Ankle Dorsiflexion Affects Hip and Knee Biomechanics During Landing

Abstract

Background:

Restricted ankle dorsiflexion range of motion (DFROM) has been linked to lower extremity biomechanics that place an athlete at higher risk for injury. Whether reduced DFROM during dynamic movements is due to restrictions in joint motion or underutilization of available ankle DFROM motion is unclear.

Hypothesis:

We hypothesized that both lesser total ankle DFROM and underutilization of available motion would lead to high-risk biomechanics (ie, greater knee abduction, reduced knee flexion).

Study Design:

Cross-sectional study.

Level of Evidence:

Level 3.

Methods:

Nineteen active female athletes (age, 20.0 ± 1.3 years; height, 1.61 ± 0.06 m; mass, 67.0 ± 10.7 kg) participated. Maximal ankle DFROM (clinical measure of ankle DFROM [DF-CLIN]) was measured in a weightbearing position with the knee flexed. Lower extremity biomechanics were measured during a drop vertical jump with 3-dimensional motion and force plate analysis. The percent of available DFROM used during landing (DF-%USED) was calculated as the peak DFROM observed during landing divided by DF-CLIN. Univariate linear regressions were performed to identify whether DF-CLIN or DF-%USED predicted knee and hip biomechanics commonly associated with injury risk.

Results:

For every 1.0° less of DF-CLIN, there was a 1.0° decrease in hip flexion excursion (r² = 0.21, P = 0.05), 1.2° decrease in peak knee flexion angles (r² = 0.37, P = 0.01), 0.9° decrease in knee flexion excursion (r² = 0.40, P = 0.004), 0.002 N·m·N⁻¹·cm⁻¹ decrease in hip extensor work (r² = 0.28, P = 0.02), and 0.001 N·m·N⁻¹·cm⁻¹ decrease in knee extensor work (r² = 0.21, P = 0.05). For every 10% less of DF-%USED, there was a 3.2° increase in peak knee abduction angles (r² = 0.26, P = 0.03) and 0.01 N·m·N⁻¹·cm⁻¹ lesser knee extensor work (r² = 0.25, P = 0.03).

Conclusion:

Lower levels of both ankle DFROM and DF-%USED are associated with biomechanics that are considered to be associated with a higher risk of sustaining injury.

Clinical Relevance:

While total ankle DFROM can predict some aberrant movement patterns, underutilization of available ankle DFROM can also lead to higher risk movement strategies. In addition to joint specific mobility training, clinicians should incorporate biomechanical interventions and technique feedback to promote the utilization of available motion.

Knee injuries are among the most prevalent injuries throughout sports. ³⁰ The development of patellofemoral pain syndrome (PFPS) and rupture of the anterior cruciate ligament (ACL) are some of the most common diagnoses across gender, especially in female athletes.^3,13 Though the mechanism of injury often differs between PFPS (insidious onset) ²⁸ and ACL rupture (acute onset), ¹⁰ similar biomechanical risk factors may put an athlete at risk for both injuries. ⁴⁰ Similarly, both injuries present as precursors for an increasing likelihood of the development of knee osteoarthritis,^25,35 which can be found in 30% to 40% of former athletes. ³⁷ Understanding the associated biomechanical risk factors of these injuries may help optimize preventative or rehabilitative strategies.

Two biomechanical patterns have been linked as potential underlying mechanisms associated with PFPS and ACL injury: dynamic lower-extremity valgus and limited knee flexion.^40,45 Dynamic lower-extremity valgus, a combined motion of hip adduction and internal rotation, knee abduction, and foot/ankle pronation, ²³ places increased aberrant forces on the knee. This movement pattern has been observed through video analysis at the time of ACL injury ¹¹ and may be predictive of future ACL injury in adolescent female athletes. ²⁹ Likewise, repetitive dynamic lower extremity valgus may alter patellofemoral mechanics and compression forces, leading to injury.^39,40 In the sagittal plane, reduced knee flexion (stiff landings) during dynamic activities has also been linked to both PFPS and ACL injury risk.^15,33 Both mechanisms, dynamic valgus and stiff-legged landing strategies, may be influenced by factors distal and proximal to the knee.

