The effects of slope-adaptive prosthetic feet on sloped gait performance and quality in unilateral transtibial prosthesis users: A scoping review

Abstract

Introduction:

In non-impaired human locomotion, sagittal-plane slope adaptation of the foot-ankle complex is a volitional function driven by neuromotor control to support upright posture and forward ambulation. Loss of this adaptation due to transtibial amputation can lead to instability and compensatory motions as most commercially-available prosthetic feet do not permit automatic slope adjustments. A selection of slope-adaptive feet (SAF) have been developed to promote biomimetic ankle motion while ambulating over slopes. This review evaluated the current literature to assess the effects of SAF prostheses on sloped gait performance in unilateral transtibial prosthesis users.

Methods:

Four databases (PubMed, Embase, CINAHL, IEEE Xplore) were searched on April 28, 2022, for relevant articles. Search keywords covered the general terms “transtibial,” “amputation,” “slope,” “adaptive,” and “gait”, and included articles comparing a SAF prosthesis to a non-SAF prosthesis condition. Data were extracted for analysis and results were grouped according to outcomes to identify trends and aid interpretation of slope adaptation effects on gait.

Results:

Of the 672 articles screened, 24 met the selection criteria and were included in this review, published between 2009 and 2022. The non-SAF condition included dynamic response feet and SAF prostheses with the adaptability function inactive. Outcomes included biomechanical variables (joint dynamics, gait symmetry, toe clearance), clinical outcome measures, and energy expenditure. All SAF demonstrated some form of foot-ankle slope gradient adaptability, but effects on other joint dynamics were inconsistent. Minimum toe clearance during incline and decline walking was greater when using SAF compared to non-SAF in all reporting studies.

Conclusions:

Results generally suggest improvements in gait quality, comfort, and safety with use of SAF compared to non-SAF during slope walking. However, variations in tested SAF and walking gradients across studies highlight the need for research to elucidate walking condition effects and advantages of specific designs.

Clinical Relevance:

Slope-adaptive prosthetic feet may improve user gait quality and comfort and enhance gait safety by increasing minimum toe clearance. Patients who encounter slopes regularly should be considered as potential users of SAF if indicated appropriately.

INTRODUCTION

In human locomotion, sagittal plane slope adaptation is used to maintain upright posture while facilitating forward and upward/downward ambulation and is achieved through alterations in joint mechanics driven by neuromotor function.^{1, 2} Much of this adaptability occurs at the ankle joint through two primary mechanisms: 1) the ankle joint adjusts its angle at initial contact and during the loading phase of stance according to the encountered slope gradient, i.e. ground compliance (Figure 1),^{1, 3, 4} and 2) management of ankle joint moments through regulation of joint rotational impedance via muscle contraction to provide adequate stability through the desired range-of-motion (RoM).⁴ While ankle impedance during stance can be measured and defined in different ways,^{5, 6} it is often characterized through a joint moment-angle curve (Figure 2) that has been used as a platform to design biomimetic function in prosthetic ankle-foot mechanisms.⁶ To that end, when the biological ankle is lost due to transtibial amputation, that individual loses a crucial component of the volitional control necessary to adapt to sloped walking. The commercially-available prosthetic feet that are most commonly prescribed do not permit automatic sagittal plane adjustment of the ankle angle and/or impedance while walking on slopes as they are passive non-articulated, solid designs or include ankle articulations with fixed rotational resistance that return to a neutral orientation when unloaded.⁷ A lack of controlled dorsiflexion range has been suggested as the cause of instability in prosthesis users when walking up slopes.^{4, 8} These limitations of commercial prosthetic feet can lead to compensatory gait patterns when walking on slopes, including increased knee flexion of the sound side⁹ and residual limb¹⁰ in early stance, decreased or absent hip extension of the sound side halfway through the gait cycle,⁸ increased prosthetic side peak hip power,¹⁰ and decreased walking speed.^{9, 10}

Figure 1.

(A) Ankle angle of a single healthy participant during sagittal plane sloped walking from initial foot contact to ipsilateral foot contact. (B) Average ankle angles at foot contact on different slope gradients. Adapted from Leroux et al., 2002, Fig. 2.

Figure 2.

Moment-angle curves of the physiologic human ankle at 1.9 m/s (left), 1.5 m/s (center), and 1.2 m/s (right). Arrows indicate direction of gait progression. Adopted from Hansen et al., 2004, Fig. 3.

