Received: 10 June 2020 

|
  Revised: 15 December 2020 
|
  Accepted: 22 December 2020

DOI: 10.1111/sms.13913

ORIGINAL ARTICLE

Applying a holistic hamstring injury prevention approach in elite

football: 12 seasons, single club study

Luis Suarez-Arrones1,2   | Fábio Yuzo Nakamura3,4  | Rafael A. Maldonado5  |

Nacho Torreno2  | Valter Di Salvo6,7  | Alberto Mendez-Villanueva8

1
Department of Sport and Informatics,
Section of Physical Education and Sport,
The aim was to investigate the preventive effect of a complex training program based
Pablo de Olavide University, Sevilla, Spain on holistic hamstring health understanding in elite professional soccer players. This
2
Performance and Health Department, study involved an elite club in Europe and was conducted over 12 seasons. The last
Basel, Switzerland
2 seasons were the intervention period, and the others were the control seasons.
3
Associate Graduate Program in Physical
During the intervention period, players performed a complex program organized into
Education UPE/UFPB, João Pessoa, Brazil
4
Research Centre in Sports Sciences,
different interventions throughout the week having as a priority the player health.
Health Sciences and Human Development, Hamstring injuries, absenteeism, injury rates, and injury burden between the con-
CIDESD, University Institute of Maia, trol and intervention seasons were compared using a rate ratio (RR) with 95% CI.
ISMAI, Maia, Portugal
5
Players had a mean exposure of 333.5 ± 18.6 hours per season with no significant
Performance Department, Girondins de
Bordeaux, Bordeaux, France differences between the intervention and control seasons. The overall injury rate was
6
Football Performance & Science 3 times lower during the two intervention seasons than during the previous seasons
Department, ASPIRE Academy, Doha, (P < .01); the match injury rate was 2.7 times lower (P < .01) and the training rate 4.3
Qatar
7
times (P < .01). Injury burden was almost 4 times lower during the two intervention
Department of Movement, Human and
Health Sciences, University of Rome “Foro seasons than during the previous seasons (P < .01), and recurrences in the control
Italico”, Rome, Italy group were 10% vs 0% in the intervention group. Hamstring injuries were reduced ~3
8
Qatar Football Association, Doha, Qatar times during the seasons in which elite football players were exposed to multicompo-
nent, complex prevention training with individual approaches based on player needs,
Correspondence
Luis Suarez-Arrones, Faculty of Sport management of training load, individualized physiotherapy treatment, and planned
Sciences, Pablo de Olavide University, Ctra. staff communication, in comparison to the control seasons without a clearly defined
deUtrera km 1, 41013, Sevilla, Spain.
Email: [email protected]
and structured injury prevention intervention.

Funding information KEYWORDS

This study was made possible by NPRP hamstring injury, injury prevention, multifactorial hamstring prevention, soccer, strength training
grant # NPRP 6-1526-3-363 from the Qatar
National Research Fund (a member of Qatar
Foundation). The statements made herein
are solely the responsibility of the authors.
The authors declare that they have no
competing interests.

© 2021 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

Scand J Med Sci Sports. 2021;31:861–874.  |

wileyonlinelibrary.com/journal/sms     861
 |
862       SUAREZ-ARRONES et al.

1  |   IN TRO D U C T ION player availability.14 Based on the foregoing, the aim of this
study was to investigate the preventive effect of a complex
Hamstring muscle tears are the most common injury subtype training program based on holistic hamstring health under-
in male football players and are associated with significant standing in elite professional soccer players.
time loss and high financial costs for the player and clubs.1
Thus, adequate prevention strategies are of major importance
within this cohort. A preventive strategy for reducing ham- 2  |  M ETHODS
string injuries, namely the use of the Nordic hamstring ex-
ercise (NHE), is described in the scientific literature.2-4 The 2.1  | Participants
strategy appears to show good to excellent results, reducing
hamstring injuries in mostly non-elite football players by Male football teams were recruited for the study from the
as much as 65% 2,4 and the rate of recurrent injuries by ap- same club (ACF Fiorentina, Serie A, Italy). This study in-
proximately 85%.4 Moreover, lower levels of eccentric ham- volved an elite club in Europe (playing regularly in the
string strength assessed during the NHE have been shown Europe League).
to increase the risk of future hamstring injuries in different
football codes,5 including soccer.6 While the effect that sys-
tematic application of the NHE could have on the prevention 2.2  |  Design study
of hamstring injuries in elite soccer populations is currently
unknown, and despite its simplicity and proposed effi- A cohort study was conducted over 12 seasons. The last 2
cacy in non-elite football players, injury and re-injury rates seasons were the intervention period (2015-2016, 2016-
have remained consistently high in professional European 2017), and the remaining seasons were the control seasons
footballers.1 (2005-2006, 2006-2007, 2007-2008, 2008-2009, 2009-2010,
Since the nature of hamstring injuries is accepted as mul- 2010-2011, 2011-2012, 2012-2013, 2013-2014, 2014-2015;
tifactorial and complex, involving interactions among vari- Figure  1). During the intervention period, the players per-
ous factors,7 a more holistic approach for hamstring health is formed a multicomponent complex prevention program
likely needed.8 In this regard, Buckthorpe et al,9 in line with organized into different interventions with the players
an emerging body of evidence-based research, have proposed throughout the week: on-field training, strength training (ST)
a holistic approach8 that translated existing knowledge on in the gymnasium (gym), individual training sessions (weak
hamstring muscles injury risk and applies this to a real-life points), physiotherapy treatment, or recovery training ses-
football context. If the proposed or similar holistic approach sions, among others. All the interventions were organized
is able to objectively and sensibly structure the content and based on good communication between the staff and medical
the multiple potential interventions to be applied according team, having as a priority the player health. In the control
to football-specific contents (eg, coaching style) and asso- seasons, players did not perform any structured and/or inte-
ciated training loads, multiple risk factors predisposing to grated prevention intervention program regularly, and no in-
hamstring injury10 and the different injury mechanisms (ie, ertial and motorized devices were implemented during the ST
sprinting, over-stretching, shooting), it could conceivably sessions in the posterior chain. In three of the control seasons,
provide flexible programming accounting for both football the players did not carry out any gym-based ST prescribed
demands on the team and player and the specific weak points by the coaching staff. The present study was approved by the
of each player. Contrary to a pre-established “one-size-fits- Institutional Research Ethics Committee (Qatar Anti-Doping
all,” general protocol,2,4,11 this approach deals with injury Lab. IRB number: E2013000004) and conformed to the rec-
prevention as hypercomplex phenomena connecting several ommendations of the Declaration of Helsinki.
interacting components8,9,12 and often involving the skills
and experience of different professionals to deliver the in-
tervention — for example, coaches, fitness coaches, doctors, 2.3  |  Hamstring injuries and exposure
and physiotherapists. However, the scientific validity of such
complex interventions has never been tested. Only training or match hamstring injuries were included
Preventing hamstring injury strain in elite soccer players in this investigation. The medical staff of the team was
is likely to benefit from a holistic approach with a complex responsible for registering all hamstring injuries for the
program, which considers multiple risk factors, their inter-re- full calendar throughout 12 seasons (same head medical
lations, and their context alongside other injuries.8,9,13 In doctor). A recordable hamstring injury was defined as an
addition, soccer clubs with an structured internal communi- injury received during training or competition that caused
cation between the coaching staff and other professionals (eg, an absence from future football participation, affecting the
medical team) are associated with fewer injuries and greater posterior side of the upper leg, irrespective of the need
SUAREZ-ARRONES et al.   
|
   863

