- Email: [email protected]
Effects of walking speed on gait stability and interlimb coordination in younger and older adults
Gait & Posture 39 (2014) 378–385
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Effects of walking speed on gait stability and interlimb coordination in younger and older adults Tal Krasovsky a,c, Anouk Lamontagne a,c, Anatol G. Feldman b,c, Mindy F. Levin a,c,* a
School of Physical and Occupational Therapy, McGill University, Montreal, Quebec, Canada Department of Physiology, Universite´ de Montre´al, Montreal, Quebec, Canada c Centre for Interdisciplinary Research in Rehabilitation, Montreal, Quebec, Canada b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 February 2013 Received in revised form 12 June 2013 Accepted 9 August 2013
Many falls in older adults occur during walking following trips. Following a trip, older adults take longer than younger adults to recover steady-state walking. Although faster gait speed may improve interlimb coordination, it may also increase fall risk in older adults. We hypothesized that older adults would take longer than younger adults to recover from an unexpected perturbation during gait especially when walking faster. Twelve younger (26.3 4.4 years) and 12 older adults (68.5 3.4 years) walked at comfortable, faster and slower speeds when movement of the dominant leg was unexpectedly arrested for 250 ms at 20% swing length. Gait stability was evaluated using the short- and longer-term response to perturbation. In both groups, walking faster diminished the occurrence of elevation and increased that of leg lowering. Older adults took longer than younger adults to recover steady-state walking at all speeds (3.36 0.11 vs. 2.89 0.08 strides) but longer-term recovery of gait stability was not related to gait speed. Arm-leg and inter-arm coordination improved with increasing gait speed in both groups, but older adults had weaker inter-leg coupling following perturbation at all speeds. Although both younger and older adults used speed appropriate responses immediately following perturbation, longer duration of recovery of steady-state walking in older adults may increase fall risk in uncontrolled situations, regardless of gait speed. Recovery from perturbation when walking faster was associated with better interlimb coordination, but not with better gait stability. This indicates that interlimb coordination and gait stability may be distinct features of locomotion. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Perturbation Central pattern generators Locomotion Aging
1. Introduction A leading cause of injury in older adults is falling, causing a heavy burden on the health-care system. Most falls in older adults occur when walking, specifically following a trip [1]. Avoidance of falling is related to gait stability, defined as the ability to maintain functional locomotion despite disturbances [2]. When walking faster, responses to unexpected perturbations need to be performed within shorter timeframes, which can increase fall risk especially among older adults [3]. Longer recovery times toward steady-state walking patterns following a perturbation may reveal lower stability properties of the unperturbed gait pattern in the absence of a fall [4]. Young healthy adults use different response strategies following a trip to prevent falling. Immediate lowering of the perturbed leg
* Corresponding author at: School of Physical and Occupational Therapy, 3654 Promenade Sir William Osler, Montreal, Quebec, Canada H3G 1Y5. Tel.: +1 514 398 3994; fax: +1 514 398 6360. E-mail address: [email protected] (M.F. Levin). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.08.011
(lowering strategy) allows a faster transition to a double-support phase. While this may be safer in the short-term [5], this strategy also requires a rapid transition of the non-perturbed leg from stance to swing, which is relatively energy-demanding [6,7] and associated with longer recovery times toward a steady-state walking pattern [8]. An alternative strategy (elevation) is to stabilize the body on the non-perturbed leg and complete the perturbed step with minimal disturbance. This strategy requires early muscle recruitment on the non-perturbed leg [9] and allows faster recovery of the steady-state walking pattern [8]. A third strategy (combined) is a combination of lowering and elevation [7]. Changing gait speed affects the available response time and may also affect strategy selection. Younger and older adults use similar leg strategies when gait speed is constrained by a treadmill [10] and when walking at their comfortable speed [8]. However, due to limitations in the magnitude and rate of muscle recruitment in older adults [9], walking faster may be associated with an increased occurrence of leg lowering, as well as a longer period of recovery toward a steady-state walking pattern. Recovery from a trip involves the re-establishment of a coordinated gait rhythm in all four limbs [8,11]. Generally,
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
379
Table 1 Anthropometric and kinematic data for young (N = 12) and older (N = 12) adults.
Age (years) Gender (M/F) Height (cm) Weight (kg) Gait speed (m/s) Swing length (cm) Swing time (s)
Slower 0.77 0.06 61.81 4.16 0.52 0.02
Young (N = 12)
Older (N = 12)
26.3 4.4 7 M/5 F 172.7 8.5 71.7 8.5 Comfortable 1.05 0.06 65.52 4.32 0.44 0.02
68.5 3.4 8 M/4 F 166.5 10.0 70.0 13.6 Comfortable 1.10 0.14 65.55 4.85 0.43 0.02
temporal coupling between movements of different limbs is weaker in older compared to younger adults [12,13], possibly due to altered processing of afferent information and/or cognitive decline [12]. These factors may also affect older adults’ ability to coordinate the response to a unilateral gait perturbation [8]. Increasing gait speed improves interlimb coupling in healthy adults [14], as well as in individuals with central nervous system damage, such as cerebral palsy [15] and stroke [4]. Recovery of interlimb coordination may be more challenging for older adults given the limited response time available at faster speeds. We evaluated the effects of gait speed on recovery of the steadystate walking pattern following a leg perturbation in younger and older adults. We hypothesized that (1) in the short-term, faster walking would involve more leg lowering responses; and in the longer-term, faster walking would result in better interlimb coordination, but longer recovery times toward a steady-state walking pattern; and (2) older adults were expected to take longer than younger adults to restore steady-state walking patterns, especially when walking faster. 2. Methods 2.1. Participants Twelve younger and 12 older adults were recruited (Table 1). Subjects were excluded if they had any orthopedic, vestibular or neurological disorders affecting gait or documented cognitive decline. All subjects signed informed consent forms approved by the local ethics committee. 2.2. Experimental procedure The experimental procedure has been described in detail elsewhere [8]. Subjects walked on a self-paced treadmill while wearing a safety harness that did not restrict trunk motion (Fig. 1A). Safety switches could stop the treadmill if needed. Comfortable gait speed was set once the subject habituated to treadmill walking. Thirty-six 40 s trials were performed in which subjects walked at one of three speeds: comfortable, 20% faster or 20% slower (12 trials each), paced by a digital metronome. Order of speed conditions was fully randomized. The dominant leg, determined as the leg preferred for kicking, was perturbed 1–2 times per trial (20 perturbations in total for each speed) by clamping a wooden rod attached to an ankle cuff. An identical rod was attached to the non-dominant leg to equalize sensation. Rods were equipped with transducers measuring perturbation force. Leg arrests of 250 ms duration occurred at early swing (20% of swing length based on mean swing length of 3 unperturbed strides) using a custom-built microcontroller. To prevent anticipation, 3–4 non-perturbed ‘‘catch’’ trials were randomly interspersed throughout all trials. Markers were placed on body landmarks, specifically on both heels (calcanei) and hands (dorsum of 2nd metacarpals). Movement kinematics were recorded at 120 samples/s using a 12-camera motion analysis
