Secondary Logo

Journal Logo

A 2-yr Biomechanically Informed ACL Injury Prevention Training Intervention in Female Field Hockey Players

Weir, Gillian1,2; Alderson, Jacqueline A.2,3; Elliott, Bruce C.2; Lee, Shina2; Devaprakash, Daniel4; Starre, Kate2; Goodman, Carmel5; Cooke, Jennifer6; Rechichi, Claire6; Armstrong, Josh2; Jackson, Ben2; Donnelly, Cyril J.2

Translational Journal of the American College of Sports Medicine: October 1, 2019 - Volume 4 - Issue 19 - p 206–214
doi: 10.1249/TJX.0000000000000105
Original Investigation

Purpose Anterior cruciate ligament (ACL) injury prevention programs have been shown to have mixed success in reducing injury rates, raising the question whether these programs are effectively targeting biomechanical mechanisms of injury. The current study examined the efficacy of a biomechanically informed ACL injury prevention training program in reducing injury risk and injury incidence and investigated its effect on athletic performance.

Participants Twenty-six elite female field hockey players participated in this study.

Methods Athletes participated in a 2-yr injury prevention training program. Injury incidence (i.e., lower limb and ACL) and athletic performance (i.e., strength, speed, and aerobic power) were measured during a control season and after two intervention seasons. Biomechanical ACL injury risk factors were recorded during unplanned sidestepping at baseline and after intensive (9 wk: 4 × 20 min·wk−1) and maintenance (16 wk: 3 × 10 min·wk−1) training phases for a subset of athletes (n = 17).

Results Training was effective in reducing ACL and lower limb injury incidence after the 2-yr program, where zero ACL injuries occurred after implementation (vs 0.4 per 1000 player hours in the control year). High-risk athletes reduced their peak knee valgus moments by 30% (P = 0.045) and demonstrated improvements in desirable muscle activation strategies after intensive training. The majority of benefits elicited in intensive training were retained during the maintenance phase. One-repetition max strength, beep test scores, and sprint times improved or were maintained over the 2-yr intervention period.

Conclusions Biomechanically informed injury prevention training was successful in reducing both biomechanical ACL injury risk factors and ACL injury incidence while maintaining and/or improving athletic performance. It is important to consider the biomechanical mechanisms of injury when designing injury prevention programs.

1Biomechanics Laboratory, University of Massachusetts Amherst, Amherst, MA

2School of Human Sciences, University of Western Australia, Crawley, AUSTRALIA

3Sports Performance Research Institute New Zealand (SPRINZ), Auckland University of Technology, Auckland, NEW ZEALAND

4 School of Allied Health Sciences, Menzies Health Institute, Griffith University, Gold Coast, AUSTRALIA

5Western Australian Institute of Sport, Mount Claremont, AUSTRALIA

6Australian Institute of Sport, Bruce, AUSTRALIA

Address for correspondence: Gillian Weir, B.Sc. (Hons), Ph.D., Biomechanics Laboratory, University of Massachusetts Amherst, Room 23A, Totman Building, 30 Eastman Lane, Amherst, MA 01003-9258 (E-mail:

Back to Top | Article Outline


It has been well documented that female athletes participating in team sports suffer anterior cruciate ligament (ACL) injuries at a four- to sixfold higher rate than their male counterparts (1–4). In addition, females competing at a higher level of competition have been shown to display significantly higher peak knee valgus (PKV) moments when compared with novice athletes (5). These injuries place a high financial and lifestyle burden on the athlete. Costs associated with surgery are upward of US$5000 (6), and these injuries remove athletes from competition for 12 months, with only 70% able to return to the same level of competition (7). Injury prevention training programs may present a cost-effective solution for this global problem. Research has investigated the effect of training programs such as balance, plyometric, resistance, flexibility, and/or a combination thereof in injury prevention training programs (8,9). However, these interventions have had mixed success in reducing ACL injury rates among general athletic populations (9,10). As some of these programs failed to measure the biomechanical mechanisms of injury in parallel with injury rates, conclusions cannot be drawn upon whether these training programs adequately targeted them. In addition, although some injury prevention training research has been successful in reducing ACL injury risk/rates in the short term (4–12 wk) (11–14), there is little to no evidence for the retention of these initial benefits in the longer term (15). Further, there are no guidelines outlining optimal intensity and duration levels for injury prevention training to be successful in reducing injury rates over a season of play. As such, research investigating the effect of a maintenance training phase after the implementation of an initial targeted injury prevention program is warranted.

