Concentric and Eccentric Time-Under-Tension during Strengthening Exercises: Validity and Reliability of Stretch-Sensor Recordings from an Elastic Exercise-Band
Bandholm T (2013) Concentric and Eccentric Time-Under-Tension during Strengthening Exercises: Validity and
Reliability of Stretch-Sensor Recordings from an Elastic Exercise-Band. PLoS ONE 8(6): e68172. doi:10.1371/journal.pone.0068172
Concentric and Eccentric Time-Under-Tension during Strengthening Exercises: Validity and Reliability of Stretch-Sensor Recordings from an Elastic Exercise- Band
Michael Skovdal Rathleff 0
Kristian Thorborg 0
Thomas Bandholm 0
Franois Hug, The University of Queensland, Australia
0 1 Orthopaedic Surgery Research Unit, Aalborg University Hospital, Aalborg, Denmark , 2 Arthroscopic Centre Amager , Copenhagen University Hospital , Amager, Copenhagen , Denmark , 3 Department of Orthopedic Surgery, Copenhagen University Hospital , Hvidovre , Denmark , 4 Department of Physical Therapy, Physical Medicine and Rehabilitation Research - Copenhagen, Copenhagen University Hospital , Hvidovre , Denmark , 5 Clinical Research Centre, Copenhagen University Hospital , Hvidovre, Copenhagen , Denmark
Background: Total, single repetition and contraction-phase specific (concentric and eccentric) time-under-tension (TUT) are important exercise-descriptors, as they are linked to the physiological and clinical response in exercise and rehabilitation. Objective: To investigate the validity and reliability of total, single repetition, and contraction-phase specific TUT during shoulder abduction exercises, based on data from a stretch-sensor attached to an elastic exercise band. Methods: A concurrent validity and interrater reliability study with two raters was conducted. Twelve participants performed five sets of 10 repetitions of shoulder abduction exercises with an elastic exercise band. Exercises were video-recorded to assess concurrent validity between TUT from stretch-sensor data and from video recordings (gold standard). Agreement between methods was calculated using Limits of Agreement (LoA), and the association was assessed by Pearson correlation coefficients. Interrater reliability was calculated using intraclass correlation coefficients (ICC 2.1). Results: Total, single repetition, and contraction-phase specific TUT - determined from video and stretch-sensor data - were highly correlated (r>0.99). Agreement between methods was high, as LoA ranged from 0.0 to 3.1 seconds for total TUT (2.6% of mean TUT), from -0.26 to 0.56 seconds for single repetition TUT (6.9%), and from -0.29 to 0.56 seconds for contraction-phase specific TUT (13.2-21.1%). Interrater reliability for total, single repetition and contraction-phase specific TUT was high (ICC>0.99). Interrater agreement was high, as LoA ranged from -2.11 to 2.56 seconds for total TUT (4.7%), from -0.46 to 0.50 seconds for single repetition TUT (9.7%) and from -0.41 to 0.44 seconds for contraction-phase specific TUT (5.2-14.5%). Conclusion: Data from a stretch-sensor attached to an elastic exercise band is a valid measure of total and single repetition time-under-tension, and the procedure is highly reliable. This method will enable clinicians and researchers to objectively quantify if home-based exercises are performed as prescribed, with respect to time-under-tension.
Funding: The Bevica Foundation paid supplied funding for the hardware used in the current study. The funders had no role in study design, data collection
and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
In the rehabilitation of patients, home-based exercises using
elastic exercise bands are often prescribed after a thorough
initial instruction by the prescribing physician or physiotherapist
. Elastic exercise bands have several qualities that make
them preferable in rehabilitation and clinical settings. They
provide adjustable resistance, and have been shown to be
effective in clinical trials on shoulder, neck, knee and hip pain
. However, adherence to home-based exercise
programmes often appears to be insufficient [5,811].
We recently showed that automatically stored data from a
stretch-sensor attached to a standard elastic exercise band
could be used to accurately identify training adherence and
quality of shoulder-abduction strength training . In the
current study, we used the same technology to investigate if
stored data from the stretch sensor could be used to quantify
time-under-tension (TUT) of shoulder-abduction strength
Total TUT refers to the total time of all concentric,
quasiisometric and eccentric contraction-phases in a single training
set . Together with load and movement velocity it is an
important strength training descriptor, as it reflects the time
factor of the strength training stimulus . Physiologically,
greater TUT has been shown to increase myofibrillar protein
synthesis more than lesser TUT after a single, work-matched,
strength training session in healthy subjects .
