Prediction of secondary ACL injury in female athletes using 2D video-based measurements obtained during dynamic tasks: a retrospective case–control study

Key Points

1. During return-to-sport testing after ACL reconstruction, there are few valid clinical measures of movement quality.
2. This retrospective case–control study investigated the ability of 2D video-based measurements of dynamic tasks to predict secondary ACL injuries in female athletes.
3. The 2D video-based measurements had fair to good predictive ability. The strongest predictor of secondary injury was exhibiting several movement patterns across tasks characterized by deep knee flexion and an upright torso (a proxy for high knee extensor moments).

BACKGROUND AND OBJECTIVE

In high-risk sports, ~25% of athletes who return to sport (RTS) after an anterior cruciate ligament (ACL) reconstruction sustain a second ACL injury, either to the same or opposite leg (1). One potential contributing factor to this high reinjury rate is inadequate RTS testing.

Current testing protocols typically assess strength and performance during dynamic tasks (e.g. jumping, landing, and cutting) but overlook movement quality (e.g. knee valgus, pelvic tilt, and trunk lean). When movement quality has been assessed, it has often been conducted in a 3D biomechanical laboratory with equipment that isn’t accessible to most clinicians. A more clinically viable approach is 2D video analysis.

This study investigated the ability of 2D video-based measurements of dynamic tasks to predict secondary ACL injuries in female athletes.

In high-risk sports, ~25% of athletes who return to sport after an anterior cruciate ligament reconstruction sustain a second ACL injury.

This finding supports the exploration of hip-dominant movement training – and other risk-reducing movement strategies investigated in the study – for secondary ACL injury prevention.

METHODS

80 female athletes (median age = 17, range = 11–26) with a history of ACL reconstruction who played high-risk sports were included in a retrospective case–control study. Of the 80 athletes, 23 cases sustained a second ACL injury (all non-contact; 9 ipsilateral, 14 contralateral). The other 57 served as controls, matched to cases by graft type, age (±5 years), sport level, and exposure hours.

Prior to RTS, all participants performed one trial each of six dynamic tasks: step down, drop jump, lateral shuffle, deceleration, triple hop, and side-step cut (see Video 1). The tasks were captured by high-speed frontal and sagittal plane video. For each task, the investigators measured five 2D angles from the video frame during which peak knee flexion occurred (see Table 1). These angles were considered surrogates for key variables typically obtained through biomechanical lab testing (2).

VIDEO 1 – 6 DYNAMIC TASKS

For each angle and each task, cutoff scores were determined that maximized sensitivity and specificity for distinguishing cases from controls. Each athlete got a composite score for each 2D angle, which was based on the number of tasks they were classified as high risk on for that angle. For example, if an athlete’s trunk-tibia angle (see Fig. 1) was deemed high risk in 4/6 tasks, her trunk-tibia angle composite score was 4. Composite scores for each 2D angle, along with time from surgery to RTS and body mass index (BMI), were used to predict secondary ACL injury.

RESULTS

Cases demonstrated significantly higher composite scores (i.e. higher-risk movement patterns) for all 2D angles except trunk lean (see Table 2). After adjusting for time from surgery to RTS and BMI, all 2D angles except trunk lean were significantly associated with increased odds of secondary injury (Table 2). The area under the receiver operating characteristic curves ranged from 0.73 to 0.81, corresponding to fair to good predictive ability. Trunk-tibia angle was the strongest predictor.

LIMITATIONS

This study had two main limitations. First, participants were female athletes aged 11–26 with a history of ACL reconstruction. The results may not generalize to other contexts (e.g. males, older athletes, or primary ACL injury risk). The cutoff scores from this dataset should be validated in other samples before being implemented in clinical practice. Second, the retrospective design relied on self-reported reinjury, which introduces the potential for recall bias. Future prospective designs are also needed for validation.

CLINICAL IMPLICATIONS

Movement quality is often overlooked in RTS testing, perhaps due to a shortage of valid clinical measures. This study addressed this gap. Although further validation is required, the 2D video-based measurements of dynamic tasks investigated had fair to good ability to predict secondary ACL injuries in female athletes.

A previous large-scale study using similar 2D techniques found no association between knee valgus and ACL injury in female athletes during single-leg squat and drop jump tasks (3). However, that study differed in two important ways: it focused on primary (not secondary) ACL injuries, and it analyzed the tasks individually. The present study’s innovation lies in its use of composite scores, which amalgamated movement quality across several tasks. These composite scores provide a more holistic description of an athlete’s movement strategy compared to an individual task, better reflecting sports’ variable demands.

In this study, the trunk-tibia angle composite score was the strongest predictor of secondary ACL injuries. Lower trunk-tibia angles (i.e. higher degrees of knee flexion with a more vertical trunk) are a surrogate for high knee extensor moments. For every additional task performed with a high-risk trunk-tibia movement strategy, odds of secondary ACL injury increased 75%. This finding supports the exploration of hip-dominant movement training – and other risk-reducing movement strategies investigated in the study – for secondary ACL injury prevention.

+STUDY REFERENCE

Straub R, Powers C (2025) Prediction of secondary ACL injury in female athletes using 2D video-based measurements obtained during dynamic tasks: A retrospective case–control study. British Journal of Sports Medicine, 59(20), 1418–1425.

SUPPORTING REFERENCE

1. Wiggins, A. J., Grandhi, R. K., Schneider, D. K., Stanfield, D., Webster, K. E., & Myer, G. D. (2016). Risk of secondary injury in younger athletes after Anterior Cruciate Ligament Reconstruction. The American Journal of Sports Medicine, 44(7), 1861–1876.
2. Straub, R. K., & Powers, C. M. (2022). Utility of 2D video analysis for assessing frontal plane trunk and pelvis motion during stepping, landing, and change in direction tasks: A validity study. International Journal of Sports Physical Therapy, 17(2).
3. Nilstad, A., Petushek, E., Mok, K.-M., Bahr, R., & Krosshaug, T. (2021). Kiss goodbye to the ‘kissing knees’: No association between frontal plane inward knee motion and risk of future non-contact ACL injury in Elite Female Athletes. Sports Biomechanics, 22(1), 65–79.