Playing surface traction influences movement strategies during a sidestep cutting task in futsal: implications for ankle performance and sprain injury risk

ABSTRACT

This descriptive laboratory crossover trial study examined the intervention of high friction synthetic vs lower friction natural sport surfaces on the ankle joint biomechanics in a sidestep cutting task. Twenty-nine male futsal players performed 5 trials of sidestep cutting task in a laboratory, recorded by an 18-camera motion capture system to obtain the ankle joint orientation, velocity and moment. Utilised friction was obtained by the peak ratio of the horizontal to vertical ground reaction force during the stance. Repeated measures (MANOVA) suggested a significant effect of the playing surface and post hoc paired t-tests revealed significantly higher utilised coefficient of friction, higher peak plantarflexion angle, lower peak eversion angle, higher peak inversion velocity, lower peak inversion moment and higher peak internal rotation moment. In performing a sidestep cutting task, futsal players demonstrated higher utilised ground friction when available friction from the playing surface was higher, resulting in higher peak inversion velocity and higher peak internal rotation moment, which may make the ankle joint more prone to sprain injury. Floorings for futsal should have an adequate coefficient of friction for agility and avoidance of the risk of slipping. Increasing the coefficient of friction may not only enhance performance but also endanger the ankle joint.

KEYWORDS:

• Ankle injury
• sports biomechanics
• sports medicine
• surface friction

Introduction

Futsal, or indoor football/soccer, is now played by 30 million people from more than 100 countries (Federation Internationale de Football Association [FIFA], 2020). The injury incidence is high in futsal and has been reported as 2 times of that in outdoor football (Junge et al., 2004) and even the highest among all sports in a 6-year injury survey conducted in the Netherlands (Schmikli et al., 2009). Like outdoor football, the most commonly injured body site in futsal is the ankle (Emery & Meeuwisse, 2006; López-Segovia & Fernández, 2019), with ligamentous sprain being the most common type of injury, accounting for as much as half (39% to 48.8%) of all injuries (Hamid et al., 2014; J. M. Serrano et al., 2013).

Futsal and outdoor football share many similarities and futsal is commonly perceived as indoor and smaller version of outdoor football (Junge & Dvorak, 2010). In fact, they also differ quite a lot and one major difference is the playing surface, which is critical in both performance and athlete safety aspects (Jastifer et al., 2019). In performance aspect, previous studies have investigated the effect of the playing surface in tennis (Cui et al., 2019), rugby (Ballal et al., 2014), ice hockey (Fortier et al., 2014), American football (Kent et al., 2015), squash (Chapman et al., 1991), dancing (Reeve et al., 2013), basketball and gymnastics (Bressel et al., 2007). In indoor sports, some common types of training and playing surfaces are wooden and synthetic (or artificial) floors (Ismail et al., 2020; Pasanen et al., 2008; C. Serrano et al., 2020), generally offering lower and higher available friction, respectively. In futsal, the playing surface was reported to have an effect on the time required to complete a change-of-direction task (Serrano et al., 2020) and this agility performance has been taken as a key indicator of the performance level of futsal players (Sekulic et al., 2019). A study on the offensive tactics of the 2010 UEFA Futsal Champion also suggested that the playing surface condition was a contributing factor towards the type of shot, leading to victory in the competition (Lapresa et al., 2013).

Whilst higher friction between the shoes and the playing surface is thought to be beneficial to agility required to perform stop and turn movements (Hennig & Sterzing, 2010), it may also increase the injury risk and such was found to be doubled in a systematic review on football (Thomson et al., 2015). Nery et al. (2016) suggested that it was probably due to the high speed, smaller field size and the rubberised and more adherent surfaces used in futsal. The most common injury type in futsal is lateral ankle sprain (Hamid et al., 2014; J. M. Serrano et al., 2013), which was caused by sudden ankle inversion torque and over-strained the lateral ligaments (Wei et al., 2015). The frictional force, or shear force, between the interface of the shoe and playing surface is very critical, as it never passes through the ankle joint centre and thus always generates a twisting torque at the ankle joint (Cong et al., 2014). Whilst this frictional force may be beneficial to performance, it should not be too much or else it may introduce higher risk of ankle sprain injury. In performing an agility manoeuvre, Pedroza et al. (2010) found that the time required improved when the coefficient of friction of the ground increased from 0.3 to 0.5 and did not further improve above 0.5 and they concluded that this amount of friction is adequate for agility in indoor sports.

