Abstract
Background
Patients with isolated ankle osteoarthritis (OA) often demonstrate disturbed ankle biomechanics during walking. Clinicians often believe that this triggers the distal foot joints to compensate these altered ankle biomechanics and that these foot joints are consequently subjected to degenerative joint diseases due to overuse.
Questions/purposes
Do patients with isolated ankle OA differ from those without ankle OA in terms of (1) ankle and foot joint kinematics and (2) ankle and foot joint kinetics as measured using three-dimensional (3-D) gait analysis? (3) Do these patients demonstrate compensatory strategies in their Chopart, Lisfranc, or first metatarsophalangeal joints in terms of increased joint kinematic and kinetic outputs?
Methods
Between 2015 and 2018, we treated 110 patients with unilateral ankle OA, and invited all of them to participate in the gait analysis laboratory. Of those, 47% (52) of patients did so, and of these, 16 patients met the inclusion criteria for this study, which were (1) diagnosis of unilateral ankle OA; (2) absence of radiographical signs of OA in the contralateral foot or lower limbs; (3) ability to walk at least 100 m without rest; and (4) being older than 18 years of age. A control group (n = 25) was recruited through intranet advertisements at the University Hospitals of Leuven. Participants were included if their age matched the age-range of the patient group and if they had no history of OA in any of the lower limb joints. Patients were slightly older (55.9 ± 11.2 years), with a slightly higher BMI (28 ± 6 kg/m2) than the control group participants (47.2 ± 4.4 years; p = 0.01 and 25 ± 3 kg/m2; p = 0.05). All participants underwent a 3-D gait analysis, during which a multisegment foot model was used to quantify the kinematic parameters (joint angles and ROM) and the kinetic parameters (rotational forces or moments), as well as power generation and absorption in the ankle, Chopart, Lisfranc, and first metatarsophalangeal joints during the stance phase of walking. Peak values were the maximum and minimum values of waveforms and the latter were time-normalized to 100% of the stance phase.
Results
Regarding joint kinematics, patients demonstrated a sagittal plane ankle, Chopart, Lisfranc, and first metatarsophalangeal joint ROM of 11.4 ± 3.1°, 9.7 ± 2.7°, 8.6 ± 2.3° and 34.6 ± 8.1°, respectively, compared with 18.0 ± 2.7° (p < 0.001), 13.9 ± 3.2° (p < 0.001), 7.1 ± 2.0° (p = 0.046) and 38.1 ± 6.5° (p = 0.15), respectively, in the control group during the stance phase of walking. With regard to joint kinetics in the patient group, we found a mean decrease of 1.3 W/kg (95% CI confidence interval 1.0 to 1.6) (control group mean: 2.4 ± 0.4 W/kg, patient group mean: 1.1 ± 0.5 W/kg) and 0.8 W/kg (95% CI 0.4 to 1.0) (control group mean: 1.5 ± 0.3 W/kg, patient group mean: 0.7 ± 0.5 W/kg) of ankle (p < 0.001) and Chopart (p < 0.001) joint peak power generation. No changes in kinetic parameters (joint moment or power) were observed in any of the distal foot joints.
Conclusion
The findings of this study showed a decrease in ankle kinematics and kinetics of patients with isolated ankle OA during walking, whereas no change in kinematic or kinetic functions were observed in the distal foot joints, demonstrating that these do not compensate for the mechanical dysfunction of the ankle.
Clinical Relevance
The current findings suggest that future experimental laboratory studies should look at whether tibiotalar joint fusion or total ankle replacement influence the biomechanical functioning of these distal joints.
Introduction
Ankle osteoarthritis (OA) is associated with stiffness and pain and often leads to maladaptive gait strategies [25]. Three-dimensional (3-D) gait analysis is the ideal methodologic tool to identify these maladaptive gait patterns and has received increasing attention with regard to ankle OA [2, 13, 16, 24, 30]. These maladaptive patterns in patients arise, for example, from compensation of the subtalar or Chopart joints, which might induce deviations in the ankle’s axes [31]. Three-dimensional gait studies of ankle OA have suggested that OA-induced stiffness leads to reduced ankle ROM during the entire stance phase, and walking speed is reduced in patients with ankle OA, which may be a function of a (mal)adaptive response to pain during walking [32].
However, to fully understand the effect of OA on the lower limb’s characteristics during walking, kinetic and mechanical function must be investigated as well. The latter has already been explored in patients with knee OA, with findings suggesting that these patients experience an increased external knee adduction moment during walking and therefore encounter greater mechanical knee loads [1, 28]. However, to our knowledge, less is known about the kinetic and mechanical function of patients with ankle OA while walking. This is primarily because multisegment kinetic foot models have been developed only recently, revealing the complicated functioning of the ankle and foot [5, 20, 26]. Developments of multisegment kinetic foot models have led to the conclusion that the ankle is the driver of the lower limb’s kinetic chain functioning because this joint takes on 50% of the total power generation during walking [6]. Therefore, investigating the kinetic behavior of the foot and ankle complex during walking would have several additional (currently hypothetical) benefits for patients with ankle OA, such as presurgical planning, individual management of nonsurgical interventions, and general follow-up. In addition, these multisegment kinetic foot models might allow us to investigate the effect of ankle OA on different foot joints. Thus, it would be possible to explore whether the more distally located foot joints compensate for loss of ankle function in patients with OA. However, before these factors can be investigated, the kinetic patterns of the foot and ankle must be explored in patients with ankle OA. Hence, we wished to explore the multisegment kinetics of the foot during walking in patients with isolated ankle OA compared with those of asymptomatic controls.
