Comparison of Limb Loading Characteristics and Subjective Functional Outcomes Between Sexes After Anterior Cruciate Ligament Reconstruction

This observational cohort study was a part of a larger point-of-care, collaborative research program in a single academic health system. Our study was conducted in a controlled university laboratory setting and approved by the university’s institutional review board for health science research. For our initial aim, the independent variable was sex (male versus female), and the dependent variable was normalized peak force limb symmetry index (LSI). For our second aim, the independent variables were limb (ACLR versus contralateral) and sex (male versus female), and the dependent variables were normalized peak force (N/kg) and unilateral cumulative load (UCL) expressed as a percentage.⁷  For our third aim, the variables of interest were the scores from the PRO measures, normalized peak force (N/kg), UCL (%), and the absolute value of LSI that was calculated from normalized peak force recorded during the bilateral bodyweight squat. Sample size was based on identifying sex differences in limb loading using data collected in our lab as a variability estimate in limb loading. We determined that 104 (52 male and 52 female) patients were necessary to identify a moderate effect (Cohen d = 0.50) between sexes with an α ≤ .05 and power (1 − β) of .80.

A total of 142 patients (50% female) volunteered to participate and were enrolled after providing written informed consent (Table 1). All patients attended their study session before clearance to RTA by their orthopaedic surgeon. We included patients if they had a primary, uncomplicated, unilateral, isolated ACLR. We excluded patients if they had a history of prior ACLR, multiligament reconstruction, contralateral ACL injury, or graft failure; had sustained other lower extremity injuries or concussions within 6 months of study participation or at any time throughout the study; or had a history of neurological disorders.

We measured limb loading during a bilateral bodyweight squat and PRO scores during a single visit. Limb dominance was also recorded and self-defined as the limb that a patient preferred using when kicking a soccer ball.

Before starting data collection, patients were instructed to perform a standardized calibration protocol from the manufacturer. Patients were weighed, stood atop a pressure mat (SB Mat; Tekscan Inc), and then their mass was converted from arbitrary units to newtons. Once the pressure mat was calibrated to the patient’s mass, we instructed patients to complete 3 sets of 3 repetitions of a squatting task with 30 seconds of rest between trials. To complete the squatting task, we instructed patients to stand with their feet shoulder-width apart and to perform a squat with approximately 90° of knee flexion or to the “height of a chair” and to return to their upright standing position (Figure 1) at a pace set by a metronome (40 beats per minute, in which patients were to be at maximum knee flexion at the bottom of the squat or at maximum knee extension at the top of the squat for each beep of the metronome). We provided patients with the opportunity to practice the squatting protocol until they felt comfortable performing the task. Vertical force was continuously measured from each limb independently using a pressure mat (SB Mat) and software (v 7.10-14; FootMat Research) sampled at 60 Hz during data collection.

Participants also completed a series of PRO measures during the visit. The International Knee Documentation Committee Subjective Knee Form (IKDC) was recorded, determining subjective knee function.²⁷  The Knee Injury and Osteoarthritis Outcome Score (KOOS; subscales include Symptom, Pain, ADLs, Sport, and Quality of Life) assessed the severity of the knee symptoms and functional disabilities experienced by the patient.²⁸  The Anterior Cruciate Ligament Return to Sport After Injury (ACL-RSI) was collected to assess the patient’s confidence, risk appraisal, and emotions related to resuming sport-related activities.²⁹  The Tegner Activity Scale (TAS) was recorded to determine patients’ perceived activity level preinjury and in their current state during their lab visit based on work and sports activities.³⁰  All PRO measures have been found to be valid and reliable measures of their respective constructs.^27–31 

All data were processed using a custom MATLAB code (v 9.12.0, MATLAB R2022a; The MathWorks Inc).

We calculated peak force (N) individually for each limb by recording the largest single force output over each of the 3 sets, and then the average of those values across the 3 sets was calculated. Once the peak load was calculated, we then normalized to each patient’s body mass (N/kg).

