Eur J Appl Physiol (2012) 112:2679–2691
DOI 10.1007/s00421-011-2248-x
ORIGINAL ARTICLE
Time course of changes in the human Achilles tendon properties
and metabolism during training and detraining in vivo
Keitaro Kubo • Toshihiro Ikebukuro •
Akira Maki • Hideaki Yata • Naoya Tsunoda
Received: 1 September 2011 / Accepted: 11 November 2011 / Published online: 22 November 2011
Ó Springer-Verlag 2011
Abstract The purpose of this study was to investigate the
time course of changes in human tendon properties and
metabolism during resistance training and detraining. Nine
men (21–27 years) completed 3 months of isometric
plantar flexion training and another 3 months of detraining.
At the beginning and on every 1 month of training and
detraining periods, the stiffness, blood circulation (blood
volume and oxygen saturation), serum procollagen type 1
C-peptide (P1P; reflects synthesis of type 1 collagen),
echointensity (reflects collagen content), and MRI signal
intensity (reflects collagen structure) of the Achilles tendon
were measured. Tendon stiffness did not change until
2 months of training, and the increase (50.3%) reached
statistical significance at the end of the training period.
After 1 month of detraining, tendon stiffness had already
decreased to pre-training level. Blood circulation in the
tendon did not change during the experimental period. P1P
increased significantly after 2 months of training. Echoin-
tensity increased significantly by 9.1% after 2 months of
training, and remained high throughout the experiment.
Communicated by Toshio Moritani.
K. Kubo (&)
Department of Life Science (Sports Sciences),
University of Tokyo, Komaba 3-8-1, Meguro-ku,
Tokyo 153-8902, Japan
e-mail: kubo@idaten.c.u-tokyo.ac.jp
T. Ikebukuro  A. Maki  N. Tsunoda
Department of Physical Education,
Kokushikan University, Tokyo, Japan
H. Yata
Sports Science Laboratory, Wako University,
Machida, Tokyo, Japan
MRI signal intensity increased by 24.2% after 2 months
and by 21.4% after 3 months of training, but decreased to
the pre-training level during the detraining period. These
results suggested that the collagen synthesis, content, and
structure of human tendons changed at the 2-month point
of training period. During detraining, the sudden decrease
in tendon stiffness might be related to changes in the
structure of collagen fibers within the tendon.
Keywords Plantar flexor  Tendon stiffness 
Collagen synthesis  Structure
Introduction
Recent studies using ultrasonography have demonstrated
that training-induced changes in the properties of human
tendons differ among the exercise protocols used (Aram-
patzis et al. 2007; Burgess et al. 2007; Kubo et al. 2007,
2009b). Furthermore, we reported that tendons adapted to
resistance training slower, but to detraining faster, than
muscle function (Kubo et al. 2010). However, the mecha-
nisms responsible for the changes in the mechanical
properties of human tendons are unknown. From a meta-
bolic point of view, previous studies using microdialysis
and biopsy techniques have demonstrated that blood flow
and type I collagen synthesis in human tendons change
during physical exercise (e.g., Boushel et al. 2000; Miller
et al. 2005). However, these invasive techniques seem to be
difficult for applying to investigate the changes in blood
circulation and collagen synthesis of human tendon in the
field of sport sciences. In particular, these invasive meth-
odologies would be inappropriate for performing the
repeated measures to investigate the time course of changes
in the variables during the training period.
123
2680
Recently, we measured the blood circulation (blood
volume and oxygen saturation) in human tendons using
three red laser lights (Kubo et al. 2008a, b, 2009a, b). With
this technique, we found that the blood volume of the
patella tendon increased significantly after 3 months of
dynamic training (tendon stiffness was unchanged),
although it did not change after static training (tendon
stiffness increased considerably) (Kubo et al. 2009b). In
addition, Langberg et al. (1999) demonstrated that con-
centrations of lactate increased after intermittent static
exercise using microdialysis technique. Other studies
showed that the application of lactate to tendon fibroblasts
increased collagen production (Klein et al. 2001; Yal-
amanchi et al. 2004). Considering these previous findings,
insufficient blood volume within the tendon would lead to
the accumulation of extra lactate after repeated isometric
contractions, which could be a factor inducing the con-
siderable increase in type 1 collagen synthesis within the
tendon after isometric training.
Many researchers have demonstrated that changes in
some serum biochemical markers represent type 1 collagen
synthesis and degradation of bone (e.g., Maimoun et al.
2006; Virtanen et al. 1993). More recently, we reported that
the amount of procollagen type 1 C-peptide (P1P; reflecting
the synthesis of type 1 collagen in other organs as well as
bone) in serum increased at 24 h after static knee extension
exercises, whereas it did not change after dynamic exercises
(Kubo et al. 2011). On the other hand, bone-specific alka-
line phosphatase (BAP; a specific marker of bone synthesis)
levels did not change after static or dynamic exercise.
Therefore, we considered the increase in P1P after static
exercise to reflect enhanced production of type 1 collagen in
tendons. Furthermore, with this technique, it may be pos-
sible to estimate type 1 collagen synthesis in human tendons
during strength training and detraining periods.
Previous studies have shown that an increase in the ech-
ointensity of muscle tissue indicated increases in fat, fluid,
and fibrosis (van Holsbeeck and Introcaso 1992; Nosaka and
Clarkson 1996; Reimers et al. 1993a, b). Reimers et al.
(1993b) reported a correlation between echointensity and
amount of intramuscular fat. In addition, Nosaka and
Clarkson (1996) demonstrated that increases in the echoin-
tensity of muscle were related to plasma enzyme levels and
edema after exercise. Concerning tendons, however, train-
ing-induced changes in echointensity would imply changes
in the amount of collagen fibers, not fat or fluid. Similarly,
the signal intensity of magnetic resonance imaging (MRI)
has previously indicated the condition of the extracellular
matrix, tissue hydration, collagen fiber structure, and pro-
teoglycan content (Erickson et al. 1993; Shalabi et al. 2002;
Soyama et al. 1995; Yoon and Halper 2005). Furthermore, T2
relaxation time of MRI (equivalent to MRI signal intensity)
has been shown to correspond to collagen orientation (Du
123
Eur J Appl Physiol (2012) 112:2679–2691
et al. 2009; Erickson et al. 1993; Nieminen et al. 2001).
Therefore, these approaches have allowed us to estimate the
training-induced changes in content and structure of collagen
fibers within the human tendon in vivo.
In the present study, we determine the time course of
changes in properties and metabolism of human tendons
during periods of resistance training and detraining. Using
‘‘non-invasive’’ measures, the findings on the collagen
synthesis, content, and structure of human tendon in vivo
would be useful to elucidate the mechanisms of training-
