Journal of Applied Biomechanics, 2015, 31, 69 -78
Human Kinetics
http://dx.doi.org/10.1123/JAB.2013-0284
© 2015 Human Kinetics, Inc.
ORIGINAL RESEARCH
Eccentric and Concentric Exercise of the Triceps Surae: An In Vivo
Study of Dynamic Muscle and Tendon Biomechanical Parameters
Saira Chaudhry, Dylan Morrissey, Roger C. Woledge, Dan L. Bader, and Hazel R.C. Screen
Queen Mary University of London
Triceps surae eccentric exercise is more effective than concentric exercise for treating Achilles tendinopathy, however the mechanisms underpinning these effects are unclear. This study compared the biomechanical characteristics of eccentric and concentric
exercises to identify differences in the tendon load response. Eleven healthy volunteers performed eccentric and concentric
exercises on a force plate, with ultrasonography, motion tracking, and EMG applied to measure Achilles tendon force, lower
limb movement, and leg muscle activation. Tendon length was ultrasonographically tracked and quantified using a novel algorithm. The Fourier transform of the ground reaction force was also calculated to investigate for tremor, or perturbations. Tendon
stiffness and extension did not vary between exercise types (P = .43). However, tendon perturbations were significantly higher
during eccentric than concentric exercises (25%–40% higher, P = .02). Furthermore, perturbations during eccentric exercises
were found to be negatively correlated with the tendon stiffness (R2 = .59). The particular efficacy of eccentric exercise does not
appear to result from variation in tendon stiffness or extension within a given session. However, varied perturbation magnitude
may have a role in mediating the observed clinical effects. This property is subject-specific, with the source and clinical timecourse of such perturbations requiring further research.
Keywords: Achilles tendon, force, extension, biomechanics, human, in vivo
Achilles tendinopathy is difficult to treat1 and prone to recurrence,2–5 with current treatments achieving only partial success.
Available data indicate that one of the most effective treatments for
Achilles tendinopathy is eccentric exercise of the triceps surae, with
a success rate of 50%–60% in sedentary patients5 and 60%–80%
in athletes.6–8 Controlled trials of this method report superior
outcomes with eccentric exercise (eccentric) when compared with
surgical interventions6 or concentric exercise (concentric) regimes.7
However, the mechanisms underpinning the observed effects remain
unknown.
Initially it was believed the tendon experienced higher forces
during eccentric than concentric, resulting in more tendon remodeling in response to higher strain.9 However, in vivo studies in healthy
humans have reported no difference in peak Achilles tendon force
during eccentric and concentric,10 instead reporting that the primary
difference between eccentric and concentric is 8–12 Hz tendon
perturbations during eccentric.11 However, these studies only considered tendon force or length changes in isolation, and to date no one
has investigated stiffness across the complete dynamic force-length
curve of the Achilles tendon during eccentric and concentric, nor
how stiffness may relate to tendon perturbation behavior. Establishing relationships between tendon stiffness and perturbations is
Saira Chaudhry is with the School of Engineering and Material Science and
the Centre for Sports and Exercise Medicine, both at Queen Mary University
of London, UK. Dylan Morrissey and Roger C. Woledge are with the Centre
for Sports and Exercise Medicine, Queen Mary University of London, UK.
Dan L. Bader and Hazel R.C. Screen are with the School of Engineering
and Material Science, Queen Mary University of London, UK. Address
author correspondence to Hazel R.C. Screen at h.r.c.screen@qmul.ac.uk.
an essential step toward understanding the mechanical effects of
exercise within tendon. The frequency of these perturbations and
their effects may vary depending on the stiffness of the tissue.12,13
To measure tendon length changes during exercise, typically,
manual marking of the muscle-tendon junction (MTJ) in successive
ultrasound frames has been adopted.14,15 However, manual analysis of MTJ location is labor intensive; a 15 second video at 17 Hz
consists of 255 frames, so tracking is generally limited to feature
monitoring in 1–2 frames during an exercise cycle. In addition,
while the human eye appears very good at image correlation, the
accuracy and reproducibility of this approach cannot be guaranteed,
and manual methods fail if movements are large and key features
move out of view. There have been a number of attempts to develop
algorithms to analyze MTJ movement automatically,16–20 typically
based on cross-correlation or Lucas-Kanade feature tracking, and
successfully tested on relatively small movements. However, the
inherent limitation of these methods, that the same feature of interest must be present in similar form in adjacent frames of interest,
persists and limits their application in the current study, as the MTJ
will sometimes move off the screen. An automatic method suitable
for large movements with intermittent key feature visibility was
developed to address these requirements.
The aim of the current study was to investigate how tendon
mechanical properties and tendon perturbations compared between
eccentric and concentric exercise cycles. We hypothesized that
tendon mechanical properties are the same during eccentric and
concentric throughout the exercise cycle. Further, we hypothesized
that the perturbations seen during eccentric correlate negatively with
tendon stiffness. To test these hypotheses, this study additionally
describes the development and testing of a novel semiautomated
high-strain tracking analysis procedure for the MTJ.
