Scand J Med Sci Sports 2002: 12: 90–98
Printed in Denmark . All rights reserved
Load-displacement properties of the human triceps surae
aponeurosis and tendon in runners and non-runners
S. Rosager, P. Aagaard, P. Dyhre-Poulsen2, K. Neergaard1, M. Kjaer, S. P. Magnusson
Team Danmark Test Center/Sports Medicine Research Unit, 1Department of Radiology, MRI Section, Bispebjerg Hospital,
Copenhagen, Denmark, 2Institute of Medical Physiology, The Panum Institute, University of Copenhagen, Copenhagen, Denmark
Accepted for publication 24 September 2001
The load-displacement and stress–strain characteristics of
the human triceps surae tendon and aponeurosis, in vivo, was
examined during graded maximal voluntary plantarflexion
efforts in runners who trained 80 km/ week or more and age-
matched non-runners. Synchronous real-time ultrasono-
graphy of triceps surae tendon and aponeurosis displace-
ment, electromyography of the gastrocnemius, soleus and
dorsiflexor muscles, and joint angular rotation were ob-
tained. Tendon cross-sectional area and ankle joint moment
arm were obtained from magnetic resonance imaging. Ten-
sile tendon force was calculated from the joint moments and
tendon moment arm and stress was obtained by dividing
force by cross-sectional area. Strain was obtained from the
displacements normalized to tendon length. Antagonist co-
activation and small amounts of ankle joint rotation sig-
nificantly affected tensile tendon force and aponeurosis and
Investigations of the effect of repeated habitual physi-
cal activity on human tendon in vivo is scarce. Never-
theless, the fact that the tendon is unable to adapt to
certain loading conditions is well recognized, and the
incidence of activity related tendon injuries is con-
siderable (Kvist, 1994).
Muscle activation results in force transmission to
bone via tendons to produce joint movement. How-
ever, tendons are not inextensible, but have non-lin-
ear spring-like characteristics. These properties
allows for the tendon to be stretched for energy stor-
age, which is subsequently converted into kinetic en-
ergy upon release (Alexander & Vernon, 1975; Ker,
Alexander, Bennet, 1988; Biewener & Roberts, 2000).
The human Achilles tendon can be subjected to ap-
preciable loads during locomotion, which causes the
tendon to deform. It has been estimated that the hu-
man Achilles tendon load may reach tensile forces
of 1400–2600 N during walking (Giddings, Beaupre,
whalen, Carter, 2000; Finni, Komi, Lukkariniemi,
1998) and 3100–5330 N during running (Scott &
Winter, 1990; Giddings et al., 2000). However, the
tendinous deformation associated with these loads is
90
COPYRIGHT C MUNKSGAARD 2002 ¡ ISSN 0905-7188
tendon displacement, respectively (P ∞ 0.01). Plantarflex-
ion moment was similar in runners (138 ∫ 27 Nm, mean ∫
SEM) and non-runners (142 ∫ 17 Nm). Tendon moment
arm was alike in non-runner (58.3 ∫ 0.2 mm) and runners
(55.1 ∫ 0.1 mm). Similarly, there was no difference in ten-
don tensile force between runners (2633 ∫ 465 N) and non-
runners (2556 ∫ 401 N). The cross-sectional area of the
Achilles tendon was larger in runners (95 ∫ 3 mm2) than
non-runners (73 ∫ 3 mm2) (P ∞ 0.01). The load-deforma-
tion data yielded similar stiffness (runners 306 ∫ 61 N/mm,
non-runners 319 ∫ 42 N/mm). The maximal strain and
stress was 4.9 ∫ 0.8% and 38.2 ∫ 9.8 MPa in non-runners
and 4.1 ∫ 0.8% and 26.3 ∫ 5.1 MPa in runners. The larger
tendon cross-sectional area in trained runners suggests that
chronic exposure to repetitive loading has resulted in a
tissue adaptation.
largely unknown. Investigations of human tendon be-
havior have largely been limited to biomechanical
testing of isolated cadaver tissue specimen (Butler,
Grood, Noyes, Zernicke, Brackett, 1984; Rack &
Ross, 1984; McGough, Debski, Taskiran, Fu, Woo,
1987; Loren & Lieber, 1995; Goldstein, Armstrong,
Chaffin, Matthews, 1987), or invasive in vivo methods
(Amis, Prochazka, Short, Trend, Ward, 1987; Fel-
lows & Rack, 1987). Therefore, the recent advance
of using real-time ultrasonography to non-invasively
determine muscle fascicle movement during muscle
contraction has provided a method for studying hu-
man aponeurosis and tendon tissue behavior during
loading in vivo (Fukashiro, Itoh, Ichinose, Kawaka-
mi, Fukunaga, 1995a; Ito, Kawakami, Ichinose, Fu-
kashiro, Fukunaga, 1998; Maganaris & Paul, 1999).
Thus far, the mechanical properties of the tibialis
anterior in the leg have been examined using ultra-
sonography during voluntary (Ito et al., 1998) and
electrically induced (Maganaris & Paul, 1999) con-
tractions, with considerably different results, which
may be attributed in part to antagonist coactivation
and tendinous deformation due to joint rotation. In

contrast to the tibialis anterior, the human Achilles
tendon is subjected to substantial stresses during hu-
man locomotion. Further, the Achilles tendon is fre-
quently associated with acute and overuse injuries re-
lated to habitul loading (Kannus, Jozsa, Natri,
Järvinen, 1997), including complete tendon ruptures
(Kannus & Józsa, 1991; Maffulli, Waterston, Squair,
Reaper, Douglas, 1999). Moreover, albeit not con-
firmed in a human model, numerous studies suggest
that exercise results in increased tendon strength and
stiffness (Viidik, 1967, 1969; Woo et al. 1980, Woo,
Gomez, Woo, Akeson, 1982; Simonsen, Klitgaard,
Bojsen-Moller, 1995). Therefore, the present investi-
gation tested whether the load-displacement and
stress–strain characteristics of the human triceps su-
rae tendon and aponeurosis, in vivo, during graded
maximal voluntary plantarflexion efforts were differ-
ent in well-trained runners and age-matched non-run-
ners.
