Cognitive-Motor Function and ACL Injury Risk: Exploring Relationships Between Clinical and Lab-Based Assessments

Abstract

Assessing cognitive-motor function following anterior cruciate ligament reconstruction (ACLR) may enable a more comprehensive assessment of readiness to return to sport. The primary purpose of this study was determining the relationship between measures of clinical cognitive-motor interference and laboratory-based second injury relevant biomechanics through a cross-sectional study.

Methods

Thirty-six individuals following ACLR (26F/10M, 19.8±1.8 years; 1.7±0.1 m; 69.6±12.8 kg, Tegner activity level 6.8±1.8, 1.5±0.6 years post ACLR) participated in the study. We assessed two clinically feasible tests of cognitive-motor function: single and dual tasks for triple hop (ST-TH, DT-TH) and medial side hop (ST-MSH, DT-MSH). Laboratory-based tests involved motion capture assessment of jump landings from a 30 cm box followed immediately by a secondary jump. Conditions included single task (anticipated secondary jump direction) and dual task (unanticipated secondary jump direction). Primary outcome variables included dual-task change (DTC) of three second ACL injury predictors (limb asymmetry in knee extensor moment at initial contact, range of knee abduction angle for the involved limb (Inv-KAbA-range), and uninvolved hip rotation net moment impulse). Correlations were used to identify associations between lab- (biomechanics) and clinic- (hop distance) derived outcomes. Secondary analyses with other risk-associated knee mechanics and patient reported outcome measures were also performed.

Results

No significant correlations were found between DTC clinical variables and DTC second ACL predictors. However, significant relationships were observed between clinical and lab outcomes when considering single task or dual task in isolation, such as single task Inv-KAbA-range correlating with ST-TH for both limbs (Inv: r=−436, p_adj=0.04, Uninv: r=−0.453, p_adj=0.04).

Conclusions

Associations between cognitive-motor interference (i.e., DTC) in clinical and lab tests were not found, but significant fair strength correlations appeared between clinical and lab variables when examining single-task and dual-task performance separately. Future prospective studies are needed to determine the predictive value of cognitive-motor function for injury rehabilitation decision-making.

Introduction

Following rehabilitation for anterior cruciate ligament reconstruction (ACLR), return-to-sport (RTS) decision-making relies on a combination of time from surgery, self-reported measures of physical and psychological function, and functional tests.^1–3 Functional tests include muscle strength, and a variety of hop or jump assessments.⁴ However, after testing these criteria and returning to sport, about 30% of young athletes sustain a second ACL injury to the reconstructed or contralateral limb.⁵ Despite numerous studies aimed at better understanding ACL injuries, the connection between meeting the RTS criteria and the risk of a second ACL injury remains unclear.⁶

Recent findings regarding the neurological alterations following lower-extremity injuries highlight the critical necessity of integrating visual-cognitive challenges within RTS assessments.^{7, 8} Notably, up to 70% of second ACL injuries happen when athletes’ attention is directed externally (e.g. an opponent, ball, or other environmental factors).⁹ Additionally, poorer visual-spatial performance is linked to riskier biomechanics during cognitively challenging movements.¹⁰ Despite the prevalence of cognitive-motor demands in ACL injury scenarios, cognitive-motor function is not routinely a criterion for RTS following ACLR. Assaying cognitive-motor function may provide more complete information for RTS decision-making and reduce secondary injury risk.^{7, 8} However, there is a gap in knowledge regarding the extent that clinically feasible tests can be adequate surrogates for injury relevant cognitive-motor interference. To address this gap, dual-task clinical hop tests (triple hop and medial-side hop combined with visual-cognitive tasks), which have established high test-retest reliability in prior research,^{11, 12} have been proposed as practical tools for assessing cognitive-motor interference.

