Collagen Damage Location in Articular Cartilage Differs if Damage is Caused by Excessive Loading Magnitude or Rate
Annals of Biomedical Engineering
Collagen Damage Location in Articular Cartilage Differs if Damage is Caused by Excessive Loading Magnitude or Rate
LORENZA HENAO-MURILLO 1 2 3
KEITA ITO 0 2 3
CORRINUS C. VAN DONKELAAR 2 3
0 Department of Orthopaedics, University Medical Center Utrecht , Utrecht , The Netherlands
1 Department of Electronics and Industrial Automation, Universidad Auto ́ noma de Manizales , Manizales , Colombia
2 Department of Biomedical Engineering, Eindhoven University of Technology , Gemini-Zuid 4.101, P.O. Box 513, 5600 MB Eindhoven , The Netherlands. Electronic mail:
3 Department of Biomedical Engineering, Eindhoven University of Technology , Gemini-Zuid 4.101, P.O. Box 513, 5600 MB Eindhoven , The Netherlands
-Collagen damage in articular cartilage is considered nearly irreversible and may be an early indication of cartilage degeneration. Surface fibrillation and internal collagen damage may both develop after overloading. This study hypothesizes that damage develops at these different locations, because the distribution of excessive strains varies with loading rate as a consequence of time-dependent cartilage properties. The objective is to explore whether collagen damage could preferentially occur superficially or internally, depending on the magnitude and rate of overloading. Bovine osteochondral plugs were compressed with a 2 mm diameter indenter to 15, 25, 35 and 45 N, and at 5, 60 and 120 mm/min. Surface fibrillation and internal collagen damage were graded by four observers, based on histology and staining of collagen damage. Results show that loading magnitude affects the degree of collagen damage, while loading rate dominates the location of network damage: low rates predominantly damage superficial collagen, while at high rates, internal collagen damage occurs. The proposed explanation for the rate-dependent location is that internal fluid flows govern the time-dependent internal tissue deformation and therewith the location of overstained and damaged areas. This supports the hypothesis that collagen damage development is influenced by the time-dependent material behaviour of cartilage.
Indentation; Mechanical loading; Cartilage dam-
Cartilage damage generally progresses into
osteoarthritis when adverse biomechanical conditions
prevail.20,33 Among the earliest signs of cartilage
degradation are tissue softening20 and surface
roughening or fibrillation.15,20 Both are associated with
disruption of the collagen network. Because of the slow
turnover of collagen in cartilage, collagen damage can
be considered an early key indicator of osteoarthritis.30
Hence, more insight into the appearance of early
collagen damage as a function of mechanical overloading
may eventually help to predict the course of the
pathology at an earlier stage and improve selection and
timing of patient-specific interventions, which is
Several studies explored the relationship between
mechanical overloading and damage to cartilage and
its collagen network. Excessive loading magnitude,
applied as a single impact or cyclically, results in a
series of changes in cartilage, including cell death,4,12,16
softening,8,29,33 loss of network interconnectivity,28
fissures3 and collagen damage.2,3,9,29,34 Interestingly, it
was also shown that collagen damage may start
internally, without visible signs of tissue damage at the
surface.9,29,34 At higher loading rates, fissure
formation,5 proteoglycan depletion,5 cell death at different
locations5,16 and micro-cracks occur.27 Superficial
cracks in the tissue, which also involve collagen
damage, have been shown to occur following excessive
loading magnitude2,3 and rate.5,27 Furthermore,
extended duration of loading enhances cell death,2,12,13
proteoglycan loss,3,12 and collagen damage.2,3,12
Thus, excessive loading magnitude, rate and
duration induce a variety of damaging effects in cartilage
and result in different appearances of collagen damage.
However, the cause of these different appearances is
yet unknown. It has been hypothesized that collagen
damage results from overstraining of the collagen
fib0090-6964/18/0400-0605/0 2018 The Author(s). This article is an open access publication
rils. Because cartilage material behaviour is
significantly time-dependent due to its biphasic nature, the
internal strain distribution will change as a function of
loading rate. At fast loading rates, fluid is not given
time to flow and tissue deformation is almost
isovolumetric. Thus, axial compression results in significant
sideways tensile straining of the cartilage and the
matrix with its internal collagen network will be strained
accordingly. However, under sustained loading or
slower loading rates, fluid redistributes and this results
in larger compressive straining of the matrix (collagen
buckling), but also smaller tensile strains. Accordingly,
this study postulates that with the difference in strain
distribution, also the locations where tensile strain in
the collagen fibers exceeds damage thresholds will be
influenced by the loading rate. Thus, it is expected that
the location of collagen damage as a consequence of
excessive loading at fast rates would differ from these
locations at slower loading rates.
