Gait & Posture 40 (2014) 499–503
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Forward and backward locomotion in individuals with dizziness Marcela Davalos-Bichara a, Maria Geraldine Zuniga a, Yuri Agrawal a, John P. Carey a, Michael C. Schubert a,b,* a b
Department of Otolaryngology, Johns Hopkins University School of Medicine, Baltimore, MD, United States Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, Baltimore, MD, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 August 2013 Received in revised form 13 May 2014 Accepted 16 June 2014
The vestibular system plays an important role in locomotion. Individuals with vestibular pathology present with gait abnormalities, which may increase their fall frequency. Backward walking (BW) has been suggested as a predictor of falls in other patient populations; however it has not been studied in individuals with dizziness. Our aims were: (1) to investigate the differences in forward walking (FW) and BW both between and within 3 groups: Healthy controls, individuals with dizziness and vestibular pathology, and individuals with dizziness without vestibular pathology, (2) describe differences in FW and BW between individuals that have fallen and those that have not. We studied 28 healthy controls (mean 53.8 17 years), 21 individuals with pathophysiology of the vestibular system (mean 68.5 13 years), and 18 individuals without a vestibular cause for their dizziness (mean 67.4 17 years). Subjects performed 2 FW and 2 BW trials over the GAITRite walkway. Data on history of falls in the preceding year were collected. We found BW was different to FW within each group. When comparing between groups and correcting for age and gender, only BW velocity (beta = 11.390, p = 0.019), cadence (beta = 8.471, p = 0.021), step time (beta = 0.067, p = 0.007) and stride time (beta = 0.137, p = 0.005) were significantly affected by having dizziness, with no differences in FW characteristics. There were no differences between FW and BW between fallers and non-fallers. BW appears to be a better biomarker than FW for identifying individuals with symptoms of dizziness; though it does not appear to characterize those who fall. ß 2014 Elsevier B.V. All rights reserved.
Keywords: Backward walking Locomotion Fall risk Vestibular dysfunction GAITRite
1. Introduction The vestibular system contributes significantly to locomotion through reflex mediation of gait and gaze stability. It is also critical for spatial orientation and navigation. Individuals with vestibular dysfunction present with gait abnormalities and those with bilateral hypofunction report 50% more falls than healthy individuals [2–4]. The high morbidity and mortality associated with falls cost the US health care system 30 billion dollars in 2010 [5]. Forward locomotion requires an accurate coordination of information from the vestibular, visual, somatosensory, and musculoskeletal systems [6]. Sensory and motor processing mediated through the vestibulospinal and vestibulo-ocular reflexes perform critical roles in locomotion [7–11]. This is
* Corresponding author at: Johns Hopkins Outpatient Center, 601N. Caroline St. Suite 6245, Baltimore, MD 21287, United States. Tel.: +1 410 955 7381; fax: +1 410 614 7222. E-mail addresses: [email protected] (M. Davalos-Bichara), [email protected] (M.G. Zuniga), [email protected] (Y. Agrawal), [email protected] (J.P. Carey), [email protected] (M.C. Schubert). http://dx.doi.org/10.1016/j.gaitpost.2014.06.008 0966-6362/ß 2014 Elsevier B.V. All rights reserved.
