Perspectives on How and Why Feet are Prescribed

Gerald Stark, BSME, CP, FAAOP

Modern prosthetic practice includes a growing array of foot designs that offer a wide spectrum of functions, indications, and costs. New designs and materials have added properties and motions that blur the lines of the traditional prosthetic foot classification. This constant innovation requires a greater knowledge of physiologic foot function and more descriptive terminology when recommending feet to patients.

PHYSIOLOGIC VERSUS PROSTHETIC FOOT FUNCTION

Prosthetic feet and ankles are often studied and compared on the basis of how they substitute for physiologic foot functions. Perry¹ lists three main functions of the physiologic foot as shock absorption, weightbearing stability, and progression. Valmassy² further discusses five functions as load bearing, leverage, shock absorption, balance, and protection.

SHOCK ABSORPTION

As the prosthetic foot makes initial contact, adapts to ground contours, and accepts weight (in loading response to early midstance), closed-chain pronation and tarsal mobility are responsible for shock absorption. The subtalar joint everts and the talus inverts as a result of the body of the talus being medial to the distal support of the calcaneus. The resulting internal rotation of the talus is followed by an internal rotation of the tibia resulting in a general collapse of the medial longitudinal arch. This is supported with an eccentric contraction of the inverters of the foot, primarily tibialis anterior, providing a degree of muscular shock absorption. Internal rotation of the ankle allows the rigid mortise of the ankle to be properly aligned to accept loading of the foot from initial contact to loading response. After forefoot contact and maximum functional subtalar eversion, midtarsal dorsiflexion contributes to shock absorption during the end of the second rocker loading response to midstance.

Prosthetic components often emulate shock absorption functions of the physiologic foot, but usually have far fewer mechanisms to do so. This is because they lack sufficient triplanar rotary motion or variable stability. The transverse rotary is approximated with the 5° of external rotation and motion allowed by the conventional prosthetic hindfoot.³ The heel lever or posterior cushion plays a major part in the first rocker arc. Perry⁴ suggests that this first rocker has been neglected when compared with the variety of forefoot keel designs.

When comparing the ankle motion curves, the plantar flexion of the single axis foot may mimic the anatomic motion. However, this is at an increased rate that does not readily dorsiflex toward neutral, similar to paralytic equinus in an orthotic user.⁴ Plantar flexion for the solid ankle cushioned heel (SACH) foot and Seattle Lite foot are roughly half that of the anatomical foot, with the shortened arc of the Flex-Foot being only one fourth of normal.⁴

WEIGHTBEARING STABILITY

Weightbearing stability refers to the advancement of the load vector from midstance to terminal stance. This loads the metatarsal heads and places an increasingly greater demand on intertarsal stability to provide a relatively stiff lever arm for late stance.⁴ This demand becomes greatest at heel rise as the foot rises up on the metatarsal heads. The increasing stiffness physiologically is a result of closed-chain supination with subtalar inversion providing the necessary support of the calcaneus under the talus with subsequent external rotation of the tibia. The talonavicular and calcaneocuboid joint axes converge to provide a locking action with the peroneus longus applying lateral pressure to the cuboid. This helps to tighten the wedge shaped cuneiform bones that function as keystones in the transverse arch and provide a rigid foot.¹ Again this helps align the ankle mortise to accept the increase load. It is interesting to note that the medial column provides flexibility with the lateral column providing rigidity.¹

Prosthetic feet do not yet offer the degree of variable stiffness provided by the physiologic foot. Designs often compromise between the softness at the second rocker and the stiffness required at terminal stance. Usually designers seem to err toward stiffness, because drop off at terminal stance is undesirable, and even potentially hazardous for the transfemoral user if it causes knee instability. With the exception of the single-axis foot, prosthetic feet usually have no mortise joint allowing easy tibial progression. Combined with a set heel height, this results in tibial progression being slowed to one half the usual rate.⁴ One disadvantage in foot designs with a cushion type heel is the prolonged "heel-only" contact. This produces an unstable knee moment until the forefoot makes contact.⁴ This could be one reason that a soft heel bumper was recommended for a transfemoral amputee to prevent an excessive (external) knee flexion moment.⁵ Prolonged heel loading can also result in more falls when patients walk on low-friction surfaces such as wet tile or ice.

