TAKING A STEP BACK TO RECONSIDER CHANGE OF DIRECTION AND ITS APPLICATION FOLLOWING ACL INJURY
– Written by Philip Graham-Smith, Qatar, Paul Jones, United Kingdom and Paul Read, Qatar
Research into changing direction has become increasingly frequent due to observations that ‘agility’ performance tests can differentiate between levels of playing ability1,2 and these movements are characterised as high-risk, potentially leading to anterior cruciate ligament (ACL) injuries. For example, rapid changes of direction are cited as a key mechanism in sports such as handball3, soccer4,5, rugby union6 and American football7.
This article revisits some fundamental concepts and provides alternative arguments pertaining to the risks and performance indicators of changing direction. The aim is to provide some clarity around key factors for consideration when developing a framework for enhanced return to play/performance after ACL injury.
What is the ‘Real’ Risk of an ACL injury?
An ACL injury is a catastrophic event which may or may not occur in a player’s career. The actual exposure of a sportsperson to this risk can be estimated based on the number of turns they are likely to make in their playing career. If we make a conservative estimate that soccer players start training and match play at the age of 10 (and it is likely to be younger), and they finish their professional career at the age of 35, this equates to 25 years of exposure. The season typically last for 9 months of the year, or approximately 40 weeks. Players may train or play 5 times per week and perform 40 changes of direction per session (conservative). Baptista et al.8 recently reported players make on average around 40 turns at angles > 90 degrees per game. Bloomfield et al.9 reported 100 turns and Withers et al.10 reported 50. Doing the math, this equates to somewhere in the region of 200,000 turns in a playing career. The majority of players will never get an ACL injury, whereas some may be unfortunate to have two (or more likely a reoccurrence of the same injury). Using this example, is it any wonder that predicting the occurrence of an ACL injury to a specific player has proven difficult? Even if we estimate that 10% of these will be performed at high intensity, we’re still looking at up to a 1 in 20,000 chance of sustaining a knee injury specifically due to changing direction.
What are the underlying causes of an ACL injury?
In mechanical terms, a structure will break or rupture when the force it is exposed to exceeds its failure tolerance. An ACL rupture is likely to occur as a result of a combination of tensile, shear and torsional loading due to concurrent movements of knee flexion, and slide and rotation of the tibia relative to the femoral head whilst the athlete accepts force and rotates about the turning foot.
It is more likely that an ACL will rupture as a result of an abnormal movement with a slightly different loading pattern that occurs maybe once in a career, probably due to a combination of factors which include a higher intensity (speed/deceleration), possibly in a fatigued state or where shoe-surface friction is excessive.
Any attempt to reduce the stresses placed on the ACL should therefore address ways in which the load it is exposed to can be reduced, preferably without any detrimental effect on performance.
Is CoD research really helping us to make the right interventions?
We know that the turning leg is most at risk in the first 17-50ms of contact and between 5-30 degrees of knee flexion11,12. Given the lack of preparation time and limited range of motion in which a player can respond to this loading, it is highly unlikely that looking at discrete body parts in isolation will help to identify ways in which to reduce the ACL load. Despite this, a large propensity of the available research has examined the mechanics of the turning leg in the hope that a golden nugget will appear highlighting that it’s the trunk, hip, knee or ankle position that somehow reduces the external knee abduction moment (a surrogate variable that is often used to quantify ACL risk).
A critical appraisal of the research in this area reveals that the methods are often far removed from reality and therefore the findings become meaningless. The most obvious and heavily debated is assessment of planned versus unplanned tasks. This is a valid discussion, but it has detracted people’s attention away from something much more relevant – a fundamental lack of appreciation and understanding of the horizontal ground reaction force (GRF). In addition, standardisation of footwear and appropriate combinations of shoe-surface interaction are rarely reported. This factor alone can have huge implications on how an athlete orientates their body as they accelerate and decelerate. It could be argued that we have been blinded by technology and an obsession over the use of integrated motion capture systems in laboratories and accepted that the lab surface and the subjects self-selected shoe is a reasonable limitation – we believe it is not! Finally, the limitations, inherent errors and trial to trial variability present when using inverse dynamics to measure ‘net’ joint moments are ignored and these variables are presented as the ‘gold standard’. The ‘net’ moments about the knee are then taken as the surrogate variable that implies injury risk or entered into more sophisticated modelling software to predict load in the ACL.
We must move on from our current practice and perceived limitations, re-evaluate the task and focus on fundamental factors that can be modified or manipulated to improve performance and reduce injury risk.
Re-evaluating the task
First and foremost, we must consider changing direction as a performance attribute. When players change direction in sport, the intention is generally to find or close down space by committing an opponent, or to make/ avoid a tackle. For example, in a sport such as basketball it is often beneficial for the attacking player aiming to take a shot to avoid making contact with an opponent and stay far enough away from arms reach.