Noncontact ACL injuries occur when the knee is in an extended posture and combined, externally applied flexion, valgus, and internal rotation moments are applied to the knee joint (16–18). In silico and in vivo studies have shown that the ACL is at greatest risk of injury during the weight acceptance (WA) phase of stance, where valgus knee moments are the greatest (19,20). Two biomechanical strategies an athlete can adapt to counter elevated peak knee moments in an effort to reduce ACL injury risk are as follows: 1) modify their technique to reduce the peak external forces applied to the knee (19,21,22) and 2) improve the muscular strength and activation to support the knee joint when these forces are elevated (23,24). This suggests that interventions should focus on the specific countermeasures of biomechanical ACL injury mechanisms when designing injury prevention interventions rather than modality alone. A meta-analysis of injury prevention programs (25) has shown that successful programs such as the Sportsmetrics, Prevent Injury and Enhance Performance, and Knee Injury Prevention Program (KIPP) have all contained plyometric and strength components, which targeted dynamic control of the hip. In addition, the Sportsmetrics and the KIPP programs both involved specific landing technique training. These components in combination may be the reason for the success of these programs.

This study aimed to verify the efficacy of a novel, biomechanically informed injury prevention program in reducing ACL injury incidence and ACL injury risk while maintaining athletic performance among elite female field hockey players. Specifically, we tested the hypotheses that; 1) ACL injury incidence would reduce after two intervention seasons when compared with a control season; 2) peak knee moments known to elevate ACL strain (18,20,26) (i.e., surrogate measure for injury risk) would be reduced, and technique and lower limb muscle activation strategies known to mitigate ACL injury risk would improve after a 9-wk initial intensive training phase (4 × 20 min sessions per week); 3) these improvements would be retained when measured after a 16-wk maintenance training phase (3 × 10 min sessions per week) of the training program; and 4) any changes brought about by injury prevention training would not have a detrimental effect on overall athletic performance.

Back to Top | Article Outline


Experimental Approach to the Problem

The biomechanically informed injury prevention training program was assessed over three consecutive seasons (control, intervention season 1, and intervention season 2) among the Australian national women’s field hockey team (Hockeyroos) (Fig. 1). The first season (2012–2013) was treated as the baseline/control season, then a 9-wk intensive training phase was implemented and immediately followed by a 16-wk maintenance training phase in the first intervention season (2013–2014), which continued through a second intervention season (2014–2015). Injury incidence was measured across all three seasons. Biomechanical injury risk factors (i.e., peak knee moments) and countermeasures (i.e., full-body kinematics and muscle activation) were assessed on three occasions: at the end of the control season and again at the completion of both phases of the training program within the first intervention season. Athletic performance was assessed on four occasions: 1) at the end of the control season, 2) after the intensive training, 3) after the first intervention season, and 4) after the second intervention season.

Figure 1

Figure 1

Back to Top | Article Outline


Twenty-six elite female hockey players (mean ± SD: age, 22.1 ± 2.3 yr; height, 1.68 ± 0.09 m; mass, 63.30 ± 7.00 kg) participated in this study. Because of retirement, player availability, and injury, not all players completed all biomechanical and performance testing sessions and/or in each seasons injury incidence measurements. Athletes with prior ACL injury (n = 3) who were completely rehabilitated were included in the study. A power analysis from previous research evaluating changes in PKV moments (27,28) and total muscle activation (TMA) (29) revealed that for 80% power with the alpha set to 0.05, a minimum of 14 subjects were required. As this was a sample of convenience and a small study on the scale of prospective injury rates, power analyses were not performed for injury incidence. Ethics approval for this study was granted from the Human Research Ethics Committee. Participants were informed of the benefits and risks of the investigation before signing an institutionally approved informed consent document to participate in the study.

Back to Top | Article Outline


During the intervention seasons, all athletes participated in injury prevention training sessions adjunct to their regular in-season warm-up and gym sessions, which were delivered by the team strength and conditioning coaches. A high coach to athlete ratio of 1:13 was implemented in attempts to maximize athlete adherence and compliance to the training protocol (30). Irrespective of the exercise genre (resistance, balance, plyometric, and technique), the overriding goal or focus of the intervention was to target four key biomechanical factors associated with ACL injury risk and/or incidence, which were as follows: 1) to increase knee flexion angle at foot strike (31,32), 2) to improve the dynamic control of the trunk and upper body (30,33,34), 3) to strengthen the hip external rotators to prevent athletes from attaining “dynamic knee valgus” postures (26), and 4) to increase the strength of the gastrocnemius muscle group (24). After review, all body-weight exercises that targeted these biomechanical factors were programmed by modality (plyometric, balance, and resistance) and intensity and delivered to the team strength and conditioning coach (Supplemental Table 1, Autonomy was then given to the strength and conditioning coach to design the intervention to best fit their cohort (i.e., elite) and training day per stimulus. During the first intervention season, a 25-wk training program was implemented and split into two phases: 1) intensive training (weeks 1–9) and 2) maintenance training (weeks 9–25). The intensive training phase consisted of 4 × 20-min sessions per week, which progressed in intensity every 2 wk. Although intensity and type of exercise remained the same, only training duration was reduced in the maintenance training phase (3 × 10-min sessions per week). The maintenance training phase was then continued for the remainder of intervention season 1 and throughout the second intervention season. During each session of the intensive training phase, attendance (present or not present) and coach ratings of compliance and athlete engagement (commitment, motivation, and perseverance) were measured on a 5-point Likert scale (35).

