Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.

com
BJSM Online First, published on July 29, 2017 as 10.1136/bjsports-2017-097754
Original article

A comprehensive strength testing protocol offers no

clinical value in predicting risk of hamstring injury:
a prospective cohort study of 413 professional
football players
Nicol van Dyk,1,2 Roald Bahr,3,4 Angus F Burnett,3 Rod Whiteley,1 Arnhild Bakken,3,4
Andrea Mosler,1,5 Abdulaziz Farooq,3 Erik Witvrouw2
1
Department of Rehabilitation, ABSTRACT has received much attention in the literature.6–10
Aspetar Orthopaedic and Sports Background Hamstring injuries remain prevalent In addition, lower limb strength balance, typi-
Medicine Hospital, Doha, Qatar
2
Department of Rehabilitation across a number of professional sports. In football, the cally expressed as hamstring-to-quadriceps (H:Q)
Sciences and Physiotherapy, incidence has even increased by 4% per year at the strength ratio, has also been associated with an
Ghent University, Ghent, Champions League level over the last decade. The role of increased risk of hamstring injury, although the
Belgium
3
muscle strength or strength ratios and their association results have been inconsistent.8 10–12 Alternative
Aspetar Orthopaedic and measures of eccentric muscle strength have been
with risk of hamstring injury remain restricted by small
Sports Medicine Hospital, Doha,
Qatar sample sizes and inconclusive results. suggested,13 and recently, a novel device was devel-
4
Oslo Sports Trauma Research Purpose The purpose of this study is to identify risk oped to objectively measure eccentric hamstring
Center, Norwegian School of factors for hamstring injury in professional football muscle strength when performing the Nordic
Sport Sciences, Oslo, Norway players in an adequately powered, prospective cohort hamstring exercise.14
5
School of Allied Health, La
Trobe University, Victoria, study. Using both established (isokinetic) and novel The findings related to muscle strength or
Australia (eccentric hamstring test device) measures of muscle strength ratios and their association with risk of
strength, we aimed to investigate the relationship injury remain restricted by small sample sizes and
Correspondence to between these strength characteristics over the entire inconclusive results.5 Further exploration of these
Nicol van Dyk, Department range of motion with risk of hamstring injury. strength ratios between hamstrings and quadriceps
of Rehabilitation, Aspetar Methods All teams (n=18) eligible to compete in muscle groups has been suggested to include a more
Orthopaedic and Sports
Medicine Hospital, PO Box the premier football league in Qatar underwent a dynamic representation,15 as well as the ‘dynamic
29222, Sport City Street, Doha, comprehensive strength assessment during their annual control ratio’ which represents the net joint torque
Qatar; periodic health evaluation at Aspetar Orthopaedic and at different angles over the entire range of motion.16
n​ icol.​vanDyk@A​ spetar.​com Sports Medicine Hospital in Doha, Qatar. Variables However, this approach has not been investigated
included isokinetic strength, Nordic hamstring exercise prospectively. Although the need for adequately
Received 6 March 2017
Revised 7 July 2017 strength and dynamic hamstring: quadriceps ratios. powered studies with appropriate design was called
Accepted 10 July 2017 Results Of the 413 players included (68.2% of all for more than a decade ago,17 few such prospective
league players), 66 suffered a hamstring injury over the studies exist. Previous methodological limitations
two seasons. Only isokinetic quadriceps concentric at include inconsistency regarding injury surveillance,
300°/s (adjusted for bodyweight) was associated with biased sampling of previous injury data and a lack
risk of hamstring injury when considered categorically. of recorded exposure for the athletes included.18
Age, body mass and playing position were also The purpose of this study was to identify risk
associated with risk of hamstring injury. None of the factors for hamstring injury in professional foot-
other 23 strength variables examined were found to be ball players in an adequately powered, prospective
associated with hamstring injury. cohort study. Using both established (isokinetic) and
Conclusion The clinical value of isolated strength novel (eccentric hamstring test device) measures of
testing is limited, and its use in musculoskeletal muscle strength, we aimed to investigate the rela-
screening to predict future hamstring injury is unfounded. tionship between these strength characteristics over
the entire range of motion with risk of hamstring
injury.
Introduction
Hamstring injuries have remained prevalent across Methods
a number of professional sports.1 In football, the Study design
incidence has even increased by 4% per year over All teams (n=18) eligible to compete in the Qatar
the last decade at the Champions League level.2 Stars League (QSL), the premier football league
Recently, the complexity involved with the occur- in Qatar, agreed to participate in the study, which
rence of injury has been highlighted,3 but clinically covered two football seasons (September 2013 to
meaningful risk factor identification remains an May 2015). Each player from the respective teams
To cite: van Dyk N,
important part of the injury prevention model.4 underwent an annual periodic health evaluation
Bahr R, Burnett AF, et al.
Br J Sports Med Published The most recent meta-analysis identified age, (PHE) at Aspetar Orthopaedic and Sports Medicine
Online First: [please include previous hamstring injury and quadriceps strength Hospital in Doha, Qatar. The PHE was performed
Day Month Year]. doi:10.1136/ as risk factors for a primary or recurrent hamstring from May to September, with the official start of
bjsports-2017-097754 injury.5 Isokinetic muscle strength in particular the season in September of each year. If players

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754    1

Copyright Article author (or their employer) 2017. Produced by BMJ Publishing Group Ltd under licence.
 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article

Figure 1 Flow chart demonstrating the movement of players and repeated measurements between different seasons.

