FLEXION MOBILIZATIONS WITTHE IMMEDIATE EFFECTS ON RPAIN IN SUBJECTS WITH LOW

Kika Konstantinou, PhD, MSc,a Nadine Foster, DPhil, PGCDavid Baxter, DPhil,d Christine Wright, C. Math,e and Ala

ABSTRACT

Objective: This study investigates the immediate effects oon spinal range of movement in individuals with low back p

been made to describe the clinical profiles of subjects who

u

6

ip

a

bl

It ha

expe

live

with

tudy

f all

(or

iro-

ithin

age-

term

hes.

However, there is limited research evidence for the efficacy

and effectiveness of specific techniques as used by physi-

otherapists for patients with LBP. In addition, it is still unclear

a SpinHaywooUniversi

b Senior Lectur

siotherapy, School of Health Sciences, Uni-which patients with LBP benefit more from specific manual

therapy approaches.

Mulligan10,11 pioneered a relatively new concept in

manual therapy; these techniques are known asmobilizations

with movement (MWMs), or as SNAGs, an acronym for

bsustained natural apophyseal glides.Q These have beenextensively described elsewhere.11 In the case of lumbar

spine MWMs, the technique involves the application of an

accessory glide along the plane of the zygapophyseal (facet)

joint in a weight-bearing position during active movement.11,p.44

Senior Lecturer in Research Methods, School of HealthSciences, University of Birmingham, UK.

f Director, Institute for Musculoskeletal Research and ClinicalImplementation, Anglo-European College of Chiropractic, Bourne-mouth, UK.Submit requests for reprints to: Kika Konstantinou, PhD, MSc,

MCSP, mMACP, Primary Care Sciences Research Centre, KeeleUniversity, Keele, Staffordshire, ST5 5BG, UK(e-mail: k.konstantinou@cphc.keele.ac.uk).Paper submitted October 13, 2006; in revised form November

11, 2006; accepted December 6, 2006.0161-4754/$32.00versity of Birmingham, UK.d Dean of Physiotherapy, Centre for Physiotherapy Research,

University of Otago, New Zealand.eCare Sciences Resec Lecturer in Phy178

Copyright D 20doi:10.1016/j.jmment of LBP in Britain. Manual therapy is a broad

that incorporates many different concepts and approacer in Therapies (Pain Management), Primaryarch Centre, Keele University, UK.Results: Mean spinal range of movement increased significantly with the MWM intervention, as compared with theplacebo (true flexion: MWMs 49.28 [SD 16.4], placebo 45.38 [SD 14.1], P = .005; total flexion: MWMs 76.78 [SD 22.4],placebo 69.78 [SD 21.5], P = .005). Mean pain scores did not change.

Conclusions: The MWMs produced statistically significant, but small, immediate spinal mobility increases but no painreduction when compared with placebo. By introducing clinical judgment into the subject selection process for the trial, 19

(73%) of 26 subjects benefited from MWMs techniques in terms of range of movement and/or pain intensity, whereas

9 (35%) subjects showed such changes with the placebo intervention. (J Manipulative Physiol Ther 2007;30:178-185)

Key Indexing Terms: Physical Therapy; Range of Motion

s been estimated that 60% to 80% of all adults

rience low back pain (LBP) at some point in their

s, although not all seek medical care.1-3 A number of

studies suggest that between 10% and 50% of patients

LBP receive physiotherapy,4-6 with a more recent s

suggesting that physiotherapists treat approximately 9% o

the subjects with LBP in Britain annually.7 Manual

manipulative) therapy is part of osteopathic and ch

practic interventions and is an area of specialization w

physiotherapy that is most commonly used in the man8,9

al Physiotherapy Specialist/Research Physiotherapist,d Hospital/ Primary Care Sciences Research Centre, Keelety, UK.Method: A small-scale explanatory study was conductedassessors blinded. After assessment by physiotherapists, 2

thought to be appropriate for treatment with MWMs, partic

intervention in a randomized order. Lumbar spinal flexion

immediately before and after each intervention, using dou07 by by National University of Health Sciences.pt.2007.01.015H MOVEMENT TECHNIQUES:ANGE OF MOVEMENT ANDBACK PAIN

E,b Alison Rushton, EdD, MSc,c

n Breen, PhD, DCf

f flexion mobilizations with movement techniques (MWMs)

ain and also their impact on pain. A preliminary attempt has

were thought to benefit from MWMs.

sing a crossover design, placebo-controlled, with subjects and

subjects with low back pain with pain on lumbar flexion,

ated. Subjects received an MWM intervention and a placebo

nd extension and pain during flexion were recorded

e inclinometry and visual analogue scales.Mulligan proposes that these spinal techniques improve

signs and symptoms by directly facilitating the restricted

priateness and effectiveness of these techniques are based on

the elbow.14-17 Although this research provides a useful

Inclusion and Exclusion CriteriaThe study took place in 2 NHS physiotherapy outpatient

departments and included subjects referred for physiotherapy

by their general practitioner or consultant for LBP with or

without leg pain. The recruitment period lasted from April

2001 to March 2002. Subjects older than 18 years were

included in the study only if the clinical impression of the

treating clinician was that they had LBP appropriate for

physiotherapy treatment and that flexion MWMs were

appropriate and potentially beneficial. A relatively high

number of subjects (n = 96) were not included in the study

after assessment because they were not thought eligible for

flexion MWMs. The subjects selected had pain primarily

upon spinal flexion and ROM limitations in the same

direction. In addition, subjects had to understand written

and spoken English so they could provide informed consent.