Back to Top | Article Outline

Injury Incidence

All lower limb injuries occurring during preseason, in-season training, and competitive games over the control (2012–2013) and intervention seasons (intervention season 1: 2013–2014; intervention season 2: 2014–2015) were collected by the same team doctor and physiotherapist using the Orchard Sport Injury Classification System (36). Total lower limb injuries (all injuries sustained to the lower limbs excluding contusions), total knee injuries (ligament, tendon, and cartilage), total knee ligament injuries, and ACL injuries were recorded. All injuries were verified by either the team doctor or physiotherapist and were defined as an event that caused a player to cease training/playing and seek medical attention. Injury incidence per 1000 player hours was calculated by dividing the number of injuries by exposure (number of athletes × number of hours of training / games per season) and multiplied by 1000 (equation 1).

To evaluate differences in injury rates per season, injuries were first calculated as the proportion of players who experienced an injury in each category. Then expected injury rates were calculated for each season as the same proportion of the total number of injured players over all 3 yr as that season’s exposure hours were of the total hours (37) (equation 2).

Back to Top | Article Outline

Biomechanical Testing

Three-dimensional (3D) motion analysis of each participant completing a previously published unplanned sidestepping protocol (30) was recorded at baseline, postintensive training, and postmaintenance training. Briefly, this protocol involved athletes responding to a 2 × 2-m projected arrow stimulus placed 10 m in front of the force platform. The protocol displayed a random series of planned and unplanned straight line run, crossover, and sidestep tasks. All tasks other than unanticipated sidestepping were used as “catch” trials to limit the predictability of the unanticipated sidestepping condition. Three successful trials of unplanned sidestepping were recorded. A trial was deemed successful if the athlete had an approach velocity of 4.5% ± 5%, and they changed direction along a line marked at 45° to the global x-axis of the laboratory. Previous research using this protocol has reported post hoc change of direction angles measured from center of mass motion to be 29.8° ± 5.1° (38). All sidestepping tasks were performed using the athletes’ self-selected dominant limb, which was consistent across testing sessions.

Participants were fitted with 42 retroreflective markers as per a customized trunk and lower limb kinematic marker set and model (30,39). Markers were affixed to the seventh cervical vertebrae, the clavicular notch, the xiphoid process, and the 10th thoracic vertebrae to define the trunk segment. The pelvis was defined by markers placed on the left and right anterior and posterior superior iliac spines. The ankle joint center was defined by anatomical markers placed on the medial and lateral malleoli. A six-marker pointer was used to digitize the medial and lateral femoral condyles, along with a functional knee axis to define the knee joint center and axes orientation. A functional method was also used to define the hip joint center (39). Four-marker semirigid clusters affixed to the thigh and lower leg were used to track knee and ankle joint centers. Marker trajectories were recorded using a 22-camera Vicon® motion analysis system (12 Vicon® MX and 10 Vicon® T40 cameras; Oxford Metrics, Oxford, UK) operating at 250 Hz. Cameras were synchronized with a 1.2 × 1.2-m force plate (AMTI, Watertown, MA) recording at 2000 Hz. These data, with a reliable full-body customized model fully compliant with International Society of Biomechanics (ISB) standards for the reporting of data (40), were used to calculate full-body kinematics and peak knee moments via inverse dynamics procedures in Vicon® Bodybuilder software through the Vicon® Nexus software pipeline (Vicon, Oxford Metrics).

Activation of nine muscles was measured with surface electromyography (sEMG) using a 16-channel telemetry system (Telemyo2400 G2, Noraxon, Scottsdale, Arizona) at 1500 Hz. These included the gluteus maximus, gluteus medius, semimembranosus (SM), biceps femoris (BF), vastus lateralis, vastus medialis, rectus femoris, medial gastrocnemius, and lateral gastrocnemius. The signal processing of these signals included, first, removing direct current offsets from the signal, then band-pass filtering between 30 and 500 Hz with a zero-lag, fourth-order Butterworth digital filter, full-wave rectification, and then a linearly envelop using a low-pass with a zero-lag, fourth-order Butterworth filter at 6 Hz (41). Muscle activation was amplitude normalized to the maximal activation observed for each muscle during a single-leg squat, a single-leg countermovement jump, or a sidestepping trial and expressed as 0%–100% maximal voluntary contraction (41). Muscle activation patterns were assessed using mean TMA and directed cocontraction ratios (DCCR) as per Heiden et al. (42). Mean TMA of the gluteal, quadriceps, hamstrings, and gastrocnemius muscle groups were calculated, as well as for all muscles crossing the knee. DCCR were calculated for flexion/extension muscle groups, medial/lateral (M/L) muscle groups, and the SM/BF muscles. All muscle activation variables were measured during precontact (PC) (50 ms before foot strike) and WA (defined from >10 N to the first trough in the unfiltered ground reaction force).

