Secondary Logo

Journal Logo

Outcome Measurement Tools

Use of Quantitative Gait Analysis for the Evaluation of Prosthetic Walking Performance

Gard, Steven A. PhD

Author Information
JPO Journal of Prosthetics and Orthotics: January 2006 - Volume 18 - Issue 6 - p P93-P104
  • Free



A quantitative gait analysis is generally considered to be any objective means that can be used to measure walking performance. The procedure can be as simple as measuring step length with a ruler or determining cadence with a stopwatch, or it can be as sophisticated as full-body motion capture with state-of-the-art instrumentation. Regardless of the methods, the measurements that are collected are used to assess the quality of the gait and to characterize the motion. Observational gait analysis involves a subjective assessment of an individual's gait, but experienced individuals are often able to visually identify many of the same gait abnormalities that can be discerned with quantitative gait analysis. However, key advantages of quantitative gait analysis for persons with lower-limb pathologies are that the results allow for easy comparison of a patient's gait characteristics to an able-bodied pattern for a relatively quick determination of abnormal movements, and it documents a patient's gait at a particular point in time so rehabilitation progress can be tracked.

Using quantitative gait analyses to fully describe a person's gait generally entails the combination of a multitude of measurements, including temporal-spatial parameters, kinematics, kinetics, and energy expenditure. When presented with large quantities of descriptive measurements, wading through all of the data and picking out relevant information can take a tremendous amount of time and effort. However, the process can generally be facilitated by involving someone who is knowledgeable about the measures and skilled in analyzing and interpreting the data. Visual gait analysis, performed first-hand or by viewing a videotape recording of the subject's gait, can greatly aid with the interpretation of the quantitative gait data.

Once regarded as a research endeavor, quantitative gait analysis has now become a clinically accepted means for documenting and evaluating the characteristics of a person's gait, particularly in the presence of pathologies that affect walking. Presently, there are numerous clinical gait analysis laboratories dedicated to the evaluation of children with cerebral palsy, myelomeningocele, or other disabling conditions that affect walking. The results from these analyses are used by physicians and therapists to determine appropriate surgical interventions or courses of treatment for the child with the intent of improving walking efficiency and appearance. Gait analyses are typically performed before and after the intervention to determine efficacy of treatment. Treatment is considered beneficial if improvement in the gait pattern is observed, evident by a reduction in abnormal movements with an evolution toward patterns that are more like those of able-bodied individuals.

Gait analysis may be useful for evaluating an amputee's prosthesis by providing objective measurements that characterize the walking pattern.1 Fundamentally, the prosthetist's goal in fitting a lower-limb amputee is to restore the ability to perform everyday activities in an easy, natural, and comfortable manner. The prosthesis user's basic requirements of the prosthetic limb are comfort, function, and appearance, the latter embracing both cosmetic appearance and appearance in use.2 The goals of the practitioner and the patient are similar, but prosthesis users may not be able to fully or accurately articulate what they perceive when they stand and walk, making it difficult for the prosthetist to make all of the necessary adjustments. Augmenting the human body with a prosthesis markedly affects the individual's mode of travel, and the task of the clinician should be to recognize optimal gait with a given device so that departures from the standard can be identified, their causes determined, and, wherever possible, corrected.3 Of all the elements affecting locomotion, those most amenable to change relate to the device; thus, in the description of walking patterns, emphasis should be placed on prosthetic design, alignment, and fit.

Subjective methods of gait assessment may be inadequate for optimizing the gaits of lower-limb amputees.4 However, the benefit of performing routine clinical quantitative gait analyses has yet to be realized for several different reasons. Cost is a primary deterrent to performing clinical gait analyses, because much time and effort are required by trained personnel for the acquisition, processing and interpretation of data. Clinicians want gait analysis techniques that can be routinely used in the diagnostic and decision-making process, and if these techniques were able to improve the quality of patient care or reduce treatment cost, they would probably find widespread clinical acceptance.5 Too, limitations in our understanding about the pathomechanics of amputee gait and the functions that need to be provided by prostheses inhibit our ability to effectively use quantitative gait data as a means to diagnose and treat observed gait deviations. Furthermore, data pertaining to amputee gait can be difficult to assess and interpret because it can be highly influenced by the particular choice of prosthetic components, socket type, and suspension, as well as by the residual anatomy, abilities and psychological well-being of the patient.

Prosthetists must be skilled in observational gait analysis to perform a rapid assessment of their client's gait in the clinic and make appropriate adjustments to the prosthesis to eliminate or reduce gait abnormalities. Quantitative gait evaluations are able to provide additional, objective information to supplement the clinical observation, and it may prove to be useful for documenting the rehabilitation progress of patients over time. The gaits of amputees vary according to a number of factors, including the level of their amputation, their state of health, and whether they have unilateral or bilateral amputations. Typically, the more distal the amputation, the better control the amputee has of his prosthesis, the more efficient the gait, and the more closely their pattern of walking resembles that of able-bodied persons. As the level of amputation occurs more proximal, the amputee must compensate for his loss to a greater extent and more abnormal gait patterns usually result. Despite the objective data, the information gained from a quantitative gait analysis will probably not enable the experienced clinician to make better decisions regarding prosthetic prescription or indicate how the prosthesis should be modified. Even though deviations from a normal gait pattern can be readily discerned in the quantitative data, diagnosing the origin of the problem can be difficult.

Despite limited clinical utility, quantitative gait analyses are useful to researchers in gait and lower-limb prosthetics. Published studies report on the analyses of different walking parameters for the purpose of characterizing amputee gait relative to able-bodied persons, and for comparing the effects of different types of prosthetic components on walking ability. These data may benefit therapists by enabling them to better recognize and identify abnormalities in gait and to develop more appropriate training regimens that would allow prosthesis users to walk more efficiently. Additionally, the information could potentially be used for establishing guidelines for prosthetic prescription. Researchers may also use the data from quantitative gait analyses to identify and address deficiencies in current prosthetic componentry that prevent amputees from walking with more normal patterns of gait and with greater efficiency. However, gait analyses are not typically performed on prosthetic users for clinical decision-making, because on individual patients the data have not been shown to be particularly enlightening or indicative of a particular prosthetic or therapeutic solution. There are probably very few cases, if any, when a prosthetist would change his mind about component selection or prosthesis adjustment based upon the results of a quantitative gait analysis. It is widely known that prosthesis users typically demonstrate gait patterns that are different from those of able-bodied individuals, with specific patterns generally being adopted according to their level of involvement and type of prosthetic components used to construct the prosthesis. However, quantitative gait analyses are not diagnostic; that is, it is generally not possible to analyze a set of gait data and be able to diagnose a condition that prosthetic prescription or modification would alleviate. At the moment, the preferred and likely better means of optimizing a person's gait requires open dialogue between the prosthetic user and his or her prosthetist, combined with the careful attention, observation and experience of the prosthetist.

A number of gait parameters are used to characterize human ambulation, and these may potentially be used to document prosthetic outcomes. Temporal-spatial parameters, such as walking speed, cadence, and step length, are useful because they provide gross indications of overall walking ability, and many of them can be made with simple clinical methods or tools. Other measures, such as joint kinematics and kinetics, ground reaction forces, and energy expenditure, require more sophisticated, expensive equipment that may not be practical for clinical purposes. The purpose of this article is to review many of the typical measures of quantitative gait analyses reported in the literature for evaluating prosthetic walking performance, particularly those that may be used as outcome measures. Although many quantitative gait studies have been performed on users of prosthetic limbs, this document is not intended to be a comprehensive review. Rather, this paper presents an overview of some of the more widely accepted gait measures that have been employed for prosthetic gait analyses and provides typical results from those types of studies.


