ARTICLE IN PRESS

1356-689X/$ -

doi:10.1016/j.m

�CorrespondE-mail add

Manual Therapy 10 (2005) 198–206

www.elsevier.com/locate/math

Original article

A normative database of lumbar spine ranges of motion

Michael Trokea,�, Ann P Moored, Frederick J Maillardetb, Elizabeth Cheekc

aHead of Research & Enterprise, York St John College, Lord Mayor’s Walk, York, YO31 7EX, UKbClinical Research Centre for Health Professions, University of Brighton, UK

cUniversity of Brighton, UKdFaculty of Management & Information Sciences, University of Brighton, UK

Received 4 June 2003; received in revised form 17 September 2004; accepted 25 October 2004

Abstract

The overall aim of the work was to develop a comprehensive normative database of indices for ranges of motion in the lumbar

spine, in an asymptomatic sample of the general population. This was a repeated measures prospective study utilizing a reliable and

valid instrument, the modified CA6000 Spine Motion Analyzer (Orthopedic Systems Inc. Union City CA & Troke/University of

Brighton). The portable equipment was used to collect data in a variety of community settings (e.g. schools, GP surgeries, offices,

leisure centres, emergency services stations).

A total of 405 asymptomatic subjects (196 female, 209 male) aged 16–90 yr from sedentary, mixed and physically demanding

occupations participated in the study and data were collected in standing, at different times of the day, following a standardized

methodology for lumbar spine motion in the sagittal, coronal and horizontal planes.

Age-related centile graphs were derived separately for male and female subjects in flexion, extension, left and right lateral flexion

and left and right axial rotation. All 12 graphs are presented as an appendix located on the Manual Therapy website

(www.elsevierscience.com/journals/math). Overall, flexion (73–401) and lateral flexion (28–141, L&R) declined 45% and 48%,

respectively, across the age range. Extension (29–61) declined the greatest at 79%. By contrast, no overall decline in axial rotational

RoMs was recorded, and the median RoM remained at 71 each way across the age spectrum examined.

A comprehensive database of indices of lumbar spine ranges of motion has thus been developed which is gender specific, age

related, drawn from a wide age range and presents data for all three planes of motion. It is considered that the new database has a

number of potential clinical and research applications.

r 2004 Elsevier Ltd. All rights reserved.

Keywords: Normative; Database; Lumbar; Spine

1. Introduction

An introductory paper has been published in Clinical

Rehabilitation (Troke et al., 2001b), in which themethodology developed for this work, along with anew approach to the analysis of lumbar spine normativedata were discussed. Summary results were alsopresented along with an example of the database graphs.The principal purpose of this current paper is to publishthe new, comprehensive, normative database of lumbar

see front matter r 2004 Elsevier Ltd. All rights reserved.

ath.2004.10.004

ing author.

ress: m.troke@yorksj.ac.uk (M. Troke).

spinal motion in full, and make the complete databaseavailable to interested clinicians and researchers. Theclinical relevance of this work is considered in Section 5.Many papers and publications offering normative

data for lumbar spine movement have used differenttypes of equipment, a variety of methodologies anddifferent methods of analysis. Some only offer dataacross a limited spectrum of age ranges, do notencompass all three planes of movement, or are notgender specific. This present work is an attempt toaddress these limitations, following the development ofthe modified CA6000 Spine Motion Analyzer (Trokeand Moore, 1995).

ARTIC

LEIN

PRES

S

Table 1

Selected comparative ranges of lumbar spinal motion including this current normative database

Year 1974 1978 1979 1985 1993 1994 1994 1995 1995 2000 2001 2001

Author(s) Kapandji White and

Panjabi

Twomey Pearcy Russell et al. Dopf et al. Greene and

Heckman

Dvorak

et al.

McGregor

et al.

Van Herp

et al.

Ng et al. Troke et al.

Reliability

reported

No No No Yes No Yes No Yes Yes No Yes Yes

Validity

reported

No No No Yes No No No No No Yes No Yes

Total subjects

No.

