A Scoping Review of Advances in Active Below-Knee Prosthetics: Integrating Biomechanical Design, Energy Efficiency, and Neuromuscular Adaptation

Abstract

Background: This scoping review systematically maps and synthesises contemporary literature on the biomechanics of active below-knee prosthetic devices, focusing on gait kinematics, kinetics, energy expenditure, and muscle activation. It further evaluates design advancements, including powered ankle–foot prostheses and variable impedance systems, that seek to emulate physiological ankle function and enhance mobility outcomes for transtibial amputees. Methods: This review followed the PRISMA-ScR guidelines. A comprehensive literature search was conducted on ScienceDirect, PubMed and IEEE Xplore for studies published between 2013 and 2023. Search terms were structured according to the Population, Intervention, Comparator, and Outcome (PICO) framework. From 971 identified articles, 27 peer-reviewed studies were found to meet the inclusion criteria between January 2013 and December 2023. Data were extracted on biomechanical parameters, prosthetic design characteristics, and participant demographics to identify prevailing trends and research gaps. This scoping review was registered with Research Registry under the following registration number: reviewregistry 2055. Results: The reviewed studies demonstrate that active below-knee prosthetic systems substantially improve gait symmetry and ankle joint range of motion compared with passive devices. However, compensatory trunk and pelvic movements persist, indicating that full restoration of natural gait mechanics remains incomplete. Metabolic efficiency varied considerably across studies, influenced by device design, control strategies, and user adaptation. Notably, the literature exhibits a pronounced gender imbalance, with only 10.7% female participants, and a reliance on controlled laboratory conditions, limiting ecological validity. Conclusions: Active prosthetic technologies represent a significant advancement in lower-limb rehabilitation. Nevertheless, complete biomechanical normalisation has yet to be achieved. Future research should focus on long-term, real-world evaluations using larger, more diverse cohorts and adaptive technologies such as variable impedance actuators and multi-level control systems to reduce asymmetrical loading and optimise gait efficiency.

1. Introduction

Walking is a highly coordinated activity that requires the integration of musculoskeletal, neural, and sensory subsystems to maintain stability and efficiency [1,2]. In individuals with unilateral transtibial amputation, the loss of the anatomical ankle–foot complex fundamentally alters this system, leading to asymmetrical gait patterns and compensatory adaptations [3,4,5,6,7].

These asymmetries typically present as increased stance duration and ground reaction forces on the sound limb, alongside reduced stance and propulsion on the prosthetic side [8,9,10]. The result is an imbalance in mechanical loading and energy distribution across the lower limbs, which, over time, contributes to secondary musculoskeletal disorders such as joint degeneration and chronic low back pain [2,11].

The gait asymmetries observed in this population can be explained biomechanically by analysing the centre of mass (COM) trajectory and limb dynamics during the stance phase. A shortened stance time on the prosthetic limb disrupts the deceleration of the COM, increasing the collisional impact at heel strike on the contralateral side [10,12]. This repetitive impact raises vertical ground reaction forces, thereby elevating joint stress and accelerating degenerative processes. Consequently, the sound limb becomes overburdened, predisposing amputees to early onset osteoarthritis in the hip and knee [13,14]. These pathomechanical changes are not only a function of prosthetic design but also of neuromuscular adaptation and user-specific compensatory control strategies.

The long-term implications of these gait deviations underscore the need for interventions that restore more symmetrical and energy-efficient walking patterns. Traditional passive prosthetic feet, although mechanically robust, provide limited capacity for energy storage and return. They cannot actively replicate the complex timing and magnitude of ankle plantarflexion required for push-off during the late stance phase. In contrast, active and semi-active prosthetic devices—equipped with actuators and microcontrollers, have been shown to improve joint kinematics, enhance propulsion, and reduce mechanical work on the intact limb [2,4,5]. These advancements mark a paradigm shift from energy-storing designs to bioinspired systems that aim to emulate physiological ankle behaviour.

Despite these technological improvements, complete biomechanical symmetry remains elusive. Studies have demonstrated that users of powered prosthetic feet often continue to rely on compensatory trunk and hip movements, leading to altered whole-body dynamics and residual asymmetries [6,7,15]. Such persistent compensations suggest that restoring symmetry involves more than mechanical replication of joint motion; it requires harmonising prosthetic control with neuromuscular function. Moreover, the energy cost of walking remains inconsistent across studies, highlighting the need for individualised prosthetic tuning and adaptive control strategies.