induced changes in their mechanical properties.
Methods
Subjects
Sixteen healthy males volunteered for this investigation.
The subjects were randomly assigned to a training group
(n = 9, age 23.4 ± 0.6 years, height 174.5 ± 2.0 cm,
weight 69.1 ± 3.2 kg, mean ± SE) and a control group
(n = 7, age 22.1 ± 0.8 years, height 175.1 ± 2.6 cm,
body mass 73.9 ± 4.1 kg). The subjects were physically
active but had not performed in any organized program of
regular exercise for at least 1 year before testing. They
were fully informed of the procedures to be utilized, as
well as the purpose of the study. Written informed consent
was obtained from all subjects. This study was approved by
the Ethics Committee for Human Experiments, Department
of Life Science (Sports Sciences), University of Tokyo.
Experimental design
In the weeks preceding the training period, the subjects
were asked to visit the laboratory to become familiar with
all testing procedures (see below). The subjects were tested
every month of the training (3 months) and detraining
(3 months) periods (seven times).
Training and detraining
The training group performed unilateral (left side) iso-
metric plantar flexion exercise in prone position. The left
ankle joint was set at 90° (anatomical position) with the
knee joint at full extension, and the foot was securely
strapped to a foot plate connected to the lever arm of a
dynamometer (Biodex-System 3, Biodex Medical, Sys-
tems, Inc, Shirey, NY, USA). The subjects trained four
times per week for 3 months and detrained for the fol-
lowing 3 months. The training protocol involved 15 con-
tractions (80% of maximal voluntary isometric strength;
MVC) of 15 s duration with a 30 s rest between each. A
measurement of MVC was made every month to adjust the
Eur J Appl Physiol (2012) 112:2679–2691
training load. After the training period (3 months), the
subjects entered a period of detraining (3 months). They
were instructed to return to their usual lifestyle and level of
physical activity during this period.
Muscle strength and neural activation level
MVC of the plantar flexor muscles was determined using a
specially designed dynamometer (Applied Office, Tokyo,
Japan). The subject lay prone on a test bench and the waist
and shoulders were secured by adjustable lap belts and held
in position. The ankle joint was set at 90° with the knee
joint at full extension, and the foot was securely strapped to
a footplate connected to the lever arm of the dynamometer.
Prior to the test, the subject performed a standardized
warm-up and submaximal contractions to become accus-
tomed to the test procedure. This task was repeated two or
three times per subject with at least three minutes between
trials. The highest value among these trials was recorded as
the muscle strength for each.
When the voluntary torque peaked, evoked twitch con-
tractions were imposed by supramaximal electrical stimu-
lations. The stimulating lead electrodes were placed on the
skin of the left popliteal fossa and oriented longitudinal to
the estimated path of the tibial nerve with the anode distal.
Rectangular pulses (triple stimuli with a 500 ls duration
for one stimulus and an interstimulus interval of 10 ms)
were delivered using a high-voltage stimulator (SEN-3301,
having a specially modified isolator SS-1963, Nihon-Ko-
den, Japan). The difference between peak twitch torque and
MVC (twitch torque) was measured. Shortly (within 1–2 s)
after MVC, the same stimulation was given to the muscle
at rest (control twitch torque). The neural activation
level (%) of the plantar flexor muscles was calculated as:
{1 - (twitch torque during MVC/control twitch tor-
que)} 9 100 as previously reported (e.g., Kubo et al. 2007).
This task was repeated two or three times per subject with at
least 3 min between trials. The highest value among these
trials was recorded as the neural activation level for each.
Mechanical properties of tendon
The subjects were instructed to develop a gradually
increasing force from a relaxed state to MVC within 5 s,
followed by a gradual relaxation within 5 s. The task was
repeated two times per subject with at least 3 min between
trials. An ultrasonic apparatus (SSD-6500, Aloka, Tokyo,
Japan) with an electronic linear array probe (7.5-MHz
wave frequency with 80 mm scanning length; UST 5047-5,
Aloka) was used to obtain longitudinal ultrasonic images of
the medial gastrocnemius muscle (MG). The probe was
longitudinally attached to the dermal surface with adhesive
tape, which prevented it from sliding. To evaluate the
2681
elongation of the Achilles tendon, the displacement of the
distal myotendinous junction of MG in the transition from
a resting state to MVC was measured. In the present study,
the Achilles tendon was defined as the distance from the
Achilles tendon insertion on the calcaneus to the distal
myotendinous junction of MG. Ultrasonic images were
recorded on videotape at 30 Hz, and synchronized with
force recordings using a clock timer for subsequent
analyses.
Tendon displacement is attributed to both angular rota-
tion and contractile tension, since any angular joint rotation
occurs in the direction of ankle plantar flexion during an
‘‘isometric’’ contraction (e.g., Magnusson et al. 2001). To
monitor ankle joint angular rotation, an electrical goni-
ometer (Penny and Giles, Biometrics Ltd, Gwent, UK) was
placed on the lateral aspect of the ankle. To correct the
measurements taken for the elongation of the Achilles
tendon, additional measurements were made under passive
conditions. Displacement of the myotendinous junction of
MG caused by rotating the ankle from 90° to 70° was
digitized in sonographs taken as described above. Thus, for
each subject, displacement of the myotendinous junction
obtained from the ultrasound images could be corrected for
that attributed to joint rotation alone (e.g., Magnusson et al.
2001). In the present study, only values corrected for
angular rotation are reported.
The measured torque (TQ) during isometric plantar
flexion was converted to tendon force (Ft) by the following
equation:
Ft ¼ TQ  MA1 ;
where MA is the moment arm length of the triceps surae
muscles at 90° of the ankle joint, which is estimated from
the lower leg length of each subject (Grieve et al. 1978). In
the present study, the tendon force and elongation values
above 50% of MVC were fitted to a linear regression
equation, the slope of which was adopted as stiffness
(Kubo et al. 1999). The area within the tendon force–
elongation loop, as a percentage of the area beneath the
curve during the ascending phase, was calculated as hys-
teresis. (Kubo et al. 2002).
The repeatability of the tendon stiffness and hysteresis
measurements was investigated on 2 separate days in our
previous study with 8 young males (Kubo et al. 2002). The
test–retest correlation coefficient was 0.89 for stiffness and
0.85 for hysteresis. The coefficient of variation was 5.6%
for stiffness and 10.2% for hysteresis.
Electromyographic activity (EMG)
Electromyographic activity was recorded during the mea-
surements of tendon properties. Bipolar surface electrodes
(5 mm in diameter) were placed over the bellies of the MG,
123
2682 Eur J Appl Physiol (2012) 112:2679–2691

