The effect of carbon fiber custom dynamic orthosis use and design on center of pressure progression and perceived smoothness in individuals with lower limb trauma

Abstract

Background:

Carbon-fiber custom dynamic orthoses are used to improve gait and limb function following lower limb trauma in specialty centers. However, the effects of commercially available orthoses on center of pressure progression and patient perception of orthosis smoothness during walking are poorly understood.

Methods:

In total, 16 participants with a unilateral lower extremity traumatic injury underwent gait analysis when walking without an orthosis, and while wearing monolithic and modular devices, in a randomized order. Device alignment, stiffness, participant rating of perceived device smoothness, center of pressure velocity, and ankle zero moment crossing were assessed.

Findings:

The modular device was approximately twice as stiff as the monolithic device. Alignment, smoothness ratings, peak magnitude of center of pressure velocity, and zero moment crossing were not different between study devices. The time to peak center of pressure velocity occurred significantly later for the modular device compared to the monolithic and no orthosis conditions, with large effect sizes observed.

Interpretation:

Commercially available orthoses commonly used to treat limb trauma affect the timing of center of pressure progression relative to walking without an orthosis. Despite multiple design differences, monolithic and modular orthoses included in this study did not differ with respect to other measures of center of pressure progression. Perceived smoothness ratings were approximately 40% greater with the study orthoses as compared to previous studies in specialty centers, which may be due to a more gradual center of pressure progression, as indicted by lower peak magnitude of center of pressure velocity with both study orthoses.

1. Introduction

Carbon-fiber custom dynamic orthoses (CDOs) are provided to individuals who have sustained traumatic lower limb injuries to improve gait, quality of life, and function (Bedigrew et al., 2014; Blair et al., 2014; Franklin et al., 2019; Hsu et al., 2017; Owens et al., 2011; Patzkowski et al., 2012; Potter et al., 2018; Sheean et al., 2016; Wilken et al., 2018). CDOs are a type of ankle foot orthosis (AFO) comprised predominantly of carbon fiber that include a proximal cuff just below the knee, a posterior carbon fiber strut, a foot plate, and may also include a heel cushion. The mechanical properties and design characteristics of a CDO, including posterior strut stiffness (Haight et al., 2015; Harper et al., 2014; Russell Esposito et al., 2014; Russell Esposito et al., 2015) and bending point (Ranz et al., 2016; Russell Esposito et al., 2017), heel cushion height and durometer (Ikeda et al., 2018), and alignment (Brown et al., 2017; Russell Esposito et al., 2021; Schmidtbauer et al., 2019), have been studied previously due to their potential influence on gait in individuals with traumatic lower limb injuries. However, available studies evaluating the effect of CDO design on limb mechanics during gait are almost exclusively from a single type of CDO provided at specialized care centers, that is not widely available (Grunst et al., 2023). It is unknown if lower limb mechanics with CDOs primarily used in civilian settings resemble those with CDOs used in specialized care centers, which must meet the demands of a physically active population.

The Intrepid Dynamic Exoskeletal Orthosis (IDEO) is the focus of most CDO-related publications (Grunst et al., 2023). The CDO is stiff, modular (MOD) to allow adjustment after fabrication, and was designed to allow military service members receiving care at specialized centers to complete a range of activities including running (Russell Esposito et al., 2017). Within the civilian setting, MOD CDOs are also used to restore function after limb trauma (e.g. Reaktiv; Fig. 1) (Williamson, n.d.-a; Williamson, n.d.-b). These MOD CDOs are often similar to the IDEO in overall design but are generally more compliant and are secured using an adjustable cuff. Monolithic devices (MONO, e.g. PhatBrace; Fig. 1), which are formed as a single structure, are generally more compliant than the MOD devices, have a more compliant foot plate, and a more compliant proximal cuff with greater padding (Shuman and Russell, 2022).

Fig. 1.

Commercially available monolithic (Left) and modular (Right) carbon fiber custom dynamic orthoses (CDOs) compared in this study.

