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

VOLUME 12 NUMBER 3 PAGES 199– 296 August 2007 Editors International Advisory Board Ann Moore PhD, GradDipPhys, FCSP, ...

Author: Editors: Ann Moore and Gwen Jull

54 downloads 781 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 12 NUMBER 3 PAGES 199– 296 August 2007

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) 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) 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) 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 Wu¡ (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.intl.elsevierhealth.com/journals/math doi:10.1016/S1356-689X(07)00094-X

ARTICLE IN PRESS

Manual Therapy 12 (2007) 199 www.elsevier.com/locate/math

Editorial

Physiotherapists: Closer together, not worlds apart Since the last issue of Manual Therapy Journal, the Fifteenth World Congress of Physical Therapy (WCPT) has taken place in Vancouver, British Columbia, Canada. The conference saw the meeting of up to 4000 physical therapists’ minds! One thousand six hundred and forty-six presentations were made at Congress by representatives from the five regions of WCPT. The largest proportion of presentations came from Europe with the Americas and the Antipodes and South East Asia also forming a large proportion of the physiotherapy contingent. Musculoskeletal physiotherapy research and development presentations formed the largest proportion of the specialities represented. The 4-yearly WCPT events provide a rich and fertile environment for the exchange of views, ideas, knowledge and research findings between physiotherapists from across the world. During Congress, an important event took place—the inaugural meeting of the International Society of Physiotherapy Journal Editors (ISPJE). Forty physiotherapy journals were represented at this event and we are delighted to say that the editors of Manual Therapy Journal were part of the development of this international society and were represented at the event. The ISPJE aims to promote the sharing of knowledge, expertise and good practice among international physiotherapy editors and raise the standards within and the academic profile of physiotherapy publication across the world. It is also hoped that by working together, we can facilitate more access to publications for those people living in developing countries. It was a privilege to be part of this innovation and we look forward to the opportunities that such collaborations can provide for Manual Therapy Journal, but also for the physiotherapy profession as a whole. WCPT once again demonstrated the goodwill and bonhomie that occurs when physiotherapists from around the world get together. Research collaborations flourished, new developments in technology and practice were shared, new research findings were taken to all corners of the earth for further dissemination by the message bearers. The conference makeup this time

1356-689X/$ - see front matter r 2007 Published by Elsevier Ltd. doi:10.1016/j.math.2007.06.001

provided a number of learning opportunities, not just from the multi-speciality content perspective, but the method of delivery, e.g. platform presentations, viper discussions, poster presentations, workshops, symposia, discussion panels, debates, pre- and post-congress courses, clinical visits and last-but-not-least, and probably most importantly, the social events and informal networking opportunities. It was indeed a magical learning environment for everyone. Importantly, WCPT highlights the positive benefits of physiotherapists from all specialities learning together and the fact that as a profession we possess many skills which are transferable across speciality boundaries and many issues and developments occurring in some areas of practice have strong applications and relevance in other areas. We can all learn from experiences that have occurred in other specialities and we should not forget to do so. Manual Therapy Journal will be publishing a Special Issue, including selected key musculoskeletal papers from Congress before the end of 2007. We are pleased to announce that Associate Professor Darren Rivett of the Editorial Board for Manual Therapy will be guest editor for this Special Issue. While we celebrate the benefits and successes of WCPT, we must now prepare for the next international event and the most important event in the musculoskeletal calendar—IFOMT 2008 which will be held in June 2008 in Rotterdam. This conference presents us with a superb opportunity to share, learn, collaborate, debate and discuss issues of importance to musculoskeletal physiotherapists worldwide. We should all look forward to hearing about the latest cutting-edge research and taking part in the philosophical debate that will inevitably occur at this most exciting event. We look forward to seeing you all in Rotterdam.

(Co-Editors) Ann Moore, Gwen Jull

ARTICLE IN PRESS

Manual Therapy 12 (2007) 200–208 www.elsevier.com/locate/math

Review

Iliotibial band friction syndrome—A systematic review Richard Ellis, Wayne Hing, Duncan Reid Health Rehabilitation Research Centre, Division of Rehabilitation and Occupation Studies, Faculty of Health & Environmental Sciences, AUT University, Private Bag 92006, Auckland, New Zealand Received 19 October 2004; received in revised form 19 July 2006; accepted 30 August 2006

Abstract Iliotibial band friction syndrome (ITBFS) is a common injury of the lateral aspect of the knee particularly in runners, cyclists and endurance sports. A number of authors suggest that ITBFS responds well to conservative treatment, however, much of this opinion appears anecdotal and not supported by evidence within the literature. The purpose of this paper is to provide a systematic review of the literature pertaining to the conservative treatment of ITBFS. A search to identify clinical papers referring to the iliotibial band (ITB) and ITBFS was conducted in a number of electronic databases using the keyword: iliotibial. The titles and abstracts of these papers were reviewed to identify papers specifically detailing conservative treatments of ITBFS. The PEDro Scale, a systematic tool used to critique randomized controlled trials (RCTs), was employed to investigate both the therapeutic effect of conservative treatment of ITBFS and also to critique the methodological quality of available RCTs examining the conservative treatment of ITBFS. With respect to the management of ITBFS, four RCTs were identified. The interventions examined included the use of nonsteroidal anti-inflammatory drugs (NSAIDs), deep friction massage, phonophoresis versus immobilization and corticosteroid injection. This review highlights both the paucity in quantity and quality of research regarding the conservative treatment of ITBFS. There seems limited evidence to suggest that the conservative treatments that have been studied offer any significant benefit in the management of ITBFS. Future research will need to re-examine those conservative therapies, which have already been examined, along with others, and will need to be of sufficient quality to enable accurate clinical judgements to be made regarding their use. r 2006 Elsevier Ltd. All rights reserved. Keywords: Iliotibial band; Iliotibial band friction syndrome; Systematic review; Conservative treatment

1. Introduction Iliotibial band friction syndrome (ITBFS) was first specifically described by Renne (1975) as a pain felt on the lateral aspect of the knee with lower limb activities such as running and cycling. Following an increase in the popularity of running and other endurance multidisciplinary sports, since the 1980s, ITBFS has become more common (Anderson, 1991; Kirk et al., 2000). The overall incidence of ITBFS can range from between 1.6% and 52% depending on which population you examine Corresponding author. Tel.: +64 9 921 9999x7800; fax: +64 9 921 9620. E-mail address: [email protected] (W. Hing).

1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.08.004

(Kirk et al., 2000; Brosseau et al., 2004). However, it is generally accepted that ITBFS is the most common running injury of the lateral knee, and has an incidence of between 1.6% and 12% (Orava, 1978; McNicol et al., 1981; Messier et al., 1995; Fredericson et al., 2000; Taunton et al., 2002). Within cycling, ITBFS is believed to account for 15–24% of overuse injuries (Farrell et al., 2003; Holmes and Pruitt, 1993). Its incidence in military recruits may range from 1% to 5.3% (Jordaan and Schwellnus, 1994; Almeida et al., 1999). ITBFS is uncommon in the inactive population (Orava, 1978). The aetiology of ITBFS is multi-factorial with representation of both intrinsic and extrinsic factor (McNicol et al., 1981; Kirk et al., 2000; Taunton et al., 2002). ITBFS in a non-traumatic overuse injury caused

ARTICLE IN PRESS R. Ellis et al. / Manual Therapy 12 (2007) 200–208

by friction/rubbing of the distal portion of the iliotibial band (ITB) over the lateral femoral epicondyle (LFE) with repeated flexion and extension of the knee. Orchard et al. (1996) describe an ‘impingement zone’ which occurs at approximately 301 of knee flexion during foot-strike and early stance phase. At approximately 301 and greater, of knee flexion, the ITB passes over and posterior to the LFE (Renne, 1975; McNicol et al., 1981; Anderson, 1991; Barber and Sutker, 1992; Aronen et al., 1993; Puniello, 1993; Messier et al., 1995; Novacheck, 1998; Fredericson et al., 2002; Farrell et al., 2003). During the impingement period, eccentric contraction of the tensor fascia lata (TFL) and gluteus maximus, to decelerate the leg whilst running, exert great tension through the ITB (Orchard et al., 1996; Kirk et al., 2000; Farrell et al., 2003). Farrell et al. (2003) described a similar impingement zone for cycling. The pathogenesis of ITBFS involves inflammation and irritation of the lateral synovial recess (Renne, 1975; Orava, 1978; McNicol et al., 1981; Ekman et al., 1994; Nemeth and Sanders, 1996; Nishimura et al., 1997; Kirk et al., 2000; Levin, 2003), as well as continued irritation of the posterior fibres of the ITB (Ekman et al., 1994; Fredericson et al., 2000; Kirk et al., 2000; Austermuehle, 2001; Fredericson et al., 2002) and inflammation of the periosteum of the LFE (McNicol et al., 1981; Noble et al., 1982; Nishimura et al., 1997; Fredericson et al., 2002), all of which describes the pathogenesis of ITBFS. Kirk et al. (2000) suggest that with repetitive soft tissue irritation there is simply not enough time for the body to repair these damaged tissues. This may lead to further irritation and injury which, in theory, would extend the area of the impingement zone and increase the risk of irritation (Levin, 2003). A number of authors have commented that ITBFS responds well to conservative treatment (Anderson, 1991; Kirk et al., 2000; Levin, 2003) with success rates reported as high as 94% (McNicol et al., 1981). A number of different treatment options are reported in the literature, however, it should be questioned whether these treatments are delivered based on sound evidence. The purpose of this paper is to perform a systematic review, evaluating the efficacy of conservative treatment of ITBFS, in order to highlight key concepts to guide evidence-based practice in the management of ITBFS. Relevant functional anatomical and biomechanical contributions to the aetiology and pathomechanics of ITBFS will also be discussed and related back to the findings of the RCTs available.

2. Methodology 2.1. Literature search strategy A search to identify clinical papers, clinical reviews and clinical trials pertaining to the ITB and ITBFS, was

201

conducted in electronic databases, subscribed to by the Auckland University of Technology (AUT) library, which included MEDLINE via PubMed (from 1966 onwards), Cumulative Index to Nursing and Allied Health Literature (CINAHL) (from 1983 onwards), The Cochrane Controlled Trials Register in the Cochrane Library (latest edition), SPORT-Discus (from 1830 onwards), Allied and Complementary Medicine Database (AMED) (from 1985 onwards), Blackwell-Synergy, Master FILE (from 1975 onwards), Expanded Academic ASAP (from 1980 onwards), Index New Zealand (INNZ) (from 1987 onwards), Lippincott 100 Nursing and Health Science Collection, Physiotherapy Evidence Database (PEDro) (from 1953 onwards), ProQuest 5000 International, ProQuest Health and Medical Complete, Web of Science (from 1945 onwards), Wiley Interscience–Life and Medical Sciences Titles. This search was conducted in August– September 2004. The ITB and ITBFS were deemed to be relatively narrow fields to search, therefore only one Medical Subject Heading (MESH) was used as a keyword: iliotibial. There was no limitation regarding date or language leading to 1260 citations being identified of which many were repeated across databases. The titles and/or abstracts of these citations were reviewed to identify papers specifically detailing the aetiology and conservative treatment of ITBFS and the anatomy and biomechanics of the ITB. The bibliographies of each paper were also used for cross-referencing to identify other relevant papers. 2.2. Study selection Inclusion criteria: The following criteria were used in order to select relevant papers to be included within this review: Type of participant: Participants to be 18 years of age and older, of either gender and have a clinical diagnosis of ITBFS for greater than 14 days duration. Type of study design: Randomized controlled trials. Type of intervention: Conservative treatment of ITBFS, i.e. non-surgical. Outcome measurements: To include at least one of the following outcome measurements: pain rating (e.g. Visual Analogue Scale (VAS)), function-specific VAS (i.e. work or sport related pain), time from diagnosis until symptom free, return to work and/or sport status. Exclusion criteria: The following criteria were used to eliminate papers from this review: papers written in nonEnglish languages, non-RCTs, RCTs which utilized non-conservative treatment, i.e. surgical interventions.


202

2.3. Review of methodological quality

paper are included in Table 2. The PEDro Scale is an 11-item scale. The various items deal with differing aspects of RCT analysis including internal validity, external validity and statistics. In order to allow quantitative analysis of the overall methodological quality of each study, seven items which relate to internal validity were identified. These seven items include the following items numbers 2, 3, 5, 6, 7, 8, 9 (refer to Table 1). The positive scores of each of these seven items is added together to calculate an Internal Validity Score (IVS) (Reid and Rivett, 2005).

Three reviewers independently assessed each of the RCTs identified for their respective methodological quality. The PEDro Scale (see Table 1), developed by The Centre of Evidence-Based Physiotherapy (CEBP) was utilized to assess each paper. The PEDro Scale is an 11-item scale, which is a validated and versatile tool used to rate RCTs for the PEDro Database (Clark et al., 1999; Maher et al., 2003). An overall score of methodological quality, or quality score (QS), was determined for each paper by each of the three reviewers as a total of positive scores for each of the 11 items. A consensus method was used to discuss and resolve discrepancies between the markings of each paper between the reviewers. The agreed QS for each

2.4. Analysis of methodological quality Based on the IVS of each paper, it is possible to make a qualitative assessment about the methodological quality. In the instance whereby the RCTs reviewed are not clinically heterogeneous, it is appropriate to use a qualitative method of analysis as quantitative analysis is made difficult in that the RCTs may not be directly comparing like interventions (Reid and Rivett, 2005; van Tulder et al., 1997). The qualitative assessment used within this review is an adaptation of those used by several authors (Karjalainen et al., 2001; Reid and Rivett, 2005) modified specifically for IVS obtained in this review using the PEDro Scale: Level 1: Strong evidence—when provided by generally consistent findings in multiple RCTs of high quality (IVS ¼ 6–7). Level 2: Moderate evidence—when provided by generally consistent findings in one RCT of high quality (i.e. IVS ¼ 6–7) and one or more lower-quality RCTs (i.e. IVSp5); Level 3: Limited evidence—when provided by generally consistent findings in one RCT of moderate quality (i.e. IVS ¼ 4–5) and one or more low-quality RCTs (i.e. IVSp3). Level 4: Insufficient evidence—when provided by generally consistent findings of one or more RCTs of limited quality (i.e. IVSp3), no RCTs available or conflicting results.

Table 1 PEDro scale (modified from Maher et al., 2003) Criteria

Score

1. Eligibility criteria were specified 2. Subjects randomly allocated to groups 3. Allocation was concealed 4. Groups similar at baseline regarding the most important prognostic factors 5. Blinding of all subjects 6. Blinding of all therapists who administered therapy 7. Blinding of all assessors who measured at least one outcome 8. Measures of at least one key outcome were obtained from more than 85% of initially allocated subjects 9. All subjects for whom outcome measures were available received treatment or control as allocated, or if this was not the case, at least one outcome measure analysed using ‘intention to treat’ analysis 10. The results of between-group statistical comparisons are reported for at least one key outcome 11. The study provides both point measures and measures or variability for at least one key outcome

No No No No

(0) (0) (0) (0)

Yes Yes Yes Yes

(1) (1) (1) (1)

No (0) No (0)

Yes (1) Yes (1)

No (0)

Yes (1)

No (0)

Yes (1)

No (0)

Yes (1)

No (0)

Yes (1)

No (0)

Yes (1)

Total

x/11

Table 2 Randomized controlled trials of the conservative treatment of ITBFS in order of PEDro score Scores for PEDro criteria

Gunter and Schwellnus (2004) Corticosteroid injection Schwellnus et al. (1991) NSAID’s Schwellnus et al. (1992) Deep transverse friction massage Bischoff et al. (1995) Phonophoresis versus immobilization

1

2

3

4

5

6

7

8

9

10

11

1 1 1 1

1 1 1 1

1 0 1 1

1 1 0 1

1 1 0 0

0 1 0 0

0 0 1 0

1 1 0 0

0 0 1 1

1 1 1 1

1 1 1 1

Note: QS ¼ overall quality score; IVS ¼ internal validity score.

QS

Methodological quality

IVS

9 8 7 7

Moderate Moderate Moderate Limited

4 4 4 3

Table 3 Randomized controlled trials of the conservative treatment of ITBFS Author

n ¼ 18 Aged between 20 and

Outcome

Results

IVS

QS

9 subjects with ITBFS


Decrease (P ¼ 0.01) in running pain from day 7 to 14 in a validated treadmill running test and VAS, intervention group compared to controls

4

9

40 mg methylprednisilone, 10 ml 1% lignocaine hydrochloride, injection deep to the ITB at knee lateral joint line

20 mg 1% lignocaine hydrochloride injection deep to the ITB at knee lateral joint line

Baseline test: a previous validated treadmillrunning test—pain whilst running (VAS)—every minute Running speed at subjects previous best speed, maintained for 30 min (after 5 min warm-up) or until the pain on VAS reached 8 ‘‘Total Daily Pain’’: evening pain (VAS) over 14 day period ‘‘Total pain during running’’ (VAS): repeated on days 7 and 14 after injection (Mean taken)


As for Gunter and Schwellnus (2004)

4

8

Physiotherapy (ITB stretch, ultrasound, deep friction massage) only

For total daily pain (VAS)—mean pain measured over 14 days

Decrease (Po0.05) in overall pain experienced over trial period. Decrease (Po0.05) in running pain from day 0 to 3 in a validated treadmill running test for (b) intervention group. Increase (Po0.05) in running distance in all groups from days 3 to 7, and in (b) intervention group from days 0 to 7

50 age 28.975, experimental group mean age 29.076.5 Duration of symptoms—within 14 days

(a) 14 subjects with Over 18 yr incl ITBFS 50 mg diclofenac, Final incl 43 pts Age range not given— physiotherapy

group 1 mean 2275, group 2 mean age 2476, group 3 mean age 2272 Duration of symptoms—group 1 mean 6.877.1 weeks, group 2 mean 6.178.1 weeks, group 3 mean 7.4713.1 weeks

(b) 16 subjects with ITBFS 400 mg ibuprofen, 500 mg paracetamol, 20 mg codeine phosphate, physiotherapy (ITB stretch, ultrasound, deep friction massage)

ARTICLE IN PRESS

Control

Control group mean

Schwellnus et al. (1991) NSAID’s

Intervention

R. Ellis et al. / Manual Therapy 12 (2007) 200–208

Gunter and Schwellnus (2004) Corticosteroid injection

Patient demographics

203

Author Schwellnus et al. (1992) Deep transverse friction massage

204

Table 3 (continued ) Patient demographics

n ¼ 20 over 18 yr Group 1 mean 2576,

group 2 mean 2975, no range for ages given Duration of symptoms group 1 23717 weeks, group 2 74795 weeks

Intervention

Control

Outcome



Deep transverse friction massage to the distal ITB from days 3 to 14 daily stretching and twice-daily ice therapy from days 0 to 14. Ultrasound to the distal ITB from days 3 to 14

Daily stretching and 2—daily ice therapy from days 0 to 14. Ultrasound to the distal ITB days 3–14

Total running pain (VAS)—treadmill test performed on days 0, 3, 7, 14 (Mean taken) As for Gunter and Schwellnus (2004)

Results

IVS

QS

4

7

3

7

Decreases in daily pain (VAS) and running pain during a validated treadmill running test in both groups. However, no significant difference between groups

n ¼ 25, Navy students Group 1 mean 22 yr, group 2 mean 23 yr

Duration of symptoms—group 1 mean 17.5 days, group 2 mean 15.0 days

13 subjects with ITBFS Phonophoresis (ultrasound through 10% hydrocortisone cream) over 2 weeks Rest, ice, stretching and ibuprofen


‘‘Number of days from diagnosis to pain-free examination’’

Knee immobilization over 2 weeks

‘‘Number of days from diagnosis to symptomfree 1 mile run’’— symptom-free ¼ ‘‘without pain or stiffness’’ Subjects were examined daily through the study

Rest, ice, stretching and ibuprofen

Less days to become pain free whilst treadmill running for intervention group compared to control group. More subjects from intervention group recovered from ITBFS in 10 days compared to control group

ARTICLE IN PRESS

Bischoff et al. (1995) Phonophoresis versus immobilization


For total daily pain (VAS)—mean pain measured over 14 days. Total running pain (VAS)—treadmill test performed on days 0, 3, 7, 14 (Mean taken)


For this review it was decided amongst the reviewers that in using a seven-item IVS, taken from the initial PEDro score (QS), a study of high methodological quality was one with an IVS of between 6 and 7, moderate quality between 4 and 5, limited quality between 0 and 3. 2.5. Analysis When RCTs were clinically and therapeutically heterogeneous, there is no available method to directly assess the relative benefit (or lack thereof) of one intervention versus another. In this instance, previous systematic reviews have decided to not include quantitative analysis for this very reason (Reid and Rivett, 2005). Therefore, it was decided not to perform any quantitative analysis, as no direct comparison could be made to determine clinical or therapeutic benefit between the RCTs and interventions examined.

3. Results 3.1. Selection of studies Four RCT’s regarding conservative management of ITBFS meeting the inclusion criteria were identified following the electronic and cross-referencing searches. These studies are summarized in Table 3. 3.2. Methodological quality The methodological quality, statistically represented by the IVS, for each paper is detailed within Table 2. Three of the four RCTs reviewed were given an IVS of four. This suggests that the authors felt that these studies were of moderate methodological quality. One of the RCTs was given an IVS of three, suggesting the authors felt this study was of limited methodological quality. All of the four RCTs satisfied the item relating to random allocation of subjects (Item 2). Otherwise, there were no clear trend towards any of the other internal validity rated items (3, 5, 6, 7, 8, 9) either being universally satisfied or not. 3.3. Study characteristics The first important point to note is that all of the four RCTs assess different therapeutic interventions. Therefore, they were clinically and therapeutically heterogeneous. See Table 3 for detail of the each study’s characteristics.

205

3.4. Efficacy Because four different therapeutic interventions were used, it is difficult to make direct comparison of therapeutic benefit using quantitative analysis. However, qualitative analysis is possible when assessing the methodological quality of the RCTs examining conservative treatment of ITBFS. Of the four RCTs identified, three had IVS’s of 4 (refer to Table 2). Using the qualitative rating system, as mentioned earlier, it appears there is limited evidence (Level 3) to support the use of conservative interventions in the treatment of ITBFS. Some discussion of the key features of these studies is reported by intervention as follows. 3.4.1. Non-steroidal anti-inflammatory drugs (NSAIDs) Schwellnus et al. (1991) conducted a RCT on 43 patients with unilateral ITBFS with pain that was severe enough to limit running or who had had to stop running as a consequence of the pain. Subjects were randomly allocated to three groups. Initial treatment to all subjects consisted of rest, ice application and medication from day 0 to 7. From day 3 to 7 all subjects received standard physiotherapy treatment consisting of ultrasound, transverse friction massage (on days 3, 5 and 7) and daily ITB stretching. The medication was delivered over the 7 days in a double blind, placebo-controlled fashion with Group 1 given a placebo anti-inflammatory medication, Group 2 an anti-inflammatory only (50 mg diclofenac) and Group 3 a combined anti-inflammatory/ analgesic (400 mg ibuprofen, 500 mg paracetamol, 20 mg codeine phosphate) medications. Outcome measures included both daily pain and running pain, each measured via the visual analogue scale (VAS). Running pain was measured by a validated treadmill test at 3 and 7 days after treatments commenced. Results of this study demonstrated that during the first week of treatment, physiotherapy in conjunction with combined anti-inflammatory/analgesic medication was the most effective management. Significant differences were seen in the combined group with decreased running pain and increased running time/distance from 0 to 7 days, compared to the other experimental groups. The combined group was also the only group to show a significant decrease in running pain at the 3-day test. It was of interest to note that there was a significant reduction in daily pain seen across all groups. 3.4.2. Deep transverse friction massage (DTFM) Schwellnus et al. (1992) commented that the use of DTFM, in the treatment of ITFBS, is often reported on the basis of anecdotal evidence that it might be effective. Schwellnus et al. (1992) also commented that it seems contradictory that friction techniques may be beneficial in an injury where the mechanism of the injury is friction. In order to test these two statements Schwellnus

ARTICLE IN PRESS 206


et al. (1992) conducted an RCT to establish the therapeutic benefit of DTFM. Twenty subjects with chronic ITBFS (414 days duration) were randomly divided into two groups. Both groups received treatment consisting of rest (apart from treadmill running exercise tests), ice twice a day and baseline physiotherapy treatment of daily stretching exercises to the ITB and 5 min of therapeutic low dose ultrasound on days 3, 5, 7 and 10. The intervention group were also given DTFM for 10 min on the treatment days whereas the control group received only the general physiotherapy treatment on the same days. Results of this study found that daily pain and treadmill running pain levels were both significantly reduced (P ¼ 0.0005) in both groups with the authors concluding that the addition of deep friction massage did not alter the therapeutic outcome of the condition. 3.4.3. Phonophoresis versus immobilization Bischoff et al. (1995) conducted an RCT comparing phonophoresis (using 10% hydrocortisone cream as the active drug) and knee immobilization, over a 2-week period in a group of navy diving students who had developed ITBFS as a result of rigorous physical training involving a significant amount of running. All subjects were of similar age (22–23 years) and had symptoms for 15–17 days prior to entering the trial. The subjects were randomly assigned to either the knee immobilization group (three panel knee immobilizer) or the phonophoresis group. All subjects received ice massage and non-steroidal anti-inflammatory medication. Outcome measures in this study were the number of days required until pain free on examination and the ability to run on a treadmill at 6.5 miles per hour. Results of this study concluded subjects in the phonophoresis group recovered from the injury in fewer than 10 days and had significantly less pain during the treadmill running test than the immobilization group. 3.4.4. Corticosteroid injection Gunter and Schwellnus (2004) conducted an RCT looking at 18 runners with an acute onset of ITBFS (o14 days duration). Subjects were randomly allocated into two groups: Group A receiving an injection of corticosteroid (40 mg methylprednisilone and 10 mg 1% lignocaine hydrochloride) deep to the distal ITB, and Group B receiving a placebo injection (20 mg 1% lignocaine hydrochloride). Subjects were instructed not to run for 14 days following the injection and to apply ice to the area twice daily at 12 h intervals for 30 min. No physiotherapy treatment was provided to subjects in this study. Outcome measures were pain measured with a VAS and an ability to perform a treadmill running test for 30 min at the subjects best recent 10 km running speed on days 7 and 14 following the injection. Although there was a clinical improvement in both

groups, a significant (P ¼ 0.01) decrease (30%) in running pain (measured with a VAS following a treadmill test) was observed in the cortisone injection group compared to control group.

4. Discussion 4.1. The conservative management of ITBFS The results of this review identified only four RCTs regarding the conservative management of ITBFS. These RCTs investigated four different types of treatments including NSAIDs deep friction massage, phonophoresis versus immobilization, and corticosteroid injection. Some discussion of the key features of these studies is pertinent. Following the qualitative statistical analysis, the authors of this review concluded that there is limited evidence to suggest that the conservative treatments analysed here are beneficial in the treatment of ITBFS. From this review, it is evident that in the majority of studies a course of physiotherapy treatment was used as baseline, which involved a combination of ice, ultrasound, deep friction massage and stretching. Indeed, it is not uncommon to find reference to the conservative treatments, within the literature pertaining to treatment of ITBFS. In light of the analysis contained within this systematic review, it seems ironic that many of these interventions are commonly used within clinical practice and their use appears to be based on no firm evidenceand research-based rationale. 4.2. Methodological quality Three of the four RCTs reviewed (Schwellnus et al., 1991, 1992; Gunter and Schwellnus, 2004) were given an IVS of four suggesting the authors felt that these studies were of moderate methodological quality. Analysis of these studies, indicate there appears to be some benefit from using NSAIDs/analgesics and corticosteroid injections and no benefit from using DTFM. The fourth RCT (Bischoff et al., 1995) examined phonophoresis versus immobilization. This study concluded that phonophoresis was more beneficial compared to immobilization. However, there was no blinding evident throughout this RCT and the present authors deemed that this study was of limited methodological quality. It is very difficult to therefore deem this study worthy of consideration when making educated judgement as to the true effectiveness of these interventions in the management of ITBFS. Of most interest was the lack of attention of all the studies to the various aspects of blinding. For example, only two studies (Schwellnus et al., 1991; Gunter and Schwellnus, 2004) satisfied the items relating to subject


blinding (Item 5). The other two RCTs either did not adequately blind the subjects or did not mention this. Only one of the four RCTs satisfied the respective items relating to therapist blinding/Item 6 (Schwellnus et al., 1991) and assessor blinding/Item 7 (Schwellnus et al., 1992). The study by Bischoff et al. (1995) did not satisfy any of the items relating to blinding, either because there was no blinding or that blinding was not mentioned. 4.3. Future research Following the extensive literature search, carried out for this review, there is an obvious paucity of research concerning the conservative management of ITBFS. Not only is there a lack in quantity of such research, upon dissection of the scarce research that is available, there seems to also be a paucity of quality. It now seems apparent that for any of the many varieties of conservative therapies, for treatment of ITBFS, that there is no research base available to conclude any clear benefit from the clinical use of any of the conservative therapies mentioned. If this is indeed the case, then future research must attempt to fill this void. From the RCTs that were available, it seems that the methodological quality of all these studies was well below a level that allowed any credible conclusions or answers to be sought. Additionally, common to all these studies was a lack of systematic blinding. It would be advisable for future research to acknowledge this problem and attempt to organize more robust methodology in order to answer the important research questions asked. Not only were the interventions heterogeneous through the four RCTs reviewed, so to were a number of other key features including outcome measures and duration of subjects symptoms. With regard to duration of symptoms, some papers looked at the more acute stages of ITBFS (i.e. Bischoff et al. (1995) and Gunter and Schwellnus (2004) within 2 weeks) compared to more chronic duration (i.e. Schwellnus et al. (1992) at approximately 2 months or greater). It would be pertinent for future research to acknowledge clearly the duration of symptoms (i.e. acute versus chronic) as it is likely that some conservative treatments may have relatively greater or lesser impact at different pathological stages throughout the course of ITBFS presentation. For example, the studies looking at corticosteroid and NSAID use (Schwellnus et al., 1991; Gunter and Schwellnus, 2004) may have more application in an early phase of ITBFS where acute inflammation may be more of a clinical problem and needing to be addressed. Further to this point, for more chronic presentations of ITBFS, it may be more appropriate to guide research to look at more rehabilitation management, such as ITB stretching, pelvic and knee muscle stabilization, DTFM, orthotics prescription, etc.

207

With respect to outcome measures, it is not only important to gain some homogeneity in intervention selection but also consistency in outcome measures selected if there is going to be quantitative analysis of therapeutic benefit of conservative treatments for ITBFS. As Reid and Rivett (2005) have stated, direct quantitative comparison, within the realms of systematic review, is very difficult when interventions, and also outcome measurements for that matter, are heterogeneous. Throughout three of the four RCTs reviewed (Schwellnus et al., 1991, 1992; Gunter and Schwellnus, 2004) the same previously validated treadmill running test was used to score running pain. This outcome measure seems to be appropriate for ITBFS and is also becoming more widely used. Perhaps a validated test like this could become a standard test in ITBFS research. From a biomechanical and pathological perspective, the knowledge base regarding ITBFS seems to be healthy. The clinical application of such theories is both possible and plausible. There now needs to be research of sufficient quality and quantity to enable these theories to be challenged and either accepted or discarded.

5. Conclusion ITBFS is a common repetitive strain injury of the lateral aspect of the knee. The pathomechanics and clinical presentation are well understood. However, trying to determine the most appropriate choice of conservative therapy has been made difficult by paucity in quality and quantity of RCTs to examine therapeutic benefit. The aetiology of ITBFS is multifactorial, with a combination of intrinsic and extrinsic factors. The causes of ITBFS are in response to the complex functional anatomy of the ITB and its action as an independent structure and indirectly through the muscles that it provides attachment to. Reviewing the efficacy of the conservative management of ITBFS has highlighted that there are a small number of RCTs investigating the effects of therapeutic interventions on ITBFS. Within the acute stage of the presenting symptoms (less than 14 days duration) corticosteroid injection alone appears to be beneficial with subjects able to return to running pain-free with 14 days of the intervention. In the more chronic presentations (greater than 14 days duration), there appears to be benefit gained from using both combined antiinflammatory/analgesic medication over anti-inflammatories alone. The inclusion of DTFM to a standard physiotherapy programme of ultrasound and stretching exercises, does not appear to produce any additional benefit. In all of the reviewed trials this generalized physiotherapy programme proved to be beneficial in



reducing both daily pain and pain experienced on treadmill running. Although this provides positive confirmation regarding the benefits of conservative treatment for ITBFS, it is unfortunate that there are no RCTs examining the benefit of these different modalities specifically or in isolation. When investigating the novel delivery of anti-inflammatory medication via phonophoresis, a significant reduction in pain was accomplished when compared to immobilization. The evidence for the use of conservative treatment in the management of ITBFS appears to be limited and of insufficient quality. The research that is available is heterogeneous and inconsistent. Further examination of the clinical effect of conservative therapies, in an ITBFS population, will be of great importance to evidencebased management of this condition and must direct future research.

References Almeida S, Williams K, Shaffer R, Brodine S. Epidemiological patterns of musculoskeletal injuries and physical training. Medicine and Science in Sports and Exercise 1999;31(8):1176–82. Anderson G. Iliotibial band friction syndrome. The Australian Journal of Science and Medicine in Sport 1991;23(3):81–3. Aronen J, Cronister R, Regan K, Hensien M. Practical, conservative management of iliotibial band syndrome. The Physician and Sports Medicine 1993;21(6):59–69. Austermuehle P. Common knee injuries in primary care. Nurse Practitioner 2001;26(10):26–40. Barber F, Sutker A. Iliotibial band syndrome. Sports Medicine 1992;14(2):144–8. Bischoff C, Prusaczyk W, Sopchick T, Pratt N, Goforth H. Comparison of phonophoresis and knee immobilization in treating iliotibial band syndrome. Sports Medicine, Training and Rehabilitation 1995;6:1–6. Brosseau L, Casimiro L, Milne S, Robinson V, Shea B, Tugwell P, et al. Deep transverse friction massage for treating tendinitis (Cochrane Review). The Cochrane Library 2004(Issue 2). Clark HD, Wells GA, Huet C, McAlister FA, Salmi LR, Fergusson D, et al. Assessing the quality of randomized trials: reliability of the Jadad scale. Controlled Clinical Trials 1999;20:448–52. Ekman E, Pope T, Martin D, Curl W. Magnetic resonance imaging of iliotibial band syndrome. The American Journal of Sports Medicine 1994;22(6):851–4. Farrell K, Reisinger K, Tillman M. Force and repetition in cycling: possible implications for iliotibial band friction syndrome. The Knee 2003;10:103–9. Fredericson M, Guillet M, DeBenedictus L. Quick solutions for iliotibial band syndrome. The Physician and Sports Medicine 2000;28(2):52. Fredericson M, White JJ, MacMahon JM, Andriacchi TP. Quantitative analysis of the relative effectiveness of 3 iliotibial band stretches. Archives of Physical Medicine and Rehabilitation 2002;83:589–92. Gunter P, Schwellnus M. Local corticosteroid injection in iliotibial band friction syndrome in runners: a randomised controlled trial. British Journal of Sports Medicine 2004;38:269–72.

Holmes J, Pruitt A. Iliotibial band syndrome in cyclists. The American Journal of Sports Medicine 1993;21(3):419–24. Jordaan G, Schwellnus M. The incidence of overuse injuries in military recruits during basic military training. Military Medicine 1994; 159(6):421–6. Karjalainen K, Malmivaara A, van Tulder M, Roine R, Jauhiainen M, Hurri H. Multidisciplinary biopsychosocial rehabilitation for subacute low back pain in working-age adults: a systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine 2001;26(3):262–9. Kirk K, Kuklo T, Klemme W. Iliotibial band friction syndrome. Orthopedics 2000;23(11):1209–15. Levin J. Run down: battling IT band syndrome in long distance runners. Biomechanics 2003;1:22–5. Maher C, Sherrington C, Herbert R, Moseley A, Elkins M. Reliability of the PEDro Scale for rating quality of randomized controlled trials. Physical Therapy 2003;83(8):713–21. McNicol K, Taunton J, Clement D. Iliotibial tract friction syndrome in athletes. Canadian Journal of Applied Sport Science 1981; 6(2):76–80. Messier S, Edwards D, Martin D, Lowery R, Cannon D, James M, et al. Etiology of iliotibial band friction syndrome in distance runners. Medicine and Science in Sports and Exercise 1995;27(7): 951–60. Nemeth W, Sanders B. The lateral synovial recess of the knee: anatomy and role in chronic iliotibial band friction syndrome. Arthroscopy 1996;12(5):574–80. Nishimura G, Yamato M, Tamai K, Takahashi J, Uetani M. MR findings in iliotibial band syndrome. Skeletal Radiology 1997; 26(9):533–7. Noble H, Hajek M, Porter M. Diagnosis and treatment of iliotibial band tightness in runners. The Physician and Sports Medicine 1982;10(4):67–74. Novacheck T. Running injuries: a biomechanical approach. Journal of Bone and Joint Surgery 1998;80(8):1220–32. Orava S. Iliotibial tract friction syndrome in athletes-an uncommon exertion syndrome on the lateral side of the knee. British Journal of Sports Medicine 1978;12(2):69–73. Orchard J, Fricker P, Abud A, Mason B. Biomechanics of iliotibial band friction syndrome in runners. American Journal of Sports Medicine 1996;24(3):375–9. Puniello M. Iliotibial band tightness and medial patellar glide in patients with patellofemoral dysfunction. The Journal of Orthopaedic and Sports Physical Therapy 1993;17(3): 144–8. Reid S, Rivett D. Manual therapy treatment of cervicogenic dizziness: a systematic review. Manual Therapy 2005;10:4–13. Renne J. The iliotibial band friction syndrome. The Journal of Bone and Joint Surgery 1975;57–A(8):1110–1. Schwellnus M, Theunissen L, Noakes T, Reinach S. Anti-inflammatory and combined anti-inflammatory/analgesic medication in the early management of iliotibial band friction syndrome. South African Medical Journal 1991;79:602–6. Schwellnus M, Mackintosh L, Mee J. Deep transverse frictions in the treatment of iliotibial band friction syndrome in athletes: a clinical trial. Physiotherapy 1992;78(8):564–8. Taunton J, Ryan M, Clement D, McKenzie D, Lloyd-Smith D, Zumbo B. A retrospective case–control analysis of 2002 running injuries. British Journal of Sports Medicine 2002;36(2): 95–101. van Tulder M, Assendelft W, Koes B, Bouter L. Method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group for spinal disorders. Spine 1997;22(20):2323–30.

ARTICLE IN PRESS


Original article

Changes in pelvic floor and diaphragm kinematics and respiratory patterns in subjects with sacroiliac joint pain following a motor learning intervention: A case series Peter B. O’Sullivan, Darren J. Beales Curtin University of Technology, School of Physiotherapy, GPO Box U1987, Perth, WA 6845, Australia Received 24 May 2005; received in revised form 16 February 2006; accepted 2 June 2006

Abstract This study was a case series design. The objectives of the study were to investigate the ability of a motor learning intervention to change aberrant pelvic floor and diaphragm kinematics and respiratory patterns observed in subjects with sacroiliac joint pain (SIJP) during the active straight leg raise (ASLR) test. The ASLR test is a valid and reliable tool to assist in the assessment of load transference through the pelvis. Irregular respiratory patterns, decreased diaphragmatic excursion and descent of the pelvic floor have been reported in subjects with SIJP during this test. To date the ability to alter these patterns has not been determined. Respiratory patterns, kinematics of the diaphragm and pelvic floor during the ASLR test and the ability to consciously elevate the pelvic floor in conjunction with changes in pain and disability levels were assessed in nine subjects with a clinical diagnosis of SIJP. Each subject then undertook an individualized motor learning intervention. The initial variables were then reassessed. Results showed that abnormal kinematics of the diaphragm and pelvic floor during the ASLR improved following intervention. Respiratory patterns were also influenced in a positive manner. An inability to consciously elevate the pelvic floor pre-treatment was reversed. These changes were associated with improvement in pain and disability scores. This study provides preliminary evidence that aberrant motor control strategies in subjects with SIJP during the ASLR can be enhanced with a motor learning intervention. Positive changes in motor control were associated with improvements in pain and disability. Randomized controlled research is required to validate these results. r 2006 Elsevier Ltd. All rights reserved. Keywords: Diaphragm; Low back pain; Motor control; Pelvic floor; Respiration; Sacroiliac joint

1. Introduction The sacroiliac joint (SIJ) and surrounding ligamentous structures are reported to be a source of symptoms in subjects with a diagnosis of non-specific chronic low back pain (Young et al., 2003). Recent research has focused on a test that investigates the ability of a subject to transfer load between the lower limb and the trunk, called the active straight leg raise (ASLR) test. The validity and reliability of this test procedure has been Corresponding author. Tel.: +61 8 9266 3629; fax: +61 8 9266 3699. E-mail address: [email protected] (P.B. O’Sullivan).


established in subjects with clinically diagnosed SIJ pain (SIJP) (Mens et al., 2001, 1999; O’Sullivan et al., 2002a). This test involves lying supine and raising the leg 5 cm off the supporting surface. The test is positive when accompanied by a primary sensation of profound heaviness of the leg (7pain), which is relieved with the application of compression across the ilium. This test is reported to be positive in a sub-group of subjects with SIJP (Mens et al., 1999; Pool-Goudzwaard et al., 2005). It has been proposed that the reduction in the sensation of heaviness with the application of compression across the ilia reflects enhanced force closure through the SIJ (Pool-Goudzwaard et al., 1998; O’Sullivan et al., 2002a).


P.B. O’Sullivan, D.J. Beales / Manual Therapy 12 (2007) 209–218

Recent research has documented motor control deficits in the presence of SIJP (O’Sullivan et al., 2002a). O’Sullivan et al. (2002a) reported in a group of SIJP subjects with a positive ASLR the presence of aberrant motor control strategies observed during the ASLR test when compared to pain-free controls. Using real time ultrasound and spirometry, the authors’ demonstrated decreased diaphragmatic motion, increased descent of the pelvic floor, increased minute ventilation and respiratory rate, and altered breathing patterns in the pain subjects during the ASLR. These aberrant motor control strategies were eliminated with the addition of manual compression through the ilia applied during the ASLR. It was hypothesized that these disruptions might represent a deficit in local motor control (pelvic floor, transverse abdominal wall) within the lumbopelvic region in these subjects. This manifested as the adoption of splinting or bracing strategies of the abdominal wall with associated disrupted patterns of respiration during the ASLR, not observed in the normal subjects (O’Sullivan et al., 2002a). Furthermore the normalization of these patterns with the application of compression supported this notion. The adoption of these splinting strategies appears to represent an underlying deficit in the motor control systems ability to provide adequate local compression, or force closure, to the SIJs during the ASLR (O’Sullivan et al., 2002a). This concept is also supported by the report that abdominal bracing is less effective than preferential activation of the transverse abdominal wall muscles for increasing the compression across the SIJs (Richardson et al., 2002). To test the validity of this hypothesis we proposed that the application of a motor learning intervention directed to the local stabilizing muscles of the pelvis would result in the normalization of the aberrant motor control strategies displayed by these subjects, with associated reductions in pain and disability. Previous studies have reported motor learning interventions to be effective in altering specific motor control deficits in the presence of chronic low back (O’Sullivan et al., 1997, 1998) and knee pain (Cowan et al., 2002), but to date no study has investigated these specific changes with SIJP during the ASLR test.

2. Methods Nine subjects (8 female and 1 male) with a clinical diagnosis of SIJP and a positive ASLR test were recruited for this study. These subjects were recruited directly from a previous study by O’Sullivan et al. (2002a) providing a series of clinical case studies. Four of the 13 subjects from the original study declined to be involved in the intervention aspect of the study as they

were already under different forms of management. The inclusion criteria included pain over the SIJ without proximal referral (Maigne et al., 1996; Young et al., 2003) present for a duration of at least 3 months and showing no signs of abating, no impairment of spinal range of motion, a positive ASLR test (Mens et al., 1999, 2001) and at least three out of five positive pain provocation tests which include: (1) posterior shear test (Ostgaard et al., 1994; Laslett et al., 2005); (2) sacral torsion test (Laslett et al., 2005); (3) sacral thrust test (Laslett et al., 2005); (4) distraction test (Laslett et al., 2005); and (5) tenderness on palpation of the long dorsal SIJ ligament (Vleeming et al., 2002). Potential subjects were excluded if they had a specific radiological diagnosis for their pain disorder, the presence of radicular pain, neurological deficits or disorders, hip joint pathology, an inflammatory disorder, a history of a significant respiratory disorder, were pregnant or less than 6 months post partum and/or had a body mass index of greater than 31 kg/m as previously described by O’Sullivan et al. (2002a, b). The demographic data for this group is displayed in Table 1. The methodology used in this paper has been previously utilized and described in detail (O’Sullivan et al., 2002a). Respiratory rate, tidal volume, diaphragmatic motion and pelvic floor kinematics were measured in resting supine, during the ASLR and during the ASLR with the application of manual compression through the ilia. In addition pelvic floor kinematics were measured while subjects were instructed to consciously elevate their pelvic floor muscles as described in detail elsewhere (Thompson and O’Sullivan, 2003). A Stead– Wells water-sealed spirometer (60 Hz, serial number: 3657, Collins, USA) was used to record respiratory rate and tidal volume from which minute ventilation was calculated. Movement of the diaphragm was recorded utilizing real-time ultrasound to visualize the leading edge of the diaphragm (Cohen et al., 1994). The probe was positioned transversely and angled superiorly below the right costal margin in the midclavicular line. Pelvic floor motion was also recorded using real-time ultrasound with the probe positioned trans-abdominally and Table 1 Demographic data of subjects Age in years Gender Duration of symptoms in months Weight (kg) Height (cm) BMI (kg/m2) Number of subjects post-partum Number of subjects post-trauma Number of subjects with a subjective complaint of bladder dysfunction Data expressed as mean (standard deviation).

34.9 (11.2) 8 female/1 male 44.1 (41.2) 64.8 (6.5) 164.3 (6.5) 24.2 (3.4) 4 9 9

ARTICLE IN PRESS P.B. O’Sullivan, D.J. Beales / Manual Therapy 12 (2007) 209–218

angled inferiorly and posteriorly (Walz and Bertermann, 1990). This allowed visualization of the bladder. Due to the anatomical relationship of the bladder to the pelvic floor (DeLancey, 1994) motion of the inferior bladder can be interpreted as motion of the pelvic floor. For both variables inbuilt electronic calipers were used to record motion. A Toshiba Sonolayer SSA-250A (serial number: 32926, Shimoishigami, Japan) real-time ultrasound unit (3.75 MHz probe) in movement mode was used for this purpose. The reliability of these methods have been previously reported (O’Sullivan et al., 2002a; Thompson and O’Sullivan, 2003; Sherburn et al., 2005; Thompson et al., 2005). Subjects then underwent a motor learning intervention tailored to their individual clinical presentation over a 12-week period. Three treating physiotherapists were involved in the study. The specifics of this intervention are discussed below. Following the intervention the dependant variables were reassessed. Measurements were carried out in a clinical practice by one of two physiotherapists, one of whom was also involved in the intervention process. In conjunction with these physiological measures the Short Form McGill Pain Questionnaire (Melzack, 1987) including a visual analogue scale for pain severity and the Oswestry Disability Index (Fairbank et al., 1980) were employed to document changes in pain and functional status following the treatment period. The study was approved by the Human Research Ethics Committee of Curtin University of Technology and written informed consent was obtained from all participants prior to testing. Initial data analysis involved visual inspection of the spirometry traces. Statistical analysis was then performed with SPSS Version 10.0 for Windows. Sonography and spirometry data were analysed using a 2 group (pre/post-treatment) by 3 condition (resting supine, ASLR, ASLR with compression) analysis of variance. Simple contrasts were performed between all possible pairs of the three conditions. Paired t-tests were performed on the pain and disability measures as well as the pelvic floor kinematics during conscious pelvic floor contraction. A critical alpha value of 0.05 was used to determine statistical significance.

