Secondary Logo

Journal Logo

Correlation of Transtibial Prosthetic Alignment Quality and Step-by-Step Variance of Gait

Fiedler, Goeran PhD; Johnson, Mariah Susan MSPO

Journal of Prosthetics and Orthotics: January 2017 - Volume 29 - Issue 1 - p 19–25
doi: 10.1097/JPO.0000000000000113
CME Article
Free
CME

Introduction The quality of lower-limb prosthetic alignment cannot easily be measured directly but may be described by its effects on gait comfort and efficiency. It is known that gait stability and step variability are correlated, as are gait stability and prosthetic alignment. This study investigated the hypothesis that prosthetic alignment and step variability are correlated. This would have the implication that step variability can be measured to assess the quality of prosthetic alignment (and possibly other relevant factors such as prosthetic fit and componentry selection).

Materials and Methods Twelve experienced users of transtibial prostheses were subjected to a protocol that introduced malalignments of their prosthetic ankle plantarflexion angle in a randomized crossover design. Perceived alignment quality was recorded via a visual analog scale. Step-by-step variability in horizontal ground reaction forces and axial ankle torsion moment was measured using a prosthesis-integrated load cell and was statistically compared with the degree of misalignment by bivariate correlation analysis.

Results Findings suggest that variance in axial torsion moment and step duration may be correlated to alignment quality.

Conclusions Subjective patient feedback is the recommendable criterion for alignment assessment in active and experienced users of prostheses. Further research is recommended before step variance may help assess prosthetic alignment quality in patients with less experience in prosthetic use.

GOERAN FIEDLER, PhD, and MARIAH SUSAN JOHNSON are affiliated with the University of Pittsburgh, School of Health and Rehabilitation Sciences, Prosthetics & Orthotics, Pennsylvania.

Funding for this work was received from the American Orthotic and Prosthetic Association.

Disclosure: The authors declare no conflict of interest.

Correspondence to: Goeran Fiedler, PhD, Department of Rehabilitation Science and Technology University of Pittsburgh, Suite 403, Bakery Square, Pittsburgh, PA 15206; email: gfiedler@pitt.edu

The alignment of an artificial limb, that is, the spatial orientation of the different components with respect to each other, is an important aspect of overall function and performance of the prosthesis. The alignment is commonly optimized in two steps, starting with the static bench alignment based on patient-specific limb shape and joint mobility, as well as manufacturer recommendations for the used componentry. This initial alignment is later dynamically optimized to facilitate efficient and comfortable locomotion with the prosthesis. A variety of tools and methods has been proposed to inform the process of alignment optimization.1–4 However, many clinicians still rely on subjective measures, such as patient feedback, gait observation, and experience, to avoid malalignments. Particularly in lower-limb prosthesis, the effects of malalignments can be severe, ranging from reduced biomechanical gait efficiency to static instability that can cause frequent stumbling or falling.5–9

Safety is a main source for concern in users of lower-limb prosthesis.10 The real or perceived instability of an artificial leg can have various detrimental effects that go beyond the injury risks associated with accidental falls.11 Patients may decide to underuse their prostheses, leading to increased risks of overusing the remaining body structure, such as the nonamputated leg or the upper limbs when crutches are used instead of prosthesis.12 Patients may also adapt a relieving posture, intended to limit load bearing on the prosthesis, which in turn contributes to a decrease in mobility, for example, gait speed and range. Active patients may be capable of compensating for reduced gait stability, which, however, involves elevated physical and mental loads6 and which therefore limits their capabilities to focus on tasks other than walking.13

Assessment of gait stability is thus an important objective in the rehabilitation of amputation patients. Various approaches to that assessment have been proposed and partly been integrated into clinical practice: Monitoring of fall incidents has been used as a direct outcome14 yet can be criticized for its nonpredictive nature. Ideally, a patient would not have to fall down first before an increased fall risk can be diagnosed. Much research work has been dedicated to the identification of fall risk factors15 that are subsequently combined into assessment scales to be used by physicians and other medical personnel.16 Although scales of that kind are commonly validated in a different population with high fall risk—elderly people—many of the principles are applicable to the population of persons with amputation as well.17 Another category of assessment methods is based on performance measures18 and instrumented gait analysis.19 Variables that have been investigated to determine fall risk include accelerometer data,20,21 functional reach range,22 gait symmetry,23 and temporal spatial variability.24,25