A deeper understanding of the neuromechanical interplay between prosthetic design and human motor control is therefore essential. Investigating how parameters such as mechanical power, joint moment, and electromyographic (EMG) activity contribute to gait performance provides valuable insight into the functional integration of prosthetic devices. Even when external gait metrics appear symmetrical, underlying neuromuscular compensations often persist, implying that users achieve a new “asymmetrical optimum” that balances stability and efficiency within their biomechanical constraints [16].

In light of these biomechanical complexities and variations in device performance, this scoping review specifically aims to systematically identify, categorise, and synthesise evidence on the effects of active below-knee prostheses across three measurable domains of gait performance in transtibial amputees:

• Mechanical outcomes—joint kinematics, joint kinetics, centre of mass behaviour, and mechanical power generation or absorption;
• Energetic outcomes—metabolic cost, mechanical work, and walking efficiency;
• Neuromuscular outcomes—electromyographic activity, muscle coordination patterns, and compensatory control strategies.

The review examines how prosthetic design parameters influence these outcomes and how these device characteristics interact with user-specific neuromuscular adaptations. The overarching objective is to map the relationships between device design choices and functional user responses, identify areas of convergence and divergence in the evidence, and highlight critical research gaps that must be addressed to optimise future development of lower-limb prosthetic technologies.

2. Materials and Methods

2.1. Study Design

This work employed a scoping review methodology to systematically map and synthesise current evidence on gait biomechanics and prosthetic function in individuals with unilateral transtibial amputation. The scoping review approach, as proposed by Arksey and O’Malley [17], was chosen to capture the breadth, depth, and heterogeneity of studies focusing on active below-knee prostheses. Unlike systematic reviews that seek to answer narrowly defined research questions, scoping reviews are well-suited to fields where emerging technologies, diverse methodologies, and rapidly evolving evidence bases exist. This scoping review was prospectively registered in Review Registry (registration number: reviewregistry 2055), and no deviations were present post-registration.

The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines to ensure methodological transparency and reproducibility. This framework facilitated structured documentation of the literature search, screening, inclusion and exclusion criteria, and synthesis of results. The methodological stages included (1) formulation of the research question; (2) identification of relevant studies; (3) selection of eligible publications; (4) data extraction and charting; and (5) collation, summarisation, and reporting of results.

2.2. Search Strategy

A comprehensive and systematic search of peer-reviewed literature was conducted across PubMed, ScienceDirect, and IEEE Xplore databases. These platforms were selected for their broad coverage of biomedical, clinical, and engineering research relevant to prosthetic development. The search encompassed studies published between January 2013 and December 2023, reflecting a decade of rapid technological advancement in powered prosthetic systems. The search cut-off of December 2023 was implemented as part of the last literature search in July 2024 to create a stable and repeatable boundary to include studies in the study.

The search strategy combined controlled vocabulary and free-text keywords structured around the Population, Intervention, Comparator, and Outcome (PICO) framework (

Table 1. A PICO table showing the population, intervention, comparator and outcome.

Population	Intervention	Biological Comparator	Prosthetic Comparator	Outcome
People with transtibial amputation	Actuated below-knee prostheses	Sound limb	Energy-storage-and-return foot	Gait pattern
Individuals with lower-limb loss	Active prosthetic systems	Non-amputee controls	Passive prosthetic foot	Walking pattern
Transtibial prosthesis users	Powered prostheses	Typical gait	Solid Ankle Cushion Heel (SACH)	Ambulation
Persons with transtibial limb loss	Bionic prosthetic devices	Normal limb function	—	Walking behaviour

). Boolean operators (AND, OR) and truncations were employed to refine results and ensure inclusivity. Specific search strings included terms such as:

• PubMed: (Active OR Powered OR bionic) AND Prosthe* AND (ankle OR foot);
• Sciencedirect: “Powered” AND (Prostheses OR Prosthesis OR Prosthetic) AND (Ankle OR Foot) AND “Below knee”;
• IEEExplorer: (“transtibial amputee*” OR “below-knee amputee*” OR “lower limb amputation”) AND (“active prosthe*” OR “powered ankle–foot” OR “bionic prosthesis”).

