SPINE Volume 26, Number 24, pp 26922700

2001, Lippincott Williams & Wilkins, Inc.

Mechanical Properties of the Human Cervical Spine as

Shown by Three-Dimensional
LoadDisplacement Curves
Manohar M. Panjabi, PhD, DTech, Joseph J. Crisco, PhD, Anita Vasavada, PhD,
Takenori Oda, MD, Jacek Cholewicki, PhD, Kimio Nibu, MD, and Eon Shin, BA

Study Design. The mechanical properties of multilevel

human cervical spines were investigated by applying
pure rotational moments to each specimen and measuring multidirectional intervertebral motions.
Objectives. To document intervertebral main and coupled motions of the cervical spine in the form of load
displacement curves.
Summary of Background Data. Although a number of
in vivo and in vitro studies have attempted to delineate
normal movement patterns of the cervical spine, none
has explored the complexity of the whole cervical spine
as a three-dimensional structure.
Methods. Sixteen human cadaveric specimens (C0
C7) were used for this study. Pure rotational moments of
flexion extension, bilateral axial torque, and bilateral lateral bending were applied using a specially designed
loading fixture. The resulting intervertebral motions were
recorded using stereophotogrammetry and depicted as a
series of load displacement curves.
Results. The resulting load displacement curves were
found to be nonlinear, and both rotation and translation
motions were coupled with main motions. With flexion
extension moment loading, the greatest degree of flexion
occurred at C1C2 (12.3), whereas the greatest degree of
extension was observed at C0 C1 (20.2). With axial moment loading, rotation at C1C2 was the largest recorded
(56.7). With lateral bending moments, the average range
of motion for all vertebral levels was 7.9.
Conclusions. The findings of the present study are
relevant to the clinical practice of examining motions of
the cervical spine in three dimensions and to the understanding of spinal trauma and degenerative diseases.
[Key words: cervical spine, biomechanics, range of motion, coupled motion] Spine 2001;26:26922700

Clinicians have long recognized that degenerative

changes in the cervical spine can adversely affect its mobility.2,14,15,41 For example, a decrease in neck mobility
has been observed in elderly patients with degenerative
lesions of the vertebral endplates. As these lesions worsened, mobility was shown to become increasingly restricted.9,47,48 Therefore, accurate information regarding the
normal movement patterns of the cervical spine should
serve as the basis for understanding abnormal conditions.53
From the Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New
Haven, Connecticut.
Supported in part by the Yale University School of Medicine Fellowship Program and NIH Grant AR-42211.
Acknowledgment date: February 6, 2001.
Acceptance date: April 24, 2001.
Device status category: 1.
Conflict of interest category: 12.

2692

A number of in vivo studies have attempted to record

normal cervical motion patterns. The earliest in vivo investigations superimposed radiographs of the cervical
spine in full flexion and extension and measured the angles between reference points to determine the complete
ranges of motion.1,3,4,43 Advances in the technology of
measurement systems have allowed investigators to
gather more detailed and accurate data on normal cervical spine movement patterns.8 Studies have also reported
significant differences in the range of flexion extension
motion between healthy subjects and those suffering
from known cervical disorders.7,14,16 In vivo investigations have recently expanded in scope to include measurements of axial rotation and lateral bending in the
cervical spine.6,18,2325,45
However, the limitations of in vivo studies are many.
The in vivo loads that are applied to the spine by the
subject are unknown, thereby precluding calculations of
cervical spine flexibility and stiffness. Additionally, the
loads can vary, depending on the motivation of individual subjects. Slippage at the skin beneath measurement
devices also contributes to inaccuracies in the data. More
accurate in vivo measurements have been reported recently, but they use roentgen stereophotogrammetry,
which is an invasive technique.19,20,46 In vivo studies
also encounter difficulties defining the neutral posture in
living subjects.4,11,42 Thus, wide variation in in vivo
measurements of three-dimensional main and coupled
motions can be expected.
In vitro models, in contrast, permit more accurate
measurements of intersegmental motions and a more
controlled environment for studying the physical properties of the spine. The earliest in vitro studies of the
cervical spine measured the range of motion (ROM) of
human autopsy specimens in full flexion and extension
without any knowledge of the loads that were applied.17,22,25,49,50 This technique later evolved to record
both the forces applied to the spine and the resultant
motions.23,45,51 Lysell,23 in a well-known in vitro study
of the cervical spine, documented the intersegmental
movement patterns and ranges of motion in flexion, extension, bilateral axial torsion, and bilateral lateral
bending and recorded coupled rotations associated with
each of these movements. This experimental protocol
has been modified in recent in vitro studies by applying
pure moments of rotation instead of linear forces.12,31,34,36,37,40,42 This modification permits greater