3. Intervention model The motor learning intervention model utilized in this study was adapted from work described elsewhere (O’Sullivan et al., 1997; Richardson et al., 1999; O’Sullivan, 2005b). This model is directed by the specific classification of a group of disorders where deficits in motor control appear to be a mechanism for increased strain and resultant ongoing pain (Elvey and O’Sullivan, 2005; O’Sullivan, 2005a). Within this management

211

model the impairments of motor control that are considered to be linked to the pain disorder are identified, and correct patterns are trained within a cognitive as well as a physical framework. The aim of the intervention was to retrain the local stabilizing muscles of the pelvis in a functional and automatic manner while gaining pain control and enhancing functional capacity. The reported functional impairments for each individual specifically directed their intervention although all patients reported pain aggravation during sustained sitting and standing. Components of the intervention paradigm are presented in Fig. 1. Each subject was initially educated that they had specific deficits in their local stabilizing muscles of the pelvis (pelvic floor, transverse fibres of the abdominal wall) that had resulted in increased strain across the pain sensitive structures of the pelvis. These deficits were identified as a potential mechanism for ongoing pain and disability. Training of each subject began in supine crook lying with a semi-full bladder utilizing transverse abdominal real-time ultrasound imaging of the pelvic floor. This was conducted as a means of providing visual feedback in order to teach each subject to achieve an elevating contraction of the pelvic floor with simultaneous co-contraction of the lower transverse abdominal wall (transverse abdominis and the transverse fibres of internal oblique) without associated breath holding and/or global bracing of the abdominal wall (Thompson et al., 2006a). Once the correct pattern of contraction had been achieved, the holding capacity of the muscles was trained for up to 30 s at a time. This stage took up to 4 weeks of training. This new motor pattern was then progressed to upright sitting, with the pelvis in slight anterior tilt, a neutral lumbar lordosis and the thorax in a relaxed neutral position. The exact sitting position was considered critical to enable pain control and facilitate automatic activation of the local stabilizing muscles (O’Sullivan et al., 2002b; O’Sullivan et al., 2006). The holding capacity in this posture was then trained so that the posture could be maintained for long periods of time, such as sitting in a non-supported chair for up to 30 min while watching TV or reading, in order to improve the endurance of the trunk postural muscles. Concurrently subjects were instructed in moving from sitting to standing while maintaining appropriate lumbopelvic alignment, to enable pain-free transfer of load during functional movement tasks. Subjects were then taught to alter their standing posture to align the thorax over the pelvis with a neutral lumbar lordosis, and avoid ‘sway’ postures known to inhibit the local stabilizing muscles of the lumbopelvic region (O’Sullivan et al., 2002b). This new posture was then trained in its holding capacity during single leg standing followed by walking. Initially subjects were instructed to walk



PHYSICAL COMPONENTS OF MOTOR LEARNING INTERVENTION Elevating contraction of the pelvic floor with simultaneous co-contraction of the lower transverse abdominal wall (transverse abdominis and the transverse fibres of internal oblique) without associated breath holding and/or global bracing of the abdominal wall

Train neutral lordosis in sitting with relaxed thoracic postures

Train load transfer such as sit to stand

Train aligned standing posture (avoiding sway standing) with neutral lumbar lordosis and relaxed thorax

Integrate postural alignment into single leg stance and walking

Train trunk loading such as bending and lifting as directed by aggravating factors reported by patient COGNITIVE COMPONENTS OF INTERVENTION Enhanced understanding of the pain mechanism Enhanced body awareness and control Learning strategies to develop pain control Enhancement of positive coping strategies and beliefs regarding disorder Self management of disorder Enhancement of functional capacity Independence from passive treatment Fig. 1. The first part of this figure presents a paradigm for clinical utility of the physical dimension of the intervention model. This is complimented by list of the cognitive components of the intervention model.

with control until they could hold the motor pattern for up to 30 min at a time. Other functional movement tasks were then identified and retrained based on pain provoking activities reported by each of the subjects. It is important to note that each subject reported pain control when they were able to adopt the new postures while maintaining their motor pattern.

Each subject was seen weekly over a period of 12 weeks and instructed to carry out a home exercise programme on a daily basis. Three subjects wore SIJ belts during the first 3 weeks of the training period until they had achieved functional activation of their local stabilizing muscles at which time they reported that they no longer required the belt. No other co-interventions were carried out during the study period.


4. Results The individual pre-treatment data for respiratory rate, tidal volume, diaphragmatic motion and pelvic floor kinematics for these subjects was extracted from our previous study (O’Sullivan et al., 2002a) and reprocessed as a new group to provide the preintervention baseline for this case intervention series. 4.1. ASLR tasks 4.1.1. Pelvic floor kinematics As there was no pelvic floor motion in resting supine this condition was not included for analyses. A significant difference was found between pre- and post-treatment (F ¼ 12.142, P ¼ 0.008) and between the remaining two conditions (F ¼ 48.700, Po0.001). There was an interaction between ASLR and ASLR with manual pelvic compression (F ¼ 12.374, P ¼ 0.008). The distinguishing feature of this interaction was the decrease in pelvic floor descent during the ASLR following the intervention (Fig. 2). No subject had any descent of the pelvic floor during the ASLR with compression post-treatment. 4.1.2. Diaphragmatic excursion A significant difference between pre- and posttreatment (F ¼ 6.105, P ¼ 0.039) was found for diaphragm excursion and between the three conditions (F ¼ 11.915, P ¼ 0.006). An interaction was distinguished between resting supine and the ASLR (F ¼ 25.928, P ¼ 0.001), and between ASLR and ASLR with manual pelvic compression (F ¼ 19.837, Resting Supine

ASLR

ASLR With Compression

P ¼ 0.002). The main feature of this interaction was increased diaphragmatic excursion during the ASLR post-intervention (Fig. 3). 4.1.3. Respiratory function Changes in minute ventilation did not reach a statistically significant difference pre- and post-treatment (F ¼ 4.966, P ¼ 0.056) nor between the conditions (F ¼ 4.008, P ¼ 0.069). However, a trend towards reduced minute ventilation during the ASLR postintervention was observed (Fig. 4A). Subanalysis of the components of minute ventilation was also undertaken. Respiratory rate was reduced in post-intervention compared to pre-intervention (F ¼ 8.563, P ¼ 0.019), however a significant difference was not identified between the three test conditions (F ¼ 1.267, P ¼ 0.339). No significant interaction was found between resting supine and the ASLR (F ¼ 2.465, P ¼ 0.155) or between the ASLR and ASLR with compression (F ¼ 2.861, P ¼ 0.129). Fig. 4b therefore denotes a trend towards decrease respiratory rate during the ASLR after the intervention. No difference in tidal volume was observed pre- and post-treatment (F ¼ 1.900, P ¼ 0.205) or between conditions (F ¼ 0.286, P ¼ 0.760). The respiratory traces across all subjects were variable. In spite of a lack of statistically significant difference, visual inspection of the respiratory traces highlighted interesting changes between the pre- to postintervention period. Three cases are depicted in Fig. 5 as examples. (Note: pre-treatment traces were previously reported in O’Sullivan et al. (2002a).) The notable feature of these traces is the marked improvement of the respiratory traces during the ASLR following intervention. The increase in respiratory rate and decrease in Resting Supine

12

ASLR


16

10 8 6 4 2 0 Pre-Treatment

Post -Treatment

Fig. 2. Mean (standard error of the mean) measurements for pelvic floor descent pre- and post-treatment. Note there is no bar for resting supine as there was no pelvic floor movement during this test condition. The graph depicts decreased descent of the pelvic floor during the ASLR post-treatment. Note there is no error bar for pelvic floor descent post-treatment during the ASLR with compression as all subjects had no descent during this task.

Mean Diaphragmatic Excursion (mm)

Mean Pelvic Floor Descent (mm)

14

213

14 12 10 8 6 4 2 0 Pre-Treatment

Post -Treatment

Fig. 3. Mean (standard error of the mean) measurements for diaphragmatic excursion pre- and post-treatment, denoting increased diaphragmatic motion during the ASLR following the intervention period.


214 Resting Supine

ASLR


Mean Minute Ventilation (L/min)

16 14 12 10 8

4.3. Pain and disability scores

6

Significant differences were found between pre- to post-treatment for the Short Form McGill Pain Questionnaire (Po0.001), the VAS for usual pain (P ¼ 0.001) and the Oswestry Low Back Pain Questionnaire (P ¼ 0.003), denoting reductions in pain and disability associated with the intervention (Fig. 6). In addition it was recorded in the treatment notes that all subjects reported reduced heaviness during the ASLR test following the intervention although this was not formally measured.

4 2 0 Pre-Treatment

(a)

Resting Supine

ASLR

Post -Treatment ASLR With Compression

Mean Respiratory Rate (breaths/min)

20 18 16 14

5. Discussion

12 10 8 6 4 2 0

(b)

conscious pelvic floor contraction with an average magnitude of 6.12 mm (SE ¼ 0.97). This change was significantly different (Po0.001). It was recorded in the treatment notes that all subjects reported improved bladder function following the intervention although this was not formally measured.

Pre-Treatment

Post -Treatment

Fig. 4. Mean (standard error of the mean) values for: (a) minute ventilation, and (b) respiratory rate, during the three test conditions before and after treatment. Both denote trends for improvement during the ASLR post-treatment.

tidal volume during the ASLR observed in Fig. 5a before the intervention match that of the resting supine condition post-intervention. Similarly the multiple breath holds displayed during the ASLR pre-intervention in Fig. 5b, denoted by the flat line in the respiratory trace, are not observed after the intervention. The erratic pattern seen in Fig. 5c during the ASLR, while not equivalent to resting supine post-intervention, has improved. 4.2. Conscious pelvic floor elevation task Pelvic floor kinematics during conscious contraction of the pelvic floor before and after intervention was only available for eight of the nine subjects due to lost data. Prior to the intervention all subjects exhibited descent of the bladder with this task. The average magnitude of this descent was 11.5 mm (SE ¼ 2.09). After the intervention all subjects demonstrated elevation during

This study provides preliminary evidence that a specific motor learning intervention for subjects with SIJP can positively change pelvic floor and diaphragm kinematics and patterns of respiration observed during the ASLR. These changes were associated with concurrent reductions in pain and disability in a group of chronically disabled pelvic pain subjects. However as this study is a case series and did not have a control group or blinded independent investigators, the findings should be viewed with caution. Randomized controlled research is required to validate these results. While it is well recognized that movement and motor control impairments commonly co-exist with lumbopelvic pain disorders (O’Sullivan et al., 2002a; Hungerford et al., 2003; Pool-Goudzwaard et al., 2005), the mere presence of these impairments does not establish cause and effect. Movement and motor control impairments are known to occur secondary to the presence of pain (Hodges and Moseley, 2003; van Dieen et al., 2003) as well as pathological and psychological processes (Frymoyer et al., 1985; Hall and Elvey, 1999; Hodges and Moseley, 2003; Marras, 2004; O’Sullivan, 2005a). Attempts to simply ‘normalize’ movement or motor control impairments in many of these disorders without consideration for their underlying mechanism may be inappropriate and ineffective. There is however growing evidence that some disorders do exist where movement and motor control impairments appear to result in abnormal tissue loading and pain, leaving them amenable to specific physical therapy intervention (O’Sullivan et al., 1997; Hides et al., 2001; Cowan et al., 2002; Stuge et al., 2004b). Furthermore there is evidence that patterns of abnormal


215

Fig. 5. Respiratory patterns of three subjects before and after treatment. Traces for Subjects A and B during the ASLR post-treatment match that of the resting supine condition. Subject C demonstrates an improved, though not fully resolved, respiratory pattern during the ASLR post-treatment. Pre-treatment traces previously published in O’Sullivan et al (2002a) (Sup ¼ resting supine, ALSR ¼ active straight leg raise, Comp ¼ ASLR with manual pelvic compression).

motor behaviour can be altered with specific exercise or motor learning interventions, leading to improvements in pain and disability in specific pain populations (O’Sullivan et al., 1998; Cowan et al., 2002). Clearly a priority for clinicians is the ability to identify specific patient groups for whom motor learning interventions are appropriate and effective. The subjects in this current study represent a subgroup with non-specific chronic pelvic pain as they had no radiological diagnosis specific for their disorder. Selection was based on specific clinical inclusion criteria.

There is growing evidence to support that this cluster of signs and symptoms are associated with pain disorders of the SIJ and its supporting structures (Mens et al., 2001; Young et al., 2003; Stuge et al., 2004a; Laslett et al., 2005; Pool-Goudzwaard et al., 2005). In our previous paper we proposed that the altered pelvic floor and diaphragm kinematics and patterns of respiration in subjects with SIJP during an ASLR that were normalized with manual pelvic compression (O’Sullivan et al., 2002a) may reflect loss of force closure within the pelvis, secondary to a deficit in the


216

Pre-Treatment

Post-Treatment

Pain and Disability Scales

100 90 80 70 60 50 40 30 20 10 0 McGill (x/45)

VAS (mm)

Oswestry (%)

Fig. 6. Mean (standard error of the mean) scores pre- and posttreatment for the Short Form McGill Pain Questionnaire, the Visual Analogue Scale for usual pain and the Oswestry Low Back Pain Questionnaire. Significant improvements were found for all three of these variables.

local muscles such as the pelvic floor and transverse abdominal wall. Biomechanical studies show that the pelvic floor and transverse abdominal wall have the capacity to locally compress or stabilize the SIJs (Snijders et al., 1993a, b; Richardson et al., 2002; PoolGoudzwaard et al., 2004; van Wingerden et al., 2004). Growing evidence suggests that dysfunction of these muscles is present in subjects with SIJP (Avery et al., 2000; O’Sullivan et al., 2002a; Hungerford et al., 2003; Pool-Goudzwaard et al., 2005). The improvement of the altered motor control patterns in the current study, following a motor learning intervention targeting these local force closure muscles, lends support for this hypothesis. Further to this the clinical reports of the reduction of the ‘heaviness’ associated with the ASLR may be suggestive of an enhanced motor control strategy for load transfer across the pelvis during the ASLR. Recent research has documented that depression of the pelvic floor is associated with generation of high levels of intra-abdominal pressure and global activation of the pelvic floor, abdominal wall and chest wall muscles (Thompson et al., 2006a). These bracing strategies have been shown less able to locally stabilize the SIJs (Richardson et al., 2002) and have been reported to be associated with reduced muscle activity of the pelvic floor during pelvic floor muscle contraction in women with bladder control disorders (Thompson et al., 2006b). Recent research has also documented a relationship between pelvic pain and bladder control disorders with increased pelvic floor muscle activation (Pool-Goudzwaard et al., 2005). Further research into the functioning of the pelvic floor muscles in conjunction with the other muscles of the abdomino-pelvic cavity is required to further identify patterns of altered motor control in subgroups with SIJP. In contrast a lifting contraction of the pelvic floor (as was trained in

this study), is associated with high levels of activation of the pelvic floor and transverse abdominal with minimal activation of the external oblique, rectus abdominis and chest wall muscles, minimal increase in intra-abdominal pressure and allows relaxed respiration (Sapsford et al., 2001; Thompson et al., 2006a). This local stabilizing strategy has been shown to enhance the stability of the SIJs (Richardson et al., 2002) and is also considered important for the control of continence (Bo et al., 1988; Thompson et al., 2006a). In light of this research, the findings of the current study support that a more local stabilizing strategy was adopted in the subjects following the training period, compared to a straining pattern prior to the intervention. This trained strategy closely reflects a normal motor control pattern associated with the ASLR under the pelvic compression condition, and that previously documented in a pain-free population during ASLR (O’Sullivan et al., 2002a). It was also interesting to note the subjective reports that bladder control symptoms reduced in subjects following the intervention period. This may be suggestive of a positive change in the motor control strategies associated with the control of intra-abdominal pressure and activation of the pelvic floor muscles associated with the control of continence. Further research is warranted to further investigate these issues. Stuge and co-workers have recently reported longterm benefits from a specific stabilizing exercise programme directed to the lumbopelvic region in subjects with post partum pelvic pain (Stuge et al., 2004a, b). Interestingly these authors reported that the normalization of the ASLR test was associated with increased functional mobility and reductions in pain in this group of subjects. These findings suggest that the change in the ASLR was predictive of outcome in these subjects. In contrast Mens et al. (2000) reported that global training of the trunk muscles did not result in reduction of pain and disability in subjects with pelvic pain. It should be noted that improvement of the motor patterns associated with the ASLR in this study did not fully resolve the pain disorder, but rather was associated with reductions in pain and disability. Furthermore the intervention had both a functional and cognitive component to it, with subjects being taught to utilize their local stabilizing muscles so as to enhance their functional capacity with pain control. These findings may support that other physical, neurophysiological and cognitive factors may also be associated with these pain disorders. Such factors may include underlying disruption to the pelvic ligaments resulting in ongoing compromise to the form closure mechanism of the pelvis, central nervous system adaptation resulting in ongoing tissue sensitization due to a chronic pain state and cognitive factors such as anxiety, fear avoidance behaviour and poor coping strategies.


Although it cannot be ruled out that the improvements observed were as a result of a change in the subjects’ natural history, this is unlikely given that the subjects had reported chronic disabling pelvic pain for an average of 44 months with no sign of abating prior to the intervention. Furthermore there is a possibility that the changes observed were due to a placebo effect secondary to the expectation of improvement with treatment. Certainly further research is indicated to repeat this study using a randomized-controlled clinical trial design with long-term follow-up and greater subject numbers. Furthermore, repeating this study with concurrent measures of IAP and muscle activity would also provide further insight into the exact motor control strategies that were utilized in these subjects before and after the intervention period. In conclusion this study provides preliminary casebased evidence that altered kinematics of the pelvic floor and diaphragm, as well as disrupted respiratory patterns, observed in subjects with SIJP can be shifted towards those patterns observed in pain-free individuals with a motor learning intervention. Furthermore the improvement of these motor patterns was associated with functional improvements and decreased symptoms. These findings may provide some insight into the relationship between motor control strategies and the ASLR test in subjects with SIJP, however further research is required to validate this.

Acknowledgments We would like to acknowledge the statistical support of Marie Blackmore and the assistance of Julie Beetham, Anita Avery, Ivan Lin, Beatrice Tucker, Jillian Crisp, Felicitas Graf and Chris Perkin.

References Avery AF, O’Sullivan PB, McCallum M. Evidence of pelvic floor muscle dysfunction in subjects with chronic sacro-iliac joint pain syndrome. In: Proceedings of the seventh scientific conference of the international federation of orthopaedic manipulative therapists, Perth, WA, Australia, 2000. p. 35–8. Bo K, Larsen S, Oseid S, Kvarstein B, Hagen RR, Jorgensen J. Knowledge about and ability to correct pelvic floor muscle exercises in women with urinary stress incontinence. Neurourology and Urodynamics 1988;7:261–2. Cohen E, Mier A, Heywood P, Murphy K, Boultbee J, Guz A. Excursion–volume relation of the right hemidiaphragm measured by ultrasonography and respiratory airflow measurements. Thorax 1994;49(9):885–9. Cowan SM, Bennell KL, Crossley KM, Hodges PW, McConnell J. Physical therapy alters recruitment of the vasti in patellofemoral pain syndrome. Medicine and Science in Sports and Exercise 2002;34(12):1879–85. DeLancey J. Functional anatomy of the pelvic floor and urinary continence mechanism. In: Schussler B, Laycock J, Norton P,

217

Stanton S, editors. Pelvic floor re-education. Principles and practice. London: Springer; 1994. p. 9–21 [chapter 1.1]. Elvey R, O’Sullivan PB. A contemporary approach to manual therapy. In: Jull GA, Boyling JD, editors. Grieve’s modern manual therapy. 3rd ed. Edinburgh: Churchill Livingstone; 2005. p. 471–93 [chapter 33]. Fairbank JC, Couper J, Davies JB, O’Brien JP. The Oswestry low back pain disability questionnaire. Physiotherapy 1980;66(8):271–3. Frymoyer JW, Rosen JC, Clements J, Pope MH. Psychologic factors in low-back-pain disability. Clinical Orthopaedics and Related Research 1985(195):178–84. Hall TM, Elvey RL. Nerve trunk pain: physical diagnosis and treatment. Manual Therapy 1999;4(2):63–73. 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. Hodges PW, Moseley GL. Pain and motor control of the lumbopelvic region: effect and possible mechanisms. Journal of Electromyography and Kinesiology 2003;13(4):361–70. Hungerford B, Gilleard W, Hodges P. Evidence of altered lumbopelvic muscle recruitment in the presence of sacroiliac joint pain. Spine 2003;28(14):1593–600. Laslett M, Aprill CN, McDonald B, Young SB. Diagnosis of sacroiliac joint pain: validity of individual provocation tests and composites of tests. Manual Therapy 2005;10(3):207–18. Maigne JY, Aivaliklis A, Pfefer F. Results of sacroiliac joint double block and value of sacroiliac pain provocation tests in 54 patients with low back pain. Spine 1996;21(16):1889–92. Marras WS. State-of-the-art research perspectives on musculoskeletal disorder causation and control: the need for an intergraded understanding of risk. Journal of Electromyography and Kinesiology 2004;14(1):1–5. Melzack R. The short-form McGill Pain Questionnaire. Pain 1987;30(2):191–7. Mens JM, Vleeming A, Snijders CJ, Stam HJ, Ginai AZ. The active straight leg raising test and mobility of the pelvic joints. European Spine Journal 1999;8(6):468–74. Mens JM, Snijders CJ, Stam HJ. Diagonal trunk muscle exercises in peripartum pelvic pain: a randomized clinical trial. Physical Therapy 2000;80(12):1164–73. Mens JM, Vleeming A, Snijders CJ, Koes BW, Stam HJ. Reliability and validity of the active straight leg raise test in posterior pelvic pain since pregnancy. Spine 2001;26(10):1167–71. Ostgaard HC, Zetherstrom G, Roos-Hansson E. The posterior pelvic pain provocation test in pregnant women. European Spine Journal 1994;3(5):258–60. O’Sullivan P. Diagnosis and classification of chronic low back pain disorders: maladaptive movement and motor control impairments as underlying mechanism. Manual Therapy 2005a;10(4):242–55. O’Sullivan PB. ‘Clinical Instability’ of the lumbar spine: its pathological basis, diagnosis and conservative management. In: Jull GA, Boyling JD, editors. Grieve’s modern manual therapy. 3rd ed. Edinburgh: Churchill Livingstone; 2005b. p. 311–31 [chapter 22]. O’Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22(24):2959–67. O’Sullivan PB, Twomey L, Allison GT. Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. The Journal of Orthopaedic and Sports Physical Therapy 1998;27(2):114–24. O’Sullivan PB, Beales DJ, Beetham JA, Cripps J, Graf F, Lin IB, et al. Altered motor control strategies in subjects with sacroiliac joint pain during the active straight-leg-raise test. Spine 2002a; 27(1):E1–8.



O’Sullivan PB, Grahamslaw KM, Kendell M, Lapenskie SC, Moller NE, Richards KV. The effect of different standing and sitting postures on trunk muscle activity in a pain-free population. Spine 2002b;27(11):1238–44. O’Sullivan PB, Dankaerts W, Burnett A, Booth R, Carlsen C, Chen D, Schultz A. Evaluation of the flexion relaxation phenomenon in the trunk muscles in sitting. Spine 2006;31(17):2009–16. Pool-Goudzwaard AL, Vleeming A, Stoeckart R, Snijders CJ, Mens JMA. Insufficient lumbopelvic stability: a clinical, anatomical and biomechanical approach to ‘‘a-specific’’ low back pain. Manual Therapy 1998;3(1):12–20. Pool-Goudzwaard A, Hoek Van Dijke G, Van Gurp M, Mulder P, Snijders C, Stoeckart R. Contribution of pelvic floor muscles to stiffness of the pelvic ring. Clinical Biomechanics 2004; 19(6):564–71. Pool-Goudzwaard AL, Slieker Ten Hove MC, Vierhout ME, Mulder PH, Pool JJ, Snijders CJ, et al. Relations between pregnancyrelated low back pain, pelvic floor activity and pelvic floor dysfunction. International Urogynecology Journal and Pelvic Floor Dysfunction 2005;16(6):468–74. Richardson CA, Jull GA, Hodges PW, Hides JA. Therapeutic exercise for spinal segmental stabilization in low back pain. Scientific basis and clinical approach. Sydney: Churchill Livingstone; 1999. Richardson CA, Snijders CJ, Hides JA, Damen L, Pas MS, Storm J. The relation between the transversus abdominis muscles, sacroiliac joint mechanics, and low back pain. Spine 2002;27(4):399–405. Sapsford RR, Hodges PW, Richardson CA, Cooper DH, Markwell SJ, Jull GA. Co-activation of the abdominal and pelvic floor muscles during voluntary exercises. Neurourology and Urodynamics 2001;20(1):31–42. Sherburn M, Murphy CA, Carroll S, Allen TJ, Galea MP. Investigation of transabdominal real-time ultrasound to visualise the muscles of the pelvic floor. Australian Journal of Physiotherapy 2005;51(3):167–70. Snijders C, Vleeming A, Stoeckart R. Transfer of lumbosacral load to iliac bones and legs. Part 1: biomechanics of self-bracing of the sacroiliac joints and its significance for treatment and exercise. Clinical Biomechanics 1993a;8(6):285–94. Snijders C, Vleeming A, Stoeckart R. Transfer of lumbosacral load to iliac bones and legs. Part 2: loading of the sacroiliac joints when lifting in a stooped posture. Clinical Biomechanics 1993b;8(6):295– 301.

Stuge B, Laerum E, Kirkesola G, Vollestad N. The efficacy of a treatment program focusing on specific stabilizing exercises for pelvic girdle pain after pregnancy: a randomized controlled trial. Spine 2004a;29(4):351–9. Stuge B, Veierod MB, Laerum E, Vollestad N. The efficacy of a treatment program focusing on specific stabilizing exercises for pelvic girdle pain after pregnancy: a two-year follow-up of a randomized clinical trial. Spine 2004b;29(10):E197–203. Thompson JA, O’Sullivan PB. Levator plate movement during voluntary pelvic floor muscle contraction in subjects with incontinence and prolapse: a cross-sectional study and review. International Urogynecology Journal and Pelvic Floor Dysfunction 2003;14(2):84–8. Thompson JA, O’Sullivan PB, Briffa NK, Neumann P. Differences in muscle activation patterns during pelvic floor muscle contraction and Valsalva manouevre. Neurourology and Urodynamics 2006a;25(2):148–55. Thompson JA, O’Sullivan PB, Briffa K, Neumann P. Altered muscle activation patterns in symptomatic women during pelvic floor muscle contraction and Valsalva manouevre. Neurourology and Urodynamics 2006b;25(3):268–76. Thompson JA, O’Sullivan PB, Briffa K, Neumann P, Court S. Assessment of pelvic floor movement using transabdominal and transperineal ultrasound. International Urogynecology Journal and Pelvic Floor Dysfunction 2005;16(4):285–92. van Dieen JH, Cholewicki J, Radebold A. Trunk muscle recruitment patterns in patients with low back pain enhance the stability of the lumbar spine. Spine 2003;28(8):834–41. van Wingerden JP, Vleeming A, Buyruk HM, Raissadat K. Stabilization of the sacroiliac joint in vivo: verification of muscular contribution to force closure of the pelvis. European Spine Journal 2004;13(3):199–205. Vleeming A, de Vries HJ, Mens JM, van Wingerden JP. Possible role of the long dorsal sacroiliac ligament in women with peripartum pelvic pain. Acta Obstetricia et Gynecologica Scandinavica 2002;81(5):430–6. Walz P, Bertermann H. Ultrasound examination of bladder and prostate. Urologia Internationalis 1990;45(4):217–30. Young S, Aprill C, Laslett M. Correlation of clinical examination characteristics with three sources of chronic low back pain. The Spine Journal 2003;3(6):460–5.

ARTICLE IN PRESS


Original article

Activation of vastus medialis oblique is not delayed in patients with osteoarthritis of the knee compared to asymptomatic participants during open kinetic chain activities John Dixona,, Tracey E. Howeb a

Teesside Centre for Rehabilitation Sciences, University of Teesside, The James Cook University Hospital, Marton Road, Middlesbrough, UK b HealthQWest, Glasgow Caledonian University, Cowcaddens Road, Glasgow, UK Received 7 July 2005; received in revised form 23 February 2006; accepted 1 June 2006

Abstract This study investigated whether the onset of electromyographic (EMG) activity of vastus medialis oblique (VMO) was delayed relative to that of vastus lateralis (VL) in patients with osteoarthritis (OA) of the knee compared to asymptomatic participants during open kinetic chain activities. An exploratory observational cross sectional study was carried out. Two groups were tested, symptomatic OA knee patients, diagnosed by an orthopaedic surgeon, ðn ¼ 17Þ, mean (SD) age 66.0 (7.6) years, and asymptomatic participants ðn ¼ 17Þ, 56.7 (8.6) years. Surface EMG activity of VMO and VL was measured, during concentric contractions extending the knee from 901 flexion, and during maximal voluntary isometric contractions at 601 knee flexion. The EMG onset times of VMO and VL were determined visually and by algorithm. The onset timing difference (OTD) between the two muscles was calculated for each subject, by subtracting the onset time of VL from VMO. Mann–Whitney U-tests revealed that the OTD between VMO and VL was not significantly different between the groups during either contraction type (both p40.05). The results of this exploratory study may have implications for rehabilitation programmes aimed at developing preferential activation of VMO compared to VL in OA knee patients. r 2006 Elsevier Ltd. All rights reserved. Keywords: Electromyography; Knee; Quadriceps; Osteoarthritis

1. Introduction Osteoarthritis (OA) is a common cause of disability in adults (Badley and Tennant, 1993). Although previously thought of being solely due to joint wear and tear, it is now believed that muscle dysfunction is an important factor in OA knee (Shrier, 2004). Impairments that may increase joint damage over time, such as arthrogenous muscle inhibition, quadriceps weakness, and slowed protective reflexes have been implicated (Hurley, 1999). Clinicians observe what appears to be selective atrophy of muscle vastus medialis oblique (VMO) in OA knee patients. There Corresponding author. Tel.: +44 1642 384125.

E-mail address: [email protected] (J. Dixon). 1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.06.012

is evidence to suggest that VMO may be preferentially affected in patellofemoral pain syndrome patients, and its activation delayed relative to that of vastus lateralis (VL) when compared to healthy subjects (Witvrouw et al., 1996; Cowan et al., 2001, 2002). Theoretically this could lead to a biomechanical imbalance at the knee joint. Despite the prevalence of OA knee, there is a scarcity of research about VMO activation in OA knee patients. This has been studied during stair climbing (Hinman et al., 2002), and during the patellar tendon reflex response (Dixon et al., 2004). In both studies the activation timing of VMO relative to VL did not differ in OA knee patients compared to asymptomatic participants. However, it has been proposed that muscle activation order may be affected by the type of muscle contraction


J. Dixon, T.E. Howe / Manual Therapy 12 (2007) 219–225

(Grabiner et al., 1994; Stensdotter et al., 2003), and voluntary muscle activation during open kinetic chain (OKC) activities remain to be investigated in this knee pathology. This type of activity is often used to evaluate quadriceps function in rehabilitation. Hence the aim of this study was to investigate onset of VMO and VL EMG activity in OA knee patients and asymptomatic participants during two OKC activities, seated nonweightbearing knee extensions (concentric contractions) extending the knee from 901 flexion, and during maximal voluntary isometric contractions (MVIC) at 601 knee flexion. We tested the hypothesis that onset of VMO EMG activity, relative to that of VL, would be delayed in OA knee patients when compared to similarly aged asymptomatic participants. We also measured torque produced by the quadriceps during MVICs using a load cell to allow a comparison with previously published studies (Tan et al., 1995; Cheing and HuiChan, 2001) reporting deficits in maximal quadriceps strength in patients with OA knee.

2. Methods An exploratory observational cross sectional study was carried out, in conjunction with a study into reflex EMG activity (Dixon et al., 2004). 2.1. Participants Ethical approval was granted by local research ethics committees. All subjects gave written and verbal informed consent before taking part in the study. Two subject groups were tested, a group of symptomatic OA knee patients, and a group of similarly aged asymptomatic participants. Asymptomatic participants comprised a convenience sample from the local area and OA patients were recruited from South Tees Hospitals NHS Trust, UK outpatients orthopaedic clinics. Diagnosis of OA knee was made by an orthopaedic surgeon, according to the American College of Rheumatology criteria (Altman et al., 1986), using clinical signs and symptoms and the presence of osteophytes determined by weightbearing radiographs. Asymptomatic participants were individuals who reported having no history of knee pain. Participants were excluded if they presented with significant cognitive, cardiorespiratory, neurological, or musculoskeletal impairments (excepting OA knee in the patient group) that limited functional ability, or reported use of medication affecting neuromuscular function, or if they could not carry out the activities. 2.2. Procedure For testing, each participant was seated on an adjustable padded ‘quadriceps chair’ (Tornvall, 1963;

Edwards et al., 1977; Dixon et al., 2004) with the posterior aspect of the knee positioned at the front edge of the seat and arms folded across the chest. The chair back-rest was set at 701 of hip flexion (Doxey and Eisenman, 1987). In bilateral OA knee patients and asymptomatic participants, data were recorded from the dominant limb, which was defined as the limb with which they would kick a ball (Cheing and Hui-Chan, 2001). The symptomatic limb of patients with unilateral OA was tested. After cleaning the skin with isopropyl alcohol, and shaving the area if necessary, active surface EMG recording electrodes (BIOPAC Inc., USA, TSD150B, Ag/Ag Cl, diameter 11.4 mm, electrode spacing 20 mm centre to centre, with a built in 350 amplification and a 3 dB bandpass of 12–500 Hz) were placed on VMO and VL at standardized sites. The electrodes were oriented in the estimated direction of the muscle fibres (Lieb and Perry, 1968). The VL electrode was sited on the muscle belly, at one-third the distance from the superior border of the patella to the greater trochanter (Mannion and Dolan, 1996), oriented 12–151 laterally from the long axis of the femur. For VMO placement was on the muscle belly 5 cm from the superior medial border patella border (Callaghan et al., 2001), oriented 50–551 medially from the long axis of the femur. Hypoallergenic conductive gel (Lectron II, Pharmaceutical Innovations Inc., USA) was applied to the electrodes to facilitate electrical contact with the skin surface. All electrodes were taped to the skin to prevent movement artifacts. A ground electrode (Red dot TM, 3M Healthcare, USA) was attached to the patella of the untested leg. 2.2.1. Concentric contraction An electrogoniometer (Type SG150, Biometrics Ltd, UK) was used to measure knee angle during the concentric contraction. This was calibrated prior to use, and attached to the knee joint of the participant with doubled sided tape. Each participant was asked to perform extension of the knee joint, from 901 knee flexion to full extension (or as far as possible) and back. The instruction ‘‘straighten your leg’’ was given. This was followed by ‘‘and relax’’ when the participant reached full extension. This was carried out ten times, each contraction separated by 20 s to minimize fatigue effects. 2.2.2. Maximal voluntary isometric contraction (MVIC) Force was measured using a load cell (TC601, Amber Instruments Ltd, UK). The load cell output was fed to a channel of the BIOPAC system and the voltage output was pre-calibrated by hanging known weights vertically from the load cell. The MVIC activity was carried out at least 3 min after the concentric contraction. The knee was positioned at 601 flexion (Kannus et al., 1987;

ARTICLE IN PRESS J. Dixon, T.E. Howe / Manual Therapy 12 (2007) 219–225

Powers et al., 1996). This position has been shown to produce a higher torque than other knee angles (Lieb and Perry, 1971). This leg position was standardized by the attachment to the chair of an adjustable arm that supported the leg of the subject at mid-calf level. The angle of flexion was verified with a goniometer and the leg support adjusted accordingly. A comfortable nonextensible ankle strap was attached to the leg of the subject at the level of the lateral malleolus. The load cell under the chair was adjusted so that the chain passed from the leg at an angle of 901. The length of the leverarm, being the distance from the centre of the knee joint to the load cell strap at the lateral malleolus, was measured to allow torque calculations. Participants carried out five maximal isometric contractions of the quadriceps with 30 s rest between each contraction to minimize fatigue. Each participant was asked to straighten the leg as forcefully as possible against the resistance of the chain for 3 s (Soderberg and Knutson, 2000) and then told to relax. The investigator gave standardized instructions of ‘ready, steady, push, push, push, relax’ to elicit a contraction of 3 s. It was confirmed that participants understood the instructions and they were given practice attempts to ensure familarization with the procedure. Verbal encouragement was given, as encouragement has been shown to increase maximal voluntary contraction values (McNair et al., 1996). The peak force was utilized as the MVIC for the purposes of the study. All data were sampled at 2048 Hz using a physiological data acquisition system (BIOPAC Inc., USA) comprising an MP100 workstation with a high-level transducer HLT100 and dedicated analysis software (AcqKnowledge 3.5.3). 2.3. Onset timing determination Onset of EMG activity, the ‘earliest rise beyond the steady state’ of the raw EMG signal (Hodges and Bui, 1996), was evaluated both visually and by computer algorithm. Both methods were used because of agerelated changes that affect the quality of the EMG signal in older people and which could affect onset times determined by algorithm, e.g. lower amplitude and poorer signal to noise ratio than younger people. For visual evaluation, the peak or trough of the first spike of the raw EMG was identified as the point of onset, by viewing the digitally stored data in the AcqKnowledge software. For computerized evaluation, an algorithm based on the work of Hodges and Bui (1996) was used to identify the EMG onset time for each electrode site (Dixon, 2004). Here, EMG data was saved as an ASCII text file with a resting baseline prior to contraction of 500 ms duration, and imported into Microsoft Excel. In Excel, the raw EMG data were full wave rectified and 50 Hz

221

low-pass filtered using a single pole first order digital filter. The algorithm calculated the mean and standard deviation (SD) of the first 100 data points (the baseline window, 48.83 ms). The SD value was then used as the threshold to determine EMG onset time. The algorithm identified onset as the time point at which the mean value of a moving window of 50 data points (24.41 ms) exceeded the mean of the baseline by 1SD of the baseline window. The mid-point of the moving window was the time determined as onset. The algorithm was also tested with 2 and 3 SDs as the threshold value, but it was found that the 1 SD threshold had best agreement with visually determined onset times using Bland and Altman limits of agreement (Dixon, 2004). The 1SD algorithm onset times were therefore used, as in previous literature (Karst and Willett, 1995). On the few occasions when the baseline window contained a marked artifact, its position was moved to ensure that artifacts did not artificially increase the SD and hence the threshold. The EMG onset time difference (OTD) between the VMO and VL was then calculated for both visually and computer evaluated results, by subtraction of the time for VL from that of VMO. A negative value therefore indicates that VMO onset is before that of VL. The median OTD (of 10 concentric and 5 MVICs) was then calculated for each subject for both activities. This was used in preference to the mean, as data for some participants were skewed. For the concentric contraction, data were excluded for one asymptomatic participant due to an electrode problem. Concentric and MVIC EMG data for one OA knee patient were excluded due to the poor signal quality possibly arising from obesity (Marks et al., 1994), which prevented successful onset determination. However this patient could perform the MVIC and their torque data were not excluded from the torque results. 2.4. Statistical analysis Data were analysed using SPSS V 10. The Mann– Whitney U-test was used to test for between group differences, as the data sets displayed either differences in variance or non-normal distributions, as has been described in other work (Karst and Willett, 1995). Values of po0.05 were regarded as statistically significant.

3. Results Descriptive characteristics of the groups are shown in Table 1. The groups were significantly different in age, mass and body mass index (Mann–Whitney U-test, all po0.05), but not in height ðp ¼ 1:00Þ, with the OA knee group being on average 9.3 years older, 9.4 kg heavier and having greater body mass index by 3.4 kg/m2.


222 Table 1 Description of participants Group

Age (years) Mean (SD)

Height (m) Mean (SD)

Mass (kg) Mean (SD)

BMI (kg/m2) Mean (SD)

Sex (M:F)

Asymptomatic OA knee

56.7 (8.6) 66.0 (7.6)

1.68 (0.09) 1.67 (0.09)

74.4 (10.9) 83.8 (12.9)

26.3 (2.7) 29.7 (3.8)

8:9 11:6

Group asymptomatic

Onset timing difference (ms)

OA knee 150

150

100

100

50

50

0

0

-50

-50

-100 N- 16 (a)

16 visual

16 16 algorithm

Onset determination method

-100

N- 17

16 visual

17 16 algorithm

Onset determination method

(b)

Fig. 1. Onset timing difference during initiation of concentric contraction (1a) and MVIC (1b). Negative value indicates VMO activated before VL.

Fig. 1 shows the results for OTD, onset of VMO EMG activity relative to that of VL, for the concentric contraction and for the MVIC, in graphic format. The median and interquartile range values are also presented in Table 2. For both contraction types, the differences between the groups in the OTD were not statistically significant for either visual or computerized evaluation methods (Mann–Whitney U-test, all p40.05). We noted greater variability in the OA knee patients than in the asymptomatic participants in both contraction types, as can be seen clearly in Fig. 1. The MVIC torque produced by the OA knee patients was on average 30% less than the asymptomatic participants (120.6 Nm compared to 84.6 Nm). This difference was statistically significant (Mann–Whitney U-test, po0.05). We were not able to report confidence intervals for the differences between the groups because the data breached the assumptions of normality and non-parametric tests were used (Bland, 2000).

4. Discussion Our aim was to investigate whether the onset of EMG activity of VMO was delayed relative to that of VL in

Table 2 Median (interquartile range) onset timing differences (ms) by contraction type and evaluation method Group

Asymptomatic OA knee p valuea a

Concentric

MVIC

Visual

Algorithm

Visual

Algorithm

2.7 (23.9) 7.9 (52.4) 0.41

2.5 (38.7) 3.6 (87.9) 0.50

4.6 (21.2) 1.5 (27.5) 0.06

13.6 (44.3) 5.3 (45.1) 0.07

Mann–Whitney U-test.

OA knee patients compared to similarly aged asymptomatic participants, during maximal voluntary isometric and concentric muscle contractions of the quadriceps. We did this by calculating the OTD by subtracting the EMG onset time of VL from VMO. We found that OTD was not significantly different between the OA knee patients and the asymptomatic participants in either activity. This agrees with the small amount of literature in this area on OA knee patients during other tasks (Hinman et al., 2002; Dixon et al., 2004). We also found that MVIC torque of OA knee patients was significantly less than the similarly aged asymptomatic group, as has been shown in previous studies (Tan et al.,


1995; Fisher and Pendergast, 1997; Cheing and HuiChan, 2001), and our torque data are comparable with these reports. We observed marked variability both within and between subjects (Fig. 1) in the VMO-VL OTD, which has been mentioned in some patellofemoral pain syndrome literature (Karst and Willett, 1995; Cowan et al., 2001) but not in great detail. We believe that this may be firstly related to the OKC activities studied here, which have been shown to elicit greater variability in onset times than closed chain type activities (Stensdotter et al., 2003). Secondly, this may also be because of agerelated changes that affect the quality of the EMG signal in older people. As the EMG signals of older people often have a lower amplitude and poorer signal to noise ratio than that of younger people, due to muscle atrophy and increased subcutaneous fat, onset determination may be more difficult. Possibly we could have used greater smoothing than a 50 Hz low pass filter, but excess smoothing can cause loss of information (Karst, 1998). Very few studies have evaluated EMG in older people, and it would be preferable if more work were carried out here. There is no universally accepted best method for computerized onset determination (Hodges and Bui, 1996; Staude and Wolf, 1999), but because these computerized methods have set standards or parameters, subjectivity is reduced. The algorithm used here is based on a well-used technique (Hodges and Bui, 1996). Some studies have used variations, such as having the threshold exceeded for a minimum of 25 ms (Hinman et al., 2002) rather than the point where the mean of the window exceeds threshold, as used here. Also, we used a first order low pass filter compared to some studies using higher order filters. However, we believe that this should produce very little difference in results, as at 50 Hz filtering the phase lag is approximately 3 ms. In addition the standard moving window technique that follows the low pass filtering averaged the data over 24 ms periods, and the VL onset times were subtracted from those of VMO, to produce a relative OTD. The lack of a significant difference between the groups could be attributable to several explanations. Firstly, the sample size ðn ¼ 17Þ could possibly result in an underpowered study. We did not carry out post hoc power calculations as it is known that any analysis finding no significant difference will be shown to have been underpowered when post hoc power calculations are carried out (Goodman and Berlin, 1994), and research must start somewhere (Bacchetti, 2002). However, we did carry out an estimation of how many participants would be required to adequately power a future study to detect differences between the groups in OTD. These post hoc calculations were carried out with a significance level of 0.05 using the larger of the group SDs from the visually evaluated results, but it should be noted that because the data breached the assumptions of normality,

223

these calculations were approximations only, and may be overestimates. Due to the variability observed, it was decided to use a value of 20 ms as the difference to be detected between the groups. The calculations showed that a study with 95 participants in each group would have 80% power to detect a difference of 20 ms in OTD for the concentric contraction. However, 142 participants in each group would be required to provide 80% power for detection of a 20 ms difference for MVIC. Further research with a larger sample size is therefore recommended, and this study provides information that will assist future studies. To avoid making a type II error from the results observed, future studies should use a much larger total sample size, and our estimations show that approximately 90–140 participants may be required per group for this type of OKC activity. The small sample size in this study is a limitation and therefore the results must be interpreted with caution. However, this was an exploratory study from which the results have enabled sample size calculations to be undertaken for subsequent studies. It is possible that the OA knee patients in this study could be a heterogeneous group, as the general OA knee population is understood to display considerable heterogeneity (Hurley, 1999). No attempt was made to categorize patients in terms of aspects of OA knee classification, or duration or level of symptoms. If this was the case in future studies, sub-group analysis could reveal whether specific sub-groups of the OA knee population do exhibit an impairment in OTD that could be clinically important. In addition, an assumption was made that the asymptomatic participants did not have radiographic OA. Although radiographs were not taken from the asymptomatic participants for ethical reasons, it is possible that they could have exhibited asymptomatic radiographic changes at the knee joint (Felson et al., 1987) which may have affected muscle activation. The OKC type muscle contraction tested here may not be representative of normal functional movement that tends to occur in the weight bearing closed kinetic chain manner (Callaghan and Oldham, 1996; Stensdotter et al., 2003). However, this is a standard method of evaluating quadriceps function. Future studies should attempt to use both open and closed kinetic chain testing. In addition, it was noticed that the MVIC activity often generated anticipatory muscle contraction in some subjects. As this anticipatory muscle activity generated force (measured by the load cell), these EMG onsets could not be validly excluded. Interestingly, using the computerized method, onset was often detected for VMO and not VL, or vice-versa, which could have added to the variability discussed above. Nevertheless, we believe that this exploratory study adds to the evidence base about neuromuscular impairments in OA knee, and will inform future research and practice. Much of our understanding of muscle


J. Dixon, T.E. Howe / Manual Therapy 12 (2007) 219–225

dysfunction in OA knee remains speculative (Hurley, 1999). This study and those already published indicate that the timing of VMO muscle activation relative to that of VL in OA knee patients does not differ from that of asymptomatic participants, in contrast to evidence in PFPS. However there are some published studies reporting that experimental knee effusion can preferentially affect VMO (Kennedy et al., 1982; Spencer et al., 1984; Torry et al., 2000). Therefore there is a now a need for the intensity of VMO and VL EMG activity to be investigated in OA knee patients, as it is possible that this is where any preferential inhibitory effect on VMO may be exhibited. The need for more research into muscle function in OA knee is clear. Subsequent to studies reporting muscle inhibition in OA knee patients (Hurley and Newham, 1993; Hurley et al., 1997), a study has found that not all OA knee patients exhibit muscle inhibition but some asymptomatic participants do (Lewek et al., 2004). Quadriceps strength and muscle inhibition do not show the expected correlation with severity of cartilage damage (Pap et al., 2004). Similarly, the findings that quadriceps strengthening may actually not benefit patients with malaligned knees (Sharma et al., 2003) shows that further research is urgently required. This seems particularly important with regard to the heterogeneity and variability within the OA knee population, if treatments are to be efficacious.

