Manual Therapy 18 (2013) 130e135

Contents lists available at SciVerse ScienceDirect

Manual Therapy
journal homepage: www.elsevier.com/math

Original article

A novel approach to the clinical evaluation of differential kinematics of the

lumbar spineq
Jonathan Mark Williams a, *, Inam Haq b, Raymond Y. Lee c
a
School of Health and Social Care, Bournemouth University, Royal London House, Christchurch Road, Bournemouth, Dorset BH1 3LT, UK
b
Brighton and Sussex Medical School, UK
c
Department of Life Sciences, University of Roehampton, UK

a r t i c l e i n f o a b s t r a c t

Article history: Clinical measurement of lumbar motion has traditionally been limited to range of motion (ROM). Despite
Received 6 April 2012 this, deficits in angular velocities and accelerations are more pronounced compared to ROM in low back
Received in revised form pain (LBP) sufferers. There is increasing interest in movement quality among manual therapists and
1 August 2012
therefore the ability to measure angular velocities and accelerations within the clinical environment is
Accepted 30 August 2012
becoming increasingly important.
The aims of this study were to (1) investigate the reliability of a clinic based inertial sensor system to
Keywords:
measure ROM along with angular velocities and accelerations in low back pain sufferers; (2) introduce
Lumbar spine
Kinematics
the feasibility and reliability of using the relationship between ROM and velocity to investigate move-
Reliability ment trajectory and irregularity.
Velocity Forty LBP sufferers completed three trials of spinal movements and lifting. The ROM curve was
differentiated and double differentiated to yield angular velocities and accelerations.
Repeated measures reliabilities were determined by comparisons of kinematic curves as well as peak
values. ROM and angular velocity relationships were investigated for their use in describing the move-
ment trajectory and irregularity.
Results show excellent similarities of ROM and angular velocity curves and moderate-to-good simi-
larities for angular acceleration curves. Peak value similarities were excellent with small error
measurements for all variables.
The quantification of ROM-angular velocity plots was reliable with small mean absolute differences in
motion irregularity scores. Such a method was able to demonstrate differences in movement irregularity.
This method provides clinicians with the ability to yield important additional movement related
information including angular velocity, acceleration and movement irregularity.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction movement. Some laboratory methods such as electromagnetic or

Measurement of lumbar motion is common in the assessment of
low back pain (LBP) (Cocchiarella and Andersson, 2000). Clinical
measures including skin distension (MacRae and Wright, 1969;
Dolan et al., 1995), finger tip-to-floor (Frost et al., 1982; Battie et al.,
1987; Gauvin et al., 1990) and inclinometers (Burdett et al., 1986;
Saur et al., 1996; Bierma-Zeinstra et al., 2001; Kachingwe and
Phillips, 2005) are unable to provide continuous data. This means
they are unable to provide information about patterns of
q This work is affiliated to the Department of Life Sciences, University of
Roehampton.
* Corresponding author. Tel.: þ44 (0) 7766141620.
E-mail addresses: jwilliams@bournemouth.ac.uk, jon__williams@hotmail.com
(J.M. Williams).
opto-electronic video based systems can provide this extra infor-
mation, but often the reporting has been limited to range of motion
(ROM) only.
It is well documented that LBP sufferers display alterations in
spinal kinematics (Shum et al., 2007a, 2007b), but this is most
pronounced in the differential kinematics of the movement, i.e.
angular velocities and accelerations (Marras and Wongsam, 1986;
Marras et al., 1995, 1999; Novy et al., 1999). Angular velocities and
accelerations have been able to accurately discriminate between
LBP and non-LBP sufferers (Marras et al., 2000) as well as distin-
guish risk of LBP within the workplace (Marras et al., 1993).
However, the reliability of the measurements of angular velocity
and acceleration need to be further established.
A few studies (Shum et al., 2005a, 2005b) have looked at the
ROM-angular velocity relationship. This relationship, known as the