Search results were exported to RefWorks reference management software, where duplicates were automatically identified and removed. To ensure reproducibility, detailed search logs were retained, including database source, search date, and full query syntax.

2.3. Inclusion and Exclusion Criteria

The inclusion and exclusion criteria were established a priori to maintain consistency and reduce selection bias.

Inclusion criteria:

• Studies involving human participants with unilateral transtibial amputation.
• Investigations examining active, semi-active, or powered ankle–foot prostheses.
• Experimental studies evaluating gait biomechanics, energy cost of walking, or muscle activity.
• Research conducted on level ground walking at self-selected or controlled speeds.

Exclusion criteria:

• Studies focusing exclusively on passive or energy-storing and returning prostheses without actuation.
• Research involving transfemoral or through-knee amputations.
• Participants with pre-existing osteoarthritis or other lower-limb pathologies affecting gait.
• Studies analysing gait on stairs or uneven terrain where confounding biomechanical factors are introduced.

Application of these criteria ensured that the synthesis focused specifically on the biomechanical and physiological implications of powered prosthetic technology for transtibial amputees. Due to the nature of this review, no risk of bias was formally assessed.

2.4. Data Charting

The data was extracted from the included peer-reviewed studies into a standardised Excel sheet. The items that were extracted were (1) bibliographic information (author, year); (2) demographics of the participants (sample size, gender, age, etiology); (3) details of the interventions (type of prosthesis, control strategy); (4) experimental setup; and (5) outcome measures (kinematics, kinetics, energy expenditure, muscle activation, user interface measures, technical performance).

2.5. Screening and Selection Process

The screening process was performed in multiple stages to ensure methodological rigour. After removing duplicates, two reviewers independently screened titles and abstracts to identify potentially relevant studies. Discrepancies were resolved through discussion and consensus, with a third reviewer consulted where necessary.

Full-text versions of shortlisted studies were retrieved and assessed against the inclusion and exclusion criteria. A total of 27 studies met all criteria and were included in the final synthesis.

Although scoping reviews are not designed to assess study quality formally, methodological transparency and reproducibility were prioritised. Selection, extraction, and synthesis procedures adhered closely to PRISMA-ScR guidance to minimise bias. The overall screening process is illustrated in Figure 1, which follows the PRISMA-ScR flow diagram convention, detailing records identified, screened, excluded, and included at each stage.

2.6. Data Synthesis and Analysis

Given the heterogeneity of study designs, sample sizes, and outcome measures, a narrative synthesis approach was employed rather than meta-analysis. Quantitative results were compared descriptively, while qualitative observations were integrated contextually. After the data extraction, the studies were tabularly categorized and contrasted based on salient characteristics such as the type of prosthesis and the main result into a tabular structure (see Appendix A and Supplementary Material Tables). A narrative synthesis was then conducted, which entailed conducting a repeated review of the tables in order to identify patterns of consistency, inconsistency and new themes in the literature.

This synthesis approach enabled the identification of converging findings and evidence gaps across different prosthetic technologies. The analysis also considered confounding factors such as walking environment, prosthesis tuning, and participant rehabilitation status.

3. Results

A total of 27 peer-reviewed studies published between 2013 and 2023 were included in this review, encompassing over 140 participants with unilateral transtibial amputation. Although several studies also included transfemoral amputees and able-bodied participants [1,10], only data pertaining to below-knee amputees were considered in this synthesis to preserve biomechanical homogeneity. The studies represented diverse methodological approaches—ranging from single-subject case studies to controlled laboratory experiments—providing comprehensive insights into gait kinematics, kinetics, energy expenditure, and muscle activation associated with active ankle–foot prostheses.

3.1. Participant Demographics and Experimental Context

Across the studies reviewed, participants’ ages ranged from 17 to 65 years [18,19], with the majority being male. A review of articles that reported gender showed that there was a significant disparity, whereby 89.3 per cent of the individuals sampled were male [13,20,21]. Only a few studies explicitly included female amputees [2,18,22,23], underscoring a notable gender disparity in the literature. Causes of amputation varied, including trauma, vascular disease, and infection, but traumatic aetiologies predominated [6,12,20,21]; however, a notable number of studies did not report aetiology [3,10,13,24,25]. Mean height and body mass across studies were comparable to non-amputee control populations, though variability existed in prosthetic alignment and limb length compensation, which may have influenced gait performance [26].

