Received: 14 September 2023

| Revised: 12 December 2023

| Accepted: 18 December 2023

DOI: 10.1111/sms.14555

ORIGINAL ARTICLE

Personalized tendon loading reduces muscle-­tendon

imbalances in male adolescent elite athletes

Theresa Domroes1,2 |Kolja Weidlich1,2 | Sebastian Bohm1,2 |

Falk Mersmann1,2 | Adamantios Arampatzis1,2

1
Department of Training and
Movement Sciences, Humboldt-­ Abstract
Universität zu Berlin, Berlin, Germany An imbalanced adaptation of muscle strength and tendon stiffness in response
2
Berlin School of Movement Science, to training may increase tendon strain (i.e., the mechanical demand on the ten-
Humboldt-­Universität zu Berlin, Berlin,
Germany
don) and consequently tendon injury risk. This study investigated if personalized
tendon loading inducing tendon strain within the effective range for adapta-
Correspondence tion (4.5%–6.5%) can reduce musculotendinous imbalances in male adolescent
Adamantios Arampatzis, Department
of Training and Movement Sciences, handball athletes (15–16 years). At four measurement time points during a com-
Humboldt-­Universität zu Berlin, petitive season, we assessed knee extensor muscle strength and patellar tendon
Philippstr 13, Haus 11, 10115 Berlin,
mechanical properties using dynamometry and ultrasonography and estimated
Germany.
Email: [email protected] the tendon's structural integrity with a peak spatial frequency (PSF) analysis of
proximal tendon ultrasound scans. A control group (n = 13) followed their usual
Funding information
training routine, an intervention group (n = 13) integrated tendon exercises into
German Federal Institute of Sport
Science their training (3x/week for ~31 weeks) with a personalized intensity correspond-
ing to an average of ~6.2% tendon strain. We found a significant time by group
interaction (p < 0.005) for knee extensor muscle strength and normalized patel-
lar tendon stiffness with significant increases over time only in the intervention
group (p < 0.001). There were no group differences or time-­dependent changes
in patellar tendon strain during maximum voluntary contractions or PSF. At
the individual level, the intervention group demonstrated lower fluctuations of
maximum patellar tendon strain during the season (p = 0.005) and a descriptively
lower frequency of athletes with high-­level tendon strain (≥9%). The findings sug-
gest that the personalized tendon loading program reduced muscle-­tendon im-
balances in male adolescent athletes, which may provide new opportunities for
tendon injury prevention.

KEYWORDS
adaptation, overuse, tendinopathy, tendon training, youth athletes

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2024 The Authors. Scandinavian Journal of Medicine & Science In Sports published by John Wiley & Sons Ltd.

Scand J Med Sci Sports. 2024;34:e14555.  wileyonlinelibrary.com/journal/sms | 1 of 14

https://doi.org/10.1111/sms.14555
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
2 of 14    DOMROES et al.

1 | I N T RO DU CT ION homogenous adaptation of muscle strength and tendon

stiffness. In case of increased tendon strain, this would
A well-­coordinated interplay of muscle and tendon is es- specifically entail loading programs that primarily target
sential for an effective execution of various movements in tendon adaptation.
sports and daily life.1–3 While muscles generate the neces- While muscles are able to adapt to a variety of mechan-
sary forces, tendons enable the storage and return of elas- ical and metabolic stimuli,22 it seems that tendon adapta-
tic strain energy during movement.3,4 Tendon elasticity tion requires a certain strain magnitude. In vitro findings
further influences the muscle's operating length and veloc- suggest that a mechanical over-­stimulation as well as an
ity and hence the force-­length-­velocity and power-­velocity under-­stimulation of tendon tissue may induce catabolic
potential as well as the efficiency-­velocity relationship of processes and matrix deterioration and that there only ex-
the muscle.1,5,6 On the other hand, tendons are quite sus- ists a certain strain range (i.e., a “sweet spot” of tendon
ceptible to overuse injuries and there is a high prevalence strain magnitude) in between that can induce net anabolic
of tendinopathies in adult as well as youth athletes.7–9 processes in the tendon.13,23 A series of systematic in vivo
One underlying factor for tendon overuse may be im- intervention studies demonstrated that cyclic loading of
balances in the changes of muscle strength and tendon the tendon with a strain magnitude between 4.5% and
mechanical properties.10 If the force generating capacity 6.5% caused improvements of tendon mechanical prop-
of the muscle increases without concomitant increases in erties, while no improvements were observed at ~3% of
the respective tendon stiffness, this results in increased strain.24–26 This tendon strain range was mostly achieved
tendon strain. Though the exact strain level for tendon at 90% of isometric maximum voluntary contractions
failure is unknown in vivo, it seems to be quite constant (MVC). This may, however, not apply to all individuals.
across tendons.11 Therefore, any increase in tendon op- Especially in populations with a high prevalence of mus-
erating strain increases the mechanical demand on the culotendinous imbalances, there exists a high variability
tendon tissue and reduces the tendon safety factor (i.e., in tendon strain during MVCs or one repetition maximum
ratio of operating tendon strain to ultimate tendon strain). (1RM) loading, so that tendon exercises at the same load
Accordingly, it has been shown in vitro that the initial relative to the MVC or 1RM may result in tendon strain
level of tendon strain during static and cyclic loading outside the effective region for tendon adaptation.27,28
was the best predictor of time or cycles to failure12 and Tendon exercises at high strain above the “sweet spot”
that tissue damage and catabolic processes outweigh tis- region may thereby even have harmful consequences
sue repair mechanisms at continuous cyclic loading with for the tendon (i.e., increase the risk for tendinopathy).21
tendon strain levels as high as 9%.13 In vivo, imbalances Accordingly, a prescription of exercise load based on an
between muscle strength and tendon stiffness have been MVC or 1RM alone does not seem to be accurate enough
observed in athletes from child-­to adulthood,14–18 result- to specifically promote tendon adaptation, particularly in
ing in high frequencies of increased tendon strain levels athletes with musculotendinous imbalances, as these af-
(i.e., an increased mechanical demand on the tendon) as fect the relationship between an MVC or 1RM and tendon
well as high fluctuations of tendon strain during a season. strain.27,28 A load prescription that considers the individ-
Thereby, an inverse relationship between tendon strain ual tendon strain during loading is consequently an in-
and the structural integrity of the proximal patellar ten- novative approach for targeted tendon adaptation, which
don was reported for male adolescent athletes, indicat- has to our knowledge not been investigated in vivo before.
ing a local disorganization of the collagenous structure With this approach, individuals with high maximum ten-
in tendons that are subjected to high tendon strain.19,20 don strain (i.e., muscle-­tendon imbalances) would train at
Further, we could recently demonstrate in a prospective lower relative intensities (e.g., a lower percentage of MVC
longitudinal study that the risk for developing tendon or 1RM) to reach the “sweet spot” for tendon adaptation,
pain is increased 2.3-­fold in athletes with tendon strain so that it can be expected that tendon adaptation is pro-
levels ≥9% during maximum effort muscle contractions.21 moted to a higher extent compared to muscle adaptation,
Considering that tendinopathies constitute a considerable which would in turn reduce musculotendinous imbal-
burden for athletic populations and that their prevalence ances. In individuals with lower maximum tendon strain
seems to increase from child-­to adulthood,8,9 early pre- (i.e., without muscle-­tendon imbalances) on the other
ventive approaches to reduce the mechanical demand on hand, the higher training intensities are presumably suit-
the tendon may have the potential to reduce tendon injury able to induce a balanced adaptation of both muscle and
risk and to restore a well-­coordinated interaction between tendon.27
muscle and tendon in athletes. This may be achieved by Therefore, the aim of the current study was to investi-
introducing exercises into the training routine that coun- gate the effects of a tendon strain-­based personalized load-
teract musculotendinous imbalances by promoting a more ing program on patellar tendon adaptation, muscle-­tendon
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    3 of 14