69
70 Chaudhry et al.
Methods
Subjects
The study was approved by the ethics and research committee of
Queen Mary University of London. Eleven healthy volunteers (6
male, 5 female; mean [SD] age 26.5 years [1.9]; body mass 65.92 kg
[10.5]; height 1.73 m [0.08]) were recruited, giving written consent.
Subjects were moderately active, defined as 2–3 hours of physical
activity a week, but had not participated in any organized program
of competitive sport over the last year and had no history of tendon
injury or any systemic disease. Subjects with current or previous
Achilles tendon pain, pathology, or surgery were excluded.
Exercise Protocol
For performing exercises, each subject was asked to stand on a step
placed on a force plate (Kistler force platform 9281B, Kistler Instruments, Winterthur, Switzerland). For eccentric, subjects started on
the ball of the foot of the right leg with the heel raised, lowering the
heel in a controlled manner (Figure 1, right to left). The exercise
was performed off the edge of the step to allow full dorsiflexion to
be reached. For concentric, subjects started with the heel below the
toes and raised the heel in a similar controlled manner (Figure 1, left
to right). Both exercises were performed at 0.5 rad∙s–1 (~3 s exercise
phase), guided by a metronome. After completing either exercise,
subjects used their other leg and a second step on an adjacent force
plate to assist in returning to the starting position before repeating
the exercise. A single data collection consisted of two cycles of
either concentric or eccentric performed consecutively. Three sets
of data were recorded for each exercise paradigm, for each subject,
in a randomized order. Time was allowed to perform familiarization
exercises before data collection, with these exercises also providing
preconditioning of the triceps surae muscle-tendon unit to ensure
minimal variation in the load-deformation curves.21 Tendon forceextension data were collected in conjunction with electromyography
(EMG) and tendon perturbations as outlined below.
Measurement of Tendon Force
Tendon force was calculated from torque around the ankle, using
inverse dynamics. The 3D ground reaction force (GrF) was captured
using a 600 × 400 mm2 force plate at 1000 Hz. An active infrared
Figure 1 — Schematic highlighting the range of motion adopted during eccentric (EL) and concentric (CL), from full plantar flexion to full dorsiflexion.
Each exercise was performed on a wooden box, placed on a force plate embedded in the ground. Electromyography (EMG) was recorded by electrodes
placed on the calf muscle (denoted with x) and joint motion was tracked by placing CODA markers on the leg joints (denoted by •). An ultrasound probe
tracked the muscle-tendon junction; markers were placed on the probe to monitor its movement.
JAB Vol. 31, No. 2, 2015
Eccentric and Concentric Exercise of Triceps Surae 71
motion analysis system (CODA, CX1, Charnwood Dynamics, Rothley, UK) with four cameras was used to determine the moments
around the ankle and knee joint by placing markers on the lateral
and medial malleoli, metatarsal 1 and 5, calcaneus, tendon (midpoint), medial and lateral tibial condyle, tibial tuberosity, femoral
epicondyle, and hip joint (Figure 1). The motion analysis global
coordinate system was aligned to the force plate. The ankle joint
rotation center was estimated as a virtual point half the distance
between the lateral and medial malleoli.22 The perpendicular distance to the ankle joint center from the line joining the calcaneus
marker and the Achilles tendon marker was taken as the moment
arm after correction for skin thickness measured by ultrasound
(US). Tendon force was calculated by dividing the externally applied
ankle joint moment by the moment arm and normalized across
subjects by body weight. Torque due to inertia around the foot was
ignored, as movements were slow. However, a complete calculation
of angular torque around the ankle, including forces exerted on the
foot by gravity and acceleration, was carried out on three subjects
to validate this assumption.
Measurement of Tendon Elongation
Advances in ultrasonography and motion analysis have enabled
simultaneous use to determine in vivo human muscle-tendon
complex dynamics.23 In the current study, a technique to measure
continual tendon elongation was developed to measure tendon
mechanical behavior throughout exercise.
Tendon length was defined as the distance between the Achilles
tendon insertion and the distal MTJ of the medial gastrocnemius
(Figure 2). The point of Achilles tendon insertion was tracked using
a single marker placed on the calcaneus. The MTJ was tracked
using combined CODA and an ultrasound system (Voluson e, GE
Healthcare, UK; wide band, multifrequency linear transducer,
frequency = 3.7–11.3 MHz, FOV = 37.4 mm, FR = 17Hz). Two
markers were placed along the ultrasound probe axis while a third
marker was placed approximately 60 mm from the midpoint of
the line intersecting the initial markers (Figure 2). This enabled
the probe position to be tracked and spatially synchronized with
the motion data.
