Content uploaded by Tobias Alt
Author content
All content in this area was uploaded by Tobias Alt on Jun 06, 2023
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rspb20
Sports Biomechanics
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rspb20
Assisted or unassisted Nordic Hamstring Exercise?
— Resistance exercise determinants at a glance
Tobias Alt, Axel J. Knicker, Yannick T. Nodler & Heiko K. Strüder
To cite this article: Tobias Alt, Axel J. Knicker, Yannick T. Nodler & Heiko K. Strüder (2021):
Assisted or unassisted Nordic Hamstring Exercise? — Resistance exercise determinants at a
glance, Sports Biomechanics, DOI: 10.1080/14763141.2021.1893376
To link to this article: https://doi.org/10.1080/14763141.2021.1893376
Published online: 16 Mar 2021.
Submit your article to this journal
View related articles
View Crossmark data
Assisted or unassisted Nordic Hamstring Exercise? — Resistance
exercise determinants at a glance
Tobias Alt
a,b
, Axel J. Knicker
b,c
, Yannick T. Nodler
b
and Heiko K. Strüder
b,c
a
Department of Biomechanics, Performance Analysis and Strength & Conditioning, Olympic Training and
Testing Centre Westphalia, Dortmund, Germany;
b
Institute of Movement and Neuroscience, German Sport
University, Cologne, Germany;
c
Research Centre for Elite Sports, Momentum, Cologne, Germany
ABSTRACT
The Nordic Hamstring Exercise (NHE) eectively strengthens the knee
exors. Typically conducted without assistance, extended knee angles
are not reached with sustained muscle activation in the presence of
insucient eccentric strength and/or fatigue. This might impair the
desired neuromuscular adaptations and assessment accuracy. This
study investigated kinetic and kinematic dierences between assisted
and unassisted NHEs (3 × 3 repetitions) performed by sixteen male
sprinters (22 years, 181 cm, 76 kg). Kinetic (peak moment, impulse) and
kinematic parameters (e.g., time under tension, range of motion to
excessive downward acceleration (ROM
DWA
) were investigated. All
analysed parameters signicantly diered between assisted and unas-
sisted NHEs (p ≤ 0.003; 0.635≤ ηp² ≤ 0.929) favouring assisted execu-
tion, except for peak moments and maximal hip exion. Repetition 1 of
assisted NHEs revealed 21% higher impulses rising to 82% during
repetition 9. Equivalent interactions of mode and repetition became
apparent for time under tension, ROM
DWA
, mean and fractional angu-
lar velocity. Unassisted NHEs elicited substantially greater inter-
repetition fatigue (rep1 vs. rep9): +79% fractional angular velocity
(d = 1.01), −41% impulse (d = 1.53), −31% ROM
DWA
(d = 0.99) and
−29% time under tension (d = 1.45). Assisted NHEs ensured higher
execution quality and lower between-participant variability by facilitat-
ing a controlled full-ROM movement. Three sets of 3 NHEs suced to
induce substantial fatigue within and across sets.
ARTICLE HISTORY
Received 18 November 2020
Accepted 16 February 2021
KEYWORDS
Eccentric resistance training;
exercise quality; hamstring
strength; kinematic analysis;
time under tension
Introduction
The Nordic Hamstring Exercise (NHE) is an effective resistance training exercise to
improve knee flexor strength, thigh muscle balance and consequently contributes to
enhanced hamstring injury prevention (Al Attar et al., 2017; Van Dyk et al., 2019). Since
2001, NHE training has received significant research interest as its supramaximal eccentric
nature and selective activation of the knee flexors cannot be emulated by any other
resistance training exercises (Brockett et al., 2001; Ebben, 2009; Timmins et al., 2015).
Most athletes—especially if not familiarised—do not possess the strength capacities to
perform the NHE across the entire range of motion while maintaining a constant knee
extension velocity and accurate hip control. An uncontrolled NHE execution in the second
CONTACT Tobias Alt tobias.alt@osp-westfalen.de
SPORTS BIOMECHANICS
https://doi.org/10.1080/14763141.2021.1893376
© 2021 Informa UK Limited, trading as Taylor & Francis Group
half of the movement might diminish or even prevent neuromuscular adaptations at
comparably long hamstring muscle length (Alt et al., 2018). Different research groups
have demonstrated that only ~50% of unassisted NHE’s range of motion (ROM) were
performed in a controlled fashion (Alt et al., 2018; Delahunt et al., 2016; Ditroilo et al.,
2013; Marshall et al., 2015). During the second half of the exercise, hamstring activation
significantly declined and the knee angular velocity continually increased from the angle at
downward acceleration (DWA) (Delahunt et al., 2016; Ditroilo et al., 2013; Marshall et al.,
2015). These findings emphasised that the physical demands of a single unassisted NHE
repetition are too high for most athletes (Marshall et al., 2015). It is suggested that an
assisted NHE execution—particularly in the presence of insufficient eccentric knee flexor
strength—might be advantageous to guide the execution of NHEs in a controlled fashion
and to ensure a consistently high hamstring activation throughout the entire movement
(Alt et al., 2018; Burrows et al., 2020; Matthews et al., 2015). Nonetheless, the evidence-
based preventive effect of unassisted NHE training is undisputed (Mjolsnes et al., 2004;
Van Dyk et al., 2019), whereas injury prevention studies implementing assisted NHEs are
lacking.
Toigo and Boutellier (2006) introduced new fundamental resistance exercise deter-
minants related to molecular and cellular muscle adaptations. They highlighted that
muscular adaptations are dependent upon the range of motion, the time under tension
(TUT), as well as the fractional and temporal distribution of one repetition. These
parameters are aggregated in the length-at-use and contraction-mode hypotheses,
meaning that muscles adapt according to their common tasks and imposed demands
(Herzog et al., 1991; Savelberg & Meijer, 2003). Concerning NHE training, the funda-
mental resistance exercise determinants to ensure a controlled exercise execution
should be the following (Alt et al., 2018; Delahunt et al., 2016; Ditroilo et al., 2013;
Marshall et al., 2015):
●knee range of motion: ~90° reaching best possible knee extension (~20° to 0°),
●time under tension: ~6–7 s lasting eccentric-only movement,
●fractional and temporal distribution: aiming at an equal distribution by maintaining
a constant angular velocity across the complete range of motion (100% ROM
DWA
).