Subjects were also excluded from participating if they had

known contraindications to manual therapy, had undergone

previous spinal surgery, or were pregnant. Figure 1 summa-

rizes the progress of all subjects screened for entry to the trial.

Sample SizeThe required sample size was estimated using the

Konstantinou et alJournal of Manipulative and Physiological Therapeutics

Spinal Flexion MWMs in LBPVolume 30, Number 3179basis, extrapolation of these results to the treatment of

subjects with LBP is not appropriate.

A recent survey in the UK specifically investigated the

use of MWM techniques in LBP management to describe

the practice of physiotherapists and explore the reported

effects of MWMs.18 The findings suggested that at least 1 in

3 therapists currently involved in LBP management uses

MWMs as part of his or her treatment approach and that

physiotherapists use their clinical decision making to select

subjects whom they feel might benefit from these techni-

ques. Therapists reported that the most common effects seen

immediately after the use of MWMs were increases in

patients active spinal range of motion (ROM) (54.4% of

survey respondents, n = 253) and pain relief (27.5%, n =

128). This was also reflected in the outcomes used to

evaluate improvement, with 47.6% (n = 198) of all

therapists most commonly using active ROM and 37%

(n = 154) most commonly using pain relief as an outcome

measure. The choice of outcomes reported by this survey

suggested that therapists choose MWMs as a form of

treatment based on the assessment of immediate changes in

spinal ROM impairments and/or pain responses.

Based on the survey findings,18 an explanatory small-

scale trial was designed in the UK National Health Service

(NHS) physiotherapy outpatient setting. The primary aim of

the study was to investigate the claimed immediate effects

of lumbar spine flexion MWMs on spinal ROM in a sample

of patients with LBP referred to physiotherapy and, secondly,

to investigate their impact on pain. In addition, a preliminary

attempt has been made to describe the group of patients

with LBP thought to be appropriate for MWM techniques,

in terms of clinical profile and response to treatment.

Flexion MWMs were investigated because they have been

reported to be the most often used in clinical practice.18

METHODS

Ethical approval for the study was obtained from the

Local Research Ethics Committees of Warwickshire and

Birmingham NHS Health Authorities, UK.

DesignThis small-scale, explanatory clinical study was a cross-whether they can bring about immediate changes in pain

status and spinal mobility in a pain-free fashion.11,p.44,45

To date, there is a lack of research on MWMs for the

spine; and the available studies have been related to

peripheral joint MWMs for the ankle12,13 and especiallymobility of the facet joints and simultaneously influencing

the mobility of the intervertebral joint. The clinical appro-over (same subject), subject- and assessor-blind, placebo-

controlled design investigating the immediate effects offlexion MWMs on 4 dependent variables: true and total

lumbar spine flexion ROM, total lumbar spine extension

ROM, and pain response during spinal flexion.Fig 1. Summary of screening process and entry of patients inthe trial.formula described by Howell19 (Fig 2 for details) for

matched-samples t tests, based on the change in lumbar

Outcome Measures and EquipmentThe main outcomes of interest were true and total lumbar

spine flexion ROM. Secondary outcomes of interest were

pain during lumbar spine flexion and total lumbar spine

extension ROM.

Active ROM for true and total lumbar spinal flexion and

total lumbar spine extension was measured with DI with the

subject standing. True lumbar flexion is the ROM measured

by placing one inclinometer on T12 and the other on S2 and

having the patient bend forward; the difference between

their readings (T12 S2) is thought to reflect the mobilityof the lumbar spine alone.21 Total lumbar flexion is the

reading from the inclinometer only on T12; this is thought to

represent lumbar spine mobility, pelvic mobility, and sof

tissue extensibility such as hamstrings, etc.21

The inclinometers used were spirit level clinical goni-

ometers with 2 points of contact and a reported accuracy, by

ured with a visual analogue scale (VAS).31-33 The VAS

Descriptive Baseline and Treatment Information

180 Journal of Manipulative and Physiological TherapeuticsKonstantinou et alMarch/April 2007Spinal Flexion MWMs in LBPtspine true flexion ROM. The true flexion ROM difference

of interest was set at 78, which is above the upper limit ofthe 95% confidence interval (CI) of the standard error of

measurement found by our double inclinometry (DI)

reliability study conducted before the trial. In the absence

of information relating to what constitutes a true difference

or a clinically important difference in ROM, 78 was chosenso that the anticipated difference in the flexion ROM

variable needed to be at least higher than the average

measurement error.20 Any such change can then be

considered with some degree of confidence as a real change

in subjects movement. The standard deviation (SD) was set

at 128, derived from the literature on normative ROM dataon asymptomatic and symptomatic populations.21-30 This

was later adjusted to 6.58 based on analysis of available datafrom the ongoing study.