Materials and methods
Subjects
Ten healthy male subjects gave written informed consent to par-
ticipate in the study. The runners consisted of five males with a
mean (∫ SD) age, body mass and height of 34 ∫ 6 years, 72.1 ∫
4.6 kg and 1.82 ∫ 0.09 m, respectively. Inclusion citeria for the
runners were that they had performed habitual distance run-
ning (⬃ 80 km per week) for the last 5 years, and they did not
report any history of symptoms related to the Achilles tendon.
The control group consisted of five males with a mean (∫ SD)
age, body mass and height of 33 ∫ 8 years, 82.2 ∫ 4.2 kg and
1.80 ∫ 0.03 m, respectively. The control subjects had engaged in
physical activity on average not more than twice per week in
the past 10 years, and these activities excluded typical ‘‘high
load’’ activities on the Achilles tendon such as running, bad-
minton, squash, soccer, etc. The control group did not report
any history of symptoms related to the Achilles tendon. The
study was approved by local local human subject ethics com-
mittee.
Experimental set-up
The experimental design and set-up has been described in detail
elsewhere (Magnusson, Aagaard, Rosager, Dyhre-Poulsen, Kja-
er, 2001). The mechanical properties of the proximal aponeur-
osis of the left leg were examined with the ankle in the neutral
position and with the tibia at a right angle to the sole of the
foot. The subjects were seated in a rigid steel frame (Fig. 1) with
the knee fully extended and the hip flexed to 90æ. The foot
rested against an adjustable plate of steel with a mechanical
axis of rotation that corresponded to the lateral malleolus. To
register plantarflexion force (N) a strain gauge load cell was
attached between the plate and the steel frame. The backrest of
the frame was adjusted to minimize displacement between the
foot and foot plate, and between the back and backrest during
maximal plantarflexion efforts.
Experimental design
After positioning the subjects performed 3–8 forceful plan-
tarflexion efforts (gradual ramps) that served as precondition-
ing of the muscle–tendon complex. Thereafter, the subjects
completed a slow isometric force ramp by gradually increasing
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Physical activity and human tendon
Fig. 1. Measurement set-up. a) The subjects were seated in a rigid
steel frame with the knee extended and the hip flexed to 90æ. The
foot rested against an adjustable foot plate of steel with a mechan-
ical axis of rotation that corresponded to the lateral malleolus. (1)
EMG activity of the medial gastrocnemius (not shown), soleus and
dorsiflexor muscles were registered. (2) Joint angular change was
monitored with an electrical goniometer. (3) To register plantarflex-
ion force (N) a strain gauge load cell was attached between the foot
plate and the steel frame. One computer was used to sample 1–3,
while a separate computer was used to sample the ultrasound data.
The two computers were interconnected electronically to ensure that
all signal sampling was synchronous. Synchronization between the
computers was achieved with a custom-built device that provided a
visual marker on the ultrasonography image and simultaneously
initiated data sampling of force, EMG and goinometer via the A/
D converter. b) The connective tissue length (Lo) of the triceps surae
complex (free tendon and aponeurosis) was obtained by measuring
the distance from the sole of the foot to the proximal point of the
fascicle aponeurosis cross point on the ultrasonography image at
rest.
plantarflexion effort over a 10-s period while proximal aponeur-
osis displacement, plantarflexion force, electromyography sig-
nals and ankle joint angle were recorded. Thereafter, two separ-
ate 4-s maximal voluntary contractions (MVC) were performed
for ankle dorsiflexor muscle strength.
Measurement of electromyographic activity
Electromyographic (EMG) activity of the medial gastro-
cnemius, soleus and dorsiflexor muscles were registered using
bipolar Ag/AgCl surface electrodes (Medicotest, Type QN-10-
A, Ølstykke, Denmark) with a 3-cm interelectrode distance.
Custom-made amplifiers with a frequency response of 20 Hz to
10 KHz and 1 : 1 preamplifiers were used for the EMG signal
measurements. The EMG signal was full-wave rectified, inte-
grated and averaged with a time constant of 200 ms and ex-
pressed as mV (Basmajian & DeLuca, 1985). The EMG was
normalized to the maximal amplitudes during the trials. The
EMG for the dorsiflexor muscles was normalized to the maxi-
mal amplitude obtained during separate the dorsiflexion efforts
(see below).
Measurement of tendon displacement
A 7.5-MHz linear array B-mode ultrasonography probe (Sono-
line Sienna, Siemens, Erlangen, Germany) with a width and
depth resolution of 0.51 mm and 0.34 mm, respectively, was used
for imaging. When the distal aponeurosis is followed in the
91

Rosager et al.
proximal (cephalic) direction it continues underneath the
gastrocnemius muscle. To measure the proximal aponeurosis
displacement the probe was placed in the saggital plane over
the most distal (caudal) part of the medial gastrocnemius head.
The ultrasonography probe was tightly fitted into a block of
Styrofoam that was taped onto the skin of the subjects. In ac-
cordance with previous reports (Maganaris & Paul 1999), pilot
work demonstrated that the ultrasonography probe did not
shift position during muscular contraction in relation to an
echo-absorptive marker placed on the skin.
The parallel echoes that can be observed on the ultrasono-
graphy image correspond to the aponeurosis between muscle
fascicles (Fukashiro et al., 1995a; Ito et al., 1998) (Fig. 2). The
cross-point of ultrasonography echoes from a fascicle and the
aponeurosis was defined as the position where the fascicle was
affixed. The displacement of this fixation point was considered
to represent the magnitude of displacement (mm) of the apo-
neurosis. The ultrasonography image was displayed in real-time
on a monitor. The S-VHS output video signal from the ultra-
sonography apparatus was fed to a computer for data collec-
tion at a rate of 50 Hz.