The present study aims to build support for the
hypothesis that time-dependent tissue behavior affects
the appearance of collagen damage, by characterizing
the extent and location of collagen damage in articular
cartilage as a function of loading magnitude and
loading rate. The collagen damage effects will be
distinguished in development of superficial clefts and
internal collagen network damage. It is hypothesized
that similar loading magnitudes applied at higher rates
will result in more internal collagen damage as a result
of the biphasic, time-dependent mechanical tissue
MATERIALS AND METHODS
72 osteochondral plugs were harvested from 16
metacarpal proximal epiphyses of 1-year old calves,
reaching the skeletal maturity,1,21 obtained from a
local slaughterhouse. Plugs were extracted using a
diamond core-drill (Einhell SB 501/1, Einhell,
Germany) of 7.5 mm inner diameter and then plugs
thickness was reduced to 5–7 mm length with a
diamond cut-off wheel (Accutom-5, Struers, Denmark).
During drilling and cutting, osteochondral plugs were
irrigated with room temperature phosphate buffered
saline (PBS) and stored in PBS at – 20 C until testing.
The plugs were thawed and equilibrated at room
temperature for 1 h in PBS. After thawing, cartilage
thickness was determined using a digital caliper, by
averaging the thickness at four locations along the
edge of the cartilage surface. The osseous part of the
osteochondral plugs were placed in a custom made
polycarbonate container with a 7.5 mm hole in the
center of the bottom plate to fix the sample. The
container was filled with PBS and covered with a lid to
avoid evaporation (Fig. 1a). Samples were slowly
press-fit to guarantee the bottom of the samples would
be in contact with the container’s hole bottom, by
applying a non-damaging compression of 25 N
(0.57 MPa at 1 mm/min), distributed over the entire
To evaluate the local effects of different loading
regimes, an impermeable indenter of 2 mm diameter
with a hemi-spherical tip (1 mm radius) was used to
apply five loading cycles of 15, 25, 35 or 45 N loading
at 5, 60 or 120 mm/min, separated by 2.4 s at rest load
of 0.05 N (Universal testing machine
BT1FB010TND30, Zwick/Roell, Germany) (Fig. 1). The
applied loading magnitudes of 15, 25, 35 and 45 N
with the 2 mm hemi-spherical indenter correspond to
stresses of approximately 4.8, 8.0, 11.1 and 14.3 MPa
respectively. These values are at the high end of
physiological loading19,35 and have been used in other
studies for unconfined compression35 and
Immediately after mechanical loading, the cartilage
was removed from the subchondral bone with a scalpel
and cut into halves. One half was stored in PBS at –
20 C, the other was embedded in Tissue-Tek
compound (Sakura Finetek, USA, Inc.), rapidly frozen in
liquid nitrogen and stored at – 30 C until histological
Cartilage samples were cryo-sectioned in 7 lm slices
(Microm HM 550, Germany), mounted on pre-coated
glass slides (SuperFrost Plus, Thermo Scientific,
Germany), dried for 1 h and stored at – 30 C.
For staining, samples were dried for 30 min at room
temperature and 1 h at 37 C, fixed with 3.7% 0.1 M
phosphate buffered (pH 7.4) formaldehyde for 5 min,
rinsed in PBS and dipped in 0.01% tween PBS. To
enhance the permeability of the extracellular matrix,
glycosaminoglycans were removed by incubating with
1% hyaluronidase in PBS (Testicular, Type I-s, EC
188.8.131.52, Sigma–Aldrich, US) for 30 min at 37 C. To
block endogenous peroxidase activity, sections were
incubated with freshly prepared 1% (v/v) peroxide in
absolute ethanol at room temperature for 30 min.
Sections were then incubated with 10% normal horse
serum (NHS) for 30 min to block nonspecific staining.
To visualize collagen damage, slices were incubated
overnight at 4 C with col2-3/4m antibody (1/800)
(Mouse monoclonal IgG1, Product Nr. 50–1011, IBEX
Pharmaceuticals Inc., Canada) in 1% bovine serum
albumin (BSA) and rinsed in PBS. Control samples
were incubated with 1% BSA. Subsequently, slices
were incubated for 1 h at room temperature with
biotin-labeled horse anti-mouse antibody (1/400) (IgG
(H + L), produced in horse, Vector Laboratories,
Inc., USA) and rinsed in PBS. Then, they were
incubated with biotin streptavidin detection system
(VectaStain Elite ABC, Vector Laboratories, Inc, USA)
reagent for 30 min and rinsed 5 min in PBS. Peroxide
was detected by incubating with 3¢,3¢ diaminobenzidine
for 7 min and rinsing thoroughly in PBS. Finally,
sections were counterstained with Mayer’s
hematoxylin, dehydrated and mounted with Entellan
(Merck, Germany). Stained sections were digitized
under inverted light microscopy at 59 magnification
(Axio Observer, Carl Zeiss, Germany).