It is often assumed that prolonged heel-only contact causes an external dorsiflexion moment as well as an excessive external knee flexion moment. If this were so, one might expect to observe higher electromyographic (EMG) readings of the quadriceps muscle for users of SACH, Seattle, or similar feet. However, observation of the EMG studies of the vastus lateralis does not show this. Review of EMG recordings from the biceps femoris actually reveals much higher EMG studies compared with the norm.⁴ This could indicate that the prolonged heel-only contact actually causes an external knee extension moment, and that the hamstring muscles are functioning to keep the knee in flexion.

PROGRESSION

During progression, the foot moves into the third rocker from terminal stance to pre-swing. This induces controlled dorsiflexion of the metacarpal phalangeal joint, which is flexed to provide a broad and stable area of support for the toes. This in turn reduces the overall pressure on the metatarsals, as the load transfers distally and controls the shape and stability of the forefoot rocker while permitting a longer forefoot roll.⁴ A pathologic foot has a shortened rocker, which results in greater load on the metatarsals and the contralateral limb. In addition to forefoot stiffening, the first ray plantarflexes with the peroneus longus, thereby helping to move the center of pressure between the first and second toes.¹

Dynamic-response feet can simulate progression by flexing with the load and carrying the arc of motion to the toes. The SACH foot's arc of motion, however, terminates before reaching the toes. Progressive stiffness is directly influenced by the composition and geometry of the forefoot keel. This may consist of multi-carbon plates, a urethane "sandwich," or a carbon footplate. The geometry of the keel also influences stiffness. The cross-sectional taper and angle or curve of the keel as well as the surrounding material provide spring stiffness. A problem often encountered with many keels integrated into foam has been that the front end of the keel pushes through the foam foot. Although cloth reinforcement has reduced this tendency, it continues to be a problem with highly active patients. Observation of ankle moments reveals that the Flex-Foot keel acts to generate twice the degree of energy return compared to a SACH foot.⁴ Along with being symmetric and thereby reducing manufacturing costs, a wide blade width accommodates a wide variety of center of pressure (COP) pathways, but may decrease efficiency overall.

TRADITIONAL FOOT CLASSIFICATIONS

Foot classifications that have been used since the 1980s included the single axis, SACH, multi-axis, and dynamic response.⁶ Perry⁴ refers only to the anatomical or single axis, biomechanic or SACH, and dynamic or Seattle Foot and Flex-Foot. The single-axis foot mimics anatomic ankle hinge movement but cannot rotate into position for load acceptance or transfer. It does offer the advantage of emulating normal plantar flexion motion with a minimized arc of motion that still exceeds normal foot drop by 50%.⁴ A soft bumper offers less resistance to plantar flexion. This may result in premature foot flat, with the reaction line being thrown anterior and the knee thrust rapidly into extension. This induced knee extension is one reason more active amputees find this foot "slow" or overly stable. A harder bumper used to counteract this would defeat the advantage of the single-axis foot by making the arc of motion and recovery to a neutral position exaggerated. It must be remembered historically that singleaxis feet were the first feet that were laboriously custom made for the patient with a toe break placed 6 mm posterior to the metatarsal heads. The foot rocker was positioned so as to augment the patient's movement. Today's single-axis designs are prefabricated and generally must depend on the alignment capability of the prosthetist to optimize the rollover characteristics.

