Original research

Efficacy of a new injury prevention programme

(FUNBALL) in young male football (soccer) players: a
cluster-­randomised controlled trial
Rilind Obërtinca ‍ ‍,1,2 Rina Meha,1 Ilir Hoxha,3,4 Bujar Shabani,5,6 Tim Meyer ‍ ‍,1
Karen aus der Fünten1

► Additional supplemental ABSTRACT

material is published online Objectives To evaluate the efficacy of a new WHAT IS ALREADY KNOWN ON THIS TOPIC
only. To view, please visit the ⇒ Youth football (soccer) is associated with a
journal online ([Link] multicomponent, exercise-­based injury prevention
org/​10.​1136/​bjsports-​2023-​ programme in football players 13–19 years old. significant injury risk.
107388). Methods Two-­arm cluster-­randomised controlled trial ⇒ Various multicomponent exercise-­based injury
with clubs as the unit of randomisation. 55 football prevention programmes may reduce the risk
1
Institute of Sports and of football-­related injuries, but evidence is
teams from Kosovo of the under 15, under 17 and
Preventive Medicine, Saarland
University, Saarbrucken, under 19 age groups were randomly assigned to the conflicting. Implementation of and adherence
Germany intervention (INT; 28 teams) or the control group (CON; to these programmes can be challenging.
2
Department of Physiotherapy, 27 teams) and were followed for one football season WHAT THIS STUDY ADDS
University ’Fehmi Agani’ in (August 2021–May 2022). The INT group performed the
Gjakova, Gjakova, Kosovo ⇒ The ‘FUNBALL’ programme is an effective
3
The Dartmouth Institute for ’FUNBALL’ programme after their usual warm-­up at least
twice per week, while the CON group followed their intervention used after the usual warm-­up
Health Policy and Clinical
Practice, Geisel School of usual training routine. The primary outcome measure was which lowers the injury incidence in male young
Medicine at Dartmouth, the overall number of football-­related injuries. Secondary football players.
Lebanon, NH, USA
outcomes were region-­specific injuries of the lower limbs ⇒ The overall injury incidence was lowered by
4
Evidence Synthesis Group, one-­third when the ‘FUNBALL’ programme was
Prishtina, Kosovo (hip/groin, thigh, knee, lower leg, ankle and foot) and
5
Faculty of Medicine, University injury severity. applied for one season.
of Prishtina, Prishtina, Kosovo Results 319 injuries occurred, 132 in the INT and 187 ⇒ Preventive benefits were also found for thigh
6
Université Claude Bernard, in the CON group. The INT group used the ’FUNBALL’ injuries, and for moderate and severe time-­loss
Lyon, France injuries.
programme in 72.2% of all training sessions, on average
2.2 times per week. There was a significantly lower ⇒ The positive effect on injury burden led to
Correspondence to better player availability.
Rilind Obërtinca, Institute incidence in the INT group regarding the overall number
of Sports and Preventive of injuries (incidence rate ratio (IRR) 0.69, 95% CI 0.55 HOW THIS STUDY MIGHT AFFECT RESEARCH,
Medicine, Saarland University, to 0.87), the number of thigh injuries (IRR 0.62, 95% CI PRACTICE OR POLICY
Saarbrucken, 66123, Germany; 0.39 to 0.98), of moderate (time loss between 7 and
​riob00001@​stud.​uni-​saarland.​ ⇒ Male adolescent football players should
de 28 days) (IRR 0.65, 95% CI 0.44 to 0.97) and of severe
be encouraged to perform the ‘FUNBALL’
injuries (time loss >28 days) (IRR 0.51, 95% CI 0.28 to
programme at least twice per week to induce
Accepted 11 March 2024 0.91).
Published Online First maximal benefits.
Conclusion The ’FUNBALL’ programme reduced
18 March 2024 ⇒ More research is needed on the efficacy of the
the incidence of football-­related injuries among male
‘FUNBALL’ programme in other age groups
adolescent football players, and its regular use for injury
(senior and veteran players) as well as in
prevention in this population is recommended.
female football players.
Trial registration number NCT05137015.
⇒ The ‘FUNBALL' programme is more football
specific compared with existing injury
prevention programmes. Future studies
INTRODUCTION should explore whether this aspect improves
Youth football (soccer) is associated with a signif- compliance and adherence compared with
icant injury risk. The overall injury incidence in previous programmes.
youth male football players has been reported
between 2.4 and 12.0 injuries per 1000 football
hours.1 2 The majority of injuries concern the lower Others aimed to reduce the overall number of
extremity,1–4 especially the thigh region.1 3–5 Severe lower limb injuries.15–18 In the above-­ mentioned
© Author(s) (or their injuries accounted for 21–37% of all injuries,1 3 cluster-­randomised controlled trials (cluster-­RCTs),
employer(s)) 2024. Re-­use
permitted under CC BY-­NC. No or 0.78 injuries per 1000 hours.6 This aligns with the highest efficacy reported was a 77% reduction
commercial re-­use. See rights injury locations and injury severity reported in in injury rates.14 Several meta-­analyses supported
and permissions. Published adult professional football players.6–8 the efficacy of IPPs.19–21 A more cautious interpre-
by BMJ. With the aim to reduce the number of football-­ tation of their efficacy emerged recently when other
To cite: Obërtinca R, related injuries, many exercise-­based injury preven- meta-­analyses included the calculation of prediction
Meha R, Hoxha I, tion programmes (IPPs) have been established. Some intervals.22 23 Despite available evidence of their
et al. Br J Sports Med of them targeted specific injuries, for example, efficacy,9–14 16–18 24–26 and the importance of good
2024;58:548–555. adductor,9 hamstring10 11 and knee injuries.12–14 compliance for injury reduction,27–29 many studies

548    Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388

 Original research

Figure 1 Consolidated Standards of Reporting Trials flow diagram of teams and players through trial.

