SPINE Volume 34, Number 6, pp E208–E214
©2009, Lippincott Williams & Wilkins
Different Ways to Balance the Spine
Subtle Changes in Sagittal Spinal Curves Affect Regional
Andrew P. Claus, MPhty,* Julie A. Hides, PhD,* G. Lorimer Moseley, PhD,†‡
and Paul W. Hodges, PhD*
Study Design. Exploratory study of regional muscle
activity in different postures.
Objective. To detail the relationship between spinal
curves and regional muscle activity.
Summary of Background Data. Sagittal balanced spi-
nal posture (C7 above S1 in the sagittal plane) is a goal for
spinal surgery and conservative ergonomics. Three com-
binations of thoracolumbar and lumbar spinal curves can
be considered sagittal balanced postures: (i) flat—at both
regions, (ii) long lordosis—lordotic at both regions, and
(iii) short lordosis—thoracic kyphosis and lumbar lordo-
sis. This study compares regional muscle activity be-
tween these 3 sagittal balanced postures in sitting, as well
as a slump posture.
Methods. Fine-wire electromyography (EMG) elec-
trodes were inserted into the lumbar multifidus (deep and
superficial), iliocostalis (lateral and medial), longissimus
thoracis, and transversus abdominis in 14 healthy male
volunteers. Fine-wire or surface EMG electrodes were
also used to record activity of the obliquus internus,
obliquus externus, and rectus abdominis muscles. Root
mean square EMG amplitude in the flat, long lordosis,
short lordosis, and slump sitting postures were normal-
ized to maximal voluntary contraction, and also to the
peak activity across the sitting postures. Muscle activity
was compared between postures with a linear mixed
Results. Of the extensor muscles, it was most notable
that activity of the deep and superficial fibers of lumbar
multifidus increased incrementally in the 3 sagittal bal-
anced postures; flat, long lordosis, and short lordosis
(P ? 0.05). Of the abdominal muscles, obliquus internus
was more active in short lordosis than the other postures
(P ? 0.05). Comparing the sagittal balanced postures, the
flat posture showed the least muscle activity (similar to
the slump posture at most muscles examined).
Conclusion. Discrete combinations of muscle activity
supported the 3 different sagittal balanced postures in
sitting, providing new detail for surgeons, researchers,
and therapists to distinguish between different sagittal
Key words: lumbar spine, sitting, multifidus, extensor
muscles, fine-wire electromyography. Spine 2009;34:
“Neutral upright sagittal spinal alignment” (sagittal bal-
ance) is a postural goal for surgical interventions and
conservative ergonomic training, but the definition of
“neutral upright sagittal spinal alignment” is vague. A
wide variety of thoracic and lumbar spinal curves satisfy
a sagittal balance criterion of C7–S1 sagittal deviation
?50 mm (asymptomatic adults in standing),1,2making it
difficult for surgeons, researchers, therapists, and patients
to know if they are examining or achieving the same pos-
tural goal. In neutral upright spinal alignment, the spine’s
propensity to bend, twist, and shear is managed by neuro-
muscular control.3–6Despite the importance of neuromus-
cular control to coordinate and protect the spine, little is
known of how thoracic and lumbar curves in sagittally
balanced postures influence regional muscle activity.
Electromyography (EMG) studies have shown that
upright spinal postures in sitting require more extensor
muscle activity than kyphotic postures (slumped).7,8A
study of upright spinal posture in standing with surface
electrodes over paraspinal muscles from C4–S2 showed
large intersubject variation in activity between regions of
the spinal extensor muscles,9but unfortunately the spi-
nal curves that subjects used were not reported. A de-
scription of spinal curves that people commonly adopt in
standing has been provided by a recent radiographic
study of 160 asymptomatic adults in standing.10The
authors suggested that balanced sagittal spinal curves
could be grouped into 4 different postures: (1) a kyphotic
thoracolumbar region with only the lower lumbar seg-
ments in a lordotic curve, (2) hypokyphotic thoracic and
hypolordotic lumbar regions (flat), (3) Inflection of
curves at the thoracolumbar region with a lordotic lum-
bar curve (short lordosis), and (4) lordotic curves at tho-
racolumbar and lumbar regions (long lordosis).10Few
studies have compared muscle activity between these
postures, but a recent surface EMG study11examined 2
of the spinal curve combinations. One posture sought to
From the *Centre of Clinical Research Excellence in Spinal Pain, Injury
and Health, The University of Queensland, School of Health and Re-
habilitation Sciences, Brisbane, Australia; †Department of Physiology,
Anatomy, and Genetics, University of Oxford, Oxford, United Kingdom;
and ‡Prince of Wales Medical Research Institute and School of Medical
Sciences, The University of New South Wales, Sydney, Australia.