Processing Data and Statistics
Indentation strain was calculated as the ratio
between the original cartilage thickness and the
displacement of the indenter from the sample surface to
the applied loading of either 0.05 N prior to each
peakloading cycle (Fig. 1c, blue squares) or the
peakloading (Fig. 1c, red squares).
The degree of cartilage damage was evaluated using
a custom made histological grading system (Table 1).
The amount of damage was classified in two
categories: macroscopic superficial damage and
microscopic internal collagen damage (visualized by col2-3/
4m staining). Each category was grouped into five
levels ranging from undamaged (score 0) to severely
damaged tissue (score 5), similar to the Mankin
score.14 Four observers received digitized images of the
stained sections and an explanation about the scoring
system (Table 1). They independently graded
macroscopic superficial and microscopic internal collagen
The summation of macroscopic superficial and
microscopic internal collagen damage was defined as
the total collagen damage in cartilage, and was
categorized as normal, mild, moderate or severe (Table 2).
The grading reliability inter- and intra- observers was
assessed with the Kappa coefficient.26 Total,
macroscopic superficial and microscopic internal collagen
damage scores from the 4 observers were averaged per
A three-way mixed ANOVA was run to test the
effects of loading magnitude and rate on strains within
the five loading cycles. Then, a two-way ANOVA was
performed to understand the effects of loading
magnitude and loading rate on the strains at the 5th
loading cycle, and also on total, macroscopic
superficial and microscopic internal collagen damage.
Outliers were assessed by inspection of a boxplot,
normality was assessed using Shapiro–Wilk’s
normality test for each cell of the design and homogeneity of
variances was assessed by Levene’s test. A square root
transformation was performed on strains and on total,
macroscopic superficial and microscopic internal
collagen damage to accomplish equality of variances.
When any of the ANOVA tests showed a significant
interaction, an analysis of simple main effects for the
corresponding levels was performed with Bonferroni
correction. All pairwise comparisons were run for each
simple main effect with reported 95% confidence
intervals and p-values Bonferroni-adjusted within each
simple main effect. When there was not a significant
interaction, an analysis of the main effect for the
corresponding levels was performed. All pairwise
comparisons were run where reported 95% confidence
intervals and p-values were Bonferroni-adjusted. Mean
differences were considered significant at the p = 0.05
level. A Spearman’s rank correlation was run to assess
the relationship between articular cartilage thickness
and total, macroscopic superficial and microscopic
internal collagen damage. Statistical tests were
produced with IBM SPSS 23 software, US.
Different collagen damage patterns could be
distinguished histologically, depending on the applied
loading protocol (Fig. 2). Macroscopic superficial
damage occurred in some samples in most groups and
was more prominent at higher loading magnitudes
(examples indicated by red arrows in Figs. 2d, 2g, and
2k). Microscopic internal damage below the cartilage
surface (indicated by black arrows in Fig. 2) occurred
predominantly in the groups that received higher
loading rates (Figs. 2f, 2j, and 2k) and penetrated to
the surface in 45% of the samples (e.g. Figs. 2g, 2h,
and 2l). In 13% of the cases, the cartilage surface
remained intact, but microscopic internal damage was
visible (e.g. Figs. 2f, and 2l).
Average indentation strains increased with each
loading cycle, both during the 0.05 N (Fig. 3, blue
bars) and the peak-loading period (Fig. 3, red bars).
Each new cycle of 0.05 N was statistically significant
different compared to the previous cycle of 0.05 N
(p < 0.005; Fig. 3). This statistically significant
difference was also observed within each peak-loading
(p < 0.005; Fig. 3). The average strains increased when
samples were subjected to lower loading rates and
higher loading magnitudes. There was no interaction
between loading magnitude and rate on strains at 5th
loading cycle (p > 0.05). There was a statistically
significant main effect of loading magnitude at 35 and
45 N, compared to 15 and 25 N (p < 0.005; Fig. 4a)
and the main effect of loading rate was also statistically
significant, comparing 5 mm/min to both 60 and
120 mm/min (p < 0.005; Fig. 4a). The increasing strain
at 0.05 N loading represents the persistence of an
indent in the cartilage, which does not restore to its
original height within the 2.4 s of 0.05 N loading.