Therefore, we asked: Do patients with isolated ankle OA differ from those without ankle OA in terms of (1) ankle and foot joint kinematics and (2) ankle and foot joint kinetics as measured using 3-D gait analysis? (3) Do these patients demonstrate compensatory strategies in their Chopart, Lisfranc, or first metatarsophalangeal joints (MTP 1) in terms of increased joint kinematic and kinetic outputs?
Patients and Methods
Participants
All participants provided informed consent, and the ethical committee of the university hospital where the study was conducted approved it (S55070; ML9038). Between 2015 and 2018, 110 patients in our treatment center with some form of ankle OA were invited to participate in the gait analysis laboratory to collect a prospective database. Of those, 47% (52) of patients enrolled. From this database, 16 patients met the inclusion criteria for the current retrospective study, which were (1) diagnosis of unilateral ankle OA by two senior orthopaedic surgeons (GM, SW), (2) absence of radiographic signs of OA in other foot or lower limb joints, (3) ability to walk at least 100 m without rest, and (4) being older than 18 years of age. Patients (n = 16) were included if they had a radiograph-confirmed Grade 4 Kellgren-Lawrence OA in only one tibiotalar joint and if they had not responded to nonoperative treatment. Therefore, these patients represented a population of patients with end-stage arthritis. Patients had either primary (n = 3) or posttraumatic ankle OA (n = 13) and had a median (range) radiographically determined Kellgren-Lawrence score of 4 (4 to 4), 1 (0 to 2), 0 (0 to 1), and 0 (0 to 2) in the tibiotalar, subtalar, talonavicular, and calcaneocuboid joints, respectively. The median (range) VAS-pain score was 7.0 (3.0 to 8.5) (Table 1).
Table 1.
Participant’s demographical, spatiotemporal, functional, and radiographical data
| Variable | Control (n = 25) | Ankle OA patients (n = 16) | p value | Mean difference (95% CI interval) |
| Age in years | 47 ± 4 | 56 ± 11 | 0.01 | 9 (4 to 14) |
| BMI in kg | 72 ± 13 | 82 ± 19 | 0.10 | 10 (0 to 21) |
| BMI in kg/m2 | 25 ± 3 | 28 ± 6 | 0.05 | 3 (0 to 6) |
| Side, left/right | 12/13 | 7/9 | ||
| Gender (M) | 10 | 9 | ||
| Pathogenesis | ||||
| Primary arthrosis | 3 | |||
| Posttraumatic | 13 | |||
| Speed (m/s) | 1.26 ± 0.13 | 0.99 ± 0.24 | 0.002 | 0.27 (0.15 to 0.39) |
| Stride length (m) | 1.32 ± 0.11 | 1.15 ± 0.16 | 0.004 | 0.17 (0.09 to 0.28) |
| Stride time (s) | 1.05 ± 0.07 | 1.20 ± 0.017 | 0.006 | 0.15 (0.11 to 0.19) |
| Step time (s) | 0.52 ± 0.04 | 0.60 ± 0.08 | 0.003 | 0.08 (0.04 to 0.12) |
| VAS pain score (10-point scale) | 6 ± 2 | |||
| Adjusted FFI pain subscore (36-point scale) | 15 ± 5 | |||
| Adjusted FFI activity subscore (36-point scale) | 16 ± 7 | |||
| Kellgren-Lawrence classifications of OA | ||||
| Tibiotalar joint | 4 ± 0 | |||
| Subtalar joint | 1 ± 1 | |||
| Talonavicular joint | 0 ± 0 | |||
| Calcaneocuboid joint | 0 ± 0 | |||
| Tibiotalar surface angle (°) | 86 ± 9 | |||
| Tibiocrural angle (°) | 11 ± 3 |
All outcomes are presented as the mean ± SD; p values presented are a result of the Student’s t-test (α = 0.05); significance (p < 0.05); FFI = Foot Function Index.
A control group, consisting of 25 healthy volunteers, was recruited through intranet advertisements at the University Hospitals of Leuven. These healthy participants were included if their age matched the age-range of the patient group (between 40 and 70 years of age) and if they had no history of OA in any of the lower limb joints based on their available medical history and subjective assessment. Patients were slightly older (55.9 ± 11.2 years), with a slightly higher BMI (28 ± 6 kg/m2) than the control group participants (47.2 ± 4.4 years; p = 0.01 and 25 ± 3 kg/m2; p = 0.05) (Table 1). These differences were not expected to influence ankle biomechanics based on the conclusions of earlier research [4].
Radiographic Data Collection and Analysis
Radiographic measurements of the patients were taken at the first consultation and at a maximum of 3 months before their gait analysis. These measurements consisted of standard weightbearing plain radiographs in the AP and mediolateral orientations and a mortise view.
Radiographic measurements were analyzed by an experienced foot and ankle-dedicated orthopaedic surgeon (SW). The following radiographic parameters were collected to characterize structural abnormalities of the rearfoot in patients with ankle OA: Kellgren-Lawrence grade [15], tibiotalar surface angle [21], and tibiocrural angle [29]. According to Kellgren and Lawrence’s classification system [15], OA is graded as Grade 0: no radiographic features of ankle OA; Grade 1: doubtful joint space narrowing and possible osteophytic lipping; Grade 2: definite osteophytes and possible joint space narrowing on AP weightbearing radiographs; Grade 3: multiple osteophytes, definite joint space narrowing, sclerosis, and possible bony deformity; or Grade 4: large osteophytes, marked joint space narrowing, severe sclerosis, and definite bony deformity.