The rate at which the center of pressure (CoP) moves anteriorly under the foot during gait has been associated with patient perception of device smoothness and device satisfaction (Ikeda et al., 2018). During normal gait, the CoP gradually progresses from the heel to the forefoot. However, CDO use can cause an abrupt CoP transition from heel to the forefoot, thus negatively impacting patient perception of the device. CDO design characteristics, such as alignment and heel cushion stiffness, can affect the peak magnitude of the velocity of the CoP (vCoP), the time when the peak vCoP occurs, and when the ankle moment crosses from a dorsiflexor to plantarflexor moment (ankle zero moment crossing; ZMC) (Ikeda et al., 2018; Russell Esposito et al., 2021). Lower rates of forward progression of the CoP, resulting in a lower vCoP and earlier ZMC, that mimics normal gait are thought to be preferred and potential CDO design targets. In addition to the lack of data evaluating CDOs primarily used in civilian settings, data comparing CoP related measures in CDO users, with and without CDOs, are lacking.

Little is known about the mechanical characteristics of commercially available CDOs, the resulting effect on CoP progression, or how they are perceived. AFO stiffness, which is commonly characterized in terms of the resistance to rotation about the physiologic ankle joint, is equal to or less than intact ankle stiffness for many orthoses (0.02–8.2 Nm/degree) (Totah et al., 2019) and has been linked to changes in limb mechanics (Totah et al., 2019). Stiffness of CDOs used after trauma can vary widely with some orthoses exceeding physiological ankle stiffness (>8 Nm/degree) (Ashcraft and Grabowski, 2022), however, changes in IDEO stiffness by as much as 20% had a limited effect on lower limb mechanics (Haight et al., 2015; Harper et al., 2014; Russell Esposito et al., 2014; Russell Esposito et al., 2015). Sagittal plane CDO alignment, which can be adjusted by directly altering the CDO using shims or by adjusting wedges under the heel, is also known to affect the limb position and CoP progression during gait (Brown et al., 2017; Ikeda et al., 2018; Russell Esposito et al., 2021; Schmidtbauer et al., 2019). However, how commercially available CDOs are aligned during patient care is poorly understood. Before systematically investigating the potential importance of design characteristics in commercially available civilian devices, it is important to determine if CoP progression generally resembles that of published data from CDOs used in specialized care centers. It is also important to determine if different CDO approaches used in clinical practice, such as representative MOD and MONO designs, differ with respect to CoP progression and patient perception. Further, studies evaluating CoP progression in CDOs used in specialized care have not assessed how wearing a CDO differs from gait after trauma without a CDO.

The purpose of this study was to determine if wearing a CDO alters CoP progression relative to walking with no CDO after lower limb trauma, and if design differences between MONO and MOD CDOs commonly provided to civilians influence vCoP measures and patient perception of CDO smoothness. We hypothesized that the MONO CDO would be more compliant than MOD and would result in lower and earlier peak vCoP, earlier ZMC, and greater ratings of CDO smoothness. We hypothesized that walking without a CDO would not be significantly different than MONO with respect to peak vCoP timing, magnitude, or ZMC but differ from MOD for these measures. Further, the data will be interpreted with the context of published literature from CDOs in specialty centers.

2. Methods

2.1. Participants

Individuals (n = 16) between the ages of 18–65 years who were at least 2 years post unilateral lower extremity traumatic injury were recruited from three participating sites (The University of Iowa Hospitals and Clinics, Iowa City, USA; The Minneapolis VA Health Care System, Minneapolis, USA; Walter Reed National Military Medical Center, Bethesda, USA). Individuals who had sustained a traumatic lower limb injury were identified using the medical records at each site. Inclusion and exclusion criteria follow previously published CDO related study (Potter et al., 2018). Inclusion criteria included: weakness of ankle plantarflexors (<4/5 on manual muscle testing), limited pain free ankle range of motion (dorsiflexion<10 deg. and plantarflexion<20 deg), mechanical pain with loading onto hindfoot/midfoot/forefoot (>4/10 on verbal numeric pain rating scale), ankle or hindfoot fusion or candidate for ankle and hindfoot fusion or a candidate for amputation secondary to ankle/foot impairment, and ability to walk 50 feet at a slow to moderate pace without using a crane or crutch. Exclusion criteria included: pain >8/10 while walking, ankle weakness as a result of spinal cord injury or central nervous system pathology, requiring a knee stabilizing device to perform activities of daily living, surgery on limb anticipated in the next 6 months, medical or psychological conditions (e.g., severe traumatic brain injury, stroke, heart disease, vestibular disorder) precluding functional testing, neurologic, musculoskeletal or other conditions limiting function of the contralateral extremity, visual or hearing impairment, body mass index >45, and pregnancy. All study activities were approved by the Institutional Review Board at each participating site and participants provided written informed consent.