F I G U R E 1   Flowchart of study
population

for medical attention or time lost from soccer activities.15 staff communication and individual management of the play-
Ultrasound and/or magnetic resonance imaging (MRI) ers (Figure S1).
were performed to verify the clinical diagnosis of team
medical staff. All hamstring injuries were registered ac-
cording to the type of injury, and recovery was considered 2.4.1  |  Strength training (ST)
finished when the player was fully recovered to train with
the team and available to be called for a match. Injury in- The ST intervention was carried out in the gym, and ST
cidence was reported in absolute numbers and as an injury sessions were delivered to the whole squad. The program's
incidence rate for number of injuries per 1000 player hours criteria and content were selected and timed according to
during matches and training sessions. Injury severity was current scientific knowledge, as well as practical and clini-
defined as the number of days that elapsed from the date cal experience of the involved staff members (coaches, fit-
of injury to the day of the player's return to full participa- ness coaches, physiotherapists, and medical doctors), on the
tion in team training and availability for match selection,15 biology of muscle injury,16 the different risk factors associ-
and was classified as minimal (1-3 days of active participa- ated with hamstring injuries (ie, reduced flexibility/mobility,
tion missed), mild (4-7 days missed), moderate (8-28 days diminished strength, altered lumbo-pelvic control)10,17 and
missed), or severe (>28 days missed). The number and du- the various mechanisms typically causing the injury (sprint-
ration of training sessions performed and the number and ing, over-stretching and shooting).1 As such, the ST ses-
duration of matches played (official and friendly games) sions integrated exercises directed to (i) improve hamstring
were recorded weekly for all the seasons by the medical muscle tissue quality, structure, and strength, (ii) correcting/
team and/or staff, using excel sheets, or specific software minimizing potential risk factors, and (iii) recreating the dif-
(SoccerSystemPro®) in the last two seasons. ferent injury mechanisms. To improve hamstring muscle tis-
sue quality, structure, and strength, at least one hip- and one
knee-dominant hamstring exercise was included. Particular
2.4  | Intervention attention was placed to exercises that targeted the biceps long
head eccentrically at outer ranges, the final goal being de-
The multicomponent complex prevention program consisted veloping stronger hamstrings at longer lengths.18 To correct/
mainly of 6 components: i) strength training including spe- minimize potential risk factors, a combination of exercises
cific exercises for the posterior chain; ii) control of on-field that activated hip extensors (eg, gluteus maximus), hip sta-
training; iii) physiotherapy treatment; iv) training load man- bilizers (eg, gluteus medium), proximal lumbo-pelvic mus-
agement; v) individual training (weak points); and vi) club culature and plantar flexors (ie, calf muscles) was employed.
 |
864       SUAREZ-ARRONES et al.

