J Oral Maxillofac Surg

60:546-551, 2002

Biomechanical Response of Retrodiscal

Tissue in the Temporomandibular Joint
Under Compression
Eiji Tanaka, DDS, PhD,* Rodrigo del Pozo, DDS,
Masaru Sugiyama, DDS, PhD, and Kazuo Tanne, DDS, PhD
Purpose: The present study was conducted to investigate the biomechanical response of bovine
retrodiscal tissue of the temporomandibular joint (TMJ) in compression.
Patients and Methods: Using 10 retrodiscal tissues obtained from 10 cattle, the viscoelastic response
of the retrodiscal tissue was evaluated by means of stress-strain analyses. These compressive strains were
produced at a high strain rate and were kept constant during 5 minutes for stressrelaxation.
Results: Although the stress-strain relationship in the retrodiscal tissue was essentially nonlinear
represented by a quadratic or power function of strain, a linear model could reasonably represent its
elastic property. In this case, the instantaneous and relaxed moduli were 1.54 and 0.21 MPa, respectively.
The stressrelaxation curve showed a marked drop in load during the initial 10 seconds, and the stress
reached a steady nonzero level. Furthermore, when using Kelvins model, a satisfactory agreement can
be obtained between the experimental and theoretical stressrelaxation curves.
Conclusion: It is concluded that bovine retrodiscal tissue has a great capacity for energy dissipation
during stressrelaxation, although it has little or no function to pull the articular disc back.
2002 American Association of Oral and Maxillofacial Surgeons
J Oral Maxillofac Surg 60:546-551, 2002

The temporomandibular joint (TMJ) is well known

as one of the load-bearing organs in the human
body. The loading placed on the TMJ is absorbed by
the TMJ disc and articular cartilage, whereas moderate loading plays an important role for TMJ
growth and development.1-3 Moreover, the retrodiscal tissue is continuous with the posterior band of

the disc, and its organization is a function of condylar position.4

The retrodiscal tissue in the TMJ contains collagen, elastic bers, and numerous blood vessels and
nerves5,6 and probably is less stiff than the disc. When
the jaws are closed, the retrodiscal tissue is formed
into a thick sheet behind the condyle. When the jaws
are widely opened, the retrodiscal tissue is expanded
in all directions.4 Although negative pressure appears
within the retrodiscal tissue as the condyle leaves the
glenoid fossa, the pressure within the attachment
becomes positive on jaw closure.4,7 Therefore, the
primary role of the retrodiscal tissue is also to provide
a volumetric compensatory mechanism for pressure
equilibration.8
Connective tissues such as skin, articular disc, articular cartilage, suture, and tendon, have been shown to
exhibit viscoelastic properties.9-14 The viscoelasticity
of the connective tissues is dependent on the collagenous portion of the matrix and the osmotic afnity of
the loaded site for displaced water.15 Of its components, collagen bers are associated with maintaining
the shape, and elastin may function in the restoration
of the tissue from load removal.16 Extracellular matrix
proteoglycans are considered to be responsible for
maintaining the resilience of the tissue and for resistance to compression force.17,18 Assessments of the

*Associate Professor, Department of Orthodontics, Hiroshima

University Faculty of Dentistry, Hiroshima, Japan.
Graduate Student, Department of Orthodontics, Hiroshima University Faculty of Dentistry, Hiroshima, Japan.
Associate Professor, The Second Department of Oral and Maxillofacial Surgery, Hiroshima University Faculty of Dentistry, Hiroshima, Japan.
Professor and Chair, Department of Orthodontics, Hiroshima
University Faculty of Dentistry, Hiroshima, Japan.
This study was supported by Grant 12771280 for Science Research from the Ministry of Education, Science and Culture in
Japan.
Address correspondence and reprint requests to Dr Tanaka:
Department of Orthodontics, Hiroshima University Faculty of Dentistry, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8553, Japan; e-mail:
etanaka@hiroshima-u.ac.jp
2002 American Association of Oral and Maxillofacial Surgeons

0278-2391/02/6005-0012$35.00/0
doi:10.1053/joms.2002.31853

546

547

TANAKA ET AL

viscoelasticity have been conducted for connective

tissues and have contributed to understanding the
biomechanical response during function. However,
the retrodiscal tissue of the TMJ was investigated in
only the histologic and morphologic aspects but not
yet in the biomechanical aspect.
The present study was designed to evaluate the
viscoelastic properties and stressrelaxation of the
bovine retrodiscal tissue of the TMJ in compression.
The aim was to determine the extent to which the
viscoelasticity could be reproduced by a simple linear
model.

