Abstract
Background:
Individuals with transtibial amputation are at increase risk of falling. The absence of an ankle joint and the associated musculature in these individuals can reduce clearance between the prosthetic foot and ground during the swing phase of gait, which may increase the risk of stumbling and in turn falling.
Objectives:
To associate minimum toe clearance during gait in the laboratory with community-based, trip-related stumbles by individuals with transtibial amputation using conventional feet.
Study design:
Prospective cohort design; following quantitative gait analysis, participants completed electronic surveys to prospectively report stumbles and falls for 1 year thereafter.
Methods:
General community with gait analysis conducted within a motion analysis laboratory and prospective tracking of stumbles occurring in the community. A volunteer sample of eight unilateral, transtibial amputees that were K3 or K4 level ambulators and current patients at a local prosthetic clinic. All participants completed the entire 1-year follow-up study. Prosthetic-side minimum toe clearance while walking on a level treadmill at self-selected speed and self-reported trip-related stumbles in the community. Minimum toe clearance was defined as a local minimum of the vertical displacement of the toe from toe-off to heelstrike relative to its position during midstance.
Results:
Prosthetic-side minimum toe clearance was more than 50% lower for participants who reported one or more trip-related stumbles on that side compared with participants who reported zero trip-related stumbles on the prosthetic side (minimum toe clearance = 12.3 ± 0.8 mm vs 25.6 ± 5.4 mm, respectively; p = 0.036).
Conclusion:
This is the first study relating laboratory-based measures to prospective stumbles by prosthesis users. The results suggest that prosthesis users with low minimum toe clearance may be at increased risk of experiencing a trip-related stumble in the community. Given that frequent stumbling increases the risk of falling, future work is warranted on the effectiveness of interventions focused on minimum toe clearance on reducing fall risk.
Clinical relevance
Interventions to increase minimum toes clearance, which could include prescription of active dorsiflexing prostheses or gait training, may help reduce the risk of trip-related falls for individuals who report a history of trip-related stumbles.
Keywords
Background
Upward of 50% of individuals with lower limb amputations may fall annually. 1 While many circumstances cause falls 2 (e.g. legs giving out, slipping on ice), trip-related falls may be particularly problematic for individuals with transtibial amputation. Trip-related falls are the leading cause of falls by community-dwelling, middle-aged, and older adults2–4 and may also be problematic for individuals with amputation. The absence of an ankle joint and the associated musculature in individuals with transtibial amputation can limit minimum toe clearance (MTC) during the swing phase of gait. MTC, defined as the vertical distance between the ground and the swing foot at approximately midswing, occurs at an instant when individuals are particularly susceptible to trip-related stumbles. 5 Prosthetic feet that actively dorsiflex during swing significantly increase MTC during treadmill walking in the laboratory 6 and, based on theoretical modeling, 7 reduce the likelihood of a trip-related stumble. 6 However, MTC has not been empirically related to trip-related stumbles in the community, where surfaces are less regular than treadmills 8 and increased attentional demands may negatively affect the ability to avoid obstacles during gait.9,10
The purpose of this study was to associate MTC measured during laboratory gait with community-based, trip-related stumbles by individuals with transtibial amputation using conventional (non-articulated) feet. The hypothesis was that individuals who prospectively report trip-related stumbles on their prosthetic limb would have significantly lower prosthetic-side MTC compared to those reporting no such stumbles. Given that stumbling increases fall risk, 11 support of the hypothesis would provide preliminary, empirical evidence of a relationship between MTC and fall risk and thus, a potentially important biomechanical variable to target in fall prevention interventions.
Methods
Participants
This study is a secondary analysis of a larger data set from a prior study approved by the Institutional Review Board at the University of Illinois at Chicago. Initial inclusion criteria included the following: unilateral transtibial amputation, K3/K4 level ambulator, 30+ years of age, prosthetic use for at least 1 year, pain-free at the time of testing, and self-reported capability to walk for 2 min on a treadmill. Of the 15 participants who consented to participate, this analysis excludes data from one who used an active dorsiflexing foot and seven who opted not to participate in the prospective portion of the study. Thus, this study considers data from N = 8 participants (age: 47.0 ± 11.8 years; height: 1.72 ± 0.09 m; mass: 78.4 ± 19.7 kg; Table 1).
Participant-specific demographics, MTC, and stumble status.
MTC: minimum toe clearance.