Not stated Composite 144 31 245 120 (a) 104 203 100 35 405

(Female No.) Not stated Not stated 72 0 118 60 (a) 42 100 50 0 196

(Male No.) Not stated Not stated 72 31 127 60 (a) 62 103 50 35 209

Ages (Yr) Not stated Not stated 20–60+ 21–37 20–69 20–35 (b) 20–70 20–70 20–77 Mean 29 16–90

Spinal

Region-

Reported

T12-sacrum T12-sacrum T12-sacrum L1-sacrum L1-sacrum T12-sacrum T12/L1-sacrum T12-sacrum T12-sacrum T12-sacrum T12-sacrum T12-sacrum

Gender

specific

No No Yes Yes Yes No Yes Yes Yes Yes Yes Yes

Age related No No Yes No Yes No No Yes Yes Yes No Yes

Measurement

Methods

Not stated Not stated Mechanical

with cadavers

Stereo-

radiographic

Isotrak CA6000 with

straps+others

Radiographic CA6000

with straps

CA6000

with straps

Isotrak Inclinometer+

rotameter

CA6000

with Pads

Results Note: Where a range of values is reported, they are shown here in descending order according to ascending age. All values are shown in degrees

Flexion 60 115–70

Combined

39–24 51 75–58 81 76–70

Combined

75–55 64–45 59–51 52 72–40

Extension 35 14–9 16 28–15 35 31–17 30–13 37–15 19 29–6

R Lat Flex 20 47–23

Combined

20–14 17 57–35

Combined

45 50–40

Combined

36–23 35–25 26–15 31 28–15

L Lat Flex 20 20–13 18 46 35–20 36–25 26–15 30 29–16

R Ax Rotn 5 18–17

Combined

16–12 4 36–26

Combined

42 12 48–32 30–23 19–13 32 7

L Ax Rotn 5 16–12 5 43 Combined 48–33 31–23 19–11 33 7

Key: T12-sacrum: Whole lumbar spine region; L1-sacrum: Excluded T/Lumbar junction.

(a) Given in cited studies—Dvorak: 41(23 Male, 18 Female), Pearcy: 11 (all Male).

(b) Given in cited studies—Dvorak: 22–50 yr, Pearcy: 25–36 yr.

M.

Tro

ke

eta

l./

Ma

nu

al

Th

erap

y1

0(

20

05

)1

98

–2

06

199

ARTICLE IN PRESSM. Troke et al. / Manual Therapy 10 (2005) 198–206200

Increasingly more sophisticated methods have beendeveloped including the use of stereo X-ray equipment(Pearcy, 1985) and different electronic methods (Hindleet al., 1990; Gomez et al., 1991; Russell et al., 1993;Dopf et al., 1994; Dvorak et al., 1995; McGregor et al.,1995; Van Herp et al., 2000). Of these, the Dopf, Dvorakand McGregor studies utilized the unmodified CA6000Spine Motion Analyzer with strap fixation. A modifiedversion of that instrument utilising a new skin padfixation system was used for this current work. Selectedpublications which offered normative data of lumbarspine motion are summarized in Table 1 and reviewedbelow.Kapandji (1974) provided maximal figures for ranges

of motion which bear comparison with this current work.It is not clear what methodologies were drawn upon toestablish the maximal RoMs quoted, although work byTanz (1953) is cited, but not specifically referenced. Thedata were not gender specific but age-related segmentalranges by Tanz were reproduced, and these have alsobeen cited by White and Panjabi (1990).Normative data reported by White and Panjabi

(1978a) were originally published in a journal paperand also in the first edition of their book (White andPanjabi, 1978b). Their figures for individual segments ofthe lumbar spine were drawn from comprehensivereferences and the ‘‘authors’ own best opinion’’, basedupon a review of the literature at the time, as well as theirown analysis. Neither the number of subjects from whichthe data were drawn, nor their gender and ages werereported and nor the methods by which the representativevalues offered by these authors were derived.In 1979, Twomey published a paper on the effects of

age on the RoMs of the lumbar spine. Results werereported on 200 cadaveric specimens, secured in amechanical test vice, of which 144 were in the adult ageranges from 20 years to over 60 years, with an equalnumber of male and female subjects. Measurementswere taken from T12 to S2, as in this current work. Theranges of motion were categorized according to age andgender, flexion, extension, left and right lateral flexionand left and right axial rotation.The work of Pearcy (1985), described extensive

