Influence of spine morphology on intervertebral disc loads and stresses in asymptomatic adults: Implications for the ideal spine

Abstract

Sagittal profiles of the spine have been hypothesized to influence spinal coupling and loads on spinal tissues.
To assess the relationship between thoracolumbar spine sagittal morphology and intervertebral disc loads and stresses.
A cross-sectional study evaluating sagittal X-ray geometry and postural loading in asymptomatic men and women.
Sixty-seven young and asymptomatic subjects (chiropractic students) formed the study group.
Morphological data derived from radiographs (anatomic angles and sagittal balance parameters) and biomechanical parameters (intervertebral disc loads and stresses) derived from a postural loading model.
An anatomically accurate, sagittal plane, upright posture, quadrilateral element model of the anterior spinal column (C2-S1) was created by digitizing lateral full-spine X-rays of 67 human subjects (51 males, 16 females). Morphological measurements of sagittal curvature and balance were compared with intervertebral disc loads and stresses obtained using a quadrilateral element postural loading model.
In this young (mean 26.7, SD 4.8 years), asymptomatic male and female population, the neutral posture spine was characterized by an average thoracic angle (T1-T12) = +43.7 degrees (SD 11.4 degrees ), lumbar angle (T12-S1) = -63.2 degrees (SD 10.0 degrees ), and pelvic angle = +49.4 degrees (SD 9.9 degrees ). Sagittal curvatures exhibited relatively broad frequency distributions, with the pelvic angle showing the least variance and the thoracic angle showing the greatest variance. Sagittal balance parameters, C7-S1 and T1-T12, showed the best average vertical alignment (5.3 mm and -0.04 mm, respectively). Anterior and posterior disc postural loads were balanced at T8-T9 and showed the greatest difference at L5-S1. Disc compressive stresses were greatest in the mid-thoracic region of the spine, whereas shear stresses were highest at L5-S1. Significant linear correlations (p < .001) were found between a number of biomechanical and morphological parameters. Notably, thoracic shear stresses and compressive stresses were correlated to T1-T12 and T4-hip axis (HA) sagittal balance, respectively, but not to sagittal angles. Lumbar shear stresses and body weight (BW) normalized shear loads were correlated with T12-S1 balance, lumbar angle, and sacral angle. BW normalized lumbar compressive loads were correlated with T12-S1 balance and sacral angle. BW normalized lumbar disc shear (compressive) loads increased (decreased) significantly with decreasing lumbar lordosis. Cervical compressive stresses and loads were correlated with all sagittal balance parameters except S1-HA and T12-S1. A neutral spine sagittal model was constructed from the 67 subjects.
The analyses suggest that sagittal spine balance and curvature are important parameters for postural load balance in healthy male and female subjects. Morphological predictors of altered disc load outcomes were sagittal balance parameters in the thoracic spine and anatomic angles in the lumbar spine.

Full text

The Spine Journal 5 (2005) 297–309
Influence of spine morphology on intervertebral disc loads and stresses
in asymptomatic adults: implications for the ideal spine
Tony S. Keller, PhD
a,
*, Christopher J. Colloca, DC
b
, Deed E. Harrison, DC
c
,
Donald D. Harrison, DC, PhD
d
, Tadeusz J. Janik, PhD
e
a
Department of Mechanical Engineering, Department of Orthopaedics and Rehabilitation, University of Vermont, 33 Cochester Avenue,
Burlington, VT 05405-0156, USA
b
Department of Kinesiology, Arizona State University, Tempe, AZ 85287, USA; State of the Art Chiropractic Center, P.C., 11011 S. 48th Street,
Phoenix, AZ, 85044, USA
c
Ruby Mountain Chiropractic Center, 123 2nd Street, Elko, NV 89801, USA
d
Biomechanics Laboratory, Universite
´du Que
´bec a Trois-Rivie
`res, 3351 Boulevard des Forges, CP 500, Que
´bec, G9A 5H7, Canada
e
CompMath R/C, Huntsville, AL 35806, USA
Received 3 March 2004; accepted 29 October 2004
Abstract BACKGROUND CONTEXT: Sagittal profiles of the spine have been hypothesized to influence
spinal coupling and loads on spinal tissues.
PURPOSE: To assess the relationship between thoracolumbar spine sagittal morphology and inter-
vertebral disc loads and stresses.
STUDY DESIGN: A cross-sectional study evaluating sagittal X-ray geometry and postural loading
in asymptomatic men and women.
PATIENT SAMPLE: Sixty-seven young and asymptomatic subjects (chiropractic students) formed
the study group.
OUTCOME MEASURES: Morphological data derived from radiographs (anatomic angles and
sagittal balance parameters) and biomechanical parameters (intervertebral disc loads and stresses)
derived from a postural loading model.
METHODS: An anatomically accurate, sagittal plane, upright posture, quadrilateral element model
of the anterior spinal column (C2-S1) was created by digitizing lateral full-spine X-rays of 67
human subjects (51 males, 16 females). Morphological measurements of sagittal curvature and
balance were compared with intervertebral disc loads and stresses obtained using a quadrilateral
element postural loading model.
RESULTS: In this young (mean 26.7, SD 4.8 years), asymptomatic male and female population,
the neutral posture spine was characterized by an average thoracic angle (T1-T12)⫽⫹43.7⬚(SD
11.4⬚), lumbar angle (T12-S1)⫽⫺63.2⬚(SD 10.0⬚), and pelvic angle⫽⫹49.4⬚(SD 9.9⬚). Sagittal
curvatures exhibited relatively broad frequency distributions, with the pelvic angle showing the
least variance and the thoracic angle showing the greatest variance. Sagittal balance parameters,
C7-S1 and T1-T12, showed the best average vertical alignment (5.3 mm and ⫺0.04 mm, respec-
tively). Anterior and posterior disc postural loads were balanced at T8-T9 and showed the greatest
difference at L5-S1. Disc compressive stresses were greatest in the mid-thoracic region of the spine,
whereas shear stresses were highest at L5-S1. Significant linear correlations (p⬍.001) were found
between a number of biomechanical and morphological parameters. Notably, thoracic shear stresses
and compressive stresses were correlated to T1-T12 and T4-hip axis (HA) sagittal balance, respec-
tively, but not to sagittal angles. Lumbar shear stresses and body weight (BW) normalized shear
FDA device/drug status: not applicable.
Support in whole or in part was received from Chiropractic Biophysics
Nonprofit, Inc. and The Foundation for the Advancement of Chiropractic
Education, nonprofit organizations. Nothing of value received from a com-
mercial entity related to this research.
This study was presented, in part, at the 30th Annual Meeting of the
International Society for the Study of the Lumbar Spine, Vancouver, British
Columbia, Canada, May 13–17, 2003.
*Corresponding author. The University of Vermont, Department of
Mechanical Engineering, 119C Votey Building, Burlington, VT 05405-
0156. Tel.: (802) 656-1936; Fax: (802) 656-4441.
E-mail address:keller@emba.uvm.edu (T.S. Keller)
1529-9430/05/$ – see front matter
쑖
2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.spinee.2004.10.050