Faster 1.33 0.1 68.94 6.1 0.39 0.01
Slower 0.79 0.11 62.18 6.15 0.52 0.04
Faster 1.30 0.20 65.47 6.99 0.38 0.02
system (Vicon, Oxford, UK) and filtered using a 6th order Butterworth filter (dual-pass, cutoff 10 Hz). 2.3. Data analysis Spatiotemporal gait parameters were calculated from heel marker positions. Indices of gait stability were described in the short (1–2 steps) and longer (>2 steps) term. The short-term response was divided according to three leg strategies (Fig. 1B). In the longer term, an index of recovery of double-support time [4,8] was used to estimate return to steady-state walking. Due to possible asymmetry in double-support times, baseline doublesupport times during unperturbed gait (the time between foot contact and subsequent toe-off of the other leg) were computed separately for each side and averaged across the 3 pre-perturbed strides. The double-support recovery index was defined as the number of gait cycles taken until recovery of baseline values on both sides (plus or minus 10%). To quantify the residual effect of the perturbation, phase shifts were calculated using timing of gait events in the arms and legs before and after perturbation [11]. The sagittal plane position of each limb in 3 pre-perturbed cycles was projected forward and the minimal time between the actual and projected event (foot contact or arm direction reversal) was divided by the pre-perturbed cycle period and multiplied by 3608. A negative value (phase advance; up to 1808) indicated that the gait event occurred prior to its unperturbed occurrence and a positive value (phase delay; up to 1808) denoted a delayed occurrence of the event. Temporal interlimb coupling following perturbation was computed using absolute phase shift differences between limb pairs: arm–leg on perturbed and non-perturbed sides, arm–arm and leg–leg pairs. Values ranged from 08, indicating full restoration of the pre-perturbed coordination pattern, to plus/minus 1808, indicating no coupling and a full shift in the coordination pattern. 2.4. Statistical analysis Student’s t-tests compared subject characteristics between groups. Repeated measures ANOVAs tested the effect of group and speed condition on baseline spatiotemporal walking characteristics, perturbation force, and short- and long-term indices of gait stability and interlimb coordination. Levels of significance in post hoc tests were adjusted using a Bonferroni correction. A circular two-way ANOVA (Harrison–Kanji test) was implemented [16] in Matlab 6.5 (Mathworks, Inc., Natick, MA, USA) for phase shifts. All tests except circular procedures were done using SPSS (version 17.0). Initial significance was P < 0.05. 3. Results Younger and older subjects did not differ in height, weight or gait speed (Table 1). Younger and older adults walked 27% and 28% slower in the slower speed condition and 27% and 18% faster in the faster speed condition respectively compared to their comfortable
380
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
Fig. 1. (A) Experimental setup. Subjects walked on a self-paced treadmill, a clamp unexpectedly arrested a wooden rod attached to their dominant ankle at 20% swing. (B) Phase plots describing leg strategies (young subject, comfortable speed), with sagittal plane velocity on the ordinate and position on the abscissa. Thin solid lines depict 3 preperturbed gait cycles, red traces depict the perturbed cycle, and dashed lines depict 2 post-perturbed cycles. A lowering strategy (top panel): lowering of the perturbed leg and a faster, shorter step in the contralateral leg. An elevation strategy (bottom panel): prolonged maintenance of the leg in the air, followed by a non-perturbed step in the contralateral leg. The combined strategy (middle) is a combination of the two. Reprinted with permission from Journal of Neurophysiology, 2012 [8].
speed, with no group differences or interactions between speed and group. None of the subjects fell during the session and no adverse effects were reported. Perturbation force (Fig. 3A) did not differ between groups, but increased with speed (F2,44 = 30.3, P < 0.01).
3.1. Effects of gait speed (Hypothesis 1) For each subject, frequency of occurrence of each leg strategy was computed (Fig. 2). At comfortable speed, use of each strategy was comparable to previous studies [8,17]. The slower speed
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
condition was associated with less lowering (F2,44 = 29.6, P < 0.01) and more leg elevation strategies (F2,44 = 11.6, P < 0.01) than the comfortable and faster speeds. In the longer-term, however, there was no main effect of speed condition on recovery of doublesupport duration. Phase shifts in the arms and legs increased with gait speed (Fig. 3C–F, perturbed leg: F2,66 = 5.89, P < 0.01; nonperturbed leg: F2,66 = 5.42, P < 0.01; ipsilateral arm: F2,66 = 16.99, P < 0.01; contralateral arm: F2,66 = 20.20, P < 0.01). Post hoc analyses revealed that for all limbs, this difference occurred between the slower and faster speeds. For all limbs except the contralateral arm, phase shifts at slower speed were also significantly smaller than those at comfortable speed. Smaller phase-shift differences in the arms and arm-leg pairs were associated with increasing gait speed (Fig. 4A–C, arm–leg pair – perturbed side: F2,44 = 42.57, P < 0.01; arm–leg pair – nonperturbed side F1,22 = 22.65, P < 0.01; arm-arm coupling: F1,22 = 37.63, P < 0.01). Leg–leg coupling did not vary with gait speed. 3.2. Effects of age (Hypothesis 2) In the short-term, young and older adults used similar strategies (Fig. 2). In the longer-term, older adults required more
strides to recover double-support duration (3.36 0.11 compared to 2.89 0.08 strides in younger adults, F1,22 = 5.89, P < 0.03, Fig. 3B). Older adults also had larger phase shifts in the legs (Fig. 3C and D, perturbed leg: 47.78 vs. 16.98, F1,66 = 8.67, P < 0.01, non-perturbed leg: 50.98 vs. 17.68, F1,66 = 9.81, P < 0.01) but phase shifts in the arms did not differ between groups ( 49.28 vs. 20.98 for the ipsilateral arm, and 41.1 vs. 19.08 for the contralateral arm). Finally, older adults had larger phase-shift differences between the legs compared to younger adults (Figure 4D, F1,22 = 8.20, P < 0.01). 4. Discussion The stability properties of gait in young and older adults were evaluated by examining responses to unilateral gait perturbations. Response to perturbation in the short- and longer-term represents the stability properties of the underlying gait pattern, which can be considered as a stable attractor[4,18]. Thus, decreased gait stability may be manifested by more frequent use of leg lowering in the short-term, and longer recovery times toward steady-state walking patterns in the longer-term. Gait stability may be quantified in different ways using responses to various external or internal perturbations [19] with the role of gait speed in gait stability changing accordingly [20].
Young
80
80
60
60
40
40
20
20
0
0
100
100
80
80
60
60
40
40
20
20
0
0
100
100
80
80
60
60
40
40
20
20
Comfortable Faster
% Cases
Slower
100
1
2
Elevaon Combined Lowering
Older
100
0
381
3
4
5
6
7
8
Subjects
9 10 11 12
0
1
2
3
4
5
6
7
8
9 10 11 12
Subjects
Fig. 2. Leg response strategies immediately following perturbation. Stacked bars represent the percentage of trials where a strategy was used for each subject. Subjects are numbered (1–12) according to treadmill gait speed (from lowest to highest).