Cervical Spine Mechanical Properties Panjabi et al 2693

control over the applied forces and is the most accurate

technique of measuring intersegmental motions.
Conspicuously lacking from the research literature is
an investigation providing a comprehensive overview of
three-dimensional motions in the cervical spine resulting
from pure moment applications. Although Moroney et
al27 reported the load displacement behavior of the cervical spine in response to compression, shear, bending,
and axial torsion, they performed their investigation using functional spinal units, i.e., two adjacent vertebrae
and their intervening soft tissues. We think it is advantageous for an in vitro study to use multilevel spine specimens to simulate in vivo conditions as closely as possible
and to measure the changes that occur at every vertebral
level. The purpose of the present study was to document
the biomechanical properties obtained from multilevel
cervical spine specimens by depicting intersegmental
main and coupled motions in the form of load
displacement curves. By measuring the intersegmental
movement patterns in flexion, extension, bilateral axial
torsion, and bilateral lateral bending, we hoped to
present kinematic data useful for understanding clinically important problems associated with cervical spine
disease and trauma.
Materials and Methods
Sixteen human, whole cervical spine specimens (one spine C0
C5, five spines C0 C6, two spines C0 C7, and eight spines
C2C7) were harvested fresh, packed in double plastic bags,
and stored at 20 degrees C until ready to be mounted and
tested. After radiographically excluding any specimens with
abnormalities beyond the normal age-related degenerative
changes, each specimen was dissected of all nonligamentous
soft tissue. The occiput and the caudal-most vertebra were
mounted in polyester resin casts (Plastic Padding, Gothenburg,
Sweden). Markers made of 1.5 mm-thick Plexiglas were affixed
to the anterior aspect of each vertebral body and the occiput.
Each marker contained at least three 0.8 mm-diameter steel
balls glued to the surface of the Plexiglas and distributed noncollinearly. Anteroposterior and lateral roentgenograms taken
of the specimens were used to establish an anatomic coordinate
system in each vertebra and to provide geometric relationships
between the coordinate system and the corresponding marker
steel balls. Thus, the steel balls served as points for stereophotogrammetric motion measurement, but the data were analyzed in the context of the bony anatomy of the spine.

Flexibility Test. The goal of this test was to determine the

three-dimensional physical properties of the spine specimen by
applying pure moments and measuring the intervertebral motions.38 A three-dimensional coordinate system was defined for
this purpose (Figure 1). The origin was the posteriorinferior
corner of the moving vertebral body. The positive y axis was
along the posterior wall of the body of the inferior vertebra
pointing up, the positive z axis was perpendicular to the y axis
pointing forward, and the positive x axis was perpendicular to
both y and z axes pointing left. We used six pure moments:
flexion (MX), extension (MX), left axial rotation (MY),
right axial rotation (MY), right lateral bending (MZ), and
left lateral bending (MZ). To generate a pure moment, ap-

Figure 1. The three-dimensional coordinate system of a moving

vertebra. The coordinate system is necessary to define the loads
applied and the motions measured. The origin is located at the
inferior-most point on the posterior wall of the body of the vertebra
in the midsagittal plane. The positive x axis is directed to the left
and perpendicular to the sagittal plane. The positive y axis is
directed superiorly, and the positive z axis is oriented anteriorly.
The broad arrows illustrate the pure moments: MX flexion;
MX extension; MY left axial torque; MY right axial
torque; MZ right bending; MZ left bending. The thin
circular arrows show the rotations: RX flexion; RX extension; RY left rotation; RY right rotation; RZ right
lateral bending; RZ left lateral bending. The thin straight
arrows show the translations: TX left; TX right; TY
superior; TY inferior; TZ anterior; TZ posterior.
(Adapted from Panjabi MM, White AA. Physical properties and
functional biomechanics of the spine. In: White AA, Panjabi MM,
eds. Clinical Biomechanics of the Spine, 2nd ed. Philadelphia:
Lippincott, 1990.)
propriate equal and opposing forces (F) were applied to the
pulleys mounted on top of the specimen (Figure 2). Design of
the loading arrangement allowed completely unconstrained
spinal motions. Each moment was applied in three equal load
increments to a maximum value of 1.00 Nm, which was judged
to be sufficient to produce physiologic motions but small
enough not to injure the specimen. Thirty seconds of creep were
allowed at each load step, and three load unload cycles were
used to precondition the specimen and minimize viscoelastic
effects. Stereophotographs were taken of the markers on the
specimen, only on the third load cycle. The photographs were
digitized, the relative three-dimensional motions of each vertebra were calculated, and load displacement curves were plotted. The neutral zone (NZ) and ROM for each level were determined from the measured load displacement curves from
the third load cycle. The NZ was defined as the rotation from
the neutral position at the beginning of the third load cycle. The
ROM was defined as the rotation from the neutral position to
the end position at the maximum load of 1.00 Nm (Figure 3).
An upward force applied at the center of gravity of the mount
balanced the weight of the upper mount, approximately 15 N.