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309298
loads were correlated with T12-S1 balance, lumbar angle, and sacral angle. BW normalized lumbar
compressive loads were correlated with T12-S1 balance and sacral angle. BW normalized lumbar disc
shear (compressive) loads increased (decreased) significantly with decreasing lumbar lordosis.
Cervical compressive stresses and loads were correlated with all sagittal balance parameters except
S1-HA and T12-S1. A neutral spine sagittal model was constructed from the 67 subjects.
CONCLUSIONS: The analyses suggest that sagittal spine balance and curvature are important
parameters for postural load balance in healthy male and female subjects. Morphological predictors
of altered disc load outcomes were sagittal balance parameters in the thoracic spine and anatomic
angles in the lumbar spine. 쑖2005 Elsevier Inc. All rights reserved.
Keywords: Biomechanical modeling; Posture; Kyphosis; Spine; X-ray; Morphology
Introduction
Alterations in spinal balance and curvature are deemed
by many investigators to be implicated in the development
of a variety of spinal disorders, including acute and chronic
low back pain [1,2], disc degeneration [3–6], spondylosis
[3,7], ossification of spinal ligaments [8,9], adolescent idio-
pathic scoliosis [10,11], Scheuermann’s kyphosis [12], im-
paired ribcage expansion [13,14], early osteoarthritis and
disc degeneration [14,15], osteoporosis and vertebral com-
pression fractures [14,16], and spondylolisthesis [17]. Before
surgical or conservative rehabilitative treatment is initiated,
factors affecting the balance and curvature of the spine must
be identified and understood. This implies a need to define
a normal or “ideal” spinal morphology (balance and curva-
ture). Hence, a number of studies have focused on document-
ing the sagittal morphology of the spine [18–26].
Interactions between low back pain and spine morphology
have been reported [27–31]. In a group of 124 normal and
low back pain subjects, Harrison and associates [1] found that
several lumbar morphological measurements, including seg-
mental and total lumbar lordosis, were capable of discrimi-
nating normal subjects from acute low back pain sufferers
(hyperlordotic) and chronic low back pain sufferers (hypolor-
dotic). Harrison and associates hypothesized that the origin
of pain in the chronic group was abnormal disc loads associ-
ated with loss of lordosis. Lumbar lordosis has also been
suggested to be an anatomic [32] and mechanical necessity
for upright posture and locomotion [33,34]. The degree of
abnormal kyphotic thoracic curvatures has also been associ-
ated with a variety of clinical syndromes. Noteworthy, in-
creased thoracic kyphosis occurs with advanced aging and
is an indicator of thoracic vertebral compression fractures
leading to pain [12,13,15,16,35].
Although sagittal profiles of the thoracolumbar spine have
been hypothesized to influence spinal coupling [6,36,37]
and loads on spinal tissues [5,7,38], this has not been
documented in a systematic manner. The aims of the current
study were to 1) measure variations in spinal morphology
on standing lateral radiographs of asymptomatic subjects,
and 2) investigate the interaction between spinal column
morphology and intervertebral disc loads and stresses.
Methods
Sixty-seven human subjects (chiropractic students) par-
ticipated in this study. Subjects were young (mean age⫽26.7,
SD 4.8 years, n⫽67) and asymptomatic (mean Numerical
Rating Scale [NRS]⫽1.1, SD 1.1) at the time of the examina-
tion. The subjects had no prior history of back pain requiring
medical attention. The average body weight (BW) and height
of the subjects was 78.8 kg (SD 16.3, range 50 to 118) and
176 cm (SD 9.6, range 146 to 191), respectively. There were
51 male subjects (mean age⫽26.6, SD 5.2 years) and 17
female subjects (mean age⫽27.1, SD 3.4 years). In addition
to providing basic demographic data, subjects reviewed the
Institutional Review Board approved study protocol and
provided informed consent for their participation.
Segmental and global measurements of spine sagittal
curvature and balance were obtained from digitized lateral
full-spine X-rays of neutral posture, upright standing human
subjects. X-Y coordinates of the anterior-superior, posterior-
superior, anterior-inferior, and posterior-inferior corners of
each vertebral body (VB) from C2-S1 were marked using
a sonic digitizer (GP-9; GTCO CalComp, Columbia, MD).
The resolution of the sonic digitizer was 0.125 mm. Addi-
tional details of the radiographic measurement procedure
are found elsewhere [39,40].
A sagittal plane, upright posture, quadrilateral element
geometric model of the anterior spinal column (C2-S1) of
each subject was created using the digitized VB coordinate
data. From the models, several segmental and global morpho-
logical measurements of anatomic angles were obtained using
a previously reported posterior tangent technique [39–41].
The posterior tangent method of Harrison uses the posterior-
superior and posterior-inferior corners of each VB to form
tangent lines. The posterior margins of the VBs are less
subject to degenerative (osteoarthritic) changes in compari-
son to the anterior margins or end plates, which makes
anatomic angle measurements more reliable and valid in
comparison to other methods [39–41]. In this study, absolute
rotation angles (ARA) between T1-T12 (thoracic kyphosis
angle) and T12-S1 (lumbar lordosis angle) were computed
(Fig. 1). The pelvic angle, defined as the angle between
the hip axis and the posterior inferior body corner of S1,
Ferguson’s angle (FA), defined as the angle between hori-
zontal and a line through the superior S1 end plate, and

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 299
Fig. 1. Quadrilateral element model (Subject 20) and posterior tangent spine curvature measurement technique after Harrison et al. [41]. Relative rotation angles
were measured using a tangent line along the posterior margin of vertebral body for each vertebral segment and for the thoracic spine area (ARA
T1-T12
). Several
sagittal balance parameters were measured, including C4-L4, C7-S1 (not shown), T1-T12, T4-S1 (not shown), T12-S1, HA-T4 (not shown), HA-T12,
HA-S1 (not shown).
several sagittal balance parameters (posterior-anterior dis-
tance between VB centroids) were also computed. Defini-
tions of each of the morphological parameters examined
in this study are summarized in Table 1. Spinal column
height, defined as the difference between the most superior
C2 VB Y-coordinate and the most inferior S1 Y-coordinate,
was also calculated.
All anatomic angle and sagittal balance measurements
were performed using Matlab (The MathWorks, Natick,
MA). Intervertebral disc (IVD) and VB centroids were com-
puted from the quadrilateral element X,Y-coordinates using
an areal analysis algorithm [42]. IVD and VB centroids were
used to construct a composite sagittal profile model of the
C2-S1 spine.
A postural loading model [16,43] was used to determine
postural loads and stresses acting on the C2-S1 intervertebral
discs. For this analysis, the quadrilateral element model ge-
ometry of each subject was used as the input to the model,
which computed postural shear and compressive loads and
stresses for each IVD segment. A local coordinate system
was defined for each segment with the shear and compression
axes oriented parallel and perpendicular, respectively, to
the line bisecting each IVD. Loads were determined at the
anterior and posterior margins of the IVD (along disc bi-
sector) and at the IVD centroid. Postural loads were based
upon the BW load above each vertebral segment with the
line of gravity (LOG) initially positioned 10 mm anterior
to the centroid of the L4 VB (refer to Fig. 1). The location
of the LOG, BW load above each segment, values for the
posterior muscle moment arms, and disc cross-sectional
areas for the C2-S1 segments were based upon previously
published data for a 70-kg, 174-cm subject [43,44], but