A synchronizing system was used to trigger motion capture
to ensure accurate ultrasound frame attribution. This consisted
of a TTL switch coupled to the digital input of the CODA active
hub, providing a triggering pulse to start data acquisition at the
commencement of the US scan with a maximum temporal lag of
55 ms (manufacturer values). The CODA active hub was electronically integrated with the force plates and EMG system, enabling
synchronization of all four devices.
Tracking of the Muscle-Tendon Joint
Tracking the MTJ throughout the US video required identifying
the precise location of the intersection of the superficial and deep
connective tissue in every frame. To reduce human error, improve
accuracy, and increase speed, an algorithm was developed and
implemented in MatLab (Version 7.9.0.529 [R2009b], 32-bit,
MathWorks, Natick, MA). To automate tracking, the projection of
the two brightest lines (Figure 3A) was established along the radial
line, and their relative orientation angle determined using the Radon
transform. Briefly, the Radon transform provides an intensity map of
an image, from which the 2D location and angular orientation of the
most intense image points can be determined. Extracting data from
the most intense points enabled the MTJ to be located from the intersection of the eight brightest lines each side of the junction (Figure
3B). From the 8 × 8 matrix of brightest lines, the 64 intersection
Figure 2 — Muscle-tendon junction (MTJ) tracking to establish Achilles tendon length. C and C’ denote the calcaneus position at full plantar flexion
and dorsiflexion, respectively. The central images indicate the movement of the MTJ during exercise, from MTJ to MTJ’ at the two extreme positions.
The distance between C and MTJ at any point in time is defined as the Achilles tendon length. A force record is also shown demonstrating the change
in force with the movement. US = ultrasound.
JAB Vol. 31, No. 2, 2015
72 Chaudhry et al.
Figure 3 — (A) Part of a single ultrasound video frame (480 × 190 pixels) showing the muscle-tendon junction (MTJ). (B) The same image with
superimposed lines. The lower lines are the eight brightest lines selected from the 231 Radon transform. These lie within the angle range 93º–103º with
intercepts within the range of 54–74 pixels. The upper lines are a similar set for the angle range 79º–89º and intercept range of 107–127. The mean
angle and mean intercepts of these lines is used to set the target range for analyzing the next frame of the video. (Note. Angles in the Radon transform
are measured with respect to the vertical in these figures, and the intercepts are with a line perpendicular to each Radon line passing through the center
of the part of the frame analyzed (white + in B). (C) The same image marked in white with the 64 points at which each pair of lines (ie, one from the
top set and one from the bottom set) approach to within 10 pixels. The black cross shows the median coordinates of these 64 points, and is the value
used for the position of the MTJ within this video frame.
points were located, and the junction taken as the median of these
(Figure 3C). Reliability of this algorithm was assessed by comparing results from three separate attempts to track each video. In
addition, it was compared with the manual method of MTJ tracking using Image J software (version 1.44o, NIH, Bethesda, MD).
Assessment of Accuracy in Ultrasound
Measurements
The accuracy of the US setup for tracking movement was further
assessed using a phantom. A metallic wand of known dimensions was immersed in a container of ultrasound gel, such that the
container of gel represented the leg and the wand represented the
tendon. Along the length of the wand, a metallic coil was placed
to represent the MTJ. Moving the wand within the container, the
coil was tracked throughout the movement using the ultrasound
system. To assess US tracking accuracy, wand and coil displacement were also measured directly using CODA markers, and the
two data sets compared.
Measurement of Muscle Activation
EMG recordings were made using Ag/AgCl dual snap electrodes
with a 20 mm interelectrode distance. After careful skin preparation, electrodes were placed on the belly of the soleus, lateral
gastrocnemius, medial gastrocnemius, and tibialis anterior muscles
following the SENIAM guidelines24 and a single, self-adhesive Ag/
AgCl snap electrode placed on the anterior tibia as a reference. EMG
signals were preamplified and band-pass filtered between 10 and
1000 Hz before sampling at 1500 Hz using a wireless EMG system
(Telemyo 2400T G2, Noraxon, Scottsdale, AZ; input impedance >
100 MΩ, common mode rejection ratio > 100 dB, base gain 500)
with 16-bit analog-to-digital resolution.
Data Analysis
Data analysis was carried out in MatLab. For each exercise task,
only cycles within 10% of the prescribed speed were retained for
analysis.
JAB Vol. 31, No. 2, 2015
Eccentric and Concentric Exercise of Triceps Surae 73
To ensure only the loading phase of each exercise was assessed,
points of minimum and maximum heel height were identified and
the portion of the record between them resampled by interpolation
to 111 uniformly spaced points. Times of single-leg support were
confirmed by absence of a force reading from the plate under the
contralateral foot. For descriptive purposes, ankle angles were
extracted during each test. After registration, GrF and EMG data
were obtained for each subject. To establish muscle activation, the
EMG signal was full-wave rectified, then smoothed using a moving
median filter with a window size of 121. To compare posterior and
anterior compartment activation, the contribution of the gastrocnemius and soleus were summed for posterior activation. Anterior
data were taken directly from the tibialis anterior.
