Abstract
Background:
The anterior cruciate ligament (ACL) is loaded under tension when the tibia translates anteriorly relative to the femur. The shape of the articular surfaces of the tibiofemoral joint may influence the amount of anterior tibial translation under compressive loading. Thus, a steep lateral tibial plateau and a shallow medial plateau are thought to be risk factors for ACL injury.
Purpose/Hypothesis:
The purpose of this study was to evaluate whether tibial plateau slope and depth influence peak ACL strain during a single-leg jump. We hypothesized that there would be a significant correlation between tibial plateau slope and depth with ACL strain.
Study Design:
Descriptive laboratory study.
Methods:
A total of 17 healthy participants (8 male, 9 female) were assessed using magnetic resonance imaging (MRI) and high-speed biplanar radiography to obtain peak ACL strain during a single-leg jump. Two orthopaedic surgeons used the sagittal plane MRI scans to measure the medial and lateral tibial plateau slopes and the medial tibial plateau depth. The intraclass correlation coefficient was used to assess measurement reliability, and the Spearman rank correlation was used to evaluate the relationship between measurements of tibial morphology and peak ACL strain during the single-leg jump.
Results:
The overall range of intraclass correlation coefficients for intra- and interrater reliability of the medial and lateral tibial plateau slopes and medial plateau depth was 0.59 to 0.97. No significant correlations were found between peak ACL strain and any of the slope or depth measurements.
Conclusion:
In this cohort of healthy participants, correlations between any of the tibial plateau measurements with peak ACL strain during a single-leg jump were not detected. These findings are consistent with prior work, suggesting that tibial plateau slope and depth may not be linked to risk for ACL rupture. However, it is possible that tibial plateau morphology may interact with other factors to increase ACL injury risk or that individuals with extreme slope angles may produce differing results.
Clinical Relevance:
This study enhances the knowledge of the loading mechanisms for the ACL and thus improves the understanding of risk factors for ACL injury.
Keywords
Injury to the anterior cruciate ligament (ACL) can occur when the tibia translates too far anteriorly relative to the femur, 11 placing insurmountable strain on the ligament. A number of different risk factors for ACL injury have been presented in the literature, with several studies focusing on the bony morphology of the knee joint.7,10,28,39 The shape of the tibiofemoral joint (ie, the slope and depth of the tibial plateaus) has garnered particular attention, as it is thought that plateau morphology may affect the articulation of the tibia with the femur and thus the range of translations and rotations which the knee undergoes.10,28,38
Several mechanisms have been proposed to explain why tibial plateau morphology may predispose certain athletes to greater risk for ACL injury. It is thought that under a compressive force, a shallow medial plateau and steep lateral plateau cause the tibia to rotate internally and translate anteriorly relative to the femur.3,38 This then strains the ACL, increasing the risk of injury.3,38 However, there remains conflicting evidence regarding the magnitude to which tibial slopes may influence strain.22,32,34 Some studies have suggested that there may be a higher prevalence of steep lateral tibial plateau slope (LTS) and shallow medial tibial plateau depth (MTD) among athletes who have sustained an ACL injury.10,13,26,28,36 Similarly, prior work has suggested that patients with steep tibial plateau slopes have a greater risk of graft failure after ACL reconstruction.13,26 In contrast, other studies have found no link between tibial plateau slope and risk for ACL injury or graft failure,7,29,35,39 with some research suggesting that there is a high incidence of steeply sloped tibial plateaus within the healthy, uninjured population. 36 Overall, it remains unclear how tibial plateau morphology contributes to increased load on the ACL during dynamic in vivo activity.
The goal of the current study was to measure the influence of tibial plateau morphology on peak ACL strain during a dynamic activity. Specifically, by examining the associations between peak ACL strain during a single-leg jump with tibial plateau slope and depth, we aimed to elucidate whether there is a potential mechanism through which the morphology of the tibia contributes to elevated load on the ACL. We hypothesized that there would be a significant correlation between peak ACL strain and tibial plateau slope and depth.
Methods
Study Participants
For this study, 17 participants (characteristics shown in Table 1) were recruited under an institutional review board–approved protocol via online posting of a recruitment flyer. 23 Participants had no history of lower-extremity injury or surgery. A previous study had found that mean (± SD) peak ACL strain during walking was 9% ± 5%. 20 Based on this standard deviation of 5%, we calculated there would be 80% power to detect nonzero peaks in ACL strain of ≥3.6% with α = .05 using data from 17 participants. Regarding tibial morphology, McLean et al 34 detected a strong correlation between ACL strain and posterior tibial plateau slope (r = 0.76, P = .007). Using this information, we determined that a minimum of 11 participants would be required to detect correlations at this level with α = .05 and 80% power.