Contractionphase specific TUT refers to the time of the individual phases
of TUT (concentric, quasi-isometric, and eccentric
contractionphases) in a single repetition [13,17]. Specific
contractionphases including their TUT are also important strength training
descriptors [13,17]. Physiologically, evidence from animal
studies suggests that eccentric contractions may govern
specific mechanisms responsible for tendon healing, while
isolated concentric contractions may not have similar effects
. However, clinical, evidence suggest that the
combination of concentric and eccentric contractions may be
more effective than eccentric contractions alone on the clinical
outcome in achilles and patellar tendinopathy . In healthy
subjects, eccentric contractions elicit larger increases in
strength and hypertrophy compared to concentric contractions
. Hence, the quantification of both total and
contractionphase specific TUT of performed strength training of the
shoulder abductors is important in order to determine if
executed training constitutes a sufficient physiological and
clinical stimulus. In particular, this is of importance if exercises
are performed at home after initial instruction by a physician or
physiotherapist. Simple tools capable of measuring TUT during
exercises with elastic exercise bands are currently unavailable.
The objective of the study was therefore to investigate the
validity and reliability of total, single repetition and
contractionphase specific TUT during shoulder abduction exercises,
based on data from a simple set-up that included a
stretchsensor attached to a standard elastic exercise band.
All participants were provided with verbal and written
information about the procedures of the study, and written
informed consent was obtained in accordance with the
Declaration of Helsinki. The Ethics Committee in the North
Denmark Region and the Danish Data Agency approved the
study (2012-2410). The participant in Figures 1 and 2 has
given written informed consent, as outlined in the PLOS
consent form, for publication of their photograph.
The study was a concurrent validity and interrater reliability
study. The study investigated if it were possible to determine
total TUT, single TUT and contraction-phase specific TUT
during shoulder abduction exercises, based on data recordings
from a stretch-sensor attached to an elastic exercise band.
Video recordings were used as a gold standard, and recorded
concordantly, so each rater could determine, from both video
recordings and stretch sensor data, the total TUT, single
repetition TUT and contraction-phase specific TUT. The
reporting of the study follows the Guidelines for Reporting
Reliability and Agreement Studies (GRRAS) .
A convenience sample of 12 healthy volunteers from the
local hospital staff was recruited. They were between 21 and
29 years of age and free of upper extremity symptoms.
Two raters with different clinical and technological
experience using elastic band exercise were recruited. Hence,
we considered the two raters representative of a larger
population of raters who, in a future setting, would use the
stretch-sensor equipment being investigated. Rater 1 was a
male research physiotherapist with three years of research
experience and 10 hours of prior experience of rating
stretchsensor data. Rater 2 was a female physiotherapist with three
years of clinical experience and no previous experience in
rating stretch-sensor data. Rater 1 already had experience in
rating video recordings and stretch-sensor data and received
no additional practice. Rater 2 was given 30 minutes of practice
in rating stretch-sensor data from participants who were not
part of the current study sample.
Stretch-sensor and elastic exercise band
The elastic stretch-sensor was based on technology
designed by Danfoss PolyPower (Nordborg, Denmark) and
custom-made for the application. It acts as an elastic capacitive
material that is stretchable in one direction. To measure the
change in capacitance of the stretch-sensor, an electrical
circuit was built with a timer. The circuit was built in such a way
that a change in capacitance of the stretch-sensor
corresponded to a change in the timer frequency, which
allowed us to measure how much the sensor was being
stretched (for further details, see Kappel et al ). The sensor
was attached to the rubber band by two clips that made the
sensor easily transferable to other elastic exercise bands,
Figure 1. The sensor was attached via an USB connector to a
small box that contains the electrical circuit, timer and data
logger. The sampling frequency of the box was 200 Hz. The
sensor is robust, and the material properties are robust to
changes during usage . A switch was mounted in the
handle of the elastic exercise band, so that the data recording
would start whenever the handle was pressed. The elastic
band was a standard blue Thera-Band exercise band, which is
commonly used in rehabilitation studies [1,4,5,25]. Before the
exercise data were collected, we tested the handle switch by
pressing it a total of 100 times. Every time the handle was
pressed, the data collection started as intended, and the
correct information regarding date and time of day was stored
on the memory card .