Measurement of angular ankle joint motion
It has been shown that passive angular rotation about a joint
results in considerable tendon displacement (Spoor, van, van
Spoor, Titulaer, Huson, 1990), and that the relationship be-
tween tendon displacement and joint angular rotation is linear
(Fukunaga, Ito, Ichinose, Kumo, Kawakami, Fukashiro, 1996).
Therefore, should any angular joint rotation occur in the direc-
tion of plantarflexion during an ‘‘isometric’’ contraction tendon
displacement will be attributed to both angular rotation and
contractile tension. Despite considerable external fixations it is
extremely difficult to completely prevent any joint angular ro-
tation during a forceful muscular contraction, especially during
plantarflexion which can yield very large forces. Thus, angular
joint movement needs to be accounted for to avoid an over-
Fig. 2. Ultrasonography of the proximal aponeurosis of the triceps
surae. The measurement was performed along the length of apo-
neurosis from the white vertical bar to the end of the US field. Note
the shift in the displacement of the aponeurosis to the left during
the graded isometric contraction effort from rest (a) to 2000 N (b)
of tensile tendon force. The bright field shown at the bottom of (a)
is the synchronization signal that initiated data sampling of force,
EMG and goniometer via the A/D converter.
92
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
estimation of tendon displacement during an isometric contrac-
tion (Magnusson et al., 2001). Therefore, to monitor ankle joint
angular rotation an electrical goniometer (Penny and Giles,
Biometrics Ltd, Gwent, UK) was placed on the lateral aspect
of the foot. The electrical goniometer end blocks were secured
with tape over the distal part of the fifth metatarsal and the
posterolateral aspect of the fibula. The ratio of tendon displace-
ment (D mm) to angular rotation (D rad) about a joint corre-
sponds to the estimated tendon moment arm (van Spoor et al.
1990; Fukunaga, Kawakami, Kuno, Funato, Fukashiro, 1997;
Ito et al. 1998). Therefore, the product of the angular rotation
(rad) and tendon moment arm can be used to estimate tendon
displacement caused by joint rotation alone. The tendon mo-
ment arm was measured from MR images (see details below).
Thus, for each subject the tendon displacement obtained from
the ultrasonography images could be corrected for that attri-
buted to joint rotation alone.
Measurement of maximal dorsiflexion moment
In addition to the plantarflexion ramp contractions the subjects
performed two dorsiflexion maximal contraction (MVC) ef-
forts. Each effort lasted for approximately 4 s with a 1-min rest
period. Dorsiflexion was performed with the subject in a seated
position with the hips and knees in 90æ of flexion. A strap was
placed over the distal part of the metatarsals and connected to
the load cell that was secured to the floor. The foot was placed
in the neutral position (90æ) for the dorsiflexion effort. The
product of the distance from the strap to the center of rotation
(obtained from MR images, see below) and the registered force
yielded the dorsiflexion moment (N¡m). The force and moment
generated due to dorsiflexor muscles coactivation during the
plantarflexion efforts was estimated assuming a linear relation-
ship between EMG amplitude of the dorsiflexor muscles and
tension (Lippold, 1952).
Data acquisition
The goniometer, EMG, force and ultrasonography image sig-
nals were recorded continuously immediately prior to and dur-
ing the 10-s plantarflexion effort. The goniometer, EMG and
force signals were sampled at 50 Hz using an A/D converter
(DT 2801 A, Data Translation), and stored on a computer for
subsequent analysis. The ultrasonography images were simul-
taneously and continuously sampled at 50 Hz on a separate
computer. The two computers were interconnected to ensure
that all signal sampling was synchronous. Synchronization be-
tween the two computers was achieved with a custom-built de-
vice that provided a visual marker on the ultrasonography im-
age and simultaneously initiated data sampling of force, EMG
and goniometer via the A/D converter.
Calculation of moment arm
Saggital plane MR images (GE Signa horizon LX 1.5T, T2
weighted FSE (TR/TE: 4000/88; FOV 12; matrix 256 ¿ 192,
slice thickness 3 mm) were obtained with the ankle in the neu-
tral position (90æ) to estimate the Achilles tendon moment arm
using a modified Reuleaux method (Voigt, 1994). Briefly, the
method assumes that the joint surface of the talus is circular
and by using geometric rules obtains a center of rotation at a
specific point distal to the joint surface of the talus. The tendon
moment arm was obtained by measuring the perpendicular dis-
tance from Achilles tendon to the center of rotation. The meas-
urement of the moment arm from the MR images was per-
formed three times for each subject and a mean was used as an
estimate for the moment arm. The mean coefficient of variation
for repeated measures across subjects has been shown to be
1.9% (Magnusson et al., 2001).

Calculation of tensile tendon force
Tendon (and aponeurosis) force of the triceps surae during the
plantarflexion effort was calculated by dividing the externally
measured moment by the tendon moment arm. Separately, the
contribution from the dorsiflexor moment was added to the
externally measured moment to compensate for coactivation
(Magnusson et al., 2001).
Measurement of tendon cross-sectional area and calculation of
tendon stress
Tendon cross-sectional area was obtained from MR images (T1
weighted SE, TR/TE: 400/15; FOV 12; matrix 512 ¿ 512, slice
thickness 6 mm). The cross-sectional area of the Achilles tendon
was measured 3 cm proximal to the Achilles tendon insertion
onto the calcaneus, which corresponds to the narrowest portion
of the Achilles tendon (Voigt, 1994) (Fig. 3). The measurement
was performed three times for each subject and the mean was
used as the cross-sectional area. The mean coefficient of vari-
ation for repeated measures across subjects has been shown to
be 5.8% (Magnusson et al., 2001).