The mean values of three unplanned sidestepping trials for each participant were used in analyses. Approach velocity and change of direction angles were measured to ensure task completion did not differ across testing sessions. Externally applied peak knee flexion, PKV, and peak knee internal rotation (PKIR) moments (normalized to body mass and height) were calculated during WA alongside trunk flexion range of motion, peak trunk lateral flexion, peak hip abduction, knee flexion range of motion, and mean knee flexion angles. Knee flexion angle and foot to center of mass displacement were measured at foot strike.

Back to Top | Article Outline

Athletic Performance Testing

Athletic performance measures were recorded during the control season and after intensive training, intervention season 1, and intervention season 2. Strength performance tests included one-repetition maximum (1RM) strength normalized to body mass for the bench press, bench pull, and back squat (43). Speed was assessed with 40-m sprint times (split at 10 and 40 m), and aerobic power was assessed using the beep test (44).

Back to Top | Article Outline

Statistical Analysis

Chi-square analysis (α = 0.05) was used to assess the difference between observed and expected injuries between the control season, intervention season 1, and intervention season 2. All performance and biomechanical-dependent variables were analyzed across time according to the intention-to-treat principle using a linear mixed model. After the intensive training phase of the intervention, a “responder analysis” was performed to determine which athletes had a positive response to training (i.e., reduction in PKV moment). A responder athlete was defined as one who returned a moderate to large effect size (d > 0.5) reduction in PKV moments (n = 5) and a nonresponder as an athlete who did not display a reduction in PKV moments (n = 11) after the intensive phase of the training intervention. Participants were entered as random factors, and an autoregressive (1) covariance structure was used to model the repeated (i.e., time) and subject variables (45). Time (baseline, postintensive training, and postmaintenance training) and response (responder and nonresponder) were input as fixed factors. Least significant difference post hoc analysis was used to assess for significant main effects and interactions. All statistical analyses were conducted in SPSS (IBM SPSS Statistics 22; SPSS Inc., Chicago, IL), with an α = 0.05. Hedge’s g effect sizes, which account for differences in sample sizes, were calculated between and within groups at baseline, after intensive training, and after maintenance training and interpreted as small (g = 0.2), moderate (g = 0.5), and large (g = 0.8) (46).

Back to Top | Article Outline


Attendance and compliance were 81% ± 25% and 88% ± 20%, respectively, with attendance only missed due to injury as advised by team medical staff. Athlete engagement was high with 89% ± 12% athlete commitment, 90% ± 11% motivation, and 92% ± 10% perseverance.

Back to Top | Article Outline

Injury Incidence

Exposure increased after the control season (6749 h), for intervention season 1 (7609 h) and intervention season 2 (7143 h), with a total of 21,501 h over all three seasons. Total knee injury incidence increased in intervention season 1 (2013–2014) but was reduced after intervention season 2 (2014–2015). Of most significance, total lower limb, knee ligament, and ACL injury incidence were reduced after the implementation of the training program after intervention season 1 and were further reduced after intervention season 2 (Table 1). Zero ACL injuries were observed after the implementation of the injury prevention training program.



Observed proportion of players with lower limb injuries were significantly lower than expected during intervention seasons 1 and 2 (χ2 = 9.87, P = 0.007, df = 2). There were no differences in observed and expected proportion of players who experienced knee (χ2 = 1.98, P = 0.370, df = 2) and knee ligament (χ2 = 4.32, P = 0.115, df = 2) injuries. Observed number of players with ACL injuries were higher than expected in the control season and lower than expected in the intervention seasons (χ2 = 7.0, P = 0.030, df = 2) (Table 1). The expected values of knee ligament and ACL injuries were less than 5, which violated the frequency assumption of the chi-square test (i.e., the P value may be less accurate). Therefore, the statistical results for observed and expected knee ligament and ACL injury proportions should be treated with caution.

Back to Top | Article Outline

Biomechanical Injury Risk Factors

There were no differences in approach velocity or cut angle between any of the three biomechanical testing sessions, and as such, it can be assumed that the sidestepping condition was the same between testing sessions (Table 2).



There were no statistically significant reductions in PKE (P = 0.060), PKV (P = 0.530), and PKIR (P = 0.480) moments across time for the total group (Fig. 2). No significant response–time interactions were observed for PKE and PKIR moments (Figs. 2A and 2C). A response–time interaction was observed for PKV moments (P = 0.047). Post hoc analyses showed that after intensive training, responders reduced their PKV moments by 30% (g = 0.668, P = 0.045). At baseline, responders displayed 44% higher PKV moments than nonresponders (g = 0.626, P = 0.007) (Fig. 2B).

Figure 2

Figure 2

There were no between-group responder and nonresponder kinematic differences for the investigated variables, and as such all data presented are for the total group (Table 2). After both phases of the intervention, there were no statistically significant changes in kinematics.

Again, there were no between-group responder and nonresponder muscle activation differences, and as such all data were combined and presented as a total group (Fig. 3). During WA, mean gluteal TMA increased by 30% after intensive training (g = 0.609, P = 0.015) (Fig. 3C). No other statistically significant differences in TMA were observed.