Materials and Methods

The experiments were carried out on 10 retrodiscal
tissues of the TMJ collected from 10 adult cattle (10
years old). The retrodiscal tissues were obtained at a
local abattoir (Japan Agriculture, Hiroshima, Japan).
The retrodiscal tissues, including the TMJ disc, were
carefully separated and immediately immersed in 154
mmol/L NaCl. In all of the cattle, the retrodiscal tissues with the TMJ disc were visually normal, and a
serial of compression tests were conducted within 6
hours after resection.
In the TMJ, the retrodiscal tissue has 3 parts: temporal, condylar, and intermediate.5 Of the 3, the temporal part extends from the back of the posterior
band to become attached to the anterior face of the
postglenoid process, and the condylar part is a thin
sheet that extends from the posteroinferior corner of
the posterior band to the inferior margin of the condylar articular surface. The intermediate part is contiguous with the temporal and condylar parts. In this
study, the specimens were dissected from temporal
and condylar parts of the retrodiscal tissue.
By using a cylindrical stainless-steel tissue punch of
approximately 6.86 mm in diameter, an experimental
specimen was derived from each retrodiscal tissue.
The actual specimens were 6.90 0.05 mm in diameter and 2.2 0.3 mm in thickness (mean SD).
Compression was applied to the specimens up
to the specied strain in a rigid frame by means
of a testing machine (EZ-test-500N; Shimadzu Corp,
Tokyo, Japan), and the compressive forces and the
distance between the compression plates were
measured instantaneously. After 5 minutes of preconditioning for each specimen by use of 0% to 20%
strains, the test of serial stressrelaxation was conducted on the 10 specimens. In a preliminary experiment, a recovery time of 5 minutes was determined
to be sufcient for the specimen to show reproducible stress-strain curves. Then, 10% and 20% strains
were applied as the stressrelaxation test. Through
the serial experiments, the retrodiscal tissues were
bathed in 154 mmol/L NaCl solution, and we were

able to keep them free of water displacement. These

compressive strains were produced at a high strain
rate of approximately 50% per second (1.04 mm/s)
and were kept constant during the 5-minute period
for stressrelaxation. To investigate the basic viscoelastic characteristics of the retrodiscal tissue, regression analyses were carried out for the elastic modulus E0 at the onset of the strain application and the
modulus E300 at the end of each 5-minute relaxation.
Furthermore, the time-dependent stressrelaxation
curves were obtained at 2 levels of strains, and a
parameter identication study was conducted with a
mechanical model to reproduce the time-dependent
stressrelaxation of the retrodiscal tissue under compression as a linear viscoelastic material.

Results
Figure 1 shows the stress-strain plots at the onset of
the stress application from the 10 specimens. The
stress-strain relation was nonlinear, and these stressstrain plots were best represented as a quadratic function. For instance, a regression analysis of a quadratic
function, a2 b, gave coefcients of a 2.75,
b 1.15 with the correlation coefcient of r 0.91.
Figure 2 shows the relaxed stress-strain plots at t
300 seconds, and a regression analysis of a power
function, ab, gave coefcients of a 0.66, b
1.67 with the correlation coefcient of r 0.95. For
the retrodiscal tissues, quadratic functions result in
the better t for the stress-strain with r 0.91 to 0.95.
With a linear regression model for the stress-strain
relation, however, the correlation coefcients were
r 0.90 for t 0 seconds and r 0.91 for t 300
seconds. These values of correlation coefcients with
linear regression models were almost similar to those
of nonlinear functions. This assumption of linearity
allows for the calculation of the elastic constants of

FIGURE 1. Experimental stress-strain relations under compression at

t 0 seconds. A curve and a dotted line are obtained from regression
analysis.

548

BIOMECHANICAL RESPONSE OF RETRODISCAL TISSUE

FIGURE 2. Experimental relaxed stress-strain relations under compression at t 300 seconds. A curve and a dotted line are obtained from
regression analysis.

FIGURE 4. Mechanical representation of Kelvins model consisting of

Maxwell model and another linear spring.