A review of the recent literature on the topic allows the conclusion that step-by-step variability is a valuable variable for the prediction of fall risk in prosthetic users.23,26,27 Data from accelerometers or instrumented walkways have been correlated to fall histories in subjects.28,29 The effect of prosthetic misalignments on various biomechanics variables has been investigated.9,30 However, the clinically relevant relationship between prosthetic misalignment and (easily measurable) step-by-step variability had not yet been investigated. If such a relationship could be identified, it would provide a useful outcome measure for the evaluation of prosthetic alignment quality and possibly other relevant factors such as prosthetic fit and componentry selection.

The objective of this work was to investigate the correlation between prosthetic alignment quality and step-by-step variability. We hypothesized that objective and subjective quality of prosthetic alignment (the first being measurable by the degree of deviation from the optimized alignment, the second being determined by the individual patient's perception31) is correlated to the variance in unilateral temporal and kinetics gait variables, specifically in horizontal ground reaction forces (GRFs) and axial torsion moments of the prosthetic leg. The rationale for investigating these variables is based on the assumption that step length and width will be less steady as a result of the misalignments, thus triggering greater step-by-step variations in the resulting forces and moments.

Back to Top | Article Outline

METHODS

SUBJECTS

A sample of persons with unilateral transtibial amputation was recruited. Inclusion criteria were an activity classification of K2 to K4, experience in prosthetic use of at least 1 year, and the stated ability to walk pain-free for at least 30 minutes. These inclusion criteria were defined to facilitate the intended data collection. However, the principle of gait stability assessment appears to be transferable to patients regardless of prosthetic history, activity, or amputation level. Exclusion criteria were an age of younger than 18 years, acute or ongoing residual limb pain, as well as use of a prosthesis that was not suitable for installation of measurement equipment (e.g., by being too short or having nonremovable cosmesis). Recruiting methods and protocol were approved by the University of Pittsburgh's institutional review board.

A sample size of 10 subjects was targeted. As this research is exploratory in nature, no comparable data existed to conduct a power analysis. The chosen sample does, however, compare favorably to sample sizes in published studies on prosthetic alignment.32 The study used a within-subjects design, where every subject serves as his/her own control, making sample size requirements less critical for statistical analysis and mainly of concern for the practical generalizability of any findings.

Back to Top | Article Outline

EQUIPMENT

A six-degrees-of-freedom prosthesis-integrated load cell (iPecs mobile gait lab, RTC Electronics, Ann Arbor, MI, USA) was used for mobile data collection. It allows continuous measurement of triaxial forces and moments independent of laboratory equipment. Data can be streamed wirelessly to a computer or stored within the unit. The data quality is comparable with conventional gait analysis data.33

Back to Top | Article Outline

PROTOCOL

At the day of the data collection session, upon providing informed written consent to participate in the study, subjects were asked to complete the 12-item PLUS-M questionnaire34 to determine mobility levels, which supplement demographic and anthropometric information that were likewise recorded, and subjects were prepared for a gait analysis. Participants' prostheses were temporarily modified by replacing the pylon adapter between socket and foot with the load cell and a respectively shorter pylon, while prosthetic foot component, alignment, and length were maintained. The exact position of the load cell with respect to the knee and ankle axes was measured and documented. Subjects were asked to don their prosthesis and walk for about 10 minutes to become familiar with the slightly changed weight and weight distribution of the prosthesis. At the same time, both the subject and a trained clinical prosthetist verified the sufficiency of the alignment. If necessary, alignment rectifications were performed. Studies on the effects of added prosthetic weight35,36 as well as experience from previous iPecs data collection37–39 suggest that the effect of the added weight is negligible (The iPecs unit adds 277 g to the weight of the prosthesis, part of which is offset by the lower weight of the shorter pylon adapter).