evidenced by a slower gait speed in benign paroxysmal positional vertigo (BPPV) [7], poor path integration in vestibular hypofunction [8,9], and increased variability in stance and swing phase characteristics in vestibular schwannoma as well as vestibular neuritis [11]. These gait aberrations may explain the increased fall risk in individuals with vestibular dysfunction, estimated to occur during locomotion nearly fifty percent of the time [12]. Although backward walking (BW) occurs much less frequently than forward walking (FW), it is still necessary for independence in daily life (e.g. removing an item from the oven), and many falls occur while moving in this direction [13,14]. The characteristics of BW have been described simply as reversal of FW [15,16]. Evidence to support this descriptor include a correlation between speed, gait cycle, and step length in both directions [15]. However, some notable differences are reported between the mechanical and temporospatial properties of FW relative to BW [15,16]. For example, compared with FW, BW is characterized by having a slower gait speed, reduced stride length, reduced swing phase, and an increase in time spent in double support [17,18]. Some argue that these differences may be explained by the anatomical constraints of the lower limbs (antero-posterior asymmetry and their multi-jointed nature) [16], while others
500
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
explain it as an inability of the central nervous system to maintain its behavioral goal of conserving the pattern of agonist and antagonist muscle action at the joints [15]. Age too appears to affect BW uniquely relative to FW, as older adults exhibit a larger magnitude of change in temporospatial characteristics during BW. Compared to younger individuals, older adults show greater decrements in gait speed, stride length, and percent of time in swing phase with a greater increment in percent of time spent in stance and double support [17,18]. BW has also has been found to have an increased variability in double support, stride length, step time and swing time when compared to younger adults [14]. Individuals with impaired mobility have even greater disparity between FW and BW. Older adults with impaired mobility and who have fallen, have slower gait speed, shorter stride length, increased time spent in double support, a wider base of support, and an increase in step-time variability in BW when compared to FW. This does not exist in older adults who have not fallen [14]. As a result, the authors suggested BW may be a clinical tool to identify risk of falling in older adults with impaired mobility [14]. In individuals with Parkinson’s disease, BW is characterized by slower gait speed, shorter strides, smaller percent of time spent in swing phase, and larger percent of time spent in double support and stance phases, compared with healthy controls [13]. BW has not yet been studied in individuals with dizziness and balance disorders. The aim of this study was to investigate the temporospatial characteristics of FW and BW in individuals with complaints of dizziness and imbalance compared to healthy controls. Additionally, we were interested in the specific role vestibular pathology may have on any differences between FW and BW; thus we subdivided the individuals into two groups – those with and without a verifiable vestibular pathology but each still reporting dizziness. We defined dizziness to include symptoms of vertigo, imbalance, lightheadedness, wooziness, and/or any headmotion induced sense of malaise. We further sought to identify differences in gait parameters during FW and BW between those individuals that have fallen and those that have not. We hypothesized that individuals with dizziness and imbalance symptoms and a history of falls would have temporospatial gait characteristics indicative of motor impairment during FW and BW with an increased variability compared to those without dizziness or a history of falls. 2. Methods Twenty-eight healthy subjects aged 23–81 years (mean 53.8 17 years), 21 individuals with vestibular dizziness (VDZ), aged 36–89 years (mean 68.5 13 years), and 18 individuals with non-vestibular dizziness (NVDZ), aged 36–94 years (mean 67.4 17 years), were included in the study. All individuals were recruited from the outpatient otolaryngology clinic at our institution. VDZ individuals included those diagnosed with a peripheral vestibular lesion by history and physical exam: Me´nie`re’s disease based on clinical history and audiologic testing; BPPV based on positive Dix-Hallpike in the case of vertical canalithiasis or roll maneuver in the case of horizontal canalithiasis; vestibular hypofunction based on abnormal headimpulse test; and superior semicircular canal dehiscence syndrome based on abnormal CT scan. NVDZ individuals included those with complaints of dizziness but with normal vestibular function based on clinical and vestibular function testing. All study participants gave informed consent. This study was approved by the Institutional Review Board. Both patient groups were screened for cognitive impairment using the Mini-Mental State Examination (MMSE). A threshold score of >24 was used as inclusion criteria. Subjects were also asked for data on history of falls and completed the
Dizziness Handicap Inventory. Each subject was assessed using the Dynamic Gait Index (DGI) (Table 1). The DGI is a validated behavioral measure of fall risk in vestibular hypofunction [19]. The temporospatial characteristics of gait were measured using the GAITRiteTM electronic walkway (CIR Systems, Inc.). The GaitRiteTM walkway contains 13,824 sensors encapsulated in a roll up carpet to produce an active area 2-feet wide and 22-feet long. Footfall information was recorded only on the central region (4.88 m) of the entire 6.7-m long walkway. The GAITRiteTM walkway is a validated and reliable tool for gait analysis comparable with video analysis techniques [20]. Participants were asked to perform two FW and two BW trials. They were instructed to walk at their self-selected speed, look straight ahead (they were not allowed to turn their head during BW), and let their arms swing naturally by their sides. Subjects started and finished the walk 1 meter before and after the GAITRiteTM walkway to prevent premature acceleration and deceleration of speed while walking on/off the GAITRiteTM. Safety during the walking trials was ensured by having an investigator walk beside the subject so as not to lead them. The patient subjects were recruited directly from the clinic and were therefore experiencing symptoms during the trials. 2.1. Statistical analysis Only valid trials were analyzed. We considered a trial valid when it included clear footsteps and in which the participant did not step off the mat or stop walking before the 1 m area beyond the end of the mat. Approximately 15% of all trials were discarded as invalid on this basis. Mean and one standard deviation (SD) error bars were calculated for the GAITRite walkway data, captured during both FW and BW. There was no statistically difference between left foot and right foot data, therefore data from both sides were pooled to increase the number of events [21]. All variables were tested for normality using the Kolmogorov–Smirnov test. FW and BW means were compared using paired-samples Student’s ttests and one-way repeated-measures ANOVA with Bonferroni post hoc tests to control for multiple comparisons. Multiple logistic regressions, correcting for age and gender, were used to further assess how the symptom of dizziness impacts gait parameters for both FW and BW. Only the VDZ and NVDZ groups were included in the falls analysis. Independent sample t-tests were used to assess the difference in temporospatial characteristics between fallers and non-fallers for both FW and BW. Levene’s test of equality of variances was used to assess if different variances should be assumed or not for each comparison. Statistical analyses were
Table 1 Characteristics of the 3 groups participating in the study.