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
Perturbation force (N)
A
**
**
B
**
60 50 40 30 20 10
Young Older
* **
p<0.05 p<0.01
Double-support recovery (cycles)
382
4.0
3.5
3.0
2.5 * Group 2.0
0
Phase shifts C
**
D
**
0 -20 -40 -60 -80
** Group
-100
0 -20 -40 -60 -80
** Group
-100
E
*
**
F
20
** 20
Contralateral arm (° )
Ipsilateral arm (° )
**
20
Non-perturbed leg (° )
Perturbed leg (° )
20
**
0 -20 -40 -60
0 -20 -40 -60
-80
-80
-100
-100 Slower
Comfortable
Faster
Speed
Slower
Comfortable
Faster
Speed
Fig. 3. Perturbation force and longer-term responses of younger (black) and older (gray) adults. (A) Average perturbation force (N). B. Gait cycles required until doublesupport time recovered to baseline 10%. (C–F) Phase shifts in the legs (C and D) and arms (E and F) following perturbation. Error bars mark 1SEM. Asterisks mark significant post hoc differences.
For example, following a slip walking faster was associated with better stability [21] while a trunk perturbation at faster speeds was associated with slower recovery of baseline center-ofmass movement [22]. Using responses to a trip perturbation as a probe to gait stability given their functional significance for older adults [1], this study focused on one aspect of this complex construct.
4.1. Gait speed effects on short- and longer-term responses to perturbation Our results showed that both younger and older adults used more lowering and less elevation strategies when increasing gait speed (Figs. 1 and 2). To use elevation, body weight needs to be maintained on the unperturbed leg [9]. This strategy may be
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
383
Phase shift differences **
Perturbed side (deg)
60
**
B *
50 40 30 20 10
Young Older
* **
**
60
p<0.05 p<0.01
Non-perturbed side (deg)
A
0
**
50 40 30 20 10 0
C
**
D
**
20
Legs (deg)
Arms (deg)
60
40
20
15
10
5
** Group 0
0 Slower
Comfortable
Faster
Speed
Slower
Comfortable
Faster
Speed
Fig. 4. Phase shift differences between limbs for younger (black) and older (gray) adults. (A and B) Arm-leg phase shift differences for the perturbed/dominant and nonperturbed/non-dominant sides. (C and D) Arm–arm and leg–leg phase shift differences. Error bars mark 1SEM. Asterisks mark significant post hoc differences.
preferred by young and older adults [23,24], since it allows faster recovery toward steady-state walking [8]. However when an unexpected perturbation occurs during faster walking, there is less time to perform this adjustment, increasing the challenge to the sensorimotor system. Under these conditions, lowering the leg can be considered as safer in the short-term, since it immediately increases the base of support [5], although it is a less energyefficient strategy [7] and associated with a longer period of instability in the longer-term. Indeed, in people with central nervous system damage (stroke), a similar perturbation resulted in frequent use of lowering regardless of gait speed [4]. Faster walking requires greater control of the increased forward body momentum, and older adults who walked faster were more likely to fall than slower walkers following an overground trip [3].Our results showed that despite the increased challenge, older adults are able to control the increased forward body momentum using speed-appropriate strategies. In the longer-term after perturbation, walking faster was associated with larger phase shifts in arms and legs in both groups. Previous results showed that leg lowering is associated with larger phase shifts in all four limbs when walking at comfortable speed [8]. Results from this study further suggest that in younger and older adults, global phase resetting occurs also at faster or slower speeds. Similar phase shifts in all four limbs may indicate that a central timing mechanism, i.e., central pattern generator (CPG), is reset to accommodate for the perturbation [25]. Feldman et al. [11]
further suggested that the CPG is driven by a hierarchically superior mechanism that determines the translation rate of the body balance in the environment, influencing gait speed. Thereby, global long-term phase shifts result from a transient change in the rate of translation of the body in response to perturbation. Our results also showed that temporal arm–leg and arm–arm coupling improved (smaller phase-shift differences) with faster gait speed. Better limb coupling is usually related to greater stability of a coordinative pattern [15], such that a perturbation will be associated with less disruption [18]. Our results suggest that the ability to maintain a consistent coordination pattern following perturbation is greater when walking faster. However, consistency of the coordinative pattern should not be equated with stability of steady-state walking since faster walking was not associated with faster recovery of double-support duration. In other words, interlimb coordination and gait stability are different aspects of locomotion which may be contrasted in their relation to gait speed. Our results further showed that short- and longer-term responses to perturbation were affected differently by gait speed. While faster walking was associated with more frequent leg lowering, no longer-term relationship was found between gait speed and the time required to recover steady-state walking (Fig. 3B). This may be related to the constraints applied to the motor system when walking at non-comfortable speeds. At comfortable speed, the lower inter-stride dependency of
384
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
spatiotemporal gait parameters suggests that gait may be more adaptable to environmental changes [26]. Thus, while strategy selection may be directly associated with gait speed, the longterm recovery of steady-state walking may be affected by other constraints increasing the recovery time at non-comfortable speeds. In a recent study using a similar paradigm, we found a relationship between functional walking performance and recovery of double-support duration in post-stroke individuals only when walking faster, and in healthy older adults when walking slower than comfortable [4], supporting the notion that non-comfortable speeds, either faster or slower, increase the challenge to the locomotor system in the longer-term. 4.1.1. Age effects on responses to perturbation at different speeds In our study, younger and older adults used similar leg strategies immediately following perturbation at different gait speeds. These results extend previous findings from constantspeed or self-paced treadmill walking [8,10] as well as overground walking with a visible obstacle [24], suggesting that the use of speed appropriate strategies immediately following perturbation is unaffected in healthy aging. However, in the longer-term following perturbation, older adults had larger leg phase shifts and took longer to recover steady-state walking. These results extend previous results suggesting that older adults may be less stable than young adults even in the absence of an external perturbation [27] and suggest that the long-term, rather than the short-term response to a perturbation may be associated with fall risk in older adults, regardless of gait speed. In contrast to our hypothesis, increasing gait speed did not significantly impair the ability of older adults to overcome perturbations, nor did decreasing gait speed improve this ability. Slower walkers among older adults are those who are more prone to falls [28], and falls during slower walking may involve more serious injury (i.e., hip fractures) [29]. Thus, it was suggested that walking slower in older adults is an attempt to increase stability [30]. However, our results show that slower walking did not assist long-term recovery in case of an unexpected perturbation, such that decreased gait speed in older adults may not be protective in terms of stability. Older adults also had less leg–leg coordination at all speeds. Leg–leg coordination is more variable in older adults even in the absence of external perturbation[13], and maintenance of interlimb coupling in older adults, especially following perturbation, may require increased attention[12]. Future studies may determine to what extent the ability to recover from a perturbation depends on cognitive resources in young and older adults. 4.2. Limitations Cadence-controlled treadmill gait may differ from overground walking and use of a metronome to stabilize gait speed may further modify gait characteristics, especially in older adults. However, metronome usage was necessary in order to stabilize gait speed across trials given that the treadmill was self-paced. Furthermore, subjects were not required to follow metronome pacing after the perturbation and did not report that the pacing was overly challenging. The duration of the perturbation was maintained across gait speeds such that the perturbation lasted for a longer percentage of swing duration in the faster speed condition. However, only 12% of the difference in gait speed was attributed to swing duration, and a control experiment (n = 3 adults; unpublished data) determined that 20% (50 ms) changes in perturbation duration did not affect leg strategy. Thus, the difference in strategy between gait speeds should not be attributed to this limitation. Future work may determine the effectiveness of specific leg strategies in young and older adults across gait speeds.