Temporal-spatial parameters are particularly useful measurements for prosthetic evaluation because they provide fundamental timing and position information about a person's gait, and they can be made relatively easily in a clinical setting with simple measurement tools. The most common temporal-spatial parameters used are walking speed, stride length, step length and cadence. Able-bodied adults tend to demonstrate similar temporal-spatial measures at freely-selected walking speeds. Persons with lower-limb amputation generally exhibit characteristics inferior to those of able-bodied persons but are generally comparable across individuals, based on their level and cause of amputation.

Walking speed probably provides a better indication of a person's walking ability than any other quantitative gait measure. The speed of walking is directly associated with the biomechanics of gait, which relies on achieving and maintaining forward momentum of the body mass, accelerating the masses of the body segments from one step to the next, transferring load from one leg to the other, and taking advantage of energy conserving mechanisms, thereby achieving an efficient, dynamic walking pattern. Most adult, able-bodied ambulators typically adopt freely-selected speeds of approximately 1.2 to 1.5 m/sec.6,7 The freely-selected walking speed is the comfortable speed that an individual naturally chooses and is believed by many to represent the speed at which the energy expended per unit distance traveled is minimized. Related evidence suggests that the body's conservation of mechanical energy is greater at freely-selected speeds than at speeds significantly faster or slower.8

Persons who walk with lower-limb prostheses generally walk slower than able-bodied persons, with those persons having more proximal levels of amputation typically adopting slower speeds. Unilateral transtibial amputees have been reported to walk at freely-selected speeds from about 1.0 to 1.3 m/sec.9,10 Transfemoral amputees typically walk at freely-selected speeds from about 0.9 to 1.2 m/sec.11–13 Because their walking speeds are generally slower, prosthetic users may not have sufficient body dynamics to enable them to take full advantage of the energy-conserving mechanisms utilized during able-bodied gait. Walking speed can be used as a means to assess prosthetic performance by determining if prosthetic modification produces a higher freely-selected speed. Presumably, an increase in a prosthetic user's freely-selected speed indicates that the prosthetic configuration is able to produce a more efficient pattern of walking. However, most of the walking parameters analyzed using quantitative gait analysis are known to vary with walking speed. Therefore, when comparing pathological gait data with that of able-bodied individuals, it is important to distinguish between variations due to walking speed and those due to the pathology.14

The stride length is the linear distance along the line of progression that a person travels during one gait cycle. Stride length, by definition, must be the same when measured on either side of the body for both normal and pathological gaits. For a unilateral amputee, the distance traversed during a stride will be the same whether it is measured on the prosthetic side or the sound side. By definition, the stride length is equal to the sum of the right and left step lengths. Able-bodied persons walking at freely-selected speeds generally adopt a stride length of about 1.4 to 1.5 m, with step lengths of approximately 0.7 to 0.75 m.7,15 In normal walking, the right and left step lengths are generally equal, but this is typically not the case for pathological gaits. For persons with unilateral amputation, the prosthetic step length is usually slightly longer than the sound limb step length. A satisfactory explanation for this difference has not been offered, but it may relate to stability during stance phase or to foot/ankle function of the prosthesis. Walking speed is equal to the product of one-half the stride length and the cadence, where cadence is simply the number of steps taken per unit time. Able-bodied ambulators typically walk with a cadence of about 110 steps/min at freely-selected speeds.7,15 Higher walking speeds are accompanied by simultaneous, predictable increases in cadence and stride length in able-bodied individuals,6,16 but amputees may demonstrate a greater reliance on one mechanism or the other to increase their speed due to constraints imposed by amputation or by their prostheses.

Temporal-spatial measures are useful for characterizing walking performance of amputees, with values that typically cluster for prosthetic users having similar amputation levels and cause of amputation. Dysvascular transtibial amputees have been shown to walk with significantly slower freely- selected walking speeds than traumatic transtibial amputees (0.85 vs. 0.99 m/sec).17 Unilateral transfemoral amputees walking with constant-friction knee components have been found to adopt a significantly slower freely-selected speed (1.0 m/sec), a lower cadence (87 steps/min), and a shorter stride length (1.36 m) compared with able-bodied subjects.12 Additionally, the amputee subjects demonstrated a significantly shorter step length with their sound leg than with their prosthesis (0.64 m vs. 0.72 m), a result that is typical of persons with unilateral amputations. The relatively shortened sound step length may have resulted from the inability of the subjects to actively plantarflex the prosthetic ankle to raise the heel, and from the inability to flex the prosthetic knee until weight was transferred to the forward limb. One study of unilateral transfemoral amputees measured a mean freely- selected speed of 1.13 m/sec, with a cadence of 90.5 steps/min.18 The stride length of the transfemoral amputee subjects was comparable to that of able-bodied individuals, indicating that cadence appeared to be the primary factor that limited walking speed. Similar results were reported in a study that found transfemoral amputees walking at faster self-selected speeds utilized longer step lengths and lower cadences relative to able-bodied individuals.11 These findings may not be particularly surprising, as transfemoral amputees often believe that it is the swing of their prosthesis that limits their ability to walk at different speeds. Quantitative gait measures have also been used to track the rehabilitation of transfemoral amputee patients; walking speed was observed to increase by 48% after prosthetic training,19 a period of approximately 6 to 7 weeks. The temporal-spatial measures of gait reported here are consistent with results from other studies of unilateral transfemoral amputees.11,20

Symmetry between temporal-spatial measures for the sound and prosthetic legs of unilateral amputees is often used as a means to evaluate prosthetic gait, but the value of using gait symmetry as a performance measure is questionable. It has been suggested that symmetry between the sound and prosthetic limbs is the best method to analyze and evaluate different prosthetic feet.21 However, the mass of prosthetic limbs is generally lower than that of an anatomical leg, the segmental positions of the centers of mass are different, and the moments of inertia are typically reduced, which would affect the swing of the prosthesis. Additionally, the prosthesis does not function as well as the sound leg, so the amputee typically compensates for the inadequacies of the prosthesis through the actions of his sound leg. Therefore, it is not only the prosthetic leg that affects the pattern of walking, but the adaptation of the anatomical leg as well. Symmetry between the sound and prosthetic legs is typically concerned with an assessment of appearance during walking (i.e., step length or timing issues), when in fact it is gait function that should be addressed. Therefore, asymmetries between the sound and prosthetic limbs should not necessarily be the basis for determining the quality of a person's gait. When an amputee is learning to walk with a prosthesis, he or she may seek out a new nonsymmetrical pattern of walking that is optimal within the constraints of his or her residual anatomy and the mechanics of the prosthesis.9 Furthermore, when a prosthetist fits a client with a prosthesis and incorporates different components or changes alignment, there will likely be trade-offs in gait performance that will affect symmetry. For example, it may be possible to improve gait symmetry through changes in prosthetic alignment,22 but doing so could adversely affect other key parameters of walking, such as speed. Conversely, some prosthetic components may improve locomotor function during gait, such as with increased shock absorption or improved forward progression, but they may reduce gait symmetry.10 Therefore, prosthetists must carefully evaluate their clients' gait to determine the effects of various interventions and prosthetic modifications, and weigh the advantages and disadvantages appropriately when trying to arrive at an optimal prosthetic solution.