5. Conclusion Our acceptance of the null hypothesis agrees with the only other studies into the activation timing of VMO and VL in OA knee (Hinman et al., 2002; Dixon et al., 2004). This evidence of this exploratory study for the lack of a preferential delay in VMO activation in OA knee patients contrasts with findings of a preferential effect in PFPS. Rehabilitation programmes for OA knee patients should not therefore be aimed at altering the timing of VMO activation relative to VL, as a growing body of evidence seems to indicate that no temporal delay exists.

Acknowledgements The authors wish to thank Vicki Whittaker for statistical advice and assistance, and all the participants who took part. This work was supported by a University of Teesside School of Health and Social Care Ph.D. Studentship. References Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, Christy W, Cooke TD, Greenwald R, Hochberg M. Development of criteria for the classification and reporting of osteoarthritis.

Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis & Rheumatism 1986;29:1039–49. Bacchetti P. Peer review of statistics in medical research: the other problem. British Medical Journal 2002;324:1271–3. Badley EM, Tennant A. Impact of disablement due to rheumatic disorders in a British population: estimates of severity and prevalence from the Calderdale Rheumatic Disablement Survey. Annals of the Rheumatic Diseases 1993;52:6–13. Bland M. An introduction to medical statistics. Oxford: Oxford University Press; 2000. Callaghan MJ, McCarthy CJ, Oldham JA. Electromyographic fatigue characteristics of the quadriceps in patellofemoral pain syndrome. Manual Therapy 2001;6:27–33. Callaghan MJ, Oldham JA. The role of quadriceps exercise in the treatment of patellofemoral pain syndrome. Sports Medicine 1996;21:384–91. Cheing GL, Hui-Chan CW. The motor dysfunction of patients with knee osteoarthritis in a Chinese population. Arthritis and Rheumatism 2001;45:62–8. Cowan SM, Bennell KL, Hodges PW, Crossley KM, McConnell J. Delayed onset of electromyographic activity of vastus medialis obliquus relative to vastus lateralis in subjects with patellofemoral pain syndrome. Archives of Physical Medicine and Rehabilitation 2001;82:183–9. Cowan SM, Hodges PW, Bennell KL, Crossley KM. Altered vastii recruitment when people with patellofemoral pain syndrome complete a postural task. Archives of Physical Medicine and Rehabilitation 2002;83:989–95. Dixon J. Quadriceps activity in osteoarthritis of the knee: An electromyographic analysis. Ph.D. Thesis, University of Teesside, 2004. Dixon J, Howe TE, Kent JR, Whittaker VJ. VMO-VL reflex latency difference between quadriceps components in osteoarthritic and asymptomatic knees. Advances in Physiotherapy 2004;6: 166–72. Doxey GE, Eisenman P. The influence of patellofemoral pain on electromyographic activity during submaximal isometric contractions. Journal of Orthopaedic & Sports Physical Therapy 1987;9:211–6. Edwards RH, Young A, Hosking GP, Jones DA. Human skeletal muscle function: description of tests and normal values. Clinical Science and Molecular Medicine 1977;52:283–90. Felson DT, Naimark A, Anderson J, Kazis L, Castelli W, Meenan RF. The prevalence of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis & Rheumatism 1987;30:914–8. Fisher NM, Pendergast DR. Reduced muscle function in patients with osteoarthritis. Scandinavian Journal of Rehabilitation Medicine 1997;29:213–21. Goodman SN, Berlin JA. The use of predicted confidence intervals when planning experiments and the misuse of power when interpreting results. Annals of Internal Medicine 1994;121:200–6. Grabiner MD, Koh TJ, Draganich LF. Neuromechanics of the patellofemoral joint. Medicine and Science in Sports and Exercise 1994;26:10–21. Hinman RS, Bennell KL, Metcalf BR, Crossley KM. Temporal activity of vastus medialis obliquus and vastus lateralis in symptomatic knee osteoarthritis. American Journal of Physical Medicine and Rehabilitation 2002;81:684–90. Hodges PW, Bui BH. A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography. Electroencephalography & Clinical Neurophysiology 1996;101:511–9. Hurley MV. The role of muscle weakness in the pathogenesis of osteoarthritis. Rheumatic Diseases Clinics of North America 1999;25:283–98.

ARTICLE IN PRESS J. Dixon, T.E. Howe / Manual Therapy 12 (2007) 219–225 Hurley MV, Newham DJ. The influence of arthrogenous muscle inhibition on quadriceps rehabilitation of patients with early, unilateral osteoarthritic knees. British Journal of Rheumatology 1993;32:127–31. Hurley MV, Scott DL, Rees J, Newham DJ. Sensorimotor changes and functional performance in patients with knee osteoarthritis. Annals of the Rheumatic Diseases 1997;56:641–8. Kannus P, Jarvinen M, Latvala K. Knee strength evaluation: a standardized test protocol and scoring scale for isokinetic and isometric strength measurement of the knee joint for evaluation of long-term healing results of various knee problems and their treatment procedures. Scandinavian Journal of Sports Science 1987;9:9–13. Karst GM. EMG onset timing. Physical Therapy 1998;78:543–6. Karst GM, Willett GM. Onset timing of electromyographic activity in the vastus medialis oblique and vastus lateralis muscles in subjects with and without patellofemoral pain syndrome. Physical Therapy 1995;75:813–23. Kennedy JC, Alexander IJ, Hayes KC. Nerve supply of the human knee and its functional importance. American Journal of Sports Medicine 1982;10:329–35. Lewek MD, Rudolph KS, Snyder-Mackler L. Quadriceps femoris muscle weakness and activation failure in patients with symptomatic knee osteoarthritis. Journal of Orthopaedic Research 2004;22:110–5. Lieb FJ, Perry J. Quadriceps function. An anatomical and mechanical study using amputated limbs. Journal of Bone and Joint Surgery 1968;50-A:1535–48. Lieb FJ, Perry J. Quadriceps function: an electromyographic study under isometric conditions. Journal of Bone & Joint Surgery 1971;53-A:749–58. Mannion AF, Dolan P. Relationship between myoelectric and mechanical manifestations of fatigue in the quadriceps femoris muscle group. European Journal of Applied Physiology and Occupational Physiology 1996;74:411–9. Marks R, Percy JS, Semple J, Kumar S. Comparison between the surface electromyogram of the quadriceps surrounding the knees of healthy women and the knees of women with osteoarthrosis. Clinical & Experimental Rheumatology 1994;12:11–5. McNair PJ, Depledge J, Brettkelly M, Stanley SN. Verbal encouragement: effects of maximum effort voluntary muscle activation. British Journal of Sports Medicine 1996;30:243–5.

225

Pap G, Machner A, Awiszus F. Strength and voluntary activation of the quadriceps femoris muscle at different severities of osteoarthritic knee damage. Journal of Orthopaedic Research 2004;22: 96–103. Powers CM, Landel R, Perry J. Timing and intensity of vastus muscle activity during functional activities in subjects with and without patellofemoral pain. Physical Therapy 1996;76: 946–55. Sharma L, Dunlop D, Cahue S, Song J, Hayes KW. Quadriceps strength and osteoarthritis progression in malaligned and lax knees. Annals of Internal Medicine 2003;138:613–9. Shrier I. Muscle dysfunction versus wear and tear as a cause of exercise related osteoarthritis: an epidemiological update. British Journal of Sports Medicine 2004;38:526–35. Soderberg GL, Knutson LM. A guide for use and interpretation of kinesiologic electromyographic data. Physical Therapy 2000;80:485–98. Spencer JD, Hayes KC, Alexander IJ. Knee joint effusion and quadriceps reflex inhibition in man. Archives of Physical Medicine and Rehabilitation 1984;65:171–7. Staude G, Wolf W. Objective motor response onset detection in surface myoelectric signals. Medical Engineering & Physics 1999;21:449–67. Stensdotter A-K, Hodges PW, Mellor R, Sundelin G, Hager-Ross C. Quadriceps activation in closed and in open kinetic chain exercise. Medicine & Science in Sports & Exercise 2003;35:2043–7. Tan J, Balci N, Sepici V, Gener FA. Isokinetic and isometric strength in osteoarthrosis of the knee: a comparative study with healthy women. American Journal of Physical Medicine and Rehabilitation 1995;74:364–9. Tornvall G. Assessment of physical capabilities with special reference to the evaluation of maximal voluntary isometric muscle strength and maximal working capacity. Acta Physiologica Scandinavica 1963;58(Suppl 201):1–102. Torry MR, Decker MJ, Viola RW, O’Connor DD, Steadman JR. Intra-articular knee joint effusion induces quadriceps avoidance gait patterns. Clinical Biomechanics 2000;15:147–59. Witvrouw E, Sneyers C, Lysens R, Victor J, Bellemans J. Reflex response times of vastus medialis oblique and vastus lateralis in normal subjects and in subjects with patellofemoral pain syndrome. Journal of Orthopaedic and Sports Physical Therapy 1996;24: 160–5.

ARTICLE IN PRESS


Original article

The validity of clinical measures of patella position Islay McEwana, Lee Herringtonb,c,, Jeanette Thomd a

Department of Exercise and Sports Science, Manchester Metropolitan University, UK Directorate of Sport, Allerton Annexe, University of Salford, Manchester M6 6PU, UK c Centre for Rehabilitation and Human Performance Research, University of Salford, Manchester, UK d Department of Exercise and Sports Science, University of Bangor, UK b

Received 4 May 2005; received in revised form 20 March 2006; accepted 27 June 2006

Abstract Patellar taping is regarded as an important element of the treatment of patellofemoral joint pain. Key to the successful use of patellar taping is the assessment of patella position. The reliability and validity of the techniques used to assess patella position has been questioned. The aim of the study was to assess the validity of the clinical assessment technique of patella medio-lateral position and patella lateral tilt against the criterion measure of MRI. Twenty-four subjects eight females and 16 males had their patella position examined in the study (mean age 24.577.9 years, range 18–42 years). The study also assessed intra-tester reliability of the technique. A good correlation was found between the findings of the clinical test for medio-lateral position and the MRI measure (r ¼ 0:611, p ¼ 0:002). All of the subjects found to have a laterally tilted patella on clinical examination had a lateral patella tilt defined by PTA of greater than 51. Those subjects with a PTA of less than 51 on clinical examination were assessed as having no degree of patella tilt. The study undertaken shows that when undertaken by an experienced manual therapist positional assessment of the patella can have strong criterion validity and intra-tester reliability. r 2006 Elsevier Ltd. All rights reserved. Keywords: Assessment; Patella; Validity; Reliability

1. Introduction Within clinical practice, it has always been considered essential to accurately diagnose the presenting dysfunction and then treat accordingly. This is be particularly true when assessing patella position prior to the application of patellar taping, a common treatment for patellofemoral pain, as the anatomical findings will to a large extent dictate the directional control the tape will provide (McConnell, 1996). Corrective taping for patella maltracking was introduced nearly two decades ago by McConnell (1986), this was one of the first treatments to have claimed to Corresponding author. Directorate of Sport, Allerton Annexe, University of Salford, Manchester M6 6PU, UK. Tel.: +00 441612952326. E-mail address: [email protected] (L. Herrington).


achieve long term success for this often difficult to treat condition (Dye, 2005). The primary theory underpinning treatment is that corrective taping alters the abnormal patella orientation, therefore normalizing patella position and in so doing relieves the stress on the patellofemoral joint (PFJ) and surrounding tissues and thus reduces pain (Dye, 2005). Although many authors have demonstrated that corrective taping relieves pain, the mechanism by which this is brought about still remains unclear (Crossley et al., 2000). Key to the success of patella taping is the accurate assessment of patella position in order that tape may be applied to counter the abnormalities of position presented (McConnell, 1996). The method described originally by McConnell (1986) to assess patella position has been investigated by a number of authors. The intertester reliability of the assessment method has been regarded as poor (Fitzgerald and McClure, 1995;

ARTICLE IN PRESS I. McEwan et al. / Manual Therapy 12 (2007) 226–230

Watson et al., 1999) along with the intra-tester reliability (Tomisch et al., 1996). Powers et al. (1999) also questioned the validity of the method. In contrast, Herrington (2002) and Herrington and Nester (2004) found the method to assess medio-lateral position of the patella described by McConnell (1986) to be reliable when used by experienced clinicians. Herrington (2002) criticized the previous studies for using relatively novice clinicians with their minimal training being the significant factor in the test’s reliability not the test itself. The study of Herrington (2002) was in contrast to that of Powers et al. (1999), with Powers et al. (1999) reporting poor validity of the assessment methods against the criterion values produced from assessment of patella position by MRI, whilst Herrington (2002) showing a strong relationship between the clinical measure and MRI findings. The contradictory findings in the literature would appear to indicate that there is a need for further study in this area. The aim of this study is firstly: to assess the validity of the clinical measures of patella position against the criterion measure of MRI, here unlike previous studies both patella medio-lateral position and lateral tilt will be assessed. Secondly the intra-tester reliability of the clinical measurements will be examined. If these assessment techniques are found to be valid and reliable then decision-making regarding the choice of taping technique will have improved efficacy. If the techniques prove to be unreliable and not valid then this will question the significance of any direction specific effects for patella taping as the direction of the effects cannot be defined clinically. 2. Method 2.1. Subjects Twenty-four subjects, eight females and 16 males were examined in the study, (mean age 24.577.9 years, range

227

18–42 years). Subjects were excluded if they had any current or previous history of knee or lower limb injury. The subjects were recruited as a sample of convenience from a University Exercise and Sports Science Department and were all physically active and regular sports participants, typical of a population from which patellofemoral pain patients would come (Taunton et al., 2002). The tests were performed in agreement with the declaration of Helsinki and all subjects gave informed written consent to participate prior to commencing study. The study was approved by the institutional research ethics committee.

2.2. Clinical measurement of medio-lateral patella position The method used was that described by McConnell (1986) and Herrington and Nester (2004). The following landmarks were located: medial and lateral epicondyles of the femur and mid point of the patella on the subject’s right knee, which was positioned and supported in 201 of knee flexion (in order to place the patella within the trochlea groove). The examiner then marked the distance from the ascribed position of the medial and then the lateral epicondyle of the femur to the mid-point of the patella on a piece of folded zinc oxide tape. The method is shown graphically in Fig. 1. The medial and lateral measurements were each repeated three times on separate pieces of tape, with re-palpation of the landmarks on each occasion. An independent assessor then measured the distance of medial and lateral points to mid point with the average of the three measurements being recorded. These measurements of the tape were then repeated one day later. The level of agreement between days was found by ICC to be r ¼ 0:92 (po0:05).

Fig. 1. Clinical technique for assessing patella position. The femoral condyles are palpated and marked on the tape along with the mid point of the patella. The distances between the condyles and the middle of the patella can be subtracted from each other to give relative medio-lateral position.


I. McEwan et al. / Manual Therapy 12 (2007) 226–230

2.3. Clinical measurement of lateral patellar tilt The method used was that described by McConnell (1986) and Fitzgerald and McClure (1995). The degree of tilt was determined by comparing the height of the medial and lateral borders of the patella. The examiner places their thumb and index finger on the medial and lateral borders of the patella. The patella would be laterally tilted if the medial border is more anterior than the lateral and vice versa for medial tilt. 2.4. Intra-tester reliability of clinical measurements At random intervals during the testing period (inbetween the other subjects) the examiner repeatedly examined the patella position of a single subject (10 times in total). These measurements were then examined to ascribe intra-tester reliability of the examiner. The physiotherapist carrying out the testing was blind to the results of each test. The physiotherapist carrying out all testing was a clinical specialist with specific manual therapy qualifications and 15 years experience of musculoskeletal medicine. 2.5. Measurement of patella position following MRI Transverse plane MRI images were performed on the participant’s right knee using an extremity coil fixed 0.2T MRI scanner (E-Scan; ESAOTE Biomedica, Genova, Italy). The knee was fixed at 20 degrees of knee flexion using a wedge whilst the participant was supine. Scans were performed using a T1 weighted 3D isotropic profile with the following scanning parameters: time to echo: 16 ms; repetition time: 38 ms; field of view: 180 mm 180 mm; matrix: 256 192. Images were sliced at 5 mm intervals (with a 10% gap) and the image which included the marker was used for analysis. The scans were digitized on a Macintosh G4 computer and analysed off-line with ImageJ image analysis software (National Institutes of Health, USA). Medio-lateral component of patella orientation was measured by finding the lateral patellar displacement (LPD) on the MRI film, using the method described by Larsen et al. (1995). The LPD quantifies the position of the patella in the frontal plane relative to the medial femoral condyle. Fig. 2 demonstrates the measurement of LPD. A single examiner was used to measure the position, who was blind to the findings of the clinical examination. The LPD was measured three times with the average being taken and compared to a repeated measurement one day later, with the level of agreement between days found by ICC to be r ¼ 0:99 (po0:05). Lateral tilt component of patella orientation was measured by finding the patella tilt angle (PTA) on the MRI film using the method described by Guzzanti et al. (1994). Fig. 3 demonstrates the measurement of PTA. A

Fig. 2. Schematic of the method used to assess medial/lateral position of the patella from MRI. Position is determined by drawing lines connecting the anterior portions of the medial and lateral femoral condyles (AB), and the maximum width of the patella (CD). A line was then drawn from the anterior portion of the lateral condyle so that it was perpendicular to line AB and bisected line CD. The distance from point E to point C represented the amount of medial or lateral orientation.

single examiner was used to measure the position, who was blind to the findings of the clinical examination. This was measured three times with the average being taken and compared to a repeated measurement one day later. The level of agreement was found by ICC to be r ¼ 0:94 (po0:05) between days.

2.6. Statistical analysis The degree of agreement between the two techniques (MRI and clinical) was quantified by calculation of Pearsons product moment. The intra-tester reliability of the clinical measurement was quantified by means of the intra-class correlation coefficient. For purposes of this study, correlation coefficients were interpreted as follows: below 0.50 was poor, 0.50–0.75 was good, and above 0.75 was excellent (Portney and Watkins, 1993). Statistical analysis was undertaken on the statistical analysis package SPSS (version 12).

ARTICLE IN PRESS I. McEwan et al. / Manual Therapy 12 (2007) 226–230

229

Table 1 Intratester reliability of the clinical measure of mediolateral patella position

Mean Standard Deviation Standard error of measurement Confidence interval (95%) ICC3, k

Lateral Distance (mm)

Medial distance (mm)

Difference (mm)

8.3 0.1

8.9 0.1

0.6 0.1

0.1

0.1

0.1

8.1–8.5

8.7–9.1

0.4–0.8

0.86

0.82

0.9

Statistical Signifcant (po0:01).

Table 2 Results from the examination of patella position by MRI and clinical method

Fig. 3. Calculation of the patella tilt angle on MRI scan (from Guzzanti et al., 1994). The PTA is identified by the intersection of a line parallel to the lateral facet of the patella (y) and the base line (x) which runs from the deepest point of the patella to the edge of the lateral condyle.

Mean Standard Deviation Range

LPD (mm)

PTA (1)

Clinical mediolateral test (difference, mm)

8.1 2.8

6.8 3.9

5.0 3.3

4.6–13.8

4–12

0–13

LPD Lateral patella displacement, PTA Patella tilt angle. Minus value equals medial patella tilt.

3. Results

3.4. Between method agreement of patella position

3.1. Intra-tester reliability

The Pearson’s product moment revealed a good correlation between the findings of the clinical test for medio-lateral position and the MRI measure for LPD (r ¼ 0:611, p ¼ 0:002). All of the subjects found to have a laterally tilted patella on clinical examination had a lateral patella tilt defined by PTA of greater than 51. Those subjects with a PTA of less than 51 on clinical examination were assessed as having no degree of patella tilt.

Table 1 shows the results of the intra-tester reliability study for the clinical measure of medio-lateral patella position.

3.2. Patella position from MRI film Table 2 shows the results from the MRI examination of patella position. All 24 subjects demonstrated some degree of lateral patella displacement as assessed by LPD. 22 of the 24 subjects had laterally tilted patellae assessed by PTA, with 2 subjects having a medially tilted patella.

3.3. Clinical assessment of patella position All subjects except one were found to have laterally displaced patellae, the one exception was found to have a centralized patella. Mean lateral patella displacement was 573 mm (range 0–13 mm). A total of 17 subjects were found to have laterally tilted patella.

4. Discussion The results of the study would appear to indicate that the clinical method used to assess patella position described shows a good and significant relationship to patella positions measured from MRI. Powers et al. (1999) found the assessment technique to have poor agreement with measures from MRI and to overestimate the degree of lateral displacement, these findings were in contrast to this study and those of Herrington (2002) with strong agreement being shown and no over-estimation of displacement. In the Powers et al. (1999) study the examiner had less than one years experience of using this technique, the length of time


I. McEwan et al. / Manual Therapy 12 (2007) 226–230

since qualification and the nature of any post-graduate training was not reported, this was in contrast to the examiner in this study, who had over 15 years clinical experience and specialist manual therapy training. The findings of the study support those of Herrington (2002) which contended that the clinical measurement of medio-lateral patella position used showed strong criterion validity. Furthermore, this study also demonstrated that the clinical measure of patella tilt outlined also showed strong criterion validity. Criterion validity establishes the ability of one test to predict the results obtained by another (George et al., 2000). This analysis is particularly useful in this case were a practical, quick and cheap test is proposed as an alternative to a time consuming and cost prohibitive procedure (MRI). Previous studies have shown that reliability when using these techniques varies considerably with training, experience and skill (Herrington, 2002; Watson et al., 1999). The clinician undertaking the measurements was an experienced manual therapist, previously reported research (Herrington, 2003) has shown manual therapy training to have a significant effect on reliability when undertaking these tests and the results of this study and its external validity must be viewed in light of such information. Alongside the external validity issue of using an experienced manual therapist to carry out the tests, the measurements could have been open to recall bias effecting the reliability findings, the study aimed to minimize this with blinding of the assessor, both to the clinical findings and those of the MRI. The study undertaken demonstrated that the majority of asymptomatic physically active individuals examined had laterally displaced and tilted patellae, which have previously been regarded as pathological (McConnell, 1986). Further study would be required in a pathological population to assess the extent of these findings within this patient group. It may prove that these tests are both reliable and valid, but lack sensitivity in detecting those individuals with PFJ pain and so have little ecological validity (George et al., 2000), this requires further investigation. Alternatively, it may be that the lateral deviation of the patella found in these normal individuals relates to a sub-optimal performance of the PFJ and would lead to potential pathology with the appropriate alteration of environmental circumstances (Dye, 2005), if this is the case then the tests may prove an appropriate screening test for PFJ pain, this again requires further investigation.

5. Conclusion Previous research has questioned the validity of the methods of assessing patella position described by McConnell (1986), this study would appear to refute those findings. The study undertaken shows that when undertaken by an experienced manual therapist positional assessment of the patella can have strong criterion validity and intra-tester reliability. References Crossley K, Cowan S, Bennell J, McConnell J. Patellar taping: is clinical success supported by scientific evidence. Manual Therapy 2000;5(3):142–50. Dye S. The pathophysiology of patellofemoral pain. Clinical Orthopaedic and Related Research 2005;436:100–10. Fitzgerald K, McClure P. Reliability of measurements obtained with four tests for patellofemoral alignment. Physical Therapy 1995;75:84–92. George K, Batterham A, Sullivan I. Validity in clinical research: a review of basic concepts and definitions. Physical Therapy in Sport 2000;1:19–27. Guzzanti V, Gigante A, Di Laazzaro A, Fabbriciani A. Patellofemoral malalignment in adolescents. American Journal of Sports Medicine 1994;22:55–60. Herrington L. The inter-tester reliability of a clinical measurement used to determine the medial lateral orientation of the patella. Manual Therapy 2002;7:163–7. Herrington L. The reliability of a clinical measurement used to determine the medial/lateral orientation of the patella World Confederation of Physical Therapy 14th International Conference, Barcelona, June 2003. Herrington L, Nester C. Q-angle undervalued? The relationship between Q angle and medio-lateral position of the patella. Clinical Biomechanics 2004;19:1070–3. Larsen B, Andersen E, Urter A, Mickelson M, Newhouse K. Patellar taping: a radiological examination of medial glide taping. American Journal of Sports Medicine 1995;23:465–71. McConnell J. The management of Chondromalacia Patellae: a long term solution. Australian Journal of Physiotherapy 1986;32:215–22. McConnell J. Management of patellofemoral problems. Manual Therapy 1996;1:60–6. Portney L, Watkins M. Foundations of clinical research: applications to practice. Norwalk, Connecticut: Appleton and Lange; 1993. Powers C, Mortenson S, Nishimoto D, Simon D. Criterion-related validity of a clinical measurement to determine the medial/lateral component of patellar orientation. Journal of Orthopaedic and Sports Physical Therapy 1999;29:372–7. Taunton J, Ryan M, Clement D, McKenzie D, Lloyd-Smith D, Zumbo B. A retrospective case-control analysis of 2002 running injuries. British Journal of Sports Medicine 2002;36:95–101. Tomisch D, Nitz A, Threlkeld J, Shapiro R. Patellofemoral alignment: reliability. Journal of Orthopaedic and Sports Physical Therapy 1996;23:200–8. Watson C, Propps M, Galt W, Redding A, Dobbs D. Reliability of McConnell’s classification of patellar orientation in symptomatic and asymptomatic subjects. Journal of Orthopaedic and Sports Physical Therapy 1999;29:378–85.

ARTICLE IN PRESS


Original article

Strain on the repaired supraspinatus tendon during manual traction and translational glide mobilization on the glenohumeral joint: A cadaveric biomechanics study Takayuki Murakia,, Mitsuhiro Aokib, Eiichi Uchiyamac, Tomoya Miyasakaa, Gen Murakamic, Shigenori Miyamotob a Graduate School of Health Sciences, Sapporo Medical University, Sapporo, Japan Department of Physical Therapy, Sapporo Medical University School of Health Sciences, Sapporo, Japan c Department of Anatomy, Sapporo Medical University School of Medicine, Sapporo, Japan

b

Received 21 July 2005; received in revised form 4 May 2006; accepted 27 June 2006

Abstract There has been no report on the mechanical effects of joint mobilization on rotator cuffs. The purpose of this study was to determine whether it is safe to use grade 3 joint mobilization techniques after rotator cuff repair. Nine fresh frozen cadaveric shoulders were used in this study. The strains on the artificially repaired supraspinatus tendon during joint mobilization were measured at 01 and 301 of shoulder abduction and were compared with those at the maximal stretching position and relaxing position. Additionally, gap distances were measured during this experiment. The strain at 301 of abduction of the repaired tendon during each joint mobilization was significantly smaller than that at 01 abduction (Po0:05). At 301 of abduction, the strain during joint mobilization was not statistically different from that of the shoulder in the relaxing position, except during the inferior glide technique. Gap distances were 0 mm at 301, while the distances were 1.06–1.46 mm at 01. Our findings suggest that joint mobilization techniques, except inferior glide, can be performed safely without significantly straining the repaired tendon at 301 of abduction, if rotator cuff repair is performed at 01 of abduction. r 2006 Elsevier Ltd. All rights reserved. Keywords: Strain; Supraspinatus tendon; Rotator cuff tear; Fresh cadaver

1. Introduction Shoulder joint contracture can occur after repair of rotator cuff tears. This can be caused by capsular contracture, tendon shortening, scars, and adhesions of the subacromial and the subscapularis-conjoined tendon regions (Warner and Greis, 1998; Hatakeyama et al., 2001a; Matsen et al., 2004). Therefore, intensive physical therapy using range of motion exercise (ROM exercise) (Cofield, 1985; Matsen et al., 2004) and joint mobilizaCorresponding author. Tel.: +81 11 611 2111; fax: +81 11 611 2150. E-mail address: [email protected] (T. Muraki).


tion (Bruzga and Speer, 1999; Mangus et al., 2002) immediately after surgery is important. Generally, in order to prevent and treat joint contracture, limited movement of the joint itself is used in ROM exercise. However, when this movement causes pain, effective stretching of the connective tissue that limits motion becomes difficult (Quillen et al., 1992). For example, when contracture of the posterior capsule occurs in the glenohumeral joint, antero-superior translation of the humeral head occurs during arm abduction and, this translation consequently leads to subacromial impingement (Harryman et al., 1990; Warner et al., 1990). In addition, contracture of the anterior capsule (Flatow et al., 1994) and the inferior


T. Muraki et al. / Manual Therapy 12 (2007) 231–239

capsule (Cofield, 1985; Hjelm et al., 1996) often lead to impingement. Therefore, in a shoulder joint with a capsular contracture, this exercise can actually worsen any injuries of the joint capsule and rotator cuff. On the other hand, joint mobilization, such as traction and glide, is used to stretch the tendon, ligament, and capsule and to improve the physiological accessory movement. Traction is the technique that distracts one articular surface perpendicular to the other, and glide techniques translationally glide one articular surface parallel to the other (Kaltenborn, 1999). These techniques are considered capable of stretching the particular connective tissues that limit joint motion without impingement, resulting in an improvement of the limited ROM and reduction in pain (Johns and Wright, 1962; Quillen et al., 1992). The effects of joint mobilization have been previously reported. Hsu et al. (2000a, b, 2002a) demonstrated that anterior–posterior and inferior translational gliding improved the range of abduction and external rotation in cadaveric glenohumeral joints. Conroy and Hayes (1998) investigated the effects of joint mobilization and compared them to general physical therapy on a patient with subacromial impingement syndrome. They then reported that joint mobilization was not effective on ROM and any functional outcome of the shoulder joint, but it provided effective pain relief. However, care should be taken when joint mobilization is used on the shoulder joint after surgery. In order to prevent and treat joint contracture after rotator cuff repair, knowledge of the mechanical stresses of joint mobilization on the repaired rotator cuff is required. Hatakeyama et al. (2001a) and Zuckerman et al. (1991) observed the effects of shoulder positions on the rotator cuff under several different conditions. Although the effects and safety of joint mobilization were studied from the physiological aspect of the shoulder (Hsu et al., 2002b; Gokeler et al., 2003), the mechanical effect of joint mobilization on the rotator cuff was unclear because connective tissues of the shoulder that had different mechanical properties were not observed individually in these studies. Therefore, the mechanical effects of joint mobilization on intact and repaired rotator cuffs should be clarified first in order to perform joint mobilization without excessive stretching stress on the rotator cuff. The purpose of this study was to measure the strain on both intact and repaired supraspinatus tendons, which is the primary site of rotator cuff tears, and the gap distance of the repair site during joint mobilization by using fresh frozen cadaveric shoulders. Furthermore, based on these results, we discuss about safely applying joint mobilization after rotator cuff repair.

2. Methods 2.1. Preparation of specimens Nine frozen shoulders (four left shoulders and five right shoulders) harvested from nine fresh cadavers were used in the experiment. The mean age at death was 80 years (71–91 years of age). Any shoulder with macroscopic evidence of rotator cuff tears or osteoarthritis was excluded. However, none of these shoulders conformed to these criteria. The shoulders, disarticulated from their thoraxes, were stored in a freezer at 20 1C. The thawing of the specimens at room temperature started 12 h prior to experimentation. Then, soft tissues—except the rotator cuff muscles, biceps brachii muscles, coracoacromial ligaments, and capsules—were carefully removed to avoid the loss of intra-articular negative pressure in the glenohumeral joint. The distal third of the humerus was exposed, and an acrylic stick was inserted perpendicular to the shaft indicating the direction of the forearm. Next, the humerus was amputated above the elbow. During the experiment, the specimens were kept moist by spraying saline solution on them every 5–10 min. The room temperature was maintained at 22 1C. 2.2. Testing apparatus A wooden jig, consisting of a wooden board and a square timber, was used for this experiment. The scapula of the specimen was fixed on the wooden jig so that the medial border of the scapula was perpendicular to the ground (Culham and Peat, 1993) (Fig. 1). Two anchors (Fastin RC threaded suture anchor, Mitek, Tokyo, Japan) were inserted into the bony insertion of the infraspinatus and subscapularis tendons to apply a compressive force of 11 N to each thread (total 22 N) against the glenoid fossa. In previous cadaveric studies, this compressive force was used as the minimum force required, preventing subluxation of the humeral head on application of translational loads (Warner et al., 1992; Tibone et al., 1998). Using this system, the humeral head was maintained in concentric position onto the glenoid fossa after joint mobilization was performed. 2.3. Measurement device The strain data on the supraspinatus tendon was obtained from a precise displacement sensor (Pulse Coder, LEVEX, Kyoto, Japan) (Fig. 2a). This Pulse Coder consisted of a coil sensor and a brass pipe. The displacement was measured by detecting the position of the brass pipe relative to the coil sensor that generated the magnetic field. Analogue data of the displacement was represented and recorded on a digital scaling meter (HV35, Allied Control, Tokyo, Japan)

ARTICLE IN PRESS T. Muraki et al. / Manual Therapy 12 (2007) 231–239

233

3SPACE sensors

Pulse corder

Fig. 1. Schematic illustration of the experimental setup. A scapula of a specimen was mounted vertically on the wooden jig. A Pulse Coder was attached to the supraspinatus tendon. Sensors of the 3SPACE device were attached to the acromion and the humerus.

Fig. 2. Photographs of the Pulse Coder. (a) The precise displacement sensor ‘‘Pulse Coder’’ used to measure the strain on tendons by changes in the length between points; (b) Pulse Coder attached to the supraspinatus tendon with sutures.

that converted the obtained data into a digital form. The non-linearity of this sensor was 0.25%/full scale, and the range of measurement was 14 mm according to the manufacturer’s instructions. The sensors with fishhook-like barbed points were attached to the greater

tuberosity and the proximal part of the supraspinatus tendon. The sensor was placed parallel to the tendon fibre (Fig. 2b). Changes in the length between the points of the coil sensor and the brass pipe allowed the sensor to measure the strain on the supraspinatus



tendon. By using a digital caliper, the accuracy of preliminary calibration of the Pulse Coder, which was attached in the supraspinatus tendon, was 0.1 mm root mean square (RMS). A similar type device, which uses the same principle, was previously used to measure strain on the supraspinatus tendon (Hatakeyama et al., 2001a, b). A six-degree-of-freedom electromagnetic tracking device (3SPACE FASTRACK, Polhemus, Colchester, Vermont) was used to monitor the precise glenohumeral angles during this measurement. This device enabled measurement of the three-dimensional position and orientation of the sensors relative to the absolute coordination generated by the source (An et al., 1988). One sensor was placed on the acromion and the other was placed on the middle portion of the humerus (Fig. 1). In this system, the angle of arm abduction and extension was defined as the angle between the plane of the glenoid fossa and the longitudinal axis of the humerus. The rotation angle was defined as the rotation of the humerus along the longitudinal axis. With a 750-mm range of measurement from the source, the positional accuracy was 0.8 mm RMS, and the angular accuracy was 0.51 RMS. The 3SPACE device kinematically monitored the glenohumeral angles during the strain measurement.

2.4. Experimental procedure The measurements were first performed on intact supraspinatus tendons to assess the difference of tendon conditions and then on the repaired tendon model. In the case of the model, the supraspinatus tendon was excised (width 2.0 cm, length 1.5 cm) (Hatakeyama et al., 2001a, b)(Fig. 3a) from the greater tuberosity, which simulated the retracted supraspinatus tendon, and was repaired with #2 polyester threads (Ethibond, Ethicon Inc. Somerville, NJ) passed through drill holes in the greater tuberosity. The threads were pulled out with a 3kg force and clamped at the outlet from the bone with the arm in 301 of external rotation at 01 of arm abduction (Hatakeyama et al., 2001a, b) (Fig. 3b). Acromioplasty and removal of the coracoacromial ligament was also performed because this technique is often used in rotator cuff repair and enabled the Pulse Coder to move in the subacromial space. All these techniques were performed by an orthopaedic surgeon who was familiar with shoulder surgery. The neutral position was defined as 301 of external rotation at 01 of abduction in the glenohumeral joint, because the scapula internally tilts 301 relative to the coronal plane in vivo (Culham and Peat, 1993; Itoi et al., 2004). For reliable comparison, the maximal stretching

Fig. 3. Photographs of the supraspinatus tendon repair. (a) Artificial supraspinatus tear (width 2.0 cm, length 1.5 cm). Small arrows indicate the site of supraspinatus tendon tear; (b) Repair of torn supraspinatus tendon. The threads were clamped with a 3-kg force at the outlet from the bone (a large arrow). Small arrows indicate the repair site.


position (adduction at extension) (Evjenth and Hamberg, 1984) and relaxing position of the supraspinatus tendon (neutral position at 301 of abduction) (Zuckerman et al., 1991; Hatakeyama et al., 2001a) were used as reference positions producing the highest and lowest tendon strain. The shoulder positions examined in this mobilization experiment were at 01 and 301 of arm abduction in the scapular plane. In order to reproduce shoulder positions during the measurement, the positions were monitored by the 3SPACE FASTRACK device. Four joint mobilization techniques, i.e. traction, inferior, anterior and posterior glide, were performed at each position by a physical therapist who had more than 5 years of experience in correcting shoulder disorders by using joint mobilization. The therapist had no information regarding the hypotheses of this study, and the strain data represented on the digital scaling meter was masked. The grade of joint mobilization was set at 3 as defined by Kaltenborn (1999). Grade 3 is used to stretch connective tissues by applying force to the final stop during joint mobilization. At these positions, the strains and gap distances on the supraspinatus tendon were measured and compared with those during joint mobilization techniques. The measurements were performed 3 times during each joint mobilization technique in each of the two positions. The holding time during each measurement was set at 20 s. In a clinical setting, joint mobilization for 20–60 s was used to stretch the connective tissue (Quillen et al., 1992; Conroy and Hayes, 1998; Mangus et al., 2002). Hsu et al. (2000a, b, 2002a, b) reported that the ROM improved after joint mobilization was done for 10–30 s. Therefore, we decided that the holding time should be set at 20 s as the minimal time that is effective in biomechanical study and clinical practice.

235

Gap distance was regarded as the distance between the proximal edge of the torn supraspinatus tendon and its distal edge. Therefore, we defined the detected displacement by the sensor, which was placed across the repair site as gap distance (DL), as Pruitt et al. (1991) defined it in a previous study regarding flexor tendon repair. In the neutral position, the gap distance was 0 mm with a 3-kg tensile force on the repair site. 2.6. Statistical analysis Intra-class correlation coefficients were calculated to determine test–retest reliability in each condition. A two-way repeated measures analysis of variance was used to determine the effects of supraspinatus tendon conditions (intact tendon and repaired tendon), shoulder positions (01 and 301 of shoulder abduction), and joint mobilization techniques (traction and inferior/ anterior/posterior glide). To detect the differences among the joint mobilization techniques, Bonferroni’s multiple comparison procedure was used. Moreover, Dunnett’s multiple comparison test was performed to compare the strain during joint mobilization techniques occurring in both the relaxing and stretching positions. The a level was set at 0.05. All statistical analyses were performed on SPSS for Windows ver.11.5J. (SPSS Japan Inc., Tokyo, Japan).

3. Results 3.1. Reliability of these measurements Intra-class correlation coefficients of these measurements in each condition ranged from 0.809 to 0.984. These values corresponded to almost perfect (Landis and Koch, 1977).

2.5. Data analysis

3.2. Strain on the intact supraspinatus tendon

First, the length between the barbed points of the coil sensor and the brass pipe on the sensors at the neutral position was recorded. Next, the longitudinal displacement of the sensors at the measurement area of the tendon was recorded when the measurement positions were held for 20 s. The displacement was defined as the length change from the neutral position. The strain on the tendon was calculated by the following formula:

The strains on the intact supraspinatus tendon are shown in Fig. 4. The strains during each joint mobilization technique at 301 of arm abduction were significantly smaller than those occurring at 01 (Po0:005). There were no significant differences among joint mobilization techniques both at 01 and 301 of arm abduction. At 01 of arm abduction, the strains during each joint mobilization technique did not show significant differences from those occurring in the stretching position, and were significantly larger than those occurring in the relaxing position (Po0:001). At 301 of arm abduction, the strains during each joint mobilization technique were significantly smaller than those occurring in the stretching position (Po0:05), while these strains did not show significant differences from those in the relaxing position.

Strainð%Þ ¼ DLðmmÞ=LðmmÞ 100. L indicates the length between the points at the neutral position, and DL indicates the displacement from L. A positive strain value indicates that the supraspinatus tendon was stretched from the neutral position, and a negative strain value indicates a slackening of the tendon.


236

10

5 Trac

Strain (%)

0

Inf Ant

-5 Post Add.Ext

-10

30 Abd -15

-20 Abduction 0°

Abduction 30°

Reference positions

Fig. 4. Strain on intact supraspinatus tendon. The values and bars represent mean strain and standard deviation, respectively. Direction of the glenohumeral mobilization and reference positions: Trac, traction; Inf, inferior glide; Ant, anterior glide; Post, posterior glide; Add. Ext, adduction at extension; 30 Abd, 301 of arm abduction with neutral rotation.

80

60 Trac

Strain (%)

40

Inf Ant

20 Post Add.Ext

0

30 Abd -20

-40 Abduction 0°

Abduction 30°

Reference positions

Fig. 5. Strain on repaired supraspinatus tendon. The values and bars represent mean strain and standard deviation, respectively. Direction of the glenohumeral mobilization and reference positions: Trac, traction; Inf, inferior glide; Ant, anterior glide; Post, posterior glide; Add. Ext, adduction at extension; 30 Abd, 301 of arm abduction with neutral rotation.

3.3. Strain on the repaired supraspinatus tendon The strains on the repaired supraspinatus tendon are shown in Fig. 5. While the strain on the repaired supraspinatus tendon during joint mobilizations was significantly larger than that on the intact tendon at 01 of abduction (Po0:05), the strain on the repaired tendon during joint mobilizations was not different from that occurring on the intact tendon at 301 of

abduction (Figs. 4 and 5). The strains during each joint mobilization technique at 301 of arm abduction were significantly smaller than those occurring at 01 (Po0:005). Among joint mobilization techniques, there were no significant differences in the strains at both 01 and 301 of arm abduction. At 01 of arm abduction, the strains during each joint mobilization technique were significantly smaller than those occurring in the stretching position (Po0:001),

ARTICLE IN PRESS T. Muraki et al. / Manual Therapy 12 (2007) 231–239 Table 1 Gap distance on the repaired supraspinatus tendon during joint mobilization, stretching position, and relaxing positiona Gap distance (mm)

Traction Inferior glide Anterior glide Posterior glide Stretching position Relaxing position a

01 of abduction

301 of abduction

1.371.2 1.571.3 1.371.3 1.171.0 4.574.0 0

0 0 0 0

The values represent mean gap distance and standard deviation.

while these strains were significantly larger than those occurring in the relaxing position (Po0:001). At 301 of arm abduction, although the strains during each joint mobilization technique were significantly smaller than those occurring in the stretching position, only inferior glide showed a significantly larger strain than that occurring in the relaxing position (Po0:01). There were no significant differences in the strains between other techniques and the relaxing position. 3.4. Gap distances The gap distances of the repaired site during joint mobilization techniques at each abduction angle are listed in Table 1. Gaps were formed in the stretching position (4.47 mm) and during all joint mobilization techniques at 01 of abduction (1.06–1.46 mm). Conversely, no gap was observed in the relaxing position and all joint mobilization techniques at 301 of abduction.

4. Discussion Application of joint mobilization techniques immediately after rotator cuff repair has not been clarified. Quillen et al. (1992) considered the stress on immature repaired tissues as a contraindication of joint mobilization. This indicated that joint mobilization should be prohibited immediately after surgery if it stresses such tissues. Bruzga and Speer (1999) recommended joint mobilization techniques, which are performed at grade 1 or 2, after surgery if these would be useful in reducing pain and promoting joint nutrition. However, they also stated that joint mobilization at grade 3 or 4 should be used only after the healing of repaired tissue. Based on these opinions, joint mobilization techniques, which exert stress on the repaired tissue, should be avoided immediately after rotator cuff repair. On the other hand, based on the concept of joint mobilization, it is hypothesized that this technique can stretch specific

237

tissues and can reduce the stress on tissues that should not be stretched (Conroy and Hayes, 1998). Accordingly, while stress on the repaired supraspinatus tendon can be avoided, a particular part of the capsule, which is responsible for joint contracture, may be stretched by joint mobilization techniques. In order to resolve this issue, the stress on particular tissues such as the supraspinatus tendon by joint mobilization should be quantitatively determined. In this study, the strain and gap distance on the repaired supraspinatus tendons were measured to estimate the stress undergone by these during joint mobilization. For ideal repair, minimization of gap formation at the repair site and maintenance of mechanical stability of the repaired tendon until solid healing occurs are important (Gerber et al., 1994). Therefore, the criteria for safely performing joint mobilization are that (1) no gap is formed, and (2) the repaired tendon relaxes more than in the neutral position in which the repair was performed. A gap was formed in the stretching position (4.5 mm) and in all joint mobilization techniques at 01 abduction (1.1–1.5 mm) in this study. In comparison with a 10-mm gap distance, which is regarded as complete gap formation (Burkhart et al., 1997a, b), joint mobilization techniques at 01 abduction corresponded to 11–15%, while the stretching position was 45%. These techniques have the risk of increasing gap formation and can lead to the failure of healing (Burkhart et al., 1998; Gerber et al., 1994, 1999). Therefore, joint mobilization at 01 abduction should be avoided immediately after tendon repair. On the other hand, during joint mobilization at 301 abduction, strains were negative compared to neutral position and no gaps were observed. The supraspinatus tendon relaxes after abducting the arm beyond 301. By using three-dimensional analysis using magnetic resonance imaging, Nakajima et al. (2004) observed that intact supraspinatus tendons were relaxed beyond 301 arm abduction. Zuckerman et al. (1991) demonstrated that the strain on repaired tendons, with both small and large tears, remained small above 301 abduction irrespective of the position of flexion/extension or rotation. Hatakeyama et al. (2001a) determined strain on repaired supraspinatus tendons under the same repair condition as ours. They concluded that arm abductions above 301 in the scapular plane seemed to be safe even immediately after the repair, because the strain decreased in these positions and the estimated tensile forces were less 0.5 kg. According to the estimation from their data, tensile forces caused by joint mobilization at 301 were 0–0.5 kg, compared to 3 kg in the neutral position. Therefore, joint mobilization techniques can be safely performed keeping the tendon relaxed and not forming a gap at the repair site if the torn tendon is repaired at 01.



However, repetitive joint mobilization at 301 abduction might lead to gap formation on the repair site. The repaired tendon has less endurance to mechanical stress than the intact tendon because these two have different mechanical properties. The failure load of the repaired tendon was reported to be 72–605 N (Rossouw et al., 1997; Hatakeyama et al., 2001a; Ma et al., 2004), while that of the intact tendon was 600– 800 N (Itoi et al., 1995). In addition, cyclic loading might lead to gap formation even if the mechanical properties of the repair site are strong, or the cyclic stress is small (Rossouw et al., 1997; Petit et al., 2003). Hatakeyama et al. (2001a) concluded that internal rotation at 301 abduction should be postponed until the rotator cuff healing progresses, because the strain and tensile force on the repair site during internal rotation significantly increased from neutral rotation. In this study, the strain during inferior glide was significantly larger than that occurring in the relaxing position. Therefore, of these four joint mobilization techniques, repeated use of the inferior glide technique should be avoided immediately after rotator cuff repair. Strain on the supraspinatus tendon is commonly measured by a small linear displacement sensor. The sensor, ‘‘Pulse corder’’ in this study, is capable of directly measuring tendon behaviour along with its fibre orientation during joint mobilization, while measurements by imaging techniques are limited. Because of distorted three-dimensional movement of imaging markers buried in the cuff tendons during shoulder motion, accurate detection of strain between each marker becomes extremely difficult. In addition, the sensor could accurately measure the strain with little resistance because preliminary calibration of the Pulse Coder, which was attached in the supraspinatus tendon, demonstrated the high accuracy of 0.1 mm RMS. There is concern that our experiment was performed on a cadaveric model. Since muscle tension in fresh cadavers is different from that of in vivo and some soft tissues were removed, the stress on the tendon due to muscle tension might also differ between them. However, the strain on the tendon in fresh cadavers might be the same or smaller than that of in vivo because physiologic muscle tone and muscle contraction stabilize the humeral head and decrease its movement (Warner et al., 1999). This study had a few limitations. First, the strain obtained from this study was not always consistent because the specimens were obtained from aged cadavers. The supraspinatus tendon in younger adults has great endurance to stress because the connective tissues in adults are more flexible and have greater failure load than that in the aged cadavers (Reeves, 1968). Therefore, we believe that our findings are applicable to younger to middle-aged adults.