Therefore, the purpose of this study was to determine the ability of low-cost, accessible cognitive-motor tests (i.e., clinical tests) to capture clinically relevant information, specifically biomechanical second ACL injury risk factors, obtained from a biomechanics laboratory (i.e., lab tests). We hypothesized that clinical measures of cognitive-motor interference would be highly correlated with motion capture laboratory-derived second ACL injury risk biomechanics. Accepting the hypothesis would provide support for clinical tests that relate to biomechanical risk factors of a second ACL injury following ACLR and support efforts to augment RTS decision-making. Rejecting the hypothesis would emphasize that hop distance during clinical tasks is not a reliable indicator of patients’ biomechanics in cognitive-motor scenarios following ACLR.

Methods

Participants

Eligibility included: 1) one prior ACL injury and reconstruction, 2) ACLR surgery occurred within 2.5 years prior to the time of testing, 3) cleared to return to sport by their clinical team, 4) no concussion or lower-body musculoskeletal injury following ACL injury, 5) no known neurologic or musculoskeletal disorder.

Thirty-six persons (Table 1) who completed rehabilitation from a primary ACLR participated. All participants provided written informed consent and assent for those under 18 years old (Montana State University Institutional Review Board protocol # SM081818). For all tasks, participants wore their preferred athletic shoes for testing. Motion capture testing of jump-landing tasks occurred prior to participants performing the hop tests.

Table 1:

Cohort characteristics (Average ± Standard deviation)

Female/Male	26/10
Age (years)	19.8±1.8
Height (m)	1.7±0.1
Mass (kg)	69.6±12.8
Time Since Surgery (years)	1.5±0.6
Tegner Score	6.8±1.8
Marx Activity Score	10.6±4.1
Graft Type
Quadriceps graft	14
Hamstring graft	12
Bone-Patella-bone graft	9
Allograft	1
Primary Sport Participation
Down-hill Skiing	13
Soccer	6
Basketball	6
Volleyball	3
Running (recreational)	3
Weightlifting	2
Other	3

Lab-based Testing: Motion Capture Assessment of Jump Landings Biomechanics

Seventy-five reflective markers attached to participants captured their motion and force plates recorded ground reaction forces during the lab tasks (Figure 1- A). The marker set followed a point-cluster distribution on the thighs and shanks.^{10, 13} A plug-in gait marker set was used for the upper body, with additional markers on the iliac crests used to assist pelvis tracking during the jumping tasks (Figure 1-A). We used a 10-camera motion capture system to record the kinematic data at 250 Hz (Motion Analysis Corp, Rohnert Park, CA) and four force plates to record ground reaction force data at 1000 Hz (OPT464508–2K; Advanced Mechanical Technology, Inc; Watertown, MA). Before starting the lab-based test, participants did two sets of 8 bodyweight squats and 5 maximal effort vertical jumps as a warm-up.

Figure 1:

(A) Picture of a participant with marker set, 3-dimensional model in Cortex, and Visual 3D for motion analysis (A, from left to right), (B) Secondary jump directions, (C) Jump-land-jump task progression.

The lab-based test was jumping from a 30 cm box placed a distance of half of their height from the force plates and had a secondary jump in one of the three directions of straight up, 45° to the right, or 45° to the left (Figure 1-B).¹⁴ The jumping task was performed in two conditions: single-task and dual-task (Figure 1-C), with five successful trials for the straight up direction collected for each participant with the average of biomechanical variables for the five trials used in analyses. The single-task (Lab-ST) condition involved an anticipated direction (up secondary jump direction) with no cognitive task. In this task, participants’ visual gaze was not constrained, and they were instructed to jump as high and hard as they could immediately upon landing from the initial jump from the box. The dual-task (Lab-DT) condition presented a secondary jump direction cue ~250 msec prior to initial contact to keep the task unanticipated.¹⁴ Timing of the cue was based on a signal from a force sensitive resistor on the box and a previously described procedure to calibrate presentation timing to individual participants’ jumping strategies to maintain a consistent ~250 msec reaction time for all participants.¹⁵ Time to react (i.e., cue presentation until initial contact) was verified post-hoc (249 ± 90 msec). The secondary tasks were conducted using E-Prime 3.0 software (Psychology Software Tools, Pittsburgh, PA) and presented on a 1.5 m television screen positioned about 5 m away from the primary landing zone. The participants were asked to look at the fixation in front of them during the entirety of the jump landing. Directional cues were random for both participant and examiner. Only trials with a secondary jump direction of straight up were analysed for this study. Successful trials were defined as reacting correctly to the arrow and not looking down during the entirety of the jump landing.

