Secondary Logo

Journal Logo


Trimline Severity Significantly Affects Rotational Stiffness of Ankle-Foot Orthosis

Bielby, Shea A.; Warrick, Tim J.; Benson, David CO; Brooks, Ronald E. CPO; Skewes, Edward CPO; Alvarez, Estefania MS; Dunning, Christopher BS; DesJardins, John D. PhD

Author Information
JPO Journal of Prosthetics and Orthotics: October 2010 - Volume 22 - Issue 4 - p 204-210
doi: 10.1097/JPO.0b013e3181f9082e
  • Free

Ankle-foot orthoses (AFOs) offer a wide range of supportive and rehabilitative ankle and foreleg apparatuses that aid individuals with acute and chronic disorders. Usually made of plastic, AFOs are worn on the lower leg and foot to support the ankle and to hold them in the correct position. It is also commonly known as the “foot-drop” brace, because an AFO has the ability to help the patients experiencing significant weakness of ankle and toe dorsiflexion.1

The primary functional goals in the design and use of AFOs can be summarized by three areas: 1) motion control and compensation for weakness, 2) maximize comfort for the patient, and 3) cosmetic attractiveness.2 These goals can be attained through either active dynamic or passive dynamic AFOs, which are either powered by external devices or rely on material characteristics for their function, respectively.3

AFOs are beneficial to a wide variety of patients with problems ranging from spina bifida to brain and spinal cord injuries, and AFOs are typically prescribed to improve gait performance.3,4 AFOs accomplish this rehabilitation in several different ways, one of the most common being through gait improvement. This improvement is because of the ability of orthosis to reduce the energetic cost of ambulation by providing support to the joint and, thus, allowing the patient to walk in a more natural manner.1

To best serve the patient, the effects of AFO systems on the ankle and foot biomechanics should be considered, so that such devices will be able to sufficiently restrict motion and compensate for any weakened muscular function the patient may have. Ankle motion occurs on three planes: the sagittal, transverse, and frontal and is composed of three motions: flexion/extension, abduction/adduction, and inversion/eversion. In these planes, weightbearing and nonweightbearing motions occur.5 When thinking about testing the mechanical properties of AFOs, ranges of motion of the ankle should be taken into consideration. The major ranges of motion of the ankle are dorsiflexion (0°–20°), plantarflexion (0°–50°), inversion (0°–20°), eversion (0°–40°), adduction (0°–20°), and abduction (0°–10°).4 The talocrural joint allows the dorsiflexion/plantarflexion motions in the saggital plane, the sutalar joint allows the inversion/eversion motions in the frontal plane, and the tarsal joints allow the adduction/abduction motions in the transverse plane.

In lieu of these three planes of motion, it is vital to test the AFO in all of them. The main purpose of prescribing an AFO is to provide appropriate control in the plane where motion is lacking because of musculoskeletal pathology. Because of this, there is a possibility that the patient may exhibit weakness in areas other than the saggital plane. This necessitates evaluation in all planes of motion, to provide proper support. Along the same lines, the orthotist does not want to overbrace the patient and, in turn, actually cause muscle weakening through the restriction of sound muscle groups over time. Thus, it is equally as important to test the frontal and transverse stiffness of the device to see how effectively the device can prevent inversion/eversion and supination/pronation along with testing plantarflexion/dorsiflexion.

Currently, solid polypropylene is the material of choice in the manufacture of custom-molded AFOs because of its ability as a thermoplastic to be deformed and reformed under heat, along with its resistance to deformity and fatigue.6 Even though polypropylene is the major material used in the manufacturing of AFOs, alternate materials can be used.

When selecting a material, traits to be considered are thickness, tensile strength and modulus, flexural modulus, and tensile elongation at break.3 Using a material that has exceptional test results in these areas will provide a durable material to construct the AFO. Although improvements have been made over the years, there are still problems with AFO design and manufacturing. Most of the manufacturing process is done by hand, which is time-consuming and tedious.3,7 The custom design process is done, however, to ensure that each patient receives a custom orthosis that fits their needs, rather than a universal model that may not provide appropriate orthosis management for the patient's pathology. The manufacturing parameters are also difficult to work with, such as thickness and curvature, especially when working by hand.6 The trimlines of the medial and lateral sides of the AFO are other major manufacturing parameters to be considered, which are best implemented through custom fitting by hand. These trimlines affect the flexibility of the AFO as a whole, especially in dorsiflexion and plantarflexion, and run from a proximal anterior location at the patient's calf to a distal anterior location at the patient's foot.6 The effect of mechanical properties of AFO on resulting ankle motion and stiffness are critical factors in the stabilization of the patient's joint.