highlighted a low programme compliance.15 30 31 Efforts have randomisation stratified by league level (Super League or
been made to optimise strategies for increasing compliance and Regional League) was performed. The stratification was chosen
adherence.32 33 Nonetheless, achieving broad-­scale effectiveness to account for possible differences in competition level. The
of IPPs remains challenging.34–37 The main perceived barriers to randomisation was performed by one researcher (RM), who was
low compliance and adherence include time constraints, physical blinded to the identities of the clubs and who was not involved
complaints (eg, fatigue and soreness) caused by exercises, lack in the intervention.
of awareness and knowledge about the programmes’ execution,
and low motivation due to the absence of football-­specific activ- Intervention
ities within the IPPs.29 32 The intervention consisted of six fundamental exercise catego-
We developed a multicomponent exercise-­based IPP specifi- ries with the intention of preventing football-­related injuries,
cally targeting youth football players. The intention was to use hence the abbreviation ‘FUNBALL’. In addition, the programme
as many football-­ specific elements as possible, based on the contained one optional game. The following mandatory exer-
assumption that they increase motivation and compliance. Exer- cise categories were included: (1) balance, (2) core stability, (3)
cise categories were based on scientific evidence that has previ- hamstring muscles eccentrics, (4) gluteal muscle activation, (5)
ously shown good efficacy in injury prevention in football. By plyometrics and (6) running/sprinting. The optional category (7)
means of a cluster-­RCT, we aimed to evaluate the efficacy of ‘games’ (three games included) reflected the intention to increase
the ‘FUNBALL’ programme to reduce injuries in male football the attractiveness of the programme (table 1). Each mandatory
players 13–19 years old. category contained two different exercises to offer more vari-
ability. The coach was free to decide which of the two to choose
METHODS for each training session. All exercises were organised in five
Study design and participants or six progressive levels with increasing physical and cognitive
The design of the study was a two-­arm, cluster-­RCT. It was chosen difficulty, and were required to be performed in order (from 1
to reduce contamination bias within clubs. The study is reported to 5/6). The exercises started on the first level and moved to the
according to the Consolidated Standards of Reporting Trials next one when exercises were executed with a proper technique
statement for cluster-­ randomised trials.38 The study protocol as assessed by the coach. The programme took about 15–20 min
was registered within ​ClinicalTrials.​gov (NCT05137015). to complete after familiarisation.
At the beginning of 2020, 21 football clubs (with 70 teams in Based on the latest evidence regarding the challenge of long-­
total) from different regions in Kosovo that fulfilled all inclusion term adherence,34 35 and in accordance with what the implemen-
criteria were invited to participate in our study, with their under tation science has proposed in relation to IPP development,33 35
15s, under 17s and under 19s male teams (figure 1). All teams it was decided among the coauthors who were involved in the
participated either in the Super League and/or Regional Leagues, development of the programme (RO, RM, TM and KadF) to
organised by the Football Federation of Kosovo. To be included, include a football coach within the team for the development
teams had to: (1) be officially registered in the above-­named and refinement of the intervention. This with the intention to
football association, (2) train at least twice per week and (3) secure the end users’ perspective throughout the whole process.
participate in regular matches of the above-­named leagues. We The coach was not part of any team later included in the study,
excluded clubs that were already using a structured IPP. All the nor in the piloting or intervention period. In addition, a psychol-
clubs that enrolled for the study were randomised either into the ogist provided input for the neurocognitive demands of the
intervention or the control group. All teams from one club were programme. Prior to its implementation, the programme was
randomised into the same treatment arm. Computer-­generated piloted on two football teams. One exercise was replaced with

Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388 549

 Original research

Table 1 Multicomponent exercise-­based programme ‘FUNBALL’ used

to prevent injuries in young football players
Number of
Exercises Repetitions/duration levels
Balance
 a. Single-­leg stance 2 sets×30 s (on each leg) 6
 b. Y-­balance 3 sets×6–8 repetitions (on 6
each leg)
Figure 2 Example of correct (left) and incorrect (right) posture
Core stability
alignment for one of the exercises provided in the programme (core
 a. Plank and side plank 2 sets×20–40 s (on each 6
stability; exercise a).
position)
 b. Straight arm plank 2 sets×8–12 repetitions 6
Hamstring muscles eccentrics the quality of programme execution. If coaches needed clarifi-
 a. Nordic hamstring 1–2 sets×3–10 repetitions 5 cation regarding the exercises, they were advised to contact the
 b. Hamstring walk-­outs 2–3 sets×30 s 5 research staff, who were continuously available throughout the
Gluteal muscle activation
study period. The coaches of the control group were instructed
to perform their training as usual. Prior to the start of the inter-
 a. Head, shoulder, hip, knee, 2 sets×6–10 repetitions 6
ankle vention, we gathered more detailed information regarding the
 b. Squat lunges 2–3 sets×8–12 repetitions 6
training ‘routine’ of control teams, by interviewing 11 of the
Plyometric
22 coaches. The aim was to collect information whether they
performed specific exercises similar to the categories used in the
 a. Forward jumps 4 sets×3 jumps 5
programme. The control group received the programme after
 b. Skater jumps 4 repetitions (2 on each leg) 5
the end of the study.
Running/sprinting
 a. Diagonal running/sprinting 3 repetitions 6
Outcome measures
 b. Forward running/sprinting 3 repetitions 6
The primary outcome measure was the overall number of
Games
football-­
related injuries that occurred during the season.
 a. Tic-­tac-­toe 3–5 games n.a. Secondary outcomes were region-­specific injuries of the lower
 b. Header game 4–5 repetitions for each player n.a. limbs (hip/groin, thigh, knee, lower leg, ankle and foot) and
 c. Dribbling game 3 games n.a. injury severity (minimal, mild, moderate and severe injuries).
n.a., not applicable.
Data collection procedures and definitions
The data collection procedures and definitions used in our study
another after the suggestions from the coaches as it was reported were in line with the consensus statement on injury definitions
as too time-­consuming. The pilot teams were not invited to and data collection procedures.39 This entailed injury definition,
participate in the study. To further address the compliance issue, injury severity, mechanism of injury, injury type and location,
we tried to make the programme as football specific as possible. and definitions for training and match exposure (online supple-
We introduced exercises requiring competition between the mental table 1). We collected data during an entire competitive
players, offered two variations for each exercise category and season from August 2021 to May 2022. During the pre-­season,
cognitive challenges in the majority of exercises. Furthermore, the research staff and research assistants collected players’ base-
the ball was included as often as possible. Previous IPPs replaced line characteristics. The baseline questionnaire included name,
the warm-­up.15 16 18 However, coaches may take this as a restric- age, weight, height, playing position, history of injuries and
tion, which may affect the long-­term compliance. Therefore, we current health conditions. Throughout the competitive season,
designed the ‘FUNBALL’ programme to be used after the usual the coaches or team’s physiotherapists reported to the research
warm-­up. In order to maintain the benefits of warm-­up, most of assistants team exposure hours, programme execution (compli-
the ‘FUNBALL’ exercises were of relatively high intensity, espe- ance) and the new injuries that occurred on a weekly basis.
cially the last three (plyometrics, running/sprinting and games). If reporting was delayed for more than 1 week, an automatic
During the pre-­ season, the programme was introduced to message was sent to them. The original plan was to record
the coaches of the intervention teams according to previous the injuries and individual exposure hours and report them
research.15 16 Within the club facilities, the research staff (led by weekly to the research team via mail. However, most coaches
first author, RO) provided instructional courses. They included reported that it was too time-­consuming. That led to a shift in
theoretical and practical training. Coaches received a detailed data reporting practices. The data exchange was subsequently
manual of the programme (online supplemental file 2) and an carried out via telephone and we collected team exposure hours
‘on pitch’ card (online supplemental file 3). They were advised instead of individual ones. When new injuries were reported,
to use the programme at least twice a week. During the coaches’ two research assistants (physiotherapists) blinded to group allo-
instructional courses, there was a focus on the key aspects of the cation contacted the injured players (or their parents if players
programme, correct postures and movement patterns. Coaches were underage) to obtain the detailed information regarding the
were explicitly instructed to pay attention to those aspects while injury and its diagnosis, by use of a standardised injury regis-
performing ‘FUNBALL’. The correct posture was illustrated and tration form.18 To increase the accuracy of the data collection,
described in detail in the manual of the programme (figure 2). thorough clarification of the protocols for injury classification
The intervention started 1 week before the clubs’ first official and injury definitions was carried out for the research assistants
match. Research staff visited the intervention teams several before the season started. The exact diagnosis was required
times, that is, three to four visits per team in season, to monitor in case the player required medical treatment. Most of severe