In addition, hip flexors mobility mainly via eccentric con- previous seasons, although never in such a systematic way
tractions was included. Recreating, and possibly overloading, every week as in previous studies.2,4
the different injury mechanisms that typically lead to a ham-
string strain was achieved by replicating partially or totally
the injury mechanism and manipulating (to overload) the in- 2.4.2  |  On-field training
tersegmental body positioning relationships and joint angles
of ankle, knee, hip, and trunk. On-field training usually consisted of a variety of exercises
Strength training was usually structured as circuit training in order to mimic the mental, physiological, and running de-
before the football exercises on the field. Players performed mands of match play, trying to reproduce strategically and
one or two sets (based on weekly matches) of a circuit con- tactically what would happen in the next game. Usually, the
sisting of 10-12 exercises (between 6-10 reps per exercise) tactical dimension was the dominant training component for
mainly focusing on the lower limbs. In addition, complemen- the design of the different soccer drills or conditioned games
tary neuromuscular training sessions were scheduled during based on the game model. This implies that tactical, techni-
the season with exercises for upper-body, core, and lum- cal, physiological and psychological elements were routinely
bo-pelvic stability; and activation-training sessions consist- trained integrated in the proposed soccer drills. The game
ing of neuromuscular training exercises in the gym or on the model during the two intervention seasons tried to highlight
field, as an initial part of the training session and before the and enhance the best features and capabilities of the play-
specific football drills. ST sessions in the gym lasted approxi- ers. During the weeks with one game, the three main training
mately 30-40 minutes, while complementary ST sessions and days (acquisition days) alternate the physical fitness compo-
collective activation-training sessions lasted 15-20 minutes. nents. This was done by either prioritizing strength in the first
An example of the organization for the strength training con- acquisition day (ie, Wednesday) using small spaces, and en-
tents during a microcycle with one or two games per week is durance and/or speed in the second and third acquisition days
shown in Table 1. (ie, Thursday and Friday) in larger spaces. When the players
The ST combined different sources of external resistance had limited exposure to high-speed running (HSR) over the
such as free-weights and other methodologies with greater week due to the use of small spaces during the football drills,
eccentric overload such as non-gravity dependent flywheel or low participation in official games, additional HSR and
inertial devices (Kbox®, Yo-Yo flywheel technology®, sprint training using individualized speed zones (based on
Versa-Pulley®) or motorized devices (Exentrix®) for a stron- maximal sprinting speed) was prescribed after training, im-
ger eccentric overload. Elastic cords/bands, sliding boards, mediately after a game, or during additional training sessions
suspension training, pneumatic cable devices, and the own for certain players (ie, regular non-starters). An example
body weight were also used. To determine the training inten- of the organization of the on-field training contents during
sity (ie, external load) to be employed during the ST, a veloc- a microcycle with one or two games per week is shown in
ity-based assessment with incremental loads in free-weights Figure 2.
and different inertias with flywheel technology and con-
ic-pulley was performed in the most employed exercises.19
The inertia/load with which the player achieved higher aver- 2.4.3  |  Physiotherapy treatment
age power was used during the training (SmartCoach Power
Encoder SPE-35, SmartCoach Europe AB, Stockholm, Physiotherapy played a main role in controlling the changes
Sweden). The inertia/load was individually readjusted every in the mechanical properties and viscoelastic characteris-
4 weeks. For the exercises that employed other source of re- tics of the muscular system due to the high and repetitive
sistance (eg, elastic cords, mini-bands, sliding), exercise in- loads during training and matches. Depending on the player's
tensity was regulated subjectively (as usually done by fitness needs, different physiotherapy techniques were applied be-
trainers). In addition to overloads related to manipulations in fore and after training, such as myofascial induction in order
external resistance and speed of execution, overloads in many to readjust the mechanical properties of the connective tis-
exercises were also achieved by manipulating motor control/ sue, the degree of pretension of the fascial system, and to
coordination, muscle lengths, and mechanical perturbations. regulate the nociceptive mechanisms. In contrast to the con-
A wide variety of exercises were employed thorough both trol seasons, during the intervention period minimal invasive
seasons aimed at bringing variability and avoiding stagna- therapy guided by ultrasound, such as intra-tissue percutane-
tion and monotony. Figures S2 and S3 illustrate some of the ous electrolysis, was employed to promote greater functional
exercises focusing on the lower limbs, and Figures  S4–S8 recovery in stiff scars or fibrosis as a preventive strategy in
exemplify some of the many exercises prescribed to players players with previous hamstring strains and tendinopathies,20
for the posterior chain. NHE was not employed during the and percutaneous neuromodulation in order to induce an im-
intervention period (2 seasons) but it was used in some of the mediate increase in balance and muscular strength.21 Almost
T A B L E 1   An example of the organization for the strength training contents during a microcycle with one (A) or two (B) games a week

(A)

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

• Game Video-Based Feedback REST • Strength Training GYMa  • Video Analysis of the Opponent • Individual Activations • Video Opponent Analysis GAME
• Players with > 60 min: • On the Pitch • Collective Activation GYM/ • On the Pitch • Individual activations
SUAREZ-ARRONES et al.

- Individual • Group Exercise: Specific Pitch • Group Exercise: Specific • On the Pitch
Training + Recovery Protocol Rondos • On the Pitch Technical Drill of Pass • Group Exercise: Generic Rondos
• Players with < 60 min: • Specific Neuromuscular • Group Exercise: Offensive • Collective Exercise: Off • Collective Exercise: Off Org/Off
- Group Exercise: Generic circuit training with Organization/Build it Up Org/Off Combined actions Evolutions
Technical Drill of Pass technical finishing Defensive Line + Midfielders with Finishing & Crossing • Collective Exercise: Off & Def
- Group Exercise: Offensive • Collective Exercise: with High Pressing • Collective Exercise: Set Pieces/ Isolated
Evolutions Defensive Organization/ • Group Exercise: Off-Def Game model Application / • Collective Exercise: Game model
- Group Exercise: Specific Ball Defensive Evolutions Transitions Conditioned & Real Game Application / Conditioned & Real
Possession/MSP • Group Exercise: Off-Def • Group Exercise: Off Org/ 11 vs 11 LSP Game 11 vs 11 with integrated
- Group Exercise: Competition Organization/SSGs by Specific MSP 7 vs 7 + JK+GK • Complementary Set Pieces
Exercise - SSGs by Positions: Positions: 6v6 + GK´s • Complementary Neuromuscular Neuromuscular Training • Complementary Neuromuscular
5v5 + GK´s + HIT short Training / Individual Trainingb  Training / Individual Training
intervals (optional)
• Complementary Neuromuscular
Training

(B)

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

• Video Game Feedback • Video Opponent Analysis • Video Opponent Analysis GAME • Video Game Feedback • Video Opponent Analysis GAME
• Players with > 60 min played: • Individual activations • Individual activations • Players with > 60 min • Individual activations
• Individual Neuromuscular • On the Pitch • On the Pitch played: • On the Pitch
Training + Recovery Protocol • Group Exercise: Specific • Group Exercise: Specific Rondos • Recovery Protocol • Group Exercise: Generic
• Players with < 60 min: Technical Drill of Pass • Collective Exercise: Def Set Pieces/ • Players with < 60 min: Rondos
• Strength Training GYMc  • Collective Exercise: Def Isolated • Group Exercise: Generic • Collective Exercise: Off &
• Group Exercise: Generic Org/ Def Evolutions • Collective Exercise: Application Technical Drill of Pass Def Set Pieces/ Isolated
Technical Drill of Pass • Group Exercise: Specific MofG/ Conditioned & Real Game • Group Exercise: Specific • Collective Exercise:
• Group Exercise: Specific Ball Ball Possession/MSP 11 vs 11 Ball Possession/MSP Specific Ball Possession/
Possession/MSP • Complementary • Group Exercise: Competition LSP
• Group Exercise: Competition Neuromuscular Training Exercise - SSGs by • Individual Training
Exercise - SSGs by Positions: / Individual Training Positions: 5v5 + GK´s + HIT (optional)
5v5 + GK´s + HIT short (optional) Short Intervals
intervals

