ACL ruptures are increasingly common in field and court sports and occur 3-6 times more frequently in girls and women than boys and men. However, even after surgical reconstruction and apparently successful rehabilitation, two-thirds of female athletes don’t return to their pre-injury levels of training and competition, a quarter experience a second ACL injury and 50-90% develop knee osteoarthritis in the subsequent 10-15 years.
Given the increasing prevalence and long-term consequences of injury, efforts to reduce ACL injury risk are a top priority for those working within women’s sport.
In field and court sports, most ACL injuries occur without player contact, typically during forceful changes of direction, deceleration, or landing manoeuvres. At the time of injury, it is common to observe dynamic knee valgus (knee collapsing inwards), limited knee flexion, a lateral and rotated trunk position, and flat foot ground contact.
At the knee, peak ACL strain occurs as a result of multi-planar forces, involving an anterior tibial drawer force, knee valgus, and internal tibial rotation moments. Biomechanical factors suggested to be associated with increased risk of a first-time ACL injury include greater knee valgus moments and peak vertical ground reaction forces; however, until recently these factors have been difficult to measure outside of the laboratory.
As women’s sport grows in participation and visibility at the elite level, the disparity in ACL injury rates between women’s and men’s competitions is increasingly apparent.
Historically, biological and anatomical factors have been to blame for the disproportionately higher rates of ACL injuries in females than in males. Although some aspects of biology are likely important – for example, a narrow intercondylar notch may impinge the ACL and smaller ACL volumes may have lower tensile strength – most biological factors such as femur Q-angles and hormonal differences lack evidence.
Strength and biomechanical factors have been widely studied, and women generally display lower muscle strength and greater ACL loading compared to men. Many practitioners misinterpret that females are inherently at high risk because of their sex, however, strength and biomechanics are largely modifiable attributes.
Women generally display lower muscle strength and greater ACL loading compared to men. However, strength and biomechanics are largely modifiable attributes.
Across almost every sport – from grassroots to elite levels of competition – girls and women have reduced access to training facilities, resources (including coaching) and medical care compared with their male counterparts. Moreover, women are underrepresented in sport and exercise medicine research, which limits our understanding of predisposing risk factors and impedes the ability of coaches and practitioners to implement an evidence-based approach to training and rehabilitation.
A meta-analysis of meta-analyses (a meta-meta-analysis?) revealed that ACL injury prevention programs reduce the risk of non-contact ACL injuries by two-thirds in female athletes. In female footballers, specifically, there is evidence that multicomponent “neuromuscular” exercise programs reduce the rates of ACL injury by 45%. However, compliance with these programs is poor and injury rates have shown limited evidence of a decline outside of these controlled studies.
Perhaps training programs that are effective in lower-level athletes are not effective or feasible in the more complex world of elite sport?
For example, less than 1 in 3 of these programs meet the guidelines for strength and 67% meet the guidelines for plyometric/power training. It might also be argued that our improved understanding of injury prevention is failing to keep pace with the increased demands of training and competition.
However, in our recent systematic review (Lima, under review), we demonstrate that ACL injury prevention programs also fail to change lower limb biomechanics during high-risk tasks, which begs the question: what is the mechanism(s) by which they prevent ACL injury?
Understanding the factors that increase an athlete’s risk of injury represents the first step towards developing targeted injury prevention interventions. Until recently, no studies had investigated ACL injury risk factors in elite women’s Australian footballers or soccer players – the former of which appear to have the highest rates of ACL injury of any sport worldwide.
Using a 10-minute field-based testing battery in preseason, we tested 322 elite female footballers’ isometric hip abductor and adductor strength, eccentric knee flexor strength, bilateral countermovement jump (CMJ) kinetics, and single leg triple-vertical hop kinematics using VALD’s ForceFrame, NordBord, ForceDecks and HumanTrak, respectively (Collings et al., 2022).  Non-contact ACL injuries that occurred in the subsequent 18 months were recorded so that we could attempt to identify how players who sustained an injury differed from those who did not.