While significant work has focused on the role of the hip in resultant biomechanics,^12,41 there are fewer investigations of how the foot and ankle complex can affect knee joint biomechanics. Reduced ankle dorsiflexion range of motion (DFROM) is a common problem among athletes, especially after injuries such as lateral ankle sprains,^1,53 and has been reported to affect an athlete’s biomechanics. In the sagittal plane, clinical measures of restricted dorsiflexion have been previously linked to reduced knee and hip flexion motion.^19-21,43,49 Similarly, restricted DFROM reduces one’s ability to absorb forces throughout the lower extremity during cutting, jumping, and landing, resulting in greater ground reaction forces and greater frontal plane motion and loading, especially at the knee.^{7,36,38,48,49} However, in these studies, ankle DFROM has been measured in a variety of ways, including weightbearing or nonweightbearing, ⁴² clinically measured with a goniometer or inclinometer, ³² or biomechanically measured during a dynamic movement with motion analysis. ²² All these measures utilize maximum values of ankle dorsiflexion, even when the maximum amount of dorsiflexion is not used during the task and submaximal control or utilization may be more appropriate to consider.

Whether it is the maximum physiological DFROM or the amount of ankle DFROM used during dynamic motions that affects lower extremity biomechanics is not as well understood and could have implications on clinical practice and the development of intervention strategies. Thus, the purpose of this study was to examine how clinical and laboratory-based DFROM measures relate to multiplanar knee and hip joint biomechanics in an active female population when landing from a jump. It was hypothesized that static clinical measures and dynamic laboratory measures would offer complementary (not identical) information, and that restricted DFROM (both clinical and laboratory based) would predict higher knee abduction angles, lower angles of knee flexion angles during landing and lower sagittal plane joint moments and energy absorption at the knee. Ultimately, we expected that lesser degrees of ankle DFROM during landing would elicit hip and knee biomechanics associated with high-risk landing strategies.

Methods

Participants

Nineteen healthy female athletes (age, 20.0 ± 1.3 years; height, 1.6 ± 0.1 m; mass, 67.0 ± 10.7 kg) volunteered to participate in this study. At the time of recruitment, all participants were (1) physically active for a minimum of 90 minutes per week, (2) without any history of knee injury or lower extremity surgery, (3) without any lower extremity injury in the past 6 months, (4) without any vestibular or balance disorder, (5) without any medical condition known to affect connective tissue, and (6) without any history of cardiovascular or pulmonary problems. This study was approved by the University of North Carolina at Greensboro Institutional Review Board, and written informed consent was obtained from each participant before participation.

Data Collection

Participants reported to the laboratory for a single testing session during which all data were collected. On arrival, participants were outfitted with compression shorts and a tight-fitting athletic top. After barefoot measurements of body height and weight were obtained, participants then performed a 5-minute warm-up on a stationary cycle ergometer (Life Cycle; Life Fitness) at a cadence of 70 to 80 RPM and a target rating of perceived exertion (RPE) of ≥3 on a Borg CR-10 RPE scale. ¹⁶

All measures were obtained on the nondominant limb (left leg for all participants), defined as the stance leg when kicking a ball for maximal distance. Ultimately, the drop vertical jump (DVJ) task was performed with a drop-off technique that has been shown to minimize between-limb asymmetries and expected no systematic difference between limbs. ⁵² The stance limb was chosen for analysis based on evidence that the knee of the stance limb is more often injured via noncontact mechanism in female athletes, ¹⁷ and there is minimal evidence or large DFROM asymmetries among healthy athletes.