Acknowledgment date: June 9, 2008. Revision date: August 26, 2008.
Acceptance date: September 8, 2008.
The manuscript submitted does not contain information about medical
this work. No benefits in any form have been or will be received from
a commercial party related directly or indirectly to the subject of this
Supported by the National Health and Medical Research Council of
University of Oxford (to G.L.M.). Reimbursement of study partici-
pants was funded by the The Dorothy Hopkins Award.
All procedures were approved by the University of Queensland Medi-
cal Research Ethics Committee, and written informed consent was
obtained from all research subjects.
Division of Physiotherapy, School of Health and Rehabilitation Sci-
tralia; E-mail: firstname.lastname@example.org
achieve lordotic curves at thoracolumbar and lumbar
lumbar region was in a lordotic curve (short lordosis).
The results showed differential muscle activity at elec-
trodes adjacent to T9 and L5, as well as obliquus inter-
nus and externus.11Unfortunately, surface EMG has
limited ability to distinguish between sources of the elec-
trical signal recorded, so the contribution of adjacent but
plexity of spinal extensor muscle anatomy12,13and the
potential for the abdominal muscles to contribute to spi-
nal support,14–16more detailed examination of regional
muscle activity associated with different spinal curves in
sagittally balanced postures is warranted.
ferentiate EMG signals between muscles and between
superficial and deep fascicles within muscles. Fine-wire
EMG data obtained during trunk rotation in asymptom-
atic subjects have shown discrete muscle activity in re-
gions of the longissimus thoracis and thoracic multifi-
dus,17as well as iliocostalis and lumbar multifidus
muscles.18Among the spinal extensor muscles, the lum-
bar multifidus show unique deterioration with back
pain,49spinal trauma48and surgical retraction,19and
deep fibers of multifidus are uniquely coordinated in
preparation for movement.20,21Furthermore, rehabilita-
tion of multifidus function plays a role in preventing
recurrence of low back pain after an initial episode,22
and may have significance for postsurgical recovery.23,24
Fine-wire electrodes are also the only tool to record ac-
tivity of deep abdominal muscles such as transversus ab-
dominis. Among the abdominal muscles, the transversus
abdominis is uniquely coordinated in preparation for
movement,25,26and this function deteriorates with back
pain.27–29The objective of this study was to examine
activity at 9 regions of the paraspinal and abdominal
muscles using fine-wire and surface EMG, comparing
activity between 3 sagittally balanced postures (flat, long
lordosis, short lordosis) and a slump posture in sitting.
Materials and Methods
Fourteen healthy men with a mean (SD) age of 22 (8) years,
height of 178 (8) cm, and weight of 71 (10) kg participated in
this study. Subjects were excluded if they had ever experienced
thoracic or lumbar spinal pain that required treatment, or rest
from normal activities for more than 2 days, or if they had a
musculoskeletal physiotherapist undertook a physical exami-
nation to determine that participants had no abnormal restric-
tion of hip mobility, spinal mobility, or evidence of a scoliosis
that would limit symmetrical performance of sitting postures.
Written informed consent was obtained and all procedures
were approved by the institutional Medical Research Ethics
Postures and Measurement
Sitting postures were examined in preference to standing be-
cause sitting allowed greater control of pelvis position and pos-
tural sway through repeated trials in lumbar postures ranging
flat—minimal curve at thoracolumbar and lumbar regions;
long lordosis—lordotic curve at thoracolumbar and lumbar
regions; short lordosis—flat/kyphotic thoracolumbar with a
and lumbar regions (Figure 1).