A. Isometric hip adductor and abductor strength (ForceFrame).
B. Eccentric knee flexor strength during the Nordic hamstring exercise (NordBord).
C. Bilateral CMJ on portable force plates (ForceDecks).
D. Single-leg triple vertical hop kinematics using a markerless motion capture system (HumanTrak).
The results revealed that a number of factors were associated with an increased risk of future non-contact ACL injury, specifically:
You can find normative data for all our test results here.
Players with stronger hip abductors relative to their adductors were at increased risk for future ACL injury (1.97x greater odds of injury for every 14% decrease in the isometric adductor to abductor strength ratio).
The mechanism(s) explaining this association is unclear, however, an imbalance between these two muscle groups may influence coordination of the hip joint during high-risk tasks like jumping and changing direction.
Landing with greater dynamic knee valgus and ipsilateral trunk flexion angles (i.e., towards the stance leg) significantly increased the risk of subsequent ACL injury. Dynamic knee valgus occurs from the combination of hip internal rotation and adduction, knee abduction, and ankle eversion.
Females typically display greater dynamic knee valgus than males during high-risk landing tasks, however, evidence for an association with ACL injury risk is less clear. Ipsilateral trunk flexion increases the knee abduction moment via increasing the moment arm of the trunk relative to the knee joint, which may also increase ACL loads.
These data suggest that strategies targeted at reducing dynamic knee valgus and lateral trunk flexion when landing on one leg may reduce the risk of future ACL injury.
Players who produced greater CMJ peak take-off forces were at significantly greater risk of subsequent ACL injury than those who produced lower forces. The quadriceps and gastrocnemius make a significant contribution to vertical ground reaction forces during jumping, and these muscles are also responsible for generating most of the load on the ACL during landing and changing direction.
Perhaps training to promote strength and hypertrophy of the quadriceps and gastrocnemius to improve jumping performance unknowingly increases the risk of ACL injury?
Despite this possibility, reducing jumping forces to protect against ACL injury risk is obviously a counterproductive approach for players who wish to perform at the highest level. Instead, female athletes who generate high CMJ forces may benefit from interventions targeted at decreasing knee joint loads or developing the muscles capable of “unloading” the ACL (i.e., the hamstrings, gluteals and soleus).
Figure 5. depicts predicted probability of sustaining an ACL injury over a range of strength and biomechanics values. Predicted probabilities derived from univariable (unadjusted) logistic regression models. Horizontal dashed line indicates base risk (5.4%), and the intersection with predicted probability indicates the value at which ACL injury risk increases/decreases relative to base risk. Values corresponding with 2%, 5.4%, 10%, 15%, and 20% predicted probabilities of ACL injury are highlighted on the curve with black dots. Bottom panel displays the distribution of individual ACL injured and uninjured data. Figure replicated from Collings et al 2022. 
When combined in a multivariable clinical prediction model, our field-based screening battery correctly identified almost 4 in 5 of the players who sustained a subsequent ACL injury.
However, the individual variables by themselves were only marginally better at predicting injury than flipping a coin. This is an important point to understand when assessing player injury risk.
Individual variables are poor indicators of injury risk because they are unable to capture the complexity of injury and account for the high variation among individuals. Much more accurate assessments of injury risk requires performing a testing battery and combining important variables into a multivariable model. However, this can only be achieved with large datasets of individuals along with their injury records. Development and validation of such models is an exciting area of research that can then be translated for use in clinical practice.
We found no convincing evidence that between-leg asymmetries in strength and biomechanics were related to ACL injury risk. Subsequently injured and uninjured players both displayed 10-20% asymmetries in lower limb strength, and 10-50% asymmetries in CMJ ground reaction force variables, suggesting that a degree of asymmetry is “normal”, at least in elite female footballers.