lateral gastrocnemius (LG), soleus (SOL), and tibialis young males (Kubo et al. 2002). The test–retest correlation
anterior (TA) muscles with a constant interelectrode dis- coefficient was 0.92 and the coefficient of variation was
tance of 25 mm. The electrodes were connected to a pre- 2.4%. In a preliminary study, the repeatability of the ten-
amplifier and differential amplifier with a bandwidth of don volume measurement was investigated on 2 separate
5–500 Hz (model 1253A, NEC Medical Systems, Tokyo, days with 8 young males. The test–retest correlation
Japan). The EMG signals were transmitted to a computer at coefficient was 0.89 and the coefficient of variation was
a sampling rate of 1 kHz. The EMG was full-wave rectified 4.1%.
and averaged for the duration of the contraction (mEMG).
In addition, the mean of mEMG in the MG, LG, and SOL Echointensity and MRI signal intensity of tendon
muscles was defined as the mEMG of plantar flexors
(mEMGPF). During the measurements of tendon properties, The cross-sectional image of the Achilles tendon was
the mEMG of TA was measured to investigate the antag- measured at 20 mm proximal to the calcaneus using
onist muscle activity of TA (co-activation level). To B-mode ultrasonography (SSD-6500, Aloka, Japan)
determine the maximal activation of TA, a maximal dorsi- (Fig. 1a). The settings (gain, focus, time gain compensa-
flexion isometric contraction was performed at the same tion) were held constant throughout the experimental per-
angle (90° of ankle joint). The mEMG value of TA was iod. The echointensity (mean gray value) of the Achilles
normalized with respect to that at the same angle at max- tendon was determined using open-source image analysis
imal effort. software (Image J, NIH, Bethesda, MD).
When tendon CSA was measured using MRI as men-
Muscle and tendon volume tioned above, a plastic tube containing 10% CuSO4 was
placed on the malleolus (Fig. 1b). The MRI signal intensity
Measurements of muscle cross-sectional area (CSA) were (mean gray value) of the Achilles tendon was determined
carried out by MRI (Resona, 1.5 Tesla System, GE). T1- for 8 slices from the proximal tuberosity of the calcaneous,
weighted spin-echo, axial-plane imaging was performed
with the following parameters; TR 850 ms, TE 25 ms,
matrix 256 9 256, field of view 250 mm, slice thickness
10 mm, and interslice gap 0 mm. The subjects were
imaged in a supine position with the knee and ankle kept at
0° (full extension) and 90° (anatomical position), respec-
tively. During the scanning, the subject lay supine with the
base of the foot resting on a polystyrene block to maintain
an ankle angle of 90°. The number of sections obtained for
each subject was 37–44. The muscles investigated were as
follows: MG, LG, and SOL. From the axial image, outlines
of each muscle were traced, and the traced images were
transferred to a computer for calculating CSA using open-
source image analysis software (Image J, NIH, Bethesda,
MD). Muscle volume was determined by multiplying the
anatomical CSA of each image by the thickness (10 mm).
Immediately following the measurement of muscle
CSA, the measurement of tendon CSA was also measured
by MRI. The images were obtained using the following
parameters: TR 850 ms, TE 25 ms, matrix 256 9 256,
field of view 250 mm, slice thickness 3 mm, and interslice
gap 0 mm. The most distal image was obtained just above,
and the most proximal image 72 mm above, the proximal
tuberosity of the calcaneous. The number of sections
obtained for each subject was 24. The average CSAs of
every three images were calculated, and tendon volume
was determined by multiplying the anatomical CSA of
each image by the thickness (3 mm).
The repeatability of muscle volume measurement was Fig. 1 Transverse ultrasonic image (a) and axial magnetic resonance
investigated on 2 separate days in our previous study with 8 image (b) of the Achilles tendon