Twenty-six subjects (15 men, 11 women) participated in

the study. This sample size was chosen so that the study

would have 90% power at a 5% significance level to detect a

difference in true lumbar spinal flexion ROM of 78 or more.Fig 2. Explanation of formula used for sample size calculations.To permit description of those subjects with LBP

thought to benefit from MWMs, the following were also

collected: subjects sex and age, their functional scores on

the self-completed Roland-Morris Disability Questionnaire

(RMDQ),34 duration of the current LBP episode (in months),

history of LBP, area of pain on body chart, easing and

aggravating factors, therapists clinical impression of the

subjects condition (diagnosis), subjects psychological

profile using the self-completed Distress and Risk Assess-

ment Method (DRAM),35 spinal level(s) treated, number of

sets, and repetitions and type of MWM techniques used by

the treating physiotherapists.

ProcedureConsenting subjects were told that the study was

investigating 2 different treatments and that they would

receive both in the same treatment session. Subjects

completed the RMDQ, and the history of the subjects back

problem was obtained. The physical examination included

assessment of active lumbar spine ROM, passive manual

examination of the lumbar spine, and any other tests the

treating clinicians felt necessary to carry out. Based on the

history and physical examination, a clinical impression was

formed by the physiotherapist and a decision was made

about whether the subject might benefit from the use ofconsisted of a 10-cm horizontal line with 2 word

descriptors at either limit of the scale: bno painQ at theleft-hand side and bmaximum painQ at the right-hand side.The subject was asked to draw a line across the scale

representing the severity of pain felt during active lumbar

spine flexion in standing position. The distance of the

mark from the lower limit of the scale was measured with

a ruler to the nearest 0.1 cm.

All measures were taken initially, then after the first

intervention, and again after the second intervention. The

same assessor, blinded to the order of intervention, took all

ROM measures for each subject.the manufacturer, of F18 (MIE Medical Research Ltd,Leeds, UK). An intratester reliability study conducted

before the clinical trial yielded intraclass correlation

coefficients of 0.81, 0.81, and 0.91 for true lumbar flexion,

total lumbar flexion, and total lumbar extension, respec-

tively. Fifteen subjects and 5 testers participated in the

reliability study. The subjects were measured on 3 occa-

sions, 5 minutes apart, by each of the 5 testers. The starting

position of the subject for the performance and measurement

of spinal movements was standardized in the same way for

both reliability study and clinical trial.

Pain intensity during lumbar spine flexion was meas-flexion MWMs. Subjects who were considered by therapists

to be eligible for the use of flexion MWMs were randomly

Upon completion of the intervention, the assessor returned

required to apply the flexion MWM technique: approx-

postplacebo. This use of baseline measurements provided a

more powerful test for a carryover effect by essentially

eliminating the variation between different subjects.39

Carryover or residual effect in crossover designs is the

possibility that part of the effect of the intervention

administered in the first period may remain during the

second period and therefore contaminate the effect of the

intervention administered in the second period.39,40 In this

study, the statistical method described by Everitt39 was used

to assess whether the effect of the first intervention

remained during the application of the second intervention.

Other possible effects are that there is a change over time

irrespective of treatment or the time of administration of the

intervention has an effect on outcome. These last 2

possibilities do not apply to this study.

As described by Everitt,39 if the first stage of analysis had

shown a significant carryover effect, it would have been

impossible to continue the study as a crossover design and it

would have been necessary to change to an independent

design using only 1 measurement from each subject.

However, based on the findings from this first stage (ie,

nonsignificant carryover effect, P = .202), the second stage

used a paired-samples t test on the data collected postMWM

intervention and postplacebo to test for an MWM effect.

VAS, visual analogue scale; RMDQ, Roland Morris disability question-

naire (scores 0-24, higher scores = higher disability); DRAM, distress and

risk assessment method; SD, standard deviation.a Scores, 0-24; higher scores = higher disability.

Konstantinou et alJournal of Manipulative and Physiological Therapeutics

Spinal Flexion MWMs in LBPVolume 30, Number 3181imately 3 minutes. This allowed the effects of time and

repeated testing to be taken into account and provided some

assessment of the bnonspecificQ effects of treatment.37,38

The placebo intervention described above was chosen

because it was considered to be nonspecific.

Data AnalysisAll data were analyzed using the Statistical Package for

the Social Sciences version 13 (SPSS Inc, Chicago, Ill).

Statistical analyses on the 2 primary outcomes (true and total

lumbar spine flexion) and the 2 secondary outcomes (pain

and total lumbar spine extension) were conducted using the

2-stage approach described by Everitt.39 The first stageand repeated the same measurement procedures of pain and

spinal ROM as for baseline. The assessor again left the room,

the second intervention was applied, and the assessor

returned and repeated the same measurements as above. At

the end of the session, the subject completed the DRAM.

InterventionsSix senior physiotherapists from the 2 physiotherapy

departments participated in the study. They had a mean of

9 years postgraduate experience, and all but 1 had been on

accredited MWM courses. In addition, before data collec-

tion, training sessions were held regarding the use and

application of MWMs and the standardization of procedures.