Accuracy of the Measurements. A rigid body with four

vertebral markers, each containing several steel balls, was used
to estimate the precision of our motion measurement system.29,38 With the markers fixed in space, the rigid-body Euler

2694 Spine Volume 26 Number 24 2001

Figure 3. The neutral zone (NZ) and range of motion (ROM) using a
load displacement curve for flexion extension rotation. In this example (1) the NZ for flexion rotation, whereas (2) the NZ for
extension rotation. Likewise, (3) the ROM for flexion rotation with
a maximum load of 1.00 Nm, whereas (4) the ROM for extension
rotation with a maximum load of 1.00 Nm. The NZ and ROM were
obtained for each of the 6 degrees of freedom of motion for all seven
levels.
0.35 mm for anteroposterior translation, and 0.22 mm for
lateral translation.
Figure 2. The experimental setup for the flexibility test. The cephalic and caudal end-vertebrae are mounted in polyester resin
mounts. The bottom mount is fixed to the test table while the top
mount carries a loading fixture. The loading fixture is designed to
apply pure moments to the specimens via three round discs: two
vertically oriented and one horizontally oriented. The vertical pulleys are used to apply moments of flexion, extension, and bilateral
bending. The horizontal pulley is used to apply axial torque.
angles and translations were calculated from 10 stereophotogrammetric images. The precision (the standard deviation of
the measurements) was found to be dependent on the plane of
rotation as follows: 0.60 for flexion extension, 0.33 for
lateral bending, and 0.17 for axial rotation. For translations
the precisions were as follows: 0.43 mm for axial translation,

Statistical Analysis. Means and standard deviations for the

ROM were computed. Motions at different cervical levels were
compared under identical loads. Each applied moment produced three rotations and three translations of each vertebra.
The vertebral motions of the different levels under each maximum applied moment (flexion, extension, leftright axial torsion, and rightleft lateral bending) were compared with use of
a factorial analysis of variance. We used the Fisher least significant difference post hoc test at a confidence level of 95% to
determine which levels were significantly different from one
another for each applied moment.
Results
The three-dimensional behavior of the entire cervical
spine is presented in the form of average load

Table 1. Neutral Zones Measured From Multidirectional Flexibility Testing ()

Flexion
Extension
Axial rotation
Lateral bending

C0C1

C1C2

C2C3

C3C4

C4C5

C5C6

C6C7

3.3 1.8
13.9 4.1
2.5 1.6
3.6 1.5

4.6 2.4
8.7 6.7
39.6 7.5
2.4 1.2

0.7 0.6
1.0 0.7
1.1 0.5
4.1 1.1

0.9 0.9
1.7 1.7
1.6 0.5
4.4 1.2

1.6 1.3
1.9 1.8
2.4 0.6
4.4 1.1

1.8 1.3
2.1 2.0
1.7 0.5
3.0 1.1

1.0 0.7
1.3 1.0
0.6 0.3
2.2 1.0

Values are mean standard deviation. Values for axial rotation and lateral bending summate both right and left sides.

Table 2. Ranges of Motion Measured From Multidirectional Flexibility Testing ()

Flexion
Extension
Axial rotation
Lateral bending

C0C1

C1C2

C2C3

C3C4

C4C5

C5C6

C6C7

7.2 2.5
20.2 4.6
9.9 3.0
9.1 1.5

12.3 2.0
12.1 6.5
56.7 4.8
6.5 2.3

3.5 1.3
2.7 1.0
3.3 0.8
9.6 1.8

4.3 2.9
3.4 2.1
5.1 1.2
9.0 1.9

5.3 3.0
4.8 1.9
6.8 1.3
9.3 1.7

5.5 2.6
4.4 2.8
5.0 1.0
6.5 1.5

3.7 2.1
3.4 1.9
2.9 0.8
5.4 1.5

Values are mean standard deviation. Values for axial rotation and lateral bending summate both right and left sides.