The SACH foot was developed in the late 1950s to address some of the obvious maintenance and availability issues of the single axis foot by incorporating the functions of the foot into an integrated design. The heel cushion compressed to simulate the eccentric lengthening of the dorsiflexors. The rigid keel simulated the stiffening affect of the plantarflexors.⁵ It has been said that the SACH foot represented the first rollover shape mentioned in today's literature by Hansen et al.⁷ Although critical of the SACH foot in many aspects, Perry⁴ also notes that the early forefoot rocker resembles the Flex-Foot. Functionally the SACH foot also helped to promote knee flexion that was important to patellar tendonbearing (PTB) interface design of the time (joint and corset interfaces were placed in full extension and relied on the stops of the joints to prevent hyperextension). The PTB design initially advocated 15° of knee flexion to prevent possible hyperextension risk with the absence of the joint stops. Because this precaution was thought to be excessive, 7° to 10° of knee flexion is now advocated.⁵ Perry⁴ has posed questions regarding premature heel rise and disruption of tibial progression. Observation shows prolonged heel cushion compression with heel-only contact (twice normal 27% gc as opposed to 12% gc normal). With a rapid midstance heel rise twice that of normal, this could be one reason why Radcliffe⁵ advocated relative plantar flexion of the foot to minimize heel cushion compression and increase foot flat stability. This was particularly emphasized for stabilizing the knee of transfemoral amputees. The key concern regarding the SACH foot became tibial progression, which was measured at 33% of the norm, as opposed to the Flex-Foot, which was 67% of the norm.⁴ Demand on the quadriceps was anticipated to have been increased, although this was not verified by energy expenditure studies by Nielsen et al.⁸ (It should be noted that these results were directly dependent on the alignment of the test subjects.)

The multi-axial foot remains popular especially for activities on uneven terrain. It can be classified according to forefoot and hindfoot designs with integrated or multi-part configurations. The mechanisms can be a simple split-toe variety, a carbon plate urethane overmolded sandwich, or a multipart design. The forefoot multiaxial designs are primarily the simple split-toe variety that helps provide third rocker stability. With increased stiffness this can also simulate plantar flexion of the first ray, and push the path of the COP slightly laterally. Hindfoot plantar flexion provides a degree of inversion-eversion response to aid the first and second rocker functions as the foot adapts to the ground during loading. The disadvantage of multiaxial feet is that the large amount of load at the ankle may apply significant transverse shear to the hindfoot mechanism. The most durable designs seem to employ mechanisms that add some rigidity and help direct and guide the COP forward.

Dynamic Response feet emerged in the 1980s with the objective of providing better loading response all the way to the toes. At first, Delrin (a nylon that can be easily machined and offers consistent spring function and toughness) was used to create the anterior keel or spring board. Initially these designs were thought to offer prolonged foot flat stability, better tibial progression, and more support distally when compared with SACH feet. Although researchers could not verify these observations with quantifiable data, many users said that the feet simply felt more lifelike. Other designs used phenolic and Fiberglas materials, but the use of high strength carbon struts greatly extended the degree of flexibility and energy return. The Flex-Foot dorsiflexes to 20°, twice the normal 10° range. This actually remedies the shortened stride length experienced by most transtibial amputees. It has also been shown to reduce vertical loading on the opposite side. This was accomplished by support being carried more distally and an accompanying plantar flexion moment. It is interesting to note that this energy return is not fully realized until the foot is partially unweighted. Therefore, its value during stance phase has been questioned, and some investigators have suggested that the main benefit of energy return in prosthetic feet is to initiate swing phase. The need for push-off has also been questioned by some experts as nonphysiologic, because the gastroc-soleus complex EMG signal dissipates at double support. Prosthetists have also noted the presence of a medial whip at unloading of the foot apparently caused by the energy return of the foot. Some patients feel that although they like the foot's springiness, they find themselves fighting the action of the foot at slower speeds, descending stairs, and even decelerating after running.

FUNCTIONAL SUBSETS

Recently many foot designs have blurred the lines of these classifications by hybridizing the properties of different classes, primarily by combining dynamic response with multi-axis attributes. It may be useful to develop a classification system based on the subsets of individual functional attributes that may be present in any foot. These subsets could include Forefoot Keel, Heel Lever, Hindfoot Roller, Flexing Strut, Forefoot Inversion/Eversion, Multiaxis Hindfoot, and Integrated Shock.

FOREFOOT KEEL
The Forefoot Keel is characteristic of the most basic dynamic response foot with any number of materials and configurations. The Forefoot Keel can be a single-bladed member or consist of multiple separate members to approximate the medial and lateral columns of the anatomical foot. Stiffness is directly dependent on the cross section, material, keel length, and geometry. Some designs use multiple layers that collapse progressively, and others use a urethane sandwich, which has a smoothing effect on the load progression.

