J. Anat.

(2006) 208, pp491–512

REVIEW
Blackwell Publishing Ltd

The effects of exercise on human articular cartilage

F. Eckstein,1 M. Hudelmaier1 and R. Putz2
1
Institute of Anatomy & Musculoskeletal Research, Paracelsus Private Medical University (PMU), Salzburg, Austria
2
Musculoskeletal Research Group, Institute of Anatomy, Ludwig-Maximilians-Universität München, Germany

Abstract
The effects of exercise on articular hyaline articular cartilage have traditionally been examined in animal models,
but until recently little information has been available on human cartilage. Magnetic resonance imaging now permits
cartilage morphology and composition to be analysed quantitatively in vivo. This review briefly describes the
methodological background of quantitative cartilage imaging and summarizes work on short-term (deformational
behaviour) and long-term (functional adaptation) effects of exercise on human articular cartilage. Current findings
suggest that human cartilage deforms very little in vivo during physiological activities and recovers from deformation
within 90 min after loading. Whereas cartilage deformation appears to become less with increasing age, sex and
physical training status do not seem to affect in vivo deformational behaviour. There is now good evidence that
cartilage undergoes some type of atrophy (thinning) under reduced loading conditions, such as with postoperative
immobilization and paraplegia. However, increased loading (as encountered by elite athletes) does not appear to
be associated with increased average cartilage thickness. Findings in twins, however, suggest a strong genetic
contribution to cartilage morphology. Potential reasons for the inability of cartilage to adapt to mechanical stimuli
include a lack of evolutionary pressure and a decoupling of mechanical competence and tissue mass.
Key words adaptation; biomechanics; cartilage; morphology; magentic resonance imaging.

Hyaline articular cartilage provides the bearing

Introduction
Diarthrodial (synovial) joints represent important organs
of the musculoskeletal system. They enable individuals
to maintain posture within the gravitational field, to
position their body relative to their surroundings, to
move, and to manipulate objects around them. When
performing these tasks, joints commonly encounter
forces of several times the body weight (Bergmann et al.
1993). Diarthrodial joints are composed of various
structures and tissues which, from a functional point of
view, act in concert. Joints are made in a manner to deal
effectively with the mechanical loads encountered over
many years of life, ideally without suffering damage
(Mow et al. 2003; Ateshian & Mow, 2005).
Correspondence
Professor Felix Eckstein, Institute of Anatomy & Musculoskeletal
Research, PMU, Strubergasse 21, A 5020 Salzburg, Austria. T: +43 662
44 20021240; F: +43 662 44 20021249; E: felix.eckstein@pmu.ac.at
Accepted for publication 16 January 2006
surface of synovial joints. This review will focus on the
morphology and function of this tissue, which – for the
sake of brevity – will simply be addressed as ‘cartilage’
throughout the remainder of the text. Cartilage is unique
as it is an avascular, aneural tissue, in which cells survive
for a lifetime, without intercellular connections (Hunziker
et al. 2002). In particular, adult articular cartilage has
no other known function than maintaining mechanical
competence. Owing to its sophisticated composition, its
high water content and its ability to withstand hydro-
static pressurization (Ateshian et al. 1994; Mow et al. 2003),
cartilage is capable of transferring enormous forces
relatively evenly from one subchondral bone plate to
the other (Mow et al. 1984, 1993). Under physiological
conditions, cartilage also provides an almost frictionless
gliding surface and is thus capable of transferring these
loads during motion (Ateshian & Mow, 2005). In order
to be able to meet these complex mechanical demands
without undergoing wear and tear, articular cartilage
displays unique morphological and biomechanical

© 2006 The Authors

Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
492 The effects of exercise on human articular cartilage, F. Eckstein et al.
properties (Mow et al. 1984, 1993; Buckwalter & Mankin,
1998b; Hunziker et al. 2002). These properties are yet
unmatched by any artificial material, despite consider-
able efforts by engineers and biologists (Buckwalter &
Mankin, 1998a; Hunziker, 2002).
Due to a lack of non-invasive methods that allow
human articular cartilage to be studied directly in vivo,
little has been known until recently about the variability
of normal cartilage morphology between subjects and
the factors which determine this. Even less is known
about the deformational behaviour of cartilage under
load in the intact joint in vivo. However, with quantita-
tive magnetic resonance imaging (qMRI) having become
available, data on these topics have begun to emerge.
This review will primarily focus on whether articular
cartilage is mechano-adaptive. In contrast to that of many
other tissues, the morphology of articular cartilage (i.e.
its thickness) is determined relatively late in postnatal
life, during adolescence, when endochondral ossifica-
tion is complete. We do not know which specific factors
prohibit the calcification front from advancing to
the joint surface and are responsible for the fact that
a layer of cartilage is maintained at the joint surface.
Given the limited number of genes available to guide
the emergence and maintenance of the morphology
of various tissues and functional systems in general, it
is tempting to hypothesize that environmental factors,
specifically mechanics, play a pivotal role in this process.
The ability of tissues to emerge and maintain their
structure in accordance with specific environmental
requirements has been termed ‘functional adaptation’
(Lamarck, 1809; Darwin, 1872; Roux, 1881; Wolff, 1892;
Pauwels, 1980; Carter et al. 1991; Huiskes et al. 2000).
Processes of functional adaptation have been regarded
as occurring during the development of the central
nervous system (e.g. the visual cortex), internal organs
(e.g. the kidney) and in tissues with primarily mechan-
ical functions, such as muscle and bone (Wolff, 1892;
Pauwels, 1980; Carter et al. 1991; Keller et al. 1992;
Booth, 1994; Huiskes et al. 2000). Physical exercise
has been shown to increase bone and muscle mass
(e.g. body building), whereas states of inactivity or
microgravity have been associated with tissue atrophy
(Keller et al. 1992; Booth, 1994). In bone, functional
adaptation to mechanics has been mathematically
characterized as a cell-mediated process in which
osteocytes act as mechanical sensors and orchestrate the
function of other (bone-forming and bone-resorbing)
cells through biochemical signalling (Huiskes et al. 2000).
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The biosynthetic activity of chondrocytes has also been
shown experimentally to be regulated by mechanical
stimuli (Kim et al. 1995; Waldman et al. 2003). Based on
these in vitro findings, mathematical models have been
developed that explain the variable thickness of carti-
lage between joints based on differences in mechanical
loading magnitude (Carter & Wong, 1988a,b, 2003;
Carter et al. 1991; Wong & Carter, 2003). To date, how-
ever, there has been little experimental evidence to
support this theory on a systemic level.
For this reason, the primary objective of this review
is to address the question of whether (and to what
extent) articular cartilage is subject to deformation
under physiological loading conditions, and what
magnitude of mechanical signals is encountered by
the cartilage matrix (and thus the chondrocytes) during
activities of daily life. The second objective to be addressed
is whether cartilage tissue can adapt to mechanical
stimuli by altering its morphology (specifically its thick-
ness) and composition (proteoglycan, collagen and
interstitial water content) to the specific mechanical
conditions on a systemic level. Because most of the
recent experiments on these questions have been
performed using MRI, we will also briefly review meth-
odological aspects of this imaging modality in the
context of quantitative cartilage analysis.
Methodological background of qMRI
of cartilage
MR pulse sequences for quantitative analysis
of cartilage morphology
An MRI pulse sequence suitable for measuring cartilage
morphology quantitatively must provide a high signal-to-
noise ratio (SNR) and contrast-to-noise ratio (CNR) for
accurate delineation of the subchondral bone interface
and articular surface, and no significant artefacts must
be present. Measurements should be obtained at
relatively short imaging times, in order to avoid motion
artefacts and to be able to measure cartilage deforma-
tion directly after exercise (Tieschky et al. 1997). Because
cartilage layers exhibit a mean thickness of only 1.3–2.5
mm throughout the human knee (Eckstein et al. 2001a;
Hudelmaier et al. 2001) and even less so in other joints
(Peterfy et al. 1995; Springer et al. 1998; Graichen et al.
2000, 2003; Al Ali et al. 2002), a high spatial resolution
is required so that a sufficient number of image points
(pixels) are available to characterize the thickness of
© 2006 The Authors
 The effects of exercise on human articular cartilage, F. Eckstein et al. 493

the tissue throughout the joint surface, including areas

with thin cartilage coverage. Increasing the resolution
by a factor of two in three dimensions requires acqui-
sition times to be increased by a factor of 64, if the SNR
is to be kept constant. Although there is no current
consensus on the optimal resolution for imaging
cartilage morphology, a 1.5-mm section thickness and
0.3-mm in-plane resolution has been commonly used at
a field strength of 1.5 T. The specific MR pulse sequences
that has been most frequently employed for cartilage
imaging is a T1-weighted spoiled gradient echo sequence
[FLASH = fast low angle shot (Frahm et al. 1986) or
SPGR = spoiled gradient recalled acquisition at steady
state]. This sequence (Fig. 1a,b) is available on most clinical
MRI systems and has been implemented either with
frequency-selective spectral fat-suppression by a
prepulse (Recht et al. 1993; Peterfy et al. 1994; Eckstein
et al. 1996a; Cicuttini et al. 2000) or with frequency-
selective water excitation (Hardy et al. 1998; Graichen
et al. 2000; Burgkart et al. 2001; Glaser et al. 2001). Both
techniques achieve effective fat-saturation, which is
required to provide a sufficient dynamic range of the
image contrast between the cartilage and its surrounding
tissues, and to eliminate artefacts at the subchondral
bone interface. New 3.0-T whole-body MR scanners
now make it possible to perform quantitative cartilage
imaging at higher field strength (Gold et al. 2004a,b;
Eckstein et al. 2005a,b; Kornaat et al. 2005). A recent
study has shown that measurements at 3.0 T are
consistent with those at 1.5 T, and that the precision
(reproducibility) of the measurements is slightly improved
when exploiting the higher field strength to obtain a
Fig. 1 (a) Coronal MR imaging (slice thickness 1.5 mm,
higher spatial resolution (1-mm slice thickness) at 3.0 T
in-plane resolution 0.31 mm × 0.31 mm) acquired with a T1-
compared with 1.5 mm at 1.5 T (Eckstein et al. 2005a). weighted spoiled gradient echo sequence (FLASH = fast low
One of the great advantages of MRI (e.g. in comparison angle shot; or SPGR = spoiled gradient recalled acquisition
with histology) is that consecutive slices are contiguous at steady state) with frequency-selective water excitation.
(b) Segmentation showing the medial tibial cartilage in
and spatially aligned, so that three-dimensional (3D) blue, the medial femoral condyle in yellow, the lateral tibia
parameters can be obtained that characterize cartilage cartilage in green, and the lateral femoral cartilage in red.
morphology appropriately (Fig. 2). These parameters (c) Sagittal dGEMRIC image kindly provided by Dr Deborah
Burstein, Beth Israel Deaconess Medical Center, Harvard
include cartilage volume, cartilage thickness (mean,
Medical School, Boston, MA, USA.
maximum, standard deviation), cartilage surface area
(or subchondral bone interface area) as a measure of
bone size, cartilage surface curvature (joint incon-
gruity) and others (Fig. 2). When reporting cartilage over time correspond. In cross-sectional studies, it is
volume, one must keep in mind that this parameter important to report cartilage thickness directly, or to
depends on both the cartilage thickness and the normalize cartilage volume to the joint surface/bone
cartilage surface area, and that only under conditions interface area, in order to provide meaningful results.
where the cartilage surface (or chondro-osseous inter- It has been shown, for instance, that gender differences
face area) is constant, do volume or thickness changes in joint surface areas are substantially larger than those