Temporal-spatial measures have also been used in research studies to determine the effects of different types of prosthetic components on amputee gait. However, the results from many studies that perform similar comparisons of components are often at odds. For example, faster freely-selected walking speeds have been reported for unilateral transtibial amputee subjects walking with a Flex Foot (Ossur, Reykjavik, Iceland) compared with when they walked with a SACH foot.23 However, no differences in walking speed were reported for transfemoral amputees walking with a Flex Foot and a SACH foot.24 A study of unilateral transtibial amputee subjects walking with a Seattle Lite Foot (Seattle Systems, Poulsbo, WA) and a SACH foot reported no statistically or clinically significant differences.25 In another study, gait analysis was used to compare six prosthetic feet—the SACH, SAFE II (Campbell Childs, White City, OR), Seattle Lite Foot (Seattle Systems), Quantum (Hosmer Dorrance, Campbell, CA), Carbon Copy II (Ohio Willow Wood, Mt. Sterling, OH), and Flex-Walk (Ossur)—in male unilateral transtibial amputees on the basis of various temporal and spatial measures of walking.21 The investigators observed statistically significant differences between vascular and traumatic amputee groups in walking speed (0.75 vs. 1.01 m/sec), cadence (82.4 vs. 94.7 steps/min), and stride length (1.1 vs. 1.4 m). Additionally, some differences in sound limb step length were observed in only the traumatic group between the different prosthetic feet. Another investigation compared the gait of transtibial amputees wearing a single-axis foot, a Seattle Lightfoot, and a Flex Foot.26 Regardless of foot type, the subjects were observed to walk slower than able-bodied control subjects (63.3 to 65.8 m/min vs. 78.5 m/min, respectively), and their stride length was shorter (1.21 to 1.26 m vs. 1.41 m, respectively). In a study that compared the SACH foot with four different dynamic elastic response feet, there were no clinically significant differences between the feet during freely- selected or fast walking on level ground, nor was symmetry affected.27 An investigation of dysvascular transtibial amputee subjects reported that a Flex Foot significantly increased walking speed compared with a SACH foot, and significantly increased stride length compared with a Seattle, Carbon Copy II (Ohio Willow Wood), and SACH foot.28 In a study that compared six different prosthetic knee joints in transfemoral amputees, no differences were observed in the freely-selected walking speed, step length, or cadence.29 However, the investigators did find that cadence-responsive knees—those incorporating hydraulic or pneumatic damping mechanisms for swing phase control—permitted higher walking speeds with less perceived effort. The investigators concluded that the more complicated prosthetic knee designs offered little benefit when walking on flat level surfaces. Other studies, however, report higher walking speeds and improved symmetry when subjects walked with knees providing swing-phase control compared with when they walked with a constant friction knee.30,31 In a study that investigated age and causative effects on transfemoral amputees walking with a locked versus unlocked prosthetic knee, older amputees who had an amputation secondary due to vascular problems were found to adopt faster freely-selected walking speeds with a locked knee than with an unlocked knee, whereas younger traumatic amputees walked faster with the unlocked knee.32

Unfortunately, the disparate results presented here for different quantitative studies that use temporal-spatial parameters to characterize walking performance of amputee subjects are typical, and as a result they are difficult to draw conclusions from for the purposes of influencing prosthetic prescription. Studies that compare different types of prosthetic components often do not agree, which may relate to slightly different methodologies being adopted by investigators, to the small numbers of research subjects that are included in the study, or to different amounts of accommodation time that subjects are permitted on each component before data collection. Different sets of inclusion/exclusion criteria used by investigators may also influence the outcome of the studies. Differences among individual amputee subjects may also relate to physical capabilities, training, confidence, and experience. Temporal-spatial measures are appealing because they are simple to acquire and easy to comprehend, but they appear to lack consistency and repeatability as a dependable outcome measure when evaluating prosthetic gait.


Kinematic measures describe the position, velocity, and acceleration of an object without regard for the forces that create the observed motion. Temporal-spatial parameters are one form of kinematic data that have been set apart from what are now considered the typical kinematic gait measures. In quantitative gait analysis, kinematics provide detailed information about the linear and angular motions of the trunk and body segments. Of particular interest are the rotations of the pelvis and the joints of the leg. Data are typically presented as graphical curves that show how the body moves over the course of a gait cycle. When matched according to walking speed, the characteristics of the kinematic curves for able-bodied ambulators are remarkably similar. Kinematic data acquired from pathological gaits are typically analyzed by utilizing a type of differential analysis, plotting able-bodied curves onto the same graphs as patient or subject data so that deviations from normal can be readily discerned. Additionally, kinematic curves from different gait analyses performed on the same subject can be plotted onto the same graph to facilitate the identification of changes that occur in a person's gait due to different interventions or testing conditions.

Kinematic data acquired from prosthetic users must be interpreted in light of the assumptions that are used to create the link segment model from which kinematic measures are derived. In the model, physical joints are assumed to exist at the junction between the various limb segments; that is, an ankle joint is assumed to connect the foot segment to the tibia segment. Many amputees do not have physical ankle joints in their prostheses, so in a set of kinematic curves what appears to be prosthetic ankle joint motion is simply the result of bending in the keel of the prosthetic foot. Additionally, joints are generally assumed to rotate about a single axis or center of rotation, which can present problems when scrutinizing data of knee kinematics acquired from a transfemoral amputee walking with a polycentric knee joint. Therefore, caution is urged when analyzing kinematic walking data acquired from a prosthesis user.

Quantitative gait studies have repeatedly shown that amputees typically demonstrate different gait kinematics than able-bodied individuals. Of course, visual analysis of an amputee's gait often indicates that the pattern of walking is different from normal, but kinematic analysis can help identify the joints in the body that are affected, and may help determine the causative source from which those abnormal movements originate. The kinematic patterns of prosthetic joints are usually very different from the anatomical joint they replace, owing to deficiencies of the technology and incomplete knowledge about the function being provided. Furthermore, more proximal levels of amputation demand greater compensatory actions to walk, so deviations from normal movement patterns are often observed in the anatomical joints of the residual limb and of the sound limb.