studies utilizing stereo radiography. The author re-ported on work with 31 male subjects who were exposedto stereo-radiographic X-rays in a specially constructedframework. Eleven subjects were measured for flexion/extension and had a mean age of 29 yrs (SDs not given).A further 10 subjects carried out lateral flexion move-ments (mean age 28 yrs) and another 10 subjects carriedout axial rotation movements (mean age 24 yrs).Measures (in degrees) were reported for each segmentfrom L1 to the sacrum, but excluded the thoraco-lumbarjunction (T12/L1).Russell et al. (1993) utilized the Isotrak electro-

magnetic equipment (Polhemus Navigation, Inc.) to

investigate normal ranges of motion for groups of malesand females aged from 20 to 69 years. The authors choseto combine left and right lateral flexion and axialrotation figures to give total ranges and like Pearcy(1985) this work reports motion of the lumbar spinetaken from L1.In 1994, the American Academy of Orthopedic

Surgeons (AAOS) published a revised and updatededition of their earlier book (AAOS, 1965) on jointmotion (Greene and Heckman, 1994). The latter bookdrew on two papers (Pearcy et al., 1984; Dvorak et al.,1991) to tabulate radiographic segmental RoMs for thelumbar spine from L1 to S1. Additionally, whendiscussing clinical measurements of the lumbar spine,the authors compared earlier results using the Schober(1937) method with the later work which utilizedinclinometers and goniometers (Loebl, 1967; Fitzgeraldet al., 1983).A study by Van Herp et al. (2000) utilized the 3-Space

Isotrak electro-magnetic system (as did Russell et al.,1993) to produce a normative database involving 50male and 50 female subjects aged from 20 to 77 years.The data reported by Van Herp and colleagues involvedthe whole of the lumbar region (as in this current work),the instrument having been applied from the T12vertebra to the sacrum. The data were arranged in 10-year age bands, were gender specific and demonstratedchanges in RoMs with advancing age.In their reliability and normative study, Ng et al.

(2001) utilized double inclinometers and rotametertechniques. Like Pearcy (1985), Ng et al. involved asmall sample (31 and 35, respectively) of young malesubjects.

1.1. Normative work with the CA6000 Spine Motion

Analyzer

The first study to be published which offerednormative values for lumbar spinal motion utilisingthe unmodified CA6000 was published by Dopf et al.(1994). In 1995, Dvorak et al. reported on data from asubstantial number of subjects across a wide age range(20–70 yr), and the data were age related and genderspecific.McGregor et al. (1995) discussed normative data for

RoMs as well as for velocity of movement. Like Dvoraket al.’s (1995) work, McGregor and colleagues’ datawere gender specific and age related in 10-year agebands.The diversity of methods used to establish lumbar

spine normative RoMs and summarized in Table 1,illustrate the significant lack of agreement and very widevariation in reported RoMs for substantially the sameregion of the spine. The values reported demonstrate theneed for a comprehensive normative database, and the

ARTICLE IN PRESSM. Troke et al. / Manual Therapy 10 (2005) 198–206 201

work highlighted in Table 1 is considered further in thediscussion section of this paper.Developmental work on the CA6000 spinal motion

analysis system has been described in a series ofconference presentations and journal papers. In thelight of difficulties encountered with the manufacturers’strap fixation, and the exaggerated results obtained withthe un-modified instrument, the programme of workbegan with the development of a new skin fixationsystem (Troke and Moore, 1995). The system comprisessemi-flexible self-adhesive pads with balance weightsand squared alloy mounting hooks. The linkage consistsof six potentiometers connected by light alloy tubularrods. The instrument is thereby secured directly onto theskin, typically over the spinous processes of T12 and S2vertebrae. This work was followed by a series ofreliability studies (Troke et al., 1996, 1998) and a studyof validity (Troke et al., 2001a), all directed towards thedevelopment of a credible clinical tool.