To address the limitations and effects of common non-adaptive prosthetic feet, slope-adaptive feet (SAF) have recently been developed and introduced into clinical practice. In this context, slope-adaptive refers to the intrinsic ability of a prosthetic foot to adjust its neutral sagittal plane angle and/or rotational impedance relative to the socket when traversing surface gradients. Achieving this goal can be accomplished in multiple ways and through different designs. Microprocessor controlled hydraulic actuators are a common method used by commercially available SAF, such as: the Elan foot (Figure 3A; Blatchford, Basingstoke, UK), the Meridium foot (Figure 3B; Ottobock, Duderstadt, Germany), the Kinnex (Freedom Innovations, Irvine, CA, USA), and the Raize (Fillauer, Chattanooga, TN, USA). The Proprio Foot (Össur, Reykjavik, Iceland) is another microprocessor foot that utilizes an electric motor to adjust the ankle position (Figure 3C). The Proprio Foot has been included in multiple studies to assess its effects on prosthesis users’ gait during sloped walking. It has been found to produce more physiologic kinematics and kinetics during ramp ascent compared to the same prosthetic foot with the adaptable mode disabled.¹¹ The Elan has also been studied during sloped walking and was shown to produce a more stable gait pattern than its comparison device.¹² In addition to commercial SAF prostheses, Nickel et al., 2014, assessed a fully mechanical prototype design that automatically adjusts to slopes with every step, as tested by a person with a unilateral transtibial amputation, reporting that the participant felt as if he needed less energy to walk on the slope with the adaptable prosthesis.¹³

Figure 3.

Example of current commercially available SAF: (A) Elan (Blatchford), (B) Meridium (Ottobock), and (C) Proprio Foot (Össur). Images adopted from blatchfordmobility.com (A), ottobock.com (B), and ossur.com (C).

The aim of this study was to review the current literature to answer the following research question: what is the evidence characterizing the effects of SAF on gait performance and quality when compared to non-slope-adaptive prosthetic feet (non-SAF) in unilateral transtibial prosthesis users? A scoping review was performed because of the potential heterogeneity in study designs and tested prosthetic componentry (commercial or experimental), that a comprehensive review on this literature has not yet been performed, and there is a need to assess for emerging clinical evidence given that commercially-available SAF are currently prescribed. As more SAF become available with further advances in technology and current designs, a summary of evidence on the effects of these prostheses can help guide clinical decision-making to determine if and when SAF would best be indicated for specific patients to address limitations of non-SAF prostheses.

METHODS

Search Strategy

The following four databases were searched for articles on April 28, 2022: PubMed, Embase, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Institute of Electrical and Electronics Engineers (IEEE) Xplore. These databases were selected to identify articles concerned with the multidisciplinary nature of the research question, which includes concepts related to medicine, allied health, prosthetics, and engineering. The search strings for each database were constructed from combining terms related to the population (persons with unilateral transtibial amputation), intervention (SAF prosthesis), and outcome (construct of gait performance) of interest. The PubMed search string was developed first (Table 1) and then adapted to the other databases (see Supplement 1).

Table 1.

Search terms with an example search string for PubMed included in the bottom row. A term for the non-SAF comparison is not included given the unique but limited designs of SAF prostheses.

	Population: Transtibial prosthesis users			Intervention: Slope-adaptive ankles	Outcome: Gait
Search Terms	Lower limb*	Artificial limb*	Amputat*	Adaptive	Kinematic*
	Lower extremit*	Artificial device*	Amputee*	Adaptation*	Kinetic*
	Ankle*	Prosthes*	Residual limb*	Adaptab*	Mobility
	Foot	Prosthetic*		Adjustab*	Gait
	Feet			Slope*	Symmetry
	Below the knee			Incline*	Walking
	Below knee			Ramp*	Ambulation
	Transtibial
	Trans-tibial
Example: PubMed Search String	((“artificial limbs”[MeSH] OR “artificial limbs” OR “artificial limb” OR “artificial device” OR “artificial devices” OR Prosthes* OR Prosthetic*) AND (“lower limb” OR “lower limbs” OR “lower extremity” OR “lower extremity”[ MeSH] OR “lower extremities” OR Ankle* OR Foot OR Feet OR “below the knee” OR “below knee” OR Transtibial OR Trans-tibial) AND (Amputat* OR “Amputation Stumps”[ MeSH] OR “amputation”[ MeSH] OR “Amputation, Traumatic”[ MeSH] OR “Amputees”[Mesh] OR Amputee* OR Residual-limb*)) AND (adaptive OR adaptation* OR adaptab* OR Adjustab* OR Slope* OR Incline* OR Ramp*) AND (kinematic* OR Kinetic* OR Mobility OR Gait OR Symmetry OR Walking OR Ambulation OR Gait[MeSH])

The asterisk (*) denotes a truncated term and “MeSH” denotes a Medical Subject Heading term.