Participants were instructed in performing shoulder
abduction exercises from 0 to 90 degrees abduction. They
were told to perform 10 repetitions at a self-determined speed,
meaning there was no restriction to perform either of the
contraction-phases at a predetermined speed or at a relative
loading (repetition maximum). However, we secured that no
slack was present in the exercise band at 0 degrees abduction
meaning that the tension in the elastic band increased
throughout the shoulder abduction with a peak at the end of the
abduction due to the elastic tension. This approach was
chosen to ensure a wide range of different speeds of exercises.
All participants performed five sets of 10 repetitions with a
2minute break between each set. The exercises were performed
in front of a white wall, and recorded with a 50mm lens on a
Canon 5D Mark II mounted on a tripod, Figure 2. Video was
recorded at a resolution of 1920x1080 pixels at 24 frames per
second. Hence, the sampling frequency was different between
the stretch-sensor and video recordings (200 Hz. vs. 24 Hz.)
and not synchronised. However this should not pose a treat to
the findings of the study, as the design resembles the validity
studies performed on handheld dynamometry versus isokinetic
dynamometry (gold standard) where synchronization is
impossible and sampling frequency is different between
Rating of the stretch sensor data was done by analysing data
from the stretch-sensor with Matlab 2011a (The MathWorks,
Nattick, USA). A custom-written Matlab programme
transformed the stretch-sensor data into an image showing the
stretch of the sensor as a function of time (see example in
Figure 3, and Matlab code in Supporting Information, Appendix
S1). Afterwards, the mouse cursor was manually used to select
the visually observed time-points that corresponded to the
contraction-specific phases. The start of the concentric phase
was defined as the last data-point before the signal started to
increase, see Figure 4. The end of the concentric phase (the
start of the quasi-isometric phase) was defined as the first
data-point after the signal stopped increasing. The end of the
quasi-isometric phase (start of the eccentric phase) was
defined as the last data-point before the rapid decrease of the
signal. The end of the eccentric phase was defined as the first
data point where there was no further decrease in the signal. In
some participants, a gradual transition from the end of the
eccentric phase to the pause at 0 degrees abduction was
observed. In these cases, the end of the eccentric phase was
defined as the first data point where the signal returned to the
mean baseline value (visually determined) in between each
The free software programme V1 Home Basic
(www.v1golfacademy.com) was used for playback of the video
recordings. The software programme allows for frame-by-frame
playback of the video recordings and thereby a precise
determination of contraction-phase specific time points. The
starting point of the concentric phase was defined as the first
frame where the hand holding the elastic exercise band moved
in the direction of abduction. The end of the concentric phase
(the start of the quasi-isometric phase) was defined as the first
frame where the hand no longer moved. The end of the
quasiisometric phase (the start of the eccentric phase) was defined
as the first frame where the hand moved in the direction of
adduction. The end of the eccentric phase was defined as the
first frame where the hand no longer moved.
The following variables were determined from both video and
stretch sensor data: (I) Total TUT, defined as the sum of the
concentric, quasi-isometric and eccentric TUT during each set
(10 repetitions) , (II) single repetition TUT, defined as the
duration of each single repetition , and (III)
contractionphase specific TUT, defined as the duration of the concentric
and eccentric phases of each repetition, respectively .
Rating of stretch-sensor data and video recordings were timed
to allow for a comparison of the time-requirements for both
methods. Quasi-isometric TUT was not included in the analysis
as subjects performed the shoulder abduction at a self-selected
speed and most subjects did not perform an quasi-isometric
Data and statistical analyses
Analysis 1: Validity
This analysis was performed to determine if total TUT, single
repetition TUT and contraction-phase specific TUT could be
determined validly from data recorded by the stretch-sensor.