Subjects were asked to walk normally and in their preferred gait speed back and forth along a hallway, covering 25 m on carpet flooring and another 25 m on concrete floor, resembling flooring conditions commonly found in real-life environments. Subjects were allowed to use walking aids if they preferred to do so. During gait trials, continuous data were collected from the integrated load cell at 100 Hz and streamed to a SD memory card within the unit. The sampling frequency allowed sufficient oversampling of the gait signal whose main frequency content was reported to be below 10 Hz.40

Between passes through the hallway, the sagittal plane alignment of the prosthesis was altered by changing the ankle plantar/dorsiflexion angle by 3°, 6°, and 9° successively. Subjects rested for at least 2 minutes between trials. After every pass, including concrete and carpet flooring, subjects were asked to rate the perceived prosthetic alignment quality on a visual analogue scale (VAS). Each VAS score was between 0 and 10, with 10 being the best. The average subject score was then computed for each ankle misalignment.

The six different ankle misalignments (three each in plantarflexion and dorsiflexion direction) as well as the normal alignment setting were introduced in a randomized order so that results would not be biased by fatigue of subjects. Alignment increments of 3° have been selected, as perturbations of this magnitude have been shown to be subtle but noticeable to prosthetic users.41,42 Possible alignment changes in axial rotation and transverse translation were not included in an effort to limit the complexity and thereby the time requirements of the protocol, as well as effects of subject's fatigue. The subjects were blinded as to what ankle misalignment was being tested.

Back to Top | Article Outline

POSTPROCESSING AND ANALYSIS

Peak GRFs in the horizontal plane (Fx and Fy) as well as peak axial torsion moments (Mz) and step duration were extracted for 10 consecutive intermediary steps for each testing condition. To identify the steps for analysis, the entire data file was sequenced by finding the points where subjects stopped and turned around. The steps (commonly between 15 and 25) between two consecutive turns were associated with one pass along the hallway on either concrete or carpet, the sequence of which had been documented in the protocol. An appropriate number of both the first and last steps from each pass were disregarded to shorten the sample to the desired 10 steps. Total range, as well as variance (Var

, where μ is the sample mean, as the base for calculating standard deviation), over the 10-step sample was extracted as dependent variables for statistical analysis. Range and variance was determined for the two horizontal GRFs and the axial moment, leaving a total of 6 variables to be compared across alignment conditions and floor surfaces. Alignment quality, as the independent variable, was expressed as the amount of deviation from the individual initial alignment (for consistency reported in degrees plantarflexion, where values smaller than 0 denote the respective increase in dorsiflexion). Perceived alignment quality, based on the subjective ratings of the VAS, was treated as dependent variable as well. Alignment quality was assumed to be independent of the floor surface, and perceived alignment quality was thus recorded only once for each perturbation.

For purposes of analysis, variance data were plotted for visual representation of any correlations with alignment quality. To investigate the hypothesis that our measurement variables and prosthetic alignment quality are correlated, bivariate correlation analysis was conducted and correlation coefficients calculated, separately for each tested flooring surface. For that purpose, alignment perturbations were categorized in four groups according to their severity: 0°, 3°, 6°, and 9° of misalignment.

Back to Top | Article Outline

RESULTS

Thirteen subjects were recruited for participation in this study. One could not complete data collection because of scheduling conflicts. Data of the remaining 12 subjects (Table 1) are reported here. Ten of these participants were men and two were women.

Table 1

Table 1

In several subjects, the initial (optimized) alignment did not coincide with a neutral position of the pyramid connector, which limited the attainable alignment range and prevented all of the planned ankle misalignments from being realized. Only five of the prostheses could be subjected to all seven different alignments (between +9° and −9° ankle flexion). In nine prostheses, it was possible to realize the five alignments between +6° and −6°. Alignment range was even further limited in the remaining three prostheses, which resulted in partially missing data.

All 12 subjects rated the perceived quality of each different ankle misalignment, including the baseline. Plotting step variance directly against subjective ratings of alignment quality showed only weak correlations. A representative example is the variable peak Mz, the axial torque during stance phase (Figure 1).