This study aimed at examining the effect of playing surface friction on the ankle joint biomechanics in male futsal players performing a sidestep cutting task. Two types of playing surfaces, wooden (parquet) and synthetic (artificial), offering lower and higher coefficients of friction were tested (Pasanen et al., 2008). We hypothesised that the synthetic surface that offered higher available friction will result in higher utilised friction, higher peak joint moments and higher peak joint kinematics in the inversion direction.

Methods

Study design

This was a descriptive laboratory study based on a blinded randomised controlled crossover trial where the intervention was comparing the effect of high friction synthetic vs lower friction natural sport surfaces on ankle biomechanics in a group of male futsal players. The randomisation was done by throwing a dice: odd > wood first and even > synthetic first.

Participants

Peak ankle inversion was selected to be the primary variable for sample size estimation, which was done in G*Power software (Germany), based on a previous study, which reported that the peak ankle inversion for experienced tennis players to perform a turning movement on acrylic and clay court is 14.3 ± 10.1 and 8.6 ± 4.4, respectively (Starbuck et al., 2016). By setting the level of significance to 0.05 and the statistical power to 0.80 in a two-tailed test on matched pairs, the estimated required sample size was calculated to be 21. Twenty-nine male futsal players (age 23.4 ± 2.0 years, height 179.1 ± 7.8 cm and body mass 82.1 ± 15.3 kg) free from lower extremity injury in the past six months participated in this study. All participants fitted identical futsal shoes during the test (Mizuno Monarcida Sala, model Q1GA1611).

They were requested to not perform any vigorous exercise nor consume any alcohol 48 hours prior to the experiment and give written inform consent before the study commenced. The University Ethics Committee approved the study.

Experimental protocol

Each participant started with a 10-minute warm-up to get familiarised with the experimental set-up and to achieve the fastest possible comfortable speed to perform the sidestep cutting task, which composed of a 6-metre straight run-up, a step on a 120 × 60 cm target area with the right foot and then a cut and a run to the forward-left direction angled at 30 degrees (second arrow from the left, Figure 1 and Kaila, 2007). Two commonly used indoor sport flooring conditions were used in a randomised order, including one synthetic flooring that provided a slip-resistant high friction surface (Taraflex® Performance, M Vinyl Sports Flooring, UK) and one wooden parquet flooring that provided a low friction surface (Junckers, Portable A3, Denmark). The available friction, as represented by the dynamic coefficient of friction (DCOF) measured by dragging a 15-kg-weighted shoe ten times over the flooring on top of a force plate (Kistler 9287BA, Switzerland), and graphical presentation of the experimental set-up are available in the study by Fong et al. (2005, 2008, 2009). Then, introducing the values into de equation, we obtain COF = (sqrt (Fx^2 + Fy^2))/sqrt (Fz^2), providing the calculated values of 0.80 ± 0.20 (unit-less, synthetic) and 0.53 ± 0.09 (unit-less, wooden), respectively. Set-up guidelines for the Kistler 9287BA user manual were followed to correctly calculate the centre of pressure when a flooring was added on top of the force plate and accounted for the changes in the depth of the origin.

Figure 1. Runway with force plates installed beneath.

Data collection

Sixteen reflective markers (14 mm diameter) were attached with double-sided adhesive tape bilaterally on the lateral and medial malleolus, talus, 1st and 5th metatarsalphalangeal joints, 1st metatarsal head and lateral and medial femoral condyles. Each participant performed sidestep cutting task until five successful trials were recorded. The marker trajectories were captured using an 18-camera motion capture system (Vicon, Oxford, UK) at a frame rate of 250 Hz, and were labelled and processed using the in-built software (Vicon Nexus 2.7.0, Oxford, UK) to obtain the ankle joint kinematics. A force plate was embedded underneath the target area for the sidestep to record the ground reaction force (GRF) at a frame rate of 2000 Hz. With the kinematics and kinetics data, together with the essential anthropometric data measured by tape and caliper and obtained from a previous study (Plagenhoef et al., 1983), the ankle joint forces and moments were calculated using a custom-written programme (Vicon Bodybuilder, Oxford, UK) following the inverse dynamics method proposed by a previous study (Vaughan et al., 1992). The utilised coefficient of friction (COF_u) was obtained by the peak ratio of the horizontal to vertical ground reaction force (Fong, Hong, & Li, 2009) during the entire stance phase. Peak joint velocity and moments occurred during the stance phase when the participant made initial contact at the centre of the force plate and pivoted in the forward-left direction angled at 30 degrees. To reduce noise, data were filtered with a 4th order low pass Butterworth filter using a cut-off frequency of 6 Hz. GRF was normalised to the participant bodyweight.