There have been many attempts at quantifying ankle biomechanics in concordance with testing AFOs. One study specific to pediatric AFOs tested the ability of brace to deform to 10° of dorsiflexion. Through testing over different time intervals, it was found that even if a patient was able to apply the amount of stress to an AFO as was seen during rigorous testing so as to deform it, recovery of the viscoelastic properties will occur if given the right amount of time (i.e., while the patient is asleep).4 In 2007, MIT developed an “anklebot” in its Laboratory for Biomechanics and Human Rehabilitation, which allowed for identification and quantification of ankle stiffness.8 Our study similarly seeks to quantify stiffness in a clinically relevant manner but is looking at the stiffness of AFOs through analysis of the changes of rotational stiffness of the AFO as a function of trimline measures, rather than that of the ankle joint. In 2003, more similarly to this study, the development of a “novel device to evaluate the stiffness of ankle-foot orthosis devices” was reported by Cappa et al.9 The article stated that “such devices (in other literature) lack in the evaluation of the stiffness of thermoplastic AFOs by means of 1-D or pseudo-2-D approach and generally have determined the stiffness values only in two fixed directions.” This inhibits real-life simulation and will, therefore, have little to no clinical significance. The results of their new design aided in prescribing a more appropriate device for the patient based on actual mechanical data, rather than the visual geometry of the AFO.3

In terms of gait, it has been found that the stiffness and original angle of the AFO has a significant impact on various stances. Miyazaki et al.10 found that AFO stiffness had a significant impact in the early stance of gait when compared with the mid-stance to late stance and that the active ankle moment in the direction of plantarflexion varied with AFO stiffness in 11 of 20 subjects. Guillebastre et al.11 found that wearing a rigid AFO disturbed walking velocity, step length, and time with an asymmetry between sides. These studies, however, focused on the AFO as a whole, and the impact of AFO stiffness in each individual plane on gait was not quantified.

In this study, we have designed and used a portable bench-top AFO testing instrument to test the hypothesis that changes in AFO malleolus trimlines will significantly affect the resulting stiffness of custom-molded, thermoplastic AFOs in all three ankle motion planes.


A manual loading AFO testing instrument was designed and constructed for use in this study (Figure 1) to replicate the 3 degrees of freedom rotational axes of the ankle. The apparatus can replicate the key ranges of motion: dorsiflexion/plantarflexion (+20°/−50°), inversion/eversion (±40°), and adduction/abduction (±40°), with the ability to isolate or combine these ranges of motion as desired. This instrument rigidly holds the calf and foot portion of the AFO, while applying rotational torques about the ankle axes, inducing foot rotations through clinically relevant ranges of motion. To measure the functional range of motion that is experienced by the AFO during each testing cycle, each axis of motion is equipped with a Novotechnik P2501 potentiometer (Novotechnik, Southborough, MA) with a measured accuracy of ±0.01°. The moments applied about the ankle axis were recorded using a Transducer Techniques SWS250 torque cell (Transducer Techniques, Temecula, CA) with a measured accuracy of ±0.020 Nm. Because the tests performed in this study required only one free axis of rotation at a time, one torque cell was mounted concentrically about the desired axis of rotation. A moment arm was attached to the open end of the torque cell to assist in the application of the moment. This testing apparatus can accommodate all sizes of AFOs because of its ability to adjust the platform vertically and the calf securing mechanism anteriorly and posteriorly to ensure that the AFO is centered at the intersection of all three axes of motion of the ankle replicated by the machine. The measurements can in turn be used to calculate the rotational stiffness of the AFO through the equation:

Figure 1.
Figure 1.:
Ankle-foot orthosis testing apparatus. Plane ZY is the sagittal plane where dorsiflexion/plantarflexion occur, XY is the horizontal plane where abduction/adduction occurs, and ZX is the frontal plane where inversion/eversion occurs.

Where M is the applied moment to the specimen, and θ is the degree of rotation experienced by the specimen at the ankle. Because of the accuracy of the measurements for both torque (M) and angle (θ), the resulting values for calculated rotational stiffness (k) will be accurate as well. Calibration of the testing apparatus was done by recording the voltage outputs of the potentiometers at several different defined angles, and a linear curve was fit to the data. The equation to this trend line was used to create a calibration equation in LabVIEW. The torque cell had a defined calibration of 13.558 Nm/V. This information was used to create a calibration curve for the torque cell in LabVIEW.