Having established appropriate subject repeatability over repeat
exercises, mean force-length curves for each subject were produced
over the three repeats of eccentric and concentric. Apparent tendon
stiffness was measured from the slope of the mean force-length
curve for each individual, placing a linear regression line through
the data.
After registration, the vertical component of the GrF vector
was further analyzed to investigate for tremor (defined throughout
the text as perturbations) in the GrF, and subsequently in the Achilles. First, the mean value was subtracted to make GrF independent
of the body mass. To obtain the magnitude and frequency of the
perturbations, power spectrum densities were calculated using a
fast Fourier transformation (FFT) method, after elimination of the
dc component. The power was subsequently summed within 1 Hz
windows across the frequency range 0–16 Hz. Finally, any correlation between tendon stiffness and the extent of perturbation in the
GrF and Achilles was investigated across subjects.
Comparing eccentric and concentric across the subject
population showed no significant differences in force-displacement
behavior during the two exercises for most subjects. Range of
motion was also similar; 47.94 ± 1.83° during eccentric and 47.05
± 1.93° during concentric, with a shortening velocity of 8.4 ± 1.5
mm/s during eccentric and 7.9 ± 1.1 mm/s during concentric.
Unsurprisingly, this resulted in no significant variation in tendon
stiffness between eccentric and concentric. Intrasubject variability
was far greater, and concentric and eccentric resulted in distinctly
different temporal loading patterns across subjects. Eccentric
and concentric mean force-length curves were produced for each
individual, with example data for six individuals shown in Figure
4. It is interesting to note that for some individuals there is a clear
difference in tendon behavior between eccentric and concentric.
There were also notable variations in tendon stiffness between
subjects. To enable a direct comparison of tendon properties
during eccentric and concentric across all subjects, normalized
force and displacement were derived. Figure 5 shows mean forcelength behavior across all subjects during eccentric and concentric,
with the time axis reversed for eccentric, for ease of comparison.
Interestingly, clear temporal differences between eccentric and
concentric were apparent for both tendon force and length (Figure
5A, B). However, these differences were lost in mean force-length
data (Figure 5C). After carrying out paired comparisons of eccentric
and concentric across all subjects, no significant differences were
observed in maximum extension (P = .07) or stiffness (P = .06).
Statistical Analysis
Normality tests were performed on the data using the Shapiro-Wilk
method (OriginPro version 8, OriginLab, Northampton, MA).
Once the data were found to be normally distributed, paired t tests
were used to examine differences in tendon force, perturbations,
muscle activation, and tendon extension for each exercise. For all
statistical tests, significance was established at P ≤ .05. Data are
presented as mean (SD). To assess accuracy of US tracking, the
correlations between tracking distances measured by ultrasound
and motion markers were determined, and the intraclass correlation
was calculated. Systematic bias between the two measures was also
investigated using a paired t test.
Results
Using the combined motion analysis and US capture techniques to
scan movement of the phantom revealed that US could determine
position to an accuracy of ±1.65 mm. The repeatability of the MTJ
tracking algorithm was found to be ±0.59 × 0.36 mm for tracking a video multiple times. Good correlation was found between
automatic and standard tracking methods with an intraclass correlation coefficient of 0.98 for the x axis (95% confidence interval
0.94–0.99) and 0.62 (95% confidence interval 0.26–0.88) for the y
axis. No systematic difference between data sets was observed (P
= .12 in x direction and P = .09 in y direction). There was less than
1% difference in the tendon forces calculated from torque around
the ankle with and without the inclusion of inertial effects, hence
inertia around the foot was ignored.
Figure 4 — Graphs of the mean Achilles tendon force-length (ATF-ATL)
relationship for six subjects, shown in separate panels. White squares report
eccentric data and black circles report concentric data. Some subjects show
different loading behavior between eccentric and concentric.
JAB Vol. 31, No. 2, 2015
74 Chaudhry et al.
Figure 5 — A comparison of (A) normalized mean Achilles tendon force (± SE) and (B) normalized mean Achilles tendon length (± SE) against time
during the loading cycle. (C) Achilles tendon force versus Achilles tendon length and (D) electromyography (EMG) data comparing activation in the
posterior muscles (circles) and anterior muscles (squares) during eccentric and concentric across all participants. Eccentric data are shown as open
circles/squares whereas concentric data are filled circles/squares.
Tendon apparent stiffness varied across the subject population,
but not between exercise types (t = .82, P = .43), with mean (± SE)
values of 62.79 N∙mm–1 (9.37) during concentric and 58.89 N∙mm–1
(9.21) during eccentric.