2. Aim and objectives for the normative database study

The aim of the study was to develop a comprehensivenormative database of indices for ranges of lumbarspinal motion in an asymptomatic sample of the generalpopulation, utilizing a reliable and valid instrument.The principal objectives were to generate these data

utilizing the modified CA6000 Spine Motion Analyzer(Orthopedic Systems Inc. Union City CA and Troke /University of Brighton), and to analyse and presentgender-specific and age-related results for all planes ofmovement which clinicians and researchers interested inspinal dysfunction could readily utilize.

3. Method

Ethical approval for the work was granted by theUniversity of Brighton Research Ethics Committee.Data collection was carried out in a variety ofcommunity locations (e.g. schools, GP surgeries, offices,leisure centres, emergency services stations, etc.). Aportable system linked to a laptop computer was utilizedwith the CA6000 instrument and the skin pad fixationsystem as described in previous papers (Troke andMoore, 1995; Troke et al., 1996, 1998).Over 400 participants were recruited for the study

with ages ranging from 16 to 90 years, 209 of whomwere male, and 196 of whom were female. In excess of 20companies, institutions or groups were involved, from avariety of sedentary, mixed and physically demandingoccupations. Data collection was carried out accordingto the standardized protocol developed during thereliability studies (Troke et al., 1996). The instrumentwas secured over T12 and S2 vertebrae, and preliminary

movements intended to overcome any initial stiffnesswere carried out. Four successive lumbar spine move-ments in each plane (sagittal, coronal and horizontal)were then recorded from a standing start position. Inorder to encompass diurnal variations, two completedata sets of lumbar spine movement data were recordedat different times of the day. Over the course of twoseparate days, a total of 24 observations were taken witheach of the 405 subjects. The methodology used for thisstudy has been described in detail in an introductorypaper published in the journal Clinical Rehabilitation

(Troke et al., 2001b).The statistical analysis of the data produced age-

related centiles as a continuum across all the subjectages, in each gender, and for each plane of movement.This is in accordance with the epidemiological andstatistical advice received, plus reference to Altman(1993), and is thought to be an innovative approach tothe presentation of spinal motion data.Prior to the construction of the centile graphs,

analyses of variance (2-way) were conducted, leadingto the calculation of Type 2,1 Intra-class CorrelationCoefficients (Shrout and Fleiss, 1979) in order to test thereliability of the data. Initial trends emerged, some ofwhich appeared to confirm established knowledge aboutthe effect of age on lumbar spine ranges of motion, andsome of which appeared to shed new light on this area.Using regression analysis techniques, age-related centileswere constructed from the data collected. The graphsare gender specific, illustrate the age-related variationsin lumbar ranges of motion as a continuum across theage spectrum, and relate to each plane of movement(sagittal flexion–extension, coronal lateral flexion to theright and to the left, and horizontal axial rotation, withthe shoulders rotating clockwise towards the right andanti-clockwise to the left).

4. Results

As examples from the complete website database(www.elsevierscience.com/journals/math), graphs forfemale flexion, extension, lateral flexion and axialrotation are illustrated in Figs. 1–4. The graphs showthe centiles as a continuum for all subjects, according togender, with individual data points shown as a scatterplot. The 50th centile (median), the 97th, 90th, 10th and3rd centiles are superimposed.The 50th centile represents the values such that half of

the subjects will have RoM values greater, and half willhave RoM values less than the median. The valuesfound between the 90th and 10th centiles representRoMs which it might be expected that 80% of thepopulation would achieve. The values beyond thesecentiles are either in the upper 10% or lower 10% of thegeneral population distribution. The values found below

ARTICLE IN PRESS

0102030405060708090

100

10 20 30 40 50 60 70 80 90

Deg

rees

Age (years)

Fig. 1. Flexion-females 3rd, 10th, 50th, 90th and 97th Centiles. r Copyright M. Troke.

0

10

20

30

40

50

10 20 30 40 50 60 70 80 90

Deg

rees

Age (years)

Fig. 2. Extension-females 3rd, 10th, 50th, 90th and 97th Centiles. r Copyright M. Troke.

0

10

20

30

40

50

10 20 30 40 50 60 70 80 90

Deg

rees

Age (years)

Fig. 3. Right lateral flexion-females 3rd, 10th, 50th, 90th and 97th Centiles. r Copyright M. Troke.