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309300
Table 1
Descriptions of the standing lateral radiograph sagittal balance (alignment) and sagittal curvature (angle) parameters
Parameter Abbreviation Description
Sagittal alignment, C4-L4 C4-L4 Perpendicular distance between C4 (centroid) and L4 (centroid) plumblines (mm)
Sagittal alignment, T1-T12 T1-T12 Perpendicular distance between T1 (centroid) and T12 (centroid) plumblines (mm)
Sagittal alignment, T12-S1 T12-S1 Perpendicular distance between T12 (centroid) and S1 (posterior-superior VB corner) plumblines (mm)
Sagittal alignment, C7-HA C7-HA Perpendicular distance between C7 (centroid) and HA plumblines (mm)
Sagittal alignment, T4-HA T4-HA Perpendicular distance between T4 (centroid) and HA plumblines (mm)
Sagittal alignment, S1-HA S1-HA Perpendicular distance between S1 (posterior-superior VB corner) and hip axis (HA) plumblines (mm)
Sagittal alignment, C7-S1 C7-S1 Perpendicular distance between C7 (centroid) and S1 (posterior-superior VB corner) plumblines (mm)
Sagittal alignment, T4-S1 T4-S1 Perpendicular distance between T4 (centroid) and S1 (posterior-superior VB corner) plumblines (mm)
Thoracic kyphosis angle ARA
T1-T12
Absolute rotation angle formed by posterior vertebral body tangent lines of T1 and T12 (degrees)
Lumbar lordosis angle ARA
T12-S1
Absolute rotation angle formed by posterior vertebral body tangent lines of T12 and S1 (degrees)
Lumbo-sacral angle ARA
L4-S1
Absolute rotation angle formed by posterior vertebral body tangent lines of L4 and S1 (degrees)
Ferguson’s sacral base angle FA Angular measurement between horizontal and a line through the superior S1 end plate (degrees)
Pelvic tilt angle PA Angular measurement between horizontal and a line from the superior acetabulum of the posterior-inferior
vertebral body of S1 (degrees)
ARA⫽absolute rotation angles; FA⫽Ferguson’s angle; HA⫽hip axis; PA⫽pelvic angle; VB⫽vertebral body.
scaled to the weight and height of the subjects in this study.
Only the anterior column of the spine was considered, and
ligamentous structures were not explicitly modeled. In this
static upright posture model, equilibrium is based solely
upon the balance of BW forces and posterior trunk muscles
forces (erector spinae muscle). Although the effects of trunk
muscle synergism (contribution of anterior muscles), passive
spinal tissues (ligament load sharing), and intra-abdominal
pressure are not considered, the model force predictions
show good agreement with other experimental and analytical
studies [16,43].
Average values of the IVD loads (shear and compression
loads normalized with respect to BW) and IVD stresses
(shear and compression) were computed for the following
regions: whole spine (C2-S1), cervical (C2-T1), thoracic
(T1-L1), and lumbosacral (L1-S1), resulting in a total of 16
IVD biomechanical parameters. A least-square, linear regres-
sion analysis was used to examine correlations between the
morphologic parameters (sagittal balance, anatomic angles)
and the three regional and whole spine biomechanical parame-
ters (BW normalized loads, stresses).
Statistical analyses
Frequency distributions (histograms) and distribution
characteristics (skew, kurtosis) were performed on the mor-
phological data. Morphological anatomic angle parameters
were also used to subdivide the subjects according to the
following criteria: Ferguson’s sacral base angle (FA), lumbo-
sacral angle (ARAL4-S1), and thoracic kyphosis (ARAT1-T12),
each normalized with respect to the lumbar lordosis angle
(ARAT12-S1). For each of these criteria, the subjects were par-
titioned into four groups constructed from the quartiles of the
measurements. The mean regional biomechanical parameters
for each quartile were then compared using a one-way analy-
sis of variance (ANOVA). The normality and homogeneity
of variance assumptions were met for all 48 analyses (3
morphological criteria×16 biomechanical parameters). Sta-
tistically significant ANOVAs were followed by Tukey’s
multiple comparison procedure to discern the nature of pair-
wise differences. Because a large number of ANOVAs were
performed, in order to protect against errors of incorrectly
finding differences when none exist, significant differences
identified by the ANOVA were only followed up by the
Tukey procedure when the ANOVA p value was less than
.0001.
Results
In this group of normal male and female subjects, the
average ratio of spinal column height (C2-S1) and body
height was 0.35 (range 0.30 to 0.42), indicating that in
the neutral, upright posture the spine was approximately one-
third of the body height. Considerable variation in neutral,
upright posture sagittal balance and anatomic angles was
observed among the subjects (Table 2). Posterior-anterior
variations in sagittal balance (minimum to maximum)
spanned a range from 59.5 mm (S1-HA) to 96.7 mm (T4-
S1). Frequency distributions for selected balance parameters
(C4-L4, T1-T12, and T12-S1) are shown in Figure 2. C4-
L4 balance showed a flattened (negative kurtosis), relatively
symmetric (low skew) frequency distribution, whereas the
T1-T12 balance was more peaked (low kurtosis) and more
skewed. T12-S1 balance was the most symmetric. C7-S1, T1-
T12, and T12-S1 sagittal balance parameters showed the
lowest average posterior-anterior misalignment (⫺5.3 mm,
⫺0.04 mm, and ⫹12.1 mm, respectively), but each exhibited
a range of over 80 mm (Table 2). The hip axis (HA) sagittal
balance parameters (C7-HA, T4-HA, S1-HA) showed the
least overall variance (SD/mean). Of the anatomic angle
measures, thoracic curvature (ARAT1-T12) and lumbo-sacral
angle (ARAL4-S1) showed the highest symmetry, whereas
pelvic angle and FA exhibited the least symmetric frequency
distributions (Table 2 and Fig. 3). Thoracic kyphosis angle
(ARAT1-T12) showed the least overall variance among the
subjects (Table 2).