© 2006 The Authors

Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
494 The effects of exercise on human articular cartilage, F. Eckstein et al.
Fig. 2 (a) Three-dimensional reconstruction of femoral and
tibial cartilage from segmentations of contiguous MR images;
(b) distribution pattern of cartilage thickness in the femur,
determined independent of the original section orientation.
The blue colour shows areas of thick cartilage, orange and red
show areas of thin cartilage.
for cartilage thickness (Faber et al. 2001), a finding that
is not evident from measuring cartilage volume
alone. Also, it has been shown that cartilage thickness
and cartilage surface areas are not closely associated
in healthy individuals (Eckstein et al. 2001b); in other
words, subjects in whom the articular cartilage occupies
a larger surface area do not necessarily have thick
cartilage and vice versa. Thus, one of these parameters
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
cannot be estimated from the other – both must be
measured as separate entities.
In order to derive quantitative data from a 3D,
contiguous image set, an anatomical structure (the
articular cartilage) must first be labelled, distinguishing
it from its immediate relations (segmentation – Fig. 1b).
Owing to the relatively low contrast in some areas of
the joint surface (joint contact areas, the vicinity of
synovial folds, tendons and ligaments, repair tissue, etc.),
fully automated segmentation of cartilage is impracti-
cal from MR images. Various semi-automated image
analysis techniques have been developed to date, each
requiring different degrees of user interaction. Verifica-
tion (and some degree of correction) by an experienced
user is generally necessary on a section-by-section basis.
The inability of current computer software reliably
to identify structures in images that are evident to the
experienced human eye may seem surprising. How-
ever, if one considers the great difficulties involved in
automated speech recognition by computers, despite
the tremendous efforts made by industry, one may
appreciate the complexity of automated recognition
in intricate image pattern identification. For these
reasons, and because many slices must be acquired of
one joint surface to obtain sufficient spatial resolution,
cartilage segmentation is a time-consuming process
currently requiring several hours of human interaction
per knee data set.
After segmentation, computation of the cartilage
volume is straightforward, by simple numerical integra-
tion of the number of voxels attributed to the cartilage
during the segmentation process (Fig. 2). More sophis-
ticated algorithms are then used to determine the
cartilage thickness (Fig. 2) and joint surface area, which
must account for out-of-plane deviations of these
parameters. Computations should therefore be made
in three dimensions, independent of the original section
orientation. Extraction of cartilage surfaces also allows
for the determination of geometric topography and
curvature characteristics of diarthrodial joints (Ateshian
et al. 1991). Mathematical descriptions of joint surfaces
and articular cartilage layers can also be applied to
derive computer models of human joints, by which the
contact areas and surface stresses in joints may be
estimated (Cohen et al. 1999, 2001) but these methods
have not been yet applied to the study of the effects of
exercise on cartilage and joint morphology.
Because these ‘global’ parameters (volume and mean
thickness for an entire cartilage plate) may be relatively
© 2006 The Authors

insensitive to regional/focal changes that affect only
small portions of the surface, several investigators have
presented techniques for displaying regional cartilage
thickness patterns (Eckstein et al. 1995, 1996a,b, 1998a;
Sittek et al. 1996; Cohen et al. 1999, 2003; McGibbon &
Trahan, 2003) (Fig. 2). Changes over time (or differences
between subjects) in regional cartilage thickness are,
however, difficult to detect from subjective comparison
of such thickness patterns, because only a limited number
of thickness intervals can be displayed. In order to track
local/regional thickness changes over time, registration
techniques have therefore been proposed (Kshirsagar
et al. 1998; Stammberger et al. 2000; Waterton et al.
2000; Lynch et al. 2001; Cohen et al. 2003; Raynauld
et al. 2003). With these methods, the bone interface or
other anatomical landmarks from two data sets are
matched so that the thickness distribution can be com-
pared on a point-by-point basis. Stammberger et al.
(2000) reported a local mismatch of cartilage thickness
for joint repositioning in the range of 0.5–1 mm. These
local errors are relatively large in comparison with the
absolute cartilage thickness in knee joint surfaces, but
this is not surprising given that an anatomically com-
plex structure is reconstructed and registered with data
obtained from a limited number of sectional images.
Validation and reproducibility (precision) of
quantitative analysis of cartilage morphology
The validity (accuracy) of qMRI of cartilage has been
addressed in numerous studies over recent years and
these have been carried out in unselected cadaver
joints, amputated joints (Peterfy et al. 1994; Cicuttini
et al. 1999) or knee joint of patients undergoing total
knee arthroplasty (TKA) (Peterfy et al. 1994; Cicuttini
et al. 1999; Burgkart et al. 2001; Graichen et al. 2004).
TKA provides a unique opportunity for validating
quantitative measurements, as patients can be imaged
prior to surgery in vivo, and the tissue can be removed
and analysed after the operation. Validation studies
have been carried out in comparison with various
reference methods, namely water displacement of
surgically retrieved tissue (either direct or by employing
Archimedes’ principle), anatomical sections obtained
with high-precision band saws, computer tomography
arthrography, A-mode ultrasound (not to be confused
with clinical B-mode ultrasound), and stereophoto-
grammetry. Most of these comparative studies have
reported close agreement between methods of meas-
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 495
uring cartilage volume, with random errors (absolute
pairwise over- or underestimation) vs. the respective
reference method of about 5–10%. Validation studies
have also been performed in other joints with thinner
cartilage, such as the metacarpophalageal joint (Peterfy
et al. 1995), the hip (McGibbon et al. 1998), the elbow
(Graichen et al. 2000) and the shoulder (Graichen et al.
2003).
Precision errors are random errors that occur when
repeated measurements of a parameter are taken under
constant conditions. Highly reproducible techniques
are required to resolve small changes (i.e. cartilage
deformation) with statistical confidence. For qMRI of
cartilage morphology, the precision depends on factors
associated with image acquisition, and factors associ-
ated with image analysis. Differences in joint position-
ing are less critical than for projectional techniques
(such as radiography), because the technique is 3D and
the relevant quantitative measures are obtained from
reconstructions of serial images rather than from
projection onto one image plane. The lowest precision
errors (CV% ∼1%) have been observed for axial pro-
tocols of the patella (Eckstein et al. 2000b). Higher
precision errors, by contrast, have been reported for
analyses of the femoral condyles in sagittal scans
(Eckstein et al. 2002b), whereas analysis of the total
femur has usually been comparable with other joint
surfaces of the knee. Precision errors of computations of
the mean cartilage thickness throughout joint surfaces
have been reported to be similar to those of cartilage
volume (Stammberger et al. 1999; Hyhlik-Dürr et al.
2000; Burgkart et al. 2001; Eckstein et al. 2002b) as
have those for quantification of cartilage surface areas
(Hohe et al. 2002; Eckstein et al. 2002b).
MR protocols for compositional cartilage imaging
In addition to measuring cartilage morphology, there
have been great efforts in using MRI to determine the
composition of cartilage, namely the glycosaminoglycan
(GAG) content, collagen content and orientation, and
the interstitial water content. Attempts to determine
the concentration of GAG include imaging of fixed
charge density by using an intravenous injection of
the charged clinical MRI contrast agent Gd(DTPA)2–. If
Gd(DTPA) 2– is allowed to penetrate into cartilage,
a process that has been estimated to last for about
90 min after injection, it distributes inversely with the
GAG concentration (Fig. 1c). Because full penetration
496 The effects of exercise on human articular cartilage, F. Eckstein et al.
is required, the technique has been termed delayed
gadolinium enhanced MRI of cartilage (dGEMRIC)
(Gray et al. 2004). When tissue is placed in a magnetic
field, magnetic moments of the protons are aligned,
resulting in a net magnetic moment. This equilibrium is
then disturbed by transmitting another magnetic field
at the same frequency as the rotations of the protons
for a very short time. The return to equilibrium of the
magnetic moments after this pulse is strongly affected
by molecular interactions of the nuclei with their sur-
roundings and can be exploited for imaging cartilage
composition (Burstein & Gray, 2003). Two time constants
are relevant in this context, the longitudinal (T1)
and transverse relaxation time (T2). When probing T1
in cartilage in the presence of fully penetrated Gd(DTPA)2–
(dGEMRIC), one can estimate the GAG content of the
tissue (Fig. 1c). dGEMRIC has been validated in basic
science and clinical studies through comparison with
biochemical and histological measures of GAG (Bashir
et al. 1997, 1999; Tiderius et al. 2003; Williams et al. 2004).
Another technique that has been successfully explored
is T2 mapping (Mosher & Dardzinski, 2004). T2 can be
obtained without the presence of a contrast agent,
but cannot be attributed to a single constituent of
cartilage composition. T2 has been shown to provide a
quantitative measure of cartilage interstitial fluid and
its interaction with the solid components of the extra-
cellular cartilage matrix, in particular with collagen
content and orientation, whereas there is little to no
sensitivity to changes in GAG concentration (Mosher &
Dardzinski, 2004). Spatially resolved cartilage T2 maps
have been shown to be correlated with the regional
water content of the deep and mid zones of cartilage
(Lusse et al. 2000) but recent data suggest that collagen
fibre anisotropy is the dominant factor related to
regional differences in T2. T2 of the superficial zone
of cartilage was shown to change with aging (Mosher
et al. 2004).
Short-term effects of exercise on articular
cartilage (deformational behaviour)
Although the mechanical properties of articular cartilage
have been thoroughly studied under in vitro conditions
(Mow et al. 1993, 2003), until recently there have been
little data on the magnitude of the in situ cartilage
deformation in intact joints, and there have been no
data for in vivo loading conditions. This information
cannot be extrapolated from in vitro studies, because
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
the magnitude of the joint loads during normal (partic-
ularly dynamic) exercise is unknown. Moreover, the effect
of boundary conditions (e.g. non-linear contact con-
ditions between incongruous joint surfaces, presence
of the synovial fluid) is difficult to take into account. In
vivo data on the deformation of articular cartilage are
necessary, however, in a number of areas. (1) Cartilage
deformational behaviour depends on the biochemical
composition of the tissue and may potentially represent
a more sensitive surrogate marker of early osteoarthritis
than morphological endpoints such as cartilage volume,
thickness or joint space narrowing (Burstein et al. 2000).
(2) Knowing the magnitude of strain in the target tissue
is important when designing cartilage transplants.
Artificial cartilage can then be designed in a way to be
able to withstand these strains ex vivo so that cartilage
transplants can be expected to meet mechanical require-
ments in vivo. (3) The magnitude of in vivo deforma-
tion of articular cartilage is related to the magnitude of
mechanical stimulation experienced by the chondro-
cytes, which is known to affect their biosynthetic
activity (Sah et al. 1989; Urban, 1994; Kim et al. 1995).
Knowledge of cartilage deformation in vivo may thus
serve as an important guideline of how to stimulate cells
optimally in tissue culture and cartilage transplants,
and also how to use mechanical signals to stimulate
cartilage in situ.
In vitro studies of the intact femoro-patellar joint
Cartilage deformation during loading cannot be readily
investigated with MRI, because it is difficult to apply
relevant loads to the joint of a living person within the
MRI scanner, and to keep the joint in a constant position
relative to the coil at the same time. Note that, given
an in-plane resolution of 300 µm, even very tiny move-
ments of the limb make it impossible to obtain an
accurate measure of cartilage thickness. To overcome
this limitation, one group (Herberhold et al. 1998) con-
structed a non-metallic compression apparatus capable
of generating loads of up to 1500 N (Fig. 3) for studying
patellofemoral compression in a clinical MRI scanner.
The time-dependent deformation of the patellofemo-
ral cartilage was studied in situ over a 4-h period under
continuous static loading with 150% body weight, with
the joint capsule being kept fully intact (Herberhold
et al. 1999). Analysis of cartilage deformation in the
central 2D slice revealed a mean reduction of 44 ± 15%
of the initial thickness in the patella, and a cartilage
© 2006 The Authors
 The effects of exercise on human articular cartilage, F. Eckstein et al. 497

