Cognitive-Motor Dual-Task Performance of the Landing Error Scoring System

We used an observational cross-sectional study design. The independent variables were LESS condition: (1) single task (Single), (2) visual-based dual task (Visual), and (3) number-based dual task (Number). The dependent variables were raw LESS scores (sagittal plane, frontal plane, and total) and individual LESS item errors.

Participants were recruited from the University of Toledo and the local community through flyers and word of mouth. A convenience sample of 21 uninjured, physically active women was enrolled. A total of 20 uninjured, physically active women (age = 22.4 ± 2.5 years, height = 1.68 ± 0.07 m, mass = 67.0 ± 13.8 kg, Tegner Activity Scale = 5.9 ± 1.1) participated, and 1 woman withdrew for personal reasons. Volunteers were eligible if they were between 18 and 30 years of age, had no history of lower extremity surgery, and self-rated their activity level as ≥5 on the Tegner Activity Scale. Exclusion criteria included a history of lower extremity injury in the 12 months before the study, stroke, migraines, severe head injury (including diagnosed concussion) in the 12 months before the study, personal or family history of neurological disorders (eg, seizures and epilepsy), current neuropathy, implanted biomedical device, cognitive/learning disorder (attention-deficit/hyperactivity disorder, dyslexia, anxiety, etc), cardiopulmonary disorder, or current use of medication (stimulants or depressants) likely to influence cognition. Individuals who met all eligibility criteria were scheduled for data collection. Data were collected on the second day of a larger 3-day study with at least 48 hours between days 1 and 2 to reduce any potential effects from the transcranial magnetic stimulation that was used on day 1. All participants provided written informed consent, and the study was approved by the University of Toledo Biomedical Institutional Review Board.

All participants followed the same testing procedures during 1 study session. The Single condition was performed first, followed by the dual-task conditions (Visual and Number) in a counterbalanced fashion. The dominant limb of each participant was operationalized as the test limb and was the only limb used to quantify the landing errors.⁴  The dominant limb was defined as the leg that would be used to kick a ball as far as possible. Two-dimensional video cameras (iPod, Generation 7, Apple, Inc) with a frame rate of 30 frames per second were placed 3.45 m in front of and to the right of the landing area at a height of 1.21 m to capture sagittal- and frontal-plane views of the landing task. The LESS procedures were based on previously described methods.⁴  The following standardized instructions were given: “Jump forward off of the box to reach the designated landing area with both feet leaving the box then landing at the same time. Immediately after landing, jump straight up as high as you can. This should be performed in a fluid motion without any pauses in between the 2 jumps.” Participants were asked to perform a minimum of 3 practice trials before the trials were recorded to ensure that the task was being executed correctly. A failed trial occurred if participants’ feet were not fully inside the marked landing area or if they did not perform the vertical jump. The LESS procedures were repeated for each condition (Single, Visual, and Number) until 5 successful trials were completed. Motor accuracy was calculated as the percentage of successful trials out of all trials performed for each condition. At the conclusion of each condition (Single, Visual, and Number), participants were given the National Aeronautics and Space Administration Task Load Index (NASA-TLX) to rate their cognitive workload relative to mental, physical, and temporal demands as well as perceived performance, effort, and frustration.¹⁷  Participants used a sliding scale from 0 (very low demand) to 100 (very high demand) to rate each subcategory of the NASA-TLX to identify which aspects of cognitive function may be influencing dual-task performance during each condition. Participants were given at least 2 minutes of rest between each of the conditions.