Acknowledgements Thanks extended to R. Guberek and R. Dannenbaum for assistance with subject recruitment and evaluation, to Revital R. ˜ a for assistance with data collection, Hacmon and Melanie C. Banin to G. Chilingaryan for statistical advice and to study participants. This work was supported by CHRP, NSERC and FQRNT (Canada). TK’s doctoral studies were supported by CIHR Locomotor Team Grant and Heart and Stroke Foundation of Canada (Focus on Stroke). MFL holds a Tier 1 Canada Research Chair in Motor Recovery and Rehabilitation. AL is supported by a FRSQ Junior 2 Salary Award. Conflict of interest statement The authors of the manuscript ‘‘Effects of walking speed on gait stability and interlimb coordination in younger and older adults’’ wish to state that none of the authors has a conflict of interest with any commercial interests. References [1] Berg WP, Alessio HM, Mills EM, Tong C. Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 1997;26: 261–8. [2] England SA, Granata KP. The influence of gait speed on local dynamic stability of walking. Gait Posture 2007;25:172–8. [3] Pavol MJ, Owings TM, Foley KT, Grabiner MD. Gait characteristics as risk factors for falling from trips induced in older adults. J Gerontol A Biol Sci Med Sci 1999;54:M583–90. [4] Krasovsky T, Lamontagne A, Feldman AG. Levin MF Reduced gait stability in high-functioning poststroke individuals. J Neurophysiol 2013;109:77–88. [5] Eng JJ, Winter DA, Patla AE. Strategies for recovery from a trip in early and late swing during human walking. Exp Brain Res 1994;102:339–49. [6] de Boer T, Wisse M, van der, Helm FCT. Mechanical analysis of the preferred strategy selection in human stumble recovery. J Biomech Eng 2010;132: 071012–018. [7] Forner Cordero A, Koopman HJFM, van der Helm. FCT Energy analysis of human stumbling: the limitations of recovery. Gait Posture 2005;21:243–54. ˜ a MC, Feldman AG, Lamontagne A, Levin MF. Stability of gait [8] Krasovsky T, Banin and interlimb coordination in older adults. J Neurophysiol 2012;107:2560–9. [9] Pijnappels M, Bobbert MF, van diee¨n JH. Control of support limb muscles in recovery after tripping in young and older subjects. Exp Brain Res 2005;160: 326–33. [10] Schillings AM, Mulder T, Duysens J. Stumbling over obstacles in older adults compared to young adults. J Neurophysiol 2005;94:1158–68. ˜ a MC, Lamontagne A, Levin MF. Changes in the [11] Feldman AG, Krasovsky T, Banin referent body location and configuration may underlie human gait, as confirmed by findings of multi-muscle activity minimizations and phase resetting. Exp Brain Res 2011;210:91–115. [12] Serrien DJ, Swinnen SP, Stelmach GE. Age-related deterioration of coordinated interlimb behavior. J Gerontol B Psychol Sci Soc Sci 2000;55: 295–303. [13] Plotnik M, Giladi N, Hausdorff J. A new measure for quantifying the bilateral coordination of human gait: effects of aging and Parkinson’s disease. Exp Brain Res 2007;181:561–70. [14] Wagenaar RC, van. Emmerik REA Resonant frequencies of arms and legs identify different walking patterns. J Biomech 2000;33:853–61. [15] Meyns P, Van Gestel L, Bruijn SM, Desloovere K, Swinnen SP, Duysens J. Is interlimb coordination during walking preserved in children with cerebral palsy. Res Dev Disabil 2012;33:1418–28. [16] CircStat. Berens P. A MATLAB toolbox for circular statistics. J Stat Soft 2009;31: 1–20. [17] Forner Cordero A, Koopman HFJM, van der Helm FCT. Multiple-step strategies to recover from stumbling perturbations. Gait Posture 2003;18:47–59. [18] Scholz JP, Kelso JAS, Scho¨ner G. Nonequilibrium phase transitions in coordinated biological motion: critical slowing down and switching time. Phys Lett A 1987;123:390–4. [19] Bruijn SM, Meijer OG, Beek PJ, van Diee¨n JH. Assessing the stability of human locomotion: a review of current measures. J Royal Soc Interface 2013;10. [20] Bruijn SM, van diee¨n JH, Meijer OG, Beek PJ. Is slow walking more stable? J Biomech 2009;42:1506–12. [21] Bhatt T, Wening JD, Pai Y-C. Influence of gait speed on stability: recovery from anterior slips and compensatory stepping. Gait Posture 2005;21:146–56. [22] Bruijn SM, Meijer OG, Beek PJ, van diee¨n JH. The effects of arm swing on human gait stability. J Exp Biol 2010;213:3945–52. [23] Weerdesteyn V, Nienhuis B, Mulder T, Duysens J. Older women strongly prefer stride lengthening to shortening in avoiding obstacles. Exp Brain Res 2005;161:39–46.
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385 [24] Chen HC, Ashton-Miller JA, Alexander NB, Schultz AB. Age effects on strategies used to avoid obstacles. Gait Posture 1994;2:139–46. [25] Hultborn H, Conway BA, Gossard JP, Brownstone R, Fedirchuk B, Schomburg ED, Enriquez-Denton M, Perreault MC. How do we approach the locomotor network in the mammalian spinal cord? Ann NY Acad Sci 1998;860:70–82. [26] Jordan K, Newell KM. The structure of variability in human walking and running is speed-dependent. Exerc Sport Sci Rev 2008;36:200–4.
385
[27] Kang HG, Dingwell JB. Effects of walking speed, strength and range of motion on gait stability in healthy older adults. J Biomech 2008;41:2899–905. [28] Luukinen H, Koski K, Laippala P, Kivela¨ SL. Predictors for recurrent falls among the home-dwelling elderly. Scand J Prim Health Care 1995;13:294–9. ˜ , Hayes WC, McMahon TA. Disturbance type and gait speed affect [29] Smeesters CA fall direction and impact location. J Biomech 2001;34:309–17. [30] Maki BE. Gait changes in older adults: Predictors of falls or indicators of fear. J Am Geriatr Soc 1997;45:313–20.