If we consider any change in direction that is 90 degrees or greater relative to the initial direction, it is a requirement that the athlete must reduce his/her speed (momentum) momentarily to zero. This is a significant factor and relates to the outcome of being able to perform the movement quickly and avoid being tackled. This is where the horizontal GRF (the friction force) and how the player decelerates is critical. If the player brakes predominantly on his/her final foot plant (the turning leg), he/she will spend more time in closer proximity to the opponent, leading to a greater chance of being tackled. If they can brake harder on the penultimate contact they can react quicker on the turning foot, and more importantly the line of the foot is parallel to the thigh meaning the braking force is taken in a stronger position for the quadriceps to decelerate13,14. In contrast, braking harder on the final contact where the foot placement angle may not be in alignment poses a greater risk of injury. Here we can see that a technical intervention may be a key factor for safer and faster changes of direction.
Breakdown of the movement
To illustrate our point, we have used a 180 degree turn, but the key principles can be applied to all forms of direction change. We have separated the movement into 4 phases; acceleration-in, preparation for turn, turn, and acceleration-out. The total time to perform the task would be the sum of each of the 4 phases.
1. Acceleration – in:
Acceleration is an important metric for heightened team sports performance and has been a recent focus for many in recent years, but it can also be assumed that practitioners are less likely to know what speeds a player can attain in a specific distance? The maximum speed that an athlete can attain prior to changing direction dictates how much braking impulse needs to be imparted. In game scenarios there is no pre-determined ‘approach’ distance, so in order to understand the loading demands we first need to evaluate our athletes’ ability to accelerate and decelerate within set distances. Using a Laveg speed gun, Graham-Smith et al.15 determined the typical speeds players can attain over a range of short distances (relative to their maximum sprint speed). This was achieved using a test where athletes were required to accelerate from specified distances and stop dead on the zero point (Figure 1). Results revealed that 54% of maximum speed is typically attained within the first 2.07m, requiring 2.93m to decelerate (within a total distance of 5m).
Regression equations were then developed to determine what % of maximum speed could be attained with a given acceleration distance, and subsequently what stopping distance would be required to decelerate from that speed (Figure 2). This approach allows practitioners to build a profile of their acceleration and deceleration capabilities. For example, a player who accelerates 10m will attain 87% of their maximum speed and this would typically require around 7.4m to decelerate and stop within a total movement distance of 17.4m. This gives us context for understanding the first part of change of direction performance.
2. Preparation for turn
This phase has two functions, to commence deceleration so that the player arrives at the final foot contact with minimal speed, and secondly to ensure that the final step length is short enough to allow double leg support. Deceleration has both technical and physical qualities attributed to it. In the study mentioned above the deceleration gradient in the acceleration-deceleration test, (denoted by the gradient of changes in mean speed relative to changes in mean stopping distances in the 10m and 5m trials) had moderate associations with combined left and right leg eccentric strength in the knee extensors and flexors, R2 = 0.281 and 0.219 respectively. From a technical perspective, greater braking impulse on the penultimate contact has been shown to relate to faster turns13,14. Positioning the centre of mass further behind the point of contact into the penultimate contact via, rearward inclination of the trunk, the leg planted in front of the body and making contact with the heel will promote greater braking forces (Figure 3). Providing the foot is planted in the same direction of the thigh, and the athlete displays good levels of eccentric strength, this will provide a sound platform to decelerate quickly. However, this strategy should not be promoted if the foot is already rotated into the turn. The length of the last step should also be short enough to permit dual foot support, as this helps to establish a firm base, increased stability and faster re-orientation of the feet prior to accelerating out.
3. The turn
A successful turn will be one which has the lowest contact time on the turning foot (thereby ensuring less time in close proximity to an opponent), and where the body doesn’t travel too far forward. If there is excessive forward trunk rotation the centre of mass travels a greater distance than is necessary and puts more load directly over the turning foot. The trunk, head and upper limbs account for approximately 60% of an athlete’s body weight; therefore, controlling the trunk movement is critical to performance and managing the joint loads at the knee. Rotating the feet when the greatest force is over the foot produces greater torsional friction and risk to the knee. Reducing torsional friction can be achieved by timing the re-orientation of the feet with weight shifts, primarily through the repositioning of the trunk. As the trunk rotates forward in the initial part of contact, the vertical force over the rear leg reduces. This is the time to re-orientate the rear foot. The greatest vertical force over the turning leg is when it reaches maximum knee flexion. Prior to rotating the turning foot, the player should reduce vertical force by repositioning the trunk into the intended direction of movement. As the rear leg is already orientated in the intended direction of travel, the purpose of the turn leg is to generate enough force to overcome inertia and shift the centre of mass ahead of the rear foot (thereby promoting faster application of horizontal propulsive force). Timing foot rotations with weight shift can be taught to help reduce torsional friction due to interaction between the shoe (boot) and surface.