Second, the safety of joint mobilization could not be determined with regard to the strain–stress curve and endurance to repetitive loading because tensile force to the tendon was not measured. In this study, instead, the strains and gap distances in the stretching and relaxing positions were observed and tensile forces on the repaired tendon were estimated from the data in a previous study (Hatakeyama et al., 2001a). Finally, our finding can be applied only to rotator cuff tears that are of the same size as or smaller than those studied here. Strain on a massive tear including the infraspinatus and/or subscapularis tendons during joint mobilization may be different from the results of this study. Therefore, further studies are required to clarify the strain on a massive tear. Further research following our study should determine whether the joint mobilization techniques could prevent joint contracture after rotator cuff repair. Investigation of strain and stress on the contracted joint capsule during mobilization techniques are necessary. In addition, the effect of repetitive joint mobilization on the repaired tendon and capsule should be also determined. These studies will contribute to the making of good decisions concerning the application of joint mobilization after rotator cuff repair.

5. Conclusion Our findings suggested that if rotator cuff repair is performed at 01 of abduction, joint mobilization techniques at 01 of arm abduction should be avoided immediately after supraspinatus tendon repair because these techniques produce large strain of the tendon and form a gap at the repair site. In the same fashion, our findings also suggest that joint mobilization techniques at 301 of arm abduction can be used without large strain of the tendon and forming a gap at the repair site. However, inferior glide, even at 301 abduction, should be postponed because relatively larger strain than that of the relaxing position may lead to failure of tendon repair. Although our findings would be useful to decide the application of joint mobilization after rotator cuff repair, further study regarding strain on the contracted joint capsule or the effect of repetitive joint mobilization would provide even more information.

Acknowledgments The authors would like to thank Daisuke Suzuki Ph.D. for his technical assistance. In addition, we would like to thank Shuhei Takauji and Mitsuo Nakamura for their assistance.


References An KN, Jacobsen MC, Berglund LJ, Chao EY. Application of a magnetic tracking device to kinesiologic studies. Journal of Biomechanics 1988;21(7):613–20. Bruzga B, Speer K. Challenges of rehabilitation after shoulder surgery. Clinics in Sports Medicine 1999;18(4):769–93. Burkhart SS, Johnson TC, Wirth MA, Athanasiou KA. Cyclic loading of transosseous rotator cuff repairs: tension overload as a possible cause of failure. Arthroscopy 1997a;13(2):172–6. Burkhart SS, Diaz Pagan JL, Wirth MA, Athanasiou KA. Cyclic loading of anchor-based rotator cuff repairs: confirmation of the tension overload phenomenon and comparison of suture anchor fixation with transosseous fixation. Arthroscopy 1997b;13(6):720–4. Burkhart SS, Wirth MA, Simonick M, Salem D, Lanctot D, Athanasiou K. Loop security as a determinant of tissue fixation security. Arthroscopy 1998;14(7):773–6. Cofield RH. Rotator cuff disease of the shoulder. Journal of Bone and Joint Surgery 1985;67A(6):974–9. Conroy DE, Hayes KW. The effect of joint mobilization as a component of comprehensive treatment for primary shoulder impingement syndrome. Journal of Orthopaedic Sports and Physical Therapy 1998;28(1):3–14. Culham E, Peat M. Functional anatomy of shoulder complex. Journal of Orthopaedic Sports and Physical Therapy 1993;18(1):342–50. Evjenth O, Hamberg J. Muscle stretching in manual therapy: a clinical manual. The extremities, vol. 1. Alfta, Sweden: Alfta Rehab Forlag; 1984 p. 40 [Chapter 2]. Flatow EL, Soslowsky LJ, Ticker JB, Pawluk RJ, Hepler M, Ark J, et al. Excursion of the rotator cuff under the acromion: patterns of subacromial contact. The American Journal of Sports Medicine 1994;22(6):779–88. Gerber C, Schneeberger AG, Beck M, Schlegel U. Mechanical strength of repairs of the rotator cuff. Journal of Bone and Joint Surgery 1994;76B(3):371–80. Gerber C, Schneeberger AG, Perren SM, Nyffeler RW. Experimental rotator cuff repair: a preliminary study. Journal of Bone and Joint Surgery 1999;81A(9):1281–90. Gokeler A, Paridon-Edauw GH, DeClercq S, Matthijs O, Dijkstra PU. Quantitative analysis of traction in the glenohumeral joint: in vivo radiographic measurement. Manual Therapy 2003;8(2):97–102. Harryman 2nd DT, Sidles JA, Clark JM, McQuade KJ, Gibb TD, Matsen 3rd FA. Translation of the humeral head on the glenoid with passive glenohumeral motion. Journal of Bone and Joint Surgery 1990;72-A(9):1334–43. Hatakeyama Y, Itoi E, Pradhan RL, Urayama M, Sato K. Effect of arm elevation and rotation on the strain in the repaired rotator cuff tendon: a cadaveric study. The American journal of Sports Medicine 2001a;29(6):788–94. Hatakeyama Y, Itoi E, Urayama M, Pradhan RL, Sato K. Effect of superior capsule and coracohumeral ligament release on strain in the repaired rotator cuff tendon: a cadaveric study. The American Journal of Sports Medicine 2001b;29(5):633–40. Hjelm R, Draper C, Spencer S. Anterior-inferior capsular length insufficiency in the painful shoulder. Journal of Orthopaedic Sports and Physical Therapy 1996;23(3):216–22. Hsu AT, Ho L, Ho S, Hedman T. Immediate response of glenohumeral abduction range of motion to a caudally directed translational mobilization: a fresh cadaver simulation. Archives of Physical Medicine and Rehabilitation 2000a;81(11):1511–6. Hsu AT, Ho L, Ho S, Hedman T. Joint position during anterior– posterior glide mobilization: its effect on glenohumeral abduction range of motion. Archives of Physical Medicine and Rehabilitation 2000b;81(2):210–4. 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

239

dorsal and ventral translational mobilization of the glenohumeral joint. Physical Therapy 2002a;82(6):544–56. 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 of Physical Medicine and Rehabilitation 2002b;83(3):360–6. Itoi E, Berglund LJ, Grabowski JJ, Schultz FM, Growney ES, Morrey BF, et al. Tensile properties of the supraspinatus tendon. Journal of Orthopaedic Research 1995;13(4):578–84. Itoi E, Morrey BF, An KN. Biomechanics of the shoulder. In: Rockwood Jr CA, Matsen 3rd FA, Wirth MA, Lippitt SB, editors. The Shoulder. 3rd ed. Philadelphia, PA: WB Saunders; 2004. p. 223–67 Chapter 6. Johns RJ, Wright V. Relative importance of various tissues in joint stiffness. Journal of Applied Physiology 1962;17(5):824–30. Kaltenborn FM. Manual mobilization of the joints: the Kaltenborn method of joint examination and treatment. The extremities, 5th ed. Oslo, Norway: Olaf Norlis bokhandel; 1999 p. 21–8 [Chapter 2]. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33(2):159–74. Ma CB, MacGillivray JD, Clabeaux J, Lee S, Otis JC. Biomechanical evaluation of arthroscopic rotator cuff stitches. Journal of Bone and Joint Surgery 2004;86A(6):1211–6. Mangus BC, Hoffman LA, Hoffman MA, Altenburger P. Basic principles of extremity joint mobilization using a Kaltenborn approach. Journal of Sport Rehabilitation 2002;11(4):235–50. Matsen 3rd FA, Titelman RM, Lippitt SB, Wirth MA, Rockwood Jr. CA. Rotator cuff. In: Rockwood Jr. CA, Matsen 3rd FA, Wirth MA, Lippitt SB, editors. The shoulder. 3rd ed. Philadelphia, PA: WB Saunders; 2004. p. 795–878 Chapter 15. Nakajima T, Hughes RE, An KN. Effects of glenohumeral rotations and translations on supraspinatus tendon morphology. Clinical Biomechanics 2004;19(6):579–85. Petit CJ, Boswell R, Mahar A, Tasto J, Pedowitz RA. Biomechanical evaluation of a new technique for rotator cuff repair. The American Journal of Sports Medicine 2003;31(6):849–53. Pruitt DL, Manske PR, Fink B. Cyclic stress analysis of flexor tendon repair. Journal of Hand Surgery 1991;16A(4):701–7. Quillen WS, Halle JS, Rouillier LH. Manual therapy: mobilization of the motion-restricted shoulder. Journal of Sport Rehabilitation 1992;1(3):237–48. Reeves B. Experiments on the tensile strength of the anterior capsular structures of the shoulder in man. Journal of Bone and Joint Surgery 1968;50B(4):858–65. Rossouw DJ, McElroy BJ, Amis AA, Emery RJ. A biomechanical evaluation of suture anchors in repair of the rotator cuff. Journal of Bone and Joint Surgery 1997;79B(3):458–61. Tibone JE, McMahon PJ, Shrader TA, Sandusky MD, Lee TQ. Glenohumeral joint translation after arthroscopic, nonablative, thermal capsuloplasty with a laser. The American Journal of Sports Medicine 1998;26(4):495–8. Warner JJ, Greis PE. The treatment of stiffness of the shoulder after repair of the rotator cuff. Instruction Course Lecture 1998;47(8):67–75. Warner JJ, Micheli LJ, Arslanian LE, Kennedy J, Kennedy R. Patterns of flexibility, laxity, and strength in normal shoulders and shoulders with instability and impingement. The American Journal of Sports Medicine 1990;18(4):366–75. Warner JJ, Deng XH, Warren RF, Torzilli PA. Static capsuloligamentous restraints to superior–inferior translation of the glenohumeral joint. The American Journal of Sports Medicine 1992;20(6):675–85. Warner JJ, Bowen MK, Deng X, Torzilli PA, Warren RF. Effect of joint compression on inferior stability of the glenohumeral joint. Journal of Shoulder and Elbow Surgery 1999;8(1):31–6. Zuckerman JD, Leblanc JM, Choueka J, Kummer F. The effect of arm position and capsular release on rotator cuff repair: a biomechanical study. Journal of Bone and Joint Surgery 1991;73B(3):402–5.

ARTICLE IN PRESS


Original article

Changes in postural activity of the trunk muscles following spinal manipulative therapy Manuela L. Ferreiraa,b,c, Paulo H. Ferreiraa,b,d, Paul W. Hodgesa, a

Division of Physiotherapy, The University of Queensland, Brisbane QLD 4072, Australia b School of Physiotherapy, The University of Sydney, Sydney, Australia c Departamento de Fisioterapia, Pontifıćie Universidade Cato´lica de Minas Gerais, Belo Horizonte, Brazil d Departamento de Fisioterapia, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil Received 6 May 2005; received in revised form 10 May 2006; accepted 27 June 2006

Abstract Spinal manipulative therapy (SMT) is common in the management of low back pain (LBP) and has been associated with changes in muscle activity, but evidence is conflicting. This study investigated the effect of SMT on trunk muscle activity in postural tasks in people with and without LBP. In 20 subjects (10 with LBP and 10 controls), EMG recordings were made with fine-wire electrodes inserted into transversus (TrA), obliquus internus (OI), and externus (OE) abdominis. Rectus abdominis (RA) and anterior deltoid EMG was recorded with surface electrodes. Standing subjects rapidly flexed an arm in response to a light, before and after a small amplitude end range rotational lumbar mobilization at L4-5. In controls, there was no change in trunk muscle EMG during the postural perturbation after SMT. In LBP subjects there was an increase in the postural response of OI and an overall increase in OE EMG. There was no change in TrA or RA EMG. These results indicate that SMT changes the functional activity of trunk muscles in people with LBP, but has no effect on control subjects. Importantly, SMT increased the activity of the oblique abdominal muscles with no change in the deep trunk muscle TrA, which is often the target of exercise interventions. r 2006 Elsevier Ltd. All rights reserved. Keywords: Spinal manipulative therapy; Motor control; Transversus abdominis; Fine-wire EMG

1. Introduction Spinal manipulative therapy (SMT), defined as manual loading of the spine using short or long leverage methods, is one of the most common approaches in the management of low back pain (LBP). The efficacy of SMT on clinical outcome measures for people with LBP has been investigated in several systematic reviews which conclude that it leads to clinically significant improvements in pain and function (van Tulder et al., 1997; Ferreira et al., 2002; Assendelft et al., 2003). Despite the positive clinical benefit, the physiological mechanisms responsible for these effects are still unclear. Corresponding author. Tel.: +61 7 3365 2008; fax: +61 7 3365 2775. E-mail address: [email protected] (P.W. Hodges).


Several possible mechanisms have been discussed in the literature. For instance, SMT has been associated with sympathoexcitatory effects (Vicenzino et al., 1998), changes in passive and active spinal range of motion (Nilsson et al., 1996; Lehman and McGill, 2001), and effects such as enhanced production of tumour necrosis factor and substance P (Brennan et al., 1992). Alternatively, it has been argued that SMT changes muscle activity. However, few studies have investigated the effects on parameters of muscle activation and results are contradictory. For instance increased (Herzog et al., 1999) and decreased (Dishman and Bulbulian, 2000; Dishman and Bulbulian, 2001) activity of the paraspinal muscles have been reported in response to SMT. Several factors may explain the inconsistency of the results. First, studies have used a variety of manipulative techniques. The term SMT is used to describe a broad

ARTICLE IN PRESS M.L. Ferreira et al. / Manual Therapy 12 (2007) 240–248

spectrum of techniques that may or may not include audible cavitation. Second, a range of muscle functions, from resting activity to reflex amplitude have been tested. Finally, most studies have involved people without LBP. Thus, from the present data it is not possible to determine whether manipulation modifies the activity of the trunk muscles in people with LBP during functional activities. One functional task that has been evaluated extensively is the postural response of the trunk muscles to a rapid limb movement. In this task trunk muscle activity is initiated prior to movement to prepare the spine for the perturbation from limb movement (Belenkii et al., 1967; Aruin and Latash, 1995; Hodges and Richardson, 1997). This task provides an ideal model to evaluate the effect of SMT as it provides a measure of the preplanned strategy used by the central nervous system to control the trunk muscles. Furthermore, people with recurrent LBP have been found to have changes in this pre-planned postural adjustment. Notably, activity of transversus abdominis (TrA), the deepest abdominal muscles, is delayed (Hodges and Richardson, 1996; Hodges and Richardson, 1998). Although impaired activity of this muscle is a relatively consistent finding in LBP, activity of superficial trunk muscles is often increased (Arendt-Nielsen et al., 1996; Radebold et al., 2000; Hodges et al., 2003a, b). We hypothesized that manipulative therapy would change the response of the trunk muscles, but on the basis of previous data it was not possible to predict whether activity would be increased or decreased. Thus, the aims of the present study were, first, to determine whether the pre-planned postural activity of the trunk muscles could be modified by SMT and second, to investigate whether the effect differed between people with and without LBP.

2. Materials and methods 2.1. Subjects Twenty subjects (10 with a history of LBP and 10 controls) volunteered for this experiment. The subject demographics are presented in Table 1 and were not different between the two groups. Subjects were excluded if they had any respiratory or neurological disorder, musculoskeletal pain elsewhere in the spine or lower limbs or if they had been pregnant in the previous 2 years. Subjects in the control group were also excluded if they had a history of LBP that had restricted function or for which they had sought medical or allied health intervention. To be included in the LBP group, volunteers needed to have a history of at least one episode of LBP that had limited function or work in the past 18 months and had

241

Table 1 Demographic details

Age (years) Height (cm) Weight (kg)

Control group mean (SD)

LBP group mean (SD)

P-valuea

33 (11) 159 (38) 68 (13)

28 (5) 171 (10) 69 (13)

P ¼ 0.21 P ¼ 0.35 P ¼ 0.95

LBP, low back pain. a P-values refers to t-test for independent samples.

an episode of LBP within the past 6 months. Patients were excluded if they had neurological signs, specific spinal pathology (e.g. malignancy, inflammatory joint or bone disease), if they had undergone back surgery in the past 12 months, or if they could not tolerate a grade IV lumbar rotary mobilization technique (no subject was excluded on the basis of this criteria). The mean (SD) pain intensity (past 7 days) was 3.2 (2.3) on a 10 cm visual analogue scale (VAS) and the disability score (Roland Morris) was 3.2 (1.8). All included patients presented with LBP of at least 3 months duration. The study was approved by the Institutional Medical Research Ethics Committee and all procedures were conducted in accordance with the declaration of Helsinki. 2.2. Electromyography EMG recordings were made with surface and intramuscular fine-wire electrodes. Fine-wire electrodes were fabricated from two strands of Teflon-coated stainless-steel wire (75 mm diameter, A-M systems, USA) threaded into a hypodermic needle (0.6 32 mm) and inserted with guidance from ultrasound imaging into the right ventro-lateral abdominal wall muscles: TrA, obliquus internus abdominis (OI), and obliquus externus abdominis (OE) half-way between the iliac crest and distal border of the rib cage in the anterior axillary line (Hodges et al., 1999). Surface electrodes were placed on left anterior deltoid muscle approximately in parallel with the muscle fibres, and over the muscle belly of the right rectus abdominis (RA). EMG data were amplified 2000 times, band-pass filtered between 20 and 1 kHz (Neurolong, Digitimer, UK) and sampled at 2 kHz using a Power-1401 and Spike-2 software (Cambridge Electronic Design, UK). 2.3. Procedure Subjects rapidly flexed or extended the left upper limb in response to visual command in a choice reaction time task. Two directions of movement were performed to limit the predictability of the response, but only the flexion data was analysed. Ten repetitions in each direction were performed in random order before and


M.L. Ferreira et al. / Manual Therapy 12 (2007) 240–248

after the spinal manipulative technique (see below). The second set of ten arm movements commenced immediately after the cessation of the manipulative technique. 2.4. Spinal manipulative technique A specific small amplitude rotational end-range (grade IV) oscillatory mobilization technique was directed to at L4–L5 with subjects in left side-lying (Maitland et al., 2001). This procedure was standardized for all patients because of the placement of electrodes on the right side of the trunk. Although it is unlikely that L4–L5 was the symptomatic level in all patients we considered this technique to be appropriate to test the aims of the experiment for a number of reasons. First, although attempts are made to focus the technique to a target segment, movement will occur over a number of levels and second, recent data suggest that that patients report significant decreases in pain even when mobilizing non-painful lumbar segments (Chiradejnant et al., 2003). The technique was repeated three times for 30 s, with oscillations at 1 Hz, with feedback from a metronome. The mobilization technique was performed at the same spinal level (L4–L5) for all subjects, even though this might not have been the level of pain in some subjects. Pain intensity was measured on a VAS before and after the application of the technique. 2.5. Data analysis In order to investigate both temporal and spatial aspects of EMG, root mean square (RMS) EMG amplitude was calculated during four 25 ms epochs before the onset of deltoid EMG and four 25 ms epochs after the onset of deltoid EMG. This analysis technique measures the postural response without problems associated with detection of EMG onset of the trunk muscles. EMG onset detection was difficult due to the high degree of baseline activity, particularly in OE after the performance of the manipulative technique. Data were normalized to the epoch with the greatest amplitude in the pre-manipulation condition. This normalization method provides a high sensitivity to compare the pre- and post-manipulation conditions for each group, but does not permit comparison of EMG amplitude between muscles or between groups, although the pattern of changes in pattern of activity could be compared between groups. Data were not normalized to EMG amplitude in a maximum voluntary contraction as these values are not possible to obtain reliably in people with LBP (Allison et al., 1998) and data were not normalized to a submaximal task as people with LBP are likely to use an abnormal strategy during submaximal tasks, making it an invalid reference. Furthermore, recent data suggests that normalization to a submaximal task may increase variability (Urquhart et al., 2005).

However, as this study was a repeated measures design with each subject acting as their own controls normalization would not affect the results, but normalization to pre-manipulation peak amplitude normalizes all data to a similar scale for graphical representation of the data. Statistical analysis involved a three-way repeated measures analysis of variance (ANOVA) with two repeated measures factors (pre- and post-mobilization and epoch) and one independent factor (group). Posthoc Duncan tests were used when appropriate. Due to the inability to compare between muscles as a result of the normalization procedure analyses were run separately for each muscle (RA, OE, OI and TrA). The alpha level was set at 0.05.

3. Results 3.1. TrA When subjects rapidly moved the left arm, a burst of EMG activity of TrA occurred in association with deltoid EMG. There was a trend for the activity to occur in an earlier epoch for TrA in the control group compared to the LBP group. Fig. 1 shows an increase in EMG activity greater than 10% above baseline EMG in epochs 3 and 4 in the control group. This increase is not as evident in the LBP group until epoch 5. When arm movement was performed after the application of spinal mobilization there was no change in the activity of TrA in the control or LBP subjects compared to the pre-mobilization condition (Fig. 1). Table 2 presents the main effects and interactions. 3.2. OI Similar to TrA, a burst of OI EMG activity occurred in conjunction with arm movement. In both groups the increase in EMG activity tended to start in the epoch prior to the onset of deltoid EMG. Following the mobilization technique there was no change in the postural response for the control group (P ¼ 0:20). Main effects and interactions are shown in Table 2. Although there was no consistent effect for the control group, OI EMG was increased after mobilization in 2 subjects. This increased the variability of the response (Fig. 2). In contrast, OI EMG activity was increased following the onset of deltoid EMG (Epochs E5: 0–25, E6: 25–50, E7: 50–75, E8: 75–100 ms) for subjects in the LBP group (all: Po0:001). Notably, mobilization did not increase the background activity prior to arm movement, only the postural adjustment associated with the limb movement. Variability was increased in this group. Three subjects had an accentuated increase in muscle activity, and two subjects had a decrease in EMG activity in epochs 1–3.

ARTICLE IN PRESS M.L. Ferreira et al. / Manual Therapy 12 (2007) 240–248 Control subjects Pre-mobilisation

LBP 1.4 Prop. pre-mobilisation EMG

243

Post-mobilisation

1.2 1.0 0.8 0.6 0.4 0.2 0 E1

E2

E3

E4

E5

E6

E7

E8

E1

E2

E3

E4

E5

E6

E7

E8

Fig. 1. Change in TrA EMG with SMT. EMG amplitude is shown for four 25-ms epochs prior to onset of deltoid EMG (E1–E4) and four 25-ms epochs after the onset of deltoid (E5–E8). The dashed line divides the epochs before and after the onset of deltoid EMG. EMG is presented as a proportion of peak activity recorded across the eight epochs during the pre-mobilization trial. Confidence intervals (95%) are shown.

Table 2 Main effects and interactions for comparison between groups, between epochs and pre- and post-manipulative technique (Manip) Main effect/interaction

TrA

Epoch Group Epochgoup Manip Manipepoch Manipgroup Manipepochgroup

OI

OE

RA

F

P

F

P

F

P

F

P

5.27 0.32 0.28 0.40 0.17 3.33 0.23

o0.01* 0.57 0.96 0.53 0.99 0.07 0.98

2.44 5.07 0.51 16.81 0.45 4.73 0.59

0.02* 0.03* 0.82 o0.01* 0.87 0.03* 0.76

0.5 29.9 0.07 11.99 0.17 16.25 0.25

0.83 o0.01* 0.99 o0.01* 0.99 o0.01* 0.97

1.03 10.88 0.42 2.61 0.02 1.17 0.11

0.42 o0.01* 0.89 0.11 1.00 0.28 1.00

*po0.05.

LBP

Control subjects Pre-mobilisation

3.0 Prop. pre-mobilisation EMG

Post-mobilisation 2.5 2.0 1.5

*

1.0

*

*

*

0.5 0.0 -0.5 E1

E2

E3

E4

E5

E6

E7

E8

E1

E2

E3

E4

E5

E6

E7

E8

Fig. 2. Change in OI EMG with SMT. EMG amplitude is shown for four 25-ms epochs prior to onset of deltoid (E1–E4) and four after the onset of deltoid. The dashed line divides the epochs before and after the onset of deltoid EMG. EMG is presented as a proportion of peak activity recorded across the eight epochs during the pre-mobilization trial. Confidence intervals (95%) are shown. An increase in EMG activity was observed postmanipulation, after the onset of deltoid in the LBP group. *Po0.001.


244

3.3. OE

3.4. RA

When subjects with and without LBP flexed the left upper limb in response to a visual stimulus, there was an increase OE EMG. There was a tendency for this burst to initiate prior to the onset of deltoid EMG in both groups. After the application of SMT, no change in OE activity was observed as a consequence of arm movement for the control subjects (P ¼ 0:68). Main effects and interactions are shown in Table 2. In two subjects there was a decrease in OE EMG in epochs 2–5 after the mobilization, which explains the increase in variability in this group. In contrast, when subjects with LBP flexed their left arm after the mobilization, an overall increase in muscle activity was observed for all epochs (Po0:01). That is, OE EMG was increased even prior to arm movement. Likewise, the group with LBP presented greater variability after the mobilization (Fig. 3).

EMG activity for RA was increased for both groups prior to the onset of deltoid activity (E4: -25 to 0). No change EMG was observed after the mobilization in the control subjects. Although not significant, there was a trend for RA activity to increase in people with LBP after manipulation (Fig. 4). Main effects and interactions are shown in Table 2.

3.5. Pain Average pain intensity immediately before the manipulation was 1.1 (1.7) and immediately after the procedure was 0.7 (2.2). There was no significant change in pain intensity following SMT (P ¼ 0:34).

LBP


Prop. pre-mobilisation EMG

3.0

Post-mobilisation

2.5 2.0 1.5 1.0 0.5

*

*

*

*

*

*

*

*

E1

E2

E3

E4

E5

E6

E7

E8

0.0 -0.5 -1.0 E1

E2

E3

E4

E5

E6

E7

E8

Fig. 3. Change in OE EMG with SMT. EMG amplitude is shown for four 25-ms epochs prior to onset of deltoid (E1–E4) and four after the onset of deltoid. The dashed line divides the epochs before and after the onset of deltoid EMG. EMG is presented as a proportion of peak activity recorded across the eight epochs during the pre-mobilization trial. Confidence intervals (95%) are shown. An overall increase in EMG activity was observed post-manipulation only in the LBP group. *Po0.001.

LBP


1.75 Prop. pre-mobilisation EMG

Post-mobilisation 1.5 1.25 1.0 0.75 0.5 0.25 0 E1

E2

E3

E4

E5

E6

E7

E8

E1

E2

E3

E4

E5

E6

E7

E8

Fig. 4. Change in RA EMG with SMT. EMG amplitude is shown for four 25-ms epochs prior to onset of deltoid (E1–E4) and four after the onset of deltoid. The dashed line divides the epochs before and after the onset of deltoid EMG. EMG is presented as a proportion of peak activity recorded across the eight epochs during the pre-mobilization trial. Confidence intervals (95%) are shown.


4. Discussion The results of this study demonstrate that pre-planned postural activity of the trunk muscles can be modified by SMT performed as a small amplitude oscillation without cavitation. This was evidenced by the increased amplitude of OE and OI EMG as a component of postural adjustment associated with arm movement after SMT. However, this change was only observed in people with LBP. In contrast to the superficial oblique abdominal muscles, SMT did not affect the postural response of TrA in either group. There are several possible mechanisms for the increase in EMG activity following SMT that are related to the effect of manual techniques on motoneuron excitability via direct effects of stimulation of joint and muscle afferents, the effect of manual techniques on pain, and other neurophysiological mechanisms. However, there has been considerable debate and findings are often inconsistent. One extensively argued possibility is that excitability of spinal muscle motoneurons is changed by afferent input from stimulation of mechanoreceptors in the joint capsules (Indahl et al., 1997), ligaments (Indahl et al., 1997), muscles (Herzog et al., 1999) and cutaneous receptors (Herzog et al., 1999). However, some studies report increased excitability (Herzog et al., 1999; Keller and Colloca, 2000), while others report decreased activity (Dishman and Bulbulian, 2000; Lehman and McGill, 2001; Lehman et al., 2001). These discrepancies may be explained to some extent by the methods that have been used to evaluate muscle activity. For instance, Keller and Colloca (2000) reported increased erector spinae activity during trunk extension in prone, Herzog et al. (1999) reported short latency excitatory muscle activity in response to manipulative thrusts, whereas Dishman and Bulbulian (2000) reported reduced Hoffman-reflex (H-reflex) amplitude in the gastrocnemius muscle in the leg, and Lehman and McGill (2001) reported decreased muscle response to a painful mechanical stimulus applied over the spinous process. Thus, few studies have measured the same parameter and it is difficult to predict from existing data how trunk muscle activity would change during a postural adjustment in association with arm movement. Short latency responses in paraspinal muscles to manipulative and mobilization techniques have been extensively reported (Herzog et al., 1999; Colloca and Keller, 2001; Shirley et al., 2002). These responses have been argued to be mediated by stimulation of muscle spindles and have short duration (100–400 ms) (Herzog et al., 1999). However, it is unclear whether these responses are associated with changes in ongoing EMG activity, although ongoing reduction in abdominal muscle activity has been identified in a single case report by Herzog et al. (1999) following initial excitation.

245

Similarly it is difficult to interpret data regarding amplitude of H-reflexes in lower leg muscles. Although these responses have been shown to be reduced (Dishman and Bulbulian, 2000), and it is argued that the motoneurons that innervate these muscles lie in the lumbosacral spinal cord, it is unlikely that these responses would reflect the properties of the distinct motoneurons to the spinal muscles. Furthermore, H-reflex amplitude is influenced by pre-synaptic effects and may not reflect changes in motoneuron excitability (Rudomin, 2002). Finally, the response of the paraspinal muscles to a mechanical stimulus to the segment may reflect changes in local reflex mechanisms and not functional activity of the trunk muscles. Studies that have investigated trunk muscle activity during trunk movements also have divergent results. Keller and Colloca (2000) reported increased activity during back extension. In contrast, Lehman and McGill (2001) found no consistent change in muscle activity during trunk movements in standing, but reported a tendency for decreased activity in some muscles in some subjects. Consistent with the data of Keller and Colloca (2000), we identified a significant increase in activity of OE and OI, and a trend for increased RA EMG. The task used in the present study involved evaluation of the activity of the trunk muscles in association with arm movement. These responses are considered to be preprogrammed by the nervous system as they are initiated in advance of limb movement (Belenkii et al., 1967; Bouisset and Zattara, 1981). As such, they reflect the strategy used by the nervous system to prepare the body for the perturbation resulting from the movement (Belenkii et al., 1967; Hodges et al., 1999). Notably, temporal and spatial parameters of the anticipatory postural adjustments are matched to the timing, direction and amplitude of the perturbation from limb movement (Bouisset and Zattara, 1981; Aruin and Latash, 1995; Hodges and Richardson, 1997; Hodges et al., 1999). Thus, changes in EMG amplitude could reflect either changes in motoneuron excitability (i.e. larger response initiated for the same descending drive), or changes in descending drive (i.e. increased output from higher centres). In terms of modification of descending drive to the motoneuron pool, if SMT modifies the afferent discharge from the peripheral mechanoreceptors this may modify the descending postural response if input regarding the status of the spine was changed. It has been shown that excitation of 1a afferents by mechanical stimulation such as vibration modifies the perception of the position of the spine (Brumagne et al., 1999) and leads to modified postural responses (Kasai et al., 2002). Although the present data do not allow differentiation between spinal or supraspinal mechanisms, the data do indicate that the postural response of the trunk muscles is modified by SMT. Importantly; the data indicate that this is not a general



response, as not all muscles responded in a similar manner. Why no change was identified in the response of TrA following SMT is unclear. However, the function of this muscle has been shown to be impaired in people with LBP (Hodges and Richardson, 1996; Hodges and Richardson, 1998), and this may not be resolved by the mechanical stimulus. This study evaluated a rotational spinal mobilization without thrust. This technique is likely to stretch the abdominal muscles, and thus stimulate muscle spindle afferents in the oblique muscles. It is possible that this stimulus may have modified excitability of abdominal motoneurons. However, the rotational mobilization technique applied in this study rotated the pelvis and spine to the right which would shorten the right OI and stretch the OE, from which EMG was recorded. As activity of both OI and OE was increased the change in length of he muscles with the procedure (opposite for these muscles) is unlikely to explain the increased in activity observed for both muscles. The present data indicate that changes following SMT were only present in people with LBP, and no changes were identified in healthy control subjects. This is consistent with previous studies. For instance, decreased paraspinal muscle activity in response to painful stimulation over the spinous process following SMT has been shown to occur only at painful segments (Lehman and McGill, 2001). Although the reason that responses were only changed in people LBP is unclear, it may relate to the function of the mechanoreceptors or due to the already abnormal control of the trunk muscles in people with pain. Several studies have reported that proprioception in the spine is reduced in people with LBP (Gill and Callaghan, 1998; Brumagne et al., 2000), thus afferent stimulation from SMT may lead to different responses. In terms of control, it is increasingly accepted that motor control of the trunk muscles is modified in people with LBP. For instance activity of the deep abdominal muscle, TrA is delayed (Hodges and Richardson, 1996) or reduced (Hodges et al., 2004), reflex responses of the paraspinal muscles are modified (Leinonen et al., 2001), co-contraction of the oblique abdominal and paraspinal muscles is increased (Radebold et al., 2000), and paraspinal muscle activity is increased during gait (Arendt-Nielsen et al., 1996) and trunk flexion (Zedka et al., 1999). Thus, changes may be apparent in this population, but not pain-free controls, due to pre-existing changes in muscle coordination. For example, activity of the quadriceps has been shown to increase as a result of reduced inhibition following SMT to the sacroiliac joint (Suter et al., 1999). Thus, the change that was apparent due to pre-existing deficit in the ability to drive the muscle, and no change would be expected if the motoneurons were not inhibited initially. Thus, the pre-existing status of the trunk muscles may determine whether activity is modified by SMT.

Previous authors have argued that changes in muscle activity may be explained by reduction in pain. For instance cervical mobilization produces a hypoalgesic effect (i.e. increased pressure pain thresholds and decreased visual analogue scores), which has been associated with a sympatho-excitatory effect and decreased superficial neck flexor muscle activity (Sterling et al., 2001). Those data are consistent with the hypothesis that SMT activates descending inhibitory pathways mediated through the midbrain periaqueductal grey region (Vicenzino et al., 1998), which could also be responsible for the motor response associated with SMT. Consistent with the present data, animal studies indicate that activation of the dorsal periaqueductal grey region induces motor facilitation (Lovick, 1992). In the lumbar spine, previous studies have identified that muscle responses to mechanical pain stimuli are changed by SMT only when it is applied to a painful spinal segment (Lehman and McGill, 2001). However, this mechanism is unlikely to explain the result of our study as pain levels did not change significantly after the SMT. A notable finding of the present study was the variability between individuals. This is consistent with several previous studies (Herzog et al., 1999; Lehman et al., 2001). Without complete understanding of the mechanism for SMT to change postural responses it is difficult to speculate on the factors contributing to the variability. However, the LBP population was heterogeneous and differences in pathology, pain intensity and functional presentation may be responsible for the differences. Furthermore, the SMT technique applied in this study was aimed at a consistent segment, L4–L5, and this is unlikely to be symptomatic in all individuals. As previous studies have found that responses are only modified by SMT at the painful segment (Lehman and McGill, 2001), this may account for some variability. In addition, although pain relief from the mobilization was not significant across the group, some individuals did report a reduction in symptoms and this may have contributed to individual variation. Was the change in muscle activity observed after SMT positive for clinical improvement? Although SMT has been shown to reduce pain in people with acute LBP (Ferreira et al., 2002), whether the change in muscle response induced by SMT is beneficial for people with chronic LBP is unclear. Increased activity of the oblique abdominals has been identified as a common strategy used by people with LBP to increase the stability of the spine (Radebold et al., 2000; van Dieen et al., 2003). Whether a further increase in activity of these muscles is beneficial is debatable. For instance, augmented activity of these muscles increases spinal loading (McGill et al., 2003), which may have negative consequences in the long term. In fact, some contemporary exercise interventions, which have been shown to be effective in


management of people with acute and chronic LBP (O’Sullivan et al., 1997; Hides et al., 2001), aim to normalize (i.e. reduce) activity of the superficial muscles (Richardson et al., 1999). Furthermore, these interventions argue that activity of deeper trunk muscles, such as TrA, which have been shown to be impaired in people with LBP (Hodges and Richardson, 1996; Hodges and Richardson, 1998), and have been shown to contribute to inter-vertebral control (Hodges et al., 2003a, b), should be augmented. However, SMT did not change the recruitment of this muscle. These results suggest manipulative therapy is unlikely to be an alternative to specific exercise interventions to improve the recruitment of deep abdominal muscles, although it has been shown that SMT can change superficial abdominal muscle recruitment. Further, biomechanical and clinical studies are required to determine whether the change in muscle activity is beneficial to the health of the spine, or leads to further effects. Three methodological issues require consideration. First, EMG was normalized to the peak activity recorded for each muscle before SMT. As a result it is not possible to compare between muscles or between subject groups. Moreover, it does not allow comparison of ongoing background activity between groups. However, it is not practical to normalize to maximum contractions in people with LBP as they do not perform true maximal efforts (Allison et al., 1998). The objective of the study was to evaluate the effect of SMT on the activity of the trunk muscles and normalization to peak activity increases the sensitivity to change in a repeated measures analysis. Second, treatment technique was not pragmatically performed, since the targeted level was not always the most symptomatic and patients were not always positioned on the non-painful side. This may have influenced the results. However, it is unlikely that this would explain the variability in response as mobilization to L4–L5 will affect adjacent segments. A follow-up study would clarify this issue. Moreover, thrust techniques may induce different changes to the mobilization technique used here, even though, it has been reported that neurophysiologic effects, including motor response, are not dependent on audible cavitation and it has been reported that similar changes occur with mobilization and manipulative thrust techniques (Herzog et al., 1999; Dishman and Bulbulian, 2000). Third, the population included in the study presented with low levels of pain and disability. These findings need to be replicated in a more disabled population.

Acknowledgements We thank Rebecca Mellor for assistance with planning of the study and data collection. PH was supported by the NHMRC of Australia.

247

References Allison GT, Godfrey P, Robinson G. EMG signal amplitude assessment during abdominal bracing and hollowing. Journal of Electromyography and Kinesiology 1998;8(1):51–7. Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P. The influence of low back pain on muscle activity and coordination during gait: a clinical and experimental study. Pain 1996;64(2): 231–40. Aruin AS, Latash ML. Directional specificity of postural muscles in feed-forward postural reactions during fast voluntary arm movements. Experimental Brain Research 1995;103:323–32. Assendelft WJ, Morton PC, Yu EI, et al. Spinal manipulative therapy for low back pain: a meta-analysis of effectiveness relative to other therapies. Annals of Internal Medicine 2003;138:871–81. Belenkii V, Gurfinkel VS, Paltsev Y. Elements of control of voluntary movements. Biofizika 1967;12(1):135–41. Bouisset S, Zattara M. A sequence of postural adjustments precedes voluntary movement. Neuroscience Letters 1981;22:263–70. Brennan P, Triano J, McGregor M. Enhanced neutrophil respiratory burst as a biological marker for manipulation forces: duration of the effect and association with substance P and tumor necrosis factor. Journal of Manipulative and Physiological Therapeutics 1992;15(2):83–9. Brumagne S, Lysens R, Swinnen S, Verschueren S. Effect of paraspinal muscle vibration on position sense of the lumbosacral spine. Spine 1999;24(13):1328–31. Brumagne S, Cordo P, Lysens R, Verschueren S, Swinnen S. The role of paraspinal muscle spindles in lumbosacral position sense in individuals with and without low back pain. Spine 2000;25(8): 989–94. Chiradejnant A, Maher C, Latimer J. Efficacy of "therapist-selected" versus "randomly selected" mobilisation techniques for the treatment of low back pain: a randomised controlled trial. Australian Journal of Physiotherapy 2003;49:233–41. Colloca CJ, Keller TS. Stiffness and neuromuscular reflex response of the human spine to posteroanterior manipulative thrusts in patients with low back pain. Journal of Manipulative and Physiological Therapeutics 2001;24(8):489–500. Dishman JD, Bulbulian R. Spinal reflex attenuation associated with spinal manipulation. Spine 2000;25(19):2519–24 discussion 2525. Dishman JD, Bulbulian R. Comparison of effects of spinal manipulation and massage on motoneuron excitability. Electromyography and Clinical Neurophysiology 2001;41(2):97–106. Ferreira M, Ferreira P, Latimer J. Does spinal manipulative therapy help people with chronic low back pain? Australian Journal of Physiotherapy 2002;48:277–84. Gill KP, Callaghan MJ. The measurement of lumbar proprioception in individuals with and without low back pain. Spine 1998;23(3): 371–7. Herzog W, Scheele D, Conway PJ. Electromyographic responses of back and limb muscles associated with spinal manipulative therapy. Spine 1999;24(2):146–52 discussion 153. Hides JA, Jull GA, Richardson CA. Long term effects of specific stabilizing exercises for first episode low back pain. Spine 2001;26:243–8. Hodges PW, Richardson CA. Inefficient muscular stabilisation of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdominis. Spine 1996;21:2640–50. Hodges PW, Richardson CA. Feedforward contraction of transversus abdominis in not influenced by the direction of arm movement. Experimental Brain Research 1997;114:362–70. Hodges PW, Richardson CA. Delayed postural contraction of of transversus abdominis associated with movement of the lower limb in people with low back pain. Journal of Spinal Disorders 1998; 11(1):46–56.



Hodges PW, Cresswell AG, Thorstensson A. Preparatory trunk motion accompanies rapid upper limb movement. Experimental Brain Research 1999;124:69–79. Hodges P, Kaigle Holm A, Holm S, Ekstrom L, Cresswell A, Hansson T, et al. Intervertebral stiffness of the spine is increased by evoked contraction of transversus abdominis and the diaphragm: in vivo porcine studies. Spine 2003a;28(23):2594–601. Hodges PW, Moseley GL, Gabrielsson AH, Gandevia SC. Acute experimental pain changes postural recruitment of the trunk muscles in pain-free humans. Experimental Brain Research 2003b; 151:262–71. Hodges PW, Saunders S, et al. Reduced tonic activity of the deep trunk muscle during locomotion in people with low back pain. Porto, Portugal: Spineweek; 2004. Indahl A, Kaigle AM, Reikeras O, Holm SH. Interaction between the porcine lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles. Spine 1997;22(24):2834–40. Kasai T, Yahagi S, Shimura K. Effect of vibration-induced postural illusion on anticipatory postural adjustment of voluntary arm movement in standing humans. Gait Posture 2002;15(1):94–100. Keller TS, Colloca CJ. Mechanical force spinal manipulation increases trunk muscle strength assessed by electromyography: a comparative clinical trial. Journal of Manipulative and Physiological Therapeutics 2000;23(9):585–95. Lehman GJ, McGill SM. Spinal manipulation causes variable spine kinematic and trunk muscle electromyographic responses. Clinical Biomechanics (Bristol, Avon) 2001;16(4):293–9. Lehman GJ, Vernon H, McGill SM. Effects of a mechanical pain stimulus on erector spinae activity before and after a spinal manipulation in patients with back pain: a preliminary investigation. Journal of Manipulative and Physiological Therapeutics 2001;24(6):402–6. Leinonen V, Kankaanpaa M, Luukkonen M, Hanninen O, Airaksinen O, Taimela S. Disc herniation-related back pain impairs feedforward control of paraspinal muscles. Spine 2001;26(16):E367–72. Lovick TA. Midbrain influences on ventrolateral medullo-spinal neurons in the rat. Experimental Brain Research 1992;90:147–52. Maitland G, Banks K, English K, Hengeveld E. Lumbar spine. Maitland’s Vertebral Manipulation. Oxford: Butterworth-Heinemann; 2001. McGill SM, Grenier S, Kavcic N, Cholewicki J. Coordination of muscle activity to assure stability of the lumbar spine. Journal of Electromyography and Kinesiology 2003;13(4):353–9.

Nilsson N, Christensen 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. O’Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22(24):2959–67. Radebold A, Cholewicki J, Panjabi MM, Patel TC. Muscle response pattern to sudden trunk loading in healthy individuals and in patients with chronic low back pain. Spine 2000;25(8):947–54. Richardson CA, Jull GA, Hodges PW, Hides JA. Therapeutic exercise for spinal segmental stabilisation in low back pain: Scientific basis and clinical approach. Edinburgh: Churchill Livingstone; 1999. Rudomin P. Selectivity of the central control of sensory information in the mammalian spinal cord. Advances in Experimental Medicine and Biology 2002;508:157–70. Shirley D, Ellis E, Lee M. The response of posteroanterior lumbar stiffness to repeated loading. Manual Therapy 2002;7(1):19–25. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent effects on pain, sympathetic nervous system activity and motor activity. Manual Therapy 2001;6(2):72–81. Suter E, McMorland G, Herzog W, Bray R. Decrease in quadriceps inhibition after sacroiliac joint manipulation in patients with anterior knee pain. Journal of Manipulative and Physiological Therapeutics 1999;22(3):149–53. Urquhart DM, Hodges PW, Allen TJ, Story IH. Abdominal muscle recruitment during a range of voluntary exercises. Manual Therapy 2005;10(2):144–53. van Dieen JH, Selen LP, Cholewicki J. Trunk muscle activation in lowback pain patients, an analysis of the literature. Journal of Electromyography and Kinesiology 2003;13(4):333–51. van Tulder M, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain - a systematic review of randomized controlled trials of the most common interventions. Spine 1997;22(18):2128–56. Vicenzino B, Collins D, Benson H, Wright A. An investigation of the interrelationship between manipulative therapy-induced hypoalgesia and sympathoexcitation. Journal of Manipulative and Physiological Therapeutics 1998;21(7):448–53. Zedka M, Prochazka A, Knight B, Gillard D, Gauthier M. Voluntary and reflex control of human back muscles during induced pain. Journal of Physiology (London) 1999;520(Pt 2):591–604.

ARTICLE IN PRESS


Original article

Assessment of fine motor control in patients with occupation-related lateral epicondylitis Darrell K. Skinnera,, Sandra L. Curwinb a

Okanagan College, 1000 KLO Road, Kelowna, BC, Canada V1Y 4X8 b 100-1314 Tower Road, Halifax, NS, Canada B3H 4S7

Received 30 October 2005; received in revised form 31 May 2006; accepted 27 June 2006

Abstract Lateral epicondylitis (LE) is a common overuse injury related to a mechanical overload of the wrist extensors’ origin; however, some patients also complain of clumsiness suggesting a possible motor control problem. The purpose of this study was to examine for differences in fine motor control ability between subjects with LE and matched control subjects. Subtests of the Purdue Pegboard Test (PPT) and the Complete Manual Dexterity Test (CMDT) were administered to 28 subjects with LE, and 28 age, gender, and hand dominance-matched control subjects. The LE group demonstrated a significant decrease in fine motor control ability on both measures, compared with the control group on both the PPT, F(1,52) ¼ 9.98, P ¼ 0.003, and the CMDT, F(1,52) ¼ 18.11, P ¼ 0.001. There appeared to be no effect for the length of time since injury. There were significant differences in fine motor control ability between individuals with LE and a matched control group for both measures used. These results suggest that tests of fine motor control should be considered in the assessment of clients with LE. The mechanism related to the deficit is unknown and warrants further research. r 2006 Elsevier Ltd. All rights reserved. Keywords: Tendonitis; Lateral epicondylitis; Motor control; Neuroplasticity; Repetitive strain

1. Introduction Lateral epicondylitis of the elbow (LE) is a common soft tissue injury which involves the tendinous origin of the wrist and finger extensors (Nirschl and Ashman, 2003). Epidemiological studies have suggested an incidence of approximately 4–7% in the general population (Assendelft et al., 1996). Workers in occupations involving repetitive hand-intensive work appear particularly at risk for developing this condition (Kivi, 1984; Dimberg et al., 1989; Kurppa et al., 1991; Chiang et al., 1993). Despite the high incidence of LE, there is little consensus on the best treatment approach to guide the clinician (Labelle et al., 1992; Bowen et al., Corresponding author. Tel.: +1 250 762 5445x4464; fax: +1 250 862 5461. E-mail address: [email protected] (D.K. Skinner).