Protocol
Laboratory measure of MTC
Here, we present details of the protocol that are relevant only to this study. For a description of the entire larger protocol, refer to Rosenblatt et al. 6 Six participants walked on a treadmill (ActiveStep, Lebanon, NH, USA) for at least 2 min at a self-selected velocity (SSV) under two conditions: 0% (level) and 5% (incline) grade, with the order of grades randomized. Due to technical issues, data from one participant (P1 in Table 1) on the 5% trial was not available. One participant (P14) preferred to complete the protocol over ground and traversed an 8-m walkway10 times at SSV (15 total steps with prosthesis). As MTC is not known to differ between treadmill and over ground walking, his data were not excluded from analysis. An incline walking condition is not available for this participant. Additionally, the MTC for participant P7 during the 0% condition at SSV (73 mm) far exceeded the values reported in the literature.6,7,12,13 Upon inspection of the data, investigators determined that the participant utilized a “high stepping gait” possibly reflecting treadmill accommodation. For this participant, we analyzed data from a trial at 0% grade and 80% SSV. We deemed this acceptable for analysis as walking speed does not significantly affect MTC.6,14
In this study, we analyze data from only the 0% grade condition for two reasons. First, more data were available for this condition, and second, a priori, we assumed this condition would be more representative of conditions encountered most often during daily locomotion. The average number of steps analyzed for the 0% condition on the treadmill was 248 ± 228 steps (range: 75–662 steps corresponding to varying durations of walking). Given the altered biomechanical demands required for inclined walking, MTC data from the 5% condition are presented for comparative purposes, for those participants where data are available, although statistical analyses were not performed.
During all conditions, an eight-camera motion capture system (Motion Analysis, Santa Anna, CA, USA) tracked the movements of a passive reflective marker placed on the shoe approximately over the second metatarsal. For each step, prosthetic-side MTC was calculated from toe-off to heelstrike as the local minimum in the vertical displacement of this marker relative its position during the previous midstance when the foot was in total contact with the treadmill belt. Thus, our measure of MTC is not a direct measure of the distance between the treadmill surface and the portion of the foot that is closest to the treadmill surface (e.g. the toe or sole of the shoe) but serves as a simple surrogate measure. In the absence of a local minimum, MTC was taken as the displacement at 50% of swing, where MTC tends to occur.5,12 For each participant, an across step mean MTC was calculated.
Prospective tracking of stumbles and falls
Trip-related stumbles and falls were tracked via electronic surveys sent every 2 weeks for 1 year. 15 The survey first asked whether, in the previous 2 weeks, the participant fell (i.e. “unintentionally came to rest on a lower surface”) or experienced “a loss of balance that did not result in a fall” (i.e. stumbled). Follow-up questions addressed the circumstances leading to the fall/stumble. Specifically, participants chose one of eight possible causes for the fall/stumble, all of which were adapted from Berg et al. 2 The eight causes included the following: trip, slip, legs give out, misplaced step, pushed or pulled by someone or something, general balance loss, ground started to move underneath my feet, and other/unknown. A trip was defined as “having the motion of the swinging limb stopped by an object or the ground.” In the event of a trip (and select other fall/stumble causes), participants reported on which limb the trip (cause of the fall/stumble) occurred. This was the only amputee-specific aspect of the survey. Participants also reported on the activity(ies) that was(were) being performed just prior to the fall or stumble, choosing from 21 activities—many of which were adapted from Berg (e.g. walking, hurrying, participating in sport, bending, reaching). In the event that participants were walking prior to the fall/stumble, then they reported whether they were walking on level ground, uphill, downhill, on a ramp, on uneven surface/ground, or on an “other/unknown” surface. The complete survey is provided in Supplementary Appendix 1. All the participants who opted to participate in this aspect of the larger study completed the entire 1-year follow-up.
Statistical analysis
Given the small sample size, a Mann–Whitney U test was used to compare MTC between those participants who reported one or more trip-related stumbles on the prosthetic limb with those who reported no such stumbles. An effect size (Cohen’s 16 d) was also calculated with small, medium, and large effects corresponding with the values of 0.2, 0.5, and 0.8, respectively. Significance was set at p ⩽0.05.
Results
Summary of prospective stumbles and falls
Three participants reported zero falls and zero stumbles during the 1-year follow-up. Two participants reported zero falls but multiple stumbles (of any type), and three participants reported one or more falls and multiple stumbles (of any type) (see Table 2 for summary). Over the course of the year, there were a total of eight falls (two trip-related falls, two slip-related falls, two legs giving out, two general balance losses).