To quantify thoracolumbar and lumbar spinal curves in the
sagittal plane, 3-dimensional surface tracking was used. Posi-
tracking system (Ascension, sensor static position absolute er-
ror was 1.8 mm, position data recorded before 15 seconds
ing); and data for 3 subjects were recorded with an optical
Figure 1. I, Thoracolumbar (T5–T10-L3) and lumbar angle (T10-L3-
S2) curve directions are shown for the sitting postures examined:
flat—flat at both regions, long lordosis—lordotic at both regions,
short lordosis—kyphotic/flat thoracolumbar and lordotic lumbar
regions, slump—kyphotic at both regions. II, Thoracolumbar and
lumbar spinal angles means (range) for each posture. III, Anterior
displacement of T1-S2 means (range) for each posture.
E209Spinal Curves and Muscle Activity•Claus et al
tracking system with (Vicon, reflector position absolute error
was 0.1 mm, position data and EMG recorded simultaneously)
using Nexus software (Vicon). Relative positions of surface
markers have been validated with radiography for relative po-
sitions of vertebrae, with a standard error of 4 mm at the
thoracolumbar region and 2 mm at the lumbar region.30Lum-
bar surface markers have also been shown to accurately repre-
sent radiographic measures for change in vertebral flexion/
extension.31The boundary between thoracic and lumbar
curves was defined as being located at the T10 vertebral seg-
ment for 2 reasons. Facet joint orientation32and spinal curves
in standing10can transition as proximally as T10. Sagittal an-
gles representing surface spinal curves at thoracolumbar and
lumbar regions of the spine were measured between segments
connecting T5–T10 and T10-L3 (thoracolumbar angle),
T10-L3 and L3-S2 (lumbar angle) (Figure 1),33with kyphotic
angles described as positive, and lordotic angles as negative.
Global sagittal balance was also measured with anterior dis-
placement of T1 relative to S2, which was similar for the 3
upright sitting postures (Figure 1).
Measures of spinal curve are continuous variables, but to
differentiate postures and categorize angles as lordotic, flat, or
kyphotic it was necessary to impose discreet limits for these
angles. The flat posture was used as a reference. When the 14
subjects performed 3 trials of the flat posture, the mean (range)
thoracolumbar angle was 2.3° (?4.9 to 5.5), and the mean
lumbar angle mean was ?0.2° (?5.5 to 4.4). On this basis,
angles ?3.0° to 7.0°, and lumbar angles ?5.0° to 5.0°. Greater
angles were categorized as kyphotic spinal curves, and lesser
angles as lordotic spinal curves. Figure 1 shows angle data in
each posture for all trials included in the study.
muscles was recorded with bipolar fine-wire electrodes fabri-
cated from Teflon-coated stainless steel wire (75-?m diameter)
or surface electrodes (Table 1). The fine-wire electrodes had 1
mm of Teflon removed from the cut ends, were bent back to
form hooks at ?1 and 2.5 mm from the ends, threaded into a
hypodermic needle, and sterilized. With the subjects positioned
abdominal muscles was imaged sonographically (8–12 MHz
transducer, GE Logic 9), and hypodermic needles (0.50 ? 70
mm or 0.50 ? 32 mm) with fine-wire electrodes were inserted
into the extensor and abdominal muscles on the subject’s right
side (Figure 2; Table 1). For the superficial abdominal muscles,
subjects were offered either fine-wire or surface electrodes.
Pairs of surface electrodes (Ag/AgCl discs, 10 mm in diameter;
either Cleartrace, Conmed, NY, or Red Dot, 3 M Health Care
Products, London, Canada were used) were placed in parallel
with fibers of the superficial abdominal muscles (20 mm inter-
electrode distance). A ground electrode was placed over the
right iliac crest.