performed PHE outside of this period and met the inclusion the QSL during the previous seasons (2012–2014) and therefore
criteria, they were still included in the study. were covered by our prospective injury surveillance.
All players over the age of 18 years and eligible to compete
in the QSL, who had provided written consent and were able to Isokinetic strength testing
perform the strength testing, were included. Players who were Knee flexion and extension muscle strength were tested using an
injured at the time of the PHE and therefore unable to perform isokinetic dynamometer (Biodex Multi-joint System 3, Biodex
the tests were excluded. If no isokinetic test was performed at Medical Systems, New York, USA). After an explanation of the
the start of a season or no exposure or injury surveillance data testing methodology, the player performed a 5–10 min warm up
were recorded over an entire season, players were also excluded. routine, cycling on a stationary exercise bike (Bike Forma, Tech-
Figure 1 depicts the inclusion methodology during the two study nogym, Cesena, Italy).
seasons. The order (ie, left, right) was randomised and maintained for
each of the three different testing modes and speeds for the each
Player information subject. All players completed the test procedure as previously
All likely non-modifiable risk factors were considered. A history described.18
of previous hamstring injury in the past 12 months,19 20 season, First, the players were tested over five repetitions of concentric
team, leg dominance, position and ethnicity were recorded. knee flexion and extension at 60°/s, followed by 10 repetitions
Player height and weight were measured, and body mass index of concentric knee flexion and extension at 300°/s. These test
(BMI) was calculated during the PHE. modes measures the concentric strength of the quadriceps (knee
History of previous injury was self-reported at the time of extension) and hamstring (knee flexion) muscles. Finally, players
screening and cross-checked with their hospital medical file for performed five repetitions of eccentric knee extension at 60°/s
the entire cohort. To minimise recall bias and account for players which measures the eccentric strength of the hamstring muscles.
not willing to reveal past injuries during screening, we also The highest peak torque value observed from all repetitions
conducted a subgroup analysis including only players playing in performed for each of the three different tests was recorded.

2 van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article
Dynamic control profile A hamstring injury was defined as acute pain in the posterior
The dynamic control profile represents the net joint torque thigh that occurred during training or match play and resulted in
(eccentric hamstrings to concentric quadriceps) over the entire immediate termination of play and inability to participate in the
range of motion during isokinetic testing.16 The specific knee next training session or match.22 These injuries were confirmed
flexion angle where the quadriceps torque was greater than through clinical examination (identifying pain on palpation, pain
the eccentric hamstring torque was identified, and the torque- with isometric contraction and pain with muscle lengthening) by
angle plots for eccentric hamstrings and concentric quadriceps the club medical team. If indicated, the clinical diagnosis was
were determined using a custom algorithm created in Labview supported by ultrasonography and MRI at the study centre. A
(V.7.0 National Instruments, Austin, Texas, USA) and exported recurrent injury was defined as a hamstring injury that occurred
in Microsoft Excel (Microsoft Office 2013, Redmond, Wash- in the same limb and within 2 months of the initial injury.23
ington, USA).
The peak torque (Nm) measurements for concentric knee
flexion and extension at 60°/s and eccentric knee extension at Statistical analyses
60°/s were used to define the dynamic control profile. Concen- Univariate analyses (independent t-tests) were performed
tric H:Q ratio (concentric hamstrings divided by concentric between the limbs of the injured and the uninjured players for
quadriceps) and dynamic control ratios (eccentric hamstrings the isokinetic strength test, Nordic hamstring exercise test and
was divided by the concentric quadriceps) were calculated at dynamic control profile. Injured limbs were compared with
angles of 30°, 40° and 50°. uninjured limbs among injured players and then to all uninjured
For each data point (angle), the torque value for concentric limbs among the uninjured players.
quadriceps was subtracted from the eccentric hamstrings. Conse- Due to the consistency in our sample, the repeated measures
quently, the ‘angle of crossover’ was identified as the point where performed over the two seasons, as well as the fact that not every
the net joint torque was equal to zero. player had the same number of measurements (ie, some subjects
would have test results including both limbs for both seasons,
while other subjects might only have been tested once), SEs
Nordic hamstring exercise testing would have increased when using general estimating equations
Before the start of season 2, players also performed one set in a traditional Cox regression model. Therefore, we performed
of three maximal repetitions on a device specially designed to a univariate Cox regression analysis in STATA (V.11.0, College
measure the Nordic hamstring exercise.14 After completion of Station, Texas, USA) using the limb as the unit of analyses,
the isokinetic test, the players were tested in a kneeling position adjusting for player identity as a cluster factor. Exposure was
on a padded board, with both ankles secured immediately above totalled as duration in hours for game and training combined
the lateral malleolus by individual ankle braces. These braces from the start to the end of each season or time to first injury.
were attached to uniaxial load cells (Delphi Measurement, Variables independently associated with hamstring strain injury
Gold Coast, Australia) with wireless data acquisition capabili- were determined from the univariate analyses. A p value of ≤0.05
ties (Mantracourt Electronics, Farringdon, UK). The device has was considered statistically significant. Effect size, which is the
been described previously14 and allows for separate measure- quantitative measure of the strength of an observed occurrence,
ments of each limb. The player was instructed to progressively was calculated and interpreted as small (0.2–0.3), medium (0.5)
lean forward at the slowest possible speed resisting the move- or large (>0.8).24
ment with both limbs while keeping the trunk and hips in a Potential risk factors were treated as continuous and cate-
neutral position and the hands held across the chest. If the force gorical variables. In the continuous analyses, all variables with
output reached a distinct peak (indicative of maximal eccentric p value ≤0.10 were considered further in a backward step-
strength), followed by a rapid decline in force that occurred wise multivariate Cox regression analysis to evaluate potential
when the player could no longer resist the effects of gravity, a predictor variables. HRs with 95% CIs are presented with exact
trial was deemed acceptable. p values, and p values of ≤0.05 were considered statistically
significant.
For the categorical analyses, the limbs of players were grouped
Injury surveillance for isokinetic strength, Nordic hamstring exercise and dynamic
All participating QSL teams are provided with medical services control profile. The OR and 95% CI were calculated for the
by the National Sports Medicine Programme, a department with groups with the lowest (<1 SD below the mean) and the highest
the Aspetar Orthopaedic and Sports Medicine Hospital. This (>1 SD above the mean) values for each variable, respectively,
centralised system with a focal point for the medical care of with the intermediate group as the reference group.25
each club competing in the QSL allowed for standardisation of
the ongoing injury surveillance through the Aspetar Injury and
Illness Surveillance Programme (AIISP).21 Results
The AIISP includes prospective injury and exposure (minutes Players
of training and match play) recording from all QSL teams. The During the two-season study period, 592 elite male soccer players
injury data were collected monthly, with regular communica- (age 25.8±4.8 years, height 177±7 cm, weight 72.4±9.3 kg,
tion with the responsible team physician/physiotherapist to BMI 23.1±2) reported for screening and were considered for
encourage timely and accurate reporting. Throughout the 2013 isokinetic testing. Players who were unable to perform the test
and 2014 season (July to May; 44 weeks), training and match due to injury, did not provide consent or had no exposure data
exposure for each team were recorded by the team physician (or recorded in either season (n=179, age 25.3±4.5 years, height
lead physiotherapist if no team physician was available). At the 177±7 cm, weight 73.5±9.8 kg, BMI 23.4±1.8) were excluded
conclusion of each season, all the data from the individual clubs from the final analyses. The remaining 413 players performed a
were collated into a central database, and discrepancies were total of 1087 isokinetic test procedures (considering both limbs)
identified and followed up at the different clubs to be resolved. over the two seasons.