Transtibial amputees tend to walk with similar kinematics as able-bodied individuals, but subtle differences can sometimes be distinguished in kinematic data. Primarily, the ankle kinematics of their prosthetic limb will differ from the normal pattern due to the inability to plantarflex in late stance phase. Knee kinematics for the sound and residual limbs may also be affected. Some unilateral transtibial amputees have reduced prosthetic-side knee flexion in early stance phase of walking at their freely-selected speeds.33 On the sound side, excessive stance-phase knee flexion angles of transtibial amputees during early stance phase have been speculated to be a compensatory action to reduce impact forces due to impaired ability to transfer body weight from the prosthetic limb onto the sound side.28 Differences between the kinematic walking data of transtibial amputees and that of able-bodied individuals may be due to both the loss of the normal foot-ankle mechanism and the compensatory actions that are required by the prosthesis user to overcome this deficiency.34

Transfemoral amputees tend to demonstrate greater deviations from able-bodied gait kinematic patterns than trans-tibial amputees. The patterns of motion on the prosthetic side of transfemoral amputees generally resemble those of able-bodied individuals, but the magnitudes of rotation tend to be reduced.35 Transfemoral amputees have been reported to walk with slightly reduced prosthetic swing phase knee flexion (45° ± 13°) compared with their sound leg (51° ± 6°) or with that typically achieved by able-bodied persons.36 Transfemoral amputees usually demonstrate a complete absence of stance-phase knee flexion during the loading response phase of gait on the prosthetic limb, particularly when walking with single-axis knees.12 Some prosthetic knees are designed to restore stance-phase knee flexion, but a few degrees of added flexion during stance can be difficult to detect using observational gait analysis due to minute changes in the appearance of the knee and the short duration of the event. However, kinematic data can be particularly useful for verifying their proper operation and use with accurate measurement of the magnitude of the stance-flexion wave.37–39

Kinematic data can be useful for identifying compensatory movements used by amputees during walking. Some trans-tibial amputees may exhibit hip-hiking—a compensatory action to lift the pelvis on the side of the swing leg to increase foot clearance—during prosthetic swing phase. Transfemoral amputees typically exhibit hip-hiking during prosthetic swing phase, and they sometimes demonstrate it during sound limb swing phase as well. As a result of hip-hiking, the pelvic obliquity patterns in persons with unilateral transtibial or transfemoral amputations are often different from normal, showing asymmetry with less excursion during prosthetic stance phase and with significantly different timing during the gait cycle.40 A persistent offset in the pelvic obliquity waveform toward the prosthetic side in transtibial and transfemoral amputee subjects may indicate that some lower-extremity amputees suffer from a dynamic leg-length discrepancy that can be detected only with instrumented gait analysis.40 This characteristic has also been observed in some children walking with known leg length discrepancies, with the pelvis continually listing toward the side of the shorter limb during gait.41 Vaulting, another compensatory movement in which the sound limb ankle plantarflexes during midstance to raise the body and increase prosthetic swing phase foot clearance, can also be readily identified in kinematic data.12,35,42 Hip-hiking and vaulting both require muscular effort to lift the body against gravity, increasing the energy required for walking and altering the normal kinematic patterns of motion. However, merely identifying deviations in the kinematic patterns of prosthesis users does not necessarily diagnose the root cause of the problem. Presumably, identifying and addressing functional deficiencies in the prosthesis would reduce these observed differences and enable more normal walking patterns.

Studies have been conducted to determine how different prosthetic components affect walking kinematics. The majority of these investigations have attempted to discriminate between the effects of different prosthetic foot designs. Generally, more compliant prosthetic feet are typically observed to produce greater apparent ankle dorsiflexion during late stance phase than stiffer feet. In one study, transtibial amputee subjects walking with six different prosthetic feet demonstrated some significant differences in the late-stance ankle dorsiflexion occurring at opposite heel contact, and in the ankle initial plantarflexion to dorsiflexion transition occurring during early to late stance phase.21 The investigators in this study pointed out that the quantitative differences between the prosthetic feet demonstrated each one was activity- and gait-specific for the individual amputee. Another study showed that a Flex Foot and Quantum foot provided significantly increased dorsiflexion in late stance phase compared with a Seattle, Carbon Copy II, or SACH foot.28 A similar study compared walking performance on a SACH foot to a Flex Foot, a Carbon Copy II, a Seattle Foot, and a STEN (Kingsley Manufacturing Co., Kelso, WA) foot, and reported that the Flex Foot increased dorsiflexion throughout the prosthetic stance phase.27 No other kinematic parameters were affected, and the investigators concluded that there were no clinically significant differences among the different prosthetic feet.

Plots of kinematic data curves that show how the leg joints and pelvis move during the gait cycle are useful for characterizing pathological gaits. Using differential analysis to compare data from prosthetic gait with that of able-bodied individuals facilitates the identification of abnormal movements because, at a glance, deviations from normal can be easily recognized. However, the literature is replete with inconsistencies in kinematic data that have been reported for similar samples of amputee subjects who were walking under comparable testing conditions. Therefore, kinematic data do not appear to be particularly useful as outcome measures for prosthetic ambulation.


Kinetic measures of walking are useful for assessing aspects of gait that are not readily evident through visual gait analysis or by analyzing kinematic data alone. Kinetic parameters include the forces, moments, and powers associated with limb movement. These measures may relate to the user's perception of the interaction between their residual limb and the prosthesis during walking.

The ability of a person to move through space requires some structure external to the body against which bodily forces can be exerted. In normal human ambulation, this structure is generally and primarily the ground. According to Newton's Third Law, a force acting on an object creates an action for which there is an equal and opposite reaction. Therefore, the forces exerted by the body through the legs resist the pull of gravity and serve to move the body forward, and they produce equal and opposite reaction forces by the ground. The ground reaction force is a 3-dimensional vector quantity, typically displayed in three orthogonal components defined by the walkway coordinate system. The profiles of the ground reaction force components reflect the dynamics of gait and are indicative of the accelerations imposed on the body's center of mass.

The orientation and magnitude of the ground reaction force vector is one of the primary determinants of the joint moments that are produced during the stance phase of walking. Joint moments are created about the ankle, knee and hip joint axes of rotation to either produce or oppose rotational motion during gait. Accurate moment calculations require knowledge about the masses of the limb segments, the location of the segmental centers of mass, and the distribution of mass within the segments. This information is generally not known for prosthetic limbs, so caution is advised when analyzing and interpreting joint moment data calculated for amputee gait.

Joint powers are probably one of the more popular parameters of interest in clinical gait analysis. The joint power during walking is equal to the product of the instantaneous joint moment and angular velocity. Joint powers can be particularly revealing when considering the contribution of the ankle and hip joints to forward progression, but limitations and assumptions of the model used to perform the power calculations can make interpretation difficult, particularly at the prosthetic ankle joint. Because passive prosthetic components cannot actively produce power, what appears to be power absorption and generation associated with ankle push-off in late prosthetic stance phase probably represents energy storage and return, respectively, from a prosthetic foot and/or ankle mechanism.

The vertical component of the ground reaction force is the largest, and most studied, of the three components, and has a characteristic double hump for able-bodied individuals walking at comfortable self-selected speeds. The first peak of the vertical ground reaction force is believed to be particularly important for analyzing shock absorption that occurs during the loading response phase of gait. Persons with unilateral transtibial amputations have been reported to walk with decreased vertical ground reaction forces under their prosthetic limb compared with able-bodied individuals, whereas the forces under their sound limb are slightly greater.43–45 Similarly, unilateral transfemoral amputees have been found to produce greater vertical ground reaction forces with their sound leg than with their prosthetic limb.35,45 Differences in the fore-aft ground reaction forces have also been observed between the prosthetic and sound sides of unilateral transtibial and transfemoral amputees, with forces under the sound limb typically having a larger magnitude.45 These differences become even more pronounced as speeds of walking increase, suggesting that amputees may walk more cautiously on their prostheses to reduce the magnitude of the transmitted force to their residual limbs. Alternatively, the bulk of residual limb soft tissue that interfaces with the socket may serve to decrease stiffness on the prosthetic side and thus automatically reduce the magnitude of ground reaction forces without conscious intervention by the prosthesis user. Some bilateral amputees with transtibial amputation of one leg and a knee disarticulation on the other were observed to walk with higher peak magnitudes of the vertical ground reaction force on their knee disarticulation side.46 Ground reaction forces may be higher for persons with a knee disarticulation due to the reduced bulk of soft tissue that provides load bearing through the residual limb, making the residual limb/prosthesis system stiffer than either a trans-tibial or transfemoral amputation would provide.