the retrodiscal tissue, the most basic biomechanical

characteristic. In this case, the elastic moduli E0 and
E300 were 1.54 and 0.21 MPa, respectively.
It was observed that regardless of the strain applied, the time-dependent stressrelaxation curves
showed a marked drop in stress during the initial 10
seconds of relaxation, as shown in Figure 3. The stress
decay basically ceased after 60 seconds, and the stress
reached a steady nonzero level as the relaxed stress
after a 5-minute relaxation. To explore the relationship between the experimental observations and the
theoretical stressrelaxation curve, Kelvins model
(Fig 4) of basic linear viscoelasticity was used. This
model has been commonly called the standard linear
model because it represents the most fundamental
relationship, including the load, the deection, and
their rst derivatives.19 Furthermore, it is an arrangement of a Maxwell model, which is a serial combination of linear spring and dashpot, and another linear

spring in parallel. The relationship between the stress

and the strain at the point of loading is formulated
as

E R
(1)
where and are time constants that mean the
relaxation times of stress and deection, respectively.
When an instantaneous strain (0) 0 is applied
instantaneously at the initial time of t 0,
0 E R0
(2)
and this equation gives as the initial condition of
equation 1. If the strain is kept constant at (t) 0
for t 0, the time-dependent behavior of the stress
(t) can be written as follows:
t
(t) ER 1 1
e 0 for t 0
(3)

where ER represents the relaxed elastic constant. The

dashpot is completely relaxed as time tends to innity. The stress-strain relation then becomes that of 2
parallel springs:
E R0
(4)
From the initial condition of equation 2, another elastic constant E0 is dened as
0 E 00
(5)
where

(6)
E 0 ER

Here, as the stress is almost fully relaxed at t 300

seconds, the relaxed elastic modulus ER can be represented by E300. Therefore, all the elastic constants
can be determined by the elastic moduli of the bovine
retrodiscal tissue, if the stress-strain relationships are
represented by a linear model. Then, by the leastsquares method, the time constants and were
determined to be 11.6 and 84.9 seconds, respectively.
Figure 5 shows the results of stressrelaxation tests at

FIGURE 3. Experimental stressrelaxation curves with points are

mean SD ( strain).

TANAKA ET AL

FIGURE 5. Experimental stressrelaxation plots obtained from retrodiscal tissues (mean 1 SD) with a theoretical curve. The theoretical
curve was calculated from the linear regression model with the time
constants and .

10% constant strain. A satisfactory agreement can be

obtained between the theoretical stressrelaxation
curve and the experimental observations, and the
agreements remain evident for the 10% strain for all of
the specimens.

Discussion
In this study, the elastic modulus of the retrodiscal
tissue at the onset of stress application was 1.54 MPa,
which was approximately 1/20th smaller than that of
the canine TMJ disc.13 Furthermore, a marked stress
decay occurred at the beginning of the 1-minute relaxation; the relaxation time for the constant strain
was 11.6 seconds in this study. Holmes and Hukins20
measured the load-relaxation under compression in
human cadaveric lumber spine and showed that the
relaxation time of the intervertebral disc was 16 minutes on average. We previously evaluated the compressive load-relaxation characteristics of the canine
TMJ discs and determined that the relaxation time
calculated was 31.2 seconds.13 The relaxation time in
the retrodiscal tissue, therefore, was much shorter
than that of the articular disc and intervertebral
disc.13,20 As a result, the relaxed modulus of the retrodiscal tissue was 0.21 MPa, and the relaxed stress
after 5 minutes of stressrelaxation was 0.015 and
0.046 MPa in the 10% and 20% strains of compression,
respectively. These values were similar to those of the
interparietal suture and periodontal ligament.14,21
This nding indicates that the energy-dissipation
mechanism exists in the retrodiscal tissue. On the
basis of previous reports, the relaxed stress in the TMJ
disc after 3 minutes of stressrelaxation was approximately half of the instantaneous stress.13,15 In the
present study, the retrodiscal tissue exhibited 90%
reduction in the instantaneous stress after 5 minutes
of stressrelaxation; therefore, the retrodiscal tissue