The clinical measure of ankle DFROM (DF-CLIN) was assessed in accordance with the procedures previously described by Bennell et al. ⁹ These methods were chosen because they utilized the closed kinetic chain position with the knee flexed, which is thought to more closely mimic the state of the knee joint during the activities that likely lead to the development of PFPS and/or ACL injury (eg, stance phase of running, repetitive landing from a jump) compared with unloaded open-chain assessment methods. Before measuring DFROM, a mark was placed on the anterior surface of the tibia, 10 cm distal to the tibial tuberosity, to ensure consistent placement of a standard bubble inclinometer. Participants then performed two 30-second repetitions of a standing calf stretch to reduce measurement variability. Standing barefoot and facing a wall, the participant’s left foot was then placed along a measuring tape secured to the floor, with the distal edge of their second toe positioned 10 cm from the wall and their heel centered. Once positioned, participants were instructed to lunge forward and attempt to touch their knee to a line on the wall without allowing their heel to break contact with the ground (Figure 1). When the participants were able to successfully make knee contact with the wall without lifting their heel, their foot was systematically moved away from the wall in 2-mm increments until they could no longer maintain heel contact while touching their knee to the wall; this process was then repeated by systematically moving their foot forward in 2-mm increments until they were able to successfully touch their knee to the wall while maintaining heel contact. At this point, the inclinometer was placed on the previously made mark on the anterior tibia and the DFROM angle was then recorded to the nearest degree. This process was repeated 3 times, and an average of the 3 trials was then calculated and used for analysis. Ankle DFROM measures were made by 1 examiner who had previously established excellent day-to-day reliability (intraclass correlation coefficient_2,k = 0.97, standard error of the mean = 0.4°).

Figure 1.

Participant setup for clinical measure of ankle dorsiflexion range of motion (DF-CLIN) measurement. DF-CLIN, clinical measure of ankle DFROM; DFROM, dorsiflexion range of motion.

Knee joint landing biomechanics were assessed during a DVJ task. To control for the potential effects of footwear on landing biomechanics, all participants were provided with standardized footwear (Adidas, Uraha 2, Adidas North America). Participants were instrumented with 4-marker clusters of optical light-emitting diode (LED) markers (PhaseSpace) so that 3-dimensional kinematic data could be obtained using an 8-camera IMPULSE motion-tracking system (PhaseSpace). Specifically, marker clusters were placed on the sacrum, lateral thigh (mid-shaft), lateral shank (mid-shaft), and foot, of the left lower extremity. A 4-segment biomechanical model (pelvis, thigh, shank, and foot) was constructed using MotionMonitor software (Version 8.77; Innovative Sports Training).

The DVJ was completed using a 0.45-m box placed 0.1 m behind the rear edge of 2 force platforms (Type 4060-130; Bertec Corporation). Standing in a neutral position along the leading edge of the box, participants were instructed to (1) drop straight down off of the box and land evenly with 1 foot on each force platform, (2) immediately jump straight into the air for maximum vertical height on landing, and (3) land evenly once again with 1 foot on each platform. Additionally, participants were required to keep their hands at ear-level throughout each trial to minimize the influence of the upper extremity on landing biomechanics and to not obscure the LED markers. To prevent experimenter bias, the investigators did not provide participants with any feedback regarding their landing mechanics. Participants performed 3 practice trials followed by 5 successful test trials in which data were recorded. A trial was considered successful if the participant (1) dropped straight down off of the box and evenly landed with 1 foot on each platform, (2) jumped for maximum vertical height immediately after landing, and (3) landed back on the force platforms after the vertical jump. Unsuccessful trials were discarded and repeated.

Data Analysis

Biomechanical data were collected and processed using MotionMonitor software. Kinetic and kinematic data were sampled at 1000 and 240 Hz, respectively, and subsequently low-pass filtered at 12 Hz using a fourth-order zero-lag Butterworth filter. Ankle and knee joint centers were determined as the midpoint between the medial and lateral malleoli and medial and lateral femoral epicondyles, respectively. Hip joint centers were determined using the Bell method. ⁶ A segmental reference system was defined within MotionMonitor, with the z-axis as the medial-lateral axis (flexion-extension), the y-axis as the distal-proximal/longitudinal axis (internal-external rotation), and the x-axis as the anterior-posterior axis (abduction-adduction). Joint motions were then quantified using Euler’s equations, with a rotational sequence of Z Y′ X′′. ³¹ Joint moments were calculated using inverse dynamics. ²⁶ These data were then exported to MATLAB (Mathworks, Inc) for data reduction using custom-written code.