The UCL is a novel measure that signifies the overall loading contribution of each limb throughout the entire squat trial. We calculated UCL as a percentage of each individual limb’s vertical force production throughout the squat trial, then averaged across the 3 squatting trials. Figure 2 is an example of how the UCL was obtained. The bold light gray line representing the ACLR limb indicates lower force values compared with the bold dark gray line representing the contralateral limb. Therefore, the UCL for the ACLR limb for this trial would be smaller than the UCL of the contralateral limb. If the force lines from each limb were perfectly positioned on top of each other, this would indicate a UCL of 50%. Stated again, a UCL of 50% would mean that each limb is contributing the same amount of force throughout the trial and that the patient is equally distributing their body mass between the ACLR and contralateral limbs.

We calculated LSI by dividing the peak force produced by the ACLR limb by that of the contralateral limb as a percentage, then averaged across the 3 trials. An LSI value of 100% is indicative of the ACLR limb producing the same or equal peak force value as the contralateral limb. We then calculated the absolute value of the LSI by conducting 1 − |LSI_PeakForce|. An absolute value LSI score of 0 is indicative of equal (symmetrical) peak force from each limb. An absolute value LSI score greater than zero is indicative of higher peak force LSI from the ACLR limb, indicating the patient places more of their body mass on their ACLR limb during the squatting task. An absolute value LSI score less than zero is indicative of higher loading contributions from the contralateral limb, indicating the patient is offloading their ACLR limb by putting more of their body mass on their contralateral limb during the squatting task.

An initial analysis was conducted using an independent-samples t test to compare normalized limb loading peak force LSI between males and females. For our second analysis, we evaluated the influence of sex on bilateral loading during squat tasks using a 2 × 2 (limb-by-sex) analysis of covariance while covarying for limb dominance (ie, whether the surgical limb was self-identified as the dominant or nondominant limb). Post hoc t tests were performed as appropriate. We conducted paired-samples t tests to compare normalized peak force limb loading values and UCL values between females and males and between the ACLR and contralateral limbs. We calculated and used Cohen d effect sizes to interpret pooled standardized mean differences, which are representative of the magnitude of observed differences. We classified effect size values as small, ≤0.29; weak, 0.30–0.49; moderate; 0.50–0.79; or strong, >0.80.³² 

The PRO data did not meet the assumptions of a normally distributed dataset; therefore, for our third analysis, we evaluated the relationship between PROs and limb loading metrics using a Spearman ρ correlation coefficient. We interpreted correlation coefficients as weak, ≤0.35; moderate, 0.36–0.67; or strong, 0.68–1.00.³³  Positive correlations indicate that, as an individual increases their limb loading or symmetry value, their PRO measures also improve. Negative correlations indicate that, as an individual offloads their limb or decreases their limb symmetry, their PRO outcomes improve. Ceiling or floor effects were evaluated and considered present if more than 15% of patients reported the best (ceiling) or worst (floor) possible score for each PRO.^34,35 

We used SPSS (v 28.0; IBM SPSS) for all statistical calculations. All analyses were performed with α ≤ .05 determined a priori.

Of our 142 participants, 91 (64.1%) exhibited peak force LSI values less than 1.0, which would indicate the majority of participants, who were on average 5.2 months postsurgery, were offloading their ACLR limb. For those individuals who had an LSI value less than 1.0, the mean ± SD LSI value for peak force was 76.8% ± 17.1%. Fifty-one individuals (35.9%) had an LSI value for peak loading greater than or equal to 1.0, indicating greater loading of the ACLR limb compared with the contralateral limb during the squatting task (Figure 3). Of our 51 individuals who exhibited an LSI value greater than or equal to 1.0, the mean ± SD LSI value for peak force was 127.3% ± 21.4%. Across males, 25 individuals (35.2%) had an LSI value greater than or equal to 1, and 46 (64.8%) had an LSI value less than 1.0. For females, 26 (36.6%) had an LSI value greater than or equal to 1.0, whereas 45 (63.4%) had an LSI value less than 1.0. We observed no significant differences between male and female peak force LSI values (t = 1.54, P = .06, Cohen d = 0.26; males, X̄ = 0.99, σ = 0.33; females, X̄ = 0.91, σ = 0.28).