HEEL LEVER
The Heel Lever emulates the first rocker, which contributes to load acceptance and plantar flexion arc characteristics. Many ankle-foot systems simply use a cushion heel that simulates plantar flexion by compressing, but this simple approach may delay the stability of foot flat. In other designs, such as the Flex-Foot Mod III, a Heel Lever projects posteriorly from the forefoot keel or midfoot attachment, and often provides stiffer support than a cushion heel. This configuration tends to reduce plantar flexion and may induce a greater knee flexion moment. Recent designs have used multiple levers, linkages, or urethane bumpers or a sandwich to simulate the progressive stiffness of the anatomical foot.

HINDFOOT ROLLER
A Hindfoot Roller mechanism used by many feet uses a rocker element mounted on a footplate to approximate the second rocker from loading response to midstance. This mechanism does not necessarily lower the arc of plantar flexion, but emphasizes the rotary motion of the second rocker, to ease the transition during the second rocker. When configured as a complete circular mechanism that wraps superiorly, the Hindfoot Roller can also function indirectly in shock absorption by emulating midtarsal dorsiflexion. Excessive rocker function in late midstance would be nonphysiologic, leading to a loss of support in late stance.

FLEXING STRUT
A Flexing Strut that extends to the proximal socket attachment originated with the Flex-Foot design. Contemporary designs usually incorporate the forefoot keel in one integrated structure, but the strut can be separate from the rest of the foot. The strut is usually a wide rectangular cross section, but can be produced in a number of U-shaped, circular, or multiple rod geometries. Using continuous fibers in the strut composition insures maximum flexibility and strength. All these Flexing Strut designs offer the greatest amount of energy return providing twice the degree of late stance plantar flexion moment, and five times the power compared with the SACH foot.⁴ The longer the continuous fibers are in the layup of the composite, the greater the amount of flexion that can occur. Unfortunately, this design also increases the height of the foot. The longest Flexing Strut designs have been shown to dorsiflex to as much as 23° in late stance, significantly more than the normal functional dorsiflexion of 10°.⁴ This increase in passive dorsiflexion compensates for the relative step length shortening that results from tibial progression in a transtibial prosthesis that is only 67% of normal (SACH foot at 33% of normal). This motion has also been shown to reduce the vertical shock to the contralateral limb due to greater support in late stance. Vertical loading on the contralateral foot averages 130% of body weight, but the use of a Flexing Strut results in approximately normal loading of the surviving foot of up to 110% of body weight.⁴ Some designs, created for athletes for racing only, use the flexing strut for toe-only gait.

FOREFOOT INVERSION-EVERSION
Forefoot Inversion-Eversion is commonly provided in strut systems by a split-toe design. Other designs are more integrated, molding different durometer materials or members together within the foot so there are not necessarily articulating parts. Some designs create a forefoot composite urethane sandwich. The disadvantage of many of these systems is that they are nonadjustable and depend on the material stiffness of the design. It is important to note that the damping characteristics of the forefoot may limit the desired energy return, or in a more favorable light, smooth late-stance forces.

MULTIAXIS HINDFOOT
A Multiaxis Hindfoot has existed historically as an articulating component with urethane rubber bumpers, bushings, spherical snubbers, or large rings damping the motion. This component can be a separate modular ankle unit that can be used with a variety of foot designs, or it may be integrated into the foot itself. Multiaxis articulating designs often need regular maintenance and servicing. Some variants extend the urethane sandwich from the forefoot to the hindfoot, thereby providing some hindfoot motion in addition to allowing midfoot torsion.

