Most foot orthoses are designed to reduce eversion and abduction of the calcaneus relative to the tibia, which are key features of foot pronation. Clinicians associate these movements with various lower limb pathology (9,13,16). Meta-analyses have estimated that foot orthoses reduce peak eversion of the calcaneus relative to the tibia by 2.12° (95% confidence interval (CI), 0.72°–3.53° ) and 2.24° (95% CI, 1.42°–3.07° (4)) and peak internal rotation of the tibia relative to the heel by 1.33° (95% CI, 0.12°–2.53° ). Rearfoot eversion velocity has also been used to investigate orthotic effect because motion velocity might relate to muscle lengthening velocity, which is known to affect muscle behavior (6) and implicitly tendon function. Reports of orthotic effect on eversion velocity tend toward no effect (8). However, angular velocity data require high-quality motion data because differentiation amplifies noise in the motion signal. Skin movement artifact may reduce the quality of angular velocity data and can be avoided by using invasive measurement of foot kinematics.
Clinical paradigms assume orthotic induced changes in rearfoot kinematics occur at the subtalar rather than the talocrural (ankle) joint (based on concepts presented by Root et al. (11), e.g., Ref. ). This follows on from traditional models that assume the function of the ankle is simply to plantar/dorsiflex the foot. The subtalar joint has long been ascribed the role of “torque converter,” converting transverse plane leg rotation into frontal plane heel motion (5). The individual contributions of the ankle and subtalar joints to the motion of the heel relative to the leg have only been reported for a small number of individuals (1,7,15). Although results of these studies cannot be generalized because of low participant numbers, they reveal that in some feet, the frontal and transverse plane motion at the ankle is greater than at the subtalar joint. This might be due to person-specific calcaneus, talus, and tibia/fibula articular geometry or relate to stiffness in ligament and joint capsule constraints. It follows that attributing the eversion effect of foot orthoses to changes at the subtalar joint alone is possibly flawed, because the orthotic effect might occur at the ankle or a combination of both joints. This is an important issue because understanding how each joint is affected by foot orthoses could provide insight concerning why there are differences in kinematic and clinical response to orthoses between individuals. For example, reducing eversion of the calcaneus relative to the tibia through changes at the subtalar joint alone might produce different stress distributions in lateral and medial ankle ligaments compared with if the ankle was the principal source of changes in eversion. Likewise, intrajoint contact pressures and strain in the interosseous ligament and sinus tarsi would differ depending upon whether orthotic effects occurred at the subtalar or ankle joint. Depending upon how the ankle and subtalar joints individually contribute to the overall motion of the calcaneus relative to the tibia, tendon moment arms could differ because movements would occur around different centers of rotation.
The aim of this study was to 1) describe the effect of an antipronation foot orthosis on the pattern and velocity of frontal and transverse plane movement between the calcaneus and tibia and 2) explore the hypothesis that any effect occurs only at the subtalar joint.
The study was approved by the Regional Ethics Committee, and five healthy male volunteers (mean age, 39.6 yr (range, 32–57 yr); mean weight, 84 kg (range, 70–112 kg); mean height, 180 cm (range, 175–183 cm)) gave written informed consent to participate.
Orthoses and footwear
The antipronation foot orthosis comprised a full-length prefabricated insole (salfordinsole-FIRM (Shore A70)) and a prefabricated 4° medial heel wedge (Shore A70) (Fig. 1). A test shoe (Champion, Roddick Lace) was used to enable near normal use of the orthosis while walking with intracortical pins in several foot bones. The upper was removed except for approximately 1 cm at the intersection of the upper and sole. Elasticated straps threaded across the foot were pulled tight and offered firm attachment of the shoe.
Before the experimental procedure, subjects performed walking trials to determine their starting position, self-selected speed, and cadence. Subsequently, each subject was taken to surgical theater for insertion of intracortical pins to the tibia, talus, and calcaneus. Self-drilling, 16-mm-diameter pins (Synthes, Bettlach, Switzerland) were inserted under dorsal local anesthetic infiltration (preserving plantar sensation) (Xylocaine and Marcain; AstraZeneca, Södertälje, Sweden) and sterile conditions. Insertion locations and pin orientation were chosen to avoid nerves/blood vessels and minimize skin impingent. Marker triads were secured to each pin.