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Effects of walking speed on gait stability and interlimb coordination in younger and older adults Tal Krasovsky a,c, Anouk Lamontagne a,c, Anatol G. Feldman b,c, Mindy F. Levin a,c,* a
School of Physical and Occupational Therapy, McGill University, Montreal, Quebec, Canada Department of Physiology, Universite´ de Montre´al, Montreal, Quebec, Canada c Centre for Interdisciplinary Research in Rehabilitation, Montreal, Quebec, Canada b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 February 2013 Received in revised form 12 June 2013 Accepted 9 August 2013
Many falls in older adults occur during walking following trips. Following a trip, older adults take longer than younger adults to recover steady-state walking. Although faster gait speed may improve interlimb coordination, it may also increase fall risk in older adults. We hypothesized that older adults would take longer than younger adults to recover from an unexpected perturbation during gait especially when walking faster. Twelve younger (26.3 4.4 years) and 12 older adults (68.5 3.4 years) walked at comfortable, faster and slower speeds when movement of the dominant leg was unexpectedly arrested for 250 ms at 20% swing length. Gait stability was evaluated using the short- and longer-term response to perturbation. In both groups, walking faster diminished the occurrence of elevation and increased that of leg lowering. Older adults took longer than younger adults to recover steady-state walking at all speeds (3.36 0.11 vs. 2.89 0.08 strides) but longer-term recovery of gait stability was not related to gait speed. Arm-leg and inter-arm coordination improved with increasing gait speed in both groups, but older adults had weaker inter-leg coupling following perturbation at all speeds. Although both younger and older adults used speed appropriate responses immediately following perturbation, longer duration of recovery of steady-state walking in older adults may increase fall risk in uncontrolled situations, regardless of gait speed. Recovery from perturbation when walking faster was associated with better interlimb coordination, but not with better gait stability. This indicates that interlimb coordination and gait stability may be distinct features of locomotion. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Perturbation Central pattern generators Locomotion Aging
1. Introduction A leading cause of injury in older adults is falling, causing a heavy burden on the health-care system. Most falls in older adults occur when walking, specifically following a trip [1]. Avoidance of falling is related to gait stability, defined as the ability to maintain functional locomotion despite disturbances [2]. When walking faster, responses to unexpected perturbations need to be performed within shorter timeframes, which can increase fall risk especially among older adults [3]. Longer recovery times toward steady-state walking patterns following a perturbation may reveal lower stability properties of the unperturbed gait pattern in the absence of a fall [4]. Young healthy adults use different response strategies following a trip to prevent falling. Immediate lowering of the perturbed leg
* Corresponding author at: School of Physical and Occupational Therapy, 3654 Promenade Sir William Osler, Montreal, Quebec, Canada H3G 1Y5. Tel.: +1 514 398 3994; fax: +1 514 398 6360. E-mail address: [email protected] (M.F. Levin). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.08.011
(lowering strategy) allows a faster transition to a double-support phase. While this may be safer in the short-term [5], this strategy also requires a rapid transition of the non-perturbed leg from stance to swing, which is relatively energy-demanding [6,7] and associated with longer recovery times toward a steady-state walking pattern [8]. An alternative strategy (elevation) is to stabilize the body on the non-perturbed leg and complete the perturbed step with minimal disturbance. This strategy requires early muscle recruitment on the non-perturbed leg [9] and allows faster recovery of the steady-state walking pattern [8]. A third strategy (combined) is a combination of lowering and elevation [7]. Changing gait speed affects the available response time and may also affect strategy selection. Younger and older adults use similar leg strategies when gait speed is constrained by a treadmill [10] and when walking at their comfortable speed [8]. However, due to limitations in the magnitude and rate of muscle recruitment in older adults [9], walking faster may be associated with an increased occurrence of leg lowering, as well as a longer period of recovery toward a steady-state walking pattern. Recovery from a trip involves the re-establishment of a coordinated gait rhythm in all four limbs [8,11]. Generally,
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
379
Table 1 Anthropometric and kinematic data for young (N = 12) and older (N = 12) adults.
Age (years) Gender (M/F) Height (cm) Weight (kg) Gait speed (m/s) Swing length (cm) Swing time (s)
Slower 0.77 0.06 61.81 4.16 0.52 0.02
Young (N = 12)
Older (N = 12)
26.3 4.4 7 M/5 F 172.7 8.5 71.7 8.5 Comfortable 1.05 0.06 65.52 4.32 0.44 0.02
68.5 3.4 8 M/4 F 166.5 10.0 70.0 13.6 Comfortable 1.10 0.14 65.55 4.85 0.43 0.02
temporal coupling between movements of different limbs is weaker in older compared to younger adults [12,13], possibly due to altered processing of afferent information and/or cognitive decline [12]. These factors may also affect older adults’ ability to coordinate the response to a unilateral gait perturbation [8]. Increasing gait speed improves interlimb coupling in healthy adults [14], as well as in individuals with central nervous system damage, such as cerebral palsy [15] and stroke [4]. Recovery of interlimb coordination may be more challenging for older adults given the limited response time available at faster speeds. We evaluated the effects of gait speed on recovery of the steadystate walking pattern following a leg perturbation in younger and older adults. We hypothesized that (1) in the short-term, faster walking would involve more leg lowering responses; and in the longer-term, faster walking would result in better interlimb coordination, but longer recovery times toward a steady-state walking pattern; and (2) older adults were expected to take longer than younger adults to restore steady-state walking patterns, especially when walking faster. 2. Methods 2.1. Participants Twelve younger and 12 older adults were recruited (Table 1). Subjects were excluded if they had any orthopedic, vestibular or neurological disorders affecting gait or documented cognitive decline. All subjects signed informed consent forms approved by the local ethics committee. 2.2. Experimental procedure The experimental procedure has been described in detail elsewhere [8]. Subjects walked on a self-paced treadmill while wearing a safety harness that did not restrict trunk motion (Fig. 1A). Safety switches could stop the treadmill if needed. Comfortable gait speed was set once the subject habituated to treadmill walking. Thirty-six 40 s trials were performed in which subjects walked at one of three speeds: comfortable, 20% faster or 20% slower (12 trials each), paced by a digital metronome. Order of speed conditions was fully randomized. The dominant leg, determined as the leg preferred for kicking, was perturbed 1–2 times per trial (20 perturbations in total for each speed) by clamping a wooden rod attached to an ankle cuff. An identical rod was attached to the non-dominant leg to equalize sensation. Rods were equipped with transducers measuring perturbation force. Leg arrests of 250 ms duration occurred at early swing (20% of swing length based on mean swing length of 3 unperturbed strides) using a custom-built microcontroller. To prevent anticipation, 3–4 non-perturbed ‘‘catch’’ trials were randomly interspersed throughout all trials. Markers were placed on body landmarks, specifically on both heels (calcanei) and hands (dorsum of 2nd metacarpals). Movement kinematics were recorded at 120 samples/s using a 12-camera motion analysis