Noteworthy, this indent was still visible in the tissue
while the tissue was collected after the loading regime.
The intra- and inter-observer reliability for scoring
histological images yielded a weighted (quadratic)
Kappa coefficient of 0.9 and 0.69, respectively. This
indicated almost perfect intra-observer and good
interobserver concordance strength.26
Few outliers were found by inspecting the boxplots
of total, macroscopic superficial and microscopic
internal collagen damage; these were kept for being
genuine values. According to Shapiro–Wilk’s
normality test, most of the damage grades were normally
distributed (p > 0.05) and Levene’s test showed
homogeneity of variances for total (p = 0.20),
macroscopic superficial (p = 0.37) and microscopic internal
(p = 0.59) collagen damage. Two-way ANOVA
showed significant interaction between loading
magnitude and loading rate for total collagen damage
(p = 0.013). There was a statistically significant
difference in mean of total damage score between all
loading magnitudes (15, 25, 35 and 45 N) at 5 mm/min
(p < 0.005), 60 mm/min (p < 0.005) and 120 mm/min
(p = 0.016), and between all loading rates (5, 60 and
120 mm/min) at 25 N (p = 0.005) and 35 N
(p < 0.005). Pairwise comparisons for each simple
main effect revealed that within each loading rate,
loading magnitude had a significant effect on total
damage (Fig. 5). In contrast, the effects of loading rate
on total damage were only significant in the groups
that received intermediate loading (25 and 35 N).
Total damage was unaffected by loading rate when
loading was 15 N, where all groups showed minor
damage, and 45 N, where all groups showed significant
damage (Fig. 5).
Two-way ANOVA analysis per type of damage
showed significant interaction between loading
magnitude and loading rate for macroscopic superficial
damage (p < 0.005; Fig. 6a), while this interaction did
not occur for microscopic internal damage (p = 0.23;
Fig. 6b). Mean differences of macroscopic superficial
damage score revealed a statistically significant
difference between all loading magnitudes (15, 25, 35 and
45 N) at 5 mm/min (p < 0.005) and 60 mm/min
(p < 0.005); and between all loading rates (5, 60 and
120 mm/min) at 45 N (p < 0.005). Pairwise
comparisons for each simple main effect showed that most
significant differences in macroscopic superficial
damage occurred between 5 and 60 mm/min when a
loading of 45 N was applied (Fig. 6a). Microscopic internal
damage score indicated a statistically significant main
effect of both loading magnitude (p < 0.005) and
loading rate (p < 0.005). Pairwise comparisons
evidenced that microscopic internal damage was
significant different between 60 and 120 mm/min when
loadings of 35 and 45 N were applied (Fig. 6b).
The average cartilage thickness was
1.05 ± 0.23 mm. Spearman’s rank correlation showed
a low positive correlation between cartilage thickness
and total damage (q = 0.340, p = 0.005), and a
moderate correlation between cartilage thickness and
microscopic internal damage (q = 0.436, p < 0.005).
The present hypothesis was that time-dependent
tissue behavior affects the appearance of collagen
damage. The rationale behind this hypothesis was that
collagen damage would occur through excessive
strains, while the strain distribution in a
time-dependent material such as articular cartilage would change
with variations in loading magnitude and rate. The
approach to test this hypothesis was to assess the
location and amount of damage that develops
macroscopically in the superficial cartilage layer and
microscopically in the internal collagen network, as a
function of various combinations of loading
magnitudes and loading rates.
In agreement with the hypothesis, results showed
that loading magnitude and loading rate are both
linked to the degree of total collagen damage in
cartilage. Their statistically significant interaction (Fig. 5)
suggests variations in appearance depending on the
specific combination of loading magnitude and loading
rate. Microscopic internal collagen damage increases
with both loading magnitude and rate, with no
statistical interaction, with suggestions of a loading
threshold to induce damage at the highest loading rate and at
lower loading magnitudes (Fig. 6b). A different trend
is present in the macroscopic superficial damage, which
increases with loading magnitude and rate for
intermediate loading magnitudes, but with significant
interaction; the loading-rate dependency is inverted for
the highest loading magnitude (Fig. 6a). Taken
together, these data show that different modes of
overloading cause distinct appearances of collagen damage
in articular cartilage.
Previous studies have demonstrated thresholds of
loading magnitude and either loading rate or loading
duration at which macroscopic
superficial3,5,10,16–18,24,32 and microscopic internal collagen
damage2,29 develops in articular cartilage. These
studies used various experimental setups, including
unconfined compression,5,17,18,24 drop towers10,32 or
channel indenters.27,28 The present study used a round
indenter to localize the loading. This has the advantage
that results are independent of sample size and of the
adverse conditions at the cut edges of the sample.