The LESS procedures were repeated for the Visual condition, but participants were given a visual-based dual task to perform simultaneously with the LESS. A 76.2-cm monitor was placed 3.78 m in front of the landing area to visually present each task option via PowerPoint 2021 (Microsoft Corporation) presentation. Participants began the task on the 30-cm box while standing with the heel of their preferred foot on a trigger. They were instructed to look at the monitor and maintain focus on the middle of the screen throughout the entire landing task to limit visual feedback and increase cognitive demands during execution of the movement. When the trigger was released, a random series of 5 numbers was displayed on the center of the screen for 250 milliseconds, which is the minimum amount of time to see and process visual information.¹⁸  Participants were told to memorize these numbers and to report them immediately after completing the motor task in the order displayed. A minimum of 3 practice trials was permitted to familiarize participants with the cognitive task. This task was considered successful when participants correctly recalled and reported the 5 numbers in the same order that they were displayed. Cognitive accuracy was calculated as the percentage of successful cognitive trials out of all completed trials. Dual-task accuracy was determined by counting the total number of trials where both cognitive and motor tasks were performed successfully and then calculated as the percentage of successful dual-task trials out of all completed trials. Only trials with a successful motor performance, regardless of cognitive accuracy, were eligible for analysis. Participants were blinded to the criteria used to determine which trials were included in the analysis to prevent them from prioritizing one task over the other. The mean difference (MD) of sagittal-plane, frontal-plane, and total LESS errors between the dual-task conditions and the Single condition (eg, Visual sagittal errors – Single sagittal errors; Number sagittal errors – Single sagittal errors) was used to define dual-task cost.

For the Number condition, an investigator (J.L.R.) called out a random number to participants before they performed the LESS. Participants were instructed to subtract 7 in their head from the number stated by the investigator and provide the answer orally before initiating the landing task and to continue the subtraction task as quickly and accurately as possible while simultaneously performing the landing task.¹¹  Trials were recorded if participants were able to orally recite the answer for at least 2 subtraction problems during the landing task, which also constituted cognitive accuracy. Motor and dual-task accuracy were calculated as described above. Only trials with successful motor performance, regardless of cognitive accuracy, were eligible for analysis.

An LESS scoring sheet was used to guide the investigator through scoring the landing task, as previously described, with higher scores corresponding to more movement errors.⁴  Videos of the trials were uploaded to KiNovea (version 0.9.5; KiNovea) to measure joint angles and more accurately identify landing errors.¹⁹  In the sagittal plane, the angle tool was used to measure knee angles at initial contact and peak knee angle to determine the magnitude of displacement. This was done for 2 items that required specific cutoffs for knee-flexion angles at initial contact (>30°) and at the point of maximum knee flexion (>45°). In the frontal plane, the line tool was used to measure true vertical using a background reference. This line was used to determine patellar position, foot rotation, stance width, and trunk lateral flexion. All videos were analyzed frame by frame to ensure accuracy in scoring of LESS criteria. All trials were scored by a single rater from the investigative team (M.M.) who was trained by another member of the study team (J.L.R.) who had >6 years of experience implementing and scoring the LESS. If the rater was unsure about the presence of an error for any item, a second member of the study team (G.N.) reviewed the trial. Sagittal-plane, frontal-plane, and total LESS scores were calculated for each condition.

A sample size estimate of 20 participants was calculated using previously reported MD values of total LESS scores between dual-task conditions (0.70 ± 1.91) with an α level of .05 and power of 0.80.¹⁵  Data normality was assessed using the Shapiro-Wilk test. Descriptive statistics were calculated for participant characteristics. Separate 1-way analyses of variance with repeated measures were used to compare sagittal-plane, frontal-plane, and total LESS scores, as well as NASA-TLX subscales, between conditions. We used dependent t tests with post hoc Tukey correction to determine which conditions differed. Cohen d effect sizes with 95% confidence intervals (CIs) were used to determine the magnitude of the differences. MDs with 95% CIs were calculated to describe the absolute differences between conditions. To compare LESS item-specific errors between conditions, we used χ² analysis. Considering landing errors were likely to differ among our sample, we used bivariate correlation coefficients to investigate the relationships between Single LESS scores and changes in LESS scores for each of the dual-task conditions to determine whether baseline performance affected dual-task cost. All analyses were evaluated at an α level of .05 and conducted using the Statistical Package for the Social Sciences (version 29; IBM Corporation).

To further supplement the findings of the primary analysis, the numbers of participants with increases in LESS errors that exceeded the standard error of measurement (SEM) and minimal detectable change (MDC) values established in previous literature were computed.²⁰ 

All data were normally distributed (P > .05). Comparisons of LESS scores between conditions are presented in Table 1. The accuracy of motor, cognitive, and dual-task performance under each condition is reported in Table 2. Comparisons of NASA-TLX scores between conditions are shown in Table 3.