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 301
Table 2
Summary of sagittal balance and anatomic angles obtained from analysis of neutral, upright standing posture radiographs. Minimum, maximum,
skew, kurtosis, and mean (standard deviation) for 67 subjects
Parameter*Minimum Maximum Skew
†
Kurtosis
†
Mean (SD)
C4-L4 (mm) ⫺24.4 72.1 ⫺0.11 ⫺0.89 23.7 (23.4)
T1-T12 (mm) ⫺57.8 ⫺27.6 ⫺0.46 ⫺0.01 ⫺0.04 (20.6)
T12-S1 (mm) ⫺27.6 55.8 ⫺0.03 0.29 12.1 (15.6)
C7-HA (mm) 13.5 77.9 ⫺0.31 ⫺0.37 35.2 (22.6)
T4-HA (mm) 16.2 115.0 ⫺0.21 0.07 64.1 (21.1)
S1-HA (mm) ⫺4.4 55.1 ⫺0.54 0.41 29.8 (11.9)
C7-S1 (mm) ⫺42.8 46.3 ⫺0.39 ⫺0.67 5.3 (23.2)
T4-S1 (mm) ⫺15.0 81.7 ⫺0.43 0.25 34.3 (19.5)
ARA
T1-T12
(degrees) 16.3 71.5 ⫺0.06 ⫺0.15 43.7 (11.4)
ARA
T12-S1
(degrees) ⫺97.4 ⫺41.1 ⫺0.31 ⫺0.06 ⫺67.4 (12.6)
ARA
LA-S1
(degrees) ⫺69.9 ⫺24.0 0.20 0.39 ⫺48.1 (9.8)
FA (degrees) 32.4 58.8 0.40 ⫺0.24 44.1 (5.7)
PA (degrees) 33.7 83.7 0.97 1.31 49.4 (9.9)
*See Table 1 and Figure 1 for abbreviations and descriptions of measurements using Harrison posterior tangent (angles) and plumbline techniques
(horizontal translation, mm). Sagittal balance parameters C4-L4, T1-T12, T12-S1, C7-HA, T4-HA, S1-HA, C7-S1, and T4-S1 are the perpendicular
distances from the respective quadrilateral element VB centroids.
†
Skewness and kurtosis indicate distribution asymmetry and flatness, respectively.
The postural loading model indicated that anterior and
posterior IVD loads were balanced at T8-T9 and showed the
greatest difference at L5-S1 (Fig. 4). The pattern of IVD
postural stresses (compression, shear) mirrored the sagittal
Fig. 2. Morphological analysis results for neutral posture sagittal balance. Histograms for C4-L4, T1-T12, and T12-S1 sagittal alignment are illustrated.
Each bar represents the number of data points between the current bin number and the adjoining higher bin.
curvatures of the spine and showed less overall variation
(2.3 fold, 1.8 fold) than the corresponding postural loads.
The average compressive postural loads acting on the IVD
centroid ranged from 13.8% of BW (C2-C3) to 94.3% BW

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309302
Fig. 3. Morphological analysis results for neutral posture sagittal profile. Histograms for ARA
T1-T12
, ARA
T12-S1
, ARA
L4-S1
, and Ferguson’s angle (FA)
curvatures are illustrated. Each bar represents the number of data points between the current bin number and the adjoining higher bin.
(T11-T12). Mean IVD compressive loads (% BW) for the
cervical (C2-T1), thoracic (T1-L1), and lumbar (L1-S1)
regions were 15.2% (SD 5.0), 61.9% (SD 6.8), and 65.1%
(SD 2.0), respectively. Compressive stresses were highest
in the mid-thoracic region of the spine, whereas shear
stresses were lowest at T6-T7 and highest at L5-S1 (Fig. 5).
Significant (p⬍.001) linear correlations (r⬎.37) were
found between a number of postural load parameters and
sagittal morphology variables (Table 3). Thoracic shear
stresses were correlated with sagittal balance variables (C7-
HA, T4-HA, C7-S1, T4-S1, C4-L4, T1-T12), but were
poorly correlated to sagittal anatomic angles. Lumbar shear

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 303
Fig. 4. Distribution of anterior and posterior intervertebral disc compressive loads along the C2-S1 spine. Symbols and error bars represent mean and
standard deviation, respectively, of the 67 subjects.
stresses were closely correlated with T12-S1 sagittal balance
and showed a significant correlation with lumbar angle
(ARAT12-S1) and sacral angle (FA). Cervical and thoracic
compressive stresses were correlated with several sagittal
balance variables, most notably C7-HA, T4-HA, and C7-
S1, but not to anatomic angles. BW normalized thoracic
shear loads were well correlated to the sagittal balance vari-
ables. BW normalized compressive loads were closely corre-
lated with sagittal balance parameters in both the cervical
and thoracic regions. FA was positively correlated with
lumbar shear loads (% BW) and lumbar compressive loads
(% BW). In general, intervertebral disc stresses and loads
were greatest in subjects with sagittal posture imbalance.
Anterior translation of cervical and thoracic segments rela-
tive to the hip axis was associated with increased IVD pos-
tural stresses and loads.
The quartile analysis revealed only a few significant dif-
ferences within the mean regional biomechanical parameters
for each quartile, and these differences were limited to
quartiles of the anatomic angle ratio ARAL4-S1/ARAT12-S1
(Q1⫽0.44 to 0.62; Q2⫽0.62 to 0.71; Q3⫽0.71 to 0.82;
Q4⫽0.82 to 1.06). Noteworthy was the finding that lumbar
shear loads (normalized to BW) decreased significantly with
increasing ARAL4-S1/ARAT12-S1 ratio (Q1 and Q2⬎Q4,
Q1⬎Q3, p⬍.0001). Conversely, lumbar compressive loads
(normalized to BW) increased significantly with increas-
ing ARAL4-S1/ARAT12-S1 ratio (Q3 and Q4⬎Q1, p⬍.0001).
These findings are graphically summarized in Figure 6.
An average neutral posture spine model was constructed
from the 67 subjects (Fig. 7). Plumblines drawn from C7
and T1 clearly illustrate the upright posture sagittal balance
of C7-S1 and T1-T12. In the composite model, the L3 VB
centroid was located 6.4 mm anterior (along posteroanterior
x-axis) to the femoral head centroid (X, Y-coordinate⫽
⫺48.7, ⫺62.7 mm). X, Y coordinates corresponding to the
C2-S1 composite model VB (and IVD) quadrilateral element
centroids are summarized in Table 4.
Discussion
Sagittal curvatures (lumbar lordosis, thoracic kyphosis)
and pelvic rotation are geometric parameters that are known
to have a significant influence on mechanical properties
during compressive loading [37,45–51]. Hence, a well-
defined, normal thoracolumbar spine sagittal morphology is
important for understanding spine biomechanics. In this study,
a postural loading model [16,43] was used to systemati-
cally investigate the relationship between sagittal morphology
and postural loads and stresses acting on the intervertebral
discs.
One of the aims of the current study was to detail the
variation in spinal column morphology on standing lateral
radiographs of asymptomatic male and female subjects. A
number of studies have performed similar segmental analy-
ses of sagittal plane alignment and balance using lateral