Data from the first landing of the DVJ was averaged across the 5 trials and used for analysis. Biomechanical variables of interest included (1) peak hip flexion, knee flexion, knee abduction, and ankle dorsiflexion angles (ie, from initial ground contact to the instant at which vertical ground reaction force first fell to less than 10 N); (2) sagittal plane hip, knee, and ankle excursion values (ie, peak angle minus angle at initial contact when vertical ground reaction force first exceeded 10 N); (3) peak hip flexion, knee flexion, knee abduction, and ankle dorsiflexion external joint moments normalized to the product of body weight (N) and height (cm); and (4) sagittal plane hip, knee, and ankle energy absorption. Additionally, we used the peak ankle dorsiflexion angle to calculate the percentage of available ankle DFROM used during landing (DF-%USED = peak ankle dorsiflexion during landing/ankle DFROM measured clinically). Eccentric work (energy absorption) values were calculated according to the methods previously described by Schmitz and Shultz. ⁴⁷ Briefly, joint powers were calculated as the product of joint angular velocity and joint moment. Eccentric work done by the extensor muscles groups was calculated by integrating the negative portion of the joint power curve as this represented energy absorption by the extensor muscles. Eccentric work done by the extensor muscle groups of each joint was reported as a positive value and normalized to the product of body weight (N) and height (cm).

Statistical Analysis

All statistical analyses were performed in SPSS, Version 26 (IBM Corp) with statistical significance set a priori at α = 0.05. Three measures of ankle DFROM were compared with the landing biomechanics: (1) peak ankle DFROM during landing, (2) DF-CLIN, and (3) DF-%USED. Because these 3 measures were expected to be correlated with one another, we examined each independently. For each measure, univariate linear regressions were performed to identify whether the DFROM measure predicted the biomechanical variables of interest. The strength of the R² values (the portion of variance explained by DFROM) was interpreted based on <0.02 as a very small effect, 0.02 to 0.15 a small effect, 0.15 to 0.25 a moderate effect, and >0.25 a large effect. ¹⁸

Results

Descriptive statistics of biomechanical measures are found in Tables 1 and 2. Greater peak ankle dorsiflexion during landing (mean: 14.5° ± 4.9°) was predictive of greater peak knee flexion angle (r² = 0.36, P = 0.007), greater knee flexion excursion (r² = 0.22, P = 0.05), lesser peak knee abduction angle (r² = −0.26,P = 0.03), and greater knee extensor work (r² = 0.37, P = 0.006) during the DVJ (Figure 2), all of which would be interpreted as moderate to large effects. For every 1.0° greater of ankle DFROM during landing, there was a 1.7° increase in peak knee flexion angle, 1.0° increase in knee flexion excursion, 0.6° decrease in knee abduction angle, and 0.002 N·m·N⁻¹·cm⁻¹ increase in knee extensor work. Peak ankle DFROM was not predictive of any peak joint moments.

Table 1.

Descriptive biomechanics of the left lower extremity during a drop vertical jump

	Peak Angle, deg	Excursion, deg	Peak Moment (N·N−1·m−1)
Hip flexion	63.3 ± 12.7	49.0 ± 15.1	0.10 ± 0.03
Knee flexion	80.1 ± 13.8	76.0 ± 10.4	0.07 ± 0.02
Knee abduction	6.7 ± 5.4	7.9 ± 4.9	0.06 ± 0.09
Ankle dorsiflexion	14.5 ± 4.9	59.7 ± 8.3	0.08 ± 0.2

Table 2.

Eccentric work of hip, knee, and ankle of the left lower extremity during a drop vertical jump

	Eccentric Work (N·m·N−1·cm−1)
Hip extensor	0.04 ± 0.02
Knee extensor	0.05 ± 0.02
Ankle plantarflexor	0.06 ± 0.01

Figure 2.

The relationship between peak ankle dorsiflexion during landing and select biomechanical variables. (a) Peak knee flexion, (b) knee flexion excursion, (c) peak knee abduction, and (d) knee extensor work.

Greater DF-CLIN (mean: 47.3° ± 7.1°) was also moderately to strongly predictive of greater hip flexion excursion (r² = 0.21, P = 0.05), peak knee flexion angle (r² = 0.37, P = 0.006), knee flexion excursion (r² = 0.40, P = 0.004), peak ankle dorsiflexion angle (r² = 0.41, P = 0.003), hip extensor work (r² = 0.28, P = 0.02), and knee extensor work (r² = 0.21, P = 0.05) (Figure 3). For every 1.0° more of DF-CLIN, there was a 1.0° increase in hip flexion excursion, 1.2° increase in peak knee flexion angles, 0.9° increase in knee flexion excursion, 0.002 N·m·N⁻¹·cm⁻¹ increase in hip extensor work, and 0.001 N·m·N⁻¹·cm⁻¹ increase in knee extensor work (r² = 0.21, P = 0.05). DF-CLIN was not predictive of any peak joint moments.

