Introduction & Background
The anterior cruciate ligament (ACL) is the most complex knee ligament as it is responsible for stabilizing the knee by inhibiting the tibia from shifting from the femur (LaBella et al., 2014). Moreover, the ACL ligament is also the most prone to injury ligament, due to being exposed to bearing considerable weight (LaBella et al., 2014). Injuries are often associated with the tear of the ligament (Cimino et al., 2010) and are experienced by the sport population including basketball and soccer, as well as recreational athletes (LaBella et al., 2014). ACL injuries are suggested to negatively impact athletes’ careers mainly because it requires a long period of rehabilitation (between 6 to 9 months) and also because many athletes do not have the ability to return to the same level of performance as prior to the injury (Nyland et al., 2016). This statement makes prevention even more important to both coaches and athletes. Up to 70% of ACL injuries are suggested to mainly occur during non-contact situations (Shimokochi et al., 2008) due to placing excessive load on the knee when cutting movements and/or single-legged landings from jumps are performed (Hewett et al., 2005). This often leads to excessive knee valgus angles (KVA), which it is believed to place greater load on the ligament and result an ACL rupture (Hewett et al., 2005). Thus, avoiding knee valgus positions when landing from a jump could reduce the risk factors.
Previous studies have investigated ACL injury prevention by mainly examining the knee (Hewett et al., 2005), hip (Barrios et al., 2016) or trunk (Sasaki et al., 2015). Some findings showed that: 1) knee motion and knee loading during landing from a jump are predictors of ACL injury risk (Hewett et al., 2005); 2) female athletes with increased knee valgus demonstrate excessive hip and knee mechanics and could eventually make them more prone to knee pathology including ACL injuries (Barrios et al., 2016); and 3) a decreased trunk angle and an increased limb angle at initial contact after a jump may affect the COM position of the body and increase the chances of suffering a lower limb injury (Sasaki et al., 2015). Nevertheless, those studies also suggested neuromuscular training as being one common ACL prevention technique, as it is believed to be effective in controlling the lower limb positions during dynamic activities, such as landing from a jump, and reducing potential risk factors (Saski et al., 2015; Hewett et al., 2005). Despite those findings and suggestions, the occurrence of ACL injuries is still constantly rising; suggesting that more research is needed in the prevention field. One neglected part of the body with regards to ACL injuries is the foot. The foot is known to form the initial parts of the lower limb kinetic mechanism during landing from a jump, by conducting forces from the ankle to the knee (Ali et al., 2010). Subsequently, the foot rotations during landing are conducted by the position of the toe, which could be internally rotated (toe-in), externally rotated (toe-out) and facing forward (toe-forward) (Dempsey et al., 2007). These rotations are characterized as ACL risk factors related to non-contact ACL injuries, raising the present study authors’ interest in the injury prevention field.
The researchers of the present study were aware of the fact that the literature is lacking of information regarding the correlation between foot position and ACL injuries. However, they still managed to find few studies that have reflected on the importance of the foot position during landing after a jump, in impelling ACL injuries. The first mentioned study was Ishida et al., (2014), which have studied the effect of foot rotation positions on knee rotation during knee valgus. Their protocol, conversely, requested the participants to only move from a standing position to a knee valgus position, which is believed to not stimulate the loading during dynamic landing sufficiently. Another study of Cortes et al., (2007), has investigated the effects of foot-landing positions during double-legged landing on lower limb extremity kinematics and muscle activation levels. Their results, however, may not be applicable for a single-legged landing, which was the main interest of the present study, as ACL injuries are suggested to occur more in the single-legged landing than in double-legged landing (). Nevertheless, the present study is mainly based on the study of Dempsey et al., (2012), due to a mutual interest involved in single-leg landing. Their results found that a toe-out position was correlated with an excessive knee valgus moment. However, only one foot rotation position was investigated in the study of Dempsey et al., (2012); making the present study the first to investigate the potential effects of multiple foot rotation positions (toe-forward and toe-in positions) on knee valgus during single-leg landing, at two different moments, initial contact and landing, respectively.
The purpose of this study was to analyze the effects of three different foot positions (toe-in/ forward and out) on knee valgus, during single-leg drop landing, aiming to potentially reduce the occurrence of ACL injuries. The present study’s hypothesis is based on the previous findings (mentioned above) of Dempsey et al., (2007) and excessive knee valgus would be expected when landing from a jump, due to a toe-out foot position.
The participants’ selection was based on the results reported by Dempsey et al., (2012), on which a power analysis was performed. Fifteen participants were initially recruited for the present study from which 4 were later excluded. One participant dropped out halfway through the study, whereas other three had intense perspiration and led to inability to attach markers on specific body sites; therefore they were dismissed from the study. Thus, the remaining data of only 11 male recreational basketball players were used in the present study. Participants had no history of ACL or lower extremity injuries in the last six months, and were required to play basketball at least once a week and a have minimum of two years of experience prior to the study.
The absence of a control group failed to provide an insight of whether this intervention is actually beneficial to non-contact ACL sufferers. In addition, only male subjects participated in the present; thus, results can only be applied to the male populations and not to the female populations which have been shown to be more prone to ACL injuries due to increased knee valgus when landing from a jump (Hewett et al., 2005; Joseph et al., 2008).