550 Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388

 Original research
injuries (92%) were diagnosed by a physician, partially by one of
Table 2 Player and injury characteristics of the intervention and
the coauthors, BS, not connected to any of the clubs assigned for
control groups
the study and blinded to the group allocation, or other doctors
not included in the study. Additionally, the research staff visited Variable Intervention group Control group
all participating teams at the end of the season to add missing or Player characteristics
to clarify unclear information by use of individual discussions  No of teams 23 22
with involved players. Data on players who dropped out or  No of players 524 503
changed the teams during the season were included until then.  Mean (SD) age (years) 15.2 (1.6) 15.3 (1.6)
Eight research assistants (two physiotherapists, five students  Mean (SD) height (cm) 171 (9.1) 172 (7.9)
of the last year of physiotherapy school, and one strength and  Mean (SD) weight (kg) 60.2 (8.6) 60.5 (8.3)
conditioning coach) blinded to group allocation registered the  Mean (SD) BMI (kg/m2) 20.4 (1.5) 20.3 (1.7)
players’ basic information and injuries on prepared Excel data-  Mean (SD) football experience* (years) 5.0 (1.8) 4.9 (1.6)
sets. We registered all injuries reported from the start of the
Exposure characteristics
intervention (1 week ahead of the season, 23 August 2021) until
 Total exposure (hours) 53 454 52 938
the last match of the season (22 May 2022). If players were
 Match exposure (hours) 9 017 8 666
already injured at the start of the study, they were included in
 Training exposure (hours) 44 437 44 272
the study; however, that injury was excluded.
Injury characteristics
 No of total injuries 132 187
Sample size  No of match injuries 65 91
A pre-­trial sample size calculation based on the data on the inci-  No of training injuries 67 96
dence of injuries in male adolescent footballers was performed.2 3  No of injured players 124 172
For the primary outcome (overall injuries), we estimated that
 Injury burden (SD) (days) 40 (3.4) 74 (5.4)
78.5% of the players in the control group will sustain an injury
*Football experience taking into account the years since the player has trained at
during the season.2 Sample size calculation (comparison of two least three times per week.
proportions) revealed that a total of 366 (183 per arm) players †Number of injury days lost per 1000 hours.
are required to achieve 80% power in detecting an estimated BMI, body mass index.
30% reduction in injury rate in the intervention group with an
alpha level of 0.05. This is based on the assumption that the
team comprises 22 players on average and taking into account an Equity, diversity and inclusion statement
estimated design effect of 2.95. For the second outcome (region-­ The study included a variety of race/ethnicities and socioeco-
specific injuries), 620 players are required based on the assump- nomic levels. The research team consists of two women and four
tion that 64% of players would report a thigh, knee or ankle men from different disciplines (physiotherapy, sports psychology,
injury during one season3 and a similar reduction in injury rate medicine, sports medicine and orthopaedics). It included two
and design effect as above. Based on an expected dropout rate junior researchers (RO and RM). As our study was conducted
of 30%, we aimed to recruit 806 (403 per arm) football players on male football players only, we cannot extrapolate findings to
(approximately 37 teams). We used G*Power software with two-­ female players. We expand on the exclusion of female players in
sided Z-­test to generate the required sample size. the discussion.