N Male (n (%)) Age in years MMSE VADL DHI DGI History of falls (n (%))
HS mean (SD)
VDZ mean (SD)
NVDZ mean (SD)
28 12 (43) 53.8 (17)
21 11 (50) 68.5 (13) 28.9 (3.1) 46.9 (20.6) 30.9 (20.1) 15.3 (4.2) 9 (43)
18 11 (61) 67.4 (17) 29.2 (1.7) 41.5 (24.1) 30 (25.6) 17.5 (5.5) 8 (44)
p-value
0.48 0.003* 0.78 0.51 0.37 0.19 1.00
HS: Healthy Subjects; VDZ: Vestibular dizziness; NVDZ: Non-vestibular dizziness; MMSE: Mini-Mental State Examination; VADL: Vestibular Activities of Daily Living; DHI: Dizziness Handicap Inventory; DGI: Dynamic Gait Index. * Statistically significant between the 3 groups (p < 0.05).
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
completed using SPSS (version 18, Chicago, IL). Alpha was set at p < 0.05 to define significance for all analyses.
501
3.4. Backward walking between groups
Our healthy control subjects were significantly younger than both patient groups (p < 0.05) but there was no difference in age between the individuals with and without a vestibular cause for their dizziness (p > 0.05, Table 1). An example of the recordings obtained from one individual in each group walking in both directions is shown in Fig. 1. The mean and SD for each gait variable for FW and BW are shown in Table 2.
Compared with the healthy control subjects, both patient groups had a slower gait velocity, lower cadence, slower stride velocity, reduced step and stride lengths, longer step and stride times, a larger base of support, and spent more time in double support (p < 0.05). We also found a greater variability (SD) for single support time, swing time, and stance time during BW (p < 0.05) (Table 2). However, after correcting for age and gender, the only variables that were significantly different between healthy and dizzy subjects were velocity (beta = 11.390, p = 0.019), cadence (beta = 8.471, p = 0.021), step time (beta = 0.067, p = 0.007) and stride time (beta = 0.137, p = 0.005).