Selection Criteria

Retrieved articles were imported into a central database (EndNote v20.0.1, Philadelphia, PA, USA). Article screening and filtering were performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) extension for Scoping Reviews.¹⁴ Duplicates were first removed, followed by screening titles and abstracts to retain only articles relevant to the topic (i.e., assessment of a SAF prosthesis), then reviewing the full text against inclusion and exclusion criteria. The inclusion criteria included: (1) use of a quantitative outcome measure to assess an aspect of in-vivo gait performance during sloped walking; (2) participants were only unilateral transtibial prosthesis users; and (3) gait performance was compared between a SAF and non-SAF prosthesis, which could include non-articulated ankles, articulated ankles with no slope adaptability, or SAF with the adaptability feature inactive. Exclusion criteria included: (1) participants were pediatric (younger than 18 years); (2) participants had reported comorbidities such as neuromuscular or musculoskeletal impairments known to affect gait; and (3) study authors did not explicitly note or test the slope-adaptable feature of the prostheses under investigation. Once the final set of articles were selected for inclusion in this review, their reference lists were reviewed to identify additional articles of relevance that were then screened and reviewed through the same process. Limits included only articles in English, studies involving human subject testing, and full-text availability (excluding only abstracts or conference proceedings).

Data Extraction

Data from each article were extracted and inserted into a spreadsheet (Excel v1206, Microsoft, Redmond, WA, USA) to characterize the body of reviewed literature and to address the research question. These data included: study authors; year of publication; study design; participant cohort(s) sample size(s); participant mean age, sex distribution, and Medicare Functional Classification (K-)level if available; test setting (e.g., clinic, laboratory); slope gradient for testing gait; SAF prosthesis tested (make and model if commercial; functional description if experimental); non-SAF comparison prosthesis tested; outcome measures; and study results.

RESULTS

The search yielded 1,032 articles total (PubMed, n=368; Embase, n=433; IEEE Explore, n=128; CINAHL, n=103), with the screening and selection results displayed in Figure 4. Following the removal of duplicates, the title and abstract of 630 articles were screened, leaving 53 articles for full text screening. After applying the selection criteria, 21 articles remained, of which the references were reviewed and yielded an additional 42 articles that appeared applicable and were subsequently reviewed. Ultimately, a total of 24 articles were included in this review.^{11–13, 15–35} The included articles were published between 2009 and 2022 (Figure 5). The range of participants was 1 to 42. The mean participant age of reporting studies (n=20)^{11–13, 15–17, 19–24, 26–28, 30–34} ranged from 29.0 to 59.7 years. For those studies reporting participant sex (n=20),^{11–13, 15–17, 19–21, 23–28, 30–34} the percentage of male participants ranged from 75% to 100%. Although only reported in 15 (63%) studies,^{11, 12, 15, 17, 21–28, 32, 33, 35} the K-level of participants was reported as K3 or K4 in all but two articles: Agrawal et al., 2015, included five K2 participants,¹⁵ and Kaluf et al., 2021, included a single K2 participant.²⁴ A table reporting individual study characteristics is available in the supplementary material (Supplement 2).

Figure 4.

PRISMA chart displaying the screening and suitability evaluation process.

Figure 5.

Frequency of published articles per year.