Statistical methods RESULTS

All statistical analyses were conducted using Stata statistical soft- Participants
ware V.17 BE (Stata Corp, Texas, USA). Descriptive statistics were The final sample consisted of 45 football teams (1027 players),
reported for baseline characteristics. Continuous variables (age, with 23 teams (524 players) in the intervention group and 22
height, weight, body mass index and football experience) were teams (503 players) in the control group (figure 1). In both clus-
reported as mean and SD and were checked for normal distribu- ters, the dropout rate was similar (17.9% in the intervention
tion. Normal distribution was determined using a histogram, QQ group and 18.2% in the control group). The players in the two
plot and Shapiro-­Wilk test. Incidence rate ratios (IRRs) with 95% groups who completed the study were similar in terms of base-
CIs were calculated according to the intention-­to-­treat principle line characteristics (table 2).
for each outcome and compared between the intervention and
control groups. We used a Poisson regression model with adjust- Exposure and injury characteristics
ment for cluster effect. Team was considered as cluster variable. During the season, 106 392 hours of football were recorded.
Two-­tailed p values were considered significant when the alpha The players in the intervention group were involved in 53 454
error had a level of less than 0.05. Training exposure was calcu- hours (44 437 training and 9017 match hours), the players in
lated by multiplying the number of training sessions, training the control group in 52 938 hours (44 272 training and 8666
time and mean training attendance rate.14 Match exposure was match hours) (table 2). 319 injuries occurred: 132 in the inter-
calculated by multiplying the number of matches, match dura- vention and 187 in the control group. The overall injury IR per
tion and the number of players on the field.14 The total football 1000 football hours for both groups was 2.99 (95% CI 2.68 to
comprised the sum of training and match exposure hours.14 39 3.34); the training injury IR was 1.83 (95% CI 1.57 to 2.14)
The injury incidence rate (IR) is presented with 95% CI and was and the match injury IR was 8.82 (95% CI 7.54 to 10.32). 296
calculated according to the formula IR=(n/e)×1000, where (n) (28.8%) of the 1027 players suffered an injury. The thigh was
is the number of soccer injuries and (e) the total exposure time the most injured region (n=80; 25.1%; IR 0.75), followed by
expressed as total hours of football exposure.16 Injury burden knee (n=62; 19.4%; IR 0.58) and ankle (n=57; 17.9%; IR
was calculated as the number of days lost to injury per 1000 0.53). Players of the age group of the under 19s sustained the
hours of football (‘injury incidence×mean absence per injury’).40 highest number of injuries (n=122; 38.2%; IR 4.49) versus the

Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388 551

 Original research

Table 3 Effectiveness of the ‘FUNBALL’ programme in adolescent male football players according to intention to treat
Intervention group Control group
Variable No of injuries (%) IR (95% CI) No of injuries (%) IR (95% CI) IRR (95% CI) P value
Total injuries 132 (100) 2.46 (2.08 to 2.92) 187 (100) 3.53 (3.06 to 4.07) 0.69 (0.55 to 0.87) 0.002
 Under 15s injuries 29 (22) 1.43 (0.99 to 2.06) 49 (26.2) 2.77 (2.09 to 3.67) 0.51 (0.32 to 0.82) 0.005
 Under 17s injuries 49 (37.1) 2.49 (1.88 to 3.30) 70 (37.4) 3.21 (2.54 to 4.05) 0.77 (0.53 to 1.11) 0.175
 Under 19s injuries 54 (40.9) 3.95 (3.03 to 5.16) 68 (36.4) 5.04 (3.97 to 6.39) 0.78 (0.54 to 1.12) 0.184
Location
 Thigh 31 (23.5) 0.57 (0.40 to 0.82) 49 (26.2) 0.92 (0.69 to 1.22) 0.62 (0.39 to 0.98) 0.042
 Knee 26 (19.7) 0.48 (0.33 to 0.71) 36 (19.3) 0.68 (0.49 to 0.94) 0.71 (0.43 to 1.18) 0.193
 Ankle 23 (17.4) 0.43 (0.28 to 0.64) 34 (18.2) 0.64 (0.45 to 0.84) 0.66 (0.39 to 1.13) 0.138
 Hip/groin 15 (11.4) 0.28 (0.16 to 0.46) 21 (11.2) 0.39 (0.25 to 0.60) 0.70 (0.36 to 1.37) 0.306
 Lower leg/Achilles tendon 6 (4.6) 0.11 (0.05 to 0.24) 10 (5.4) 0.18 (0.10 to 0.35) 0.59 (0.21 to 1.63) 0.313
 Foot/toe 7 (5.3) 0.13 (0.06 to 0.27) 9 (4.8) 0.17 (0.08 to 0.32) 0.77 (0.28 to 2.06) 0.605
 Forearm 5 (3.8) 0.09 (0.03 to 0.22) 6 (3.2) 0.11 (0.05 to 0.25) 0.82 (0.25 to 2.70) 0.751
 Hand/finger/thumb 5 (3.8) 0.09 (0.03 to 0.22) 6 (3.2) 0.11 (0.05 to 0.25) 0.82 (0.25 to 2.70) 0.751
 Head/face/neck 5 (3.8) 0.09 (0.03 to 0.22) 5 (2.7) 0.09 (0.03 to 0.22) 0.99 (0.28 to 3.42) 0.988
 Lower back/sacrum/pelvis 4 (3) 0.07 (0.02 to 0.19) 4 (2.1) 0.07 (0.02 to 0.20) 0.99 (0.24 to 3.95) 0.989
 Shoulder/clavicle 2 (1.5) 0.03 (0.00 to 0.14) 4 (2.1) 0.07 (0.02 to 0.20) 0.49 (0.90 to 2.70) 0.417
 Elbow 1 (0.8) 0.01 (0.00 to 0.13) 1 (0.5) 0.01 (0.00 to 0.13) 0.99 (0.06 to 15.83) 0.995
 Wrist 1 (0.8) 0.01 (0.00 to 0.13) 1 (0.5) 0.01 (0.00 to 0.13) 0.99 (0.06 to 15.83) 0.995
 Abdomen 1 (0.8) 0.01 (0.00 to 0.13) 1 (0.5) 0.01 (0.00 to 0.13) 0.99 (0.06 to 15.83) 0.995
Injury mechanism
 Trauma 114 (86.4) 2.13 (1.77 to 2.56) 165 (88.2) 3.11 (2.67 to 3.63) 0.68 (0.53 to 0.86) 0.002
 Overuse 18 (13.6) 0.33 (0.21 to 0.53) 22 (11.8) 0.41 (0.27 to 0.63) 0.81 (0.43 to 1.51) 0.508
Injury occurrence
 Training 67 (50.8) 1.50 (1.18 to 1.91) 96 (51.3) 2.16 (1.17 to 2.64) 0.69 (0.50 to 0.94) 0.022
 Match 65 (49.2) 7.20 (5.65 to 9.19) 91 (48.7) 10.50 (8.55 to 12.89) 0.68 (0.49 to 0.94) 0.021
Injury severity
 Minimal (1–3 days) 18 (13.6) 0.33 (0.21 to 0.53) 22 (11.8) 0.41 (0.27 to 0.63) 0.81 (0.43 to 1.51) 0.508
 Mild (4–7 days) 56 (42.4) 1.04 (0.80 to 1.36) 70 (37.4) 1.32 (1.04 to 1.67) 0.79 (0.55 to 1.12) 0.194
 Moderate (8–28 days) 41 (31.1) 0.76 (0.56 to 1.04) 62 (33.2) 1.17 (0.91 to 1.50) 0.65 (0.44 to 0.97) 0.035
 Severe (>28 days) 17 (12.9) 0.31 (0.19 to 0.51) 33 (17.6) 0.62 (0.44 to 0.87) 0.51 (0.28 to 0.91) 0.024
IRs are reported per 1000 hours of football play and are unadjusted.
IRRs are adjusted for team.
IR, incidence rate; IRR, incidence rate ratio.