Considering the EMG data (Figure 6D), muscle activation
during the cycle and maximal muscle activation values were significantly higher in concentric than eccentric (P < .01) for both the
anterior compartment (concentric = 0.030 mV [SE 0.004], eccentric
= 0.025 mV [SE 0.003]) and the posterior compartment (concentric =
0.140 mV [SE 0.014], eccentric = 0.118mV [SE 0.011]) of the triceps
surae muscle group.
A greater amplitude of higher frequency force perturbations
were present during eccentric (P = .02) at 10 Hz and showed negative
correlation with tendon stiffness at 10 Hz. Summary data showing
the FFT of tendon force (Figure 6C) demonstrated that the main
amplitude of the power component was typically seen at frequencies below 5 Hz, with a peak around 2–3 Hz, representative of low
frequency movements. However, a range of vibration frequencies of
JAB Vol. 31, No. 2, 2015
Eccentric and Concentric Exercise of Triceps Surae 75
A
C
B
D
Figure 6 — (A) Example data set showing normalized Achilles tendon force (ATF) and normalized ground reaction force (GrF) for one subject during
eccentric. The data are detrended, removing low frequency force change corresponding with movement, to focus on perturbations. (B) The corresponding fast Fourier transformation (FFT) of the GrF data. (C) Mean FFT of Achilles tendon force across all subjects for eccentric (empty circles) and
concentric (filled squares) (± SE). (D) The correlation between magnitude of perturbations at 10 Hz and tendon stiffness during eccentric across the
subject population. A negative correlation was observed (R2 = .59).
up to 15 Hz were present in all subjects. Statistically significant differences between eccentric and concentric were observed at around
10 Hz, where tremor during eccentric was greater. Further, the magnitude of 10 Hz perturbations was negatively correlated with tendon
stiffness (Figure 6D). However, not every subject showed a higher
power of perturbations at around 10 Hz during eccentric (7 out of 11).
Discussion
The current study shows that eccentric and concentric do not differ
in terms of tendon force and stiffness behavior during exercise.
However, there is a clear difference in the magnitude of perturbations generated by these exercises, with perturbations at around 10
JAB Vol. 31, No. 2, 2015
76 Chaudhry et al.
Hz generated specifically during eccentric exercise. Furthermore,
data highlights that the magnitude of tendon perturbations is
inversely correlated to tendon stiffness.
To collate these findings, this paper presents a novel semiautomated method of measuring of tendon extension during high-strain
movement, which increased the efficiency of analyzing one video
of 200 frames 10 times as compared with manual tracking. These
findings also provide a solution, independent of the human eye,
able to track large dynamic movements when the region of interest
moves out of the field of view. Such an algorithm can be very useful
in different biomechanical studies based around MTJ tracking.
Systematic error is always to be considered with any such
measurement technique. However, by placing markers on the
US probe to account for its movement, the use of US to measure
tendon extension has previously been validated and considered reliable.23 However, the US image tracking technique developed and
tested during this study can still be sensitive to temporal lag and
image quality, and skill and practice are required to track the MTJ
accurately. Temporal lag was less than 2% of the movement cycle
and unlikely to significantly affect data. Blurred or low intensity
images prevented accurate tracking, thus videos that did not meet
the quality standards were not included in the analysis. However,
of 88 videos collected, only 8 were below the required standard. In
addition, while the technique was shown to be reproducible for one
participant, individual to individual variability may occur, owing
to factors such as differences in individual muscle architecture
and changes in tissue shape during movement. These may alter the
apparent location of the MTJ which would in turn influence strain
and stiffness measures. Furthermore, the use of a constant value for
tendon thickness may increase interindividual variability. However,
using identical test procedures and a single tester in the current
study minimized error and enabled eccentric and concentric to be
successfully compared. Indeed, no significant differences in tendon
mechanical properties or muscle activity data were observed, even
when the same subject was tested on two consecutive days, as long
as the ROM, subject foot placement, and US probe position were
consistent between the measures. This was ensured by carefully
marking the positions for the second day of measures.
At the start of a concentric movement from full dorsiflexion,
the calf muscles are activated, increasing in activation as they
accelerate the subject upwards. Tendon force is at its peak at the
start of the concentric movement, possibly as the triceps surae
moment arm is furthest from the ankle joint axis for sagittal plane
movement at this point. As the subject rises to plantar flexion, force
reduces. In contrast, subjects start an eccentric movement with the
heel at its highest point, and lower ‘under control’. The movement
is controlled (resisted) by lengthening of the activated calf muscle
and stretch of the Achilles tendon. Maximal force occurs at the
end of the movement, again at full dorsiflexion, when maximum
force is required to decelerate the subject against gravity. It should
be noted that the tendon was always loaded during the movement
and therefore the force-extension behavior seen is similar to that
evident in the linear region of an in vitro test curve, with muscle
contraction likely acting to strain the tendon beyond the low stiffness toe region seen when testing tendon in vitro. Normalizing
force curves to heel height enables the change in force during each
movement to be readily compared. From these data, significantly
higher forces in the tendon were observed during concentric for a
small range of ankle angles when the foot was fully dorsiflexed,
and the forces passing through the tendon are maximum. However,
no overall significant differences in force during the loading cycle
were observed between eccentric and concentric. Such a finding
is consistent with a previous study,10 comparing similar eccentric
and concentric exercises. However, Rees et al10 only reported the
maximum forces during eccentric and concentric. This study has
additionally reported tendon stiffness and its relationship with 10
Hz perturbations.
