1. Introduction
Being a critical load-bearing and movement joint for the lower extremity, the knee joint is often afflicted by diseases, and a variety of complicated factors are associated with the development, maintenance, and health of the joint tissues [
1,
2]. It is not surprising that a significant level of research is devoted to the knee joint. For patients with mild knee disease, joint replacement and medication (associated with a high risk of side effects) are not necessarily required to reduce knee joint load and pain [
3]. Therefore, there is a need for conservative treatments to reduce peak values of tibiofemoral (TF) contact force or knee joint reaction force (JRF), such as valgus knee bracing [
4], variable-stiffness shoes [
5], lateral wedges [
6], and gait modification/retraining [
7,
8].
Gait modification includes adjustments in trunk lean, walking speed, and the foot progression angle (FPA). It has been reported that gait modifications have a significant effect on ground reaction force (GRF) and knee adduction moment (KAM) during stance phases [
9,
10,
11], the effects of which depend on participant type (age, gender, and with osteoarthritis or healthy), gait modification type, and implementation level of gait modification. Trunk lean is reported to be highly efficient in reducing peak KAM values, while potential adverse effects including lower back discomfort, imbalance, and difficulty in maintenance for a long time [
12,
13]. Although the toe-in and toe-out gait patterns could alleviate the progression of knee OA (osteoarthritis), the detailed mechanism is still vague [
14,
15]. Gait modification has showed positive benefits at the knee joint, but it might be associated with negative effects on other joints; the longer-term effects of the toe-in and toe-out gait modifications in patients with knee OA is unknown. Previous work reported that peak value of KAM is positively related to walking speed; however, whilst peak KAM reduces with decreasing walking speed, the load duration on the knee joint increases and the effect of gait modifications on neuromuscular function is less [
16,
17]. Bennet et al. [
18] reported that toe-in and toe-out with wider step width gaits seem to be viable options to reduce peak adduction moment but that sagittal and transverse knee loadings should be monitored. Controlling FPA implies modifying the gait by either toe-in or toe-out adjustment. Individualized toe-in and toe-out would lead to immediate quadricep and hamstring co-contraction increases in healthy adults. However, after adjusting the gait for a short period of time, the co-activations of these muscles were not significant [
11].
The effective outcome of FPA gait modification is mainly adjusting the peak values of KAM in a short period of time [
10,
11]. Because it is very difficult to directly measure TF contact forces, KAM has often been used for predicting total TF force and has been considered as a proxy for medial TF contact force [
9,
16,
19]. From literature, it is not clear whether KAM should be used as a surrogate measure for total TF force or medial TF contact force. Higher KAM is related to radiographic structure changes at knee joints and cartilage degeneration [
1,
20]. Some research work also has suggested that the knee flexion moment (KFM) also plays an important role in knee loading and that it should also be taken into account [
14,
21]. Richards et al. [
14] suggested that the combination of KAM and KFM yields an improved surrogate measure for TF contact force than KAM alone; however, it improved only the first peak values not the second peak values of knee contact force during the stance phase. While FPA gait modification showed that the first peak KAM is reduced, it remains unclear whether these treatments subsequently increased the peak KFM and TF contact force. Previous studies showed that the reduced peak KFM also carries the risk of causing quadricep weakness, which might be associated with developing knee osteoarthritis (OA) [
7,
21]. Increasing quadricep strength could improve functional outcomes and could alleviate pain in patients with OA [
22]. However, it is still unclear whether an increase in KFM is related to quadriceps strength and whether it would increase the KAM and/or the TF contact force, which further affects the progression of OA [
20]. On the other hand, modified gait patterns will involve changes at more than one body segment and one joint. Positive benefits at the TF joint might be linked with negative impact on other joints concurrently [
11]. Furthermore, muscle coordination also plays a key role in contributing TF contact force [
23,
24,
25,
26]. Miller et al. [
27] conducted only computer simulations to predict contact forces by gait modification; they reported that gait modification can reduce the peak KAM from normal walking. From the literature review, it was found that there was little information on the study of Chinese subjects through gait modifications, especially lacking analysis by using musculoskeletal modelling. To address the above issues, quantitatively comparing differences in lower-extremity kinematics, dynamics, and electromyography caused by FPA gait modification is important and will be conducted herein.
Therefore, the first objective of this paper was to study the effects of toe-in and toe-out gaits through experimental measurement and computer modelling at normal walking speeds on lower-extremity joint angles, GRF, electromyography (EMG) of muscles crossing the knee joint, lower-extremity joint moment, and TF contact force during level walking. The second objective was to evaluate whether KAM can be used as a surrogate measure for medial or total TF contact force.