3.1. History of falls
4. Discussion
3. Results
Only the patient groups were assessed for differences in FW and BW characteristics between individuals that reported a fall in the previous year and those that did not. The number of falls between the patient groups was similar; 43% of individuals with a vestibular diagnosis and 44% of those without a known vestibular cause for their dizziness. The ages of individuals who reported a fall were not different than the ages of the non-fallers (mean 71.3 15.1 vs. 65.5 14.4, respectively) (p > 0.05). In addition, there were no differences between fallers and non-fallers in any of the variables tested for FW or BW (p > 0.05). 3.2. Forward and backward walking within each group Healthy Control Subjects: The healthy control subjects had a faster velocity, cadence, stride velocity, larger step length, larger stride length, smaller base of support, less time spent in double support and in stance, and shorter step and stride times in FW than in BW (p < 0.05). There was an increased variability (i.e. greater SD) in every gait variable tested during BW when compared to FW within this group. Vestibular Dizziness: Individuals with a vestibular cause for their dizziness had a faster velocity, higher cadence, faster stride velocity, larger step length and stride length, smaller base of support, and less time spent in double support in FW than in BW (p < 0.05). There was no statistical difference between FW and BW in either step time or stride time (p > 0.05; Table 2, gray-shaded area). This patient group showed increased gait variability in BW when compared to FW for all parameters tested (p < 0.05) except for step length and stride velocity (p > 0.05; Table 2, grayshaded area). Non-Vestibular Dizziness: Individuals with dizziness but no vestibular pathology had a faster velocity, higher cadence, faster stride velocity, larger step length and stride length, smaller base of support, spent less time in double support, and had reduced step and stride times during FW compared with BW (p < 0.05). The patient group showed an increased gait variability in BW when compared to FW for all of the parameters tested (p < 0.05), except for step length, base of support, and stride velocity (p > 0.05; Table 2, gray-shaded area). 3.3. Forward walking between groups Although we found some differences between the groups in FW (Table 2), these were no longer significant after correcting for age and gender (p > 0.05).
[(Fig._1)TD$IG]
4.1. Effect of dizziness on backward walking An important finding of this study is that amongst a heterogeneous population of individuals with complaints of dizziness and imbalance (with or without identifiable vestibular pathology), only BW was found to demonstrate detectable differences in gait characteristics in individuals compared to healthy controls when correcting for age and gender. Of the variables measured during BW, our data suggest that subjects with dizziness have significantly lower velocity, lower cadence, reduced step time, and reduced stride time compared to healthy controls. We believe these gait performance data are related to having dizziness, an impairment that likely costs the individual greater energy consumption (reduced efficiency) when performing less common tasks. Motor learning occurs through repeated exposure to correction of error [22]. Backward walking is a less common direction of locomotion; thereby less time (less exposure) is spent learning how to do it efficiently. Compounding this reduced exposure is the symptom of dizziness that the brain considers to be an impairment, due to the aberrant vestibular (i.e. reflexive, spatial) information. In this scenario, we propose dizziness has a greater impact (negative) during performance of a less common motor task. There is indirect evidence to support this theory in the amount of energy required for certain tasks. It has been suggested that BW has higher energy expenditure than FW based on higher mean EMG activity [15,23]. Additionally during BW there are greater cardiorespiratory demands with higher oxygen consumption and increased heart rate [24,25]. These data suggest that BW might be a biomarker for individuals with symptoms of dizziness. Our data further suggest it is the subjective perception of dizziness and not the objective evidence of vestibular pathophysiology that is a greater contributor to abnormal gait while walking backwards. 4.2. Effect of dizziness on forward walking Prior studies have investigated FW gait in individuals with vestibular pathology and have shown differences relative to healthy controls [7–9,11,26]. Although our data show a similar trend, we did not detect a difference in FW between healthy control and patient subjects. Our patient population included a diverse array of vestibular and non-vestibular pathologies and may explain the apparent discrepancy; previous studies have described such differences in more homogeneous populations. Previous studies have shown an increased variability of time during stance, swing and stride during FW in individuals with vestibular pathologies [23,26]. We corroborate this finding and additionally report a similar increased variability during BW in individuals with dizziness. 4.3. Fall risk
Fig. 1. Illustrated is GAITRite walkway data capture of Forward (FW) and Backward Walking (BW) from one subject each from the three groups; Healthy Control, Vestibular Dizziness, and Non-Vestibular dizziness.