Faster 1.33 0.1 68.94 6.1 0.39 0.01
Slower 0.79 0.11 62.18 6.15 0.52 0.04
Faster 1.30 0.20 65.47 6.99 0.38 0.02
system (Vicon, Oxford, UK) and filtered using a 6th order Butterworth filter (dual-pass, cutoff 10 Hz). 2.3. Data analysis Spatiotemporal gait parameters were calculated from heel marker positions. Indices of gait stability were described in the short (1–2 steps) and longer (>2 steps) term. The short-term response was divided according to three leg strategies (Fig. 1B). In the longer term, an index of recovery of double-support time [4,8] was used to estimate return to steady-state walking. Due to possible asymmetry in double-support times, baseline doublesupport times during unperturbed gait (the time between foot contact and subsequent toe-off of the other leg) were computed separately for each side and averaged across the 3 pre-perturbed strides. The double-support recovery index was defined as the number of gait cycles taken until recovery of baseline values on both sides (plus or minus 10%). To quantify the residual effect of the perturbation, phase shifts were calculated using timing of gait events in the arms and legs before and after perturbation [11]. The sagittal plane position of each limb in 3 pre-perturbed cycles was projected forward and the minimal time between the actual and projected event (foot contact or arm direction reversal) was divided by the pre-perturbed cycle period and multiplied by 3608. A negative value (phase advance; up to 1808) indicated that the gait event occurred prior to its unperturbed occurrence and a positive value (phase delay; up to 1808) denoted a delayed occurrence of the event. Temporal interlimb coupling following perturbation was computed using absolute phase shift differences between limb pairs: arm–leg on perturbed and non-perturbed sides, arm–arm and leg–leg pairs. Values ranged from 08, indicating full restoration of the pre-perturbed coordination pattern, to plus/minus 1808, indicating no coupling and a full shift in the coordination pattern. 2.4. Statistical analysis Student’s t-tests compared subject characteristics between groups. Repeated measures ANOVAs tested the effect of group and speed condition on baseline spatiotemporal walking characteristics, perturbation force, and short- and long-term indices of gait stability and interlimb coordination. Levels of significance in post hoc tests were adjusted using a Bonferroni correction. A circular two-way ANOVA (Harrison–Kanji test) was implemented [16] in Matlab 6.5 (Mathworks, Inc., Natick, MA, USA) for phase shifts. All tests except circular procedures were done using SPSS (version 17.0). Initial significance was P < 0.05. 3. Results Younger and older subjects did not differ in height, weight or gait speed (Table 1). Younger and older adults walked 27% and 28% slower in the slower speed condition and 27% and 18% faster in the faster speed condition respectively compared to their comfortable
380
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
Fig. 1. (A) Experimental setup. Subjects walked on a self-paced treadmill, a clamp unexpectedly arrested a wooden rod attached to their dominant ankle at 20% swing. (B) Phase plots describing leg strategies (young subject, comfortable speed), with sagittal plane velocity on the ordinate and position on the abscissa. Thin solid lines depict 3 preperturbed gait cycles, red traces depict the perturbed cycle, and dashed lines depict 2 post-perturbed cycles. A lowering strategy (top panel): lowering of the perturbed leg and a faster, shorter step in the contralateral leg. An elevation strategy (bottom panel): prolonged maintenance of the leg in the air, followed by a non-perturbed step in the contralateral leg. The combined strategy (middle) is a combination of the two. Reprinted with permission from Journal of Neurophysiology, 2012 [8].
speed, with no group differences or interactions between speed and group. None of the subjects fell during the session and no adverse effects were reported. Perturbation force (Fig. 3A) did not differ between groups, but increased with speed (F2,44 = 30.3, P < 0.01).
3.1. Effects of gait speed (Hypothesis 1) For each subject, frequency of occurrence of each leg strategy was computed (Fig. 2). At comfortable speed, use of each strategy was comparable to previous studies [8,17]. The slower speed
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
condition was associated with less lowering (F2,44 = 29.6, P < 0.01) and more leg elevation strategies (F2,44 = 11.6, P < 0.01) than the comfortable and faster speeds. In the longer-term, however, there was no main effect of speed condition on recovery of doublesupport duration. Phase shifts in the arms and legs increased with gait speed (Fig. 3C–F, perturbed leg: F2,66 = 5.89, P < 0.01; nonperturbed leg: F2,66 = 5.42, P < 0.01; ipsilateral arm: F2,66 = 16.99, P < 0.01; contralateral arm: F2,66 = 20.20, P < 0.01). Post hoc analyses revealed that for all limbs, this difference occurred between the slower and faster speeds. For all limbs except the contralateral arm, phase shifts at slower speed were also significantly smaller than those at comfortable speed. Smaller phase-shift differences in the arms and arm-leg pairs were associated with increasing gait speed (Fig. 4A–C, arm–leg pair – perturbed side: F2,44 = 42.57, P < 0.01; arm–leg pair – nonperturbed side F1,22 = 22.65, P < 0.01; arm-arm coupling: F1,22 = 37.63, P < 0.01). Leg–leg coupling did not vary with gait speed. 3.2. Effects of age (Hypothesis 2) In the short-term, young and older adults used similar strategies (Fig. 2). In the longer-term, older adults required more
strides to recover double-support duration (3.36 0.11 compared to 2.89 0.08 strides in younger adults, F1,22 = 5.89, P < 0.03, Fig. 3B). Older adults also had larger phase shifts in the legs (Fig. 3C and D, perturbed leg: 47.78 vs. 16.98, F1,66 = 8.67, P < 0.01, non-perturbed leg: 50.98 vs. 17.68, F1,66 = 9.81, P < 0.01) but phase shifts in the arms did not differ between groups ( 49.28 vs. 20.98 for the ipsilateral arm, and 41.1 vs. 19.08 for the contralateral arm). Finally, older adults had larger phase-shift differences between the legs compared to younger adults (Figure 4D, F1,22 = 8.20, P < 0.01). 4. Discussion The stability properties of gait in young and older adults were evaluated by examining responses to unilateral gait perturbations. Response to perturbation in the short- and longer-term represents the stability properties of the underlying gait pattern, which can be considered as a stable attractor[4,18]. Thus, decreased gait stability may be manifested by more frequent use of leg lowering in the short-term, and longer recovery times toward steady-state walking patterns in the longer-term. Gait stability may be quantified in different ways using responses to various external or internal perturbations [19] with the role of gait speed in gait stability changing accordingly [20].
Young
80
80
60
60
40
40
20
20
0
0
100
100
80
80
60
60
40
40
20
20
0
0
100
100
80
80
60
60
40
40
20
20
Comfortable Faster
% Cases
Slower
100
1
2
Elevaon Combined Lowering
Older
100
0
381
3
4
5
6
7
8
Subjects
9 10 11 12
0
1
2
3
4
5
6
7
8
9 10 11 12
Subjects
Fig. 2. Leg response strategies immediately following perturbation. Stacked bars represent the percentage of trials where a strategy was used for each subject. Subjects are numbered (1–12) according to treadmill gait speed (from lowest to highest).