Cervical Spine Mechanical Properties Panjabi et al 2695

Figure 4. Load displacement

curves at each of the seven levels due to the application of flexion and extension moments. The
main motion, defined as the rotation in the direction of the applied moment, is indicated by the
heavier line. The standard deviations are plotted for the main motion only. Also shown are the
coupled motions, defined as all
rotations and translations other
than the main rotation. Rotations
and translations that were not
significantly different from zero
are not plotted. See Figure 1 for
nomenclature.

displacement curves and standard deviation bars. The six

moments are grouped into three pairs: flexion
extension, leftright axial torsion, and leftright lateral
bending. This results in 21 separate graphs (three moment pairs at seven levels), each of which contains six
curves representing 6 degrees of freedom intersegmental
motion (Figures 4, 5, and 6).

With extension moment loading, the largest main motion (extension rotation) occurred at C0 C1 (20.2),
compared with all other levels (P 0.005). Extension
moment loading also produced coupled translations in
the sagittal plane, which were generally equivalent to
those induced under flexion moment loading except that
they were directed posteriorly.

FlexionExtension Moment Loading

In flexion moment loading, the main motion (flexion rotation) at the cephalic cervical levels (C0 C1 and C1
C2) was significantly greater than the motions of the
caudal vertebral levels (P 0.005) (Figure 4). Flexion
rotations at C0 C1 and C1C2 averaged 7.2 and 12.3
respectively, with a moment of 1.00 Nm. Coupled translations in the sagittal plane were directed anteriorly and
occurred at all levels decreasing in magnitude caudally.

Axial Torsion Moment Loading

With the application of right and left axial torsion moments, the main motion (axial rotation) was greatest at
C1C2. The axial rotation averaged 28.4 to each side at
this level, which was significantly greater than the mean
rotations at C0 C1 (4.9) and the caudal levels, which
exhibited a mean axial rotation of 2.6 at each intersegmental level (P 0.001) (Figure 5). Significant coupled

2696 Spine Volume 26 Number 24 2001

Figure 5. Load displacement

curves at each of the seven levels due to the application of right
and left axial torques. The main
motion is indicated by the
heavier line. The standard deviations are plotted for the main motion only. Also shown are the
coupled motions, defined as all
rotations and translations other
than the main rotation. Rotations
and translations that were not
significantly different from zero
are not plotted. Note that the y
axis scale for C1C2 is different
from the other graphs. See Figure 1 for nomenclature.

extension rotation was observed at C0 C1 (11.7) and

C1C2 (3.5). In the lower region some coupled flexion
was observed. At C0 C1 and C1C2, the lateral bending
was to the opposite side of the applied torque (1.8 and
3.1, respectively), i.e., coupled right lateral bending was
associated with left axial torque, and vice versa. However, the remainder of the intersegmental levels exhibited
coupled lateral bending toward the same side as the applied axial torque.
Coupled translations occurred in the sagittal plane at
C0 C1 and C1C2 and were directed posteriorly with no
regard to the direction of the applied torque. The posterior
translations averaged 11.4 mm. Small translation average
about 1.2 mm and were observed at other levels.

Lateral Bending Moment Loading

With the application of left and right lateral bending
moments, the main motion (lateral bending rotation)
was found to be greatest at C2C3, averaging 4.8 to one
side, although this value was not significantly different
from the lateral bending at other vertebral levels (P
0.05) (Figure 6). Coupled flexion extension rotations
were small at all vertebral levels (1.2) except at C1
C2, where a mean extension of 3.0 was observed. Coupled axial rotations were to the same side as the applied
lateral bending moment. Coupled axial rotation was
largest at C1C2 (5.3), followed by C0 C1 (2.4). At all
other levels the coupled axial rotations were 1.8. Lateral translations were observed to be in the same direc-

Cervical Spine Mechanical Properties Panjabi et al 2697

Figure 6. Load displacement

curves at each of the seven levels due to the application of right
and left lateral bending moments.
The main motion is indicated by
the heavier line. The standard
deviations are plotted for the
main motion only. Also shown
are the coupled motions, defined
as all rotations and translations
other than the main rotation. Rotations and translations that
were not significantly different
from zero are not plotted. See
Figure 1 for nomenclature.

tion as the applied lateral bending moment, i.e., the right

lateral bending produced right translation. Coupled lateral translations were found to be greatest at C1C2 (3.2
mm). All other translation motions ranged from 1.2 to
2.5 mm.
NZ and ROM
The NZ and ROM for each intersegmental level were
calculated from the load displacement data. The largest
NZ for flexion occurred at C1C2 (4.6), for extension
at C0 C1 (13.9), for axial torsion at C1C2 (39.6),
and for lateral bending at C3C4 and C4 C5 (4.4 at
both levels). In general, the NZ tended to be largest
where the cervical spine exhibited the greatest ROM (Tables 1 and 2).