The flexion MWMs were applied either centrally or

unilaterally, at the discretion of the treating therapist, to the

symptomatic spinal level(s). The treatment consisted of

mobilizing between 1 and 3 levels, using 2 to 3 sets of 4 to

6 repetitions on each spinal level. Level of symptom

irritability guided the physiotherapist in the application of

the technique, reflecting normal clinical practice (eg,

subjects with nonirritable symptoms are thought to require

more repetitions or sets, whereas those with irritable, painful

symptoms require less, as described by Maitland36,p.107 and

Mulligan.11,p.46

The placebo intervention consisted of each subject lying

on the treatment couch in a comfortable position. This could

be either side lying, prone, crook lying, or supine. Subjects

were told that this position would enable the muscles around

the spine to relax and therefore reduce pain. The subject

remained in this position for the same amount of timeallocated to 1 of 2 groups by coin toss as they presented for

treatment. This was done by the treating clinician. The

assessor, blinded to intervention order, performed the base-

line measurements of spinal ROM and asked the subject to

rate their pain during flexion. After the baseline measure-

ments, the assessor left the room and the treating clinician

applied the intervention as dictated by the randomization.tested for a carryover effect by using an independent-samples

t test on the data collected at baseline, postintervention, andTable 1. Summary of demographic and descriptive clinicalinformation

Variable

Characteristics

Mean SD Range

Age (y) 38.3 11.7 18-61

Duration of symptoms (mo) 26.8 47.9 0.1-240

Pain intensity of current episode

at its worst (VAS, 0-10 scale)

7 2 3-10

Pain intensity during flexion at

assessment (VAS, 0-10 scale)

5.2 1.9 1.7-9.8

RMDQa 11.4 4.7 4-22

Sex 11 female, 15 male

Previous LBP history 12 yes, 14 no

Pain pattern (body chart) 20 LBP, 3 LBP

and leg pain, 3 leg

pain only

Clinical impression/diagnosis 8 arthrogenic,

3 discogenic,

12 description of

signs (neural,

flexion stiffness),

3 others

DRAM (2 missing values) 8 normal, 11 at risk,

4 distressed/somatic,

1 distressed/depressiveLevels of significance were set at 5% for primary

outcomes41,42 (true and total lumbar flexion) and at 1%

presented with improvements in flexion-extension ROM

CI 38.5-50.3 61.8-77.1 16.3-27.5 4.2-6.2

Postplacebo intervention

Mean (SD) 45.3 (14.1) 69.7 (21.5) 21.2 (11.1) 4.3 (2.2)

182 Journal of Manipulative and Physiological TherapeuticsKonstantinou et alMarch/April 2007Spinal Flexion MWMs in LBPfor secondary outcomes41,42 (lumbar spine extension, pain).

CI 39.6-51.0 61.0-78.4 15.1-27.3 3.1-5.5

PostMWM intervention

Mean (SD) 49.2 (16.4) 76.7 (22.4) 24.0 (11.0) 4.2 (2.5)

CI 42.6-55.8 67.7-85.8 18.0-30.0 2.8-5.6

Difference (MWM placebo)Mean (SD) 3.9 (6.5) 7.0 (9.3) 2.8 (6.8) -0.1 (0.3)

CI 1.3-6.5 3.3-10.8 0.9-6.5 0.8-0.6P value

(2 tailed)

.005 .005 .045 .800

Mean difference = MWM intervention placebo intervention.a 95% CI.b 99% CI.Table 2. Summaries for values at baseline, postplacebo, andpost-MWMs, and P values from the tests on the effects of MWMscompared with placebo

Summary

Outcome measure

True

flexiona

ROM (8)

Total

flexiona

ROM (8)

Total

extensionb

ROM (8)

Pain

scoresb

(cm)

Baseline measurement

Mean (SD) 44.4 (14.7) 69.5 (19.0) 21.9 (10.2) 5.2 (1.9)Summaries of the data were provided through mean values,

SDs, and CIs for the mean (95% for the primary outcomes,

99% for the 2 secondary outcomes).

Cutoff scores were defined to categorize subjects into

those who seemed to have a real ROM change (ie, beyond

that which could be explained by measurement error) and

those who seemed to experience a clinically meaningful

reduction in pain intensity. In this study, a change N78 wasconsidered as representing a real change in the subjects

flexion-extension ROM. The cutoff score for significant

changes in pain intensity was based on Farrar,43 who

suggested a change of 2 points (cm) or more on the VAS to

be clinically important.

RESULTS

Twenty-six subjects participated in the study and

received both interventions. Baseline characteristics are

presented in Table 1. All data were normally distributed.

The first stage of analysis tested for carryover effect, and

this was found to be nonsignificant (P N .20).Table 2 presents the mean values, SDs, and appropriate

CIs for measurements at baseline, postplacebo, and post

MWM intervention. Differences in the mean values are

presented as an indication of MWM effects, and the

P values indicate significance or otherwise of these effects.