The current literature reveals a substantial lack of information about the resistance exercise
determinants of NHE training. To our knowledge, three cross-sectional studies (Ditroilo et al.,
2013; Marshall et al., 2015; Sarabon et al., 2019) and two longitudinal studies (Alt et al., 2018;
Delahunt et al., 2016) ensured a comprehensive and precise exercise prescription by imple-
menting kinematic analyses. Alt et al. (2018) introduced assisted ‘isokinetic’ (constant velo-
city) NHE training sessions and found altered resistance exercise determinants (e.g., lower
peak moments, higher impulses, ROM
knee
and TUT) and other accompanying parameters
(e.g., higher ROM
DWA
, lower hip flexion) compared to unassisted sessions. However, no
information is available about the time course of these parameters within a NHE training
session consisting of multiple sets and repetitions. Hence, there is no evidence whether an
assisted or unassisted NHE should be preferentially performed and which of the different
existing training regimens is advisable: 3 sets of 3 repetitions (Alt et al., 2018), 2 to 6 sets of 5
repetitions (Delahunt et al., 2016; Ditroilo et al., 2013; Marshall et al., 2015) up to 3 sets of 12
repetitions (Delahunt et al., 2016; Mjolsnes et al., 2004). This knowledge is deemed to be
2T. ALT ET AL.
essential for coaches, physiotherapists and scientists to design, perform and adjust reasonable
and individual NHE training programmes for athletes of different physical performance levels
and for patients in different stages of the rehabilitation process.
Thus, the aims of the present secondary analysis of an already published cohort (Alt
et al., 2018) were (1) to establish the kinetic and kinematic differences between assisted
and unassisted ‘isokinetic’ NHE execution and (2) to investigate the time course of NHE
performance parameters within NHE training sessions containing 3 × 3 repetitions. Our
hypotheses are as follows: (1) assisted NHE training is advantageous concerning resis-
tance exercise determinants and (2) 3 sets of 3 ‘isokinetic’ eccentric-only repetitions
suffice to induce significant inter-repetition fatigue which impairs exercise quality.
Methods
Participants
Sixteen regional to national class male sprinters (21.6 ± 2.5 years, 181.4 ± 7.1 cm,
75.7 ± 9.8 kg) with different training history (2–17 years), training volume (4–12 h/
week) and performance level (100 m season’s best: 10.99–12.90 s) gave their written
informed consent to voluntarily participate in the study. All of them were familiar with
lower extremity resistance training, but not with NHE training. They had neither thigh
muscle nor knee injuries within the last three years. The local ethics commission of the
German Sport University, Cologne confirmed that the requirements of the Declaration of
Helsinki were met. Within the intervention period all participants maintained their
normal physical activity (e.g., sprint and resistance training) which did not demonstrably
affect the relevant outcome parameters (Alt et al., 2018). Thirteen participants performed
all 108 NHE repetitions (54 assisted and 54 unassisted repetitions), whereas three omitted
one NHE session due to reasons which were not related to the NHE training.
Instruments
Prior to the first session, the anthropometric data of all participants were measured
according to the Hanavan model (Hanavan, 1964). All NHE training sessions were
executed on the isometrically operating dynamometer IsoMed2000 (D&R Ferstl
GmbH, Hemau, Germany) recording raw data at 200 Hz. All relevant test parameters
were synchronised and derived from the gathered moment-time curves by a self-
developed software written in C++. The participants’ motion was captured from above
and laterally by two synchronised high-speed cameras (acA640-120gc, Basler AG,
Ahrensburg, Germany) at 100 fps (TEMPLO 8.2.358, Contemplas, Kempten,
Germany). Retroreflective markers (Ø = 15–25 mm) were attached on three bony land-
marks of the right side of the body (acromion, trochanter major, lateral femoral epicon-
dyle) and on a distinctive position on the lever arm representing the location of the lateral
malleolus (Figure 1a, b). Two headlights (Bar Fly 200, Kino Flo® Lightning Systems,
Burbank, California, USA) improved the markers’ visibility. The measuring volume was
statically calibrated with a rigid frame of 912 × 590 x 870 mm (L x W x H). Kinematic
analyses were conducted with Vicon Peak Motus (V10.0.1, New York, New York, USA).
SPORTS BIOMECHANICS 3
Testing procedure
Data were derived from an earlier study (Alt et al., 2018) and re-analysed to investigate
the kinetic and kinematic differences of assisted and unassisted NHEs. Prior to the
analysed sessions, each participant performed six NHE training sessions (3 sessions per
week with a 47 h inter-session rest), whereas the gradually alternating sequence of
Figure 1. Representative illustrations of (a) assisted (grey lines) and (b) unassisted (black lines) NHE
execution as well as their corresponding (c) moment-time and (d) angular velocity-time histories
across 3 exemplary sets with 3 repetitions each.
4T. ALT ET AL.
assisted (A) and unassisted (U) sessions was stratified (week 1: A A A; week 2: A U A).
The analysed sessions were the first two training sessions of week 3 (U A U) so that
a sufficient familiarisation to NHE execution was ensured by the preceding six accom-
modation visits (Alt et al., 2018). Each session adhered to the identical procedure. After
determining their mass, participants performed an individual warm-up (jogging,
dynamic stretching) for ten minutes. The general warm-up was followed by a specific
preparation at a doorway pull-up bar (Denqbar DQ-0161, Pirna, Germany) (2 x 5
repetitions, 3 min rest). These eccentric-only repetitions were executed at submaximal
intensity (~60%) with a slow self-selected velocity until the DWA angle was reached. The
return into the starting position was assisted by the examiner to avoid any muscular
activity. Afterwards, 3 sets of 3 ‘isokinetic’ NHEs each were performed on the dynam-
ometer (Figure 1a, b). The inter-repetition and inter-set rest periods were set at 10 sec-
onds and 5 minutes, respectively. Each repetition was performed solely eccentrically
while the return to the starting position was to be executed without any hamstring muscle
activation by a push of the arms combined with knee and hip flexion (Alt et al., 2018).