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754 3

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article
Strength measurements as potential risk factors
Table 1 Characteristics of injured (n=66) and uninjured players
The results of the univariate analyses are shown tables 3 and 4
(n=347)
for isokinetic strength and Nordic hamstring exercise, respec-
Injured (n=66) Uninjured (n=347) p Value tively. Among injured players (n=66), there were no differences
Age, years 27.9 (4.3) 25.9 (4.9) 0.002 in strength between the injured and uninjured limbs.
Weight, kg 72.2 (7.7) 72.6 (9.2) 0.86 Comparing injured limbs (n=69) to the uninjured limbs
Height, cm 175.8 (6.7) 176.8 (6.8) 0.30 (n=1018), the parameter estimates of the univariate Cox regres-
Body mass index, kg/m2 23.4 (1.9) 23.1 (2.0) 0.33 sion analyses of the isokinetic strength test variables are presented
Previous Injury, n (%) 21 (31.8) 117 (30.9) 0.89 in table 3, expressed as HRs per one-unit (1 Nm/kg) strength
Player position, n (%) 0.02 change. In the continuous analyses, none of the 11 strength vari-
 Goalkeeper 1 (1.4) 44 (11.6) ables were found to be significantly associated with an increased
 Defender 28 (42.4) 122 (32.3)
risk of hamstring injury (table 3). The categorical analyses identi-
fied the greater strength group (>1 SD above mean, 2.2–3.7 Nm/
 Midfielder 25 (37.9.) 140 (37.0)
kg) for quadriceps concentric torque at 300°/s (normalised to
 Forward 12 (18.2) 72 (19.0)
bodyweight) as being at increased risk for injury (table 2).
Limb dominance, n (%) 0.39
No significant differences were found for any of the Nordic
 Left 16 (24.2) 71 (18.8)
hamstring exercise test variables between injured and uninjured
 Right 50 (75.8) 307 (81.2)
limbs (table 4).
Ethnicity, n (%) 0.16 When profiling dynamic control, no difference was observed
 Arab 35 (53.0) 225 (59.5) either in the angle of cross over between the injured limbs
 Black 23 (34.8) 109 (28.8) (n=56) and uninjured limbs (n=752) (injured limbs: 45°±8°
 Asian 2 (3.0) 9 (2.4) (SD), uninjured limbs: 44°±7°) or in the dynamic control ratio
 Caucasian 6 (9.1) 35 (9.38.6) (figure 2).
Data are shown as mean values with SD or percentages.

Discussion
The main finding of this prospective two-season cohort study
New hamstring strain injuries of 413 football players, the largest to date, was that none of
Over the two seasons, 413 unique players (68.2% of all QSL the 24 strength variables examined differed between injured and
players) competed for 544 player seasons (132 players competed healthy players. The only exception was that the group with the
both seasons) (figure 1). In total, 66 of the 413 players sustained highest quadriceps concentric torque at 300°/s (>1 SD below
69 index hamstring injuries. The three players who had more the mean) had an increased risk of hamstring injury. Age, body
than one injury were retained in the analyses (none of these mass and playing position (ie, being a goalkeeper) were associ-
injuries met the criteria for reinjury), and all injured players in ated with injury risk.
season 1 had their previous injury status adjusted accordingly in
season 2.
Modifiable risk factors
The comparison of the strength measures between the injured
Non-strength-related risk factors and uninjured groups (tables 3 and 4) clearly demonstrate that
There were no differences in height, ethnicity, limb domi- it is not possible to distinguish between the groups clinically. In
nance and body composition between injured and uninjured the categorical analyses, the greater strength group for concen-
groups (table 1). Previous hamstring injury was reported by tric quadriceps strength at 300°/s (adjusted for bodyweight) was
31% of the entire cohort (n=413) with no significant difference found to be significant (see table 2). Although our finding of a
between injured and uninjured players. Also in the subgroup of weak association with quadriceps strength is supported by the
players with injury history based on injury surveillance during meta-analyses performed by Freckleton and Pizzari,5 the small
the previous season (n=336), a history of previous injury did not effect size of 0.2 and the fact that there was no group differ-
represent an increased risk of new hamstring injury. ence in strength indicates that this holds little clinical value. The
Univariate analyses identified age and position as potential risk smallest detectable difference for concentric quadriceps peak
factors for hamstring injury (table 1). Goalkeepers were signifi- torque is reported between 11.9% and 20%,26 27 and therefore,
cantly less likely to sustain a hamstring injury than defenders, the differences reported in this study are likely equivalent to
midfielders or forwards. When age was considered as a cate- test–retest variability. Comparison to previous findings is diffi-
gorical variable, players in the younger age group (<1 SD below cult, such as the testing protocol, inclusion criteria, duration of
the mean, 18–21 years) had a lower risk of injury than the inter- the follow-up period and injury definition. It is, however, more
mediate age group (table 2). Players who weighed more (>1 SD striking that only 1 out of 24 strength variables evaluated (11
above the mean, 81.8–104.5 kg) were at lower risk for injury isokinetic strength test, 5 Nordic hamstring exercise test and
compared with the intermediate weight group (table 2). 8 dynamic control profile measures) was weakly associated with
In the multivariate Cox regression analysis, age (HR 1.07 per an increased risk of hamstring injury.
1-year increase in age, 95% CI 1.03 to 1.12) and position (HR The Nordic hamstring exercise has received much attention
5.79 for outfield players vs goalkeepers, 95% CI 1.44 to 23.32) in the literature, and its value as a preventative tool is well estab-
were retained from the univariate analyses and were significantly lished.28 Further investigation has been done to examine the use
associated with hamstring injury risk. of this exercise as a test to determine risk of hamstring injury,
The OR (95% CI) was calculated for the group of players and initially, no significant association was found between a
with the lowest (>1 SD below the mean) and the highest (>1 SD simple visual assessment of test performance and hamstring
above mean) values for each variable, respectively, with the injury risk.13 Despite this, a novel device has recently been
intermediate group of players as reference group. developed to accurately measure eccentric hamstring strength