INTEGRATED SHOCK ABSORBERS
Many prosthetic feet also incorporate shock absorbers in a parallel or series configuration. A series configuration is usually found with a damper more proximal to the spring-like foot, whereas a parallel design has a damper and spring at the same level.⁹ It should be noted that a true shock absorber in engineering terms has a damper and spring in parallel where the spring functions to reset the damper once energy has been dissipated. The stiffness of the series configuration is limited by the softest component in the chain or the damper. The stiffness is truly a sum of the damper and spring together where the damper absorbs energy and prevents recoil of the spring and the spring resets the damper and prevents it from "bottoming out."⁹ The telescoping nature of many shock absorbers may be considered nonphysiologic, because the overall length shortens. Some designs use a rotary motion linkage at the ankle or behind the forefoot to minimize this affect, and attempt to imitate the shock attenuation mechanisms of the physiologic ankle/foot.

Future components are sure to continue this blending of qualities to provide greater foot function and movement. In the near future, advances in composite geometry and material design will be able to provide more variable stiffness or flexibility characteristics for the forefoot and/or flexible strut. Longer-range expectations would be to create a foot in which the patient can actively vary the amount of shock absorption, stiffness, heel height, or third rocker support at the toes. With microprocessor and electronic control theories, this may be possible within the foot and ankle itself.

PROSTHETIC CHOICES

How does the prosthetist select these various designs and match the various functions with the patient's needs? Ideally the process would involve systematically evaluating peerreviewed research, using a clinical outcome-generated decision tree, and then scientifically determining the optimal selection. However, the information presently available does not yet permit this methodology. Currently, prosthetists use a process based primarily on empirical observation. This includes functional assessment and patient feedback concerning the transition in rollover during stance. Studies regarding rollover shape from Northwestern University show that prosthetists can greatly affect the rollover shape by altering component alignment, and tend to equalize the rollover among all designs by using slide, tilt, or a combination of these movements. In an informal survey of five prosthetists, who included innovators, early adopters, early majority, late majority, and traditionalists,¹⁰ the author noted that four or five ankle-foot designs were typically preferred. These specific components were selected because of clinical confidence based on performance, durability, and predictability, among other qualities. It is interesting to note that all the prosthetists indicated that although cost was a consideration, providing the patient the component that best matched his or her needs was the overriding concern. Such performance assessment can be based on a variety of functional foot selection factors including cadence speed, uneven terrain, stability and balance, amputation level, weight, size of foot, special functions, effect of alignment, product warranty and maintenance, and cost.

CADENCE SPEEDS
Cadence influences keel stiffness at higher speeds and responsiveness to gait changes. It is generally believed that faster ambulators benefit more from stiffer, more dynamic foot designs. Currently, the Medicare 0–4 K level activity codes are often used for this assessment. Although this scale is widely accepted, it is geared toward the low-functioning user and lacks finer gait increments. This may be a result of the predominantly geriatric Medicare population.

UNEVEN TERRAIN
The ability of the prosthetic foot to accommodate uneven terrain is desirable for patients who may enjoy outdoor activities or who require stability on irregular surfaces. It is important to realize that patient falls or stumbles may be related to lack of compliance at the forefoot just as much as the lack of hind-foot inversion/eversion. Some prosthetists feel that the more proximal the amputation, the more ankle motion should be provided to aid in stability by keeping the foot flatter on the floor. Ideally, the multiaxial mechanism should be adjustable so the desired stiffness can be set. This is not possible with integrated designs, and adjustable bumpers may also add the disadvantage of more maintenance.

STABILITY AND BALANCE
Stability and balance obviously affect foot choice. Often a more flexible, compliant foot is desired. The patient should not be hampered or slowed by the foot choice, but this cannot be verified until dynamic alignment, when the patient actually walks on the recommended component.

AMPUTATION LEVEL
Amputation level also influences foot choice. Transfemoral patients must control the prosthesis with their hip. Softer feet are usually selected to allow greater foot flat position, but alignment also can have a great effect. It should be noted that dynamic response feet may help amputees with higher level amputations to initiate swing phase motion.

WEIGHT
Minimization of weight has always been a consideration in foot design. It is the most distal portion of the prosthesis, and therefore arguably the most weight sensitive. Use of metal linkages, urethane, and shock absorbers all contribute to increased weight. alternative designs contain fewer pieces and emphasize light weight. Some practitioners have speculated that feet could actually be too light, especially when inertia is needed for swing through with transfemoral prosthesis. (Consider what it might be like for the transfemoral amputee with an extremely light foot to try to walk forward into a very stiff wind, or to advance the knee when walking in tall grass.) Another benefit to weight in the foot is to enable the amputee to perceive the position of the prosthesis as it is moved through space.