Abbreviations: GK, goal keeper; HIT, high intensity running interval training; LSP, large side possession; MF, midfielders; MofG, model of game; MSP, medium side possession; SSG, small side game.
a
Example of possible exercises used in a session like this = 2 × Figure 4.
  

b
|

Example of possible exercises in this type of session = 2 × Figure 6.

c
Example of possible exercises used in this type of session = 2 × [No 1 (Figure S8); No 4 (Figure S3); No 2 (Figure S2)+No1 (Figure S3); No 1 (Figure S5) + No 4 (Figure S6)].
   865
 |
866       SUAREZ-ARRONES et al.

daily, the players were checked and treated by the physio- sessions and games on the field). In addition, running activity
therapists. Based on the player's need, the physiotherapists in parameters were monitored using GPS devices for the ex-
conjunction with the medical doctor decided if any treatment ternal load during the on-field training sessions and games.
should be made with the player, and what kind of treatment. As an example, characterization of the weekly external load
profile using some of the different parameters monitored is
shown in Figure 2. During the season the players completed
2.4.4  |  Training load management a subjective well-being questionnaire every morning assess-
ing their fatigue, sleep quality, and muscle soreness in order
Individual management of player load was employed to manage the training load of the day. Using the match ef-
throughout the season in order to optimize performance. A fort as a reference, training load was planned and prescribed
specific software was used integrating all the information aiming to keep the weekly load pattern as stable as possible
(SoccerSystemPro®). Internal training load was calculated (ie, avoiding big changes in acute loads compared to chronic
using the session rating of perceived exertion (S-RPE) 22 by loads) in order to have a good balance between fitness, fa-
two fitness coaches after each training session (all the neu- tigue, recovery, and injury risk. This was especially empha-
romuscular training sessions in the gym and all the training sized during the periods in which players could train with

F I G U R E 2   An example of the organization for the on-field training contents during a microcycle with one or two games per week. TD: total
distance covered; DC: distance covered; #Acc + Dec: number of accelerations + decelerations. Compensatory training for players without, or
with low participation in the game. The duration of the training on Tuesday and Wednesday in the microcycle with two games was based on their
participation in the previous game
SUAREZ-ARRONES et al.   
|
   867

higher volumes and intensities (ie, pre-season, weeks with the previous season) during a typical microcycle is shown in
only one game). As expected, approximately half of the play- Figure 3. This individual work was carried out in combina-
ers belonging to the squad participated for only a few min- tion with the ST of the whole team (Figure 4). Additionally,
utes or did not participate at all in the official match. Based the same player was subjected to: (i) an individual activation
on this, compensatory training sessions to simulate (at least before an on-field training session in large spaces with a high
partially) the volume and intensity of the missing match load volume of HSR (Figure  5), (ii) and individual ST after the
were prescribed (sometimes it was a friendly game with play- on-field training (Figure 6), and (iii) an individual activation
ers from the reserve team and the academy). before an official game (Figure 7).

2.4.5  |  Individual training (weak points) 2.4.6  |  Club staff communication

Individual training sessions targeted primarily players with a A structured and planned internal communication was organ-
history of hamstring injuries, as previous hamstring injuries ized daily. For the first time in the club, two daily meetings
have been identified as probably the most predisposing fac- were scheduled involving a medical doctor, physiotherapists,
tor for suffering a hamstring injury.10 Thus, the individual and fitness coaches. The first meeting was prior to the tech-
training was mainly organized as secondary prevention for nical staff meeting (coaches and fitness coaches) in order to
a specific group of players with a theoretically heightened establish which players would have full participation in the
risk of suffering a hamstring injury episode. The individual training, and which players needed individualized care. The
training sessions were mostly carried out in the gym and fo- second meeting was at the end of the workday in order to col-
cused on strengthening the posterior chain. Sessions’ crite- lect information on the status of the players and feedback on
ria and content were selected and organized according to the the trainings and treatments.
player's weak points. Specifically, sessions were organized
around one or more of the following aspects: imbalances or
asymmetries in posterior chain strength, eccentric tensions 2.5  |  Statistical analysis
in old scars or fibrosis, changes in the muscle morphology
or architecture in the hamstring or gluteal muscles, or in a Descriptive statistics for playing exposure parameters and
specific region along the length of the muscle, mobility lim- characteristics of the hamstring injuries are presented as
its in the contralateral hip flexors and lumbo-pelvic control. means and standard deviations. In addition, the severity and
Sometimes, HSR interval training after strengthening of the location of the injury are described using frequencies and
hamstring muscles and under fatigue conditions was also percentages. Injury rate was calculated as the number of in-
included in the individual training (ie, runs at 90%-95% of juries per 1000 h with corresponding 95% confidence inter-
maximal sprinting speed). The individual training was usu- vals (CI). Injury burden was calculated as the number of lay
ally planned after the football field training, or during an off days per 1000 h (injury rate × mean absence per injury).
additional training session, and lasted 10-25 min. As an ex- Injury rates and injury burden between the control and inter-
ample, the organization of the different training interventions vention seasons were compared using a rate ratio (RR) with
with a high-risk player (a biceps femoris long head injury in 95% CI. The significance level was set at P < .05.

F I G U R E 3   An example of the organization of the different training interventions with a high-risk player, who suffered a biceps femoris long
head injury in the previous season, during a microcycle with one game per week. ST, strength training; Ind, individual; Col., collective
 |
868       SUAREZ-ARRONES et al.

