Discussion

This study was conducted to analyze the relationship between estimated ACL loading and biomechanical variables during a variety of simulated DVJ trials. The assessment of ACL loading was performed with a whole-body musculoskeletal FE model. This was the first study to estimate absolute ACL strain and force during an in vivo landing using a specimen-specific, validated FE model. The hypothesis was that ACL loading would be higher in asymmetrical landing in the frontal plane, especially with (1) higher vertical and lateral ground-reaction forces, (2) lower gluteus medius force, and (3) laterally tilted and shifted pelvic motion. The hypothesis was investigated using reconstructed waveforms from PCs that predicted the trials with high ACL strain in a logistic regression model. Our hypothesis was supported, as the peak vertical and lateral ground-reaction force was higher, the gluteus medius force was lower, and the lateral pelvic tilt and hip adduction (indicative of lateral pelvic shift ⁵¹ ) were higher in the high-risk waveforms. Therefore, this study demonstrated that estimated ACL loading was greater during simulations of the participants’ landings that exhibited asymmetry in the frontal plane than in landings that were symmetrical.

The ACL strain and force during DVJs seen in the current study were lower compared with previously reported failure loads in young female and male in vitro specimens, which were 15% to 30% and 1266 to 2160 N, respectively. ^9,10,52 As expected, the ACL strain and force estimates in this study were substantially lower than measures at failure, as data from laboratory-controlled DVJ trials were used to drive the models. For 15 of the 37 trials, peak strain was observed at IC, when the knee was extended the most. This is consistent with previous fluoroscopic studies. ^12,13 However, the trials that ACL loading was higher than the third quartile presented peak forces at approximately 30 to 60 milliseconds after IC, which is consistent with previously reported ACL failure time from video analysis and cadaveric landing simulations. ^6,30,31,50

The 6 DOF tibiofemoral joint kinematics revealed that higher knee abduction, internal tibial rotation, and translations toward the anterior, medial, and superior directions were linked to higher ACL strain. The observed increase in tibial translation in the anterior and superior directions indicated that the relative position of the tibial and femoral condyle became more distant in the anteroposterior DOF and closer in proximity in the inferosuperior DOF. This indicated that the femoral condyle rolled posterior and inferior down the tibial plateau. Furthermore, increased knee abduction under a compressive load supported the presence of a continuous compression on the lateral tibial plateau.

With increased internal rotation, the lateral femoral condyle must roll in an inferior and posterior direction at a greater magnitude than the medial femoral condyle to demonstrate these kinematic combinations. These observations indicate that a pivot-shift mechanism occurred, in which the knee abduction created a compression on the lateral plateau of the tibia and induced anterior tibial translation and internal tibial rotation partially because of the posterior tibial slope. ³⁵ Specifically, the previous validation study of the model presented that the local angle of the tibial plateau slope, which was calculated on each surface element of the cartilage model, initiated at the tibial eminence and reached its maximal angulation of 22° at the posterior edge of the tibia (see Navacchia et al ³⁹ ). In the trials with high ACL loading, ACL loads were presented even with the increased knee flexion angle in the later phase of landing. This is consistent with previous in vitro studies indicating that the ACL is loaded under anterior tibial force and knee abduction moment at the knee flexion angle of 60° to 90°. ^16,33

The present findings showed that ACL strain was higher during asymmetrical landings, as the waveforms of high ACL strain trials demonstrated higher peak vertical and lateral ground-reaction forces as well as lower gluteus medius force and laterally tilted and shifted pelvic motion. Knee abduction angle during DVJ is a predictor of ACL injury. ^3,25 The present study confirmed that athletes with a high knee abduction angle have higher ACL strain during the DVJ. Note that this study does not prove a single effect of knee abduction angle but rather that other featured biomechanical variables are linked in high ACL strain trials. A recently presented ACL injury reduction program includes trunk and hip joint exercises to stabilize body control and reduce ACL injury risk. ^22,23,42,48 Although these previous studies could not quantify the effect of such training on ACL loading, the present study supports the results from the training program, suggesting that they have the potential to decrease ACL loading during landing.

Generally, lower muscle forces were observed in high ACL strain waveforms, except for the iliacus, psoas, vastus intermedius, and gastrocnemius medialis. The vastus intermedius and gastrocnemius medialis might contribute to an increase in ACL loading as antagonists of the ACL. ⁴¹ However, a causal analysis was not performed in this study, and the causal effect of these muscles on ACL loading remains unclear.

This study had some limitations. First, the FE analysis was conducted separately from the muscle force estimation step. Since the ligament forces and contact forces on the knee joint resist external joint moments, especially knee abduction/adduction and internal/external rotation moments, there are interactions between ligament forces and muscle forces. As some previous studies have reported, ^20,40 concurrent estimation of muscle, ligament, and joint contact forces would more adequately unveil the causal interaction between these forces. Second, only 1 specimen-specific FE model was used for all the landing trials, and body weight and size were not scaled to each participant.

While the methodology in this study was aimed at simulating ACL loading during an in vivo landing with physiologic loading, the results do not represent the participants’ actual ACL strain/force during landing. The ACL loads simulated in this study represent the loads in the ligament for a single model landing with various strategies. Patient-specific models that account for personalized anatomy are needed to overcome this limitation. However, in vivo patient-specific models with validated ligament properties have not yet been reported in the literature. Promising work has been done to correlate magnetic resonance images with in vivo ligamentous material properties ³⁸ and may assist in the development of these patient-specific models in future studies. In addition, the validation step of the FE model was performed relative to a knee flexion angle of 25°. ³⁹ Therefore, a limitation of the current modeling paradigm is that the results have not been explicitly validated throughout the full range of flexion expressed by each range of motion. Finally, ground-reaction force was applied on the ankle joint center but not on the foot, since the ankle joint was not validated and the movement was kinematically driven.