TUT determined from the video recordings was used as the
gold standard. One rater (Rater 1) determined the total,
singlerepetition and contraction-phase specific TUT from all five
exercise sets from three randomly chosen participants, from
both video recordings and stretch-sensor data. This
corresponded to 15 exercise sets consisting of 150
contractionspecific phases per method for both concentric and eccentric
phases. The rater was blinded as to which subject he was
Figure 5. Bland-Altman plots showing the agreement between stretch-sensor recordings and video recordings.
concentric contraction-phase. B: eccentric contraction-phase C: single repetition. D: total time-under-tension:.
rating and which of the five exercise sets he was rating. Firstly,
all rating of video recordings was done, followed by rating of
the stretch-sensor data. During rating of the video recordings
and stretch-sensor data, the rater was randomly presented with
one of the five exercise sets from the three different
participants, until all 15 exercise sets had been rated.
For the statistical analyses of validity, paired t-tests were
used to test for systematic bias between the two methods.
Systematic bias is reported as the mean difference between
methods and graphically presented as Bland-Altman plot in
Figure 5. Pearson correlation coefficients were used to express
the degree of linear association between the two methods.
Finally, Limits of Agreements (LoA) were used to express the
agreement between the two methods . The LoA were
calculated as the mean difference between methods 1.96
times the standard deviation of the differences in total, single
repetition and contraction-phase specific TUT between
stretchsensor data and video recordings. The LoA were presented as
a range indicating the maximal potential difference between the
two methods in 95% of the ratings and relative agreement was
calculated as mean TUT divided by LoA .
Heteroscedasticity was visually assessed and there we no
trends towards heteroscedasticity.
Analysis 2: Interrater reliability of stretch-sensor data
This analysis was performed to determine the interrater
reliability and agreement between independent raters, rating
the stretch-sensor data. Raters 1 and 2 rated all 60 exercise
sets from all 12 participants and determined TUT for each of
the 600 single repetitions.
For the statistical analysis, a two-way random effects model
(2.1), single measures, absolute agreement, and intraclass
correlation coefficients (ICC) were used to express interrater
reliability, while LoA were used to express agreement between
raters . LoAs were calculated and presented in the same
way as in the validity analysis, but based on the difference in
TUT between raters. The reliability analysis included total TUT
for all 10 repetitions (n=60), total concentric and eccentric
contraction-phase specific TUT for all 10 repetitions in each
exercise set (n=60), single repetition TUT (n=600) and
contraction-specific TUT for all repetitions (n=600).
We used a non-inferiority sample size calculation to
determine the sample-size required for the validity analysis.
This approach was chosen, as the primary goal of the validity
analysis was to show that there were no clinically relevant
differences between total TUT determined from video
recordings and stretch-sensor data. The data used for the
sample-size calculation were collected in a pilot study. Using a
non-inferiority limit of 2 seconds of total TUT (3.3% of the total
TUT in the pilot study) and a standard deviation of 1.5 seconds
at 5% significance and 80% power, it was necessary to rate 14
exercise sets. We chose to increase the number of exercise
sets to 15, as this enabled us to use all exercise sets from
three randomly selected participants. The sample-size for the
Table 1. Agreement between total, single repetition, and contraction-phase specific time-under-tension, determined from
video recordings and stretch-sensor data.
Table 2. Interrater reliability and 95% limits of agreement between Raters 1 and 2 for total, single and contraction-phase
reliability analysis included 60 data points for total TUT, 60 data
points for total concentric and eccentric contraction-phase and
600 data points for single repetition TUT and contraction
specific TUT .
Analysis 1: Validity
Total TUT, single repetition TUT and contraction-phase
specific TUT-determined from video recordings and stretch
sensor data were highly correlated (r>0.99). Moreover, the
agreement between the two methods was high, as the LoA
ranged from 0.0 to 3.1 seconds for total TUT (2.6% of mean
TUT), from -0.26 to 0.56 seconds for single repetition TUT
(6.9% of mean TUT), and from -0.29 to 0.56 seconds for
contraction-phase specific TUT (13.2-21.1% of mean TUT),
Table 1 Figure 5. Generally, TUT determined from stretch
sensor data was systematically of longer duration, compared
with those determined from video recordings (p < 0.00001).