(A1) (B1) (A2) (B2) (A3) (B3)

(A5)
(B5)

(A4) (B4)

(A6) (B6)

(A7) (B7) (A8) (B8)

(A9) (B9)

(A10) (B10)

F I G U R E 4   An example of a collective strength training in the GYM. The training consisted of 2 × 10 exercises (between 6-10 reps per
exercise) mainly focusing on the lower limbs

3  |   R ES U LTS were no significant differences between match and train-

ing exposure between the intervention and control seasons
The mean ± SD age, body mass, height, and fat mass (Faulkner) (Table 2).
of the soccer players were 27.8 ± 2.7 years, 77.8 ± 4.4 kg,
1.83 ± 0.09 m, and fat mass (Faulkner) 10.1 ± 0.5, respec-
tively. All the players had more than 8 years of experience in 3.2  |  Hamstring Injuries
a professional soccer environment.
Seasons 2009-2010, 2012-2013, and 2013-2014 were not During the 9 seasons analyzed, 55 hamstring injuries were
included in the data analysis as not all the information related recorded in 228 players, of which 39 (70.9%) occurred during
to the training exposure was available. The hamstring injuries match play and 16 (29.1%) during training. On average, 30%
during these 3 seasons were 8.0 ± 1.0, with no differences in of players sustained at least one hamstring injury during a
comparisons with the other 7 seasons selected as a control season in the control group vs 10% in the intervention group.
group (7.9 ± 1.6). Recurrences in the control group were 10% vs 0% in the in-
tervention group. Details on the hamstring injury pattern are
presented in Table 2.
3.1  |  Exposure to training and matches The overall injury incidence rate was 3 times lower during
the two intervention seasons than during the seven previous
Players involved in the study had a mean exposure during seasons (RR 3.0; 95% CI 2.9-3.1, P  <  .0001); the match
all seasons of 333.5  ±  18.6 (range from 306.1 to 359.1) injury rate being 2.7 times lower (RR 2.7; 95% CI 0.0-0.0,
hours per season. The mean training and match exposure P  <  .0001) and the training rate 4.3 times (RR 4.3; 95%
were 254.7  ±  20.6 (range from 233.3 to 282.3) hours and CI 0.0-0.0, P  <  .0001). Injury burden was almost 4 times
78.8 ± 8.7 (range from 69 to 93) hours, respectively. There lower during the two intervention seasons than during the
SUAREZ-ARRONES et al.   
|
   869

F I G U R E 5   An example of different exercises used during an individual activation for the posterior chain before an on-field training in larger
spaces with high volume of high-speed running

(A2) (B2)
(A1) (B1)

(A3) (B3) (A4) (B4)

F I G U R E 6   An example of different exercises used during an individual strength training after the on-field training with a player who suffered
a biceps femoris long head injury in the previous season

seven previous seasons (RR 3.6; 95% CI 3.6-3.7, P < .0001; approaches based on players’ needs, optimal management of
Table 2). training load, individualized physiotherapy treatment, and
optimal staff communication, in comparison to the control
seasons without a clearly defined and structured injury pre-
4  |   D IS C U SS ION vention intervention. Furthermore, during the two interven-
tion seasons there were no re-injuries. Hence, it is suggested
The main finding of this study was that the incidence of that complex training approaches could be effective in re-
hamstring injuries was reduced ~3 times during the sea- ducing the injury occurrence in elite level football athletes.
sons in which elite football players were exposed to mul- In the current study, hamstring injury rates during the 7
ticomponent, complex prevention training with individual control seasons ranged from 0.60 to 1.20 (0.93 on average),
 |
870       SUAREZ-ARRONES et al.

F I G U R E 7   An example of different exercises used during an individual activation before a game

which lie within the lower ranges of previously reported val- that comparisons between different countries, leagues, and
ues 0.90-1.30 (1.20 on average) in a comparable population performance levels are complicated since training interven-
(ie, professional football clubs in Europe).1 Similarly, the tions, prophylactic measures, and preventive strategies that
hamstring injury burden was typically lower, between 8 to could potentially be implemented are heavily influenced by
19  days (10.2 on average) during the control season com- the amount of football practice (both games and training) a
pared with the European clubs, 12-29  days (19.7 on av- particular team is exposed to. For example, the total exposure
erage).1 It is important to note that in contrast to Ekstrand time per season reported for football teams from Norway and
et al1 with data from 36 different clubs, in our case the same Iceland (~5100 h and ~2030 h, respectively), differs substan-
person (the club's Head of Medicine) was the only person tially from our control and intervention groups (~8200 h) and
responsible for collecting all injury-related data during the from elite European clubs (~6700 h).1
12 seasons of the study. Even though the injury rate and in- The percentage of hamstring re-injury in the con-
jury burden were not particularly elevated during the control trol seasons in the present study (10%) was slightly lower
seasons, the two intervention seasons yielded a significant than previously reported (13%) in elite European clubs.1
reduction in these markers, 0.31 and 3.00 for injury rate and These numbers are far lower than the 39% of re-injuries re-
burden, respectively. Similar data in high-level professional ported among the teams that did not follow the 2 seasons
football players are currently lacking so comparisons are not of eccentric training with the Nordic hamstring exercise
possible. In lower level competitions (Norway and Iceland in Norwegian and Iceland football teams.2 The teams that
league2), a 2 season intervention consisting of eccentric used the eccentric training with the Nordic hamstring exer-
strength training with the Nordic hamstring exercise resulted cise had a similarly high proportion of re-injuries (36%).3
in a significant reduction in the overall incidence of ham- Our approach to improving hamstring protection resulted
string injury rates/1000 h compared with teams that did not in no re-injuries (0%) during the 2 seasons of intervention.
use the program (0.22 vs 0.62). Thus, both injury figures, the In this regard, the training/preventive approach adopted by
intervention (0.22) and the control group (0.62) were lower Arnason et al2 was characterized by targeting a single risk
compared with our intervention seasons (0.31) and control factor: knee flexor eccentric strength with a single exercise
seasons in the current study (0.93) and in European profes- (the Nordic hamstring). Although several investigations have
sional football players (1.20).1 It should be acknowledged quantified muscle use during the Nordic hamstring exercise
SUAREZ-ARRONES et al.   
|
   871

