Secondary Logo

Journal Logo


Effects of Shoe Heel Height on the Roll-Over Shapes of Prosthetic Ankle-Foot Systems: Implications for Heel-Height-Adjustable Components

Hansen, Andrew H. PhD; Childress, Dudley S. PhD

Author Information
JPO Journal of Prosthetics and Orthotics: January 2009 - Volume 21 - Issue 1 - p 48-54
doi: 10.1097/JPO.0b013e318191c902
  • Free

Able-bodied ambulators frequently wear shoes of different heel heights with little or no difficulty. Several studies have examined the effects of shoe heel height on the gait of able-bodied persons.1–3 The results of these studies suggest that able-bodied persons primarily adapt their ankle motion to accommodate shoes of different heel heights with other subtle changes occurring at the knee. A study of the effects of heel height on ankle-foot roll-over shape (i.e., the effective rocker shape created by the ankle-foot system during walking) in young able-bodied women showed that the ankle adaptations to shoe heel height led to roll-over shapes that were similar in geometry and orientation (Figure 1).4 The roll-over shapes in the Hansen and Childress4 study were parallel and offset vertically because the combined effect of ankle plantarflexion and the higher heeled shoes increased the ankle’s height above the floor during walking (and the ankle is the origin of the shank-based coordinate system). The radii and forward shifts of the best-fit circular arcs to the roll-over shapes did not change as a function of heel height or walking speed. Other studies have also suggested an apparent invariance of the knee-ankle-foot roll-over shape to walking speed5 and walking with added weights.6 This apparent invariance of the roll-over shape suggests that it may be a goal of the able-bodied locomotor system for level walking.

Figure 1.
Figure 1.:
Drawing of the adaptation that occurs with the able-bodied ankle-foot system when wearing shoes of different heel heights.4 A, Roll-over shape of the ankle-foot system in a low-heeled shoe. B, Roll-over shape of the ankle-foot system in a high-heeled shoe. C, Superimposed ankle-foot systems (and roll-over shapes) in the low- and high-heeled shoes. The adaptation results in ankle-foot roll-over shapes that are parallel (figure reprinted with permission from J Rehabil Res Dev 2004;41:547–554).

Although able-bodied persons seem to adapt automatically to shoes of widely different heel heights, lower limb prosthesis users (e.g., transtibial or transfemoral prosthesis users) are often restricted to the use of a narrow range of heel heights. When the shoe heel height is changed beyond this small range, an alignment change in the prosthesis is usually necessary to accommodate the higher or lower heel height. Kapp and Cummings7 stated “Shoe heel height is probably the single most important factor of shoe fit as regards to prosthetic foot function …” and “Once a prosthesis has been aligned and fabricated, the patient should not significantly increase shoe heel heights unless an appropriate wedge is added inside the shoe.” Some prosthetic systems have been developed to allow the prosthesis user the ability to change the “resting” level of dorsiflexion of the ankle-foot system— or “set point”—to adapt to changes in shoe heel height. However, these systems require manual methods for setting the alignment and are not automatic as in the case of the intact able-bodied ankle-foot system.

The roll-over shapes of several ankle-foot prostheses were examined in this study using a modified quasi-static testing procedure described previously.8 This procedure was found to give similar results as feet tested under dynamic walking conditions.8 Prosthetic ankle-foot systems were tested while using a no-heel shoe and a low-heel shoe. The hypothesis regarding the prosthetic ankle-foot testing was that these systems would not adapt to changes in shoe heel height, resulting in ankle-foot roll-over shapes that would not be parallel to one another in the shank-based reference frame without alignment changes at the ankle.


The following prosthetic ankle-foot systems were tested in a random order: C-Walk (Otto Bock Healthcare, Duderstadt, Germany), TruStep (College Park Industries, Fraser, MI, USA), Dynamic Response Foot with Multiflex Ankle (Endolite North America, Centerville, OH, USA), Masterstep (Ossur, Reykjavik, Iceland), SACH (Kingsley Mfg. Co., Costa Mesa, CA, USA), Single Axis (Otto Bock Healthcare), and Total Concept™ (Century XXII Innovations Inc., Jackson, MI, USA). Side-view photographs of each foot were taken using a digital camera. Feet were selected based on convenience sampling of prosthetic feet available within our laboratory and were chosen to represent a wide variety of prosthetic feet commercially available (e.g., energy storage and return, multi-axial, SACH, single-axis, and heel-height adjustable).