imbalances and tendon micromorphological structure 10 weeks between measurements in the control group and
during a competitive season in male adolescent elite 14, 14, and 6 weeks in the intervention group. The train-
handball athletes, as it has been previously shown that ing intervention started approximately 3 weeks after M1
musculotendinous imbalances are quite common in this so that the intervention duration was on average 31 weeks.
population.18,29 We hypothesized that fluctuations of pa- All measurements were performed on the dominant leg,
tellar tendon strain over the season as an indicator for which was defined as the leg used to kick a ball. Biological
individual imbalances in the adaptation of muscle and maturity of the participants was predicted as the years to
tendon will be lower in the intervention group in compar- peak height velocity (PHV) using age and sitting height in
ison to the control group. the recalibrated prediction equation suggested by Moore
et al.30 The prevalence of patellar tendon pain and pain-­
related disability was assessed with the validated German
2 | M ET H ODS version of the VISA-­ P questionnaire31 considering the
symptoms of the past 2 weeks. Participants were catego-
2.1 | Participants and experimental rized as symptomatic with a score ≤ 87 points, as this rep-
design resents the minimum clinically important difference31
from the maximum score of 100 points.
To investigate the effects of a personalized tendon exercise Two participants of the control group were excluded
intervention, we recruited male adolescent elite handball from the study due to unavoidable movement artifacts in
players (aged 15–16 years) for a control and an interven- the ultrasound recordings of the tendon, which did not
tion group. The necessary sample size was calculated in allow an accurate analysis of tendon properties. One par-
a power analysis (G*Power, version 3.1.9.7). Based on a ticipant of the intervention group did not participate in
large effect (d = 1.4), which we observed previously for pre-­ M2, one participant of each group did not participate in
post changes in tendon stiffness after a non-­personalized M3 and M4, leading to a total of 50 observations in the
tendon exercise intervention in adolescent athletes,20 and control group and 49 observations in the intervention
a power of 0.9, we calculated a sample size of eight per group.
group. Considering a potential drop-­out, 15 participants
were recruited for the control group. In the following sea-
son, 13 athletes from the same team (i.e., same age group), 2.2 | Assessment of tendon
including four athletes that already participated in the micromorphology
control group, were recruited for the intervention group.
All participants had a weekly training duration of at least The assessment of tendon micromorphology was based
10 h (approx. 60% handball specific training, 40% strength on a spatial frequency analysis of ultrasound recordings of
and athletic training; excluding competition). Exclusion the proximal part of the patellar tendon. The ultrasound
criteria were any neurological or musculoskeletal impair- recordings took place prior to the measurement of muscle
ments of the lower extremity. Participants with patellar strength and tendon mechanical properties to avoid the
tendon pain were included if maximum voluntary muscle influence of possible acute loading-­related responses. The
strength testing was possible. All participants and their participants were positioned supine with their knee flexed
legal guardians gave written informed consent to the ex- to 90° (measured with a goniometer), as tendon slack is
perimental procedures, which were approved by the eth- commonly removed in this angle, while strain induced by
ics committee of the Humboldt-­Universität zu Berlin (ref. passively generated forces is still low.19
no. HU-­KSBF-­EK_2020_0005). Two short sequences were captured with an ultrasound
The control group received only their sport-­specific probe (ultrasound system: MyLab60; Esaote, Genova, Italy;
training, while the intervention group integrated a person- probe: linear array LA523, 13 MHz, depth: 3.0 cm) placed
alized tendon loading program into their regular training. over the patellar tendon parallel to its longitudinal axis
In both groups, knee extensor muscle strength, patellar below the most distal apex of the patella. Images were an-
tendon mechanical properties and micromorphology as alyzed with a custom-­written MATLAB interface (version
well as the prevalence of tendon pain and disability were R2016a; MathWorks, Natick, MA, United States). A polyg-
assessed at four measurement time points during a com- onal region of interest (ROI) was defined, spanning from
petitive season (M1, M2 and M3: in-­season; M4: transition the deep insertion of the patella to the central portion of
period). The exact time span between measurements dif- the tendon with a length of 40% of the tendon's rest length
fered between groups and individuals due to constraints (measured as described below) and a height that covered
in the scheduling of the measurements (e.g., holidays, the full thickness of the tendon. All 32 × 32 pixel kernels
training camps, tournaments) with on average 12, 13, and contained within the ROI were analyzed as suggested by
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
4 of 14    DOMROES et al.

Bashford and colleagues.32 In short, a 2D Fast Fourier trans- direction of force application perpendicular to the shank
form and a high-­pass filter with a radial frequency response and that the knee joint angle reached 60° during contrac-
and half-­power cutoff frequency of 1.23 mm−1 were applied. tions (0° represents full knee extension; measured with a
The filtered kernels were zero padded in both directions to a goniometer), as this is approximately the optimal angle for
size of 128 × 128 pixels to increase frequency resolution. The maximum force generation of the knee extensors.33 Ten
distance of the peak spatial frequency (PSF) to the spectral submaximal isometric knee extension contractions with
origin in the frequency spectrum of all analyzed kernels was increasing effort served as additional warm-­ up, famil-
averaged and the mean PSF of both recorded trials was then iarization and preconditioning of the tendon. Afterwards,
used for further analysis. Thereby, low PSF values represent the participants performed two MVCs with 2–3 min rest
a more isotropic speckle pattern in the ultrasound images, between trials. Force data was recorded with a custom-­
indicating a low packing density and alignment of the colla- written MATLAB interface (version R2016a) at 200 Hz
gen bundles as is common for tendinopathic tendons.32 and filtered with a second-­order Butterworth filter with a
cutoff frequency of 6 Hz. Subsequently, a moving average
filter with a window size of 50 ms was used for the deter-
2.3 | Assessment of knee extensor mination of maximum force values. The recorded force
muscle strength was then multiplied with the lever arm of the applied force
(i.e., distance from the lateral epicondyle to the middle
Muscle strength of the knee extensors was assessed as of the strap around the shank) to get the knee joint mo-
the maximum knee extension moment during isometric ment, which was corrected for moments of gravity based
MVCs using a custom-­built mobile measurement system on the data provided by Dempster34 for the estimation of
(Figure 1). After a self-­selected warm-­up which included the mass and center of mass of the foot and shank. The
submaximal and five maximal jumps, the participants MVC-­trial with the highest moment was then used for fur-
were fixed in a seated position with a nonelastic belt with ther analysis.
a trunk angle of about 90°. A force sensor (2 kN; Biovision,
Wehrheim, Germany) in series with a non-­elastic rope was
attached with a strap to the participants' lower shank. The 2.4 | Assessment of tendon
rope was adjusted so that the force sensor was aligned in the mechanical properties