4 van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article

Table 2 Comparison between uninjured and injured players with potential risk factors treated as categorical variables using univariate Cox
regression analyses
Risk as a categorical variable
>1 SD below the mean >1 SD above the mean
HR (95% CI) p Value HR (95% CI) p Value
Age, years 0.15 (0.04 to 0.61) 0.008 1.21 (0.72 to 2.03)  0.48
Weight, kg 0.95 (0.54 to 1.66) 0.90 0.39 (0.17 to 0.89) 0.024
 
Height, cm 0.93 (0.50 to 1.73) 0.81 0.64 (0.28 to 1.44)  0.28
Body mass index, kg/m2 0.80 (0.41 to 1.54) 0.50 0.79 (0.42 to 1.49)  0.46
Quadriceps
Concentric at 60°/s 0.82 (0.41 to 1.63) 0.57 1.00 (0.53 to 1.91)  0.10
 BW adjusted 1.10 (0.56 to 2.18) 0.78 1.46 (0.77 to 2.79)  0.25
Concentric at 300°/s 0.43 (0.17 to 1.13) 0.09 1.37 (0.81 to 2.34)  0.25
 BW adjusted 0.90 (0.43 to 1.90) 0.79 2.06 (1.21 to 3.51) 0.008
 
Hamstrings
Concentric at 60°/s 0.78 (0.37 to 1.67) 0.53 0.72 (0.35 to 1.49)  0.38
 BW adjusted 0.92 (0.45 to 1.89) 0.82 0.88 (0.45 to 1.89)  0.71
Concentric at 300°/s 0.41 (0.14 to 1.18) 0.10 0.72 (0.37 to 1.40)  0.33
 BW adjusted 0.75 (0.33 to 1.69) 0.49 1.49 (0.82 to 2.70)  0.19
Eccentric at 60°/s 0.68 (0.34 to 1.38) 0.29 0.48 (0.21 to 1.14)  0.10
 BW adjusted 0.83 (0.42 to 1.64) 0.59 1.02 (0.52 to 2.03)  0.95
Hamstrings eccentric 60°/s to quadriceps concentric 300°/s 0.85 (0.43 to 1.70) 0.66 0.70 (0.33 to 1.52)  0.37
Peak force (N) 0.87 (0.30 to 2.50) 0.79 1.51 (0.43 to 3.06)  0.78
 BW adjusted (N/kg) 0.83 (0.29 to 2.39) 0.73 1.39 (0.57 to 2.39)  0.47
Peak force imbalance (N) 0.40 (0.54 to 2.99) 0.37 0.87 (0.25 to 2.99)  0.82
Average force (N) 0.60 (0.18 to 1.96) 0.40 1.06 (0.42 to 2.68)  0.90
 BW adjusted (N/kg) 0.79 (0.28 to 2.28) 0.67 1.05 (0.41 to 2.66)  0.93
Angle of crossover 0.89 (0.42 to 1.89) 0.76 1.42 (0.70 to 2.89)  0.34
Dynamic control ratio 0.75 (0.29 to 1.95) 0.55 1.79 (0.91 to 3.51)  0.09
Dynamic control ratio at various degrees in ROM
 Concentric at 30° 0.95 (0.67 to 1.42) 0.71 1.43 (0.78 to 3.41)  0.44
 Concentric at 40° 0.89 (0.56 to 1.31) 0.49 1.74 (0.91 to 2.98)  0.17
 Concentric at 50° 1.60 (1.00 to 2.43) 0.50 1.81 (0.95 to 3.46)  0.07
 Eccentric at 30° 1.16 (0.30 to 4.47) 0.83 1.78 (0.79 to 4.00)  0.16
 Eccentric at 40° 0.90 (0.38 to 2.10) 0.80 1.32 (0.64 to 2.69)  0.45
 Eccentric at 50° 0.71 (0.31 to 1.62) 0.41 1.38 (0.65 to 2.92)  0.40
 Overall H:Q ratio 0.42 (0.57 to 3.08) 0.39 1.71 (0.80 to 3.66)  0.80
The odds ratio (OR; 95% confidence interval[CI]) was calculated for the group of players with the lowest (>1 SD below the mean) and the highest (>1 SD above mean) values for
each variable, respectively, with the intermediate group of players as reference group BW, body weight; H:Q, hamstring to quadriceps; ROM, range of motion.