EMG data were amplified 2000 ? (Neurolog Digitimer,
UK) and data from 9 subjects were sampled at 2000 Hz
(Spike2.6, CED, Cambridge, UK). Due to error in data collec-
of the EMG spectrum lies above 100 Hz, careful evaluation of
the EMG amplitude showed that between postures the 100 Hz
data changed in a similar manner to the 2000 Hz data, so the
100 Hz data were retained in the analysis.
To achieve the required spinal curves, subjects sat on a stool
adjusted to the height of their popliteal crease. They were then
shown pictures of each posture (similar to Figure 1) with verbal
description of the spinal curves required. For the flat, long
lordosis and short lordosis postures, manual guidance and
feedback pertaining to the pelvis position and spinal curves
were provided. For the long and short lordosis postures, par-
ticipants were taught to tilt the upper aspect of the sacrum
Table 1. Electrode Placement
Deep multifidus at L4
Origin—insertion of relevant fibres
Laminae, mamillary processes and
facet joint capsules—2
vertebral segments inferiorly
Upper lumbar segments spinous
processes—sacrum and ilia
Fine-wire electrodes placement
?30 mm lateral to spinous process,
and directed medially to contact the
L4 lamina21n ? 12
?30 mm lateral to the spinous process;
?10 mm below the skin surface21
n ? 13
?80 mm lateral to L2 spinous process
(previously examined with sEMG54)
n ? 13
?60 mm lateral to T11 spinous process,
and lateral to the transverse
process18n ? 14
?40 mm lateral to T11 spinous process;
directed toward the dorsal aspect of
the transverse process17n ? 14
Between the anterior superior iliac
spine and the ribcage55n ? 13
Between the anterior superior iliac
spine and the ribcage55n ? 7
Surface electrodes placement
Superficial multifidus at L4
Iliocostalis L2 (lateral)Ribs 11 and 12—iliac crest/lumbo-
Iliocostalis T11 (medial) Ribs 6–9—iliac crest/lumbo-dorsal
Longissimus thoracis Transverse processes of mid-
thoracic vertebral segments—
Lumbodorsal fascia—linea albaTransversus abdominis
Obliquus internusIliac crest—aponeurosis to linea
Transverse arrangement ?30 mm
medial to the anterior superior
iliac spine n ? 7
Lateral electrode proximal to
medial ?45 deg to transverse
plane between the iliac crest
and the ribcage n ? 7
Sagittal arrangement ?20 mm
infero-lateral to the umbilicus
n ? 9
Obliquus externus Ribs 9–12—iliac crest and inguinal
Between the anterior superior iliac
spine and the ribcage55n ? 7
Rectus abdominis Pubis—costal cartilage of ribs 5–7
and xiphoid process
?20 mm infero-lateral to the umbilicus
n ? 5
E210Spine•Volume 34•Number 6•2009
forward, sitting toward the front of their ischia and perineum,
similar to the pelvis position on a bicycle saddle.33Sitting pos-
tures were performed in random order for three 45-second
and generally face forward during trials. Between trials, sub-
jects stood briefly to minimize the effects of fatigue or task
A series of 3 maximal voluntary contractions (MVC) for 3
seconds against manual resistance were recorded.34In supine
lying, subjects flexed the trunk to recruit the rectus abdominis
muscle, and rotated the trunk to the left for the obliquus exter-
nus muscle. In sitting, subjects performed a maximal forced
expiratory maneuver for the obliquus internus and transversus
abdominis muscles. In prone lying, subjects extended the trunk
to recruit the 5 spinal extensor muscles that were recorded.
Activities at rest, in supine and in prone, were also recorded to
determine the amplitude of baseline activity/noise.
Spinal curve measures from 3-dimensional tracking data were
exported and analyzed (Matlab 6, Mathworks) for the 3 trials
in each posture. Trials in which spinal angles failed to achieve
the required curve directions (Figure 1) were excluded, leaving
data for 12 subjects for the flat and long lordosis postures, 13
subjects for short lordosis, and 10 subjects in the slump posture
(slump data were excluded from 4 subjects who achieved a
EMG data were exported, a 5-second sample was selected
from each trial to avoid artifact in the EMG traces (assessor
blinded to posture), and root mean square amplitude was cal-
culated (Matlab 6, Mathworks) for each trial. Baseline root
mean square EMG amplitude at rest was subtracted from all
trials (sitting posture and the MVC trials), and 2 separate nor-
malization procedures were used. To indicate the absolute am-
plitude of EMG activity in the sitting postures and data were
expressed as a percentage of MVC. However, normalization of
data in comparison with normalization to a submaximal
task.35Hence, data were also expressed as a percentage of the
peak activity recorded in sitting.
EMG amplitude of data normalized to peak activity recorded
in sitting for each muscle were compared between postures
with a linear mixed model analysis (SPSS version 15, IL). Test
was significant (P ? 0.001), so for each of the 9 muscles, pair-
wise comparisons of EMG amplitude between postures were
undertaken with Bonferroni adjustment for multiple compari-
sons. The level of significance was set at ? ?0.05.