2.2. Testing protocol

Study participants were evaluated while walking without a CDO (NoCDO) and with a MONO and MOD CDO (Fig. 1), in a randomized order. The MONO orthosis (PhatBrace, Bio-Mechanical Composites Inc., Des Moines, IA, USA) was fabricated with a single carbon fiber strut running from the proximal cuff, down the posterior aspect of the leg, and through the foot plate. The carbon fiber strut was augmented with a fiberglass overlay that forms the remainder of the upper cuff and completes the full-length foot plate. The proximal cuff was padded and secured using two Velcro straps. The MOD orthosis (Reaktiv, FabTech Systems, Everett, WA, USA) included a proximal cuff with patellar tendon bearing design, a stiff posterior strut, and semi-rigid full-length footplate. The MOD proximal cuff includes a rigid carbon fiber anterior shell secured using a cable based ratcheting system (BOA Technology Inc., Denver CO, USA). Participant casting and fitting with the study devices was completed by Certified Prosthetist Orthotists (CPOs) based on a standardized fitting manual reflecting manufacturers recommendations. Participants accommodated to the MONO orthosis and MOD orthosis for three months prior to testing in each condition.

Kinematic and kinetic data were collected using motion capture systems with at least 12 cameras (Vicon Motion Systems Ltd., Denver, CO, USA or Qualisys AB, Gothenburg, Sweden) and at least three force plates (AMTI, Watertown, MA, USA) at a collection frequency of 1200 Hz, using methods with demonstrated reliability (Kaufman et al., 2016). Marker and force data were interpolated and smoothed using 4th order Butterworth low-pass filter with cut-off frequencies of 6 and 50 Hz, respectively. Data were collected as participants walked across a level walkway at a standardized speed based on leg length, with a Froude number of 0.16 (Wilken et al., 2012). A marker placed on the C7 vertebra was used to track walking speed and verbal feedback was provided to the study participants, if required, to ensure they were within +/− 5% of the target speed. A 57-marker set was used to track foot, shank, thigh, and pelvis segment positions in a manner consistent with previous studies (Ikeda et al., 2018; Russell Esposito et al., 2021; Wilken et al., 2012). After accommodating to the CDO, participants were asked to rate their experience walking with each CDO for perceived smoothness of gait. Perceived smoothness was assessed using a modified version of the Socket Comfort score with scores ranging from 0 = least smooth to 10 = most smooth (Hanspal et al., 2003; Ikeda et al., 2018).

Data from a minimum of five walking trials (ipsilateral heel strike to contralateral heel strike) were analyzed using Visual-3D software (C-motion Inc., Germantown, MD, USA) and then exported to a custom MATLAB program (The MathWorks, Inc., Natick, MA, USA). Two descriptive variables, stiffness and alignment, were assessed to compare the design characteristics between the two study CDOs. Three primary biomechanical outcome variables were assessed: time to peak vCoP, magnitude of the peak vCoP, and ankle ZMC. CDO alignment was calculated using the average ankle angle from 80 to 90% gait cycle, representing the angle between the plantar surface of the shoe and long axis of the tibia when the limb was in an unloaded condition. CDO stiffness was assessed using a custom-made mechanical testing device to quantify orthosis rotational stiffness about the ankle joint. The time to peak vCoP and magnitude of the peak vCoP were determined in a manner consistent with previous studies (Ikeda et al., 2018; Russell Esposito et al., 2021). The timing of the ankle ZMC was determined when the ankle joint moment transitioned from a dorsiflexor moment to a plantarflexor moment, and is presented in terms of percent stance.