Primary and Secondary Study Outcomes
Our primary study outcome was the difference between patients with ankle OA and those without OA in terms of kinematics (ROM during stance phase) in the ankle, Chopart, Lisfranc, and MTP 1 joints. We assessed this by subtracting the maximum value from the minimum value in terms of joint angles during the stance phase of walking, which was measured using a multisegment foot model during 3-D gait analysis.
Our secondary study outcome was the difference between patients with ankle OA and those without OA in terms of kinetics (internal joint moment [Nmm/kg], angular velocity [°/s] and joint power [W/k]) in the ankle, Chopart, Lisfranc, and MTP 1 joints. We assessed this by measuring the joint kinetics using a multisegment foot model during 3-D gait analysis.
Collection and Analysis of Multisegment Foot Kinetic Data
Gait was analyzed at a clinical motion analysis laboratory equipped with a 10-meter walkway surrounded by a passive optoelectronic motion analysis system (Vicon Motion System, Ltd., Oxford Metrics, Oxford, UK) consisting of 10 T-10 cameras (1 megapixel, captures 10-bit grayscale using 1120 * 896 pixels, sampled at 100 Hz) to track the motion of markers. The maximum absolute difference of 3-D coordinate representation was < 0.88 mm in a dynamic condition. In the middle of the walkway, a force plate (Advanced Mechanical Technology Inc, Watertown, MA, USA) was integrated with a pressure plate (Footscan, dimension 0.5 m x 0.4 m, 4096 sensors, 2.8 sensors per cm2, RSscan International, Olen, Belgium) placed on top. Data of the force and pressure plates were synchronized using the RSscan 3D Box© (RSscan International).
Retroreflective markers (Ø = 10 mm) were placed on the participants according to the marker placement protocol of the Rizzoli foot model [17]. The measurement of kinetics in multiple foot joints using this multisegment foot model has been validated in a similar patient population in an earlier study [9]. A detailed explanation of the used model and marker placement is given in a supplementary file to enhance readability (see Supplemental Digital Content, http://links.lww.com/CORR/A406). Participants performed five representative walking trials at a self-selected, comfortable speed. For each trial, the stance phase during which the targeted feet of all participants contacted the pressure platform was selected.
To analyze the ankle and foot kinetics of each participant, first, we projected the position of the foot markers on the floor to determine the boundaries of the respective foot segments: rearfoot, midfoot, forefoot, and toes. The intersegmental angle calculations in this model were as follows: ankle (shank-calcaneus angle), Chopart (calcaneus-midfoot angle), Lisfranc (midfoot-metatarsus angle), first metatarsophalangeal (MTP 1) angle, and calcaneus-metatarsus segment angle. The latter was included to investigate motion of the calcaneus and forefoot. The following joint center definitions were applied: ankle (midpoint between the medial and lateral malleoli), Chopart (midpoint between the cuboid and navicular bone), Lisfranc (base of the second metatarsal), and MTP 1 (projection of the MTP 1 marker halfway to the floor). Next, ground reaction forces and moments were distributed over the different segments of the Rizzoli foot model using the proportionality scheme validated by Eerdekens et al. [10] in a similar study population. For every time frame, the resulting pressure in each of these segments, compared with the total pressure, provided the proportion of the total ground reaction force to each corresponding segment. Then, we calculated inertial parameters based on the mass of each segment and their geometric solids [14]. The mass of the foot was distributed at a 30/30/30/10 (rearfoot/midfoot/forefoot/hallux) percent rate. Finally, joint kinetics were computed starting from the MTP1 joint and progressing proximally, using Newton-Euler equations using an in-house custom inverse dynamic analysis program (ACEP_Manager, Leuven, Belgium). The following outcome variables of the four foot joints were investigated: ROM during the stance phase, internal plantarflexion and dorsiflexion moments (waveforms and peaks), angular plantarflexion and dorsiflexion velocity (waveforms and peaks), joint power generation and absorption (waveforms and peak), and the relative segmental influence of positive (amount of generation) and negative (amount of absorption) mechanical work (which is the product of joint force and displacement [J/kg]). Peak discrete variables were considered to be the maximum and minimum values of the waveforms during the stance phase. For the waveforms, time was normalized to 100% of the stance phase.
Statistical Analysis
An unpaired t-test (α = 0.05) and mean difference (with 95% CIs) were calculated in SPSS 20.0 (IBM Corp, Armonk, NY, USA) to analyze group differences between the zero-dimensional parameters of the control and patient groups. Differences among the full kinematic and kinetic waveform patterns were assessed using a one-dimensional statistical parametric mapping unpaired t-test (α = 0.05), and the assessment was performed with an open-source code (www.spm1d.org) in Matlab (R2014a, 8.3.0.532, Mathworks Inc, Natick, MA, USA).