Measurement of tendon length and calculation of tendon strain
For the purpose of this study the connective tissue length (Lo)
of the triceps surae complex (free tendon and aponeurosis) was
obtained by measuring the distance from the sole of the foot to
the point of the fascicle aponeurosis cross point on the ultra-
sonography image at rest (see Fig. 2). Strain (ä) was calculated
by dividing aponeurosis cross-point displacement (mm) during
plantarflexion by Lo.
Calculation of tendon stiffness and Young’s modulus
A quadratic fit was applied to the load-displacement data for
each person. Then the mechanical stiffness (D?F/D?mm) was cal-
culated from the load-displacement relationship in the final 10%
of the force range, i.e. from 90 to 100% (Magnusson et al., 2001)
Fig. 3. Transverse magnetic resonance imaging of one subject. The
cross-sectional area of the Achilles tendon was measured 3 cm proxi-
mal to the Achilles tendon insertion onto the calcaneus.
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Physical activity and human tendon
Fig. 4. The load-displacement data for one subject.
(Figs 4 and 5). Stiffness was calculated with correction for ankle
rotation and coactivation of dorsiflexor and plantarflexor
muscles (Magnusson et al., 2001). Similarly, the Young’s modu-
lus (MPa) of the Achilles tendon and aponeurosis was estimated
by dividing the stress (s, N/m2) by the strain (ä) obtained at the
level at the gastrocnemius in the final 10% of stresses (90–100%).
Data reduction and statistics
All signals were considered at baseline with the subject com-
pletely relaxed in the experimental set-up. The ultrasonography
images were identified and analyzed at tendon force intervals of
200 N. The displacement for each ultrasonography image was
measured three times and the mean was used for analysis. We
have shown previously that analysis of two separate ramp con-
tractions yield the same results (Magnusson et al., 2001). There-
fore, only one ramp contraction was analyzed. Non-parametric
Mann–Whitney tests were used to examine if differences existed
between the runners and non-runners. An alpha level of 0.05
was considered significant. Results are reported as population
means ∫ SEM (standard error of the mean).
Results
The tendon moment arm was similar for the non-
runners (58.3 ∫ 0.2 mm) and runners (55.1 ∫ 0.1 mm).
However, there was a significant difference in the
cross-sectional area of the Achilles tendon between
the non-runners and runners (P ⬍ 0.01, Fig. 6). When
coactivation was accounted for there was a significant
increase in calculated tensile tendon force (P ⬍ 0.01).
The calculated tensile tendon force for non-runners
was 2473 ∫ 361 N, which corresponded to 97 ∫ 1% of
the estimated tensile tendon force when tibialis an-
terior muscle coactivation was accounted for (2556 ∫
401 N). The tensile tendon force for the runners was
2524 ∫ 484 N, which corresponded to 95 ∫ 4% of esti-
mated force when tibialis anterior coactivation was
accounted for (2633 ∫ 465 N).
93

Rosager et al.
Fig. 5. The mean load-displacement data for the runners and non-
runners. The mechanical stiffness (DF/Dmm) was calculated from
the load-displacement relationship in the final 10% of the force
range, i.e. from 90 to 100% based on the quadratic fit.
Fig. 6. Group mean ∫ SEM. The cross-sectional area of the Achilles
tendon. Runners had significantly greater cross-sectional area com-
pared to non-runners, P ⬍0.01.
The angular joint rotation during the graded plan-
tarflexion effort was 3.4 ∫ 0.4æ in the direction of
plantarflexion for the non-runners and 3.2 ∫ 0.9æ for
the runners. The measured maximal aponeurosis dis-
placement was 14.5 ∫ 2.1 mm for the non-runners and
13.1 ∫ 2.5 mm for the runners. When the maximal
tendon and aponeurosis displacement was corrected
for joint rotation, there was a significant decrease in
displacement (runners, 10.2 ∫ 1.9 mm; non-runners,
10.8 ∫ 1.9 mm) (P ⬍ 0.01). The average corrected
load-displacement data for runners and non-runners
are shown in Fig. 5. The joint rotation and coacti-
vation corrected load-displacement curves yielded a
significant quadratic relationship (non-runners, R2 Ω
0.98 ∫ 0.01; runners, R2 Ω 0.96 ∫ 0.02) (P ⬍ 0.01). The
was no difference in the final 10% stiffness of the
94
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
force range between the groups (runners; 306 ∫ 61 N
mmª1, non-runners; 319 ∫ 42 N mmª1, Fig. 5).
The average stress–strain characteristics for runners
and non-runners is shown in Fig. 7 for tensile tendon
forces common to all subjects within the respective
group. There were no differences in the stress–strain
characteristics between the groups. For non-runners,
the displacement and the cross-sectional area of the
Achilles tendon yielded an estimated maximal strain
and stress of 4.9 ∫ 0.8% and 38.2 ∫ 9.8 MPa, respec-
tively, with a resulting Young’s modulus of 1017 ∫ 159
MPa. For runners, the respective strain, stress and
Young’s modulus were 4.1 ∫ 0.8%, 26.3 ∫ 5.1 MPa
and 728 ∫ 119 MPa.
There was no difference in the maximal isometric
plantarflexion moment about the ankle joint between
the runners (138 ∫ 27 Nm) and non-runners (142 ∫
17 Nm). For the runners the corresponding EMG ac-
tivity was 252 ∫ 38 mV for the gastrocnemius muscle
and 124 ∫ 36 mV for the soleus muscle. For the non-
runners the corresponding EMG activity was 301 ∫
143 mV and 72 ∫ 13 mV, respectively. There was a sig-
nificant difference in the maximal isometric dorsi-
flexion moment about the ankle joint between the
runners (31 ∫ 3 Nm) and non-runners (43 ∫ 3 Nm)
(P ⬍ 0.01). The corresponding EMG activity of the
tibialis muscle was 298 ∫ 48 mV in runners and 348 ∫
58 mV in non-runners. During the plantarflexion ef-
fort the coactivation level of the tibialis anterior
muscle was 13 ∫ 5% in the controls and 21 ∫ 13% for
the runners.