We observed a significant limb-by-sex interaction for normalized limb loading peak force (F_1,139 = 5.71, P = .02; Figure 4). Post hoc analysis revealed females offloaded their ACLR limb compared with their contralateral limb (t = 3.55, P < .001, Cohen d = 0.42; Table 2). Males, in general, offloaded their ACLR limb compared with their contralateral limb to a lesser extent than females (t = 1.49, P = .07, Cohen d = 0.18; Table 2).

We did not observe a limb-by-sex interaction for UCL (F_1,139 = 2.48, P = .12). However, there was a main effect for limb, wherein the UCL of the ACLR limb (48.8% ± 10.0%) was loaded significantly less than the contralateral limb (51.2% ± 9.0%; F_1,139 = 11.84, P < .001, Cohen d = 0.22). There was no difference in UCL across sexes (F_1,139 = 0.032, P = .86).

Descriptive statistics for limb loading metrics and all PRO measures are found in Table 3. We observed weak significant negative relationships between the contralateral limb for normalized limb loading peak force and UCL and PROs (P < .05; Table 4). There were weak significant positive relationships between the ACLR limb for normalized limb loading peak force and UCL and PROs (P < .05; Table 4). Ceiling effects, defined as >15% of patients scoring the best possible score, were observed for only the following PRO measures: KOOS Pain (18.3% of patients scored 100), KOOS ADLs (36.6% of patients scored 100), KOOS Sport (15.5% of patients scored 100), and TAS preinjury (22.5% of patients scored a 10). No floor effects were observed for any PRO measures.

The purpose of our study was to examine limb loading symmetry during a bilateral bodyweight squatting task between sexes, compare side-to-side limb loading metrics between sexes, and explore the presence of a relationship with perceived function in patients recovering from unilateral ACLR. We observed that patients approximately 5 months post-ACLR, on average, offloaded their ACLR limb compared with their contralateral limb and that females offloaded their ACLR limb more than males during the bilateral bodyweight squatting task. Our findings align with previously conducted studies in which patients after ACLR offloaded their surgical limb compared with their contralateral limb during tasks such as a sit-to-stand, squat, and stop jump.^8,9,36  Our study builds upon previous findings by highlighting a unique offloading adaptation from females after ACLR compared with their male counterparts. Our findings could provide clinicians insights into how to optimize or potentially better tailor their patients’ rehabilitation protocol based on their patients’ sex.

The differences in our findings for the normalized peak force and UCL metrics were unexpected. Our hypothesis was that there would be significant difference across both metrics. However, there were significant differences for the normalized peak force metric and no significant differences for the UCL metric (Table 2). These results indicate that it is possible that the differences between limb loading metrics may be exaggerated when looking at a singular time point such as a single peak. The UCL metric, though not statistically significant in this sample, demonstrated patients’ tendency to offload their ACLR limb compared with their contralateral limb with a small effect size (Table 2). The UCL metric is a unique comprehensive measure that allows for analysis in differences across all parts of the squatting trial, whereas the peak force metric is only a single moment in time. The lack of significance in the UCL metric potentially indicates that not all phases of the bilateral squat are asymmetric. It is also possible that the UCL metric may indicate that the variation in side-to-side loading or patients shifting their weight may be accounting for the variation in the time-course data. More research needs to be done with motion-capture analysis to determine which phases of the squat patients are offloading their ACLR limb.

We also observed that patients, on average, offloaded their ACLR limb compared with their contralateral limb during a bilateral squatting task, which aligns with previously conducted research.^7–9,36  The magnitudes of differences between limb loading discrepancies for the ACLR versus the contralateral limb in previously conducted studies were larger than the ones observed in this study.^7–9,36  This difference in magnitude could be due to differences in bilateral bodyweight squatting methodologies (eg, metronome use, tactile feedback, repetition scheme) and instrumentation (eg, motion capture, force plates, standardized footwear); however, the results are in congruence. Even though there was a difference in the magnitude of differences found compared with previous studies, patients participating in the current study were instructed to complete an inherently symmetric task, yet they still executed the task asymmetrically. The offloading from the ACLR limb observed in individuals during an easy and safe bilateral squatting activity is cause for concern. If this movement pattern adaptation is not corrected, this could perpetuate long-term when patients are participating in more dynamic and ballistic activities like running, jumping, and cutting maneuvers.^6,37  Poor biomechanics, specifically offloading the ACL limb, during these high-demand activities could increase one’s risk for reinjury, and, when compounded over time, could catalyze the progression toward the development of early onset knee osteoarthritis.^4,8,38 