123
Eur J Appl Physiol (2012) 112:2679–2691
and then normalized to 10% CuSO4 (tendon signal inten-
sity/CuSO4 signal intensity) (Carroll et al. 2008, 2011).
In a preliminary study, the repeatability of the echoin-
tensity and MRI signal intensity measurements was
investigated on 2 separate days with 8 young males. The
test–retest correlation coefficient was 0.88 for the echoin-
tensity and 0.86 for the MRI signal intensity. The coeffi-
cient of variation was 5.2% for the echointensity and 4.8
for the MRI signal intensity.
Blood volume and oxygen saturation of tendon
During quiet resting, we measured the blood volume (total
hemoglobin: THb) and oxygen saturation (StO2) of the
Achilles tendon using red laser lights (BOM-L1TRSF,
Omega Wave, Tokyo, Japan). A probe (SF-DS, Omega
Wave, Tokyo, Japan) was positioned 30 mm proximal to
the calcaneus. Stained dots ensured the same positioning of
the probe in each test during the experimental period. This
instrument uses three red laser lights (635, 650, and
690 nm), and calculates the relative tissue levels of oxy-
hemoglobin (OxyHb), deoxyhemoglobin (DeoxyHb) and
THb. The distance between the light source and photode-
tector was 5 mm. According to the findings of Kashima
(2003), measuring depth was estimated as 3–5 mm when
the distance between the light source and photodetector
was 5 mm. The details of this technique and principles of
this instrument have been described elsewhere (Kashima
2003; Kubo et al. 2008a, b). Briefly, two-point detection
and the differential calculation method were used to mea-
sure THb and StO2 only in the deep region of the tissue
(measurement depth of 3–5 mm). THb and StO2 at specific
tissue depths could be measured by changing the location
of the two detectors. The offset value of THb was reduced,
and highly sensitive measurements were achieved using the
two-point detection method. In the present study, the units
of OxyHb, DeoxyHb, and THb are expressed in lmol/L,
although this does not represent actual physical volume.
The StO2 was calculated from OxyHb and THb values
using the following formula: StO2 (%) = 100 9 OxyHb/
THb.
The repeatability of the THb and StO2 measurements for
tendon during resting was investigated on 6 separate days
in our previous study involving one subject (Kubo et al.
2008b). The coefficient of variation was 8.8% for THb and
1.6% for StO2.
Serum biochemical markers of type 1 collagen
synthesis
Blood samples (20 mL) of the subjects were taken from the
antecubital vein by a medical practitioner using veni-
puncture. The details of this measurement have been
2683
described elsewhere (Kubo et al. 2011). Briefly, the blood
samples were centrifuged at 3,000 rpm for 10 min, and
then the serum was separated and stored at -20°C until
analysis. In the present study, we measured serum con-
centrations of two markers; BAP and P1P (also called
P1CP). BAP is a specific marker of bone synthesis
(Garnero and Delmas 1993; Risteli and Risteli 1993) and
was measured by an enzyme immunoassay kit (Osteolinks
BAP, Quidel, San Diego, CA, USA). Intra-assay and inter-
assay CVs were less than 7 and 9%, respectively. P1P
reflects the synthesis of type 1 collagen in other organs as
well as bone (Jackson et al. 2003; Langberg et al. 2000),
and was measured by an enzyme immunoassay kit (Pro-
collagen Type 1 C-peptide EIA Kit, Takara Bio Inc,
Tokyo, Japan). The intra-assay and inter-assay CVs were
less than 8% for both tests.
The repeatability of the BAP and P1P measurements
was investigated on 2 separate days in our previous study
with 8 young males (Kubo et al. 2011). The coefficient of
variation was 2.7% for BAP and 7.2% for P1P.
Statistics
Values are reported as mean ± standard error (SE). A two-
way ANOVA with repeated measures {2 (groups) 9 7 (test
times)} was used to analyze the data. The F ratios for main
effects and interactions were considered significant at
P \ 0.05. Significant differences among means at P \ 0.05
were detected using a Tukey post hoc test. To assess the
relationships among measured parameters, the Pearson
product–moment correlation was computed.
Results
MVC, activation level (using the interpolated twitch tech-
nique) and mEMGPF increased significantly by 27.6 ± 9.1,
6.1 ± 2.2 and 44.2 ± 13.1%, respectively, after 3 months
of training (Table 1). However, these values did not change
during the 3 months of detraining. No change in the co-
activation level was found during training and detraining
periods (P = 0.189). The volume of plantar flexor muscles
increased by 2.9 ± 0.9% after 2 months and 4.1 ± 1.0%
after 3 months, and decreased to the pre-training level after
1 month of detraining (Table 1).
The average tendon force–elongation relationships at
each time point during training and detraining are shown in
Fig. 2. Although the maximal elongation of tendon
remained unchanged during 3 months of training, these
values during detraining were significantly greater com-
pared to the pre-training level (Table 2). Tendon stiffness
did not change until 2 months of training, but the increase
(50.3 ± 12.2%) reached statistical significance at the end
123
2684 Eur J Appl Physiol (2012) 112:2679–2691