Statistical analysis

The dependent variables were the utilised coefficient of friction and the peak joint orientation, velocity and moment at the ankle joint in plantarflexion/dorsiflexion, inversion/eversion and internal/external rotation planes. The Shapiro-Wilk test was performed to check normality. If normality was achieved (e.g., p > 0.05), one-way multivariate analysis of variance (MANOVA) with repeated measures was applied to show any significant effect of the flooring condition. If a significant effect was found, post hoc paired t-tests were performed on each independent variable. If normality was not achieved, the parametric procedures were replaced by the corresponding non-parametric procedures, e.g., Friedman MANOVA and Mann-Whitney U test. The statistical significance level was set at p < 0.05 level. All statistical analysis procedures were performed using statistical analysis software (SPSS 16.0, IBM Corporation, Armonk, NY, USA).

Results

The Shapiro-Wilk test suggested that normality was achieved and therefore, parametric statistical analysis procedures were performed. MANOVA with repeated measures showed a significant effect of the flooring condition (Wilks’ Lambda = 0.051, F(19, 10) = 9.764, p < 0.001). Significant differences in UCoF values were obtained when performing the SSCM over the two different surfaces. The test disclosed a significantly higher UCoF on the high friction surface with respect to the lower friction surface.

Angular kinematics of the right ankle demonstrated that there were significant differences between the two tested surface conditions on plantar flexion and eversion peak angles. Furthermore, there was a significant increase in inversion angular velocity from the lower friction surface (1273.6 ± 466.2º/s) to higher friction flooring (1439.7 ± 679.0º/s). All other kinematic variables did not display any further significant differences.

Kinetic data analysis identified a significant reduction in peak ankle inversion moment from the low friction surface (42.3 ± 17.4 Nm) to higher friction flooring (23.0 ± 11.1 Nm). In addition, there was a significant increase found in peak internal rotation moment from the low friction surface (72.4 ± 30.7 Nm) to the higher friction surface (109.6 ± 43.6 Nm). However, there were no additional significant kinetic differences identified.

Descriptive statistics for utilised friction and the peak ankle joint orientation, velocity and moment for the two flooring conditions with the result of the paired t-tests are available in Table 1.

Table 1. Utilised friction and the peak ankle joint orientation, velocity and moment for the two flooring conditions

	Synthetic flooring	Wooden parquet flooring	p-value
Utilised friction
- Utilised coefficient of friction (COFu, unit-less)	0.57 ± 0.09	0.36 ± 0.05	<0.001*
Peak ankle joint orientation
- Plantarflexion (degree)	53.3 ± 7.6	48.1 ± 7.1	<0.001*
- Dorsiflexion (degree)	0.8 ± 9.2	0.1 ± 8.1	0.349
- Inversion (degree)	28.8 ± 7.3	28.6 ± 6.6	0.776
- Eversion (degree)	4.2 ± 7.1	6.7 ± 7.3	0.003*
- Internal rotation (degree)	10.6 ± 16.4	14.4 ± 4.9	0.224
- External rotation (degree)	7.3 ± 15.7	3.1 ± 6.3	0.156
Peak ankle joint velocity
- Plantarflexion velocity (degree/s)	1353.3 ± 544.3	1305.2 ± 480.1	0.398
- Dorsiflexion velocity (degree/s)	1890.2 ± 1344.9	1848.9 ± 1439.1	0.907
- Inversion velocity (degree/s)	1439.7 ± 679.0	1273.6 ± 466.2	0.047*
- Eversion velocity (degree/s)	800.0 ± 251.9	938.9 ± 468.1	0.142
- Internal rotation velocity (degree/s)	762.1 ± 365.3	707.4 ± 279.4	0.276
- External rotation velocity (degree/s)	1151.6 ± 418.5	1124.3 ± 442.8	0.710
Peak ankle joint moment
- Plantarflexion moment (Nm)	5.1 ± 13.6	10.7 ± 16.5	0.138
- Dorsiflexion moment (Nm)	207.8 ± 43.3	194.8 ± 73.9	0.297
- Inversion moment (Nm)	23.0 ± 11.1	42.3 ± 17.4	<0.001*
- Eversion moment (Nm)	17.9 ± 21.6	15.3 ± 18.6	0.521
- Internal rotation moment (Nm)	109.6 ± 43.6	72.4 ± 30.7	<0.001*
- External rotation moment (Nm)	6.8 ± 8.8	6.4 ± 6.3	0.841

*Significant difference, p < 0.05.