Data normalized to MVC for each muscle and posture
(mean and 95% CI) are shown in Table 2. In general, the
deep and superficial fibers of multifidus and the longissi-
mus thoracis muscle were the most active, with ?10%
MVC in one of the postures. The Iliocostalis at T11 and
at L2, transversus abdominis, obliquus internus, and
obliquus externus muscles showed activity levels up to
3% to 4% MVC, and rectus abdominis was active at
?1% MVC in the sitting postures.
Comparisons between postures are shown in Figure 3
for extensor muscles and Figure 4 for abdominal mus-
cles, with EMG amplitude normalized to peak activity in
sitting. For the deep and superficial fibers of the multifi-
dus muscle, activity in the flat and slump postures were
similar (P ? 1.00), but activity was greater in the long
lordosis and greatest in the short lordosis posture (all
comparisons P ? 0.05). For iliocostalis at L2, activity in
the short lordosis posture was greater than flat and
slump postures (P ? 0.05), but other comparisons were
not different. The longissimus thoracis muscle was least
active in the slump posture (P ? 0.05), but other com-
parisons were not different. For the obliquus internus
muscle, the flat and long lordosis postures were similar
(P ? 1.00), activity was least in slump, and activity was
Figure 2. Intramuscular EMG
electrodes were inserted into
deep and superficial muscle fi-
bers of multifidus at L4, iliocos-
talis adjacent to L2 and T11, lon-
gissimus at T11, and transversus
abdominis on the right side.
Obliquus internus, obliquus ex-
ternus, and rectus abdominis
muscle activity were measured
with either surface EMG or intra-
Table 2. EMG Results Normalized to MVC
Mean % MVC (95% CI) in Each Posture
Iliocostalis L2 (lateral)
Iliocostalis T11 (medial)
E211Spinal Curves and Muscle Activity•Claus et al
0.05). The transversus abdominis and obliquus externus
muscles were less active in the slump than the long lor-
dosis and short lordosis postures (P ? 0.05), but other
comparisons were not different. The EMG amplitude of
iliocostalis adjacent to T11 and rectus abdominis mus-
cles did not differ between postures.
This study provides the most specific measurement of
surface spinal curves and associated regional muscle ac-
tivity to date. The results show that thoracolumbar and
lumbar spinal curves in sagittally balanced postures are
activity. Notably, activity of the deep and the superficial
3 sagittally balanced postures; flat, long lordosis, and
short lordosis. The activity of the iliocostalis muscle ad-
jacent to L2 was greater in the short lordosis than in the
flat posture, but activity of the iliocostalis adjacent to
T11 and of the longissimus thoracis muscles did not dif-
fer between the sagittally balanced postures. Consistent
with the anatomy of the lumbar multifidus muscle (fas-
cicles crossing 2–5 motion segments12), the results dem-
onstrate that the lumbar multifidus has a unique role
and/or support of a lumbar lordosis.
To speculate on the role of lumbar multifidus (and
other intrinsic muscles) to mediate spinal curves, origins
and insertions of the deep-medial and superficial-lateral
fascicles12bear obvious resemblance to the attachment
sites for screws and wires to secure posterior internal
fixators. Intrinsic muscular support might act segmen-
tally to mediate bending, twisting, and shearing of a mo-
tion segment,36,37or regionally to mediate distribution
Figure 3. Extensor muscle EMG normalized to peak activity in
sitting. *P ? 0.05; error bars—95% confidence intervals. Highest
maximal voluntary contractions normalization results from Table 1
are shown to indicate magnitude of peak activity.
Figure 4. Abdominal muscle EMG normalized to peak activity in
sitting. *Significant difference at P ? 0.05; error bars—95% con-
fidence intervals. Highest maximal voluntary contractions normal-
ization results from Table 1 are shown to indicate magnitude of
E212Spine•Volume 34•Number 6•2009
of movement among a chain of mobile segments. The
incremental change in multifidus activity shown to occur
with change in thoracolumbar and lumbar curves may
represent a combination of these segmental and regional
functions of the multifidus muscles.
EMG was that obliquus internus was more active in
short lordosis than in the other postures. Surface elec-
trodes were used in 7 subjects (activity from obliquus
internus and transversus abdominis) and the lower re-
activity of transversus abdominis.34Thus, the contribu-
tion of transversus abdominis cannot be excluded. Mus-
cle activity ?5% MVC at obliquus internus, obliquus
externus, transversus abdominis, and iliocostalis adja-
cent to L2 suggests that these muscles play a modest role
in supporting the long lordosis and short lordosis pos-
tures more than slump.