As can be seen, the mean difference between MWMs andand/or pain scores (z78, z2 points on VAS); these subjectswere considered as having responded favorably to the

MWM intervention. With the placebo intervention, 9 of the

26 subjects (35%) showed immediate improvements in

ROM and/or pain outcomes. Table 3 presents the number

and percentages of subjects who improved for each

outcome measure.

Description of Subjects,Profiles

The additional information collected on the clinical

profiles of subjects included in the study indicated that half

reported a first-time LBP episode and that most reported

pain confined to the low back with symptoms of more than

3 months duration. Most subjects presented with intermit-

tent pain of moderate to high intensity when at its worst,

scored moderately on the functional disability scale, and

were classified as bnormalQ or bat riskQ for psychologicaldistress. Most of the subjects reported activities and/orplacebo intervention was 3.98 for true lumbar spine flexion,78 for total lumbar spine flexion, and 2.88 for total lumbarspine extension. For pain, the overall mean difference

between MWMs and placebo was 0.1 cm.

The second stage of analysis indicated statistically

significant mean differences in true (P = .005) and total

lumbar flexion (P = .005) but not in lumbar extension (P =

.045) or pain intensity during lumbar flexion (P = .800).

According to the cutoff scores defined, immediately after

the MWM application, 19 (73%) of the 26 subjects

MWMs 11 (38.5%) 14 (53.8%) 8 (30.8%) 11 (42.3%)

Placebo 6 (23%) 5 (19.2%) 3 (11.5%) 6 (23%)Table 3. Number (and percentage) of subjects with improvementsaccording to cutoff scores for ROM and pain (z78 in ROM, z2points on VAS) based on raw data

Intervention

True

lumbar

flexion

Total

lumbar

flexion

Total

lumbar

extension Painpostures of flexion as aggravating factors and indicated rest

and cessation of the aggravating activity or posture as easing

factors (Table 4). During assessment, all subjects had pain

during lumbar flexion.

Nonrespondents to TreatmentOverall, 7 subjects did not change or were made worse

with the application of MWMs. Their data suggest that they

had longer symptom duration, had higher disability scores,

had pain that referred distally, were more often female, and

were reporting mostly a first-time LBP episode. Because of

the small number of subjects who did not benefit, statistical

comparison was not appropriate.

Konstantinou et alJournal of Manipulative and Physiological Therapeutics

Spinal Flexion MWMs in LBPVolume 30, Number 3183Table 4. Summary of characteristics of respondents and non-respondents to treatment

Variable

Respondents

(n = 19)

Nonrespondents

(n = 7)

Duration of

symptoms4 (mo)25.7 (mean)

(0.1 [3 d]-240)

29.9 (2-60)

RMDQa,4 10.6 (4-18) 13.6 (8-22)Pain intensity of

current episode

at its worst

7.0, SD 2.3 (3-10) 7.0, SD 0.8

(6-8)

Pain intensity during

flexion at

assessment

5.1, SD 1.9

(1.7-9.8)

5.3, SD 1.8

(1.7-6.8)

Sex4 6 female, 13 male 5 female, 2 maleLBP history4 10 yes, 9 no 2 yes, 5 noAggravating factors

reported by patients49 flexion,

5 extension,

3 flexion and

extension, 2 other

2 flexion,

4 flexion and

extension, 1 other

Pain pattern4(body chart)

12 LBP, 7 leg pain 3 LBP, 4 leg pain

Clinical impression/ 5 arthrogenic, 3 arthrogenic,DISCUSSION

The study was designed to test the efficacy of flexion

MWMs in terms of those clinical effects attributed to these

techniques in clinical practice. The findings suggested that

flexion MWMs produced a statistically significant immedi-

ate improvement in ROM compared with the placebo

intervention for true and total lumbar spine flexion, but

not for total lumbar spine extension or pain scores.

There is little previous information to guide decision

making about what degree of change in lumbar spine ROM

may be considered clinically important, despite the fact that

in physiotherapy practice, ROM is regularly assessed and

considered to be an important goal.6,8,44,45 To define a

meaningful change for this trial, any score in lumbar ROM

above the average measurement error (78, as recorded in theDI reliability study) was considered to represent a real

change in subject spinal mobility. Both true and total flexion

ROM measurements were well correlated at baseline as well

as after the MWMs intervention; this correlation is in

agreement with previous research.46

There are several potential reasons for the ROM increase

with the application of MWMs compared with the placebo.

diagnosis4 1 discogenic,11 description of

signs, 2 other

2 discogenic,

1 description

of signs, 1 other

DRAM

(2 missing values)

5 normal, 9 at risk,

3 distress/somatic

3 normal, 2 at risk,

1 distress/somatic,

1 distress/

depressive

The asterisk denotes difference between characteristics; however, this

was not tested statistically.