It is also important to note that 51 of the 142 participants (36%) had an LSI value of 1.0 or greater. It is currently unclear which physical or psychological characteristics these patients might have had that allowed them to evenly distribute or overload their ACLR limb during this squatting task. Authors of a recent study found that under 3 conditions, natural, instructed, and feedback, patients after ACLR went from being asymmetric in their loading during the natural condition to symmetrically loaded with the instructed and feedback conditions.⁷  This finding indicates that, with the proper training and instruction, individuals after ACLR can evenly distribute their weight across their limbs. Authors of another previously conducted study found that time was a factor contributing to loading distribution: individuals who were >24 months post-ACLR overloaded their surgical limb compared with their contralateral limb.²⁷  One additional possible explanation could be that the patients in the current study were in a controlled laboratory setting doing a variety of lower extremity assessments on their surgical limb that could potentially have altered their focus during the task and changed their natural kinematics by increasing the loading on their ACLR limb.

To our knowledge, our study is the first to examine the effect of sex on limb loading metrics during a bilateral bodyweight squat in patients recovering from ACLR. We observed that females offloaded their ACLR limb more and were more asymmetric during the bilateral squatting task compared with their male counterparts. Our results align with those of related studies whose authors investigated the influence of biological sex on 3D kinematic and kinetic variables during a variety of tasks such as walking and jump landings.^15–17  Results of these studies, evaluating explosive tasks, suggested that females had worse movement quality and decreased muscle activity compared with their male counterparts after ACLR.^15,16  However, during a mild walking task the opposite conclusion was made. Males with a noncontact mechanism of injury were found to offload their ACLR limb more than their female counterparts and males who sustained a contact mechanism of injury.³⁹  The difference between the explosive tasks and walking indicates that an intermediary task such as a bilateral squatting task could be an appropriate area of intervention. Whereas a squat is not as physically demanding as a jump-landing explosive task, it can be more demanding than leisure walking when progressed appropriately. Given the nature of a bilateral squatting task, it could have utility in highlighting loading asymmetries between sexes. Additionally, it is possible that females after ACLR cope differently in their biomechanical movement patterns compared with their male counterparts. In our study, the presence of sex differences for limb loading metrics at approximately 5 months post-ACLR indicates that clinicians may intervene during structured ACLR rehabilitation. The impact of the observed sex differences in limb loading on postoperative rehabilitation, RTA decision-making, and reinjury risk is an area of future research.

Weak relationships found between limb loading metrics and PROs demonstrate that as a patient becomes more symmetric their subjective knee function improves. One significant relationship was observed for the normalized peak force variable: for patients who applied less force on the contralateral limb, their KOOS Symptom score improved. This finding is similar to those of previous researchers who found that individuals after ACLR who were considered symptomatic via KOOS scores offloaded their ACLR limb compared with their contralateral limb during a walking task.^25,27  We also observed that as the contralateral limb UCL decreased and the ACLR limb UCL increased, KOOS Symptom and ADLs subscale scores improved. The KOOS ADLs subscale questions address the degrees of difficulty when performing everyday tasks (eg, rising from sitting, getting in or out of a car, getting on or off the toilet) that frequently mimic the motion of a bodyweight bilateral squat.²⁸  These relationships observed potentially indicate that regardless of limb, patients who load their limbs more unevenly have a lower perception of their ability to complete ADLs. A similar relationship was observed in which individuals who had greater normalized limb loading peak force from their ACLR limb were currently participating in higher levels of physical activity as determined by the TAS. The only relationship between any loading metric and the ACL-RSI was a negative relationship between the ACL-RSI score and the normalized peak force from the contralateral limb. This is somewhat contradictory to the finding of a previous study in which no relationship was observed between limb loading during walking and kinesiophobia as determined by the Tampa Scale for Kinesiophobia.⁴⁰  This is likely due to the difference in limb loading requirements during gait compared with a squatting task. Ceiling effects for the KOOS Pain, KOOS ADLs, KOOS Sport, and TAS preinjury were observed and should be taken into consideration.