T A B L E 2   Comparison of the intervention and control groups of interest; professional football players.13,25 Significant in-
Intervention Group Control Group creases in muscle use were detected mainly in the semitendi-
(n = 49 players) (n = 179 players) nosus and short head of the biceps femoris muscles.25 Albeit
(n = 2 teams/seasons) (n = 7 teams/seasons) speculative, it is possible that even in the case that eccentric
Exposure per season (h) strength developed via the Nordic hamstring exercise had an
initial protective effect (ie, reduction in injury occurrence),
Total 333.4 ± 2.2 333.5 ± 21.5
the lack of other important stimuli in other muscles and/or
Matches 80.3 ± 1.1 78.4 ± 10.0
movements might limit the long-term hamstring protection
Training 253.2 ± 3.3 255.1 ± 23.8
(ie, high recurrence rate). For instance, in professional foot-
Hamstring Injuries per season ball players the Nordic hamstring exercise has shown negli-
Total 2.5 ± 0.7 7.7 ± 1.6 gible involvement of the long head of the biceps femoris,25
Matches 2 ± 0 5.6 ± 1.0 which is the most commonly injured muscle belly in football
Training 0.5 ± 0.7 2.3 ± 1.0 players.1,11 In addition to this, the return to play process and
Soccer 22.0 ± 1.4 84.4 ± 29.2 hamstring injury prevention in professional soccer requires
absenteeism a complex and holistic approach, which considers multiple
(days) per factors and their inter-relations implemented effectively.7,12
season Thus, in addition to strength training, the integration of all the
Soccer 9.1 ± 2.0 11.2 ± 4.3 other factors proposed in the current intervention (individual
absenteeism approaches based on players’ needs, optimal management of
(days) per
training load, individualized physiotherapy treatment, and
injury
optimal staff communication) might be of vital importance
Recurrent 0 1.0 ± 0.6
to successfully return players to play and decrease the re-
Injuries
currence rates. Nevertheless, similar to the injury rates and
Severity
as mentioned above, these hamstring re-injury comparisons
Minimal 0 0 between different countries, leagues, and performance levels
(1-3 d)
should be interpreted with caution.
Mild (4-7 d) 2 (50) 27 (50) In addition to the multiple actions directed toward increas-
Moderate 2 (50) 25 (46) ing hamstring protection taken in the present intervention (see
(8-28 d) methods), the strength training implemented here was based
Severe 0 2 (4) on the integration (complementing the football-specific train-
(> 28 d) ing) and content (multifactorial and performance-driven).
Location of The selection of content was based on a detailed and func-
Injury tional analysis of the different risk factors additional to the
BFl 3 (75) 36 (67) knee flexor eccentric strength, such as hip extensor strength
BFs 0 0 (avoiding functional imbalances between hamstring and glu-
ST 1 (25) 12 (22) teus maximus), intra- and intermuscular coordination, hori-
SM 0 6 (11) zontal strength, hamstring outer ranges strength, hamstring
Hamstring and ankle flexibility/mobility, hip stabilizers strength, ec-
Injury centric training at high velocities or proximal lumbo-pelvic
rate/1000 h stability, potentially associated with the hamstring injury 9,26-
29
Total 0.31 (0.30-0.32) 0.93 (0.91-0.93) ; and the main mechanisms causing the injury (high speed,
Matches 1.02 (0.99-1.05) 2.78 (2.75-2.81) acceleration, stretching/lengthening actions, forward trunk
Training 0.08 (0.07-0.09) 0.35 (0.34-0.36)
movements). Subsequently, the contents were organized on
a daily basis to synergize (eg, activation exercises before a
Hamstring 2.78 (2.76-2.81) 10.12 (10.10-10.14)
Injury Burden session requiring high speed) the football-specific training
content, therefore ensuring optimal physical fitness readiness
Abbreviations: BFl, biceps femoris long head; BFs, biceps femoris short head;
to complete the most important training content (ie, football
ST, semitendinosus; SM, semimembranosus.
**P < .001.
training). Moreover, non-modifiable factors such as age5,30
and previous injury history5,30 were dealt with at an individ-
ual level (see in the intervention section, methods), mainly
with recreationally active people,23,24 only two of those stud- by addressing other potential companion modifiable factors.
ies have, to date, examined changes in muscle use via func- Somehow surprisingly, compliance, specifically the
tional MRI during the Nordic hamstring in the population lack of it with the highly scientifically efficacious Nordic
 |
872       SUAREZ-ARRONES et al.