549
has a great capacity for energy dissipation during
stressrelaxation, although it has little or no function
to pull back the articular disc. The major factor for
energy dissipation may be the presence of a large
amount of loose connective tissue in the retrodiscal
tissue. The moisture content of the retrodiscal tissue
is greater than that of the TMJ disc, so the capacity to
reduce high loads via diffusion of water may be
greater in retrodiscal tissue relative to the articular
disc.
Typically, the stress-strain relation in the connective tissue was nonlinear. Haut and Little22 developed
a constitute equation for collagen bers based on the
quasilinear viscoelasticity law suggested by Fung19
and found that the elastic stress-strain relation for
collagen bers in rat tails was represented as a square
power. Previous studies, using the interparietal suture,14 the intervertebral disc,20 and the articular
disc12,13,23 as materials, supported their assumption.
In this study, the instantaneous and relaxed stresses in
the retrodiscal tissue were proportional to the square
of the strain. The possible explanation for this shape
of stress-strain curve is that the relationship is a nonlinear expression of the viscoelastic properties of the
retrodiscal tissue. When a viscoelastic specimen is
deformed, the water is squeezed out of the specimen
as the orientation of the collagen bers is rearranged.24 However, a satisfactory agreement can be
also obtained between experimental and theoretical
data from a standard linear model (Kelvins model).
Thus, for 2 strain levels of 10% and 20%, the retrodiscal tissue can be represented as a linear viscoelastic
material.
The viscoelasticity of the retrodiscal tissue is affected by extrinsic and intrinsic factors. Experience
from TMJ surgery has suggested that when the inferior layer of the retrodiscal tissue between the disc
and the condyle is damaged, it loses its rmness and
becomes more elastic, allowing the disc to be
displaced anteriorly or anteromedially.25 In the histologic analysis of the retrodiscal tissue in normal subjects, the asymptomatic joints showed a predominance of loose and dense connective tissue. In
contrast, chodroblast-like cells can be present when
the retrodiscal tissue replaces the articular disc above
the condyle during altered disc position.26 The viscoelastic property of the retrodiscal tissue is also
affected by aging as an intrinsic factor. Pereira et al27
investigated the histology of the retrodiscal tissue in
younger and older persons and reported that a significantly lower density of broblasts was observed in
the elderly persons in the lateral and central portions
of the posterior disc attachment of the TMJs. In addition, elderly persons demonstrated a signicantly
higher presence of dense connective tissue in the
central third of the posterior disc attachment. Their

550
study shows that the retrodiscal tissue, which may
eventually function as an articular disc during altered
disc position, is susceptible to age-related connective
tissue changes. Further study should be conducted to
investigate the changes of the viscoelasticity in the
retrodiscal tissue due to such extrinsic and intrinsic
factors. These results will encourage other studies to
be carried out because of their importance for a better
understanding of changes of biomechanical environments in the TMJ.
The present study showed for the rst time the
biomechanical behavior of the retrodiscal tissues in
the TMJ. The retrodiscal tissues are considered to play
an important role in lling the posterior joint spaces
and controlling the disc position during jaw opening.4,7 Therefore, these tissues are subjected to tension with normal jaw opening, although various
stresses, such as tensile, compressive, and shear
stresses, are produced within the tissues.4,7,28 In this
study, the biomechanical responses of the retrodiscal
tissues, however, were evaluated under compression,
not tension. A previous study suggested that when
the condyle is nearing the fully closed position, the
condylar part of retrodiscal tissue may be under compressive load.28 In addition, with the anterior disc
displacement, the retodiscal tissues are subjected to
continuous compression.29 However, it is true that
tension is mainly induced in the retrodiscal tissues
under normal function. Therefore, the present result
may not sufciently reect the biomechanical characteristics under normal function, and the biomechanical examination of the retrodiscal tissues under
tension may have useful information for better understanding the biomechanical environment and response under normal function.
We conclude that the bovine retrodiscal tissue exhibits near-linear elastic characteristics at each time
within a 5-minute period as well as articular disc, and
the load decay essentially ceases 60 seconds after
compression. After 5 minutes of stressrelaxation, the
retrodiscal tissue exhibited 90% reduction in the instantaneous stress. Therefore, bovine retrodiscal tissue has a great capacity for energy dissipation during
stressrelaxation, although it has little or no function
to pull back the articular disc.
Acknowledgment
We thank Prof Masao Tanaka (Division of Mechanical Science,
Osaka University Graduate School of Engineering Science, Osaka)
for his valuable advice and excellent technical assistance.

References
1. Nickel JC, McLauchlan KR: In vitro measurement of the stressdistribution properties of the pig temporomandibular joint
disc. Arch Oral Biol 39:439, 1994