A bilateral bodyweight squat is a commonly used exercise during rehabilitation sessions and when performing ADLs. The ability to measure limb loading objectively and precisely during a commonly performed movement (ie, bilateral squat) gives clinicians tools to practice personalized medicine. All PRO measures used in the current study are freely available online for clinical use. The implementation of PRO measures could give clinicians insight as to their patients’ readiness to RTA but also could be used as an indicator of the potential adoption of asymmetric limb loading movement patterns that may need intervention. Personalizing a patient’s rehabilitation protocol to fit their individualized needs using a translatable clinical measure allows for more targeted interventions and ultimately better outcomes. During traditional RTA testing batteries, clinicians commonly use strength measures, functional tasks, and subjective functional PROs.¹  Few researchers have incorporated a squatting task variation in an RTA protocol, and few have used a bilateral bodyweight squat in their arsenal of tests.^1,41  Researchers have suggested that the optimal time for individuals to RTA can be anywhere between 9 and 12 months after surgery.^1,41  The average time after ACLR for the individuals in our study was approximately 5 months, indicating that patients may need to be reexamined closer to the time of their clearance to RTA. It is currently unclear whether these loading asymmetries persist past 5 months after ACLR. The evidence provided in our study highlights the need to further investigate the influence of limb loading metrics after ACLR during serial assessment and explore the feasibility of implementing a squatting task into an RTA protocol.

Participants in this study underwent rehabilitation with their preferred clinician; therefore, we did not control for postoperative rehabilitation. Additionally, rehabilitation plans were not recorded, nor was compliance in completing their recommended protocol, although all patients were given the same postoperative general protocol by their surgeons. Our study required 1 visit to the laboratory at a time point when rehabilitation may not have been complete; therefore, the findings of this study should be considered an interim functional analysis as patients are recovering from ACLR. Given the cross-sectional study design, it is possible that patients had differences in their limb loading symmetries before injury. A single testing session may not be representative of overall patient function after ACLR; the importance of repeat testing to track progress, evaluating the efficacy of exercise interventions, and making informed health care decisions is immeasurable.

Throughout the testing session we standardized squat pace using a metronome, which could have influenced preferred squatting pace and may be less generalizable to functional scenarios during activity or sport. However, it has been observed that there is a moderate to strong relationship between the biomechanical movement profiles during a bilateral squatting task and a drop-landing task in individuals after ACLR.⁸  This relationship suggests that movement patterns during a squatting task could carry over during a high-demand dynamic jump-landing task, indicating that a squatting task could be a useful proxy to evaluate the quality of movement patterns in a safe manner. Additionally, a limitation of this study was a lack of kinematic data collection. The integration of motion-capture capabilities would elucidate during which phase of the squatting trial patients are offloading their ACLR limb more than their contralateral limb. Future researchers should consider implementing kinematic data collection when investigating limb loading characteristics during a bilateral squatting task.

Approximately 5 months postsurgery, patients recovering from ACLR were observed offloading their ACLR limb to their contralateral limb. Females offloaded their ACLR limb more than males. From these findings, clinicians may consider confirming that their patients, particularly their female patients, are not offloading their ACLR limb during functional tasks, such as a squat. This would help facilitate good biomechanics and minimize the adoption of poor movement patterns that may increase reinjury risks. Additionally, the relationship, albeit small, between limb loading metrics and PROs suggests that individuals who are asymmetrically loading their limbs during a bodyweight bilateral squatting task perceive a decreased ability to perform their ADLs. The evidence provided in our study highlights the need to further investigate the influence of limb loading metrics after ACLR and explore the feasibility of implementing a bilateral squatting task into an RTA protocol.

The authors would like to acknowledge Ms Haleigh Hopper, Mr Clay Ford, Dr Garret Burks, Mr Alex Gioia, Ms Hannah Orens, Mr Austin Kobza, and Ms Gianna Buckley for their assistance during data collection, entry, and processing. The results of this study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data omission. This study was supported by funding from the Innovative, Developmental, and Exploratory Award research and development fund from the School of Education and Human Development at the University of Virginia. This study was also supported by funding from the graduate student summer research fellowship from the American College of Sports Medicine Biomechanics Interest Group.