All of the studies reported results of testing in a laboratory apart from one.³⁰ Notably, two articles that collected gait data from laboratory testing also collected self-report information based on the subjects’ experiences with the experimental SAF during a home trial.^{23, 24} Across all tested SAF, the commercially-available Proprio Foot (Össur) was the most commonly investigated prosthesis with 9 articles,^{11, 15, 19–21, 25, 27, 32, 35} in which a microprocessor-controlled motor adapts the ankle angle to changes in terrain. Regarding type, microprocessor-controlled hydraulic feet that adapt ankle impedance were the most investigated type of SAF, specifically: Elan (Blatchford),^{12, 20, 25, 28, 31} Meridium (Ottobock),^{20, 22, 33} Kinnex (Freedom Innovations),^{23, 24} and a prototype foot described by Bartlett et al., 2021.¹⁸ Three passive (impedance adaptation not controlled through a microprocessor) hydraulic devices were also investigated: Echelon (Blatchford, Basingstoke, UK),^{16, 25, 26, 30, 31} Odyssey (College Park, Warren, MI, USA),¹⁷ and Kinterra (Freedom Innovations, Irvine, CA, USA).²⁴ Finally, the same mechanical SAF (not commercially available) that adapts ankle angle setpoint was investigated in 3 separate articles.^{13, 29, 34} The main type of non-SAF comparison device tested was an energy storage and return (ESAR) foot (n=14).^{13, 16–18, 20–26, 30, 31, 33}, while six studies tested the SAF with its slope adaptability feature disabled as a comparison.^{11, 12, 19, 27, 28, 35}

The categories of outcome measures and frequency of use among the reviewed studies are summarized in Figure 6. Note that only outcome measures related to sloped gait are included in this analysis given the purpose of this review; outcome measures unrelated to gait or concerned solely with level ground walking, stair ambulation, or transfers were omitted. The most commonly used outcome variables were spatiotemporal measures of gait (e.g. plantar center of pressure (CoP) forward progression, gait speed)^{12, 15, 18, 24–28, 31–33} and prosthetic ankle ‘joint’ dynamics (angle range-of-motion, moment, power) during walking,^{11, 13, 18, 20, 22, 24, 25, 27, 29, 33, 34} with 11 articles for each category. Joint dynamics were also measured for the impaired limb anatomical joints in 8 articles,^{11, 12, 22, 24, 27, 28, 33, 34} while 3 of these 10 also reported contralateral side joint dynamics.^{11, 28, 33} Patient-reported outcome measures (PROMs) were used in 7 articles, capturing perceived exertion, balance confidence, mobility capability, socket comfort, and prosthesis use.^{13, 17, 19, 21, 23, 24, 26} Bilateral limb symmetry was estimated in 4 articles, where it was calculated as symmetry in external (mechanical) work,^{15, 25} symmetry of joint moments,²⁸ or symmetry of body center of mass excursion during prosthetic and sound side steps.¹⁶ Three articles measured energy expenditure calculated as metabolic cost of transport.^{16, 19, 21}

Figure 6.

Frequency of outcome categories used to measure gait performance and quality across the articles included in this review.

Additional outcome measures included a performance-based outcome measure (PBOM) quantifying slope mobility^{21, 24} and peak/rate of residuum-socket interface pressure.^{30, 35} Table 2 describes the specific outcome measures of each outcome category and the corresponding results from each study suggesting how the SAF performed relative to the non-SAF comparison.

Table 2.

Outcome measures and corresponding results from each study categorized by how the SAF performed relative to the non-SAF comparison. To note, no study reported the non-SAF outperforming the SAF. Some studies are listed in both columns describing effects if they report that the SAF outperformed the non-SAF and was also comparable for certain test conditions or outcomes.