2.3. Statistical analysis

Descriptive statistics including the mean and standard deviation (SD) were calculated for all variables. Paired t-tests were used to compare alignment and stiffness between the MONO and MOD conditions. Mixed model analysis of variance (ANOVAs) tests were performed to test the main effects of the condition (NoCDO, MONO, MOD) for each outcome variable (peak vCoP magnitude, time to peak vCoP, and ZMC). Significance level was set at α ≤ 0.05. Post-hoc paired t-tests with Bonferroni-Holm corrections were performed to evaluate pairwise comparisons. Smoothness scores and stiffness for the MOD and MONO devices were compared using paired t-tests. Cohen’s d, was calculated to represent the magnitude of effect size for each pairwise comparison (J C, 1988). Cohen’s d was interpreted as d ≤ 0.20 representing a small effect, d = 0.20–0.8 representing a moderate effect, and d ≥ 0.8 representing a large effect. Statistical analysis was performed using SPSS version 28 (IBM Corp, Armonk, NY, USA). Further, the results are presented relative to previously published literature on CDO use from a specialized care center.

3. Results

3.1. Demographics

In total, 16 individuals with mean (SD) age of 42.9 (11.0) years, height of 1.8 (0.1) m, and body mass of 101.8 (18.3) kg completed testing in all study conditions. Participants were 5.8 (5.8) years post injury.

3.2. Alignment

The MONO and MOD conditions were not significantly different (p = 0.189, Table 1). A small effect size was observed for the pairwise comparison between the MONO and MOD (d = 0.34).

Table 1.

Means (SD) of the orthoses alignment, stiffness, perceived smoothness scores, magnitude of the center of pressure velocity (vCOP), timing to peak vCOP, and zero moment crossing (ZMC) on the affected limb for the three conditions. N/A = not applicable.

Condition	Alignment (degree) Mean (SD)	Stiffness (Nm/degree) Mean (SD)	Smoothness Mean (SD)	vCOP magnitude (m/s) Mean (SD)	Timing to peak vCOP (% stance) Mean (SD)	ZMC (% stance) Mean (SD)
NoCDO	N/A	N/A	N/A	0.8 (0.3)	13.0 (4.4)	24.2 (9.0)
MONO	5.2 (2.7)	4.6 (2.4)	8.1 (1.4)	0.7 (0.2)	14.5 (4.7)	26.4 (6.1)
MOD	5.9 (2.0)	8.7 (2.7)*	7.6 (2.0)	0.9 (0.3)	26.1 (11.5) *	26.7 (5.8)

Significantly different from NoCDO – in bold.

Smoothness rating was based on numerical scale of 0 = least smooth and 10 = most smooth.

^*Modular (MOD) significantly different from Monolithic (MONO).

3.3. Stiffness

The MONO and MOD conditions were significantly different (p < 0.001, Table 1). A large effect size was observed for the pairwise comparison between the MOD and MONO (d = 1.24).

3.4. Perceived smoothness

Perceived smoothness was not statistically different between MONO and MOD conditions (p = 0.388, Table 1). A small effect size was observed for the comparison between the MOD and MONO conditions (d = 0.22).

3.5. vCOP magnitude

Peak vCOP magnitudes were not statistically different between the three conditions (p = 0.245, Table 1). Small effect sizes were observed for all three pairwise comparisons (MOD vs. NoCDO: d = 0.10, NoCDO vs. MONO: d = 0.39, MOD and MONO: d = 0.49).