Experiments were conducted under controlled laboratory conditions, utilising either treadmill walking [1,7,20,26,27,28] or overground walking [3,29,30]. Participants typically walked at self-selected speeds, though some studies implemented fixed-speed protocols to standardise comparison between devices [6,31]. Evidence suggests that treadmill and overground walking yield comparable biomechanical outcomes [32,33], validating the use of treadmill testing as a representative experimental approach for gait analysis. However, the lack of ecological environments (e.g., slopes, uneven surfaces, fatigue states) limits the generalisability of findings to real-world ambulation, with only a few studies venturing beyond level ground [6,24,30].

3.2. Gait Kinematics

Most reviewed studies consistently reported that powered ankle–foot prostheses improved kinematic symmetry when compared to passive devices. Specifically, powered prostheses enabled greater ankle plantarflexion during push-off and increased ankle range of motion throughout the gait cycle [3,34,35]. This enhancement facilitated smoother transitions between stance and swing phases, contributing to more natural foot roll-over characteristics [4].

For instance, Gabert et al. [1] demonstrated that a compact polycentric robotic ankle–foot prosthesis restored ankle motion trajectories closely resembling those of non-amputees, particularly during terminal stance. Similarly, De Pauw et al. [3] found improved symmetry in ankle angle and velocity profiles when participants used active prosthetic feet compared with conventional energy-storing and return (ESAR) designs. These improvements were accompanied by more consistent step lengths and reduced mediolateral sway, suggesting improved stability and gait efficiency.

However, despite these kinematic gains, complete restoration of normative gait remained elusive. Persistent pelvic obliquity and trunk lean were noted across several studies [3,10,24], reflecting compensatory strategies to maintain balance and forward progression. Esposito et al. [6] reported that while powered prostheses improved symmetry at the ankle, compensatory trunk movements persisted. This is further supported by the findings of increased whole-body momentum [31]. This highlights that enhanced distal joint control may not necessarily translate to full-body biomechanical normalisation.

3.3. Gait Kinetics

From a kinetic perspective, active prostheses influenced joint moments and ground reaction forces in ways that suggest partial restoration of physiological mechanics. Various studies observed a reduction in peak vertical ground reaction forces on the sound limb when powered prostheses were used, indicating decreased overloading compared with passive configurations [2,6,15]. This redistribution of load has critical implications for mitigating the long-term risk of osteoarthritis in the contralateral knee [6].

During late stance, powered devices generated positive mechanical power through active plantarflexion, augmenting forward propulsion and reducing mechanical work demands on the residual limb [2,22,33]. Wolf et al. [2] reported that users of powered prostheses achieved smoother centre of pressure progression and reduced braking impulses during heel strike, resulting in lower collisional losses. These findings collectively suggest that powered prostheses can contribute to improved mechanical efficiency by emulating the natural push-off mechanism.

However, inter-study comparisons reveal substantial heterogeneity in kinetic outcomes, largely attributable to differences in device tuning, control algorithms, and user adaptation. Davidson et al. [12] found that altering prosthetic power output did not consistently reduce collisional work at heel strike, challenging the assumption that higher push-off necessarily improves walking economy. This underscores the importance of user-specific prosthetic calibration and highlights the nonlinear relationship between mechanical output and overall metabolic benefit.

3.4. Energy Expenditure

Energy expenditure during ambulation remains one of the most critical determinants of prosthetic efficacy. Amputees typically experience a 30–40% increase in metabolic cost compared with able-bodied individuals. The reviewed studies demonstrate that active prosthetic devices can reduce metabolic cost to varying degrees, depending on the level of actuation and control sophistication.

Esposito et al. [22] and Ingraham et al. [20] reported significant reductions in oxygen consumption and metabolic rate when users employed powered prosthetic feet capable of dynamic energy storage and return. For example, Esposito et al. [22] found a 16% lower metabolic rate with a powered prosthesis on level ground. However, other studies did not find any statistically significant group-level differences in metabolic cost during powered and passive conditions [7,21,23,28].