Outcome Category	Outcome Measure / Variable	Articles sorted by SAF versus non-SAF result
		SAF outperformed Non-SAF	SAF comparable to Non-SAF
Joint Dynamics	Prosthetic ankle ‘joint’ ROM (degrees)	Fradet et al. 201011, Nickel et al. 201413, Bartlett et al. 202118, Davot et al. 202120, Ernst et al. 202222, Kaluf et al. 202124, Ko et al 201425, Lamers et al. 201927, Nickel et al. 201229, Schmalz et al. 201933, Williams et al. 200934	--
	Prosthetic ankle ‘joint’ peak moment and power (Nm/kg)	Nickel et al. 201413, Davot et al. 202120, Ernst et al. 202222, Nickel et al. 201229, Schmalz et al. 201933	--
	Anatomical joint ROMs (degrees)	Fradet et al. 201011, Struchkov and Buckley 201612, Ernst et al. 202222, Kaluf et al. 202124	Fradet et al. 201011, Lamers et al. 201927, Schmalz et al. 201933, Williams et al. 200934
	Anatomical joint peak moments and powers (Nm/kg)	Fradet et al. 201011, Struchkov and Buckley 201612, Ernst et al. 202222, McGrath et al. 201828, Schmalz et al. 201933	Fradet et al. 201011, Lamers et al. 201927
Bilateral Limb Symmetry	Individual limb work symmetry (J/J, expressed as %)	Agrawal et al. 201515, Ko et al. 201425	Agrawal et al. 201515, Ko et al. 201425
	Degree of asymmetry of net support moment (Nm/Nm)	McGrath et al. 201828	McGrath et al. 201828
	Body center-of-mass trajectory symmetry (computed from ratio of modeled Fourier coefficients)	--	Askew et al. 201916
Spatiotemporal Parameters	Minimum toe clearance (mm)	Bartlett et al. 202118, Lamers et al. 201927, Riveras et al. 202031, Rosenblatt et al. 201432	--
	Walking speed (m/s)	McGrath et al. 201828	Kaluf et al. 202124, Schmalz et al. 201933
	Stance time (%gait cycle)	Ko et al. 201425	Ko et al. 201425
	Single-limb support shank to vertical angle (degrees)	--	Koehler-McNicholas et al. 201726
	Time to foot flat (s)	Bartlett et al. 202118	Struchkov and Buckley 201612
	Plantar CoP forward progression (%foot length)	--	Agrawal et al. 201515, Lamers et al. 201927
Residuum-Socket Interface Pressure	Peak pressure (kPa/kg)	Wolf et al. 200935	Wolf et al. 200935
	Von Mises stress (kPa)	Portnoy et al. 201230	--
Patient Reported Outcome Measures	Rating of perceived exertion (modified Borg)	Nickel et al. 201413, Darter and Wilken 201419	Nickel et al. 201413, Darter and Wilken 201419, Koehler-McNicholas et al. 201726
	5-point Likert scale rating sloped walking difficulty	Atar et al. 202217	Atar et al. 202217
	ABC	--	Kaluf et al. 202023
	PEQ-MS	Kaluf et al. 202023	--
	PLUS-M	--	Kaluf et al. 202023
	LCI-5	--	Delussu et al. 201321
	SCS	Nickel et al. 201413, Kaluf et al. 202023, Kaluf et al. 202124	McNicholas et al. 201726
	Houghton scale	--	Delussu et al. 201321
Performance Based Outcome Measures	Hill Assessment Index	--	Delussu et al. 201321, Kaluf et al. 202124
Energy Expenditure	Metabolic cost (J or ml O2/kg/min)	Askew et al. 201916	Darter and Wilken 201419, Delussu et al. 201321

ROM: Range of motion. CoP: Center of pressure. ABC: Activities Specific Balance Confidence Scale. PEQ-MS: Prosthesis Evaluation Questionnaire–Mobility Subscale. PLUS-M: Prosthetic Limb User Survey of Mobility. LCI-5: Locomotor Capability Index-5. SCS: Socket Comfort Score.

DISCUSSION

The aim of this study was to review the current literature to assess the effects of SAF prostheses on gait performance and quality when compared to non-SAF in unilateral transtibial prosthesis users. The 24 studies included in this review cover 9 different SAF, each with distinct designs and control systems to achieve the goal of adaptability, across a range of outcome measures spanning biomechanical variables to perceived mobility function. As the purpose of this scoping review was to generally assess the effect of current SAF technology on sloped gait, the results were grouped into outcome category for synthesis while highlighting results attributable only to unique SAF prostheses. Notably, none of the results across the reviewed articles suggested that the non-SAF comparison outperformed the SAF for sloped ambulation (Table 2), and therefore these results describe only where the SAF outperformed the non-SAF or their performance was comparable.

Biomechanics

Joint dynamics were commonly quantified across the reviewed studies,^{11–13, 18, 20, 22, 24, 25, 27–29, 33, 34} primarily to verify the kinematic slope adaptability function for both microprocessor-controlled^{11, 18, 20, 22, 24, 25, 27, 33} and mechanical designs^{13, 29, 34} through comparison of the measured ankle angle to the slope gradient. For all these studies the ability of the SAF ankle angle to match the ground surface angle during the loading phase (i.e., ground compliance) was confirmed. While such kinematic adaptation would theoretically set the shank in a more favorable position for loading to support transition over the stance limb and hence upright posture,² the reported effects of this adaptability on more proximal and contralateral joints were mixed.