Measurements of the ground reaction forces have been used to analyze the effects of different shock absorbing components on gait and may provide some indirect indication of the force and pressure sensation that prosthetic users perceive on their residual limbs. Studies have demonstrated reduced peak forces under the prosthesis during running and other high impact activities when shock-absorbing components are placed in the prosthesis. In one study, differences in ground reaction forces and moments of force during the gait of a transtibial amputee were measured using the Re-Flex Vertical Shock Pylon (VSP; Ossur) as designed and with the shock absorber immobilized.47 Investigators found that when subjects walked with the VSP, the vertical and the fore-aft components of the ground reaction force on the prosthetic side were slightly decreased during the loading response phase, a period when shock absorption is critical. In another study, quantitative gait measures were used to analyze the effect of the Re-Flex VSP on two transtibial amputees who were walking and jogging in place.48 Few biomechanical differences were found for walking when comparing trials with and without the VSP, but increases in vertical ground reaction forces were noted during fast walking and jogging with the VSP. Though these results are counterintuitive, they probably relate to the ability of prosthesis users to readily adapt to prosthetic modifications. Subjects may have applied greater impact forces to their prostheses when using the VSP because of its increased shock absorption capabilities. In a study of unilateral transtibial amputee subjects walking at freely-selected slow and fast speeds with and without an Endolite TT (Telescopic-Torsion) Pylon (Endolite North America, Centerville, OH), few quantitative changes were found in kinematic and kinetic gait parameters.10 However, the investigators did observe a reduction in a force transient associated with impact loading during the prosthetic loading response phase, an effect that was more evident at walking speeds above about 1.3 m/sec when shock forces are greater.

In a study involving transtibial amputee subjects walking with a SACH foot and four different types of dynamic elastic response feet, similar ground reaction force patterns were observed among all of the feet.27 However, there was an increased dorsiflexion moment that occurred about the ankle joint during stance with a Flex Foot. In a similar study, measurements of the vertical ground reaction force were used to compare the gaits of unilateral transtibial amputees walking with a SACH foot, a Seattle foot and a Jaipur (Vhagwan Mahaveer Viklang Sahayata Samiti, Jaipur, India) foot.49 The SACH foot was found to provide better shock absorption (i.e., a smaller impact force peak) than the other two feet. Conversely, another study reported better shock absorption in the Seattle foot compared to the SACH.50 Prosthetic foot type does not generally appear to affect loading characteristics in the prosthetic limb, but it does appear to influence the ground reaction forces under the sound limb.25, 28, 51–53

Amputee subjects consistently demonstrate reduced ankle power in late prosthetic stance phase compared with able-bodied ambulators. In one study, ankle joint powers were analyzed in unilateral transtibial amputees to determine energy return from a SACH foot and two other energy-storing feet.9 The data showed that the energy-storing feet were able to return 20% to 30% more energy than the SACH foot. Another study showed greater energy absorption and return at the ankle by transtibial amputees walking with a Flex Foot compared with a SACH foot. Energy return by the Flex Foot was greater than with the other feet, but it had no effect on the pattern or magnitude of the knee and hip joint powers compared with the SACH foot.33 In a study that compared the gaits of unilateral amputees walking with a SAFE II foot and a Flex Foot, the Flex Foot provided greater power absorption (i.e., energy storage) during early to midstance, and a trend toward a greater plantarflexion moment and power generation (i.e., energy return) in late stance.54 Other studies have reported similar results, indicating that dynamic elastic response feet store and return more energy than SACH feet or other types of less dynamic feet.53,55–57 How this increased energy returned by dynamic elastic response feet is utilized by the prosthetic user for walking has not been sufficiently explained. In response to the reduced ankle powers associated with the prosthetic side, it appears that transtibial amputees increase power generation by the anatomical knee and hip joints of their residual limbs.58

Individuals with amputations may modify the way they walk based on the perception of the forces, moments, and pressures transmitted between their residual limbs and prostheses. A change in gait that is visually apparent can be confirmed with kinematic gait measures. However, kinematics provide no direct indication of the forces and moments imposed on the residual limb by the prosthesis. Kinetic measures of walking are useful because they convey information that cannot be discerned visually by an observer, and they may directly relate to what the prosthetic user perceives while they walk. Nonetheless, research studies have failed to demonstrate consistent results in kinetic measures—ground reaction forces, joint moments, and powers—from various investigations of amputee gait and prosthetic components. Some of the variability undoubtedly relates directly to the inconsistency of the kinematic data, which has already been mentioned. Therefore, kinetic measures do not appear to be reliable clinical outcome measures for assessing prosthetic gait performance.


When considering the different measurements employed in conducting a quantitative gait analysis, it is difficult to point to any single measure as being more important than the others. Temporal-spatial parameters provide fundamental gait characterization; kinematic data are useful for determining patterns of individual joint motion during gait; and kinetic data indicate how force is being transmitted to the leg, the influence of that force on joint rotation, and how the joints ultimately affect motion (i.e., joint moments and powers). These measures are typically taken together and used to characterize an individual's gait, but they do not directly indicate a person's efficiency or perceived effort during the activity. However, energy expenditure measurements may provide some of the most informative data for assessing walking efficiency because they represent an aggregate measure of overall walking performance that corresponds with the perceived effort required by an individual to ambulate. Because amputees are anatomically deficient they may require some form of compensatory action, or combination thereof, to achieve adequate, functional ambulation. The loss of one of the major joints of the lower extremity may be almost fully compensated for by exaggerated motions at other levels, but the loss of two or more major joints of the lower extremity makes effective compensation nearly impossible, and the cost of locomotion in terms of energy is increased with an inevitable drain upon ambulation economy.59

Energy expenditure may be expressed as a rate (i.e., energy expended per unit time) or as an energy cost (i.e., energy expended per unit distance). The energy expenditure rate is equal to the product of the energy cost and the steady-state walking speed. Energy expenditure measurements are typically made by indirect means using a face mask to capture respirations to monitor O2 consumption and CO2 production as subjects ambulate at a steady-state speed.

Some researchers have claimed that individuals possess a single speed of walking that corresponds to a minimum level of energy cost and at or near this most comfortable speed there is minimal use of muscles with maximal reliance on gravity and pendulum action. However, human beings are unlikely to have a natural or habitual mode of walking, although we may have a preferred mode in a given environmental situation.60 Most able-bodied adults are able to comfortably walk at speeds in the range of about 1.0 to 1.7 m/sec.61 Over this range, the energy cost curves are relatively flat, indicating that virtually uniform efficiency exists for normal gait throughout the comfortable range of walking speeds.