Sagittal-plane (F_2,57 = 4.167, P = .02) and total (F_2,57 = 4.829, P = .01) LESS scores differed between conditions. The Visual condition resulted in large increases in sagittal-plane errors (MD = 1.07 errors; 95% CI, 0.32–1.82 errors; P = .02, d = 0.91; 95% CI, 0.26–1.56) and total errors (MD = 1.67 errors; 95% CI, 0.64–2.70 errors; P = .008, d = 1.03; 95% CI, 0.37–1.69) compared with the Single condition. The sagittal-plane, frontal-plane, and total errors in the Number condition did not differ from those in the Single or Visual condition (P > .05).

For the Visual condition, 17 (85%) and 11 (55%) participants had increases in LESS errors that exceeded the SEM and MDC values, respectively, established in previous literature.²⁰  For the Number condition, 11 (55%) and 7 (35%) participants had increases in LESS errors exceeding the SEM and MDC for total LESS errors, respectively, that were established by previous literature.²⁰ 

The proportion of errors committed for individual sagittal-plane (Figure 1) and frontal-plane (Figure 2) items did not differ between conditions (P > .05; Table 4).

Scatter plots representing Single condition scores (baseline) and dual-task cost for each condition can be found in Figure 3A and B. Greater sagittal-plane errors at baseline were associated with less dual-task cost during the Number (r = −0.486, P = .03) and Visual (r = −0.490, P = .03) conditions. Greater frontal-plane errors at baseline were associated with less dual-task cost during the Number (r = −0.504, P = .02) and Visual (r = −0.606, P = .005) conditions. Greater total errors at baseline were associated with less dual-task cost during the Visual condition (r = −0.580, P = .007).

Our purpose was to expand the scope of dual-task research to include a clinically accessible means of assessing dual-task cost using the LESS. Our findings support our primary hypothesis that the LESS would be able to detect a dual-task cost with the addition of multimodal cognitive loads. The addition of a visual-based cognitive load increased sagittal-plane (MD = 1.07; 95% CI, 0.32–1.82) and total (MD = 1.67; 95% CI, 0.64–2.70) LESS errors, but the number-based cognitive task did not. Although we observed changes in landing errors between conditions, a clear pattern of item-specific adaptations did not emerge. Although additional work is warranted to elucidate the movement-related adaptations unique to different forms of cognitive-motor interference, our findings provide initial evidence supporting a clinical model of dual-task assessment using the LESS.

To date, only 1 other study has implemented cognitive loads with the LESS, and the researchers did not observe changes in LESS scores despite a decline in cognitive performance.¹⁵  A potential reason that our findings may have differed could be that we studied a female cohort rather than both men and women. In individuals with ACL reconstruction and healthy active individuals, women perform the LESS with more landing errors related to medial knee displacement and more total errors than their male counterparts, further highlighting injury-risk profiles for women.^16,21  Therefore, a strength of this study is the primary focus on women given their increased risk for ACL injury, and LESS differences may not be accounted for when men and women are pooled together. However, future research comparing dual-task cost between men and women is still needed to assay the effects of dual tasks between sexes and better understand if the increased injury risk in women is exacerbated by increased cognitive loading.

Another major difference between our study and previous research was that different types of cognitive tasks were used.¹⁵  Biese et al expressed that limited difficulty of the cognitive tasks could have played a role in their findings.¹⁵  In our study, we recorded participants’ cognitive workload via the NASA-TLX for each condition to determine how the perceived difficulty of each task might have influenced their landing errors. Participants rated the Number and Visual conditions as having greater cognitive workload across most of the categories than the Single condition (Table 3). The increased cognitive load during the dual tasks may have exacerbated the landing errors, as they required a high cognitive demand and were challenging enough to divide the participants’ attention away from the landing task. Although the Visual condition was not perceived to be as cognitively challenging as the Number condition, the added complexity of removing visual feedback from the landing area likely increased landing errors. Baseline LESS scores in our study were slightly lower than previously reported (4.90 ± 1.68 versus 5.5 ± 1.66).¹⁵  Our findings suggested that fewer landing errors at baseline leave opportunity for greater change in LESS scores when a cognitive load is added. This appears to highlight the need for clinicians to be cognizant of individuals’ abilities to attenuate the influence of cognitive loads while maintaining appropriate neuromuscular control, which may aid in identifying those who could benefit from cognitive-motor training.