RMDQ, Roland Morris disability questionnaire; DRAM, distress and risk

assessment method; LBP, low back pain; SD, standard deviation.a Scores, 0-24; higher scores = higher disability.The placebo intervention consisted of a comfortable non

weight-bearing position adopted by the subject. The subjects

were led to believe that this resting position was a form of

treatment that might reduce their pain on subsequent

movement. It may therefore be suggested that the increases

in ROM with the application of MWMs were due to the

repeated flexion movement involved in the MWMs,

whereas the placebo did not involve movement. However,

designing a placebo that closely resembles an active

intervention without producing physiological effects is

exceptionally difficult in manual therapy.47,48 The placebo

in this study was designed to account for time and

measurement effects while considered nonspecific. The case

may be that repeating a spinal movement, although painful,

may lead to ROM increases simply by stretching effects.

However, not all subjects demonstrated increases in ROM,

although all of them performed repeated spinal flexion as

part of the treatment and measuring procedures, with a small

number of subjects getting worse. In addition, the statistical

analysis did not indicate any significant carryover treatment

effect, which may suggest that repeated movement was not

necessarily the explanation for ROM increases. Other

possible explanations could be the conditioning and expect-

ancy related to the credibility of the intervention and the

triggering of central inhibitory systems.37 Neurophysiolog-

ical muscular effects may play a role in ROM changes; total

flexion ROM (which combines pelvic mobility and ham-

strings extensibility) was greater than the true flexion ROM.

This suggestion seems to have some support from limited

experimental research to date.49

The only published study to date exploring the effect

of posteroanterior mobilizations on spinal ROM in a

population with LBP did not report any significant ROM

improvements but demonstrated reduction in pain scores.50

In the current study, despite the overall immediate,

positive differences in spinal flexion ROM, no inferences

can be made about the clinical relevance of the small

differences observed.

One reason why differences in pain did not reach

statistical significance may have been the sample size in

this study, which was calculated using estimated differences

in ROM. To detect a difference of 2 points on the VAS for

the study population (power 90%, significance 5%), we

estimate that a sample of 46 would have been required.

The preliminary investigation to describe the clinical

profiles of subjects who were thought to benefit from

MWMs suggested that the treating physiotherapists consid-

ered subjects appropriate for treatment with flexion MWMs

if they had mainly local LBP symptoms that were

predominately intermittent, aggravated by flexion activities

and/or postures; pain produced on active lumbar flexion;

and restriction of spinal flexion. These subjects seemed to

have moderate functional restrictions but not high psycho-logical distress as measured with the DRAM. Most could be

considered as having chronic LBP because they reported

designed to mirror clinical practice, that is, based on

Scand 1998;S281:28-31.2. Waxman R, Tennant A, Helliwell P. A prospective follow-

941-7.6. van der Valk RWA, Dekker J, van Baar ME. Physical therapy

for patients with back pain. Physiotherapy 1995;81:345-51.7. Maniadakis N, Gray A. The economic burden of back pain in

the UK. Pain 2000;84:95-103.8. Foster NE, Thompson KA, Baxter GD, Allen JA. Management

of nonspecific low back pain by physiotherapists in Britain andIreland. Spine 1999;24:1332-42.

9. Gracey JH, McDonough SM, Baxter GD. Physiotherapymanagement of low back pain. A survey of current practicein Northern Ireland. Spine 2002;27:406-11.

10. Mulligan BR. Manual therapy. bNags,Q bSnags,Q bMWMSQ etc.3rd ed. New Zealand7 Plane View Services Ltd; 1995.

11. Mulligan BR. Manual therapy. bNags,Q bSnags,Q bMWMSQ etc.4th ed. New Zealand7 Plane View Services Ltd; 1999.

12. OBrien T, Vicenzino B. A study of the effects of Mulligansmobilisation with movement treatment of lateral ankle pain

184 Journal of Manipulative and Physiological TherapeuticsKonstantinou et alMarch/April 2007Spinal Flexion MWMs in LBPindividual clinical judgment from the assessment findings.

This, in turn, is based on training, experience, and, to some

extent, treatment preferences.51,52 It is not unreasonable to

suppose that different clinicians may have selected different

types of subjects for the study or may have differently

influenced the studys results. However, in the absence of

robust classification systems or well-described clinical

subgroups within the population of nonspecific LBP, it is

difficult to standardize the selection process except in general

terms. Two different NHS sites participated in the trial, and 6

senior physiotherapists were responsible for patient selection

and treatment; it is believed that this increases the external

validity of the results to some degree.

Limitations of the study include the relatively small

number of subjects (in terms of examining the effects upon

pain and description of subjects profile) and the lack of

investigation of the credibility of the placebo intervention as

a form of treatment for pain relief. It is fully acknowledged

that the placebo used in this study may not have been very

credible to patients; however, it was thought that it would be

appropriate as a totally nonspecific intervention because it

was not expected to produce any physiological effects such

as increases in spinal mobility or pain relief. The main

criticism of crossover designs is that the effect of the first

intervention may have remained during the application of

the second intervention because the 2 treatments were

applied within the same session. The researchers addressed

this through the use of baseline measurements for each

intervention and by conducting the recommended statistical

tests for carryover effect.

CONCLUSION

This study investigated the efficacy of spinal flexion

MWM techniques in terms of ROM and pain scores in

subjects with LBP selected for treatment with these

techniques. There were statistically significant differences

in spinal ROM but not in pain. Differences in ROM however

were quite small, and they may not be clinically important.