Results
Joint Kinematics
The mean (± SD) ankle ROM in the sagittal plane (11.4 ± 3.1°), frontal plane (4.7 ± 1.8°) and transverse plane (7.0 ± 2.7°) in the patient group were all lower compared with the control group with a mean difference of 6.6° (95% CI 4.8 to 8.5; p < 0.001), 2.4° (95% CI 1.3 to 3.5; p < 0.001) and 4.0° (95% CI 2.3 to 5.7; p < 0.001), respectively. The mean Chopart joint ROM in the sagittal plane (9.7 ± 2.7°) and transverse plane (4.4 ± 1.6°) in the patient group were all lower compared with the control group with a mean difference of 4.2° (95% CI 2.2 to 6.2; p < 0.001) and 4.4° (95% CI 3.1 to 5.7; p < 0.001), respectively, whereas the frontal plane ROM not differ between the patient group (4.9 ± 1.5°) and control group (5.5 ± 1.3°; p = 0.17). We further observed a Lisfranc joint sagittal plane ROM of 8.6 ± 2.3° in the patient group and 7.1 ± 2.0° (p = 0.046) in the control group. We observed no differences between the sagittal plane ROM of the MTP 1 joint in the patient group (34.6 ± 8.1°) and control group (38.1 ± 6.5°; p = 0.15) (Table 2).
Table 2.
ROM (°) during stance phase of walking (mean ± SD)
| Variable | Control | Ankle OA patients | p | Mean difference (95% confidence interval) |
| Ankle | ||||
| Sagittal | 18 ± 3 | 11 ± 3 | < 0.001 | 7 (5 to 9) |
| Frontal | 7 ± 2 | 5 ± 2 | < 0.001 | 2 (1 to 4) |
| Transverse | 11 ± 3 | 7 ± 3 | < 0.001 | 4 (2 to 6) |
| Chopart | ||||
| Sagittal | 14 ± 3 | 10 ± 3 | < 0.001 | 4 (2 to 6) |
| Frontal | 5.5 ± 1.3 | 4.9 ± 1.5 | 0.17 | 0.6 (-0.3 to 1.5) |
| Transverse | 8.8 ± 2.3 | 4.4 ± 1.6 | < 0.001 | 4.4 (3.1 to 5.7) |
| Lisfranc | ||||
| Sagittal | 7.1 ± 2.0 | 8.6 ± 2.3 | 0.046 | 1.5 (0.0 to 3.0) |
| Frontal | 3.3 ± 0.9 | 5.0 ± 1.7 | < 0.001 | 1.7 (0.9 to 2.5) |
| Transverse | 5.5 ± 2.4 | 5.1 ± 2.1 | 0.60 | 0.4 (-1.1 to 1.9) |
| MTP 1 | ||||
| Sagittal | 38 ± 7 | 35 ± 8 | 0.15 | 3 (-1 to 8) |
| Calcaneus-metatarsus segment | ||||
| Frontal | 11 ± 2 | 7 ± 3 | < 0.001 | 4 (2 to 5) |
P values represent the outcome of the Student t-test (α = 0.05); significance = p < 0.05; MTP 1 = First metatarsophalangeal joint.
Examining the kinematic waveforms, we found that patients with ankle OA experienced stiffness most often during the second part of the stance phase (greater than 50% of the stance phase). There were differences in the sagittal and frontal planes of movement of the Lisfranc joint: patients had lower angles than control participants did, possibly because of the often-observed pes planovalgus position in these patients [27] (Fig. 1).
Fig. 1 (A-E).
These images show kinematic waveforms during the stance phase of walking in the patient group (dark gray line) and control group (light gray line) in the sagittal, frontal and transverse plane of the (A) ankle, (B) Chopart and (C) Lisfranc joints. (D) Shows the frontal plane kinematic waveform of the Calcaneus – Metatarsus segment and (E) shows the sagittal plane kinematic waveform of the first metatarsophalangeal (MTP 1) joint. SDs are visualized as a dark gray cloud for patients with ankle OA and light gray cloud for control participants. Statistical differences (p < 0.05) between waveforms according to the one-dimensional Statistical Parametrical t-test are presented underneath each figure as a black box; PF = plantarflexion; DF = dorsiflexion; Inv = inversion; Ev = eversion; Add = adduction; Abd = abduction.
Joint Kinetics
The mean (± SD) ankle joint peak internal moment (1269 ± 262 Nmm/kg), peak plantarflexion (PF) velocity (106 ± 36°/s), peak dorsiflexion velocity (52 ± 18°/s), peak power generation (1.1 ± 0.5 W/kg) and peak power absorption (0.2 ± 0.2 W/kg) in the patient group were all lowered compared with the control group with a mean difference of respectively 252 Nmm/kg (95% CI 125 to 381; p < 0.001), 120°/s (95% CI 80 to 160; p < 0.001), 39°/s (95% CI 23 to 54; p < 0.001), 1.3 W/kg (95% CI 1.0 to 1.6; p < 0.001) and 0.3 W/kg (95% CI 0.1 to 0.4; p < 0.001) (Table 3).
Table 3.
