SPINE Volume 27, Number 8, pp E215–E220

©2002, Lippincott Williams & Wilkins, Inc.

Development and Validation of a New Technique for

Assessing Lumbar Spine Motion

Sai-wing Lee, PhD,* Kris W. N. Wong, PDPT,† Man-kwong Chan, MBBS, FRCR,‡
Hon-ming Yeung, BSc,‡ Jeffrey L. F. Chiu, MBBS, FRCR,‡ and
John C. Y. Leong, FRCS, FRCSE, FRACS§

Furthermore, an optoelectric motion analysis system has

Study Design. Dynamic lumbar flexion– extension mo- been used to assess the three-dimensional motion of the
tions were assessed by an electrogoniometer and a lumbar spine.5,8 Despite the good accuracy of these as-
videofluoroscopy unit simultaneously.
sessment methods, the clinical value of the measurement
Objective. To develop and validate a new technique
for the assessment of lumbar spine motion. is limited because only the lumbar spine’s range of mo-
Summary of Background Data. Spine instability, a tion is measured. It is hardly possible to identify the spe-
clinical condition that is common but difficult to diagnose, cific location of the intervertebral dysfunction, although
has been suggested to involve a characteristic change in this is particularly important for surgical intervention for
the relation between vertebrae during motion. Assess-
instability.
ment of lumbar instability using functional radiographs is
controversial. Information regarding dynamic spine kine- Conversely, functional radiographs are used com-
matics in vivo is limited. monly to assess the segmental disorders of the lumbar
Methods. A lumbar spine motion analysis system was spine in clinical practice.2,10,18 Reportedly, high preci-
developed, and its reliability was assessed. Simultaneous sion in the assessment of intervertebral motion is ob-
total flexion range of motion and segmental motion of the
tained successfully with roentgen stereophotogrammet-
lumbar spine were assessed in 30 healthy volunteers.
Lumbar images were captured in 10° intervals during ric analysis or its modified method.12 However, these
flexion– extension. Intervertebral flexion– extension of methods are difficult to use in routine clinical practice.
each vertebral level was calculated. On the other hand, the reliability of assessing interverte-
Results. In flexion, the lumbar vertebrae flexed with a bral displacement with plain radiographs is questionable
descending order from L1 to L5 throughout the motion.
because the radiographs are taken frequently for certain
Conversely, the concavity of lumbar lordosis increased
steadily in extension. No statistically significant difference postures without standardization. Mayer et al15 showed
in the pattern of motion was found between genders. that the interobserver variance in active range of motion
Conclusions. The results from this study showed that assessments of the lumbar spine is significantly large.
the newly developed technique is reliable. It may have Ensink et al4 also has reported that the total lumbar
potential value for evaluating spine instability in clinical
range of motion measured by an inclinometer varies sig-
practice. [Key words: assessment, genders, lumbar spine,
motion] Spine 2002;27:E215–E220 nificantly even within a day! Errors in the assessment of
the intervertebral displacement at the end range of either
flexion or extension is therefore subject to large varia-
Analysis of lumbar spine motion can improve under- tion. As a result, comparison of assessments using func-
standing of instability and related surgical interventions. tional radiography is unreliable.
For example, using a three-dimensional motion analysis To overcome these drawbacks, the authors proposed
system, Lu et al13 validated the surgical effect of multi- assessment of the lumbar spine’s coherent motion, in-
level fenestration and discectomy in vitro. Glossop and cluding both the gross total lumbar spine motion and the
Hu7 investigated intraoperative spine motion during intervertebral levels motion, using an electrogoniometer
pedicle screw insertion. and a videofluoroscopy (VF) unit, respectively. The dy-
Inclinometers have been proposed as accurate and namic motion of the lumbar spine could be then assessed
practical for clinical assessment of lumbar spine mo- and compared at a number of standardized positions.
tion.16,20 With the advancement of technology, Dvorák Therefore, the objectives of this study were to develop
et al3 and McGregor et al17 have used an electrogoniom- and validate a new assessment method for lumbar spine
eter to assess the continuous motion of the lumbar spine. motion analysis, and to assess the flexion– extension of
the lumbar spine in a group of healthy volunteers.
From the *Department of Rehabilitation Sciences, The Hong Kong
Polytechnic University, Hunghom, Kowloon, Hong Kong, the Depart- Methods
ments of †Physiotherapy and ‡Radiology and Imaging, Queen Eliza-
beth Hospital, Kowloon, Hong Kong, and the §Department of Ortho- Development of a New Lumbar Spine Motion Analysis
paedic Surgery, Queen Mary Hospital, The University of Hong Kong, System. A lumbar spine motion analysis system has been de-
Pokfulum, Hong Kong. veloped using a spinal electrogoniometer, a VF unit, and a
Acknowledgment date: July 23, 2001. tailor-made image digitizing system (Figure 1). The spinal elec-
Acceptance date: November 6, 2001.
The manuscript submitted does not contain information about medical
trogoniometer is an in-house– developed device comprising a
devices. series of metal link bars and three high-precision potentiome-
No funds were received in support of this study. ters installed at the movable joints of the link bars (Figure 2). Its