Also in agreement with previous studies,25,26 our data highlighted comparatively lower EMG activity during eccentric than
concentric. Muscle activation within the medial gastrocnemius,
lateral gastrocnemius, and soleus were all significantly lower
during eccentric (P < .01). By contrast, there was no significant
difference in the antagonistic tibialis anterior muscle activation
between eccentric and concentric. Such a response from the tibialis
anterior also follows previously reported findings.27 However, there
is variability considering data from a range of different studies, as
a result of different test methodologies.28,29
Individual variations in tendon properties, particularly stiffness, were evident across the group of participants. However, no
significant differences in tendon stiffness between eccentric and
concentric were reported. Repeats by an individual were similar,
and reported tendon properties were within the range of previously
published data.15,30,31 The variation in Achilles tendon stiffness
between individuals could be partly dependent on previous training;27 although all subjects in the study performed only moderate
activity, the effect of specific training cannot be ruled out. In
addition, measurements were undertaken following a clinicallyrelevant protocol, where no control for the muscle activation was
considered, a factor previously shown to affect stiffness measurements.32,33 However, to get a true understanding of the differences
between eccentric and concentric, it was necessary to carry out
measurements during the typical protocol and not under isometric
conditions. By reporting whole curves, we have shown for the first
time the extent of variability in tendon loading curves between
individuals, which could be relevant to understanding the variable
patient responses to therapy. We have also shown temporal differences between eccentric and concentric for the first time, which
could be of further interest.
Strain is an important stimulus for tendon repair. A number
of studies have indicated that mechanical loading is an essential
stimulus for tendon repair,34,35 initiating tenocyte signaling mechanisms in a strain dependent manner.36 Results from the current study
indicate that eccentric and concentric, performed under equal load
and at the same speed, result in different temporal strains within the
tendon (Figure 5A, B). However, if the data are plotted for normalized position (Figure 5C), this difference disappears. Consequently,
differences in strain alone are unlikely to be the differentiating
trigger for repair during eccentric exercise, although the situation
may differ in patients with tendinopathy.
By contrast, tendon perturbations have been shown to differ
between eccentric and concentric, and were higher during eccentric,
particularly around 10 Hz (P < .05). Significantly higher perturbations during eccentric, in the 8–12 Hz range, have been reported
previously.11 However, the origin of these perturbations has not previously been considered. The current study assessed the correlation
between perturbations and tendon stiffness and found that while
perturbation frequency does not change with tendon stiffness, the
magnitude of tendon perturbations at 10 Hz is inversely correlated
to tendon stiffness. This suggests the perturbations are not a result
of mechanical resonance, as previous studies suggest a correlation
between perturbation frequency and tendon stiffness should be
present in such a situation.12,13,37,38 However, the negative correla-
JAB Vol. 31, No. 2, 2015
Eccentric and Concentric Exercise of Triceps Surae 77
tion between tendon stiffness and perturbation magnitude at 10 Hz
during eccentric does show some influence of tendon mechanics on
resulting perturbations. This relationship is interesting, as tendon
stiffness has previously been reported to decrease in tendinopathic
tissue,39 which would imply that 10 Hz perturbations would be
greater in injured tendon.40 Indeed, an increased magnitude perturbation has recently been reported.41
It is important to note that not every subject showed 10 Hz
perturbations, implying that not every patient performing eccentric
would receive this stimulus. If this stimulus has some relationship
with tendinopathy, this may partly explain why not all subjects
respond to eccentric. However, further analysis of perturbations is
needed to draw firm conclusions. It is still unclear whether perturbations are a beneficial stimulus for tendon repair or, in fact, indicate
a muscle weakness or poorly synchronized muscle activity in the
triceps surae, both of which may be risk factors for tendinopathy.
It may be that eccentric treats tendinopathy by training the triceps
surae to function more effectively and minimize perturbation, with
a progressive load and speed challenge incorporated as the triceps
surae adapts to the training stimuli.
There is a lack of consensus concerning the management of
tendinopathy, which has limited options for developing new and
more effective treatments. This study has given further insight
into the biomechanical behavior of the triceps surae during physiologically representative eccentric and concentric exercises. It has
highlighted the need to carry out controlled training studies, where
the effects of carefully controlled eccentric and concentric exercise
are examined with a simultaneous investigation of the underlying
mechanisms.