Ref. [27] observed that there is no significant interaction between net prosthesis work rate and metabolic rate and postulated that a threshold effect is present after which extra mechanical support produces nonlinear returns. This observation aligns with the concept of metabolic optimisation, wherein users adapt gait patterns to balance energy efficiency and stability rather than maximising propulsion. The interindividual variability was always noted to be high in these studies [7,20].

3.5. Muscle Activation and Neuromuscular Adaptation

The electromyographic (EMG) data across studies revealed adaptive neuromuscular responses to active prosthetic use. Kim et al. (2021) [7] observed that powered ankle–foot prostheses reduced activation levels in hip extensors and hamstrings, reflecting decreased compensatory demand during late stance and swing phases. Conversely, De Marchis et al. [10] reported that users exhibited increased activity in proximal stabilisers such as the gluteus medius, which facilitated balance and mediolateral control.

Increased biceps femoris activation during terminal stance was reported in several studies [7,10], signifying a compensatory mechanism for stabilising the knee joint and synchronising hip–ankle coordination. This adaptation may reflect the downstream consequences of residual limb muscular limitations, such as the need for greater knee stability/hip control following the loss of the primary ankle plantarflexors. Importantly, higher gluteus medius activity correlated with lower overall energy expenditure, suggesting that neuromuscular efficiency plays a critical role in gait economy.

Despite these adaptations, EMG patterns remained asymmetrical between limbs, particularly during stance transitions. The persistence of asymmetry highlights that prosthetic assistance alone cannot fully normalise muscle coordination, necessitating rehabilitation strategies that combine mechanical enhancement with neuromuscular retraining.

4. Discussion

4.1. Overview of Key Findings

This review provides a comprehensive synthesis of recent research addressing the biomechanics of active below-knee prosthetic systems. The findings collectively indicate that powered ankle–foot devices contribute to significant improvements in gait kinematics, partial restoration of ankle joint power, and modest reductions in metabolic cost. However, the persistence of compensatory trunk and pelvic movements suggests that achieving complete biomechanical normalisation remains an ongoing challenge. Furthermore, substantial variability across studies, stemming from differences in device type, control strategy, and user adaptation, underscores the complexity of translating mechanical improvements into physiological efficiency.

Despite these challenges, the evolution of prosthetic technology from purely passive to powered systems represents a major milestone in lower-limb rehabilitation. By enabling active control of ankle motion, powered prostheses approximate the biomechanical role of the intact limb more effectively than their passive counterparts [2,36,37]. Yet, as demonstrated in several studies [6,12,15], users continue to rely on compensatory mechanisms that shift the biomechanical burden proximally to the hip and trunk. These residual asymmetries not only elevate energy cost but may also predispose individuals to long-term degenerative conditions, particularly in the contralateral limb [13].

4.2. Integration of Biomechanical Theory

The observed benefits and limitations of active prosthetic devices can be interpreted through a biomechanical lens. The natural human ankle performs critical roles in energy absorption, storage, and return during walking, acting as both a shock absorber and a power generator. Passive prostheses fail to replicate this bidirectional energy flow, leading to higher mechanical work demands on the proximal joints. Active prosthetic feet, equipped with actuators and sensors, are designed to emulate the timing and magnitude of biological ankle power output, especially during terminal stance.

Biomechanical theory posits that gait symmetry depends on the coordinated transfer of mechanical power across the lower-limb joints. When the prosthetic ankle fails to provide sufficient push-off power, compensatory hip extension and knee flexion increase to maintain forward momentum. Studies such as Gabert et al. [1] and De Pauw et al. [3] have shown that restoring ankle power can reduce this compensatory workload. However, the non-linear relationship between ankle work and overall metabolic cost, as noted by Davidson et al. [12], suggests that merely increasing prosthetic output does not guarantee efficiency gains. Instead, optimal outcomes likely depend on matching mechanical output to individual gait dynamics and user-specific neuromuscular control patterns.

4.3. Design Innovations and Mechanical Control

A recurring theme across the literature is the pivotal role of mechanical design and control systems in determining prosthetic performance. Recent research highlights the emergence of Variable Impedance Actuators (VIAs) as a promising approach for replicating the adaptive stiffness characteristics of the biological ankle joint [38]. VIAs allow real-time modulation of joint compliance in response to changing gait demands, providing both stability during stance and elasticity during push-off.