With correct alignment of both the athlete’s feet and legs, their body is now facing the intended direction of travel and the trunk is inclined forwards over the rear foot (now effectively the front foot). All that remains is for the player to accelerate away.
The importance of eccentric strength
Eccentric strength in the quadriceps and hamstrings have shown associations with the ability to change direction quicker13,14,16,17,18 and more specifically to decelerate15,19. Eccentric strength in the quadriceps and hamstrings has an additional benefit as it helps to provide dynamic stability of the knee when subjected to high shear forces, brought about by rapid decelerations. Graham-Smith et al.20 suggested that dynamic stability of the knee musculature could be assessed using the ‘angle of crossover’ from an isokinetic dynamometer. This metric represents the angle where the eccentric torque of the hamstrings is equal to the concentric torque of the quadriceps. Within a cohort of professional football players, the average angle of crossover was 31 degrees (0 being full extension) with a range of 16 to 55 degrees. It was suggested that if the angle of crossover is closer to mid-range then there is a greater ‘safe’ range where hamstrings can resist more than the quadriceps can generate.
Implications for ACL rehabilitation and return to play/performance
The decision to release a player back into competitive situations requires practitioners to have the confidence that the player can withstand the demands of the game.
Many rehabilitation exercises are performed in a vertical direction, for example, double and single leg drop and hold, progressing to reactive drop jumps (double and single legs). While this approach helps the athlete to accept load vertically, we cannot overlook exercises that also build in horizontal braking forces. Hopping is a good progression, but typically the centre of mass is above the foot on landing. This means that eccentric loading is still mainly directed vertically, although we cannot dismiss that within the knee joint itself shear forces will be present.
Whilst slow speed forward, backward and lateral movements should be encouraged to reintroduce the athlete to more game related movements, later stage rehabilitation and return to ‘performance’ must incorporate drills that expose the athlete to greater levels of horizontal braking forces. In addition, it would be prudent to develop eccentric strength in the quadriceps and hamstrings as a pre-requisite to give the athlete confidence in their ability to decelerate.
Deceleration drills should be introduced as the athlete makes good progress in their ability to accelerate and attain speeds of over 7.5m/s.
Using the regression equations in Figure 2, progressive drills can be developed to mark out appropriate acceleration and deceleration distances, gradually reducing the stopping distance. For example, if a 10m acceleration distance was marked out, the practitioner would aim to progressively decrease the stopping distance to ~ 7.4 m, which would be acceptable. Technical interventions could play a vital role in offsetting reinjury, something that until recently21 does not seem to have been promoted in rehabilitation programmes or within the available literature. Emphasis should be on keeping the feet aligned in the direction of travel and gradually encouraging the athlete to lean backwards and plant the leg further in front of the body. The practitioner should also observe to see if the athlete is able to confidently decelerate on both injured and non-injured legs.
Concurrently with deceleration drills, technical interventions can be introduced for changing direction at angles equal to or greater than 90 degrees, but at relatively slow speeds. Technique should focus on a short last step for dual foot support and synchronising feet re-orientation with unloading as a result of weight shift from trunk movement.
Speeds and deceleration into the turn can then gradually increase whilst still adopting confident and ‘safer’ technique.
A word of caution when using total time alone to assess change of direction performance as part of an athletes return to play criteria. Deficiencies in strength and control when turning off either leg can be masked by compensation strategies, leading to similar performance times.
Common sense tells us that the overall performance time is a function of acceleration and deceleration abilities of both limbs. King et al.22 has confirmed this recently and identified that turning off the ACLR leg is potentially less hazardous as a result of stronger braking off the non-injured limb in the penultimate contact. In this regard, if a direct assessment of penultimate and final contact braking forces and impulses is not possible, simply comparing contact times on the turning leg along with a qualitative analysis of technique (as described above) will at least give the practitioner some indication of the athlete’s ability to return to play. For example, if the overall performance times are similar, poor decelerative ability of the ACLR limb in the penultimate contact is likely to equate to a longer contact time on the non-injured limb in the turn (because there will be more speed and load to accept in the final contact). Contact times can be assessed accurately by filming the movement in high speed mode (120 / 240 fps) on iPhone, Casio Exilim or GoPro and counting the number of frames in contact when reviewing through free software such as Quintic Player or various mobile phone applications such as Dartfish Express.
Philip Graham-Smith Ph.D.
Head of Biomechanics & Innovation
Paul Jones Ph.D.
Lecturer in Sports Biomechanics and Strength & Conditioning
University of Salford,
Paul J. Read Ph.D.
Clinical Research Scientist
Aspetar – Orthopaedic and Sports Medicine Hospital
Buttifant D, Graham K, Cross K. Agility and speed in soccer players are two different performance parameters. In: Science and football IV. Eds: Spinks W., Reilly T., Murphy A., editors. London, Routledge 2002; pp 329-332.