Figure 3.

The relationship between clinical measure of ankle dorsiflexion range of motion (DF-CLIN) and select biomechanical variables. (a) Hip flexion excursion, (b) peak knee flexion, (c) knee flexion excursion, (d) peak ankle dorsiflexion, (e) hip extensor work, and (f) knee extensor work.

Greater DF-%USED (mean: 30.5% ± 8.6%) was predictive of lesser peak knee abduction angle (r² = 0.26, P = 0.03) and greater knee extensor work (r² = 0.25, P = 0.03) (Figure 4), but not of any peak kinematics or joint moments. For every 10% more of DF-%USED, there was a 3.2° reduction in peak knee abduction angles and 0.01 N·m·N⁻¹·cm⁻¹ increase in knee extensor work.

Figure 4.

The relationship between the percentage of available ankle dorsiflexion used during landing (DF-%USED) and select biomechanical variables during the drop vertical jump. (a) Peak knee abduction range of motion and (b) knee extensor work. DFROM, dorsiflexion range of motion.

Discussion

Ankle DFROM restrictions are common after injury and/or immobilization to the ankle and foot. ¹ The findings of the current study suggest that restricted ankle DFROM is associated with knee and hip biomechanics that may lead to differential loading patterns and subsequent injury. Specifically, lower maximum ranges of ankle DFROM is associated with reduced hip and knee flexion range of motion and knee extensor work during landing from a jump. Additionally, underutilizing available ankle DFROM during landing is associated with aberrant frontal plane movement such as increased knee abduction motion. Thus, both the attainment of maximum physiologic joint range of motion and control and utilization of available ankle DFROM warrant clinical attention.

Data from this study are consistent with previous investigations indicating abnormal sagittal and frontal plane movement patterns resulting from reduced ankle dorsiflexion. During squatting or landing activities, the body’s center of mass is descending, forcing flexion of the hip, knee, and ankle joints. Evidence indicates that instead of compensating for restricted ankle DFROM by increasing flexion displacements of other proximal joints, athletes display a pattern of reduced flexion excursions throughout the lower extremity.^19-21,43,49 While this pattern may improve athletic performance (by decreasing the landing time associated with lower extremity joint flexion), it may have significant ramifications on risk for future injury. Stiff-legged landing patterns have been closely linked to knee injuries,^14,34 potentially because of the greater vertical ground reaction forces, knee extensor moments, and tibiofemoral anterior shear and compression forces associated with this movement pattern.^2,50 Similarly, reduced ankle DFROM can promote aberrant frontal plane motions that may further heighten an athlete’s injury risk, as prior studies have connected ankle DFROM restrictions with increased frontal plane knee motion during squatting^7,36 and landing from a jump.^48,49 Thus, it appears that the compensatory motions resulting from ankle DFROM deficits can be exhibited as foot/ankle pronation, knee abduction, or hip adduction.^{7,27,36,43,48,49} Promoting optimal, healthy ankle DFROM may lead to improved lower extremity biomechanics, ultimately reducing injury risk.

The current study intentionally measured ankle DFROM in a weightbearing flexed-knee position, to best simulate functional loading patterns that occur during gait, jumping, and landing. Other studies have used nonweightbearing ankle DFROM measurements.^7,21,48 Weightbearing measures, like the ones we employed in this study, allow for larger external applied torque and peak ranges 2 to 4 times that of nonweightbearing measures.^5,42 Nonweightbearing procedures also likely affect the extent to which subtalar motion or eversion contributes to the measure. Rabin and Kozol ⁴² reported only a moderate correlation between weightbearing and nonweightbearing measures, suggesting that the 2 methods may provide different information. To that end, Dill et al ¹⁹ reported no differences in lower extremity biomechanics during squatting and jumping tasks between groups characterized by nonweightbearing ankle DFROM measures; however, connections with sagittal plane knee and ankle angles were observed when characterized by weightbearing ankle DFROM measures. Additionally, knee position is an important factor in weightbearing measures considering that the gastrocnemius can restrict ankle DFROM when the knee is in an extended position. Flexing the knee to 20° can remove the influence of the gastrocnemius and more closely simulate functional movement patterns. ⁵ As such, nonweightbearing measures may help guide clinicians in understanding physiologic motion, but a weightbearing measure of ankle DFROM, with the knee in a flexed position, should be considered by clinicians and researchers when attempting to simulate functional movement patterns.