1356-689X/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.math.2012.08.003
 J.M. Williams et al. / Manual Therapy 18 (2013) 130e135
spatial relationship, can be visualised by a ROM-velocity plot. The
diagram provides a useful clinical picture of examining the spine
from a dynamical perspective, in which the emphasis is on repre-
senting the kinematic control of the system (Burgess-Limerick
et al., 1993; Li et al., 1999). Visualising motion control is impor-
tant because various studies have shown that low back pain is
associated with altered control (Van Dillen et al., 2009; Hodges
et al., 2009; Xu et al., 2010). One novel use of this visualisation is
to examine the degree of movement irregularity, which cannot be
revealed by routine ROM assessment or the more familiar ROM-
time plot.
In the clinic the above approach can be easily achieved with the
use of inertial sensors. Similar sensors have been used to study gait,
movements of the upper and lower limb as well as the spine (Tong
and Granat, 1999; Lee et al., 2003; Zhou et al., 2008; Saber-Sheikh
et al., 2010; Cuesta-Vargas et al., 2010; Theobald et al., 2012; Ha
et al., in press). However before such sensors are recommended
for clinical use it is imperative to determine the reliability in LBP
sufferers within a clinical setting, including the angular velocities
and accelerations. Furthermore the feasibility of analysing and
quantifying movement irregularity through spatial relationship
plots should be established.
The aim of this study was to (1) investigate the similarity
between repeated movements, known as repeated measures reli-
ability, of an inertial sensor system to measure ROM, angular
velocity and acceleration of LBP sufferers; (2) to describe the use of
spatial plots to study the movement irregularity of spinal
movements.
2. Methods
2.1. Participants
Forty LBP sufferers were initially recruited (twenty acute and
twenty chronic) from general practitioner referrals to local therapy
departments. Referrals were screened for inclusion and exclusion
criteria. Inclusion criteria included pain confined to between the
lower ribs and inferior gluteal folds, 18e55 years old, seeking
healthcare for their LBP and evoked pain on at least 3 of the test
movements. Exclusion criteria included a history of tumours, spinal
fractures, surgery, rheumatological or neurological diseases and
any neurological signs or symptoms. Acute pain was pain present
for less than three weeks and chronic defined as pain present on at
least three days per week for greater than twelve months. Four
acute low back pain (ALBP) and eight chronic low back pain (CLBP)
sufferers were excluded as they did not demonstrate three or more
movements which evoked pain. Sixteen acute LBP sufferers (male
10; age 42.6  7.2 years; BMI 27.7  3.5 kg/m2; duration 12.0  7.3
days) and twelve chronic LBP sufferers (male 7; age 34.5  10.0
years; BMI 27.4  4.8 kg/m2; duration 4030.2  2992.2 days)
recorded their mean pain scores from the week preceding data
collection, using a visual analogue scale (VAS) (61.5  18.9 mm
ALBP; 45.7  22.1 mm CLBP) and a tampa scale of kinesiophobia
questionnaire (39.3  4.1 ALBP; 38.3  7.5 CLBP).
This study was approved by the National Research Ethics Service
of the National Health Service and written informed consent was
obtained following explanation of procedures and risks.
2.2. Instrumentation
Two wired inertial sensors were used to measure lumbar
kinematics (3DM-GX3-25, Microstrain, VT, USA). Each sensor con-
tained tri-axial gyroscopes (300 s1), accelerometers (5 g) and
magnetometers. Each sensor was cased with dimensions 44 mm
(h)  25 mm (w)  11 mm (d), and weight of 18 g. The sensors have
131
a reported accuracy of 0.5 during static testing and 2.0 during
dynamic testing.
One sensor was placed over the S1 spinous process and the
second over the L1 spinous process. The sensors were attached
using double sided tape with the wires secured to the trunk so as
not to move the sensors erroneously. The sensors were connected
to a purpose built datalogger and software (ThetaMetrix, UK) and
data were collected at 100 Hz. The relative orientations between
the L1 and S1 sensor were determined from the direction cosine
matrices (Grood and Suntay, 1983; Burnett et al., 1998; Lee et al.,
2003). Flexion, left side bending and left rotation were consid-
ered positive and the opposite movements negative.
2.3. Procedure
Participants stood upright and completed three trials of flexion,
extension, left side bending, right side bending, left rotation, right
rotation and a box lift. The box (460  260  300 mm) was posi-
tioned on markers ensuring identical placement and weighed 3 kg.
The movement order was identical for all participants and the
sensors remained attached throughout.
2.4. Data analysis
All raw data were transferred to Matlab for processing (Math-
works, R2008b). The ROM data for each movement were deter-
mined and smoothed using a fourth order zero-lag Butterworth
filter with a cut off frequency of 1 Hz (Tsang et al., 2011). The ROM
data were differentiated to yield the velocity and double differen-
tiated to calculate acceleration. The ROM curves were time-
normalised so that each movement had the same time base from
0% to 100%. This normalisation process takes into account the
differing rates at which the participants completed the trials and
allows the direct comparison of the kinematic pattern. The coeffi-
cient of multiple correlation (CMC) (Li and Caldwell, 1999; Williams
et al., 2010) and root mean square error (RMSE) were calculated
from the normalised range of motion, angular velocity and accel-
eration curves to determine the similarities between the three
movement trials. The peak ROM, peak angular velocity and peak
angular acceleration values were obtained and intra-class coeffi-
cients (ICC3,1) calculated along with the mean absolute differences
between peak values for the three trials. Two peak values were
obtained, positive and negative. Positive velocity and acceleration
relate to flexion, left side bending and left rotation movement,
whereas negative velocity and acceleration relate to movement in
the opposite direction.
Angular velocity-ROM plots were used to reveal the spatial
relationship where the overall shape of the plot describes the
trajectory of the movement and dynamic control. Movement
irregularity was determined from the plots by separating the plot
into quartiles (see Fig. 2). The quartiles were defined as the region
from motion onset to peak angular velocity (quartile 1); peak
angular velocity to end ROM or maximum angular displacement
(quartile 2); maximum angular displacement to minimum angular
velocity (quartile 3) and minimum angular velocity to minimum
angular displacement or return to upright standing (quartile 4).
Each quartile was fitted with a 4th order polynomial and the
beginning and end of the polynomial was fixed to match the
collected data. The polynomial was used to determine the RMS
difference between the actual data and the polynomial, providing
quantification of motion irregularity for each quartile. The RMS was
normalised to the maximum velocity of each individual to provide
a value representing the percentage of maximum velocity. Scores of
motion irregularity (RMS) were determined for each quartile across
132 J.M. Williams et al. / Manual Therapy 18 (2013) 130e135