The energy cost has been shown to increase with more proximal levels of leg amputation, but the rate of energy expenditure appears to be relatively uniform across able-bodied and amputee individuals at freely-selected walking speeds.61–64 How is this possible? Because amputees naturally adopt slower speeds when they walk. By lowering their walking speeds, prosthesis users are able to keep their rate of energy expenditure within normal limits, but the reduced speed consequently increases their energy cost. Significantly higher heart rates have been observed in transtibial amputees compared with normal subjects when they walked at their freely-selected speeds, but the oxygen consumption rates are comparable between the two groups.65 At freely-selected speeds, unilateral transtibial amputees have been reported to expend approximately 9% more energy than able-bodied individuals, unilateral transfemoral amputees approximately 49% more, and bilateral transfemoral amputees nearly 280% more.66 The residual limb length has been determined to have a significant effect on metabolic cost when transtibial amputees are stratified by long and short stump length, but prosthesis mass does not appear to significantly alter ambulation energy consumption when other variables are controlled.65 Transtibial amputees with longer residual limbs have been shown to expend less energy during walking than those with shorter residual limbs.65,67

Energy expenditure studies have been used to evaluate and compare different types of prosthetic components, particularly prosthetic feet. The inherent assumption underlying these studies is that dynamic elastic response feet should return more energy at the end of stance phase than other types of feet, and presumably the increased energy returned may somehow be effectively utilized to reduce the metabolic energy required by the prosthesis user to ambulate. Some studies report that dynamic elastic response feet reduce the energy required to walk compared with a SACH foot.24,68 However, other studies have found that there are no differences in the energy rate or cost for transtibial amputees walking with the SACH foot and different types of dynamic elastic response feet.25,27,69 Transfemoral amputees have been reported to reduce their energy expenditure when they walk with a Flex Foot compared with a SACH foot.24 In a comparison of vascular versus traumatic unilateral amputees walking with six different prosthetic feet, significantly greater energy cost was measured in vascular amputees compared with traumatic amputees.21 However, there were no significant differences in energy cost between the different prosthetic feet for either group. The inconsistency in energy expenditure measures between studies such as these that have similar experimental protocols is of concern, and is currently without a satisfactory explanation.

Other types of prosthetic components have also been shown to influence the energy expenditure of walking in prosthesis users. The Re-Flex VSP has been observed to significantly reduce energy cost, increase gait efficiency, and decrease exercise intensity in subjects with unilateral trans-tibial amputations.70 However, the benefits were found to be speed dependent, with the advantages becoming more apparent at speeds above about 1.1 m/sec. Similarly, the Endolite TT Pylon has been reported to reduce energy expenditure in transtibial amputees at fast walking speeds.71 These observed effects of shock absorbing components on energy expenditure may be unanticipated, but one possible explanation is that the increased shock absorption may make faster walking speeds more comfortable, so the amputee compensates less in order to reduce shock and thereby decreases muscular effort and reduces metabolic demand. Studies that have compared the Contoured Adducted Trochanteric-Controlled Alignment Method (CAT-CAM) socket to a quadrilateral socket in transfemoral amputees have reported that subjects expend considerably less energy when walking with the CAT-CAM socket.72,73 Surprisingly, no significant differences were found in the energy expenditure of a group of unilateral transfemoral amputees walking with their prosthetic knee locked and unlocked.74

Explanations for the observed differences in energy expenditure between different types of prosthetic components will require further investigation and a better understanding about prosthetic walking mechanics. Compared with temporal-spatial, kinematic, and kinetic data, the energy expenditure measures are appealing as an outcome measure for prosthetic gait because they directly relate to walking efficiency and to subject perception.


Quantitative gait analyses are useful for characterizing the motions of walking and for documenting progress as a person undergoes rehabilitation. They can be beneficial for verifying visual observation, providing objective measurements that substantiate a subjective assessment. Assuming that the able-bodied pattern of walking represents the most efficient bipedal form of ambulation given the constraints of human anatomy, then quantitative gait analyses may be of tremendous benefit for identifying and addressing differences in a prosthetic user's gait from normal. However, at present, quantitative gait analysis does not appear to be particularly useful for assisting the prosthetist with an individual's prosthetic prescription or for diagnosing problems with the prosthesis, because at the moment we possess insufficient knowledge about how to adequately address functional deficiencies in prosthetic ambulation. Further research is required to identify the functional requirements of able-bodied walking, which can be used to establish appropriate criteria for evaluating prosthetic gait.

Many studies that have compared different types of prosthetic components generally report little or no difference in the data obtained from quantitative gait analyses, and when improvements are observed, they tend to be incremental in nature and are often not statistically significant. It is unlikely that one component will demonstrate a dramatically improved pattern of walking in amputee subjects, even when the most sophisticated technologies are compared with the simplest component designs. These results suggest that the types of components being compared may actually be quite similar in function, even though their modes of operation may be significantly different. If different prosthetic component designs produce gait characteristics in subjects that are similar, but are consistently inferior to that of able-bodied persons, then perhaps the primary causative factors behind the poor walking performance are being overlooked. Significant improvements in prosthetic gait performance may not be realized until a new generation of components is developed that more accurately reproduces the functions used in able-bodied walking.

There should be greater consideration for the testing environment when evaluating prosthetic walking performance. Well-lighted motion analysis laboratories with level walkways are probably not the best conditions for evaluating the effects of prosthetic components that may be designed for improving ambulation on other types of walking surfaces or in more demanding circumstances. More meaningful results, with increased validity, would probably be obtained by developing and utilizing gait measurement systems that enable different components to be compared in “real world” situations outside of the gait laboratory.

Many of the published quantitative gait studies seem to indicate that amputees are able to readily adapt to changes in their prostheses. Most prosthetists have probably noticed this effect in the clinic—they fit their client with a new prosthetic component, and visual gait analysis indicates no discernible difference from a previous prosthetic configuration. Experienced prosthetic users are able to readily adapt to minor, and major, prosthetic modifications. In those cases, their gait does not usually improve and they will often display the same gait pattern as before. Therefore, greater attention must be paid to gait training for new amputees, and retraining for more experienced users to break their bad habits and to instill proper form. For prosthetic users to achieve maximum benefit from new technology, they must be taught how to walk with their prostheses in such a manner that they take full advantage of the design features. Good gait requires that the user develop trust, security and confidence in his or her prosthesis, learning one's capabilities and identifying limitations, all of which take time and experience.

Even though statistically significant differences among components are usually not detected using quantitative gait measures, subjects often express clear preference for one component over another, suggesting that very subtle changes in gait may be detected by the user and be perceived as significant.27 Further exploration and analysis are required to unravel the complex relationship that exists among quantitative gait data, clinical observation and patient perception of the prosthesis.75 The inability to detect changes between prosthetic configurations using quantitative gait analysis is not a limitation of current motion measurement systems; they are able to measure body motion and forces with sufficient accuracy. The problems we are currently struggling with concern our lack of understanding about how to best restore ambulation ability in someone with a leg amputation, how to provide sufficient function through prosthetic design and with appropriate selection of componentry, and how to best use and incorporate quantitative gait data with visual observation and subjective feedback to effect substantive improvement in the function, aesthetics, and efficiency of prosthetic gait.