Discussion
The present study examined the three-dimensional intersegmental motions of the cervical spine by documenting
movements associated with flexion, extension, bilateral
axial torsion, and bilateral lateral bending loads. The
findings of this study are relevant to the clinical practice
of examining motions of the cervical spine in three dimensions and to the understanding of spinal trauma and
degenerative diseases.
Weaknesses and Strengths of the Multilevel
In Vitro Model
There are limitations to studying cervical spine motions
with an in vitro model. An obvious criticism is the lack of

2698 Spine Volume 26 Number 24 2001

Table 3. Comparison of Average Ranges of Motion of Flexion Extension Rotation ()

In vivo studies
Penning (1978)
Dvorak et al (1988)*5
Dvorak et al (1988)*5
Lind et al (1989)21
Dvorak et al (1993)
Holmes et al (1994)
In vitro studies
Panjabi et al (1986)39
Panjabi et al (1988)
Moroney et al (1988)
Present study

C0C1

C1C2

C2C3

C3C4

C4C5

C5C6

C6C7

30.0

14.0

30.0
12.0
15.0
13.0
14.1

12.0
10.0
12.0
10.0
12.0
7.7

18.0
15.0
17.0
14.0
17.2
13.5

20.0
19.0
21.0
16.0
21.1
17.9

20.0
20.0
23.0
15.0
22.6
15.6

15.0
19.0
21.0
11.0
21.4
12.5

24.5

27.4

22.4

24.4

9.9

9.7
6.2

9.9

9.7
7.7

9.9

9.7
10.1

9.9

9.7
9.9

9.9

9.7
7.1

* Dvorak et al studied active (upper values) and passive (lower values) examinations of the cervical spine range of motion in a single study.

musculature associated with in vitro studies. The cervical

spine musculature has been shown to exert significant
stabilizing forces on the spine.28 This function was assumed by the balance weight in the present study. But the
lack of musculature did not result in excessive motions,
as seen by the comparison of our study and prior in vivo
results.
We used multilevel specimens to approximate in vivo
conditions as closely as possible. The study by Moroney
et al,27 although comprehensive in scope, was limited by
the use of isolated functional spinal units, two adjacent
vertebrae, and the connecting ligamentous tissues. Because the spine is a structure composed of multiple functional spinal units in series, combining the behaviors of
individual functional spinal units may approximate its
total behavior. However, such studies violate the continuity of longitudinal structures, such as the anterior and
posterior longitudinal ligaments, and may not represent
the normal lordosis of the cervical spine. In using a mul-

tilevel in vitro model, we were able to simulate the in vivo

condition more closely.
Comparisons With Previous Studies
To test the validity of our experimental model, we compared our results with data from other studies, both in
vivo and in vitro. For flexion extension rotation (Table
3), our values of 27.4 and 24.4 at C0 C1 and C1C2,
respectively, were in between the values reported by Panjabi et al35 (24.5, 22.4) and Penning43 (30.0, 30.0).
We concede that the intersegmental motions measured at
C1C2 of the present study were larger than most of the
values from the in vivo studies, but in vitro studies cannot simulate the presence of the chin in flexion motion
and skin folds in extension motion, which can compromise the full range of flexion extension motion. In axial
rotation the results of our data, as well as those of Lysell,23 Panjabi et al,35 and Moroney et al,27 closely approximated those of the in vivo studies (Table 4). Al-

Table 4. Comparison of Average Ranges of Motion of Right Left Axial Torsion Rotation ()
C0C1