Data Processing for Lab Tests

Data processing was performed in Visual 3D (HAS-Motion; Kingston, Ontario) and MATLAB (version R2022b; MathWorks, Natick, MA). A CODA pelvis was employed, and hip joint centers were estimated using the Visual3D implementation of the prediction equations from Bell et al.¹⁶ The knee joint centers were defined as the midpoints between markers on the femoral epicondyles, and the ankle joint centers as the midpoints between markers on the malleoli. Segments were defined with the z-axis directed from distal to proximal joint centers, the y-axis oriented anteriorly, and the x-axis oriented to the participants’ right. Kinematic and force plate data were filtered using fourth-order, zero-lag, low-pass Butterworth filters with cutoff frequencies set at 15 Hz.¹⁷ Kinematic data were computed using a Quasi-Newton optimization method and an inverse kinematics model with limiting the degrees of freedom to 3 rotational degrees at the hip and knee and 2 rotational degrees at the ankle. Kinetic data were computed using the model with the inertial and ground reaction force data to solve for the internal net joint forces and moments.¹⁵ For kinetic variables, forces normalized to the mass, and moments normalized to both mass and height of each participant.

Joint kinematics were estimated using X-Y-Z Cardan rotation with X-, Y-, Z-axis rotations corresponding to (+) flexion/ (−) extension, (+) adduction/ (−) abduction, and (+) internal/ (−) external rotation, respectively, for the hip and knee joints. Internally defined net reaction moments resolved in the proximal joint segment were also calculated, with the knee and hip following similar conventions as previously stated for the kinematics. Initial contact was the frame where the vertical ground reaction force first exceeded 10 N.

Lab-based Biomechanical Outcomes

Previously reported biomechanical predictors of second ACL injuries were calculated for the landing from the initial jump.¹⁸ We calculated the range in knee abduction angle for the involved limb (Inv-KAbA-range) within a window starting from the initial contact to the minimum of the body’s center of mass, with larger values reflecting more frontal plane motion and interpreted as riskier.¹⁸ Limb asymmetry in knee extensor moment at initial contact was calculated and normalized to body weight and height (KEM-Asym), with negative values reflecting a greater knee extension moment in the uninvolved limb compared to the involved limb and interpreted as riskier for second injury.¹⁸ Uninvolved hip rotation net moment impulse was calculated during the initial 10% of the chosen window (Uni-HRot-Imp), with larger values reflecting less external rotator moments and interpreted as riskier.¹⁸ To quantify cognitive-motor interference, we calculated dual-task change (DTC = DT – ST) for each of the variables. A pure difference, rather than a percent change, was used due to the kinematic/kinetic outcomes spanning zero and therefore being overly sensitive when dividing by values near zero.

Clinic-Based Testing: Hop Distance

The clinical tasks were the triple hop (TH)¹¹ and medial side hop (MSH)¹² for distance, assessed without biomechanics laboratory equipment. Participants were asked to do these tasks with (DT) and without visual-cognitive challenges (ST), as described below.