Subjects completed multiple practice walks to ensure they acclimatized to walking with the pins and shoes. Subjects then performed five walking trials in each of the shoes only and shoes plus orthotic condition. A 12-camera motion capture system (240 Hz; Qualisys, Göteborg, Sweden) was used to track movement of the reflective markers. To establish local coordinate frames, additional markers were placed on the floor at the tip of the second toe and at the center of the heel, on both malleoli and both femoral condyles during standing trials.
After pin removal, insertion sites were cleaned and covered with sterile dressings. Subjects were provided with antibiotic (Heracillin, AstraZeneca) and pain medication (Citodon, AstraZeneca) if required. No clinical complications were reported.
All calculations were conducted using Visual 3D (C-Motion, Inc., Rockville, MD). Local coordinate frames for the tibia, talus, and calcaneus were defined relative to the three markers attached to each bone pin. The vertical (z) axis of the local tibial frame was defined using the knee and ankle centers (defined using anatomical markers on femoral condyles and malleoli in the static trial). The anterior/posterior tibial axis (y) was perpendicular to a plane defined by the femoral condyle and malleoli markers. The medial/lateral tibial axis (x) was perpendicular to the other two tibial axes. For the talus and calcaneus, local frame orientation was set such that in relaxed standing, the y (anterior/posterior), x (medial/lateral), and z (inferior/superior) axes were parallel to those of a foot reference frame. The (y) axis of the foot reference frame was defined by a marker on the floor behind the center of the heel and a marker on the floor at the apex of the second toe during standing trials. The (x) axis was in the global transverse plane (XY) but perpendicular to the foot reference y axis (z was perpendicular to x and y). Joint rotations were calculated using Cardan angles (sequences x, y, and z) and angular velocity derived through differentiation of motion data from each trial. All data were normalized to 0%–100% of stance phase; 0° was the joint position in relaxed standing.
Data analysis replied upon descriptive statistics because of the low sample size. A change in the initial peak in eversion of the calcaneus relative to the tibia was the primary measure of orthotic effect. For each subject, measures of separate ankle and subtalar inversion/eversion were derived to understand the contribution of each joint to the motion of the calcaneus relative to the tibia. The decision as to whether any orthotic effect was systematic across the sample was based upon the 95% CI of the change in inversion or eversion at each joint. If the 95% CI for the sample crossed 0°, then the change was judged not to be systematic.
Descriptive analysis of other parameters was undertaken where there were apparent differences because of the orthosis in both frontal and transverse planes (e.g., peak ankle inversion, range of eversion, duration of eversion, and position of the joints at 50% stance). This was based on reports of person-specific response to orthoses (14), the fact that effects at the ankle had not previously been reported and were therefore unknown, and because the link between rearfoot eversion (our primary outcome) and pathology is not strongly proven.
Frontal plane effects
The mean range of and/or peak in initial calcaneus–tibia eversion was reduced by the foot orthosis in all participants (Fig. 2, Table 1). The 95% CI of the change in the initial peak in eversion did not cross 0°, indicating a systematic effect across subjects. Effects were subject specific in nature and scale and occurred at different periods of stance. Reductions in the initial peak of eversion ranged from 0.7° (subject 1) to 3.6° (subject 3). Changes at the separate ankle and subtalar joints were subject specific, and the 95% CI for these changes crossed 0°, indicating no systematic effect across subjects (Table 1). However, for the increased ankle inversion, the lower threshold for the 95% CI was only marginally below 0° (−0.1°), and only one of the five subjects displayed reduced ankle inversion (subject 2, reduced by 0.4°). Nonsystematic effects were apparent in other respects. Subject 4 demonstrated reduced peak subtalar eversion coupled with increased peak in ankle inversion in early stance, whereas subject 1 experienced increased peak subtalar eversion coupled with increased peak ankle inversion. Subject 3 demonstrated changes primarily at the subtalar joint, whereas subject 4 showed change at both ankle and subtalar joints. For subjects 2 and 5, changes at each joint were <0.5°.