All participants kneeled on the lounger with the knee axis’ orientation matching the
dynamometer’s axis. Prior to each repetition both heels were pushed against two pads
attached to the lever arm (Figure 1a, b). The participants were asked to meet the
following execution quality criteria: (1) to execute eccentric NHEs with a constant knee
extension velocity of ~15°/s across the largest possible ROM
knee
(~90°) targeting a time
under tension of ~6.5 seconds per repetition and (2) to realise minimal hip flexion whilst
(3) the hands were positioned in front of the shoulders (Alt et al., 2018). Minimal hip
flexion was striven to increase the eccentric load on the knee flexors by maximising the
lever arm of the centre of mass to the knee joint. To realise the best possible ‘isokinetic’
NHE execution velocity, the participants received a sagittal visual live feedback provided
by a webcam recording at 30 fps (C200, Logitech, Apples, Switzerland) and a stick figure
moving at target speed shown on a 17” monitor (710 N, SAMSUNG, Seoul, South Korea).
Assisted NHE training sets were conducted via rope-controlled resistance by the exam-
iner which was transferred to the back part of a climbing harness (Petzl, Crolles, France)
worn by the participants (Figure 1a, 2). During every single assisted NHE repetition, the
examiner adapted the resistance to the climbing harness according to the actual move-
ment speed of the participant so that the required constant knee extension velocity was
optimally achieved. Resistance was increased if participants moved faster than the
projected stick figure and contrariwise.
Data processing
Raw data of NHE repetitions were stored as ASCII files before a recursive 5
th
order
Butterworth low-pass filter with a cut-off frequency of 6 Hz was applied. The beginning
and the end of a NHE repetition was determined by the first derivative of the moment-
time and knee flexion angle-time curves. Knee and hip joint angles were set at 0° for full
extension and increased with flexion. Figure 1 illustrates representative moment-time (c)
and angular velocity-time curves (d) of assisted and unassisted NHE sets. Peak moment
(PM), angle of peak moment (APM), impulse (J), time under tension (TUT), range of
motion of the knee joint (ROM
knee
), range of motion to downward acceleration
(ROM
DWA
), minimal knee flexion (KF
min
) and maximal hip flexion angles (HF
max
) as
SPORTS BIOMECHANICS 5
well as mean (ω
mean
) and fractional knee extension velocity within 60–30° knee flexion
(ω
mean
60–30° KF) were selected as relevant kinetic and kinematic parameters character-
ising NHE execution quality. The knee angle at DWA was identified as coinciding with
the highest angular acceleration in the knee extension velocity-knee flexion angle curve
(Delahunt et al., 2016; Ditroilo et al., 2013). At the beginning of the NHE training period,
we recommended to place the knee joint in front of the shank support (Figure 1a, b) to
enable that the articular cartilage of the tibia head can roll underneath the patella
throughout NHE execution. However, due to inter-individual differences in perception
of convenience, some participants (n = 4) chose a knee position on the lounger or on the
edge of it. Misalignment of the dynamometer and the knee joint axis was mathematically
corrected by calculating the resultant joint moments for each training session. Figure 2
demonstrates the agreement between resultant values and theoretical values derived from
inertial segment properties based on the Hanavan model (Hanavan, 1964). All kinetic
NHE parameters were normalised to body mass allowing inter- and intra-individual
comparison. Percentage differences of each parameter across sets and repetitions served
to estimate fatigue.
Statistical analysis
For all data, normal distribution were confirmed by the Kolmogorov-Smirnov (α ≤ 0.05).
Two-way repeated measures analyses of variance identified the effects of execution mode
(assisted vs. unassisted) and repetition number (repetition 1 to 9) on kinetic and kine-
matic NHE parameters. Bonferroni post hoc tests determined the actual p-values
between the nine NHE repetitions within each execution mode. Variance homogeneity
was assessed by the Levene’s test to examine the between-participant variability of the
two NHE execution modes. The level of significance was set at α ≤ 0.05 for all statistical
tests. All statistical tests were calculated with SPSS V.23.0 (SPSS Inc., Chicago, Illinois,
USA). Effect sizes are indicated as Cohen’s d (≥0.8 large; 0.8–0.5 moderate, 0.5–0.2 small;
Figure 2. Knee moment curves of a single eccentric-only NHE repetition of different execution
modalities: optimal unassisted execution (black), two assisted repetitions (light and dark grey) and
idealistic anthropometric model (grey dashed range). The magnitude of provided rope assistance can
be quantified by the area between the dashed curve and the respective grey solid line.
6T. ALT ET AL.
<0.2 negligible) and partial eta-squared (ηp²) (≥0.26 large; 0.26–0.13 moderate, 0.13–0.02
small; <0.02 negligible) specifying the meaningfulness of the respective differences
(Cohen, 1988).
Results
Kinetic and kinematic dierences between assisted and unassisted NHE
Table 1 lists all kinetic and kinematic parameters of assisted and unassisted NHEs across 3
sets of 3 repetitions each. Both execution modes significantly differed for all analysed
parameters (p ≤ 0.003; 0.635≤ ηp² ≤ 0.929). Peak moments as well as angles of peak
moment of assisted NHEs were 17% to 21% and 9° to 16° lower compared to unassisted
executions. In contrast, assisted NHEs revealed significantly larger values for the impulse
(+21% to +82%) (Figure 3a) and ROM
knee
(+6° to +13°). The higher impulses were
associated with a significantly greater TUT (+45% to +93%) (Figure 3b) and constantly
lower ω
mean
(−5°/s to −8°/s), whereas the greater range of motion was related to a higher
range of motion at the end of the exercise (−7° to −14° KF
min
). Due to a lower ω
mean
60–30°
KF (−15°/s to −37°/s) (Figure 3c) the ROM
DWA
was substantially higher (+23% to +47%)
when assisted NHEs were performed (Table 1). Furthermore, assisted NHE induced a 7° to
9° smaller maximal hip flexion than the unassisted execution mode (Figure 1a, b). Between-
participant variability of assisted NHEs was significantly lower for all parameters compared
to unassisted NHEs, except for APM and HF
max
(Table 1).
Time course of NHE performance parameters from 3 sets of 3 repetitions
Peak moment, impulse and time under tension showed moderate to large effects of NHE
repetition in both assisted and unassisted execution mode (p ≤ 0.002; 0.246≤ ηp² ≤ 0.465)
(Table 1). Concerning the peak moment this effect was more pronounced in assisted
NHEs—especially at repetition 6 and 9—, whereas the impulse (Figure 3a), time under
tension (Figure 3b) as well as the fractional angular velocity within 60–30° knee flexion
(Figure 3c) of unassisted NHEs were significantly reduced right away from repetition 2
onwards. The impulse of assisted NHEs decreased 11% from repetition 1 to 9 compared
to an inter-repetition fatigue of 41% during unassisted execution. The same was observed
for the TUT (−5% vs. −29%) and ω
mean
60–30° KF (+1% vs. +79%). In addition to these
three parameters, ROM
DWA
(0% vs. −31%) and ω
mean
(+9% vs. +30%) demonstrated
large interactions between mode and repetition (p ≤ 0.004; 0.272≤ ηp² ≤ 0.347) (Table 1).