C1C2

C2C3

C3C4

C4C5

C5C6

C6C7

8.0
2.0

8.0

83.0
81.0

76.0

6.0
6.0
7.4
9.0

13.0
13.0
5.8
7.0

13.4
13.6
4.2
7.0

14.0
13.8
5.4
6.0

10.8
10.8
6.4
6.0

14.6

9.9

77.8

56.7

6.0

3.7
3.2

9.8

3.7
5.1

10.3

3.7
6.8

8.0

3.7
5.1

5.7

3.7
2.9

In vivo studies
Dvorak et al (1987)
Penning and Wilmink (1987)
Mimura et al (1989)
Iai et al (1993)
In vitro studies
Lysell (1969)
Panjabi et al (1988)
Moroney et al (1988)
Present study

Table 5. Comparison of Average Ranges of Motion of Right Left Lateral Bending Rotation ()

In vivo studies
Penning (1978)
In vitro studies
Panjabi et al (1988)
Moroney et al (1988)
Present study

C0C1

C1C2

C2C3

C3C4

C4C5

C5C6

C6C7

5.0

6.0

6.0

6.0

6.0

6.0

11.0

9.1

13.4

6.5

9.4
9.5

9.4
9.1

9.4
9.3

9.4
6.5

9.4
5.4

Cervical Spine Mechanical Properties Panjabi et al 2699

though our measured ROM for C6 C7 was smaller than

the values determined by in vivo studies, the presence of
powerful musculature anchored to the spinous process of
C7 may have accounted for this difference. In lateral
bending our values were in between those of Penning,43
Panjabi et al,35 and Moroney et al27 (Table 5).
Main and Coupled Motions in the Cervical Spine
The phenomenon of coupled motions has been well documented. In vitro, it occurs in the thoracic and lumbar
spines,34,38 but it is most dramatic in the cervical
spine.23,27 Clinically, two kinds of coupled motions are
especially well known in the cervical spine: axial rotation
in the same direction as the applied lateral bending and
lateral bending in the same direction as the applied axial
rotation.10,26,44,52 Our study supports these data and
documents other coupled rotations as well as coupled
translations.
With flexion and extension moment the greatest main
rotation was extension at C0 C1 (Figure 4). The coupled motions were unremarkable, except for translations
in the sagittal plane. With torsional loading the largest
main rotation (axial rotation) was observed at C1C2, as
expected. The largest coupled sagittal plane rotation (extension) and translation (posterior) occurred at C0 C1,
irrespective of the direction of the torsion (Figure 5). The
extension decreased and changed to flexion at about C4
C5. Although these coupled sagittal plane rotations were
small, the findings suggest that the cervical spine bears a
load on the vertebral body on the posterior edge above
C4 C5 and on the anterior edge below it. These same
findings were reported in an in vivo cervical spine study
by Mimura et al26 and may explain why osteophytes
have been observed to develop anteriorly at segments
below C5 and posteriorly above it.13 The coupled lateral
bending, in the same direction as the axial rotation, was
present at all levels. Finally, the lateral bending moment
elicited coupled axial rotation in the same direction as
the lateral bending (Figure 6). There were also coupled
extensions at C0 C1 and C1C2, again without any regard to the direction of the applied lateral bending
moment.
Neutral Zones in the Cervical Spine
It has been theorized that the NZ may be a more sensitive
parameter of spinal instability than the ROM.30,32,33
Our study, for the first time, provides comprehensive NZ
data for the entire cervical spine and in all three planes of
motion. Such a normative database can be useful for
determining abnormal NZ in disease and trauma. The
NZ is a region of low stiffness of the osteoligamentous
spine. Thus, in this part of the ROM little ligamentous
resistance is offered to external loads. Thus, the role of
muscle forces in stabilizing the spine within the NZ is
very important. We found the largest NZ to be in axial
rotation at C1C2. This is most likely the result of special
anatomic features, e.g., the lack of an intervertebral disc,
nearly flat transversally oriented facet articulations, and
loose capsular ligaments. The role of muscles at this level

may therefore be important. There are a large number of

muscles attached to C2, which supports this concept.
Muscular dysfunction at this level may lead to clinical
problems.
Key Points
In vivo studies currently lack the capacity for
documenting three-dimensional intervertebral motions. Comprehensive in vitro studies, such as the
present one, are the only studies that can provide
such a data set.
With flexion extension moment loading, coupled translations in the sagittal plane were anteriorly directed for flexion and posteriorly directed
for extension at all intersegmental levels.
With axial torque loading the cervical spine exhibited the largest main rotation at C1C2 and the
largest coupled extension at C0 C1. Coupled lateral bending was present at all levels, in the same
direction as the applied torque.
Lateral bending moment elicited the largest main
lateral bending in the middle region and at C0 C1.
Coupled axial rotation was in the same direction as
the lateral bending at all intersegmental segments.
There were also coupled extensions at C0 C1 and
C1C2.