Discussion and implications

Peak coefficient of friction

The proposed hypothesis that higher COFu of the shoe-surface interaction enhances the ankle performance strategy during a sidestep cutting task is further supported (Dowling et al., 2010). Previous research has been made on indoor surfaces such as synthetic materials, where the minimum traction to optimally execute a rapid directional change was set to a COFu of 0.6 (Keshvari & Senner, 2015). This magnitude is in accordance with the current study, suggesting that polyurethane flooring offers enough traction to perform the pivoting action whilst dissipating the risk factor of foot fixation. Morio et al. (2017) raised the threshold to 0.7, stating it as the minimum amount of COFu at which grip perception allows enough confidence to the individual to perform the cutting tasks more safely. On the other hand, the natural flooring, as expected, offered a lower COFu of 0.4, which does not provide much slip resistance, suggesting that it could decrease the performance of explosive pivoting actions during gameplay due to an accidental slip (Pasanen et al., 2008), which, in the worst of the cases, could cause a ground contact injury.

A study by Worobets and Wannop (2015), which investigated the effects of traction on performance, determined that the higher the friction offered, the greater the level of performance attained during agility movements. Whilst a 20% reduction in traction is enough to significantly decrease the performance of ballistic manoeuvres including sidestep cutting task, the current study results demonstrated a reduction of 21% on COFu from the synthetic surface to natural wood parquet. This would suggest that the meaningfully lower COFu on the natural surface led to a series of significant biomechanical alterations (Vidal et al., 2019) on ankle movement strategies, which may result in a decrease in performance of the futsal players (Dowling et al., 2010).

Peak inversion moment

Contrary to expectation hypothesis, significantly higher peak ankle inversion moments were found in the natural flooring when compared to synthetic. This unexpected outcome could possibly be explained by the low COFu reported from the parquet flooring, where unnoticed sliding might have increased the inversion moment, which consequently entails the risk of exceeding the physiological limits of the ankle ligaments (Kristianslund et al., 2011). This possible lack of proprioception from the participants on a lower friction surface could have resulted in a change in ankle kinetics and consequently the possibility of altering the intended ankle motion technique when executing the pivoting tasks.

Previous literature has found excessive ankle joint moments to induce major affliction to the ligaments, leading to ankle sprains (Lee et al., 2000). The ankle inversion moment obtained on the high friction surface matches 23 Nm found in the study by Wei et al. (2015). In their study, this quantity was reflected as the highest inversion moment before exposure to development of grade I ankle sprain. Surprisingly, this suggests that with the natural playing surface, with a lesser shoe-surface traction, the ankle joint could be excessively exposed to a major ankle inversion moment as the mean recorded was 42 Nm, almost doubling the considered maximum safety limit. However, a possible explanation for this discrepancy in results may be found on the methodological approach and the different surfaces employed, suggesting that friction does have an impact on the performance of ballistic movements. Nonetheless, these are conjectures as there is some conflict within the available literature as different ranges of ankle inversion moment thresholds have been proposed as safe. For instance, Begeman et al. (1993) established 35 Nm as a maximal ankle inversion moment safety during dynamic loading conditions, whilst Lee et al. (2000) reported, without any ankle incidence, the inversion moment during cutting tasks to be 80 Nm. The unforeseen result may indicate that the group of young futsal players who performed the cutting task on this study could have been exposed to the development of ankle ligament trauma due to excessive ankle inversion, suggesting that wooden flooring may not be as safe to the ankle joint as it was thought.