References
1. Maffulli N, Khan KM, Puddu G. Overuse tendon conditions: time to
change a confusing terminology. Arthroscopy. 1998;14(8):840–843.
PubMed doi:10.1016/S0749-8063(98)70021-0
2. Kvist M. Achilles tendon injuries in athletes. Ann Chir Gynaecol.
1991;80(2):188–201. PubMed
3. Jarvinen TA, Kannus P, Paavola M, Jarvinen TL, Jozsa L, Jarvinen M.
Achilles tendon injuries. Curr Opin Rheumatol. 2001;13(2):150–155.
PubMed doi:10.1097/00002281-200103000-00009
4. James SL, Bates BT, Osternig LR. Injuries to runners. Am J Sports
Med. 1978;6(2):40–50. PubMed doi:10.1177/036354657800600202
5. Sayana MK, Maffulli N. Eccentric calf muscle training in non-athletic
patients with Achilles tendinopathy. J Sci Med Sport. 2007;10(1):52–
58. PubMed doi:10.1016/j.jsams.2006.05.008
6. Alfredson H, Pietilä T, Jonsson P, Lorentzon R. Heavy-load eccentric
calf muscle training for the treatment of chronic Achilles tendinosis.
Am J Sports Med. 1998;26(3):360–366. PubMed
7. Mafi N, Lorentzon R, Alfredson H. Superior short-term results
with eccentric calf muscle training compared to concentric training in a randomized prospective multicenter study on patients with
chronic Achilles tendinosis. Knee Surg Sports Traumatol Arthrosc.
2001;9(1):42–47. PubMed doi:10.1007/s001670000148
8. Fahlstrom M, Jonsson P, Lorentzon R, Alfredson H. Chronic Achilles tendon pain treated with eccentric calf-muscle training. Knee
Surg Sports Traumatol Arthrosc. 2003;11(5):327–333. PubMed
doi:10.1007/s00167-003-0418-z
9. Stanish WD, Rubinovich RM, Curwin S. Eccentric exercise in chronic
tendinitis. Clin Orthop Relat Res. 1986; (208):65–68. PubMed
10. Rees JD, Lichtwark GA, Wolman RL, Wilson AM. The mechanism
for efficacy of eccentric loading in Achilles tendon injury; an in vivo
study in humans. Rheumatology (Oxford). 2008;47(10):1493–1497.
PubMed doi:10.1093/rheumatology/ken262
11. Henriksen M, Aaboe J, Bliddal H, Langberg H. Biomechanical characteristics of the eccentric Achilles tendon exercise. J
Biomech. 2009;42(16):2702–2707. PubMed doi:10.1016/j.jbiomech.2009.08.009
12. Stiles RN, Randall JE. Mechanical factors in human tremor frequency.
J Appl Physiol. 1967;23(3):324–330. PubMed
13. McAuley JH, Marsden CD. Physiological and pathological tremors
and rhythmic central motor control. Brain. 2000;123(Pt 8):1545–1567.
PubMed doi:10.1093/brain/123.8.1545
14. Maganaris CN. Tensile properties of in vivo human tendinous tissue.
J Biomech. 2002;35(8):1019–1027. PubMed doi:10.1016/S00219290(02)00047-7
15. Maganaris CN, Paul JP. Tensile properties of the in vivo human gastrocnemius tendon. J Biomech. 2002;35(12):1639–1646. PubMed
doi:10.1016/S0021-9290(02)00240-3
16. Korstanje JW, Selles RW, Stam HJ, Hovius SE, Bosch JG. Development and validation of ultrasound speckle tracking to quantify
tendon displacement. J Biomech. 2010;43(7):1373–1379. PubMed
doi:10.1016/j.jbiomech.2010.01.001
17. Loram ID, Maganaris CN, Lakie M. Use of ultrasound to make noninvasive in vivo measurement of continuous changes in human muscle
contractile length. J Appl Physiol. 2006;100(4):1311–1323. PubMed
doi:10.1152/japplphysiol.01229.2005
18. Magnusson SP, Hansen P, Aagaard P, et al. Differential strain patterns
of the human gastrocnemius aponeurosis and free tendon, in vivo. Acta
Physiol Scand. 2003;177(2):185–195. PubMed doi:10.1046/j.1365201X.2003.01048.x
19. Gerus P, Rao G, Berton E. A method to characterize in vivo tendon
force-strain relationship by combining ultrasonography, motion capture and loading rates. J Biomech. 2011;44(12):2333–2336. PubMed
doi:10.1016/j.jbiomech.2011.05.021
20. Pearson SJ, Ritchings T, Mohamed AS. The use of normalized crosscorrelation analysis for automatic tendon excursion measurement in
dynamic ultrasound imaging. J Appl Biomech. 2013;29(2):165–173.