E215
E216 Spine • Volume 27 • Number 8 • 2002

Figure 1. Schematic diagram of the lumbar spinal motion analysis

system.

function is to assess the continuous anatomic motion of the

lumbar spine in the sagittal plane. The analog signals of the
potentiometers then are digitized and stored in a portable com-
puter by means of DataQ, an analog-to-digital conversion sys-
tem, and Windaq, Version 1.71, a data acquisition software
program, at a sampling rate of 100 Hz.
The VF used in this study is a standard clinical unit manu-
factured by the General Electronic Ltd. The radiographic ex-
posure factors of the unit, such as kilovoltage and milliampere
per second, are automatically regulated. During screening, the
images are displayed with a high-resolution monitor (1260
lines per second) and can be recorded with a super VHS video
recorder. To assess the coherent motion of the lumbar spine, at
both the intervertebral and anatomic levels, it is necessary to Figure 3. The newly designed chest and pelvic harnesses.
synchronize the video output signals of the VF and the spinal
electrogoniometer data. Using a cable, the computer is con-
ion– extension using a DC30! monochrome frame-grabber
nected to an input socket of the video recorder. When the com-
(Pinnacle Systems, CA) and Adobe Premiere 4.2 computer soft-
puter is triggered to collect the spinal electrogoniometer data,
ware. To correct the pincushion distortion of the VF, a special
the computer immediately generates a signal and transmits it to
design calibration board has been developed for image analy-
the video recorder, producing a blank screen in the recording.
sis. This board contains 66 columns times 66 rows of steel
When the videotape is rewound and played back, the blank
metal grid, with each column measuring 1 mm in length. Before
screen is traced. This marks the commencement of the data
each measurement, the board is hung midway between the
collection in the electrogoniometer. As a result, the frame num-
image intensifier and the examination table. The calibration
bers of the radiographic images of the lumbar spine can be
board is screened with the VF unit, and the image is recorded.
matched with the electrogoniometer data.
During the assessment, the subjects are asked to wear a
For processing of the radiographic images, the lumbar spine
harness at the upper back and another harness at the back of
motion is captured at every 10° interval of lumbar spine flex-
the pelvis (Figure 3). The electrogoniometer then is attached to
these harnesses and connected to the computer. The subjects
are asked to stand upright between the image intensifier and the
examination table (Figure 4). The intensifier then is positioned
by the side of the subjects’ backs. The lumbar spines are
screened to confirm the correct focus of the image intensifier.
The radiograph beam field of the VF unit also is collimated to
obtain optimal sharpness of the image. The magnification of
the image intensifier is adjusted so that the size of the imaging
field is large enough to view the whole lumbar spine.
Before actual screening, the subjects are asked to practice
flexion– extension of the lumbar spine a few times with correc-
tion. The electrogoniometer and the video recorder then are
activated. The subjects are asked to perform flexion, extension,
and return to the neutral position at their own speed three
times, with rest between. To capture the whole lumbar spine
motion, the image intensifier is controlled by the radiologist
and moved along the lateral side of the back during image
Figure 2. Goniometer and DataQ system developed in-house. capturing while the subjects are performing lumbar flexion and
 Lumbar Spine Motion Analysis • Lee et al E217