Table 1
Mean (sd) values of kinematic variables recorded by the inertial system.

Disp ( ) þve Vel ( s1) ve Vel ( s1) þve Acc ( s2) ve Acc ( s2)

ALBP CLBP ALBP CLBP ALBP CLBP ALBP CLBP ALBP CLBP
Flex 35.1 (14.5) 48.1 (11.4) 20.1 (15.8) 28.2 (9.6) 20.1 (11.8) 29.0 (10.0) 29.8 (18.8) 34.1 (28.7) 26.6 (16.7) 41.3 (21.0)
Ext 11.8 (8.8) 16.5 (5.1) 10.6 (6.3) 15.0 (6.3) 6.2 (3.0) 10.0 (4.3) 18.9 (10.1) 25.1 (12.0) 14.6 (8.1) 26.1 (14.5)
LSF 11.5 (3.7) 14.6 (6.9) 8.2 (4.4) 9.6 (5.8) 11.2 (4.3) 11.6 (5.8) 15.3 (8.5) 18.9 (11.4) 17.5 (7.6) 17.6 (10.8)
RSF 15.0 (4.1) 15.9 (4.6) 13.4 (4.8) 13.6 (5.9) 8.8 (3.7) 10.2 (5.1) 20.9 (9.3) 20.6 (11.8) 17.2 (8.2) 20.7 (11.3)
LRot 7.1 (3.5) 4.4 (3.6) 5.0 (2.6) 4.5 (3.1) 5.1 (2.6) 5.0 (3.9) 11.5 (4.5) 13.5 (9.2) 10.6 (4.6) 13.5 (8.9)
RRot 10.6 (4.5) 8.3 (4.1) 7.0 (3.0) 6.1 (4.2) 5.6 (2.5) 6.4 (4.8) 12.2 (7.0) 14.9 (10.1) 12.7 (5.7) 14.2 (9.9)
Lift 29.9 (15.0) 43.8 (12.6) 21.5 (14.0) 33.8 (13.9) 26.6 (13.5) 36.3 (14.5) 44.1 (23.3) 59.7 (29.8) 42.8 (24.6) 58.5 (30.5)