Quantitative gait analysis is recognized as being useful for providing an objective assessment about the way a person walks. Studies of prosthetic users reported in the literature tend to indicate a lack of consistency in quantitative gait measures, even in similar populations of amputee subjects who are walking with comparable prosthetic configurations. Therefore, the value of using these data individually for outcome measures is questionable. Energy expenditure measures, as a gross indicator of walking performance, tend to show some promise as reliable outcome measures for the evaluation of prosthetic gait, though it is not possible to readily identify specific gait abnormalities that may be evident in temporal-spatial, kinematic or kinetic data. For the time being, quantitative gait analysis may be best used in the research laboratory as opposed to the clinic, but it is important that we continue to strive to effectively integrate these measurements with the experience and skill of the prosthetist and the subjective feedback of the prosthetic user.


1. Gage JR, Hicks R. Gait analysis in prosthetics. J Gerontol 1985;39:662–666.
2. Radcliffe C. Functional considerations in the fitting of above-knee prostheses. In: Wilson AB, ed. Selected Articles from Artificial Limbs. Huntington, NY: Robert E. Krieger Publishing Co. Inc.;1970:5–30.
3. Edelstein JE. Prosthetic and orthotic gait. In: Smidt GL, ed. Gait in Rehabilitation. New York: Churchill Livingston Inc.;1990:281–300.
4. Saleh M. Alignment and gait optimization in lower-limb amputees. In: Murdoch, G, Donovan RG, eds. Amputation Surgery & Lower Limb Prosthetics. Oxford: Blackwell Scientific Publications;1988:357-366.
5. Kleissen RFM. Factors in clinical application of gait analysis.. J Rehabil Sci 1990;3:106–109.
6. Murray MP. Gait as a total pattern of movement. Am J Phys Med 1967;46:290–333.
7. Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;8:383–392.
8. Miff SC. Gait initiation and termination in non-disabled ambulators and in people with unilateral lower limb loss. [Thesis]. Evanston, IL: Northwestern University, 2004.
9. Winter DA, Sienko SE. Biomechanics of below-knee amputee gait. J Biomech 1988;21:361–367.
10. Gard SA, Konz RJ. The effect of a shock-absorbing pylon on the gait of persons with unilateral transtibial amputation.. J Rehabil Res Dev 2003;40:109–124.
11. James U, Oberg K. Prosthetic gait pattern in unilateral above-knee amputees. Scand J Rehabil Med 1973;5:35–50.
12. Murray MP, Sepic SB, Gardner GM, Mollinger LA. Gait patterns of above-knee amputees using constant friction knee components. Bull Prosthet Res 1980;17:35–45.
13. Boonstra AM, Fidler V, Eisma WH. Walking speed of normal subjects and amputees: aspects of validity of gait analysis. Prosthet Orthot Int 1993;17:78–82.
14. Andriacchi TP, Ogle JA, Galante JO. Walking speed as a basis for normal and abnormal gait measurements. J Biomech 1977;10:261–268.
15. Murray MP, Drought AB, Kory RC. Walking Patterns of Normal Men. J Bone Joint Surg Am 1964;46:335–360.
16. Inman VT, Ralston HJ, Todd F. Human Walking, Baltimore, MD: Williams & Wilkins; 1981.
17. Hermodsson Y, Ekdahl C, Persson BM, Roxendal G. Gait in male trans-tibial amputees: a comparative study with healthy subjects in relation to walking speed. Prosthet Orthot Int 1994;18:68–77.
18. Drillis R. Objective recording and biomechanics of pathological gait. Ann N Y Acad Sci 1958;74:86–109.
19. Gauthier-Gagnon C, Gravel D, St-Amand H, et al. Changes in ground reaction forces during prosthetic training of people with transfemoral amputations: a pilot study. J Prosthet Orthot 2000;12:72–77.
20. Jaegers SM, Arendzen JH, de Jongh HJ. Prosthetic gait of unilateral transfemoral amputees: a kinematic study. Arch Phys Med Rehabil 1995;76:736–743.
21. Barth DG, Schumacher L, Sienko-Thomas S. Gait analysis and energy cost of below-knee amputees wearing six different prosthetic feet. J Prosthet Orthot 1992;4:63–75.
22. Hannah RE, Morrison JB, Chapman AE. Prostheses alignment: effect on gait of persons with below-knee amputations. Arch Phys Med Rehabil 1984;65:159–162.
23. Nielsen DH, Shurr DG, Golden JC, Meier KG. Comparison of energy cost and gait efficiency during ambulation in below-knee amputees using different prosthetic feet: a preliminary report. J Prosthet Orthot 1989;1:24–31.
24. Macfarlane PA, Nielsen DH, Shurr DG, et al. Transfemoral amputee physiological requirements: comparisons between SACH foot walking and Flex-foot walking. J Prosthet Orthot 1997;9:138–143.
25. Lehmann JF, Price R, Boswell-Bessette S, et al. Comprehensive analysis of dynamic elastic response feet: Seattle Ankle/Lite Foot versus SACH foot. Arch Phys Med Rehabil 1993;74:853–861.
26. Perry J, Boyd LA, Rao SS, Mulroy SJ. Prosthetic weight acceptance mechanics in transtibial amputees wearing the Single Axis, Seattle Lite, and Flex Foot. IEEE Trans Rehabil Eng 1997;5:283–289.
27. Torburn L, Perry J, Ayyappa E, Shanfield SL. Below-knee amputee gait with dynamic elastic response prosthetic feet: a pilot study.. J Rehabil Res Dev 1990;27:369–384.
28. Snyder RD, Powers CM, Fontaine C, Perry J. The effect of five prosthetic feet on the gait and loading of the sound limb in dysvascular below-knee amputees.. J Rehabil Res Dev 1995;32:309–315.
29. Godfrey CM, Jousse AT, Brett R, Butler JF. A comparison of some gait characteristics with six knee joints. Orthot Prosthet 1975;29:33–38.
30. Murray MP, Mollinger LA, Sepic SB, et al. Gait patterns in above-knee amputee patients: hydraulic swing control vs constant-friction knee components. Arch Phys Med Rehabil 1983;64:339–345.
31. Boonstra AM, Schrama J, Fidler V, Eisma WH. Energy cost during ambulation in transfemoral amputees: a knee joint with a mechanical swing phase control vs a knee joint with a pneumatic swing phase control. Scand J Rehabil Med 1995;27:77–81.
32. Isakov E, Susak Z, Becker E. Energy expenditure and cardiac response in above-knee amputees while using prostheses with open and locked knee mechanisms. Scand J Rehabil Med Suppl 1985;12:108–111.
33. Gitter A, Czerniecki JM, DeGroot DM. Biomechanical analysis of the influence of prosthetic feet on below-knee amputee walking. Am J Phys Med Rehabil 1991;70:142–148.
34. Breakey JW. Gait of unilateral below-knee amputees. Orthot Prosthet 1976;30:17–24.
35. Lewallen R, Quanbury AO, Ross K, Letts R. A biomechanical study of normal and amputee gait, Champaign, Illinois: Human Kinetics Publishers; 1985.
36. Zuniga EN, Leavitt LA, Calvert JC, et al. Gait patterns in above-knee amputees. Arch Phys Med Rehabil 1972;53:373–382.
37. Judge GW, Fisher L. A bouncy knee for above-knee amputees. Eng Med 1981;10:27–31.