For the actual protocol, participants followed a 5-minute warm-up after which they were familiarized with the drop-landing task before any data collection. For the actual trials, participants were then required to maintain a single-legged stance using their dominant leg on a 30 cm-high platform, drop land on a force plate using the same leg and maintain balance in that position for two seconds. Three different foot rotation positions (toe-forward, toe-in and toe-out) randomly represented were executed for the drop-landings. A successful trial was considered only when participants landed with the entire foot on the force plate without jumping and when they fully maintained the balance for at least two additional seconds. A total of three successful trials were collected of each foot rotations positions.
More information could have been congregated with regards to prevention of non-contact ACL injuries if the non-dominant leg would have also been assessed in the present study. It is suggested that the non-dominant leg displays less knee valgus angle than the dominant-leg when landing from a jump due to the inability to overcome excessive load and/or moments (Ludwig et al., 2017), making the non-dominant leg more prone to an ACL injury.
The effects of toe-in and toe-forward foot positions on maximum KVM may not be as qualitative as expected because all trials were conducted in a controlled laboratory environment and not on a basketball court (field), where the positions could be less pronounced. Additionally, basketballs could have also been used as references for the participants when performing the drop-landings tasks, in order to potentially make the data even more qualitative. For example, the participants could have first caught the ball and then performed the drop-landing task while holding the ball. The participants would have then imitated a specific action known as “rebound”, which requires the players to jump, catch the ball in mid-air and then land with the ball in their hand as quick as possible. Thus, their attention could have been focused more on the basketball and not on the task itself and could potentially influence the participants to land in a more natural position.
Retroreflective markers were attached to specific body sites (sacrum, anterior superior iliac spine, greater trochanter of femur, mid-thigh, medial and lateral knee epicondyles, tibial tuberosity, head of fibula, anterior aspect of shin, medial and lateral malleoli, bottom of calcaneus, Achilles tendon insertion point of calcaneus, head and base of first metatarsal, head and of fifth metatarsal and toe) of the dominant leg, with the use of tape in order to increase the (Leukoplast @, BSN medical, GmbH, Hamburg, Germany).
The kinematic 3-dimensional data was captured using eight digital cameras and the Cortex software (version 1.1.4368, Motion Analysis Corp, Santa Rosa, CA, USA) sampled at 200 Hz. Subsequently, a force plate (Kistler type 9287CA, Winterthur, Switzerland) was used for collecting kinetic data (ground reaction forces) sampled at 1000 Hz.
Any potential footwear limitation was eliminated by requiring all participants to wear the exact same model of basketball shoes (Nike Zoom, Attero, Nike, Inc., Beaverton, OR, USA).
For data analysis, a similar design as per Dempsey et al., (2007; 2012) was used in the present stud, including the UWA Lower Body Models (Besier et al., 2003b). This consisted of creating a 3-dimensional lower limb segment-marker-set of the pelvis, thigh, shank and foot (Figure 1), in Visual 3D (v5.00.33, C-Motion, Germantown, PA, USA). This was achieved by placing retroreflective markers on specific anatomical landmarks following a modified marker-set as presented in the study of Dempsey et al.,(2007). The pelvis segment was accurately defined as the CODA model (Charnwood Dynamics Ltd., Leicestershire, UK) was used. This model is suggested to be the most accurate method due to its ability to predict the location of the of the hip joint centre with 95% certainty (Bell et al., 1989).
The quality of both kinematic and kinetic data was smoothed by using a low-pass Butterworth digital filter at cut-off frequencies of 9 Hz and 50 Hz, respectively. A specific movement event was also created when the vertical GRF was greater than 10N in order to define the initial contact (IC). KVA at IC and maximum KVM during the landing phase were the two variables that were separately calculated using inverse dynamic equations as per de Leva (1996). These calculations provided additional information with regards to external joint moments when landing from a jump.
Rotations were used with respect to both pelvis and shank during the landing phase, in order to highlight any potential difference in foot rotations with respect to both body parts. This allowed the researchers to further investigate whether the shank or the pelvis is more correlated with the position of the foot. Additionally, foot rotation angles were only analyzed with respect to knee valgus. A combination of sagittal and frontal planes could have been considered because it could have provided more inclusive data which could lead to further understand the foot-landing effects in ACL injury prevention.
Descriptive statistics (mean and standard deviation) of KVA at IC, maximum KVM at landing and foot rotation angles at landing were used for further statistical analysis. A Shapiro-Wilk test was also carried out and highlighted normally distribution of the data (p>0.05). Subsequently, a one-way repeated measure ANOVA was conducted in Statistical Package for the Social Sciences (SPSS). Statistical significance was accepted when p value was less or equal to 0.05 (p?0.05). Bonferonni post-hoc test were also conducted with a family-wise set at a=0.05 in order to compare all the foot positions with each other. Since the sample size was quite modest, consisting of only 11 participants, an effect size was calculated as ?2 and values of 0.01 (small), 0.09 (medium) and 0.25 (large) interpreted.
Results showed that there were no significant differences in the foot rotation angles when measured from the shank (p=0.81). In contrast, significant differences were observed (p