Traditional and Visual-Cognitive Triple Hop¹¹

The order of ST-TH and DT-TH was randomized. In the ST, the participants hopped three consecutively on the same leg, maximizing distance and maintain balance on the final hop for three seconds. In the visual-cognitive task, we used the FitLight system (FITLIGHT Sports Corp.) following the same protocol as Farraye et. al.¹¹ Participants did two practice trials. In the visual-cognitive task, participants were required to respond to specific-colored peripheral lights as their “GO” signal and hopping forward three times while keeping their focus on a light in front of them and then recalling a sequence of colors flashed by the central light during the hopping (Figure 2-A). The peripheral lights randomly displayed different colors, with one designated as the “GO” signal (Figure 2-B). A successful trial required the participant to trigger the reaction time FitLight while executing the same TH task as described for the ST condition. If the participant lost balance, added extra hops, or switched legs during the trial, it was considered a physical miss, and the trial was repeated. Three successful trials were completed on each limb, starting with the uninjured side.

Figure 2:

(A) Time-lapse picture of a participant reacting to “GO” color on the peripheral lights and initialized triple hop (TH) while looked at the light in front of them to memorize the three colors in order, (B) Set up of the triple hop task, (C) Picture of a participant performing a medial side hop (MSH) and focused on the task on TV, (D) Overview of the visual-cognitive task presented on TV during MSH. (TH: Triple Hop, MSH: Medial Side Hop)

Traditional and Visual-Cognitive Medial Side Hop¹²

Participants also performed a MSH for distance as a ST (ST-MSH), and with simultaneous visual-cognitive task as a DT (DT-MSH) in a randomized order (Figure 2-C). For the ST, participants stood on one leg, hands on hips and, hopped medially as far as possible, holding the landing for 30 seconds, and starting on their uninjured limb. After one practice trial for each limb, three successful trials were completed for each limb, with success defined by maintaining balance without foot shifts, opposite foot use, or hand movement. In the visual-cognitive task, participants stood 2 m away from a 1.5 m television in a relaxed position. They were shown a circular array of eight letters (D, F, J, K) around a central dot on the screen for 100ms (Figure 2-D). After the letters disappeared, a line appeared for 1.9s, prompting participants to recall the letter at the corresponding location. A 10s practice trial and two 30s trials were completed of the visual-cognitive task in isolation (i.e., no hopping) prior to completing the dual task. Then participants were asked to do the same visual-cognitive task while they performed the MSH for distance (DT-MSH).

Clinical Test Outcomes: Hop Distance

To avoid the need for specialized biomechanics lab equipment, clinical test outcomes focused solely on hop distance. The maximum distance among 3 successful trials normalized to leg length (%L) were used for analyses. DTC for triple hop (DTC-TH) and medial side hop (DTC-MSH) distances for involved and uninvolved limbs were primary outcomes of the clinical tests. Reliability for these metrics are ICC_3,1 = 0.62 for DTC-MSH and ICC_3,1 = 0.94 for DTC-TH based on reanalysis of data of prior studies for the DTC metric.^{11, 12} For consistency with the lab-based outcomes, we used a difference score rather than percent change to capture the cognitive-motor interference; however, difference scores and percent changes yielded high correlations (r = 0.95–0.97) for hop distances.

Supplemental Outcomes

Our primary analysis focused on the ability for clinical test outcomes to correlate with previously reported biomechanical predictors of second ACL injuries measured via motion capture in a biomechanics laboratory.¹⁸ To provide additional perspective regarding the clinical relevance of the clinical cognitive-motor tests, we added exploratory outcomes focused on primary ACL injury risk factors of the involved limb such as peak knee flexion angle (pKFA), peak knee abduction angle (pKAbA), and peak knee abduction moment (defined externally and normalized to body weight and height (pKAbM)) as well as a metric of overall task performance (i.e., jump height) for the lab tasks. Additionally, we performed correlations between the clinical cognitive-motor test outcomes and patient reported outcomes (PROs) related to physical function, quality of life, psychological readiness, and fear. Specifically, participants completed the following surveys: International Knee Documentation Committee subjective knee evaluation form (IKDC),¹⁹ Knee Injury and Osteoarthritis Outcome Score (KOOS),²⁰ shortened Tampa Scale of Kinesiophobia (TSK-11)²¹, ACL Return to Sport after Injury (ACL-RSI)²², and the Forgotten Joint Score-12 knee (FJS-12).²³ Analyses for these supplemental outcomes are provided in Supplemental Table 1 and Supplemental Figures 1–3.