hamstring protocol, has been indicated as a cause to explain of the risk factors are dichotomous (eg, low or high) and a nu-
the failure in reducing the occurrence of hamstring injuries in anced, so multi-layered appraisal is typically needed before
professional football.31 In this regard, Goode et al32 suggested an intervention can be proposed. Overall, this might illustrate
that future studies and training programs should focus on the the inherent difficulty of implementing experimental results
implementation of eccentric strength training, highlighting derived from single-factor interventions without considering
the successful compliance. In the same line, Bahr et al31 con- the multiple potential factors that combine to determine ham-
cluded that more attention is needed on the execution of the string injury risk in an elite football setting.
NHE program at the highest level of male football in Europe,
as the adoption is too low. In the present study, compliance
with the holistic intervention was absolute, except when 4.1  | Limitations
illness or injury caused a player to be absent from training
participation. From coaches and technical staff members, Due to the hierarchical dominance of football-specific
our complex program was intrinsically ingrained into the training and matches over any other aspect, the contents of
football-training program and delivered to the entire team. the other multiple components of the proposed intervention
Considering the inherent complexity of a football season (eg, cannot be pre-standardized. As such, a similar intervention
multiple domestic and international competitions, extensive might well fail to produce the same results as reported here
traveling or potential ad hoc changes in staff and players) if football-specific contents, the strength training, train-
and associated competitive calendars, pre-established, fixed ing load management, physiotherapy treatment, individual
“prevention” protocols are believed to be unrealistic. In our training, or the structured and planned communication are
case, we did not use a pre-prescribed set of fixed sets and reps not taken into account. The way players were progressed or
of a particular exercise labeled as the “hamstring prevention identified to be prescribed with particular exercises, load
protocol,” which was an optional or recommended program management, and/or physiotherapy treatments can always
where each athlete was individually responsible for his own be open to interpretation. Different staffs can interpret the
training as in previous studies.11,33 In our case, all the training same information very differently, which will result in the
sessions were prescribed, executed, and supervised by mem- implementation of potentially dissimilar interventions.
bers of the coaching staff and were integrated into the daily This study did not follow a recommended randomized con-
routine of the professional players. trolled trial (RCT) design. Although we recognize the ne-
It is important to note that, at individual level, each of cessity of implementing the complex program proposed in
the controlled factors in the present study represents a po- this study in future RCTs, the difficulties of implementing
tentially relative risk factor and do not represent a quantita- this design in an elite level football club need to be high-
tive measure. Interactions among potential modifiable and lighted (ie, in the current case, competing in the Europa
non-modifiable risk factors are complex and are likely to League). Additionally, considering that we conducted the
be mutually reinforcing. Compound risk situations are very experiment trying to keep the design as ecological and
likely to occur in elite soccer. For example, a hamstring “real life” as possible, the lower injury rates observed dur-
injury in an elite football setting might be attributed to the ing the two seasons of intervention could not be attributed
combination of high residual muscle fatigue due to an inap- to a specific reason. The different components of the cur-
propriately placed posterior chain strength training, a very rent intervention may have affected multiple variables/fac-
demanding competitive game (which leads to high volume tors that could potentially impact the injury rates observed
of sprinting distance), and a lack of proper channels of com- here. As such, it is not possible to isolate the role that each
munication between the medical and coaching staff. On the individual component could have on the reduced injury
other hand, interactions among the different risk factors can rates observed. Thus, future studies should try to statisti-
be also be mutually excluding. This was shown, for example, cally control and evaluate the impact of each of the several
in Australian Football rules and soccer, where the effect of elements included in such multicomponent and complex
age (a non-modifiable risk factor) was counteracted by high interventions.
levels of eccentric strength.5,6 Additionally, some factors are
likely to behave in a U-shape, with low and high inputs being
equally (potentially) harmful. For example, both relatively 5  |  PERSPECTIVE
low and high accumulated sprinting distances covered in
training sessions plus games are likely to be risk factors.34 The present study showed that the incidence of hamstring in-
For some factors, timing is equally (if not more) important juries was reduced ~3 times after multicomponent and com-
than their actual magnitude; it is not the same to deal with a plex prevention training during two soccer seasons, when
potential sprinting distance overload 48 hours before the next compared to the previous 9 seasons without any systematic
competitive game than 72 hours. Additionally, virtually none intervention to promote injury prevention. Furthermore, there
SUAREZ-ARRONES et al.   
|
   873

were no re-injuries during the two intervention seasons, thus 7. Mendiguchia J, Martinez-ruiz E, Edouard P, et al. A multifactorial,
reinforcing the necessity to design multiple and individually criteria-based progressive algorithm for hamstring injury treat-
ment. Med Sci Sports Exerc. 2017;49(7):1482-1492.
adjusted training and treatment for the lower limbs and pos-
8. Oakley AJ, Jennings J, Bishop CJ. Holistic hamstring health:
terior chain. Future RCTs might be needed to compare the not just the Nordic hamstring exercise. Br J Sports Med.
impact of the current intervention with the NHE program. 2018;52(13):816-817.
Lastly, considering the relatively high complexity associated 9. Buckthorpe M, Gimpel M, Wright S, Sturdy T, Stride M. Hamstring
with the multi-layered intervention presented here, which muscle injuries in elite football: translating research into practice.
requires a rather large number of highly specialized profes- Br J Sports Med. 2018;52(10):628-629.
sionals, its implementation in sub-elite and amateur football 10. Green B, Bourne MN, van Dyk N, Pizzari T. Recalibrating the risk
of hamstring strain injury (HSI): a 2020 systematic review and me-
settings, where hamstring injuries are also frequent, might
ta-analysis of risk factors for index and recurrent hamstring strain
not be realistic.
injury in sport. Br J Sports Med. 2020;54(18):1081-1088.
11. Askling C, Karlsson J, Thorstensson A. Hamstring injury occur-
ACKNOWLEDGMENTS rence in elite soccer players after preseason strength training with
The authors gratefully acknowledge Dr Paolo Manetti (medi- eccentric overload. Scand J Med Sci Sports. 2003;13(4):244-250.
cal officer (2002-2008) and head of sports medicine (2008- 12. Buckthorpe M, Wright S, Bruce-Low S, et al. Recommendations
2017) in ACF Fiorentina) for his technical support to this for hamstring injury prevention in elite football: translating re-
research during the experiments and data provided. The au- search into practice. Br J Sports Med. 2019;53(7):449-456.
13. Mendez-Villanueva A, Suarez-Arrones L, Rodas G, et al. MRI-
thors thank the ACF Fiorentina players, the other coaches
based regional muscle use during hamstring strengthening exer-
(Paulo Sousa, Victor Sanchez, Manuel Cordeiro, Damir cises in elite soccer players. PLoS One. 2016;11(9):e0161356.
Blokar), medical staff members (Luca Pengue, Luca Lonero, 14. Ekstrand J, Lundqvist D, Davison M, D'Hooghe M, Pensgaard
Stefano Dainelli, Simone Michelassi, Francesco Tonarelli, AM. Communication quality between the medical team and
Cristian Petri), and board members. the head coach/manager is associated with injury burden and
player availability in elite football clubs. Br J Sports Med.
CONFLICT OF INTEREST 2019;53(5):304-308.
The authors declare no conflict of interest. 15. Fuller CW, Ekstrand J, Junge A, et al. Consensus statement on in-
jury definitions and data collection procedures in studies of foot-
ball (soccer) injuries. Br J Sports Med. 2006;40(3):193-201.
ORCID 16. Jarvinen TA, Jarvinen TL, Kaariainen M, Kalimo H, Jarvinen
Luis Suarez-Arrones  https://orcid. M. Muscle injuries: biology and treatment. Am J Sports Med.
org/0000-0002-5432-4786 2005;33(5):745-764.
17. Sherry MA, Best TM. A comparison of 2 rehabilitation programs
R E F E R E NC E S in the treatment of acute hamstring strains. J Orthop Sports Phys
1. Ekstrand J, Walden M, Hagglund M. Hamstring injuries have in- Ther. 2004;34(3):116-125.
creased by 4% annually in men's professional football, since 2001: 18. Proske U, Morgan DL. Muscle damage from eccentric exercise:
a 13-year longitudinal analysis of the UEFA Elite Club injury mechanism, mechanical signs, adaptation and clinical applications.
study. Br J Sports Med. 2016;50(12):731-737. J Physiol. 2001;537(Pt 2):333-345.
2. Arnason A, Andersen TE, Holme I, Engebretsen L, Bahr R. 19. Suarez-Arrones L, Saez de Villarreal E, Núñez FJ, et al. In-season
Prevention of hamstring strains in elite soccer: an intervention eccentric-overload training in elite soccer players: Effects on
study. Scand J Med Sci Sports. 2008;18(1):40-48. body composition, strength and sprint performance. PLoS One.
3. van Dyk N, Behan FP, Whiteley R. Including the Nordic hamstring 2018;13(10):e0205332.
exercise in injury prevention programmes halves the rate of ham- 20. Abat F, Diesel WJ, Gelber PE, Polidori F, Monllau JC, Sanchez-
string injuries: a systematic review and meta-analysis of 8459 ath- Ibanez JM. Effectiveness of the Intratissue Percutaneous
letes. Br J Sports Med. 2019;53(21):1362-1370. Electrolysis (EPI(R)) technique and isoinertial eccentric exercise
4. Petersen J, Thorborg K, Nielsen MB, Budtz-Jorgensen E, Holmich in the treatment of patellar tendinopathy at two years follow-up.
P. Preventive effect of eccentric training on acute hamstring inju- Muscles Ligaments Tendons J. 2014;4(2):188-193.
ries in men's soccer: a cluster-randomized controlled trial. Am J 21. de la Cruz-Torres B, Barrera-Garcia-Martin I, Albornoz-Cabello
Sports Med. 2011;39(11):2296-2303. M. Immediate effects of ultrasound-guided percutaneous neuro-
5. Opar DA, Williams MD, Timmins RG, Hickey J, Duhig SJ, modulation versus physical exercise on performance of the flexor
Shield AJ. Eccentric hamstring strength and hamstring in- hallucis longus muscle in professional dancers: a randomised clin-
jury risk in Australian footballers. Med Sci Sports Exerc. ical trial. Acupuncture Med. 2019;37(2):91-97.
2015;47(4):857-865. 22. Foster C, Florhaug JA, Franklin J, et al. A new approach to moni-
6. Timmins RG, Bourne MN, Shield AJ, Williams MD, Lorenzen toring exercise training. J Strength Cond Res. 2001;15(1):109-115.
C, Opar DA. Short biceps femoris fascicles and eccentric knee 23. Bourne MN, Williams MD, Opar DA, Al Najjar A, Kerr GK,
flexor weakness increase the risk of hamstring injury in elite Shield AJ. Impact of exercise selection on hamstring muscle acti-
football (soccer): a prospective cohort study. Br J Sports Med. vation. Br J Sports Med. 2017;51(13):1021-1028.
2016;50(24):1524-1535.
 |
874       SUAREZ-ARRONES et al.