Fig. 7. Group mean ∫ SEM. The stress–strain characteristics for
the triceps surae aponeurosis and tendon in runners and non-run-
ners. There was no difference between runners and non-runners.
Note that in contrast to the similar Achilles tensile tendon forces in
the two groups, the greater cross-sectional area of the runners have
served to lower the average stress on the tendon.

Discussion
The main finding of the present investigation was that
the Achilles tendon cross-sectional area was markedly
larger in runners than non-runners, which suggests
that the chronic exposure of the Achilles tendon to
repetitive loading has resulted in significant tissue ad-
aptation. In contrast, the load-displacement and
stress–strain characteristics of the human triceps su-
rae aponeurosis and tendon, in vivo, during graded
maximal voluntary plantarflexion efforts did not dif-
fer in well-trained runners and age-matched non-run-
ners. From a methodological standpoint the data
confirmed previous reports (Magnusson et al., 2001),
that accounting for minor joint rotation with respect
to tendinous displacement, and small magnitudes of
antagonist coactivation with respect to tensile tendon
force, both significantly affect the results obtained.
The Achilles tendon tensile force in the present
study was on average 2556 N for non-runners and
2633 for runners. To the best of our knowledge no
data exist on triceps surae aponeurosis and tendon
displacement in well-trained runners and non-run-
ners. Therefore, it is relevant to address if the load-
displacement data obtained during the isometric
plantarflexion effort in the present study are obtained
under loading magnitudes comparable to those dur-
ing functional activities. It has been estimated that
the human Achilles tendon tensile force may reach up
to 2600 N during walking (Giddings et al., 2000) and
3100–5330 N during running (Scott & Winter, 1990;
Giddings et al., 2000). Others (Finni et al., 1998)
have measured Achilles tensile tendon forces of 1430
N during walking, and tendon forces of 1895–3786 N
during various jumping activities (Fukashiro, Komi,
Jarvinen, Miyashita, 1995b). We have previously re-
ported Achilles tensile tendon forces of 3255 N using
the same measurement technique (Magnusson et al.,
2001). In the present study the tendon force during
the maximal isometric plantarflexion effort was
⬃2600 N, which suggests that the triceps surae dis-
placement data was obtained at tendon loading com-
parable, in fact, to those imposed on the Achilles ten-
don during human locomotion and jumping.
Although there are reports of exercise-induced ten-
don hypertrophy in animal models (Woo et al., 1980;
Birch, McLaughlin, Smith, Goodship, 1999) it lacks
confirmation in humans. Birch et al. (1999) examined
the tendon cross-sectional area using ultrasono-
graphy in horses subjected to high intensity and low
intensity exercise over a short- (5 months) or long-
term (18 months) training period. The short-term
training resulted in tendon-specific hypertrophy due
to high intensity training compared to low intensity
which, curiously enough, was not observed in the
long term. Woo et al. (1980) also demonstrated ten-
don hypertrophy in swine after 12 months of tread-
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Physical activity and human tendon
mill running with an average increase in cross-sec-
tional area of 21% for the lateral extensors. Further,
in animal models an age-related increase in cross-sec-
tional area has been shown (Birch et al., 1999; Naka-
gawa, Hayashi, Yamamoto, Nagashima, 1996). Using
ultrasonography, Kallinen & Suominen (1994) exam-
ined the Achilles tendon in a group of elderly (⬎ 70
years) endurance and power-sport athletes compared
to age-matched controls. Although the accuracy of
the technique was questioned, the mean width, but
not the cross-sectional area, was greater in the ath-
letes. Using high-resolution MRI (see Fig. 3) the re-
sults of the present study showed that runners who
had trained extensively for a long period of time (⬎
5 years) had markedly greater (22%) Achilles tendon
cross-sectional area than that of the age-matched
controls. It was shown recently that the Achilles ten-
don cross-sectional area was related positively to
body weight (Beyer, Abrahamsen, Beergaard, Kjaer,
Aagaard, Magnusson, 2001). In the present study the
subjects were age- but not weight-matched. However,
despite the lesser body weight of the runners they dis-
played a greater cross-sectional area than non-run-
ners, which suggests that the training stimulus con-
tributed more to the tendon hypertrophy than body
weight alone.
Within the confines of a cross-sectional obser-
vation, the greater cross-sectional area in runners of
the present study suggests that the repetitive tendon
loading associated with running has resulted in a hy-
pertrophic adaptation of the Achilles tendon. It has
been shown that tendon stiffness is related positively
to maximal muscle force production (Scott & Loeb,
1995), such that stronger muscles have greater stiff-
ness. However, the lack of a difference in maximal
plantarflexion moment and tensile tendon force be-
tween runners and non-runners suggest that, while
running does not increase planterflexion strength, the
repetitive loading is an adequate stimulus to induce
tendon hypertrophy. In contrast to other tissue, ten-
don is generally considered to be metabolically slow
(Prockop, Kivirikko, Tuderman, Guzman, 1979), and
data on human tendon metabolism are sparse. How-
ever, it was shown recently in a human model that
type I collagen synthesis in the peritendinous tissue of
the Achilles tendon, as measured with a microdialysis
technique, was elevated immediately and in the sub-
sequent days after endurance running (Langberg,
Skovgaard, Petersen, Bulow, Kjaer, 1999; Langberg,
Skovgaard, Asp, Kjaer, 2000). Furthermore, a 12-
week training study showed that collagen synthesis
was elevated while degradation remained unchanged,
indicating that physical training resulted in a net syn-
thesis of tendon-related collagen (Langberg, Rosend-
al, Kjaer, 2001). It is posssible that such an increase
in net synthesis may account for the larger cross-sec-
tional area in the runners in the present study.