The feet were mechanically tested using a modified version of the quasi-static roll-over shape method described in Hansen et al.8 In this method, the foot was attached to a weight via a pylon (Figure 2). The foot was mounted such that the pylon was orthogonal to the top surface of the foot. Two markers were placed along the side of the pylon and in line with the bolt hole of the prosthetic foot (in a sagittal plane projection). The apparatus was then pressed onto a force platform and rolled over the force platform in a slow manner. A visual display of the vertical force being applied to the force platform was used as feedback to the technician conducting the test so he could act to maintain the force near the weight of the person who would use the foot (i.e., between approximately 90% and 110% of the weight of the person). The same technician conducted all of the loading tests in this study. The trajectories of the two markers and center of pressure were measured during the trial in laboratory coordinates using an eight camera motion analysis system (Motion Analysis Corporation, Santa Rosa, CA, USA) and a force platform (Advanced Mechanical Technology, Inc., Watertown, MA, USA), respectively. To find the roll-over shape of the prosthetic ankle-foot system, the center of pressure was transformed into a shank-based coordinate system. To create the shank-based coordinate system, a unit vector was created between the lower and upper markers on the pylon. This vector was used as the y axis for the shank-based coordinate system. A perpendicular unit vector to the y axis that was in the sagittal plane with positive pointing “forward” was used as the x axis of the shank-based coordinate system (Figures 2B and C). The roll-over shapes indicated center of pressure in shank-based coordinates for shank angles of −15 to 25 degrees with respect to the vertical.

Figure 2.
Figure 2.:
Modified quasi-static roll-over prosthetic ankle-foot testing method (similar to the protocol in Ref. 8). The prosthetic ankle-foot system is connected to a weight using an orthogonal pylon and is rolled over a force platform. The force is maintained near the level of the user’s body weight through loading situations ranging from (A) heel contact to (B) midstance to (C) opposite heel contact. The technician controls force manually while using visual feedback of the vertical force applied to the force platform.

Prosthetic feet were tested first without shoes. The resulting roll-over shapes were superimposed onto the side-view photographs of the feet. Prosthetic feet were then tested with shoes. The first shoe had essentially no difference between the forefoot and rearfoot sole thicknesses. The second shoe had a small difference between forefoot and rearfoot thicknesses (approximately 20 mm). Both shoes are shown in Figure 3. The shoes were chosen to represent a small change in heel height. Larger shoe heel height changes could be expected to yield greater differences. Shoes were chosen to have the appropriate size to be used with the prosthetic feet in the study, i.e., relatively tight fits that would be acceptable in a clinical situation. No-heel wedges were used in the study. No alignment changes were made to accommodate the shoes. Tests were conducted to determine the change in roll-over shape that would occur because of changes in shoe heel height only.

Figure 3.
Figure 3.:
Shoes used on the prosthetic ankle-foot systems. The top shoe has zero difference between heel and forefoot sole thicknesses. The bottom shoe has a difference between the heel and forefoot sole thicknesses of approximately 20 mm.

The Total Concept prosthetic ankle-foot system was tested in a neutral position (i.e., bottom of the foot at a 90-degree angle with the pylon) with both shoes and in a plantarflexed position with both shoes. The amount of plantarflexion was set such that it made the pylon vertical in the low-heeled shoe (with the shoe sitting on a level surface). The Total Concept prosthetic ankle-foot system was tested in different alignments because it has an adjustable ankle joint built into it to accommodate shoes of different heel heights. Some of the other feet tested also had mechanisms that allowed for changes because of heel height but they required the use of tools for adjustment. Three trials were performed for each foot-shoe combination and the average roll-over shape was calculated.