A key reason for maintaining C7 over S1 is to mini-
mize muscular effort to support the spine.38–40Sagittal
imbalance in standing, with anterior displacement of C7
relative to S2 ?50 mm occurs in kyphotic patients,2,38
scoliotic patients, and can occur in asymptomatic popu-
lations,41especially with increasing age.42If the objec-
tive of an efficient posture is to balance the spine with
the flat posture was midway between the lumbar curves
in lordotic and slump postures. Although surface track-
ing accurately represents changes in lumbar flexion and
extension, there is evidence that radiographic measures
show larger angles of lordosis.31Hence, flat surface mea-
sures may involve a small degree of lordosis at the verte-
bral bodies. Furthermore, the flat posture demonstrated
the lowest muscle activity of the 3 upright postures, with
only longissimus thoracis and obliquus internus muscles
more active than in the slump posture. Provided that
lumbar extension range is available for standing and
gait, the flat posture may satisfy clinical descriptions of
sagittal balance and provide the most efficient mechani-
cal balance for the spine in sitting.
In contrast with flat, the short lordosis posture dem-
onstrated the highest levels of muscle activity, particu-
larly at the lumbar multifidus, with 16.8% (7.9) MVC.
This raises an important question, can people sustain the
% MVC observed in the short lordosis sitting posture?
Evidence to quantify endurance at specific extensor mus-
cles could not be found, but 5% MVC of the spinal
extensor muscles (surface EMG) can be sustained for 30
minutes or more.43At other regions of the body 10%
MVC is sustainable for an hour at triceps surae44and
elbow flexors.45Endurance at 20% MVC varies from 1
to 30 minutes at the elbow flexors45,46and cranio-
cervical flexor muscles.47Based on those data, sustained
lordosis in sitting postures may exceed the endurance
capacity of the lumbar multifidus muscle. For people
whose function of the lumbar multifidus is compromised
after spinal pain, injury, or surgery,19,48,49the long lor-
dosis and short lordosis postures could be even harder to
sustain. To sustain lordotic spinal curves in sitting might
be an unreasonable stress on the lumbar multifidus mus-
cles, contributing to fatigue or pain. Alternatively, lor-
dotic sitting postures might maintain or aid rehabilita-
tion of the lumbar multifidus muscles. The advantages
and disadvantages of specific sagittally balanced pos-
tures warrant investigation.
The current results should be considered in light of
several factors. First, this was a study of a common,
functional, and commonly problematic task, sitting
without backrest support. It is uncertain whether the
muscle activity results would be comparable in standing
postures. Second, individuals with inherently kyphotic
or lordotic lumbar curves (research participants were
screened for ability to adopt all required spinal curves)
may require greater or lesser activity than our partici-
pants showed in the lordotic postures. Third, this was a
study of “global sagittal balance” of the spine, as op-
posed to “segmental sagittal balance,” which would re-
quire radiographic measures of segmental positions.50
Fourth, aspects beyond the scope of this study include
the activity of other paraspinal muscles (such as quadra-
tus lumborum and psoas major) and gender differences.51
These limitations aside, the results of this study distin-
guish 3 sagittally balanced postures by spinal curves and
associated muscle activity, provide new detail of multifi-
dus muscle function, and a basis to examine whether
particular “neutral upright sagittal spinal alignment”
may be advantageous.
The results provide new detail to distinguish between
sagittally balanced postures. Subtle changes in thoraco-
lumbar and lumbar spinal curves in sitting are associated
with varying magnitudes of muscle activity, particularly
in the deep and superficial regions of lumbar multifidus.
The highest activity levels at multifidus and obliquus in-
ternus abdominis muscles occurred in the short lordosis
posture. The lowest activity levels were observed at most
muscles in the flat posture.
● EMG of 9 spinal extensor and abdominal mus-
cles was used to compare activity in 3 different sag-
ittal balanced sitting postures.
● Activity at the lumbar multifidus muscles ad-
justed with subtle changes of lumbar and thoraco-
lumbar spinal curves.
● The results provide basis to distinguish between
sagittal balanced spinal postures.