Flex, Flexion; Ext, Extension; LSF, Left side flexion; RSF, Right side flexion; LRot, Left rotation; RRot, Right rotation; Disp, Displacement; þve, positive; ve, negative; Vel,
Velocity; Acc, Acceleration.

each movement and ICC3,1 and mean absolute difference between angular acceleration and movement irregularity through spatial
trials were used to determine reliability of such a method. plot analysis.
The ROM and angular-velocity plots, as well as peak values
demonstrate good reliability and small errors, however the results
3. Results
for angular-acceleration plots show slightly lower results in
comparison. This may suggest greater acceleration movement
Mean (sd) for displacement, velocity and acceleration for each
variation from the participants; however it may also reflect the
movement can be found in Table 1.
sensitivity of the double derivative technique to yield accelerations.
The mean CMC values were found to be good across all repeated
The peak acceleration ICC values were good suggesting this variable
range of motion curves for the ALBP (0.83e0.97  0.02e0.09) and
is highly reliable for use in quantifying lumbar acceleration in LBP
CLBP (0.72e0.94  0.03e0.26) groups. This was also seen for
sufferers. Moreover the mean absolute difference in peak value was
angular velocity curves (0.83e0.94  0.04e0.10 ALBP and 0.71e
small, especially for movements other than lifting.
0.91  0.05e0.18 CLBP).
The mean CMC values for angular acceleration curves show
moderate consistency across the different movements tested
(0.61e0.77  0.10e0.18 ALBP and 0.52e0.73  0.10e0.26 CLBP). The
RMSE magnitudes across all kinematic variables were small for all
repeated movements (Fig. 1).
Mean ICC values were shown to be good across all movements
for both groups (0.71e0.99) and mean absolute differences of
repeated peak measures were small, <3.7 and <5.6 s1 for ROM
and angular velocity and moderate <13.5 s2 for angular acceler-
ation (Fig. 1).
The similarity in irregularity scores between repeated trials was
moderate-to-good for all the movements investigated (0.49e0.86
ALBP and 0.50e0.83 CLBP group). Mean absolute differences
between repeated movements were small (1.2e5.5) suggesting
good reliability of such a method.
Fig. 2 shows a typical spatial plot exhibiting motion irregularity.
It can be visualised from graph 2(a) that the second quartile
displays greater irregularity especially at more than 40 ROM. It is
also around this region that more irregularity is evident for quartile
3 resulting in a score of 2.6. If this graph is compared to 2(b) then it
is evident that quartiles 1 and 4 are again those with fewest
irregularities, as mirrored by the group results, however the second
quartile demonstrates much greater irregularity (score ¼ 8.5). This
is in part due to some jerkiness visualised in the final 5 of the
movement. Quartile 3 also scores highly due to motion irregularity
in the first 5 of returning from flexion.
The mean numerical motion irregularity score for each quartile
for each group as a whole is presented in Table 2, and shows that
significant between group difference in motion irregularity values
during extension, left and right side flexion, left rotation and lifting.
The quartile with the greatest irregularity was consistently the
second for all movements except left rotation and lifting in the CLBP
group.