95

Rosager et al.
Investigations of human tendon behavior have been
confined mainly to biomechanical testing of isolated
cadaver tissue specimens which may be influenced by
a variety of factors, including storage method, fix-
ation technique, strain rate, actual tendon, donor age
and whether the loading is due to failure under
physiological conditions. Testing of isolated tissue
specimens have reported stress, strain and Young’s
modulus to range from 32 to 111 MPa, 2–12% and
421–1724 MPa, respectively, for various isolated hu-
man tendons specimens (Gratz, 1931; VanBrocklin &
Ellis, 1965; Benedict, Walker, Harris, 1968; Blan-
ton & Biggs, 1970; Butler et al., 1984; Loren & Lieb-
er, 1995; McGough et al., 1996; Schechtman & Bad-
er, 1997). In vivo measures of stiffness for the tibialis
anterior aponeurosis and tendon has been reported
to be 32–161 N mmª1 with a Young’s modulus of
530–1200 MPa (Ito et al., 1998; Maganaris et al.,
1999). The stiffness and Young’s modulus for the
combined human triceps surae aponeurosis and ten-
don were reported recently to be 467 N mmª1 and
1474 MPa (Magnusson et al., 2001). The stiffness
(306–319 N mmª1) and modulus (728–1017 MPa) in
the present study were somewhat lower than reported
previously (Magnusson et al., 2001), which may be
due to the lower force production (plantarflexion
strength) in the present study. Nevertheless, the
values resemble those obtained for several mam-
malian (Bennett, Ker, Dimery, Alexander, 1986) and
human tendons (Butler et al., 1984). Using an ultra-
sonography method Kubo et al. (2000) showed that
well-trained runners had greater stiffness of the ten-
don and aponeurosis of the vastus lateralis than that
of untrained subjects. In contrast, the present study
could not show a difference in the stiffness and
modulus of the triceps surae tendon and aponeurosis
in the runners and non-runners. It is possible that the
discrepancy is related to the different muscle–tendon
units studied. Theoretically, an increase in tendon
thickness would yield a greater stiffness for a given
deformation. Although there was a marked difference
in cross-sectional area there was no difference in the
load-deformation data (Fig. 5). However, at the same
time it cannot be ruled out that the training simul-
taneously resulted in altered intrinsic properties, such
as increased collagen cross-linkage. It should be
noted that the lack of difference between the groups
may be due to the considerable variation in the ten-
don properties in the present study, which has also
been noted in several investigations (Viidik, 1967,
1969; Birch et al., 1999; Kubo, Kawakami, Fukuna-
ga, 1999; Maganaris & Paul, 1999; Magnusson et al.,
2001). For example, in the present study the ranges
of stiffness and stress values were 174–530 N mmª1
and 19–46 MPa for the runners, and 221–369 N mmª1
and 22–71 MPa for non-runners. Such seemingly
large intersubject variations within groups of the
96
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
same gender, and similar age, height, weight and
loading history suggest that genetics influence tendon
properties. Therefore, the effect of training on the
mechanical properties of tendon may be better ad-
dressed in within-subject design studies, i.e. longitudi-
nal training studies.
It should be noted that the deformation and strain
values in the present study represent those of the
combined aponeurosis and free tendon. Whether the
free tendon and the aponeurosis have similar mech-
anical properties has not been established firmly.
There are animal data to suggest that the aponeurosis
and free tendon have similar (Trestik & Lieber, 1993;
Scott & Loeb, 1995) and dissimilar mechanical prop-
erties (Lieber, Leonard, Brown, Trestik, 1991; Delga-
do-Lezama, Raya, Munoz-Martinez, 1997). Based on
human in vivo data it has been suggested that the ten-
don of the tibialis anterior undergoes less strain than
that of the aponeurosis (Maganaris & Paul 2000). To
what extent the human free tendon and the aponeur-
osis of the triceps surae in vivo have similar mechan-
ical properties remains to be established.
The failure stress for tendon is generally considered
to be approximately 100 MPa (Ker et al., 1988). The
safety factor, i.e. the stress during high loading activ-
ity divided by the failure stress, is about 8 for the
majority of tendons (Ker et al., 1988), but it may be
as low as 1–2 (Biewener & Roberts, 2000). In the
present study the safety factor was 4 for the runners
and 2.6 for the non-runners, assuming a failure stress
of 100 MPa. Thus, it is conceivable that the Achilles
tendon approaches the safety limit during typical ec-
centric activities that results in greater tendon strain
and stress. It was recently shown that tendon sub-
jected to a similar maximal tensile tendon forces in
vivo had markedly different tendon stress (50%) due
to dissimilar tendon cross-sectional area (Magnusson
et al., 2001). In the present study it can be seen that
despite similar tensile tendon forces (Fig. 5) there is a
reduction in tendon stress due to the greater tendon
cross-sectional area in the runners compared to the
non-runners (Fig. 7). Such a reduction in stress may
provide for a greater ‘‘safety factor’’ for tendon in-
jury. Interestingly, in animal models it has been
shown that aging is associated with an increase in
tendon cross-sectional area (Nakagawa et al., 1996;
Birch et al., 1999). Furthermore, it was shown re-
cently in humans that the Achilles tendon undergoes
hypertrophy with aging (Beyer et al., 2001). In the
present study the effect of exercise appeared to yield
a tendon hypertrophy response. It remains unknown
if the combined effect of an age (⬍ 50 year) and exercise
related tendon hypertrophy serve to lower the stress on
the tendon and thereby the injury risk. At the same
time it should be noted that the cross-sectional design
and the limited sample size present inherent limi-
tations. The process of natural selection cannot be ex-

cluded. That is, runners have a large Achilles tendon
cross-sectional area prior to exposing the tendon to
chronic loading exposure. Thereby they have reduced
stress on the tendon and diminshed risk of injury.