Figure 1

Figure 1

The data demonstrate a strong correlation between perceived alignment quality and objective degree of misalignment (Figure 2). Bivariate correlation analysis resulted in a correlation coefficient R2 of 0.98 (p = 0.001). Other correlations between measurement variables and alignment quality were not significant (Table 2). Given the strong correlation between perceived and actual alignment quality in this sample, only one of them (actual alignment quality) was used as independent variable in subsequent analyses.

Figure 2

Figure 2

Table 2

Table 2

Of the five subjects who successfully underwent all six different ankle misalignments, one was a woman and four were men. The demographic and anthropometric characteristics were representative of the overall sample.

In this subsample, the variance of the variable Fy range was found to be correlated with alignment quality in that higher variances were measured for severe misalignments than for less severe or neutral misalignments similarly than perceived alignment quality. A notable exception exists at the −9° plantarflexion setting (Figure 3). R2 was computed as 0.472 (p = 0.339) with all alignments included. When the outlier for the −9° alignment was removed, R2 increased to 0.869 (p = 0.025).

Figure 3

Figure 3

When the most extreme ankle maladjustments (−9° and +9°) were disregarded for analysis, nine subjects completed all trials. This resulted in a somewhat stronger trend toward increased standard deviation with greater ankle malalignment, as shown for the variable step duration variance (Figure 4). Correlation between step duration and alignment quality in this sample was strong (R2 = 0.732) yet not statistically significant (p = 0.160).

Figure 4

Figure 4

Back to Top | Article Outline

DISCUSSION

Hypothesized correlations between prosthetic alignment quality and measurable kinetics variables of step-by-step variance could be found only in a subset of the collected data, after the exclusion of subjects who did not complete the entire protocol. Investigated correlations for the whole sample were mostly weak or insignificant, thus providing insufficient support for the study hypothesis.

Analysis and interpretation of the presented data were mainly motivated by the clinical question of whether the investigated method would allow an objective quantification of individual prosthetic alignment quality. For that purpose, measurement variables would have to reflect the different changes made to the alignment. The conventional understanding of the binary nature between acceptability and unacceptability of prosthetic alignment is represented by the line of acceptability in Figure 2. Assuming that only the highest rated setting is considered acceptable, the acceptability threshold was located between the first and second most positive scores. The convention that “neutral” (e.g., optimized) alignment was the only acceptable one was upheld in drawing that line for the interpretation of the plotted force and moment variances. Whereas for some variables the distinction could not be clearly made, such as shown in Figure 3, where the measured step variance for one of the more extreme misalignments resembled the optimal alignment setting, the hypothesis appeared better supported by other variables, such as the variance in step duration (Figure 4) and torsion moment during stance (Figure 1 and Table 2).

The VAS scores in Figure 2 show a clear patient preference for the neutral ankle alignment, with successively worse ratings corresponding to both increased dorsiflexion and plantarflexion. Investigating this correlation was not part of the study hypothesis, yet the graph seems to confirm the commonly accepted notion that acceptability of alignment to the users of prostheses is a valid and useful criterion in assessing the quality of a prosthesis' alignment. Measured against this criterion, the predictive power of the investigated variables of step variability appears to be clearly limited. However, it should be considered that the subjective criteria of perceived alignment quality may have been affected by a number of factors that are difficult to account for: The subjects participating in this study were mostly very experienced in using prosthetic limbs and were therefore more likely to provide pertinent assessments of their alignment than, for instance, patients who are being fitted with their first prosthesis after limb loss. Furthermore, although subjects were not told which of the alignment changes were implemented at any given step of the protocol, in an effort to blind them to the intervention, they were able to observe the changes that were done to their ankle alignment. Given the possible familiarity with the function principle of pyramid adapters, this may have biased them in their subsequent assessment of the alignment quality. This could have affected the VAS scores as subjects would naturally want to rate the quality of extreme misalignments lower. Against that background, it appears reasonable to consider the merit of the objective assessment criteria that have been investigated in this study.

Bivariate correlation analysis yielded the statistically significant correlation depicted in Figure 1. However, the respective correlation coefficient of 0.3 suggests a weak relationship, which limits the relevance of this finding for clinical purposes. Without a strong correlation, it is practically impossible to accurately discern between different levels of alignment quality based on the measurement variable.