when performing this exercise.14 Subsequent studies positively a standalone Nordic hamstring exercise test to individual risk
identified players with inferior eccentric strength as being at for injury.35
increased risk for hamstring injury,29 30 while Bourne et al31 The H:Q ratio as conventional and dynamic entities of mixed
found no increased risk with lower eccentric strength in rugby isokinetic strength has been identified previously as risk factors
union players. These previous studies determined best fit cut-off for hamstring injury,7 8 11 36 with some debate over how these
values for eccentric hamstring strength and assessed the risk of ratios are interpreted. Essentially, we should consider that
injury based on these in multivariate models. In the present scaling these data may not appropriately represent the lower
study, we compared eccentric strength in the Nordic hamstring and higher end of the range (ie, the slope of the relationship
exercise test between injured and uninjured groups and also between the two variables is not equal to 1). Furthermore, when
assessed risk in the group with inferior performance. However, we divide two normally distributed variables, the resulting ratio
none of the variables related to the Nordic hamstring exercise is unlikely to be normally distributed itself.37 Two large previous
test were found to be significantly associated with an increased investigations have reported contradicting results, which make it
risk of hamstring injury. We do not dispute that the Nordic difficult to determine whether these strength ratios are valuable
hamstring exercise may be a useful injury prevention tool.32 or not.8 18 Therefore, the dynamic control profile, as described
33
It may alter the influence of non-modifiable risk factors
by Graham-Smith et al,16 was also included to explore the rela-
such as age or previous injury29 or lead to protective muscle
tionship between hamstring and quadriceps strength throughout
architecture adaptations.30 34 However, the present findings
the entire test range of motion.
urge caution if the clinician attempts to relate the results of
No association was found between the dynamic H:Q ratios
(knee flexion (‘hamstring’) eccentric peak torque at 60°/s to

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754 5

6
Original article

Table 3 Univariate comparison of isokinetic strength tests between the injured and the uninjured limb in the injured players, all uninjured limbs in the uninjured players and Cox regression analysis
demonstrating parameter estimates (95% CIs) for all isokinetic strength variables when comparing injured to uninjured limbs.
Injured players Uninjured players Univariate Cox regression
Injured limb Uninjured limb Uninjured limbs
(n=69) (n=69) Difference 95% CI p Value (n=948) Difference 95% CI p Value HR 95% CI p Value
Quadriceps
 Concentric at 60°/s 235.3 (46.3) 239.2 (46.7) 3.9 −12.1 to 19.9 0.62 234.0 (46.9) −1.3 −13.0 to 10.4 0.81 1.00 1.00 to 1.01 0.96
  BW adjusted 3.28 (0.6) 3.34 (0.6) 0.06 −0.15 to 0.27 0.55 3.23 (0.6) −0.05 −0.20 to 0.10 0.55 1.09 0.68 to 1.74 0.71
 Concentric at 300°/s 139.3 (30.9) 134.2 (28.6) −5.1 −15.4 to 5.2 0.32 134.9 (26.3) −4.4 −11.1 to 2.3 0.25 1.01 1.00 to 1.02 0.17
  BW adjusted 1.93 (0.4) 1.86 (0.4) −0.07 −0.21 to 0.07 0.28 1.86 (0.3) −0.07 −0.15 to 0.01 0.13 2.05 0.96 to 4.37 0.06
Hamstrings
 Concentric at 60°/s 126.6 (23.9) 124.9 (25.8) −1.7 −10.3 to 6.9 0.70 126.1 (27.8) −0.5 −7.4 to 6.4 0.88 1.00 0.99 to 1.01 0.97
  BW adjusted 1.76 (0.3) 1.74 (0.3) −0.02 −0.12 to 0.08 0.68 1.74 (0.3) −0.02 −0.10 to 0.06 0.54 1.16 0.55 to 2.46 0.70
 Concentric at 300°/s 97.6 (17.6) 94.8 (17.6) −2.8 −8.9 to 3.3 0.35 96.5 (20.5) −1.1 −6.2 to 4.0 0.63 1.00 0.99 to 1.01 0.60
  BW adjusted 1.36 (0.2) 1.32 (0.2) −0.04 −0.1 to 0.03 0.34 1.33 (0.3) −0.03 −0.10 to 0.04 0.35 1.62 0.60 to 4.34 0.34
 Eccentric at 60°/s 206.3 (40.1) 203.1 (40.6) −3.2 −17.1 to 10.7 0.65 203.2 (43.7) −3.1 −14.0 to 7.8 0.55 1.00 1.00 to 1.01 0.48
  BW adjusted 2.86 (0.5) 2.82 (0.5) −0.04 −0.21 to 0.13 0.65 2.80 (0.5) 0.98 0.94 to 1.02 0.32 1.30 0.81 to 2.11 0.28
H:Q ratio (functional)
 Hamstrings eccentric 60°/s to 1.52 (0.3) 1.55 (0.3) 0.03 −0.07 to 0.13 0.49 1.53 (0.3) 0.01 −0.07 to 0.09 0.72 0.90 0.41 to 1.96 0.79
quadriceps concentric 300°/s
Absolute values and values adjusted for body weight (BW) are shown as mean force values in Newton-metre (Nm) with SD.
H:Q, hamstring to quadriceps.
Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article

Table 4 Univariate comparison of Nordic hamstring exercise test results between the injured and the uninjured limb in the injured players, all uninjured limbs in the uninjured players and Cox regression
analysis demonstrating parameter estimates (95% CIs) for all strength variables included when comparing injured to uninjured limbs. Absolute values and values adjusted for body weight (BW) are shown

p Value
0.36
0.21
0.50
0.47
0.22
Univariate Cox regression

 0.98 to 1.01
 0.88 to 1.76
 0.98 to 1.01
 1.00 to 1.01
 0.88 to 1.77
95% CI
1.00
1.25
1.00
1.00
1.24
HR
p Value
0.32
0.20
0.59
0.40
0.20
Figure 2 Dynamic control ratio at 30°, 40° and 50° for injured (closed
−31.29 to 5.29
−0.51 to 0.01
−4.7 to 10.29
−28.30 to 7.10
−0.49 to 0.01

symbols) and uninjured players (open symbols) throughout the test

range of motion for the two test modes (concentric and eccentric). Data
95% CI

represent group means with SD.