Sheppard J. and Young W. Agility Literature Review: Classifications, Training and Testing. Journal of Sports Sciences 2006; 24: 919-32.
Olsen OE, Myklebust G, Engebretse, L, Bahr R. Injury mechanisms for anterior cruciate ligament injuries in team handball: a systematic video analysis. American Journal of Sports Medicine 2004; 32:1002-1012.
Faude O, Junge A, Kindermann W, Dvorak J. Injuries in female soccer players. A prospective study in the German national league. American Journal of Sports Medicine 2005; 33: 1694-1700.
Waldén M, Krosshaug T, Bjørneboe J, Andersen TE, Faul O, Hagglund M. Three distinct mechanisms predominate in non-contact anterior cruciate ligament injuries in male professional football players: a systematic video analysis of 39 cases. British Journal of Sports Medicine 2015; 49: 1452–60.
Montgomery C., Blackburn J, Withers D, Tierney G, Moran C, Simms C. Mechanisms of ACL injury in professional rugby union: a systematic video analysis of 36 cases. British Journal of Sports Medicine 2018; 52; 994-1001.
Johnson JT, Mandelbaum BR, Schub D, Rodeo SA, Matava MJ, Silvers-Granelli HJ, Cole BJ, ElAttrache NS, McAdams TR, Brophy RH. Video analysis of anterior cruciate ligament tears in professional American football athletes. American Journal of Sports Medicine 2018; 46;: 862-868.
Baptista I, Johansen D, Seabra A, Pettersen SA. Position specific player load during match-play in a professional football club. PLoS One 2018; 13: e0198115.
Bloomfield J, Polman R, O'Donoghue P. Physical Demands of Different Positions in FA Premier League Soccer. Journal of Sports Science & Medicine 2007; 6: 63–70.
Withers RT, Maricic Z. Wasilewski S, Kelly L. Match analyses of Australian professional soccer players. Journal of Human Movement Studies 1982; 8, 159-176.
Koga H, Nakamae A, Shima Y, Iwasa J, Myklebust G, Engebretsen L, Bahr R, Krosshaug T. Mechanisms for noncontact anterior cruciate ligament injuries: knee joint kinematics in 10 injury situations from female team handball and basketball. American Journal of Sports Medicine 2010; 38: 2218-25.
Krosshaug T, Nakamae A, Boden BP, Engebretsen L, Smith G, Slauterbeck JR, Hewett TE, Bahr R. Mechanisms of anterior cruciate ligament injury in basketball: video analysis of 39 cases. American Journal of Sports Medicine 2007; 35: 359-67.
Graham-Smith P, Atkinson L, Barlow R, Jones P. Braking characteristics and load distribution in 180 degree turns. Proceedings of the UKSCA annual conference, Wyboston Lake, June 2009.
DosʼSantos T, Thomas C, Jones PA, Comfort P. Mechanical Determinants of Faster Change of Direction Speed Performance in Male Athletes. Journal of Strength and Conditioning Research 2017; 31: 696-705.
Graham-Smith P, Rumpf M, Jones PA. Assessment of deceleration ability and relationship to approach speed and eccentric strength. ISBS Proceedings Archive: 2018; 36: Iss. 1, Article 3.
Jones P, Bampouras T, Marrin K. An investigation into the physical determinants of change of direction speed. Journal of Sports Medicine and Physical Fitness 2009; 49: 97-104.
Jones PA, Thomas C, Dos’Santos T, McMahon JJ, Graham-Smith P. The role of eccentric strength in 180° turns in female soccer players. Sports 2019; 5, 42: 1-11.
Spiteri T, Nimphius S, Hart NH, Specos C, Sheppard JM, Newton RU. Contribution of Strength Characteristics to Change of Direction and Agility Performance in Female Basketball Athletes. Journal of Strength and Conditioning Research 2014; 28: 2415-2423
Naylor J, Greig M. A hierarchical model of factors influencing a battery of agility tests. Journal of Sports Medicine and Physical Fitness 2015; 55: 1329-35.
Graham-Smith P, Jones PA, Comfort P, Munro AG. Assessing knee flexor and extensor muscle balance. International Journal of Athletic Therapy & Training 2013; 18: 1-5.
Dos’Santos T, McBurnie A, Comfort P, Jones P. The effects of six-weeks change of direction speed and technique modification training on cutting performance and movement quality in male youth soccer players. Sports 2019; 7, 205; doi:10.3390/sports7090205
King E, Richter C, Franklyn-Miller A, Daniels K, Wadey R, Jackson M, Moran R, Strike S. Biomechanical but not timed performance asymmetries persist between limbs 9 months after ACL reconstruction during planned and unplanned change of direction. Journal of Biomechanics 2019; 81; 93-103