The relationship between prosthetic alignment and horizontal GRFs has been investigated by different researchers before. Our findings compare to those of Geil and Lay,43 who investigated the effect of alignment changes on the distribution of plantar pressure under the feet. Although their research showed clear effects for angular alignment changes in the frontal plane, results were inconsistent across the sample for changes in the sagittal plane similar to the ones used in our protocol. Fridman et al.5 compared the effects of internal and external rotation on various gait parameters, finding that prosthetic step length is most affected by misalignments. Although they did not report variances within subjects, some agreement with the finding of our study that step duration variance responds to alignment changes may be inferred. Zahedi et al.,44 in a book chapter, described the effects of alignment efforts on step variability, stating that greater variability exists in kinetic parameters with misalignments. These findings were partly confirmed by our results, although several of the investigated variables, such as Fx and Mz, did not appear to be affected in the same way in our study as in Zahedi et al. Wurdeman et al.45 described significant effects of inappropriate prosthetic description on step-by-step variability in leg flexion angles using the largest Lyapunov exponent as comparison variable. Although their results indicate that step variability is affected by alignment, they are not directly comparable with our results owing to the different variables evaluated.

Limitations that should be considered in interpreting the findings of this study include the possible effect that insufficient blinding to the interventions may have had on the natural gait pattern of subjects. Being aware of the magnitude of the alignment changes could have made subjects more cautious walking on the prosthesis and, as such, affected the step-by-step variability. The VAS instrument used to gauge alignment quality may have been insufficient to gauge discomfort or pain that may be associated with prosthetic malalignments, which may have obscured an important confounding variable. However, none of the participants reported a level of pain during the experiment that would have affected collection. The selected sample may also not have been representative of the target population, as all subjects were experienced prosthetic users and as such might have had a more finely developed sense for alignment quality than a novice. Across the sample, there was no uniform response to alignment change. Each subject appeared to have his/her own coping method to adjust to ankle misalignment, causing large standard deviations in the data. Allowing subjects to use walking aids—two participants decided to walk with a cane─may have contributed to that effect. Differences in foot type and configuration, if applicable, are likely to play a role as well. In general, a stiffer foot design would be expected to be more sensitive to alignment perturbations, generating untimely peak ankle moments and thereby disrupting the gait pattern. Future research should investigate the impact of foot design and other individual patient characteristics on step variability after alignment perturbations.

Technical difficulties in realizing all the intended misalignments with the standard adapters contained in the subjects' prostheses led to a reduction in sample size beyond the initially recruited 13 subjects. Only data from five were used in the analysis of the full range of all seven ankle maladjustments. When the most extreme maladjustments (−9° and +9° of ankle flexion respectively) were eliminated for a post hoc analysis, data from nine of the initial subjects could be included. Because of the extreme nature of the −9° and +9° maladjustments, a practitioner would be unlikely to mistake these alignments as neutral and ideal. Therefore, analyzing a range of ankle maladjustments between −6° and +6° of plantarflexion is arguably clinically most relevant.

Back to Top | Article Outline

CONCLUSION

The correlations found between step-by-step variance in some gait variables, such as axial moments and step duration, and prosthetic alignment quality were too weak for most clinical applications. Further research is recommended to investigate these relationships before implementing the respective measurements in the clinical assessment of prosthetic alignment. Studying step-by-step variance in relation to gait quality, while accounting for subjects' prosthetic experience, activity level, and possibly foot design, could provide further insight into the effect of alignment changes on step variance. In patients that resemble the subjects who made up this study sample, subjective assessment is still the most robust variable in determining alignment quality.