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309304
Fig. 5. Distribution of intervertebral compressive and shear postural stresses along the C2-S1 spine (IVD centroids). Symbols and error bars represent mean
and standard deviation, respectively, of the 67 subjects.
standing radiographs of the lumbar spine [24,52,53] and
thoracolumbar spine [20,21,25,29,54–56] of asymptomatic
subjects. Examination of the literature indicates that there
is large variability in sagittal balance and alignment param-
eters in asymptomatic subjects, which do not appear to be
race-related [52,54] or gender-related [20,21,24,25,29].
However, weak age-related alterations in sagittal alignment
have been reported where adolescents tend to stand with
greater posterior sagittal balance [20] and seniors have for-
ward sagittal balance with increased thoracic kyphosis [57].
Table 3
Correlation (correlation coefficients) of sagittal balance parameters, anatomic angles, and regional intervertebral disc postural loads (% BW) and stresses
Parameter C7-HA T4-HA S1-HA C7-S1 T4-S1 C4-L4 T1-T12 T12-S1 ARA
T12-S1
Shear stress (C2-S1) ⫺.643 ⫺.461 .371 ⫺.817 ⫺.724 ⫺.815 ⫺.622 ⫺.341 .103
Lumbar ⫺.233 ⫺.296 .183 ⫺.321 ⫺.432 ⫺.055 .299 ⫺.890 ⫺.437
Thoracic ⫺.732 ⫺.522 .120 ⫺.775 ⫺.637 ⫺.844 ⫺.951 .147 .274
Cervical .039 .147 .389 ⫺.161 ⫺.077 ⫺.303 ⫺.092 ⫺.060 .207
Compressive stress (C2-S1) .565 .544 .015 .544 .580 .531 .433 .249 ⫺.212
Lumbar ⫺.035 .052 .177 ⫺.124 ⫺.051 ⫺.163 ⫺.143 .045 ⫺.058
Thoracic .434 .529 .119 .363 .500 .310 .212 .306 ⫺.194
Cervical .751 .539 .187 .828 .698 .885 .783 .147 ⫺.250
Shear load (C2-S1) ⫺.715 ⫺.691 .223 ⫺.811 ⫺.883 ⫺.631 ⫺.364 ⫺.728 ⫺.090
Lumbar ⫺.203 ⫺.283 .148 ⫺.274 ⫺.397 .008 .345 ⫺.890 ⫺.487
Thoracic ⫺.733 ⫺.628 .043 ⫺.737 ⫺.706 ⫺.801 ⫺.891 .096 .426
Cervical .090 .209 .370 ⫺.102 .001 ⫺.282 ⫺.089 .030 .225
Compressive load (C2-S1) .748 .737 ⫺.194 .829 .916 .756 .450 .643 ⫺.263
Lumbar .161 .187 ⫺.307 .314 .390 .134 ⫺.227 .775 .267
Thoracic .679 .750 ⫺.122 .724 .886 .627 .314 .691 ⫺.282
Cervical .812 .569 ⫺.270 .929 .779 .988 .851 .190 ⫺.276
ARA⫽absolute rotation angles; BW⫽body weight; HA⫽hip axis.
Correlations greater than .37 or less than ⫺.37 were significant at p⬍.001 or more.
Our study focuses on a group of similar age asymptomatic
male and female subjects and, like previous investigations,
considerable variation in neutral, upright posture sagittal
balance and anatomic angles was observed. In addition, in-
teractions between spinal morphology and intervertebral disc
loads and stresses were observed. Discussion of these inter-
actions follows.
To begin, our biomechanical and morphological analyses
indicated that lumbar shear loads increased with decreasing
lumbar lordosis (decreasing ARAL4-S1/ARAT12-S1 ratio).

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 305
Fig. 6. Box plot summary of the quartile analysis for the ratio ARA
L4-S1
/ARA
T12-S1
. Increasing ARA
L4-S1
/ARA
T12-S1
ratio corresponds to increasing lumbar
lordosis. (A) The first quartile (Q1) and second quartile (Q2) have significantly greater (p⬍.0001) mean lumbar shear load (BW) responses than the fourth
quartile (Q4). Q1 has a significantly greater (p⬍.0001) mean lumbar shear load (BW) response than the third quartile (Q3). (B) Q1 has a significantly
lower (p⬍.0001) mean lumbar compressive load (BW) response than Q3 or Q4.
Using cadaveric lumbar spines tested in axial load, Umehara
et al. [5] also found increased posterior shear stresses as
a consequence of an 8⬚reduction of the L4-S1 lordosis. The
change in loading from compression to shear, with loss of
the distal lordosis, may have clinical relevance to low back
pain and disc degeneration. For example, in asymptomatic
normal subjects, the distal lumbar lordosis (L4-S1) com-
prises approximately 65% of total lordosis [20,21,24,29,54]
and, mechanically, the normal lumbar spine is best suited
to withstand compressive loads [23,58]. In comparison to
normal subjects, chronic low back patients with and without
lower lumbar degenerative discs have been found to have a
statistically significant reduction in the distal lumbar lordosis
[1,29,31,55]. In subjects with reduced distal lumbar lordo-
sis, the increased shear loads may well trigger an increase
in matrix metalloproteinases leading to sensitization of noci-
ceptive neurons and eventual degradation of disc tissue
[59,60].
Second, of clinical importance was the finding that lumbar
shear stresses were closely correlated with T12-S1 sagittal
balance, with lumbar lordosis, and with sacral angle (FA).
Jackson and associates [55] found that hyperlordosis of the
lumbar spine was present in 30 patients with Grades I and II
L5-S1 spondylolisthesis. Hypothesizing that increased shear

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309306
Fig. 7. Neutral spine ensemble average sagittal profile model (vertebral
body centroids) derived from the quadrilateral element models of all 67
subjects examined in this study. Sagittal balance parameters, T1-T12 and
C7-S1, are shown illustrating the vertical alignment of these segments in
the upright posture.
stresses were the cause of spondylolisthesis, Rajnics and
colleagues [61] found increased lumbar lordosis and sacral
angle in patients with spondylolisthesis matched to normal
controls. Using a sagittal balance measurement from L1
centroid to posterior superior of S1, Kawakami et al. [62]
identified an anterior displacement of greater than 35 mm to
adversely affect the recovery in 47 surgically treated patients
with degenerative spondylolisthesis. In fact, anterior to
posterior thoracic translations can cause a net change in
thoracic kyphosis of 26⬚and 31⬚in sacral angle [63]. Thus, the
increased shear stresses resulting from a forward displaced
sagittal balance, increased sacral angle, and lumbar angle
could be a likely mechanism for the development, progres-
sion, or exacerbation of spondylolisthesis.
Third, the model prediction of increased IVD postural
stresses and loads associated with anterior translation of
cervical and thoracic segments relative to the hip axis may
have implications for conditions such as upper and low back
pain [63], post-fusion adjacent segment disc degeneration
[4], osteoporotic deformity [16], and age-related increases
in anterior sagittal postural balance [57]. In general, our
model predictions indicated that intervertebral disc stresses
and loads were greatest in subjects with sagittal posture
imbalance. This result agrees qualitatively with previously
published studies on sagittal postural displacements in the
cervical [64] and thoracolumbar regions [38].
The predicted sagittal spine/posture induced load
changes may expose the IVD to an abnormal stress environ-
ment. Apparently, the IVD is well suited for this situation be-
cause the IVD stresses showed less overall variation (2.3
Table 4
X-Y coordinates of the neutral posture ensemble average model. Vertebral
body (VB) and intervertebral disc (IVD) centroid coordinates defined
relative to the posterior-inferior corner of the S1 VB (digitizing tablet
X,Y-coordinate origin)
VB Centroid IVD Centroid
X-coor. Y-coor. X-coor. Y-coor.
Segment (mm) (mm) (mm) (mm)
C2 ⫺26.34 583.58
C3 ⫺25.18 562.96 ⫺26.36 575.25
C4 ⫺23.05 544.86 ⫺24.34 553.73
C5 ⫺20.52 527.39 ⫺21.94 536.03
C6 ⫺17.83 509.94 ⫺19.46 518.67
C7 ⫺13.59 491.80 ⫺16.51 501.01
T1 ⫺6.81 472.76 ⫺11.03 482.65
T2 1.49 452.90 ⫺2.78 463.16
T3 9.16 432.31 5.77 443.06
T4 15.36 410.84 12.94 422.11
T5 20.00 388.11 18.60 399.86
T6 22.70 364.75 22.18 376.63
T7 23.29 340.54 23.88 352.79
T8 21.13 315.73 22.81 328.24
T9 16.90 290.37 19.14 303.29
T10 11.10 263.92 13.95 277.57
T11 3.48 235.71 7.26 250.24
T12 ⫺6.77 205.81 ⫺2.11 221.32
L1 ⫺18.83 174.14 ⫺14.09 190.80
L2 ⫺31.86 139.94 ⫺27.09 157.90
L3 ⫺42.37 103.01 ⫺39.59 122.06
L4 ⫺46.73 64.66 ⫺47.58 83.76
L5 ⫺41.20 27.12 ⫺47.89 44.96
S1 ⫺20.14 ⫺2.13 ⫺35.53 8.31
fold, 1.8 fold) than the corresponding postural loads. None-
theless, because the pattern of IVD postural stresses mirrored
the sagittal curvatures and sagittal displacement of the spine, a
failure of the IVD’s hydrostatic mechanism under these
sustained loads could occur. In the sagittal plane, progression
of kyphotic deformities follows mechanical loading (Hueter-
Volkmann law) where increased compression loads are
known to cause deformity accentuation [12]. However, in-
creased dorsal kyphosis as a result of aging has been shown to
be closely related to the integrity of the IVD as well [65].
Our findings, of increased thoracic IVD compression load
with increased thoracic kyphosis and increased IVD com-
pression and shear loads with increased sagittal balance
displacement, might prove to be a logical mechanism for
IVD deformity causation and progression in the thoracic
region.
The spine morphological data and quadrilateral element
data presented in this study should be useful for future
analytical and numerical studies of the normal and pathologi-
cal spines. One of the areas of interest should be that
of compensated/uncompensated spinal balance in terms of
sagittal alignment parameters. In general, as the thoracic
kyphosis increases the lumbar lordosis is also seen to in-
crease [55]. As an example, in Scheuermann’s kyphosis,
hyperlordosis of the lumbar spine is known to occur with

