical rehabilitation: assessment and Treatment. 5th ed. Philadelphia: F.A. Davis Company; 2007 chapter 24, p. 1025. Sloane PD. Dizziness in primary care: results from the National Ambulatory Medical Care Survey. Journal of Family Practice 1989;29(1):33–8. Simons DG, Travell JG, Simons LS. 2nd ed. Myofascial pain and dysfunction: the trigger point manual volume 1. Upper half of body. Williams and Wilkins; 1999. Wijetunga MN, Schatz IJ. /http://www.emedicine.com/med/ topic299.htmS; 2005. [accessed on 11/6/06]. Williams D, Wilson TG. The diagnosis of the major and minor syndromes of basilar insufﬁciency. Brain 1962;85:741–4. Yi-Kai L, Yun-Kun Z, Shi-Zhen Z. Changes and implications of blood ﬂow velocity of the vertebral artery during the rotation and extension of the head. Journal of Manipulative and Physiological Therapeutics 1999;22:91–5.

Manual Therapy (2008) 13(4), 372

Available online at www.sciencedirect.com

Diary of events

5th Chiropractic Osteopathy Physiotherapy Annual Undergraduate/Pre-Registration Research Conference ‘‘Moving Forward Through Research and Practice’’ 25th October 2008, Bournemouth, UK. Keymote Speaker: Prof. Gordon Waddell, Rehabilitation What Works, For Who and When? Details on how to submit an abstract and register available at www.aecc.ac.uk

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

2nd World Congress on Manual Therapy and Sport Rehabilitation, The Spine II, in Roma Italy 6th–8th of February 2009 www.newmaster.it

If you wish to advertise a course/conference, please contact: Karen Beeton, Associate Head of School (Professional Development), School of Health and Emergency Professions, University of Hertfordshire, College Lane, Hatﬁeld, Herts AL10 9AB, UK. There is no charge for this service.

372