Difference

knee extension (‘quadriceps’) concentric peak torque at 300°/s)

−0.25

−0.24

(table 2), a finding supported by a previous meta-analysis.5

−13.0

2.8
−10.6

Additionally, there was no difference in the dynamic control

profile over the entire range of motion or the angle of crossover.
Uninjured players

Figure 2 clearly demonstrates that the ratios for both the injured
Uninjured limbs

and uninjured players were indistinguishable. Although this

4.20 (1.01)

3.97 (0.98)
299.1 (70.9)

33.6 (29.8)
281.9 (68.7)

dynamic ratio might be considered a valuable tool when inter-

(n=432)

preting isokinetic strength in previously injured players, it adds

little value for the clinician to establish risk of hamstring injury
for the individual. Our findings suggest that H:Q ratios in any
form are ineffective in risk factor identification.
p Value
0.55
0.75
0.75
0.76
0.79

Mean body mass did not differ between groups, but players
with greater body mass (>1 SD above the mean, >82 kg) did
Absolute values and values adjusted for body weight (BW) are shown as mean force values in Newton(N) with SD.

have a lower risk for a hamstring injury than the intermediate

−43.74 to 23.74

−10.66 to 18.66
−38.66 to 28.26

weight group. However, the absolute difference between the

−0.60 to 0.44

−0.58 to 0.44

injured and uninjured players was less than 1 kg. This finding is
therefore also not clinically useful, and as previously reported,5
95% CI

13 25
we would not consider body mass as an important risk
factor for hamstring injury.
Difference

−0.08

−0.07

Non-modifiable risk factors

−10.0

4.0
−5.2

Age has consistently been identified as a risk factor for hamstring

injury,5 10 20 25 38 as in the current study, where we observed a 7%
increased risk of hamstring injury per year added. The injury risk
Uninjured limb

4.37 (0.97)

4.14 (0.97)
302.1 (60.8)

34.8 (33.1)
287.3 (62.7)

of the youngest group (<22 years) was 85% less than the inter-
mediate group, while we detected no difference in risk between
(n=29)

the intermediate and the oldest groups (table 2). This supports
the results of previous studies where fewer hamstring injuries
as mean force values in Newton(N) with SD.

were found in the younger age group.39 Playing positions were

Injured players

4.45 (1.01)

4.21 (0.98)
Injured limb

retained in the multivariate analyses to investigate whether it

312.1 (67.3)

30.8 (21.4)
292.5 (64.5)

might influence age, as goalkeepers were on average 1 year older

(n=29)

than the other playing positions combined. It did not, however,

change the risk for these players significantly, even though
outfielders were five times more likely to get injured compared
with goalkeepers.
Peak force imbalance (N)

Surprisingly, a history of previous injury was not found to be

 BW adjusted (N/kg)

 BW adjusted (N/kg)

associated with risk of hamstring injury, although it is consis-

Average force (N)

tently identified as a risk factor in the literature.5 40 This finding

Peak force (N)

was further explored, attempting to reduce the risk of recall bias

in self-reporting injuries (or unwillingness to disclose previous
injuries), by analysing a subgroup of players where the previous
injury status was known through detailed injury surveillance.