Figure 4. Subject stands upright between the image intensifier

and the examination table.

extension. The anatomic motion of the lumbar spine is stored

in the notebook computer while the images of the segmental
spine motion are recorded on a video tape. The retrieved data
are analyzed by the processor discussed in relation to the data
analysis system. The calibration board image then is retrieved,
and the distortion is corrected by 32-bit Morpher (Fujimiya,
Japan) a commercially produced computer software program.
After this procedure, all the captured images of the subjects are
processed for further image analysis.
An in-house image analysis program was developed in-
house using Visual C!! Dibview 4.0. During the image anal-
ysis, this program is run and the processed images of the sub-
jects displayed, frame by frame, on a high-resolution monitor.
Selecting the image when the spine is in the neutral position, the
user then is instructed by a dialogue box to draw straight lines
along the four borders of each vertebral body, from L1 to L5
vertebrae, according to the method devised by Dvorák et al.2,3
With the borders of the vertebral bodies marked, the template
of each vertebra is generated. The position of each vertebra is Figure 5. a, Each vertebra of the motion segment is enclosed by
denoted by the intersection points of the four lines. The coor- four lines, which are tangential to the vertebral outline. b, Deter-
dinates of these points are calculated and saved into data files. mination of the vertebral rotation Rz and the translations, Ax, Ay,
Then the rest of the captured images are retrieved and the Bx, and By, at the vertebral corners of A and B, which have moved
templates of the vertebra redisplayed in the monitor. Users are to A' and B'. The horizontal and vertical translation of each lumbar
asked to drag the templates to overlap the corresponding ver- vertebra then are presented as the translation of the posteroinfe-
tebrae by means of the computer-pointing device. The coor- rior corner of the vertebral body relative to its adjacent distal
segment, and to that of the same levels in the neutral position.2,3
dinates of the lumbar spine in different captured images are
calculated. Following the analysis method devised by
Cholewicki et al1 and Dvorák et al,2 intervertebral flexion–
extension (IVFE) between lumbar vertebrae are then calcu- tebrae, wrapped with a thin layer of polyethylene food wrap.
lated (Figure 5). The model then was repeatedly assessed by the VF from 10° of
flexion to 10° of extension, in 2° intervals. Image analysis was
Evaluation of Errors and Repeatability of the Newly De- performed on the images obtained, and the IVFE of the L3
veloped System. To assess the accuracy of the spinal electro- vertebra was calculated relative to the L4 vertebra. The root
goniometer, the device was mounted on a computerized milling mean square errors were calculated because the errors included
machine with an accuracy of 0.001 mm. The two ends of the both positive and negative values. Because it is relatively com-
electrogoniometer were separated 40 times in 15-mm intervals. mon for subjects to perform flexion– extension of the lumbar
Flexion– extension measurements of the electrogoniometer spine beyond the sagittal plane, errors resulting from out-of-
then were compared with the results of the computerized mill- plane motion were assessed in the model again. It was assumed
ing machine. Assessment of the image analysis system’s errors that out-of-plane motion in the segmental motion of the lum-
was performed using a spine model comprising L3 and L4 bar spine would be 2° in axial rotation. The model was assessed
vertebrae with a single axis metallic hinge joint sandwiched repeatedly by means of the VF unit when the model was fixed in
between. To simulate the scattering effect that the soft tissue of the various combinations of 5° IVFE and 2° axial rotation.
the back in vivo has on the radiograph beam, a piece of The repeatability of marking the radiographic images was
100-mm fresh pork roll was circled around the model of ver- investigated in five subjects. The four corners of each lumbar
E218 Spine • Volume 27 • Number 8 • 2002