4. Discussion
The results of this study show that inertial sensors are capable of
measuring lumbar kinematics including ROM, angular velocity,
Fig. 1. Root mean square error values for displacement-, velocity- and acceleratione
time curves along with absolute mean difference in peak values of displacement,
velocity and acceleration (Flex, Flexion; Ext, Extension; LSF, Left Side-Flexion; RSF,
Right Side-Flexion; Lrot, Left Rotation; Rrot, Right Rotation).
 J.M. Williams et al. / Manual Therapy 18 (2013) 130e135
Fig. 2. Quantification of spatial plots. (a) Flexion trial of individual participant. (b) Flexion trial of different individual.
The results of this study are comparable to those available in the
literature (Marras and Wongsam, 1986; Esola et al., 1996; Granata
et al., 1997; Marras and Granata, 1997; McClure et al., 1997;
Marras et al., 2000, 2001; Pal et al., 2007; Shum et al., 2007a;
Milosavljevic et al., 2008). The differences that do exist may be due
to the different characteristics in the participants, with non-LBP
participants displaying greater angular velocities and accelera-
tions. These differences may be due to greater intra-participant
movement variation possibly due to having or anticipating move-
ment evoked pain. This may be especially true for angular velocities
and accelerations which would be significantly affected by less
smooth movements.
The results of this study demonstrate that quantification of
motion irregularity using spatial plots is possible and the results, as
a score of movement irregularity, are reliable. This is the first time
such movement profiles have been quantified and applied to LBP
sufferers and it is evident that this new information enables the
identification of which section of the movement is affected. The
ability of the spatial plots to display movement coordination and
control is of great use to the clinician. The quantification of
movement irregularity demonstrates a consistent pattern across
the LBP sufferers where more irregular motions are evident in the
quartile leading up to end of range. This may possibly be due to
attempts to minimise provocation of pain, as the individual adjusts
or ‘explores’ the movement close to the terminal range in an
attempt to find the most comfortable path. This results in devia-
tions in velocity behaviour and causes greater scores in the irreg-
ularity quantification. However, it may also be the case that this loss
of movement smoothness represents an impairment of spinal
function due to perhaps alteration in proprioceptive input and
133
therefore represents an attempt to increase afferent information to
guide the movement pattern.
Previous research has already provided extensive information
about how manual therapy may improve spinal ROM (Goodsell
et al., 2000; Lee et al., 2005; Powers et al., 2008). However, it
has been established that ROM does not correlate well with
function in LBP sufferers (Parks et al., 2003), and that LBP sufferers
have large deficits in angular-velocities and accelerations (Marras
and Wongsam, 1986; Marras et al., 1995, 1999; Novy et al., 1999;
Shum et al., 2007a, 2007b). Furthermore clinicians are increasingly
interested in movement coordination and control (O’Sullivan,
2000, 2005; Shum et al., 2005a; Dankaerts et al., 2007). These
methods of quantifying movement behaviour will enhance the
understanding of movement coordination and control as well as
enabling clinicians to study, in detail, the effect of interventions
targeting movement control. The current work allows further
research to look at the effects of manual therapy with new
perspectives. For example, it would be able to show how manual
therapy affects the angular velocity and acceleration of the spine.