24. Bourne MN, Opar DA, Williams MD, Al Najjar A, Shield AJ. 32. Goode AP, Reiman MP, Harris L, et al. Eccentric training for pre-
Muscle activation patterns in the Nordic hamstring exercise: Impact vention of hamstring injuries may depend on intervention com-
of prior strain injury. Scand J Med Sci Sports. 2016;26(6):666-674. pliance: a systematic review and meta-analysis. Br J Sports Med.
25. Fernandez-Gonzalo R, Tesch P, Linnehan R, et al. Individual mus- 2015;49(6):349-356.
cle use in hamstring exercises by soccer players assessed using 33. Engebretsen AH, Myklebust G, Holme I, Engebretsen L, Bahr R.
functional MRI. Int J Sports Med. 2016;37(7):559-564. Prevention of injuries among male soccer players: a prospective,
26. Schuermans J, Danneels L, Van Tiggelen D, Palmans T, Witvrouw randomized intervention study targeting players with previous inju-
E. Proximal neuromuscular control protects against hamstring inju- ries or reduced function. Am J Sports Med. 2008;36(6):1052-1060.
ries in male soccer players: a prospective study with electromyog- 34. Malone S, Owen A, Mendes B, Hughes B, Collins K, Gabbett TJ.
raphy time-series analysis during maximal sprinting. Am J Sports High-speed running and sprinting as an injury risk factor in soccer:
Med. 2017;45(6):1315-1325. can well-developed physical qualities reduce the risk? J Sci Med
27. Malone S, Roe M, Doran DA, Gabbett TJ, Collins K. High chronic Sport. 2018;21(3):257-262.
training loads and exposure to bouts of maximal velocity run-
ning reduce injury risk in elite Gaelic football. J Sci Med Sport.
2017;20(3):250-254. SUPPORTING INFORMATION
28. Mendiguchia J, Conceição F, Edouard P, et al. Sprint versus isolated Additional supporting information may be found online in
eccentric training: comparative effects on hamstring architecture and the Supporting Information section.
performance in soccer players. PLoS One. 2020;15(2):e0228283.
29. Edouard P, Mendiguchia J, Lahti J, et al. Sprint acceleration me-
chanics in fatigue conditions: compensatory role of gluteal muscles
in horizontal force production and potential protection of hamstring How to cite this article: Suarez-Arrones L,
muscles. Front Physiol. 2018;9:1706. Nakamura FY, Maldonado RA, Torreno N, Di Salvo
30. Hagglund M, Walden M, Ekstrand J. Risk factors for lower extrem- V, Mendez-Villanueva A. Applying a holistic
ity muscle injury in professional soccer: the UEFA Injury Study.
hamstring injury prevention approach in elite football:
Am J Sports Med. 2013;41(2):327-335.
31. Bahr R, Thorborg K, Ekstrand J. Evidence-based hamstring injury
12 seasons, single club study. Scand J Med Sci Sport.
prevention is not adopted by the majority of Champions League or 2021;31:861–874. https://doi.org/10.1111/sms.13913
Norwegian Premier League football teams: the Nordic Hamstring
survey. Br J Sports Med. 2015;49(22):1466-1471.