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
Perturbation force (N)
A
**
**
B
**
60 50 40 30 20 10
Young Older
* **
p<0.05 p<0.01
Double-support recovery (cycles)
382
4.0
3.5
3.0
2.5 * Group 2.0
0
Phase shifts C
**
D
**
0 -20 -40 -60 -80
** Group
-100
0 -20 -40 -60 -80
** Group
-100
E
*
**
F
20
** 20
Contralateral arm (° )
Ipsilateral arm (° )
**
20
Non-perturbed leg (° )
Perturbed leg (° )
20
**
0 -20 -40 -60
0 -20 -40 -60
-80
-80
-100
-100 Slower
Comfortable
Faster
Speed
Slower
Comfortable
Faster
Speed
Fig. 3. Perturbation force and longer-term responses of younger (black) and older (gray) adults. (A) Average perturbation force (N). B. Gait cycles required until doublesupport time recovered to baseline 10%. (C–F) Phase shifts in the legs (C and D) and arms (E and F) following perturbation. Error bars mark 1SEM. Asterisks mark significant post hoc differences.
For example, following a slip walking faster was associated with better stability [21] while a trunk perturbation at faster speeds was associated with slower recovery of baseline center-ofmass movement [22]. Using responses to a trip perturbation as a probe to gait stability given their functional significance for older adults [1], this study focused on one aspect of this complex construct.
4.1. Gait speed effects on short- and longer-term responses to perturbation Our results showed that both younger and older adults used more lowering and less elevation strategies when increasing gait speed (Figs. 1 and 2). To use elevation, body weight needs to be maintained on the unperturbed leg [9]. This strategy may be
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
383
Phase shift differences **
Perturbed side (deg)
60
**
B *
50 40 30 20 10
Young Older
* **
**
60
p<0.05 p<0.01
Non-perturbed side (deg)
A
0
**
50 40 30 20 10 0
C
**
D
**
20
Legs (deg)
Arms (deg)
60
40
20
15
10
5
** Group 0
0 Slower
Comfortable
Faster
Speed
Slower
Comfortable
Faster
Speed
Fig. 4. Phase shift differences between limbs for younger (black) and older (gray) adults. (A and B) Arm-leg phase shift differences for the perturbed/dominant and nonperturbed/non-dominant sides. (C and D) Arm–arm and leg–leg phase shift differences. Error bars mark 1SEM. Asterisks mark significant post hoc differences.
preferred by young and older adults [23,24], since it allows faster recovery toward steady-state walking [8]. However when an unexpected perturbation occurs during faster walking, there is less time to perform this adjustment, increasing the challenge to the sensorimotor system. Under these conditions, lowering the leg can be considered as safer in the short-term, since it immediately increases the base of support [5], although it is a less energyefficient strategy [7] and associated with a longer period of instability in the longer-term. Indeed, in people with central nervous system damage (stroke), a similar perturbation resulted in frequent use of lowering regardless of gait speed [4]. Faster walking requires greater control of the increased forward body momentum, and older adults who walked faster were more likely to fall than slower walkers following an overground trip [3].Our results showed that despite the increased challenge, older adults are able to control the increased forward body momentum using speed-appropriate strategies. In the longer-term after perturbation, walking faster was associated with larger phase shifts in arms and legs in both groups. Previous results showed that leg lowering is associated with larger phase shifts in all four limbs when walking at comfortable speed [8]. Results from this study further suggest that in younger and older adults, global phase resetting occurs also at faster or slower speeds. Similar phase shifts in all four limbs may indicate that a central timing mechanism, i.e., central pattern generator (CPG), is reset to accommodate for the perturbation [25]. Feldman et al. [11]
further suggested that the CPG is driven by a hierarchically superior mechanism that determines the translation rate of the body balance in the environment, influencing gait speed. Thereby, global long-term phase shifts result from a transient change in the rate of translation of the body in response to perturbation. Our results also showed that temporal arm–leg and arm–arm coupling improved (smaller phase-shift differences) with faster gait speed. Better limb coupling is usually related to greater stability of a coordinative pattern [15], such that a perturbation will be associated with less disruption [18]. Our results suggest that the ability to maintain a consistent coordination pattern following perturbation is greater when walking faster. However, consistency of the coordinative pattern should not be equated with stability of steady-state walking since faster walking was not associated with faster recovery of double-support duration. In other words, interlimb coordination and gait stability are different aspects of locomotion which may be contrasted in their relation to gait speed. Our results further showed that short- and longer-term responses to perturbation were affected differently by gait speed. While faster walking was associated with more frequent leg lowering, no longer-term relationship was found between gait speed and the time required to recover steady-state walking (Fig. 3B). This may be related to the constraints applied to the motor system when walking at non-comfortable speeds. At comfortable speed, the lower inter-stride dependency of
384
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385
spatiotemporal gait parameters suggests that gait may be more adaptable to environmental changes [26]. Thus, while strategy selection may be directly associated with gait speed, the longterm recovery of steady-state walking may be affected by other constraints increasing the recovery time at non-comfortable speeds. In a recent study using a similar paradigm, we found a relationship between functional walking performance and recovery of double-support duration in post-stroke individuals only when walking faster, and in healthy older adults when walking slower than comfortable [4], supporting the notion that non-comfortable speeds, either faster or slower, increase the challenge to the locomotor system in the longer-term. 4.1.1. Age effects on responses to perturbation at different speeds In our study, younger and older adults used similar leg strategies immediately following perturbation at different gait speeds. These results extend previous findings from constantspeed or self-paced treadmill walking [8,10] as well as overground walking with a visible obstacle [24], suggesting that the use of speed appropriate strategies immediately following perturbation is unaffected in healthy aging. However, in the longer-term following perturbation, older adults had larger leg phase shifts and took longer to recover steady-state walking. These results extend previous results suggesting that older adults may be less stable than young adults even in the absence of an external perturbation [27] and suggest that the long-term, rather than the short-term response to a perturbation may be associated with fall risk in older adults, regardless of gait speed. In contrast to our hypothesis, increasing gait speed did not significantly impair the ability of older adults to overcome perturbations, nor did decreasing gait speed improve this ability. Slower walkers among older adults are those who are more prone to falls [28], and falls during slower walking may involve more serious injury (i.e., hip fractures) [29]. Thus, it was suggested that walking slower in older adults is an attempt to increase stability [30]. However, our results show that slower walking did not assist long-term recovery in case of an unexpected perturbation, such that decreased gait speed in older adults may not be protective in terms of stability. Older adults also had less leg–leg coordination at all speeds. Leg–leg coordination is more variable in older adults even in the absence of external perturbation[13], and maintenance of interlimb coupling in older adults, especially following perturbation, may require increased attention[12]. Future studies may determine to what extent the ability to recover from a perturbation depends on cognitive resources in young and older adults. 4.2. Limitations Cadence-controlled treadmill gait may differ from overground walking and use of a metronome to stabilize gait speed may further modify gait characteristics, especially in older adults. However, metronome usage was necessary in order to stabilize gait speed across trials given that the treadmill was self-paced. Furthermore, subjects were not required to follow metronome pacing after the perturbation and did not report that the pacing was overly challenging. The duration of the perturbation was maintained across gait speeds such that the perturbation lasted for a longer percentage of swing duration in the faster speed condition. However, only 12% of the difference in gait speed was attributed to swing duration, and a control experiment (n = 3 adults; unpublished data) determined that 20% (50 ms) changes in perturbation duration did not affect leg strategy. Thus, the difference in strategy between gait speeds should not be attributed to this limitation. Future work may determine the effectiveness of specific leg strategies in young and older adults across gait speeds.