Perspectives
Tendon pathology represents a considerable portion
of all soft tissue injuries in sports medicine. However,
relatively little is known with respect to human ten-
don adaptation to physical activity. Mechanical prop-
erties of human tendinous tissue have been limited
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Physical activity and human tendon
previously to cadaver tissue specimens; however, re-
cent advances in ultrasonography make it possible to
view muscle–tendon tissue in ‘‘real-time’’ during
movement and contractions. The present technique,
which accounts for small amounts tendinous dis-
placement due to joint rotation and small amounts of
antagonist cocontraction, allows for for future non-
invasive investigations into the effect of human ten-
donous adaptation to various forms of physical activ-
ities and interventions.
Key words: Achilles tendon; adaption; deformation;
strain; stress.

References

Alexander RM, Vernon A. The mech-
anics of hopping by kangaroos (Macro-
podidae). J Zool 1975: 177: 265–303.
Amis A, Prochazka A, Short D, Trend PS,
Ward A. Relative displacements in
muscle and tendon during human arm
movements. J Physiol (Lond) 1987:
389: 37–44.
Basmajian JV, DeLuca CJ. Muscles alive.
Their functions revealed by electromy-
ography, 5th edn. Baltimore, MD: Wil-
liams & Wilkins, 1985: 94–97.
Benedict JV, Walker LB, Harris EH.
Stress–strain characteristics and tensile
strength of unembalmed human ten-
don. J Biomech 1968: 1: 53–63.
Bennett MB, Ker RF, Dimery NJ, Alex-
ander RM. Mechanical properties of
various mammalian tendons. J Zool
1986: 209: 537–548.
Beyer N, Abrahamsen K, Neergaard K,
Kjaer M, Aagaard P, Magnusson SP.
Cross-sectional area of the Achilles ten-
don in young and elderly women. Med
Sci Sports Exerc 2001: 33: S126.
Biewener AA, Roberts TJ. Muscle and
tendon contributions to force, work,
and elastic energy savings: a compara-
tive perspective. Exerc Sport Sci Rev
2000: 28: 99–107.
Birch HL, McLaughlin L, Smith RK,
Goodship AE. Treadmill exercise-in-
duced tendon hypertrophy. assessment
of tendons with different mechanical
functions. Equine Vet J Suppl 1999: 30:
222–226.
Blanton PL, Biggs NL. Ultimate tensile
strength of fetal and adult human ten-
dons. J Biomech 1970: 3: 181–189.
Butler DL, Grood ES, Noyes FR, Zer-
nicke RF, Brackett K. Effects of struc-
ture and strain measurement technique
on the material properties of young hu-
man tendons and fascia. J Biomech
1984: 17: 579–596.
Delgado-Lezama R, Raya JG, Munoz-
Martinez EJ. Methods to find aponeur-
osis and tendon stiffness and the onset
of muscle contraction. J Neurosci Meth
1997: 78: 125–132.
Fellows SJ, Rack PM. Changes in the
length of the human biceps brachii
muscle during elbow movements. J
Physiol (Lond) 1987: 383: 405–412.
Finni T, Komi PV, Lukkariniemi J. Achil-
les tendon loading during walking: ap-
plication of a novel optic fiber tech-
nique. Eur J Appl Physiol 1998: 77:
289–291.
Fukashiro S, Ito M, Ichinose Y, Kawaka-
mi Y, Fukunaga T. Ultrasonography
gives directly but noninvasively elastic
characteristic of human tendon in vivo.
Eur J Appl Physiol 1995a: 71: 555–557.
Fukashiro S, Komi PV, Järvinen M, Mi-
yashita M. In vivo Achilles tendon
loading during jumping in humans.
Eur J Appl Physiol 1995b: 71: 453–458.
Fukunaga T, Ito M, Ichinose Y, Kuno S,
Kawakami Y, Fukashiro S. Tendinous
movement of a human muscle during
voluntary contractions determined by
real-time ultrasonography. J Appl Phy-
siol 1996: 81: 1430–1433.
Fukunaga T, Kawakami Y, Kuno S, Fun-
ato K, Fukashiro S. Muscle architec-
ture and function in humans. J Bio-
mech 1997: 30: 457–463.
Giddings VL, Beaupre GS, Whalen RT,
Carter DR. Calcaneal loading during
walking and running. Med Sci Sports
Exerc 2000: 32: 627–634.
Goldstein SA, Armstrong TJ, Chaffin
DB, Matthews LS. Analysis of cumula-
tive strain in tendons and tendon
sheaths. J Biomech 1987: 20: 1–6.
Gratz M. Tensile strenght and elasticity
tests on human fascia lata. J Bone Joint
Surg [Br ] 1931: 13: 334–340.
Ito M, Kawakami Y, Ichinose Y, Fuka-
shiro S, Fukunaga T. Nonisometric be-
havior of fascicles during isometric
contractions of a human muscle. J
Appl Physiol 1998: 85: 1230–1235.
Kallinen M, Suominen H. Ultrasono-
graphic measurements of the Achilles
tendon in elderly athletes and seden-
tary men. Acta Radiol 1994: 35: 560–
563.
Kannus P, Józsa L. Histopathological
changes preceding spontaneous rupture
of a tendon. A controlled study of 891
patients. J Bone Joint Surg Am 1991:
73: 1507–1525.
Kannus P, Józsa L, Natri A, Järvinen M.
Effects of training, immobilization and
remobilization on tendons. Scand J
Med Sci Sports 1997: 7: 67–71.