3.6. Time to peak vCOP

The time to peak vCOP was significantly different between the three conditions (p < 0.001, Table 1, Fig. 3). The mean time to peak vCoP was greater in the MOD compared to the NoCDO (p = 0.001), and the MONO (p = 0.004) conditions. No difference in time to peak vCOP was observed between NoCDO and MONO conditions (p = 0.694). Large effect sizes were observed between the MOD and NoCDO (d = 1.0), and MOD and MONO (d = 0.86) conditions. However, the effect size between the NoCDO and MONO conditions was small (d = 0.28). Representative data from one study participant is included in the supplemental material (Supplemental fig. 1).

Fig. 3.

Mean time to peak center of pressure velocity (vCoP) for study subjects (n = 16) for the three study conditions (NoCDO = no orthosis, MOD = modular, and MONO = monolithic).

3.7. Zero-moment crossing (ZMC)

The mean ZMC values were not statistically different between the three conditions (p > 0.349, Table 1). Small effect sizes were observed for all three pairwise comparisons (MOD vs. MONO: d = 0.06, MOD vs. NoCDO: d = 0.27, NoCDO vs. MONO: d = 0.33). Representative data from one study participant is included in the supplemental material (Supplemental fig. 2).

4. Discussion

Studies evaluating the effect of CDO design on CoP progression and perceived smoothness during walking in individuals with limb trauma are limited, as are studies comparing vCoP characteristics with and without CDOs. This study examined how two different CDOs provided to civilians after limb trauma are perceived and how they affect vCoP as compared to no CDO, and each other. Our results show that the design characteristics of the MONO and MOD CDOs studied here have limited effect on CoP progression during walking as compared to no CDO, and that the rate of CoP progression with the two study CDOs were more gradual than reported in prior studies with devices used in specialized care centers.

Despite the significant difference in stiffness and other apparent design differences, perceived smoothness ratings did not differ between the MONO and MOD CDOs. However, smoothness scores with the two study CDOs were almost 40% greater than those reported previously with CDO used in specialized care centers (Ikeda et al., 2018). The improved perceived smoothness may be related to the less abrupt CoP forward progression, as evident from lower peak vCoP magnitude values with both the MONO and MOD CDOs. Lower peak vCOP magnitudes are considered desirable as they represent a more gradual transition of CoP from the heel to toe. The mean peak vCoP magnitudes for the study CDOs were over 20% lower than those previously reported with CDOs used in specialized care centers (Ikeda et al., 2018; Russell Esposito et al., 2021). The mean peak vCoP magnitude values were less than or equal to 0.9 m/s for both the MONO and MOD devices, and not significantly different than no CDO. Multiple factors could contribute to the higher vCOP magnitude found with CDO used in specialized care centers (>1.17 m/s), including a design tailored for performance of high-impact activities such as load carriage or running (Owens et al., 2011; Schmidtbauer et al., 2019). Even though the two study CDOs had a much lower vCoP than the CDO used in specialized care centers, and resembled the no CDO condition, efforts to further reduce vCoP magnitude and improve perceived smoothness in civilian devices may prove beneficial. Evaluation of factors influencing perceived smoothness is important, as perceived smoothness scores may be associated with device preference, compliance, and satisfaction (Russell Esposito et al., 2021; Wanamaker et al., 2017).

The time to peak vCoP differed between the two CDO designs, with the time to peak vCoP with the MOD nearly double that of the MONO and no CDO conditions, which were not significantly different. These findings, when combined with no difference in perceived smoothness between devices, suggest that the timing of vCoP may not be linked to perceived smoothness but instead to device stiffness, and requires further investigation. Additionally, there were apparent differences in the timeseries data between the three study conditions (Fig. 2). MOD maintained the higher vCoP for a longer duration, as compared to the MONO and no CDO conditions. A less abrupt heel to forefoot transition with the study CDOs may decrease the importance of the specific timing of the peak vCoP and contributed to the more variable time to peak vCOP in the MOD condition. The mean time to peak vCoP was, however, similar to the time to ZMC with the MOD CDO, which most closely resembles CDOs used in specialized care centers.

Fig. 2.

Plot of the mean velocity of the center of pressure (vCoP) for study participants (n = 16).

NoCDO = no orthosis, MOD = modular, and MONO = monolithic.