Sagittal plane kinetics of ankle, Chopart, Lisfranc and MTP1 joints (mean ± SD)
| Variable | Control | Ankle OA patients | p | Mean difference (95% CI) |
| Ankle | ||||
| Peak internal moment (Nmm/kg) | 1521 ± 131 | 1269 ± 262 | < 0.001 | 252 (125 to 381) |
| Peak PF angular velocity (°/s) | 226 ± 73 | 106 ± 36 | < 0.001 | 120 (80 to 160) |
| Peak DF angular velocity (°/s) | 91 ± 26 | 52 ± 18 | < 0.001 | 39 (23 to 54) |
| Peak positive power (W/kg) | 2.4 ± 0.4 | 1.1 ± 0.5 | < 0.001 | 1.3 (1.0 to 1.6) |
| Peak negative power (W/kg) | 0.5 ± 0.2 | 0.2 ± 0.2 | < 0.001 | 0.3 (0.1 to 0.4) |
| Chopart | ||||
| Peak internal moment (Nmm/kg) | 1004 ± 115 | 860 ± 177 | 0.004 | 144 (49 to 239) |
| Peak PF angular velocity (°/s) | 178 ± 40 | 104 ± 40 | < 0.001 | 74 (47 to 100) |
| Peak DF angular velocity (°/s) | 51 ± 28 | 35 ± 12 | 0.04 | 16 (1 to 31) |
| Peak positive power (W/kg) | 1.5 ± 0.3 | 0.7 ± 0.5 | < 0.001 | 0.8 (0.4 to 1.0) |
| Peak negative power (W/kg) | 0.4 ± 0.2 | 0.4 ± 0.2 | 0.81 | 0.0 (-0.1 to 0.1) |
| Lisfranc | ||||
| Peak internal moment (Nmm/kg) | 480 ± 83 | 407 ± 137 | 0.04 | 73 (2 to 144) |
| Peak PF angular velocity (°/s) | 48 ± 20 | 62 ± 26 | < 0.001 | 110 (95 to 125) |
| Peak DF angular velocity (°/s) | 56 ± 25 | 72 ± 35 | 0.12 | 16 (3 to 36) |
| Peak positive power (W/kg) | 0.3 ± 0.1 | 0.3 ± 0.2 | 0.77 | 0.0 (-0.1 to 0.1) |
| Peak negative power (W/kg) | 0.04 ± 0.03 | 0.05 ± 0.05 | 0.53 | 0.1 (-0.1 to 0.3) |
| MTP 1 | ||||
| Peak internal moment (Nmm/kg) | 181 ± 67 | 78 ± 82 | < 0.001 | 103 (54 to 151) |
| Peak PF angular velocity (°/s) | 248 ± 95 | 209 ± 127 | 0.29 | 43 (-28 to 115) |
| Peak DF angular velocity (°/s) | 397 ± 96 | 302 ± 123 | 0.01 | 96 (26 to 166) |
| Peak positive power (W/kg) | 0.1 ± 0.1 | 0.04 ± 0.06 | 0.01 | 0.06 (0.01 to 0.13) |
| Peak negative power (W/kg) | 0.8 ± 0.4 | 0.4 ± 0.5 | 0.002 | 0.4 (0.2 to 0.8) |
P values represent the outcome of the Student t-test (α = 0.05; significance = p < 0.05); MTP 1 = first metatarsophalangeal joint.
The mean (± SD) Chopart joint peak internal moment (860 ± 177 Nmm/kg), peak PF velocity (104 ± 40°/s), peak dorsiflexion velocity (35 ± 12°/s) and peak power generation (0.7 ± 0.5 W/kg) in the patient group were all lowered compared with the control group with a mean difference of respectively 144 Nmm/kg (95% CI 49 to 239; p = 0.004), 74°/s (95% CI 47 to 100; p < 0.001), 15.9°/s (95% CI 0.6 to 31.1; p = 0.04) and 0.8 W/kg (95% CI 0.4 to 1.0; p < 0.001). No differences were observed concerning the peak Chopart joint power absorption between the patient group (0.4 ± 0.2 W/kg) and the control group (0.4 ± 0.2 W/kg; p = 0.81) (Table 3).
In the Lisfranc joint, a mean difference of 73 Nmm/kg (95% CI 2 to 144) and 110°/s (95% CI 95 to 125) was observed between the patient group (407 ± 137 Nmm/kg and 62 ± 26°/s) and control group (480 ± 83 Nmm/kg and 48 ± 20°/s), respectively, for the peak internal moment (p = 0.04) and the peak PF velocity (p < 0.001) (Table 3).
The mean (± SD) MTP 1 joint peak internal moment (78 ± 82 Nmm/kg), peak dorsiflexion velocity (302 ± 123°/s), peak power generation (0.04 ± 0.06 W/kg) and peak power absorption (0.4 ± 0.5 W/kg) in the patient group were all lowered compared with the control group with a mean difference of respectively 103 Nmm/kg (95% CI 54 to 151; p < 0.001), 96°/s (95% CI 26 to 166; p = 0.01), 0.06 W/kg (95% CI 0.01 to 0.13; p = 0.01) and 0.4 W/kg (95% CI 0.2 to 0.8; p = 0.002). No difference was observed concerning the peak MTP 1 joint peak PF velocity between the patient group (209 ± 127°/s) and the control group (248 ± 95°/s; p = 0.29) (Table 3). Compared with the joint kinematic waveforms (Fig. 1), differences in kinetic patterns could already be observed at the beginning of the stance phase (Fig. 2). Patients with ankle OA demonstrated reduced ankle power absorption from 0% to 5% and from 50% to 70% of the stance phase and generated less ankle power at push-off, or 85% to 100% of the stance phase, because of reduced ankle joint moments and plantarflexion angular velocity during the same phase of stance (Fig. 2). Ankle OA predominantly altered the kinetics of the ankle, Chopart joint, and MTP 1 joint at the end of the stance phase. There were small differences in the Lisfranc joints of patients and controls between 78% and 81% of the stance phase (Fig. 2).
Fig. 2 (A-D).
These images show kinetic waveforms during the stance phase of walking in patients with ankle OA (dashed line) and control participants (full line), for the (A) ankle, (B) Chopart, (C) Lisfranc, and (D) first metatarsophalangeal (MTP 1) joints. SDs are visualized as a dark gray cloud for patients and a light gray cloud for control participants. Statistical differences (p < 0.05) between waveforms according to the one-dimensional Statistical Parametric Mapping t-test are presented underneath each figure as a black box; PF = plantarflexion; DF = dorsiflexion; Gen = generation; Abs = absorption.