The authors thank Linda-Joy Lee, David MacDonald,
Bo ¨rje Rehn, and Ross Darnell for assistance in planning
methods, data collection, and analysis for this study; and
Dorothy Hopkins award for financial assistance with re-
imbursing research subjects.
E213Spinal Curves and Muscle Activity•Claus et al
from the occiput to the pelvis in asymptomatic adults: a review and resyn-
thesis of the literature. J Neurosurg Spine 2007;6:104–12.
2. Booth KC, Bridwell KH, Lenke LG, et al. Complications and predictive
factors for the successful treatment of flatback deformity (fixed sagittal im-
balance). Spine 1999;24:1712–20.
3. Morris JM, Lucas DB, Bresler B. Role of the trunk in stability of the spine.
J Bone Joint Surg 1961;43:327–51.
4. Crisco JJ, Panjabi MM. Euler stability of the human ligamentous lumbar
spine. Part II: experiment. Clin Biomech (Bristol, Avon) 1992;7:27–32.
5. Cholewicki J, McGill SM. Mechanical stability of the in vivo lumbar spine:
implications for injury and chronic low back pain. Clin Biomech (Bristol,
6. Adams MA, Dolan P. Recent advances in lumbar spinal mechanics and their
clinical significance. Clin Biomech (Bristol, Avon) 1995;10:3–19.
7. Floyd WF, Silver PHS. The function of the erector spinae muscle in certain
movements and postures in man. J Physiol 1955;129:184–203.
8. O’Sullivan PB, Grahamslaw KM, Kendell M, et al. The effect of different
standing and sitting postures on trunk muscle activity in a pain-free popula-
tion. Spine 2002;27:1238–44.
9. Joseph J, McColl I. Electromyography of muscles of posture: posterior ver-
tebral muscles in males. J Physiol 1961;157:33–7.
10. Roussouly P, Gollogly S, Berthonnaud E, et al. Classification of the normal
variation in the sagittal alignment of the human lumbar spine and pelvis in
the standing position. Spine 2005;30:346–53.
11. O’Sullivan PB, Dankaerts W, Burnett AF, et al. Effect of different upright
sitting postures on spinal-pelvic curvature and trunk muscle activation in a
pain-free population. Spine 2006;31:E707–12.
12. Bogduk N, Macintosh JE, Pearcy MJ. A universal model of the lumbar back
muscles in the upright position. Spine 1992;17:897–913.
13. Kalimo H, Rantanen J, Viljanen T, et al. Lumbar muscles: structure and
function. Ann Med 1989;21:353–9.
14. Hodges PW. Is there a role for transversus abdominis in lumbo-pelvic sta-
bility? Man Ther 1999;4:74–86.
15. Hodges PW, Gandevia SC. Changes in intra-abdominal pressure during pos-
tural and respiratory activation of the human diaphragm. J Appl Physiol
16. Hodges PW, Eriksson AE, Shirley D, et al. Intra-abdominal pressure in-
creases stiffness of the lumbar spine. J Biomech 2005;38:1873–80.
17. Lee LJ, Coppieters MW, Hodges PW. Differential activation of the thoracic
multifidus and longissimus thoracis during trunk rotation. Spine 2005;30:
muscles of the back in man. J Anat 1962;96:509–20.
19. Taylor H, McGregor AH, Medhi-Zadeh S, et al. The impact of self-retaining
retractors on the paraspinal muscles during posterior spinal surgery. Spine
20. Moseley GL, Hodges PW, Gandevia SC. External perturbation of the trunk
in standing humans differentially activates components of the medial back
muscles. J Physiol 2003;547:581–7.
21. Moseley GL, Hodges PW, Gandevia SC. Deep and superficial fibers of the
lumbar multifidus muscle are differentially active during voluntary arm
movements. Spine 2002;27:E29–36.
22. Hides JA, Jull GA, Richardson CA. Long-term effects of specific stabilizing
exercises for first-episode low back pain. Spine 2001;26:E243–8.
23. Datta G, Gnanalingham KK, Peterson D, et al. Back pain and disability after
lumbar laminectomy: is there a relationship to muscle retraction? Neurosur-
gery 2004;54:1413–20; discussion 20.
after surgery for a lumbar intervertebral disc herniation. Spine 1993;18:
25. Hodges PW, Richardson CA. Feedforward contraction of transversus abdo-
minis is not influenced by the direction of arm movement. Exp Brain Res
26. Hodges PW, Richardson CA. Transversus abdominis and the superficial
abdominal muscles are controlled independently in a postural task. Neurosci
27. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar
spine associated with low back pain. A motor control evaluation of trans-
versus abdominis. Spine 1996;21:2640–50.