These systematic differences ranged from 1.56 seconds for
total TUT (2.8%), 0.16 seconds (2.8%) for single repetition TUT
and from 0.090.20 seconds (4.7-6.9%) for contraction-phase
Analysis 2: Interrater reliability
The ICC for interrater reliability for total, single repetition and
contraction-phase specific TUT was above 0.99, Table 2. The
agreement between raters for total TUT ranged from -2.11 to
2.56 seconds (4.7% of mean TUT). The agreement for single
repetitions was between -0.46 to 0.50 seconds (9.7% of mean
TUT) and from -0.41 to 0.44 seconds for the contraction
phases (5.2-14.5% of mean TUT). There were small systematic
differences between raters (p<0.05) which ranged from 0.02
seconds (0.4%) for single repetition TUT to 0.04 seconds
(2.2%) at total concentric TUT.
Time requirements for rating video recordings and
Both contraction-phase specific TUT and total TUT from
video recordings took significantly longer time to calculate
compared to calculating TUT when using stretch-sensor data
(p<0.0001). Rating the video recordings of 10 repetitions of
contraction-phase specific TUT took, on average, 17: 35
minutes. In comparison, it took, on average, 55 seconds to
determine total TUT for 10 repetitions from stretch-sensor
recordings and 3: 10 minutes to rate contraction-phase specific
TUT for 10 repetitions.
Currently, there are no tools available to determine
timeunder-tension during home-based exercises, performed using
elastic exercise bands. Such a tool is especially relevant, as
patients most often perform home-based exercises alone,
without being under the surveillance of the prescribing
physician or therapist. Therefore, we investigated if data from a
new stretch-sensor attached to a standard elastic exercise
band could be used to validly and reliably calculate total TUT,
single TUT and contraction-phase specific TUT. The current
data show that total and single repetition TUT can be validly
and reliably determined from stretch-sensor data. We
previously validated the ability of this tool to identify specific
exercise scenarios with different ranges of motion and
contraction speeds when subjects performed dynamic shoulder
abductions at a relative intensity of 12 repetition maximum
(RM). Combining TUT with range of motion, and relative
intensity determined by a test of repetitions maximum allows
for a detailed description of the specific strength training
Practical relevance of the results
For illustration purposes, imagine two subjects, both enrolled
in a research study investigating the effectiveness of a
homebased strength-training intervention. Previous work indicates
that total TUT in addition to concentric and eccentric
contractions are important for the intervention to be able to
induce a clinical effect [28,29]. The eccentric contractions, in
particular, seem to be important [18,19]. The two subjects are
both instructed to follow the exercise prescription of three
training sessions per week. Each prescribed training session
should consist of shoulder abduction from 0 degrees of
abduction to 90 degrees of abduction in three sets of 10
repetitions at twelve RM. Each prescribed repetition should
consist of a three seconds concentric phase, two seconds
quasi-isometric phase at 90 degrees of shoulder abduction,
and finally three seconds of eccentric phase.
Both patients adhere to the prescribed programme and
relative load, but one patient does perform the exercises too
fast; two seconds of concentric phase, no quasi-isometric
phase at 90 degrees, and then a one second eccentric phase.
After 12 weeks of the programme, both patients will have
performed an identical number of exercise sets and repetitions
and the patient who adhered to the prescribed TUT will have
performed exercises corresponding to a TUT of 8640 seconds
(144 minutes). However, the patient who did not adhere to the
prescribed TUT, will have performed exercises corresponding
to only 37.5% of the total training stimulus and 33.3% of the
eccentric stimulus of the patient who adhered. It seems
plausible that such variations in training volume (TUT) and
training specificity (contraction-phase specific TUT) may lead to
differences in the physiological and clinical response even
though the same number of repetitions were performed by the
two patients [16,2830]. If exercise adherence were recorded
using an exercise diary in the theoretical example above, both
patients would have appeared equally adherent to the
prescribed exercise programme.
If instead the patients had trained using the elastic exercise
band and stretch-sensor investigated in the current study, the
treating physician or physiotherapist could have measured TUT
for sets, single repetitions and specific contraction-phases.
Afterwards, it would have been possible to calculate how close
the measured TUT corresponded to e.g. the prescribed
eccentric TUT and thereby quantify adherence as a percentage
of the prescribed total eccentric TUT during the programme.