BIOMECHANICAL RESPONSE OF RETRODISCAL TISSUE

2. Hinton RJ: Adaptive response of the articular eminence and
mandibular fossa to altered function of the lower jaw: an
overview, in Carlson DS, McNamara JA Jr, Ribbens KA (eds):
Developmental Aspects of Temporomandibular Joint Disorders. Ann Arbor, MI, Center for Human Growth and Development, University of Michigan, 1988, pp 279-298
3. Copray JCVM, Jansen HWB, Duterloo HS: Growth and
growth pressure of mandibular condylar and some primary
cartilages of the rat in vitro. Am J Orthod Dentofac Orthop
90:12, 1986
4. Scapino RP: The posterior attachments: its structure, function,
and appearance in TMJ imaging studies. Part 1. J Craniomandib
Disord Facial Oral Pain 5:83, 1991
5. Scapino RP: Histopathology associated with malposition of
the human temporomandibular joint disc. Oral Surg 55:382,
1983
6. Kino K, Omura Y, Amagasa T: Reconsideration of the bilaminar
zone in the retrodiskal area of the temporomandibular joint.
Oral Surg Oral Med Oral Pathol 75:410, 1993
7. Osborn J: The disc of the human temporomandibular joint:
Design, function and failure. J Oral Rehabil 12:279, 1985
8. Wilkinson TM, Crowley CM: A histologic study of retrodiscal
tissues of the human temporomandibular joint in the open and
closed position. J Orofac Pain 8:7, 1994
9. Woo SL-Y: Biomechanics of tendons and ligaments, in SchmidSchonbein GW, Woo SL-Y, Zweifach BW (eds): Frontiers in
Biomechanics. New York, NY, Springer, 1986, pp 180-195
10. Oloyede A, Flachsmann R, Broom ND: The dramatic inuence
of loading velocity on the compressive response of articular
cartilage. Connect Tissue Res 27:211, 1992
11. Reihsner R, Menzel EJ: On the orthogonal anisotropy of human
skin as a function of anatomical region. Connect Tissue Res
34:145, 1996
12. Kuboki T, Shinoda M, Orsini MG, et al: Viscoelastic properties
of the pig temporomandibular joint articular soft tissues of the
condyle and disc. J Dent Res 76:1760, 1997
13. Tanaka E, Tanaka M, Miyawaki Y, et al: Viscoelastic properties
of canine temporomandibular joint disc in compressive loadrelaxation. Arch Oral Biol 44:1021, 1999
14. Tanaka E, Miyawaki Y, del Pozo R, et al: Changes in the
biomechanical properties of the rat interparietal suture incident to continuous tensile force application. Arch Oral Biol
45:1059, 2000
15. Scapino RP, Canham PB, Finlay HM, et al: The behaviour of
collagen bres in stress relaxation and stress distribution in the
jaw-joint disc of rabbits. Arch Oral Biol 41:1039, 1996
16. Mills DK, Fiandaca D, Scapino RP: Morphologic, microscopic
and immunohistochemical investigations into the function of
the primate TMJ disc. J Orofac Pain 8:136, 1994
17. Nakano T, Scott PG: Changes in the chemical composition of
the bovine temporomandibular joint disc with age. Arch Oral
Biol 41:845, 1996
18. Sindelar BJ, Evanko SP, Alonzo T, et al: Effects of intraoral splint
wear on proteoglycans in the temporomandibular joint disc.
Arch Biochem Biophys 379:64, 2000
19. Fung YC: Stress-strain history relations of soft tissues in simple
elongation, in Fung YC, Perrone N, Anliker M (eds): Biomechanics: Foundations and Objectives. Englewood Cliffs, NJ,
Prentice Hall, 1972, pp 181-208
20. Holmes AD, Hukins DWL: Analysis of load-relaxation in compressed segments of lumber spine. Med Eng Phys 18:99,
1996
21. Tanne K, Inoue Y, Sakuda M: Biomechanical behavior of the
periodontium before and after orthodontic tooth movement.
Angle Orthod 65:123, 1994
22. Haut RC, Little RW: A constitutive equation for collagen bers.
J Biomech 5:421, 1972
23. Chin LPY, Aker FD, Zarrinnia K: The viscoelastic properties of
the human temporomandibular joint disc. J Oral Maxillofac
Surg 54:315, 1996
24. Mow VC, Kwan MK, Lai WM, et al: A nite deformation
theory for nonlinearly permeable soft hydrated biological
tissues, in Schmid-Schonbein GW, Woo SL-Y, Zweifach BW

TANAKA ET AL
(eds): Frontiers in Biomechanics. New York, NY, Springer,
1986, pp 153-179
25. Eriksson L, Westesson P-L, Macher D, et al: Creation of disc
displacement in human temporomandibular joint autopsy
specimens. J Oral Maxillofac Surg 50:869, 1992
26. Pereira FJ, Lundh H, Westesson PL: Age-related changes of the
retrodiscal tissues in the temporomandibular joint. J Oral Maxillofac Surg 54:55, 1996
27. Pereira FJ, Lundh H, Eriksson L, et al: Microscopic changes in

551
the retrodiscal tissues of painful temporomandibular joints.
J Oral Maxillofac Surg 54:461, 1996
28. Scapino RP: Morphology and mechanism of the jaw joint
(chapter 2), in McNeill C (ed): Science and Practice of Occlusion. Chicago, IL, Quintessence Publishing, 1997, pp 23-40
29. Scapino RP, Mills DK: Disc displacement internal derangements (chapter 18), in McNeill C (ed): Science and Practice of
Occlusion. Chicago, IL, Quintessence Publishing, 1997, pp
220-234