Statistical Analysis

Pearson correlations were used for normally distributed variables and Spearman correlations for variables not normally distributed according to the Anderson-Darling Normality test to identify significant associations between lab and clinical measures. The same approach was used for the supplemental outcomes. P-values were adjusted within each set of correlations (e.g. DTC lab-based and clinical variables, DT lab-based and clinical variables, and ST lab-based and clinical variables) between biomechanic variables and hop distance with Benjamini-Hochberg method to control the false discovery rate. For significant correlations, follow-up regression models were used to test for DTC lab-based variables as dependent variables with the DTC clinic-based variables as predictors and assess potential confounding factors (i.e., gender and jump height) included as covariates. The Benjamini-Hochberg method was used for the p-values of regressions for the same set of variables with significant results. The necessary sample size to provide 80% statistical power and α = 0.05 to detect relationships with magnitudes of at least moderate strength (r = 0.5) was estimated to be n = 29 by using G*Power.

Results

The results are based on 36 clinical-ST and 36 clinical-DT data sets and 36 lab-ST data sets and 35 lab-DT data sets (number of missed trials and collected data in Supplemental Table 2). We excluded lab-DT data for one participant due to data collection issues (trials used: 160 out of 180 (36 participants × 5 trials) trials for lab-DT, and 179 out of 180 trials for lab-ST. Descriptive statistics and DTC for these trials are provided in Table 2.

Table 2:

Average ± Standard deviation of clinical tests and lab drop vertical jump tests variables

CLINICAL VALUES	ST	DT	DTC (DT – ST)
MSH-Inv [%L]	1.08±0.30	1.07±0.30	−0.005±0.110
MSH-Uninv [%L]	1.12±0.33	1.1±0.3	−0.024±0.100
TH-Inv [%L]	4.91±0.99	4.55±0.89	−0.36±0.35*
TH-Uninv [%L]	5.03±1.03	4.73±0.86	−0.3±0.5*
LAB VALUES
Uni-HRot-Imp [N.m.s/kg]	0.005±0.004	0.005±0.004	0.00005±0.0031
KEM-Asym [%BW-HT]	−0.36±0.83	−0.36±0.79	−0.073±0.64
Inv-KAbA-range [°]	15.9±4.4	15.5±4.8	−0.25±2.03

^*:Significant difference of p-value <0.05) (MSH: Medial Side Hop, TH: Triple Hop, Inv: Involved limb, Uninv: Uninvolved limb, Uni-HRot-Imp [N.m.s/kg]: Uninvolved hip rotation net moment impulse, KEM-Asym [%BW-HT]: Limb asymmetry in knee extensor moment at initial contact normalized to body weight and height, Inv-KAbA-range [°]: range in knee abduction angle for the involved limb, N: Newton, m: meters, s: second, BW: body weight, HT: height, %L: normalized to the leg length

Second ACL Injury Predictors

There were no significant correlations between DTC clinical variables and DTC second ACL injury predictors (Figure 3). When considering the ST conditions in isolation, clinical and lab outcomes had several significant relationships. Inv-KAbA-range was related to ST-TH hop distance for both involved and uninvolved limbs (Inv: r = −0.44, p_adj = 0.04, Uninv: r = −0.45, p_adj = 0.04 Figure 4-A), and ST-MSH-Inv hop distance (r = −0.42, p_adj = 0.04, Figure 4-A), with farther clinical hop distance related to less frontal plane motion in the lab-based jump landing. In post hoc models, these results stayed significant with gender as a covariate (ST-TH-Inv: p_adj = 0.034, ST-TH-Uninv: p_adj = 0.02, ST-MSH-Inv: p_adj = 0.037). However, when adding lab testing jump height as a covariate, only the regression between Inv-KAbA-range and ST-TH-Uninv remained significant (p_adj = 0.029), with other previously significant relationships trending after controlling for multiple comparisons (ST-TH-Inv: p_adj = 0.061, ST-MSH-Inv: p_adj = 0.074). DT-TH-Inv was correlated with Lab-DT-Uni-HRot-Imp (r = 0.432, p = 0.01, Figure 4-B) before adjusting the p-value; however, it no longer remained significant after correcting for multiple comparisons (p_adj = 0.12).