Another way of measuring TUT during home-based
exercises is to use webcams connected to a computer. This
approach has been used in the field of tele-rehabilitation of
patients with stroke [31,32]. However, this solution requires that
patients remember to turn on the webcams before starting to
exercise. Moreover, a computer and knowledge on how to
operate it is mandatory. Even if a computer solution could be
established, our data showed that determining TUT from video
recordings is much more time-consuming than from
stretchsensor data. Accelerometer based solutions could also be used
in some situations where the external load (for example a
dumbbell) is known . This allows for a detailed description
of measuring force and power, which are also important
strength training descriptors . The stretch-sensor attached
to an exercise-band seems to be extremely user-friendly, as it
does not require specific equipment and skills, data collection
is fully automatic and is exercise-integrated. In addition, it
allows for identification of specific shoulder abduction exercise
scenarios with different ranges of motion and contraction
speeds when subjects performed dynamic shoulder abductions
at a relative intensity of 12 repetition maximum (RM). Finally,
and of great importance, the investigated method allow patients
to perform prescribed exercises away from home, such as at
work or when traveling.
Improving reliability and agreement
All ratings were done manually, using the definition of
concentric, quasi-isometric and eccentric phases, stated in
Methods, and visualised in Figures 3 and 4. The transition
from the end of the eccentric phase to the pause at 0 degrees
abduction is often a gradual transition. In some cases, this
makes it difficult to pinpoint the exact end of the eccentric
phase. This may also be the reason why we observed the
lowest reliability and agreement for single repetition eccentric
TUT. However, some of this error is apparently reduced when
summarising the single contraction TUT to TUT for a set of 10
repetitions, as the reliability for eccentric TUT in a set of 10
repetitions increases considerably. The clinical implication of
this finding is that when using the stretch-sensor to investigate
the adherence to prescribed eccentric TUT in physical
medicine and rehabilitation, one should use the total eccentric
TUT as opposed to evaluate each individual eccentric TUT for
A possible solution to increase reliability of the TUT of the
in more detail. Based on the research conducted in healthy
eccentric phase could be to apply an automatic algorithm to
detect the time spent in each of the three contraction-phases.
physiological stimulus, may hold true for patients as well. Using
One could also argue that the achieved degree of reliability and
a stretch-sensor attached to an elastic exercise band would
agreement using the manual rating procedure leaves little room
allow for similar studies to be conducted in various patient
for improvement, other than using the same rater for multiple
ratings from the same patient. The reason for this is that
intratester ratings are generally reported to be more reliable than
inter-tester rating . However, a likely benefit from using an
automatic algorithm would be decreased time-requirements for
rating data. The current manual detection took on average 3:
10min, while an automatic algorithm would likely decrease
rating to a few seconds.
As a clinical gold standard for measuring TUT, we used
video recordings with a sampling rate of 24 Hz. This is lower
than the 200 Hz, used by the stretch sensor. It may have
introduced some random error in the determination of the start
and stopping points of the contraction phases, as the
may initiate between two image frames. The
position transducers, used by Tran et al.  to measure TUT,
may have introduced less random error. However, given the
high degree of agreement between procedures found in the
transducers, as opposed to standard video recordings, would
most likely not be of clinical relevance
The example of the two patients above appearing to
demonstrate equal adherence to the exercise programme,
even though they display a large difference in total and
eccentric TUT during an intervention period, suggests that
future research should aim to describe adherence to exercise
band can be used to calculate total and single repetition times
under tension. The procedure is valid and highly reliable
between raters. The current method will enable clinicians and
researchers to objectively quantify if home-based exercises are
performed correctly with respect to prescribed contraction
phases and time-under-tension.
Appendix S1. Matlab code for generating images as seen
in Figures 3 and 4. (DOCX)
Camilla Rams Rathleff is acknowledged for her help in rating
data. Hans Jrgen Krebs, Louise Simonsen and Julie Louise
Kristensen are acknowledged for their help in collecting data.
Conceived and designed the experiments: MSR KT TB.
Performed the experiments: MSR KT TB. Analyzed the data:
MSR KT TB. Contributed reagents/materials/analysis tools:
MSR KT TB. Wrote the manuscript: MSR KT TB.