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 307
the cervical spine adopting a curvature that is the net differ-
ence between the two lower curves (compensated) [66].
However, in patients with spondylolisthesis, total kyphosis
is shown to decrease with increased lumbar lordosis. An
analysis of postural load and IVD stresses assists in under-
standing why different deformities are compensated (or un-
compensated) and may help to explain when postural
geometry/load imbalance is associated with increased risk
of deformity.
Investigators have reported interactions between low back
pain and biomechanical changes of the lumbar spine, includ-
ing changes in lumbar curvature [1,29,31,55,67] and inter-
vertebral disc height [28]. Clinical evaluation of spine
morphology and postural load balance in acute and chronic
back pain patients is therefore needed for at least two primary
reasons: 1) to isolate morphological parameters that might
provide clinicians with diagnostic tools to decide when an
individual might be at risk of a first time experience of
pain requiring treatment intervention, and 2) to identify areas
of likely deformity progression/development in an individ-
ual spine. The postural disc load and spinal morphology
interactions identified in this study have clinical implications
for the ideal sagittal spine geometry.
When developing a biomechanical model, it is often nec-
essary to make a number of assumptions in order to avoid
as much excess complexity as is practical within the con-
straints of validity. The quadrilateral element model used
in this study, and applied to the neutral, standing posture, is
a single muscle model designed to provide an overview of
the loading profile of a complex mechanical system. This
static model has been used in previous studies to examine
IVD and VB loads and stresses in conjunction with body
height changes [43], thoracic deformity [16], and anterior
translated postures [68]. These studies provide additional
details and validation of the model, and indicate that the
magnitude of predicted cervical, thoracic, and lumbar loads
and stresses is consistent with previous experimental stud-
ies. Several model-specific assumptions, however, warrant
additional elaboration. Firstly, the X-ray-based quadrilateral
element model does not incorporate the load sharing charac-
teristics of the passive spinal tissues, although it is known
that these tissues do contribute to the restorative moment
of the spine [69,70]. Secondly, stresses were computed using
a local coordinate system oriented with respect to the IVD
bisector. Thirdly, and perhaps most noteworthy, is our as-
sumption that the location of the center of mass (COM)
is constant, acting along a fixed LOG or “line of gravity”.
According to a computed tomography study conducted by
Pearsall et al. [71], the COM shows a curvilinear change from
T4 to L5 with an average deviation of the COM relative to
L4 of 5.9 mm. Small deviations in the COM have a very
small overall effect on the LOG-based shear and compressive
force predictions for segments L5 and above, but are
more problematic for force estimates at L5-S1 where the
deviation of the COM from a fixed LOG is marked (about
20 mm). Previous biomechanical studies indicate that the
upright posture L5-S1 anterior-posterior shear is generally
less than or equal to one-fifth of compressive force [72,73],
which is substantially lower than the ratio obtained in
this study (mean L5-S1 shear/compression ratio⫽0.53).
However, because the main outcome measures reported in
the current study were based on a region-by-region analysis
of IVD loads, the overall influence of COM variations on
the biomechanical results and conclusions is minor.
Conclusions
In this study we have presented a mathematical analysis
of postural disc loads and sagittal balance parameters based
on digitized radiographs of asymptomatic subjects. Postural
load analyses and identification of spinal geometry can
be used for clinical comparisons, and may be used to
discriminate between normal subjects and those with low
back pain. The sagittal plane quadrilateral element geometric
model data will also assist researchers in developing math-
ematical models. This study provides descriptive morpho-
logical and biomechanical information on asymptomatic
subjects. Our analyses suggest that sagittal spine balance
and anatomic curvature influence postural loading and load
balance of the intervertebral disc in healthy male and female
subjects. Specifically, the model predicted altered and in-
creased loads as a result of variances in the ratio of distal
lumbar lordosis to total lumbar lordosis, lumbar lordosis
to thoracic kyphosis, and sagittal plane translations of the
cervical and thoracic regions relative to the hip axis and
sacrum. These deviations might prove to be undesirable for
optimum spinal structure and function. We speculate that
much of this variability is related to rotation and translation
displacements in the thorax and head, which may be coupled
to the variations in sagittal alignment observed in this
and previous studies.
Acknowledgments
The authors would like to thank the following funding
agencies for their support of this research: Chiropractic Bio-
physics NonProfit, Inc. and the Foundation for the Advance-
ment of Chiropractic Education. Statistical support provided
by Burt Holland is also greatly appreciated.
References
[1] Harrison DD, Cailliet R, Janik TJ, Troyanovich SJ, Harrison DE,
Holland B. Elliptical modeling of the sagittal lumbar lordosis and
segmental rotation angles as a method to discriminate between normal
and low back pain subjects. J Spinal Disord 1998;11:430–9.
[2] Steinberg EL, Luger E, Arbel R, Menachem A, Dekel S. A compara-
tive roentgenographic analysis of the lumbar spine in male army
recruits with and without lower back pain. Clin Radiol 2003;58:985–9.