A series of lightweight, 4.7625-mm thick polypropylene, rigid, solid-ankle AFOs were fabricated from an identical patient's lower leg casting (patient age 15 years) by the Orthotics and Prosthetics Department of Shriners Hospital of Greenville, SC. All AFOs were tested serially for measures of rotational stiffness in three “trimline” configurations: “conservative,” “moderate,” and “aggressive” (Figure 2). AFOs have a medial and lateral trimline around the malleoli that can be modified to meet a specific patient pathology and, therefore, do not have to be the same length. A trimline essentially determines the amount of material present to resist rotation about the ankle. Initially, the average conservative trimline lengths measured approximately 110 mm on the medial side and 100 mm on the lateral side. The average moderate trimline lengths measured 91 mm on the medial side and 44 mm on the lateral side. The average aggressive trimline lengths measured 74 and 54 mm for medial and lateral sides, respectively. Trimline length was measured using callipers at a 45° angle from the back of the heel on either the medial or the lateral side of the ankle, depending on which side was being trimmed. Figure 3 displays the relationship found between trimline depth and stiffness.

Figure 2.
Figure 2.:
Averages of the trimline lengths of each of the trimlines with respect to the side of the ankle-foot orthosis that the material was trimmed from, medial or lateral.
Figure 3.
Figure 3.:
Trend fitting of the average stiffness and the average trimline length, revealing a linear fit, where trimline length is measured 45° from the back of the heel.

For each testing iteration, an AFO from each of the trimline configuration was tested in each of the following ranges of motion: dorsiflexion/plantarflexion (±4°), inversion/eversion (±7°), and adduction/abduction (±7°). Each specimen was tested five times to ensure reproducibility of the testing cycle. Care was taken to ensure minimal damage, and plastic deformation was inflicted on the AFO during serial testing by testing on the aforementioned ranges that are well within the elastic limits of the specimen. The stiffness of the AFOs about each rotational axis was calculated using LabVIEW by graphing the angle of rotation against the moment applied and fitting a linear slope to the curve as was consistent with data analysis methods used by Bregman et al.12

A general linear model was fitted using Minitab 15 software to compare the effects on the average stiffness with respect to trimline location, plane of motion, and specimen number. Using these results, a two-way analysis of variance (ANOVA) was done to quantify the interaction between trimline and plane of motion with respect to the average stiffness measurements. ANOVA assumes independence of samples, and although the same AFO was repeatedly tested in this study, testing conditions assured that plastic deformation was kept to a minimum, allowing for the independent samples assumption. All the stiffness data collected were then sorted according to the type of trimline and plane of motion, making up a total of nine categories (conservative trimline with dorsiflexion/plantarflexion, conservative trimline with inversion/eversion, etc.). A one-way ANOVA with Tukey comparative analysis was then examined for statistical difference between the nine different categories of data.

Before the collection of experimental data, reproducibility of the testing apparatus was evaluated. Five testers tested the same specimen five times along all three ranges of motion. The results were plotted in groups according to range of motion (inversion/eversion, plantarflexion/dorsiflexion, and abduction/adduction). The slope of each of these lines was recorded along with their R2 values to ensure accuracy. The overall reproducibility for the testing apparatus was then evaluated, combining all the results from each testing interval (25 total).


Figure 4 shows the average rotational stiffness values for the three AFO trimline configurations about each of the three ankle rotation axes. The one-way ANOVA showed that all stiffness measures were statistically significantly different between all trimlines (p < 0.05) with the exception of the moderate to aggressive trimlines in the adduction/abduction range of motion. All stiffness values changed dramatically between conservative and aggressive trimlines, with a total 82.9% decrease in stiffness in dorsiflexion/plantarflexion, a total 65.0% decrease in stiffness in inversion/eversion, and a total 76.3% decrease in stiffness in adduction/abduction. Changes in dorsiflexion/plantarflexion consistently affected the most by changes in trimlines with 56.0% decrease and 61.1% decrease between the conservative to moderate and moderate to aggressive trimlines, respectively.

Figure 4.
Figure 4.:
A comparison of rotational stiffness between the three trimline configurations as a result of the one-way analysis of variance and Tukey comparitive analysis. *Statistically significant changes in the ankle-foot orthosis stiffness between the trimline stiffness measure and the conservative measure on that axis, and +statistically significant changes between the moderate measure and the trimline stiffness measure on that respective axis (n = 4).