By dynamically adjusting damping and stiffness, VIAs enable smoother energy exchange between the prosthetic limb and the ground, effectively reducing impact forces and improving energy return. This adaptability has the potential to mitigate residual limb discomfort and fatigue—two major barriers to long-term prosthesis use. Windrich et al. [38] demonstrated that VIA-based designs can significantly improve comfort and gait symmetry by allowing a prosthesis to adjust mechanical impedance across different phases of walking.

Complementing mechanical innovations are advances in prosthetic control algorithms, which range from finite state machines to adaptive myoelectric systems. High-level controllers integrate environmental sensors and inertial data to detect walking conditions, while low-level impedance controllers modulate torque and stiffness accordingly. This multilevel control architecture mirrors the human neuromotor hierarchy, where high-level intent translates into low-level joint control. Studies such as Simon et al. [21] and Kim et al. [13] underscore the potential of electromyographic (EMG) control, which enables more intuitive and voluntary operation by directly translating muscle signals into prosthetic actuation commands. However, signal noise, latency, and inter-user variability remain key technical challenges limiting widespread clinical adoption.

4.4. Neuromuscular and Energetic Implications

Despite improvements in mechanical assistance, powered prostheses do not fully normalise neuromuscular coordination between limbs. The persistence of asymmetrical EMG activation patterns, particularly increased activity of the biceps femoris and gluteus medius, suggests that prosthetic use induces a unique neuromuscular adaptation pattern rather than restoring pre-amputation function [7,15]. These adaptations likely represent a compromise between maintaining stability and conserving energy.

The relationship between mechanical power assistance and metabolic cost appears complex. While active prosthetic systems can provide external mechanical energy, the user’s neuromuscular system must still coordinate the timing and control of that energy to maintain balance. Quesada et al. [27] demonstrated that excessive mechanical push-off may actually increase metabolic demand if it disrupts synchrony between the prosthetic and biological limbs. Conversely, Au et al. [39] found that moderate power assistance optimised metabolic economy by reducing eccentric braking and concentric propulsive work on the sound limb. These findings underscore the importance of individualised prosthetic tuning, wherein mechanical and physiological parameters are harmonised to achieve optimal gait efficiency.

Active prosthetic feet focus on replicating the sagittal plane ankle movements, while neglecting active frontal and transverse plane ankle movements. Though these movements provide minimal contribution in shock absorption and propulsion, they play a significant role in controlling mediolateral movements and maintaining the hip-foot relationship. Ankle eversion and inversion have been known for contributing to the reduction in mediolateral sway, and in the absence of this active movement, the hip muscles compensate for this absence [40]. The trend in powered prosthetic feet of using passive ESAR for mediolateral control or completely eliminating mediolateral control may be a contributing factor. Therefore, future studies must investigate the influence of replicating active mediolateral control in regulating gluteus medius activity.

The long-term adaptation appears to play a crucial role. Users who undergo extended familiarisation periods exhibit improved coordination and reduced asymmetry compared to short-term users [7]. This supports the concept that neuromuscular retraining is integral to achieving functional integration of powered prostheses. Hence, rehabilitation protocols should not only focus on mechanical fitting but also include gait retraining programmes that reinforce symmetrical motor patterns and adaptive muscle activation strategies.

4.5. Clinical and Translational Implications

From a clinical standpoint, the findings of this review underscore the potential of active below-knee prosthetics to transform rehabilitation outcomes for transtibial amputees. By providing active ankle power and adaptive stiffness, these devices can improve walking performance, reduce compensatory strain, and enhance overall mobility. However, their clinical implementation must consider several key factors:

• Variations in residual limb length, muscle strength, and balance necessitate custom calibration of prosthetic stiffness and torque output.
• Effective integration of powered devices requires prolonged acclimation and guided physiotherapy to promote neuromuscular adaptation.
• Future designs should account for real-world terrains and activities, ensuring robustness across uneven and compliant surfaces.
• Addressing the significant underrepresentation of female participants is essential for developing gender-responsive prosthetic systems that account for anatomical and biomechanical differences.