Figure 3:

Correlogram for dual-task change (DTC) for second and primary ACL injury predictors, overall performance in drop vertical jump, and clinical tests (TH: Triple Hop, MSH: Medial Side Hop). Numeric values shown are correlation coefficients, and the darker blue color is more positive correlation while darker red color is more negative correlation.

Figure 4:

(A) Correlogram for single-task variables (ST) for second and primary ACL injury predictors, and overall performance in drop vertical jump in ST jump landing and clinical tests. (B) Correlogram for dual-task variables (DT) for second and primary ACL injury predictors, and overall performance in drop vertical jump in DT jump landing and clinical tests (*: p-adj of correlation <0.05, **: p-adj of regression with gender as a covariate <0.05, ***: p-adj of regression with gender and jump height as covariates <0.05).

Summary of Supplemental Analyses

In summary, DTC for clinical variables did not significantly correlate with DTC for primary ACL predictors, DTC for overall performance (Supplemental Figure 1), or PROs (Supplemental Figure 2). In ST conditions, most clinical test variables correlated with primary ACL injury predictors, with greater distances in clinical tasks linked to increased knee flexion and abduction angles, and decreased abduction moments (Supplemental Figure 3-A). Similar correlations were observed in DT conditions (Supplemental Figure 3-B). The results related to ST-pKAbA and ST-pKAbM with ST-TH-Uninv stayed significant after running the regression modeling and adding both gender and jump height as covariates with adjusting p-values, but the rest of correlations did not remain significant (Supplemental Figure 3). Greater distances in clinical tasks were associated with greater jump heights in both ST and DT jump-landing tasks. PROs showed significant correlations with ST and DT metrics for the clinical hop tests before correcting for multiple comparisons, suggesting that farther hop distances were linked to less fear (TSK11) and higher function (IKDC), however these correlations were not significant after correcting for multiple comparisons (Supplemental Figure 2).

Discussion

This study aimed to determine the potential correlation between clinical dual-task tests relative to laboratory-acquired second ACL injury risk biomechanics for cognitive-motor interference. The results of this study did not support our primary hypothesis that cognitive-motor interference (i.e., DTC) during two clinical cognitive-motor tests would be related to cognitive-motor interference of biomechanical second ACL injury predictors. However, significant, fair strength correlations existed between clinical and lab metrics when focusing on ST and DT performance in isolation, rather than the within-subject DTC. These correlations existed between clinically accessible outcomes (i.e., hop distance) and select biomechanical predictors of second and primary ACL injury. The collective results highlight that dual-task performance is a potentially more robust clinical metric to identify biomechanical risk than dual-task change metrics to describe cognitive-motor function. Additional prospective studies are needed to clarify the predictive value of clinical tests of cognitive-motor function with rehabilitation outcomes and injury risk.