2001; Smidt et al., 2003). LE is considered a self-limiting condition that will resolve gradually with adequate rest and time, however, there is a smaller subpopulation of patients with chronic or recurrent problems and increased disability who present a significant treatment challenge (Burgess, 1990; Boyer and Hastings, 1999; Haahr and Andersen, 2003). The ‘‘overuse’’ or ‘‘biomechanical’’ model of LE focuses on the repetitive mechanical overloading of the tendon beyond its adaptive and reparative capacity as the primary cause of signs and symptoms (Jarvinen et al., 1997; Melborn, 1998; Moore, 2002). Few signs of inflammation are typically present in chronic cases of tendonitis, and it is now appreciated that the traditional soft tissue healing model with a prolonged inflammatory response does not fully explain the pathology involved (Assendelft et al., 1996; Green et al., 2002; Smidt et al., 2002). The effect of repetitive movement on the


D.K. Skinner, S.L. Curwin / Manual Therapy 12 (2007) 249–255

somatosensory cortex has interested researchers. Highly stereotypical repetitive movement patterns have been shown in animal models to produce cortical changes which may lead to impaired motor performance (Byl et al., 1996a, b, 1997; Byl and Melnick, 1997). It is possible that these changes in movement strategies and altered muscle recruitment patterns, whether due to pain or central alterations in motor control, may result in increased loading of the already compromised structures within the muscles and tendon attachments (Byl et al., 2000a, b; Barr et al., 2004; Byl, 2004; Ervilha et al., 2004). Patients in the clinical setting often describe a feeling of ‘‘clumsiness’’ associated with LE, suggesting the possible existence of a fine motor control problem. Few researchers have measured motor control in patients with tendonitis. Viikari-Juntura et al. (1994) compared reaction time, movement time, visual attention, and visuospatial ability in 26 meat cutters and packers with a history of two or more episodes of wrist tenosynovitis to a control group, and found no significant differences. Subjects in the experimental group had a history of tenosynovitis, however, were asymptomatic at the time of testing (Viikari-Juntura et al., 1994). These results suggest that poor manual dexterity may not be a predictor of wrist tenosynovitis. Pienimaki et al. (1997) measured the gross motor control of patients with chronic unilateral LE, and concluded that reaction speed and speed of movement were decreased bilaterally in patients when compared to age and gender-matched control subjects; however, were unable to explain the results (Pienimaki et al., 1997). The purpose of the present study was to systematically examine whether there were differences in fine motor control ability between individuals with LE and control subjects matched on age, gender, and hand dominance, using valid and reliable measures. This study was reviewed and approved by the Ethics Boards of the University of Alberta, and Okanagan University College.

2. Subjects Subjects in the LE group (n ¼ 28) were injured workers between the ages of 30 and 53 years with LE, attending a Cumulative Activity-Related Disorder (CARD) programme at Millard Health in Edmonton (n ¼ 15) or Orion Health in Calgary (n ¼ 13). All the subjects in the LE group were on a Worker’s Compensation claim due to a work-related repetitive strain injury of the upper extremity. A work site visit was completed by a staff of the CARD programme to confirm the presence of identifiable risk factors to substantiate the claim. The length of time since injury was self-reported by the subject as the time in weeks since they first experienced symptoms related to their

claim. The mean time since injury for subjects in the LE group was 30.54 weeks (SD 36.69 weeks). The 28 control subjects were individuals also between 30 and 53 years, attending physiotherapy treatment at Sun City Physiotherapy (n ¼ 21), or Columbia Health (n ¼ 7) in Kelowna, for a non-upper extremity-related condition. The procedure for subject recruitment and reasons for attrition are illustrated in a flow chart (Fig. 1).

3. Methods The diagnosis of LE was confirmed clinically by: (1) positive Cowen’s test (pain reproduced with resisted wrist extension with the forearm in pronation), (2) lateral elbow pain with an extended grasp, (3) accompanying localized tenderness to palpate near the origin of the extensor carpi radialis brevis and extensor carpi ulnaris tendons just distal to the lateral epicondyle (Burgess, 1990; Boyer and Hastings, 1999). All three of these tests were required to be positive for inclusion in this study as this gave the strongest reassurance of the presence of LE. Subjects with cervical radiculopathy were excluded by clinical examination by a physiotherapist including a clinical neurological examination of the cervical spine. Subjects with referred pain or elbow pathology, other than LE, were excluded by clinical examination of the shoulder and elbow joints. Subjects with medical factors affecting motor control including: carpal tunnel syndrome and peripheral nerve conditions, multiple sclerosis, Parkinson’s disease, brain tumours, cerebral vascular accidents, peripheral neuropathies and any other known neurological condition were excluded from the LE or control groups. Individuals on medications such as tricyclic antidepressants, neuroleptic and antipsychotic drugs, or those in withdrawal from alcohol or street drugs were also excluded. The presence of medical conditions and medications was determined by chart review and subject interview. Any subject who complained of pain or discomfort during testing was excluded from this study. The outcome measures used in this study were the subtest of the Purdue Pegboard Test (PPT) Model 32020, and the One-handed Turning and Placing test of the Complete Manual Dexterity Test (CMDT) Model 32023A. Both tests are manufactured by Lafayette Instrument (3700 Sagmore Parkway N, PO Box 5729, Lafayette, IN 47903). The PPT consists of a rectangular board with two vertical lines of pin holes, and four shallow wells containing pins at the top of the board. The One-handed Pin Placement assesses a person’s ability to place as many pins as possible over a 30 s period using predominantly a finger tip to tip grip. The total number of pins placed by subjects over three trials of the One-handed Pin Placement test was recorded. The CMDT consists of a folding board with 60 wells, into

ARTICLE IN PRESS D.K. Skinner, S.L. Curwin / Manual Therapy 12 (2007) 249–255

251

Informed Consent

Lateral Epicondylitis (n = 38)

Control Group (n = 31)

Assessed for Eligibility

Assessed for Eligibility

1 subject excluded due to possible radial tunnel syndrome 1 subject excluded due to language difficulties

Testing of Motor Control (n = 36)

Testing of Motor Control (n = 31)

1 subject excluded due to elbow discomfort 1 subject excluded due to shoulder discomfort

1 subject excluded due to symptoms of possible ulnar neuropathy

Matching (n = 34)

Matching (n = 30)

6 subjects excluded as no age or gender match in the control group

2 subjects excluded as no age or gender match in the LE group

Analysed (n = 28)

Analysed (n = 28)

12 males, 16 females Mean age 41.93 (SD 6.44)

12 males, 16 females Mean age 42.36 (SD 6.44) Fig. 1. Subject enrolment.

which medium sized cylindrical blocks are placed. The One-handed Turning and Placing test involves using the fingers to turn over the block, and then reaching forward to place each block consecutively into a well. The total time in seconds to complete four trials of the One-handed Turning and Placing test was recorded. The experience of many therapists who regularly use the PPT is that the original norms supplied with the test are often higher than those test scores observed in clinical practice. Hamm and Curtis (1980) have also recommended that a clinical sample be compared to its own control group (Hamm and Curtis, 1980). The literature suggested that age, gender, and hand dominance were probable confounding variables which could influence the test results and therefore were controlled for in this study.

4. Matching The LE and control groups were matched on the basis of age, gender, and hand dominance. Both genders were included in this study and there were 12 males and 16 females in both the LE and control groups. Matching for age was achieved by individually selecting the control subject with the closest available match for age to the individual with LE. Twenty out of 28 of the subjects (71%) in the LE group had LE in their dominant limb. The effect of hand dominance was controlled by matching the affected limb (whether dominant or non-dominant) of the subject with LE to the corresponding dominant or non-dominant limb of the control subject. The dominant limb was defined as the hand that the individual chooses to write with. The affected limb was the limb that the subject was attending



treatment for. Matching of the LE and control groups by right handedness or left handedness was not performed, as the intent was to control for the possible effect of hand dominance, not the effect of cerebral lateralization. In reality, only one subject in the LE group was left-hand dominant and the inclusion or exclusion of this subject would not have substantially altered the results of this study.

5. Data analyses For reasons described in Fig. 1, 10 subjects in the LE group and three subjects in the control group were excluded from the data. In particular, two subjects in the LE group were excluded due to discomfort during testing. Statistical analyses of the data were performed using the SPSS Ver 12.0 software (SPSS for Windows, SPSS, Chicago, IL). The level for statistical significance was set at Po0.05. Standard descriptive statistical analyses were used to describe the characteristics of both groups. An independent t-test was used to compare the mean ages of the LE and control groups to illustrate that there was no significant difference between the mean ages of the two groups. Two-way ANOVAs were used for each measure to test for the main effect for the factors of hand dominance and group, and the interaction effect between group and affected limb. The LE group was then divided into two subgroups: those who had LE for 12 weeks or less (acute group, n ¼ 13), and those individuals who had LE for greater than 12 weeks (chronic group, n ¼ 15). Chronicity is often defined as pain lasting more than 3 or 6 months, or simply longer than the expected course of healing for an acute disease process (Russo and Brose, 1998; Starcke, 2005). The cut-off point of 12 weeks was used in this study for the purpose of examining if there was a greater deficit of fine motor control in the more chronic cases of LE. A one-way ANOVA was performed with these two subgroups.

6. Results There were no significant differences between the mean ages of the LE group (41.9876.44 years) and the control group (42.3676.44 years) (t2 ¼ 0.25, P ¼ 0.80). The gender distribution was identical between the groups by design. A two-way ANOVA indicated no main effect for hand dominance for either the Purdue scores F(1,52) ¼ 1.66, P ¼ 0.20, or the CMDT scores F(1,52) ¼ 1.52, P ¼ 0.22, and no interaction effect between the group and the affected limb for the Purdue F(1,52) ¼ 0.003, P ¼ 0.96, or the CMDT F(1,52) ¼ 0.10, P ¼ 0.75. Therefore, the scores on the PPT and CMDT for the dominant and non-dominant limbs were collapsed to represent only two groups, n ¼ 28 (Table 1). Subjects in the LE group placed on average 5.00 fewer pins over three trials of the PPT, and were on average 51.03 s slower to complete four trials of the CMDT, than individuals in the control group. Statistical comparison indicated a significant difference between the mean scores of the LE group and the control group, for both the PPT F(1,52) ¼ 9.98, P ¼ 0.003, and the CMDT F(1,52) ¼ 18.11, P ¼ 0.001. In summary, the LE group demonstrated a significant decrease in fine motor control ability on both the PPT and CMDT compared with the age- and gendermatched control groups. There was no significant difference between the acute LE group and the chronic LE group for both the PPT, F(1,26) ¼ 0.087, P ¼ 0.77; and the CMDT, F(1,26) ¼ 0.094, P ¼ 0.76.

7. Discussion Repetitive strain injuries of the wrist and forearm are common work-related conditions that can be disabling and costly to treat (Barr and Barbe, 2002). The commonly assumed model of injury for work-related LE is a biomechanical overloading of the tendons of the common extensor origin; however, resistance of some cases of LE to various forms of conservative treatment

Table 1 Mean test scores and standard deviations for lateral epicondylitis and control groups N

Purdue* lateral epicondylitis control CMDT** lateral epicondylitis control

28 28 28 28

Mean

41.43a 46.43a 354.82b 303.79b

SD

6.84 3.65 46.91 38.49

95% confidence interval for the mean Lower

Upper

38.78 45.01 336.63 288.86

44.08 47.84 373.01 318.71

*Po.005; **Po.001. a Cumulative score of the number of pins placed in three 30 s trials. b Cumulative score in seconds to complete four trials of the Complete Manual Dexterity Test (CMDT).

ARTICLE IN PRESS D.K. Skinner, S.L. Curwin / Manual Therapy 12 (2007) 249–255

may imply that this explanation is not true for all cases. Complaints of clumsiness from some patients with LE suggests that there may be alterations in motor control ability in these individuals (Byl et al., 1996a, b, 1997; Byl and Melnick, 1997). The mean age of the LE subjects (41.9876.44 years) and the propensity for the dominant limb to be most often affected (71%) was consistent with the findings of previous epidemiological studies, suggesting that the sample was likely representative of the larger population of individuals with LE (Coonrad and Hooper, 1973; Assendelft et al., 1996). The results of this study demonstrated a significant decrease in fine motor control in subjects with LE compared to the control group. Results from both the PPT and the CMDT showed similar decreases in fine motor control increasing the concurrent validity of this study. The design of this present study did not allow determination of whether LE preceded the observed deficits, or whether the deficits in fine motor control preceded the development of the LE. Although the sample size was small in both the acute and chronic subgroups, comparison of the two groups suggested that the time since injury did not affect fine motor control performance, and raises some questions regarding the possible time course of the difficulty in motor control. Further research with an estimated sample size of approximately 25 subjects in each group would be needed to confirm the current findings. The results from animal research by Barbe and Barr suggest that multiple pathomechanisms may exist during the development of tendonitis, and that simultaneous pathophysiological changes may occur in the musculoskeletal system along with neurological system adaptation (Barr and Barbe, 2002; Barbe et al., 2003). The mechanisms involved in the decrease in fine motor control in the subjects with LE in this study are not known. The literature related to this area does, however, offer some possible theories. Experimental primate studies (Jenkins et al., 1990; Byl et al., 1996a, b), and human studies with musicians (Elbert et al., 1998; Byl et al., 2000a b; Pantev et al., 2001), have demonstrated that reorganization of the cortical representation of the somatosensory cortex (Brodman’s area 3b) can occur with repetitive use. These neuroplastic changes potentially may result in decreased fine motor control and increased load on the tendon structures. The difficulties in fine motor control found in subjects with LE in this study, however, does not explicitly mean that this is due to cortical remodelling. Other possible explanations for the results of this study include altered motor performance due to the effect of a chronically painful condition, or individual behavioural responses (Lund et al., 1991; Smeulders et al., 2001; Flor, 2003; Giamberardino, 2003). In this present study, pain did not appear to be a significant factor during testing as only two subjects in the LE group complained of

253

discomfort during testing and were excluded. The verbal instructions given to each subject are standardized to help minimize differences in individual performance due to communication and interpretation.

8. Limitations The selection of the LE group was a non-randomized convenience sample, and the results of this study can only be generalized to a population of similar characteristics. Sampling bias was minimized through the use of age, gender, and dominance-matched control subjects. The examiner was not blinded to the subject group assignment, and this may possibly have influenced the subjects’ performance. Strength or endurance limitations may affect the results of this study; however, both tests are short and non-strenuous tests of finger dexterity. Motivational factors are a potential factor in any test of human performance. The Purdue Pegboard and the CMDT tests are well standardized and structured as to the verbal instructions given to each subject; however, the response of subjects to the standardized instructions could vary.

9. Clinical implications Failure to recognize or address problems with fine motor control may be one explanation for the persistence of difficult cases of LE that are resistant to existing treatment approaches usually aimed at the peripheral musculotendinous structures. The PPT and the CMDT are commonly used clinical assessment tools, and a practical method for clinicians working with patients with LE to similarly identify if there are problems with motor control. This factor could then be addressed in the patient’s rehabilitation programme. The PPT and the CMDT appear equally sensitive in detecting differences in fine motor control ability in these subjects with LE; however, the PPT is less expensive and time consuming for the practicing clinician. The PPT and CMDT are measures of impairment, and it is not known how the observed deficits would affect the individual functionally. A study examining the relationship between and the degree of impairment and the level of functional disability would be helpful. The results of this study do not explain or imply causation for the decrease in fine motor control; however, this study may offer some explanation of why some subjects complain of clumsiness in their affected limb. The possibility of dysfunction in neuromuscular control should be considered in assessment and treatment of repetitive strain disorders (Byl et al., 2000a, b; McKenzie et al., 2003; Barr et al., 2004). Measurement of fine motor control may help select subjects with LE who



could benefit from specific motor control training. It is not known as yet, however, if improvements in fine motor control will have a direct impact on the resolution of LE, and this warrants further research.

Acknowledgements I am very thankful to my advisor, Sandra Curwin Ph.D., for her encouragement and advice throughout this research project. I thank my committee members and Bonnie Dobbs, Ph.D. for the important contribution of their time and experience. The support of management and clinicians from the Workers Compensation Board of Alberta, Orion Health, Sun City Physiotherapy, and Columbia Health was invaluable in making this research project possible.

References Assendelft WJ, Hay EM, et al. Corticosteroid injections for lateral epicondylitis: a systematic overview. British Journal of General Practice 1996;46(405):209–16. Barbe MF, Barr AE, et al. Chronic repetitive reaching and grasping results in decreased motor performance and widespread tissue responses in a rat model of MSD. Journal of Orthopaedic Research 2003;21(1):167–76. Barr AE, Barbe MF. Pathophysiological tissue changes associated with repetitive movement: a review of the evidence. Physical Therapy 2002;82(2):173–87. Barr AE, Barbe MF, et al. Work-related musculoskeletal disorders of the hand and wrist: epidemiology, pathophysiology, and sensorimotor changes. Journal of Orthopaedic Sports and Physical Therapy 2004;34(10):610–27. Bowen RE, Dorey FJ, et al. Efficacy of nonoperative treatment for lateral epicondylitis. American Journal of Orthopaedics 2001;30(8): 642–6. Boyer MI, Hastings II H. Lateral tennis elbow: is there any science out there? Journal of Shoulder Elbow Surgery 1999;8(5):481–91. Burgess RC. Tennis elbow. Journal of Ky Medical Association 1990;88(7):349–54. Byl N. Focal hand dystonia may result from aberrant neuroplasticity. Advances in Neurology 2004;94:19–28. Byl NN, Melnick M. The neural consequences of repetition: clinical implications of a learning hypothesis. Journal of Hand Therapy 1997;10(2):160–74. Byl N, Wilson F, et al. Sensory dysfunction associated with repetitive strain injuries of tendinitis and focal hand dystonia: a comparative study. Journal of Orthopaedic Sports and Physical Therapy 1996a;23(4):234–44. Byl NN, Merzenich MM, et al. A primate genesis model of focal dystonia and repetitive strain injury: I. Learning-induced dedifferentiation of the representation of the hand in the primary somatosensory cortex in adult monkeys. Neurology 1996b;47(2): 508–20. Byl NN, Merzenich MM, et al. A primate model for studying focal dystonia and repetitive strain injury: effects on the primary somatosensory cortex. Physical Therapy 1997;77(3):269–84. Byl N, Nagarajan SS, et al. Effect of sensory discrimination training on structure and function in a musician with focal hand dystonia. Physical Therapy Case Report 2000a;3:94–113.

Byl NN, McKenzie A, et al. Differences in somatosensory hand organization in a healthy flutist and a flutist with focal hand dystonia: a case report. Journal of Hand Therapy 2000b;13(4): 302–9. Chiang HC, Ko YC, et al. Prevalence of shoulder and upperlimb disorders among workers in the fish-processing industry. Scandinavian Journal of Work Environment and Health 1993; 19(2):126–31. Coonrad RW, Hooper WR. Tennis elbow: its course, natural history, conservative and surgical management. Journal of Bone Joint Surgery America 1973;55(6):1177–82. Dimberg L, Olafsson A, et al. The correlation between work environment and the occurrence of cervicobrachial symptoms. Journal of Occupational Medicine 1989;31(5):447–53. Elbert T, Candia V, et al. Alteration of digital representations in somatosensory cortex in focal hand dystonia. Neuroreport 1998;9(16):3571–5. Ervilha UF, Arendt-Nielsen L, et al. The effect of muscle pain on elbow flexion and coativation tasks. Experiments in Brain Research 2004;156(2):174–82. Flor H. Cortical reorganisation and chronic pain: implications for rehabilitation. Journal of Rehabilitation Medicine 2003; 35(Suppl 41):66–72. Giamberardino MA. Referred muscle pain/hyperalgesia and central sensitisation. Journal of Rehabilitation Medicine 2003;35 (Suppl 41):85–8. Green S, Buchbinder R, et al. Non-steroidal anti-inflammatory drugs (NSAIDs) for treating lateral elbow pain in adults. Cochrane Database System Review 2002(2):CD003686. Haahr JP, Andersen JH. Prognostic factors in lateral epicondylitis: a randomized trial with one-year follow-up in 266 new cases treated with minimal occupational intervention or the usual approach in general practice. Rheumatology (Oxford) 2003;42(10): 1216–25. Hamm N, Curtis D. Normative data for the purdue pegboard on a sample of adult candidates for vocational rehabilitation. Perceptual and Motor Skills 1980;50:309–10. Jarvinen M, Jozsa L, et al. Histopathological findings in chronic tendon disorders. Scandinavian Journal of Medicine and Science in Sports 1997;7(2):86–95. Jenkins WM, Merzenich MM, et al. Functional reorganization of primary somatosensory cortex in adult owl monkeys after behaviorally controlled tactile stimulation. Journal of Neurophysiology 1990;63(1):82–104. Kivi P. Rheumatic disorders of the upper limbs associated with repetitive occupational tasks in Finland in 1975–1979. Scandinavian Journal of Rheumatology 1984;13(2):101–7. Kurppa K, Viikari-Juntura E, et al. Incidence of tenosynovitis or peritendinitis and epicondylitis in a meat-processing factory. Scandinavian Journal of Work Environment and Health 1991; 17(1):32–7. Labelle H, Guibert R, et al. Lack of scientific evidence for the treatment of lateral epicondylitis of the elbow. An attempted meta-analysis. Journal of Bone and Joint Surgery—British Volume 1992;74(5):646–51. Lund JP, Donga R, et al. The pain-adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of Physiology and Pharmacology 1991;69(5):683–94. McKenzie AL, Nagarajan SS, et al. Somatosensory representation of the digits and clinical performance in patients with focal hand dystonia. American Journal of Physical Medicine and Rehabilitation 2003;82(10):737–49. Melborn JM. Cumulative trauma disorders and repetitive strain injuries. Clinical Orthopaedics and Related Research 1998; 351:107–28.

ARTICLE IN PRESS D.K. Skinner, S.L. Curwin / Manual Therapy 12 (2007) 249–255 Moore JS. Biomechanical models for the pathogenesis of specific distal upper extremity disorders. American Journal of Industrial Medicine 2002;41(5):353–69. Nirschl RP, Ashman ES. Elbow tendinopathy: tennis elbow. Clinical Sports in Medicine 2003;22(4):813–36. Pantev C, Engelien A, et al. Representational cortex in musicians. Plastic alterations in response to musical practice. Annals of the New York Academy of Sciences 2001;930:300–14. Pienimaki TT, Kauranen K, et al. Bilaterally decreased motor performance of arms in patients with chronic tennis elbow. Archives of Physical Medicine and Rehabilitation 1997; 78(10):1092–5. Russo CM, Brose WG. Chronic pain. Annual Review of Medicine 1998;49:123–33.

255

Smeulders MJ, Kreulen M, et al. Fine motor assessment in chronic wrist pain: the role of adapted motor control. Clinical Rehabilitation 2001;15(2):133–41. Smidt N, van der Windt DA, et al. Corticosteroid injections, physiotherapy, or a wait-and-see policy for lateral epicondylitis: a randomised controlled trial. Lancet 2002;359(9307):657–62. Smidt N, Assendelft WJ, et al. Effectiveness of physiotherapy for lateral epicondylitis: a systematic review. Annali Medici 2003; 35(1):51–62. Starcke AM. Managing the elusive beast of chronic pain: a challenge for PTs. Magazine of Physical Therapy 2005;13(11):48–52. Viikari-Juntura E, Hietanen M, et al. Psychomotor capacity and occurrence of wrist tenosynovitis. Journal of Occupational Medicine 1994;36(1):57–60.

ARTICLE IN PRESS


Original article

The diagnostic validity of the cervical flexion–rotation test in C1/2-related cervicogenic headache Mark Ogince, Toby Hall, Kim Robinson, A.M. Blackmore School of Physiotherapy, Curtin University of Technology, C/o 54 Bunya Street, Noranda, Perth, WA 6062, Australia Received 8 March 2005; received in revised form 21 March 2006; accepted 27 June 2006

Abstract This single-blind comparative group design aimed to investigate the sensitivity and specificity of the cervical flexion–rotation test in the diagnosis of C1/2-related cervicogenic headache. This study tested 23 cervicogenic headache, 23 asymptomatic controls and 12 migraine with aura subjects, all aged 18–66 years. In stage 1, an experienced manipulative physiotherapist who did not partake in the flexion–rotation test procedure identified C1/2 dysfunction using passive segmental mobility tests in the cervicogenic headache group. Those with C1/2 dysfunction participated in stage 2. In stage 2, using the flexion–rotation test, subjects were tested by two experienced manipulative physiotherapists blinded to the subjects’ group allocation. Each therapist stated whether the test was positive or not based on the therapist’s interpretation of range of motion. The sensitivity and specificity of the flexion–rotation test was 91% and 90%, respectively (Po.001), with an overall diagnostic accuracy of 91% (Po.001). The cervical flexion–rotation test significantly assists in the differential diagnosis of cervicogenic headache and in the identification of movement impairment at the C1/2 segment in patients with cervicogenic headache. r 2006 Published by Elsevier Ltd. Keywords: Sensitivity; Specificity; Manual examination; C1/2 segment

1. Introduction Cervicogenic headache has been identified as a distinct subgroup by the International Headache Society (IHS) (Headache Classification Committee of the International Headache Society, 2004) and is caused by disease or dysfunction of structures in the neck (Edmeads, 2001). The diagnostic criteria for cervicogenic headache outlined by the IHS include subjective features together with evidence of impairment of cervical function on physical examination. Such impairment includes atlanto-axial motion segment (C1/2 level) dysfunction (Hall and Robinson, 2004) identified by the flexion–rotation test. Corresponding author. Tel./fax: +61 8 9375 3224.

E-mail address: [email protected] (M. Ogince). 1356-689X/$ - see front matter r 2006 Published by Elsevier Ltd. doi:10.1016/j.math.2006.06.016

Headache presents a diagnostic challenge due to similarities of signs and symptoms among the many types of headache (Nicholson and Gaston, 2001). In particular, distinguishing between cervicogenic headache and migraine is problematic (Lewit, 1977, Sjaastad and Bovim, 1991, Vernon et al., 1992; Blau and MacGregor, 1994). The IHS subjective classification criteria are commonly used to classify cervicogenic headache however this process fails to identify the segmental source of pain, which is important for the application of manual therapy treatment. Radiological examinations are not effective (Jensen et al., 1990; Edmeads, 2001), and nerve block procedures are often impractical, particularly in the upper cervical region (Bogduk et al., 1985). Impaired neck mobility is a diagnostic criterion in cervicogenic headache (Mersky and Bogduk, 1994; Sjaastad et al., 1998), however several studies have

ARTICLE IN PRESS M. Ogince et al. / Manual Therapy 12 (2007) 256–262

found that active cervical mobility is unreliable in differential diagnosis (Jull et al., 1988; Jensen et al., 1990; Treleaven et al., 1994; Sandmark and Nisell, 1995; Placzek et al., 1999; Hall and Robinson, 2004). Conversely, manual examination has been shown to detect symptomatic cervical joint dysfunction in a number of studies of cervical headache (Jull et al., 1988; Jaeger, 1989; Jensen et al., 1990; Watson and Trott, 1993; Treleaven et al., 1994; Whittingham et al., 1994; Schoensee et al., 1995). Clinically, these tests, described by Maitland et al. (2001), are used as the current reference standard; however, they involve a high degree of skill on the part of the therapist. Additionally, most external measures of cervical motion incorporate movements of both the upper and lower cervical regions simultaneously (Amiri et al., 2003). However, as cervicogenic headache has a primary involvement in the upper cervical segments, measurement of rotation purportedly biased to the upper cervical region could be a relevant clinical measure in the differential diagnosis of cervicogenic headache (Amiri et al., 2003). Furthermore, determining the dysfunctional cervical segment facilitates a more accurate treatment approach. The cervical flexion–rotation test is an easily applied method of manual examination that is said to provide a means of determining the presence of joint dysfunction at the C1/2 level (Stratton and Bryan, 1994). The flexion– rotation test is conducted with the cervical spine fully flexed in an attempt to block as much rotational movement as possible above and below C1/2. The head is then rotated to the left and the right. If firm resistance is encountered and range is limited before the expected end range, then this is said to be significant, with a presumptive diagnosis of limited rotation of the atlas on the axis (Stratton and Bryan, 1994). Anecdotally, pain provocation during the flexion–rotation test is also a feature of a positive test result, however, pain is not a feature in asymptomatic subjects (Hall and Robinson, 2004). It is deducible from the frequently reported overlapping characteristics seen with cervicogenic headache and migraine that in the clinical realm many cervicogenic patients are currently misdiagnosed as having migraine headache and migraine patients misdiagnosed as having cervicogenic headache (Sjaastad and Bovim, 1991). Consequently, it is likely that treatment is unsubstantiated and a poor prognosis will follow. Thus, there appears to be a need to identify physical tests that are valid, reliable and sensitive in assisting the diagnosis of cervicogenic headache. Accordingly, the aims of this study were to determine the sensitivity and specificity of the cervical flexion– rotation test. An additional aim was to determine if a relationship exists between cervicogenic headache severity and the extent of restriction demonstrated by the flexion–rotation test.

257

2. Methods 2.1. Subjects A single-blind comparative group design was used to determine differences between asymptomatic subjects, migraine with aura subjects and those with C1/2-related cervicogenic headache. Based on the previous findings of Hall and Robinson (2004), in order to detect a 101 difference of rotation with the cervical spine in flexion, with alpha at .05, power of 80% and a standard deviation of 81, this study required at least 10 subjects per group. In total, 23 cervicogenic headache subjects (3 males, 20 females, mean age ¼ 46 years), 23 asymptomatic controls (8 males, 15 females, mean age ¼ 40 years) and 12 migraines with aura subjects (9 males, 3 females, mean age ¼ 37 years) participated in the study. Subjects ranged in age from 18 to 66 years. This study had approval by the appropriate Human Research Ethics Committee.

3. Materials The cervical range of motion device (CROM) (Performance Attainment Associates. 958 Lydia Drive, Roseville, Minnesota, USA. 55113) was modified to measure cervical rotation in a fully flexed cervical spine position. Two Velcro straps were fixed to the subject’s head, traversing the transverse and coronal planes, respectively (Fig. 1). The CROM goniometer was attached to the centre of the coronal Velcro strap to measure cervical rotation in maximal flexion (Fig. 1). The CROM and modified CROM have been shown to have good intratester and intertester reliability (Capuano-Pucci and Rheault, 1991; Hall and Robinson, 2004; Rheault and Albright, 1992). A headache questionnaire

Fig. 1. The cervical range of motion device (modified to measure cervical rotation in flexion) and the flexion–rotation test.


M. Ogince et al. / Manual Therapy 12 (2007) 256–262

(Niere and Robinson, 1997) was used to assess the severity of headache. 3.1. Procedures Subjects were recruited from medical specialists and by advertisements. In response, 325 headache and 23 asymptomatic subjects expressed interest in participation. Using the IHS (Headache Classification Committee of the International Headache Society, 1988) subjective diagnostic criteria as well as criteria outlined by Sjaastad et al. (1998), Bogduk (1994) and Lord et al. (1994), telephone interviews were conducted to identify subjects with cervicogenic headache, migraine with aura headache and asymptomatic controls. The inclusion criteria for the cervicogenic subjects were unilateral or side dominant headache without sideshift, (Sjaastad et al., 1998) headache associated with neck pain or reported stiffness (Bogduk, 1994; Sjaastad et al., 1998) neck symptoms preceding or co-existent with the onset of headache (Lord et al., 1994; Sjaastad et al., 1998) and pain precipitated or aggravated by specific neck movements or sustained posture (Headache Classification Committee of the International Headache Society, 1988). Additionally, passive segmental mobility tests reveal symptomatic C1/2 dysfunction (Maitland et al., 2001) as well as headache frequency at least an average of one per week and history of episodic semicontinuous or continuous headache for at least the previous 3 months. The exclusion criteria for the cervicogenic headache group were diagnostic criteria of headache that are not of cervical origin (Headache Classification Committee of the International Headache Society, 1988). Asymptomatic subjects had no history of subjective features of cervicogenic headache, migraine, migraine with aura headache, episodic headache, and neck pain or stiffness. The following exclusion criteria related to all 3 groups; dizziness on cervical spine movement, inability to tolerate the flexion–rotation test position, failure to provide informed consent, insufficient fluency in English, known congenital fusion, history of cervical surgery, cervical or cranial trauma, rheumatoid arthritis and Downs Syndrome. The inclusion criteria for the migraine with aura group were based on the IHS classification (Headache Classification Committee of the International Headache Society, 1988). The exclusion criteria for this group were neck pain, discomfort or stiffness. Of the 325 symptomatic subjects interviewed, 82% reported overlapping cervicogenic and migraine symptoms and were rejected from the study. The remaining symptomatic subjects consisted of 12 migraine with aura and 46 cervicogenic subjects. As a sample of convenience 23 cervicogenic, 23 asymptomatic and 12 migraine with aura subjects who matched the criteria

were recruited into the study. Telephone interviews and the allocation of subjects to their respective groups were conducted by the principal investigator who did not partake in the flexion–rotation test procedures. Cervicogenic headache subjects were required to attend a preliminary session to determine the presence of C1/2 dysfunction. This was defined as stage 1. Thereafter, all subjects attended a single session for data collection, defined as stage 2. Prior to the start of the study all subjects were required to provide informed consent. In stage 1, an experienced teacher of manipulative physiotherapy, with 12 years experience who did not take part in the flexion rotation test procedure, identified 23 out of 34 cervicogenic headache subjects with C1/2 as the dominant level of dysfunction using passive segmental mobility tests. This method of assessment has been shown to be a valid means of identifying the symptomatic cervical level in a previous study (Jull et al., 1988). These 23 subjects were then allocated to the cervicogenic headache group. The flexion–rotation test was not used in this part of the assessment. Segmental mobility was assessed using passive accessory and physiological intervertebral movements (PAIVMs and PPIVMs) as described by Maitland et al. (2001). In congruence with the clinical setting, assessing segmental dysfunction using PAIVMs and PPIVMs served as the reference standard in this study. In stage 2, using the flexion–rotation test subjects were tested by two experienced manipulative physiotherapists who had 16 and 13 years postgraduate manipulative therapy experience and were blinded to the subjects’ group allocation. The flexion–rotation test was conducted with the subject relaxed and recumbent. The cervical spine was fully flexed with the occiput resting against the examiner’s abdomen (Fig. 1). The head was then rotated to the left and right. Each therapist was required to interpret the flexion–rotation test results and state whether the test was positive or not, based on the therapist’s interpretation of range of motion. The range of motion was considered limited when the therapist determined a firm end feel and based on the therapists interpretation there was a minimum of a 101 reduction from the expected range. Pain provocation during the flexion–rotation test was not tested in this study. Thereafter, the goniometer was fixed and the flexion– rotation test was repeated twice in each direction, with one therapist ascertaining range in reverse order and measurements recorded on each occasion. An interval of 30 s elapsed between each trial. Additionally, each therapist was blinded to each other’s interpretation and recordings. At the time of assessment, cervicogenic and migraine subjects completed a headache questionnaire (Niere and Robinson, 1997) detailing their history of headache frequency, intensity and duration. This questionnaire


has been shown to be reliable and enables an index of headache severity to be calculated (Niere and Robinson, 1997). Additionally, the intensity of headache at the time of the assessment was determined on a 10 cm visual analogue scale (VAS). The results of the questionnaire and VAS were used to determine if the degree of restriction of rotation in flexion was related to the severity of the headache symptoms in the cervicogenic headache group.

259

Table 1 The frequencies used to calculate sensitivity and specificity Cervicogenic headache Positive Cervical flexion Rotation test a

Positive Negative Total

a

21/21 2/2a 23/23a

Total

Negative 3/4a 32/31a 35/35a

24/25a 34/33a 58

Therapist 1/Therapist 2.

3.2. Statistical analysis

4. Results The average range of unilateral rotation for both sides was 391 (SD ¼ 6.9), 391 (SD ¼ 6.5) and 201 (SD ¼ 11) for the migraine with aura, asymptomatic and cervicogenic headache groups, respectively. For the cervicogenic headache group the average range of unilateral rotation refers to the most restricted side. The difference between groups was significant (Po.001) (Fig. 2). The results indicate that the range of rotation was significantly reduced in the cervicogenic headache group when

Range of Motion (degrees)

50

All data were analysed using SPSS Version 11.0 statistical software (SPSS, Inc., Chicago, IL). In all cases alpha was set at the .05 level. An analysis of variance (ANOVA) and planned orthogonal comparisons were used to analyse range of rotation with the cervical spine in flexion between the 3 subject groups. An ANCOVA was used to determine whether age and gender accounted for the difference in range of rotation with the cervical spine in flexion between the 3 groups. The sensitivity and specificity of the flexion–rotation test were analysed using cross tabulation and were determined with a receiver operating characteristic (ROC) curve. To calculate the sensitivity and specificity the migraine with aura and asymptomatic groups were combined and then compared to the cervicogenic headache group. The frequencies used to calculate sensitivity and specificity are given in Table 1. The dichotomous variables used to determine the sensitivity and specificity were the therapists’ identification of the presence or absence of C1/2 dysfunction and thereby cervicogenic headache. The ROC curve was created with the flexion–rotation range of motion values. The intertester reliability of the flexion–rotation test was calculated from a cross-tabulation using kappa. A variable headache severity index was calculated using the method described by Niere and Robinson (1997). The greater the score on the headache severity index the greater the severity of headache. To determine the relationship of the headache severity index score and the VAS score to the range of rotation in flexion, Pearson’s correlation analyses were used.

40

30

20

10

0 Cervicogenic

Migraine with aura

Asymptomatic

Fig. 2. The mean range of motion and 95% confidence interval of unilateral cervical rotation in maximal flexion to each side for each subject group.

compared to the migraine with aura and asymptomatic subjects (Po.001). There was no significant difference in range between the migraine with aura subjects and asymptomatic subjects (P ¼ :971) (Fig. 2). There was a significant negative correlation between age and ROM (r ¼ :404, P ¼ :002), and the difference in ROM between males (M ¼ 35:1, SD ¼ 9.2) and females (M ¼ 29:7, SD ¼ 13.7) was close to significance, tð52Þ ¼ 1:76, P ¼ :084. Therefore, age and gender were both included as covariates in an ANCOVA, and a significant result on ANCOVA (Po.001) revealed that neither age nor gender accounted for the differences between groups. Sensitivity is the test’s ability to obtain a positive test when the target condition is really present (Portney and Watkins, 1993). Specificity is the test’s ability to obtain a negative test when the condition is really absent (Portney and Watkins, 1993). Positive predictive value is the likelihood that person who tests positive actually has the disease and negative predictive value is the probability that a person who tests negative is actually disease-free (Portney and Watkins, 1993). A positive likelihood ratio indicates the increase in odds favouring the condition given a positive test result and a negative


260

Table 2 The sensitivity, specificity, positive and negative predictive values as well as likelihood ratios of the cervical flexion–rotation test

Sensitivity (%) Specificity (%) Positive predictive value (%) Negative predictive value (%) Positive likelihood ratio Negative likelihood ratio

Therapist 1

Therapist 2

91.3 91.4 87.5 94.1 10.65 0.095

91.3 88.6 84.0 93.9 7.99 0.098

the ROC curve indicated that the test value, which provides the highest sensitivity and the lowest 1-specificy, is 321 (cut-off score). That is, if the flexion rotation-test value is less than or equal to 321 the test is positive. In examining the relationship between the headache severity index score and the range of rotation in flexion, there was no significant correlation rð23Þ ¼ :24, P ¼ :27. Additionally, when examining the relationship between severity of headache at the time of assessment on a VAS to range of rotation in flexion, there was no significant correlation rð23Þ ¼ :09, P ¼ :960.

1.00

5. Discussion

Sensitivity

0.75

0.50

0.25

0.00 0.00

0.2

0.50

0.75

1.00

1 – Specificity Fig. 3. The diagnostic accuracy of the cervical flexion–rotation test. The area under the curve is .91 (Po.001, CI ¼ 82%, CI ¼ 100%).

likelihood ratio indicates the change in odds favouring the condition given a negative test result (Irwig et al., 1994). Table 2 outlines the sensitivity, specificity, positive predictive value, and negative predictive value as well as likelihood ratios for each therapist. The mean sensitivity and specificity for the two therapists were 91% and 90%, respectively (Po.001). Cross tabulation revealed that 98.3% of the time both testers agreed on their identification of C1/2 movement restriction and the absence or presence of cervicogenic headache (Po.001). The kappa value for the cervical flexion–rotation test was .81, indicating excellent agreement (Landis and Koch, 1977). The ROC curve (Fig. 3) shows the relationship between sensitivity and specificity. The area under the curve represents the ability of the test to discriminate between the diseased and nondiseased state. An ROC curve revealed that presented with a randomly chosen pair of patients, using the flexion–rotation test, the clinician is able to make the correct diagnosis 91% of the time (Po.001) (Fig. 3). Additionally, coordinates on

This study found the range of cervical rotation in flexion was significantly reduced in the presence of C1/2 dominant cervicogenic headache when compared to a control group of either asymptomatic or migraine with aura. These results concur with Hall and Robinson (2004) who evaluated the flexion–rotation test comparing only cervicogenic headache subjects and asymptomatic controls. Hall and Robinson (2004) demonstrated that the average range of unilateral rotation to each side was 281 and 451 for the cervicogenic and asymptomatic groups, respectively. This study showed that average range of unilateral rotation was 201 and 391 for the cervicogenic and asymptomatic groups, respectively. Additionally, the range of rotation in flexion in asymptomatic controls is comparable with Amiri et al. (2003) who found the average range of unilateral rotation was 421. This study also found there was no difference in range of flexion between the migraine with aura and the asymptomatic subjects. Migraine with aura has been described as a disturbance of brain function or a neurovascular event (Sanchez-del-Rio and Reuter, 2004) and consequently does not involve impairment of the cervical spine. Hence, this result was not unexpected and helps to confirm the lack of cervical involvement in headache with aura patients. The results of this study suggest that the clinical method purported to bias rotation to the upper cervical region is an accurate and reliable clinical measure in the identification of dysfunction at the C1/2 level and in the differential diagnosis of cervicogenic headache. Thus, in accordance with Amiri et al. (2003) it would be reasonable to recommend that the flexion–rotation test be used in the assessment of patients with C1/2 dysfunction, for purposes of differential diagnosis and assessment of treatment outcomes. This is the first study to establish the sensitivity and specificity of the cervical flexion–rotation test in cervicogenic headache diagnosis. The average sensitivity and specificity was 91% and 90%, respectively, with an


overall diagnostic accuracy of 91%. This indicates that the cervical flexion–rotation test has very high accuracy in determining the presence of C1/2 involvement in cervicogenic headache and for headache differential diagnosis. Furthermore, the average kappa statistic of .81 suggests excellent therapist agreement (Landis and Koch, 1977). Bogduk (1997) concludes that for most good clinical tests in physical examination, the kappa value should range between .4 and .6. This study has also determined a range of 321 as the cut-off value at which the flexion rotation-test is deemed positive. This enables the clinician to confidently interpret the results of the test in clinical practice. Consequently, these results help to establish the flexion–rotation test as a reliable measure that assists in differential headache diagnosis and determining the presence of C1/2 dysfunction. It is important to note that at the time of testing each therapist interpreted the flexion–rotation test results by stating whether the test was positive or not, based on the therapist’s interpretation of range of motion. That is, the range of motion was considered limited when the therapist determined a firm end feel and based on the therapists interpretation there was a minimum of a 101 reduction from the expected range. Since it is impractical to use a CROM device in the clinical environment this method more accurately reflects those used by physiotherapists in the clinical environment. In contrast to Hall and Robinson (2004), this study found that the severity of cervicogenic headache is not related to the degree of restriction of rotation with the cervical spine in maximal flexion. Given that the sample of cervicogenic headache subjects was similar in both studies it appears that this difference may be due to alternative factors. Silberstein et al. (2001) contend that many patients are not good observers of their own complaints, even when those complaints are chronic. Certainly, the authors observed that whilst subjects completed the questionnaire they expressed difficulty in recalling the intensity, frequency and duration of symptoms. This may have skewed the results of the questionnaire and thus account for the different results in this study. Using the VAS to assess the influence of intensity of headache on range of rotation at the time of testing, this study indicated that the severity of headache at the time of assessment is not related to the degree of restriction with the cervical spine in flexion. Clearly, this appears to be advantageous as the diagnostic accuracy of the flexion–rotation test is not influenced by the patient’s head pain at the time of assessment. When comparing the cervicogenic (n ¼ 23) and asymptomatic groups (n ¼ 23), we were only able to recruit 12 subjects in the migraine with aura group (n ¼ 12). This sample corresponds to Ziegler and Hassanein (1990) who showed that at most, only 30%

261

of the migraine population have migraine with aura. Of the symptomatic subjects interviewed, 82% reported overlapping cervicogenic and migraine symptoms and thus were rejected from the study. This data corresponds with Sjaastad and Bovim (1991), Fishbain et al. (2001) and Nicholson and Gaston (2001) who report frequent overlapping characteristics seen with cervicogenic and migraine patients. This study’s population was limited to cervicogenic and migraine with aura patients who have symptoms characteristic of no overlap. Based on the study sample, it appears that those who participated represent a small percentage of the symptomatic population who present in the clinical setting. The study’s case control design and spectrum bias is by no means a limitation, as in order to determine the diagnostic accuracy of physical test the subjects must represent either a purely positive disease state, that is, cervicogenic headache or symptoms manifest of a disease free state. Thus, this study has demonstrated the cervical flexion–rotation test significantly assists in C1/2 cervicogenic headache diagnosis when the subjective criteria follow those of the IHS. Consequently, this study sets the groundwork for future studies to evaluate the cervical flexion–rotation test on a subject population whose symptoms overlap cervicogenic and migraine headache. The authors acknowledge a number of limitations of the study. In particular, the use of a single assessor to identify C1/2 dysfunction using manual diagnosis. At the time of this study no credible alternative reference standard was available for determining the segmental level of involvement for the high cervical spine. The only other potential reference standard, double blind anaesthetic blocks, has not been attempted in the high cervical spine. In addition, selection bias was unavoidable in obtaining a large sample size. Not all subjects presenting with the relevant condition were included in order of entry neither was the selection random.

6. Conclusion This study has demonstrated that the cervical flexion– rotation test is extremely reliable and has high sensitivity and specificity in detecting the presence of C1/2 rotation restriction in patients with cervicogenic headache. The cut of value for a positive test is range of rotation less than 321. The flexion–rotation test is a simple, noninvasive test that can be easily applied in the clinical setting.

Acknowledgements The authors thank Mr. Wim Dankaerts for assisting with the testing procedures. The authors also acknowl-


M. Ogince et al. / Manual Therapy 12 (2007) 256–262

edge the School of Physiotherapy at Curtin University of Technology for providing assistance to present the research findings at the WCPT Conference in Barcelona 2003 as well as the Australian Physiotherapy Association WA Branch for awarding the MJ Rosen Scholarship to present this research at the Annual Conference of the APTA in Chicago 2004.