The results of the reproducibility analysis reveal that measured variability was low. The standard deviations for the dorsiflexion/plantarflexion, inversion/eversion, and abduction/adduction planes were found to be 0.460, 0.207, and 0.068 Nm/°. The general linear model also showed that the difference in average stiffness between the four different specimens was not significant (p = 0.921).

Figure 3 shows the correlation between the average stiffness measurements and the trimline lengths for each of the categories: conservative, moderate, and aggressive. This linear fit is helpful when looking at stiffness measures and trimline depths in between the values that were investigated in this study. The graphs reveal the general trend that because more material is trimmed away (the smaller the trimline length), the less stiff the specimen is, as was expected. This trend is instrumental in determining the length of the trimline required to achieve the specific AFO stiffness needed by the patient.


In a clinical setting, eligible patients exhibiting ankle/foot disorders are prescribed an AFO to provide the support that they are lacking. In addition to personalized curvature, thickness, and AFO size, trimlines are also modified for the individual patient's needs in hopes to produce an AFO with sufficient support for the injury or illness.

Through our research efforts, we have improved the quality of the mechanism by which AFOs are tested. AFOs are tailored to fit individual patient needs and can range from stiff and rigid to flexible and articulated but are rarely quantified for functional measures of mechanical stiffness or range of motion. Changes in AFO stiffness can be achieved by trimming away more or less material during fabrication. A common challenge with the AFO is where to make the trimline in relation to the malleolus (or ankle) to achieve an appropriate stiffness for the patient's needs.

This study reveals that there is a statistically significant change between the conservative, moderate, and aggressive trimlines in the realm of stiffness performance. Figure 4 shows that all planes relative to their respective trimline were statistically different except for the aggressive and moderate trimlines along the adduction/abduction plane. The realization, through this study, that each trimline produces a significantly different stiffness is vital to the applicability of AFOs to different situations. For example, if a stiffness of 2 Nm/° was needed in the dorsiflexion/plantarflexion range of motion, this study has shown that using a conservative rather than a moderate trimline has a significant impact on the stiffness exhibited during wear (Figure 3). These results add another dimension of versatility and customizability to the specimens, allowing for a more personalized treatment.

Although this knowledge allows clinicians to have a better understanding of, generally, what stiffness and trimline would best benefit their patient, the results of these trials are also broad and are a good foundation for further customizing the AFO to meet the needs of the patient according to specific pathologies and functional demands. Information gathered in this study may contribute to the future development of a system for clinicians to determine the appropriate trimline for the patient before production of the AFO begins to help to save time and money.

Currently, hand-made AFOs are trimmed until the clinician “feels” that the AFO will provide appropriate support for the patient. The next objective in this study is to analyze and record AFO stiffnesses for particular AFO prescriptions to build a database indicating the appropriate stiffness for a specific patient pathology. Knowledge of the patient's specific pathology allows the physician to analyze the extent to which the patient is lacking support, and in which directions of motion, it is lacking, be it dorsiflexion, inversion, etc. From there, the clinician would be able to approximate the stiffness required for the patient and then be able to identify the trimline required to replicate that stiffness through the use of equations similar to those found in Figure 3.

There are several limitations in this study that must be noted. Only one patient's AFO was used, resulting in the exclusion of the impact of patient's size and foot shape on the overall performance of the AFO. In addition, testing of the AFOs occurred with the footplate fixed to the base of the testing apparatus. Although this may not completely mimic the motions an AFO experiences when donned by a patient during dynamic conditions, it gives a reasonable approximation for the stiffness the AFO is subjected to during the gait phases. Trimline measurements were taken at a 45° angle from the base of the heel to maintain consistency with practices used at Shriners Hospital, which prevents the further understanding of how variation along the entire trimline length affects the overall rotational stiffness of the AFO. Finally, analysis of the AFO rotational stiffness was based on the linear models as opposed to nonlinear models, which may more accurately predict rotational stiffness of an AFO. Despite these limitations, this study serves as a foundation for further research in this area.

Through the development of this testing device and subsequent analysis, the results obtained during this study may be helpful in the preliminary development of a trimline/stiffness analysis system for clinicians to use when evaluating patients with varying pathologies and the prescription of an appropriate AFO to fit their specific needs.