Figure 3

Figure 3

Before the intervention, M/L DCCR were directed toward muscles with lateral moment arms during PC and WA. By contrast, after the intensive training phase, M/L DCCR were directed toward muscles with medial moment arms during PC (g = 0.596, P = 0.046) and WA (g = 0.561, P = 0.049). Similarly, SM/BF DCCR was laterally directed toward BF before training, yet after the intensive training phase, SM/BF DCCR was medially directed toward SM during PC (g = 0.720, P = 0.015) and WA (g = 0.609, P = 0.045). After maintenance training, SM/BF DCCR returned to a lateral activation strategy during PC (g = 0.574, P = 0.018) and WA (g = 0.194, P = 0.035) and was not significantly different to activation recorded before the intervention (Figs. 3B and 3D).

Back to Top | Article Outline

Athletic Performance

Improved 1RM strength was observed for bench press (+[INCREMENT]8%, g = 0.99, P = 0.001) and bench pull (+[INCREMENT]10%, g = 0.67, P = 0.013) after intervention season 1 relative to the control season, and these strength gains were maintained after intervention season 2. Back squat 1RM improved from the control season after intervention season 2 (+[INCREMENT]4%, g = 0.27, P = 0.046) (Table 3).



Ten-meter split time for 40-m sprint efforts improved relative to the control season after intervention season 1 (−[INCREMENT]2%, g = 0.63, P = 0.013). Times increased (+2%, g = 1.02, P = 0.011) from intervention season 1 to intervention season 2. However, there were no differences in 10-m split time from the control season to intervention season 2 (P = 0.353). Improvements in 40-m sprint time were observed between postintensive training and postintervention season 1 (−[INCREMENT]3%, g = 0.80, P = 0.026) and at postintervention season 2 from the control season (−[INCREMENT]0.2%, g = 0.5, P = 0.004) (Table 3). Beep test scores improved by 8% after intervention season 1 (g = 1.47, P < 0.001), reduced by 2% from intervention season 1 to intervention season 2 (g = 0.45, P = 0.004), but were still 6% better than the control season (g = 1.59, P = 0.002) (Table 3).

Back to Top | Article Outline


Biomechanically informed injury prevention training successfully reduced lower limb and ACL injury incidence among elite-level female field hockey players. In support of these findings, biomechanical risk factors (i.e., PKV moments) and neuromuscular countermeasures in general improved during the intensive training phase and were retained during the maintenance phase. In addition, although time was taken away from skills/performance training, selected performance measures were not compromised, and for some measures, improvements were observed. Overall, results support the efficacy of biomechanically informed and focused injury prevention training as both an ACL and a lower limb injury prevention training prescription.

Although training exposure increased after the control season, both total lower limb injury incidence (control season: 23; intervention season 2: 5.2) and ACL injury incidence (control season: 3; intervention season 2: 0) were reduced (Table 1). In support of our first hypothesis, observed ACL injuries were lower than expected in the two intervention seasons as compared with the control season (P = 0.03), with zero ACL injuries occurring after implementation. In addition, total lower limb injuries reduced after the second intervention season (P = 0.001). Although this was not an intentional outcome of this program, this finding highlights the importance of understanding the biomechanical mechanisms underpinning the observed reductions in injury rates.

In partial support of our second hypothesis, 9 wk of intensive training was successful in reducing PKV moments among the elite female hockey players within the responder group. Interestingly, responder athletes possessed PKV moments that were 44% higher in magnitude to athletes within the nonresponder group before the training program, indicating that they may have been “high risk” at the program commencement. Similarly, Myer and colleagues (47), through logistic regression, identified high-risk and low-risk athletes in a cohort of 18 high school female athletes and found that high-risk athletes reduced their PKV moments by 13% after 7 wk of training, whereas the low-risk and control groups did not demonstrate any meaningful reduction. Taken in combination with previous findings, our results suggest that athletes who have high knee loading may be able to reduce these loads through training; however, athletes who have relatively low knee loading at baseline will not experience a similar benefit (by magnitude). Identifying athletes who have high PKV moments during dynamic tasks such as sidestepping may, therefore, be an important factor when assessing the efficacy of injury prevention training programs to avoid data “washout” and to help enhance coach/athlete compliance to training (47,48). In support of our third hypothesis, the initial stimulus from intensive training among responder athletes was sufficient to maintain reductions in the high PKV moments from 9 to 25 wk.