The current guidelines in prosthetic rehabilitation often lack a foundation in robust evidence, indicating a need for standardised procedures that extend beyond gait training [41]. Promising paradigms, such as dual-task training to improve automaticity and balance under cognitive load, are underexplored in the amputee population and represent a crucial area for future research [41]. Therefore, clinicians should adopt an evidence-based, user-centred approach, integrating mechanical tuning, gait analysis, and rehabilitative feedback loops to optimise outcomes. The translation of laboratory findings into clinical practice also requires greater emphasis on longitudinal monitoring of joint health, energy expenditure, and musculoskeletal strain to mitigate secondary complications.

4.6. Limitations of Current Research

While the reviewed literature has advanced understanding of powered prosthetic performance, several limitations constrain interpretive generalisation. The small sample sizes (typically fewer than 10 participants per study) limit statistical power and the ability to capture population-level variability. Most investigations were conducted in controlled laboratory environments, which, while valuable for internal validity, fail to replicate the complex perturbations and surface variations encountered in daily life.

Moreover, inconsistencies in outcome measures, such as differing definitions of gait symmetry, normalisation techniques, and data processing algorithms—impede direct cross-study comparison. There is also a paucity of longitudinal data exploring the cumulative effects of powered prosthesis use on joint health and energy metabolism over time. Future research should adopt standardised biomechanical protocols, employ larger, more diverse cohorts, and integrate real-world monitoring technologies such as wearable inertial sensors to capture continuous gait data in naturalistic settings.

5. Conclusions

A growing body of research has examined the biomechanical and neuromuscular challenges faced by individuals with transtibial amputation; however, considerably fewer studies have explored how targeted physiotherapy and rehabilitation strategies enhance long-term adaptation to prosthetic devices. The literature referenced by the reviewer illustrates this gap clearly. Recent work on cognitive–motor interference in lower-limb prosthesis users [42] provides valuable insight into postural stability and dual-task performance, yet its scope remains largely assessment-driven. The study quantifies balance disturbances under cognitive load and highlights underlying deficits in automaticity, but it does not evaluate whether structured rehabilitation, such as progressive motor retraining or dual-task physiotherapy, can ameliorate these impairments. Consequently, while such studies inform clinicians about the nature of postural challenges during everyday mobility, they do not clarify how rehabilitation interventions might improve prosthesis adaptation or functional independence.

In contrast, the broader neurological rehabilitation literature provides compelling evidence that dual-task gait training, when delivered with appropriate timing, progression, and cognitive complexity, can significantly enhance gait automaticity, stability, and motor control [42,43]. These studies demonstrate that integrating cognitive and physical tasks within physiotherapy can accelerate motor learning and improve real-world functional outcomes. However, although theoretically relevant, these findings have not been translated adequately to the amputee population. Lower-limb amputees present unique biomechanical constraints, including altered limb loading, socket–residual limb interactions, and compensatory muscle activation, that differ substantially from neurological disorders. As a result, neurorehabilitation principles cannot be assumed to have equivalent effects in prosthesis users without direct empirical evaluation.

The present scoping review reinforces this gap: across the 27 included studies, the vast majority focused on device-specific performance under controlled laboratory conditions rather than rehabilitation-specific outcomes. Studies typically assessed gait kinematics, kinetics, metabolic cost, or muscle activation when comparing passive versus active prosthetic devices, but did not integrate or examine physiotherapy interventions, gait retraining programmes, or dual-task training paradigms. Participants were predominantly long-term prosthesis users who had already completed routine physiotherapy for strengthening and basic gait training, which limits the ability to attribute observed improvements to rehabilitative processes. As a result, current evidence reflects the mechanical benefits of prosthetic design, rather than the functional gains that may emerge from structured rehabilitation tailored to prosthesis adaptation.

None of the reviewed studies evaluated real-world community reintegration, dynamic environmental challenges, or long-term adaptation under variable cognitive and physical demands. These omissions are clinically significant, as community ambulation requires continuous integration of attention, balance, environmental negotiation, and device control, demands that physiotherapy is well-positioned to address. The limited exploration of these aspects highlights the need for future research that integrates targeted rehabilitation protocols, including dual-task gait training, motor control exercises, and progressive skill-based training, alongside biomechanical assessments of prosthetic devices.