Appreciating the extent to which the available DFROM is utilized during movement may also help inform treatment options and clinical decision-making. Our data show that less than 50% of available range of ankle dorsiflexion is used during the initial landing phase of a standard 2-legged DVJ. Thus, it may not be that an athlete does not have adequate DFROM, but rather that an athlete does not fully use their available range of motion. This suggests that a focus on motor control (utilizing more dorsiflexion during landing, getting the knees out over the toes) may be of more significant and immediate benefit than a focus on increasing the end-range extensibility of the joint. The greater the amount of ankle DFROM observed during landing may promote increased knee and hip flexion motion and reduce aberrant knee abduction motion, ultimately resulting in a safer movement pattern.

There has been limited work on modifying foot/ankle movement patterns during jumping and landing, yet evidence at other joints may be applied to the ankle. Changing inherent movement patterns can be challenging, especially a complex movement that is reliant on synergistic movement from the ankle, knee, hip, and pelvis/trunk. Movement retraining at the ankle can use progressive loading from simple weightbearing exercises (squatting, step up/down activities) to more dynamic activities (landing, cutting) to encourage greater ankle dorsiflexion and subsequent knee and hip flexion. This should be supervised to guide the athlete into true sagittal plane ankle motion with minimal frontal or transverse plane motion at the ankle. Augmented feedback, in the form of verbal, visual, or auditory cueing, has shown promise in improving high-risk biomechanical movement patterns for lower extremity injury. ⁴ Clinicians can provide this feedback with either an internal (attention to the athlete’s body) or external (attention to the environment) focus. Internal verbal cues used during jump landings like “bend your ankles” may promote greater flexion at all lower extremity joints. External verbal cues are less intuitive, yet some evidence would suggest that an external focus may promote higher levels of change than an internal focus.^8,24 Using the cueing strategy “push yourself as hard as possible off the ground after landing” has shown to increase lower extremity flexion and reduce knee abduction during a plyometric jump landing, with improvements retained a week after testing. ⁵¹

Limitations

There are limitations with this study’s population and measurement methods. The study population was limited to a relatively small sample of healthy college-aged women with normal physiological DFROM. Future studies should attempt to study the effects of DFROM in both healthy and injured populations stratified by age and sex. While women are more likely to sustain ACL injury and PFPS, men are not excluded from risk. With a larger sample, a multivariate analysis including both the clinical and utilization measure of DFROM may help identify relationships between the 2 measures and resultant biomechanics. Additionally, we did not control for time of the menstrual cycle, which has been shown to influence knee joint biomechanics. ⁴⁶ However, our study was limited to within subject associations, and we are not aware of any data to suggest that the joint would be affected differently that the other. This should be confirmed in future research.

Despite the moderate to strong associations observed between dorsiflexion ROM and biomechanics (explaining 20%-40% of variance), there remains a substantial proportion of variance unexplained. It is well accepted that lower extremity landing biomechanics can be influenced by multiple factors. For example, we did not assess potential compensatory motions of foot pronation in the frontal and transverse planes that could also potentially influence knee motion. ⁴⁴ Future multivariate studies should examine the relative contribution of DFROM among these other factors.

Conclusion

Reduced ankle DFROM is associated with lower extremity biomechanics during a jump landing task in healthy female athletes. Specifically, lower levels of ankle dorsiflexion were associated with corresponding lower levels of knee and hip flexion and hip and knee extensor work absorption and aberrant frontal plane knee motion. Moreover, it appears as though physically active female athletes do not utilize the entire physiologically available ankle DFROM during jump landings. Thus, clinical focus may need to shift toward motor control strategies once adequate physiologic motion is available.