References
1. Bakke S. Rontgenologische beobachtungen uber die bewegungen der
halswirbelsaule. Acta Radiol Suppl (Stockh) 1931;13:76.
2. Bland J. Disorders of the Cervical Spine. Philadelphia: Saunders, 1987.
3. Buetti-Ba uml C. Funktionelle roentgendiagnostik der halswirbelsaule. Arch
Atlas 1954;70:19 23.
4. De Seze C, Dijian A, Abdelmoula M. Etude radiologique de la dynamique
cervical dans le plain sagittal. (Une contribution radiophysiologique a letude
pathogenique des artheoses cervicales). Rev Rhum 1951;18:37 46.
5. Dvorak J, Froehlich D, Penning L, et al. Functional radiographic diagnosis of
the cervical spine: flexion/extension. Spine 1988;13:748 55.
6. Dvorak J, Hayek J, Zehnder R. CTfunctional diagnostics of the rotatory
instability of the upper cervical spine. Part 2. An evaluation on healthy adults
and patients with suspected instability. Spine 1987;12:726 31.
7. Dvorak J, Panjabi MM, Grob D, et al. Clinical validation of functional
flexion/extension radiographs of the cervical spine. Spine 1993;18:120 7.
8. Dvorak J, Panjabi MM, Novotny JE, et al. In vivo flexion/extension of the
normal cervical spine. J Orthop Res 1991;9:828 34.
9. Fenlin JM. Pathology of degenerative disease of the cervical spine. Orthop
Clin North Am 1971;2:371 87.
10. Fielding JW. Cineroentgenography of the normal cervical spine. J Bone Joint
Surg Am 1957;39:1280 8.
11. Fineman S, Borrelli FJ, Rubinstein BM, et al. The cervical spine: transformation of the normal lordotic pattern into a linear pattern in neutral posture.
J Bone Joint Surg Am 1963;45:1179.
12. Goel VK, Clark CR, McGowan D, et al. An in vitro study of the kinematics
of the normal, injured and stabilized cervical spine. J Biomech 1984;17:363
76.
13. Hayashi K, Miyagawa S, Yabuki T, et al. A morphological study of the
cervical spondylosis: mode of development of osteophytes. Seikeigeka 1983;
34:8639.
14. Hino H, Abumi K, Kanayama M, et al. Dynamic motion analysis of normal
and unstable cervical spines using cineradiography: an in vivo study. Spine
1999;24:163 8.
15. Hirsch C, Schajowicz F, Galante JO. Structural changes in the cervical spine.
Acta Orthop Scand 1968;39:303.
16. Holmes A, Wang C, Han ZH, et al. The range and nature of flexionextension motion in the cervical spine. Spine 1994;19:250510.