References Amiri M, Jull G, Bullock-Saxton J. Measuring range of active cervical rotation in a position of full head flexion using the 3D Fastrak measurement system: an intra-tester reliability study. Manual Therapy 2003;8:176–9. Blau J, MacGregor E. Migraine and the neck. Headache 1994;34:88–90. Bogduk N. Cervical causes of headache and dizziness. In: Boyling J, Palastanga N, editors. Grieves modern manual therapy: the vertebral column. Edinburgh: Churchill Livingstone; 1994. p. 317–32. Bogduk N. Musculoskeletal pain: toward precision diagnosis. In: Proceedings of the eighth world congress on pain. Seattle: IASP Press; 1997. p. 507–26. Bogduk N, Corrigan B, Kelly P. Cervical headache. Medical Journal of Australia 1985;143:202–7. Capuano-Pucci D, Rheault W. Intratester reliability of the cervical range of motion device. Archives of Physical Medicine and Rehabilitation 1991;72:338–40. Edmeads J. Disorders of the neck: cervicogenic headache. In: Silberstein SD, Lipton RB, Dalessio DJ, editors. Wolff’s headache and other head pain. Oxford: Oxford University Press; 2001. p. 447–58. Fishbain D, Cutler R, Cole B, Rosomoff H, Rosomoff R. International Headache Society headache diagnostic patterns in pain facility patients. Clinical Journal of Pain 2001;17:78–93. Hall T, Robinson K. The flexion–rotation test and active cervical mobility—a comparative measurement study in cervicogenic headache. Manual Therapy 2004;9:197–202. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and fascial pain. Cephalalgia 1988;8: 1–96. International Headache Society. The International classification of headache disorders 2nd edition. Cephalalgia 2004;24(Suppl 1):9–160. Irwig L, Tosteson A, Gatsonis C. Guidelines for metanalyses evaluating diagnostic tests. Annals of Internal Medicine 1994;120: 667–76. Jaeger B. Are cervicogenic headaches due to myofascial pain and cervical spine dysfunction? Cephalalgia 1989;9:157–64. Jensen O, Nielsen F, Vosmar L. An open study comparing manual therapy with the use of cold packs in the treatment of post concussional headache. Cephalalgia 1990;10:241–9. Jull G, Bogduk N, Marsland A. The accuracy of manual diagnosis for cervical zygapophyseal joint pain syndromes. Medical Journal of Australia 1988;148:233–6. Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74.

Lewit K. Pain arising in the posterior arch of the Atlas. European Neurology 1977;16:263–9. Lord S, Barnsely L, Wallis B, Bogduk N. Third occipital headache: a prevalence study. Journal of Neurology, Neurosurgery and Psychiatry 1994;57:1187–90. Maitland G, Hengeveld E, Banks K, English K. Vertebral manipulation, 6th ed. London: Butterworth Heinemann; 2001. Mersky H, Bogduk N. Classification of chronic pain. 2nd ed. Seattle: IASP Press; 1994. Nicholson G, Gaston J. Cervical headache. Journal of Orthopedic and Sports Physical Therapy 2001;31:184–93. Niere K, Robinson P. Determination of manipulative physiotherapy treatment outcome in headache patients. Manual Therapy 1997;2: 199–205. Placzek J, Pagett B, Roubal P, Jones B, McMichael H, et al. The influence of cervical spine on chronic headache in women: a pilot study. The Journal of Manual and Manipulative Therapy 1999;7: 33–9. Portney G, Watkins M. Foundation of clinical research: applications to practice. London: Appleton and Lange; 1993. Rheault W, Albright B. Intertester reliability of the cervical range of motion device. Journal of Orthopaedic and Sports Physical Therapy 1992;15:147–50. Sanchez-del-Rio M, Reuter U. Migraine aura: new information on underlying mechanisms. Current Opinion in Neurology 2004;17(3): 289–93. Sandmark H, Nisell R. Validity of five common manual neck pain provocation tests. Scandinavian Journal of Rehabilitation Medicine 1995;27:131–6. Schoensee H, Jensen G, Nicholson G, Gossman M, Katholi C. The effect of mobilisation on cervical headaches. Journal of Orthopaedic and Sports Physical Therapy 1995;21:181–96. Silberstein S, Lipton R, Dalessio D. Overview, diagnosis and classification of headache. In: Silberstein S, Lipton R, Dalessio D, editors. Wolff’s headache and other head pain. Oxford: Oxford University Press; 2001. p. 6–26. Sjaastad O, Bovim G. Cervicogenic headache: the differentiation from common migraine, an overview. Functional Neurology 1991;6: 93–100. Sjaastad O, Fredriksen T, Pfaffenrath V. Cervicogenic headache: diagnostic criteria. Headache 1998;38:442–5. Stratton S, Bryan J. Dysfunction, evaluation, and treatment of the cervical spine and thoracic inlet. In: Donatelli R, Wooden M, editors. Orthopaedic physical therapy. New York: Churchill Livingstone; 1994. p. 77–122. Treleaven J, Jull G, Atkinson L. Cervical musculoskeletal dysfunction in post-concussional headache. Cephalalgia 1994;14:273–9. Vernon H, Steiman I, Hagino C. Cervicogenic dysfunction in muscle contraction headache and migraine: a descriptive study. Journal of Manipulative and Physiological Therapeutics 1992;15:418–29. Watson D, Trott P. Cervical headache. An investigation of natural head posture and upper cervical flexor muscle performance. Cephalalgia 1993;13:272–84. Whittingham W, Ellis W, Molyneux T. The effect of manipulation (toggle recoil techniques) for headaches with upper cervical joint dysfunction: a pilot study. Journal of Manipulative and Physiological Therapeutics 1994;17:369–75. Ziegler D, Hassanein R. Specific headache phenomena: their frequency and coincidence. Headache 1990;30:152–6.

ARTICLE IN PRESS


Original article

Accuracy and reliability of observational motion analysis in identifying shoulder symptoms Brendan W. Hickeya, Stephan Milosavljevica,, Melanie L. Bellb, Peter D. Milburna a

Centre for Physiotherapy Research, School of Physiotherapy, University of Otago, P.O. Box 56, Dunedin, New Zealand b Department of Preventive and Social Medicine, University of Otago, P.O. Box 56, Dunedin, New Zealand Received 22 December 2004; received in revised form 20 February 2006; accepted 8 May 2006

Abstract Introduction: Aberrations in shoulder movement patterns are believed to be associated with the presence of shoulder symptoms. However, the detection of movement aberrations has not been rigorously investigated. It is possible that manipulative physiotherapists use the clinical history to prejudge the existence of aberrations, rather than the actual observation of the movement pattern itself. There is a need to determine whether physiotherapists, in the absence of a clinical history, can relate observed anomalies of shoulder movement to the presence of symptoms and to determine the reliability for observation of such anomalous shoulder movement. Methods: The sample comprised of 9 symptomatic subjects recruited from four physiotherapy clinics in Christchurch, New Zealand and a further 11 asymptomatic subjects recruited from Christchurch’s general population. They were videotaped performing shoulder flexion, abduction, and scapular plane abduction. The video-recordings were evaluated by 11 manipulative physiotherapists who did not know which subjects were symptomatic and who were thus required to judge the symptomatic status as: asymptomatic, symptomatic left, symptomatic right or symptomatic both. Additionally, each physiotherapist completed a survey on each of the 20 subjects regarding the type of movement anomaly that was perceived (e.g. too much scapular elevation, too little glenohumeral movement, etc). Classification accuracy (percentage of correct responses) and agreement (k) among physiotherapists were computed. Results: Out of the 220 responses by the physiotherapists regarding symptomatic status, 58% were correct, with 68% asymptomatic, 71% symptomatic left and 30% symptomatic right subjects correctly classified. Reliability analysis showed k statistics for all subjects was 0.23, for asymptomatic subjects 0.22, symptomatic left 0.34, and symptomatic right 0.17. Only five subjects had two or more evaluators agree on the type of anomalous movement. Conclusions: Although movement analysis is considered an integral part of a physiotherapist’s skill this research has shown that a sample of experienced manipulative physiotherapists had difficulty in determining the symptomatic status of patients with clinically diagnosed shoulder complaints by movement analysis alone. r 2006 Elsevier Ltd. All rights reserved. Keywords: Shoulder; Movement; Accuracy; Reliability

1. Introduction Shoulder pain is a common musculoskeletal complaint with an estimated lifetime prevalence of 10% in the adult population (van der Heijden, 1999). Although Corresponding author. Tel.: +64 3 479 7460; fax: +64 3 479 8414.

E-mail address: [email protected] (S. Milosavljevic). 1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.05.005

aberrations in shoulder movement patterns are thought to be associated with shoulder symptoms (Sahrmann, 2002) it is not clear how this interaction occurs. It is common clinical practice for physiotherapists to use bilateral assessment and comparison of scapulo-humeral movement to predict the dysfunctional status of the shoulder yet there is little evidence to indicate that such observation is either associated with shoulder dysfunction or can be agreed upon between physiotherapists.


B.W. Hickey et al. / Manual Therapy 12 (2007) 263–270

There is a need to determine whether manipulative physiotherapists, in the absence of a clinical history, can accurately observe whether a given subject has shoulder symptoms, can decide which shoulder is symptomatic and, can reliably agree on the observed anomaly of shoulder movement.

2. Background Shoulder complex movement has been described as a synchronous contribution of the glenohumeral, acromioclavicular, sternoclavicular, and scapulothoracic joints (Inman et al., 1944; Culham and Peat, 1993) and the kinematic relationship between the scapula and humerus has been researched extensively. Although earlier literature considered that the humerus and scapula moved in a 2:1 ratio, respectively (Codman, 1934) recent research has demonstrated variable humerus-to-scapula movement ratios, ranging from 0.91:1 to 7.51:1 depending on the phase of movement (Bagg and Forrest, 1988; Talkhani and Kelly, 1997; Mandalidis et al., 1999). Three-dimensional (3D) changes in scapular orientation have also been observed during shoulder elevation including changes in lateral and external rotation and posterior tilting (Ludewig et al., 1996). Concentric and eccentric activity, external loads, changes in velocity, different spinal postures, and age are also thought to influence the scapular and humeral relationship (Ludewig and Cook, 1996; Kebaetse et al., 1999; Pascoal et al., 2000; Talkhani and Kelly, 2001; Borstad and Ludewig, 2002; Sugamoto et al., 2002; Finley and Lee, 2003). In symptomatic individuals, changes in scapular orientation including decreased posterior tilt and decreased external rotation (‘scapular winging’) have been noted during elevation in the scapular plane (Warner et al., 1992; Lukasiewicz et al., 1999; Ludewig and Cook, 2000; Borstad and Ludewig, 2002; He´bert et al., 2002). Such reductions are thought to reduce the available sub-acromial space during shoulder elevation, and increase the risk of sub-acromial impingement (Kamkar et al., 1993). As movement timing changes have been associated with disorders within the stabilizing and activating muscular complex of the shoulder (Kibler, 1998; Sahrmann, 2002) the evaluation of movement timing and the active correction of asymmetry are considered important for successful rehabilitation (Hayes et al., 2001; Sahrmann, 2002; Magarey and Jones, 2003). Although detailed descriptions of clinical assessment of shoulder movement exist, the ability of physiotherapists to diagnose movement disorders within the shoulder complex has yet to be rigorously examined (Kapandji, 1982; Magarey and Jones, 2003). The issues

of reliability and accuracy of diagnosis of movement disorder in particular need to be ascertained. Poor reliability, high costs and limited clinical applicability of scapular orientation measurement with 3D digital tracking devices have led to the use of videotape recording of movement as a tool for shoulder complex movement analysis (Barnett et al., 1999; Yanai and Hay, 2000; Johnson et al., 2001; Karduna et al., 2001; Ackermann et al., 2002; Magarey and Jones, 2003; Lowe, 2004). However, only two studies have assessed how well physiotherapists can visually evaluate a shoulder complex for movement anomalies (Babyar, 1996; Kibler et al., 2002). A four category classification system for qualitatively describing video-recorded scapular movement anomalies was used by Kibler et al. (2002) on 20 symptomatic and six asymptomatic subjects during scapular and frontal plane abduction. Fair-to-moderate agreement for intra- and inter-rater reliability (Landis and Koch, 1977) was noted for two orthopaedic surgeons (k ¼ 0:59 and k ¼ 0:31), and two physical therapists (k ¼ 0:49 and k ¼ 0:42). However, despite the use of both symptomatic and asymptomatic subjects, no attempt was made to relate observed movement patterns to the presence of symptoms. In order to consider how well physiotherapists can visually identify the presence of symptomatic shoulder movement disorders the aims of this study are: (1) to determine whether the observation skills of manipulative physiotherapists alone allow them to correctly decide; (a) if a given subject has shoulder symptoms, (b) which shoulder is symptomatic, and (2) to describe the nature of the observed aberration and whether physiotherapists can agree on this description.

3. Methods Following approval from the Canterbury Regional Ethics Committee 9 subjects with symptomatic shoulders were recruited from the clinical caseloads of the manipulative physiotherapists working in four private physiotherapy clinics in Christchurch, New Zealand. The recruiting physiotherapists were asked to evaluate patients with shoulder symptoms using their standard clinical examination and to provisionally recruit them if they met the inclusion criteria described below. They were also screened by the principal investigator (BWH) to determine eligibility. Inclusion criteria (1) shoulder symptoms below the acromion for greater than 1 month,

ARTICLE IN PRESS B.W. Hickey et al. / Manual Therapy 12 (2007) 263–270

(2) currently on active physiotherapy treatment for their shoulder, (3) aged between 18 and 55 and, (4) capable of at least 1501 of shoulder flexion and 1401 of shoulder abduction in both shoulders, measured against the vertical plane. (5) For asymptomatic subjects—no shoulder symptoms requiring treatment in the previous year. Exclusion criteria To exclude shoulder dysfunction due to spinal or nonmechanical causes, and to minimize for asymmetric postural anomalies and surface marks. (1) clinical reproduction or evidence of shoulder symptoms with any cervical spine movement, (2) a history of systemic inflammatory disease, (3) a history of neurological symptoms including descriptions of numbness, tingling and/or other sensory disturbance in the shoulder and upper limb in the presence of upper limb weakness, (4) a scoliosis producing a visible rib hump in neutral standing posture, (5) any prior history of shoulder surgery, (6) any hearing loss limiting their ability to understand verbal instructions, and (7) any tattoos on the posterior aspect of the trunk between the cervico-thoracic and thoraco-lumbar junctions. A further 11 asymptomatic subjects with no history of previous shoulder complaint in the previous 12 months were recruited from the general population of the Christchurch metropolitan area. These subjects were chosen so that their ages and anthropometric characteristics were similar to the symptomatic subjects. Following pilot testing all subject recording took place in the same room at one of the participating clinics, in order to maintain consistency of visual background and to control for environmental factors. Sex, age, hand dominance, height, weight and body mass index (BMI) were recorded for each subject. Subjects were then instructed on the following three shoulder movements to be completed for evaluation: shoulder flexion (about a coronal axis), scapular plane abduction (approximated as between 301 and 401 anterior to the frontal plane), and shoulder abduction (about a sagittal axis). Following instruction and movement familiarization each subject was positioned for video recording. The subject stood barefoot, with feet placed on a predetermined mark and arms by their side. Each subject stood equidistant from the overhead lighting and closest corners of the room. A video camera (Sony Digital Handicam 8 mm, PCR-TRV320E PAL) was placed on a tripod 1.2 m high and 2 m behind the subjects. The field of view was adjusted to include from

265

the occiput to the thoraco-lumbar junction, and the olecranon processes on either side. Once recording had begun, each subject used the investigator’s verbal cues to initiate and then repeat the shoulder flexion, abduction and scapular plane abduction movements. The between-subject order in which these movements were video-recorded was randomized prior to recording to counter any possible within group summative effects of movement on symptom reproduction. So that the physiotherapists could not determine symptomatic status by order of presentation, the master videotape was constructed with the order of presentation randomized for the presence or absence of symptoms. The master tape was copied onto 11 standard VHS tapes and given to the 11 participating manipulative physiotherapists. These physiotherapists were selected from the staff and graduates of the University of Otago and were based in Dunedin, New Zealand. The advantage of a considerable geographical distance between the two cities thus minimized the likelihood of evaluator recognition of subjects. The physiotherapists had all completed graduate-level qualifications in manipulative physiotherapy between 1985 and 2003 and had been practising as a physiotherapist for a minimum of 5 years. They were instructed not to confer with each other prior to or after their videotape evaluation session, and to perform their evaluation session in isolation. Each therapist was provided with subject evaluation forms and an information sheet which detailed the purpose of the study, the terminology to be used for classifying any identified anomalous movements, and instructions on completing the evaluation forms. The nomenclature and description of the movement anomalies (see Table 1) were adapted from a reference text on movement impairment syndromes (Sahrmann, 2002). Upon viewing the video, each evaluator was required to nominate: (1) whether they thought the subject was symptomatic or not; (2) which shoulder(s) they considered were symptomatic, if any, and; (3) to state what movement anomalies they used to judge the shoulder(s) as having symptoms. If an evaluator judged a subject to be symptomatic, they were instructed to choose the one anomalous movement that most strongly caused them to believe the subject had a symptomatic shoulder. If the subject was thought to be asymptomatic, they were instructed not to complete the remainder of the form. The evaluators were allowed to observe each subject’s recording as often as required in order to make their best possible decision.

4. Statistical analyses Our quantitative data consisted of the rating (asymptomatic, symptomatic right, symptomatic left, sympto-



Table 1 Definitions of scapular movement for evaluators Movement

Definition

Scapular lateral rotation

Combined lateral and superior movement of the inferior angle of the scapula, through an axis located perpendicular to the frontal plane Superior translation of the whole scapula, with minimal to no medial or lateral movement Lateral translation of the vertebral border of the scapula, with minimal to no superior or rotation movement Tilting of the scapula on an oblique angle, such that the acromion moves anterior and inferior, and the inferior angle of the scapula moves posteriorly away from the rib cage Internal rotation of the scapula, such that the whole vertebral border of the scapula moves posteriorly away from the ribcage Excessive movement at the glenohumeral joint, in the absence of scapular movement anomalies

Scapular elevation Scapular abduction Scapular anterior tilt Scapular winging Hypermobile glenohumeral range of motion Hypomobile glenohumeral range of motion Other

Diminished movement at the glenohumeral joint, in the absence of scapular movement anomalies Another movement anomaly not covered by the criteria already listed

(Adapted from Sahrmann, 2002).

matic both) that each of the 11 physiotherapists gave to each of the 20 subjects, as well as their true status, as determined by the recruiting physiotherapists and the primary investigator (BWH). All statistical analyses were performed using SAS statistical software (SAS Institute, 1999).To compute chance adjusted agreement (k) between multiple evaluators the SAS MAGREE macro (SAS Technical Support, 2004) was used. Generalized linear mixed (logistic) models, as implemented in the SAS macro GLIMMIX were used to model accuracy, i.e. the proportion of correct responses. These models allow for the lack of independence in the data, i.e. that there are multiple observations for each evaluator and for each subject, by including evaluator and subject as random effects. For more on these models see Brown and Prescott (1999). Backward selection for the fixed effect variables was used to determine the best model, starting with the symptomatic status (left, right, neither), age, BMI and sex. Statistical significance of Po0:05 was used. The qualitative data on the nature of the anomalous movement, as judged by each of the physiotherapists, was summarized by frequencies of the survey responses.

5. Results Table 2 shows subject characteristics. Although the asymptomatic subjects were on average 4.7 years younger and had a wider weight distribution, the comparison of height, weight and BMI revealed similar mean scores. Four of the 220 evaluation sheets submitted by the 11 physiotherapists were incompletely recorded for choosing a subject’s symptomatic status, and these responses were not used in the analyses. Fig. 1 shows the symptomatic status chosen by each of the 11 physiotherapists for each subject, as well as the true

status. Individual physiotherapist accuracy ranged from 45% to 75%. Table 3 shows the proportions and confidence intervals of evaluator accuracy by subject status, as estimated by the logistic mixed model. Evaluators were most accurate classifying symptomatic left subjects. The proportion of these responses that were correct was estimated to be 71% (i.e. 71% of evaluators classified a subject as being symptomatic on their left side after viewing video footage of a symptomatic left subject). This was comparable to the asymptomatic group, with a proportion of 68% correctly classifying the subject. The right-sided symptomatic group was accurately judged by only 30% of responses. The evaluator accuracy was not statistically significantly different between asymptomatic and left-sided symptomatic groups (P ¼ 0:8302). However, evaluators were statistically significantly more accurate for these groups compared to the right-sided symptomatic group (P ¼ 0:0281 for left vs right and P ¼ 0:0070 for asymptomatic vs right). Subject age, BMI and sex were not significant predictors of accuracy in the model. Table 4 shows the k values for inter-rater agreement by response for the evaluators. The overall level of agreement was k ¼ 0:23. When examined relative to subject status, the asymptomatic responses demonstrated a k value of 0.22, the symptomatic left responses had a k value of 0.34 and the symptomatic right responses had a k value of 0.17. It should be noted that these values only represent the agreement of responses by evaluators, and do not take into account whether these responses were correct. There were only a small number of instances where two or more evaluators agreed within a subject on the nature of anomaly. Two subjects had three evaluators who agreed, two subjects had two evaluators agree, and one subject had two separate sets of agreement within


267

Table 2 Subject demographic means and standard deviations, mean (SD)

Age (years) Height (metres) Weight (kg) Body mass index (kg/m2) Symptomatic side

Total (n ¼ 20)

Symptomatic (n ¼ 9)

Asymptomatic (n ¼ 11)

32.7 1.72 77.3 26.1

35.1 (14.4) 1.73 (0.10) 76.3 (9.0) 25.9 (3.5) Right ¼ 6 Left ¼ 3

30.7 (10.8) 1.72 (0.08) 78.1 (16.9) 26.2 (4.1) Not applicable

(12.4) (0.08) (13.6) (3.8)

Fig. 1. Distribution of evaluator ratings for individual subjects. (+) Marks the subject’s actual symptomatic status. (o) Represents one evaluator’s choice for classifying the subject’s status.

Table 3 Accuracy of evaluators as estimated by the logistic mixed model Status

Proportion (95% CI)

Asymptomatic Symptomatic right Symptomatic left

68.0% (53.2–79.9) 30.1% (15.7–49.9) 71.0% (43.3–88.7)

Total

57.5% (43.6–70.3)

Table 4 k statistics for agreement amongst evaluators by response Status

Kappa (95% CI)

Asymptomatic Symptomatic right Symptomatic left Symptomatic both

0.22 0.17 0.34 0.01

Overall

0.23 (0.19–0.27)

(0.16–0.28) (0.11–0.23) (0.29–0.40) (0.05 to 0.07)

their dataset (which did not agree with each other). Thirty-six responses regarding the nature of a subject’s anomalous movement checked more than one criterion

Fig. 2. Frequency of anomalous movements selected by evaluators. Key: MSE, Too much scapular elevation; LSE, Too little scapular elevation; MSA, Too much scapular abduction; LSA, Too little scapular abduction; MSLR, Too much scapular lateral rotation; LSLR, Too little scapular lateral rotation; MSW, Too much scapular winging; MSAT, Too much scapular anterior tilt; LGHM, Too little glenohumeral movement; MGHM, Too much glenohumeral movement; Other.

to describe the anomalous movement (up to five criteria for some subjects), despite instructions to choose only one criterion. These responses were included in the frequencies of anomalies. These results are displayed in Fig. 2, as the number of instances each scapular anomaly was chosen by the evaluators.

6. Discussion We have modelled the accuracy, as defined by the proportion of four possible responses that were correct for 11 physiotherapists rating 20 subjects, by video recording. Additionally we modelled the inter-rater reliability (chance corrected agreement) as defined by the k statistic. The type of movement anomaly was also recorded. Examination of the overall accuracy of the evaluators demonstrated correct classification in 57.5% of the



subjects. Unlike the k statistic, there are no guidelines for defining nominal categories of high, good, moderate, fair or poor for accuracy. Although this is better than chance (25% for four possible responses), this is not a clinically acceptable level of accuracy. There were some trends in evaluator accuracy dependent on the subject’s symptom status. Asymptomatic subjects were correctly identified in over twothirds (71%) of instances, which indicates this group of manipulative physiotherapists is fairly capable of accurately identifying asymptomatic shoulder movement in an adult population. Within the symptomatic group there was a significant divergence of accuracy dependent on which shoulder was symptomatic. Although it is possible that the symptomatic left group had shoulder pathologies that more readily demonstrated anomalous movement patterns, it is more likely that it was due to chance as there were only three left-sided symptomatic subjects. As all subjects were right-handed it is also possible that hand dominance plays a role in the perception of anomalous movement. It has been suggested that anomalies will occur in the static orientation of the scapula predominantly on the dominant shoulder (Sobush et al., 1996) and also that anomalies in static scapular orientation relate to dysfunctional movement patterns (Sahrmann, 2002). If movement pattern differences exist at the dominant shoulder, symptomatic movement anomalies on the dominant side may not be easily visualized by physiotherapists due to the confluence of relevant and irrelevant scapular positional anomalies. These hypotheses require a larger study to clarify the reasons for this divergence in accuracy. The overall k value for agreement regarding symptom status was 0.23, which can be considered only fair agreement (Landis and Koch, 1977). A comparison to the four pre-determined movement pattern categories by Kibler et al. (2002) shows their values of agreement between two physical therapists achieved moderate agreement (k ¼ 0:42). While a direct comparison cannot be made as Kibler et al. (2002) did not ask the evaluators to gauge symptom status, it would be reasonable to expect that with a much larger group of evaluators as in the present study, the levels of agreement would diminish somewhat. Overall, this group of manipulative physiotherapists was capable of agreeing on subject status only mildly better than chance. Only five subjects were agreed on by two or more evaluators for accurate symptom status and the nature of anomalous movement. It is clear from these results that evaluators, when given the freedom to classify anomalous shoulder movements via multiple choices of theorized movement dysfunction, find it difficult to come to any agreement. This was further demonstrated by the large number of responses which chose multiple

anomalies, despite instructions to check only one anomaly from the list. Out of the 220 responses, there were 109 symptomatic responses, and of these there were 35 (32%) with multiple responses. Five of the evaluators did not submit any multiple responses, three submitted one or two multiple responses, and three submitted the remainder of the multiple responses. We chose to include all responses in our summary because we felt there was valuable information within them, although it does mean that care should be taken in interpreting the results. It is unknown why these evaluators recorded their evaluations this way, whether they were poor at following instructions, the instructions were not clear enough, or they felt that the symptoms could not be narrowed to one movement anomaly. A possible limitation of this study is the use of a classification system of movement anomalies adapted from the work of Sahrmann (2002). It is possible that some of the participating manipulative physiotherapists were not familiar with this classification system, used other observational criteria for analysis, and/or did not believe that only one movement anomaly would occur in a dysfunctional shoulder. Although in common clinical use, with some evidence for the presence of such movement disorders, the reliability and validity of this classification system for clinically identifying true disorders of shoulder movement has yet to be determined. We compared our qualitative movement results with other published research. Although ‘too much scapular elevation’ was the most frequent response it has only been identified as an anomalous movement in one study (Babyar, 1996) yet included as a clinical syndrome category by Kibler et al. (2002). ‘Too much scapular winging’ was also a common choice and its relationship with shoulder pathology has been well established (Ludewig and Cook, 2000; He´bert et al., 2002). ‘Too little glenohumeral movement’ and ‘too much scapular lateral rotation’ were also reported by our evaluators yet decreased glenohumeral movement is only alluded to by Babyar (1996), and excessive scapular lateral rotation has only been reported by Graichen et al. (2001). Surprisingly, ‘too much scapular anterior tilt’ was not a common choice, despite the evidence for such a relationship (Mottram, 1997; Kibler, 1998; Lukasiewicz et al., 1999; Ludewig and Cook, 2000). The overall frequency and variation of responses suggests manipulative physiotherapists may not be looking for the few scapular movement anomalies which have more reliably been shown to be relevant in symptomatic shoulders, such as scapular winging and scapular anterior tilt. Thus it may be timely for further review of the literature by those who practise clinical evaluation and management of musculoskeletal dysfunction in the shoulder. It may be that increased attention to the disorders of scapula position known to occur during dysfunctional shoulder


complex movement will be a future issue in successful management of the problematic shoulder. A high number of responses were marked ‘Other’ as a response for the nature of anomalous movement. The descriptions tended to report anomalies in terms of the poor control of the scapulothoracic muscles (e.g. ‘poor motor control serratus anterior/lower traps’) or as some variation in speed or eccentric control (e.g. ‘faster return left scapula during abduction’). It is apparent that some physiotherapists preferred to define movement anomalies in terms of possible neuromuscular causes, instead of their biomechanical consequences. We did not to arrive at an inclusion diagnosis for the symptomatic subjects by radiological investigation (such as X-ray, MRI and/or ultrasound), and further investigation by orthopaedic referral was also not done. The diagnosis and suitability of subjects with primary shoulder dysfunction was made by the recruiting physiotherapist in conjunction with the screening process of the first author. Thus it is possible that some of these subjects had conditions that did not lend themselves to the presence of anomalous shoulder movement, given that anomalous movement in symptomatic shoulders has been most consistently demonstrated in subjects with clinically diagnosed shoulder impingement (Lukasiewicz et al., 1999; Ludewig and Cook, 2000; He´bert et al., 2002). Some evaluators may also have felt limited in their ability to accurately judge the presence of relevant movement anomalies given no further clinical history on the subjects was provided, however a goal of this study was to ascertain what accuracy could be achieved in the absence of other information. Care should be taken in generalizing these results. Each of the evaluating physiotherapists had graduatelevel manipulative physiotherapy training, and each was affiliated with the University of Otago. Newer graduates or musculoskeletal physiotherapists without manipulative physiotherapy training may demonstrate different accuracy from the results presented here, although we expect them to be fairly similar or less accurate as a whole.

7. Conclusions Movement analysis is considered an integral part of a physiotherapist’s skill. Yet this research has shown that a sample of experienced manipulative physiotherapists had difficulty determining the symptomatic status of a sample of patients with clinically diagnosed shoulder complaints by movement analysis alone. Paradoxically, based on this movement analysis, they were relatively successful in determining whether a person was asymptomatic. The implications therefore are that they can detect asymmetric movement but are poor at determin-

269

ing a relationship between such asymmetry and which shoulder is symptomatic. Given that they were not aware of the subjects’ symptomatic status and clinical history this may not be surprising. However the low level of accuracy allows us to determine that movement analysis per se cannot be disassociated from the clinical history. Manipulative physiotherapists who place a high level of importance on movement analysis to determine the dysfunctional status of a given shoulder complaint need to consider whether the anomalous movement observations they make are relevant to the presenting shoulder complex disorder. It is easy to associate or ‘‘fit’’ a given movement pattern(s) once the clinical history is known. However this research indicates that it is difficult to reliably ascertain shoulder movement anomalies in the absence of a clinical history.

Acknowledgements The authors would like to thank the Barrington Physiotherapy Clinic, Riccarton Physiotherapy, and Physiosouth for acting as sources for study subjects and the subjects who participated in the research.

References Ackermann B, Adams R, Marshall E. The effect of scapula taping on electromyographic activity and musical performance in professional violinists. Australian Journal of Physiotherapy 2002;48(3):197–203. Babyar SR. Excessive scapular motion in individuals recovering from painful and stiff shoulders: causes and treatment strategies. Physical Therapy 1996;76(3):226–38. Bagg SD, Forrest WJ. A biomechanical analysis of scapular rotation during arm abduction in the scapular plane. American Journal of Physical Medicine and Rehabilitation 1988;67(6):238–45. Barnett ND, Duncan RDD, Johnson GR. The measurement of three dimensional scapulohumeral kinematics–—a study of reliability. Clinical Biomechanics 1999;14(4):287–90. Borstad JD, Ludewig PM. Comparison of scapular kinematics between elevation and lowering of the arm in the scapular plane. Clinical Biomechanics 2002;17(9–10):650–9. Brown H, Prescott R. Applied mixed models in medicine. Chinchester: Wiley; 1999 Chapter 3. Codman EA. The shoulder: rupture of the supraspinatus tendon and other lesions in or about the subacromial bursa. Privately printed 1934, Boston. Culham E, Peat M. Functional anatomy of the shoulder complex. Journal of Orthopaedic and Sports Physical Therapy 1993;18(1):342–50. Finley MA, Lee RY. Effect of sitting posture on 3-dimensional scapular kinematics measured by skin-mounted electromagnetic tracking sensors. Archives of Physical Medicine and Rehabilitation 2003;84(4):563–8. Graichen H, Stammberger T, Bonel H, Weidemann E, Englmeier K-H, Reiser M, et al. Three-dimensional analysis of shoulder girdle and supraspinatus motion patterns in patients with impingement syndrome. Journal of Orthopaedic Research 2001;19(6):1192–8.



Hayes K, Walton JR, Szomor ZL, Murrell GAC. Reliability of five methods for assessing shoulder range of motion. Australian Journal of Physiotherapy 2001;47(4):289–94. He´bert LJ, Moffet H, McFadyen BJ, Dionne CE. Scapular behavior in shoulder impingement syndrome. Archives of Physical Medicine and Rehabilitation 2002;83(1):60–9. Inman VT, Saunders JBDC, Abbott LC. Observations on the function of the shoulder joint. Journal of Bone and Joint Surgery— American Volume 1944;26(1):1–30. Johnson MP, McClure PW, Karduna AR. New method to assess scapular upward rotation in subjects with shoulder pathology. Journal of Orthopaedic and Sports Physical Therapy 2001;31(2):81–9. Kamkar A, Irrgang JI, Whitney SL. Nonoperative management of secondary shoulder impingement syndrome. Journal of Orthopaedic and Sports Physical Therapy 1993;17(5):212–24. Kapandji IA. The physiology of joints. Annotated diagrams of the mechanics of the human joints, vol 1. Upper Limb, 5th ed. Churchill Livingstone, Edinburgh, 1982. Karduna AR, McClure PW, Michener LA, Sennett B. Dynamic measurements of three-dimensional scapular kinematics: a validation study. Journal of Biomechanical Engineering 2001;123(4): 184–90. Kebaetse M, McClure P, Pratt NA. Thoracic position effect on shoulder range of motion, strength, and three-dimensional scapular kinematics. Archives of Physical Medicine and Rehabilitation 1999;80(8):945–50. Kibler WB. The role of the scapula in athletic shoulder function. American Journal of Sports Medicine 1998;26(2):325–37. Kibler WB, Uhl TL, Maddux JWQ, Brooks PV, Zeller B, McMullen J. Qualitative clinical evaluation of scapular dysfunction: a reliability study. Journal of Shoulder and Elbow Surgery 2002;11(6): 550–6. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33(1):159–74. Lowe BD. Accuracy and validity of observational estimates of shoulder and elbow posture. Applied Ergonomics 2004;35(2): 159–71. Ludewig PM, Cook TM. The effect of head position on scapular orientation and muscle activity during shoulder elevation. Journal of Occupational Rehabilitation 1996;6(3):147–58. Ludewig PM, Cook TM. Alterations in shoulder kinematics and associated muscle activity in people with symptoms of shoulder impingement. Physical Therapy 2000;80(3):276–91. Ludewig PM, Cook TM, Nawoczenski DA. Three-dimensional scapular orientation and muscle activity at selected positions of

humeral elevation. Journal of Orthopaedic and Sports Physical Therapy 1996;24(2):57–65. Lukasiewicz AC, McClure P, Michener L, Pratt N, Sennett B. Comparison of 3-dimensional scapular position and orientation between subjects with and without shoulder impingement. Journal of Orthopaedic and Sports Physical Therapy 1999;29(10):574–86. Magarey ME, Jones MA. Dynamic evaluation and early management of altered motor control around the shoulder complex. Manual Therapy 2003;8(4):195–206. Mandalidis DG, McGlone BS, Quigley RF, McInerney D, O’Brien M. Digital fluoroscopic assessment of the scapulohumeral rhythm. Surgical and Radiologic Anatomy 1999;21(4):241–6. Mottram SL. Dynamic stability of the scapula. Manual Therapy 1997;2(3):123–31. Pascoal AG, van der Helm FCT, Correia PP, Carita I. Effects of different arm external loads on the scapulo-humeral rhythm. Clinical Biomechanics 2000;15(Supplement 1):S21–4. Sahrmann SA. Diagnosis and treatment of movement impairment syndromes. St. Louis, Missouri: Mosby; 2002. p. 193–9, Chapter 5. SAS Institute. SAS/STAT V8 software user’s guide, Cary. North Carolina: SAS Institute; 1999. SAS Technical Support, Data Analysis Sample Programs. /http:// ftp.sas.com/techsup/download/stat/S Retrieved 10 June, 2004. Sobush DC, Simoneau GG, Dietz KE, Levene J, Grossman RE, Smith WB. The Lennie test for measuring scapular position in healthy young adult females: a reliability and validity study. Journal of Orthopaedic and Sports Physical Therapy 1996;23(1):39–50. Sugamoto K, Harada T, Machida A, Inui H, Miyamoto T, Takeuchi E, et al. Scapulohumeral rhythm: relationship between motion velocity and rhythm. Clinical Orthopaedics and Related Research 2002;401:119–24. Talkhani IS, Kelly CP. Scapulo thoracic rhythm in normal male volunteers. Biomedical Sciences Instrumentation 1997;34:327–31. Talkhani IS, Kelly CP. Movement analysis of asymptomatic normal shoulders: a preliminary study. Journal of Shoulder and Elbow Surgery 2001;10(6):580–4. van der Heijden GJ. Shoulder disorders: state of the art review. Baillieres Clinical Rheumatology 1999;13(2):287–309. Warner JJP, Micheli LJ, Arslanian LE, Kennedy J, Kennedy R. Scapulothoracic motion in normal shoulders and shoulders with glenohumeral instability and impingement syndrome: a study using Moire topography. Clinical Orthopaedics and Related Research 1992;285:191–9. Yanai T, Hay JG. Shoulder impingement in front-crawl swimming: II. Analysis of stroking technique. Medicine and Science in Sports and Exercise 2000;32(1):30–40.

ARTICLE IN PRESS


Original article

The influence of specific training on trunk muscle recruitment patterns in healthy subjects during stabilization exercises Veerle K. Stevens, Pascal L. Coorevits, Katie G. Bouche, Nele N. Mahieu, Guy G. Vanderstraeten, Lieven A. Danneels Department of Rehabilitation Sciences and Physiotherapy, Faculty of Medicine and Health Sciences, Ghent University, De Pintelaan 185, 6K3, B9000 Ghent, Belgium Received 14 February 2005; received in revised form 12 May 2006; accepted 10 July 2006

Abstract Low back pain is a major problem involving high medical costs, therefore effective prevention strategies are essential. Stabilization exercises seem to facilitate the neuromuscular control of the lumbar spine and may be useful in prevention programs. To investigate whether specific lumbar stabilization training has an effect on muscle recruitment patterns in a healthy population, in the present study 30 subjects were recruited to perform two types of testing exercises, i.e. bridging exercises and exercises in fourpoint kneeling, both before and after training. Surface electromyographic data of different abdominal and back muscles were obtained. After training, analysis of the relative muscle activity levels (percentage of maximal voluntary isometric contraction) showed a higher activity of the local (segmental-stabilizing) abdominal muscles, but not of the local back muscles; minimal changes in global (torque-producing) muscle activity also occurred. Analysis of the local/global relative muscle activity ratios revealed higher ratios during all exercises after training, although not all differences were significant. These results indicate that muscle recruitment patterns can be changed in healthy subjects by means of a training program that focuses on neuromuscular control. Additional studies are needed to evaluate this type of training as a prevention strategy. r 2006 Elsevier Ltd. All rights reserved. Keywords: Stabilization exercise; Prevention; Lumbar stabilization training; Surface electromyography

1. Introduction In 1994 52% of Belgian hospital nurses reported musculoskeletal job-related problems that lasted longer than 1 day; low back pain (LBP) was the major cause (53.3%) (Clarijs et al., 1998). Effective primary and secondary prevention strategies are needed to address this problem because the costs for, e.g. health insurance, employers and society, as well as the reduced quality-oflife of the patients, are substantial. Short- and long-term results indicate that specific lumbar stabilizing therapy can decrease the number of recurrent pain episodes (Hides et al., 2001) and recurrent Corresponding author. Tel.: 32 9 2402996; fax: 32 9 2403811.

E-mail address: [email protected] (V.K. Stevens). 1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.07.009

treatment periods (O’Sullivan et al., 1997;Rasmussen-Barr et al., 2003). Specific lumbar-stabilizing therapy involves changing muscle recruitment patterns. Symptomatic chronic LBP patients with clinical evidence of spondylolysis or spondylolisthesis were shown to be able to activate the deep abdominal muscles without significant coactivation of the rectus abdominis muscle (RA) when performing an abdominal drawing-in manoeuvre after a 10-week intervention (O’Sullivan et al., 1998). Lumbar stabilization training that paid no specific attention to the local muscles showed no changes in relative electromyographic (EMG) amplitudes during more complex stabilization exercises after 12 weeks of training (Arokoski et al., 2004). Apart from these studies on patients, the effect of stabilization training has not yet been investigated in a healthy population in relation to primary prevention.


V.K. Stevens et al. / Manual Therapy 12 (2007) 271–279

Local muscles of the trunk, such as the lumbar multifidus (MF), with their vertebrae to vertebrae attachments (as described by Macintosh and Bogduk, 1987), are supposed to control the fine-tuning of the positions of adjacent vertebrae (segmental stabilization) (Bergmark, 1989; Richardson et al., 1999; Hodges and Moseley, 2003). Because of their connection through the thoracolumbar fascia, the transversus abdominis (TA) and the inferior fibres of the internal oblique (IO) also have direct attachment to the lumbar vertebrae and can therefore also be considered as local muscles (Hodges, 1999). Unlike the local muscles, the global muscles are supposed to be important for torque production and general trunk stability, because they are not directly attached to the spine (Bergmark, 1989). The global muscle system includes the RA, the external oblique (EO), the gluteus maximus, the latissimus dorsi and the thoracic part of the iliocostalis lumborum muscles (ICLT) (Richardson et al., 1999). To train functional stability, once an optimal local activation has been achieved, the interplay between local and global muscles is thought to be necessary (Hodges and Moseley, 2003; Richardson et al., 2004). Because biomechanical and muscle research has shown no clear distinction between the contribution of the local and global muscles to spine stability, this functional classification based on anatomic findings needs to be considered with some caution (Arokoski et al., 2001; Cholewicki and Van Vliet, 2002; Kavcic et al., 2004a; Stevens et al., 2006). Not only is the local muscle system important, but also the controlled co-operation between the two muscle systems can provide a stable structure. Consequently, the ratio of the local muscle activity to the global muscle activity needs to be further elucidated. According to Edgerton et al. (1996) EMG ratios can be a sensitive discriminator of altered recruitment patterns and muscle dysfunction. In order to highlight differences in the synergistic activity of local versus global muscles (e.g. IO versus RA; lumbar versus thoracic erector spinae muscles), ratios of muscle activity levels during various stabilization exercises have been investigated in healthy subjects (Marshall and Murphy, 2005) and in LBP patients (O’Sullivan et al., 1998; van Diee¨n et al., 2003). The purpose of the current study was to evaluate the benefit derived from specific stabilization training for a prevention program by investigating whether this training had an effect on muscle recruitment patterns in healthy subjects. The training was an isolated local muscle contraction (first phase) followed by integrating local co-contraction in different movements starting from various positions. The exercises used for the evaluation were bridging exercises and exercises in four-point kneeling, both of which are often used in clinical practice to train lumbar stability. The specific

attention paid to the local muscles during the intervention aimed to increase local muscle activity and consequently change the local/global ratio.

2. Methods 2.1. Study design This was a cross-sectional study. The baseline EMG test session was followed by a 3-month intervention period and then a second EMG test session. 2.2. Subjects Thirty healthy subjects (15 men and women) voluntarily participated in this study. Their mean age was 19.6 (range 19–23) years, mean height was 176.5 (range 157– 194) cm and mean weight was 66.9 (range 42–84) kg. All subjects gave written informed consent. The study was approved by the Ghent University Ethics Committee. 2.3. Procedures 2.3.1. Intervention period The two EMG test sessions were separated by an intervention period of 3 months. The subjects were instructed in accordance with the principles often used in training lumbar stability (Richardson and Jull, 1995; Richardson et al., 1999; O’Sullivan, 2000). Instruction in the basic anatomy of the TA, MF and other abdominal and back muscles was aimed at emphasizing the difference between the local and global trunk muscles and to help avoid ‘substitution’ strategies of the global muscles. In the first phase of the training, local muscle activity was facilitated without substitution strategies of the global muscles and with focus on the continuation of normal breathing during the exercises. Subsequently, the time for holding the position and the number of repetitions were increased, and different postures (supine, four-point kneeling, prone, sitting and standing) were added (Richardson et al., 1999). Once an accurate and sustained contraction of the local muscles was achieved in different postures (10 contractions with 10-s holds), the exercises progressed to the second phase which involved applying low load to the muscles through controlled movements of the upper and lower extremities (Richardson and Jull, 1995). The aim during the third phase was to integrate the motor skill into normal static tasks and dynamic functional tasks (Richardson et al., 1999). During the 3-month intervention period, eight guided training sessions took place, each lasting 30 min; the subjects were also asked to perform the exercises for about 15 min each day at home as part of the intervention. During the intervention

ARTICLE IN PRESS V.K. Stevens et al. / Manual Therapy 12 (2007) 271–279

period, no specific attention was paid to the exercises performed in the test sessions. 2.3.2. Test sessions Before and after the specific stabilization training, the subjects performed two types of testing exercises: bridging exercises and exercises in four-point kneeling. For both, the only instruction given during the testing was to maintain the lumbar neutral spine position. At the moment of the first test session, the subjects had no knowledge or experience of stabilization principles. 2.4. Equipment and measurements 2.4.1. Electromyography The skin was prepared by shaving excess hair and rubbing the skin with alcohol to reduce impedance (typically p10 kO). Disposable Ag/AgCl surface electrodes (Blue Sensor, Medicotest A/S, Ølstykke, Denmark) were attached parallel to the muscle-fibre orientation, as described previously (Danneels et al., 2001b, 2002), bilaterally over the following so-called local trunk muscles: the inferior fibres of the IO (midway between the anterior iliac spine and symphysis pubis, above the inguinal ligament) and the lumbar MF (lateral to the midline of the body, above and below a line connecting both posterior superior iliac spines). Although the focus of stabilization training was on the TA, it was expected that this would be reflected in the surface EMG of the inferior fibres of the IO. Marshall and Murphy (2003) showed that medially and inferiorly to the anterior superior iliac spine, the fibres of the TA and IO are blended, so that a distinction between the muscle signals cannot be made at this location; also, at this site the direction of the fascicles of both muscles is similar (inferomedial) (Urquhart et al., 2005). Moreover, both the TA and the inferior fibres of the IO play a similar role in compressing the sacroiliac joint and consequently increasing the control of that region (Richardson et al., 2002). The selected so-called global trunk muscles were the EO (15 cm lateral to the umbilicus), the RA (3 cm lateral to the umbilicus), and the ICLT (above and below the

273

L1 level, midway between the midline and the lateral aspect of the body). The maximal inter-electrode spacing between the recording electrodes was 2.5 cm as recommended by Ng et al. (1998), and each electrode had a pick-up area of approximately 1.0 cm2. To reduce the variability due to the electrode position, a personalized template ensured the exact reapplication of the electrodes (Danneels et al., 2001a). 2.4.2. Exercises during the test procedure Maximal voluntary isometric contractions (MVIC) of the muscles were measured before the experimental tasks. These exercises were performed to provide a basis for EMG signal amplitude normalization. Three different isometric exercises against manual resistance were performed according to the description of Danneels et al. (2001b), and each exercise was registered three times during 3 s. After the MVICs, the subjects were asked to start the experimental exercises. Six exercises (often used in clinical practice to train lumbar stability) were performed. The first group of exercises was executed in supine position, knees bent (601 flexion) and feet on the floor. A bridging exercise, either simple or accompanied by leg movements, was performed (exercises 1 to 3 in Table 1). The second group of exercises was performed in four-point kneeling (exercises 4 to 6 in Table 1). At the start of each exercise, the examiner determined the subject’s lumbar neutral spine position and the subjects were asked to hold this position throughout the exercise. In four-point kneeling, the neutral spine position was set about halfway between full extension and a flat spine (Danneels et al., 2002); in supine position the anterior and posterior superior iliac spines were in line (Richardson et al., 2004). The exercises were performed in a random sequence. In order to standardize the position of the subject and the equipment, markers were placed on the floor. The dynamic phases (i.e. lifting and lowering of the pelvis and the extremities) lasted 2 s. The mid-phase (i.e. extended leg/arm and lifted pelvis) was held for 5 s. The rhythm of 60 beats/min was set by a metronome. For each exercise three trials were performed. To prevent muscular fatigue, an interval of at

Table 1 Exercises Group 1 Exercise 1 Exercise 2 Exercise 3

Bridging in supine position Bridging in supine position Ball bridge Unilateral bridging: bridging with extension of the left/right leg

Group 2 Exercise 4 Exercise 5

Four-point kneeling Single-leg lift, performed by extending the left/right leg out to the horizontal and returning it to the starting position The leg extension of exercise 4 coupled with the simultaneous raising of the contralateral arm to the horizontal before returning the extended leg and arm to the original position This exercise is basically the same as exercise 5, but with the addition of moving the trunk/pelvis in a backward direction (i.e. away from the hands), which increases the angle of hip flexion of the loaded leg by 30 1

Exercise 6



least 15 s was allowed between the exercises; during these periods the exercises were explained. Before the second test session, an ultrasound (US) evaluation was carried out to assess whether the subjects could produce an isolated contraction of the TA. In a clinical setting the tonic contraction of the MF is easy to palpate, but the difference between contraction of the IO and the TA is not always easy to detect. Therefore, the subjects were placed in a supine position lying crooked, and were then asked to draw in their lower abdomen slowly and gently, without moving the spine. The transducer was placed on the anterolateral aspect of the abdominal wall, at the level of the umbilicus. A Siemens Sonoline SL-1 ultrasound imaging device was used with a linear array probe with a wave frequency of 7.5 MHz. The criterion was a slow and controlled tensioning of the anterior fascial attachment of the TA in a lateral direction. The TA was slightly thickened and the IO and EO remained relatively inactive (Richardson et al., 2004). Because the aim was to evaluate the effect of training local muscle co-contraction on the performance of stabilization exercises, this additional US evaluation was useful to understand the reasons for changes in EMG activity. This US study revealed that 5 subjects were not able to contract the TA in isolation from the global muscle system and were thus excluded. 2.5. Data analysis The raw surface EMG signals were measured at a bandwidth of 10–500 Hz, using a differential amplifier (MyoSystem 1400, Noraxon Inc, Scottsdale, USA). The overall gain was 1000 and the common mode rate rejection ratio was 115 dB. The signals were analogue/ digitally (A/D) (12-bit resolution) converted at 1000 Hz and stored in a personal computer. The stored data were full-wave rectified and smoothed with a root mean square (RMS) with a window of 150 millis. For each of the muscles and for each testing session, the RMS was calculated for the 3 repetitions of the different exercises. The mean RMS of the three MVIC trials for every muscle was used to provide a basis for EMG signal amplitude normalization of the data of the exercises. The static phases of the exercises were analysed using an interval of 4700 ms after the defined starting point of the holding position. Noraxon MyoResearch software 2.10 was used for these analyses. The effect on muscle recruitment patterns was investigated in two ways. First, the changes in muscle activity of each muscle as a result of the training were investigated. Second, the difference in the ratio of local muscle activity to global muscle activity (separately for abdominal and back muscles) before and after training was evaluated. This assessment was based on two ratios of the abdominal muscles (the IO/RA and the IO/EO) and one of the back muscles (the MF/ICLT).