Fig. 3 The non-metallic compression

apparatus that fits into the extremity coil
of a clinical MRI scanner is capable of
generating loads of up to 1500 N using
a pneumatic pressure piston, and can
accommodate a human patellofemoral
joint at a 60° flexion angle. The patella
and pressure piston are guided between
Delrin trays, so that the cartilage
deformation could be monitored using
a fast two-dimensional MR imaging
sequence with an acquisition time of
< 1 min. Images on the right show the
status of the femoropatellar cartilage
before compression (t = 0 min) and after
120 min and 240 min of compression,
respectively.

thickness change of 30 ± 10% in the femoral trochlea
after 3.5 h of static loading with 150% body weight
(Figs 3 and 4a). The deformation of the patellar
cartilage exceeded that of the trochlea, this being
consistent with the differences in cartilage mechanical
properties reported between the patella and femoral
trochlear (Froimson et al. 1997). The maximal cartilage
deformation observed was 57 ± 15% in the patella
and 44 ± 7% in the trochlea (Herberhold et al. 1999).
Cartilage thickness decreased in an approximately
exponential manner and cartilage deformation ceased
in the central slice after 3.5 h (equilibrium). Interest-
ingly, however, only a small fraction of the final (3.5 h)
deformation was reached during the first 1 min of
static loading (3% cartilage thickness change in the
patella and 1.3% in the trochlea), and the deformation
was still of little magnitude after ∼8 min of loading
(11% cartilage thickness change in the patella and 9%
in the trochlea; Fig. 4b). This revealed that only about
4–7% of the final (near equilibrium) deformation is
reached with the first 1 min of loading, and that only
25 –30% of this final deformation is reached during the
first 8 min of load application.
The patellar cartilage volume (as measured over
the entire patella, but not in the central 2D slice) was
reduced by 8 ± 5% after 14 min and by 29 ± 3.2% after
3.5 h of loading (Fig. 4c). After 3.5 h, deformation
had not yet ceased and equilibrium had not yet been
reached, because the load-bearing area continued to
increase proximally and distally from the site of central
contact, in which equilibrium had been reached (see
above). Given an interstitial water content of cartilage
of about 80%, these data suggest that more than 50%
of the interstitial fluid was displaced from the patellar
cartilage matrix during 3.5 h of static compression.
Based on the assumptions that the solid matrix is incom-
pressible (Mow et al. 2003), that all volume changes are
due to fluid flow and that during compression the fluid
flow will occur throughout the articular surface into
the joint cavity, the rate of interstitial fluid loss from
the matrix was estimated to be 1.3 (± 0.5) mm3 min−1
per cm2 surface area (fluid flux = 0.217 ± 0.083 µm s−1)
for the first 14 min of loading, and 0.22 (± 0.04) mm3
(min cm2)−1 (0.037 ± 0.007 µm s−1) in the terminal phase
of the experiment (> 120 min). Note that these values
provide only a mean throughout the joint surface and

© 2006 The Authors

Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
498 The effects of exercise on human articular cartilage, F. Eckstein et al.
Fig. 4 In situ compression of patellofemoral cartilage using
the compression device shown in Fig. 3. (a) Graph showing the
mean reduction of thickness in a central axial 2D slice through
the patellofemoral contact zone over 3.5 h of static loading
with 150% body weight; (b) only a small fraction of the final
thickness is reached during the first few minutes of static
loading; (c) patellar cartilage volume (measured over the
entire patella) was reduced by approximately 30% after 3.5 h
of loading.
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
that the flux is assumed to be highly variable through-
out it during the compression experiment. Also, the
thickness changes were found to be highly inhomoge-
neous throughout the joint surface and load-bearing
area, with the maximal deformation (the lateral
patellar facet) being identical to the site of the maximal
cartilage thickness. An experiment using pressure-
sensitive FUJI film, conducted in the same specimens
after imaging, showed pressure maxima of 3.6 ± 1.3 MPa,
in the lateral patellar facet, the pressure distribution
being very similar to the pattern of cartilage deformation.
In one of the specimens imaging was continued after
the 3.5-h experiment and after removal of the load.
The cartilage displayed almost full recovery (98%) after
approximately 4 h. The rate of fluid re-uptake by the
cartilage matrix was 0.9 mm3 (min cm2)−1 (0.15 µm s−1)
within the first 14 min of recovery, and 0.14 mm3 (min
cm2)−1 (0.023 µm s−1) in the terminal phase of the recovery
period.
It may appear surprising that cartilage deforms so
little for the first few minutes under high loads. How-
ever, the primary function of cartilage is not to absorb
energy through deformation (a function that is performed
by the muscles, the tendons and the joint as an organ),
but to distribute the load equally to the subchondral
bone plate and to provide minimal friction during
motion. This function can be compared with the inflated
tyre of a bike, which should not deform, in order to
avoid damage to the wheel (the subchondral bone)
and to provide minimum resistance during rolling. The
energy absorption, by contrast, is provided by a spring
(the joint, with its tendons and muscles), and energy
absorption is warranted by controlled joint movement
and negative acceleration in it. It may be argued that
deformation is higher when high-impact loading occurs,
but it should be remembered that cartilage consists of
80–90% interstitial fluid, which is incompressible, and
which has little time to escape from the load-bearing
area in high-impact loading. This can be appreciated
when landing on one’s front or back on water after a
jump from a 10-m tower. Given that the water has little
time to escape under these high-impact conditions, little
energy absorption is provided, and this even more so in
the cartilage, in which the fluid is bound to the matrix.
In vivo deformation of the patellar cartilage
In order to determine in vivo cartilage deformation
shortly after exercise, we have quantified the cartilage
© 2006 The Authors
 The effects of exercise on human articular cartilage, F. Eckstein et al. 499

volume of healthy volunteers after 1 h of physical rest

(no weight-bearing in the scanner) and then 3–7 min
after 50 knee bends (Eckstein et al. 1998b). Note that so
far it has been impossible to determine cartilage defor-
mation reliably during loading in vivo for the reasons
mentioned in the first paragraph, and that it also can-
not be determined immediately after the exercise, as
some time is required to reposition the patient in the
magnet, to obtain a preliminary view for slice position
and alignment, and actually to acquire the 3D image
data. After the knee bends we observed a reduction
in cartilage volume (compression) in the patella of
2.4–8.6% (mean = 6.0%) (Fig. 5) (Eckstein et al. 1998b).
Note that this is the mean deformation across the
entire patellar cartilage and that a certain amount of
inhomogeneity may be present, with some areas encoun-
tering higher and some lower deformation (see below).
When asking the same subjects to perform 100 knee
bends (Eckstein et al. 1999), the level of deformation
was of 2.4 – 8.5% (mean = 5.0%) and was not signific-
antly different from that after 50 knee bends (Fig. 5).
We then examined the time required for recovery of
the cartilage under non-weight-bearing conditions
(Eckstein et al. 1999). A period of 45 min was required
to compensate approximately 50% of the deformation
observed after knee bends and a period of 90 min to
attain the pre-exercise volume before the knee bends
(Eckstein et al. 1999) (Fig. 5). The fluid re-uptake appeared
to be almost linear throughout the observation period, Fig. 5 Patellar cartilage deformation after six sets of 50 knee
bends at 15-min intervals (top), and during recovery after
and the average fluid flux during the recovery period
100 knee bends (bottom).
was estimated to be approximately 0.027 µm s−1. This
compares well with the values observed during terminal macromolecular composition of the cartilage occurred
recovery after long-term static loading (Eckstein et al. following exercise.
1999) (see above). We demonstrated further that When comparing static (90° squatting for 20 s) with
multiple sets of 50 knee bends, with intervals of 15 min of dynamic loading (30 deep knee bends to 120°), dif-
rest in between, maintained the level of deformation ferences were found both in the magnitude and in the
measured after the first set of knee bends (around pattern of cartilage deformation throughout the joint
5–6%), but did not lead to further cartilage deforma- surface (Eckstein et al. 2000a). The reduction in cartilage
tion beyond this value (Eckstein et al. 1999) (Fig. 5). thickness after 30 knee bends (5.9 ± 2.1%) was not less
Liess et al. (2002) studied 20 healthy volunteers after than after 50 or 100 knee bends, with the range being
performing 60 knee bends. MR images of the patellar 2.9–9.6%. After static loading, we also observed a
cartilage were acquired immediately following exercise significant reduction of the patellar cartilage volume
and then after 45 min of rest. Within 45 min of loading, of 4.7 ± 1.6% (range 2.4–6.5%). The maximal deforma-
patellar cartilage volume increased by 5.4 ± 1.5%. At tion was observed in the central aspect of the lateral facet,
the same time, T2 maps of the patellar cartilage were which is also the site of maximal cartilage thickness
acquired. During the 45-min recovery period, T2 increased (Fig. 6). The cartilage volume changes recorded after
by 2.6 ± 1.0% (P < 0.05). The authors concluded that dynamic loading were significantly higher than those
small physiological changes in the water content of after static loading (P < 0.05), probably because the
patellar cartilage and concomitant changes of the patellar joint surface area involved in load-bearing is

© 2006 The Authors

Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
500 The effects of exercise on human articular cartilage, F. Eckstein et al.

Fig. 6 In vivo deformation patterns of

patellar cartilage after different
physiological activities (see text). A
posterior view onto the right patellar
cartilage surface (proximal pole of the
patella on top, medial side on the left):
average of differences in cartilage
thickness before and after various
activities averaged over 12 volunteers.
Red and orange colours show areas of
high deformation, blue colour areas of
little deformation.

much higher for dynamic exercise with a wide range of
knee angles being taken, and the correlation coefficient
for deformation after the two activities was r = 0.69 (P <
0.05). In a recent paper we compared several activities
(Eckstein et al. 2005c), including (1) 30 deep knee bends,
(2) static loading (see above), (3) normal walking at
ground level for 5 min, (4) running 200 m and walking
up and down 54 steps over a total time of 4 min, and
(5) cycling for 10 min on a training bike at a frequency
of 80 Hz (Fig. 6). The cartilage deformation of the patella
was 5.9 ± 2.1% after the 30 knee bends, 2.8 ± 0.8% after
walking, 5.0 ± 1.3% after running and 4.5 ± 1.6% after
cycling (Fig. 6), with all changes being significant at P <
0.01. In order to investigate the pattern of cartilage defor-
mation throughout the patella as well, we employed
the matching algorithm described by Stammberger et al.