Summary of number of stumbles and falls by participant.
There were a total of 44 all-cause stumbles reported over the course of the year. In total, 18 of these stumbles (41%) occurred during walking. Of those stumbles that occurred during walking, 83% occurred while walking on level ground, 11% occurred while walking on uneven ground/surface, and 6% occurred while walking downhill. The number of stumbles was biased by one participant (P7) who reported 18 stumbles, with 13 having an “unknown/other” cause. The direct cause of the other five stumbles included two general balance losses, two slips (slipping limb was unknown), and one trip on the non-prosthetic limb. The underlying cause for the high number of stumbles by this participant is unclear.
After excluding the 13 unknown-cause stumbles by participant P7, trip-related stumbles on the prosthetic side constituted the largest percentage of remaining stumbles (31%; n = 10). The 10 trip-related stumbles on the prosthetic side were reported by three participants; two of these three participants also reported stumbles due to other causes (Figure 1; Table 2). In addition to the 10 trip-related stumbles that were reported, participants reported four slip-related stumbles, four stumbles due to general balance loss, four stumbles due to legs giving out, and nine stumbles due to the remaining six causes.
Categorization of participants based on prospective reporting of stumbles. There were five participants who reported stumbles of any type and three who did not report any stumble. Prosthetic-side MTC was compared between (1) participants who reported a trip-related stumble on the prosthetic side (white boxes) and all other subjects (dark gray boxes) and (2) participants who reported a non-trip-related stumble on the prosthetic side (boxes with single-framed borders) and participants who did not report these type of stumbles (boxes with double-framed border). For reference, participant codes (from Tables 1 and 2) are included in their corresponding box.
Relating MTC and prospective stumbles
Prosthetic-side MTC was ~50% lower for the group of three participants reporting trip-related stumbles on that side compared with the remaining five participants who either reported no stumbles on that side or if stumbles were reported on that side then the stumbles were not trip-related (12.3 ± 0.8 vs 25.6 ± 5.4 mm, respectively; p = 0.036; see Figure 1 for groupings). This difference represents a large effect (Cohen’s d = 2.7). Prosthetic-side MTC was not different between the group of four participants reporting non-trip-related stumbles on their prosthetic side and those reporting no stumbles on their prosthetic side (19.8 ± 6.3 vs 22.2 ± 11.1 mm, respectively; p = 0.69; Cohen’s d = 0.22; see Figure 1 for groupings). With regard to inclined walking, it can be seen that two of three participants who reported zero trip-related stumbles on the prosthetic side had MTC during the 5% condition that was considerably larger than that of participants reporting such stumbles (Figure 2)
The effect of incline walking on MTC. Data from six subjects who completed both 0% and 5% grade walking on the treadmill are presented. In general, prosthetic-side MTC decreased during incline walking, which has been reported. A larger study is needed to quantify the extent to which MTC during the 5% condition differs between those subjects who self-reported prosthetic-side, trip-related stumbles in the community (dotted lines with open symbols), and those that did not report these stumbles (solid lines and filled black symbols).
Discussion
The purpose of this study was to associate MTC in the laboratory with community-based, trip-related stumbles by individuals with transtibial amputation using conventional (non-articulated) feet. The results support the hypothesis that individuals with transtibial amputation who prospectively report trip-related stumbles with their prosthetic would have lower prosthetic-side MTC compared to those reporting no such stumbles. Given a relationship between low MTC and community-based, trip-related stumbles, and that individuals with lower limb amputation show impaired ability to successfully recover from a laboratory-induced trip,17,18 low MTC on the prosthetic limb may partly explain increased fall risk in this population.
Despite the logical assumption that low MTC should increase the risk of stumbling, several features of our results require highlighting. First, and most important, the relationship between MTC and prospective, community-based stumbles has not been reported for this or any population. Only one study reported a similar result, associating MTC in able-bodied older adults and retrospective trip-related falls.13,19 The results of that study were somewhat unexpected in that individuals with a history of trip-related falls had significantly higher MTC, thought to reflect a compensatory mechanism in response to a history of falling. It remains unknown whether high MTC existed prior to experiencing trip-related falls. This is not trivial. Given that a history of falls increases future fall risk, 20 the ability to identify at-risk individuals and intervene before the index fall (i.e. that which leads to future falls) could delay, or prevent, future falls from occurring.