The force-­elongation relationship of the patellar tendon

was determined by recording tendon elongation with ul-
trasound (MyLab60; Esaote, Genova, Italy) during five
isometric ramp contractions in the same setup used for
the MVCs. During the contractions, the participants re-
ceived visual feedback to increase effort steadily from
rest to maximum within 5 s. Tendon elongation was re-
corded by fixing a 10 cm ultrasound probe (probe: LA923,
7.5 MHz; 25 Hz image frequency) over the longitudinal
axis of the patellar tendon with a modified knee brace. A
manually released 5 Volt trigger signal was used to syn-
chronize the ultrasound and force data. The displacement
of the deep insertion of the tendon at the patellar apex
and the tibial tuberosity in the ultrasound images was
tracked with a semi-­automatic tracking software (Tracker
F I G U R E 1 Mobile measurement system. The force generated Video Analysis and Modeling Tool V. 5.1.5; Open Source
during isometric knee extensions at 60° knee joint angle (0° Physics, Aptos, CA, United States) to determine tendon
represents full knee extension) was recorded with a force sensor.
elongation. Tendon rest length was determined from the
To obtain the forces acting on the tendon, measured forces were
ultrasound images as the length of the slack tendon prior
multiplied with the lever arm of the applied force (i.e., distance
to the onset of the contraction (i.e., in a relaxed state at 60°
from the lateral epicondyle to the middle of the strap around
the shank), corrected for moments of gravity and divided by knee joint angle). It was measured with a spline fit through
the patellar tendon moment arm (based on anthropometry). the deep insertion marks and two additional points along
Patellar tendon elongation during contractions was captured with the border of the tendon and averaged for all five trials.
ultrasound and time synchronized with a manually released 5 Volt The forces acting on the tendon were obtained by divid-
trigger signal. ing the knee joint moment by the patellar tendon moment
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    5 of 14

arm which was estimated based on anthropometry14 and evidence-­based effective strain magnitude (i.e., 4.5%–6.5%
adjusted to the knee joint angle with the data reported by tendon strain24,25) for tendon adaptation. Accordingly, the
Herzog and Read.35 The force-­elongation relationship for training load was personalized for each participant based
each single ramp trial was saved as piecewise-­linear poly- on their maximum patellar tendon strain during MVCs
nomial function in MATLAB (version R2019b; “interp1” (calculated as described above). With the simplified as-
function with “pp” option), using all the measured data sumption of a linear relationship between tendon force
of the respective ramp as input. To achieve a high reli- and elongation, we determined the relative force needed
ability and observer independence,36 we averaged the five for each participant to induce 5.5% of tendon strain. The
force-­elongation curves. Therefore, at each measured data maximum relative training load was limited to 90% of
point, all single ramp trials were evaluated using their re- maximum force to ensure an accurate execution of the
spective piecewise-­linear polynomials and then averaged training protocol. This means that all participants who
(using the “fncmb” function in MATLAB). The resulting would, according to our calculation, need a training in-
mean piecewise-­linear polynomial was evaluated up to tensity above 90% MVC to reach 5.5% tendon strain (i.e.,
80% of maximum tendon force (i.e., the force calculated participants with maximum tendon strain ≤6.1%), also
from the highest MVC trial of each participant), as this trained at 90% of their MVC, which has been shown to be
was the highest relative force all participants were capa- an adequate stimulus for the balanced adaptation of both
ble of achieving during the ramp contractions. Finally, muscle and tendon.24,25
a second-­order polynomial fit passing through zero was To ensure an adequate transferability of the strain val-
applied to the averaged patellar tendon force-­elongation ues determined in the experimental setup and the strain
data points. Patellar tendon stiffness was calculated from achieved during training, isometric knee extensions were
this polynomial function as the quotient between 50% performed in a training device similar to the experimen-
and 80% of the maximum tendon force. Tendon stiffness tal setup (Figure 2A). Therefore, a non-­elastic band was
was further normalized by multiplying it with the tendon fixed to the lower shank and adjusted so that the knee
rest length. Accordingly, normalized tendon stiffness rep- joint angle reached 60° during contractions. A digital scale
resents the slope of the tendon force-­strain relationship. mounted in series with the band served to monitor the
Maximum tendon strain was calculated by extrapolating training load. For the determination of the personalized
tendon elongation to the maximum tendon force based on training load to reach the target patellar tendon strain,
the polynomial fit of the force-­elongation curve and divid- every participant performed an MVC in the training de-
ing it by tendon rest length. vice. From the maximum load recorded in the training
device and the prescribed relative loading intensity (i.e., %
MVC), the absolute load to be applied during training was
2.5 | Tendon exercise intervention then determined for each participant individually. Both
MVCs, in the measurement system and training device,
The participants in the intervention group received an were performed at the same knee joint angle, therefore we
exercise program which was designed to provide an assumed similar muscle forces and tendon strain. Based

F I G U R E 2 (A) Training set-­up and (B) loading profile for the personalized tendon exercises. Isometric knee extension contractions
were conducted at 60° knee joint angle (0° represents full knee extension). The individual load was controlled with a digital scale fixed
to an adjustable non-­elastic band. Training load was prescribed individually from an MVC performed in the training device, that was
adjusted every 2 weeks, and the relative load determined based on tendon strain at the four measurement time points and two additional
measurements (~7-­week intervals).
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
6 of 14    DOMROES et al.