Characteristics of the Study Participants Overall and by Sex a
Data are presented as mean ± SD unless otherwise indicated.
All participants underwent magnetic resonance imaging (MRI) and high-speed biplanar radiography on a single self-selected knee.
Magnetic Resonance Imaging
MRI scans were acquired on a 3.0-T scanner (Trio Tim; Siemens Medical Solutions USA) with the knee in a relaxed, extended position. Images in the sagittal, axial, and coronal planes were acquired using a double-echo steady-state sequence (voxel size: 0.3 × 0.3 × 1 mm; flip angle: 25°; repetition time/echo time: 17/6 ms) and an 8-channel knee coil.1,9,15,19,37
Sagittal MRI scans were used to measure the medial tibial plateau slope (MTS) and LTS as well as the MTD. Tibial plateau slope and depth were measured based on previously validated techniques.27,30 Specifically, the outer borders of the tibia were defined as the most medial and lateral MRI slices in which the cortical bone could be visualized. The central-most slice, used for measuring the long axis of the tibia, was defined as the slice that fell midway between the tibial borders. The MTS and LTS were measured on the slices that fell halfway between the central slice and the respective borders of the tibia. The long axis of the tibia (Figure 1A) was established by drawing 2 circles within the tibial cortex—one within the tibial plateau and the other within the tibial shaft such that its center was on the circumference of the first circle. The long axis was then defined as the line connecting the centroids of the 2 circles. The MTS and LTS (Figure 1B) were defined by drawing a line connecting the anterior and posterior sides of the proximal borders of the tibial cortex. MTS and LTS were then measured relative to the line orthogonal to the long axis. MTD (Figure 1C) was measured as the length of the line orthogonal to and originating from the medial tibial plateau to the edge of the cortex at its deepest point.

(A) The long axis of the tibia was defined by drawing 2 circles within the cortex on the central-most sagittal plane slice of the tibia. The first circle was drawn within the plateau, and the second was drawn within the tibial shaft with its center on the circumference of the first. The tibial long axis was then defined as the line connecting the centroids of the 2 circles. A second line was defined orthogonal to the long axis, which was used subsequently for the slope measurements. (B) The medial and lateral tibial slopes (indicated by the arrows) were each defined as the angle between the proximal border of the plateau (dotted line) and the line orthogonal to the long axis (dashed line). Medial and lateral slopes were measured on the central-most sagittal plane slice of the medial and lateral plateaus, respectively. (C) Tibial plateau depth was measured as the length of the line orthogonal to and originating from the medial tibial plateau (dotted line) to the edge of the cortex at its deepest point, as indicated by the arrows. Medial tibial plateau depth was measured on the central-most sagittal plane slice of the medial tibial plateau.
Two fellowship-trained orthopaedic surgeons (S.T. and J.R.W.; raters 1 and 2, respectively) independently performed the measurements of slope and depth for all participants. Importantly, on each iteration of measurements, the surgeons selected the appropriate slices; defined the tibial long axis; and measured the MTS, LTS, and MTD.
Three-dimensional (3D) models of the knee joint, including the femur, tibia, and attachment site footprints of the ACL, were created via manual segmentation of the MRI scans using solid-modeling software (Rhinoceros 4.0, Robert McNeel and Associates & Geomagic, 3D Systems) as previously described (Figure 2, A-C).15,42 Segmentations were reviewed by a fellowship-trained musculoskeletal radiologist (C.E.S.).

(A) The femur and tibia (green lines) were traced on a sagittal-plane magnetic resonance imaging (MRI) scan; the anterior cruciate ligament attachment sites on the tibia and femur (red lines) were then traced on all 3 MRI planes and then cross-registered together onto the sagittal plane. (B) These tracings were then compiled into a wireframe model of the bones and attachment sites. (C) The wireframe models were used to generate a 3-dimensional model of the knee joint. (D) The bone model was then registered onto both frames of the high-speed biplanar radiographs simultaneously to recreate the relative position of the knee joint at the time of imaging.