Using spatial plots, it would be possible to visualise how manual
therapy affects control or movement irregularities in different
quartiles of spine motion.
The inertial sensors are easy to use and provide information that
could be analysed immediately after clinical assessment. Although
the current method involves some complexities in data processing,
the algorithms can easily be incorporated into carefully designed
computer software providing the clinician with immediate infor-
mation about ROM, angular velocity, angular accelerations and
movement irregularity through spatial plot analysis. Indeed this is
currently being explored in on-going development.
134 J.M. Williams et al. / Manual Therapy 18 (2013) 130e135
Table 2
Quantification of movement irregularity per quartile.
ALBP
Flexion
1 3.1
2 6.4
3 4.4
4 3.5
Extension
1 6.0
2 13.1
3 7.2
4 3.1
LSF
1 2.0
2 8.0
3 5.5
4 1.3
RSF
1 1.4
2 7.1
3 4.1
4 1.1
LRot
1 6.8
2 11.4
3 9.3
4 11.4
RRot
1 5.1
2 10.5
3 7.4
4 5.6
Lift
1 3.4
2 4.7
3 4.1
4 2.2
*P ¼ 0.05 or less.
5. Conclusion
The results of this study show that the inertial sensor system is
capable of providing a method of measuring ROM, angular veloci-
ties and accelerations in LBP sufferers within a clinical environ-
ment. The additional kinematic information can be incorporated to
study the spatial relationships providing a measure of movement
irregularity. These data processing methods can be simplified by
their inclusion into automated computer software. Therefore it is
possible to study detailed kinematics of the lumbar region which
may yield important clinical information regarding movement
quality and control.
Acknowledgements
The authors would like to thank Henry Robinson, ThetaMetrix
for his technical assistance and the Private Physiotherapy Education
Foundation for their financial assistance.
References
Battie MC, Bigos SJ, Sheehy A, Wortley MD. Spinal flexibility and individual factors
that influence it. Physical Therapy 1987;67(5):653e8.
Bierma-Zeinstra SMA, van Gool JJCM, Bernsen RMD, Njoo KH. Measuring the sacral
inclination angle in clinical practice: is there an alternative to radiographs.
Journal of Manipulative and Physiological Therapeutics 2001;24(8):505e8.
Burdett RG, Brown KE, Fall MP. Reliability and validity of four instruments for
measuring lumbar spine and pelvic positions. Physical Therapy 1986;66(5):
677e84.
Burgess-Limerick R, Abernethy B, Neal RJ. Relative phase quantifies interjoint
coordination. Journal of Biomechanics 1993;26(1):91e4.
Burnett AF, Barrett CJ, Marshall RN, Elliot BC, Day RE. Three-dimensional
measurement of lumbar spine kinematics for fast bowlers in cricket. Clinical
Biomechanics 1998;13(8):574e83.
CLBP Cocchiarella L, Andersson GBJ. Guides to the evaluation of permanent impairment.
5th ed. AMA Bookstore; 2000.
3.7 Cuesta-Vargas AI, Galan-Mercant A, Williams JM. The use of inertial sensors system
for human motion analysis. Physical Therapy Reviews 2010;15(6):462e73.
4.9
Dankaerts W, O’Sullivan PB, Burnett AF, Straker LM. The use of a mechanism-based
3.7
classification system to evaluate and direct management of a patient with non-
2.4
specific chronic low back pain and motor control impairment. Manual Therapy
2007;12(2):181e91.
6.1 Dolan P, Mannion AF, Adams MA. ‘Schober test’ measurements do not correlate well
6.3* with angular movements of the lumbar spine. In: Paper presented at the
5.5 International Society for the Study of the Lumbar Spine, Helsinki, Finland; 1995.
4.0 Esola MA, McClure PW, Fitzgerald GK, Siegler S. Analysis of lumbar spine and hip
motion during forward bending in subjects with and without a history of low
5.6* back pain. Spine 1996;21(1):71e8.
5.8 Frost M, Stuckey S, Smalley LA, Dorman G. Reliability of measuring trunk motions in