Table 1. IVFE of the Lumbar Spine at Different Ranges of Motion and Segmental Levels
Range of Motion (°) L1–L2 L2–L3 L3–L4 L4–L5 L5–S1

#10 #2.64 " 2.18 #2.47 " 2.06 #1.83 " 1.95 #1.30 " 4.33 #2.38 " 3.76
0 0 0 0 0 0
10 3.18 " 2.66 2.75 " 1.95 2.23 " 1.43 1.58 " 1.50 0.72 " 1.32
20 5.46 " 2.87 5.35 " 2.09 4.43 " 1.90 3.00 " 1.84 1.05 " 1.59
30 7.93 " 3.27 7.55 " 2.44 6.36 " 2.35 4.28 " 2.14 1.89 " 2.14
40 10.02 " 3.31 9.64 " 2.67 8.18 " 2.64 5.94 " 2.62 2.79 " 2.40
Positive values $ flexion; negative values $ extension.

vertebra and the two superior corners of the sacrum were Results
marked with the back in the neutral position. The images then
were marked on Days 0, 10, and 20 for each subject. The Evaluation of the Errors and Repeatability of the
coordinates of the vertebral corners in all the segments then Newly Developed System
were compared among the three sessions of measurements by The errors in measuring flexion– extension of the lumbar
means of intraclass correlation coefficient ICC. Using the newly spine using the electrogoniometer were 0.61° " 0.28°.
developed lumbar spine motion system, these five subjects were The average error in measuring the IVFE in the spine
asked further to perform flexion– extension of the lumbar model was 0.4° " 0.3° (range, 0 – 0.8°). The average root
spine. Their VF images then were assessed from 40° of flexion mean square errors in the IVFE, resulting from plane
to 10° of extension, in 10° intervals, and the IVFE was calcu- motion was 0.7° " 0.2°. The average ICC value in mark-
lated at each segmental level of the lumbar spine. The lumbar ing the VF images in in vivo was 0.88, ranging from 0.81
spine motion of these subjects was reassessed on Days 10 and to 1. Similarly, the average ICC value in measurement of
20 after the initial assessment. The ICC values of the measure-
the lumbar spine IVFE in vivo was 0.86 (range,
ments also were calculated.
0.81– 0.96).
Assessment of Lumbar Spine Motion in
Assessment of Lumbar Spine Motion in Healthy Volun- Healthy Volunteers
teers. For this study, 30 healthy volunteers, ages 20 to 30 This study involved 30 healthy volunteers (16 men and
years, were recruited, and both genders were accepted. How- 14 woman). Their mean age was 22.7 years (range,
ever, subjects were excluded if they had experienced low back 20 –29 years). The maximum range of lumbar flexion
pain 1 year before the study, spine surgery, or a query of preg-
was 53° " 10.2°, and the maximum range of lumbar
nancy in female subjects. Written consent was obtained from
extension was 15.4° " 8.3°. Table 1 summarizes the
all the subjects. The estimated radiation dose given by the as-
sessment procedure was only 1.5 mSv for male subjects and 2.3 IVFE for each vertebral level. At L1–L2, the IVFE in-
mSv for female subjects, which is less than the radiation expo- creased steadily from 10° of extension to 40° of flexion
sure for a single plain radiograph of the lumbar spine (2.4 (Figure 6). A linear pattern of the IVFE curve was ob-
mSv). Ethical approval of this study was obtained from both served. The IVFE curve at L2–L3 aligned immediately
The Hong Kong Polytechnic University and the Queen Eliza- next to L1–L2, and a linear pattern of the curve also was
beth Hospital. observed. Similarly, the IVFE curve at L3–L4, L4 –L5,
During the assessment, the subjects were asked to wear a and L5–S1 increased steadily from 10° of extension to
harness at the upper back and another harness at the back of 40° of flexion, except for 10° of extension at L5–S1. In
the pelvis. The electrogoniometer then was attached to these general, a linear pattern of the IVFE curve at different
harnesses and connected to the computer. The subjects were
levels was found. Moreover, the IVFE decreased in de-
asked to stand upright within the VF unit. A quick screening of
the low back was performed to ensure the correct positioning
of the subject. The electrogoniometer and video recorder then
were activated. The subjects were asked to perform lumbar
flexion– extension and return to the neutral position at their
own speed three times. The lumbar spine motion of the subjects
was assessed with the developed method. Intervertebral flex-
ion– extension was calculated in 10° intervals, from 40° of flex-
ion to 10° of extension. Using the linear regression method, the
change in IVFE throughout the assessment range of motion was
calculated at each vertebral level and in different gender groups.
Comparison of the change in IVFE throughout the assessment
range of motion was performed with one-way analysis of vari-
ance (ANOVA). The level of significance was set at 0.05. For
any statistically significant findings obtained in the ANOVA,
Scheffe post hoc tests were performed to identify the differences Figure 6. Intervertebral flexion– extension of the lumbar spine in
among the IVFE at each vertebral level. different ranges of motion.
 Lumbar Spine Motion Analysis • Lee et al E219