High-Speed Biplanar Radiography
Participants underwent high-speed biplanar radiography (matrix size: 1152 × 1152 pixels, frame rate: 120 Hz, pulse width: 1.0-1.5 ms) while performing a single-leg jump, as previously described.15,17-20,23 The participants were positioned within the field of view of the radiographs. They were then instructed to jump vertically, taking off and landing on the same leg without holding onto any external supports.15,23 Ground-reaction forces (GRFs) were acquired throughout the jump via an instrumented treadmill (Bertec) sampled at 1200 Hz. The GRFs were used to identify the start and end of a single-leg jump trial. A jump was determined to be valid if both the tibia and femur remained at least partially inside the field of view throughout the motion. 20 To ensure that the total estimated effective radiation dose remained <0.14 mSv per participant,15,23 each experiment used a radiographic protocol not exceeding 110 kVp/200 mA (a limit that was established in conjunction with the institutional office for radiation safety).15,18,20 One valid trial per participant (nominally 1-3 attempts) was collected.
Anterior Cruciate Ligament Strain
To obtain ACL strain measurements, the 3D bone models and radiographic images were imported into custom software written in MATLAB (MathWorks) that optimizes the position of the 3D bone models to align their projections with the bony contours in each frame of the radiographic images (Figure 2D). 20 This process reproduces the relative in vivo positions of the bones, making it possible to measure ACL strain.
ACL strain (ϵ), defined by the equation below, is the change in the centroid-to-centroid distance (l) between the ACL attachment sites on the femur and tibia, measured relative to a reference length (l0), which represents the length at the slack-taut transition point 21 :
In this study, the position of the knee on the MRI scan was selected as the reference because it was assumed to be a position where the ligament was minimally loaded.15,41,44 ACL strain was measured throughout the entire jump cycle. A jump cycle was defined from toe-off (when GRF last registered on the force plate before the flight phase of the jump) to the time of maximum GRF during the landing phase of the jump. The peak ACL strain value during the jump for each participant was recorded and used for analysis. These techniques have been previously shown to reproduce the relative positions of the bones to within 70 µm and identify the locations of the ACL attachment site centroids to within 0.3 mm.1,20,40
To evaluate how robust our results were in utilizing the knee position on MRI to determine the reference length (l0) for calculating ACL strain, we perturbed the reference length such that each participant's l0 was multiplied by a random number (Gaussian distribution generated with mean = 1, standard deviation = 0.05; equivalent to a 5% change in reference length). This new set of l0 values was used to calculate the resultant peak strain for each participant. This simulation was repeated 100 times, and the Spearman rank correlation of peak strain with MTS/LTS/MTD (using measurements from rater 1) was computed for each. The number of simulations that resulted in P < .05 was calculated as well as the mean rS and P values across the 100 sets of strain measurements using the perturbed reference length.
Statistical Analysis
The reliability of the morphological measurements was evaluated using the intraclass correlation coefficient (ICC). ICCs based on 1-way analysis of variance were used to evaluate the intrarater reliability on a randomly-selected set of 5 participants who were measured 3 times by each rater. ICCs based on 2-way analysis of variance were used to validate the interrater reliability on the total group of 17 participants who were evaluated. Spearman rank correlation analysis was used to evaluate the relationship between each of the morphological measurements (MTS, LTS, and MTD) with peak ACL strain. This test was selected as it allows for the evaluation of monotonic relationships between variables. The correlation coefficient (rS) values were computed separately for each rater. Furthermore, linear regression models were used to determine whether there was an effect of sex on the slope of each of the tibial plateau measurements with peak ACL strain. Correlations were deemed significant if a test for correlation equal to zero reached a value of P < .05. All statistical analyses were performed using RStudio Version 4.2.0 (The R Project for Statistical Computing) and supervised by an experienced faculty statistician (A.S.K.).
Results
MTS, LTS, MTD, and ACL strain were measured in all 17 participants. The ICC for interrater reliability of the MTS was 0.59; ICCs for all other inter- and intrarater reliability measures for tibial plateau slope and depth were >0.8 (Table 2). The mean morphological measurements from rater 1 were 3.3°± 2.9° for MTS, 4.1°± 5.1° for LTS, and 2.1 ± 0.6 mm for MTD. Based on the Spearman rank correlation analysis, no significant relationship was found between peak ACL strain and any of the tibial plateau slope or depth measurements for either rater (P > .46 for all measurements) (Figure 3). For both raters, a significant correlation was identified between MTS and LTS (rater 1: rS = 0.8, P < .001; rater 2: rS = 0.49, P = .04). There was no differential effect of sex, such that for both men and women, no correlation was detected between tibial plateau slope and depth and peak ACL strain.