Wilson et al.34 using a similar indenter, found
subsuperficial microscopic collagen fiber damage after
25 N indentation, sometimes penetrating to the
surface, in agreement with the present results (Figs. 2f–2h,
2j). The same was observed by Chen et al.2,3 under
compression with a flat-ended indenter at 5 MPa for
120 min and under confined compression at 1 MPa for
24 h and 5 MPa for 1 h. Unlike other studies, Chen
et al.2,3 reported collagen damage to occur only at the
surface. However, they used isolated cartilage rather
than osteochondral plugs, and attachment to bone is
known to affect the mechanical conditions in the
cartilage and its collagen network significantly.10 Thibault
et al.29 found that stress rates between 2 and 5 MPs/s
produce collagen damage, which is mainly
concentrated in the deep zone, sometimes extending into the
Macroscopic superficial damage in the present study
was mostly limited up to level 3 severity (superficial
damage). Only few cases, all at higher loading rates
and magnitudes, presented slightly deeper clefts. Full
depth fissures may occur as the clefts propagate over
longer time, or as a result of more extreme loading
magnitudes and rates than those used in the present
study.10,32 The present results concur with other
studies which showed that the formation of fissures and
cracks increased with loading magnitude28 and loading
rate,5,17,18,24,27 with fissures starting from the
superficial zone5,17,18,24,27,28,32 and propagating into the
transitional and deep zone.10,27
Thus, the general responses at both the microscopic
and macroscopic level in the present study were in
agreement with previous work. The additional insight
from the present study is the differential effect of
loading magnitude and rate on the macroscopic
superficial and microscopic internal collagen damage.
One possible explanation for this differential effect is
that the failure mode of collagen type II may be
loading-rate dependent. Strain-rate dependent damage
has been reported for fiber-reinforced polymers in
general25 as well as for collagen type I,7 showing that
collagen fiber damage started at 10% strain under high
strain rates and at 7% strain during quasi-static
loading. The present study imposed different strain
rates to the tissue surface, but did neither monitor
strains nor strain-rates in the tissue in the direction of
the local fibers. Therefore, it is not possible to conclude
to what extend strain-rate dependent damage of
collagen type II could explain the observed effects in this
The second, more likely explanation for the
observed strain-rate dependent collagen damage in
cartilage is that the strains experienced by the collagen
at the surface and internally in the tissue are influenced
in a different way by the biphasic response of the
cartilage (illustrated in Fig. 7). During slow
indentation, fluid is given time to flow away from the area
under the indenter (arrows in Fig. 7), whereas fluid
remains in place during fast loading rates. As a
consequence of the fluid loss, the average indentation
depth increased more for lower loading rates than for
higher loading rates. This is evidenced by the
indentation depth at the 0.05 N loading period (Figs. 3, 4, 7,
bottom). The indent in the cartilage is also visible by
the naked eye after the experiment. The deformation of
the cartilage surface and its parallel collagen fibers
directly follows the imposed deformation by the
indenter. Thus, lower indentation rates result in deeper
indentations, larger surface stretching and
consequently more straining and potentially more damage of
the tangential superficial collagen fibers.22 This
explains the inverse relationship between loading rate
and macroscopic superficial collagen damage (Figs. 6a
and 7, right column). The same biphasic mechanism
results in an opposite time-dependent strain response
deeper inside the tissue. At faster loading rates, i.e., at
shorter loading durations, fluid is not given time to
flow relative to the matrix, and the tissue bulges
sideways in the area under the indenter (asterisk in Fig. 7).
This causes internal strains in the matrix and in the
internal collagen network (Fig. 7, left column). The
intermediate zone with its dispersed collagen fiber
network is the most susceptible to damage. Thus, at
intermediate loading rates and magnitudes, this zone is
selectively damaged (Figs. 2f, and 2g), in agreement
with former studies. With increasing loading rate
(Figs. 2j, and 2k), magnitude (Fig. 2h) or both
(Fig. 2l), the damage extends into the deeper zones.
Here the staining appears more intense, presumably
because the strong anisotropic nature of the tissue in
the deeper zone causes a larger number of fibers to
become damaged at the same time, when a critical
strain threshold in the tissue is crossed.