Ker RF, Alexander RM, Bennett MB.
Why are mammalian tendons so thick.
J Zool 1988: 216: 309–324.
Kubo K, Kanehisa H, Kawakami Y, Fu-
kunaga T. Elastic properties of muscle–
tendon complex in long-distance run-
ners. Eur J Appl Physiol 2000: 81: 181–
187.
Kubo K, Kawakami Y, Fukunaga T. In-
fluence of elastic properties of tendon
structures on jump performance in
humans. J Appl Physiol 1999: 87:
2090–2096.
Kvist M. Achilles tendon injuries in ath-
letes. Sports Med 1994: 18: 173–201.
Langberg H, Rosendal L, Kjaer M.
Training induced changes in peritendi-
nous type I collagen turnover deter-
mined by microdialysis in humans. J
Physiol (Lond) 2001: 297–302.
Langberg H, Skovgaard D, Asp S, Kjaer
M. Time pattern of exercise-induced
changes in type I collagen turnover
after prolonged endurance exercise in
humans. Calcif Tissue Int 2000: 67: 41–
44.

97

Rosager et al.
Langberg H, Skovgaard D, Petersen LJ,
Bulow J, Kjaer M. Type I collagen syn-
thesis and degradation in peritendinous
tissue after exercise determined by
microdialysis in humans. J Physiol
1999: 521: 299–306.
Lieber RL, Leonard ME, Brown CG, Tre-
stik CL. Frog semitendinosis tendon
load-strain and stress–strain properties
during passive loading. Am J Physiol
1991: 261: 86–92.
Lippold OC. The relationship between in-
tegrated action potentials in a human
muscle and its isometric tension. J Phy-
siol (Lond) 1952: 177: 492–499.
Loren GJ, Lieber RL. Tendon bio-
mechanical properties enhance human
wrist muscle specialization. J Biomech
1995: 28: 791–799.
Maffulli N, Waterston SW, Squair J, Re-
aper J, Douglas AS. Changing inci-
dence of Achilles tendon rupture in
Scotland: a 15-year study. Clin J Sport
Med 1999: 9: 157–160.
Maganaris CN, Paul JP. In vivo human
tendon mechanical properties. J Physi-
ol (Lond) 1999: 521: 307–313.
Maganaris CN, Paul JP. Load-elongation
characteristics of in vivo human tendon
and aponeurosis. J Exp Biol 2000: 203:
751–756.
Magnusson SP, Aagaard P, Rosager S,
Dyhre-Pouelsen P, Kjaer M. Load-dis-
placement properties of the human tri-
ceps surae aponeurosis in vivo. J Physi-
ol (Lond) 2001: 531: 277–288.
98
McGough RL, Debski RE, Taskiran E,
Fu FH, Woo SL. Mechanical prop-
erties of the long head of the biceps
tendon. Knee Surg Sports Traumatol
Arthrose 1996: 3: 226–229.
Nakagawa Y, Hayashi K, Yamamoto N,
Nagashima K. Age-related changes in
biomechanical properties of the Achil-
les tendon in rabbits. Eur J Appl Physi-
ol 1996: 73: 7–10.
Prockop DJ, Kivirikko KI, Tuderman L,
Guzman NA. The biosynthesis of col-
lagen and its disorders (second of two
parts). N Engl J Med 1979: 301: 77–85.
Rack PM, Ross HF. The tendon of flexor
pollicis longus: its effects on the muscu-
lar control of force and position at the
human thumb. J Physiol (Lond) 1984:
351: 99–110.
Schechtman H, Bader DL. In vitro fatigue
of human tendons. J Biomech 1997: 30:
829–835.
Scott SH, Loeb GE. Mechanical prop-
erties of aponeurosis and tendon of the
cat soleus muscle during whole-muscle
isometric contractions. J Morph 1995:
224: 73–86.
Scott SH, Winter DA. Internal forces of
chronic running injury sites. Med Sci
Sports Exerc 1990: 22: 357–369.
Simonsen EB, Klitgaard H, Bojsen-Moll-
er F. The influence of strength training,
swim training and ageing on the Achil-
les tendon and m. soleus of the rat. J
Sports Sci 1995: 13: 291–295.
16000838, 2002, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1034/j.1600-0838.2002.120205.x by Central Library Of Semmelweis, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
van Spoor CWLJ, Meskers CG, Titulaer
AF, Huson A. Estimation of instan-
taneous moment arms of lower-leg
muscles. J Biomech 1990: 23: 1247–
1259.
Trestik CL, Lieber RL. Relationship be-
tween Achilles tendon mechanical
properties and gastrocnemius muscle
function. J Biomech Eng 1993: 115:
225–230.
VanBrocklin JD, Ellis DG. A study of the
mechanical behavior of toe extensor
tendons under applied stress. Arch
Phys Med Rehabil 1965: 46: 369–373.
Viidik A. Tensile strength properties of
Achilles tendon systems in trained and
untrained rabbits. Acta Orthop Scand
1969: 40: 261–272.
Viidik A. The effect of training on the
tensile strength of isolated rabbit ten-
dons. Scand J Plastic Reconstr Surg
1967: 1: 141–147.
Voigt M. Muscle-tendon elasticity in vivo
and its significance for human per-
formance. Thesis. Denmark: Univer-
sity of Copenhagen, 1994.
Woo SL, Gomez MA, Woo YK, Akeson
WH. Mechanical properties of tendon
and ligaments. The relationship of im-
mobilization and exercise on tissue re-
modeling. Biorheol 1982: 19: 397–408.
Woo SL, Ritter MA, Amiel D, et al. The
biomechanical and biochemical prop-
erties of swine tendons: long term ef-
fects of exercise on the digital exten-
sors. Connect Tissue Res 1980: 7: 177–
183.