The rejection of the DTC correlation hypothesis may be grounded in properties of the difference score metric. Specifically, there was a high degree of association between single- and dual-task performance (i.e., r > 0.7 for all clinical and lab test in our study, Supplemental Table 3). The stronger the correlation, the less predictable variance in change/cost measures (e.g., DTC),²⁴ leading to null correlations with other measures. Therefore, DTC measures may generally be less reliable than the ST or DT performance in isolation, especially to the extent that ST and DT performance are correlated.²⁴ Considering the clinical tasks used in this study,^{11, 12} the test-retest reliability of MSH task was ICC = 0.88 for ST-MSH, ICC = 0.86 for DT-MSH, and ICC= 0.64 for DTC-MSH, and ICC = 0.96 for ST-TH, ICC = 0.92 for DT-TH, and ICC= 0.94 for DTC-TH. A pronounced decrease in reliability of MSH measuring DTC instead of measuring the ST and DT separately. Differences in the level of difficulty of the individual cognitive and motor tasks may influence the extent that test-retest reliability is influenced for DTC metrics. More work is needed to elucidate what outcome(s) of cognitive-motor tests (e.g., DTC or dual-task performance) are most robust and insightful to maximize clinical utility. A potential, albeit unexplored, area for ACLR RTS would be considering performance on multiple visual-cognitive augmented RTS assessments (Figure 5). By isolating shared variance of DT performance across multiple tests of cognitive-motor function, the underlying construct, rather than test-specific performance, may be more robustly assessed. This approach has yielded substantial benefits in improving reliability and identifying individual differences in other fields, particularly in contrast to difference score metrics.²⁴

Figure 5:

Prospective theoretical framework for comprehensive evaluation of patient cognitive-motor function. Complementary perspectives from single-task, dual-task, and dual-task cost outcomes, each defined by composite factors determined from performance on multiple clinically tests. A. Evaluate motor capacity using various hop and strength tests (e.g., ST-MSH: traditional single-task medial side hop; ST-TH: traditional single-task triple hop, isokinetic strength), with hypothetical weightings defined by w_ST¹, etc. B. Assess ‘real-world’ performance via a composite metric defined by performance on multiple cognitive-motor tests (e.g. DT-MSH: Dual-task Medial Side Hop weighted by w_DT¹; DT-TH: Dual Task Triple Hop weighted by w_DT²; additional tests having respective weightings (w_DTⁿ)). The approach may circumvent reliability and test-specific pitfalls of relying on dual-task change/cost metrics from isolated clinical tests to ultimately be more robust in understand cognitive-motor function in rehabilitation programs. C. Probe for maladaptive visual-compensatory strategies using dual-task change outcomes across multiple cognitive-motor tests. Notably, insights afforded by B and C are likely enhanced by first ensuring adequate performance on A. Additionally, optimal sets of clinical tests and their relative weightings for each composite factor would need to be determined through future research.

Importantly, restricting assessment to even just DT performance likely provides an incomplete picture of patient function. For example, low DTC may be achieved by both the highest performers (e.g., high performance on both ST and DT conditions leading to no DT-ST change) and lowest performers (e.g., low performance on both ST and DT conditions leading to no DT-ST change), with no ability to distinguish between these scenarios if only considering a DTC outcome. Therefore, we theorize that complementary perspectives may be gained from expanding this composite approach for cognitive and motor single-task performance (i.e., assessing idealized performance capacity), cognitive-motor dual-task performance (i.e., assessing ‘real-world’ performance), and cognitive-motor interference (i.e., DTC to probe compensation strategies) to more fully evaluate patient function and inform personalized rehabilitation and RTS decisions (Figure 5). This multifactorial strategy aligns with recent RTS frameworks that emphasize the importance of integrating physical, psychological, contextual, and sport-specific factors to more fully evaluate patient readiness and risk.²⁵ Notably, considerable research is needed to determine optimal sets of tests, timing of tests, and ultimately the clinical value of this approach for achieving appreciable impact on ACLR outcomes.