Moderate interactions of mode and repetition became apparent for APM, ROM
knee
and
KF
min
(p ≤ 0.049; 0.131≤ ηp² ≤ 0.157). Maximal hip flexion was affected by neither
repetition nor mode*repetition (p > 0.05; 0.056≤ ηp² ≤ 0.081).
Discussion & implications
The discussion is organised according to the aims of the study: (1) to establish the kinetic
and kinematic differences between assisted and unassisted ‘isokinetic’ NHE execution
and (2) to investigate the time course of NHE performance parameters within NHE
training sessions containing 3 × 3 repetitions. The hypothesis was that assisted NHE
SPORTS BIOMECHANICS 7
Table 1. Kinetic and kinematic NHE determinants (mean ± SD) obtained from 3 sets of each 3 repetitions of assisted and unassisted NHE. Significant differences (p ≤ 0.05)
between execution modes, repetitions, the interaction of both factors and variance homogeneity are highlighted. Significant differences (p ≤ 0.05) between repetitions
within one execution mode are emphasised by corresponding repetition numbers in italics.
PM [Nm/kg] APM [°] J [Nm∙s/kg] TUT [s] ROM
knee
[°]
Assisted Unassisted Assisted Unassisted Assisted Unassisted Assisted Unassisted Assisted Unassisted
Set 1 rep 1 3.61 ± 0.35 4.49 ± 0.35 38.4 ± 8.6 46.9 ± 7.2 18.1 ± 2.8 15.0 ± 5.2 6.98 ± 0.64 4.83 ± 1.17 80.2 ± 8.9 74.2 ± 7.7
rep 2 3.61 ± 0.35 4.38 ± 0.48 38.7 ± 7.1 49.6 ± 8.0 17.1 ± 2.0 13.2 ± 4.2
1
6.67 ± 0.49 4.56 ± 1.07
1
80.0 ± 7.6 73.6 ± 7.6
rep 3 3.51 ± 0.35 4.40 ± 0.41 38.0 ± 8.4 52.3 ± 9.3
1
16.6 ± 1.7 11.1 ± 3.3
1
6.62 ± 0.42 3.95 ± 0.75
1 2
83.1 ± 7.4 72.1 ± 8.8
Set 2 rep 4 3.62 ± 0.33 4.35 ± 0.53 37.5 ± 5.3 48.5 ± 9.1
3
18.0 ± 2.1
3
13.2 ± 5.6
1 3
6.96 ± 0.41
3
4.55 ± 1.24
3
80.9 ± 5.6 74.5 ± 9.9
rep 5 3.53 ± 0.32
4
4.37 ± 0.45 37.3 ± 8.4 52.1 ± 10.6
1
17.4 ± 1.8 11.8 ± 4.3
1
6.82 ± 0.43 4.16 ± 1.00
1 2 4
81.9 ± 5.8 74.5 ± 11.3
rep 6 3.45 ± 0.33
1 2 4
4.35 ± 0.53 38.9 ± 8.0 51.5 ± 9.9
1
16.7 ± 1.8
4
10.9 ± 3.4
1 2 4
6.68 ± 0.42 3.90 ± 0.76
1 2 4
82.4 ± 5.1 73.3 ± 9.4
Set 3 rep 7 3.59 ± 0.37
6
4.33 ± 0.53
1
36.7 ± 7.8 49.4 ± 7.0 17.9 ± 2.2
3 6
12.6 ± 5.3
1 6
6.86 ± 0.48 4.29 ± 1.18
1 2 6
80.9 ± 6.6 71.1 ± 7.9
rep 8 3.54 ± 0.35 4.33 ± 0.51
1
37.8 ± 8.7 50.9 ± 7.9
1
17.5 ± 1.8
6
10.7 ± 4.0
1 4 7
6.95 ± 0.41
2
3.80 ± 1.05
1 2 4 5 7
81.4 ± 5.8 72.4 ± 7.4
rep 9 3.40 ± 0.33
1 2 4 5 7 8
4.25 ± 0.51
1
36.9 ± 9.8 53.3 ± 9.2
1 4
16.2 ± 1.6
1 4 5 7 8
8.9 ± 2.6
1 2 3 4 5 6 7 8
6.60 ± 0.46
4 8
3.43 ± 0.78
1 2 3 4 5 6 7 8
82.0 ± 6.2 69.1 ± 10.2
Mode p < 0.001; ηp² = 0.862 large p < 0.001; ηp² = 0.760 large p < 0.001; ηp² = 0.781 large p < 0.001; ηp² = 0.929 large p=0.003; ηp² = 0.681
large
Repetition p = 0.002; ηp² = 0.246 moderate p = 0.254; ηp² = 0.083 p < 0.001; ηp² = 0.465 large p < 0.001; ηp² = 0.391 large p = 0.230; ηp² = 0.082
Interaction p = 0.245; ηp² = 0.085 p = 0.020; ηp² = 0.157
moderate
p = 0.012; ηp² = 0.204 moderate p = 0.002; ηp² = 0.247 moderate p = 0.028; ηp² = 0.141
moderate
Variance
homogeneity
p < 0.001; F = 13.452 p = 0.151; F = 2.072 p < 0.001; F = 66.075 p < 0.001; F = 67.280 p = 0.012; F = 6.410
ROM
DWA
[%] KF
min
[°] HF
max
[°] ω
mean
[°/s] ω
mean
60–30° KF [°/s]
Assisted Unassisted Assisted Unassisted Assisted Unassisted Assisted Unassisted Assisted Unassisted
Set 1 rep 1 100.0 ± 0.0 79.4 ± 20.5 28.7 ± 8.1 35.3 ± 7.1 8.6 ± 8.2 16.2 ± 11.0 11.4 ± 1.6 16.1 ± 4.2 11.4 ± 2.7 26.8 ± 16.3
rep 2 100.0 ± 0.0 71.6 ± 20.1
1
29.6 ± 8.5 36.9 ± 6.8 9.8 ± 7.7 17.4 ± 9.5 12.3 ± 1.2 16.9 ± 4.6 12.5 ± 2.1 33.2 ± 21.6
1
rep 3 100.0 ± 0.0 62.3 ± 27.6
1
27.6 ± 7.8
2
38.1 ± 8.2 8.8 ± 7.7 17.2 ± 7.3 13.0 ± 1.5
1
18.7 ± 4.3
1 2
13.1 ± 2.1 38.9 ± 23.7
1
Set 2 rep 4 100.0 ± 0.0 73.0 ± 26.4 27.3 ± 4.7 36.2 ± 9.3 8.0 ± 6.8 16.6 ± 10.6 12.1 ± 1.3
3
17.1 ± 4.2
3
11.6 ± 2.1 31.6 ± 19.9
1 3
rep 5 100.0 ± 0.0 69.7 ± 27.7
1
27.5 ± 5.3 37.0 ± 11.3 9.6 ± 8.1 17.4 ± 7.3 12.2 ± 1.5 18.5 ± 4.9
1 4
11.8 ± 2.1 36.8 ± 23.2
1
rep 6 100.0 ± 0.0 56.8 ± 24.2
1 2 4
27.5 ± 5.2 37.8 ± 9.6 10.4 ± 7.5 17.3 ± 8.8 12.4 ± 1.6 19.5 ± 5.2
1 2 4
11.8 ± 2.4 38.1 ± 23.9
1 4
Set 3 rep 7 100.0 ± 0.0 72.0 ± 26.5
6
27.5 ± 4.6 37.9 ± 8.3 8.5 ± 6.1 16.0 ± 10.2 11.9 ± 1.2
3
17.7 ± 5.3
1 5 6
11.4 ± 2.1 32.8 ± 22.1
1 3 6
rep 8 96.0 ± 15.9 57.1 ± 27.9
1 2 4 7
29.1 ± 5.1 37.9 ± 8.2 10.0 ± 7.2 17.7 ± 8.5 12.1 ± 1.3 20.0 ± 5.1
1 2 4 7
11.8 ± 2.1 42.3 ± 26.8
1 4 7
rep 9 100.0 ± 0.0 53.4 ± 26.6