38. Blumentritt S, Scherer HW, Wellershaus U, Michael JW. Design principles, biomechanical data and clinical experience with a polycentric knee offering controlled stance phase knee flexion: a preliminary report. J Prosthet Orthot 1997;9:18–24.
39. Koehler SR, Gard SA, Meier MR, et al. Stance-phase knee flexion in persons with unilateral transfemoral amputations walking on an Otto Bock 3R60 EBS Knee: A preliminary report. Paper presented at the 9th Annual Meeting of the Gait and Clinical Movement Analysis Society (GCMAS), Lexington, KY, April 21–24, 2004.
40. Michaud SB, Gard SA, Childress DS. A preliminary investigation of pelvic obliquity patterns during gait in persons with transtibial and transfemoral amputation.. J Rehabil Res Dev 2000;37:1–10.
41. Song KM, Halliday SE, Little DG. The effect of limb-length discrepancy on gait. J Bone Joint Surg Am 1997;79:1690–1698.
42. Cappozzo A, Figura F, Gazzani F, et al. Angular displacements in the upper body of AK amputees during level walking. Prosthet Orthot Int 1982;6:131–138.
43. Lewallen R, Dyck G, Quanbury A, et al. Gait kinematics in below-knee child amputees: a force plate analysis. J Pediatr Orthop 1986;6:291–298.
44. Engsberg JR, Lee AG, Patterson JL, Harder JA. External loading comparisons between able-bodied and below-knee-amputee children during walking. Arch Phys Med Rehabil 1991;72:657–661.
45. Michaud SB. Prosthetic vs. standard walking: Quantitative observations. [Thesis] Evanston, IL: Northwestern University, 1997.
46. Pinzur MS, Smith D, Tornow D, et al. Gait analysis of dysvascular below-knee and contralateral through-knee bilateral amputees: a preliminary report. Orthopedics 1993;16:875–879.
47. Mooney J, Hill S, Supan TJ, Barth DG. Comparison of floor reaction and rotational forces in the gait of a transtibial amputee using a Re-Flex VSP Flex Foot design: A pilot study. Paper presented at the 21st Annual Meeting & Scientific Symposium of the AAOP, New Orleans, LA, March 21–25, 1995.
48. Miller LA, Childress DS. Analysis of a vertical compliance prosthetic foot.. J Rehabil Res Dev 1997;34:52–57.
49. Arya AP, Lees A, Nirula HC, Klenerman L. A biomechanical comparison of the SACH, Seattle and Jaipur feet using ground reaction forces. Prosthet Orthot Int 1995;19:37–45.
50. Murray D, Hartvikson W, Anton H, et al. With a spring in one's step. Clin Prosthet Orthot 1988;12:128–135.
51. Powers CM, Torburn L, Perry J, Ayyappa E. Influence of prosthetic foot design on sound limb loading in adults with unilateral below-knee amputations. Arch Phys Med Rehabil 1994;75:825–829.
52. Czerniecki JM, Gitter AJ. Gait analysis in the amputee: has it helped the amputee or contributed to the development of improved prosthetic components? Gait Posture 1996;4:258–268.
53. van der Linden ML, Solomonidis SE, Spence WD, et al. A methodology for studying the effects of various types of prosthetic feet on the biomechanics of trans-femoral amputee gait. J Biomech 1999;32:877–889.
54. Underwood HA, Tokuno CD, Eng JJ. A comparison of two prosthetic feet on the multi-joint and multi-plane kinetic gait compensations in individuals with a unilateral trans-tibial amputation. Clin Biomech (Bristol, Avon) 2004;19:609–616.
55. Czerniecki JM, Gitter A, Munro C. Joint moment and muscle power output characteristics of below knee amputees during running: the influence of energy storing prosthetic feet. J Biomech 1991;24:63–75.
56. Macfarlane PA, Nielsen DH, Shurr DG. Mechanical gait analysis of transfemoral amputees: SACH Foot versus the Flex-foot. J Prosthet Orthot 1997;9:144–151.
57. Postema K, Hermens HJ, de Vries J, et al. Energy storage and release of prosthetic feet, I: biomechanical analysis related to user benefits. Prosthet Orthot Int 1997;21:17–27.
58. Sadeghi H, Allard P, Duhaime PM. Muscle power compensatory mechanisms in below-knee amputee gait. Am J Phys Med Rehabil 2001;80:25–32.
59. Saunders JB, Inman VT, Eberhart HD. The major determinants in normal and pathological gait. J Bone Joint Surg 1953;35A:543–558.
60. Grieve DW. Gait patterns and the speed of walking. Bio-Med Engr 1968;3:119–122.
61. Waters RL, Yakura JS. The energy expenditure of normal and pathologic gait. Crit Rev Phys Rehab Med 1989;1:183–209.
62. James U. Oxygen uptake and heart rate during prosthetic walking in healthy male unilateral above-knee amputees. Scand J Rehabil Med 1973;5:71–80.
63. Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. J Bone Joint Surg Am 1976;58:42–46.
64. Jaegers SM, Vos LD, Rispens P, Hof AL. The relationship between comfortable and most metabolically efficient walking speed in persons with unilateral above-knee amputation. Arch Phys Med Rehabil 1993;74:521–525.
65. Gailey RS, Wenger MA, Raya M, et al. Energy expenditure of trans-tibial amputees during ambulation at self-selected pace. Prosthet Orthot Int 1994;18:84–91.
66. Huang CT, Jackson JR, Moore NB, et al. Amputation: energy cost of ambulation. Arch Phys Med Rehabil 1979;60:18–24.
67. Gonzalez EG, Corcoran PJ, Reyes RL. Energy expenditure in below-knee amputees: correlation with stump length. Arch Phys Med Rehabil 1974;55:111–119.
68. Schneider K, Hart T, Zernicke RF, et al. Dynamics of below-knee child amputee gait: SACH foot versus Flex foot. J Biomech 1993;26:1191–1204.
69. Torburn L, Powers CM, Guiterrez R, Perry J. Energy expenditure during ambulation in dysvascular and traumatic below-knee amputees: a comparison of five prosthetic feet.. J Rehabil Res Dev 1995;32:111–119.
70. Hsu MJ, Nielsen DH, Yack HJ, Shurr DG. Physiological measurements of walking and running in people with transtibial amputations with 3 different prostheses. J Orthop Sports Phys Ther 1999;29:526–533.
71. Buckley JG, Jones SF, Birch KM. Oxygen consumption during ambulation: comparison of using a prosthesis fitted with and without a tele-torsion device. Arch Phys Med Rehabil 2002;83:576–580.
72. Gailey RS, Lawrence D, Burditt C, et al. The CAT-CAM socket and quadrilateral socket: a comparison of energy cost during ambulation. Prosthet Orthot Int 1993;17:95–100.
73. Flandry F, Beskin J, Chambers RB, et al. The effect of the CAT-CAM above-knee prosthesis on functional rehabilitation. Clin Orthop 1989;:249–262.
74. Traugh GH, Corcoran PJ, Reyes RL. Energy expenditure of ambulation in patients with above-knee amputations. Arch Phys Med Rehabil 1975;56:67–71.
75. Hafner BJ, Sanders JE, Czerniecki J, Fergason J. Energy storage and return prostheses: does patient perception correlate with biomechanical analysis? Clin Biomech (Bristol, Avon) 2002;17:325–344.
© 2006 American Academy of Orthotists & Prosthetists