on the most effective loading protocol for the tendon from the respective constants for intercept and slope. In the inter-
our earlier systematic research,24–26 the training consisted vention group, β0 + β2 and β1 + β3 represent the constants for
of five sets of four isometric knee extension contractions intercept and slope, respectively. This allows testing of the
at the individual training load with a contraction duration significance of changes over time in the control group (slope
of 3 s (Figure 2B). The participants were instructed to train constant β1), baseline differences between groups (inter-
both legs with the same relative load and to complete the cept constant β2), and group differences in time-­dependent
training three times per week. The training frequency was changes (slope constant β3; henceforth referred to as time
chosen since we found in previous intervention studies by group interaction). Additionally, we tested separately
with adolescent athletes that tendon exercises performed for the intervention group if the slope differed significantly
three times per week elicited significant adaptations of from zero. Note that age was only tested for baseline differ-
tendon stiffness,20 while no systematic changes of ten- ences. To account for individual and group differences in
don stiffness were observed with training two times per the time span between measurements, the time points of
week.29 the measurements were coded in weeks for each partici-
To account for strength adaptations during the training pant individually. As the inclusion of random effects in the
period, a new MVC was recorded in the training device linear mixed-­effects model accounts for individual differ-
every 2 weeks. With regards to possible changes in tendon ences in intercept and slope, the residuals of the model can
mechanical properties, we additionally determined ten- be used as a measure of individual fluctuations of muscle
don strain (during three isometric ramp contractions) at and tendon properties over time. Hence, we determined the
an interval of about 7 weeks between the regular measure- estimated standard deviations of the residuals of the linear
ment time points to adjust the prescribed training load mixed model for the contextually relevant parameters. To
accordingly. Since M3 and M4 were only 6 weeks apart in assess group differences in the standard deviation of the re-
the intervention group, these intermediate measurements siduals, we formulated two linear mixed models according
were performed between M1, M2 and M3, respectively. to equation 1. One model assumed homogeneity of the re-
sidual variance (no group differences) while the other model
allowed for group differences in the residual variance. These
2.6 | Statistics two models were then compared by a likelihood ratio test.
From differences in this model comparison, it can then be
A linear mixed-­ effects model was formulated using inferred whether the standard deviation of the residuals
the nlme package in RStudio (version 4.1.2; RStudio, differed between groups. To investigate whether there ex-
PBC, Boston, MA, United States) with group-­specific y-­ ists an association between initial tendon strain levels and
intercept, slope and variance of the residuals to analyze changes in tendon stiffness, the correlation of maximum
time-­and group-­dependent changes. The advantage of tendon strain at M1 and relative changes in normalized ten-
linear mixed models is that they can handle missing data don stiffness from M1 to M4 was analyzed for both groups
and are robust against violations of the normality assump- separately using the Pearson correlation coefficient (r). The
tion,37 which was not given for age, body height, body significance level for all statistical tests was set to α = 0.05.
mass, and tendon rest length according to the Shapiro–
Wilk test applied to the normalized residuals. The model
equation was 3 | RESULTS

yij = β0 + β1 tij + gi β2 + gi β3 tij + b0i + b1i tij + rij (1) There were no significant baseline differences between
groups in age (p = 0.464), biological maturity (estimated
where i is the index for participant, j is the index for the as offset from PHV; p = 0.530), body height (p = 0.433),
measurement session, gi = 0 in the control group and 1 in and body mass (p = 0.913; Table 1). Biological maturity,
the intervention group, tij is the time point (in weeks) for body height and mass increased significantly over time in
the measurement session j of the participant i, β0 is the y-­ the control (p < 0.001, p = 0.011 and 0.003) and interven-
intercept constant for the control group, β1 is the slope con- tion group (p < 0.001) with no time by group interaction
stant for the control group, β2 is the y-­intercept constant for (p = 0.328, 0.565, and 0.154). Average VISA-­P scores and
the difference between control and intervention group, β3 the number of symptomatic athletes are shown in Table 1.
is the slope constant for the difference between control and The percentage of symptomatic athletes during the meas-
intervention group, b0i is the participant-­specific y-­intercept urement period ranged from 0% to 38.5% in the control
(random effect), b1i is the participant-­specific slope (random group and 0% to 8.3% in the intervention group. The
effect), and rij is the residual. Accordingly, the terms includ- tendon exercise intensity relative to the individual MVC
ing g equal zero in the control group and β0 and β1 represent in the intervention group was on average (mean ± SD)
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    7 of 14

T A B L E 1 Anthropometric data, average VISA-­P scores and number of symptomatic athletes for both groups and each measurement
time point (M1-­4) over a competitive season. Note: The numbers are mean ± standard deviation of the given experimental data. Maturity
offset refers to the estimated years to peak height velocity. Symptomatic was defined as VISA-­P score ≤ 87.

Control Intervention

M1 M2 M3 M4 M1 M2 M3 M4
(n = 13) (n = 13) (n = 12) (n = 12) (n = 13) (n = 12) (n = 12) (n = 12)
Age (years) 16.0 ± 0.6 16.2 ± 0.6 16.5 ± 0.6 16.7 ± 0.6 15.8 ± 0.6 16.2 ± 0.6 16.3 ± 0.5 16.4 ± 0.5
a,b
Maturity offset (years) 2.45 ± 0.84 2.61 ± 0.79 2.87 ± 0.83 2.97 ± 0.83 2.25 ± 0.65 2.54 ± 0.67 2.70 ± 0.64 2.77 ± 0.71
Body height (cm)a,b 186 ± 7 186 ± 7 187 ± 7 187 ± 8 189 ± 8 189 ± 8 189 ± 8 190 ± 9
a,b
Body mass (kg) 80.4 ± 10.3 81.0 ± 10.2 83.1 ± 10.8 82.5 ± 10.1 80.0 ± 10.6 80.8 ± 11.8 81.8 ± 12.4 82.8 ± 12.2
VISA-­P score 97.5 ± 8.9 93.0 ± 9.5 100 ± 0 100 ± 0 100 ± 0 98.5 ± 4.6 100 ± 0 100 ± 0
No. of sympt. athletes 1 5 0 0 0 1 0 0
(7.6%) (38.5%) (0.0%) (0.0%) (0.0%) (8.3%) (0.0%) (0.0%)
a
Significant change over time in the control group.
b
Significant change over time in the intervention group (p < 0.05); there was no time by group interaction (p > 0.05). Note that age was only tested for baseline
differences between groups.