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309308
[3] Schlegel JD, Smith JA, Schleusener RL. Lumbar motion segment
pathology adjacent to thoracolumbar, lumbar, and lumbosacral fu-
sions. Spine 1996;21:970–81.
[4] Kumar MN, Baklanov A, Chopin D. Correlation between sagittal
plane changes and adjacent segment degeneration following lumbar
spine fusion. Eur Spine J 2001;10:314–9.
[5] Umehara S, Zindrick MR, Patwardhan AG, et al. The biomechanical
effect of postoperative hypolordosis in instrumented lumbar fusion on
instrumented and adjacent spinal segments. Spine 2000;25:1617–24.
[6] Akamaru T, Kawahara N, Tim YS, et al. Adjacent segment motion
after a simulated lumbar fusion in different sagittal alignments: a
biomechanical analysis. Spine 2003;28:1560–6.
[7] Oda I, Cunningham BW, Buckley RA, et al. Does spinal kyphotic
deformity influence the biomechanical characteristics of the adjacent
motion segments? An in vivo animal model. Spine 1999;24:2139–46.
[8] Fukuyama S, Nakamura T, Ikeda T, Takagi K. The effect of mechanical
stress on hypertrophy of the lumbar ligamentum flavum. J Spinal
Disord 1995;8:126–30.
[9] Nathan M, Pope MH, Grobler LJ. Osteophyte formation in the verte-
bral column: a review of the etiologic factors–Part II. Contemp Orthop
1994;29:113–9.
[10] Deacon P, Flood BM, Dickson RA. Idiopathic scoliosis in three
dimensions: a radiographic and morphometric analysis. J Bone Joint
Surg Br 1984;66:509–12.
[11] Dickson RA. The aetiology of spinal deformities. Lancet 1988;1:
1151–5.
[12] Wenger DR, Frick SL. Scheuermann kyphosis. Spine 1999;24:2630–9.
[13] Culham EG, Jimenez HA, King CE. Thoracic kyphosis, rib mobility,
and lung volumes in normal women and women with osteoporosis.
Spine 1994;19:1250–5.
[14] Abdel-Hamid OA, Bassiouni H, Koutri R, Nijs J, Geusens P,
Dequeker J. Aging of the thoracic spine: distinction between wedging
in osteoarthritis and fracture in osteoporosis: a cross-sectional and
longitudinal study. Bone 1994;15:437–42.
[15] Lauerman WC, Platenberg RC, Cain JE, Deeney VF. Age-related
disk degeneration: preliminary report of a naturally occurring baboon
model. J Spinal Disord 1992;5:170–4.
[16] Keller TS, Harrison DE, Colloca CJ, Harrison DD, Janik TJ. Predic-
tion of osteoporotic spinal deformity. Spine 2003;28:455–62.
[17] Antoniades SB, Hammerberg KW, Dewald RL. Sagittal plane config-
uration of the sacrum in spondylolisthesis. Spine 2000;25:1085–91.
[18] Voutsinas SA, MacEwen GD. Sagittal profiles of the spine. Clin
Orthop 1986;210:235–42.
[19] Repanti M, Korovessis PG, Stamatakis MV, Spastris P, Kosti P. Evolu-
tion of disc degeneration in lumbar spine: a comparative histological
study between herniated and postmortem retrieved disc specimens.
J Spinal Disord 1998;11:41–5.
[20] Vedantam R, Lenke LG, Keeney JA, Bridwell KH. Comparison of
standing sagittal spinal alignment in asymptomatic adolescents and
adults. Spine 1998;23:211–5.
[21] Gelb DE, Lenke LG, Bridwell KH, Blanke K, McEnery KW. An
analysis of sagittal spinal alignment in 100 asymptomatic middle and
older aged volunteers. Spine 1995;20:1351–8.
[22] Jackson RP, Kanemura T, Kawakami N, Hales C. Lumbopelvic lordo-
sis and pelvic balance on repeated standing lateral radiographs of
adult volunteers and untreated patients with constant low back pain.
Spine 2000;25:575–86.
[23] Rajnics P, PomeroV, Templier A, Lavaste F, Illes T. Computer-assisted
assessment of spinal sagittal plane radiographs. J Spinal Disord 2001;
14:135–42.
[24] Janik TJ, Harrison DD, Cailliet R, Troyanovich SJ, Harrison DE. Can
the sagittal lumbar curvature be closely approximated by an ellipse?
J Orthop Res 1998;16:766–70.
[25] Harrison DE, Janik TJ, Harrison DD, Cailliet R, Harmon SF. Can
the thoracic kyphosis be modeled with a simple geometric shape?: the
results of circular and elliptical modeling in 80 asymptomatic patients.
J Spinal Disord Tech 2002;15:213–20.
[26] Harrison DD, Harrison DE, Janik TJ, Cailliet R, Haas J. Do alterations
in vertebral and disc dimensions affect an elliptical model of thoracic
kyphosis? Spine 2003;28:463–9.
[27] Kostuik JP, Hall BB. Spinal fusions to the sacrum in adults with
scoliosis. Spine 1983;8:489–500.
[28] Frymoyer JW, Newberg A, Pope MH, Wilder DG, Clements J,
MacPherson B. Spine radiographs in patients with low-back pain:
an epidemiological study in men. J Bone Joint Surg [Am ] 1984;66:
1048–55.
[29] Jackson RP, McManus AC. Radiographic analysis of sagittal plane
alignment and balance in standing volunteers and patients with low
back pain matched for age, sex, and size. A prospective controlled
clinical study. Spine 1994;19:1611–8.
[30] Harrison DD, Colloca CJ, Troyanovich SJ, Harrison DE. Torque
misuse revisited. J Manipulative Physiol Ther 1998;21:649–55.
[31] Korovessis P, Dimas A, Iliopoulos P, Lambiris E. Correlative analysis
of lateral vertebral radiographic variables and medical outcomes study
short-form health survey: a comparative study in asymptomatic volun-
teers versus patients with low back pain. J Spinal Disord Tech 2002;
15:384–90.
[32] Adams MA, Bogduk N, Burton K, Dolan P. The biomechanics of
back pain. Edinburgh: Churchill Livingstone; 2002. p. 107–9.
[33] Gracovetsky S, Farfan H. The optimum spine. Spine 1986;11:543–73.
[34] Farfan HF. Form and function of the musculoskeletal system as re-
vealed by mathematical analysis of the lumbar spine. An essay. Spine
1995;20:1462–74.
[35] Singer KP, Edmondston SJ, Day RE, Breidahl WH. Computer-assisted
curvature assessment and Cobb angle determination of the thoracic
kyphosis: an in vivo and in vitro comparison. Spine 1994;19:1381–4.
[36] Panjabi M, Abumi K, Duranceau J, Oxland T. Spinal stability and
intersegmental muscle forces: a biomechanical model. Spine 1989;14:
194–200.
[37] Cholewicki J, Crisco JJ III, Oxland TR, Yamamoto I, Panjabi MM.
Effects of posture and structure on three-dimensional coupled rotations
in the lumbar spine: a biomechanical analysis. Spine 1996;21:2421–8.
[38] Kiefer A, Shirazi-Adl A, Parnianpour M. Synergy of the human spine
in neutral postures. Eur Spine J 1998;7:471–9.
[39] Harrison DE, Harrison DD, Cailliet R, Troyanovich SJ, Janik TJ,
Holland B. Cobb method or Harrison posterior tangent method: which
to choose for lateral cervical radiographic analysis. Spine 2000;25:
2072–8.
[40] Harrison DE, Cailliet R, Harrison DD, Janik TJ, Holland B. Reliability
of centroid, Cobb, and Harrison posterior tangent methods: which to
choose for analysis of thoracic kyphosis. Spine 2001;26:E227–34.
[41] Harrison DE, Harrison DD, Cailliet R, Janik TJ, Holland B. Radio-
graphic analysis of lumbar lordosis: centroid, Cobb, TRALL, and
Harrison posterior tangent methods. Spine 2001;26:E235–42.
[42] Nagurka ML, Hayes WC. An interactive graphics package for calculat-
ing cross-sectional properties of complex shapes. J Biomech 1980;13:
59–64.
[43] Keller TS, Nathan M. Height change caused by creep in intervertebral
discs: a sagittal plane model. J Spinal Disord 1999;12:313–24.
[44] Ruff S. Brief acceleration: less than one second. Ger Aviat Med World
War II 1950;1:584–97.
[45] Crisco JJ, Panjabi MM. The intersegmental and multisegmental mus-
cles of the lumbar spine: a biomechanical model comparing lateral
stabilizing potential. Spine 1991;16:793–9.
[46] Dulhunty JA. Anthropometrical and mechanical considerations in
determining normal parameters for the sagittal lumbar spine.
J Manipulative Physiol Ther 1997;20:92–102.
[47] Granata KP, Wilson SE. Trunk posture and spinal stability. Clin
Biomech 2001;16:650–9.
[48] Kiefer A, Shirazi-Adl A, Parnianpour M. Stability of the human spine
in neutral postures. Eur Spine J 1997;6:45–53.
[49] Lavaste F, Skalli W, Robin S, Roy-Camille R, Mazel C. Three-dimen-
sional geometrical and mechanical modelling of the lumbar spine.
J Biomech 1992;25:1153–64.