1 2 4 7
27.0 ± 6.9 41.0 ± 10.3
1
9.2 ± 8.0 17.2 ± 8.0 12.5 ± 1.5 20.9 ± 5.9
1 2 3 4 5 7
11.6 ± 2.3 48.0 ± 29.4
1 2 4 5 7
Mode p < 0.001; ηp² = 0.722 large p < 0.001; ηp² = 0.688 large p < 0.001; ηp² = 0.660 large p < 0.001; ηp² = 0.717 large p < 0.001; ηp² = 0.635 large
Repetition p = 0.004; ηp² = 0.272 large p = 0.273; ηp² = 0.077 p = 0.282; ηp² = 0.081 p < 0.001; ηp² = 0.347 large p = 0.006; ηp² = 0.244 moderate
Interaction p = 0.013; ηp² = 0.227 moderate p = 0.049; ηp² = 0.131 moderate p = 0.482; ηp² = 0.056 p = 0.020; ηp² = 0.203 moderate p = 0.007; ηp² = 0.241 moderate
Variance
homogeneity
p < 0.001; F = 486.299 p < 0.001; F = 13.251 p = 0.085; F = 2.990 p < 0.001; F = 88.821 p < 0.001; F = 235.416
Footnote: PM (peak moment), APM (angle of peak moment), J (impulse (M*t)), TUT (time under tension), ROM
knee
(range of motion of the knee joint), ROM
DWA
(percentage of range of motion to
downward acceleration in relation to ROM
knee
), KF
min
(mimimal knee flexion angle), HF
max
(maximal hip flexion angle), ω
mean
(mean angular velocity of the knee joint), ω
mean
60–30° KF (mean angular
velocity within 60–30° knee flexion)
8T. ALT ET AL.
training is beneficial concerning resistance exercise determinants and that 3 sets of 3
‘isokinetic’ repetitions suffice to induce significant fatigue.
Assisted vs. unassisted NHE execution mode
The rationale for applying rope-assisted NHEs at a constant knee extension velocity was
to meet the demands of the length-at-use and contraction-mode hypotheses (Herzog
et al., 1991; Savelberg & Meijer, 2003). In order that muscles adapt according to their
common tasks and imposed demands, NHEs should contain a high fractional and
temporal portion at comparably long muscle length (Toigo & Boutellier, 2006).
Adaptations at long fascicle length are of major importance because a long hamstring
fascicle length has been hypothesised as an important preventive characteristic for ham-
string injuries (Timmins et al., 2015). Furthermore, the slow angular velocity of 15°/s and
the long concomitant time under tension of ~6–7 s complied with the recommendations
for optimal muscle-tendon adaptation (Bohm et al., 2015).
As Table 1 emphasises, assisted NHE executions met the required exercise determinants,
whereas unassisted NHEs revealed large between-participant variabilities and fundamental
deviations of TUT, ROM
DWA
and ω
mean
60–30° KF. When executing unassisted NHEs,
participants were not able to maintain a constant ω
mean
across the full ROM
knee
(Figure 1d)
(Alt et al., 2018). Right from the first repetition, significant differences between the two
analysed NHE execution modes became apparent in all parameters (Table 1). With rising
repetition number, the difference between assisted and unassisted NHEs gradually
increased up to +82% (J) and +93% (TUT) (Figure 3a, b). These results underline that
most participants did not have the strength capacities to conduct an unassisted ‘isokinetic’
NHE across the full ROM
knee
. Additionally, this fact is supported by the large standard
deviations of the impulse and ω
mean
60–30° KF (Figure 3a, c) demonstrating vast inter-
individual differences in NHE execution quality. Only if these inter-individual differences
are quantified, subsequent adaptations can be interpreted in a meaningful way. To our
knowledge, no longitudinal study is available so far which provided this information.
Assisted NHEs revealed favourable resistance exercise determinants concerning eight out
of ten analysed parameters (Table 2) with respect to ensuring high execution quality and
hamstring function assessment accuracy. Maximal hip flexion cannot be explicitly assigned
to one execution mode because recent studies suggested that NHE with flexed hip joints
will further promote strength adaptation at long hamstring muscle length (Hegyi et al.,
2019; Sarabon et al., 2019). Therefore, it is not yet clear if NHE execution with a flexed hip
joint is preferable compared to best possible hip extension. Extended hip joints increase the
eccentric load on the knee flexors by maximising the lever arm of the centre of mass to the
knee joint. The only resistance exercise determinant which favoured unassisted NHEs was
the intensity as peak moments were 17% to 21% lower during assisted executions (Table 1).