These findings underscore that meaningful progress in prosthetic mobility cannot rely solely on advances in mechanical design. Instead, optimal functional outcomes are likely to arise from a combined approach that pairs technological innovation with evidence-based rehabilitation strategies that support neuromuscular adaptation, cognitive–motor integration, and real-world functional capacity. This represents a critical gap in the current literature and an important direction for future research.

5.1. Implications for Prosthetic Design

The review highlights a paradigm shift from static, passive devices toward adaptive, powered prosthetic systems capable of responding dynamically to user intent and environmental conditions. This transition mirrors broader developments in bioinspired engineering, where mechanical structures are designed to emulate the variability and adaptability of biological tissues.

Emerging technologies such as VIAs and EMG driven control systems offer promising avenues for future design. VIAs enable modulation of stiffness and damping throughout the gait cycle, allowing the prosthesis to adjust energy storage and release in synchrony with the user’s movement. This technology addresses the limitation of fixed-stiffness devices that cannot accommodate changes in walking speed or terrain. Likewise, EMG-driven controllers introduce an element of volitional control, translating residual muscle activity into actuator commands that more closely mirror natural motor intent [13,38,44].

Nevertheless, successful implementation of these technologies will depend on solving several practical challenges, including signal noise reduction, latency optimisation, power efficiency, and device miniaturisation. Integration of machine learning algorithms to interpret EMG patterns and adapt control parameters in real time represents a promising next step toward autonomous, self-tuning prosthetic systems. Furthermore, incorporating real-world adaptive feedback—such as terrain sensing or gait phase recognition—may enable future prosthetic devices to predict and respond to environmental variations, further enhancing user safety and comfort.

5.2. Clinical and Rehabilitation Perspectives

Clinically, powered prostheses offer a unique opportunity to enhance functional mobility and reduce secondary complications among transtibial amputees. The observed decrease in contralateral joint loading and partial restoration of gait symmetry have important implications for mitigating osteoarthritis progression and chronic lower back pain, both of which are prevalent in long-term prosthesis users [6,12,14].

However, achieving these benefits in practice requires more than mechanical fitting. Comprehensive rehabilitation programmes that integrate physiotherapy, neuromuscular re-education, and adaptive gait training are critical for helping users exploit the full functional potential of powered prostheses. These programmes should incorporate dynamic gait analysis, enabling clinicians to fine-tune prosthetic parameters based on real-time biomechanical feedback. Additionally, the inclusion of gender-specific prosthetic fitting protocols is essential, given the underrepresentation of women in existing research and potential differences in gait biomechanics due to pelvic and muscular morphology [45,46].

Long-term monitoring is also necessary to understand how powered prosthesis use affects joint health and energy efficiency over time. Integration of wearable gait sensors and telemetric monitoring systems could allow clinicians to track joint kinetics, step symmetry, and metabolic indicators in naturalistic settings, bridging the current gap between laboratory studies and daily use.

5.3. Research Priorities and Methodological Recommendations

The reviewed literature reveals several key priorities for future research:

• Most current evidence is derived from short-term laboratory trials; long-term studies are needed to capture adaptation trajectories, fatigue effects, and sustained biomechanical changes.
• Future trials should recruit representative populations, including adequate female participation, older adults, and users from low-resource settings, to enhance external validity.
• Adoption of unified biomechanical metrics, such as joint power normalisation, symmetry indices, and metabolic cost per stride, will facilitate cross-study comparisons and meta-analytical synthesis.
• Combining EMG, motion capture, and metabolic data can provide a holistic understanding of how mechanical and neural factors co-adapt during powered prosthesis use.
• Effective innovation requires cooperation between biomedical engineers, physiotherapists, and clinicians to ensure that technological development aligns with user needs and clinical feasibility.

The transition from passive to active prosthetic systems marks a transformative period in lower-limb rehabilitation. While current powered devices substantially improve functional gait characteristics, they do not yet replicate the integrative control of the human neuromuscular system. Bridging this gap will require innovations that merge biomechanical engineering, neural control, and rehabilitation science into a unified framework.

Future generations of prosthetic technology must not only replicate human movement but also adapt to the user’s physiology, learning, predicting, and optimising performance dynamically. By combining mechanical sophistication with biological insight, it may ultimately be possible to achieve truly symmetrical, efficient, and sustainable gait restoration for individuals with transtibial amputation.