While some studies observed more biomimetic (i.e., closer to able-bodied) joint angles and moments for the anatomical joints when using a SAF compared to a non-SAF,^{11, 12, 22, 24, 28, 33} others showed no differences between the two device types.^{11, 27, 34} The Proprio Foot with its adaptive function active was found to reduce residual peak knee flexion moments at loading response and increase hip flexion at heel strike during incline walking, which were closer to able-bodied dynamics, but demonstrated no differences during decline walking, suggesting a gradient-dependent effect.¹¹ Another study investigating the Proprio Foot found no differences between the inactive and active SAF, though data was only gathered on incline walking.²⁷ Generally, microprocessor-controlled hydraulic SAF were reported to reduce residual peak knee flexion throughout loading response when walking on declines compared to non-SAF, which more closely matched control participant dynamics.^{12, 24, 33} Though some studies did not include incline data,^{12, 33} the single study that did include ascent walking found more biomimetic residual limb knee flexion for both the decline and incline walking.²⁴ Moreover, one study demonstrated a decrease in residual knee and SAF ankle moments during stance while ascending and descending slopes,²² and there was evidence of a reduction in the sound limb support moment integral (sum of integration of the joint moments with respect to time across stance),²⁸ to suggest reduced limb loading for ascent and descent. While the non-commercial mechanical design was confirmed to demonstrate slope adaptability with more biomimetic prosthetic ankle dynamics,^{13, 29, 34} there were no significant differences in kinematics and kinetics for any anatomical joint on inclines or declines when compared to a non-SAF.³⁴ Overall, while all SAF could demonstrate improved ground compliance and there appeared to be a trend toward reduced joint loading with SAF, that effect may be subject to the interaction between the type of device (and associated methods of angle and impedance accommodation) and the slope gradient direction of walking (ascent or descent). Reductions in joint loading may have implications for long-term musculoskeletal health that include reduced risk of secondary injury and development of osteoarthritis^36–39 and so the trending evidence toward benefits of SAF in this regard may be clinically important and warrants further study.

Given the potential for SAF to affect contralateral limb dynamics and the implications of reduced sound limb reliance to protect against secondary injuries in the long-term,^38–40 several studies focused on observations of bilateral limb symmetry. Three studies investigated kinetic (i.e., individual limb work^{15, 25} and net support moment²⁸ to support and move the body center of mass) symmetry with one study observing symmetry in body center-of-mass trajectory¹⁶ during incline and/or decline walking. Similar to the observations of individual limb dynamics, these results were somewhat mixed. While one study (n=1) observed increased kinetic symmetry with a SAF compared to ESAR for inclines and declines,²⁵ others observed increased symmetry on decline but not incline walking (comparing SAF to a SACH¹⁵ and ESAR^{15, 28} foot), or found no significant difference on declines but trended towards improved symmetry with the SAF compared to ESAR.¹⁶ Generally, there appeared to be trending evidence to suggest that SAF may generate increased kinetic symmetry during sloped walking, but this effect may again be dependent on the gradient direction, while acknowledging here the limited observations on incline walking.

Regarding gait safety, minimum toe clearance (MTC) is a relevant clinical outcome given its relationship to fall risk, whereas increased toe clearance would minimize potential for colliding the prosthetic limb with the ground thereby reducing the potential for stumbles.^41–44 As related to SAF, the ability of these devices to comply with the walking gradient and maintain or actively produce dorsiflexion when entering swing would theoretically minimize potential for toe catch. Compared to non-SAF, MTC was observed to be greater for the Proprio Foot,^{27, 32} microprocessor-controlled hydraulic designs,^{18, 31} and passive hydraulic designs.³¹ Apart from the Proprio Foot which was only tested on inclines, the increase in MTC for SAF was observed during both incline and decline walking to confirm the functionality of this unique feature of SAF and its potential benefit to gait safety. Notably, MTC was not quantified for the experimental mechanical designs^{13, 29, 34} and so it is not possible to extend these findings to those devices. Overall, there was consensus evidence across the reviewed studies to suggest that commercially-available SAF can increase MTC when ascending and descending slopes across different SAF designs.