M. Troke et al. / Manual Therapy 10 (2005) 198–206202

the 3rd or above the 97th centiles could be regarded asat the extremes of RoMs to be expected in the generalpopulation.The median flexion RoM value for male subjects

declined from 731 to 401 across the age spectrum of16–90 yr. The median value for females was similar,from 681 to 401. In extension, the male median RoMdeclined with age from 291 to 71 and the female median

RoM similarly declined from 281 to 61. These overallranges of motion are summarized in Table 2.In lateral flexion both male and female median values

were very similar, ranging from 281 to 141, left or right.In axial rotation, the median RoM value, both for maleand female subjects, to both the left and the rightremained at approximately 71 in each direction, acrossthe whole age spectrum studied.

ARTICLE IN PRESS

0

5

10

15

20

25

30

10 20 30 40 50 60 70 80 90

Deg

rees

Age (years)

Fig. 4. Left axial rotation-females 3rd, 10th, 50th, 90th and 97th Centiles. r Copyright M. Troke.

Table 2

Maximum and minimum median ranges of lumbar spinal motion across all subjects (overall age range of subjects 16–90 yr)

Movement Male Female

Max Min Max Min

(median of values) (deg) (median of values) (deg)

Flexion 73 40 68 40

Extension 29 7 28 6

Right lateral flexion 28 15 27 14

Left lateral flexion 28 16 28 18

Right axial rotation 7 7 8 8

Left axial rotation 7 7 6 6

M. Troke et al. / Manual Therapy 10 (2005) 198–206 203

Whilst the apparent effect on RoMs of increasing ageis self-evident, it will be noted that this change (where itoccurs) is generally linear. Nevertheless, it is alsonoteworthy that extension appears to decline as a slightcurve and that all the axial rotation RoMs remainsubstantially constant across the age spectrum.On comparing Fig. 1 with the male equivalent, there

appears to be slightly greater variability in RoMs, andtherefore the centiles are more closely spaced with malesubjects. Female subjects have lower RoMs at 16 yr, butappear to equal male subjects in the final decade. Thedecline in RoMs is linear, and between ages 16 and 90 yrthe reduction is 45%.In extension, the male and female values and centile

patterns are almost identical. The slightly curvedchanges suggest a steeper decline in earlier years, whichlevels out somewhat in later life. Overall, the reductionacross the age ranges is 79%, and is the greatest declineof any of the primary movements.In lateral flexion, the decline in RoMs is linear at 48%

overall. There appears to be less variability with malesubjects than with females. Overall patterns betweenmovement to the left and right are very similar.In axial rotation, no overall decline is apparent in

RoMs across the age spectrum. The 50th centile value isvery similar for both men and women, and overallvariability is also comparable.

The presentation of the results of this study as centilegraphs has been designed to make direct reading ofranges immediately accessible, if comparison needed tobe made with other subjects or patients of a given genderand age, in a given plane of motion. It can be seen thatthe centiles offer a comprehensive reference for ranges ofmotion of the lumbar spine for each sex and in all threeplanes of motion, in either a clinical or research arena.

5. Discussion

5.1. Current work with the modified CA6000

There was considerable agreement between the overallmaximum and minimum RoMs achieved by the maleand female participants in this study. This is in contrastto many previous studies which have found differencesbetween overall mobility in spinal motion between thesexes. Results from this database suggest that there wasno discernable difference in RoMs between males andfemales. A possible exception was that the median valuefor the youngest of the female subjects was approxi-mately 51 less in flexion than the equivalent male figure.This difference is less than 10% and may therefore beconsidered by some clinicians as within the bounds ofnormal variation.