Two biomechanical strategies capable of reducing ACL injury risk in sport were assessed within this study: 1) reducing external loads applied to the knee joint by modifying one’s technique and 2) increasing muscle support to counter external loads applied to the knee joint. In this study, we found no statistically significant differences in full-body sidestepping kinematics (technique). However, there were significant improvements in gluteal TMA, medial DCCR, and SM/BF DCCR after intensive training. Improvements in gluteal TMA and M/L DCCR were retained after maintenance training. Sidestepping is a complex dynamic movement with a vast kinematic solution space, and although we were able to observe kinetic and neuromuscular changes, this did not translate to a measurable kinematic effect. Simply, there may have not been a single unilateral kinematic change by this sample to explain the observed reductions in peak knee moments. This is not unusual or unexpected as simulation research has shown that for the same unplanned sidestepping task, an athlete can choose from 511 kinematic solutions to effectively reduce their PKV moments (34). As such, athletes may have responded to each of the training goals differently; however, all resulted in reduced or similar levels of knee loading in parallel with elevated musculature support at the hip and knee. This has been shown in other interventions using training to target trunk and hip neuromuscular control where the researchers observed increased hip abductor strength (27) and increased energy absorption at the hip (49).

A notable consideration in intervention-based research is training stimulus. In a retention study, Padua and colleagues (15) found that after 3 months of detraining, a 9-month training group maintained benefits, whereas a 3-month training group did not. This suggests that training stimulus required to elicit safe movement patterns may require more than the traditionally prescribed 2 × 15-min sessions per week for 6–12 wk as reported in ACL intervention literature. This is in contrast to the high dose intensive training phase in the present study which saw 4 × 20-min sessions for 9 wk. It is therefore important to ensure there is appropriate stimulus in the intensive training phase before understanding the retention effect of maintenance training.

A low coach–athlete ratio of 1:13 and 81% ± 25% attendance and 88% ± 20% compliance were exhibited in the present study. These results reflect those of other successful injury prevention programs such as the Sportsmetrics (13), KIPP (50), and Prevent Injury and Enhance Performance (14) programs. This is in contrast with unsuccessful interventions where attendance/compliance was low and coach to athlete ratios were high (30,51). By delivering a training “message” rather than a specific prescription of individual exercises, coaches are provided with education and choice surrounding the specifics of program implementation that on face value appears to have improved athlete compliance (52). In addition, to support our fourth hypothesis, there was no detriment to athletic performance when removing time from training to implement injury prevention training, which is a positive outcome for coach perceptions toward injury prevention implementation. These findings highlight three key factors in the design of injury prevention training research: 1) coach autonomy, 2) low coach to athlete ratios, and 3) high athlete attendance and compliance to gain optimal exposure. These research design factors appear to be crucial in understanding the effect of preventative training on the biomechanical factors associated with ACL injury risk and the subsequent reduction in ACL injury rates.

There were two notable limitations to this study. The first is the absence of a control group to compare these findings as our sample was limited to an Olympic female hockey team of which there is only one in Australia. We acknowledge that this limits the interpretation of the findings of this study; however, the combination of injury incidence and biomechanical risk factors of injury provides rationale for future work to investigate this training paradigm using control groups and among different populations. Although all athletes completed both phases of training, competing scheduling demands resulted in access to only 17 players for 3D biomechanical testing. Finally, due to retirement, injury, and availability over the study timeframe, six players were lost to follow-up after the maintenance phase of the intervention training and testing.

Back to Top | Article Outline


This study provides evidence for the efficacy of a biomechanically informed ACL injury prevention training paradigm. Rather than prescribing training modality (i.e., balance/plyometric/strength/technique), the injury prevention program presented in this manuscript is broken down into exercises within each modality that target specific biomechanical countermeasures associated with ACL injury risk. This allows training programs to be designed a) to specifically target an individual athlete’s malingering biomechanical patterns and b) to fit into any sporting environment, be it elite or community level. These findings provide evidence that injury prevention and athletic performance are not mutually exclusive and provide the necessary framework for coaches and sport science staff to translate into their program environment and structure.

The authors wish to thank Hockey Australia and the Hockeyroos for their continued support and involvement in this research project.