under 17s (n=119; 37.3%; IR 2.87) and the under 15s (n=78; Efficacy of the intervention programme
24.5%; IR 2.06) (table 3). Further injury characteristics data are For the primary outcome investigated, there was a significantly
presented in tables 2 and 3. lower incidence in the intervention group for the overall number
of injuries (IRR 0.69, 95% CI 0.55 to 0.87, p=0.002). Secondary
outcomes that reached significantly lower incidences in the
Compliance with the ‘FUNBALL’ programme and training
‘routine’ of the control teams intervention group were thigh injuries (IRR 0.62, 95% CI 0.39
The intervention group used the ‘FUNBALL’ programme in to 0.98, p=0.042), moderate injuries (IRR 0.65, 95% CI 0.44 to
72.2% of all training sessions, on average 2.2 times per week 0.97, p=0.035) and severe injuries (IRR 0.51, 95% CI 0.28 to
(online supplemental table 2). The average player attendance for 0.91, p=0.024). Moreover, a significantly lower incidence was
training sessions was 17.2 in the intervention group and 17.5 found for match (IRR 0.68, 95% CI 0.49 to 0.94, p=0.021),
in the control group. All the interviewed coaches (n=11; 50%) training (IRR 0.69, 95% CI 0.50 to 0.94, p=0.022) and trau-
of the control teams reported that they used exercises of similar matic injuries (IRR 0.68, 95% CI 0.53 to 0.86, p=0.002). The
categories that are contained in the ‘FUNBALL’ programme. subgroup analysis according to age groups showed a significantly
The coaches of the under 15s (n=4; 18.2%) reported they lower incidence for the overall number of injuries among the
perform balance, core stability and running/sprinting exercises
under 15 players (IRR 0.51, 95% CI 0.32 to 0.82, p=0.005).
in their training. The coaches of the under 17s and under 19s
The incidence of knee and ankle injuries did not reach signif-
teams (n=7; 31.8%) reported that they employ core stability,
hamstring eccentric, plyometric and running/sprinting exercises, icance (table 3). The injury burden was 40 days lost per 1000
but very rarely balance exercises. The majority of them applied hours in the intervention group and 74 days lost per 1000 hours
these exercises at least once a week. However, their use was not in the control group (table 2). No harmful events associated with
structured with regard to the number of repetitions, duration the use of the programme, for example, injuries during their
and types of exercises. execution, were reported by the coaches.