Higher intensity leads to higher tendon loading (Bohm et al., 2015) and due to a higher
fractional angular velocity (Figure 3c) to higher muscle-tendon strain rates. These are
reasonable adaptation triggers provided that the time under tension and muscle activation
are not impaired (Delahunt et al., 2016; Ditroilo et al., 2013; Marshall et al., 2015). Thus,
appropriate NHE training should combine a high intensity, high impulses, as well as a high
temporal and fractional distribution within a range of motion near full knee extension. If
these guidelines cannot be fulfilled by unassisted NHE execution, as in the case of
SPORTS BIOMECHANICS 9
unfamiliarised, unexperienced and weak athletes and patients, assisted NHEs should be the
method of choice to execute the exercise properly and to acquire adequate strength
capacities (Alt et al., 2018; Burrows et al., 2020; Matthews et al., 2015). Depending on the
individual moment-angle characteristics of the athlete, assistance can be provided across
Figure 3. Inter-repetition courses of the impulse (a), the time under tension (b) and the mean angular
velocity within 60–30° knee flexion (ω
mean
) (c) obtained from 3 sets of each 3 repetitions of assisted
(grey) and unassisted (black) NHE. The associated significant effects between execution mode and
repetition number are indicated in Table 1.
10 T. ALT ET AL.
the entire NHE repetition (Figure 2, light grey line) or within suited portions of the
movement (dark grey line).
Based on the present results (Table 2), the hypothesis that assisted NHE training is
advantageous concerning resistance exercise determinants can be confirmed.
Fatigue-induced eects of 3 sets of 3 NHE repetitions
Current literature introduced different NHE training regimens: 3 sets of 3 repetitions (Alt
et al., 2018), 2 to 6 sets of 5 repetitions (Delahunt et al., 2016; Ditroilo et al., 2013;
Marshall et al., 2015) up to 3 sets of 12 repetitions (Delahunt et al., 2016). However, there
is no clear evidence how many repetitions suffice to induce a significant fatigue. Marshall
et al. (2015) investigated 6 sets of 5 repetitions each conducted at an average cadence of
30°/s. They found no significant differences of fractional angular velocity, but determined
reduced isokinetic peak moments (−8% to −17%) after set 2 to 6. To our knowledge, this
study was the first which evaluated inter-repetition and inter-set fatigue effects by
assessing NHE kinetics and kinematics.
Independent of NHE execution mode, the peak moment, impulse, time under tension,
ROM
DWA
, ω
mean
and ω
mean
60–30° KF were significantly affected by repetition-induced
exhaustion (Table 1). While the peak moments of unassisted NHEs remained almost
unchanged, assisted NHEs suffered from a stepwise inter-repetition and inter-set decre-
ment (Figure 1c). Unassisted NHEs revealed this phenomenon regarding inter alia the
mean angular velocity (Figure 1d), impulse (Figure 3a), time under tension (Figure 3b)
and the fractional angular velocity (Figure 3c). From repetition 1 to 3, the impulse
(−26%), time under tension (−18%) and fractional angular velocity (+45%) of unassisted
NHEs demonstrated significant changes. Although the inter-set rest of 5 minutes lead to
a certain degree of recovery, these three parameters became fundamentally altered up to
repetition 9 (−41%; −29%; +79%). Despite rope-assistance, peak moments, impulses and
time under tension were significantly reduced comparing repetition 1 to 9 (Table 1). This
effect emphasises the high physical demands of NHEs on eccentric hamstring strength
(Alt et al., 2018; Brockett et al.; Ebben, 2009; Timmins et al., 2015). Other kinematic
parameters, such as the knee and hip flexion angles remained nearly unchanged under-
lining that movement execution was not affected by fatigue.
Regarding these results, it is doubtful if exercise quality and intensity can be main-
tained at a sufficiently high level when performing 3 sets of 8–12 NHE repetitions per
Table 2. Benefits of assisted and unassisted NHEs with respect to ensuring a high
execution quality and to meeting the required exercise determinants as well as
parameters which cannot be explicitly assigned.
Assisted NHE Draw Unassisted NHE
Angle of peak moment (APM) Maximal hip flexion (HF
max
) Peak moment (PM)
Impulse (J)
Time under tension (TUT)
Range of motion (ROM
knee
)
Downward acceleration (ROM
DWA
)
Minimal knee flexion (KF
min
)
Mean angular velocity (ω
mean
)
Fractional angular velocity (ω
mean
60–30° KF)
SPORTS BIOMECHANICS 11
training session. Unfortunately, no kinetic and kinematic performance parameters of
such high volume NHE training regimen are available (Delahunt et al., 2016; Mjolsnes
et al., 2004). However, the results of a recent study suggest that strength increases
following a high volume NHE training (4 weeks of 2 x 4 x 8-10 repetitions) are not
higher compared to a low volume regimen (4 weeks of 1 x 2 x 4 repetitions) (Presland
et al., 2018). Furthermore, it is well-known that optimal tendon adaptation requires high
intensities (~90%) and a sufficiently high time under tension (~3 s) (Bohm et al., 2015).
Regarding these specific demands, low volume NHE training (2 sets of 4 reps per session)
is not inferior to high volume regimen (4–5 sets of 8–10 reps) (Presland et al., 2018).
Accordingly, the hypothesis that 3 sets of 3 ‘isokinetic’ NHE repetitions suffice to
induce significant fatigue can be confirmed for assisted execution. For unassisted NHEs,
already 2 sets of 3 repetitions lead to significant fatiguing effects.
Limitations & perspectives
The present results rely on a heterogenous group of sprinters varying in training history,
training volume and performance level. A study incorporating homogenous samples of
weak and strong athletes could identify if assisted or unassisted NHEs are more effective
in promoting eccentric hamstring strength.