FOOT SIZE AND KEEL LENGTH
Another characteristic is the relative size or keel of the foot. Larger feet usually have a proportionately longer keel, but anecdotally, the patient often feels the keel is too long. Larger patients may prefer a softer or shorter multiaxis forefoot keel. Conversely, with smaller feet, the keel length may terminate prematurely. In modern endoskeletal feet this is generally a nonadjustable feature.

ALIGNMENT
Often keel length stiffness and rollover can be compensated with A-P linear shift, but this is not always possible, depending on the alignment components incorporated into the prosthesis. Although a linear shift is possible with pyramid systems when both a proximal and distal pyramid are incorporated, there is a tendency to use distal plantar/dorsiflexion exclusively to achieve proper rollover performance.

WARRANTY AND MAINTENANCE
Issues of warranty and maintenance have always been a factor in the selection of a prosthetic foot. Some practitioners try a variety of feet and return the ones that were rejected by the amputee. This approach requires the support and understanding of the foot manufacturer, but gives the patient and practitioner direct experience with each foot. The functional quality of the prosthetic foot is subject to interpretation by the patient who may choose a design that seems counterintuitive to the prosthetist. Designs that function well but require frequent servicing and maintenance are generally avoided, particularly when the patient lives many miles away from the prosthetic laboratory.

COST AND REIMBURSEMENT
There is a wide spectrum of cost for different prosthetic ankle-foot mechanisms based on function and manufacturing complexity. Because the market for prosthetic feet is relatively small, development and production is costly and smaller prosthetic manufacturers may require many years to recoup their development costs. Fiscal constraints such as third-party reimbursement policies may preclude provision of the prosthetic foot considered optimal for a specific individual.

SUMMARY

Combining an understanding of the normal human foot with knowledge of how prosthetic feet try to emulate those functions assists in the design, evaluation, prescription, and use of a prosthetic foot and ankle system. Based on available research, education, empirical observation, and patient feedback, clinicians can select an appropriate foot and optimize its use for an individual amputee. Evaluating a specific foot based on its functional subsets may help practitioners and scientists develop a more accurate method of matching patient needs and prosthetic component performance.

Correspondence to: Gerald Stark, BSME, CP, FAAOP, 421 Cheever, Geneva, IL 60134;
e-mail: .

GERALD STARK, BSME, CP, FAAOP is affiliated with Northwestern University, Chicago, Illinois.

References:

Perry J. Gait Analysis: Normal and Pathological Function. Thorofare, NJ: Slack; 1992.
Valmassy R. Clinical Biomechanics of the Lower Extremities. St. Louis: Mosby-Year Book; 1996.
Northwestern University Prosthetic Orthotic Center. Northwestern University Transtibial Instruction Manual. Northwestern University Prosthetic Orthotic Center, Fall 1991.
Perry J. "Amputee Gait." In: Smith D, Michael J, Bowker J, eds. Atlas of Amputations and Limb Deficiencies. 3rd ed. Rosemont, IL: American Academy of Orthopedic Surgeons; 2004, pp 367–384.
Radcliffe C. Functional considerations in the fitting of above knee prostheses. Artif Limb 1955;2:35–60.
Michael J. Energy storing feet: a clinical comparison. Clin Prosthet Orthot 1987;11:154–168.
Hansen A, Childress D, Miff S, et al. The human ankle during walking: implications for design of biomimetic ankle prostheses. J Biomech 2004;37:1467–1474.
Nielsen D, Shurr D, Golden J, Meier K. Comparison of energy cost and gait efficiency during ambulation in below-knee amputees using different prosthetic feet: a preliminary report. J Prosthet Orthot 1989;1:24–31.
Gard SA, Konz RJ. The influence of prosthetic shock absorbing pylons on transtibial amputee. J Proc AAOP, 2001.
Stark G. Jumping into new technology. O&P Almanac 2003; 52(7):48–51.