3.6 centimeters. Physical Therapy 1982;62(10):1431e7.
2.6 Gauvin MG, Riddle DL, Rothstein JM. Reliability of clinical measurements of forward
bending using the modified fingertip-to-floor method. Physical Therapy 1990;
70(7):443e7.
3.5*
Goodsell M, Lee M, Latimer J. Short-term effects of lumbar posteroanterior mobi-
7.2
lization in individuals with low back pain. Journal of Manipulative and Physi-
4.2 ological Therapeutics 2000;23(5):332e42.
2.9* Granata KP, Marras WS, Davis KG. Biomechanical assessment of lifting dynamics,
muscle activity and spinal loads while using three different styles of lifting belt.
16.7* Clinical Biomechanics 1997;12(2):107e15.
14.3 Grood E, Suntay M. A joint coordinate system for the clinical description of three-
13.3 dimensional motions: application to the knee. Journal of Biomechanical Engi-
11.5 neering 1983;105(2):136e44.
Ha TH, Saber-Sheikh K, Moore AP, Jones M. Measurement of lumbar spine range of
6.4 motion using inertial sensors e a protocol validity study. Manual Therapy, in press.
14.3 Hodges P, van den Hoorn W, Dawson A, Cholewicki J. Changes in the mechanical
11.5 properties of the trunk in low back pain may be associated with recurrence.
Journal of Biomechanics 2009;42(1):61e6.
8.8
Kachingwe AF, Phillips BJ. Inter- and intrarater reliability of a back range of motion
instrument. Archives of Physical Medicine and Rehabilitation 2005;86(12):
2.9 2347e53.
4.3 Lee RYW, Laprade J, Fung EHK. A real-time gyroscopic system for three-dimensional
3.8 measurement of lumbar spine motion. Medical Engineering and Physics 2003;
4.5* 25(10):817e24.
Lee RY, Tsung B, Tong P, Evans J. Posteroanterior mobilization reduces the bending
stiffness and the pain response in the lumbar spine. In: Proceedings of the 2nd
international conference on movement dysfunction, Edinburgh, UK; 2005.
Li L, van den Bogert EC, Caldwell GE, van Emmerik RE, Hamill J. Coordination
patterns of walking and running at similar speed and stride frequency. Human
Movement Science 1999;18(1):67e85.
Li L, Caldwell GE. Coefficient of cross correlation and the time domain correspon-
dence. Journal of Electromyography and Kinesiology 1999;9(6):385e9.
MacRae IF, Wright V. Measurement of back movement. Annals of the Rheumatic
Diseases 1969;28(6):583e9.
Marras WS, Wongsam PE. Flexibility and velocity of the normal and impaired lumbar
spine. Archives of Physical Medicine and Rehabilitation 1986;67(4):213e7.
Marras WS, Lavender SA, Rajulu SL, Allread WG, Fathallah FA, Ferguson SA. The role
of dynamic three-dimensional trunk motion in occupationally-related low back
disorders. The effects of workplace factors, trunk position, and trunk motion
characteristics on risk of injury. Spine 1993;18(5):617e28.
Marras WS, Parnianpour M, Ferguson SA, Kim J-Y, Crowell RR, Bose S, et al. The
classification of anatomic- and symptom-based low back disorders using
motion measure models. Spine 1995;20(23):2531e46.
Marras WS, Granata K. Spine loading during trunk lateral bending motions. Journal
of Biomechanics 1997;30(7):697e703.
Marras WS, Ferguson SA, Gupta P, Bose S, Parnianpour M, Kim J-Y, et al. The
quantification of low back disorder using motion measures. Methodology and
validation. Spine 1999;24(20):2091e100.
Marras WS, Lewis KEK, Ferguson SA, Parnianpour M. Impairment magnification
during dynamic trunk motion. Spine 2000;25(5):587e95.
Marras WS, Davis KG, Ferguson SA, Lucas BR, Gupta P. Spine loading characteristics
of patients with low back pain compared with asymptomatic individuals. Spine
2001;26(23):2566e74.
McClure P, Esola M, Schreier R, Siegler S. Kinematic analysis of lumbar and hip
motion while rising from a forward, flexed position in patients with and
without a history of low back pain. Spine 1997;22(5):552e8.