ARTICLE IN PRESSM. Troke et al. / Manual Therapy 10 (2005) 198–206204

The use of centile graphs to present the completenormative data sets is thought to be an innovative wayof presenting spinal motion data whilst making theoverall results readily accessible and easy to use. Theage-related data can be related directly to the individualresearch subject or clinical patient RoM valuesachieved, and used as part of a broader clinicalassessment of possible spinal pathology or dysfunction.Centiles were also used in order to avoid artificial age

bands and in order to present the individual’s results asa continuous correlation between range of motion andage. The centiles were calculated using regressiontechniques with the 50th centile (median value) beingthe line of best fit. The reference intervals areconsequential upon this approach.Utilization of age-related centiles also enabled the use

of reference intervals (e.g. 80% of the populationbetween the 10th and 90th centiles) to determine thelocation of an individual’s results as within the referenceinterval, or not, as the case may be. Use of the term‘reference interval’ can be as an alternative to describe‘normal’ or ‘normative’ data in the context of range ofmotion (Bland and Altman, 1995). In this study,reference interval is defined as a range of valuesobtained from a majority of normal subjects.Extensive speculation as to the possible reasons for

the substantially similar RoMs achieved by male andfemale subjects in this study, and for the age-relatedchanges, or lack of change, in the lumbar spinal RoMsshown in the new database, is perhaps more suited to afuture paper. However, given the availability of the newdatabase, further research may now be able to offergreater insight into lumbar spinal motion, and offersubstantiated propositions for these similarities, and forthe age-related changes found in this work.

5.2. Earlier work with the unmodified CA6000

Dopf et al. (1994) made comparisons with othermethods of measuring spinal motion including theSchober (1937) method for measuring flexion, andinclinometers for measurements in other planes. TheRoMs reported for horizontal axial rotation were clearlyexaggerated when compared to the Kapandji (1974)data, White and Panjabi (1978a, b), Twomey’s (1979)paper and Pearcy’s (1985) data. It is perhaps unfortu-nate that Dopf et al. (1994) did not acknowledge thelimitations of the early versions of the CA6000 whenreporting their findings.In a subsequent study, Dvorak et al. (1995) used a

much more constrictive method when measuring sub-jects in lateral flexion than this current work or anyother studies which utilised the CA6000. Unlike Dopf etal. (1994) however, Dvorak et al. acknowledged that theresults obtained for axial rotation were exaggerated andthat the strap fixation supplied by the manufacturer was

a likely source of error. Insofar as the paper provided aninteresting insight into various aspects of lumbar spinalmovement, comparison of results with the new norma-tive database reported here requires caution in the lightof Dvorak and colleagues’ own reservations.A similar criticism to that which could be applied to

the RoMs reported by Dopf et al. (1994) in axialrotation could also be applied to McGregor et al.’s(1995) results. In comparison to earlier work (e.g.Panjabi et al., 1994) the axial rotation RoMs againappeared exaggerated.

5.3. Clinical applicability of the new database

The modified CA6000 instrument is a safe, non-invasive means of establishing reliable and valid indicesof regional lumbar spine motion. The new normativedatabase offers complementary data on lumbar rangesof motion which might be expected amongst anasymptomatic population. It has been established aseasy to use in different clinical and communitylocations, economic of operator’s time and readilyportable. The new database is intended to be quickand easy to interpret by clinicians, and where appro-priate to be used in explanations to patients as a meansof offering feedback following treatment intervention.In considering the possible clinical applicability of the

instrument and the new database, it should be acknowl-edged immediately, that RoM indices and patterns ofmovement graphs can only be part of the whole clinicalpicture and the clinical relevance of such data is still amatter for debate. The graphical representations pro-vided by the instrument combined with the newnormative database could, however, be seen as usefuladjuncts in initial assessment in the context of specificspinal conditions, monitoring the progress of rehabilita-tion and the efficacy of treatment regimes.It has been suggested anecdotally that research

colleagues investigating low back pain (LBP) syndromesin general, as well as more specific conditions, havepossibly been reluctant to utilize objective measures oflumbar spinal RoMs in their studies. This has been insome measure due to the lack of confidence in the resultsproduced by some instruments purporting to quantifylumbar spine motion.Long-term monitoring of the progress (or otherwise)

of rheumatic diseases such as ankylosing spondylitis alsosuggest themselves as examples where reliable and validRoMs data on lumbar spinal motion would beadvantageous. Additionally, studies are already progres-sing on assessing risks associated with specific manualhandling tasks, applications in the field of occupationalhealth, and post-surgical outcomes—all areas of currentresearch seen as of high priority.It is expected that future use of the data will itself

promote further opportunities to evaluate treatment

ARTICLE IN PRESSM. Troke et al. / Manual Therapy 10 (2005) 198–206 205

efficacy, monitor diseases and quantify outcomes inclinical and research settings—and relate findings to theage-related data illustrated in the new normativedatabase.