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754 7

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article
Again, no significant association was found. In contrast, a retro-
spective investigation of the same league over previous seasons
What are the findings?
did identify previous injury as a risk factor.18 To interpret these
►► Isolated strength measurements cannot assist the clinician in
contrasting results, consider that two large randomised control
predicting risk of hamstring injury.
trials were being conducted at the Aspetar Orthopaedic and
►► Functional and dynamic measures of isokinetic H:Q ratio are
Sports Medicine Hospital concurrent with this prospective
not useful in determining individual risk of injury.
study.41 Both these studies incorporated a structured crite-
►► Age and playing position are non-modifiable risk factors
ria-based rehabilitation programme, including a large number of
associated with an increased risk of hamstring injury.
QSL players. The first randomised controlled trial reported a low
►► Population-based risk of injury in previously injured players
re-injury rate of 6%23 compared with other football leagues.42
might be reduced by implementation of criteria based
A reduced risk in previously injured players has been reported
rehabilitation programmes.
before43 due to the effect of successful intervention programmes.
Our finding suggests a similar effect, where the introduction of
systematic criteria-based rehabilitation programmes may have
reduced the risk associated with previous injury.
How might it impact on clinical practice in the future?
Strengths and limitations ►► Greater consideration might be given to the use of strength
To detect strong to moderate associations in prospective cohort
testing and strength ratios in identifying players at risk
studies, it is suggested that 30–40 injury cases are needed, while
for hamstring injury.
for small to moderate associations to be detected, 200 injury
►► Muscle strength continues to be identified in the causation
cases are needed.17 Importantly, even though this is one of the
of hamstring injury and must be considered in prevention
largest prospective studies to date on risk factors for hamstring
measures.
injury, we were not able to include enough cases to detect small
►► The small association found with strength highlights again
associations, illustrating the difficulty in performing adequately
the multifactorial nature of hamstring injury.
powered investigations. However, the effect sizes calculated for
each strength variable never exceeded 0.3, failing to reach clin-
ical significance.
Intraseason variability in the repeated measures over the 2-year associated with hamstring injury when considered categorically,
observation period may have limited our ability to identify an although the magnitude of this difference is likely too small to
association between strength and injury risk. Injury surveillance be clinically relevant. The use of strength measurements and
was carefully monitored during each season; however, we cannot different strength characteristics in musculoskeletal screening to
discount that the clinical criteria used to confirm a hamstring predict future hamstring injury is unfounded.
injury may have involved other differential diagnoses.
All tests performed used the same isokinetic testing system Contributors ND: principal investigator, data collection and analyses, writing,
with highly experienced assessors in a multinational, multilan- editing and outline of the manuscript. EW: first principal supervisor of PhD
programme, concept and outline as well as writing and editing of the manuscript.
guage clinical setting for professional athletes. Although every
RB: second main supervisor, concept and outline as well as writing and editing of the
effort was made to ensure players understood the test procedure manuscript. AFB: statistical analyses and data analyses, writing of the manuscript.
and instructions, it is possible that some players did not compre- RW: investigator and coauthor, particularly with the data collection and writing
hend the instructions fully. However, this reflects a ‘real-world’ of the manuscript. AB: organisation and planning of study, data collection and
scenario, increasing the external validity of the study. writing of manuscript. AM: study design and ethical approval, data collection and
organisation, and writing of the manuscript. AF: statistical analyses.
We also acknowledge the homogeneity of our study population
of professional male football players, which limits the general- Competing interests None declared.
isation of these findings to other sports, age groups or female Patient consent Obtained.
players. Other factors such as training culture and possible Ethics approval Institutional Review Board, Anti-Doping Lab, Qatar.
prevention strategies within different teams or climate specific Provenance and peer review Not commissioned; externally peer reviewed.
to the Middle East region could have influenced the results.
Data sharing statement Patient-level data and/or full data set and/or and/or
statistical code are available from the corresponding author.
Clinical implications © Article author(s) (or their employer(s) unless otherwise stated in the text of the
It is clear that isolated strength variables, including the new article) 2017. All rights reserved. No commercial use is permitted unless otherwise
test device for the Nordic hamstring exercise and a more expressly granted.
comprehensive interpretation of strength ratios, have limited
clinical application in identifying individual players at risk. Our References
continued pursuit of risk factor identification through perfor- 1 Opar DA, Williams MD, Shield AJ, et al. Hamstring strain injuries. Sports Medicine
2012;42:209–26.
mance tests seems incongruous. Muscle strength continues to 2 Ekstrand J, Waldén M, Hägglund M. Hamstring injuries have increased by 4% annually
form part of a multifactorial complex model that may lead to in men's professional football, since 2001: a 13-year longitudinal analysis of the UEFA
injury. However, our findings should urge the clinician to exer- Elite Club injury study. Br J Sports Med 2016;50:731–7.
cise caution and not translate the results of these commonly used 3 Bittencourt NF, Meeuwisse WH, Mendonça LD, et al. Complex systems approach for
sports injuries: moving from risk factor identification to injury pattern recognition-
screening tests into risk of injury for the individual player.
narrative review and new concept. Br J Sports Med 2016;50:1309–14.
4 van Mechelen W, Hlobil H, Kemper HC. Incidence, severity, aetiology and prevention of
Conclusion sports injuries. A review of concepts. Sports Med 1992;14:82–99.
5 Freckleton G, Pizzari T. Risk factors for hamstring muscle strain injury in sport: a
None of the other 24 strength-related variables were found to systematic review and meta-analysis. Br J Sports Med 2013;47:351–8.
be associated with hamstring injury. Increased concentric quad- 6 Bennell K, Wajswelner H, Lew P, et al. Isokinetic strength testing does not predict
riceps strength when adjusted for body weight was the only hamstring injury in Australian rules footballers. Br J Sports Med 1998;32:309–14.