Intraclass Correlation Coefficients for the Tibial Plateau Slope and Depth Measurements a
LTS, lateral tibial plateau slope; MTD, medial tibial plateau depth; MTS, medial tibial plateau slope.

The least-squares linear fit line (solid line) and pointwise 95% CI (shaded area) for measurements performed by rater 1 between peak anterior cruciate ligament (ACL) strain and (A) medial tibial plateau slope, (B) lateral tibial plateau slope, and (C) medial tibial plateau depth. No significant correlation was detected between peak ACL strain and medial slope, lateral slope, or medial depth for either rater.
Mean rS and P values for the correlations of 100 sets of peak ACL strain measurements using the perturbed reference lengths with MTS, LTS, and MTD measurements are shown in Table 3. Of the 100 simulations of altered reference length, <5 produced a significant effect for each morphological measurement.
Results of Spearman Rank Correlation Analysis of 100 Simulations of Peak Strain Using Perturbed Reference Length a
MTS, medial tibial plateau slope; LTS, lateral tibial plateau slope; MTD, medial tibial plateau depth.
Measurements from rater 1 were used.
Discussion
Several prior studies have focused on the morphology of the tibiofemoral joint, particularly the slope and depth of the tibial plateaus, as a risk factor for ACL injury.7,10,28,39 However, there are limited data in the literature characterizing the influence of tibial plateau morphology on ACL strains during dynamic activity. Thus, we performed an in vivo investigation into tibiofemoral shape and elongation of the ACL. In this cohort of participants with no history of ACL injury, we did not detect any correlations between tibial plateau slope or depth with peak ACL strain during a single-leg jump.
We measured the peak ACL strain during a jump and correlated it with MTS, LTS, and MTD as measured from sagittal-plane MRI. Tibial slope is often measured in vivo using lateral view radiographs due to their clinical utility. 30 However, radiographic techniques may be susceptible to errors arising from an inability to control the orientation of the tibia within the image 31 and difficulty with distinguishing the medial and lateral tibial plateaus. 30 An advantage of MRI is that the medial and lateral tibial plateaus can be clearly delineated from one another.
Here, we demonstrated that 2 orthopaedic surgeons can consistently measure tibial plateau slope and depth from MRI (ICCs, 0.59-0.97). These ICCs are comparable to those of other studies utilizing a similar measurement technique using MRI.27,30 While the interrater ICC for MTS was low (0.59) compared with the other ICCs, the intrarater ICCs for the MTS measurements were 0.93 and 0.80 for raters 1 and 2, respectively. Furthermore, neither rater detected a correlation between MTS and peak ACL strain. Thus, while there were some differences in measurements between the 2 raters, both analyses produced comparable results regarding the relationship between MTS and ACL strain.
The findings of this study are based on results from healthy participants with no history of lower extremity injury or surgery. Yet, the slope and depth measurements presented here are in line with those of other studies that applied a similar measurement methodology in patients who have sustained an ACL injury.10,28 Importantly, as there are several methods for measuring the morphology of the tibial plateaus (ie, using 3D methods, radiographs, or soft tissue surfaces), it may be difficult to compare results across studies.14,34,43 In particular, a 2023 study by Garra et al 24 that used the same measurement methodology as the present study suggested that compared with MRI, radiographic measurements on average overestimate the MTS and LTS by 6° and 4°, respectively. Furthermore, these authors suggested that a cutoff point for tibial plateau slope correction should be closer to 7° when measured from MRI compared with the 12° cutoff, which is typically estimated for radiographic measurements.12,24 This distinction between cutoffs highlights the importance of accounting for imaging modality when interpreting tibial plateau slope measurements.
In the present study, we failed to detect a correlation between tibial plateau slope and depth with ACL strain during a single-leg jump. However, this data set included only healthy participants who had not sustained an ACL injury. Furthermore, from these results, we cannot rule out the possibility that patients with tibial plateau slope and depth outside the range measured here may present differently, such that there could be a critical threshold at which tibial plateau slope contributes to elevated ACL strain. Nonetheless, these findings suggest that tibial plateau morphology in isolation may not be the driving factor for elevating strain on the ACL, a conclusion corroborated by other studies.2,22,25 Regardless, we cannot rule out the possibility that a combination of bony or other risk factors may compound to elevate the likelihood of sustaining an ACL injury.