Peak inversion velocity

Ankle peak inversion angular velocity was significantly higher over the synthetic surface when compared to natural. This may indicate that the ankle joint encountered a greater explosive peak inversion moment and subsequent brusque kinematic changes (Mok et al., 2011). Sudden inversion and internal rotation of the ankle joint with low levels of plantarflexion are considered as ankle sprain injury mechanisms (Panagiotakis et al., 2017). The joint would have to resist during violent dynamic movements a large magnitude of moments in a very brief period of time (Mok et al., 2011).

Previous research by Fong, Hong, Shima et al. (2009) suggested that an inversion velocity over 632º/s could potentially increase the likelihood of injuring the ankle joint as their case study demonstrated. Additionally, Kristianslund et al. (2011) found an even lower inversion velocity of 559º/s in a lateral ankle sprain case where the subject performed a sidestep cutting task. However, these values are the concurrent findings, as a peak inversion angular velocity of 1440º/s was recorded without any sign of musculoskeletal incidence. Discrepancies between studies could be attributed to the different types of sports footwear used, the sport being investigated, methodological variations and participant gender. A study by Mok et al. (2011) found in two injury cases that an angular velocity of 1397º/s could be negative for the health of the joint. This could suggest that the higher friction flooring exposes the joint to injury development, denoting that the natural wooden surface, which was found to be 1274º/s, could be safer for the practice of futsal dynamic movements. Ankle angular velocities have not been extensively examined in the futsal literature, indicating that these high velocities may be due to the nature of their fast-game, where futsal players have adapted to attain greater ankle angular velocities as a normal sporting motion.

A study by Girard et al. (2007) suggested that surfaces that provide a high coefficient of friction allow players to undergo greater accelerations and decelerations when compared to lower friction floorings, potentially resulting in a higher speed game, a key characteristic of futsal, subsequently indicating that it may increase the performance of sidestep cutting task and speed up the team gameplay dynamics.

Peak internal rotation moment

Larger ankle internal rotation moments could facilitate rotating the upper body and opposite limb externally (Lee et al., 2000; Mok et al., 2011), suggesting a higher level of pivoting performance during gameplay. Synthetic floorings might be more suitable for sudden changes in the direction as it reported a significantly higher peak ankle internal rotation moment when compared to the natural lower friction surface, which would benefit supination motion. However, previous studies have denoted the important role of internal rotation in the mechanism of ankle supination sprain (Kristianslund et al., 2011; Mok et al., 2011; Wei et al., 2015). These two significant ankle moments were the result of the difference in friction offered by the surfaces, perhaps not strategic but involuntary/unnoticed sliding coping counteractions from the joint (Nigg et al., 2009). Excessive internal rotation in combination with inversion and plantarflexion motion has been demonstrated to raise the anterior talofibular ligament (ATFL) strain to over 20% (Wei et al., 2015), being detrimental for the ankle health, as it could vigorously stretch the ligaments fibres and tear them, injuring the joint (Panagiotakis et al., 2017).

There was a high discrepancy between the results obtained by Wei et al. (2015) and this study as they reported 11 Nm to be one of the highest internal rotation moments, whilst the highest in the current study was 109 Nm on the synthetic surface, however, without any reported incidences. A possible explanation for this discrepancy could be attributed to the different methodological techniques as the study by Wei et al. (2015) was based on a computer model simulation, whereas the current study was based on physical laboratory testing, where limitations such as an error from human fallibility or equipment miscalibration could have caused an overestimation of joint moment calculations. The results suggest that the tested synthetic surface may not be as safe as the natural for the practice of sports as peak internal rotation moments were recorded.

Coaching implications

As coaching implications, training over synthetic floorings might be more suitable for sudden changes in the direction as it reported a significantly higher ankle internal rotation moment when compared to the natural lower friction surface, which would benefit supination motion. Additionally, coaches and athletes are suggested to implement proprioceptive and balance training programmes to their exercise routines on these common types of playing surfaces, which will aid the performance of the demanding and dynamic movements, as a better proprioceptive interpretation of the surface traction would allow the athlete to adapt faster to new unfamiliar surfaces when playing away.

Conclusion

This investigation compared two different indoor playing surfaces. The synthetic surface had a higher available ground friction than a wooden parquet playing surface. Futsal players are exploiting higher utilised ground friction of the surface, allowing peak inversion velocity and peak internal rotation moment during a sidestep cutting task on a synthetic playing surface. These changes may generate higher twisting torque at the ankle joint, which may influence risk and severity of ankle sprain injury.

Disclosure statement

No potential conflict of interest was reported by the author(s).