The roll-over shapes were fit with the lower arc of a circle as described in Hansen et al.5 The fitting routine returned a radius and forward shift (arc center in the fore-aft direction) of the best-fit circular arc to the roll-over shapes. The radius of the roll-over shape of each prosthetic foot in use with the no-heel shoe was determined. This radius was then assumed during the fitting of the best-fit arc to the roll-over shape of the foot being used with the low-heel shoe. Forcing the same radius for each condition removed the inherent coupling between radius and forward shift in the circular fitting routine.5 Perfectly parallel roll-over shapes would have the same forward shifts of their best-fit circular arcs. For this reason, differences in the forward shifts of the arcs between shoe conditions were interpreted as differences in the alignment of the foot-shoe combinations. From earlier descriptions of circular arc alignment,9,10 a roll-over shape that is translated backward in the shank-based coordinate system is similar to a roll-over shape that is “dorsiflexed” in this same coordinate system. Therefore, we hypothesized that roll-over shapes would have smaller forward shifts when using the low-heel shoe compared with those found when using the no-heel shoe when no alignment adjustments were made to accommodate shoe heel height. We expected similar forward shifts when the prosthetic foot is aligned to accommodate shoe heel height.


The photographs of the various prosthetic ankle-foot systems tested in the study are shown in Figure 4 with their unshod roll-over shapes superimposed on the images. The roll-over shapes of the various prosthetic ankle-foot systems with the two different shoes (no alignment adjustment to accommodate the different shoe heel heights) are shown in Figure 5. The roll-over shapes of the Total Concept prosthetic ankle-foot system are shown in Figure 6 with the following configurations: neutral position in the no-heeled shoe and plantarflexed position in the heeled shoe. The roll-over shapes seem to support the hypothesis that prosthetic ankle-foot systems in this study cannot adapt automatically to shoes of different heel heights without a change in alignment.

Figure 4.
Figure 4.:
Photographs of the prosthetic feet used in this experiment with roll-over shapes of these feet superimposed. Feet used were (A) Otto Bock C-Walk, (B) College Park TruStep, (C) Endolite Dynamic Response Foot with Multiflex Ankle, (D) Ossur Masterstep, (E) Kingsley SACH, (F) Otto Bock Single-axis, and (G) Century XXII Total Concept. Note that scales are not the same for all plots.
Figure 5.
Figure 5.:
Roll-over shapes of the prosthetic ankle-foot systems when tested in a neutral alignment and with each of the two shoes shown in Figure 3: (A) Otto Bock C-Walk, (B) College Park TruStep, (C) Endolite Dynamic Response Foot with Multiflex Ankle, (D) Ossur Masterstep, (E) Kingsley SACH, (F) Otto Bock Single-axis, and (G) Century XXII Total Concept. The roll-over shapes shown using circles are for the no-heel shoe (white shoe in Figure 3) and the roll-over shapes shown using squares are for the low-heel shoe (black shoe in Figure 3). Note that scales are not the same for all plots.
Figure 6.
Figure 6.:
Roll-over shapes of the Century XXII Total Concept prosthetic ankle-foot (AF) system when wearing the no-heel shoe (in a neutral alignment) and the low-heel shoe (with a plantarflexed alignment to accommodate the shoe heel height). This figure and Figure 5 illustrate why persons using lower-limb prostheses have difficulties changing between shoes of different heel heights. This figure also illustrates how alignment can be used to accommodate changes in shoe heel height by helping to maintain a properly oriented roll-over shape (compare this figure with the drawing in Figure 1).

When alignment adjustments were not made to accommodate the difference in shoe heel heights, forward shifts were lower for feet used with the low-heel shoe compared with those found for the feet used with the no-heel shoe (Figure 7). When the alignment adjustment was made to accommodate the heel height difference, the forward shifts were nearly equal, supporting the hypothesis.

Figure 7.
Figure 7.:
Forward shifts of the best-fit circular arcs to the roll-over shapes. For all of the feet that were not aligned to accommodate the shoe, large differences are seen in the forward shift of the best-fit arc. In these cases, the heel essentially shifts the best-fit arc backward. In the case in which the Century XXII Total Concept was aligned for the shoe, the forward shift of the best-fit arc when using the low heel shoe was nearly identical to that seen in the same foot when using the no-heel shoe and the initial alignment.