Given that low MTC in prosthesis users appears to lead to trip-related stumbles, and that frequent stumbling increases fall risk, 11 then interventions to increase MTC, before the occurrence of an index fall, could effectively reduce the risk of trip-related falls for individuals with low MTC. One potential such intervention could include providing active dorsiflexing prostheses to persons with transtibial amputation. These prostheses have been shown to increase MTC with minimal changes to kinematics at the instant of MTC. 6 On the other hand, they may affect kinematics and kinetics at other points of the gait cycle21,22 or result in unanticipated, patient-specific, biomechanical adaptations, either of which could increase the risk of stumbling. An alternative intervention, independent of componentry, could include gait training. Given that MTC is affected by kinematics not only of the ankle but also of the hip and knee, 12 training individuals to walk with increased hip and/or knee flexion could act to increase MTC. The efficacy of these or any other interventions to improve MTC and reduce stumbles and fall risk has yet to be established.
A second important feature of our results relates to the specificity. In particular, a significant relationship was absent between MTC and non-trip-related stumbles, but present between MTC and trip-related stumbles. The existence of a plausible, modifiable relationship between MTC and trip-related stumbles and falls is predicated on this difference. For example, were a relationship to exist between MTC measured during gait and stumbles occurring, for example, while standing and turning, this relationship would likely be mediated by secondary factors; increasing MTC would be unlikely to reduce stumbles during turning. The extent to which targeting a biomechanical variable measured during a particular activity in the laboratory or clinic helps to reduce stumbles (or falls) may depend on the strength of the relationship between the variable and community-based stumbles (or falls) during that same activity.
Following this logic, an important question remains: is level walking at SSV the optimal condition for measuring MTC if one is interested in understanding behaviors that lead to trip-related stumbles in the community? Given that inclined walking may be more destabilizing 23 and more difficult 24 for persons with transtibial amputation, and that the biomechanical demands of the task may differ considerably from level walking, 25 it is reasonable to assume that MTC during the 5% incline condition should better reflect behaviors leading to stumbles in the community. However, based on the previous specificity argument, this assumes that, in the community, stumbles frequently occur while walking on inclines. On the other hand, such events may actually be rare. Our participants reported zero stumbles while walking uphill or on ramps. Possible, non-mutually exclusive explanations for the absence of such stumbles could include: (1) participants did not encounter inclines in the community (e.g. the topography of the Chicago-land area is generally flat), (2) participants were unable to accurately perceive differences between inclined and level walking in the community, and (3) due to the increased demands, participants avoided walking on inclines. Regardless, our results suggest that MTC during inclined walking in the laboratory may not be a strong indicator of the causes of prospective, community-based stumbles, at least in our cohort. Further work is needed to evaluate the repeatability of these findings and their generalizability to other, larger samples of prosthesis users.
Study limitations
Several study limitations should be noted. First, given the small sample size, we may be under-powered to detect a relationship between prosthetic-side MTC and any-cause stumbles, although, as discussed, our results are expected and logical. To minimize the influence of small sample size, we included data from two participants (P7 and P14) who did not walk on a treadmill at 0% grade and at SSV (see section “Protocol”). However, inclusion is justified, and given the large effect size, even excluding these subjects would not result in p >0.05. Second, based on the definition, self-report of stumbles may be more subjective than that of falls, which may explain the high inter-subject variability in this measure. For example, the high number of stumbles reported by participant P7 may reflect individual hypervigilance with regard to balance loss. This subject may have reported events that other participants ignored or deemed unimportant to report. Increasing sample size would reduce this variability. Finally, the results of this study are limited to participants that are 30 years of age and older and may not generalize to younger prosthesis users. Despite these limitations, we observed a significant effect of MTC on trip-related stumbles. Given that this is the first study to quantify the effect of MTC on prospective stumbles and, given the strength of the effect, these preliminary findings justify the need for a larger study to relate MTC and the causes of community-based stumbles and falls by individuals with transtibial amputation.
Conclusion
Prosthetic users with low MTC may be at increased risk of experiencing a trip-related stumble in the community. While a fall need not follow a stumble, given that increased stumbling increases fall risk, 11 interventions targeting individuals with low MTC may help reduce future fall risk.
Footnotes
Author contribution
All authors contributed equally in the preparation of this article.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Össur, the manufacturer of the ProprioFoot—an active dorsiflexing prosthesis. Össur did not play a role in the collection, analysis, or interpretation of data.