Table 1 Measured variables of muscle for the training group

Pre 1 month 2 months Post De-1 month De-2 months De-3 months

Maximum voluntary 130 (11) 146 (8)* 155 (8)*** 160 (8)*** 152 (9)** 149 (6)** 149 (7)**
contraction (Nm)
Activation level (%) 93.3 (1.8) 96.7 (0.8)* 98.2 (0.5)** 98.7 (0.4)*** 97.5 (0.7)** 96.5 (0.8)** 97.0 (0.6)*
mEMGPF (mV) 0.41 (0.05) 0.48 (0.05)* 0.50 (0.05)** 0.57 (0.04)** 0.53 (0.05)** 0.47 (0.04)* 0.48 (0.04)*
Co-activation level (%) 9.4 (0.7) 9.5 (0.7) 10.7 (0.9) 10.4 (0.5) 11.5 (1.1) 9.9 (0.7) 9.2 (0.7)
Muscle volume (cm3) 859 (49) 871 (47) 883 (46) *** 894 (47) *** 867 (46) 866 (47) 868 (45)
Values represent mean (SE)
Pre before training, 1 month after 1 month of training, 2 months after 2 months of training, Post after 3 months of training, De-1 month after
1 month of detraining, De-2 months after 2 months of detraining, De-3 months after 3 months of detraining
* Significantly different from Pre (* P \ 0.05, ** P \ 0.01, *** P \ 0.001)

hysteresis was found during the training and detraining

periods (P = 0.651; Table 2). In addition, no changes in
average tendon CSA every three slices (Fig. 4) and tendon
volume (P = 0.381; Table 2) were found over the course
of the experimental period.
THb and StO2 did not change during the experiment
(THb, P = 0.318; StO2 P = 0.069) (Table 2). Similarly,
no changes in P1P and BAP were found during training or
detraining (P1P, P = 0.133; BAP, P = 0.900) (Table 2).
However, the relative increase in P1P (to pre-training level)
after 2 months of training was significant (Fig. 3b).
Echointensity increased by 9.1 ± 2.5% after 2 months
of training, and remained high throughout the experimental
period, except at 3 months of detraining (Table 2; Fig. 3c).
MRI signal intensity increased by 24.2 ± 6.3% after
2 months and by 21.4 ± 5.8% after 3 months of training,
but decreased to the pre-training level during the detraining
period (Table 2; Fig. 3d). The relative changes in tendon
stiffness were significantly correlated with those in the
MRI signal intensities at 2 months of training, and 1 month
(P = 0.07) and 2 months of detraining (Fig. 5), but not to
those in the echointensities at all the measured periods
(Fig. 6).
In the control group, no changes in the measured vari-
ables of muscle (Table 3) and tendon (Table 4) were found
over the course of the experiment.