The above effects are explained by the time allowed
for the fluid to move through the tissue. This not only
depends on loading rate, but also on local tissue
permeability and proteoglycan density, as proteoglycans
attract fluid and prevent it from moving. As the effects
of fluid flow are dependent on these variables, the
depth-dependent cartilage composition as well as
variability in cartilage contents between species, joints
or locations within joints will influence the quantitative
relation between loading rate, matrix straining and
collagen damage. Theoretically, the loading rate
threshold above which damage may occur in cartilage,
will be lower for tissue with lower permeability and
higher proteoglycan density. For low or high loading
rates, loading rate will overrule the effect of
permeability and damage development will occur as
presented in Fig. 7.
Recently, strain-rate dependent collagen damage in
cartilage was predicted by a computational model.23 In
a similar range of indentation magnitudes and rates, it
was found that collagen damage at the surface would
increase when either loading magnitude or loading rate
was increased. Also in agreement with the present
study is that an interaction between loading rate and
magnitude seemed to exist for the internal collagen
damage. Only at intermediate loading magnitudes and
for intermediate rates, significant collagen damage
occurred internally before it became apparent at the
surface. However, in the computational predictions it
was found that this internal collagen damage would
reduce upon further increase in loading rate (Fig. 8 in
P a´rraga Quiroga et al.),23 whereas it increases in the
present experimental data. Thus, the present data
provide insight that should be used to improve the
damage predictions by the computational model.
The present study used indentation with a 2 mm
diameter round-ended indenter. This loading condition
was chosen to be in line with former studies,8,23,34
where indentation was chosen as a mechanism to
invoke significant collagen strains both at the surface
and inside the cartilage. The advantage of indentation
is that the result is independent of sample size and of
edge-effects due to sample processing. The
physiological loading condition where a rounded condyle
compresses a rather flat tibia plateau is in between loading
with an indenter and unconfined compression. The
indenter radius is important for the amount of stress
and strain induced in the area below the indenter.
Because of the indenter’s round shape, it is difficult to
calculate the exact stress level applied to the sample.
Presumably, applied stresses reach 5–14 MPa, which is
at the high end of physiological loading and
comparable to stresses applied by Quinn et al. in unconfined
Finally, tissue response varies between samples.
Some samples which are subjected to high loading
magnitudes and fast loading rates do not show signs of
macroscopic superficial or microscopic internal
collagen damage, whereas some samples in milder loading
groups do. This may be explained by biological
variation between metacarpal proximal epiphysis joints of
1 year-old calves. In particular, differences in cartilage
thickness have been proposed to affect cartilage
damage development.34 In the present study a moderate
correlation between cartilage thickness and
microscopic internal damage was found (q = 0.436,
p = 0.0003), suggesting that thickness has an effect but
is not the predominant parameter to explain
differences between samples. Variations due to the
experimental analyses cannot be excluded. For instance, the
selected histological slide for evaluation of damage
may not be the most central one under the indenter
and slices may have different orientations with respect
to the split line direction. Such suboptimal histology
would result in a slight underestimation of the actual
damage. Thus, the presented results can be considered
To conclude, this study confirms that loading
magnitude and loading rate both affect the degree of
collagen damage in articular cartilage. It demonstrates
that macroscopic superficial damage and microscopic
internal collagen damage respond differently to
variations in loading magnitude and rate. Macroscopic
superficial damage is highest when the indentation
depth is largest, i.e., under high loading magnitudes
and slow loading rates. Microscopic internal collagen
damage occurs when the internal deformation in the
cartilage is largest. This internal deformation is
governed by the time water is allowed to flow through the
matrix. At faster deformations, less water flow can
occur relative to the matrix compared to slower
deformations. Consequently, faster compression
results in more internal straining of the hydrated tissue,
and the collagen network is more likely to become
overstrained. Damage at the surface is less affected by
water displacement, and more dependent on the actual
local deformation of the tissue by the indenter. Such
differential effects of the loading regime on
macroscopic superficial and microscopic internal damage
have not been demonstrated before. Further
understanding of the mechanically complex, time-dependent
mechanisms that result in cartilage damage are
important for understanding the etiology and
progression of osteoarthritis.
This study was supported by the grant program
‘‘Programa de Formacio´ n Doctoral Francisco Jose´ de
Caldas Generacio´ n del Bicentenario’’ awarded by the
Francisco Jos e´ de Caldas Institute for the
Development of Science and Technology (COLCIENCIAS).
LASPAU ID 20110290.
This article is distributed under the terms of the
Creative Commons Attribution 4.0 International
which permits unrestricted use, distribution, and
reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons license, and
indicate if changes were made.