© 2006 The Authors

Journal compilation © 2006 Anatomical Society of Great Britain and Ireland

Fig. 7 Scheme showing the state of patellar cartilage
deformation (cartilage thickness change) during normal daily
activity and the physiological window of cartilage
deformation between non-weight-bearing conditions and
heavy exercise, such as deep knee bends.
(2000). Cartilage thickness difference maps for the 12
volunteers were displayed using grey value coding.
Averages of the deformation patterns of the 12 volunteers
(for each activity) were then derived, in order to reduce
the noise. They were then recoded in colour intervals,
in order to visualize better the effects of short-term
exercise throughout the patellar surface (Fig. 6). During
squatting and walking, changes were confined to limited
regions of the patellar surface, consistent with the con-
tact areas involved for these activities (Hehne, 1990).
Activities where there was a larger range of knee
motion (running including stairs, cycling, knee bends)
involved a more widespread area of deformation.
When measuring patellar cartilage of volunteers in the
evening after a day of normal activity (no period of rest
prior to imaging) and then after spending the night
sleeping (and without weight-bearing) in the MRI unit,
Sitoci et al. (2003) found that cartilage volume increased
significantly, but only by 2.2% overnight. This finding
is consistent with that of a 3% deformation observed
when walking after a period of non-load-bearing. In
summary, these data demonstrate that during normal
daily activity (walking etc.), the patellar cartilage is at
a state of approximately 2–3% average compression
vs. non-weight-bearing conditions, and that intense
exercise may add another 2–3% of average compres-
sion on top of those encountered during normal daily
activity vs. non-weight-bearing conditions (Fig. 7).
In vivo deformation of femorotibial cartilage
Femorotibial cartilage deformation is more challeng-
ing to investigate compared with the patella, because
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 501
the cartilages are thinner and the precision errors (CV
2–3%) are higher than for patellar cartilage. Waterton
et al. (2000) studied volunteers in the morning and then
after a day of mainly standing activity. They reported
no change in overall femoral cartilage volume and
thickness but cartilage thinning in the femorotibial
contact zones. They observed an increase in thickness
in areas of the femoral condyles and trochlea that were
supposedly not involved in loading during standing,
and they hypothesized that this resulted from negative
intra-articular pressure during joint extension by the
quadriceps muscles. It was alternatively suggested that
interstitial fluid was displaced from load-bearing to
non-load-bearing areas within the cartilage. To inves-
tigate femorotibial cartilage deformation after more
intense activities, we acquired two coronal scans after
a period of 60 min of physical rest in a first session (in
order to reduce the precision error involved), and then
one coronal scan after 30 knee bends in 12 healthy
subjects (Eckstein et al. 2005c). Other activities investig-
ated in the same volunteers included 12 knee bends
performed on one leg only, 2 min of static loading of
the femorotibial joint of one leg at 15° flexion, and ten
jumps from a chair (40 cm height) onto one leg (Fig. 8).
No significant change in cartilage volume was observed
in femorotibial cartilage after two-legged knee
bends (except for the lateral tibia) or one-legged knee
bends (Eckstein et al. 2005c). Significant changes were
observed in the medial and lateral tibial cartilage after
impact loading (jumps from 40 cm height), but not in
the medial or lateral femoral condyle (Fig. 8). Changes
with borderline significance were found in the medial
tibial and lateral femoral condyles after the static
loading exercise (Eckstein et al. 2005c).
When investigating seven female runners after
performing a full distance marathon under competi-
tive conditions (Boston marathon 2005), no significant
deformation of tibiofemoral or patellofemoral carti-
lage was observed 90 min after the race (Kunz et al.
2005). These findings do not exclude the possibility
that some deformation may have been present imme-
diately after the race, but these data show that no pro-
longed deformation occurs after a very intense activity
involving several thousand load cycles. In the same
sample, Williams et al. (2005) found a decrease of the
dGEMRIC index in the medial and lateral femoral con-
dyles 1 day after the race (up to 18%), followed by a
return towards pre-race values at 1 week and recovery
to near pre-race values by 6 weeks post-race. The
502 The effects of exercise on human articular cartilage, F. Eckstein et al.
Fig. 8 Graphs showing the magnitude of tibiofemoral
cartilage deformation (change in mean thickness) and level of
significance after various types of activities: MT = medial tibia;
cMF = central medial femur (condyle); LT = lateral tibia;
cLF = central lateral femur (condyle).
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
tibial cartilage, by contrast, displayed no change in the
dGEMRIC index at day 1; but a significant increase at
week 1. Six weeks after the race, values had returned
to pre-race levels. Because no change in cartilage
thickness was observed immediately after the race
(Kunz et al. 2005) and therefore also is unlikely to have
occurred 1 day, 1 week or 6 weeks after the race, these
data indicate that there may be a loss of GAG after
intense physical activity, and that this may evoke a
biological response that increases cell metabolism and
production of GAG.
Mosher et al. (2005) obtained T2 maps of weight-
bearing femoral and tibial articular cartilage in young
healthy men before and immediately after 30 min of
running. They found no statistically significant change
in T2 profiles of tibial cartilage, but a significant decrease
in T2 of the superficial 40% of weight-bearing femoral
cartilage after exercise. The authors concluded that
these results support the hypothesis that cartilage
compression results in greater anisotropy of superficial
collagen fibres.
In vivo deformation of cartilage as related to cartilage
mechanical properties
As in vivo cartilage deformation is a complex event that
is determined by (1) the load applied to the joint, (2)
the load distribution within the joint during the specific
activity and (3) cartilage mechanical properties, and
because measurements of it are additionally confounded
by precision errors, one cannot equate the deforma-
tional behaviour of cartilage directly with its material
properties. However, in a study of cartilage recovery
after 100 knee bends (Eckstein et al. 1999) we found
that the fluid flux during recovery (no weight bearing)
was highly correlated (r = 0.87) with the magnitude of
deformation observed after the knee bends. Because
the recovery was found to be approximately linear
throughout the 90-min observation period, and because
no external forces acted on the joint during the recovery
period, it is reasonable to assume that the individual
fluid flux throughout the surface observed during
cartilage recovery after knee bends reflects the specific
mechanical properties of the cartilage in this individual.
The high correlation of the flux during recovery with
the magnitude of deformation after knee bends
suggests therefore that MRI-based measurement of
in vivo deformation gives at least some insight into the
individual mechanical properties of the cartilage in vivo.
© 2006 The Authors

When investigating non-osteoarthritic subjects aged
50 –75 years, Hudelmaier et al. (2001) found a smaller
degree of patellar cartilage deformation than in younger
subjects aged 20–30 years. A potential explanation
for the reduced deformation is that older individuals
exhibit different motor strategies (Papa & Cappozzo,
2000) and subject their knee joints to smaller loads
during knee bending. It has, however, been demonstrated
that due to non-enzymatic glycolisation, collagen cross-
links increase with aging (pentosidine), both in animals
and in humans (Takahashi et al. 1995; Brama et al. 1999),
a process that has been found to render the cartilage
matrix stiffer than in the young (Chen et al. 2002;
Verzijl et al. 2002; Saudek & Kay, 2003). The finding of
reduced cartilage deformation after exercise in the
50 –75-year-old subjects therefore lends support to the
concept that changes in cartilage mechanical properties
can be detected in vivo, by examining deformational
behaviour of cartilage after exercise. No difference in
cartilage deformational behaviour between women
and men was observed in this study (Hudelmaier et al.
2001), either in the 20 –30-year-olds or in the 50–75-
year-olds.
In one of the studies previously mentioned (Eckstein
et al. 2005c) we tested the hypothesis that the in vivo
deformation of patellar cartilage is smaller in pro-
fessional athletes than in non-athletic volunteers. This
hypothesis was based on the observation that cartilage
composition and mechanical properties functionally
adapt to mechanical stimulation in some animal experi-
ments (reviewed by Vanwanseele et al. 2002b), whereas
another study showed that continuous training of
dogs did not alter the compositional or mechanical
properties of articular cartilage, even though the
animal had been training throughout life (Newton
et al. 1997). These different outcomes of animal models
may be potentially explained by differences in suscept-
ibility to mechanical stimuli at different levels of skeletal
maturity. However, there is as yet no convincing evidence
that human cartilage composition can be changed by
mechanical stimulation, and in particular changed
to a degree that it affects its in vivo deformational
behaviour. We therefore compared patellar cartilage
deformational behaviour in 14 men who had never
performed regular training of muscle strength with
seven professional weightlifters, and seven professional
bobsleigh sprinters. The weightlifters were regional
champions and included one world champion. The
bobsleigh sprinters were national finalists and included
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 503
three world champions and one Olympic medallist. The
reduction in patellar cartilage volume after 30 knee
bends was 4.1 ± 2.6% in the non-athletic participants,
2.9% ± 1.9% in the weightlifters and 3.9 ± 1.8% in the
bobsleigh sprinters. Although there was a trend for the
weightlifters to display a lower magnitude of deforma-
tion, the difference between the groups was not statis-
tically significant.
To date, there are no data on in vivo cartilage deforma-
tion in osteoarthritic patients, and obviously there are
ethical problems in subjecting patients with knee
pain to intense exercise protocols. However, cartilage is
known to become more compliant (less stiff) with osteo-
arthritis, and standardized in vivo loading protocols
may eventually be used clinically to evaluate functional
properties over the course of joint disease, or even to
study the effects of different types of therapy on carti-
lage quality (Burstein & Gray, 2003).
Long-term effects of exercise on articular
cartilage (functional adaptation)
Intersubject variability, side differences and
correlation with anthropometric measures
Numerous studies have reported a high degree of
intersubject variability in cartilage volume, thickness
and surface areas of the human knee (Jones et al. 2000;
Eckstein et al. 2001a,b; Hudelmaier et al. 2001, 2003;
Cicuttini et al. 2002, 2003; Burgkart et al. 2003), and
also in other joints of the body (Springer et al. 1998;
Graichen et al. 2000, 2003; Al Ali et al. 2002). Hudel-
maier et al. (2003) showed that muscle cross-sectional
areas (MCSAs) of the thigh and calves can be reproduc-
ibly measured using a spin echo magnetic resonance
sequence and are more highly correlated with cartilage
morphology than body height or body weight. Also,
MCSAs contributed significant, independent informa-
tion in multiple regression models, with cartilage
morphology as the dependent varaible and body weight
and height as independent variables. These models
predicted approximately 75% of the variability in
cartilage volume (Hudelmaier et al. 2003). No signifi-
cant differences were found in knee cartilage mor-
phology of the left and right limbs of volunteers, and
the left–right differences were substantially lower
than the intersubject variability. This applied to carti-
lage volume, thickness and surface area (Eckstein et al.
2002c). The side differences were not associated with
504 The effects of exercise on human articular cartilage, F. Eckstein et al.
lower limb dominance, i.e. subjects with dominance
of the right lower limb did not display systematically
higher cartilage volume, thickness or surface areas
on that side. However, side differences in MCSAs were
moderately correlated with side differences in cartilage
volume, thickness and surface areas, i.e. subjects with
higher thigh MCSAs on the right tended to have higher
cartilage volume, thickness and joint surface areas in
the right knee (Eckstein et al. 2002c).
Cicuttini et al. (1999) reported sex differences of
cartilage volume in adults, and Jones et al. (2000) in
children and adolescents. The sex differences remained
significant after adjusting for age, body weight, height
and femoral (condylar) bone volume. Faber et al. (2001)
also observed significant sex differences of cartilage
volume in the medial tibia (+43% in men) and lateral
tibia (+47%), and smaller (albeit significant) differences
in the patella (+20%) and femur (+27%). It was shown,
however, that gender differences in cartilage volume
originated mainly from differences in the joint surface
area size (total knee = +23% in men; P < 0.01), but to a
lesser extent from differences in cartilage thickness
(total knee = +8% in men; difference not statistically
significant) (Faber et al. 2001). Eckstein et al. (2001a)
reported that after matching men and women with
identical body weight and height, men did not display
significantly higher cartilage thickness than women,
but still had significantly larger joint surface areas.
This suggests that even when being matched for body
height and weight, women have smaller knee joint sur-
faces than men. Eckstein et al. (2004) further examined
whether gender differences in joint surface areas of
the ankle joint were smaller than in the knee (based
on the notion that women suffer more often from knee
osteoarthritis than men, but men more often from
osteoarthritic changes in the ankle). However, the authors
found the sex differences to be similar for both joints.
A source of controversy has been whether cartilage
thinning occurs during the normal aging process
(possibly as a result of reduction in mechanical loading)
or whether the decrease only affects people with
arthritis. In human specimens, Meachim (1971) found
no significant decrease of cartilage thickness in the
human shoulder with age, whereas in the patella
Meachim et al. (1977) observed cartilage thinning, in
particular in women over the age of 50 years. These
changes were attributed to osteoarthritis rather than
to aging as such. Karvonen et al. (1994) concluded from
local measurements in MR images that age accounted
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
for a significant linear decrease in knee cartilage thick-
ness both in the presence and in the absence of osteo-
arthritis. These findings were, however, confined to
the site of most frequent femorotibial contact of the
lateral and medial femoral condyle but did not apply to
the patella, the tibia or the posterior aspects of the
femoral condyles. Hudelmaier et al. (2001) examined
men and women aged 50–75 years with no history of
knee pain, trauma or surgery. In the patella, they found
no significant difference in cartilage thickness com-
pared with men aged 18–40 years (−6%), but a signific-
antly lower thickness (−12%; P < 0.05) compared with
women aged 18–40 years. Interestingly, women also
displayed a larger decrease of the quadriceps cross-
sectional area with age than men. For other joint surfaces
of the knee they estimated the amount of cartilage
thinning with age to be approximately 4% per decade
and rates were similar for both sexes (Hudelmaier et al.
2003).
Reduced loading conditions
As stated above, mechanical stimuli are known to
represent potent regulators of muscle and bone tissue
mass. Animal studies have suggested that cartilage
thickness decreases during immobilization but investiga-
tions in animals with increased levels of exercise have
produced inconclusive and partly contradictory
results (Helminen et al. 1992; Newton et al. 1997; for
a recent review see Vanwanseele et al. 2002b). In a
cross-sectional study, Vanwanseele et al. (2002a) exam-
ined the knee joints of paraplegic patients at 6, 12 and
24 months after injury and found the cartilage thick-
ness to be significantly reduced in relation to healthy
subjects of the same sex. The authors confirmed a
reduction in cartilage thickness in paraplegic patients
in a 12-month longitudinal study (Vanwanseele et al.
2003), in which they reported an annual reduction in
cartilage thickness of > 10% in all compartments of the
knee, whereas no significant changes of cartilage
thickness were found in the shoulder (Vanwanseele
et al. 2004). Hinterwimmer et al. (2004) recently showed
that even during short-term reduced loading condi-
tions (7 weeks of partial weight bearing at the knee,
after a surgical intervention at the ankle) there was a
significant degree of cartilage thinning (Fig. 9). Partial
unloading was associated with an average reduction of
the quadriceps muscle cross-sectional area (as measured
with MRI) of 11% (Fig. 9), but no significant increase
© 2006 The Authors