The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Agel J, Arendt EA, Bershadsky B. Anterior cruciate ligament injury in national collegiate athletic association basketball and soccer. Am J Sports Med. 2005;33(4):524–31.
2. Arendt EA, Agel J, Dick R. Anterior cruciate ligament injury patterns among collegiate men and women. J Athl Train. 1999;34(2):86.
3. Gianotti SM, Marshall SW, Hume PA, Bunt L. Incidence of anterior cruciate ligament injury and other knee ligament injuries: a national population-based study. J Sci Med Sport. 2009;12(6):622–7.
4. Gornitzky AL, Lott A, Yellin JL, Fabricant PD, Lawrence JT, Ganley TJ. Sport-specific yearly risk and incidence of anterior cruciate ligament tears in high school athletes: a systematic review and meta-analysis. Am J Sports Med. 2016;44(10):2716–23.
5. Sigward S, Powers CM. The influence of experience on knee mechanics during side-step cutting in females. Clin Biomech (Bristol, Avon). 2006;21(7):740–7.
6. Cole DW, Ginn TA, Chen GJ, et al. Cost comparison of anterior cruciate ligament reconstruction: autograft versus allograft. Arthroscopy. 2005;21(7):786–90.
7. Filbay SR, Ackerman IN, Russell TG, Macri EM, Crossley KM. Health-related quality of life after anterior cruciate ligament reconstruction: a systematic review. Am J Sports Med. 2014;42(5):1247–55.
8. Webster KE, Hewett TE. Meta-analysis of meta-analyses of anterior cruciate ligament injury reduction training programs. J Orthop Res. 2018;36(10):2696–708.
9. Noyes FR, Barber Westin SD. Anterior cruciate ligament injury prevention training in female athletes: a systematic review of injury reduction and results of athletic performance tests. Sports Health. 2012;4(1):36–46.
10. Stojanovic MD, Ostojic SM. Preventing ACL injuries in team-sport athletes: a systematic review of training interventions. Res Sports Med. 2012;20(3–4):223–38.
11. Caraffa A, Cerulli G, Projetti M, Aisa G, Rizzo A. Prevention of anterior cruciate ligament injuries in soccer. Knee Surg Sports Traumatol Arthrosc. 1996;4(1):19–21.
12. Cochrane JL, Lloyd DG, Besier TF, Elliott BC, Doyle TL, Ackland TR. Training affects knee kinematics and kinetics in cutting maneuvers in sport. Med Sci Sports Exerc. 2010;42(8):1535–44.
13. Hewett TE, Lindenfeld TN, Riccobene JV, Noyes FR. The effect of neuromuscular training on the incidence of knee injury in female athletes. Am J Sports Med. 1999;27(6):699–706.
14. Mandelbaum BR, Silvers HJ, Watanabe DS, et al. Effectiveness of a neuromuscular and proprioceptive training program in preventing anterior cruciate ligament injuries in female athletes. Am J Sports Med. 2005;33(7):1003–10.
15. Padua DA, DiStefano LJ, Marshall SW, Beutler AI, de la Motte SJ, DiStefano MJ. Retention of movement pattern changes after a lower extremity injury prevention program is affected by program duration. Am J Sports Med. 2012;40(2):300–6.
16. Besier TF, Lloyd DG, Cochrane JL, Ackland TR. External loading of the knee joint during running and cutting maneuvers. Med Sci Sports Exerc. 2001;33(7):1168–75.
17. Krosshaug T, Nakamae A, Boden BP, et al. Mechanisms of anterior cruciate ligament injury in basketball: video analysis of 39 cases. Am J Sports Med. 2007;35(3):359–67.
18. Markolf KL, Burchfield DM, Shapiro MM, Shepard MF, Finerman GA, Slauterbeck JL. Combined knee loading states that generate high anterior cruciate ligament forces. J Orthop Res. 1995;13(6):930–5.
19. Dempsey AR, Lloyd DG, Elliott BC, Steele JR, Munro BJ, Russo KA. The effect of technique change on knee loads during sidestep cutting. Med Sci Sports Exerc. 2007;39(10):1765–73.
20. McLean SG, Huang X, Su A, Van Den Bogert AJ. Sagittal plane biomechanics cannot injure the ACL during sidestep cutting. Clin Biomech (Bristol, Avon). 2004;19(8):828–38.
21. Chaudhari AM, Hearn BK, Andriacchi TP. Sport-dependent variations in arm position during single-limb landing influence knee loading. Am J Sports Med. 2005;33(6):824–30.
22. Donnelly CJ, Elliott BC, Ackland TR, et al. An anterior cruciate ligament injury prevention framework: incorporating the recent evidence. Res Sports Med. 2012;20(3–4):239–62.
23. Besier TF, Lloyd DG, Ackland TR. Muscle activation strategies at the knee during running and cutting maneuvers. Med Sci Sports Exerc. 2003;35(1):119–27.
24. Morgan KD, Donnelly CJ, Reinbolt JA. Elevated gastrocnemius forces compensate for decreased hamstrings forces during the weight-acceptance phase of single-leg jump landing: implications for anterior cruciate ligament injury risk. J Biomech. 2014;47(13):3295–302.
25. Noyes FR, Barber-Westin SD. Neuromuscular retraining intervention programs: do they reduce noncontact anterior cruciate ligament injury rates in adolescent female athletes? Arthroscopy. 2014;30(2):245–55.
26. Hewett TE, Myer GD, Ford KR, et al. Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes. Am J Sports Med. 2005;33(4):492–501.