5.4. Further developments with the instrument and the

database

It might be desirable to replicate this study in anumber of diverse locations nationally, with the aim ofreflecting differing living environments. Statistical (ac-tuarial) advice would, however, be advisable in thisrespect to avoid expanding the database to an unneces-sary extent without the benefit of clear advantages interms of statistical power and the generalizablity of theresults.All the subjects for this study were of caucasian

background, although people from other backgroundswere not specifically excluded from volunteering for thestudies. It might therefore be argued that to be trulyrepresentative of the overall diversity of cultural andethnic backgrounds to be found in the UK, people witha variety of ethnic backgrounds should be included in anexpanded database. This might appear to be a counsel ofperfection, which would be theoretically desirable butlogistically impractical. It might therefore be morefeasible to carry out small-scale studies involvingparticipants from specific ethnic backgrounds (e.g. theIndian sub-continent, Africa, the Far East), andcompare their results with the existing database toestablish if any significant difference exists.In considering the possible influence of coupled

motion characteristics on primary movement a detailedanalysis of the three-dimensional data collected wasbeyond the scope of this work. However, the raw datawere recorded in three dimensions as part of the normalcollection process. Given recent developments in cali-bration and computer modelling of the instrument’smeasures of coupled motion (Higgison, 2004, pers.comm.), subsequent analysis could focus on the coupledmotion characteristics displayed by this group ofsubjects. The coupled RoMs and patterns of movementcould be directly related to the individual participants,and also analysed as centile graphs. The results couldalso be presented as parallel data to the primarymovement centiles illustrated in Figs. 1–12 in the websiteappendix (www.elsevierscience.com/journals/math).

6. Conclusions

As has been previously noted, there has beenconsiderable diversity in approach to the collection ofnormative data for motion of the lumbar spine.Similarly, there has been considerable variation in the

analysis or presentation of results, making the compar-ison of different studies very difficult.The innovative, clear, consistent and uniform pre-

sentation of the normative values for spinal motion datain all three planes of movement has been achieved in thiswork. It is anticipated that the results may be ofparticular value to clinicians who are involved inresearch with, or the management of, patients withlow back pain syndromes. The database offers compre-hensive indices of spinal RoMs, and may also be ofinterest to clinicians utilizing other forms of spinalmeasurement equipment.Additionally, the database may be of interest to

clinicians and researchers working in occupationalhealth, primary, secondary or intermediate care settings.In due course the database may be used in themonitoring of rehabilitation programmes, the evalua-tion of treatment regimes and as an outcome measurefor interventions.

Acknowledgement

The authors gratefully acknowledge the support ofSouth Thames Regional Health Authority for this work.

References

AAOS. Joint motion, method of measuring and recording. Chicago:

American Academy of Orthopedic Surgeons; 1965.

Altman DG. Construction of age-related reference centiles using

absolute residuals. Statistics in Medicine 1993;12:917–24.

Bland JM, Altman DG. Comparing two methods of clinical

measurement: a personal history. International Journal of Epide-

miology 1995;24(Suppl. 1):S7–S14.

Dopf CA, Schlomo SM, Geiger DF, Mayer PJ. Analysis of spine

motion variability using a computerised goniometer compared to

physical examination. Spine 1994;19(5):586–95.

Dvorak J, Panjabi MM, Chang DG, Theiler R, Grob D. Functional

radiographic diagnosis of the lumbar spine. Spine

1991;16(5):562–71.

Dvorak J, Vajda EG, Grob D, Panjabi MM. Normal motion of the

lumbar spine as related to age and gender. European Spine Journal

1995;4:18–23.

Fitzgerald GK, Wynveen KJ, Rhealt W, Rothschild B. Objective

measurement with establishment of normal values for lumbar

spinal range of motion. Physical Therapy 1983;63(11):1776–81.