Fig. 9 Graph showing the average change in quadriceps cross-
sectional area and in cartilage thickness during 7 weeks of
partial weight bearing of the knee (sole contact) in a group of
20 volunteers following a fracture of the ankle joint.
or decrease in cartilage volume or thickness was
observed in the contra-lateral limb that was subject to
increased loading during the 7-week period. These
data indicate that cartilage undergoes some process of
atrophy in the absence of mechanical stimulation. The
findings may have important implications for the clini-
cal management of the postoperative period in bone
and joint surgery, and also for long-term space travel in
the context of the international space station or a jour-
ney to Mars. These observations in particular raise the
question as to what extent morphological changes
(cartilage thinning) during reduced weight-bearing
conditions are reversible.
In immature beagle dogs, Kiviranta et al. (1994)
reported that after 11 weeks of immobilization and
15 weeks of remobilization, the GAG content was fully
restored in most joint compartments, but not in the
peripheral regions of the femoral condyles. They also
found an incomplete restoration of femoral cartilage
thickness after remobilization and reasoned that there
was an inability of previously unloaded cartilage to with-
stand physiological loads during the remobilization
process (Kiviranta et al. 1994). The same group reported
a decrease in indentation stiffness after 11 weeks of
immobilization, which was not fully restored during a
50-week remobilization period (Haapala et al. 1999). The
authors claimed that extended immobilization of a joint
may cause long-lasting biochemical and biomechanical
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 505
alterations of the cartilage, specifically the proteo-
glycans, that may jeopardize its morphological integrity
and mechanical competence (Haapala et al. 1999).
Hudelmaier et al. (2001) examined a single volunteer
after 6 weeks of complete non-load-bearing and restric-
tion of knee joint motion to 0–30° of flexion. They
found a 14% lower patellar cartilage thickness on the
immobilized side, but no side differences in tibial carti-
lage morphology. At the same time, side differences in
quadriceps cross-sectional area amounted to 36%
after immobilization. During a 24-month remobilization
period with intensive physiotherapy and weight train-
ing over the first 7 months, side differences in patellar
cartilage thickness and quadriceps cross-sectional
area reduced to 2 and 9%, respectively, within the
first 9 months, and to 8 and 4%, respectively, over the
entire observation period (24 months) (Fig. 10). Patellar
cartilage deformation was 9% 4 weeks after immobiliza-
tion (the first time point at which the participant was
able to perform deep knee bends) and was lower at later
time points (1–6% at 6–18 months). These findings
indicate that the stiffness of the patellar cartilage may
be reduced after immobilization, probably as a conse-
quence of biochemical alterations of the cartilage.
Increased loading conditions
In a cross-sectional study of 92 children (age 9–18 years),
Jones et al. (2000) reported a significant association
between cartilage volume and self-reported level of
exercise in children. The authors observed that physical
Fig. 10 Graph showing the side differences of quadriceps
cross-sectional area (CSA) and mean patellar cartilage
thickness (in percentage left vs. right) during a 24-month
remobilization period in one subject, after 6 weeks of
immobilization of the left limb, with no weight bearing and
restriction of knee movement from 0° to 30° of flexion.
506 The effects of exercise on human articular cartilage, F. Eckstein et al.
activity was a significant explanatory factor for cartilage
volume at all sites in the knee (r2 = 7–14%, P < 0.05)
and that the most consistent association was with
vigorous activity in the 2 weeks prior to imaging. In a
longitudinal study of 74 children of the same cohort
(observation period 1.3 –1.9 years) the authors (Jones
et al. 2003) found an association of average intensity of
sport (again evaluated by responses to a questionnaire).
Subjects above the median were found to gain approx-
imately twice as much cartilage as those below the
median at tibial (but not patellar) cartilage. From these
studies, the authors concluded that cartilage develop-
ment is amenable to modification. Note, however, that
analysis of cartilage volume does not permit one to
separate the effect on cartilage development (thick-
ness) and bone growth (epiphyseal joint surface area).
When investigating mature adults, however, Eckstein
et al. (2002a) found no difference in cartilage thickness
in triathletes, who had been training for at least 10 h
per week over the last 3 years and had also been physi-
cally active throughout childhood and adolescence in
comparison with individuals who had never been physi-
cally active (< 1 h sport per week throughout life), had
no job that involved physical activity, and had a normal
body mass index (Eckstein et al. 2002a). The latter find-
ings suggest that the relatively great variability in
cartilage thickness observed between subjects is not
readily explained by variability in mechanical loading his-
tory (Carter et al. 1991). Interestingly, however, triath-
letes displayed somewhat larger knee joint surface
areas (+9%, P < 0.01 in men; +7%, P = 0.08 in women).
These observations indicate that the biological mecha-
nism to reduce high stress at the articular surface may
be by an increase in the area of the load-bearing
surface rather than an increase in cartilage thickness
(Eckstein et al. 2002a). A potential reason for this is that
beyond a certain thickness the nutritive situation of
the cartilage becomes critical, and/or that the stress
distribution (load partitioning) within the cartilage
becomes unfavourable with thicker cartilage. With
thicker cartilage, there is more space for the interstitial
fluid to escape laterally from the site of contact and
hydrostatic pressurization is reduced. With larger con-
tact areas, by contrast, the force is distributed onto a
wider area, keeping the stress at the joint surface within
reasonable limits, and the mechanism of hydrostatic
pressurization of the interstitial fluid is enhanced.
Tiderius et al. (2003) compared the dGEMRIC index in
(1) non-exercising individuals, (2) individuals physically
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
exercising twice per week and (3) elite runners. The
index was found to be significantly higher in individuals
who exercised regularly, and still higher in the elite
runners. The authors concluded that human knee
cartilage may adapt to exercise by increasing the GAG
content, but that a higher proportion of extracellular
body water (a larger distribution volume for the contrast
agent applied) may also to some extent explain the
higher dGEMRIC index values in elite runners. Recently,
the same group (Roos & Dahlberg, 2005) reported that
the dGEMRIC index increased over 4 months in 30
subjects (who had undergone partial medial meniscectomy
3–5 years previously) during a supervised, three-times
weekly exercise protocol. There also was a relatively
high correlation (r = 0.74) between change in T1 (Gd)
and the self-reported change in physical activity level,
over the 4-month period. These results indicate that
compositional adaptation (increase in GAG content) of
cartilage may occur over relatively short time intervals
in response to exercise and mechanical loading.
Gratzke et al. (2002) compared knee cartilage
morphology in 14 elite athletes (seven weightlifters
and seven bobsleigh sprinters) with 14 men who had
never performed strength training (for a more detailed
description of the participants see above). All athletes
had been actively training throughout adolescence, well
before reaching skeletal maturity. The weightlifters
displayed 26% higher (P < 0.01) extensor forces, as
measured with a Cybex dynamometer, and 30% higher
(P < 0.001) extensor MCSAs than the non-athletic
volunteers. The same applied to the bobsleigh sprinters
(+ 43%/+23%, P < 0.001, respectively). The cartilage
thickness, articular joint surface area and chondro-
osseous interface area were not significantly greater in
the group of athletes compared with the group of non-
athletic volunteers, except for the cartilage thickness of
the patella. This was 14% greater (P < 0.01) in the weight-
lifters and 17% greater (P < 0.01) in the sprinters.
Nevertheless, there was still a wide overlap in patellar
cartilage thickness values between the athletes and
non-athletic participants. The findings of this (and
previously cited) studies indicate that differences in
mechanical environment may not readily explain
the relatively large variability of cartilage morphology
observed between human subjects, despite enormous
interindividual differences in mechanical loading
history throughout the relevant developmental period
(adolescence) in which endochondral ossification is
completed and cartilage form is determined. These
© 2006 The Authors