The relative segmental influence on positive work was similar between patients and control individuals; the ankle in both cohorts took on 52% of the total positive work during the stance phase of walking. The Chopart, Lisfranc, and MTP 1 joints accounted for 30%, 16%, and 2% of the total positive work, respectively, in control participants, compared with 25% (Chopart joint), 22% (Lisfranc joint), and 1% (MTP 1 joint) in patients with ankle OA (Fig. 3). We observed distinct differences when examining the relative segmental influence on the total negative work; patients took on only 25% of the total negative work in the ankle, compared with 39% in control participants. Conversely, the Chopart joints in patients with ankle OA accounted for 45% of the total negative work, compared with only 24% in control participants (Fig. 3).
Fig. 3.
This figure compares the relative influence of the ankle and Chopart, Lisfranc, and first metatarsophalangeal joints on the total positive (left) and negative (right) work between the patient and the control group. Values are presented below the figure (%).
Discussion
It is often believed that in patients with end-stage isolated ankle OA, the distal foot joints are triggered to compensate for the altered ankle joint biomechanics during walking and that these foot joints are consequently subjected to degenerative joint diseases due to overuse. This study therefore compared the ankle and foot joint kinematics and kinetics during walking in these patients with healthy volunteers by means of a multisegment foot model in 3-D gait analysis. In addition, these findings gave us the opportunity to investigate whether the distal foot joints indeed compensate for ankle joint dysfunction in these patients. The key findings suggested that in the patient group, both the kinematic and kinetic parameters are decreased compared with the control group and that the distal foot joints did not demonstrate a compensatory strategy.
Limitations
Comparisons between the patient and control group seemed limited at first due to the differences in age and body mass. Yet, the current age difference of around 8.7 years between the two groups has been investigated to not affect ankle joint biomechanics during walking, especially since the maturation of the joints is completed at these ages [4]. The BMI and speed differences between the control group and patients are a typical representation of the clinical situation as patients with isolated end-stage ankle OA often have an increased BMI [18] and show lowered walking speeds [23]. Matching these parameters would therefore in our opinion make it far more difficult to valorize the findings outside of the study results because in real-life situations, these parameters differ as well. We also acknowledge that no radiographs were collected for the control group and thus the absence of ankle OA in this group was solely confirmed using the available clinical examination and history of these volunteers.
Another limitation to the current study is that the subtalar joint cannot be defined separately because of the limitations of multisegment kinetic foot modelling [20], but in vivo studies demonstrated that the subtalar joint does not particularly influence the sagittal plane kinetics [19]. We included only patients with a well-defined joint pathology, making it perhaps difficult the generalize these results beyond the study cohort. Another limitation was that only barefoot walking data were collected, although in real life, these patients often prefer to wear shoes while walking. We opted not to include shod walking analyses in this study because multisegment kinetic foot modelling is not yet considered accurate in shod conditions. Future studies could, however, make use of new developments concerning in-shoe measurements to investigate the effect of shoes on the distal foot joints in these patients, similar to a recent study in patients with hemophilic ankle arthropathy [8].
Kinematics
The ankle’s ROM during the stance phase of walking was lower for patients than for control individuals. Similar findings were observed in studies on ankle OA [12, 22, 32], and these confirmed the OA-induced stiffness of the ankle in our patient population. Regarding more distally located foot joints, we observed lower ROM of the Chopart joint during the stance phase. Contrarily, the Lisfranc joint in these patients had greater ROM. This might be a strategy to produce sufficient frontal plane movement during the late stance phase of walking to compensate for loss of frontal plane movement in the ankle and Chopart joint. Despite the latter observation, the overall kinematic findings suggest that the distal located foot joints do not compensate for the loss of ankle joint mobility in patients with isolated ankle OA. Clinicians could therefore perhaps note that they potentially do not have to focus on increasing the joint mobility of these distal foot joints if their aim is to improve the gait functionality in these patients. Future clinical RCT studies are needed to confirm this hypothesis.
Kinetics
We observed the key kinetic differences in the more proximally located joints, that is, the ankle and Chopart joint, as the joints kinetics were considerably lowered compared with the control group. It has already been reported that ankle peak power is lower in these patients during walking; Brodsky et al. [3] reported an ankle joint peak power generation of 1.3 ± 0.5 W/kg, compared with our result of 1.1 ± 0.5 W/kg [3]. Yet, the aforementioned study used a one-segment foot model to measure ankle kinetics, which is known to overestimate the ankle kinetics [11]. The Chopart joint demonstrated a similar amount of power absorption to that of control individuals and consequently did not compensate for the loss of power absorption of the ankle. This can also be confirmed when looking at other studies in which it was found that the Chopart joint absorbs around 0.35 W/kg [7] compared with the currently observed 0.4 W/kg. Based on current kinetic findings, patients with isolated ankle OA may tend to adapt a proximal strategy in which they rely more on the knees and hips to propulse while walking barefoot, that is, a hip strategy. This was already confirmed in patients with excessive muscle co-contractions at the ankle joint [33], but still needs to be confirmed in patients with isolated ankle OA.