8 van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

Original article
7 Sugiura Y, Saito T, Sakuraba K, et al. Strength deficits identified with concentric action 25 Arnason A, Sigurdsson SB, Gudmundsson A, et al. Risk factors for injuries in football.
of the hip extensors and eccentric action of the hamstrings predispose to hamstring Am J Sports Med 2004;32:5S–16.
injury in elite sprinters. J Orthop Sports Phys Ther 2008;38:457–64. 26 Otten R, Whiteley R, Mitchell T. Effect of subject restraint and resistance pad
8 Croisier JL, Ganteaume S, Binet J, et al. Strength imbalances and prevention of placement on isokinetic knee flexor and extensor strength: implications for testing
hamstring injury in professional soccer players: a prospective study. Am J Sports Med and rehabilitation. Sports Health 2013;5:137–42.
2008;36:1469–75. 27 McCleary RW, Andersen JC. Test-retest reliability of reciprocal isokinetic knee
9 Cameron M, Adams R, Maher C. Motor control and strength as predictors of extension and flexion peak torque measurements. J Athl Train 1992;27:362–5.
hamstring injury in elite players of Australian football. Physical Therapy in Sport 28 Al Attar WSA, Soomro N, Sinclair PJ, et al. Effect of injury prevention programs that
2003;4:159–66. Include the Nordic hamstring exercise on hamstring injury rates in soccer players: a
10 Orchard J, Marsden J, Lord S, et al. Preseason hamstring muscle weakness systematic review and meta-analysis. Sports Med 2017;47:907–16.
associated with hamstring muscle injury in Australian footballers. Am J Sports Med 29 Opar DA, Williams MD, Timmins RG, et al. Eccentric hamstring strength and hamstring
1997;25:81–5. injury risk in australian footballers. Med Sci Sports Exerc 2015;47:857–65.
11 Yeung SS, Suen AM, Yeung EW. A prospective cohort study of hamstring injuries in 30 Timmins RG, Bourne MN, Shield AJ, et al. Short biceps femoris fascicles and eccentric
competitive sprinters: preseason muscle imbalance as a possible risk factor. Br J Sports knee flexor weakness increase the risk of hamstring injury in elite football (soccer): a
Med 2009;43:589–94. prospective cohort study. Br J Sports Med 2016;50:1524–35.
12 Henderson G, Barnes CA, Portas MD. Factors associated with increased propensity 31 Bourne MN, Opar DA, Williams MD, et al. Eccentric knee flexor strength and
for hamstring injury in English Premier League soccer players. J Sci Med Sport risk of hamstring injuries in Rugby Union: a prospective study. Am J Sports Med
2010;13:397–402. 2015;43:2663–70.
13 Engebretsen AH, Myklebust G, Holme I, et al. Intrinsic risk factors for hamstring 32 Petersen J, Thorborg K, Nielsen MB, et al. Preventive effect of eccentric training on
injuries among male soccer players: a prospective cohort study. Am J Sports Med acute hamstring injuries in men's soccer: a cluster-randomized controlled trial. Am J
2010;38:1147–53. Sports Med 2011;39:2296–303.
14 Opar DA, Piatkowski T, Williams MD, et al. A novel device using the Nordic hamstring 33 van der Horst N, Smits DW, Petersen J, et al. The preventive effect of the nordic
exercise to assess eccentric knee flexor strength: a reliability and retrospective injury hamstring exercise on hamstring injuries in amateur soccer players: a randomized
study. J Orthop Sports Phys Ther 2013;43:636–40. controlled trial. Am J Sports Med 2015;43:1316–23.
15 Aagaard P, Simonsen EB, Magnusson SP, et al. A new concept for isokinetic hamstring: 34 Bourne MN, Williams MD, Opar DA, et al. Impact of exercise selection on hamstring
quadriceps muscle strength ratio. Am J Sports Med 1998;26:231–7. muscle activation. Br J Sports Med 2017;51:469–77.
16 Graham-Smith P, Jones PA, Comfort P, et al. Assessment of knee flexor and extensor 35 Bahr R. Why screening tests to predict injury do not work-and probably never will…:
muscle balance. International Journal of Athletic Therapy and Training 2013;18:1–5. a critical review. Br J Sports Med 2016;50:776–80.
17 Bahr R, Holme I. Risk factors for sports injuries—a methodological approach. Br J 36 Fousekis K, Tsepis E, Poulmedis P, et al. Intrinsic risk factors of non-contact quadriceps
Sports Med 2003;37:384–92. and hamstring strains in soccer: a prospective study of 100 professional players. Br J
18 van Dyk N, Bahr R, Whiteley R, et al. Hamstring and quadriceps isokinetic strength Sports Med 2011;45:709–14.
deficits are weak risk factors for hamstring strain injuries: a 4-year cohort study. Am J 37 Atkinson G, Batterham AM. The use of ratios and percentage changes in sports
Sports Med 2016;44:1789–95. medicine: time for a rethink? Int J Sports Med 2012;33:505–6.
19 Opar DA, Williams MD, Timmins RG, et al. The effect of previous hamstring strain 38 Gabbe BJ, Bennell KL, Finch CF, et al. Predictors of hamstring injury at the elite level of
injuries on the change in eccentric hamstring strength during preseason training in Australian football. Scand J Med Sci Sports 2006;16:7–13.
elite Australian footballers. Am J Sports Med 2015;43:377–84. 39 Woods C, Hawkins RD, Maltby S, et al. The Football Association Medical Research
20 Hägglund M, Waldén M, Ekstrand J. Previous injury as a risk factor for injury in Programme: an audit of injuries in professional football—analysis of hamstring
elite football: a prospective study over two consecutive seasons. Br J Sports Med injuries. Br J Sports Med 2004;38:36–41.
2006;40:767–72. 40 van Beijsterveldt AMC, van de Port IGL, Vereijken AJ, et al. Risk factors for hamstring
21 Bakken A, Targett S, Bere T, et al. Interseason variability of a functional movement injuries in male soccer players: a systematic review of prospective studies: risk factors
test, the 9+ screening battery, in professional male football players. Br J Sports Med for hamstring injuries. Scand J Med Sci Sports 2013;23:253–62.
2017;51:1081–6. 41 Hamilton B, Tol JL, Almusa E, et al. Platelet-rich plasma does not enhance return
22 Eirale C, Farooq A, Smiley FA, et al. Epidemiology of football injuries in Asia: a to play in hamstring injuries: a randomised controlled trial. Br J Sports Med
prospective study in Qatar. J Sci Med Sport 2013;16:113–7. 2015;49:943–50.
23 Tol JL, Hamilton B, Eirale C, et al. At return to play following hamstring injury the 42 Ekstrand J, Hägglund M, Waldén M. Epidemiology of muscle injuries in professional
majority of professional football players have residual isokinetic deficits. Br J Sports football (soccer). Am J Sports Med 2011;39:1226–32.
Med 2014;48:1364–9. 43 Bahr R, Lian O, Bahr IA. A twofold reduction in the incidence of acute ankle sprains in
24 Cohen J, Cohen P, West SG, et al. Applied multiple regression/correlation analysis for volleyball after the introduction of an injury prevention program: a prospective cohort
the behavioral sciences. Routledge 2013;735. study. Scand J Med Sci Sports 1997;7:172–7.

van Dyk N, et al. Br J Sports Med 2017;0:1–9. doi:10.1136/bjsports-2017-097754 9

 Downloaded from http://bjsm.bmj.com/ on July 30, 2017 - Published by group.bmj.com

A comprehensive strength testing protocol

offers no clinical value in predicting risk of
hamstring injury: a prospective cohort study
of 413 professional football players
Nicol van Dyk, Roald Bahr, Angus F Burnett, Rod Whiteley, Arnhild
Bakken, Andrea Mosler, Abdulaziz Farooq and Erik Witvrouw

Br J Sports Med published online July 29, 2017

Updated information and services can be found at:

http://bjsm.bmj.com/content/early/2017/07/29/bjsports-2017-097754

These include:

References This article cites 42 articles, 15 of which you can access for free at:
http://bjsm.bmj.com/content/early/2017/07/29/bjsports-2017-097754
#BIBL
Email alerting Receive free email alerts when new articles cite this article. Sign up in the
service box at the top right corner of the online article.

Notes

To request permissions go to:

http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to:

http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to:

http://group.bmj.com/subscribe/