findings, however, refer to the average cartilage
thickness throughout joint surfaces, and adaptation at
a regional level cannot be discounted. Why there was
a trend towards increased patellar cartilage thickness
in athletes with high muscle strength, and a trend
towards larger joint surface areas in athletes with high
endurance is currently unclear, but these findings will
have to be confirmed in larger samples and in athletes
with different types of specializations. Future studies
should also include regional analyses of cartilage thick-
ness, which have so far not been presented in the con-
text of the response of cartilage to exercise.
Genetic influence
In the interpretation of the cumulative data reviewed
in this article, environmental (mechanical) factors appear
to play only a small (if any) role in determining cartilage
morphology in adults, although a wide variability of
cartilage thickness is observed in the population,
even in the absence of arthritic changes (Eckstein et al.
2001a; Burgkart et al. 2003; Hudelmaier et al. 2003).
One potential explanation for the intersubject differ-
ences observed in cartilage morphology is genetic dif-
ferences between individuals.
Antoniades et al. (2001) examined the minimal joint
space width of hip radiographs of 222 monozygotic (MZ)
and 240 dizygotic (DZ) twins. They found that genetic
factors accounted for most of the variation in joint space
width, but were not able to make direct measurements
of cartilage thickness. Hunter et al. (2003) examined 31
MZ and 37 DZ twin pairs to assess the relative contribu-
tion of genetic and environmental factors to knee cartilage
volume. The heritability was estimated to range from
61 to 76% for different compartments of the knee. How-
ever, the authors used an MR sequence (T2-weighted,
fat-saturated sagittal gradient echo) that has not been
validated for the purpose of cartilage volume measure-
ments, and cartilage volume is strongly dependent on
bone size, so that the genetic contribution to cartilage
(thickness) formation is difficult to estimate.
Zhai et al. (2004) investigated the heritability of cartilage
volume and bone size by examining sibling pairs. They
estimated the heritability of cartilage volume to be
65 – 84% for various knee compartments, and that of
bone size to be 57–85%. Again, however, cartilage
thickness was not determined directly.
Siedek et al. (2002) examined 12 MZ twin pairs to deter-
mine knee cartilage thickness and joint surface areas as
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 507
separate entities. They observed a remarkable similarity
in cartilage morphology between twins (coefficient
of variation 3.2% for cartilage thickness and 2.2% for
joint surface area), in comparison with the substantially
larger variability between subjects for a young reference
population of 117 subjects (CV 12 and 11% for cartilage
thickness in women and men, respectively, and 10 and
9% for joint surface areas). These findings suggest that
not only bone size (or joint surface areas), but also
cartilage thickness appears to be strongly determined
by genetics. The lack of correlation in cartilage thick-
ness between the knee and ankle (Eckstein et al. 2004),
however, indicates that different genes may be respon-
sible for cartilage thickness in different joints of the
human body.
Conclusion
Taken together, these findings indicate that differences
in cartilage form between individuals cannot be readily
explained by functional adaptation to mechanics, and
that the emergence and maintenance of cartilage form
appears to depend on information from the genome.
Although hyaline articular cartilage appears to display
atrophic changes (thinning) during unloading and may
exhibit compositional changes (increase in GAG) after
exercise, it seems to differ from other musculoskeletal
tissues with load-bearing function as it cannot increase
tissue mass postnatally as a result of mechanical stimu-
lation (Lamarck, 1809; Darwin, 1872; Roux, 1881; Wolff,
1892; Pauwels, 1980; Carter et al. 1991; Huiskes et al.
2000). The specific signals that stop the ossification
front from progressing to the joint surface (and that
preserve a cartilage layer of given thickness) remain
enigmatic. However, our results clearly suggest that,
contrary to general expectation, mechanical feedback
does not play a relevant role in this process and, in con-
trast to bone, does not serve to regulate the complex
biochemical metabolic machinery towards lasting
optimality of cartilage form (Huiskes et al. 2000). In
attempting to address the question of why (in contrast
to muscle and bone) cartilage does not appear to
adapt its mass to mechanical stimulation, the following
aspects need to be considered. (1) Loss of cartilage and
joint function is a problem generally encountered by
individuals after their reproductive period and – in con-
trast to bone fracture and muscle weakness – this may
have not created evolutionary pressures for the tissue
to be able to adapt to mechanical usage. (2) Whereas
508 The effects of exercise on human articular cartilage, F. Eckstein et al.
too much bone or too much muscle are metabolically
expensive and provide a clear disadvantage for fast
locomotion, a slight increase in cartilage thickness has
no known negative consequences on metabolism and
speed of locomotion. Therefore, there may have been
less of an evolutionary pressure to adapt cartilage
morphology to the immediate mechanical demands
encountered by the individual. (3) Lastly, whereas ‘more’
muscle provides more tensile strength, and ‘more’ bone
provides higher structural compressive and bending
strength and hence better protection against fractures,
‘more’ cartilage is not known to be associated with
improved mechanical competence of joints. There may
thus exist a decoupling between functional competence
and tissue mass. Hydrostatic pressurization provides a
mechanism by which cartilage is able to distribute joint
forces evenly onto the subchondral bone (Ateshian et al.
1994; Mow et al. 2003), protect itself from mechanical
damage, and probably also provides almost frictionless
surfaces during dynamic motion (Krishnan et al. 2003).
This mechanism is probably more difficult to accom-
plish in a thick cartilage layer, in which the interstitial
fluid can escape laterally from the site of contact. These
aspects may provide some explanation as to why carti-
lage is less (if at all) responsive to increased mechanical
loading. However, the inability of cartilage to adapt to
its mechanical environment may be coupled with its
inability to repair following mechanical and other insults
(Hunziker et al. 2002). This inability may be essential
in understanding the deleterious consequences of
cartilage loss in degenerative joint disease and the
large burden of osteoarthritis on the aging population
(Yelin & Callahan, 1995).
References
Al Ali D, Graichen H, Faber S, et al. (2002) Quantitative carti-
lage imaging of the human hind foot: precision and inter-
subject variability. J Orthop Res 20, 249 –256.
Antoniades L, Spector TD, MacGregor AJ (2001) The genetic
contribution to hip joint morphometry and relationship to
hip cartilage thickness. Osteoarthritis Cartilage 9, 593–595.
Ateshian GA, Soslowsky LJ, Mow VC (1991) Quantitation of
articular surface topography and cartilage thickness in
knee joints using stereophotogrammetry. J Biomech 24,
761–776.
Ateshian GA, Lai WM, Zhu WB, Mow VC (1994) An asymptotic
solution for the contact of two biphasic cartilage layers. J
Biomech 27, 1347–1360.
Ateshian G, Mow VC (2005) Friction, lubrication, and wear of
articular cartilage and diathrodial joints. In Basic Orthopaedic
Biomechanics and Mechanobiology, 3rd edn (eds Mow VC,
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
Huskes R), pp. 447– 494. Philadelphia: Lippincott, Williams &
Wilkins.
Bashir A, Gray ML, Boutin RD, Burstein D (1997) Glyco-
saminoglycan in articular cartilage: in vivo assessment with
delayed Gd (DTPA) (2-)-enhanced MR imaging. Radiology
205, 551–558.
Bashir A, Gray ML, Hartke J, Burstein D (1999) Nondestructive
imaging of human cartilage glycosaminoglycan concentra-
tion by MRI. Magn Reson Med 41, 857– 865.
Bergmann G, Graichen F, Rohlmann A (1993) Hip joint loading
during walking and running, measured in two patients. J
Biomech 26, 969 – 990.
Booth FW (1994) Terrestrial applications of bone and muscle
research in microgravity. Adv Space Res 14, 373 –376.
Brama PA, Tekoppele JM, Bank RA, Van Weeren PR, Barneveld A
(1999) Influence of site and age on biochemical characteristics
of the collagen network of equine articular cartilage. Am J
Vet Res 60, 341–345.
Buckwalter JA, Mankin HJ (1998a) Articular cartilage: degen-
eration and osteoarthritis, repair, regeneration, and trans-
plantation. Instr Course Lect 47, 487–504.
Buckwalter JA, Mankin HJ (1998b) Articular cartilage: tissue
design and chondrocyte–matrix interactions. Instr Course
Lect 47, 477–486.
Burgkart R, Glaser C, Hyhlik-Durr A, et al. (2001) Magnetic
resonance imaging-based assessment of cartilage loss in severe
osteoarthritis: accuracy, precision, and diagnostic value.
Arthritis Rheum 44, 2072–2077.
Burgkart R, Glaser C, Hinterwimmer S, et al. (2003) Feasibility
of T and Z scores from magnetic resonance imaging data for
quantification of cartilage loss in osteoarthritis. Arthritis
Rheum 48, 2829 –2835.
Burstein D, Bashir A, Gray ML (2000) MRI techniques in early
stages of cartilage disease. Invest Radiol 35, 622– 638.
Burstein D, Gray M (2003) New MRI techniques for imag-
ing cartilage. J Bone Joint Surg Am 85, Supplement 2, 70–
77.
Carter DR, Wong M (1988a) Mechanical stresses and endo-
chondral ossification in the chondroepiphysis. J Orthop Res
6, 148 –154.
Carter DR, Wong M (1988b) The role of mechanical loading
histories in the development of diarthrodial joints. J Orthop
Res 6, 804 – 816.
Carter DR, Wong M, Orr TE (1991) Musculoskeletal ontogeny,
phylogeny, and functional adaptation. J Biomech 24 (Suppl. 1),
3 –16.
Carter DR, Wong M (2003) Modelling cartilage mechanobiology.
Philos Trans R Soc Lond B Biol Sci 358, 1461–1471.
Chen AC, Temple MM, Ng DM, et al. (2002) Induction of
advanced glycation end products and alterations of the
tensile properties of articular cartilage. Arthritis Rheum 46,
3212–3217.
Cicuttini F, Forbes A, Morris K, et al. (1999) Gender differences
in knee cartilage volume as measured by magnetic resonance
imaging. Osteoarthritis Cartilage 7, 265 –271.
Cicuttini F, Forbes A, Asbeutah A, Morris K, Stuckey S (2000)
Comparison and reproducibility of fast and conventional
spoiled gradient-echo magnetic resonance sequences in the
determination of knee cartilage volume. J Orthop Res 18,
580 –584.
© 2006 The Authors