2.6. Statistical analysis Statistical analysis was performed using the SPSS 11.0 software package (SPSS Inc., Chicago, IL) for Windows. Given the symmetry of the task during the singlebridging and the ball-bridging exercise (PX0.1), the EMG values of the muscles of the left and the right side were averaged. Because there were no significant differences (P40.05) between the ipsilateral and contralateral muscle activity values during the asymmetric bridging exercises, they were also averaged. Ipsilateral referred to the side of the extended leg and contralateral to the other side. A multivariate analysis of variance (MANOVA) was used to evaluate the changes in muscle activity as a result of specific stabilization training (factors muscle [5 muscles], time [before and after training], and side [only in the asymmetric exercises]). In the event of several significant interactions, least significance difference tests, adjusted by a (although conservative) Bonferroni test to protect against type I errors, were used to analyse the significant differences between the individual muscles in each exercise. Consequently, the level for statistical significance was set at a ¼ 0:002 for exercises 1 to 4 (two-factor interaction) and at a ¼ 0:0008 for exercises 5 and 6 (three-factor interaction). To analyse the difference in the ratio of local muscle activity to global muscle activity before and after training, paired sample t-tests were used and a was set at 0.05.

3. Results 3.1. Changes in relative muscle activity Figs. 1–3 present the relative pre- and post-training EMG levels (% MVIC) of the different muscles and their respective P-values. 3.1.1. Local muscle activity After training, the local IO showed a significantly higher relative muscle activity (Pp0.001) during all the bridging exercises. In contrast to the bilateral changes of the local IO during the bridging exercises, after training only the relative muscle activity of the ipsilateral IO increased significantly during the asymmetric fourpoint kneeling exercise 5 (Pp0.001) and exercise 6 (P ¼ 0:001). 3.1.2. Global muscle activity After training, during the symmetric exercises 1 and 2 the relative muscle activity of the global RA was also significantly higher (Pp0.001). For the global EO and the local and global back muscles, no significant differences between the EMG


275

Relative muscle activity (% MVIC)

70 60 50 40

< 0.001 < 0.001

0.81

0.19 0.63

Ex1 - Pre Ex1 - Post Ex2 - Pre Ex2 - Post

30 20

0.01 0.93 0.26

10

0.001

0.001

0 IO

MF

RA Muscle

EO

ICLT

Fig. 1. Mean values, SD and P-values (between pre- and post-training for each exercise) of the relative EMG activity during the symmetric exercises (a ¼ 0:002). Ex, exercise; Pre, before training; Post, after training; MVIC, maximal voluntary isometric contraction; IO, internal oblique; MF, lumbar multifidus; RA, rectus abdominis; EO, external oblique; ICLT, iliocostalis lumborum pars thoracis.

70

Relative muscle activity (%MVIC)

< 0.001 60

0.39

50 0.11 0.38 0.91 0.38

40 30

0.007 < 0.001 0.001

Ex3 - Pre Ex3 - Post Ex4 - Pre 0.29 Ex4 - Post Ex5 - Pre Ex5 - Post Ex6 - Pre 0.93 Ex6 - Post

0.02 0.11

0.007 0.003

0.06

20 0.005 0.020.06 0.03

10 0 IO

MF

RA Muscle

EO

ICLT

Fig. 2. Mean values, SD and P-values (between pre- and post-training for each exercise) of the ipsilateral relative EMG activity during the asymmetric exercises (a ¼ 0:002 in exercises 3 and 4; a ¼ 0:0008 in exercises 5 and 6). Ex, exercise; Pre, before training; Post, after training; MVIC, maximal voluntary isometric contraction; IO, internal oblique; MF, lumbar multifidus; RA, rectus abdominis; EO, external oblique; ICLT, iliocostalis lumborum pars thoracis.

levels before and after training were found for any of the exercises. 3.2. Changes in ratios Because co-operation between the local and global muscle systems is particularly important in creating a stable structure, changes in the local/global muscle activity ratio after stabilization training were also evaluated. To detect changes in this ratio, paired sample t-tests were used; the level for statistical significance was set at a ¼ 0:05. After training, the ratio of the local to global muscle activity was higher in all exercises (Table 2). However,

not in all exercises the increase in the IO/RA and the MF/ICLT ratios was significant. All ratios were significantly higher (Pp0.01) in the single bridging exercise (exercise 1). In the ball bridge exercise (exercise 2) a significant difference (Pp0.001) was found for the abdominal muscles (IO/RA ratio and IO/EO ratio), but not for the back muscles (MF/ICLT ratio). In the unilateral bridging exercise (exercise 3), the difference between the ratios before and after training was significant (Pp0.02), except for the MF/ICLT ratio and the ipsilateral IO/RA ratio. In the exercises in fourpoint kneeling (exercises 4 – 6), the difference in the ratio local to global muscles before and after training was significant (Pp0.05) only for the ipsilateral muscles.


276

Relative muscle activity (% MVIC)

70 0.05

60

0.007 50 40

0.45 0.28

30 0.001

0.52

Ex3 - Pre Ex3 - Post Ex4 - Pre Ex4 - Post Ex5 - Pre Ex5 - Post Ex6 - Pre Ex6 - Post

0.43 0.29

0.44 0.09

0.05

0.01 0.06

0.07

0.38

0.41

20 0.02 0.10 0.01

10

0.03

0 IO

MF

RA Muscle

EO

ICLT

Fig. 3. Mean values, SD and P-values (between pre- and post-training for each exercise) of the contralateral relative EMG activity during the asymmetric exercises (a ¼ 0:002 in exercises 3 and 4; a ¼ 0:0008 in exercises 5 and 6). Ex, exercise; Pre, before training; Post, after training; MVIC, maximal voluntary isometric contraction; IO, internal oblique; MF, lumbar multifidus; RA, rectus abdominis; EO, external oblique; ICLT, iliocostalis lumborum pars thoracis.

However, the ratio of the contralateral oblique abdominal muscles also increased significantly (P ¼ 0:04) in exercise 5.

4. Discussion 4.1. Changes in relative muscle activity Stabilization training involves isolated local muscle contraction and an integration of the local and global muscle systems during particular movement patterns (O’Sullivan, 2000). It was thought that such training with specific attention paid to the TA and MF (Richardson et al., 1999; O’Sullivan, 2000) would significantly increase the relative muscle activity of the local muscles in healthy subjects. 4.1.1. Local muscle activity The results of the present study indicate that, since the relative local IO abdominal muscle activity was increased on both sides during bridging exercises and ipsilaterally during four-point kneeling exercises, abdominal muscle activity can be changed after a lumbar stabilization training program in healthy subjects. In contrast, no significant change in relative muscle activity of the local back muscle MF was found after training. One reason for this finding is that it may be more difficult to produce an isolated contraction of the MF during training. This idea is supported by the clinical experience that, in general, subjects find it easier to concentrate on drawing in the lower abdomen than on focusing on the lower back muscle contraction. Also, perhaps it was not possible to train both the deep and

the superficial fibres of the MF during the intervention period. Moseley et al. (2002, 2003, 2004) demonstrated a different activation of the deep and superficial fibres of the MF anticipating different loading conditions in standing. Since it has been shown that recording the muscle activity of the deep fibres of the MF by surface electrodes may be difficult (Stokes et al., 2003), this technique may not have been sufficiently accurate to detect any changes in activity of the deep MF fibres. Whatever the reason, the training strategy used in the present study was unable to influence the muscle activation patterns of the local back muscle. 4.1.2. Global muscle activity Not only did the local (so-called segmental-stabilizing abdominal) muscle activity levels change, but also the relative activity of some global (so-called torqueproducing and general-stabilizing) muscles changed. After training, the activity of the RA was significantly higher during the symmetric bridging exercises. Co-contraction with other abdominal muscles is often reported when subjects are trying to contract the TA (Richardson et al., 1999). Beith et al. (2001) concluded that while performing an abdominal hollowing manoeuvre, because elimination of activity in the EO muscles may be too difficult or even impossible for some to achieve, it may not always be a feasible goal. Studies on the effect of an isometric contraction of all the abdominal wall muscles (known as an abdominal brace manoeuvre) showed a considerably higher relative muscle activity of the RA in exercises 1, 3, 4 and 5 (Kavcic et al., 2004b). However, the results of the local and global relative muscle activity changes in the present study are limited to only those subjects who have been


277

Table 2 Mean values, standard deviations (SD) and P-values of the ratio local muscle activity/global muscle activity (mean) during the different exercises Exercise

1

2

Pre

IO/RA IO/EO MF/ICLT Exercise

P

Post

Mean

SD

Mean

SD

3.31 1.19 1.21

1.76 0.91 0.50

7.62 2.93 1.81

6.23 2.18 1.12

P

Post SD

Mean

SD

Ipsilateral IO/RA IO/EO MF/ICLT

8.26 2.70 1.00

3.40 1.96 0.36

9.90 3.35 1.48

6.28 1.95 0.78

Contralateral IO/RA IO/EO MF/ICLT

3.34 0.77 1.48

1.86 0.38 0.79

4.37 1.53 2.48

2.91 1.22 1.92

Mean

SD

Mean

SD

3.19 0.80 1.27

2.19 0.54 0.56

7.38 2.34 1.59

5.93 1.85 0.97

Pre

0.001* o0.001* 0.09

P

Post

Mean

SD

Mean

SD

0.20 0.02* 0.15

4.04 0.43 2.09

3.03 0.32 1.36

5.88 0.95 3.21

4.65 0.61 2.25

0.04* o0.001* 0.02*

0.01* 0.01* 0.31

10.02 1.83 0.86

5.07 1.50 0.39

10.77 2.43 1.25

9.66 1.72 1.99

0.72 0.11 0.32

5

6

Pre

Contralateral IO/RA IO/EO MF/ICLT

P

Post

4

Mean

Ipsilateral IO/RA IO/EO MF/ICLT

0.001* o0.001* 0.01*

3 Pre

Exercise

Pre

P

Post

Mean

SD

Mean

SD

3.42 0.30 1.63

3.35 0.22 1.05

5.27 0.67 2.58

3.49 0.50 1.50

10.13 1.65 0.89

6.47 1.33 0.65

11.89 2.43 0.92

9.77 1.75 1.38

Pre

P

Post

Mean

SD

Mean

SD

0.02* o0.001* 0.004*

3.36 0.40 1.11

2.84 0.25 0.37

4.99 0.88 1.64

3.83 0.68 1.05

0.05* 0.001* 0.01*

0.44 0.04* 0.91

8.08 1.57 1.01

5.65 1.09 0.28

8.23 2.17 1.27

6.36 1.70 1.21

0.92 0.06 0.31

IO, internal oblique; MF, lumbar multifidus; RA, rectus abdominis; EO, external oblique; ICLT, iliocostalis lumborum pars thoracis. *P-value significant at a ¼ 0.05 level.

shown able to isolate TA contraction (investigated using real-time ultrasound). The absence of co-contraction of the more global abdominal muscles during the exercises in four-point kneeling compared with the bridging exercises, might be explained by the difference in posture and level of difficulty between the two groups of exercises. The fourpoint kneeling position provides increased awareness of the abdominal wall due to the gravitational stretch on the muscles, and allows complete relaxation of the abdominal wall. This position may increase the sensitivity of the stretch receptors and might enhance the stimulus to contract abdominal muscles separately (Richardson and Jull, 1995; Richardson et al., 1999;Beith et al., 2001). This stretch of the abdominal wall does not exist in the supine position, which might make it harder to recruit the deep abdominals separately in this position.

4.2. Changes in ratios In general, analysis of the relative muscle activity levels showed significant changes only in the local IO after stabilization training. However, analysis of the local to global muscle activity ratios revealed that all ratios increased after training. This shows that muscle activity patterns can be changed in a healthy population if stabilization training is performed with specific attention paid to the so-called local muscles (Richardson et al., 1999;O’Sullivan, 2000). The ratio between the local and global muscle activity increased due to a greater increase in local muscle activity compared with global muscle activity. This increase in the local/global ratio was significant in most of the bridging exercises. In the four-point kneeling exercises, the local muscles at the side of the extended leg seemed to be activated to higher intensities than the



global muscles. The increase in activity was also apparent at the contralateral side, but the difference between the ratios was not significant. Similar stabilization training in symptomatic chronic LBP patients with clinical evidence of spondylolysis or spondylolisthesis also resulted in a significant increase in the IO/RA ratio during an abdominal drawing in manoeuvre (O’Sullivan et al., 1998). Simulations based on MVIC contractions in several directions predicted that an increase of the IO/RA ratio would be effective in increasing spinal stability (Van Diee¨n et al., 2003). In the present study, the relative muscle activity ratios increased during all exercises independent of the type of exercise or the surface on which the exercise was performed. Marshall and Murphy (2005) also demonstrated no change in the IO/RA ratio between stabilization exercises performed on and off a swiss ball. A limitation of the present study is that only static phases of stabilization exercises were evaluated. However, during the intervention period, the exercises were progressed to non-neutral positions (Akuthota and Nadler, 2004) and to dynamic functional movements with upper and lower extremity movements (Richardson et al., 2004). In addition to the evaluation of the static phases, further studies could also investigate the more advanced dynamic movements. The results of our relatively young group of participants may not be representative for the whole population. However, the present study was primarily designed to evaluate the effects of a basic stabilization package that could be used in a prevention programme.

5. Conclusion In the present study, healthy subjects learned isolated local muscle contractions (controlled by ultrasound) and their integration into basic static stabilization exercises. After training, analysis of the relative muscle activity levels showed a higher activity of the local abdominal muscles, but not of the local back muscles; minimal changes in global muscle activity also occurred. Analysis of the local/global relative muscle activity ratios revealed higher ratios during all exercises after training, although not all differences were significant. This indicates that muscle recruitment patterns can be changed in healthy subjects after a training programme that focuses on neuromuscular control, which could be useful in prevention programmes. More studies are needed to substantiate these results.

Acknowledgements The authors thank Ms. Evelien De Burck and Ms. Wendy Van Loo for their assistance with collection of

the data, Prof. Georges van Maele for statistical advice, and Ms. Iris Wojtowicz for linguistic corrections.

References Akuthota V, Nadler SF. Core strengthening. Archives of Physical Medicine and Rehabilitation 2004;85:S86–92. Arokoski JP, Valta T, Airaksinen O, Kankaanpaä¨ M. Back and abdominal muscle function during stabilization exercises. Archives of Physical Medicine and Rehabilitation 2001;82:1089–98. Arokoski JP, Valta T, Kankaanpaä¨ M, Airaksinen O. Activation of lumbar paraspinal and abdominal muscles during therapeutic exercises in chronic low back pain patients. Archives of Physical Medicine and Rehabilitation 2004;85:823–32. Beith ID, Synnott E, Newman A. Abdominal muscle activity during the abdominal hollowing manoeuvre in the four-point kneeling and prone positions. Manual Therapy 2001;6(2):82–7. Bergmark A. Stability of the lumbar spine. A study in mechanical engineering. Acta Orthopaedica Scandinavica 1989;230(Suppl.):20–4. Cholewicki J, Van Vliet JJ. Relative contribution of trunk muscles to the stability of the lumbar spine during isometric exertions. Clinical Biomechanics 2002;17:99–105. Clarijs JP, Zinzen E, Van Roy P, Duquet W, Caboor D, Verlinden M, et al. Multi- en interdisciplinaire evaluatie van cervicale en lumbale wervelkolomproblematiek bij ziekenhuis verpleegkundigen, met ontwikkeling en toepassing van een primair preventieprogramma. Eindverslag ST/03/029 Wetenschappelijk ondersteuningsprogramma voor de gezondheidsbescherming van de werknemer (1994– 1998); 1998. Danneels LA, Cagnie BJ, Vanderstraeten GG, Cambier DC, Witvrouw EE, De Cuyper HJ. Intra-operator and inter-operator reliability of surface electromyography in the clinical evaluation of back muscles. Manual Therapy 2001a;6(3):145–53. Danneels LA, Vanderstraeten GG, Cambier DC, Witvrouw EE, Stevens VK, De Cuyper HJ. A functional subdivision of hip, abdominal, and back muscles during asymmetric lifting. Spine 2001b;26(6):E114–21. Danneels LA, Coorevits PL, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE, et al. Differences in electromyographic activity in multifidus muscle and the iliocostalis lumborum between healthy subjects and patients with subacute and chronic low back pain. European Spine Journal 2002;11:13–9. Edgerton V, Wolf S, Levendowski D, Roy R. Theoretical basis for patterning EMG amplitudes to assess muscle dysfunction. Medical Science Sports Exercise 1996;28:744–51. Hides JA, Jull GA, Richardson CA. Long-term effects of specific stabilizing exercises for first-episode low back pain. Spine 2001;26(11):E242–8. Hodges PW. Is there a role for transversus abdominis in lumbo-pelvic stability? Manual Therapy 1999;4(2):74–86. Hodges PW, Moseley GL. Pain and motor control of the lumbopelvic region: effect and possible mechanisms. Journal of Electromyography and Kinesiology 2003;13:361–70. Kavcic N, Grenier S, McGill SM. Determining the stabilizing role of individual torso muscles during rehabilitation exercises. Spine 2004a;29(11):1254–65. Kavcic N, Grenier S, McGill SM. Quantifying tissue loads and spine stability while performing commonly prescribed low back stabilization exercises. Spine 2004b;29(20):2319–29. Macintosh JE, Bogduk N. Volvo award in basic science. The morphology of the lumbar erector spinae. Spine 1987;12(7):658–68. Marshall PW, Murphy BA. The validity and reliability of surface EMG to assess the neuromuscular response of the abdominal muscles to rapid limb movement. Journal of Electromyography and Kinesiology 2003;13:477–89.

ARTICLE IN PRESS V.K. Stevens et al. / Manual Therapy 12 (2007) 271–279 Marshall PW, Murphy BA. Core stability exercises on and off a Swiss ball. Archives of Physical Medicine and Rehabilitation 2005;86:242–9. Moseley GL, Hodges PW, Gandevia SC. Deep and superficial fibers of the lumbar multifidus muscle are differentially active during voluntary arm movements. Spine 2002;27:E29–36. Moseley GL, Hodges PW, Gandevia SC. External perturbation of the trunk in standing humans differentially activates components of the medial back muscles. Journal of Physiology 2003;547:581–7. Moseley GL, Nicholas MK, Hodges PW. Does anticipation of back pain predispose to back trouble? Brain 2004;127:2339–47. Ng JK, Kippers V, Richardson CA. Muscle fibre orientation of abdominal muscles and suggested surface EMG electrode positions. Electromyography and Clinical Neurophysiology 1998;38(1):51–8. O’Sullivan PB. Lumbar segmental ‘instability’: clinical presentation and specific stabilizing exercise management. Manual Therapy 2000;5(1):2–12. O’Sullivan PB, Twomey L, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22(24):2959–67. O’Sullivan PB, Twomey L, Allison GT. Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. JOSPT 1998;27(2):114–24. Rasmussen-Barr E, Nilsson-Wikmar L, Arvidsson I. Stabilizing training compared with manual treatment in subacute and chronic low-back pain. Manual Therapy 2003;8(4):233–41.

279

Richardson CA, Jull GA. Muscle control—pain control. What exercises would you prescribe? Manual Therapy 1995;1:2–10. Richardson CA, Jull GA, Hides JA, Hodges PW. Therapeutic exercise for spinal stabilisation. Scientific basis and practical techniques. London: Churchill Livingstone, Harcourt Brace and Company Limited; 1999. Richardson CA, Snijders CJ, Hides JA, Damen L, Pas MS, Storm J. The relation between the transversus abdominis muscles, sacro-iliac joint mechanisms, and low back pain. Spine 2002;27(4):399–405. 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, Harcourt Brace and Company Limited; 2004. Stevens VK, Vleeming A, Bouche KG, Mahieu NN, Vanderstraeten MD, Danneels LA. Electromyographic activity of trunk and hip muscles during stabilization exercises in four-point kneeling in healthy volunteers. European Spine Journal 2006, in press. Stokes IAF, Henry SM, Single RM. Surface EMG electrodes do not accurately record from lumbar multifidus muscles. Clinical Biomechanics 2003;18:9–13. Urquhart DM, Barker PJ, Hodges PW, Story IH, Briggs CA. Regional morphology of the transversus abdominis and obliquus internus and externus abdominis muscles. Clinical Biomechanics 2005;20: 233–41. Van Diee¨n JH, Cholewicki J, Radebold A. Trunk muscle recruitment patterns in patients with low back pain enhance the stability of the lumbar spine. Spine 2003;28(8):834–41.

ARTICLE IN PRESS


Case report

Improved contraction of the transversus abdominis immediately following spinal manipulation: A case study using real-time ultrasound imaging Norman W. Gilla,, Deydre S. Teyhenb, Ian E. Leea a

US Army-Baylor University Post Professional Doctoral Program in Orthopaedic and Manual Physical Therapy, Brooke Army Medical Center, San Antonio, TX 78233, USA b US Army-Baylor University Doctoral Program in Physical Therapy, Army Medical Department Center and School, San Antonio, TX, USA Received 28 November 2005; received in revised form 24 February 2006; accepted 27 June 2006

1. Introduction The authors encountered a clinical dilemma when attempting to apply a clinical prediction rule for manipulation (Flynn et al., 2002; Childs et al., 2004) to a patient with a history and physical examination consistent with clinical lumbar instability (Hicks et al., 2005). Although the patient met four of five criteria predicting short-term success with manipulation, the presence of symptoms suggestive of underlying clinical instability remains a relative contraindication to thrust manipulation (Greenman, 1996; Maitland, 2001). Could the application of both manipulation and stabilization be logically justified in this patient? Despite the widespread use of spinal manipulation, the biological mechanisms by which it produces a beneficial effect in certain patients are not fully understood (Pickar, 2002). There is evidence which supports a reflexogenic effect from manipulation in the paraspinal muscles as one possible mechanism (Herzog et al., 1995, 1999; Lehman et al., 2001; Pickar, 2002). Specifically, several researchers have identified altered motor neuron pool excitability following spinal manipulation (Murphy et al., 1995; Floman et al., 1997; Herzog et al., 1999; Dishman and Bulbulian, 2000). The effect on neural pathways associated with manipulation has been suggested as one possible mechanism that may improve muscle performance (Pickar, 2002) and patient symptoms. Corresponding author.

E-mail address: [email protected] (N.W. Gill). 1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.06.014

Support exists for an association between spinal manipulation and improved muscle function in the quadriceps (Suter et al., 1999), the erector spinae (Keller and Colloca, 2000), and the deep neck flexors (Sterling et al., 2001). Therefore, it is reasonable to hypothesize that spinal manipulation, by a reflexogenic mechanism, may improve the performance of the deep trunk stabilizers. In turn, improved relaxation and contractility of the lumbar multifidus and the transversus abdominis (TrA) theoretically could lead to improved functional stability of the spine through enhancement of the neurological and active subsystems as defined by Panjabi (1992a, b). This single case study describes changes observed in the TrA musculature pre- to post-manipulation in a patient that presented with a clinical paradox (symptoms suggestive of clinical lumbar instability but also meeting the clinical prediction rule to succeed with lumbar manipulation therapy). Real-time ultrasound imaging (USI) was used to describe the changes in the TrA musculature.

2. Methods 2.1. Patient The patient was a 43-year-old male with a 30-day history of right low back pain and diffuse, posterior right thigh pain to the knee (Fig. 1). His symptoms had

ARTICLE IN PRESS N.W. Gill et al. / Manual Therapy 12 (2007) 280–285

developed insidiously while running. He reported a history of mild recurrent low back pain for 10 years. The patient’s most notable examination findings were constant pain in the thigh and a painful catch in flexion and in extension. Upon screening for a concurrent research study, he was found to have signs suggestive of clinical instability (Table 1). However, on examination he was also noted to have four of five criteria that predict shortterm success with spinal manipulation (Table 2). The patient consented to receive USI both before and after spinal manipulation. 2.2. Pre-manipulation training A previously developed technique (Teyhen et al., 2005) was used to train an isometric contraction of the TrA muscle through the abdominal drawing-in manoeuvre (ADIM). The patient was first instructed in the ADIM via traditional training techniques (Richardson et al., 1999). The patient was instructed to gently pull his abdominals in toward his spine as he exhaled and then to maintain this contracted state for 10 s. The initial training also included tactile cueing with the patient’s fingers palpating 1 inch medial and inferior to the anterior superior iliac spine to help confirm contraction of the TrA. After the patient understood the intent of the ADIM, he underwent training using real-time USI as biofeedback. In order to obtain a

281

ceiling effect, training continued until the patient was able to perform three isolated TrA contractions with no further increase in the thickness of the TrA during the ADIM. The training session lasted for approximately 15 min. 2.3. Ultrasound instrumentation and measurement technique Ultrasound images of the lateral abdominal muscles (Fig. 2) were obtained both pre- and post-manipulation using the Sonosite 180 Plus (Sonosite Inc. Bothell, WA) with a 2–5 MHz curvilinear array in B mode. The patient was positioned in a supine crook-lying position. The transducer was placed approximately 1 inch superior to the iliac crest along the mid-axillary line in the transverse plane. Once an adequate image of the TrA was obtained, a skin marker (a single line bisecting the length of the transducer head) was used to standardize location of the transducer between pre- and postmanipulation images. Measurements of muscle thickness of the TrA were obtained pre- and post-manipulation with the patient at rest and during the ADIM. To standardize the influence of respiration (De Troyer et al., 1990; Hodges et al., 1997; Hodges and Gandevia, 2000a, b) on the thickness of the TrA, the images were collected upon completion of the exhalation as determined by visual inspection of

Fig. 1. Pain body diagram as annotated on initial evaluation. P1, P2, P3 ¼ pain #1, pain #2, pain #3; C ¼ constant; I ¼ intermittent; V ¼ variable; Sup. ¼ superficial; check mark ¼ area cleared.


N.W. Gill et al. / Manual Therapy 12 (2007) 280–285

Table 1 Criteria used during concurrent study for patients with symptoms suggestive of clinical instability Criteria

Present on examination

1 of 5 aberrant movement signs Painful arc on flexion Painful arc on return from flexion Reversal of lumbo-pelvic rhythm Gower’s sign Deviation in motion

Yes Yes Yes Yes No No

Prone instability test Hypermobility on lumbar spring test Recurrent low back pain Straight leg raise 4901 Age o40 years of age

Yes Yes (L3) Yes No No

Based, in part, on criteria predictive of success with lumbar stabilization exercises; from Hicks et al (2005).

Table 2 Criteria for clinical predictive rule for regional lumbopelvic manipulation Criteria

Present on examination

Symptoms o16 days FABQ (w) o19 No symptoms distal to knee At least one level hypomobility in lumbar spine At least one hip internal rotation 4351

No Yes Yes Yes (L5) Yes

From Flynn et al. (2002). FABQ (w) ¼ Fear Avoidance Behaviour Questionnaire Work Subscale.

2.4. Manipulation The manipulation performed was a regional lumbopelvic manipulation detailed elsewhere (Flynn et al., 2002; Childs et al., 2004). This regional technique was selected since it was the specific technique used to develop (Flynn et al., 2002) and validate (Childs et al., 2004) the clinical predictive rule criteria that matched our patient. The manipulation was attempted on the right side first but yielded no cavitation. The patient was repositioned and the left side was then manipulated. Again, no cavitation was heard or felt. Finally, a repeat manipulation on the right yielded a cavitation and the intervention was considered complete. The patient was immediately repositioned on the table and the postmanipulation measurements were captured using the previously describe protocol. Care was taken to place the transducer head in the exact location that was used to capture the pre-manipulation images. 2.5. Data analysis The statistical significance of the changes in muscle thickness was analysed by considering a previously reported standard error of the measure (SEM). The SEM for the measurement technique described was established to be 0.031 cm (Teyhen et al., 2005). Therefore, a significant change in muscular thickness was defined as a change greater than two SEMs (0.062 cm); thus the authors could be 95% confident that any change measured was due to an actual change in the phenomenon being measured and not measurement error.

3. Results

Fig. 2. Ultrasound image of the lateral abdominal wall (resting image taken post-manipulation). White solid arrows indicate thickness. SST ¼ superficial soft tissue; EO ¼ external oblique; IO ¼ internal oblique; TrA ¼ transversus abdominis; AC ¼ abdominal contents.

the ultrasound image. The procedures used to measure muscle thickness of the TrA are described elsewhere (Teyhen et al., 2005).

Changes in the TrA function were noted immediately post-manipulation. The thickness of the TrA muscle during the resting and contracted states both pre- and post-manipulation is provided in Table 3. The total restto-contract difference post-manipulation was 0.38 cm, nearly doubling the total rest-to-contract thickness premanipulation (0.21 cm). Our patient demonstrated a statistically significant change in all three measurements related to the TrA thickness: a decrease in resting thickness of the TrA, an increase in maximal thickness during the ADIM, and an increase in total rest-tocontract thickness of the TrA post-manipulation. The changes in muscular function noted after the manipulation were also accompanied by modest clinical improvements. The patient’s painful catch moving into extension was abolished and his constant thigh pain became intermittent. However, his painful catch with forward bending was only transiently improved.

ARTICLE IN PRESS N.W. Gill et al. / Manual Therapy 12 (2007) 280–285

283

Table 3 Muscular thickness of the transversus abdominis (TrA); measured in centimeters

Resting TrA Max contract during ADIM Rest-to-contract difference

Pre-manipulation

Post-manipulation

Difference

0.51 0.72 0.21

0.42 0.80 0.38

0.09* +0.08* +0.17*

Significance (*)42SEMs (0.062 cm); SEM ¼ standard error of the measurement; ADIM ¼ abdominal drawing-in manoeuvre.

4. Discussion Our results document a dramatic change in one patient’s ability to perform a preferential TrA contraction during an ADIM immediately following spinal manipulation. It is notable that the observed improvement occurred after an established ceiling effect in the performance of the ADIM was reached. This improvement in the ability to contract the TrA was also associated with improvements in some clinical exam findings. Manipulating a patient with a presentation suggestive of clinical instability appears counterintuitive. However, there exists the logical possibility that in some patients with clinical instability, a high-velocity, lowamplitude thrust procedure may be a reasonable treatment option. There are at least three theoretical mechanisms of action for spinal manipulation including a mechanical effect on joint arthrokinematics, a neuroendocrine effect (such as endorphin release), and a neurophysiological or reflexogenic effect (Herzog et al., 1999). The reflexogenic effect describes the process of a stretch reflex elicited by manipulation in either the joint mechanoreceptors or muscle spindles that eventually leads to an inhibition, depression, or attenuation of the alpha motor neuron pool thus causing muscle relaxation that breaks the ‘‘spasm–ischemia–pain– spasm’’ cycle (Murphy et al., 1995). This has been demonstrated in the porcine model (Indahl et al., 1997). In further support of this theory, Lehman et al. (2001) found an exaggerated surface electromyographic (EMG) response to painful stimuli in patients with chronic low back pain, which was attenuated following spinal manipulation. In another study by Lehman and McGill (2001) with low back pain patients, they found a decrease in muscle amplitude on surface EMG during quiet stance post-manipulation in their most acute subjects. More specific to our study was a case they highlighted of a subject who showed a dramatic decrease in surface EMG activity (41%) during quiet stance postmanipulation in the area of the right internal oblique which was qualitatively ‘‘in spasm’’ pre-manipulation (Lehman et al., 2001). This may be similar to the decrease in the resting tone of the TrA post-manipulation in our patient.

Two groups of researchers noted a decrease in the S1 alpha motor neuron pool excitability as measured by the H-reflex in asymptomatic subjects receiving spinal manipulation (Murphy et al., 1995; Dishman and Bulbulian, 2000). However, Floman et al. (1997) found a facilitation instead of an inhibition of the H-reflex after manipulation in patients with a unilateral herniated nucleus pulposus and an initially diminished H-reflex. Therefore, activated inhibitory and facilitory pathways may work synergistically to break the ‘‘spasm–ischemia–pain–spasm’’ cycle. In our patient, improved relaxation of the muscle in the resting state coupled with increased excitability during contraction (during the ADIM) may reflect changes in motor neuron pool excitability following manipulation. An unexplored, yet potentially critical link exists between the reflexogenic effect of manipulation and spinal stabilization theory. The three components of spinal stabilization include the passive (non-contractile) subsystem, active (musculotendinous) subsystem, and neural control subsystems (Panjabi, 1992a, b, 2003). Clinicians often view instability as a dysfunction of only the passive subsystem. However, a dysfunction arising in the neural control or active subsystems could alone, or together, alter the stabilization system (Panjabi, 1992a, b, 2003; van Dieen et al., 2003). It is possible, then, that neural inhibition of key spinal stabilization muscles, such as the TrA and lumbar multifidus, could cause signs and symptoms consistent with instability despite the passive constraints being normal. Manipulation in this case, via a possible reflexogenic mechanism, may restore or improve the neural control and active subsystems and thereby minimize the clinical complaints associated with instability. The authors propose this is analogous to a ‘‘control+alt+delete’’ procedure performed on a malfunctioning software programme. The relationship between inhibition of primary muscular stabilizers and pain, joint dysfunction, and instability has already been demonstrated in the lumbar spine (Hodges and Richardson, 1996; Richardson et al., 2002; Hungerford et al., 2003). This concept has been extended to other regions including the cervical spine (Sterling et al., 2001), the shoulder (Magarey and Jones, 2003), and the knee (Suter et al., 1999; Cowan et al.,


N.W. Gill et al. / Manual Therapy 12 (2007) 280–285

2001). Demonstrating a relationship between muscle inhibition and spinal manipulation, Suter et al. (1999) measured a decrease in quadriceps inhibition in patients with anterior knee pain after manipulation of the sacroiliac joint. Furthermore, Keller and Colloca (2000) used surface EMG activity to assess the isometric strength of the erector spinae post-manipulation and noted increased gains compared to a sham manipulation. Therefore, it is likely that a regional lumbopelvic manipulation could reduce muscle inhibition and/or facilitate optimal function of the intrinsic lumbar stabilizers via a reflexogenic mechanism; improved performance of the TrA and multifidus muscles would, in turn, result in improved stability to the spine. There is growing evidence that a combination of manipulation and lumbar stabilization is effective (Aure et al., 2003; Niemisto et al., 2003; Childs et al., 2004). This could be due, in part, to the potential for improved functioning of the TrA and multifidus muscles after manipulation. The immediate effect of manipulation may decrease pain, restore motion, and reset the neural control and active (muscle) subsystems. This would facilitate the training of isolated TrA and multifidus contractions. In essence, the manipulation would provide a neurophysiological ‘‘window of opportunity’’ to maximize the effectiveness of the first phase of lumbar stabilization. This is then followed by advanced and functional lumbar stabilization training to facilitate motor re-patterning for the long-term rehabilitative and preventive effect (O’Sullivan et al., 1997, 1998; Hides et al., 2001). A similar suggestion (application of mobilization prior to muscle retraining) has been suggested by Sterling et al. (2001) in the retraining of the deep neck flexor muscles. Direct support for this theory was measured in a recent trial by Childs et al. (2004) in which those patients receiving two manipulations and simple motion exercises during the first week followed by 3 weeks of lumbar stabilization exercises experienced greater improvement in short-term and long-term function than those receiving lumbar stabilization exercises alone. Threats to internal validity abound in a case study and therefore it is not possible to ascribe any changes documented here to a cause and effect relationship. Training effect is one significant threat in this case study but the training effect was mitigated by imaging the lateral abdominal muscles pre-manipulation after a ceiling effect was noted. Furthermore, the proximity of the pre- and post-measurements helps to minimize the influence of a training effect. The authors are currently investigating the effects of manipulation on motor control of the TrA and the multifidus using a randomized control trial. It is important to note that the patients of interest are those with signs of clinical instability and not those with frank or gross instability.

5. Conclusion In this single case study of a patient with symptoms suggestive of clinical instability, improvement in the contraction of the TrA was found immediately following a regional lumbo-pelvic manipulation. It is theoretically feasible that the manipulation provided a neurophysiological ‘‘window of opportunity’’ in which to maximize the preferential activation of the key spinal stabilization muscles. Future researchers should investigate the effects of manipulation on neuromuscular control of the deep trunk muscles and on symptoms suggestive of clinical instability.

Disclaimer The opinions or assertions contained here in are the private views of the Authors and are not to be construed as official or as reflecting the views of the Departments of the Army or Defense.

References Aure OF, Nilsen JH, Vasseljen O. Manual therapy and exercise therapy in patients with chronic low back pain: a randomized, controlled trial with 1-year follow-up. Spine 2003;28(6):525–32. Childs JD, Fritz JM, Flynn TW, Irrgang JJ, Johnson KK, Majkowski GR, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Annals of Internal Medicine 2004;141(12):920–8. Cowan SM, Bennell KL, Hodges PW, Crossley KM, McConnell J. Delayed onset of electromyographic activity of vastus medialis obliquus relative to vastus lateralis in subjects with patellofemoral pain syndrome. Archives of Physical Medicine and Rehabilitation 2001;82:183–9. De Troyer A, Estenne M, Ninane V, Van Gansbeke D, Gorini M. Transversus abdominis muscle function in humans. Journal of Applied Physiology 1990;68(3):1010–6. Dishman JD, Bulbulian R. Spinal reflex attenuation associated with spinal manipulation. Spine 2000;25(19):2519–25. Floman Y, Liram N, Gilai AN. Spinal manipulation results in immediate H-reflex changes in patients with unilateral disc herniation. European Spine Journal 1997;6(6):398–401. Flynn T, Fritz J, Whitman J, Wainner R, Magel J, Rendeiro D, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine 2002;27(24):2835–43. Greenman PE. Principles of manual medicine, second ed. Philadelphia: Lippincott Williams & Wilkins; 1996. Herzog W, Conway PJ, Zhang YT, Gal J, Guimaraes AC. Reflex responses associated with manipulative treatments on the thoracic spine: a pilot study. Journal of Manipulative and Physiological Therapeutics 1995;18(4):233–6. Herzog W, Scheele D, Conway PJ. Electromyographic responses of back and limb muscles associated with spinal manipulative therapy. Spine 1999;24(2):146–53. Hicks GE, Fritz JM, Delitto A, McGill SM. Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program.

ARTICLE IN PRESS N.W. Gill et al. / Manual Therapy 12 (2007) 280–285 Archives of Physical Medicine and Rehabilitation 2005;86(9): 1753–62. 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. Hodges PW, Gandevia SC. Activation of the human diaphragm during a repetitive postural task. Journal of Physiology 2000a;522:165–75. Hodges PW, Gandevia SC. Changes in intra-abdominal pressure during postural and respiratory activation of the human diaphragm. Journal of Applied Physiology 2000b;89(3):967–76. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 1996;21(22):2640–50. Hodges PW, Gandevia SC, Richardson CA. Contractions of specific abdominal muscles in postural tasks are affected by respiratory maneuvers. Journal of Applied Physiology 1997;83(3):753–60. Hungerford B, Gilleard W, Hodges P. Evidence of altered lumbopelvic muscle recruitment in the presence of sacroiliac joint pain. Spine 2003;28(14):1593–600. Indahl A, Kaigle AM, Reikeras O, Holm SH. Interaction between the porcine lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles. Spine 1997;22(24):2834–40. Keller TS, Colloca CJ. Mechanical force spinal manipulation increases trunk muscle strength assessed by electromyography: a comparative clinical trial. Journal of Manipulative and Physiological Therapeutics 2000;23(9):585–95. Lehman GJ, McGill SM. Spinal manipulation causes variable spine kinematic and trunk muscle electromyographic responses. Clinical Biomechanics 2001;16(4):293–9. Lehman GJ, Vernon H, McGill SM. Effects of a mechanical pain stimulus on erector spinae activity before and after a spinal manipulation in patients with back pain: a preliminary investigation. Journal of Manipulative and Physiological Therapeutics 2001;24(6):402–6. Magarey ME, Jones MA. Dynamic evaluation and early management of altered motor control around the shoulder complex. Manual Therapy 2003;8(4):195–206. Maitland GD. Vertebral manipulation, sixth ed. Oxford: ButterworthHeinemann; 2001. Murphy BA, Dawson NJ, Slack JR. Sacroiliac joint manipulation decreases the H-reflex. Electromyography and Clinical Neurophysiology 1995;35(2):87–94.

285

Niemisto L, Lahtinen-Suopanki T, Rissanen P, Lindgren KA, Sarna S, Hurri H. A randomized trial of combined manipulation, stabilizing exercises, and physician consultation compared to physician consultation alone for chronic low back pain. Spine 2003;28(19): 2185–91. O’Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22(24):2959–67. O’Sullivan PB, Twomey LT, Allison GT. Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. Journal of Orthopaedic & Sports Physical Therapy 1998;27(2):114–24. Panjabi M. The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, and enhancement. Journal of Spinal Disorders 1992a;5(4):383–9. Panjabi M. The stabilizing system of the spine. Part II. Neutral zone and instability hypothesis. Journal of Spinal Disorders 1992b;5(4):390–7. Panjabi M. Clinical spinal instability and low back pain. Journal of Electromyography and Kinesiology 2003;13(4):371–9. Pickar JG. Neurophysiological effects of spinal manipulation. Spine Journal 2002;2(5):357–71. Richardson CA, Jull G, Hodges P, Hides JA. Therapeutic exercise for spinal segmental stabilization in low back pain. London: Churchill Livingstone; 1999. Richardson CA, Snijders CJ, Hides JA, Damen L, Pas MS, Storm J. The relation between the transversus abdominis muscles, sacroiliac joint mechanics, and low back pain. Spine 2002;27(4):399–405. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent effects on pain, sympathetic nervous system activity and motor activity. Manual Therapy 2001;6(2):72–81. Suter E, McMorland G, Herzog W, Bray R. Decrease in quadriceps inhibition after sacroiliac joint manipulation in patients with anterior knee pain. Journal of Manipulative and Physiological Therapeutics 1999;22(3):149–53. Teyhen DS, Miltenberger CE, Deiters HM, Del Toro YM, Pulliam JN, Childs JD, et al. The use of ultrasound imaging of the abdominal drawing-in maneuver in subjects with low back pain. Journal of Orthopaedic & Sports Physical Therapy 2005;35(6):346–55. van Dieen JH, Selen LP, Cholewicki J. Trunk muscle activation in lowback pain patients: an analysis of the literature. Journal of Electromyography and Kinesiology 2003;13(4):333–51.

ARTICLE IN PRESS


Technical and measurement report

Measurement of segmental cervical multifidus contraction by ultrasonography in asymptomatic adults Jo-Ping Leea, Wen-Yih I. Tsengb, Yio-Wha Shauc, Chung-Li Wangd, Hsing-Kuo Wanga, Shwu-Fen Wanga, a

School and Graduate Institute of Physical Therapy, College of Medicine, National Taiwan University, Floor 3, No.17, Xuzhou Road, Zhongzheng District, Taipei City 100, Taipei, Taiwan, ROC b Center for Optoelectronic Biomedicine, College of Medicine, National Taiwan University, Taipei, Taiwan, ROC c Institute of Applied Mechanics, College of Engineering, National Taiwan University, Taipei, Taiwan, ROC d Department of Orthopedic Surgery, National Taiwan University Hospital, Taipei, Taiwan, ROC Received 15 June 2005; received in revised form 12 May 2006; accepted 10 July 2006

Abstract The deep muscles that play significant roles in maintaining segmental stability have been measured using ultrasonography (US). However, few studies have been carried out to determine the reliability and validity of US for measuring the cervical multifidus during contraction. The aims of this investigation were to evaluate the reliability of the dimensions of the cervical multifidus as measured using US and compare the US measurements with those determined with magnetic resonance imaging (MRI), the gold standard. Ten asymptomatic adult subjects (age, 28.573.5 years) participated in testing–retesting of muscle dimensions at rest and during isometric head extension with the cranio-cervical spine maintained in a flexed position against individual maximum resistance. Ten asymptomatic adult subjects (age, 28.174.1 years) participated in testing to compare US and MRI measurements of the thickness, width, area, and shape ratio of the cervical multifidus at the C4, C5, and C6 levels. US measurements of muscle thickness at the C4, C5, and C6 levels at rest were 0.6770.14, 0.7070.20 and 0.7370.09 cm, respectively; the corresponding measurements as determined by MRI were 0.7070.12, 0.6770.15 and 0.7070.06 cm. The within-subject coefficient of variation (CVw) for thickness at rest and during contraction was less than 10%, indicating acceptable reliability. US measurement of thickness had better reliability and validity. The range of limit of agreement for muscle thickness at the C4, C5, and C6 levels was 0.20 to 0.20 cm, and the range of R2 was 0.42–0.64. The thickness of the cervical multifidus muscle increased significantly during contraction (1.1370.20, 1.1970.20 and 1.1770.12 cm for the C4, C5, and C6 levels, Po0.05). However, no significant differences were noted among the three levels. The results indicate that US can detect changes in segmental cervical multifidus during contraction. r 2006 Elsevier Ltd. All rights reserved. Keywords: Ultrasonography; Cervical multifidus; Validity; Reliability

1. Introduction Neck pain, which has been thought to be caused by neck trauma or occupational repetitive injuries, is extremely prevalent in modern society (Cote et al., 1998; Cote et al., 2000). In a Canadian study of adults, 64% of respondents reported that they had experienced neck pain during their life time and 54% reported that Corresponding author. Tel.: +2 33228139; fax: +2 332218161.