The foot orthosis also reduced calcaneus eversion relative to the tibia at 50% of stance for subjects 1, 2, and 3, but not subject 4 or 5 (Fig. 2). Subject 4 showed reduced eversion from 60% to 100% stance. Reductions in eversion at 50% of stance were achieved via changes almost entirely at the ankle joint in subject 1 (+4.0° inversion position). By contrast, subjects 2 and 3 showed greater changes at the subtalar than the ankle joint (1.4° less subtalar eversion, subject 2; 3.1° less subtalar eversion, subject 3). Subject 4 demonstrated 2° less everted position at both ankle (at 70% stance) and subtalar joints (80% stance).
The peak and/or period of calcaneus–tibia eversion velocity during the first 10% of stance was reduced by the foot orthosis in subjects 1–4 (Fig. 3). Subject 4 experienced increased ankle inversion velocity (22°·s−1 increased to 37.4°·s−1), and subject 1 experienced ankle inversion rather than eversion velocity (−23°·s−1 eversion velocity at 13% stance changed to 5.2°·s−1 inversion velocity). Subject 2 showed evidence of reduced peak ankle (−24.7°·s−1 reduced to −13.8°·s−1 at 25% stance) and subtalar eversion velocity (−108.6°·s−1 reduced to −66.3°·s−1 after initial contact). For subject 3, the orthosis prevented the rapid transition from subtalar eversion to inversion velocity in 0%–10% stance. At the ankle, the inversion velocity was increased (from 30.7°·s−1 to 39.8°·s−1), but subsequent eversion velocity also increased (from −21.5°·s−1 to −34.1°·s−1). Subject 5 displayed no changes in frontal plane velocity.
Transverse plane effects
With the exception of subject 5, the orthosis reduced the range of and/or peak in abduction of the calcaneus relative to the tibia. The timing of changes was subject specific (Fig. 4). Subject 5 displayed a small increase (maximum increase, 0.7°) in calcaneus–tibia abduction throughout stance. The effect for other subjects was not present throughout all stance. For subject 4, the ankle was unaffected at initial contact, but the subtalar joint was less adducted (3.2° compared to 5.8°). By contrast, for subject 1, peak ankle abduction was reduced by 2.7°, and the subtalar joint showed a trend toward more abduction (+2.6° abduction at 25% stance). Opposite ankle and subtalar effects were also evident for subjects 2 and 3.
Changes in transverse plane angular velocity were not systematic (Fig. 5). For subjects 1, 2, 4, and 5, there were no significant changes. Subject 3 demonstrated reduced calcaneus–tibia abduction velocity in early stance (from −70.7°·s−1 to −28.6°·s−1).
The antipronation orthosis reduced the peak and range of rearfoot eversion and, to a lesser degree, abduction relative to the leg in all subjects. However, contrary to clinical paradigms, these changes were inconsistent between subjects and occurred to varying degrees at different times of stance. Eversion and abduction velocity were affected less consistently than angular motion. Overall, the kinematic changes were generally small and concurred with the only other report of orthotic effect using bone anchored markers (14). The reduced eversion of the calcaneus relative to the leg (mean, −2.0°; Table 1) was within the range defined through meta-analysis of the literature (from Ref. (5): 95% CI, 1.42°–3.07°; from Ref. (9): 95% CI, 0.72°–3.53°). The smaller transverse plane effects concur with Mills et al. (8).
In contrast with clinical paradigms, the orthosis induced kinematic changes at both ankle and subtalar joints. However, these effects were not systematic across subjects. At heel strike, load is first applied to the subtalar joint via the calcaneus, and its response to the external loads, the distribution of which are changed by foot orthoses (10), will influence the load transferred to the ankle via the talus. Subtalar kinematics that occur in response to external loads will be influenced by constraints imposed by osseous contact forces, ligament and capsular restraints, and tendon forces, and these will be person specific. Many of the structures at the subtalar joint also cross the ankle, and thus, the properties and kinematic response of the ankle will affect the subtalar kinematic response. It seems plausible that feet with distinctly different ankle and subtalar structures, such as “pes planus” and “pes cavus” foot types, might be associated with specific patterns of response to a foot orthotic. Furthermore, because the initial effect of an orthotic is to influence load distribution under the sole of the foot, the orthotic effect could be sensitive to the underlying pattern of plantar load distribution. Plantar pressure patterns are known to differ between different foot types) (2).