28. Hodges PW, Moseley GL, Gabrielsson A, et al. Experimental muscle pain
changes feedforward postural responses of the trunk muscles. Exp Brain Res
29. Hodges PW, Richardson CA. Delayed postural contraction of transversus
abdominis in low back pain associated with movement of the lower limb.
J Spinal Disord 1998;11:46–56.
30. Bryant JT, Reid JG, Smith BL, et al. Method for determining vertebral body
positions in the sagittal plane using skin markers. Spine 1989;14:258–65.
during flexion/extension and free lifts. Spine 1990;15:1333–9.
32. Singer KP, Edmondston SJ, Day RE, et al. Computer-assisted curvature as-
sessment and Cobb angle determination of the thoracic kyphosis. An in vivo
and in vitro comparison. Spine 1994;19:1381–4.
33. Claus A, Hides J, Moseley GL, et al. Is ideal sitting posture real? Measure-
ment of spinal curves in four sitting postures. Man Ther. In press.
34. Hodges P, Cresswell A, Thorstensson A. Preparatory trunk muscle motion
accompanies rapid upper limb movement. Exp Brain Res 1999;124:69–79.
35. Allison GT, Godfrey P, Robinson G. EMG signal amplitude assessment
during abdominal bracing and hollowing. J Electromyogr Kinesiol 1998;8:
36. Panjabi MM. The stabilizing system of the spine. Part I. Function, dysfunc-
tion, adaptation, and enhancement. J Spinal Disord 1992;5:383–9.
37. Panjabi MM. The stabilizing system of the spine. Part II. Neutral zone and
instability hypothesis. J Spinal Disord 1992;5:390–6.
38. Farcy JP, Schwab FJ. Management of flatback and related kyphotic decom-
pensation syndromes. Spine 1997;22:2452–7.
39. Kim YJ, Bridwell KH, Lenke LG, et al. An analysis of sagittal spinal align-
ment following long adult lumbar instrumentation and fusion to L5 or S1:
can we predict ideal lumbar lordosis? Spine 2006;31:2343–52.
and metabolic energy expenditure during standing. Spine 2007;32:1605–11.
41. El Fegoun AB, Schwab F, Gamez L, et al. Center of gravity and radiographic
posture analysis: a preliminary review of adult volunteers and adult patients
affected by scoliosis. Spine 2005;30:1535–40.
42. Schwab F, Lafage V, Boyce R, et al. Gravity line analysis in adult volunteers:
age-related correlation with spinal parameters, pelvic parameters, and foot
position. Spine 2006;31:E959–67.
43. van Dieen JH, Westebring-van der Putten EP, Kingma I, et al. Low-level
activity of the trunk extensor muscles causes electromyographic manifes-
tations of fatigue in absence of decreased oxygenation. J Electromyogr
Kinesiol. In press.
activation: within- and between-synergist substitution in the triceps surae
muscles. Eur J Appl Physiol 2004;91:204–16.
45. Bjorksten M, Jonsson B. Endurance limit of force in long-term intermittent
static contractions. Scand J Work Environ Health 1977;3:23–7.
46. Caldwell LS, Smith RP. Pain and endurance of isometric muscle contrac-
tions. J Eng Psychol 1966;5:25–32.
47. O’Leary S, Jull G, Kim M, et al. Cranio-cervical flexor muscle impairment at
maximal, moderate, and low loads is a feature of neck pain. Man Ther
48. Hodges P, KaigleHolm A, Hansson T, et al. Rapid atrophy of the lumbar
multifidus follows experimental disc or nerve root injury. Spine 2006;31:
49. Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not auto-
matic after resolution of acute, first-episode low back pain. Spine 1996;21:
50. Angevine PD, Bridwell KH. Sagittal imbalance. Neurosurg Clin N Am 2006;
51. Vialle R, Levassor N, Rillardon L, et al. Radiographic analysis of the sagittal
alignment and balance of the spine in asymptomatic subjects. J Bone Joint
Surg Am 2005;87:260–7.
E214 Spine•Volume 34•Number 6•2009