Discussion

In the present study, we tried to estimate changes in

Fig. 2 The relation between tendon force and tendon elongation during
training (a) and detraining (b). Pre before training, 1 mon after 1 month
of training, 2 mon after 2 months of training, Post after 3 months of
training, De-1 mon after 1 month of detraining, De-2 mon after
2 months of detraining, De-3 mon after 3 months of detraining
of the training period. After 1 month of detraining, tendon
stiffness had already decreased to pre-training level
(Table 2; Fig. 3a). No significant change in tendon
metabolism in human tendons (blood circulation, collagen
synthesis, etc.) in order to elucidate the mechanism of
change in the mechanical properties of tendons during
resistance training and detraining.
First, THb and StO2 of the Achilles tendon did not
change during the experimental period. This result relating
to the blood volume of tendons was consistent with that of
a longitudinal experiment (Kubo et al. 2009b). Our recent

123
 Eur J Appl Physiol (2012) 112:2679–2691
Table 2 Measured variables of tendon for the training group
Pre 1 month 2 months Post De-1 month De-2 months De-3 months

Maximal elongation (mm) 18.8 (0.9) 20.8 (0.9) 21.0 (0.8) 18.5 (0.6) 23.1 (1.2)*** 23.1 (1.1)*** 23.4 (0.7)***
Stiffness (N/m) 183 (15) 191 (11) 210 (18) 277 (37)** 206 (23) 192 (19) 186 (10)
Hysteresis (%) 18.1 (4.0) 20.8 (3.8) 17.5 (2.7) 14.5 (3.6) 16.7 (4.1) 22.2 (4.0) 15.7 (34)
Tendon volume (mm3) 5,216 (154) 5,305 (130) 5,350 (138) 5,358 (154) 5,300 (135) 5,311 (129) 5,321 (140)
THb (lmol/L) 12.2 (0.5) 11.9 (0.6) 12.3 (0.5) 11.6 (0.3) 11.6 (0.4) 11.1 (0.3) 12.0 (0.6)
StO2 (%) 62.7 (1.1) 61.6 (1.2) 62.0 (1.3) 60.3 (1.2) 58.9 (0.8) 60.0 (0.7) 62.1 (1.7)
BAP (U/L) 31.0 (4.5) 31.1 (4.0) 31.2 (5.3) 31.7 (3.0) 29.4 (3.3) 31.4 (4.0) 31.7 (4.2)
PlP (ng/mL) 940 (139) 1,022 (150) 1,202 (139) 1,104 (138) 1,021 (126) 1,054 (133) 1,035 (152)
Echointensity (mean gray value) 54.7 (1.7) 56.7 (1.9) 59.5 (1.8)* 59.1 (1.3)* 59.2 (2.1)* 58.9 (1.8)* 57.9 (2.1)
MRI signal density (relative mean gray value) 0.69 (0.06) 0.76 (0.08) 0.85 (0.07)** 0.83 (0.07)* 0.76 (0.06) 0.73 (0.06) 0.78 (0.06)
Values represent mean (SE)
Pre before training, 1 month after 1 month of training, 2 months after 2 months of training, Post after 3 months of training, De-1 month after 1 month of detraining, De-2 months after 2 months
of detraining, De-3 months after 3 months of detraining
* Significantly different from Pre (* P \ 0.05, ** P \ 0.01, *** P \ 0.001)
icantly different from Pre (*P \ 0.05, **P \ 0.01)
detraining, De-3 mon after 3 months of detraining. Asterisk Signif-
1 mon after 1 month of detraining, De-2 mon after 2 months of
2 mon after 2 months of training, Post after 3 months of training, De-
training group. Pre before training, 1 mon after 1 month of training,
Achilles tendons during the training and detraining periods for the
type 1 C-peptide (b), echointensity (c) and MRI signal intensity (d) of
Fig. 3 Relative changes in tendon stiffness (a), serum procollagen
training, although it increased significantly after dynamic
did not change after 3 months of isometric knee extension
study showed that the blood volume of the patellar tendon
123

2685
2686
Fig. 4 Average values of the cross-sectional area of Achilles tendons
every three images during the training and detraining periods for the
training group. Pre before training, 1 mon after 1 month of training,
2 mon after 2 months of training, Post after 3 months of training, De-
1 mon after 1 month of detraining, De-2 mon after 2 months of
detraining, De-3 mon after 3 months of detraining
training (Kubo et al. 2009b). Furthermore, we reported that
blood volume and oxygen saturation values for the Achilles
tendon did not change after repeated isometric plantar
flexions with long duration, but increased significantly after
repeated ballistic contractions (Kubo et al. 2009a). To
repair damage involving mechanical loading of the tendon,
it would be necessary to increase blood volume of the
tendon to supply the necessary oxygen. Indeed, the oxygen
consumption of tendon increased significantly after the
repeated muscle contractions (Boushel et al. 2000; Kubo
et al. 2008a). On the other hand, previous studies showed
that the application of lactate to tendon fibroblasts
increased collagen production (Klein et al. 2001; Yal-
amanchi et al. 2004). Considering the present and previous
results, it is likely that insufficient blood volume and
oxygen saturation within the tendon after longer-duration
isometric contractions would lead to the accumulation of
‘‘surplus’’ lactate, which could be a factor inducing the
marked increase in type 1 collagen synthesis within the
tendon (this point will be discussed later).
Second, we tried to clarify the changes in synthesis of
type 1 collagen in tendons by comparing the dynamics of
two biochemical markers (BAP and P1P). In the present
study, serum BAP levels (reflecting newly synthesized
bone) did not change during the experimental period. This
result implied that the unilateral plantar flexion exercise in
the present study had no effect on bone metabolism. More
recently, we reported that P1P (reflecting synthesis of type
1 collagen in other organs as well as bone) increased at
24 h after static knee extension exercises (Kubo et al.
2011). Taking the present result into account together with
123
Eur J Appl Physiol (2012) 112:2679–2691
our recent finding, we can say that the production of type 1
collagen in the Achilles tendon was enhanced at the
2 months point of the training period. In the present study,
however, P1P decreased at the end of resistance training
(3 months of training). As far as we know, few studies have
examined the time course of changes in biochemical
markers during training (Langberg et al. 2001). Langberg
et al. (2001) reported that tissue and plasma P1CP levels
increased significantly after 4 weeks of resistance training,
but tended to decrease after 11 weeks. Accordingly, it is
likely that the enhanced collagen synthesis due to chronic
exercise (i.e., resistance training) reach a limit during
3 months of training.
Third, the echointensity of tendons increased during the
training period, and remained high during the detraining
period. Some previous researchers demonstrated increases
in the echointensity of muscle tissue to indicate increases in
fat, fluid, and fibrosis (van Holsbeeck and Introcaso 1992;
Nosaka and Clarkson 1996; Reimers et al. 1993a, b). For
example, Reimers et al. (1993a) reported an average
increase in muscle echointensity of 0.7%/year in women
and 1.1%/year in men. This result implied that the amount
of intramuscular fat increased with age. On the other hand,
Reimers et al. (1993b) showed using biopsy technique that
the echointensity of muscle was correlated highly with
muscular lipomatosis, but not fibrosis. Given the limited
training period (3 months), however, it might be wrong to
assume that the lipomatosis within the tendon increases
after training. Concerning changes in fluid, previous studies
demonstrated that increases in the echointensity of muscle
were related to plasma enzyme levels and edema within the
muscle after exercise (e.g., Nosaka and Clarkson 1996). In
the present study, however, the echointensity of the ten-
dons remained high until 2 months of detraining, compared
to the pre-training level. Considering these points, it seems
reasonable to suppose that an increase in the echointensity
of tendons implies an increase in the amount of collagen
fibers but not fat or fluid within the tendon. Indeed, Kon-
gsgaard et al. (2009) demonstrated in vivo using biopsy
techniques that the collagen content of human patellar
tendons tended to increase after 12 weeks of resistance
training.
Fourth, the MRI signal intensity of the tendons increased
significantly during training, but decreased to the pre-
training level during the detraining period. These results
were different from those of echointensity. If the increase
in MRI signal intensity of tendon mostly represents an
increase in the amount of collagen fibers, relative increases
in MRI signal intensity would be correlated to those in
echointensity. In the present study, however, no significant
correlation was found between these parameters at any
time point (data not shown). These results implied that
the changes in MRI signal intensity and echointensity
Eur J Appl Physiol (2012) 112:2679–2691 2687