7. Andersen LL , Saervoll CA , Mortensen OS , Poulsen OM , Hannerz H et al. ( 2011 ) Effectiveness of small daily amounts of progressive resistance training for frequent neck/shoulder pain: randomised controlled trial . Pain 152 : 440 - 446 . doi:10.1016/j.pain. 2010 .11.016. PubMed: 21177034 .
8. McCarthy CJ , Mills PM , Pullen R , Roberts C , Silman A et al. ( 2004 ) Supplementing a home exercise programme with a class-based exercise programme is more effective than home exercise alone in the treatment of knee osteoarthritis . Rheumatology (Oxford) 43 : 880 - 886 . doi:10.1093/rheumatology/keh188. PubMed: 15113993 .
9. Dunstan DW , Daly RM , Owen N , Jolley D , Vulikh E et al. ( 2005 ) Homebased resistance training is not sufficient to maintain improved glycemic control following supervised training in older individuals with type 2 diabetes . Diabetes Care 28 : 3 - 9 . doi:10.2337/diacare. 28.suppl_1. S3 . PubMed: 15616225.
10. Ravaud P , Giraudeau B , Logeart I , Larguier JS , Rolland D et al. ( 2004 ) Management of osteoarthritis (OA) with an unsupervised home based exercise programme and/or patient administered assessment tools. A Clusters Randomised Control Trial With A 2x2 factorial design . Annals of the Rheumatic Diseases 63 : 703 - 708 .
11. McLean SM , Burton M , Bradley L , Littlewood C ( 2010 ) Interventions for enhancing adherence with physiotherapy: a systematic review . Man Therapy 15 : 514 - 521 . doi:10.1016/j.math. 2010 .05.012. PubMed: 20630793 .
12. Rathleff MS , Bandholm T , Ahrendt P , Olesen JL , Thorborg K ( 2013 ) Novel stretch-sensor technology allows quantification of adherence and quality of home-exercises: a validation study . Br J Sports Med. PubMed: 23467964.
13. Toigo M , Boutellier U ( 2006 ) New fundamental resistance exercise determinants of molecular and cellular muscle adaptations . Eur J Appl Physiol 97 : 643 - 663 . doi:10.1007/s00421- 006 - 0238 -1. PubMed: 16845551 .
14. Mohamad NI , Nosaka K , Cronin J ( 2011 ) Maximizing Hypertrophy: Possible Contribution of Stretching in the Interset Rest Period . Strength Conditioning J 33: 81 . doi:10.1519/SSC.0b013e3181fe7164.
15. Schoenfeld BJ ( 2010 ) The mechanisms of muscle hypertrophy and their application to resistance training . J Strength Cond Res , 24 : 2857 - 72 / National Strength & Conditioning Association 24 : 2857 - 2872 . PubMed: 20847704
16. Burd NA , Andrews RJ , West DW , Little JP , Cochran AJ et al. ( 2012 ) Muscle time under tension during resistance exercise stimulates differential muscle protein sub-fractional synthetic responses in men . J Physiol 590 : 351 - 362 . PubMed: 22106173 .
17. Tran QT , Docherty D , Behm D ( 2006 ) The effects of varying time under tension and volume load on acute neuromuscular responses . Eur J Appl Physiol 98 : 402 - 410 . doi:10.1007/s00421- 006 - 0297 -3. PubMed: 16969639 .
18. Kaux JF , Drion P , Libertiaux V , Colige A , Hoffmann A et al. ( 2013 ) Eccentric training improves tendon biomechanical properties: A rat model . J Orthop Res Off Publ Orthopaedic Research Society , 31 : 119 - 24 . PubMed: 22847600 .
19. Heinemeier KM , Olesen JL , Schjerling P , Haddad F , Langberg H et al. ( 2007 ) Short-term strength training and the expression of myostatin and IGF-I isoforms in rat muscle and tendon: differential effects of specific contraction types . J Appl Physiol 102 : 573 - 581 . PubMed: 17038487 .
20. Heinemeier KM , Olesen JL , Haddad F , Langberg H , Kjaer M et al. ( 2007 ) Expression of collagen and related growth factors in rat tendon and skeletal muscle in response to specific contraction types . J Physiol 582 : 1303 - 1316 . doi:10.1113/jphysiol.2007.127639. PubMed: 17540706 .