The strong interaction between the mechanical behavior of the subtalar and ankle joints and the multiple (often common) factors/structures affecting each joint may explain the subject-specific responses observed. For example, in one subject, a relatively stiff subtalar joint may exhibit minor modifications to its kinematics because of the foot orthosis. In this case, the external forces altered by the orthosis may be largely transferred to the ankle. Depending upon its own mechanical constraints, the ankle may demonstrate a greater kinematic change than the subtalar joint. In another foot, however, a more compliant subtalar joint might allow for a greater change in its kinematics. However, depending upon the nature of the change in forces under the foot created by the orthosis (which itself will be variable between people, dependent upon dynamic foot shape and tissue properties, orthosis shape, and material properties), this might still result in transfer of forces to the ankle comparable with those in the case of the stiffer subtalar joint. Thus, a large kinematic response at the ankle might also be observed, even in the case of a large subtalar response. Furthermore, even if a subtalar joint demonstrates little kinematic response, there may also be little kinematic response at the ankle if the constraints are sufficient to resist the forces altered by the foot orthosis (see subject 2, ankle response was negligible and subtalar response the least of the five subjects). In this case, the forces would be more strongly referred to the tibia and knee, and we know this can occur (3). It should also be remembered that joint stiffness is dynamic and varies as joint position and motion velocity change.
There are many interrelated variables affecting subtalar and ankle kinematics, and therefore, a systematic kinematic response to a foot orthosis seems unlikely. The role of the orthosis cannot be to adjust foot position or motion to achieve a single optimum kinematic pathway or foot position as current orthotic paradigms suggest (based on concepts described in Root et al. ). Rather, the orthosis influences the external loads that determine the person-specific foot kinematics, and as a result, the orthosis causes the joints of the foot to pass through adjusted versions of a person’s underlying kinematic pathways.
It is not clear whether the small alterations in joint kinematics we report can lead to changes in clinical symptoms. The changes observed are presumably representative of those that would occur over many repetitive gait cycles each day, and there may be potential for accumulative benefit from even small kinematic changes. Also, we have only reported on effects across two rearfoot joints, whereas most foot structures span multiple joints. Furthermore, we do not know what feature (or features) of foot kinematics needs to change to elicit a clinical response, and in this study, changes were highly person specific. Peak heel eversion and range of heel eversion are the most commonly reported parameters (8), but their clinical relevance is unproven. Identifying which parameter a foot orthosis must change remains a significant challenge because plantar loading, joint moments, and joint motion cannot be modified separately by the orthosis, and thus, changes in each will always be coupled.
Understanding how the observed changes in separate ankle and subtalar kinematics might lead to clinical benefit is also complex. Any orthotic-induced change in the stress experienced by injured tissues might vary depending upon how each of the ankle and subtalar joints respond to the orthosis. For example, the moment arms of tendons that span both the ankle and subtalar joints will differ depending upon where movement is occurring and where centers of rotation lie. A change in ankle kinematics might affect stress in a specific tendon or ligament, or joint contact forces, differently than if the same kinematic change occurred at the subtalar joint.
The data are from a small sample of asymptomatic individuals, and thus, the orthosis has been tested in a limited range of types of feet. Indeed, the subject who showed the least response (subject 5) displayed minimal rearfoot motion and perhaps reflects a foot that might not normally receive an antipronation foot orthosis. Also, it is inappropriate to draw conclusions about how foot orthoses result in clinical benefit from such small asymptomatic samples. However, the validity of the data (direct measurement of bone motion) and measurement of separate ankle and subtalar kinematics offers new insight into how orthoses can affect feet. This research is therefore suited to identifying new avenues for understanding orthotic effect. The effects of a foot orthoses are linked to the footwear in which they are used. The forces applied to the base and sides of the orthoses are via the sole and shoe upper, respectively. In this study, the upper was largely removed, and although not evaluated formally, the shoe had reduced stiffness. The effect of an orthosis will be strongly related to its geometry and material properties, and results here will not apply to orthoses with significantly different properties.