Fig. 5 The relationship

between relative changes (to the
pre-training level) in tendon
stiffness and MRI signal
intensity in all the measured
periods. 1 mon after 1 month of
training, 2 mon after 2 months
of training, Post after 3 months
of training, De-1 mon after
1 month of detraining, De-2
mon after 2 months of
detraining, De-3 mon after
3 months of detraining

represented different phenomena. According to previous
findings (Shalabi et al. 2002; Yoon and Halper 2005), a
change in MRI signal intensity indicated changes in tissue
hydration, proteoglycan and glycosaminoglycan content,
and collagen fiber structure. In particular, changes in T2
relaxation time (equivalent to MRI signal intensity) rep-
resented changes in the architecture of collagen fibers
within the tissues (Du et al. 2009; Erickson et al. 1993;
Nagy and Dyson 2011). To date, only a few reports have
been published regarding changes in the MRI signal
intensity of human tendons in vivo (Carroll et al. 2008,
2011). More recently, Carroll et al. (2011) showed that the
MRI signal intensity of patellar tendons did not change
after 12 weeks of knee extension training in elderly indi-
viduals. Furthermore, resistance training for 12 weeks did
not alter the fibril morphology, collagen content, or cross-
linking of patellar tendon in young male with patellar
tendinopathy (Kongsgaard et al. 2009). Possible reasons for
the difference between these previous and present findings
are the average age (young vs. elderly; Carroll et al. 2011)

123
2688 Eur J Appl Physiol (2012) 112:2679–2691

Fig. 6 The relationship

between relative changes (to the
pre-training level) in tendon
stiffness and echointenisty in all
the measured periods. 1 mon
after 1 month of training, 2 mon
after 2 months of training, Post
after 3 months of training, De-1
mon after 1 month of detraining,
De-2 mon after 2 months of
detraining, De-3 mon after
3 months of detraining

and state (healthy vs. tendinopathy; Kongsgaard et al.
2009) of subjects. In the present study, correlation coeffi-
cients between the relative increases in tendon stiffness and
MRI signal intensity were significant after 2 months of
training (r = 0.684, P \ 0.05), but not at 3 months (r =
-0.206, P [ 0.05). Considering these findings, we may say
that changes in the structure of collagen fibers within the
tendon mainly contribute to the change in tendon stiffness
in the middle of the training period (2 months).
At 1 month of detraining, maximal elongation of the
tendons was greater than that post-training. This result
agreed with our recent finding (Kubo et al. 2010). In
addition, tendon stiffness had already returned to the pre-
training level at 1 month of detraining. Similarly, previous
studies also showed that the tendon elongation at a given
force level increased significantly after only 3 weeks of
bed rest (de Boer et al. 2007; Kubo et al. 2004). Therefore,
the elongation of tendon would increase immediately once
the mechanical stress to the tendon during daily life and/or
resistance training was removed. During the detraining
period, MRI signal intensity returned to the pre-training
level, although echointensity (reflects mainly fibrosis

123
 Eur J Appl Physiol (2012) 112:2679–2691
Table 3 Measured variables of muscle for the control group
Pre 1 month 2 months Post De-1 month De-2 months De-3 months
Maximum voluntary contraction (Nm) 135 (14) 138 (11) 133 (9) 136 (12) 138 (10) 134 (11) 137 (13)
Activation level (%) 95.1 (1.9) 94.0 (2.1) 93.3 (2.5) 95.8 (1.7) 94.2 (2.1) 93.9 (2.5) 94.8 (1.7)
mEMGPF (mV) 0.43 (0.05) 0.40 (0.05) 0.52 (0.08) 0.49 (0.06) 0.48 (0.05) 0.44 (0.04) 0.49 (0.06)
Co-activation level (%) 12.1 (1.8) 13.3 (2.0) 12.2 (2.0) 13.5 (2.0) 12.6 (1.4) 12.2 (1.3) 12.2 (1.4)
Muscle volume (cm3) 906 (62) 909 (59) 908 (61) 916 (62) 899 (55) 902 (58) 913 (60)
Values represent mean (SE)
Pre before training, 1 month after 1 month of training, 2 months after 2 months of training, Post after 3 months of training, De-1 month after 1 month of detraining, De-2 months after 2 months
of detraining, De-3 months after 3 months of detraining
Table 4 Measured variables of tendon for the control group Mean (SE)
Pre 1 month 2 months Post De-1 month De-2 months De-3 months
Maximal elongation (mm) 21.4 (0.9) 22.0 (0.7) 20.9 (1.0) 21.2 (0.9) 21.8 (0.8) 20.7 (0.9) 21.7 (1.0)
Stiffness (N/m) 182 (24) 177 (20) 185 (25) 189 (28) 175 (19) 186 (22) 179 (23)
Hysteresis (%) 18.7 (3.3) 20.2 (4.1) 19.5 (4.0) 18.6 (3.9) 17.5 (4.6) 19.1 (3.5) 17.9 (3.9)
Tendon volume (mm3) 5,388 (141) 5,376 (132) 5,401 (145) 5,354 (138) 5,397 (133) 5,409 (142) 5,369 (140)
THb (l mol/L) 11.9 (0.3) 12.3 (1.0) 11.6 (0.8) 12.2 (0.3) 11.7 (0.7) 11.4 (0.7) 12.3 (0.5)
StO2 (%) 60.6 (1.3) 61.2 (1.3) 61.2 (1.2) 60.3 (1.5) 58.5 (0.6) 57.7 (0.7) 60.7 (1.0)
BAP (U/L) 28.9 (2.9) 28.8 (3.1) 27.1 (2.0) 29.1 (2.3) 26.4 (2.4) 24.4 (1.6) 24.3 (1.2)
PlP (ng/mL) 989 (156) 1,006 (122) 908 (145) 946 (101) 885 (132) 899 (128) 947 (117)
Echointensity (mean gray value) 60.6 (3.6) 58.8 (2.5) 62.2 (2.0) 62.8 (2.4) 63.7 (2.8) 61.1 (2.2) 61.7 (2.0)
MPJ signal density (relative mean gray value) 0.73 (0.08) 0.70 (0.06) 0.74 (0.10) 0.69 (0.07) 0.73 (0.09) 0.71 (0.08) 0.74 (0.08)
Values represent mean (SE)
Pre before training, 1 month after 1 month of training, 2 months after 2 months of training, Post after 3 months of training, De-1 month after 1 month of detraining, De-2 months after 2 months
of detraining, De-3 months after 3 months of detraining
123