E-mail address: [email protected] (S.-F. Wang). 1356-689X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2006.07.008

neck pain had occurred during the previous 6 months (Cote et al., 1998). Moreover, nearly 5% of the respondents reported that they had been considerably disabled because of neck pain during the previous 6 months (Cote et al., 1998). These findings highlight the need to better understand the pathogenesis of the cervical spine. Novel measurement techniques can help to advance this understanding. The deep muscles play major roles in maintaining segmental stability, (Panjabi et al., 1989; Kristjansson and Jonsson, 2002; Kristjansson et al., 2003) and

ARTICLE IN PRESS J.-P. Lee et al. / Manual Therapy 12 (2007) 286–294

evidence suggests an association between deep muscle dysfunction and spinal pain (Hides et al., 1994, 1996; Hayashi et al., 2002; Kristjansson, 2004). The importance of deep lumbar muscles (Panjabi et al., 1989; Hides et al., 1994; Hides et al., 1996) and abdominal muscles (Hodges and Richardson, 1996; Hodges and Richardson, 1997) for controlling back stability has been demonstrated in biomechanical models (Panjabi et al., 1989) and clinical experiments (Hides et al., 1994, 1996; Hodges and Richardson, 1996, 1997). Furthermore, in patients with low back pain, intervertebral segmental muscle atrophy has been found (Hides et al., 1994, 1996; Campbell et al., 1998) with use of magnetic resonance imaging (MRI) (Campbell et al., 1998) or ultrasonography (US) (Hides et al., 1994, 1996). Similarly, muscle atrophy has been noted after neck injury (Hayashi et al., 2002; Kristjansson, 2004). In addition, altered configuration of the cervical lordosis has been found in patients with neck disorders (Ueki et al., 1995; Kristjansson and Jonsson, 2002; Kristjansson et al., 2003). One possible explanation for this altered configuration is the inability of the deep cervical segmental muscles to maintain cervical alignment (Kristjansson, 2004). MRI is considered to be the gold standard (Westbrook and Kaut, 1993; Hides et al., 1995; Esformes and Narici, 2002) for measuring the lumbar multifidus at rest (Hides et al., 1994, 1996; Stokes et al., 2005), and US has also been demonstrated to be a valid and reliable method. US has the advantages of a lower cost (Esformes and Narici, 2002; Kristjansson, 2004) and the ability to provide non-invasive visualization of the change in architecture of the deep cervical (Esformes and Narici, 2002; Hodges et al., 2003; Kristjansson, 2004) and lumbar muscles (Ito et al., 1998). Furthermore, good reliability was obtained in measuring the thickness of the transversus abdominis muscle during the hollowing maneuver (McMeeken et al., 2004). However, to our knowledge, there have been no reports on the validity and reliability of US for measuring the cervical multifidus muscle during contraction. Changes in the dimensions of the cervical multifidus during contraction with maximum resistance are also unknown. US was suggested to be the reliable method for detecting isometric contractions of low maximal voluntary contraction (Hodges et al., 2003), but the changes in muscle

287

dimensions and force differed among the muscles tested (Hodges et al., 2003; McMeeken et al., 2004). The purposes of this study were to determine the reliability of US for measuring changes in the dimensions of the cervical multifidus muscle during rest and contraction and to evaluate the validity of US as compared with MRI. The specific aims were to (1) investigate the intra-rater, intersession reliability of US for measuring the thickness, width, area, and shape ratio of the cervical multifidus at the C4, C5, and C6 levels at rest and during isometric head extension with craniocervical spine maintained in a flexed position against individual maximum resistance; (2) compare US and MRI measurements of the thickness, width, area, and shape ratio of the cervical multifidus at the C4, C5, and C6 levels under static condition; and (3) determine the change in the dimensions of the cervical multifidus at the C4, C5, and C6 levels during contraction.

2. Subjects and methods 2.1. Participants After Institutional Review Board (IRB) approval, 17 asymptomatic volunteers (age, 26.7573.8 years) were recruited (Table 1). Most of the participants exercised routinely (mean of 1.470.7 times per week, for a mean of 1.571.0 h each time). Criteria for exclusion were a history of trauma to the cervical, thoracic, or lumbar spine, previous surgery of the spine, and neck pain within the previous three months. All participants gave their informed consent after the nature of the procedures had been fully explained. Reliability and validity testings were carried out with 10 participants each; thus four subjects participated in both groups. 2.2. US US was performed using a real-time ultrasound apparatus with a 10 MHz linear array transducer (HDI 5000; ATL Ultrasound, Bothell, WA, USA). With this system, the axial resolution of the image was 0.7 mm at a depth of 3 cm and the horizontal resolution was 1.0 mm at a depth of 4–6 cm. The measurement protocol using US was designed through several pilot trials and was

Table 1 Demographic details of participants Total

Age (years) Height (cm) Body mass (kg)

Reliability

Validity

Men (n ¼ 11)

Women (n ¼ 6)

Men (n ¼ 7)

Women (n ¼ 3)

Men (n ¼ 6)

Women (n ¼ 4)

26.0974.0 179.977.6 65.675.8

27.1774.0 157.172.6 56.376.4

27.574.7 168.576.3 67.075.6

29.073.5 156.672.8 54.877.6

28.1773.8 170.876.4 70.073.6

27.0574.7 158.772.3 58.372.0


J.-P. Lee et al. / Manual Therapy 12 (2007) 286–294

based on a fundamental knowledge of cervical anatomy and US. The bifurcation of the spinal process at the C2–C6 cervical levels was identified manually and on the US images in the following order. The C2 level was identified as the first bifurcation of the spinal process palpated from the occipital bone downward. The C7 level was palpated as the most prominent process among the cervical levels where no bifurcation was found on the US image. The C4 level was identified as the second bifurcation of the spinal process palpated downward from the C2 level, and the C5 and C6 levels were identified as the first and second bifurcation of the spinal process caudal to the C4 level. The cervical spinal processes were palpated and marked by the rater, a physical therapist, with US guidance to identify the bifurcation of the spinal process. This method of identifying the segmental spinal processes has been shown to be acceptably reliable (intra-rater withinsubject coefficient of variation [CVw] of 6.4–10.7%) (Wu et al., 2005). 2.3. MRI MRI was performed with a 1.5 Tesla scanner (Magnetom Sonata, Siemens, Erlangen, Germany). A flexible surface coil, 20 50 cm, was used as a receiver coil and was fixed over the posterior aspect of the participant’s neck. The participant was placed in the same prone position as for US. The locations of the vertebral levels were determined from parallel images in the mid-sagittal plane. To position the slices of MRI at the same location as those obtained by US, multiple transverse slices were scanned from the superior border of the C3–C4 intervertebral disk to the

inferior border of the C6–C7 intervertebral disk at 5-mm intervals (Fig. 1A). From these images, the rater, an experienced radiologist, determined the C4, C5, and C6 levels by identifying the bifurcation of the spinal process and the articular process (Fig. 1A). T1-weighted images were obtained with use of half-Fourier single-shot turbo spin-echo acquisition (HASTE); the parameters were TR/TE/flip angle of 1500/114 ms/901; the matrix size was 256 256, the field of view was 21 cm, the slice thickness was 5 mm, and the number of average was 20. 2.4. Reliability of US Reliability testing was carried out in 10 asymptomatic participants (four women and six men; age, 28.1074.1 years) (Table 1). To standardize the position of the participants during measurement, each participant was asked to sit upright in a customized chair and rest their arms on their legs. A head belt was fixed over the forehead of the participant to maintain the head and neck in a neutral position (Fig. 2). The participant was asked to extend both knees, and the heels of their feet were placed on a 15-cm height board to eliminate the overflowing force of the lower extremities under the conditions of rest and the isometric head extension with the cranio-cervical spine maintained in a flexed position against maximum resistance. After each cervical level was identified, participants were asked to perform isometric extension of the head using maximum force, against a board placed behind the head at occipital level, with the cranio-cervical spine maintained in a flexed position. Participants practiced this isometric head extension maneuver twice under the supervision of the rater, to ensure that the head and neck underwent a pure horizontal movement without

Fig. 1. Images obtained from magnetic resonance imaging (MRI). (A) Sagittal view of the cervical spine. Multiple transverse slices were scanned from the superior border of the C3–C4 intervertebral disk to the inferior border of the C6–C7 intervertebral disk at 5-mm intervals. (B) The square indicates the image view relevant to the ultrasonography (US) view. The region within white line is the area of the right side of the cervical multifidus.


cranio-cervical spine moving toward flexion or extension. Because the participants were asymptomatic, each individual maximum force was assumed to be the same for each participant. The maximum contraction could activate the neural pathway and neuromuscular junction as well as the contractile and non-contractile tissue of the muscles. The two practice isometric head extension helped to maintain the condition of the cervical muscle before measurement. US of the multifidus muscle on the right at the C4, C5, and C6 levels, in sequence, were recorded for the conditions of rest and isometric head extension contraction. Proper positioning of the participant was ensured after each test, and the same procedure was repeated twice in a 20-min interval by the same rater.

289

The multifidus muscle is located lateral to the junction of the spinal process and lamina; dorsolateral to the lamina of vertebrae, medial to the articular process, and ventral to the interfaces of the fasciae of the semispinalis cervicis (SSC) muscle (Fig. 3A) (Gerhardt and Frommhold, 1991). The orientation of the transducer was adjusted to bring it horizontal to the transverse plane. To obtain the clearest image, the US transducer was held by a custom-designed device and was tilted slightly (up or down) until the clearest interface lines could be observed. An US image was obtained for each cervical level, and the boundary of the muscle was identified and measured manually using a custom-written C++ computer graphic program. The muscle thickness, width, and area of the cervical multifidus muscle were measured at the C4–C6 levels. The muscle thickness was measured as the largest distance from the dorsal to the ventral boundary of the multifidus, and the width was measured as the largest distance between the medial and lateral boundaries of the muscle (Fig. 3). The area was measured as the region along the boundary lateral to the junction of the spinal process and lamina; dorsolateral to the lamina of the vertebrae, medial to the articular process, and ventral to the interfaces of the fasciae of the SSC muscle. The shape ratio was calculated as the lateral diameter (width) divided by the ventrodorsal diameter (thickness). 2.5. Validity of US

Fig. 2. Experimental setup for the isometric head extension with the cranio-cervical spine maintained in a flexed position.

Validity testing was carried out in 10 asymptomatic participants (three women and seven men; age, 26.273.5 years). Each participant was asked to lie in a comfortable and relaxed prone position on an examination table, with both arms resting symmetrically beside the body. A head support was used to maintain the head in the neutral position. Flexion and extension of the neck were controlled by maintaining the posterior part

Fig. 3. Ultrasonography (US) images of the transverse plane of the right side of the cervical multifidus at the C5 level at rest (A) and during contraction (the isometric head extension with the cranio-cervical spine maintained in a flexed position) (B). The borders of the multifidus muscle are: medially, the spinous process (SP); dorsolaterally, the semispinalis cervicis (SSC); laterally, the articular process (AP); and inferiorly, the lamina. The region within the dotted line indicates the area of the multifidus muscle. T ¼ thickness, and W ¼ width.


290

of the head and the thoracic region on a horizontal level with use of an oil-filled inclinometer. The spinal process level was identified, and MRI of the cervical multifidus at each level was carried out, followed by US. The multifidus was located lateral to the junction of the spinal process and lamina; dorsolateral to the lamina of vertebrae, medial to the articular process, and ventral to the interfaces of the fasciae of the SSC muscle (Fig. 3A) (Gerhardt and Frommhold, 1991). As with testing for reliability, the transducer was held by a custom-designed device and was tilted slightly (up or down) until the clearest interface lines could be observed. The orientation of the transducer was adjusted to bring it perpendicular to the horizontal plane, calibrated by a vertical line using a plume. US and MRI images obtained for each cervical level and the boundary of the muscle were identified and measured manually using a custom-written C++ computer graphic program. 2.6. Statistical analysis The reliability of intra-rater, intersession was described using CVw and between-subject coefficient of variation (CVb). The variance within each subject was presented as the percentage of the average coefficient of variation (CV) for each subject. The variance between subjects was presented as the percentage of the average CV from each measurement (Rainoldi et al., 1999). Thus, vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi , , ffi u n uP m 2 P t xij xi ðm 1Þ xi i¼1

CVw ¼

CVb ¼

3. Results 3.1. Reliability and validity of US The CVw for muscle thickness, width, area, and shape ratio of the cervical multifidus muscle at rest were within 10%, with the exception of the shape ratio at the C4 level (CVw ¼ 11.39%). During contraction, the CVw was within 10% for thickness at the C4–C6 levels, width at the C4 and C5 levels, and area at the C4 level. Hence, the CVw was less than 10% for the measurement of muscle thickness at the C4–C6 levels, both at rest and during contraction (Table 2). There were significant relationships between the muscle thickness at the C4–C6 levels as measured by US and MRI (Table 3). Data from the three cervical levels were pooled for regression analysis and determination of R2 values. The R2 values for thickness at the C4, C5, and C6 levels and pooled value were 0.42, 0.64, 0.43 (Table 3, Po0:05), and 0.60 (y ¼ 1:0239x20:0005, Po0.05). The R2 values for width were 0.02, 0.11, 0.22 (Table 3, P40:05), and 0.02 (y ¼ 0:1275x þ 1:5184, P40.05); the R2 values for area 0.39 (Table 3, Po0:05), 0.11, 0.29, (P40.05), and 0.25 (y ¼ 0:6351x þ 0:4244, P40.05); and the R2 values for shape ratio were 0.13, 0.28, 0.02 (Table 3, P40:05) and 0.05 (y ¼ 0:2247xþ 2:0092, P40.05). 3.2. Change in muscle dimensions

j¼1

, n s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi m n 2 P P xij xj xj ðn 1Þ j¼1

SPSS software (version 10.0; SPSS Inc., Chicago, Illinois, USA).

i¼1

m

.

In these equations, n denotes the number of subjects (n ¼ 10), m denotes the number of repeated measurements (m ¼ 2), and X ij denotes the US measurement for the ith subject in the jth repeated measurement. A limit of agreement was used to evaluate the differences between the geometric parameters of the cervical multifidus as measured on US and MRI. Additionally, the relationships of the values obtained with US and MRI were determined by regression analysis. The changes in the thickness of the cervical multifidus muscle at the C4–6 levels and at two grades of resistance (rest and contraction) were evaluated by two-way analysis of variance (ANOVA) with repeated measurement (three cervical levels [C4, C5, and C6] two grades of resistance [rest and contraction]) with use of

There were no significant interactions between the measurement of muscle thickness at the three cervical levels and the two grades of resistance (F(2, 18) ¼ 2.620, P40.05). The main effect of contraction did differ significantly (F(1, 9) ¼ 22.875, Po0.05, Table 2; Fig. 3A and B) and the main effect of cervical level did not differ significantly (F(2, 18) ¼ 1.353, P40.05). The muscle thickness increased significantly during isometric head extension with the cranio-cervical spine maintained in a flexed position against maximum resistance, indicated that the muscle thickness changes during contraction (Table 2, Fig. 4).

4. Discussion In the present study, we attempted to determine whether US was a reliable and valid method for measuring the dimensions of the cervical multifidus at rest and during contraction. Intra-rater, intersession reliability for measurement of muscle thickness was acceptable during both rest and contraction. Compared


291

Table 2 Coefficients of variation (within subject [CVw] and between subjects [CVb] for measurements of the dimensions of the cervical multifidus at rest and during contractiona CVw (%)b

Measurement

CVb (%)b

Session 1

Session 2

Rest Thickness (cm) C4 C5 C6

0.7270.10 0.6870.08 0.7770.09

0.7070.07 0.7170.12 0.8270.12

7.89 (4.27–11.49) 7.22 (4.55–9.88) 6.49 (3.81–9.16)

12.78 (12.21–13.35) 14.57 (13.83–15.30) 14.21 (13.39–15.03)

Width (cm) C4 C5 C6

1.7070.20 1.8070.25 1.9070.38

1.7670.38 1.8370.34 2.0770.40

8.22 (1.45–14.99) 3.88 (1.82–5.93) 4.57 (0.77–8.37)

16.99 (14.15–19.81) 16.42 (12.95–19.88) 19.85 (10.16–29.54)

Area (cm2) C4 C5 C6

0.9270.16 0.9670.16 1.2070.29

0.9170.05 1.0470.27 1.3970.33

7.89 (4.27–11.49) 3.92 (0.02–7.88) 6.49 (3.81–9.16)

19.11 (17.11–21.10) 21.87 (19.87–23.85) 24.35 (18.32–29.54)

Shape ratio (lateral/ventrodorsal dimension) C4 2.4070.47 C5 2.6770.52 C6 2.5470.63

2.5370.60 2.6470.66 2.5870.56

11.39 (1.82–20.96) 7.47 (1.25–13.69) 5.64 (1.50–9.78)

Contraction Thickness (cm) C4 C5 C6

1.1370.20c 1.197020c 1.1770.12c

1.1870.16 1.1670.15 1.1370.13

6.86 (3.63–10.11) 8.73 (1.64–15.83) 6.72 (3.58–9.85)

15.87 (13.84–17.88) 15.26 (11.86–18.64) 11.83 (10.81–12.85)

Width (cm) C4 C5 C6

1.4370.34 1.3570.33 1.3470.42

1.4270.53 1.3970.55 1.4670.48

7.65 (3.71–19.81) 8.67 (3.77–13.57) 10.43 (0.32–20.54)

19.73 (14.74–24.72) 18.64 (13.82–23.46) 20.80 (5.77–35.81)

Area (cm2) C4 C5 C6

1.3570.30 1.3170.31 1.3370.31

1.3770.41 1.2670.29 1.2770.26

8.92 (2.33–15.51) 10.72 (4.64–16.79) 10.64 (6.46–14.81)

23.89 (15.47–32.29) 16.46 (7.8–25.11) 13.87 (9.65–18.08)

1.2270.51 1.2470.58 1.3270.49

12.89 (3.91–21.86) 11.56 (4.56–18.56) 11.16 (0.61–22.94)

16.79 (4.46–29.10) 21.92 (2.42–46.27) 21.63 (9.47–33.78)

Shape ratio (lateral/ventrodorsal dimension) C4 1.2870.36 C5 1.1870.44 C6 1.1770.39

21.71 (5.08–38.32) 22.24 (4.38–40.08) 12.47 (15.4840.77)

a

Measurements given as the mean and standard deviation (SD). Value presented as mean ((mean1 SD)(mean+1 SD)). c Significantly different from the value measured at rest, Po0.05. b

with MRI, US had an acceptable validity for measuring the thickness of the cervical multifidus muscle but not for measuring its width and area. Additionally, US detected significant increases in muscle thickness during contraction. A significant linear relationship of measurements was found between the two modalities for muscle thickness but not for muscle width, area, or shape ratio. In a previous study, no difference was found between US and MRI measurements of the cross-sectional area of the lumbar multifidus (Hides et al., 1995). The cervical multifidus muscle is much smaller than the lumbar multifidus, and the variation in measurement may have

had a significant influence on the R2 value. In addition, the lateral boundary of the cervical multifidus is not distinct on either US (Kristjansson, 2004) or MRI because its attachment, the articular processes of the cervical vertebrae (Anderson et al., 2005), are the common insertion areas of the semispinalis capitis muscle (Platzer, 2005) or another passing muscle fiber (Stokes et al., 2005). Thus, the lack of a significant relationship between the US and MRI measurements of width, area, and shape ratio may have been due to the size and anatomic structure of the cervical multifidus. Another possible reason for the lack of a significant relationship is the anatomical alignment of the cervical


292

Table 3 Limit of agreement and R2 values for the dimensions of the cervical multifidus muscle at the C4, C5, and C6 levels, as measured by ultrasonography (US) and magnetic resonance imaging (MRI)a US

MRI

Mean differenceb

R2

p

Thickness (cm) C4 C5 C6

0.6770.14 0.7070.20 0.7370.09

0.7070.12 0.6770.15 0.7070.06

0.0170.16 0.00370.20 0.0270.14

0.42 0.64 0.43

0.04 0.005 0.03

Width (cm) C4 C5 C6

1.7270.28 1.6570.20 1.7470.19

1.4370.16 1.4170.23 1.6470.40

0.2970.62 0.2470.70 0.0970.70

0.02 0.11 0.22

0.69 0.33 0.16

Area (cm2) C4 C5 C6

0.9970.35 0.9570.30 1.0470.28

0.8870.20 0.8270.22 0.9970.28

0.1070.54 0.1270.60 0.0470.54

0.39 0.11 0.29

0.05 0.34 0.10

2.0870.46 2.1870.57 2.4070.61

0.4271.06 0.3371.17 0.1171.80

0.13 0.28 0.02

0.31 0.11 0.66

Shape ratio (lateral/ventrodorsal dimension) C4 2.5070.48 C5 2.5170.63 C6 2.5070.57 a

Values given as the mean and standard deviation (SD). Value given as the mean and two SDs.

b

thickness (cm)

1.5

∗

rest contraction ∗

∗

1

0.5

0 1

2 level

3

Fig. 4. Comparison of thickness measured at rest and during contraction (the isometric head extension with the cranio-cervical spine maintained in a flexed position) at the three cervical levels. *Significant difference between the values measured at rest and during contraction, Po0.05.

multifidus in relation to the resolution of the US image. The muscle tissue of the cervical multifidus was aligned from the spinal process to the articular process, which was perpendicular to the traveling direction of the soundwave (Kremkau, 1989). This alignment made the thickness at the muscle interface easily identifiable because the axial resolution of US is better than the horizontal resolution (Kremkau, 1989). Also, the presence of fat tissue around the muscle fascia results in slightly different tissue boundaries seen on images from the two modalities. Fat tissue between the layers of muscle fascia may influence the accuracy of the US measurement. Fat tissue could not be easily distinguished from muscles on US because both were

identified by a hyperecho (Hides et al., 1992; Stokes et al., 2005). In contrast, because the relaxation time for fat tissue is shorter than that for water or protein, fat tissue could be distinguished more clearly from muscle on MRI (Westbrook and Kaut, 1993). Thus, the use of US to measure the cervical multifidus in sedentary adults with possible fat infiltration requires further verification. The different scanning planes of US and MRI may also cause differences in the measurements for the two modalities. All participants were placed in the same comfortable and relaxed prone position on an examination table, with both arms resting symmetrically beside the body and the head maintained in the neutral position by a head support. However, due to lordosis, as Hides et al. (1995) discussed, if MRI slices are not manually positioned perpendicular to the muscle at each vertebral segment, the images will pass obliquely through the muscle and show different cross-sectional areas. Indeed, in the cervical region, lordosis also plays a role in the measurement of muscle thickness. Even if osseous landmarks, such as lamina and articular processes, are used to locate the plane of scanning by US and MRI, the scanning plane by US would not fine adjusted to be perpendicular to the muscle fiber to obtain a clearest US image, which may explain the differences between the US and MRI measurements. The measurement for the centre segment of the cervical lordosis (C4) was estimated with less error; this was supported by our data (Table 3). The present study was the first, to our knowledge, to document contraction of the cervical multifidus with use of US. Several cervical muscles, including the splenius


capitis (Soltani et al., 1996), semispinalis capitis (Rezasoltan et al., 1998; Rezasoltan et al., 2002), and multifidus (Kristjansson, 2004) have been measured with US during rest, with acceptable reliability. Previous studies have also demonstrated acceptable reliability of US measurements of the lumbar multifidus during rest (Hides et al., 1992; Hides et al., 1995). A significant change in the area of the semispinalis capitis muscle has been observed during contraction (Rezasoltan et al., 2002). Good reliability (ICC ¼ 0.98) of the US measurement of thickness of the transversus abdominis during contraction has also been reported (McMeeken et al., 2004), and there was a significant change in thickness during abdominal holding (Hodges et al., 2003) and the abdominal hollowing maneuver (McMeeken et al., 2004). Through a strict protocol, we were able to obtain acceptable reliability of measurements during contraction of the cervical multifidus, which is smaller than its lumbar counterpart. Contraction was achieved with the isometric head extension, which has been shown in this study to activate the cervical multifidus in adults. A limitation to using US to measure muscle thickness is the assumption of constant sound velocity in different tissues, including skin, fat tissue, fascia, and muscles (Kremkau, 1989). Therefore, the muscle thickness measured by US is an estimated value. To reduce these errors, either normalization (dividing the value during contraction by the initial value of the muscle geometry) or continuous US measurements should be considered in a future study to investigate the change pattern of the muscle under contraction. 5. Conclusion Using noninvasive, real-time US with a standard protocol, the thickness of the cervical multifidus muscle at the C4, C5, and C6 levels in asymptomatic young adults could be quantified during contraction. The application of this methodology for individuals with chronic neck pain should be further explored. References Anderson JS, Hsu AW, Vasavada AN. Morphology, architecture, and biomechanics of human cervical multifidus. Spine 2005;30:E86–91. Campbell WW, Vasconcelos O, Laine FJ. Focal atrophy of the multifidus muscle in lumbaosacral radiculopathy. Muscle & Nerve 1998;21:1350–3. Cote P, Cassidy JD, Carroll L. The Saskatchewan Health and Back Pain Survey. The prevalence of neck pain and related disability in Saskatchewan adults. Spine 1998;23:1689–98. Cote P, Cassidy JD, Carroll L. The factors associated with neck pain and its related disability in the Saskatchewan population. Spine 2000;25:1109–17. Esformes R, Narici MV. Measurement of human muscle volume using ultrasonography. European Journal of Applied Physiology 2002; 87:90–2.

293

Gerhardt P, Frommhold W. Altas of anatomic correlations in CT and MRI. New York: Thieme Medical Publishers, Inc.; 1991. Hayashi N, Masumoto T, Abe O, Aoki S, Ohtomo K, Tajiri Y. Accuracy of abnormal paraspinal muscle findings on contrastenhanced MR images as indirect signs of unilateral cervical rootavulsion injury. Radiology 2002;223:397–402. Hides JA, Cooper DH, Stokes MJ. Diagnostic ultrasound imaging for measurement of the lumbar multifidus muscle in normal young adults. Physiotherapy Theory and Practice 1992;8:19–26. Hides JA, Stokes IA, Saide M, Jull GA, Cooper DH. Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 1994;19:165–72. Hides JA, Richardson CA, Jull GA. Magnetic resonance imaging and ultrasonogrpahy of the lumbar multifidus. Spine 1995;20:54–8. Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine 1996;21:2763–9. Hodges PW, Richardson C. Feedforward contraction of transversus abdominis is not influenced by the direction of arm movement. Experimental Brain Research 1997;114:362–70. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 1996;21:2640–50. Hodges PW, Pengel LHM, Herbert RD, Gandevia SC. Measurement of muscle contraction with ultrasound imaging. Muscle & Nerve 2003;27:682–92. Ito M, Kawakami Y, Ichinose Y, Fukashiro S, Fukunaga T. Nonisometric behavior of fascicles during isometric contractions of human muscle. Journal of Applied Physiology 1998;85:1230–5. Kremkau FW. Diagnostic ultrasound: principle, instruments, and exercise. Philadelphia: W.B. Saunders Company; 1989. Kristjansson E. Reliability of ultrasonography for the cervical multifidus muscle in asymptomatic and symptomatic subjects. Manual Therapy 2004;9:83–8. Kristjansson E, Jonsson Jr. H. Is the sagittal configuration of the cervical spine changed in women with chronic whiplash syndrome? A comparative computer-assisted radiographic assessment. Journal of Manipulative Physiological and Therapeutics 2002;25:550–5. Kristjansson E, Leivseth G, Brinckmann P, Frobin W. Increased sagittal plane segmental motion in the lower cervical spine in women with chronic whiplash-associated disorders, grades I–II: a case–control study using a new measurement protocol. Spine 2003; 28:2215–21. McMeeken JM, Beith ID, Newham D J, Milligan P, Critchley DJ. The relationship between EMG and change in thickness of transversus abdominis. Clinical Biomechanics (Bristol., Avon.) 2004;19: 337–42. Panjabi M, Abumi K, Duranceau J, Oxland T. Spinal stability and intersegmental muscle forces. A biomechanical model. Spine 1989;14:194–9. Platzer W. Locomotor system, color atlas of human anatomy, vol. 1. New York: Thime Stuttgart; 2005. Rainoldi A, Galardi G, Maderna L, Comi G, Lo Conte L, Merletti R. Repeatability of surface EMG variables during voluntary isometric contractions of the biceps brachii muscle. Journal of Electromyography Kinesiology 1999;9:105–9. Rezasoltan A, Kallinen M, Malkia E, Vihko V. Neck semispinalis capitis muscle size in sitting and prone positions measured by realtime ultrasonography. Clinical Rehabilitation 1998;12:36–44. Rezasoltan A, Ylinen JJ, Vihko V. Isometric cervical extension force and dimensions of semispinalis capitis muscle. Journal of Rehabilitation Research and Development 2002;39:423–8. Soltani AR, Kallinen M, Malkia E, Vihko V. Ultrasonogrpahy of the neck splenius capitis muscle. Acta Radiologica 1996;37:647–50. Stokes M, Rankin G, Newham DJ. Ultrasound imaging of lumbar multifidus muscle: normal reference ranges for measurements and


J.-P. Lee et al. / Manual Therapy 12 (2007) 286–294

practical guidance on the technique. Manual Therapy 2005;10: 116–26. Ueki J, DeBruin PF, Pride NB. In vivo assessment of diaphragm contraction by ultrasound in normal subjects. Thorax 1995;50: 1157–61.

Westbrook C, Kaut C. MRI in practice. London: Oxford Blackwell Scientific Publications; 1993. Wu JP, Lee JP, Mao HY, Yu MH, Wang SF. Reliability of locating cervical spinal levels by palpation and ultrasonography. Formosan Journal of Physical Therapy 2005;30:80–7.

Manual Therapy (2007) 12(3), 295

Diary of events

First international Fascia Research Congress Basic Science and Implication for Conventional and Complementary Health Care 4–5 October 2007 The Conference Center, Harvard Medical School Boston MA http://www.fascia2007.com

10th International Conference in Mechanical Diagnosis and Therapy — The Evidence Mounts 23–25 March 2007 Queenstown, New Zealand Honorary Chairman: Robin McKenzie Presented by: The McKenzie Institute International For more information visit: www.mckenziemdt.org

5th International Course on the Hand October 21–25, 2007 Target audience: colleagues of the following disciplines; physical medicine and rehabilitation, plastic- and hand surgery, physical- and occupational therapy and other health care professionals, interested in the topic of the hand Lectures include: Prof. Dr. S.E.R Hovius, Ton A.R. Schreuders PT, PhD and G. Van Strein MSc Accreditation applied for at the EACCME (Accreditation Council) of the European Union of Medical Specialists (UEMS) More information and registration: website: www.vitalmedbodrum.com E-mail: [email protected]

4th Low Back Pain Symposium April 30–May 3, 2007 Target audience: colleagues of the following disciplines; physical medicine and rehabilitation, orthopaedic surgery, neurosurgery, physical-, occupational-, manual- and Mensendieck therapy. Moreover, company doctors, medical advisors of insurance companies and other health care professionals interested in the topic of low back pain. Chairmen: Prof. Dr. Henk J. Stam, Prof. David Niv M.D. FIPP, Prof. Dr. Mehmet Zileli Accreditation applied for at the EACCME (Accreditation Council) of the European Union of Medical Specialists (UEMS) More information and registration: website: www.vitalmedbodrum.com E-mail: [email protected]

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

6–10 July 2007, Singapore Changing Pain and Movement Behaviour – A Classification Based Approach to the Management of Chronic Low Back Pain Disorder by A/P Peter O’Sullivan for information on the workshop, please e-mail [email protected]

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]

1st World Congress on Manual Therapy July 27th–August 1st, 2007 Mangalore, India Organising Chair: Prof. Umasankar Mohanty, President, Manual Therapy Foundation of Indias For Registration and more information Website: www.wcomt.com E-mail: [email protected]

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.

295

ARTICLE IN PRESS

Manual Therapy 12 (2007) e1 www.elsevier.com/locate/math

Letter to the Editor Letter to the Editor regarding a study titled ‘‘Diagnosis of sacroiliac joint pain: Validity of individual provocation tests and composite of tests’’ [Manual Therapy 10 (2005) 207–218] To the Editors: I always appreciate it when new evidence comes out that may help clinicians detect when patients have sacroiliac joint dysfunction. I am unsure however whether this paper adds anything new to the existing literature. The difficulty lies in the patient’s that were selected for this study and their credibility as ‘‘trustworthy’’ respondents. The inherent problem of this study is in choosing a patient population where susceptibility bias is intrinsically possible. According to Table 1 the mean duration of symptoms for this cohort of patients was 42.1 months or over 3 years, the time off work was nearly 18 months or one and a half years, and


finally the Roland-Morris pain questionnaire’s mean score was about 76%! This data suggests a distorted assembly and one that I could hardly have faith in when trying to determine if an 80% subjective improvement in pain reduction occurred after sacroiliac joint injection. The target or intended patient population of this study should resemble characteristics of low back pain patients frequently seen in the clinic and not one that is reminiscent of a cohort of workman’s compensation or litigating patients. I realize that there is no ‘‘perfect’’ study and I applaud the authors for trying to help improve the diagnostic accuracy of detecting sacroiliac joint dysfunction. Best Regards, Michael T. Cibulka Maryville University, Program in Physical Therapy, St. Louis, MO 63105, USA

ARTICLE IN PRESS


Book review Neuromusculoskeletal Examination and Assessment. Nicola J. Petty. Elsevier, Churchill Livingstone, Kidlington, UK (2006) (413pp., price £27.99, ISBN: 044310204X) The primary aim of this book is to provide a systematic approach to the assessment of the neuromusculoskeletal system. As the third edition in the series it covers the same scope as previous editions however the layout is clearer with improved tables and figures directing the reader to pertinent points. The book devotes a chapter to each region of the neuromusculoskeletal system with separate chapters on subjective examination and physical examination. There is some repetition of the subjective and physical assessments section in each chapter however this only helps to reiterate salient points and limits page turning. This edition explores clinical reasoning in more depth than its predecessor, both towards planning the physical examination and for follow on treatment. The author has also made subtle changes in this edition which helps the book move away from it’s ‘‘joint’’ bias and embraces the neuromuscular components of assessment and reasoning. This is particularly evident in the clinical

doi:10.1016/j.math.2006.09.005

reasoning exercises. Throughout, references have been updated however there is limited critique of the quality of the evidence presented. The physical examination section is an excellent toolbox for undergraduates and practitioners with concise instructions and updated photographs albeit with recognition that the tests need to be flexibly implemented to suit both patient and therapist. Comment on the validity and reliability of testing procedures would have been useful although it is helpfully explained, particularly for undergraduates, that no one test can be truly diagnostic for any single pathologic entity. Overall this is a very practically based book which guides the reader through the assessment process with useful prompts to assist clinical reasoning. It is therefore a valuable resource for undergraduates, new graduates and also as a reference text for clinicians working in any neuromusculoskeletal setting.

Caroline Miller University Hospital Birmingham Foundation Trust, UK

ARTICLE IN PRESS


Book Review Maintaining Body Balance Flexibility and Stability. Chaitow L. Elsevier-Churchill Livingstone, Kidlington, UK (2004) (194pp., price £21.99, ISBN 0443073511). Due to daily load our body physically adapts to certain situations. Sometimes the body adapts well, sometimes poorly. Leon Chaitow shows us that through an application of simple techniques we can learn a better way of physical adaptation. The book is written for patients as well as practitioner or therapist. It clearly says in the preface that it is not a substitute for professional attention and treatment, but can be used as support and guidance material. It contains more than 180 pages of well-illustrated techniques and exercises with separate exercise sheets inside the back cover. The author describes a series of exercises to influence the soft tissues. The basic idea is that, in the case of articular problems, attention should first be given to the soft tissues, in order to normalize joint function. It is the soft tissues that support and move the joint. The author, being a registered osteopathic practitioner himself, uses osteopathy as the primary source of many of the

doi:10.1016/j.math.2006.09.004

methods described. According to Chaitow the book also relies on work of medical physicians, physiotherapists, exercise physiologists, chiropractors, massage therapists and others, however no references are given throughout the text. The first chapter is the most theoretical of all. Different forms of muscle energy technique (MET), like reciprocal inhibition and postisometric relaxation, are explained. The other chapters describe methods on testing muscle length, treatment and self-treatment MET, trigger points, flexibility, stability, balance, positional release technique and strain and counter strain. The book is written in a way that every patient can understand with clear step-by-step instructions on each exercise method. For the therapist it is a work which gives a lot of practical information and inspiration for treatment. Ulrike Van Daele Higher Institute of Physiotherapy, Hogeschool Antwerpen, Belgium Faculty of Physical Education and Physiotherapy, Vrije Universiteit Brussel, Brussel, Belgium

ARTICLE IN PRESS


Book Review Meaningful Motion: Biomechanics for Occupational Therapists. 1st ed., S.J. Spaulding, Elsevier, Amsterdam (2005). 187pp., ISBN:0443074399. This book has been specifically written for members of the occupational therapy community, as mentioned in the preface, however, because understanding motion is one of the cornerstones of our profession too, this book could be useful for students, teachers and clinicians in the field of physical therapy/manual therapy. Spaulding summarizes in the first section of her book the recent and seminal research findings in the area of biomechanics, motor control and learning. The second section of the book gives the integration of the knowledge from the first section into the specific areas in which occupational therapists work: balance training, the environment, ergonomics and leisure. Perhaps the most interesting parts of the book are the comprehensive case studies, one in each chapter. The end of each chapter includes questions about the case and a list of questions in the form of laboratory exercises. This gives

doi:10.1016/j.math.2007.01.002

the reader the opportunity to think further about the material in the chapter and will help to apply the principles of biomechanics and motor control to practice. Despite finding the case studies invaluable, as a musculoskeletal physical therapist/manual therapist, this reviewer would love to see more patient cases with musculoskeletal disorders. Most cases describe patients with neurological disorders, which is perfectly normal for occupational health practitioners. In summary, this is a well-written book that deserves to be read, not only by occupational therapy students and practicing occupational therapists. This useful workbook could also broaden the understanding and advancing the practice of physical therapy students and practicing physical therapists/manual therapists.

Simon Brumagne Faculty of Kinesiology and Rehabilitation Sciences, University of Leuven, Leuven, Belgium E-mail address: [email protected]

ARTICLE IN PRESS


Book Review Clinical Application of Neuromuscular Techniques. Practical Case Study Exercises, L. Chaitow, J. DeLany. Elsevier/Churchill Livingstone, New York, NY (2005). (289pp., £21,99), ISBN: 0443100004. Clinical Application of Neuromuscular Techniques, Practical Case Study Exercises by Leon Chaitow and Judith DeLany provides a collection of 34 patient cases authored by 19 different ‘‘Neuromuscular Therapists’’ presenting a wide range of clinical presentations. Cases are systematically presented under the headings of Profile, Health or Family History, Presenting Complaint(s), Significant Contributing Factors, Clinical Evidence and Previous Treatment, Questions for Reader, Examination, Clinical Impression and Treatment or Action Suggested, and Key Points with some cases also including sections on Further Reading, References and Websites Worth a Visit. Interspersed through most cases are additional boxes with supportive academic content pertaining to aspects of the case and very good red flag boxes highlighting clinical features that warrant immediate concern and medical consultation. The intention of this book is to assist readers’ integration of neuromuscular techniques and complementary medicine (e.g. diet) into their practice. This is achieved through inclusion of posture, lifestyle, biomechanical and psychosocial considerations through all cases with corresponding management suggestions. Consistent with the book’s focus, physical impairment of muscle, typically in the form of trigger points, but also with respect to posture, habits of movement, muscle length and to some respect motor control, are thoroughly considered. While joint impairment is included, this is considerably more superficial in both assessment

doi:10.1016/j.math.2007.01.003

and management with a notable lack of contemporary content and referencing. The case authors’ clinical reasoning is essentially a summary of management suggestions, not explicitly linked to supporting or negating evidence, with no discussion of recommended outcome measures to monitor or thoughts on progression of treatment. While listing of management options in this manner is very useful, case reasoning could be developed and articulated better. An interesting, but also frustrating feature of this book is its extensive use of highlighting key words in red throughout its cases. Highlighted words can then be looked up in the index of the two companion books ‘‘Clinical Application of Neuromuscular Techniques, Volumes 1 and 2’’ for definition and explanation. Many of the words highlighted refer to assessment or management procedures that are somewhat specialised to ‘‘Neuromuscular Therapists’’ and their philosophy of practice, making the cases less accessible to manual therapists without this background. Musculoskeletal physiotherapy does not feature very positively, with the majority of the cases either highlighting the poor results of previous physiotherapy management received or making recommendations for osteopathic, chiropractic but not physiotherapy referral. As such, I anticipate this book is likely to be of greatest interest to ‘‘Neuromuscular Therapists’’. Mark Jones Postgraduate Coursework Masters Programs in Physiotherapy, School of Health Sciences, University of South Australia, Australia E-mail address: [email protected]

ARTICLE IN PRESS


Book review Rehabilitation of the Spine. A Practitioners Manual, 2nd ed. Liebenson C. Elsevier/Churchill Livingstone, Amsterdam/New York (2007) (972pp). ISBN:0781729971. With over 900 pages and 39 chapters, the size of this text is a reminder of the diversity of opinion, and methods of management, of spinal pain disorders. The author, Craig Liebenson has written or co-authored 14 chapters, and the remainder have been written by invited contributors. Much of the text is focused on the lumbar spine with relatively less attention given to cervical spine pain. Surprisingly, not one chapter is devoted to disorders of the thoracic spine and their management. The introductory chapters provide an overview of some fundamental concepts in spinal pain including the anatomical sources of pain, mechanisms of injury, pain physiology and the biopsychosocial model. These chapters are well written and supported by relevant literature. The chapters on assessment were somewhat disappointing as the focus was on the diagnostic triage of spinal pain, including screening for psychosocial ‘yellow flags’. The only chapter on physical examination from a physical therapy perspective was in relation to muscle imbalance, and the remaining 3 chapters in this section reviewed physical performance testing. What appears to be missing here is an overview of the clinical examination process for patients with spinal pain, and the clinical reasoning process, which would guide the interpretation of the information obtained. The absence of this information makes the application of the following chapters on physical treatment difficult to interpret. This would be particularly the case for students of the physical therapy professions, or clinicians looking for a stronger framework to support their approach to patient management. In the next section of the text there are 9 chapters which describe ‘acute care management’ which include treatment approaches such as massage, muscle release

doi:10.1016/j.math.2007.01.004

techniques, rehabilitation of breathing patterns, the McKenzie method, neural mobilisation and joint manipulation. Specifically how these treatment methods relate to the management of acute pain is not stated, and it seems that a number would be equally or better suited to the management of non-acute conditions. The following 10 chapters on ‘recovery care management’ mostly present various approaches to trunk muscle strengthening around the ‘stabilisation’ theme. While there is a lot of information presented and some useful ideas for exercise prescription, these chapters are somewhat repetitive, and again, in many cases do not provide a good framework in relation to the types of patient presentation where these approaches may be most useful. Equally disappointing were the chapters on ‘integrated management’ which surprisingly presented similar material to the preceding chapters, and with case studies linked to anatomical diagnoses such as ‘facet syndrome’ and ‘discogenic radiculopathy’. This approach was clearly inconsistent with the introductory chapters which highlight the problems with anatomical diagnoses, and emphasise a broader approach to the evaluation of ‘non-specific’ spinal pain. For clinicians with a sound approach to clinical reasoning and treatment prescription, this text may provide a range of new ideas or treatment techniques which could be integrated into patient management. However, for the inexperienced clinician or manual therapy student the range of treatment approaches presented may be somewhat overwhelming, and of less value in relation to the common problem of deciding which approach to treatment is likely to be most useful.

Steve Edmondston School of Physiotherapy, Curtin University of Technology, Perth, Australia E-mail address: [email protected]

ARTICLE IN PRESS


Book review Physiotherapeutic Management of Lumbar Spine Pathology, MacDonald David, Jemmett Rick. first ed. Novont Health (2005). (199pp.), ISBN: 9024234344. I was pleased to be asked to review this book as I am always on the lookout for books that underpin and extend my clinical practice using an up to date evidence base. However, the title was a little misleading as rather than a concise text on spinal pathology, it focused on the area of segmental dysfunction. Nevertheless, I found it an easy book to read and as the authors’ state, it is a ‘‘useful reference tool for evidence-based management of patients with segmental dysfunction’’. The first three chapters deal with an overview of relevant anatomy, biomechanics and spinal pathology including motor control dysfunction with an up to date reference list to back them up. The subsequent chapters deal with the assessment and management of segmental dysfunction with the authors proposing their own model

doi:10.1016/j.math.2007.01.005

of segmental dysfunction. These chapters are reinforced by clear photos to demonstrate points and techniques, although there is less research evidence presented in this part of the book to support the techniques. Chapter seven presents case studies, which are of benefit to clinicians and students new to this area, although there is a lack of objective measurement tools used throughout this section. Chapter 8 consists of invited peer review commentaries, which is uncommon in texts, but provides a balanced view of the text as a whole. Overall, the authors have summarised lumbar dysfunction and present a useful reference for clinicians and/or students in clinical practice.

Lyndsay Alexander School of Health Sciences, Robert Gordon University, Aberdeen, UK E-mail address: [email protected].

ARTICLE IN PRESS


Book review Principles of Neuromusculoskeletal Treatment and Management: A Guide for Therapist, N.J. Petty. Churchill Livingstone, New York, NY (2004). h47.99, (368pp.), ISBN: 0443070628. The aim of this book is to make explicit the underlying principles behind treatment and management of patients with neuromusculoskeletal disorders. In this extensive book, the author has succeeded in her aim. It is well written, easy to follow and it gives a lot of information about the state of the art in assessment and treatment. This book is divided into nine chapters in which items as assessment, function and dysfunction of joint, muscle and nerve, principles of treatment strategies and finally a chapter on principles of patient management by A. Moore. Each chapter is completed with an extensive reference list. Although most of the content is used within curricula of education in physical therapy and or manipulative therapy, this book brings all the knowledge together, with clear figures and illustrations. The content is based

doi:10.1016/j.math.2007.01.006

on fundamental or basic research, and on that basis it tries to add knowledge to an evidence-based approach of patient management. Unfortunately, it does not answer the question whether a chosen approach in treatment of muscle, joint or nerve is really effective. The author states that ‘the best treatment is the one that improves the patient’s sign and symptoms in the shortest period of time.’ The reader will be curious about effectiveness and efficacy of chosen treatment strategies, but that was a priori, not the primary aim. However, in evidence-based management, this knowledge could make the clinical reasoning process more clear. This book is highly recommended for physical therapists and manipulative therapists as a reference book and for students as a textbook.

Jan M. Pool EMGO Institute, VU University Medical Center, Amsterdam, The Netherlands E-mail address: [email protected]

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Recommend Documents