Back to Top | Article Outline

REFERENCES

1. Blumentritt S. A new biomechanical method for determination of static prosthetic alignment. Prosthet Orthot Int 1997;21:107–113.
2. Hansen AH, Childress DS, Knox EH. Prosthetic foot roll-over shapes with implications for alignment of trans-tibial prostheses. Prosthet Orthot Int 2000;24:205–215.
3. Lin MC, Wu YC, Edwards M. Vertical alignment axis for transtibial prostheses: a simplified alignment method. J Formos Med Assoc 2000;99:39–44.
4. Talaty M, Esquenazi A. Determination of dynamic prosthetic alignment using forceline visualization. J Prosthet Orthot 2013;25:15–21.
5. Fridman A, Ona I, Isakov E. The influence of prosthetic foot alignment on trans-tibial amputee gait. Prosthet Orthot Int 2003;27:17–22.
6. Schmalz T, Blumentritt S, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: the influence of prosthetic alignment and different prosthetic components. Gait Posture 2002;16:255–263.
7. Yang L, Solomonidis SE, Spence WD, Paul JP. The influence of limb alignment on the gait of above-knee amputees. J Biomech 1991;24:981–997.
8. Boone DA, Kobayashi T, Chou TG, et al. Influence of malalignment on socket reaction moments during gait in amputees with transtibial prostheses. Gait Posture 2013;37:620–626.
9. Kobayashi T, Orendurff MS, Zhang M, Boone DA. Effect of transtibial prosthesis alignment changes on out-of-plane socket reaction moments during walking in amputees. J Biomech 2012;45:2603–2609.
10. Miller WC, Deathe AB, Speechley M, Koval J. The influence of falling, fear of falling, and balance confidence on prosthetic mobility and social activity among individuals with a lower extremity amputation. Arch Phys Med Rehabil 2001;82:1238–1244.
11. Berry D, Olson MD, Larntz K. Perceived stability, function, and satisfaction among transfemoral amputees using microprocessor and nonmicroprocessor controlled prosthetic knees: a multicenter survey. J Prosthet Orthot 2009;21:32–42.
12. Gailey R, Allen K, Castles J, et al. Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use. J Rehabil Res Dev 2008;45:15–29.
13. Morgan S, Kelly V, Hafner B. The impact of transfemoral amputation on the cognitive load associated with level walking. Presented at: 40th Annual Academy Meeting AAOP; February 26, 2014; Chicago, IL.
14. Miller WC, Speechley M, Deathe B. The prevalence and risk factors of falling and fear of falling among lower extremity amputees. Arch Phys Med Rehabil 2001;82:1031–1037.
15. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319:1701–1707.
16. Perell KL, Nelson A, Goldman RL, et al. Fall risk assessment measures: an analytic review. J Gerontol A Biol Sci Med Sci 2001;56:M761–M766.
17. Gailey RS, Roach KE, Applegate EB, et al. The amputee mobility predictor: an instrument to assess determinants of the lower-limb amputee's ability to ambulate. Arch Phys Med Rehabil 2002;83:613–627.
18. Dite W, Connor HJ, Curtis HC. Clinical identification of multiple fall risk early after unilateral transtibial amputation. Arch Phys Med Rehabil 2007;88:109–114.
19. Kobayashi T, Orendurff MS, Boone DA. Effect of alignment changes on socket reaction moments during gait in transfemoral and knee-disarticulation prostheses: case series. J Biomech 2013;46:2539–2545.
20. Bautmans I, Jansen B, Van Keymolen B, Mets T. Reliability and clinical correlates of 3D-accelerometry based gait analysis outcomes according to age and fall-risk. Gait Posture 2011;33:366–372.
21. Tura A, Raggi M, Rocchi L, et al. Gait symmetry and regularity in transfemoral amputees assessed by trunk accelerations. J Neuroeng Rehabil 2010;7:4.
22. Duncan PW, Weiner DK, Chandler J, Studenski S. Functional reach: a new clinical measure of balance. J Gerontol 1990;45:M192–M197.
23. Vanicek N, Strike S, McNaughton L, Polman R. Gait patterns in transtibial amputee fallers vs. non-fallers: biomechanical differences during level walking. Gait Posture 2009;29:415–420.