T.S. Keller et al. / The Spine Journal 5 (2005) 297–309 309
[50] Shirazi-Adl A, Parnianpour M. Role of posture in mechanics of the
lumbar spine in compression. J Spinal Disord 1996;9:277–86.
[51] Shirazi-Adl A, Parnianpour M. Effect of changes in lordosis on me-
chanics of the lumbar spine-lumbar curvature in lifting. J Spinal
Disord 1999;12:436–47.
[52] Chen YL. Geometric measurements of the lumbar spine in Chinese
men during trunk flexion. Spine 1999;24:666–9.
[53] George SZ, Hicks GE, Nevitt MA, Cauley JA, Vogt MT. The relation-
ship between lumbar lordosis and radiologic variables and lumbar
lordosis and clinical variables in elderly, African-American women.
J Spinal Disord Tech 2003;16:200–6.
[54] Korovessis PG, Stamatakis MV, Baikousis AG. Reciprocal angulation
of vertebral bodies in the sagittal plane in an asymptomatic Greek
population. Spine 1998;23:700–4.
[55] Jackson RP, Peterson MD, McManus AC, Hales C. Compensatory
spinopelvic balance over the hip axis and better reliability in measur-
ing lordosis to the pelvic radius on standing lateral radiographs of
adult volunteers and patients. Spine 1998;23:1750–67.
[56] Jackson RP, Hales C. Congruent spinopelvic alignment on standing
lateral radiographs of adult volunteers. Spine 2000;25:2808–15.
[57] Hammerberg EM, Wood KB. Sagittal profile of the elderly. J Spinal
Disord Tech 2003;16:44–50.
[58] Shirazi-Adl A, Parnianpour M. Load-bearing and stress analysis of
the human spine under a novel wrapping compression loading. Clin
Biomech 2000;15:718–25.
[59] Goupille P, Jayson MI, Valat JP, Freemont AJ. Matrix metalloprotei-
nases: the clue to intervertebral disc degeneration? Spine 1998;23:
1612–26.
[60] Weiler C, Nerlich AG, Zipperer J, Bachmeier BE, Boos N. 2002 SSE
Award Competition in Basic Science: expression of major matrix
metalloproteinases is associated with intervertebral disc degradation
and resorption. Eur Spine J 2002;11:308–20.
[61] Rajnics P, Templier A, Skalli W, Lavaste F, Illes T. The association
of sagittal spinal and pelvic parameters in asymptomatic persons
and patients with isthmic spondylolisthesis. J Spinal Disord Tech
2002;15:24–30.
[62] Kawakami M, Tamaki T, Ando M, Yamada H, Hashizume H,
Yoshida M. Lumbar sagittal balance influences the clinical outcome
after decompression and posterolateral spinal fusion for degenera-
tive lumbar spondylolisthesis. Spine 2002;27:59–64.
[63] Harrison DE, Cailliet R, Harrison DD, Janik TJ. How do anterior/
posterior translations of the thoracic cage affect the sagittal lumbar
spine, pelvic tilt, and thoracic kyphosis? Eur Spine J 2002;11:287–93.
[64] Harrison DE, Jones EW, Janik TJ, Harrison DD. Evaluation of axial
and flexural stresses in the vertebral body cortex and trabecular bone
in lordosis and two sagittal cervical translation configurations with an
elliptical shell model. J Manipulative Physiol Ther 2002;25:391–401.
[65] Manns RA, Haddaway MJ, McCall IW, Cassar PV, Davie MW. The
relative contribution of disc and vertebral morphometry to the angle of
kyphosis in asymptomatic subjects. Clin Radiol 1996;51:258–62.
[66] Loder RT. Profiles of the cervical, thoracic, and lumbosacral spine in
children and adolescents with lumbosacral spondylolisthesis. J Spinal
Disord 2001;14:465–71.
[67] Itoi E. Roentgenographic analysis of posture in spinal osteoporotics.
Spine 1991;16:750–6.
[68] Harrison DE, Colloca CJ, Harrison DD, Janik TJ, Haas JW, Keller
TS. Anterior thoracic posture increases lumbar intervertebral disc
loading. Eur Spine J (in press). [Published on-line: DOI: 10.1007/
s00586-004-0734-0].
[69] McGill SM, Norman RW. Partitioning of the L4–L5 dynamic moment
into disc, ligamentous, and muscular components during lifting. Spine
1986;11:666–77.
[70] Nussbaum MA, Chaffin DB. Development and evaluation of a scalable
and deformable geometric model of the human torso. Clin Biomech
1996;11:25–34.
[71] Pearsall DJ, Reid JG, Livingston LA. Segmental inertia parameters
of the human trunk as determined from computer tomography. Ann
Biomed Eng 1996;24:198–210.
[72] Falhallah FA, Marras WS, Parnianpour M. An assessment of complex
loads during dynamic lifting tasks. Spine 1998;23:706–16.
[73] Davis JR, Mirka GA. Transverse-contour modeling of trunk muscle-
distributed forces and spinal loads during lifting and twisting. Spine
2000;25:180–9.