67% ± 11%, 65% ± 8%, and 61% ± 13% for M1, M2, and time in the control (p = 0.603 and 0.883) and intervention
M3 respectively, with an overall range from 47% to 90%. group (p = 0.967 and 0.364) and no time by group interac-
Based on the experimentally determined force-­elongation tion (p = 0.686 and 0.637).
relationship, this corresponded to a tendon strain during The intervention group demonstrated lower fluctu-
training of 6.1% ± 0.4%, 6.3% ± 0.5%, and 6.2% ± 0.5%, with ations (residuals) of maximum patellar tendon strain
an overall range from 4.9% to 7.0%. (p = 0.005; Figure 3C) during the season. There were no
We found a tendency for a higher maximum knee joint group differences in the fluctuations of maximum knee
moment in the control group compared to the intervention joint moment (p = 0.170, Figure 3A), normalized knee
group at baseline (p = 0.059; Figure 3A) as well as a signifi- joint moment (p = 0.187) and maximum patellar ten-
cantly higher knee joint moment normalized to body mass don force (p = 0.159), as well as normalized patellar ten-
and higher maximum force applied to the tendon in the don stiffness (p = 0.339; Figure 3B) and PSF (p = 0.161;
control group (p = 0.011 and p = 0.031; Table 2). There were Figure 3D). Descriptively, the average frequency of ath-
no significant changes of maximum knee joint moment, letes with high-­level patellar tendon strain (i.e., ≥ 9%) was
normalized knee joint moment and maximum tendon lower in the intervention group compared to the control
force over time in the control group (p = 0.789, 0.740 and group (27% vs. 42%; Figure 4). Further, we found a sig-
0.845), but there was a significant time by group interac- nificant positive correlation between maximum patellar
tion (p < 0.001) and significant increases over time in the tendon strain at M1 and relative changes in normalized
intervention group in all three parameters (p < 0.001). patellar tendon stiffness from M1 to M4 in the interven-
Patellar tendon stiffness (Table 2) and normalized ten- tion group (r = 0.619, p = 0.032) but not in the control
don stiffness (Figure 3B) were significantly higher in the group (r = 0.091, p = 0.778; Figure 5).
control group at baseline (p = 0.021 and 0.015) with no
significant time-­dependent changes in the control group
(p = 0.414 and 0.416). There was a significant time by group 4 | DISC USSION
interaction for both patellar tendon stiffness (p = 0.002)
and normalized patellar tendon stiffness (p = 0.002) and The present study investigated for the first time the effects
a significant increase over time in the intervention group of a personalized tendon loading program on imbalances
(p < 0.001). Patellar tendon rest length and maximum of knee extensor muscle strength and patellar tendon me-
tendon elongation (Table 2) were not significantly differ- chanical properties as well as patellar tendon micromor-
ent between groups (p = 0.953 and 0.235) and there were phology during a competitive season in male adolescent
no significant changes over time in the control group elite handball athletes. As hypothesized, we found sig-
(p = 0.628 and 0.635) or intervention group (p = 0.102 and nificantly lower fluctuations of maximum patellar tendon
0.760) and no time by group interaction (p = 0.623 and strain during MVCs over time in the intervention group
p = 0.838). We found no group differences of maximum compared to the control group that followed their usual
patellar tendon strain during MVCs (p = 0.111; Figure 3C) training schedule. On a descriptive level, the frequency of
and PSF (p = 0.729; Figure 3D), as well as no changes over athletes with high-­level patellar tendon strain (i.e., ≥ 9%)
16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
|DOMROES et al.
  
8 of 14
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    9 of 14

F I G U R E 3 (A) Maximum resultant knee joint moment, (B) normalized patellar tendon stiffness, (C) maximum patellar tendon strain
and (D) peak spatial frequency (PSF) for both groups and each measurement time point (M1-­4) over the season. The given experimental data
is shown in box plots with individual data depicted in colors. The estimated standard deviation of the residuals of the linear mixed model
as a measure of parameter fluctuations over time is shown in the column on the right. The p-­values for baseline differences between groups
(pGroup), changes over time in the control group (pTime_Con), time by group interactions (pInteraction), changes over time in the intervention
group (pTime_Int) as well as group differences of the standard deviation of the residuals (pResid) are given. * marks significant p-­values (<0.05).

was also lower in the intervention group. Further, knee where a generic (non-­personalized) tendon exercise inter-
extensor muscle strength and patellar tendon stiffness in- vention decreased the number of athletes with high-­level
creased significantly in the intervention group, but not in patellar tendon strain. It has been shown in vitro that the
the control group. These findings indicate that the integra- repeated exposure to high tendon strain increases the risk
tion of individualized patellar tendon exercises into the for tendon injury,12,13 as mechanical damage and degen-
regular training has the potential to reduce muscle-­tendon erative processes seem to exceed tissue repair mecha-
imbalances in highly trained male adolescent athletes. nisms.13 Although there is still more in vivo data needed
In vivo, tendon strain is dependent on the level of force to confirm the exact relationship between tendon strain
that acts on the tendon and normalized tendon stiffness. and tendon degeneration and injury, we recently observed
As ultimate tendon strain is considered to be quite con- higher patellar tendon strain in future symptomatic ath-
stant in tendons,11 tendon strain during maximum effort letes compared to athletes that remained asymptomatic as
muscle contractions can serve as a marker for the me- well as a 2.3-­fold increased injury risk for athletes with
chanical demand on the tendon, and high levels of tendon maximum patellar tendon strain ≥9%.21 Further, we found
strain reflect imbalances between muscle strength and indications for structural degeneration of the tendon in
tendon stiffness.27 In the current study, maximum patellar adolescent athletes who were subjected to high tendon
tendon strain values during MVCs were on average 8.9% strain.19,20 Thus, the current findings suggest that the im-
in the control and 8.3% in the intervention group without plementation of regular tendon exercises into the training
significant differences between groups. Maximum patellar may be an effective preventive measure to reduce tendon
tendon strain did not change significantly over the eight-­ overuse injury risk.
month study period in both the control and intervention We observed significant increases in normalized pa-
group. However, the fluctuations of maximum patellar tellar tendon stiffness by on average 28% and maximum
tendon strain during the season were significantly lower knee extension moment by 26% over the season in the
in the intervention group, indicating a more balanced ad- intervention group, while both parameters showed no
aptation between muscle strength and tendon stiffness significant changes in the control group (−8% and 1.4%,
at an individual level in this group. Thereby, the overall respectively). It needs to be noted that the groups in the
range of maximum patellar tendon strain was 6.2%–10.4% current study were not matched with regard to muscle
in the intervention group compared to 5.9%–13.1% in the strength and tendon stiffness at baseline with the inter-
control group. High fluctuations of tendon strain are in- vention group demonstrating on average a 13% lower
dicative for a non-­uniform adaptation of muscle strength knee extension moment and 18% lower normalized
and tendon stiffness over time that may cause temporar- tendon stiffness. Therefore, it cannot be excluded that
ily increased tendon strain. Accordingly, previous studies other factors besides the tendon exercises contributed to
reported higher fluctuations of tendon strain in athletic the observed changes in muscle and tendon properties
populations compared to non-­athletic controls14,16,17 and in the intervention group. However, it should be con-
an increased frequency of high-­level tendon strain in ath- sidered that, when examining the individual changes of
letes.14,15,17 This is substantiated by the high fluctuations normalized stiffness (Figure 3B), also all intervention
of maximum patellar tendon strain in the control group participants with a baseline stiffness within the inter-
of the current study. The reduction of these fluctuations quartile range of the control group data demonstrated a
in the intervention group indicates that the personalized clear increase over time. Further, our findings coincide
tendon loading approach is effective in promoting a more well with earlier reports for the Achilles tendon24–26 that
homogenous development of muscle strength and tendon tendon exercises conducted at tendon strain values be-
stiffness in athletes. tween 4.5% and 6.5% and with a long contraction dura-
In line with this, we observed a lower average fre- tion are effective in promoting tendon adaptation. This
quency of athletes with high-­level patellar tendon strain suggests that the adequate strain region for tendon ad-
(i.e., ≥ 9%) in the intervention group (27%) compared to aptation seems to be similar for the Achilles and patellar
the control group (42%). Though merely descriptive, this tendon and that this also applies to highly trained ado-
observation coincides with recent findings of our group,20 lescent athletes. Taken together, our results indicate that
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
10 of 14    DOMROES et al.

T A B L E 2 Maximum resultant knee joint moment normalized to body mass, corresponding force applied to the patellar tendon, patellar
tendon stiffness and maximum tendon elongation during maximum isometric knee extension contractions as well as patellar tendon rest
length for both groups and each measurement time point (M1-­4) over a competitive season. All values are mean ± standard deviation of the
given experimental data.