24. Parker K, Hanada E, Adderson J. Gait variability and regularity of people with transtibial amputations. Gait Posture 2013;37:269–273.
25. Kendell C, Lemaire ED, Dudek NL, Kofman J. Indicators of dynamic stability in transtibial prosthesis users. Gait Posture 2010;31:375–379.
26. Lin SJ, Winston KD, Mitchell J, et al. Physical activity, functional capacity, and step variability during walking in people with lower-limb amputation. Gait Posture 2014;40:140–144.
27. Howcroft J, Kofman J, Lemaire ED. Review of fall risk assessment in geriatric populations using inertial sensors. J Neuroeng Rehabil 2013;10:1.
28. Ambrose AF, Paul G, Hausdorff JM. Risk factors for falls among older adults: a review of the literature. Maturitas 2013;75:51–61.
29. Weiss A, Brozgol M, Dorfman M, et al. Does the evaluation of gait quality during daily life provide insight into fall risk? A novel approach using 3-day accelerometer recordings. Neurorehabil Neural Repair 2013;27:742–752.
30. Blumentritt S, Schmalz T, Jarasch R. Significance of static prosthesis alignment for standing and walking of patients with lower limb amputation. Orthopade 2001;30:161–168.
31. Boone DA, Kobayashi T, Chou TG, et al. Perception of socket alignment perturbations in amputees with transtibial prostheses. J Rehabil Res Dev 2012;49:843–853.
32. Neumann ES. State-of-the-science review of transtibial prosthesis alignment perturbation. J Prosthet Orthot 2009;21:175–193.
33. Fiedler G, Slavens B, Briggs D, Hafner BJ. Criterion and construct validity of prosthesis-integrated measurement of joint moment data in persons with transtibial amputation. J Appl Biomech 2014;30:431–438.
34. Hafner B. Prosthetic Limb Users Survey of Mobility (PLUS-M™) 12-Item Short Form (v1.2). Seattle, WA: 2015.
35. Gailey RS, Nash MS, Atchley TA, et al. The effects of prosthesis mass on metabolic cost of ambulation in non-vascular trans-tibial amputees. Prosthet Orthot Int 1997;21:9–16.
36. Selles RW, Bussmann JB, Klip LM, et al. Adaptations to mass perturbations in transtibial amputees: kinetic or kinematic invariance? Arch Phys Med Rehabil 2004;85:2046–2052.
37. Fiedler G, Slavens BA, O'Connor KM, et al. Effects of physical exertion on trans-tibial prosthesis users' ability to accommodate alignment perturbations. Prosthet Orthot Int 2014;40:75–82.
38. Fiedler G, Slavens BA, Hafner BJ, et al. Leg laterality differences in persons with bilateral transtibial amputation: a pilot study using prosthesis-integrated load cells. J Prosthet Orthot 2013;25:168–176.
39. Fiedler G. Prosthesis-integrated load cells as alternatives to conventional gait analysis methods—capabilities and limitations. In: ISPO 2013, Symposium on Load Measurement for Evidence Based Practice. Hyderabad, India: International Society of Prosthetics & Orthotics; 2013.
40. Antonsson EK, Mann RW. The frequency content of gait. J Biomech 1985;18:39–47.
41. Chow DH, Holmes AD, Lee CK, Sin SW. The effect of prosthesis alignment on the symmetry of gait in subjects with unilateral transtibial amputation. Prosthet Orthot Int 2006;30:114–128.
42. Sin SW, Chow DH, Cheng JC. Significance of non-level walking on transtibial prosthesis fitting with particular reference to the effects of anterior-posterior alignment. J Rehabil Res Dev 2001;38:1–6.
43. Geil MD, Lay A. Plantar foot pressure responses to changes during dynamic trans-tibial prosthetic alignment in a clinical setting. Prosthet Orthot Int 2004;28:105–114.
44. Zahedi M, Spence W, Solomonidis S. The influence of alignment on prosthetic gait. In: Murdoch G, RG D, (eds.). Amputation Surgery and Lower Limb Prosthetics. Oxford: Blackwell Scientific; 1998:367–378.
45. Wurdeman SR, Myers SA, Jacobsen AL, Stergiou N. Adaptation and prosthesis effects on stride-to-stride fluctuations in amputee gait. PLoS One 2014;9:e100125.
Keywords:

artificial limbs; prosthetic alignment; lower-limb prosthesis; step variability; gait stability; mobile gait analysis; load cell

© 2017 by the American Academy of Orthotists and Prosthetists.