Cicuttini FM, Wluka AE, Wang Y, et al. (2002) Compartment
differences in knee cartilage volume in healthy adults. J
Rheumatol 29, 554–556.
Cicuttini FM, Wluka AE, Forbes A, Wolfe R (2003) Comparison
of tibial cartilage volume and radiologic grade of the tibio-
femoral joint. Arthritis Rheum 48, 682– 688.
Cohen ZA, McCarthy DM, Kwak SD, et al. (1999) Knee cartilage
topography, thickness, and contact areas from MRI: in-vitro
calibration and in-vivo measurements. Osteoarthritis Cartilage
7, 95–109.
Cohen ZA, Roglic H, Grelsamer RP, et al. (2001) Patellofemoral
stresses during open and closed kinetic chain exercises. An
analysis using computer simulation. Am J Sports Med 29,
480–487.
Cohen ZA, Mow VC, Henry JH, Levine WN, Ateshian GA (2003)
Templates of the cartilage layers of the patellofemoral joint
and their use in the assessment of osteoarthritic cartilage
damage. Osteoarthritis Cartilage 11, 569 –579.
Darwin C (1872) The Origin of Species, 6th edn. New York:
New America Library.
Eckstein F, Sittek H, Milz S, et al. (1995) The potential of
magnetic resonance imaging (MRI) for quantifying articular
cartilage thickness – a methodological study. Clin Biomech
(Bristol, Avon) 10, 434 – 440.
Eckstein F, Gavazzeni A, Sittek H, et al. (1996a) Determination
of knee joint cartilage thickness using three-dimensional
magnetic resonance chondro-crassometry (3D MR-CCM).
Magn Reson Med 36, 256 –265.
Eckstein F, Sittek H, Gavazzeni A, et al. (1996b) Magnetic
resonance chondro-crassometry (MR CCM): a method for
accurate determination of articular cartilage thickness? Magn
Reson Med 35, 89 –96.
Eckstein F, Schnier M, Haubner M, et al. (1998a) Accuracy of
cartilage volume and thickness measurements with mag-
netic resonance imaging. Clin Orthop 352, 137–148.
Eckstein F, Tieschky M, Faber SC, et al. (1998b) Effect of phys-
ical exercise on cartilage volume and thickness in vivo: MR
imaging study. Radiology 207, 243 –248.
Eckstein F, Tieschky M, Faber S, Englmeier KH, Reiser M (1999)
Functional analysis of articular cartilage deformation,
recovery, and fluid flow following dynamic exercise in vivo.
Anat Embryol (Berl) 200, 419 – 424.
Eckstein F, Lemberger B, Stammberger T, Englmeier KH,
Reiser M (2000a) Patellar cartilage deformation in vivo after
static versus dynamic loading. J Biomech 33, 819 – 825.
Eckstein F, Lemberger B, Stammberger T, Englmeier KH,
Reiser M (2000b) Patellar cartilage deformation in vivo after
static versus dynamic loading. J Biomech 33, 819 – 825.
Eckstein F, Reiser M, Englmeier KH, Putz R (2001a) In vivo mor-
phometry and functional analysis of human articular cartilage
with quantitative magnetic resonance imaging – from image
to data, from data to theory. Anat Embryol (Berl) 203, 147–173.
Eckstein F, Winzheimer M, Hohe J, Englmeier KH, Reiser M
(2001b) Interindividual variability and correlation among
morphological parameters of knee joint cartilage plates:
analysis with three-dimensional MR imaging. Osteoarthritis
Cartilage 9, 101–111.
Eckstein F, Faber S, Muhlbauer R, et al. (2002a) Functional
adaptation of human joints to mechanical stimuli. Osteoar-
thritis Cartilage 10, 44 – 50.
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 509
Eckstein F, Heudorfer L, Faber SC, et al. (2002b) Long-term and
resegmentation precision of quantitative cartilage MR
imaging (qMRI). Osteoarthritis Cartilage 10, 922–928.
Eckstein F, Muller S, Faber SC, et al. (2002c) Side differences of
knee joint cartilage volume, thickness, and surface area,
and correlation with lower limb dominance – an MRI-based
study. Osteoarthritis Cartilage 10, 914 –921.
Eckstein F, Siedek V, Glaser C, et al. (2004) Correlation and sex
differences between ankle and knee cartilage morphology
determined by quantitative magnetic resonance imaging.
Ann Rheum Dis 63, 1490 –1495.
Eckstein F, Charles HC, Buck RJ, et al. (2005a) Accuracy and
precision of quantitative assessment of cartilage morpho-
logy by magnetic resonance imaging at 3.0T. Arthritis Rheum
52, 3132–3136.
Eckstein F, Hudelmaier M, Wirth W, et al. (2005b) Double echo
steady state (DESS) Magnetic resonance imaging of knee
articular cartilage at 3 Tesla – a pilot study for the osteoarthritis
initiative. Ann Rheum Dis, doi: 10.113b/ard.2005.039370.
Eckstein F, Lemberger B, Gratzke C, et al. (2005c) In vivo cartilage
deformation after different types of activity and its dependence
on physical training status. Ann Rheum Dis 64, 291–295.
Faber SC, Eckstein F, Lukasz S, et al. (2001) Gender differences
in knee joint cartilage thickness, volume and articular sur-
face areas: assessment with quantitative three-dimensional
MR imaging. Skeletal Radiol 30, 144–150.
Frahm J, Haase A, Matthaei D (1986) Rapid three-dimensional
MR imaging using the FLASH technique. J Comput Assist
Tomogr 10, 363 –368.
Froimson MI, Ratcliffe A, Gardner TR, Mow VC (1997) Differ-
ences in patellofemoral joint cartilage material properties
and their significance to the etiology of cartilage surface
fibrillation. Osteoarthritis Cartilage 5, 377–386.
Glaser C, Faber S, Eckstein F, et al. (2001) Optimization and
validation of a rapid high-resolution T1-w 3D FLASH water
excitation MRI sequence for the quantitative assessment
of articular cartilage volume and thickness. Magn Reson
Imaging 19, 177–185.
Gold GE, Han E, Stainsby J, et al. (2004a) Musculoskeletal MRI
at 3.0 T: relaxation times and image contrast. AJR Am J
Roentgenol 183, 343 –351.
Gold GE, Suh B, Sawyer-Glover A, Beaulieu C (2004b) Muscu-
loskeletal MRI at 3.0 T: initial clinical experience. AJR Am J
Roentgenol 183, 1479 –1486.
Graichen H, Eisenhart-Rothe R, Vogl T, Englmeier KH, Eckstein F
(2004) Quantitative assessment of cartilage status in osteo-
arthritis by quantitative magnetic resonance imaging:
technical validation for use in analysis of cartilage volume
and further morphologic parameters. Arthritis Rheum 50,
811– 816.
Graichen H, Jakob J, Eisenhart-Rothe R, et al. (2003) Validation
of cartilage volume and thickness measurements in the human
shoulder with quantitative magnetic resonance imaging.
Osteoarthritis Cartilage 11, 475– 482.
Graichen H, Springer V, Flaman T, et al. (2000) Validation
of high-resolution water-excitation magnetic resonance
imaging for quantitative assessment of thin cartilage layers.
Osteoarthritis Cartilage 8, 106 –114.
Gratzke C, Glaser C, Englmeier KH, Reiser M, Eckstein F (2002)
Comparison of cartilage morphology in professional weight-
510 The effects of exercise on human articular cartilage, F. Eckstein et al.
lifters and sprinters with normal volunteers suggests that human
articular cartilage cannot adapt to functional stimulation.
Osteoarthritis Cartilage 10 [(Suppl. A)], S11 [abstract].
Gray ML, Eckstein F, Peterfy C, et al. (2004) Toward imaging
biomarkers for osteoarthritis. Clin Orthop 427, S175–S181.
Haapala J, Arokoski JP, Hyttinen MM, et al. (1999) Remobili-
zation does not fully restore immobilization induced articu-
lar cartilage atrophy. Clin Orthop 00, 218–229.
Hardy PA, Recht MP, Piraino DW (1998) Fat suppressed MRI of
articular cartilage with a spatial-spectral excitation pulse. J
Magn Reson Imaging 8, 1279–1287.
Hehne HJ (1990) Biomechanics of the patellofemoral joint and
its clinical relevance. Clin Orthop 00, 73–85.
Helminen HJ, Kiviranta I, Saamanen AM, et al. (1992) Effect of
motion and load on articular cartilage in animal models. In
Articular Cartilage and Osteoarthritis (eds Kuettner KE,
Roth A), pp. 501–510. New York: Raven.
Herberhold C, Stammberger T, Faber S, et al. (1998) An MR-
based technique for quantifying the deformation of articu-
lar cartilage during mechanical loading in an intact cadaver
joint. Magn Reson Med 39, 843–850.
Herberhold C, Faber S, Stammberger T, et al. (1999) In situ
measurement of articular cartilage deformation in intact
femoropatellar joints under static loading. J Biomech 32,
1287–1295.
Hinterwimmer S, Krammer M, Krotz M, et al. (2004) Cartilage
atrophy in the knees of patients after seven weeks of partial
load bearing. Arthritis Rheum 50, 2516–2520.
Hohe J, Ateshian G, Reiser M, Englmeier KH, Eckstein F (2002)
Surface size, curvature analysis, and assessment of knee
joint incongruity with MRI in vivo. Magn Reson Med 47,
554–561.
Hudelmaier M, Glaser C, Hohe J, et al. (2001) Age-related
changes in the morphology and deformational behavior of
knee joint cartilage. Arthritis Rheum 44, 2556–2561.
Hudelmaier M, Glaser C, Englmeier KH, et al. (2003) Correla-
tion of knee-joint cartilage morphology with muscle cross-
sectional areas vs. anthropometric variables. Anat Rec 270A,
175–184.
Huiskes R, Ruimerman R, Van Lenthe GH, Janssen JD (2000)
Effects of mechanical forces on maintenance and adapta-
tion of form in trabecular bone. Nature 405, 704–706.
Hunter DJ, Snieder H, March L, Sambrook PN (2003) Genetic
contribution to cartilage volume in women: a classical twin
study. Rheumatology (Oxford) 42, 1495–1500.
Hunziker EB (2002) Articular cartilage repair: basic science and
clinical progress. A review of the current status and pros-
pects. Osteoarthritis Cartilage 10, 432–463.
Hunziker EB, Quinn TM, Hauselmann HJ (2002) Quantitative
structural organization of normal adult human articular car-
tilage. Osteoarthritis Cartilage 10, 564–572.
Hyhlik-Dürr A, Faber S, Burgkart R, et al. (2000) Precision of
tibial cartilage morphometry with a coronal water-excitation
MR sequence. Eur Radiol 10, 297–303.
Jones G, Glisson M, Hynes K, Cicuttini F (2000) Sex and site dif-
ferences in cartilage development: a possible explanation
for variations in knee osteoarthritis in later life. Arthritis
Rheum 43, 2543–2549.
Jones G, Ding C, Glisson M, et al. (2003) Knee articular carti-
lage development in children: a longitudinal study of the
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
effect of sex, growth, body composition, and physical activ-
ity. Pediatr Res 54, 230–236.
Karvonen RL, Negendank WG, Teitge RA, et al. (1994) Factors
affecting articular cartilage thickness in osteoarthritis and
aging. J Rheumatol 21, 1310 –1318.
Keller TS, Strauss AM, Szpalski M (1992) Prevention of bone
loss and muscle atrophy during manned space flight. Micro-
gravity Q 2, 89 –102.
Kim YJ, Bonassar LJ, Grodzinsky AJ (1995) The role of cartilage
streaming potential, fluid flow and pressure in the stimula-
tion of chondrocyte biosynthesis during dynamic compres-
sion. J Biomech 28, 1055 –1066.
Kiviranta I, Tammi M, Jurvelin J, et al. (1994) Articular carti-
lage thickness and glycosaminoglycan distribution in the
young canine knee joint after remobilization of the immo-
bilized limb. J Orthop Res 12, 161–167.
Kornaat PR, Reeder SB, Koo S, et al. (2005) MR imaging of
articular cartilage at 1.5T and 3.0T: comparison of SPGR and
SSFP sequences. Osteoarthritis Cartilage 13, 338–344.
Krishnan R, Park S, Eckstein F, Ateshian GA (2003) Inhomoge-
neous cartilage properties enhance superficial interstitial
fluid support and frictional properties, but do not provide a
homogeneous state of stress. J Biomech Eng 125, 569–577.
Kshirsagar AA, Watson PJ, Tyler JA, Hall LD (1998) Measure-
ment of localized cartilage volume and thickness of human
knee joints by computer analysis of three-dimensional
magnetic resonance images. Invest Radiol 33, 289 –299.
Kunz M, Williams A, McKenzie C, Burstein D, Eckstein F (2005)
No prolonged deformation of knee joint cartilage after the
Boston marathon. Osteoarthritis Cart 13 (Suppl. A), 124
[abstract].
Lamarck JB (1809) Philosophie Zoologique. Paris: Bailleres.
Liess C, Lüsse S, Karger N, Heller M, Glüer CC (2002) Detection
of changes in cartilage water content using MRI T2-mapping
in vivo. Osteoarthritis Cartilage 10, 907–913.
Lusse S, Claassen H, Gehrke T, et al. (2000) Evaluation of water
content by spatially resolved transverse relaxation times of
human articular cartilage. Magn Reson Imaging 18, 423 –430.
Lynch JA, Zaim S, Zhao J, Peterfy CG, Genant HK (2001) Auto-
matic measurement of subtle changes in articular cartilage
from MRI of teh knee by combining 3D image registration
and segmentation. Proc SPIE (Int Soc for Opt Engineering)
4322, 431– 439.
McGibbon CA, Dupuy DE, Palmer WE, Krebs DE (1998) Carti-
lage and subchondral bone thickness distribution with MR
imaging. Acad Radiol 5, 20 –25.
McGibbon CA, Trahan CA (2003) Measurement accuracy of
focal cartilage defects from MRI and correlation of MRI
graded lesions with histology: a preliminary study. Osteo-
arthritis Cartilage 11, 483– 493.
Meachim G (1971) Effect of age on the thickness of adult articular
cartilage at he shoulder joint. Ann Rheum Dis 30, 43 –46.
Meachim G, Bentley G, Baker R (1977) Effect of age on thickness
of adult patellar articular cartilage. Ann Rheum Dis 36, 563–568.
Mosher TJ, Dardzinski BJ (2004) Cartilage MRI T2 relaxation
time mapping: overview and applications. Semin Muscu-
loskelet Radiol 8, 355–368.
Mosher TJ, Liu Y, Yang QX, et al. (2004) Age dependency of
cartilage magnetic resonance imaging T2 relaxation times
in asymptomatic women. Arthritis Rheum 50, 2820 –2828.
© 2006 The Authors