While there was converging evidence of MTC benefits with SAF, results on other spatiotemporal variables were not often reported and less convincing. Both average walking speed^{24, 33} and peak shank-to-vertical angle during stance²⁶ were no different between the SAF and its non-SAF comparison in three independent studies. One study did find increased walking speed with the SAF, but only on ascent walking.²⁸ Prosthetic limb single support time was found to increase during decline walking for various SAF designs compared to an ESAR foot in an n of 1 study, but demonstrated no difference during incline walking.²⁵ Given ability for ground compliance, time to foot flat is also a relevant variable as early foot flat can provide a stable base of support during the loading phase and support body center of mass advancement over the prosthetic limb.^45–47 While one study reported that the SAF condition produced time to foot flat that was closer to able-bodied timing for both inclines and declines,¹⁸ another study observed no differences during decline walking (incline walking was not tested).¹² Given the limited and varied reporting on spatiotemporal variables, this review is unable to infer much from these results.

Patient-Reported and Performance-Based Outcome Measures

Though clinical outcome measures have not been utilized as frequently as biomechanical measures, they provide unique clinical insight given their accessibility to clinicians and applicability to clinical settings. Moreover, patient-reported outcome measures can allow clinicians to appreciate user perceptions that are not easily captured through biomechanical/performance-based variables but could impact rehabilitation outcomes. Socket comfort as measured with the Socket Comfort Score (SCS) was reported to be higher (more comfortable) for all tested categories of SAF (microprocessor-controlled hydraulic,^{23, 24} passive hydraulic,²⁴ and mechanical¹³) on both incline and decline walking than the non-SAF comparison. Only one study observed no differences in SCS between a passive hydraulic SAF and ESAR and multiaxial feet during incline and decline walking.²⁶ The increased socket comfort with a SAF may be reflective of the ability of the device to comply with the walking surface gradient and align the proximal anatomy in a more favorable position to potentially limit localized pressures on the residuum. In fact, one study observed lower peak pressure at the posterior midpoint of the residuum-socket interface and a smaller pressure time integral during incline walking with the SAF compared to non-SAF (Proprio Foot with and without adaptability function).³⁵ Another study observed lower peak stresses and loading rates applied to the cut end of the residuum tibia when walking on inclines and declines with a SAF (Echelon) compared to an ESAR foot.³⁰ Importantly, pressure on the residual limb can not only affect comfort but is also related to soft tissue trauma.⁴⁸ Therefore, the converging evidence on socket comfort and residuum-applied loads would suggest benefits of SAF to user-perceived comfort and a need to further study its influence on residuum health. Importantly, effects on comfort and interface pressure would also be dependent on the socket fit and suspension system.

A modified Borg Rating of Perceived Exertion (RPE) scale was used in three separate studies to assess the participants’ perceived level of exertion when walking on slopes, with varying results. While one study observed a lower RPE for a mechanical SAF compared to an ESAR when walking on declines but not inclines,¹³ another comparing a microprocessor-controlled SAF (Proprio Foot) to the same foot with the adaptability disabled observed lower RPE for the SAF condition on incline walking but not decline walking.¹⁹ A third study found no differences in RPE between a passive hydraulic SAF (Echelon) and ESAR and multiaxial feet for either incline or decline walking.²⁶ Finally, one study that used a custom Likert scale to assess level of perceived difficulty ascending and descending slopes observed lower difficulty on declines when walking with a passive hydraulic SAF compared to an ESAR, but no differences were observed on inclines.¹⁷ Similar to other results, there appears to be a trend toward lower rating of exertion and difficulty in ascending and descending slopes with SAF, but this effect is likely dependent on an interaction between the slope gradient direction and SAF design, including factors such as device weight. However, notably, when considered with potential for increased perceived comfort, a trend toward lower perceived exertion could encourage use of SAF.

Other patient-reported outcome measures were used in one study and included the Prosthesis Evaluation Questionnaire–Mobility Subscale (PEQ-MS), Activities Specific Balance Confidence Scale (ABC), and Prosthetic Limb User Survey of Mobility (PLUS-M) to evaluate ambulation capability.²³ However, only results from the PEQ-MS were significant, with improved mobility capability using the SAF (Kinnex) compared to an ESAR foot.²³

The only performance-based outcome measure used across the reviewed studies was the Hill Assessment Index, an ordinal measure quantifying ramp ambulation ability and performance, and two studies reported no differences between the SAF and the comparison device.^{21, 24} As there was limited used of performance-based outcome measures to assess slope walking, it is unclear how SAF ambulation would differ from non-SAF on more general and perhaps global measures of gait and mobility.