tients with degenerative spondylolisthesis, but not in pa-

tients with low back pain. Harada et al9 also investigated
the intervertebral motion of L3 to S1 in 10 healthy males
volunteers using cineradiography. Motion profiles at dif-
ferent vertebral levels were compared.
Unfortunately, assessment of intervertebral motion in
these studies was limited to only a few lumbar segments
(e.g., L3 to S1), or subjects were not allowed to perform
full lumbar flexion– extension range of motion naturally.
This was because the screening field size of the cineradio-
graphic machine is small whereas the excursion of the
whole lumbar spine is large during full lumbar flexion–
extension range of motion. The image of the spine easily
escaped beyond the screening field. As compared with
previous studies, the image intensifier of the current
Figure 7. The slopes of intervertebral flexion– extension curve for study therefore was not fixed, and was driven by a radi-
different lumbar motion segments (*P % 0.05). ologist so that full lumbar spine range of motion would
be tracked. As a result, the dynamic motion of the whole
lumbar spinal column could be assessed. Subjects were
scending order from L1–L2 to L5–S1 at different points free to perform lumbar flexion– extension in a natural
of flexion range of motion. However, the IVFE values manner. The analysis of the VF images had been confined
were found to be about the same at different levels in from 40° of flexion to 10° of extension because this was
extension, except for the IVFE at L5–S1, which relatively the common total flexion range of motion of all the sub-
less than that of the others. jects. The developed technique could investigate the VF
Using linear regression, the slopes of the IVFE curve at images in a more extensive lumbar movement without
different levels and in different genders were calculated. the practical limitations in previous studies.
No statistically significant difference in the slope was Another drawback in previous studies is that the inter-
found between the genders, as shown by ANOVA. vertebral motion was assessed at certain fixed time points,
Therefore, the slope of the IVFE curves at different levels whereas time has been chosen to be the external reference.
for all the subjects are plotted in Figure 7. The slope Comparison of the results sourced from either fixed frame
decreased gradually from L1–L2 to L5–S1. Comparison numbers or certain fixed seconds,17,19 between subjects is
among the slopes of different levels was made with questionable. Although Harada et al9 tried to standardized
ANOVA, and statistical significance difference was the time of the lumbar movement within 6 seconds, it
found (P % 0.05). Therefore, Scheffe post hoc tests were was obvious that even a small difference in time could
performed, and statistical significant differences were lead to large variation of the motion analysis results. The
found among all vertebral levels (P % 0.05). current authors therefore proposed assessment of the in-
tervertebral motion at a certain fixed anatomic range of
Discussion
motion of the lumbar spine, which is not a time-
Previously, lumbar spine motion commonly was studied dependent parameter. Hence, the assessment method of
by means of radiographic methods. However, spine pos- spinal motion analysis can be standardized so that the
ture is seldom controlled during radiographic imaging. pattern or quality of spine motion can then be quantified.
As a result, interpretation and comparison of the results Despite the interesting findings of Kanayama et al,11
are difficult. This is particularly true in assessment of Okawa et al,19 and Harada et al,9 the errors and reliabil-
dynamic intervertebral spine motion. ity of their spine motion analysis methods have not been
Fielding6 was the pioneer who used cineradiography detailed in their reports. Although the errors in our
to assess the dynamic motion of the cervical spine. With newly developed method may not be the least, we found
recent advancement of technology, it has become possi- that it is crucial to assess and report these in this report.
ble to capture the dynamic motion of the spine and assess It is hoped that in the future this may encourage other
it frame by frame using computer-aided assessment tech- researchers to develop further methods in lumbar spine
nique. Using cineradiography, Kanayama et al11 investi- motion analysis based on the current protocol.
gated the segmental motion of L3 to S1 in eight healthy In the studies of Kanayama et al,11 lower lumbar spine
male subjects. Specific intervertebral motion pattern was motion was found to be stepwise from the upper level,
observed in both flexion and extension of the lumbar with intersegmental motion lags during flexion. Harada
spine. Okawa et al19 investigated whole lumbar spine et al9 also reported that lumbar flexion was initiated
motion using VF at five frames per second in 29 subjects, from upper to lower segments, with phase lags. How-
including 13 healthy volunteers, 8 patients with low ever, the current results do not agree with this finding
back pain, and 8 patients with degenerative spondylolis- because IVFE was found in all levels of lower lumbar
thesis. A different motion pattern was identified in pa- spine. As shown in Figure 7, the slopes of the IVFE curve
E220 Spine • Volume 27 • Number 8 • 2002