The addition of a visual constraint to the landing also induced dual-task costs in both sagittal-plane and total LESS scores. Similar changes in motor performance caused by a visual dual task and visual occlusion have been reported.^13,22  During visual dual tasks, knee-flexion angles and dynamic postural stability are reduced, and knee-abduction angles are increased during jump landings in recreationally active individuals.^13,23,24  Concurring with our results, visual distraction away from the landing tasks increases biomechanical errors associated with stiffer landing patterns. Given that vision is a necessary sensory feedback modality to appropriately plan and execute movement, reducing visual input requires individuals to rely on proprioceptive and vestibular systems to successfully perform the motor task. Future interventions, including external focus of attention in the already established LESS and landing training protocols, can improve landing mechanics in physically active females as external focus allows for the motor system to self-organize without reliance of visual attention solely on the landing pattern.^24–26 

Performance of the Visual condition increased mental demand and effort compared with the Single condition based on the NASA-TLX. This suggests that visual distraction sufficiently increased cognitive loading enough to induce landing errors in our sample. Although the Visual condition was not as cognitively challenging as the Number condition, 17 (85%) participants and 11 (55%) participants had increases in LESS errors that exceeded the SEM and the MDC values, respectively, reported in a systematic review of the reliability of the LESS (range, 0.19–0.52 and 0.53–1.44, respectively).²⁰  The errors exceeding these values would indicate a meaningful increase in errors during Visual dual tasks and further illustrate the importance of visual feedback relative to lower extremity injury risk. Training landing mechanics while restricting visual feedback may be preferred for future neuromuscular training programs, as sport and physical activity do not permit individuals to constantly be looking at the ground or their lower extremities during movement and encourages individuals to use visual feedback from their environment to inform decision-making. Future research should be done using visual distractions to better understand the role of vision and decision-making in sport relative to injury risk.

The number-based task induced increases in LESS errors, but the increases were not statistically different. This finding is in line with our hypothesis because the Number condition provoked more landing errors but not to the same magnitude as the Visual condition. Number-based dual tasks have been reported to change both sagittal- and frontal-plane biomechanics in uninjured men and women. Dai et al reported increased ground reaction forces and decreased knee flexion and jump height because of backward counting of 1s and 7s.⁶  Although the increases were not statistically different, our results may suggest that when attention is split between the number-based task and the jump landing, some individuals produce landing errors in multiple planes. Eleven (55%) and 7 (35%) participants exceeded the SEM and MDC for total LESS errors, respectively. It is possible that the number-based task exceeded the threshold of cognitive functioning in a portion of our sample, and, therefore, meaningful increases in motor errors occurred as attention was split during the dual task, possibly increasing their risk of injury.

During the Number condition, participants reported the highest degree of mental demand, effort, and frustration but the lowest performance demand compared with the Single condition. Interestingly, their perception of performance demand did not appear to translate entirely to landing errors committed given the lack of statistical differences. The discrepancy between perceptions and performance likely speaks to participants’ decisions to prioritize one task over the other. Task prioritization has been a consistent theme in the dual-task literature, as many researchers have described a phenomenon where individuals often sacrifice cognitive performance to preserve motor performance.^15,27,28  Given that the subtraction task was frustrating and required much effort and mental demand, participants may not have cared about the accuracy of the cognitive task and ranked it lower on the NASA-TLX. This seems to be supported, as cognitive and dual-task performance were the lowest in the Number condition and motor accuracy was relatively preserved. However, it is possible that some participants were concerned about successfully and accurately completing the subtraction task, and their prioritization of this cognitive task increased their LESS errors.

Although no differences existed in landing errors between conditions based on individual items, several approached statistical significance (P ≤ .10) and exceeded a 20% increase in errors from baseline. Most of these errors occurred in the sagittal plane (less trunk flexion at initial contact, less trunk flexion displacement, and less overall joint displacement) and included a more medially positioned patella at initial contact. These items had the greatest proportion of errors during the Visual condition, suggesting that a subset of our sample performed stiffer landings with more knee valgus when completing the LESS with a constrained visual field. Although not statistically different, traditional LESS scoring appears to be able to identify the dual-task cost of landing errors in both sagittal and frontal planes during the Number and Visual conditions. Variability in item-specific findings may suggest that women respond to cognitive loads differently, supporting the need for a more individualized approach to identify injury risk using dual-task LESS assessments.