The study lends some support to the anecdotal evidence that

MWM techniques may produce immediate change in spinal

movement for some subjects. It also provides preliminarymore than 3 months duration of symptoms. More than half

were experiencing their first episode of LBP.

A point that merits consideration is the physiotherapists

who participated in this clinical trial, as they were responsible

for subject selection and treatment. All 6 therapists had

experience in LBP management, and all but 1 had attended a

formal course on the MWM concept and were using these

techniques in their normal practices. In addition, educational

sessions were held before data collection regarding the use

and application of MWMs. Subject selection in the trial wasinformation on the clinical characteristics of the population

with LBP thought to benefit from these techniques.up study of low back pain in the community. Spine 2000;25:2085-90.

3. COST B13 Working Group on Guidelines for the Managementof Acute Low Back Pain in Primary Care (2004) 8.

4. Croft P. Low back pain in the community and in hospitals. Areport to the Clinical Standards Advisory Group of theDepartment of Health. Prepared by the Arthritis and Rheuma-tism Council, Epidemiology Research Unit, Manchester; 1994.

5. Faas A, van Eijk JTM, Chavannes AW, Gubbels JW.A randomised trial of exercise therapy in patients with acutelow back pain: efficacy on sickness absence. Spine 1995;20:ACKNOWLEDGMENT

This article was developed as part of a PhD program at

Coventry University. The authors would like to thank the

physiotherapy managers, the participating physiotherapists,

and the subjects who took part. The authors would also like

to acknowledge the State Scholarships Foundation (IKY),

Republic of Greece, and the Chartered Society of Physi-

otherapy Charitable Trust Fund for financial assistance.

REFERENCES1. Andersson GBJ. Epidemiology of low back pain. Acta Orthop

Practical Applications

! Flexion MWMs seemed to be more effective inincreasing active lumbar spine flexion when

compared with a placebo intervention in subjects

with LBP.

! The ROM changes however were small, and it isunclear whether they have any clinical signifi-

cance.

! Mobilizations with movement techniques mayresult in immediate improvements in ROM and

pain levels for some subjects with LBP.using a case study design. Man Ther 1998;3:78-84.13. Collins N, Teys P, Vicenzino B. The initial effects of a

Mulligans mobilisation with movement technique on dorsi-

31. Huskisson EC. Visual analogue scales. In: Melzack R,

Konstantinou et alJournal of Manipulative and Physiological Therapeutics

Spinal Flexion MWMs in LBPVolume 30, Number 31859:77-82.14. Vicenzino B, Wright A. Effects of a novel manipulative

physiotherapy technique on tennis elbow: a single case study.Man Ther 1995;1:30-5.

15. Vicenzino B, Paungmali A, Buratowski S, Wright A. Specificmanipulative therapy treatment for chronic lateral epicondy-lalgia produces uniquely characteristic hypoalgesia. Man Ther2001;6:205-12.

16. Abbott JH, Patla CE, Jensen RH. The initial effects of an elbowmobilisation with movement technique on grip strength insubjects with lateral epicondylalgia. Man Ther 2001;6:163-9.

17. Bisset L, Beller E, Jull G, Brooks P, Darnell R, Vicenzino B.Mobilisation with movement and exercise, corticosteroidinjection, or wait and see for tennis elbow: randomised trial.BMJ 2006;333:939.

18. Konstantinou K, Foster NE, Rushton A, Baxter GD. Theuse and reported effects of mobilisation with movementtechniques in low back pain management: a cross-sectionaldescriptive survey of physiotherapists in Britain. Man Ther2002;7:206-14.

19. Howell DC. Statistical methods for psychology. 4th ed. USA7Duxbury Press; 1997. p. 450-508 [ch 14].

20. Petersen CM, Johnson RD, Schuit D, Hayes KW. Intraobserverand interobserver reliability of asymptomatic subjects thor-acolumbar range of motion using the OSI CA 6000 spinemotion analyzer. J Orthop Sports Phys Ther 1994;20:207-12.

21. Keeley J, Mayer T, Cox R, Gatchel R, Smith J, Mooney V.Quantification of lumbar function. Part 5: reliability of range ofmotion measures in the sagittal plane and an in vivo torsorotation measurement technique. Spine 1986;11:31-5.

22. Gomez T, Beach G, Cooke C, Hrudey W, Goyert P. Normativedatabase for trunk range of motion, strength, velocity, andendurance with the Isostation B-200 Lumbar Dynamometer.Spine 1991;16:15-21.

23. Russell P, Pearcy MJ, Unsworth A. Measurement of the rangeand coupled movements observed in the lumbar spine. Br JRheumatol 1993;32:490-7.

24. Alaranta H, Hurri H, Heliovaara M, Soukka A, Harju R.Flexibility of the spine: normative values of goniometric andtape measurements. Scand J Rehabil Med 1994;26:147-54.

25. Dopf CA, Mandel SS, Geiger DF, Mayer PJ. Analysis of spinemotion variability using a computerised goniometer comparedto physical examination. A prospective clinical study. Spine1994;19:586-95.