Clinical Relevance
The results from this study could be used to test and improve existing rehabilitation programs that aim at enhancing gait in these patients. A specific exercise program in which the plantar flexors are trained might potentially improve the currently observed reduced ankle power generation during the push-off phase of stance. Improving functionality of the distal foot joints by increasing the currently lowered kinematic and kinetic parameters could perhaps also further unburden the ankle dysfunction. There is also currently increasing clinical interest in biomechanical classifications of orthopaedic and degenerative joint diseases. The current study provides a broad fundamental quantification of the biomechanical patterns of the foot and ankle in this specific pathology. It could therefore be interesting to compare these results with the biomechanical phenomena in other degenerative joint diseases to detect specific differences otherwise not found using current classification methods, such as radiographic imaging.
Conclusion
We found a decrease in ankle kinematics and kinetics during walking in patients with isolated ankle OA compared with a control group. We also found that the kinematics and kinetics were reduced in the distal foot joints of these patients and concluded that these did not compensate the ankle joint dysfunction during walking. Based on these preliminary findings, it might be possible for clinicians to consider adjusting treatment plans without additionally risking adjacent joint degeneration in the distal foot joints due to overuse or compensation. Future studies should investigate the effect of tibiotalar joint fusion or total ankle replacement on the biomechanics of these distal joints. These could perhaps also simultaneously measure muscle activity with a surface EMG to better depict the overall functioning of these joints during walking. Future laboratory studies should also perhaps combine a multisegment foot model and lower limb model to measure simultaneously the foot, ankle, knee and hip to investigate the interaction between these joints during walking.
Footnotes
The institution of one or more of the authors (ME) has received, during the study period, funding from Belgische Vereniging voor Orthopedie en Traumatologie.
Each author certifies that he has no commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Each author certifies that his institution approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.
This work was performed at UZ Leuven, Clinical Motion Analysis Laboratorium, Pellenberg, Belgium.
References
- 1.Astephen JL, Deluzio KJ, Caldwell GE, Dunbar MJ, Hubley-Kozey CL. Gait and neuromuscular pattern changes are associated with differences in knee osteoarthritis severity levels. J Biomech. 2008;41:868–76. [DOI] [PubMed] [Google Scholar]
- 2.Barton T, Lintz F, Winson I. Biomechanical changes associated with the osteoarthritic, arthrodesed, and prosthetic ankle joint. Foot Ankle Surg. 2011;17:52–57. [DOI] [PubMed] [Google Scholar]
- 3.Brodsky JW, Kane JM, Coleman S, Bariteau J, Tenenbaum S. Abnormalities of gait caused by ankle arthritis are improved by ankle arthrodesis. Bone Joint J. 2016;98:1369–1375. [DOI] [PubMed] [Google Scholar]
- 4.Chehab EF, Andriacchi TP, Favre J. Speed, age, sex, and body mass index provide a rigorous basis for comparing the kinematic and kinetic profiles of the lower extremity during walking. J Biomech. 2017;58:11–20. [DOI] [PubMed] [Google Scholar]
- 5.Deschamps K, Eerdekens M, Desmet D, Matricali GA, Wuite S, Staes F. Estimation of foot joint kinetics in three and four segment foot models using an existing proportionality scheme: Application in paediatric barefoot walking. J Biomech. 2017;61:168–175. [DOI] [PubMed] [Google Scholar]
- 6.Ebrahimi A, Goldberg SR, Stanhope SJ. Changes in relative work of the lower extremity joints and distal foot with walking speed. J Biomech. 2017;58:212–216. [DOI] [PubMed] [Google Scholar]
- 7.Eerdekens M, Deschamps K, Staes F. The impact of walking speed on the kinetic behaviour of different foot joints. Gait Posture. 2019;68:375–381. [DOI] [PubMed] [Google Scholar]
- 8.Eerdekens M, Peerlinck K, Staes F, Hermans C, Lobet S, Deschamps K. The biomechanical behaviour of ankle and foot joints during walking with shoes in patients with haemophilia. Haemophilia. [Published online ahead of print May 4, 2020]. DOI: 10.1111/hae.14017. [DOI] [PubMed]
- 9.Eerdekens M, Staes F, Giovanni MA, Deschamps K. Clinical applicability of an existing proportionality scheme in three-segment kinetic foot models. Ann Biomed Eng. 2019;48:547–557. [DOI] [PubMed] [Google Scholar]
- 10.Eerdekens M, Staes F, Matricali GA, Wuite S, Peerlinck K, Deschamps K. Quantifying clinical misinterpretations associated to one-segment kinetic foot modelling in both a healthy and patient population. Clin Biomech. 2019;67:160–165. [DOI] [PubMed] [Google Scholar]
- 11.Egloff C, Hügle T, Valderrabano V. Biomechanics and pathomechanisms of osteoarthritis. Swiss Med Wkly. 2012;142:1–14. [DOI] [PubMed] [Google Scholar]