Control Intervention

M1 M2 M3 M4 M1 M2 M3 M4
(n = 13) (n = 13) (n = 12) (n = 12) (n = 13) (n = 12) (n = 12) (n = 12)
Norm. moment (Nm/kg)a,b,c 4.09 ± 0.47 4.08 ± 0.46 4.08 ± 0.46 3.99 ± 0.39 3.60 ± 0.51 4.06 ± 0.49 4.13 ± 0.50 4.22 ± 0.25
a,b,c
Tendon force (N) 5555 ± 936 5562 ± 763 5703 ± 984 5519 ± 673 4812 ± 709 5484 ± 897 5633 ± 803 5819 ± 689
Tendon stiffness (N/mm)a,b,c 1930 ± 360 1802 ± 372 2051 ± 435 1733 ± 374 1593 ± 461 1604 ± 468 1977 ± 363 1989 ± 440
Max. tendon elongation 4.6 ± 0.8 4.6 ± 1.0 4.1 ± 0.9 4.5 ± 0.8 4.1 ± 0.8 4.4 ± 0.8 4.1 ± 0.6 4.0 ± 0.6
(mm)
Tendon rest length (mm) 50.3 ± 3.2 50.2 ± 3.1 50.2 ± 3.3 50.1 ± 3.3 50.5 ± 4.8 49.9 ± 5.0 49.9 ± 4.8 50.0 ± 4.9
a
Significant baseline difference between groups.
b
Significant time by group interaction.
c
Significant change over time in the intervention group, p < 0.05.

changes in normalized patellar tendon stiffness in the in-

tervention but not in the control group, which indicates
that increases in patellar tendon stiffness were higher in
individuals with initially high patellar tendon strain val-
ues in the intervention group. These findings are in line
with a previous study where a strong correlation between
maximum patellar tendon strain at baseline and increases
in patellar tendon stiffness was found for male adolescent
athletes who performed functional tendon exercises.20 It
can be argued that especially athletes with increased pa-
tellar tendon strain and accordingly an increased risk for
tendon overuse injuries may benefit from exercise inter-
ventions that target tendon adaptation.
F I G U R E 4 Frequency of athletes demonstrating high-­level While our results indicate that personalized patellar
maximum patellar tendon strain (≥9%) for the control (black) and tendon exercises have the potential to individually pro-
intervention group (white) in the four measurement sessions (M1-­ mote a more balanced adaptation of knee extensor muscle
4) over the season.
strength and patellar tendon stiffness in male adolescent
athletes, some athletes in the intervention group still
demonstrated high-­level tendon strain (≥9%) during the
the tendon exercise intervention had a positive effect on season and at the end of the intervention (Figure 4). This
muscle and tendon adaptation in general, while at the may partly be explained by interindividual differences in
same time reducing musculotendinous imbalances and tendon plasticity. Accordingly, Passini and colleagues38
fluctuations of maximum patellar tendon strain at an could for example demonstrate genetically determined
individual level. differences in the activity of the mechanosensitive ion
Theoretically, tendon strain does not change as long channel PIEZO1, which influences the adaptation of the
as increases in tendon stiffness are matched by increases tendon to mechanical loading. Further, the training load
in muscle strength and a reduction in tendon strain and content during regular training and competition
can only be achieved if increases in tendon stiffness are differed between individuals due to position-­specific re-
comparatively higher than increases in muscle strength. quirements, participation in multiple teams (e.g., older
Therefore, the observed reduction of the frequency of age group or national team), or injuries not related to the
high-­level patellar tendon strain may be attributed to pre- study, which may have influenced differences in muscle
dominant changes of tendon stiffness compared to mus- and tendon adaptation. However, it needs to be pointed
cle strength in individuals with high tendon strain values. out that peak strain values were never higher than 10.4%
Interestingly, we found a significant correlation between in the intervention group while reaching 13.1% in the con-
maximum patellar tendon strain at M1 and relative trol group.
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    11 of 14

F I G U R E 5 Association of maximum patellar tendon strain at the first measurement time point (M1) and the relative changes of
normalized patellar tendon stiffness from the first to fourth measurement time point (M1 to M4) for the control (black, solid line) and
intervention group (white, dashed line). r Pearson correlation coefficient; * significant association (p < 0.05).

One consequence of the repeated exposure to high-­ muscle-­tendon adaptation during the season in both legs.
level tendon strain may be the deterioration of the Therefore, we cannot exclude that possible asymmetries
structural integrity of tendinous tissue.12,13,19 Therefore, between the two legs affected the outcomes in the non-
we performed a PSF analysis as a marker for the mi- dominant leg. While there exists to our knowledge no
crostructural integrity of the proximal patellar tendon. study on leg asymmetries of tendon mechanical proper-
We found no significant changes of PSF over time, as ties in handball players, habitual side-­specific loading in
well as no significant group differences. Low PSF values fencers and badminton players does not seem to induce
are indicative for a more disorganized microstructure systematic differences in maximum tendon strain during
of the tendon which is characteristic for tendinopathic MVCs between legs, despite differences in muscle strength
tendons.32 Accordingly, PSF values of 1.4–1.8 mm−1 (in- and tendon mechanical properties.40 Nevertheless, a per-
terquartile range) have been reported for the patellar sonalized loading for both legs may further improve the
tendon of athletes with tendinopathy and values be- adaptational response.
tween 1.7 and 2.0 mm−1 for asymptomatic controls.39 In Further, it needs to be mentioned that the identifi-
the current study, the majority of the PSF values (90% of cation of muscle-­ tendon imbalances and the respec-
the observations in intervention group and 84% in the tive exercise prescription required the determination
control group) falls within the physiological range of of maximum tendon strain,27 which was derived from
healthy tendons. This suggests that most of the partici- extrapolating tendon elongation to the experimentally
pants had no severe impairments of the structural integ- assessed maximum tendon force in the current study.
rity at the proximal patellar tendon, which may explain This resulted from constraints in the execution of the
the absence of group differences or changes over time. ramp trials as most of the participants had difficulties
Even though we observed a high prevalence of increased reaching their MVC force levels while performing an
tendon strain especially in the control group, it might evenly ramped contraction. We are quite confident that
be conceivable that detectable changes in tendon micro- the extrapolation of the data over a small portion of the
morphology may only manifest with long-­term exposure force-­elongation curve (80%–100% MVC) did not affect
to high-­level tendon strain. our main outcomes since the determination of the force-­
One limitation of our study that needs to be pointed elongation relationship in the current study was rela-
out is that we prescribed the same exercise intensity for tively robust due to the averaging of five trials. Thereby,
both legs based on the data derived only from the domi- the R2 for the second-­order polynomial fit (that was used
nant leg. Due to the time-­consuming measurement pro- to extrapolate tendon elongation) was on average >0.98
cedures, we were not able to investigate tendon strain and in every measurement time point and was in no case
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
12 of 14    DOMROES et al.