4. Discussion
This study used the experimental measurements and musculoskeletal modeling to study and compare the effects of toe-in and toe-out gaits at normal walking speeds on the lower-extremity kinematics, dynamics, and EMGs. The results showed that the toe-in and toe-out gaits indeed influenced these values, especially on the joint moments and EMGs. This study also showed that the peak value of JRFs does not decrease with reducing peak KAM, which is consistent with previous research findings [
35].
Gait modification changed moments and angle of joints; for example, the toe-in gait changed joint rotation angles at the ankle and hip joints, as shown in
Table 1, and knee adduction moments at knee joint, as shown in
Figure 5. The obtained joint angles and joint moments during the natural gait are slightly different from those reported by Hamner et al. [
25], who used the IK and residual reduction algorithm (RRA) tool to calculate joint angles during a running gait. Both the toe-in and toe-out gaits have no statistically significant effects on reducing the peak values of GRFs in the three directions except for the second peak ML force in the toe-out gait, as shown in
Table 2.
EMG is used to detect and assess muscle coordination and activation levels; EMG signal across different gaits may be able to provide insight into muscular force production and function of individual muscles during movement. The use of EMG has been controversial because of considerable variations between subjects; however, it still has reference value in individual cases where a baseline has been established. The EMG profiles recorded for the knee-crossing muscles revealed there were differences in EMG activations among the three gait patterns. As shown in
Table 3, the toe-in and toe-out gaits resulted in higher values for quadriceps and peroneus longus; however, they reduced EMG values for biceps femoris and gastrocnemius lateralis. The toe-in and toe-out gaits resulted in the opposite effect on the activations of tibialis anterior, semitendinosus, and gastrocnemius medialis.
Many studies suggested that an elevated KAM value represents an elevated TF contact force or medial TF contact force, which has been considered as one of the important contributors to knee osteoarthritis [
36,
37,
38]. The toe-out gait increased the first peak KAM and decreased the second peak KAM, as shown in
Table 4, which is consistent with the existing literature [
9]. The second peak value of the KAM, as shown in
Figure 5, is higher than the first peak which is inconsistent with those in the literature during the natural and toe-in gaits. It might be due to variations in the placement of markers and because healthy individuals were measured in this paper while other studies measured subjects with OA or TKA [
14,
17,
35]. However, the results in
Figure 4 and
Figure 6 showed that the toe-in and toe-out gaits had little effects on the GRFs and TF joint contact forces (
p > 0.05) but that it has significant influence on the KAM and KFM during the stance phase (
p < 0.05), as shown in
Figure 5. The toe-in and toe-out gaits, on the one hand, are able to relieve pain by laterally changing the center of contact location [
31,
39,
40]. On the other hand, a shift in the bilateral contact point locations caused by the gait patterns also has a direct effect on the lower limb joint moment and redistribution of TF contact forces [
41]. Based on the results of significant changes in KAM, KAM or a combination of KAM and KFM can be used as surrogate measures for TF medial contact force but not total contact force.
Although very good outcomes were obtained in this paper, there are a number of shortcomings that should be addressed in future studies. First, participants only walking at a single speed were studied; previous work showed that peak value of KAM is positively related to walking speed [
16,
17]. Therefore, the results may not be applicable to the healthy or knee OA population, especially to people with knee arthroplasty. Second, the musculoskeletal model used in OpenSim was scaled using weight of the subject and 16 markers. The simulation of the joint moments and knee JRFs could be improved by using more markers. Third, the musculoskeletal model does not include ligament and soft tissues, which also are associated with TF contact forces directly. Finally, the difference in lower-limb kinematics, dynamics, and EMG caused by FPA gait modification were compared in this paper but long-term effects were not assessed. Further work is needed to address these limitations, the TF contact force may alter over time [
14].
The results presented in the paper were based on an exploratory study with a relatively small sample size; however, FPA modification results might be used by physiotherapists to help OA patients to relieve pain and could provide loading information for engineers when they design knee replacement implants for patients who may have toe-in or toe-out gaits. It seems more reasonable to assume that toe-in and toe-out gaits could relieve patient pain by redistributing TF contact forces of medial and lateral condylar due to the change of lateral contact center and the shift of bilateral contact points. When exploring new conservative treatment methods of OA in the future, it might be more reasonable to study the joint contact forces, contact locations, and contact areas in addition to TF joint KAM and KFM.