Interestingly, we did observe that “non-directional” asymmetry in eccentric knee flexor strength on the NordBord was associated with an increased risk of subsequent ACL injury, however, the stronger leg was injured just as frequently as the weaker leg. Eccentric knee flexor weakness is commonly observed following ACL reconstruction and is likely a consequence of surgery involving hamstring grafts rather than an independent risk factor for first-time injury.
Figure 7. shows between-leg asymmetry in hip strength, knee flexor strength, and CMJ kinetics for players who sustained an ACL injury (ACL injured players) and players who did not (uninjured group). Uninjured players (n < 262) are represented by density curve, with shaded areas indicating increasing amounts of asymmetry. Players who sustained an ACL injury are presented with black dots. Significant variables indicated with a P value. Figure replicated from Collings et al 2022. 
In our study, female footballers with any history of ACL injury were 9.7 times more likely to sustain a subsequent ACL injury than those without. Prior ACL injury has consistently been shown to be the strongest predictor of a subsequent ACL injury in female field and court sports.
Although the mechanisms underpinning high rates of injury recurrence are not fully understood, we have previously shown that female athletes display deficits in lower limb structure and function that persist for many years after injury.
Interestingly, previously injured athletes tend to exhibit a distinct physical profile characterised by between-leg asymmetries in eccentric knee flexor strength, isometric hip abduction strength, and peak landing forces, which distinguishes them from uninjured players with 93% accuracy.
Figure 8. depicts receiver operating characteristic curves for univariable and multivariable logistic regression models summarizing their ability to discriminate between players with and without a history of anterior cruciate ligament reconstruction (ACLR). Figure replicated from Collings et al 2021. 
Our recent systematic review and meta-analysis also revealed chronic deficits in several force-plate derived measures of vertical jump performance lasting well beyond the completion of apparently successful rehabilitation. During bilateral jumping tasks, individuals with a history of ACLR display an offloading strategy whereby they favour their uninjured contralateral limb, likely as a protective mechanism to shield the injured limb from excessive force.
During bilateral jumping tasks, individuals with a history of ACLR display an offloading strategy whereby they favour their uninjured contralateral limb
During unilateral jumping, the ACLR limb displayed a lack of explosive strength with deficits seen across all jump types and force-plate derived metrics. More work is needed to understand if these deficits increase the risk of re-injury, and whether they can be redressed via targeted training.
Figure 9. shows an overview of meta-analyses from Dutaillis et al 2023.  On the left, standardised means differences (SMD; circles) are plotted with 95% confidence intervals (95% CI; error bars) with a negative effect representing a lower value in the anterior cruciate ligament reconstructed (ACLR) limb or individual. Green circles represent comparisons of ACLR limb/individual to uninjured controls. Blue circles represent comparisons of ACLR limb to uninjured contralateral limb. On the right, values for SMD, 95% CI, heterogeneity (I2), number of ACLR (n.injured) and contralateral/control (n.uninjured) limbs/individuals, and references (Ref) for each meta-analysis are reported.
The success of female athletes on the world stage has fuelled a long overdue rise in the professionalism and profile of female sports. However, girls and women still face many barriers to sports participation and high-performance including unacceptably high rates of ACL injuries.
Despite significant research efforts over several decades, ACL injury prevention strategies are still heavily influenced by clinical assumptions and expert opinion rather than empirical evidence. Innovative, field-based human measurement technologies that allow for rapid, reliable and objective measures of strength and biomechanics provide an opportunity to significantly improve clinical decision-making.
Appropriately developed and validated clinical prediction models built using combination of these field-based tests show promise in being able to identify which athletes are at risk of ACL injury, and perhaps one day, might inform the design of personalised risk reduction strategies.
Ultimately, solving the ACL injury crisis will likely require the development and application of robust tools to understand player injury risk and the implementation of targeted training interventions to address the underlying causes of injury (e.g. injury mechanisms and causal risk factors).