at L1–L2 and L2–L3 were nearly double or even triple 2. Dvoák J, Panjabi MM, Novotny JE, et al. Clinical validation of functional
flexion– extension roentgenograms of the lumbar spine. Spine 1991;16:943–50.
those of L4 –L5 and L5–S1, respectively. This implies 3. Dvoák J, Vajda EG, Grob D, et al. Normal motion of the lumbar spine as
that every segment of the lumbar spine does move, but related to age and gender. Eur Spine J 1995;4:18 – 423.
that the motion is unevenly contributed at different lev- 4. Ensink FB, Saur PM, Frese K, et al. Lumbar range of motion: influence of
time of day and individual factors on measurements. Spine 1996;21:1339– 43.
els. The observation of phase lags between levels in pre- 5. Esola MA, McClure PW, Fitzgerald GK, et al. Analysis of lumbar spine and
vious studies may be explained by greater flexion of the hip motion during forward bending in subjects with and without a history of
upper levels as compared with the lower levels. This re- low back pain. Spine 1996;21:71– 8.
6. Fielding JW. Cineroetgenography of the normal cervical spine. J Bone Joint
sults in an apparent stepwise motion from upper to lower Surg [Am] 1957;39:1280 – 8.
levels of the lumbar spine in flexion. It has been suggested 7. Glossop N, Hu R. Assessment of vertebral body motion during spine surgery.
that lower segments of the lumbar spine are the key seg- Spine 1997;22:903–9.
8. Gracovetsky S, Newman N, Pawlowsky M, et al. A database for estimating
ments initiating extension.9,11 However, our current normal spinal motion derived from noninvasive measurements. Spine 1995;
results show that IVFE was about the same in different 20:1036 – 46.
levels in extension. This implies that extension is contrib- 9. Harada M, Abumi K, Ito M, et al. Cineradiographic motion analysis of
normal lumbar spine during forward and backward flexion. Spine 2000;25:
uted relatively evenly by different levels, and that the 1932–7.
concavity of lumbar lordosis will increase steadily in 10. Hayes MA, Howard TC, Gruel CR, et al. Roentgenographic evaluation of
extension. lumbar spine flexion– extension in asymptomatic individuals. Spine 1989;
14:327–31.
Generally, in the current study, a linear pattern of the 11. Kanayama M, Abumi K, Kaneda K, et al. Phase lag of the intersegmental
IVFE curve at different levels was found, and the IVFE motion in flexion– extension of the lumbar and lumbosacral spine: an in vivo
decreased in order from L1–L2 to L5–S1 at different study. Spine 1996;21:1416 –22.
12. Leivseth G, Brinckmann P, Frobin W, et al. Assessment of sagittal plane
points of range of motion in flexion. However, Luk et segmental motion in the lumbar spine: a comparison between distortion-
al14 observed a different pattern in their study using lat- compensated and stereophotogrammetric roentgen analysis. Spine 1998;23:
eral radiograph. This could be because their method of 2648 –55.
13. Lu WW, Luk KD, Ruan DK, et al. Stability of the whole lumbar spine after
imaging differed completely from that in the current multilevel fenestration and discectomy. Spine 1999;24:1277– 82.
study. According to the description, their films were 14. Luk KD, Chow DH, Evans JH, et al. Lumbar spinal mobility after short
taken in side-lying position. During flexion, the subject anterior interbody fusion. Spine 1995;20:813– 8.
15. Mayer RS, Chen IH, Lavender SA, et al. Variance in the measurement of
flexed the lumbar spine maximally, with the arms going sagittal lumbar spine range of motion among examiners, subjects, and in-
around the hips and knees. Extension of the lumbar spine struments. Spine 1995;20:1489 –93.
was achieved with manual assistance. As compared with 16. Mayer TG, Kondraske G, Beals SB, et al. Spinal range of motion: accuracy
and sources of error with inclinometric measurement. Spine 1997;22:1976 – 84.
our method, the subjects in the current study were free to 17. McGregor AH, McCarthy ID, Hughes SP. Motion characteristics of the
move from the neutral position to full flexion and then to lumbar spine in the normal population. Spine 1995;20:2421– 8.
full extension in the standing position. A more natural 18. Miyasaka K, Ohmori K, Suzuki K, et al. Radiographic analysis of lumbar
motion in relation to lumbosacral stability: investigation of moderate and
and active movement of the lumbar spine was assessed. maximum motion. Spine 2000;25:732–7.
In conclusion, the newly developed technique for as- 19. Okawa A, Shinomiya K, Komori H, et al. Dynamic motion study of the
sessing dynamic lumbar motion is reliable. Development whole lumbar spine by videofluoroscopy. Spine 1998;23:1743–9.
20. Saur PM, Ensink FB, Frese K, et al. Lumbar range of motion: reliability and
of a database of lumbar spine motion in the general pop- validity of the inclinometer technique in the clinical measurement of trunk
ulation is important because once the database is estab- flexibility. Spine 1996;21:1332– 8.
lished, comparison of pathologic spine motion disorder
will be possible. A precise clinical diagnosis of spine mo-
tion disorder, such as spine “instability,” can then be Address reprint requests to
made. As a result, effective treatment can be introduced. Sai-wing Lee, PhD
Further studies comparing lumbar spine motion in dif- Department of Rehabilitation Sciences
ferent age groups and patient profiles are suggested. Hong Kong Polytechnic University
Hunghom, Kowloon
References Hong Kong
E-mail address: rsswlee@polyu.edu.hk
1. Cholewicki J, McGill SM, Wells RP, et al. Method for measuring vertebral
kinematics from videofluoroscopy. Clin Biomech 1991;6:73– 8.