552 Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388

 Original research
DISCUSSION statistical power for these particular injury types) may explain
Principal findings why ‘FUNBALL’ did not show a significant preventive effect in
The main finding of this study among male young football players several secondary outcomes, especially in reducing knee injuries.
is a lower overall injury incidence by one-­third in the group that The efficacy of ‘FUNBALL’ differed between age groups.
used the ‘FUNBALL’ programme. Also, training and match inju- The highest efficacy was found among the under 15 players in
ries were lower in the intervention group when considered sepa- comparison with under 17 and under 19 players (IRR 0.51 vs
rately. Further relevant findings were the programme’s efficacy 0.77 and 0.78). The reason for this might be the previously
in reducing the incidences of one of the most frequently affected mentioned fact by the interviewed coaches that they use similar
injury regions (thigh), injuries causing the longest time loss in categories of our programme in their training routine, espe-
football (moderate and severe injuries) and the injury burden. cially in the older age groups. Therefore, the significant lower
Thus, players’ availability was higher in the teams of the inter- injury incidence due to the use of ‘FUNBALL’ might be mainly
vention group. attributed to the large effect in the youngest age group. There
were no indications that differing compliance with the conduc-
tion of the programme was a relevant confounder.
Efficacy of the programme and comparison with previous
research
The ‘FUNBALL’ intervention proved to be successful in a Strengths and limitations
number of aspects. The inclusion of evidence-­ based exer- Our study has several strengths. First, the IPP was investigated
cise categories for prevention of football-­related injuries may through a large cluster-­randomised trial. We followed good prac-
be one of the main reasons. The first two categories included tice by cluster-­randomising the clubs to avoid contamination
balance and core stability exercises. Previous studies reported between the control and intervention groups and by blinding the
on the efficacy of balance training in reducing ankle ligament injury data collection assistants. In-­season, we regularly visited
injuries in football,41 42 and the association between impaired the clubs without previous announcement to monitor the imple-
core stability and the development of lower extremity injuries mentation of the programme. Moreover, we were in contact
in healthy athletes.43 Hamstring eccentrics were also included with players and their parents with regard to detailed injury
in our programme. Their efficacy in preventing hamstring inju- information in addition to the data provided by coaches or the
ries is well-­known.10 11 Even though there is limited evidence teams’ physiotherapists. Finally, we collected detailed informa-
regarding the role of gluteal activation for injury prevention, tion from the coaches of the control group regarding the exer-
there is evidence that reduced activity represents a risk factor cises that they usually perform during the season with a focus on
for hamstring injuries.44 Moreover, the crucial role of gluteal exercises similar to those used in our intervention programme.
muscles in maintaining a correct knee position, that is, avoiding a This provided a possibility of a more accurate assessment of the
dynamic knee valgus, during activities such as walking, running, efficacy found in our study since an unintentional use of similar
jumping and landing has been reported.45 Incorporating plyo- exercises would have lowered the effect of the investigated
metric exercises in IPPs has been shown to effectively decrease programme.
the risk of anterior cruciate ligament injuries.46 Finally, and for This study also has some limitations. Despite the inclusion of
the first time in connection with IPPs, we introduced sprinting a football coach, we lacked the input of footballers themselves in
exercises to mitigate hamstring injury risk.47 Combining many the process of developing the intervention. We knew in advance
exercise categories makes it (more) difficult to understand which that most of the participating clubs lacked female teams. Thus,
categories provide the highest benefit for reducing injury risk. it was a conscious decision to confine the study to male teams
A comparison with existing studies is difficult as only very few only. This impacts the strength of clinical recommendations for
of them considered our specific age group and male players. The the programme implementation. We relied on an older version
preventive effect on the overall injury incidence is in accordance of the data collection methodology39 as the planning of the study
with two large RCTs investigating the efficacy of ‘FIFA11+’ in took place before a more sophisticated version50 was available.
youth female and male football players, respectively.16 25 Similar The older version lacks some details, especially with regard to
to the ‘FIFA11+’ study conducted in females,16 ‘FUNBALL’ ‘overuse/growth-­ related injuries’. Collecting team exposure
reached a significantly positive effect on overall and severe hours instead of individual exposure hours as it was originally
injuries, furthermore on thigh injuries. This may be expected planned is a further limitation, since playing and training time
as this type of injury occurs more often in male footballers.6 alike can vary greatly among players.16
Owoeye et al25 investigated youth male football players. They After the start of the study, some barriers appeared in both
reported an even higher efficacy if the ‘FIFA11’ programme was groups. Four coaches of the intervention teams decided to stop
employed. The efficacy rate was higher for overall and match the programme implementation. For them, the small number of
injuries compared with our findings. Their figures were 41% and coaching staff within the team as well as the limited time for
65%, respectively, as compared with 31% and 32% in our study. training was the main reason for terminating the programme.
In contrast to the ‘FUNBALL’ study, neither of the two above-­ In both groups, several coaches presented low motivation for
mentioned ‘FIFA11+’ studies reached significant effects with providing the exposure hours and injuries that occurred. Some
regard to training injuries.16 25 Additionally, ‘FUNBALL’ lowered coaches did not report the data on a weekly basis. We excluded
the injury burden and the number of injuries lasting >8 days by teams from the study if they did not provide the data for a period
about 50%. This can be a highly important point, knowing that of 4 weeks. Moreover, the decision of when to progress to the
a team with lower injury burden and less severe injuries has a next exercise level was left to the coaches without any guidance
better chance of improved team performance.40 Injury patterns from the study assistants. In some cases, we recognised a big
and frequencies differ among different age groups and sexes. difference. Some clubs moved rapidly, within the first weeks of
Forearm fractures are quite common in children, whereas ante- the study, to the most advanced levels, while other clubs still
rior cruciate ligament ruptures are more common in females used the initial levels. Finally, the additional time that is required
aged 16 years and above.48 49 This (together with lacking to perform the programme (15–20 min) may be considered as a

Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388 553

 Original research
downside, which however should be weighed against less injured solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all
players. The vast majority of the limiting factors listed above liability and responsibility arising from any reliance placed on the content.
Where the content includes any translated material, BMJ does not warrant the
potentially impact the programme’s success. accuracy and reliability of the translations (including but not limited to local
regulations, clinical guidelines, terminology, drug names and drug dosages), and
Clinical implications, applicability and future research is not responsible for any error and/or omissions arising from translation and
Reducing football-­related injuries holds many benefits both indi- adaptation or otherwise.
vidually for the players as well as for the team. A lower number Open access This is an open access article distributed in accordance with the
of injuries, apart from the health benefits, will contribute to the Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which
permits others to distribute, remix, adapt, build upon this work non-­commercially,
performance of the teams and the financial-­related aspects, but and license their derivative works on different terms, provided the original work is
it will also increase the likelihood that the young footballers will properly cited, appropriate credit is given, any changes made indicated, and the use
reach their highest potential. Early adaptation to preventative is non-­commercial. See: [Link]
exercise might, thus, be highly valuable especially at younger
ORCID iDs
ages, as they may serve as a blueprint for an application later in Rilind Obërtinca [Link]
the career. The ‘FUNBALL’ was investigated among male adoles- Tim Meyer [Link]
cent football players (aged 13–19 years). Its efficacy in other age
groups (seniors and veterans) or female football players was not REFERENCES
investigated in our study. This calls for future studies to eval- 1 Light N, Johnson A, Williams S, et al. Injuries in youth football and the relationship to
uate the efficacy within these groups. Furthermore, it is recom- player maturation: an analysis of Time‐Loss injuries during four seasons in an English
mended investigating the efficacy of the ‘FUNBALL’ in an even elite male football academy. Scand J Med Sci Sports 2021;31:1324–34.
2 Wik EH, Lolli L, Chamari K, et al. Injury patterns differ with age in male youth football:
larger cohort and possibly over a longer period of time. This a four-­season prospective study of 1111 time-­loss injuries in an elite National
will enable making a comprehensive evaluation of its potential Academy. Br J Sports Med 2021;55:794–800.
in reducing severe injuries that are less frequent such as anterior 3 Renshaw A, Goodwin PC. Injury incidence in a Premier League youth soccer Academy
cruciate ligament ruptures. using the consensus statement: a prospective cohort study. BMJ Open Sport Exerc
Med 2016;2:e000132.
4 Tears C, Chesterton P, Wijnbergen M. The elite Player performance plan: the impact
CONCLUSIONS of a new national youth development strategy on injury characteristics in a Premier
The ‘FUNBALL’ programme was effective in lowering the overall League Football Academy. J Sports Sci 2018;36:2181–8.
injury incidence by 31% in male adolescent football players over 5 Raya-­González J, Suárez-­Arrones L, Navandar A, et al. Injury profile of elite male
young soccer players in a Spanish professional Soccer Club: a prospective study during
an entire season. This also referred to thigh injuries as one of
4 consecutive seasons. J Sport Rehabil 2020;29:801–7.
the most frequent football-­related injury types, and to moderate 6 Robles-­Palazón FJ, López-­Valenciano A, De Ste Croix M, et al. Epidemiology of injuries
and severe injuries, which cause longest absence from football. in male and female youth football players: a systematic review and meta-­analysis. J
Therefore, we recommend its implementation in male adoles- Sport Health Sci 2022;11:681–95.
cent football players. 7 aus der Fünten K, Tröß T, Hadji A, et al. Epidemiology of football injuries of the
German Bundesliga: a media-­based, prospective analysis over 7 consecutive seasons.
Sports Med - Open 2023;9.
Correction notice This article has been corrected since it published Online First. 8 Ekstrand J, Hägglund M, Waldén M. Injury incidence and injury patterns in
The title and affiliation number 3 have been updated. professional football: the UEFA injury study. Br J Sports Med 2011;45:553–8.
Acknowledgements We acknowledge all the players and their parents for 9 Harøy J, Clarsen B, Wiger EG, et al. The Adductor strengthening programme prevents
participating in the study. We are grateful to the study assistants (Shpetim Kadriu, groin problems among male football players: a cluster-­randomised controlled trial. Br J
Lekë Tërmkolli, Doruntinë Salihu, Jonida Sahiti, Blina Obërtinca, Shpëtim Tërnava, Sports Med 2019;53:150–7.
Faruk Dragusha and Beriana Makolli), the Football Federation of Kosovo and 10 van der Horst N, Smits D-­W, Petersen J, et al. The preventive effect of the Nordic
representatives of the participating football clubs (board members, coaches and hamstring exercise on hamstring injuries in amateur soccer players: a randomized
medical staff). We thank Professor Dr Sabine Schaefer and Elon Berisha for their controlled trial. Am J Sports Med 2015;43:1316–23.
contribution to the development of the intervention, and Professor Dr Stefan 11 Petersen J, Thorborg K, Nielsen MB, et al. Preventive effect of eccentric training on
Wagenpfeil for his support in the statistical analysis. acute hamstring injuries in men’s soccer: a cluster-­randomized controlled trial. Am J
Sports Med 2011;39:2296–303.
Contributors RO and KadF were responsible for the conception and design of
12 Gilchrist J, Mandelbaum BR, Melancon H, et al. A randomized controlled trial to
the study. RO, RM, TM and KadF were responsible for the development of the
prevent noncontact anterior cruciate ligament injury in female collegiate soccer
intervention programme. RO coordinated the study. RM was responsible for database
players. Am J Sports Med 2008;36:1476–83.
management and contributed to the data preparation and merged the data files. IH
13 Waldén M, Atroshi I, Magnusson H, et al. Prevention of acute knee injuries in
and RO conducted the data analyses. BS and KadF checked the plausibility of injury
adolescent female football players: cluster randomised controlled trial. BMJ
information. RO wrote the first draft of the paper with inputs from KadF, and all
2012;344:e3042.
authors contributed to the final manuscript. RO is the guarantor of the manuscript.
14 Kiani A, Hellquist E, Ahlqvist K, et al. Prevention of soccer-­related knee injuries in
Funding The authors have not declared a specific grant for this research from any teenaged girls. Arch Intern Med 2010;170:43–9.
funding agency in the public, commercial or not-­for-­profit sectors. 15 Steffen K, Myklebust G, Olsen OE, et al. Preventing injuries in female youth football -
a cluster-­randomized controlled trial. Scand J Med Sci Sports 2008;18:605–14.
Competing interests TM is the chairman of UEFA’s and the German FA’s (DFB)
16 Soligard T, Myklebust G, Steffen K, et al. Comprehensive warm-­up programme to
medical committee.
prevent injuries in young female footballers: cluster randomised controlled trial. BMJ
Patient and public involvement Patients and/or the public were involved in the 2008;337:a2469.
design, or conduct, or reporting, or dissemination plans of this research. Refer to the 17 Emery CA, Meeuwisse WH. The effectiveness of a neuromuscular prevention strategy
Methods section for further details. to reduce injuries in youth soccer: a cluster-­randomised controlled trial. Br J Sports
Patient consent for publication Not applicable. Med 2010;44:555–62.
18 Rössler R, Junge A, Bizzini M, et al. A multinational cluster randomised controlled
Ethics approval This study involves human participants and was approved by trial to assess the efficacy of ’11+ kids’: a warm-­up programme to prevent injuries in
the Kosovo Chamber of Physiotherapists (identifier 2020/368). Participants gave children’s football. Sports Med 2018;48:1493–504.
informed consent to participate in the study before taking part. 19 Crossley KM, Patterson BE, Culvenor AG, et al. Making football safer for women:
Provenance and peer review Not commissioned; externally peer reviewed. a systematic review and meta-­analysis of injury prevention programmes in 11 773
female football (soccer) players. Br J Sports Med 2020;54:1089–98.
Data availability statement Data are available upon reasonable request. 20 Lemes IR, Pinto RZ, Lage VN, et al. Do exercise-­based prevention programmes reduce
Supplemental material This content has been supplied by the author(s). non-­contact musculoskeletal injuries in football (soccer)? A systematic review and
It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not meta-­analysis with 13 355 athletes and more than 1 million exposure hours. Br J
have been peer-­reviewed. Any opinions or recommendations discussed are Sports Med 2021;55:1170–8.