1Ahsan, T. , F. Harwood , K. B. McGowan , D. Amiel , and R. L. Sah . Kinetics of collagen crosslinking in adult bovine articular cartilage . Osteoarthr. Cartil . 13 : 709 - 715 , 2005 .
2Chen, C. T. , M. Bhargava , P. M. Lin , and P. A. Torzilli .
Time , stress, and location dependent chondrocyte death and collagen damage in cyclically loaded articular cartilage . J. Orthop. Res . 21 : 888 - 898 , 2003 .
3Chen, C. T. , N. Burton-Wurster , G. Lust , R. A. Bank , and J. M. Tekoppele . Compositional and metabolic changes in damaged cartilage are peak-stress, stress-rate, and loadingduration dependent . J. Orthop. Res . 17 : 870 - 879 , 1999 .
4de Vries, S. A. H. , M. C. van Turnhout , C. W. J. Oomens , A. Erdemir , K. Ito , and C. C. van Donkelaar. Deformation thresholds for chondrocyte death and the protective effect of the pericellular matrix . Tissue Eng. A 20 ( 13 , 14): 1870 - 1876 , 2014 .
5Ewers , B. J. , D. Dvoracek-Driksna , M. W. Orth , and R. C. Haut . The extent of matrix damage and chondrocyte death in mechanically traumatized articular cartilage explants depends on rate of loading . J. Orthop. Res . 19 : 779 - 784 , 2001 .
6Gardiner , B. S. , F. G. Woodhouse , T. F. Besier , A. J. Grodzinsky , D. G. Lloyd , L. Zhang , and D. W. Smith. Predicting knee osteoarthritis . Ann. Biomed. Eng. 44 ( 1 ): 222 - 233 , 2016 .
7Haut, R. C. Age-dependent influence of strain rate on the tensile failure of rat-tail tendon . J. Biomech. Eng . 105 ( 3 ): 296 - 299 , 1983 .
8Hosseini, S. M. , M. B. Veldink , K. Ito , and C. C. van Donkelaar. Is collagen fiber damage the cause of early softening in articular cartilage? Osteoarthr . Cartil. 21 : 136 - 143 , 2013 .
9Hosseini, S. M. , W. Wilson , K. Ito , and C. C. van Donkelaar . A numerical model to study mechanically induced initiation and progression of damage in articular cartilage . Osteoarthr. Cartil . 22 ( 1 ): 95 - 103 , 2014 .
10Jeffrey, J. E. , D. W. Gregory , and R. M. Aspden . Matrix damage and chondrocyte viability following a single impact load on articular cartilage . Arch. Biochem. Biophys . 322 ( 1 ): 87 - 96 , 1995 .
11Kraus , V. B. , F. J. Blanco , M. Englund , M. A. Karsdal , and L. S. Lohmander . Call for standardize definitions of osteoarthritis and risk stratification for clinical trials and clinical use . Osteoarthr. Cartil . 23 : 1233 - 1241 , 2015 .
12Lin, P. M. , C. T. C. Chen , and P. A. Torzilli . Increased stromelysin-1 (MMP-3), proteoglycan degradation (3B3- and 7D4) and collagen damage in cyclically load-injured articular cartilage . Osteoarthr. Cartil . 12 : 485 - 496 , 2004 .
13Lucchinetti , E., C. S. Adams , W. E. Horton , and P. A. Torzilli . Cartilage viability after repetitive loading: a preliminary report . Osteoarthr. Cartil . 10 : 71 - 81 , 2002 .
14Mankin, H. J. , H. Dorfman , L. Lippiello , and A. Zarins . Biochemical and metabolic abnormalities in articular cartilage from osteo-arthritic human hips. II. Correlation of morphology with biochemical and metabolic data . J. Bone Joint Surg. Am . 53 : 523 - 537 , 1971 .
15Mansour, J. M. Biomechanics of cartilage . In: Kinesiology. The Mechanics and Pathomechanics of Human Movement, edited by C. A . Oatis . Philadelphia: Lippincott Williams and Wilkins Lippincott , a Wolters Kluwer business, 2009 , pp. 69 - 83 .
16Milentijevic , D. , and P. A. Torzilli . Influence of stress rate on water loss, matrix deformation and chondrocyte viability in impacted articular cartilage . J. Biomech . 38 : 493 - 502 , 2005 .
17Morel , V. , C. Berutto , and T. M. Quinn . Effects of damage in the articular surface on the cartilage response to injurious compression in vitro . J. Biomech . 39 : 924 - 930 , 2006 .
18Morel , V. , and T. M. Quinn . Cartilage injury by ramp compression near the gel diffusion rate . J. Orthop. Res . 22 : 145 - 151 , 2004 .