While tibial plateau morphology has been found to differ based on sex, 5 we did not detect a differential effect of sex between slopes (tibial plateau slope/depth with peak ACL strain). It is important to note that we collected a relatively small sample size for making comparisons based on sex, as this was not a main objective of the present study. Nonetheless, these results suggest that tibial plateau morphology alone may not account for sex-based differences in ACL injury rates. However, future studies should be conducted to evaluate potential sex-based differences.
Multiple studies have found a strong inverse relationship between knee flexion angle and ACL strain, whereby peak strain occurs when the knee is in an extended position.6,15,23 In addition, when the quadriceps contract, they apply an anteriorly-directed force on the tibia, which in turn loads the ACL, an effect that is greatest when the knee is at low flexion angles.11,16,17,19 Viewing the findings of the present study in the context of previous work, it is possible that tibial plateau slope and depth may only contribute to elevating ACL strain to the point of rupture when combined with other key loading factors, such as kinematics and muscular forces.
A strength of the present investigation is the ability to test for correlations between participant-specific measurements of both knee morphology (plateau slope and depth) and of dynamic ACL strain, often used as a marker of ACL loading.8,33 Our finding that there was not a significant influence of tibial morphology on ACL strains during a single-leg jump is in line with prior cadaveric studies. These studies demonstrated that, during combined anteroposterior and axial loading, increasing the slope of the tibial plateaus shifts the resting position of the tibiofemoral joint anteriorly but does not significantly increase the forces experienced by the ACL.22,25 In the present study, we examined ACL strains during a single-leg jump. Subsequent studies may examine the relationship between tibial plateau morphology and ACL strain patterns during other dynamic activities.
We recognize that a limitation of ACL strain measurements, both in vivo and ex vivo, is the difficulty associated with determining the appropriate reference length of the ACL. 4 As such, we tested the correlations (to MTS, LTS, and MTD) with 100 sets of peak ACL strain values, which were calculated by applying controlled, random changes to the ligament's reference length, l0 (see Table 3). With a level of significance set at P < .05, we would expect at most 5 of these simulations to produce significant results for each of the morphological measurements simply due to random error. This was true for the present results, suggesting that the few instances that produced significant correlations did so due to chance. Furthermore, this simulation bolsters the conclusions made about the relationships between peak strain and MTS/LTS/MTD, showing that they are robust to perturbation in the reference length of the ACL.
Conclusion
We investigated the role of tibial plateau slope and depth on ACL injuries using peak in vivo ACL strain as our outcome measure. We did not detect a significant relationship between the slope or depth of the tibial plateaus with peak ACL strain during a single-leg jump. These findings are consistent with prior work indicating that tibial plateau slope and depth may not be linked to risk for ACL rupture.7,39 However, it is possible that tibial plateau morphology may interact with other factors to increase ACL injury risk or that those with extreme slope angles may produce differing results. Overall, this study provides novel data relating tibial plateau slope and depth to in vivo ACL strain.
Footnotes
Acknowledgements
The authors thank Stephanie Danyluk for her assistance with participant recruitment, Jean Shaffer and Raven Boykin at the Duke Center for Advanced Magnetic Resonance Development for assistance with MRI acquisition, and Donald T. Kirkendall for his assistance with manuscript preparation.
Final revision submitted December 24, 2023; accepted February 2, 2024.
One or more of the authors has declared the following potential conflict of interest or source of funding: This study received support from the U.S. National Institutes of Health (grants AR065527, AR079184, and AR074800 to L.E.D.). S.T. has received grant support from Arthrex, education payments from Arthrex and Smith+Nephew, and hospitality payments from Stryker. A.A. has received education payments from Arthrex; consulting fees from Bioventus, Arthrex, DJO, and Limacorporate; nonconsulting fees from Arthrex; hospitality payments from Lima USA and Stryker; royalties from Arthrex; and has stock/stock options in Anika Therapeutics. J.R.W. has received education payments from Arthrex, consulting fees from Geistlich Pharma and Vericel, nonconsulting fees from Vericel and Arthrex, and hospitality payments from Aesculap Biologics. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was obtained from Duke Health (ref No. Pro00042239-CR-10.1).