554 Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388

 Original research
21 van Dyk N, Behan FP, Whiteley R. Including the Nordic hamstring exercise in injury 36 Lindblom H, Sonesson S, Torvaldsson K, et al. Extended knee control programme
prevention programmes halves the rate of hamstring injuries: a systematic review and LOWERS weekly hamstring, knee and ankle injury prevalence compared with
meta-­analysis of 8459 athletes. Br J Sports Med 2019;53:1362–70. an Adductor strength programme or self-­selected injury prevention exercises in
22 Obërtinca R, Hoxha I, Meha R, et al. Efficacy of multi-­component exercise-­based injury adolescent and adult amateur football players: a two-­armed cluster-­randomised trial
prevention programs on injury risk among footballers of all age groups. A systematic with an additional comparison arm. Br J Sports Med 2023;57:83–90.
review and meta-­analysis. Sports Med 2023;53:837–48. 37 Bizzini M, Junge A, Dvorak J. Implementation of the FIFA 11+ football warm up
23 Impellizzeri FM, McCall A, van Smeden M. Why methods matter in a meta-­analysis: a program: how to approach and convince the football associations to invest in
reappraisal showed inconclusive injury preventive effect of Nordic hamstring exercise. prevention. Br J Sports Med 2013;47:803–6.
J Clin Epidemiol 2021;140:111–24. 38 Schulz KF, Altman DG, Moher D, et al. CONSORT 2010 statement: updated guidelines
24 Silvers-­Granelli H, Mandelbaum B, Adeniji O, et al. Efficacy of the FIFA 11+ for reporting parallel group randomised trials. BMJ 2010;340:c332.
injury prevention program in the collegiate male soccer Player. Am J Sports Med 39 Fuller CW, Ekstrand J, Junge A, et al. Consensus statement on injury definitions and
2015;43:2628–37. data collection procedures in studies of football (soccer) injuries. Br J Sports Med
25 Owoeye OBA, Akinbo SRA, Tella BA, et al. Efficacy of the FIFA 11+ warm-­up 2006;40:193–201.
programme in male youth football: a cluster randomised controlled trial. J Sports Sci 40 Hägglund M, Waldén M, Magnusson H, et al. Injuries affect team performance
Med 2014;13:321–8. negatively in professional football: an 11-­year follow-­up of the UEFA champions
26 Nuhu A, Jelsma J, Dunleavy K, et al. Effect of the FIFA 11+ soccer specific warm up league injury study. Br J Sports Med 2013;47:738–42.
programme on the incidence of injuries: a cluster-­randomised controlled trial. PLoS 41 Hrysomallis C. Relationship between balance ability, training and sports injury risk.
One 2021;16:e0251839. Sports Med 2007;37:547–56.
27 Steffen K, Emery CA, Romiti M, et al. High adherence to a neuromuscular injury 42 Al Attar WSA, Khaledi EH, Bakhsh JM, et al. Injury prevention programs that
prevention programme (FIFA 11+) improves functional balance and reduces injury risk include balance training exercises reduce ankle injury rates among soccer players: a
in Canadian youth female football players: a cluster randomised trial. Br J Sports Med systematic review. J Physiother 2022;68:165–73.
2013;47:794–802. 43 De Blaiser C, Roosen P, Willems T, et al. Is core stability a risk factor for lower
28 van Reijen M, Vriend I, van Mechelen W, et al. Compliance with sport injury extremity injuries in an athletic population? A systematic review. Phys Ther Sport
prevention interventions in randomised controlled trials: a systematic review. Sports 2018;30:48–56.
Med 2016;46:1125–39. 44 Franettovich Smith MM, Bonacci J, Mendis MD, et al. Gluteus medius activation during
29 Soligard T, Nilstad A, Steffen K, et al. Compliance with a comprehensive warm-­up running is a risk factor for season hamstring injuries in elite footballers. J Sci Med
programme to prevent injuries in youth football. Br J Sports Med 2010;44:787–93. Sport 2017;20:159–63.
30 Hammes D, Aus der Fünten K, Kaiser S, et al. Injury prevention in male veteran 45 Rinaldi VG, Prill R, Jahnke S, et al. The influence of Gluteal muscle strength deficits on
football players – a randomised controlled trial using ’FIFA 11 J Sports Sci dynamic knee Valgus: a scoping review. J Exp Orthop 2022;9:81.
2015;33:873–81. 46 Al Attar WSA, Bakhsh JM, Khaledi EH, et al. Injury prevention programs
31 O’Brien J, Young W, Finch CF. The use and modification of injury prevention exercises that include plyometric exercises reduce the incidence of anterior Cruciate
by professional youth soccer teams. Scand J Med Sci Sports 2017;27:1337–46. ligament injury: a systematic review of cluster randomised trials. J Physiother
32 Whalan M, Lovell R, Steele JR, et al. Rescheduling part 2 of the 11+ reduces injury 2022;68:255–61.
burden and increases compliance in semi-­professional football. Scand J Med Sci 47 Edouard P, Mendiguchia J, Guex K, et al. Sprinting: a key piece of the hamstring injury
Sports 2019;29:1941–51. risk management puzzle. Br J Sports Med 2023;57:4–6.
33 Bruder AM, Crossley KM, Mosler AB, et al. Co-­creation of a sport-­specific anterior 48 Faude O, Rößler R, Junge A. Football injuries in children and adolescent players: are
cruciate ligament injury risk reduction program for women: a concept mapping there clues for prevention Sports Med 2013;43:819–37.
approach. J Sci Med Sport 2020;23:353–60. 49 Beck NA, Lawrence JTR, Nordin JD, et al. ACL tears in school-­aged children and
34 Bizzini M, Dvorak J. FIFA 11+: an effective programme to prevent football adolescents over 20 years. Pediatrics 2017;139:e20161877.
injuries in various player groups worldwide—a narrative review. Br J Sports Med 50 Bahr R, Clarsen B, Derman W, et al. International Olympic Committee consensus
2015;49:577–9. statement: methods for recording and reporting of epidemiological data on injury
35 Owoeye OBA, McKay CD, Verhagen E, et al. Advancing adherence research in sport and illness in sport 2020 (including STROBE extension for sport injury and illness
injury prevention. Br J Sports Med 2018;52:1078–9. surveillance (STROBE-­SIIS)). Br J Sports Med 2020;54:372–89.

Obërtinca R, et al. Br J Sports Med 2024;58:548–555. doi:10.1136/bjsports-2023-107388 555