- 12.Flavin R, Coleman SC, Tenenbaum S, Brodsky JW. Comparison of gait after total ankle arthroplasty and ankle arthrodesis. Foot Ankle Int. 2013;34:1340–1348. [DOI] [PubMed] [Google Scholar]
- 13.Hanavan EP. A mathematical model of the human body. AMRL-TR-64-102. Aerospace. 1964:1–149. [PubMed] [Google Scholar]
- 14.Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16:494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kellgren JH, Lawrence JS, Saltzman CL, Salamon ML, Blanchard GM, Huff T, Hayes A, Buckwalter J a, Amendola A, Corazza F, Parenti-Castelli V, Stagni R, Cappello A, O’Connor JJ, Leardini A, Sharp JT, Wolfe F, Lassere M, Boers M, Heijde D Van Der, Paulus H, Rau R, Strand V, de Vries G, Roy K, Chester V, Queen R, Goldberg AJ, Zaidi R, Thomson C, Doré CJ, Skene SS, Cro S, Round J, Molloy A, Davies M, Karski M, Kim L, Cooke P, Tenenbaum S, Coleman SC, Brodsky JW, Novak AC, Mayich DJ, Perry SD, Daniels TR, Brodsky JW, Khazzam M, Long JT, Marks RM, Harris GF, Philippe P, Paul C, Mark M, Jean-Noel C, Delphine P, Thierry J, Flavin R, Coleman SC, Tenenbaum S, Brodsky JW, Sun SF, Hsu CW, Sun HP, Chou YJ, Li HJ, Wang JL, Lasanianos NG, Lyras DN, Mouzopoulos G, Tsutseos N, Garnavos C, Karatosun V, Unver B, Ozden A, Ozay Z, Gunal I, Valderrabano V, Horisberger M, Russell I, Dougall H, Hintermann B. Directing clinical care using lower extremity biomechanics in patients with ankle osteoarthritis and ankle arthroplasty. Foot Ankle Int. 2014;6:1–7. [Google Scholar]
- 16.Leardini A, Benedetti MG, Berti L, Bettinelli D, Nativo R, Giannini S. Rear-foot, mid-foot and fore-foot motion during the stance phase of gait. Gait Posture. 2007;25:453–462. [DOI] [PubMed] [Google Scholar]
- 17.Lundgren P, Nester C, Liu A, Arndt A, Jones R, Stacoff A, Wolf P, Lundberg A. Invasive in vivo measurement of rear-, mid- and forefoot motion during walking. Gait Posture. 2008;28:93–100. [DOI] [PubMed] [Google Scholar]
- 18.MacWilliams BA, Cowley M, Nicholson DE. Foot kinematics and kinetics during adolescent gait. Gait Posture. 2003;17:214–224. [DOI] [PubMed] [Google Scholar]
- 19.Nosewicz TL, Knupp M, Bolliger L, Hintermann B. The reliability and validity of radiographic measurements for determining the three-dimensional position of the talus in varus and valgus osteoarthritic ankles. Skeletal Radiol. 2012;41:1567-1573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Nüesch C, Valderrabano V, Huber C, Von Tscharner V, Pagenstert G. Gait patterns of asymmetric ankle osteoarthritis patients. Clin Biomech. 2012;27:613–618. [DOI] [PubMed] [Google Scholar]
- 21.Queen RM, Carter JE, Adams SB, Easley ME, DeOrio JK, Nunley JA. Coronal plane ankle alignment, gait, and end-stage ankle osteoarthritis. Osteoarthritis Cartilage. 2011;19:1338–1342. [DOI] [PubMed] [Google Scholar]
- 22.Saltzman CL, Zimmerman MB, O’Rourke M, Brown TD, Buckwalter JA, Johnston R. Impact of comorbidities on the measurement of health in patients with ankle osteoarthritis. J Bone Joint Surg Am. 2006;88:2366-2372 [DOI] [PubMed] [Google Scholar]
- 23.Saraswat P, MacWilliams BA, Davis RB, D’Astous JL. Kinematics and kinetics of normal and planovalgus feet during walking. Gait Posture. 2014;39:339–345. [DOI] [PubMed] [Google Scholar]
- 24.Schaefer KL, Sangeorzan BJ, Fassbind MJ, Ledoux WR. The comparative morphology of idiopathic ankle osteoarthritis. J Bone Joint Surg Am. 2012;94:2239-2246 [DOI] [PubMed] [Google Scholar]
- 25.Simic M, Hinman RS, Wrigley T V, Bennell KL, Hunt MA. Gait modification strategies for altering medial knee joint load: A systematic review. Arthritis Care Res. 2011;63:405–426. [DOI] [PubMed] [Google Scholar]
- 26.Sung KH, Kwon SS, Moon SJ, Lee SY. Radiographic evaluation of the normal ankle joint in children and adolescent. J Orthop Sci. 2018;23:658-664 [DOI] [PubMed] [Google Scholar]
- 27.Tenenbaum S, Coleman SC, Brodsky JW. Improvement in gait following combined ankle and subtalar arthrodesis. J Bone Joint Surg Am. 2014;96:1863–1869. [DOI] [PubMed] [Google Scholar]
- 28.Tetsworth K, Paley D. Malalignment and degenerative arthropathy. Orthop Clin North Am. 1994;25:367-377 [PubMed] [Google Scholar]
- 29.Valderrabano V, Nigg BM, von Tscharner V, Stefanyshyn DJ, Goepfert B, Hintermann B. Gait analysis in ankle osteoarthritis and total ankle replacement. Clin Biomech. 2007;22:894–904. [DOI] [PubMed] [Google Scholar]
- 30.Wang R, Gutierrez-farewik EM. Compensatory strategies during walking in response to excessive muscle co-contraction at the ankle joint. Gait Posture. 2014;39:926–932. [DOI] [PubMed] [Google Scholar]
- 31.Wu G, Siegler S, Allard P, Kirtley C, Leardini A, Rosenbaum D, Whittle M, D'Lima DD, Cristofolini L, Witte H, Schmid O, Stokes I, Standardization and Terminology Committee of the International Society of Biomechanics. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion--part I: ankle, hip, and spine. International Society of Biomechanics. J Biomech. 2002;35:543-548. [DOI] [PubMed] [Google Scholar]