Energy Expenditure

A commonly observed outcome in prosthetic research is energy expenditure given its implications to endurance and perceived exertion, although its ecological validity is under question.⁴⁹ Energy expenditure comparing the SAF to non-SAF conditions was observed in three separate studies. Two studies on the Proprio Foot reported no significant difference between conditions for incline walking.^{19, 21} While one of these studies did report a lower metabolic cost during decline walking compared to the participants’ daily-use ESAR foot, the difference was not found to be significant.¹⁹ The other study did observe reduced energy expenditure when walking with a passive hydraulic SAF (Echelon) compared to an ESAR foot on declines (inclines were not tested).¹⁶ While there may be trending evidence to suggest reduced energy expenditure with SAF on declines that aligns with the findings on perceived exertion, the available evidence is minimal and warrants further study.

Limitations

There are limitations to this scoping review that should be considered when interpreting the results. Regarding the methodology, only a single reviewer retrieved and evaluated the articles for inclusion, though concerns regarding eligibility of articles for inclusion were discussed by both authors who also contributed to synthesis and interpretation of results. Further, the review scope was limited to unilateral transtibial prosthesis users, thereby restricting the generalizability of the findings to summarize the current state of the literature on this topic. Finally, this scoping review focused on measured gait performance and quality, excluding outcomes related to qualitative measures or other common features of daily ambulation (stairs, transfers, standing balance).

Regarding limitations of the body of literature examined in this review, the included studies had on average small sample sizes, with the single largest sample size being 42 but limited to collecting only patient-reported information.¹⁷ While common in prosthetics research given the practicality of in-vivo studies, this limitation often results in underpowered studies, which may partially explain the lack of statistical significance for some comparisons that were deemed to be equivalent results. Conversely, results from some studies were classified as favorable to SAF although a statistical analysis was not performed due to the acknowledged small samples. Although a recognized challenge, future research should consider larger sample sizes perhaps through multisite studies to enhance confidence in the true effect of SAF on sloped gait. Additionally, participants in the reviewed studies were primarily young, active males, which further limited the generalizability and impact of the results, especially given the need for research on women with limb loss.^50–52 Given the suggested benefits of SAF to walking safety as noted in this review, including ground compliance and increased MTC, future work may consider more studies on less active prosthesis users with reduced mobility.

Importantly, there was no consistent, standardized terminology describing SAF prostheses and their function used by the clinical and research communities at the time of this review, which presented certain challenges. First, while the search terms attempted to cover known terminology associated with the intended function of SAF, there is a possibility that articles were not retrieved from the database search due to missing less commonly used keywords. Second, this issue may have also confounded results comparing SAF and non-SAF performance. While investigators of these studies tested a SAF prosthesis, the function by which they achieved slope adaptability (impedance, set angle) varied and would affect the generalized results and conclusions of this review. Studies on this topic should continue to include as much detail as possible regarding the slope adaptive function of SAF to assist with interpreting results. Moreover, differences in socket design and suspension would also present a confounding influence to gait outcomes and this was not often considered in these studies. Similar to the inconsistency in terminology and SAF design functionality, there is also little consistency in outcomes and testing conditions across studies, rendering synthesis challenging. While there was an assortment of outcomes assessed, these mainly focused on the lower extremity and did not include upper body (e.g., trunk) dynamics despite its relevance to compensatory mechanisms and balance.⁵³ In addition to the range of outcome measures that addressed similar constructs, even the tested slope gradient varied between 3° to 10° and included one or both testing ascent and descent. Discrepancies between study results might therefore be related to differences in outcomes and slope gradient, suggesting future research carefully consider these methodological choices.

CONCLUSION

Studies investigating SAF prostheses with the ability to adjust joint set angles and impedance have generally reported potential benefits of SAF for unilateral transtibial prosthesis users when ascending and descending slopes. These benefits may include lower residual knee peak moments, increased bilateral limb symmetry, and improved comfort that might lead to reduced risk of secondary conditions, as well increased toe clearance that could potentially reduce fall risk. However, while there are trends pointing to clinical benefits of SAF compared to non-SAF prostheses, the state of the literature is constrained by a limited number of studies, small sample sizes, and variations in test protocol, SAF designs, and outcome measures, thereby suggesting a need for future research and standardizations of clinical study methods.