This study examined bilateral NHE kinetics. Prospective studies should quantify the
contribution of each thigh to the generated eccentric work. Shank orientation should be
modified from a negative to a positive inclination or to a horizontal orientation (Sarabon
et al., 2019) to realise full knee extension (Figure 1a, b). Even during assisted NHE
a minimal knee flexion of 28° remained (Table 1) due to the participants’ fear of slipping
off the edge. If the shank is more horizontally aligned, a greater ROM
knee
up to almost full
knee extension will be possible. Future studies should determine individual relationships
between NHE training stress and NHE performance as well as related isokinetic para-
meters to recognise responders and non-responders. Finally, the transfer of NHE-
induced improved hamstring strength to sport-specific loading and relevant performance
parameters (e.g., joint power peak moments, impulses) during e.g., sprinting is of major
interest.
Assisted NHE execution mode was favourable to ensure high execution quality and to
meet the required exercise determinants. However, assisted NHE training requires
additional equipment and is therefore less feasible for on-field training. It must still be
investigated whether assisted NHE training promotes muscle-tendon adaptations at
comparably long hamstring length and whether it reduces injury risk to a greater extent
than unassisted NHEs. This knowledge is deemed to be essential for coaches, phy-
siotherapists and scientists to design, perform and adjust reasonable and individual
NHE training programmes for athletes of different physical performance levels and for
patients in different stages of the rehabilitation process (Al Attar et al., 2017). It might
contribute to a targeted injury prevention of hamstring strain injuries and related injuries
such as anterior cruciate ligament tears (Van Dyk et al., 2019). Future studies should
expand the previous kinetic, kinematic and electromyographic analyses to investigate the
effects of different NHE training modalities (assisted vs. unassisted; bilateral vs. unilat-
eral; neutral vs. flexed hip; unloaded vs. loaded; fast vs. slow velocity; constant-velocity vs.
accelerated vs. decelerated NHEs) on eccentric hamstring strength and fascicle length
12 T. ALT ET AL.
(Presland et al., 2018; Sarabon et al., 2019; Timmins et al., 2015) to develop individualised
training regimen adjusted to different profiles of requirements (e.g., high vs. low physi-
cal performance level).
Practical implications
NHE execution on specific devices (Hegyi et al., 2019; Opar et al., 2013; Sarabon et al., 2019)
or isometric dynamometers (Alt et al., 2018) is recommended to assess exercise determinants,
performance parameters and exercise quality. If not available, wall bars, doorway pull-up bars
or any other solid and rigid object can serve to provide adequate counter-pressure for the
heels. Despite extensive research since 2001, standardised NHE execution is lacking as most
studies execute partner-assisted NHE training (Delahunt et al., 2016; Ditroilo et al., 2013;
Mjolsnes et al., 2004). This procedure has two major disadvantages that impede its scientific
demands. Firstly, the applied resistance from hands to shanks/heels is mostly insufficient for
mature athletes to perform a NHE across the full ROM
knee
. Beyond inadequate strength
capacities of the athletes this deficient fixation results in ROM
DWA
of below 30% (Delahunt
et al., 2016; Ditroilo et al., 2013; Marshall et al., 2015). Secondly, looking from a more
scientific perspective, it is not possible to determine the generated forces/moments quantify-
ing e.g., training stimuli and fatigue and derive adequate exercise determinants.
Assisted NHE execution is recommended to acquire proper exercise quality by lower
inter-individual variability and to reduce the fear of uncontrolled falling, especially within
early training stages of inexperienced athletes or patients. It is advisable to conduct NHEs
always across the full ROM
knee
close to full knee extension (~20° to 0° KF). The intensity of
effort is important to increase hamstring strength and fascicle length (Roig et al., 2009;
Timmins et al., 2016). That’s why a nearly constant movement velocity during the second
half of the NHE’s ROM
knee
(=100% ROM
DWA
) should be particularly paramount where
peak moments occur (Figures 1d, 2, 3c). In order to achieve this guideline, assistance can be
provided by a partner being located in front of the athlete and adjusting the pressure of his
hands to the athlete`s shoulders according to the respective movement velocity.
The use of additional weights during unassisted NHE execution with extended hip
joints should be exclusively limited to experienced athletes to avoid injuries caused by
excessive strain rates (Hegyi et al., 2019). Unilateral NHE training is only recommended
if an uncontrolled falling is restricted by any sort of assistance. Due to increasing inter-set
fatigue during unassisted NHEs which highly varied between participants (Figure 3a-c),
it might be more reasonable to use muscular failure (Toigo & Boutellier, 2006)—
associated with a large increase of angular velocity at the end of a repetition (Figure
1d, rep6 and rep9)—as completion criterion of a set rather than a prescribed repetition
number. This can contribute to more individualised NHE training regimen respecting
actual daily performance state.
Conclusion
Assisted NHEs were favourable to ensure a high execution quality to meet the required
exercise determinants. Furthermore, between-participant variability was significantly lower
in assisted than unassisted NHEs. Only if the athlete’s/patient’s physical performance level
is adequate, controlled full-ROM NHEs in unassisted fashion are recommended to ensure
SPORTS BIOMECHANICS 13
a consistently high hamstring activation throughout the entire movement and subsequent
adaptations at comparably long hamstring muscle length. If these guidelines are fulfilled,
exercise intensity can be reasonably estimated and reproduced. An exercise volume of 3 sets
of 3 eccentric-only NHE repetitions induced substantial fatigue within and across sets and
can be considered as adequate exercise stimulus Further research is needed to establish
which NHE execution modalities elicit the best adaptations.
Acknowledgments
The authors would like to thank all participants who volunteered to participate in this study and
demonstrated great motivation and commitment. Dr. Werner Groß-Alt deserves our sincere
gratitude because his skills in informatics made these insights possible.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
Al Attar, W. S., Soomro, N., Sinclair, P. J., Pappas, E., & Sanders, R. H. (2017). Effect of injury
prevention programs that include the nordic hamstring exercise on hamstring injury rates in
soccer players: A systematic review and meta-analysis. Sports Medicine, 47(5), 907–916. https://
doi.org/10.1007/s40279-016-0638-2
Alt, T., Nodler, Y. T., Severin, J., Knicker, A. J., & Strüder, H. K. (2018). Velocity-specific and
time-dependent adaptations following a standardized nordic hamstring exercise training.