19Morrel, K. C. , W. A. Hodge , D. E. Krebs , and R. W. Mann . Corroboration of in vivo cartilage pressures with implications for synovial joint tribology and osteoarthritis causation . Proc. Natl. Acad. Sci. USA 102 ( 41 ): 14819 - 14824 , 2005 .
20Mow , V. C. , W. Y. Gu , and F. H. Chen . Structure and function of articular cartilage and meniscus . In: Basic Orthopaedic Biomechanics and Mechano-Biology , edited by V. C. Mow , and R. Huiskes . Philadelphia: Lippincott Williams and Wilkins, 2005 , pp. 181 - 258 .
21Otsuki, S. , S. P. Grogan , S. Miyaki , M. Kinoshita , H. Asahara , and M. K. Lotz . Tissue neogenesis and STRO-1 expression in immature and mature articular cartilage . J. Orthop. Res . 28 ( 1 ): 96 - 102 , 2010 .
22P a´rraga Quiroga , J. M. , W. Wilson , K. Ito , and C. C. van Donkelaar. Relative contribution of articular cartilage's constitutive components to load support depending on strain rate . Biomech. Model. Mechanobiol . 16 : 151 - 158 , 2017 .
23P a´rraga Quiroga , J. M. , W. Wilson , K. Ito , and C. C. van Donkelaar. The effect of loading rate on the development of early damage in articular cartilage . Biomech. Model. Mechanobiol . 16 : 263 - 273 , 2017 .
24Quinn, T. M. , R. G. Allen , B. J. Schalet , P. Perumbuli , and E. B. Hunziker . Matrix and cell injury due to sub-impact loading of adult bovine articular cartilage explants: effects of strain rate and peak stress . J. Orthop. Res . 19 : 242 - 249 , 2001 .
25Ray , B. C. , and D. Rathore . A review on mechanical behavior of FRP composites at different loading speeds . Crit. Rev. Solid State Mater. Sci . 40 ( 2 ): 119 - 135 , 2015 .
26Sim, J. , and C. C. Wright . The Kappa statistic in reliability studies: use, interpretation, and sample size requirements . Phys. Ther . 85 : 257 - 268 , 2005 .
27Thambyah , A. , G. Zhang , W. Kim, and N. D. Broom . Impact induced failure of cartilage-on-bone following creep loading: a microstructural and fracture mechanics study . J. Mech. Behav. Biomed. Mater . 14 : 239 - 247 , 2012 .
28Thambyah , A. , J. Y. Zhao , S. L. Bevill , and N. D. Broom . Macro-, micro - and ultrastructural investigation of how degeneration influences the response of cartilage to loading . J. Mech. Behav. Biomed. Mater . 5 ( 1 ): 206 - 215 , 2012 .
29Thibault, M. , A. R. Poole , and M. D. Buschmann . Cyclic compression of cartilage/bone explants in vitro leads to physical weakening, mechanical breakdown of collagen and release of matrix fragments . J. Orthop. Res . 20 : 1265 - 1273 , 2002 .
30Tiku, M. L. , and B. Madhan . Preserving the longevity of long-lived type II collagen and its implication for cartilage therapeutics . Ageing Res. Rev . 28 : 62 - 71 , 2016 .
31Torzilli, P. A. , R. Grigiene , J. Borrelli , Jr, and D. L. Helfet . Effect of impact load on articular cartilage: cell metabolism and viability, and matrix water content . J. Biomech. Eng . 121 : 433 - 441 , 1999 .
32Verteramo , A. , and B. Seedhom . Effect of a single impact loading on the structure and mechanical properties of articular cartilage . J. Biomech . 40 : 3580 - 3589 , 2007 .
33Willie , B. M. , T. Pap , C. Perka , C. O. Schmidt , F. Eckstein , A. Arampatzis , H.-C. Hege, H. Madry , A. Vortkamp , and G. N. Duda . Overload of joints and its role in osteoarthritis. Towards understanding and preventing progression of primary osteoarthritis . Z. Rheumatol . 74 : 618 - 621 , 2015 .
34Wilson, W. , C. van Burken , C. C. van Donkelaar , P. Buma , B. van Rietbergen , and R. Huiskes . Causes of mechanically induced collagen damage in articular cartilage . J. Orthop. Res . 24 ( 2 ): 220 - 228 , 2006 .
35Zimmerman, N. B. , D. G. Smith , L. A. Pottenger , and D. R. Cooperman . Mechanical disruption of human patellar cartilage by repetitive loading in vitro . Clin. Orthop. Relat. Res . 229 : 302 - 307 , 1987 .