2689
2690
within the tissues) remained high compared to pre-training.
Furthermore, the relative changes in tendon stiffness were
correlated to those in MRI signal intensities at 1 month
(P = 0.07) and 2 months of detraining (P \ 0.05). Con-
sidering these findings, the decrease in tendon stiffness
during the detraining period would indicate that the struc-
ture of collagen fibers within the tendon returned to the pre-
training level, while the content of collagen fibers remained
high. In other words, the mechanisms which resulted in the
increase in tendon elongation during detraining mainly
involved changes in the structure of collagen fibers within
the tendon.
In conclusion, our results demonstrated that tendon
stiffness adapted to resistance training slower, but to
detraining faster, than muscle strength. These results sup-
ported our recent finding (Kubo et al. 2010). Furthermore,
from a metabolic point of view, collagen synthesis (esti-
mated from serum biochemical markers), content (esti-
mated from echointensity), and structure (estimated from
MRI signal intensity) in the human tendons changed at the
2-month point of the training period. During detraining, the
sudden decrease in tendon stiffness might be related to
changes in the structure of collagen fibers within the ten-
don. Regardless, further investigations using animals are
needed to clarify this point. The methodology used to
estimate blood circulation, collagen synthesis and structure
of human tendons in vivo should help to elucidate the
mechanisms of tendon plasticity.
Acknowledgments This study was supported by a Grant-in-Aid for
Young Scientists (A) (21680047 to K. Kubo) from Japan Society for
the Promotion of Science and Yamaha Motor Foundation for Sports.
The authors thank the staff of Kouikai Clinic for their technical
assistance with MRI measurements. The authors also thank Mr. Yuki
K and Mr. Sato H (Kishimoto Clinical Laboratory Group) for their
conscientious work on the analyses of serum biochemical markers.
References
Arampatzis A, Karamanidis K, Albracht K (2007) Adaptational
responses of the human Achilles tendon by modulation of the
applied cyclic strain magnitude. J Exp Biol 210:2743–2753
Boushel R, Langberg H, Green S, Skovgaard D, Bulow J, Kjaer M
(2000) Blood flow and oxygenation in peritendinous tissue and
calf muscle during dynamic exercise in humans. J Physiol
524:305–313
Burgess KE, Connick MJ, Graham-Smith P, Pearson SJ (2007)
Plyometric vs. isometric training influences on tendon properties
and muscle output. J Strength Cond Res 21:986–989
Carroll CC, Dickinson JM, Haus JM, Lee GA, Hollon CJ, Aagaard P,
Magnusson SP, Trappe TA (2008) Influence of aging on the in
vivo properties of human patellar tendon. J Appl Physiol
105:1907–1915
Carroll CC, Dickinson JM, LeMoine JK, Haus JM, Weinheimer EM,
Hollon CJ, Aagaard P, Magnusson SP, Trappe TA (2011)
Influence of acetaminophen and ibuprofen on in vivo patellar
123
Eur J Appl Physiol (2012) 112:2679–2691
tendon adaptations to knee extensor resistance exercise in older
adults. J Appl Physiol (in press)
de Boer MD, Maganaris CN, Seynnes OR, Rennie MJ, Narici MV
(2007) Time course of muscular, neural and tendinous adapta-
tions to 23 day unilateral lower-limb suspension in young men.
J Physiol 583:1079–1091
Du J, Pak BC, Znamirowski R, Statum S, Takahashi A, Chung C,
Bydder GM (2009) Magic angle effect in magnetic resonance
imaging of the Achilles tendon and enthesis. Mag Reson Imag
27:557–564
Erickson SJ, Prost RW, Timins ME (1993) The ‘‘magic angle’’ effect:
background physics and clinical relevance. Radiology
188:23–25
Greive DW, Pheasant S, Cavagna, PR (1978) Prediction of gastroc-
nemius length from knee and ankle joint posture. In: Asmussen
E, Jorgensen K (eds) Biomechanics VI-A, International series in
biomechanics. University Park Press, Baltimore, pp 405–412
Garnero P, Delmas PD (1993) Assessment of the serum levels of bone
alkaline phosphatase with a new immunoradiometric assay in
patients with metabolic bone disease. J Clil Endocrinol Met
77:1046–1053
Jackson BF, Smith RKW, Price JS (2003) A molecular markers of
type 1 collagen metabolism reflects changes in connective tissue
remodeling associated with injury to the equine superficial
digital flexor tendon. Equine Ver J 35:211–213
Kashima S (2003) Spectroscopic measurement of blood volume and
its oxygenation in a small volume of tissue using red laser lights
and differential calculation between two point detections. Optics
Laser Technol 35:485–489
Klein MB, Pham H, Yalamanchi N, Chang J (2001) Flexor tendon
wound healing in vitro: the effects of lactate on tendon cell
proliferation and collagen production. J Hand Surg 26:847–854
Kongsgaard M, Kovanen V, Aagaard P, Doessing S, Hansen P,
Laursen AH, Kaldau NC, Kjaer M, Magnusson SP (2009)
Corticosteroid injections, eccentric decline squat training and
heavy slow resistance training in patellar tendinopathy. Scand J
Med Sci Sports 19:790–802
Kubo K, Kawakami Y, Fukunaga T (1999) Influence of elastic
properties of tendon structures on jump performance in humans.
J Appl Physiol 87:2090–2096
Kubo K, Kanehisa H, Fukunaga T (2002) Effects of resistance and
stretching training programs on the viscoelastic properties of
tendon structures in vivo. J Physiol 538:219–226
Kubo K, Akima H, Ushiyama J, Tabata I, Fukuoka H, Kanehisa H,
Fukunaga T (2004) Effects of resistance training during bed rest
on the viscoelastic properties of tendon structures in lower limb.
Scand J Med Sci Sports 14:296–302
Kubo K, Morimoto M, Komuro T, Yata H, Tsunoda N, Kanehisa H,
Fukunaga T (2007) Effects of plyometric and weight training on
muscle-tendon complex and jump performance. Med Sci Sports
Exer 39:1801–1810
Kubo K, Ikebukuro T, Tsunoda N, Kanehisa H (2008a) Changes in
oxygen consumption of human muscle and tendon following
repeat muscle contractions. Eur J Appl Physiol 104:859–866
Kubo K, Ikebukuro T, Tsunoda N, Kanehisa H (2008b) Noninvasive
measures of blood volume and oxygen saturation of human
Achilles tendon by red laser lights. Acta Physiol 193:257–264
Kubo K, Ikebukuro T, Yaeshima K, Kanehisa H (2009a) Effects of
different duration contractions on elasticity, blood volume, and
oxygen saturation of human tendon in vivo. Eur J Appl Physiol
106:445–455
Kubo K, Ikebukuro T, Yaeshima K, Yata H, Tsunoda N, Kanehisa H
(2009b) Effects of static and dynamic training on the stiffness
and blood volume of tendon in vivo. J Appl Physiol
106:412–417
Eur J Appl Physiol (2012) 112:2679–2691
Kubo K, Ikebukuro T, Yata H, Tsunoda N, Kanehisa H (2010) Time
course of changes in muscle and tendon properties during
strength training and detraining. J Strength Cond Res 24:322–331
Kubo K, Yuki K, Ikebukuro T (2011) Changes in bone alkaline
phosphatase and procollagen type-1 C-peptide after static and
dynamic exercises. Res Quart Exer Sport (in press)
Langberg H, Skovgaard D, Karamouzis M, Bulow J, Kjaer M (1999)
Metabolism and inflammatory mediators in the peritendinous
space measured by microdialysis during intermittent isometric
exercise in humans. J Physiol 515:919–927
Langberg H, Skovgaard D, Asp S, Kjaer M (2000) Time pattern of
exercise-induced changes in type 1 collagen turnover after
prolonged endurance exercise in humans. Calcif Tissue Int
67:41–44
Langberg H, Rosendal L, Kjaer M (2001) Training-induced changes
in peritendinous type 1 collagen turnover determined by
microdialysis in humans. J Physiol 534:297–302
Magnusson SP, Aagaard P, Rosager S, Poulsen PD, Kjaer M (2001)
Load-displacement properties of the human triceps surae
aponeurosis in vivo. J Physiol 531:277–288
Maimoun L, Manetta J, Couret I, Dupuy AM, Mariano-Goulart D,
Micallef JP, Peruchon E, Rossi M (2006) The intensity level of
physical exercise and the bone metabolism response. Int J Sports
Med 27:105–111
Miller BF, Olesen JL, Hansen M, Dossing S, Crameri RM, Welling
RJ, Langeberg H, Flyvbjerg A, Kjaer M, Babraj JA, Smith K,
Rennie MJ (2005) Coordinated collagen and muscle protein
synthesis in human patella tendon and quadriceps muscle after
exercise. J Physiol 567:1021–1033
Nagy A, Dyson S (2011) Anatomical, magnetic resonance imaging
and histological findings in the accessory ligament of the deep
digital flexor tendon of forelimbs in nonlame horses. Equine Ver
J 43:309–316
Nieminen MT, Rieppo J, Toyras J, Hakumaki JM, Silvennoinen J,
Hyttinen MM, Helminen HJ, Jurvelin JS (2001) T2 relaxation
2691
reveals spatial collagen architecture in articular cartilage: a
comparative quantitative MRI and polarized light microscopic
study. Mag Reson Med 46:487–493
Nosaka K, Clarkson PM (1996) Changes in indicators of inflamma-
tion after eccentric exercise of the elbow flexors. Med Sci Sports
Exer 28:953–961
Reimers CD, Fleckenstein JL, Witt TN, Muller-Felber W, Pongratz
DE (1993a) Muscular ultrasound in idiopathic inflammatory
myopathies of adults. J Neurol Sci 116:82–92
Reimers K, Reimers CD, Wagner S, Paetzke I, Pongratz DE (1993b)
Skeletal muscle sonography. A correlative study of echogenicity
and morphology. J Ultrasound Med 12:73–77
Risteli L, Risteli J (1993) Biochemical markers of bone metabolism.
Ann Med 25:385–393
Shalabi A, Kristoffersen-Wiberg M, Papadogiannakis N, Aspelin P,
Movin T (2002) Dynamic contrast-enhanced MR imaging and
histopathology in chronic Achilles tendinosis. A longitudinal
MR study of 15 patients. Acta Radiol 43:198–206
Soyama N, Kuratsu J, Ushio Y (1995) Correlation between magnetic
resonance images and histology in meningiomas: T2-weighted
images indicate collagen contents in tissue. Neurol Med Chir
35:438–441
van Holsbeeck M, Introcaso JH (1992) Musculoskeletal ultrasonog-
raphy. Radiol Clin Am 30:907–925
Virtanen P, Viitasalo JT, Vuori J, Vaananen K, Takala TES (1993)
Effect of concentric exercise on serum muscle and collagen
markers. J Appl Physiol 75:1272–1277
Yalamanchi N, Klein MB, Pham HM, Longaker MT, Chang J (2004)
Flexor tendon wound healing in vitro: lactate up-regulation of
TGF-ß expression and functional activity. Plastic Reconstr Surg
113:625–632
Yoon JH, Halper J (2005) Tendon proteoglycans: biochemistry and
function. J Musculoskelet Neuronal Interact 5:22–34
123