The reproducibility results obtained by this study ensure that future studies are able to be accurately performed without the danger of having a wide variability between measurements taken by different people or between different AFO subjects. Bregman et al. and Sumiya et al. conducted similar studies to determine the stiffness of AFOs and emphasized the importance of repeatability of their experiments. Both studies involved the design of a testing apparatus and using an applied force to calculate the stiffness of an AFO and also had high levels of reproducibility. Sumiya et al. also conducted a similar study, but only one plane of motion was evaluated with different values of moments about the ankle joint. The standard deviation of the stiffness measurements obtained from our apparatus ranged 0.460 Nm/° for dorsiflexion and plantarflexion, 0.207 Nm/° for inversion and eversion, and 0.068 Nm/° for adduction and abduction, whereas the standard deviations were larger for the other study. This is because our apparatus analyzed all three planes of motion rather than just one, and reproducibility was analyzed across five different testers, whereas the study conducted by Sumiya et al. used only one experimenter to calculate reproducibility. Lower torques were also explored during our testing (no >5 Nm), whereas the testing for the study conducted by Sumiya et al. went up to a magnitude of 30 Nm, which requires nonlinear analysis because of the deformations incurred on the specimen.12,13

The study conducted by Bregman et al. used a G-study approach for their analysis of reliability/reproducibility during testing. Despite this different approach, their biggest contribution to error variance was found when comparing different testers, as was found with our study. The article states that “AFO characteristics can be measured in a reliable manner, when the appropriate study design is used,” suggesting that having a strict uniform testing procedure can minimize this variability between testers.12

The linear fit found relating the trimline depth to the average stiffness along each of the three planes was found to be a very good model and can be used to predict behavior of trimline depths in between the values tested during this study. These models will serve as a baseline for comparison of different AFOs undergoing the same testing regimen.

Future work for this study will include the development of a manufacturing algorithm, which inputs the desired AFO clinical prescription and outputs material type, trimline location, and 3D stiffness of the AFO. In addition, an analysis of the effect of material thickness and footwear support on the resulting AFO rotational stiffness will be performed along with the use of finite element analysis of AFOs to see computer-generated models of stress locations and deformations on the AFO to add another practical and clinical dimension to the study.


The authors thank the BioE850 Spring 2007 design group and the 2008–2009 SC Life Undergraduate group for their design development efforts on this project.


1.Brehm MA, Harlaar J, Schwartz M. Effect of ankle-foot orthoses on walking efficiency and gait in children with cerebral palsy. J Rehabil Med 2008;40:529–534.
2.Blaya JA. Force-Controllable Ankle Foot Orthosis (AFO) to Assist Drop Foot Gait. MS Thesis. Cambridge, MA: Massachusetts Institute of Technology; 2000.
3.Faustini MC, Neptune RR, Crawford RH, Stanhope SJ. Manufacture of passive dynamic ankle-foot orthoses using selective laser sintering. IEEE Trans Biomed Eng 2008;55:784–790.
4.Lunsford TR, Ramm T, Miller JA. Viscoelastic properties of plastic pediatric AFOs. J Prosthet Orthot 1994;6:3–9.
5.Nordin M, Frankel V. Basic Biomechanics of the Musculoskeletal System. 3rd ed. Baltimore, MD: Lippincott Williams & Wilkins; 2001.
6.Murdoch G, ed. The Advance in Orthotics. London, England: Edward Arnold Publishers Ltd; 1976.
7.Nagaya M. Shoehorn-type ankle-foot orthoses: prediction of flexibility. Arch Phys Med Rehabil 1997;78:82–84.
8.Roy A, Krebs HI, Patterson SL, et al. Measurement of human ankle stiffness using the anklebot. Rehabil Robot 2007;13:356–363.
9.Cappa P, Patanè F, Pierro MM. A novel device to evaluate the stiffness of ankle-foot orthosis devices. J Biomech Eng 2003;125:913–917.
10.Miyazaki S, Yamamoto S, Kubota T. Effect of ankle-foot orthosis on active ankle moment in patients with hemiparesis. Med Biol Eng Comput 1997;35:381–385.
11.Guillebastre B, Calmels P, Rougier P. Effects of rigid and dynamic ankle-foot orthoses on normal gait. Foot Ankle Int 2009;30:51–56.
12.Bregman DJ, Rozumalski A, Koops D, et al. A new method for evaluating ankle foot orthosis characteristics: BRUCE. Gait Posture 2009;30:144–149.
13.Sumiya T, Suzuki Y, Kasahara T. Stiffness control in posterior-type plastic ankle-foot orthoses: effect of ankle trimline Part 1: a device for measuring ankle moment. Prosthet Orthot Int 1996;20:129–131.

AFO; ankle stiffness

© 2010 American Academy of Orthotists & Prosthetists