lower than 0.95, which indicates that the measured data FUNDING INFORMATION
is well reflected by the polynomial. Also, when running This study was conducted within the scope of the research
the statistical analysis with the tendon strain measured project “Resistance Training in Youth Athletes” that was
at 80% of maximum tendon force, the main findings funded by the German Federal Institute of Sport Science
were also unchanged. The model comparison, that indi- (ZMVI1-­081901/20–23).
cates differences in the standard deviation of the residu-
als between groups, was also significant (p = 0.002) and CONFLICT OF INTEREST STATEMENT
there were no significant changes over time (p = 0.520) The authors declare that the research was conducted
or time by group interaction (p = 0.512). in the absence of any commercial or financial relation-
In conclusion, this novel tendon exercise approach ships that could be construed as a potential conflict of
of personalizing patellar tendon training load to match interest.
the strain region for effective tendon adaptation led to
increases in knee extensor muscle strength and patellar DATA AVAILABILITY STATEMENT
tendon stiffness and a reduction of muscle-­tendon imbal- The data that support the findings of this study are avail-
ances and high-­level tendon strain in highly trained male able from the corresponding author upon reasonable
adolescent athletes. With regards to the injury risk associ- request.
ated with increased tendon strain and the increasing prev-
alence of tendinopathies with maturation8,9 this training ORCID
approach may be a useful tool for the early prevention of Theresa Domroes https://orcid.
tendon overuse injuries. org/0000-0001-9941-6716
Kolja Weidlich https://orcid.org/0000-0002-4398-5773
Sebastian Bohm https://orcid.org/0000-0002-5720-3672
4.1 | Perspectives Falk Mersmann https://orcid.org/0000-0001-7180-7109
Adamantios Arampatzis https://orcid.
Especially in elite-­level sports with a high incidence org/0000-0002-4985-0335
of tendinopathy, a regular screening of the athletes'
muscle-­tendon properties might help to identify individ- REFERENCES
uals with an increased risk for tendon overuse injuries 1. Bohm S, Mersmann F, Santuz A, Arampatzis A. Enthalpy effi-
and to prescribe preventive personalized tendon exer- ciency of the soleus muscle contributes to improvements in run-
cises. In the current study, we estimated the respective ning economy. Proc R Soc B. 2021;288:20202784. doi:10.1098/
rspb.2020.2784
training load from the maximum patellar tendon strain
2. Krupenevich RL, Beck ON, Sawicki GS, Franz JR. Reduced
during MVCs. Due to the marked individual differences
Achilles tendon stiffness disrupts calf muscle neurome-
in maximum patellar tendon strain during the MVCs, chanics in elderly gait. Gerontology. 2022;68(3):241-251.
the resulting relative training load to reach patellar ten- doi:10.1159/000516910
don strain of on average 6.2% during exercises varied 3. Kharazi M, Bohm S, Theodorakis C, Mersmann F, Arampatzis
highly between individuals (47%–90% of the MVC). The A. Quantifying mechanical loading and elastic strain energy of
high individual variability in the percentage of MVC for the human Achilles tendon during walking and running. Sci
the same target patellar tendon strain emphasizes that Rep. 2021;11(1):5830. doi:10.1038/s41598-­021-­84847-­w
4. Roberts TJ, Azizi E. Flexible mechanisms: the diverse roles
the regulation of tendon exercises based on the MVC
of biological springs in vertebrate movement. J Exp Biol.
may be less effective in providing an individually ade-
2011;214(3):353-361. doi:10.1242/jeb.038588
quate stimulus for tendon adaptation. Individuals with 5. Nikolaidou ME, Marzilger R, Bohm S, Mersmann F, Arampatzis
high tendon strain values during MVCs may even be A. Operating length and velocity of human M. Vastus lat-
exposed to critically increased tendon strain levels dur- eralis fascicles during vertical jumping. Royal Soc open Sci.
ing exercise. Therefore, the current approach of patellar 2017;4(5):1-15. doi:10.1098/rsos.170185
tendon exercises based on an individual experimental 6. Bohm S, Mersmann F, Santuz A, Schroll A, Arampatzis A.
assessment of the target patellar tendon strain opens up Muscle-­specific economy of force generation and efficiency of
work production during human running. Elife. 2021;10:e67182.
new opportunities for suitable tendon loading especially
doi:10.7554/eLife.67182
in populations with a high prevalence of musculotendi- 7. Lian ØB, Engebretsen L, Bahr R. Prevalence of jump-
nous imbalances. er's knee among elite athletes from different sports: a
cross-­sectional study. Am J Sport Med. 2005;33(4):561-567.
ACKNOWLEDGMENTS doi:10.1177/0363546504270454
Open Access funding enabled and organized by Projekt 8. Simpson M, Rio E, Cook J. At what age do children and
DEAL. adolescents develop lower limb tendon pathology or
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
DOMROES et al.    13 of 14