Acknowledgements Thanks extended to R. Guberek and R. Dannenbaum for assistance with subject recruitment and evaluation, to Revital R. ˜ a for assistance with data collection, Hacmon and Melanie C. Banin to G. Chilingaryan for statistical advice and to study participants. This work was supported by CHRP, NSERC and FQRNT (Canada). TK’s doctoral studies were supported by CIHR Locomotor Team Grant and Heart and Stroke Foundation of Canada (Focus on Stroke). MFL holds a Tier 1 Canada Research Chair in Motor Recovery and Rehabilitation. AL is supported by a FRSQ Junior 2 Salary Award. Conflict of interest statement The authors of the manuscript ‘‘Effects of walking speed on gait stability and interlimb coordination in younger and older adults’’ wish to state that none of the authors has a conflict of interest with any commercial interests. References [1] Berg WP, Alessio HM, Mills EM, Tong C. Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 1997;26: 261–8. [2] England SA, Granata KP. The influence of gait speed on local dynamic stability of walking. Gait Posture 2007;25:172–8. [3] Pavol MJ, Owings TM, Foley KT, Grabiner MD. Gait characteristics as risk factors for falling from trips induced in older adults. J Gerontol A Biol Sci Med Sci 1999;54:M583–90. [4] Krasovsky T, Lamontagne A, Feldman AG. Levin MF Reduced gait stability in high-functioning poststroke individuals. J Neurophysiol 2013;109:77–88. [5] Eng JJ, Winter DA, Patla AE. Strategies for recovery from a trip in early and late swing during human walking. Exp Brain Res 1994;102:339–49. [6] de Boer T, Wisse M, van der, Helm FCT. Mechanical analysis of the preferred strategy selection in human stumble recovery. J Biomech Eng 2010;132: 071012–018. [7] Forner Cordero A, Koopman HJFM, van der Helm. FCT Energy analysis of human stumbling: the limitations of recovery. Gait Posture 2005;21:243–54. ˜ a MC, Feldman AG, Lamontagne A, Levin MF. Stability of gait [8] Krasovsky T, Banin and interlimb coordination in older adults. J Neurophysiol 2012;107:2560–9. [9] Pijnappels M, Bobbert MF, van diee¨n JH. Control of support limb muscles in recovery after tripping in young and older subjects. Exp Brain Res 2005;160: 326–33. [10] Schillings AM, Mulder T, Duysens J. Stumbling over obstacles in older adults compared to young adults. J Neurophysiol 2005;94:1158–68. ˜ a MC, Lamontagne A, Levin MF. Changes in the [11] Feldman AG, Krasovsky T, Banin referent body location and configuration may underlie human gait, as confirmed by findings of multi-muscle activity minimizations and phase resetting. Exp Brain Res 2011;210:91–115. [12] Serrien DJ, Swinnen SP, Stelmach GE. Age-related deterioration of coordinated interlimb behavior. J Gerontol B Psychol Sci Soc Sci 2000;55: 295–303. [13] Plotnik M, Giladi N, Hausdorff J. A new measure for quantifying the bilateral coordination of human gait: effects of aging and Parkinson’s disease. Exp Brain Res 2007;181:561–70. [14] Wagenaar RC, van. Emmerik REA Resonant frequencies of arms and legs identify different walking patterns. J Biomech 2000;33:853–61. [15] Meyns P, Van Gestel L, Bruijn SM, Desloovere K, Swinnen SP, Duysens J. Is interlimb coordination during walking preserved in children with cerebral palsy. Res Dev Disabil 2012;33:1418–28. [16] CircStat. Berens P. A MATLAB toolbox for circular statistics. J Stat Soft 2009;31: 1–20. [17] Forner Cordero A, Koopman HFJM, van der Helm FCT. Multiple-step strategies to recover from stumbling perturbations. Gait Posture 2003;18:47–59. [18] Scholz JP, Kelso JAS, Scho¨ner G. Nonequilibrium phase transitions in coordinated biological motion: critical slowing down and switching time. Phys Lett A 1987;123:390–4. [19] Bruijn SM, Meijer OG, Beek PJ, van Diee¨n JH. Assessing the stability of human locomotion: a review of current measures. J Royal Soc Interface 2013;10. [20] Bruijn SM, van diee¨n JH, Meijer OG, Beek PJ. Is slow walking more stable? J Biomech 2009;42:1506–12. [21] Bhatt T, Wening JD, Pai Y-C. Influence of gait speed on stability: recovery from anterior slips and compensatory stepping. Gait Posture 2005;21:146–56. [22] Bruijn SM, Meijer OG, Beek PJ, van diee¨n JH. The effects of arm swing on human gait stability. J Exp Biol 2010;213:3945–52. [23] Weerdesteyn V, Nienhuis B, Mulder T, Duysens J. Older women strongly prefer stride lengthening to shortening in avoiding obstacles. Exp Brain Res 2005;161:39–46.
T. Krasovsky et al. / Gait & Posture 39 (2014) 378–385 [24] Chen HC, Ashton-Miller JA, Alexander NB, Schultz AB. Age effects on strategies used to avoid obstacles. Gait Posture 1994;2:139–46. [25] Hultborn H, Conway BA, Gossard JP, Brownstone R, Fedirchuk B, Schomburg ED, Enriquez-Denton M, Perreault MC. How do we approach the locomotor network in the mammalian spinal cord? Ann NY Acad Sci 1998;860:70–82. [26] Jordan K, Newell KM. The structure of variability in human walking and running is speed-dependent. Exerc Sport Sci Rev 2008;36:200–4.
385
[27] Kang HG, Dingwell JB. Effects of walking speed, strength and range of motion on gait stability in healthy older adults. J Biomech 2008;41:2899–905. [28] Luukinen H, Koski K, Laippala P, Kivela¨ SL. Predictors for recurrent falls among the home-dwelling elderly. Scand J Prim Health Care 1995;13:294–9. ˜ , Hayes WC, McMahon TA. Disturbance type and gait speed affect [29] Smeesters CA fall direction and impact location. J Biomech 2001;34:309–17. [30] Maki BE. Gait changes in older adults: Predictors of falls or indicators of fear. J Am Geriatr Soc 1997;45:313–20.