Individuals with lower LESS errors during the Single condition demonstrated greater dual-task cost in the Number and Visual conditions. The inverse relationship observed indicates that those who performed the Single condition with minimal errors have a greater capacity for errors when dual tasks are incorporated into the landing assessment. Alternatively, those who performed the LESS with more errors during the Single condition likely could not perform the landing much worse, and, therefore, the magnitude of the dual-task cost is not as large. These relationships support the need for implementation of dual tasks in clinical assessments to highlight those demonstrating risk factors for lower extremity injury. Although identifying gross motor errors with the LESS is easy, an additional cognitive task may better represent real landing performance when attention is not fully dedicated to the landing task. Therefore, finding easily implemented and low-cost cognitive challenges, similar to the number- and visual-based dual tasks in the present study, is imperative for providing clinicians with better injury-risk screenings for tasks that better represent sport performance where high cognitive demands are present.

Both cognitive tasks used in this study were able to degrade motor performance measured via LESS scoring. The challenge of replicating a visual-based dual-task assessment similar to ours in the clinic is that it does require the use of equipment (eg, foot switch, computer, and television screen), which could limit its accessibility. However, other unanticipated visual stimuli using different-colored cones or having a clinician pointing in different directions can be easily implemented. The number-based task may be a simpler, yet challenging, task that could easily be used in a clinical setting. Both presented tasks have strong preliminary evidence for use in clinical dual-task assessment.

We recommend that clinicians include dual-task paradigms in already established injury-risk screenings that are performed before the start of the athlete’s season. Landing assessments are a key part of injury-risk screenings, but some individuals at greater injury risk may be overlooked due to the traditional screening being anticipated and lacking additional cognitive loads. As demonstrated in our study, dual-task paradigms may exacerbate landing errors, highlighting those at greater risk of injury during landings, and can identify potential candidates for further neuromuscular training.

Our study had several limitations. We only observed female participants, and thus our study should only contribute to the current knowledge of female biomechanics and is not generalizable to men. Another limitation was that we were unaware of participants’ baseline neurocognitive function. Those with worse baseline neurocognitive scores have demonstrated landing biomechanics associated with ACL injury.^10,29  Variable baseline neurocognition may have affected the magnitude of dual-task cost during the LESS conditions. We also only analyzed trials where the motor task was performed successfully regardless of cognitive accuracy. It is possible that in scored trials where both tasks were performed successfully, the landing error results could have changed, as participants would not have underperformed at any aspect of the dual-task assessment. However, we did not tell our participants which trials would be analyzed to avoid prioritization of either the cognitive or motor task. We also did not randomize the order of the LESS conditions. The Single condition served as the participants’ familiarization to the task, so we only counterbalanced the dual-task conditions to reduce the influence of an order effect. However, by not randomizing all 3 conditions, we could have introduced a learning effect for the landing task. We also controlled participants’ gaze only during the Visual condition. To strengthen this assessment, future research should be done to consider another condition where participants complete a visual dual-task condition with and without gaze fixation to investigate the changes more elegantly in landing errors as multistep cognitive processing is gradually induced. Finally, some of the LESS items did not meet all assumptions for the χ² analysis. However, even for the items that did meet all assumptions, our study was likely underpowered to demonstrate a real difference between conditions, and a larger sample would be needed for future item analyses.

The LESS is a viable assessment to combine with either a number-based or visual distraction to assess dual-task performance in uninjured, physically active women. The Visual condition used in this study affected unique aspects of landing patterns in the sagittal plane as well as total landing errors. Although summed LESS scores differed between conditions, analysis of individual LESS items suggests that women adapt to cognitive loading differently, supporting a need for individualized assessment. Collectively, our findings support the use of the LESS as a clinical model of dual-task performance in physically active women.

This work was supported by the Gordy Graham Research Assistance Award from the Great Lakes Athletic Trainers’ Association (G.N. and Emma Nicholson). We acknowledge Emma Nicholson, MS, AT, ATC, for her efforts with recruitment and data collection for this project.