2700 Spine Volume 26 Number 24 2001

17. Hughes AW. Die drehbewegungen der menschlichen wirbelsa ule und die
sogenannten musculi rotatores (Theile). Arch Anat Entwickl-Gesch. 265,
1892.
18. Iai H, Moriya H, Goto S, et al. Three-dimensional motion analysis of the
upper cervical spine during axial rotation. Spine 1993;18:2388 92.
19. Johnsson R, Selvik G, Stromqvist B, et al. Mobility of the lower lumbar spine
after posterolateral fusion determined by roentgen stereophotogrammetric
analysis. Spine 1990;15:34750.
20. Johnsson R, Axelsson P, Gunnarsson G, et al. Stability of lumbar fusion with
transpedicular fixation determined by roentgen stereophotogrammetric
analysis. Spine 1999;24:68790.
21. Lind B, Sihlbom H, Nordwall A, et al. Normal range of motion of the cervical
spine. Arch Phys Med Rehabil 1989;70:6925.
22. Lovett RW. The mechanism of the normal spine and its relation to scoliosis.
Boston Med Surg J 1905;153:349 58.
23. Lysell E. Motion in the cervical spine. Acta Orthop Scand Suppl 1969;123:
1 61.
24. McClure P, Siegler S, Nobilini R. Three-dimensional flexibility characteristics of the human cervical spine in vivo. Spine 1998;23:216 23.
25. Meyer GH. Die statik und mechanik der menschlichen knochengeru stes.
Leipzig: Engelmann, 1873.
26. Mimura M, Moriya H, Watanabe T, et al. Three-dimensional motion analysis of the cervical spine with special reference to the axial rotation. Spine
1989;14:11359.
27. Moroney SP, Schultz AB, Miller JAA, et al. Load-displacement properties of
lower cervical spine motion segments. J Biomech 1988;21:769 79.
28. Nolan JP, Sherk HH. Biomechanical evaluation of the extensor musculature
of the cervical spine. Spine 1988;13:9 11.
29. Oxland TR, Lin RM, Panjabi MM. Three-dimensional mechanical properties of the thoracolumbar junction. J Orthop Res 1992;10:573 80.
30. Oxland TR, Panjabi MM. The onset and progression of spinal injury: a
demonstration of neutral zone sensitivity. J Biomech 1992;10:116572.
31. Panjabi MM. Biomechanical evaluation of spinal fixation devices: I. A conceptual framework. Spine 1988;13:1129 34.
32. Panjabi MM. The stabilizing system of the spine: I. Function, dysfunction,
adaptation, and enhancement. J Spinal Disord 1992;4:3839.
33. Panjabi MM. The stabilizing system of the spine: II. Neutral zone and instability hypothesis. J Spinal Disord 1992;4:390 6.
34. Panjabi MM, Brand RA, White AA. Mechanical properties of the human
thoracic spine: as shown by three-dimensional load-displacement curves.
J Bone Joint Surg Am 1976;58:64252.
35. Panjabi MM, Dvorak J, Duranceau J, et al. Three-dimensional movements of
the upper cervical spine. Spine 1988;13:726 30.
36. Panjabi MM, Krag MH, White AA, et al. Effects of preload on load displacement curves of the lumbar spine. Orthop Clin North Am 1977;8:18192.
37. Panjabi MM, Oda T, Crisco JJ, et al. Posture affects motion coupling patterns of the upper cervical spine. J Orthop Res 1993;11:52536.
38. Panjabi MM, Oxland TR, Yamamoto I, et al. Mechanical behavior of the

39.

40.
41.
42.
43.
44.
45.
46.
47.
48.

49.
50.

51.
52.
53.

human lumbar and lumbosacral spine as shown by three-dimensional loaddisplacement curves. J Bone Joint Surg Am 1994;76:41324.
Panjabi MM, Summers DJ, Pelker RR, et al. Three-dimensional loaddisplacement curves due to forces on the cervical spine. J Orthop Res 1986;
4:152 61.
Panjabi MM, White AA, Johnson RM. Cervical spine mechanics as a function of transection of components. J Biomech 1975;8:32736.
Panjabi MM, White AA, Keller D, et al. Stability of the cervical spine under
tension. J Biomech 1978;11:189 97.
Panjabi MM, Yamamoto I, Oxland TM, et al. How does posture affect
coupling in the lumbar spine? Spine 1989;14:100211.
Penning L. Normal movements of the cervical spine. Am J Roentgenol 1978;
30:31726.
Penning L, Wilmink JT. Rotation of the cervical spine. Spine 1987;12:
732 8.
Sedlin ED, Hirsch C. Factors affecting the determining of the physical properties of femoral cortical bone. Acta Orthop Scand 1966;37:29 48.
Selvik G. Roentgen stereophotogrammetry: a method for the study of the
kinematics of the skeletal system. Acta Orthop Scand (suppl 232) 1989.
Ten Have HA, Eulderink F. Degenerative changes in the cervical spine and
their relationship to its mobility. J Pathol 1980;132:13359.
Ten Have HA, Eulderink F. Mobility and degenerative changes of the aging
cervical spine: a macroscopic and statistical study. Gerontology 1981;27:42
50.
Volkmann AW. Von der drehbewegung des ko rpers. Virchows Arch Path
Anat 1872;56:467.
Weber EH. Anatomisch-physiologische untersuchung u ber einige einrichtungen im mechanismus der menschlichen wirbelsa ule. Arch Anat Physiol
1827;240.
Werne S. Studies in spontaneous atlas dislocation. Acta Orthop Scand (suppl
23), 1957.
White AA, Panjabi MM. The basic kinematics of the human spine: a review
of past and current knowledge. Spine 1978;3:1220.
White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd ed. Philadelphia: Lippincott, 1990.

Address correspondence to
Manohar M. Panjabi, PhD
Biomechanics Research Laboratory
Department of Orthopaedics and Rehabilitation
Yale University School of Medicine
P.O. Box 208071
New Haven, CT 06520-8071
E-mail: manohar.panjabi@yale.edu