Gomez T, Beach G, Cooke C, Rudley WH, Goyert P. Normative

database for trunk range of motion, strength, velocity and

endurance with the Isostation B200 lumbar dynamometer. Spine

1991;16(1):15–21.

Greene WB, Heckman JD. The clinical measurement of joint motion.

Chicago: American Academy of Orthopedic Surgeons; 1994.

Higgison D. Oxford Brookes University; 2004, personal communica-

tion.

Hindle RJ, Pearcy MJ, Cross AT, Miller DHT. Three dimensional

kinematics of the human back. Clinical Biomechanics

1990;5(4):218–28.

Kapandji IA. The Physiology of the Joints. 2nd ed. vol. 3. Trunk &

Vertebral Column. Edinburgh: Churchill Livingstone; 1974.

ARTICLE IN PRESSM. Troke et al. / Manual Therapy 10 (2005) 198–206206

Loebl WY. Measurement of spinal posture and range of spinal

movement. Annals of Physical Medicine 1967;IX(3):103–10.

McGregor AH, McCarthy ID, Hughes SP. Motion characteristics of

the lumbar spine in the normal population. Spine 1995;

20(22):2421–8.

Ng JK-F, Kippers V, Richardson CA, Parnianpour M. Range of

motion and lordosis of the lumbar spine. Spine 2001;26(1):

53–60.

Panjabi MM, Oxland TR, Yamamoto I, Crisco JJ. Mechanical

behaviour of the human lumbar and lumbosacral spine as shown

by three-dimensional load-displacement curves. Journal of Bone

and Joint Surgery 1994;76A(3):413–24.

Pearcy MJ, Portek I, Shepherd J. Three-dimensional X-ray

analysis of normal movement in the lumbar spine. Spine 1984;9(3):

294–7.

Pearcy MJ. Stereo radiography of lumbar spine motion. Acta

Orthopaedica 1985;56(212 (Suppl)):1–41.

Russell P, Pearcy MJ, Unsworth A. Measurement of the range and

coupled movements observed in the lumbar spine. British Journal

of Rheumatology 1993;32(6):490–7.

Schober P. The lumbar vertebral column and backache. Munich

Medical Wschr 1937;84:336.

Shrout PE, Fleiss JL. Intra-class correlations: Uses in assessing rater

reliability. Psychological Bulletin 1979;86:420–8.

Tanz SS. Motion of the lumbar spine: A roentgenologic study.

American Journal of Roentgenology 1953;69(3):399–412.

Troke M, Moore AP. The development of a new form of instrument

fixation for the OSI CA6000 spine motion analyzer. Manual

Therapy 1995;1(1):43–6.

Troke M, Moore AP, Cheek E. Intra-operator and inter-operator

reliability of the OSI CA6000 Spine Motion Analyzer with a new

skin fixation system. Manual Therapy 1996;1(2):92–8.

Troke M, Moore AP, Cheek E. Reliability of the OSI CA6000 spine

motion analyzer with a new skin fixation system when used on the

thoracic spine. Manual Therapy 1998;3(1):27–33.

Troke M, Moore AP, Maillardet FJ, Cheek E, Wells NS, Sidhu PS.

X-ray validation of measures recorded by the modified CA6000

spine motion analysis system. Physiotherapy 2001a;87(2):90–1.

Troke M, Moore AP, Maillardet FJ, Hough A, Cheek E. A new,

comprehensive normative database of lumbar spine ranges of

motion. Clinical Rehabilitation 2001b;15(4):371–9.

Twomey L. The Effects of age on the ranges of motions of the lumbar

region. Australian Journal of Physiotherapy 1979;25(6):257–63.

Van Herp G, Rowe PJ, Salter P, Paul JP. Three dimensional

kinematics: a study of 100 healthy subjects aged 20–60+yrs.

Rheumatology 2000;39:1337–40.

White A, Panjabi MM. Basic kinematics of the human spine. Spine

1978a;3(1):12–20.

White A, Panjabi MM. Clinical Biomechanics of the Spine.

Philadelphia: Williams & Wilkins; 1978b.

White A, Panjabi MM. Clinical biomechanics of the spine, second ed.

Philadelphia, Lippincott: Williams & Wilkins; 1990.