The time to ZMC, the point at which the ankle transitions from a dorsiflexor moment to a plantarflexor moment, did not differ between no CDO, MONO and MOD conditions. The measure, which is thought to represent the CoP transition from heel to the forefoot of the CDO, has been previously linked to device preference and perceived smoothness (Ikeda et al., 2018). CDO configurations that restore the timing of forward progression of CoP under the foot are generally preferred (Russell Esposito et al., 2021). The close approximation of ankle ZMC with the two study CDOs may have contributed to the high perceived smoothness, consistent with a prior study (Ikeda et al., 2018). Further, consistently low vCOP values through the early stance (Fig. 2), and disassociation between time to peak vCOP and ZMC in the MONO condition suggests a smooth CoP transition from heel to the forefoot. Although the mean ZMC values across the three study conditions were within 2% of stance of each other, the standard deviation for each condition was >5%, opening the potential for further refinement of device configuration and more detailed study of how altering ZMC in each device influences patient perception. Alignment has been identified as an important factor influencing limb mechanics (Grunst et al., 2023) including ZMC, whether through direct changes in device alignment or by altering limb alignment by manipulating wedging under the heel of the CDO (Brown et al., 2017; Grunst et al., 2023; Ikeda et al., 2018).

The study CDOs represent two types of CDOs which are widely used in civilian clinical practice and have multiple design differences including cuff, strut, and footplate design, and stiffness. Despite these differences, the alignment did not differ between MONO and MOD. The CDO casting and fabrication processes were standardized, however, clinical providers were able to modify the device alignment using heel wedges or alignment shims to meet the patient’s needs. Even with this freedom to modify the CDO for each participant, both CDO alignments differed by less than one degree on average and were similar to the preferred alignment reported in previous work, received high ratings for smoothness, and resulted in similarly low vCoP peak values (Brown et al., 2017). The similarity in outcomes between the study CDOs and CDOs from prior work indicates the importance of alignment in determining the optimal CDO configuration. Despite the importance of alignment, alignment data are rarely provided in AFO related publications, limiting the ability to compare to other studies (Owen et al., 2018; Ridgewell et al., 2010). Despite the substantive differences between the study CDOs and previously studied CDOs, the two study CDOs resulted in high levels of perceived smoothness, low vCOP and ZMC values that approximate those of gait without a CDO.

4.1. Limitations

Study limitations include a modest sample size and the multiple design differences between study CDOs. While the MONO and MOD CDOs represent CDOs designs used in civilian clinical practice, the design differences make it difficult to attribute differences in dependent measures to specific design characteristics. The similarities in CoP characteristics between the two devices may be due to the interaction of multiple design characteristics, such as strut stiffness, footplate geometry, and cuff design to achieve a similar clinical goal which is restored limb function, and likely perceived smoothness. Further, while MONO and MOD CDOs represent CDOs designs used in civilian clinical practice, the findings are not likely generalizable to all other civilian CDOs. Additionally, we did not directly compare the two study CDOs with CDOs used at specialized care centers. However, the methods used here are identical to previously published studies and thus facilitate potential comparison to previous literature.

5. Conclusion

Despite multiple design differences, patient perception of smoothness did not differ between MONO and MOD CDOs, but they were perceived as smoother than previously studied CDOs (Ikeda et al., 2018). The greater perceived smoothness for both CDOs were observed with similar alignments, low peak vCoP magnitudes, and ankle ZMC values which resembled gait without the CDO. The timing of CoP progression under the foot, as assessed by time to peak vCoP, was greater for the stiffer MOD CDO, as compared to the more compliant MONO CDO and no CDO conditions. The interrelated metrics of patient perception and CoP characteristics provide a benchmark for future CDO designs and efforts to further improve patient perception and vCoP parameters following lower limb trauma.

Highlights.

• Orthosis design can influence forces on the limb during gait.

• Effect of commercial orthoses on center of pressure progression is not understood.

• Study orthoses were perceived as being smoother than reported in prior literature.

• Magnitude of peak center of pressure velocity was also lower with study orthoses.