PubMed
21. Maganaris CN, Paul JP. In vivo human tendon mechanical properties.
J Physiol. 1999;521(Pt 1):307–313. PubMed doi:10.1111/j.14697793.1999.00307.x
22. Wu G, Siegler S, Allard P, et al. ISB recommendation on definitions
of joint coordinate system of various joints for the reporting of human
joint motion–part I: ankle, hip, and spine. International Society of
Biomechanics. J Biomech. 2002;35(4):543–548. PubMed doi:10.1016/
S0021-9290(01)00222-6
23. Lichtwark GA, Wilson AM. In vivo mechanical properties of the
human Achilles tendon during one-legged hopping. J Exp Biol.
2005;208(Pt 24):4715–4725. PubMed doi:10.1242/jeb.01950
24. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of
recommendations for SEMG sensors and sensor placement procedures.
J Electromyogr Kinesiol. 2000;10(5):361–374. PubMed doi:10.1016/
S1050-6411(00)00027-4
25. Enoka RM. Eccentric contractions require unique activation strategies by the nervous system. J Appl Physiol. 1996;81(6):2339–2346.
PubMed
26. Henriksen M, Aaboe J, Bliddal H, Langberg H. Biomechanical characteristics of the eccentric Achilles tendon exercise. J
Biomech. 2009;42(16):2702–2707. PubMed doi:10.1016/j.jbiomech.2009.08.009
27. Reeves ND, Maganaris CN, Narici MV. Effect of strength training
on human patella tendon mechanical properties of older individuals.
JAB Vol. 31, No. 2, 2015
78 Chaudhry et al.
J Physiol. 2003;548(Pt 3):971–981. PubMed doi:10.1113/
jphysiol.2002.035576
28. Kellis E, Baltzopoulos V. Muscle activation differences between
eccentric and concentric isokinetic exercise. Med Sci Sports Exerc.
1998;30(11):1616–1623. PubMed doi:10.1097/00005768-19981100000010
29. Reeves N, Narici M. Behavior of human muscle fascicles during
shortening and lengthening contractions in vivo. J Appl Physiol.
2003;95(3):1090–1096. PubMed
30. Finni T, Hodgson JA, Lai AM, Edgerton VR, Sinha S. Nonuniform
strain of human soleus aponeurosis-tendon complex during submaximal voluntary contractions in vivo. J Appl Physiol. 2003;95(2):829–
837. PubMed
31. Muramatsu T, Muraoka T, Takeshita D, Kawakami Y, Hirano Y, Fukunaga T. Mechanical properties of tendon and aponeurosis of human
gastrocnemius muscle in vivo. J Appl Physiol. 2001;90(5):1671–1678.
PubMed
32. Hof AL. In vivo measurement of the series elasticity release curve of
human triceps surae muscle. J Biomech. 1998;31(9):793–800. PubMed
doi:10.1016/S0021-9290(98)00062-1
33. Hof AL. Muscle mechanics and neuromuscular control. J Biomech. 2003;36(7):1031–1038. PubMed doi:10.1016/S00219290(03)00036-8
34. Takai S, Woo SL, Horibe S, Tung DK, Gelberman RH. The effects of
frequency and duration of controlled passive mobilization on tendon
healing. J Orthop Res. 1991;9(5):705–713. PubMed doi:10.1002/
jor.1100090510
35. Wang N, Butler JP, Ingber DE. Mechanotransduction across the cell
surface and through the cytoskeleton. Science. 1993;260(5111):1124–
1127. PubMed doi:10.1126/science.7684161
36. Scott A, Danielson P, Abraham T, Fong G, Sampaio AV, Underhill
TM. Mechanical force modulates scleraxis expression in bioartificial
tendons. J Musculoskelet Neuronal Interact. 2011;11(2):124–132.
PubMed
37. Halliday AM, Redfearn JW. An analysis of the frequencies of finger
tremor in healthy subjects. J Physiol. 1956;134(3):600–611. PubMed
doi:10.1113/jphysiol.1956.sp005668
38. Joyce GC, Rack PM. The effects of load and force on tremor at the
normal human elbow joint. J Physiol. 1974;240(2):375–396. PubMed
doi:10.1113/jphysiol.1974.sp010615
39. Arya S, Kulig K. Tendinopathy alters mechanical and material properties of the Achilles tendon. J Appl Physiol. 2010;108(3):670–675.
PubMed doi:10.1152/japplphysiol.00259.2009
40. Grigg NL, Wearing SC, O’Toole JM, Smeathers JE. Achilles Tendinopathy modulates force frequency characteristics of eccentric
exercise. Med Sci Sports Exerc. 2013;45(3):520–526. PubMed
41. Grigg NL, Wearing SC, O’Toole JM, Smeathers JE. The effect of
exercise repetition on the frequency characteristics of motor output
force: implications for Achilles tendinopathy rehabilitation. J Sci Med
Sport. 2014;17(1):13–17. PubMed doi:10.1016/j.jsams.2013.03.014
JAB Vol. 31, No. 2, 2015