21. Malliaras P , Barton CJ , Reeves ND , Langberg H ( 2013 ) Achilles and patellar tendinopathy loading programmes : a systematic review comparing clinical outcomes and identifying potential mechanisms for effectiveness . Sports Med 43 : 267 - 286 . doi:10.1007/s40279- 013 - 0019 - z. PubMed: 23494258.
22. Roig M , O'Brien K , Kirk G , Murray R , McKinnon P et al. ( 2009 ) The effects of eccentric versus concentric resistance training on muscle strength and mass in healthy adults: a systematic review with metaanalysis . Br J Sports Med 43 : 556 - 568 . doi:10.1136/bjsm.2008.051417. PubMed: 18981046 .
23. Kottner J , Audige L , Brorson S , Donner A , Gajewski BJ et al. ( 2011 ) Guidelines for Reporting Reliability and Agreement Studies (GRRAS) were proposed . Int J Nurs Stud 48 : 661 - 671 . doi:10.1016/j.ijnurstu. 2011 .01.016. PubMed: 21514934 .
24. Kappel SL , Rathleff MS , Hermann D , Simonsen O , Karstoft H et al. ( 2012 ) A Novel Method for Measuring In-Shoe Navicular Drop During Gait . Sensors.
25. Walther M , Werner A , Stahlschmidt T , Woelfel R , Gohlke F ( 2004 ) The subacromial impingement syndrome of the shoulder treated by conventional physiotherapy, self-training, and a shoulder brace: results of a prospective, randomized study . J Shoulder Elbow Surg / American Shoulder and Elbow Surgeons [et al] 13 : 417 - 423
26. Bland JM , Altman DG ( 1986 ) Statistical methods for assessing agreement between two methods of clinical measurement . Lancet 1 : 307 - 310 . PubMed: 2868172 .
27. Hopkins WG ( 2000 ) Measures of reliability in sports medicine and science . Sports Med 30 : 1 - 15 . doi: 10.2165/ 00007256 - 200030010 -00001. PubMed: 10907753 .
28. Yu J , Park D , Lee G ( 2013 ) Effect of Eccentric Strengthening on Pain , Muscle Strength , Endurance, and Functional Fitness Factors in Male Patients with Achilles Tendinopathy . Am J Phys Med Rehabil , 92 : 68 - 76 / Association of Academic Physiatrists. PubMed: 23044702
29. Mafi N , Lorentzon R , Alfredson H ( 2001 ) 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 9 : 42 - 47 . doi: 10.1007/s001670000148. PubMed: 11269583 .
30. Buitrago S , Wirtz N , Yue Z , Kleinder H , Mester J ( 2012 ) Effects of load and training modes on physiological and metabolic responses in resistance exercise . Eur J Appl Physiol 112 : 2739 - 2748 . doi:10.1007/ s00421- 011 - 2249 -9. PubMed: 22116573 .
31. Or CK , Karsh BT , Severtson DJ , Burke LJ , Brown RL et al. ( 2011 ) Factors affecting home care patients' acceptance of a web-based interactive self-management technology . J Am Med Inform Assoc 18 : 51 - 59 . doi:10.1136/jamia.2010.007336. PubMed: 21131605 .
32. Brown SH , Lewis CA , McCarthy JM , Doyle ST , Hurvitz EA ( 2010 ) The effects of Internet-based home training on upper limb function in adults with cerebral palsy . Neurorehabil Neural Repair 24 : 575 - 583 . doi: 10.1177/1545968310361956. PubMed: 20581338 .
33. Comstock BA , Solomon-Hill G , Flanagan SD , Earp JE , Luk HY et al. ( 2011 ) Validity of the Myotest(R) in measuring force and power production in the squat and bench press . J Strength Conditioning Res / National Strength & Conditioning Association 25 : 2293 - 2297
34. Eliasziw M , Young SL , Woodbury MG , Fryday-Field K ( 1994 ) Statistical methodology for the concurrent assessment of interrater and intrarater reliability: using goniometric measurements as an example . Phys Ther 74 : 777 - 788 . PubMed: 8047565 .