27. Myer GD, Brent JL, Ford KR, Hewett TE. A pilot study to determine the effect of trunk and hip focused neuromuscular training on hip and knee isokinetic strength. Br J Sports Med. 2008;42(7):614–9.
28. Myer GD, Ford KR, Palumbo OP, Hewett TE. Neuromuscular training improves performance and lower-extremity biomechanics in female athletes. J Strength Cond Res. 2005;19(1):51–60.
29. Donnelly CJ, Elliott BC, Doyle TL, Finch CF, Dempsey AR, Lloyd DG. Changes in muscle activation following balance and technique training and a season of Australian football. J Sci Med Sport. 2015;18(3):348–52.
30. Donnelly CJ, Elliott BC, Doyle TL, Finch CF, Dempsey AR, Lloyd DG. Changes in knee joint biomechanics following balance and technique training and a season of Australian football. Br J Sports Med. 2012; bjsports–2011.
31. Hashemi J, Breighner R, Chandrashekar N, et al. Hip extension, knee flexion paradox: a new mechanism for non-contact ACL injury. J Biomech. 2011;44(4):577–85.
32. Krosshaug T, Slauterbeck JR, Engebretsen L, Bahr R. Biomechanical analysis of anterior cruciate ligament injury mechanisms: three-dimensional motion reconstruction from video sequences. Scand J Med Sci Sports. 2007;17(5):508–19.
33. Zazulak BT, Hewett TE, Reeves NP, Goldberg B, Cholewicki J. Deficits in neuromuscular control of the trunk predict knee injury risk. Am J Sports Med. 2007;35(7):1123–30.
34. Donnelly CJ, Lloyd DG, Elliott BC, Reinbolt JA. Optimizing whole-body kinematics to minimize valgus knee loading during sidestepping: implications for ACL injury risk. J Biomech. 2012;45(8):1491–7.
35. Jackson B, Dimmock JA, Taylor IM, Hagger MS. The tripartite efficacy framework in client–therapist rehabilitation interactions: implications for relationship quality and client engagement. Rehabil Psychol. 2012;57(4):308.
36. Rae K, Orchard J. The Orchard sports injury classification system (OSICS) version 10. Clin J Sport Med. 2007;17(3):201–4.
37. Phillips LH, Standen PJ, Batt ME. Effects of seasonal change in rugby league on the incidence of injury. Br J Sports Med. 1998;32(2):144–8.
38. Dempsey AR, Lloyd DG, Elliott BC, Steele JR, Munro BJ. Changing sidestep cutting technique reduces knee valgus loading. Am J Sports Med. 2009;37(11):2194–200.
39. Besier TF, Sturnieks DL, Alderson JA, Lloyd DG. Repeatability of gait data using a functional hip joint centre and a mean helical knee axis. J Biomech. 2003;36(8):1159–68.
40. Wu G, Siegler S, Allard P, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion—part I: ankle, hip, and spine. J Biomech. 2002;35(4):543–8.
41. Devaprakash D, Weir GJ, Dunne JJ, Alderson JA, Donnelly CJ. The influence of digital filter type, amplitude normalisation method, and co-contraction algorithm on clinically relevant surface electromyography data during clinical movement assessments. J Electromyogr Kinesiol. 2016;31:126–35.
42. Heiden TL, Lloyd DG, Ackland TR. Knee joint kinematics, kinetics and muscle co-contraction in knee osteoarthritis patient gait. Clin Biomech (Bristol, Avon). 2009;24(10):833–41.
43. Kraemer W, Ratamess N, Fry A, French D. Physiological assessment of human fitness. In: Maud PJ, Foster C, editors. Strength Training: Development and Evaluation of Methodology. Champain (IL): Human Kinetics; 2006.
44. Lager L, Lambert J. A maximal multistage 20-m shuttle run test to predict. Eur J Appl Physiol O. 1982;49:1–12.
45. West BT, Welch KB, Galecki AT. Linear Mixed Models: A Practical Guide Using Statistical Software. Chapman and Hall/CRC; 2014.
46. Cohen J. Statistical Power Analyses for the Social Sciences. Hillsdale (NJ): Lawrence Erlbaum Associates; 1988.
47. Myer GD, Ford KR, Brent JL, Hewett TE. Differential neuromuscular training effects on ACL injury risk factors in “high-risk” versus “low-risk” athletes. BMC Musculoskelet Disord. 2007;8(1):39.
48. Hewett TE, Ford KR, Xu YY, Khoury J, Myer GD. Effectiveness of neuromuscular training based on the neuromuscular risk profile. Am J Sports Med. 2017;45(9):2142–7.
49. Pollard CD, Sigward SM, Powers CM. ACL injury prevention training results in modification of hip and knee mechanics during a drop-landing task. Orthop J Sports Med. 2017;5(9):2325967117726267.
50. LaBella CR, Huxford MR, Grissom J, Kim K-Y, Peng J, Christoffel KK. Effect of neuromuscular warm-up on injuries in female soccer and basketball athletes in urban public high schools: cluster randomized controlled trial. Arch Pediatr Adolesc Med. 2011;165(11):1033–40.
51. Myklebust G, Engebretsen L, Braekken IH, Skjølberg A, Olsen OE, Bahr R. Prevention of anterior cruciate ligament injuries in female team handball players: a prospective intervention study over three seasons. Clin J Sport Med. 2003;13(2):71–8.
52. Chan DK, Hagger MS. Self-determined forms of motivation predict sport injury prevention and rehabilitation intentions. J Sci Med Sport. 2012;15(5):398–406.

Supplemental Digital Content

Back to Top | Article Outline
© 2019 American College of Sports Medicine