Mosher TJ, Smith HE, Collins C, et al. (2005) Change in knee
cartilage T2 at MR imaging after running: a feasibility study.
Radiology 234, 245–249.
Mow VC, Holmes MH, Lai WM (1984) Fluid transport and
mechanical properties of articular cartilage: a review. J Bio-
mech 17, 377–394.
Mow VC, Ateshian GA, Spilker RL (1993) Biomechanics of
diarthrodial joints: a review of twenty years of progress.
J Biomech Eng 115, 460 – 467.
Mow VC, Gu WY, Chen FH (2003) Structure and function of
articular cartilage and meniscus. In Basic Orthopaedic
Biomechanics and Mechanobiology, 3rd edn (eds Mow VC,
Huskes R), pp. 181–258. Philadelphia: Lippincott, Williams &
Wilkins.
Newton PM, Mow VC, Gardner TR, Buckwalter JA, Albright JP
(1997) Winner of the 1996 Cabaud Award. The effect of life-
long exercise on canine articular cartilage. Am J Sports Med
25, 282–287.
Papa E, Cappozzo A (2000) Sit-to-stand motor strategies inves-
tigated in able-bodied young and elderly subjects. J Bio-
mech 33, 1113–1122.
Pauwels F (1980) Biomechanics of the Locomotion Apparatus.
Berlin: Springer.
Peterfy CG, Van Dijke CF, Janzen DL, et al. (1994) Quanti-
fication of articular cartilage in the knee with pulsed
saturation transfer subtraction and fat-suppressed MR
imaging: optimization and validation. Radiology 192, 485 –
491.
Peterfy CG, Van Dijke CF, LUY, et al. (1995) Quantification
of the volume of articular cartilage in the metacarpopha-
langeal joints of the hand: accuracy and precision of
three-dimensional MR imaging. AJR Am J Roentgenol 165,
371–375.
Raynauld JP, Kauffmann C, Beaudoin G, et al. (2003) Reliabil-
ity of a quantification imaging system using magnetic reso-
nance images to measure cartilage thickness and volume in
human normal and osteoarthritic knees. Osteoarthritis Car-
tilage 11, 351–360.
Recht MP, Kramer J, Marcelis S, et al. (1993) Abnormalities of
articular cartilage in the knee: analysis of available MR tech-
niques. Radiology 187, 473– 478.
Roos EM, Dahlberg L (2005) Positive effects of moderate exer-
cise on glycosaminoglycan content in knee cartilage: a four-
month, randomized, controlled trial in patients at risk of
osteoarthritis. Arthritis Rheum 52, 3507–3514.
Roux W (1881) Der Kampf der Teile Im Organismus. Leipzig:
Engelmann.
Sah RL, Kim YJ, Doong JY, et al. (1989) Biosynthetic response
of cartilage explants to dynamic compression. J Orthop Res
7, 619–636.
Saudek DM, Kay J (2003) Advanced glycation endproducts and
osteoarthritis. Curr Rheumatol Rep 5, 33 – 40.
Siedek V, Glaser C, Englmeier KH, Reiser M, Eckstein F (2002)
MRI-based analysis of knee and ankle cartilage in mono-
cygotic twins suggests that its morphology is strongly deter-
mined by genetics. Osteoarthritis Cartilage 10 (Suppl. A),
S56 (PS 110) [abstract].
Sitoci KH, Hudelmaier M, Glaser C, et al. (2003) Nocturnal
changes of cartilage morphology in healthy subjects. Osteo-
arthritis Cartilage 11 (Suppl. A), S95 [abstract].
© 2006 The Authors
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
The effects of exercise on human articular cartilage, F. Eckstein et al. 511
Sittek H, Eckstein F, Gavazzeni A, et al. (1996) Assessment of
normal patellar cartilage volume and thickness using MRI:
an analysis of currently available pulse sequences. Skeletal
Radiol 25, 55–62.
Springer V, Graichen H, Stammberger T, et al. (1998) [Non-
invasive analysis of cartilage volume and cartilage thickness
in the human elbow joint using MRI]. Anat Anz 180, 331–
338.
Stammberger T, Eckstein F, Englmeier KH, Reiser M (1999)
Determination of 3D cartilage thickness data from MR
imaging: computational method and reproducibility in the
living. Magn Reson Med 41, 529 –536.
Stammberger T, Hohe J, Englmeier KH, Reiser M, Eckstein F
(2000) Elastic registration of 3D cartilage surfaces from MR
image data for detecting local changes in cartilage thick-
ness. Magn Reson Med 44, 592– 601.
Takahashi M, Hoshino H, Kushida K, Inoue T (1995) Direct
measurement of crosslinks, pyridinoline, deoxypyridinoline,
and pentosidine, in the hydrolysate of tissues using high-
performance liquid chromatography. Anal Biochem 232,
158–162.
Tiderius CJ, Olsson LE, Leander P, Ekberg O, Dahlberg L (2003)
Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC)
in early knee osteoarthritis. Magn Reson Med 49, 488–
492.
Tieschky M, Faber S, Haubner M, et al. (1997) Repeatability of
patellar cartilage thickness patterns in the living, using a
fat-suppressed magnetic resonance imaging sequence with
short acquisition time and three-dimensional data process-
ing. J Orthop Res 15, 808– 813.
Urban JP (1994) The chondrocyte: a cell under pressure. Br J
Rheumatol 33, 901–908.
Vanwanseele B, Eckstein F, Knecht H, Stussi E, Spaepen A
(2002a) Knee cartilage of spinal cord-injured patients dis-
plays progressive thinning in the absence of normal joint
loading and movement. Arthritis Rheum 46, 2073–2078.
Vanwanseele B, Lucchinietti E, Stussi E (2002b) The effects of
immobilization on the characteristics of articular cartilage:
current concepts and future directions. Osteoarthritis Carti-
lage 10, 408 – 419.
Vanwanseele B, Eckstein F, Knecht H, Spaepen A, Stussi E
(2003) Longitudinal analysis of cartilage atrophy in the
knees of patients with spinal cord injury. Arthritis Rheum
48, 3377–3381.
Vanwanseele B, Eckstein F, Hadwighorst H, et al. (2004) In
vivo precision of quantitative shoulder cartilage measure-
ments, and changes after spinal cord injury. Magn Reson
Med 51, 1026–1030.
Verzijl N, Degroot J, Ben ZC, et al. (2002) Crosslinking by
advanced glycation end products increases the stiffness
of the collagen network in human articular cartilage: a pos-
sible mechanism through which age is a risk factor for osteo-
arthritis. Arthritis Rheum 46, 114–123.
Waldman SD, Spiteri CG, Grynpas MD, Pilliar RM, Kandel RA
(2003) Long-term intermittent shear deformation improves
the quality of cartilaginous tissue formed in vitro. J Orthop
Res 21, 590 –596.
Waterton JC, Solloway S, Foster JE, et al. (2000) Diurnal vari-
ation in the femoral articular cartilage of the knee in young
adult humans. Magn Reson Med 43, 126 –132.
512 The effects of exercise on human articular cartilage, F. Eckstein et al.
Williams A, Gillis A, McKenzie C, et al. (2004) Glycosamino-
glycan distribution in cartilage as determined by delayed
gadolinium-enhanced MRI of cartilage (dGEMRIC):
potential clinical applications. AJR Am J Roentgenol 182,
167–172.
Williams A, Krishnan N, McKenzie C, Burstein D (2005) Mole-
cular imaging of cartilage effects of running a marathon on
dGEMRIC of the knee. Osteoarthritis Cartilage in press.
Wolff J (1892) Das Gesetz der Transformation der Knochen.
Berlin: Hirschwald (reprinted Schattauer, Stuttgart 1991).
Journal compilation © 2006 Anatomical Society of Great Britain and Ireland
Wong M, Carter DR (2003) Articular cartilage functional histo-
morphology and mechanobiology: a research perspective.
Bone 33, 1–13.
Yelin E, Callahan LF (1995) The economic cost and social and
psychological impact of musculoskeletal conditions. National
Arthritis Data Work Groups. Arthritis Rheum 38, 1351–1362.
Zhai G, Stankovich J, Ding C, et al. (2004) The genetic contri-
bution to muscle strength, knee pain, cartilage volume,
bone size, and radiographic osteoarthritis: a sibpair study.
Arthritis Rheum 50, 805 – 810.
© 2006 The Authors