tendinopathy? A systematic review and meta-­analysis. Sports 23. Pizzolato C, Lloyd DG, Zheng MH, et al. Finding the sweet
Med. 2016;46(4):545-557. doi:10.1007/s40279-­015-­0438-­0 spot via personalised Achilles tendon training: the future is
9. Florit D, Pedret C, Casals M, Malliaras P, Sugimoto D, Rodas G. within reach. Br J Sports Med. 2019;53(1):11-12. doi:10.1136/
Incidence of tendinopathy in team sports in a multidisciplinary bjsports-­2018-­099020
sports club over 8 seasons. J Sports Sci Med. 2019;18(4):780-788. 24. Arampatzis A, Karamanidis K, Albracht K. Adaptational re-
10. Mersmann F, Bohm S, Arampatzis A. Imbalances in the de- sponses of the human Achilles tendon by modulation of the
velopment of muscle and tendon as risk factor for tendinop- applied cyclic strain magnitude. J Exp Biol. 2007;210(15):2743-
athies in youth athletes: a review of current evidence and 2753. doi:10.1242/jeb.003814
concepts of prevention. Front Physiol. 2017;8:1-18. doi:10.3389/ 25. Arampatzis A, Peper A, Bierbaum S, Albracht K. Plasticity of
fphys.2017.00987 human Achilles tendon mechanical and morphological prop-
11. LaCroix AS, Duenwald-­ Kuehl SE, Lakes RS, Vanderby R. erties in response to cyclic strain. J Biomech. 2010;43(16):3073-
Relationship between tendon stiffness and failure: a meta- 3079. doi:10.1016/j.jbiomech.2010.08.014
analysis. J Appl Physiol. 2013;115(1):43-51. doi:10.1152/ 26. Bohm S, Mersmann F, Tettke M, Kraft M, Arampatzis A.
japplphysiol.01449.2012 Human Achilles tendon plasticity in response to cyclic strain:
12. Wren TAL, Lindsey DP, Beaupré GS, Carter DR. Effects of creep effect of rate and duration. J Exp Biol. 2014;217(22):4010-4017.
and cyclic loading on the mechanical properties and failure of doi:10.1242/jeb.112268
human Achilles tendons. Ann Biomed Eng. 2003;31(6):710-717. 27. Arampatzis A, Mersmann F, Bohm S. Individualized muscle-­
doi:10.1114/1.1569267 tendon assessment and training. Front Physiol. 2020;11(June):1-
13. Wang T, Lin Z, Day RE, et al. Programmable mechanical stim- 7. doi:10.3389/fphys.2020.00723
ulation influences tendon homeostasis in a bioreactor sys- 28. Weidlich K, Mersmann F, Domroes T, Schroll A, Bohm S,
tem. Biotechnol Bioeng. 2013;110(5):1495-1507. doi:10.1002/ Arampatzis A. Quantification of patellar tendon strain and op-
bit.24809 portunities for personalized tendon loading during back squats.
14. Mersmann F, Bohm S, Schroll A, Marzilger R, Arampatzis A. Sci Rep. 2023;13(1):8661. doi:10.1038/s41598-­023-­35441-­9
Athletic training affects the uniformity of muscle and tendon 29. Mersmann F, Laube G, Marzilger R, Bohm S, Schroll A,
adaptation during adolescence. J Appl Physiol. 2016;121(4):893- Arampatzis A. A functional high-­ load exercise intervention
899. doi:10.1152/japplphysiol.00493.2016 for the patellar tendon reduces tendon pain prevalence during
15. Charcharis G, Mersmann F, Bohm S, Arampatzis A. a competitive season in adolescent handball players. Front
Morphological and mechanical properties of the quadri- Physiol. 2021;12(March):282. doi:10.3389/fphys.2021.626225
ceps femoris muscle-­tendon unit from adolescence to adult- 30. Moore SA, McKay HA, Macdonald H, et al. Enhancing a
hood: effects of age and athletic training. Front Physiol. somatic maturity prediction model. Med Sci Sports Exerc.
2019;10(August):1-12. doi:10.3389/fphys.2019.01082 2015;47(8):1755-1764. doi:10.1249/MSS.0000000000000588
16. Karamanidis K, Epro G. Monitoring muscle-­tendon adapta- 31. Lohrer H, Nauck T. Cross-­ cultural adaptation and valida-
tion over several years of athletic training and competition tion of the VISA-­ P questionnaire for German-­ speaking pa-
in elite track and field jumpers. Front Physiol. 2020;11:1-13. tients with patellar tendinopathy. J Ortho SportsPhy Therapy.
doi:10.3389/fphys.2020.607544 2011;41(3):180-190. doi:10.2519/jospt.2011.3354
17. Pentidis N, Mersmann F, Bohm S, et al. Development of muscle– 32. Bashford GR, Tomsen N, Arya S, Burnfield JM, Kulig K.
tendon adaptation in preadolescent gymnasts and untrained Tendinopathy discrimination by use of spatial frequency
peers: a 12-­month longitudinal study. Med Sci Sports Exerc. parameters in ultrasound b-­mode images. IEEE Trans Med
2021;53(12):2565-2576. doi:10.1249/mss.0000000000002742 Imaging. 2008;27(5):608-615. doi:10.1109/TMI.2007.912389
18. Domroes T, Weidlich K, Bohm S, Arampatzis A, Mersmann 33. Herzog W, Abrahamse SK, ter Keurs HEDJ. Theoretical de-
F. Effect of sex on muscle–tendon imbalances and tendon termination of force-­length relations of intact human skeletal
micromorphology in adolescent athletes—a longitudinal muscles using the cross-­bridge model. Pflügers Arch Euro J
consideration. Scand Med Sci Sports. 2023;33(12):2561-2572. Physiol. 1990;416(1–2):113-119. doi:10.1007/BF00370231
doi:10.1111/sms.14483 34. Dempster WT. Space Requirements of the Seated Operator:
19. Mersmann F, Pentidis N, Tsai MS, Schroll A, Arampatzis A. Geometrical, Kinematic, and Mechanical Aspects of the Body
Patellar tendon strain associates to tendon structural abnor- with Special Referece to the Limbs. Wright-­Patterson Air Force
malities in adolescent athletes. Front Physiol. 2019;10(JUL):1- Base; 1955.
10. doi:10.3389/fphys.2019.00963 35. Herzog W, Read LJ. Lines of action and moment arms of the
20. Mersmann F, Domroes T, Pentidis N, et al. Prevention of major force-­carrying structures crossing the human knee joint.
strain-­induced impairments of patellar tendon micromorphol- J Anat. 1993;182:213-230.
ogy in adolescent athletes. Scandinavian J Med and Sci Sport. 36. Schulze F, Mersmann F, Bohm S, Arampatzis A. A wide num-
2021;31(8):1708-1718. doi:10.1111/sms.13979 ber of trials is required to achieve acceptable reliability for
21. Mersmann F, Domroes T, Tsai MS, et al. Longitudinal evi- measurement patellar tendon elongation in vivo. Gait Posture.
dence for high-­level patellar tendon strain as a risk factor for 2012;35(2):334-338. doi:10.1016/j.gaitpost.2011.09.107
tendinopathy in adolescent athletes. Sports Medicine -­Open. 37. Jacqmin-­Gadda H, Sibillot S, Proust C, Molina JM, Thiébaut
2023;9(1):83. doi:10.1186/s40798-­023-­00627-­y R. Robustness of the linear mixed model to misspecified error
22. Duchateau J, Stragier S, Baudry S, Carpentier A. Strength train- distribution. Computational Stat Data Anal. 2007;51(10):5142-
ing: in search of optimal strategies to maximize neuromuscular 5154. doi:10.1016/j.csda.2006.05.021
performance. Exerc Sport Sci Rev. 2021;49(1):2-14. doi:10.1249/ 38. Passini FS, Jaeger PK, Saab AS, et al. Shear-­ stress sens-
JES.0000000000000234 ing by PIEZO1 regulates tendon stiffness in rodents and
 |

16000838, 2024, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/sms.14555 by Cochrane Lithuania, Wiley Online Library on [24/11/2024]. 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
14 of 14    DOMROES et al.

influences jumping performance in humans. Nature Biomed

Eng. 2021;5:1457-1471. doi:10.1038/s41551-­021-­00716-­x How to cite this article: Domroes T, Weidlich K,
39. Kulig K, Landel R, Chang YJ, et al. Patellar tendon morphology Bohm S, Mersmann F, Arampatzis A. Personalized
in volleyball athletes with and without patellar tendinopathy.
tendon loading reduces muscle-­tendon imbalances
Scand J Med Sci Sport. 2013;23(2):81-88. doi:10.1111/sms.12021
40. Couppé C, Svensson RB, Skovlund SV, et al. Habitual side-­
in male adolescent elite athletes. Scand J Med Sci
specific loading leads to structural, mechanical, and compo- Sports. 2024;34:e14555. doi:10.1111/sms.14555
sitional changes in the patellar tendon of young and senior
lifelong male athletes. J Appl Physiol. 2021;131(4):1187-1199.
doi:10.1152/japplphysiol.00202.2021