The results of this article support the hypothesis that typical prosthetic ankle-foot systems cannot adapt automatically to changes in shoe heel height. These systems can be altered using changes in alignment to properly orient the roll-over shape of the ankle-foot-shoe system. However, these changes require manual adjustments and, with most systems, require tools to make the adjustments. The use of heel-height-adjustable prosthetic ankle-foot systems (ones that do not require the use of tools) may be advised for persons who would prefer to use shoes of different heel heights and who understand the necessary alignment changes needed for switching between these shoes. Further work is needed to examine the consequence of using shoes of different heel heights without changing the alignment of the prosthesis to accommodate these heel heights.

Limitations of this study include the fact that results are based on mechanical loading tests that are quasi-static in nature. However, earlier work by Hansen et al.8 showed that prosthetic ankle-foot roll-over shapes are similar when using this method as when using dynamic loading reflective of walking. Another limitation of the study is that only one sample of each prosthetic ankle-foot system was used in the study. However, the effects of alignment on the prosthetic ankle-foot systems were consistent across all seven components, which are significantly different in their mechanical construction. It is unlikely that variability within prosthetic ankle-foot system type would change the conclusions of the study. Lastly, the adjustment of only one foot’s alignment in the study makes statistical analysis impossible.

The inability of prosthetic ankle-foot systems to adapt is perhaps due to the fact that these systems have set points that cannot be altered without changes in alignment. The able-bodied lower limb system has set points that can be set “on the fly” allowing automatic adaptation. Future designs of ankle-foot prostheses could make attempts at systems that can adapt automatically to changes in shoe heel height, perhaps using an invariant roll-over shape as a design constraint. This feature could allow persons using prostheses the ability to walk using a variety of shoes (and to walk without shoes) comfortably and without needing manual adjustments of their prostheses. The use of similarly oriented roll-over shapes would presumably yield similar loading moments to the residual limb socket, which may be perceived as a similar comfort level by the prosthesis user. The ability to automatically adjust for different heel heights may also lead to automatic adaptability for walking on different ramped surfaces.9,11


Prosthetic ankle-foot systems tested could not adapt to small changes in shoe heel height without experiencing changes in the orientation (measured as forward shifting) of their roll-over shapes. An alignment change can be made to properly orient the roll-over shape of the ankle-foot-shoe system when wearing shoes of different heel heights, but this change requires a manual adjustment of the prosthesis.


The authors acknowledge the use of the VA Chicago Motion Analysis Research Laboratory of the Jesse Brown VA Medical Center, Chicago, Illinois.


1.Opila-Correia KA. Kinematics of high-heeled gait. Arch Phys Med Rehabil 1990;71:304–309.
2.Opila-Correia KA. Kinematics of high-heeled gait with consideration for age and experience of wearers. Arch Phys Med Rehabil 1990;71:905–909. Lateur BJ, Giaconi RM, Questad K, et al. Footwear and posture: compensatory strategies for heel height. Am J Phys Med Rehabil 1991;70:246–254.
4.Hansen A, Childress D. Effects of shoe heel height on biologic roll-over characteristics during walking. J Rehabil Res Dev 2004;41:547–554.
5.Hansen A, Childress D, Knox E. Roll-over shapes of human locomotor systems: effects of walking speed. Clin Biomech 2004;19:407–414.
6.Hansen A, Childress D. Effects of adding weight to the torso on roll-over characteristics of walking. J Rehabil Res Dev 2005;42:381–390.
7.Kapp S, Cummings D. Prosthetic management. In: Bowker JH, Michael JW, eds. Atlas of Limb Prosthetics: Surgical, Prosthetic, and Rehabilitation Principles. St. Louis: Mosby-Year Book, Inc.; 1992:453–478.
8.Hansen A, Childress D, Knox E. Prosthetic foot roll-over shapes with implications for alignment of trans-tibial prostheses. Prosthet Orthot Int 2000;24:205–215.
9.Hansen AH. Roll-over characteristics of human walking: with applications for artificial limbs [dissertation]. Evanston, IL: Northwestern University; 2002.
10.Hansen AH, Meier MR, Sam M, et al. Alignment of trans-tibial prostheses based on roll-over shape. Prosthet Orthot Int 2003;27:89–99.
11.Hansen A, Childress D, Miff S. Roll-over characteristics of human walking on inclined surfaces. Hum Mov Sci 2004;23:807–821.

prosthetic foot; shoes; heel height; roll-over shape

© 2009 American Academy of Orthotists & Prosthetists