The antipronation foot orthosis produced small and unsystematic reductions in eversion and abduction of the heel relative to the leg at various times during stance. This was achieved via complex changes at the ankle and subtalar joints that were specific to each subject tested. These changes contradict existing orthotic paradigms and are indicative of a strong interaction between the ankle and subtalar joints.
No funding was received for the work contained in this article.
The Intellectual Property (IP) for the insoles tested in this work is owned by the University of Salford and licensed to a University spin off company. Nester is a Director of that company. No payment was received by Nester for the work in the article submitted.
Liu, Jones, Lundgren, Kundberg, Arndt, and Wolf declare no conflict of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Arndt A, Wolf P, Liu A, et al.. Intrinsic foot kinematics measured in vivo during the stance phase of slow running. J Biomech
. 2007; 40 (12): 2672–8.
2. Burns J, Crosbie J, Hunt A, Ouvrier R. The effect of pes cavus on foot pain and plantar pressure. Clin Biomech (Bristol, Avon)
. 2005; 20 (9): 877–82.
3. Chen YC, Lou SZ, Huang CY, Su FC. Effects of foot orthoses on gait patterns of flat feet patients. Clin Biomech (Bristol, Avon)
. 2010; 25 (3): 265–70.
4. Cheung RTH, Chung RCK, Ng GFY. Efficacies of different external controls for excessive foot pronation: a meta-analysis. Br J Sports Med
. 2011; 45: 743–51.
5. Close JR, Inman VT, Poor PM, Todd F. The function of the subtalar joint. Clin Orthop
. 1967; 50: 159–79.
6. Hill AV. The heat of shortening and the dynamic constants of muscle. Proc R Soc Lond
. 1938; B126: 136–95.
7. Lundgren P, Nester C, Liu A, et al.. Invasive in vivo measurement of rear-, mid- and forefoot motion during walking. Gait Posture
. 2008; 28 (1): 93–100.
8. Mills K, Blanch P, Chapman AR, McPoil TG, Vicenzino B. Foot orthoses and gait: a systematic review and meta-analysis of literature pertaining to potential mechanisms. Br J Sports Med
. 2010; 44 (14): 1035–46.
9. Moen MH, Tol JL, Weir A, Steunebrink M, De Winter TC. Medial tibial stress syndrome: a critical review. Sports Med
. 2009; 39 (7): 523–46.
10. Redmond AC, Landorf KB, Keenan AM. Contoured, prefabricated foot orthoses demonstrate comparable mechanical properties to contoured, customised foot orthoses: a plantar pressure study. J Foot Ankle Res
. 2009; 2: 20.
11. Root ML, Orien WP, Weed JH, Hughes RJ. Normal and Abnormal Function of the Foot
. Los Angeles: Clinical Biomechanics Corp; 1977. p. 478.
12. Ryan M, Grau S, Krauss I, Maiwald C, Taunton J, Horstmann T. Kinematic analysis of runners with achilles mid-portion tendinopathy. Foot Ankle Int
. 2009; 30 (12): 1190–5.
13. Shanthikumar S, Low Z, Falvey E, McCrory P, Franklyn-Miller A. The effect of gait velocity on calcaneal balance at heel strike; Implications for orthotic prescription in injury prevention. Gait Posture
. 2010; 31 (1): 9–12.
14. Stacoff A, Reinschmidt C, Nigg BM, et al.. Effects of foot orthoses on skeletal motion during running. Clin Biomech (Bristol, Avon)
. 2000; 15 (1): 54–64.
15. Westblad P, Hashimoto T, Winson I, Lundberg A, Arndt A. Differences in ankle–joint complex motion during the stance phase of walking as measured by superficial or bone anchored markers. Foot Ankle Int
. 2002; 23 (9): 856–63.
16. Wyndow N, Cowan SM, Wrigley TV, Crossley KM. Neuromotor control of the lower limb in Achilles tendinopathy: implications for foot orthotic therapy. Sports Med
. 2010;1; 40 (9): 715–27.