Scandinavian Journal of Medicine & Science in Sports, 28(1), 65–76. https://doi.org/10.1111/
sms.12868
Bohm, S., Mersmann, F., & Arampatzis, A. (2015). Human tendon adaptation in response to
mechanical loading: A systematic review and meta-analysis of exercise intervention studies
on healthy adults. Sports Medicine - Open, 1(1), 7. https://doi.org/10.1186/s40798-015-0009-
9
Brockett, C. L., Morgan, D. L., & Proske, U. (2001). Human hamstring muscles adapt to eccentric
exercise by changing optimum length. Medicine and Science in Sports and Exercise, 33(5),
783–790. https://doi.org/10.1097/00005768-200105000-00017
Burrows, A. P., Cleather, D., Mahaffey, R., & Cimadoro, G. (2020). Kinetic and electromyographic
responses to traditional and assisted nordic hamstring exercise. Journal of Strength and
Conditioning Research, 34(10), 2715–2724. https://doi.org/10.1519/JSC.0000000000003689
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Erlbaum.
Delahunt, E., McGroarty, M., De Vito, G., & Ditroilo, M. (2016). Nordic hamstring exercise
training alters knee joint kinematics and hamstring activation patterns in young men.
European Journal of Applied Physiology, 116(4), 663–672. https://doi.org/10.1007/s00421-015-
3325-3
Ditroilo, M., De Vito, G., & Delahunt, E. (2013). Kinematic and electromyographic analysis of the
nordic hamstring exercise. Journal of Electromyography and Kinesiology, 23(5), 1111–1118.
https://doi.org/10.1016/j.jelekin.2013.05.008
Ebben, W. P. (2009). Hamstring activation during lower body resistance training exercises.
International Journal of Sports Physiology and Performance, 4(1), 84–96. https://doi.org/10.
1123/ijspp.4.1.84
Hanavan, E. P. (1964). A mathematical model of the human body. AMRL-TR-64-102, AD-608-463.
Aerospace Medical Research Laboratories, Wright-Patterson Air Force Base, Ohio.
14 T. ALT ET AL.
Hegyi, A., Lahti, J., Giacomo, J. P., Gerus, P., Cronin, N. J., & Morin, J. B. (2019). Impact of hip
flexion angle on unilateral and bilateral nordic hamstring exercise torque and high-density
electromyography activity. The Journal of Orthopaedic and Sports Physical Therapy, 49(8),
584–592. https://doi.org/10.2519/jospt.2019.8801
Herzog, W., Guimaraes, A. C., Anton, M. G., & Carter-Erdman, K. A. (1991). Moment-length
relations of rectus femoris muscles of speed skaters/cyclists and runners. Medicine &
Science in Sports & Exercise, 23(11), 1289–1296. https://doi.org/10.1249/00005768-
199111000-00015
Marshall, P. W., Lovell, R., Knox, M. F., Brennan, S. L., & Siegler, J. C. (2015). Hamstring fatigue
and muscle activation changes during six sets of nordic hamstring exercise in amateur soccer
players. Journal of Strength and Conditioning Research, 29(11), 3124–3133. https://doi.org/10.
1519/JSC.0000000000000966
Matthews, M. J., Jones, P., Cohen, D., & Matthews, H. (2015). The assisted nordic hamstring curl.
Strength and Conditioning Journal, 37(1), 84–87. https://doi.org/10.1519/SSC.
0000000000000084
Mjolsnes, R., Arnason, A., Osthagen, T., Raastad, T., & Bahr, R. (2004). A 10-week randomized
trial comparing eccentric vs. concentric hamstring strength training in well-trained soccer
players. Scandinavian Journal of Medicine and Science in Sports, 14(5), 311–317. https://doi.
org/10.1046/j.1600-0838.2003.367.x
Opar, D. A., Piatkowski, T., Williams, M. D., & Shield, A. J. (2013). A novel device using the
Nordic hamstring exercise to assess eccentric knee flexor strength: A reliability and retro-
spective injury study. Journal of Orthopaedic & Sports Physical Therapy, 43(9), 636–640.
https://doi.org/10.2519/jospt.2013.4837
Presland, J. D., Timmins, R. G., Bourne, M. N., Williams, M. D., & Opar, D. A. (2018). The effect of
Nordic hamstring exercise training volume on biceps femoris long head architectural
adaptation. Scandinavian Journal of Medicine & Science in Sports, 28(7), 1775–1783. https://
doi.org/10.1111/sms.13085
Roig, M., O’Brien, K., Kirk, G., Murray, R., McKinnon, P., Shadgan, B., & Reid, W. D. (2009). The
effects of eccentric versus concentric resistance training on muscle strength and mass in healthy
adults: A systematic review with meta-analysis. British Journal of Sports Medicine, 43(8),
556–568. https://doi.org/10.1136/bjsm.2008.051417
Sarabon, N., Marusic, J., Markovic, G., & Kozinc, Z. (2019). Kinematic and electromyographic
analysis of variations in Nordic hamstring exercise. PLoS One, 14(10), e0223437. https://doi.org/
10.1371/journal.pone.0223437
Savelberg, H. H., & Meijer, K. (2003). Contribution of mono- and biarticular muscles to extending
knee joint moments in runners and cyclists. Journal of Applied Physiology, 94(6), 2241–2248.
https://doi.org/10.1152/japplphysiol.01001.2002
Timmins, R. G., Bourne, M. N., Shield, A. J., Williams, M. D., Lorenzen, C., & Opar, D. A. (2015).
Short biceps femoris fascicles and eccentric knee flexor weakness increase the risk of hamstring
injury in elite football (soccer): A prospective cohort study. British Journal of Sports Medicine, 50
(24), 1524–1535. https://doi.org/10.1136/bjsports-2015-095362
Timmins, R. G., Shield, A. J., Williams, M. D., Lorenzen, C., & Opar, D. A. (2016). Architectural
adaptations of muscle to training and injury: A narrative review outlining the contributions by
fascicle length, pennation angle and muscle thickness. British Journal of Sports Medicine, 50(23),
1467–1472. https://doi.org/10.1136/bjsports-2015-094881
Toigo, M., & Boutellier, U. (2006). New fundamental resistance exercise determinants of molecular
and cellular muscle adaptations. European Journal of Applied Physiology, 97(6), 643–663.
https://doi.org/10.1007/s00421-006-0238-1
van Dyk, N., Behan, F. P., & Whiteley, R. (2019). Including the Nordic hamstring exercise in injury
prevention programmes halves the rate of hamstring injuries: A systematic review and
meta-analysis of 8459 athletes. British Journal of Sports Medicine, 53(21), 1362–1370. https://
doi.org/10.1136/bjsports-2018-100045
SPORTS BIOMECHANICS 15