There were several fair-to-moderate strength correlations between clinical variables and second ACL injury predictors in ST and DT conditions separately (Figure 4). ST-TH for both limbs and ST-MSH-Inv were correlated with Lab-ST-Inv-KAbA-range with better performance in clinical tests related to reduced risk factor of Lab-ST-Inv-KAbA-range (Figure 4-A). DT lab-based variables did not have any significant correlations with DT clinical variables after correcting for multiple comparisons (Figure 4-B). The variable response to clinical versus lab dual-task challenges (i.e., significant for ST but not for DT) may be partially due to differences in the specific DT demands elicited in the clinic versus lab scenarios. Although both clinic and lab conditions introduced visual-cognitive challenges, the lab test emphasized a reactive motor response to the directional cue in contrast to the Go/No-Go challenge of the clinical hop tests. The varying cognitive-motor response may highlight context- (e.g., task objective, movement) and domain-specific response (e.g., rapid decision-making, response inhibition) to cognitive-motor challenges. Notably, prior work has demonstrated that different cognitive tasks can elicit divergent neuromuscular and biomechanical responses. For example, Fischer et al.¹⁵ reported that visual constraints led to altered jump landing biomechanics, whereas a visual-spatial memory task did not elicit biomechanical adaptations. Additionally, the selective response to specific aspects of cognitive-motor tasks may be exacerbated in different populations (e.g., affected by neuroplasticity following ACLR). Individuals following ACLR exhibit increased postural instability during visual-cognitive dual-task conditions compared to motor dual-task or eyes-closed conditions, suggesting a heightened reliance on visual input for postural control.²⁶ Furthermore, ACLR patients demonstrated greater cognitive-motor interference than non-injured controls during drop vertical jumps combined with secondary cognitive tasks.²⁷ This interference was revealed by impaired performance in both cognitive and motor domains. Particularly, these deficits were independent of isolated cognitive ability, suggesting specific challenges in cognitive-motor integration following ACLR.²⁷ Notably, rapid decision making and inhibitory control challenges appeared to have a more salient impact on ACLR individuals’ biomechanics compared to a memory challenge.²⁸ These studies underscore the importance of considering the specific nature of cognitive challenges when assessing cognitive-motor interactions in ACLR populations. Future work is needed to determine whether more closely aligned cognitive demands would increase the strength of relationships between clinical and lab conditions, and also what the most salient dual-task demands are regarding injury risk prediction.

Shortcomings of focusing on hop distances alone may be reflected in our correlations, suggesting the need for additional metrics during jumps. Notably, jump height was a significant covariate for a number of the relationships with clinical hop distances, which corroborates the extent that hop distance may reflect overall power/strength rather than more nuanced kinematic/kinetic variables (Supplemental Figure 3). Burland et al.²⁹ showed that there was increased variability and no agreement between hop distance symmetry and knee biomechanics on a trial-by-trial basis during TH, indicating that passing clinical tests does not necessarily reflect movement quality. Therefore, it is important for clinicians to consider that hop distance symmetry and performance may not reveal underlying biomechanical issues after ACLR, and hop distance during clinical tasks is not a reliable indicator of patients’ biomechanics. Studies have demonstrated that force-sensing insoles are closely linked with knee extension moment symmetry during stop jumps in ACLR patients and may enhance the ability to diagnose neuromuscular control deficits before RTS.³⁰ Enhancing clinical assessments with practical means to capture aspects of patients’ kinematics and kinetics during the task may strengthen agreement with the kinematic/kinetic lab-based outcomes and cognitive-motor interference.

It is important to acknowledge the limitations of this study. This paper focused on specific jump landing tasks; however, these are not representative of all movements related to second ACL injury. Therefore, future studies can focus on different tasks such as single-leg landing or more similar cognitive-motor challenge tasks for both clinical and lab assessments. For the clinical tests, we focused on the hop distances, however there are some other factors that can be used in future projects such as reaction time, cognitive errors (task prioritization), and physical errors (landing stability). Furthermore, investigating the optimal timing for integrating cognitive-motor function assessments into the rehabilitation process and return to sport decision-making is an avenue worth exploring.

Conclusion

This study aimed to determine the ability of clinical dual-task tests to capture changes in dual-task changes in biomechanical risk factors during laboratory testing. Correlations were more pronounced when examining ST and DT performances separately, rather than focusing solely on within-subject dual-task changes. Our results underscore the importance of considering individual performance on clinical tests within specific conditions, particularly in relation to hop performance. Furthermore, correlations between PROs and clinical test performance highlight the potential clinical relevance of subjective experiences in the rehabilitation process.