26. McGregor AH, McCarthy ID, Hughes SP. Motion character-istics of the lumbar spine in the normal population. Spine1995;20:2421-8.

27. Vachalathiti R, Crosbie J, Smith R. Effects of age, gender andspeed on 3-dimensional lumbar spine kinematics. Aust JPhysiother 1995;41:245-53.

28. Esola MA, McClure PW, Fitzgerald GK, Siegler S. Analysis oflumbar spine and hip motion during forward bending insubjects with and without a history of low back pain. Spine1996;21:71-8.

29. Hasten DL, Lea RD, Johnston FA. Lumbar range of motion inmale heavy laborers on the applied rehabilitation concepts(ARCON) system. Spine 1996;21:2230-4.

30. McGregor AH, McCarthy ID, Dore CJ, Hughes SP. Quanti-tative assessment of the motion of the lumbar spine in the lowback pain population and the effect of different spinalpathologies on this motion. Eur Spine J 1997;6:308-15.editor. Pain measurement and assessment. NY7 Raven Press;1983. p. 33-7.

32. Melzack R, Katz J. Pain measurement in persons in pain. In:Wall PD, Melzack R, editors. The textbook of pain. 4th ed.Hong Kong7 Harcourt Publishers Limited; 1999. p. 409-26.

33. Scrimshaw SV, Maher C. Responsiveness of visual analogueand McGill pain scale measures. J Manipulative Physiol Ther2001;24:501-4.

34. Stratford PW, Binkley J, Solomon P, Finch E, Gill C, MorelandJ. Defining the minimum level of detectable change for theRoland-Morris questionnaire. Phys Ther 1996;76:359-65.

35. Main CJ, Wood PLR, Hollis S, Spanswick CC, Waddell G. Thedistress and risk assessment method. A simple patientclassification to identify distress and evaluate the risk of pooroutcome. Spine 1992;17:42-52.

36. Maitland GD. Vertebral manipulation. 5th ed. UK7 ButterworthHeinemann; 1986.

37. Wall PD. The placebo and the placebo response. In: Wall PD,Melzack R, editors. The textbook of pain. 4th ed. Hong Kong7Churchill Livingstone; 1999. p. 1419-30 [ch 62].

38. Mengshoel AM. Physiotherapy and the placebo effect. PhysTher Rev 2000;5:161-5.

39. Everitt BS. Statistical methods for medical investigations. 2nded. NY7 Edward Arnold; 1994. p. 77-104 [ch 6].

40. Kenward MG, Jones B. The analysis of data from 2 2cross-over trials with baseline measurements. Stat Med 1987;6:911-26.

41. Prentice RL. On the role of analysis of secondary outcomes inclinical trials. Control Clin Trials 1997;18:561-7.

42. Davis CE. Secondary end points can be validly analysed, evenif the primary end point does not provide clear statisticalsignificance. Control Clin Trials 1997;18:557-60.

43. Farrar JT. What is clinically meaningful: outcome measures inpain clinical trials. Clin J Pain 2000;16:106-12.

44. Battie MC, Cherkin DC, Dunn R, Ciol MA, Wheeler KJ.Managing low back pain: attitudes and treatment preferencesof physical therapists. Phys Ther 1994;74:219-26.

45. Jette AM, Smith K, Haley SM, Davis KD. Physical therapyepisodes of care for patients with low back pain. Phys Ther1994;74:101-15.

46. Rainville J, Sobel JB, Hartigan C. Comparison of totallumbosacral flexion and true lumbar flexion measured by adual inclinometer technique. Spine 1994;19:2698-701.

47. Hawk C, Azad A, Phongphua C, Long CR. Preliminary studyof the effects of a placebo chiropractic treatment with shamadjustments. J Manipulative Physiol Ther 1999;22:436-43.

48. Pichar JG, Wheeler JD. Response of muscle proprioceptors tospinal manipulative-like loads in the anesthetised cat.J Manipulative Physiol Ther 2001;24:2-11.

49. Indahl A, Kaigle AM, Reikeras O, Holm SH. Interactionbetween the porcine lumbar intervertebral disc, zygapophysialjoints, and paraspinal muscles. Spine 1997;22:2834-40.

50. Goodsell M, Lee M, Latimer J. Short-term effects of lumbarposteroanterior mobilisation in individuals with low-back pain.J Manipulative Physiol Ther 2000;23:332-42.

51. Turner PA, Whitfield TWA. Physiotherapists use of evidencebased practice: a cross-national survey. Physiother Res Int1997;2:17-29.

52. Turner PA, Whitfield TWA. Physiotherapists reasons forselection of treatment techniques: a cross-national survey.Physiother Theory Pract 1999;15:235-46.flexion and pain in subacute ankle sprains. Man Ther 2004;

Flexion Mobilizations With Movement Techniques: the Immediate Effects on Range of Movement and Pain in Subjects With Low Back PainMethodsDesignInclusion and Exclusion CriteriaSample SizeOutcome Measures and EquipmentDescriptive Baseline and Treatment InformationProcedureInterventionsData Analysis

ResultsDescription of Subjects' ProfilesNonrespondents to Treatment

DiscussionConclusionAcknowledgmentReferences