Milosavljevic S, Pal P, Bain D, Johnson G. Kinematic and temporal interactions of the
lumbar spine and hip during trunk extension in healthy male subjects. Euro-
pean Spine Journal 2008;17(1):122e8.
Novy DM, Simmonds MJ, Olson SL, Lee CE, Jones SC. Physical performance: differ-
ences in men and women with and without low back pain. Archives of Physical
Medicine and Rehabilitation 1999;80(2):195e8.
O’Sullivan PB. Lumbar segmental ‘instability’: clinical presentation and specific
stabilizing exercise management. Manual Therapy 2000;5(1):2e12.
O’Sullivan PB. Diagnosis and classification of chronic low back pain disorders:
maladaptive movement and motor control impairments as underlying mech-
anism. Manual Therapy 2005;10(4):242e55.
 J.M. Williams et al. / Manual Therapy 18 (2013) 130e135
Pal P, Milosavljevic S, Sole G, Johnson G. Hip and lumbar continuous motion char-
acteristics during flexion and return in young healthy males. European Spine
Journal 2007;16(6):741e7.
Parks KA, Crichton KS, Goldford RJ, McGill SM. A comparison of lumbar range of
motion and functional ability scores in patients with low back pain: assessment
for range of motion validity. Spine 2003;28(4):380e4.
Powers CM, Beneck GJ, Kulig K, Landel RF, Fredericson M. Effects of a single session
of posterior-to-anterior spinal mobilization and press-up exercise on pain
response and lumbar spine extension in people with nonspecific low back pain.
Physical Therapy 2008;88(4):485e93.
Saber-Sheikh K, Bryant EC, Glazzard C, Hamel A, Lee RYW. Feasibility of
using inertial sensors to assess human movement. Manual Therapy 2010;15(1):
122e5.
Saur PMM, Ensink F-BM, Frese K, Seeger D, Hildebrandt J. Lumbar range of motion:
reliability and validity of the inclinometer technique in the clinical measure-
ment of trunk flexibility. Spine 1996;21(11):1332e8.
Shum GLK, Crosbie J, Lee RYW. Effect of low back pain on the kinematics and joint
coordination of the lumbar spine and hip during sit-to-stand and stand-to-sit.
Spine 2005a;30(17):1998e2004.
Shum GLK, Crosbie J, Lee RYW. Symptomatic and asymptomatic movement coor-
dination of the lumbar spine and hip during an everyday activity. Spine 2005b;
30(23):E697e702.
135
Shum GLK, Crosbie J, Lee RYW. Movement coordination of the lumbar spine and hip
during a picking up activity in low back pain subjects. European Spine Journal
2007a;16(6):749e58.
Shum GLK, Crosbie J, Lee RYW. Three-dimensional kinetics of the lumbar spine and
hips in low back pain patients during sit-to-stand and stand-to-sit. Spine
2007b;32(7):E211e9.
Theobald PS, Jones MD, Williams JM. Do inertial sensors represent a viable method to
reliably measure cervical spine range of motion? Manual Therapy 2012;17(1):92e6.
Tong K, Granat MH. A practical gait analysis system using gyroscopes. Medical
Engineering and Physics 1999;21(2):87e94.
Tsang S, Szeto G, Lee R. Movement kinematics and muscular activation patterns of
the cervicothoracic spine of people with chronic neck pain. In: Paper presented
at the International Society of Biomechanics Congress, Brussels, Belgium; 2011.
Van Dillen LR, Maluf KS, Sahrmann SA. Further examination of modifying patient-
preferred movement and alignment strategies in patients with low back pain
during symptomatic tasks. Manual Therapy 2009;14(1):52e60.
Williams JM, Haq I, Lee RYW. Dynamic measurement of lumbar curvature using
fibre-optic sensors. Medical Engineering and Physics 2010;32(9):1043e9.
Xu Y, Choi J, Reeves NP, Cholewicki J. Optimal control of the spine system. Journal of
Biomechanical Engineering 2010;132(5):051004.
Zhou H, Stone T, Hu H, Harris N. Use of multiple wearable inertial sensors in upper
limb motion tracking. Medical Engineering and Physics 2008;30(1):123e33.