We did not find any differences in BW or FW in those who fell between individuals with (VDZ) and those without (NVDZ)
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
502
Table 2 Spatiotemporal characteristics during forward and backward walking for the 3 groups expressed as mean and 1 Standard Deviation. Healthy subjects
Velocity (cm/s) Cadence (steps/min) Step Time Stride Length (cm) Step Length (cm) Stride time Base of Support (cm) Single support phase (%) Double support phase (%) Swing time (%) Stance time (%) Stride Velocity Step Time SD Stride Length SD Stride Time SD Step Length SD Base of support SD Stride Velocity SD Single support time SD Double support time SD Swing Time SD Stance time SD
Forward mean (SD)
Backward mean (SD)
118.96 (19.47)a,b,c 106.81 (9.83)a,b 0.57 (0.05)a,b 133.74 (17.44)a,b,c 66.75 (8.86)a,b,c 1.13 (0.11)a,b 8.31 (2.97)a,c 35.24 (2.23)a,c 29.67 (4.35)a,c 35.23 (2.19)a,c 64.77 (2.19)a,c 119.33 (19.19)a,b,c 0.01 (0.01)a,c 2.29 (1.88)a,c 0.02 (0.02)a,c 1.67 (1.09)a,b,c 1.68 (1.25)a,b 2.97 (2.59)a 0.01 (0.16)a,b 0.27 (0.02)a 0.014 (0.016)a,b,c 0.016 (0.018)a,c
81.48 101.97 0.60 96.00 47.86 1.19 18.90 33.82 32.80 33.83 66.17 82.08 0.03 6.04 0.04 3.82 2.25 5.49 0.21 0.03 0.02 0.03
(25.59)b,c (13.24)b (0.076)b (26.96)b,c (13.49)b,c (0.15)b,c (3.62)b (3.89)b,c (7.57)b,c (3.87)b,c (3.88)b,c (25.85)b (0.01)c (3.01) (0.02)c (1.70) (1.12) (2.93) (0.11)b,c (0.03)c (0.12)b,c (0.02)b,c
Vestibular dizziness (VDZ)
Non-vestibular dizziness (NVDZ)
Forward mean (SD)
Backward mean (SD)
Forward mean (SD)
98.34 (19.63)a 100.36 (10.01)a 0.60 (0.06) 117.86 (18.17)a 58.66 (9.18)a 1.20 (0.12) 9.35 (4.37)a 34.08 (2.29)a,d 32.03 (4.14)a,d 34.07 (2.19)a,d 65.94 (2.19)a,d 99.14 (19.61)a 0.04 (0.02)a 3.30 (2.02)a 0.04 (0.25)a 2.47 (1.64) 2.62 (1.85)a 3.67 (1.95) 0.02 (0.02)a 0.02 (0.02)a 0.02 (0.01)a 0.025 (0.02)a
55.01 (19.63) 92.32 (12.71) 0.66 (0.11) 70.98 (21.75) 35.33 (11.09) 1.33 (0.22) 21.98 (5.25)d 29.76 (4.44) 40.46 (8.54) 29.76 (4.44) 70.23 (4.45) 55.41 (19.87) 0.03 (0.02) 6.23 (3.18) 0.59 (0.03)d 4.28 (1.72) 2.46 (1.20) 4.90 (2.96) 0.33 (0.18) 0.05 (0.033)d 0.033 (0.02) 0.025 (0.04)
98.59 103.23 0.59 112.78 56.24 1.17 10.42 32.29 35.26 32.29 67.71 99.08 0.06 3.74 0.05 2.74 1.99 3.55 0.02 0.43 0.02 0.04
(37.82)a (11.63)a (0.074)a (39.47)a (19.75)a (0.14)a (2.93)a (4.62)a (8.88)a (4.59)a (4.59)a (38.05)a (0.06)a (2.90) (0.08)a (1.82) (1.33) (2.06) (0.02)a (0.08)a (0.02)a (0.08)a
Backward mean (SD) 46.92 (30. 82) 97.01 (22.43) 0.65 (0.144) 63.02 (33.77) 28.87 (19. 60) 1.31 (0.29) 19.40 (4.52) 27.17 (6.80) 46.91 (14.09) 27.15 (6.71) 27.15 (6.71) 60.03 (66.54) 0.03 (0.06) 9.53 (20.96) 0.09 (0.09) 8.98 (28.53) 2.03 (1.09) 29.58 (148.52) 0.041 (0.03) 0.094 (0.12) 0.04 (0.03) 0.08 (0.08)
Gray: No significant difference between FW and BW within the same group. In FW, there was no significant difference found between healthy and dizzy subjects (VDZ and NVDZ) for any variable after correcting for age and gender. In BW, after correcting for age and gender, the only variables that were significantly different between healthy and dizzy subjects (VDZ and NVDZ) were velocity, cadence, step time, and stride time. a Statistically significant difference between forward walking (FW) and backward walking (BW) within the same group (p < 0.05). b Statistically significant difference within the same direction between healthy subjects and both patients with dizziness (VDZ and NVDZ). c Statistically significant difference within the same direction between HS and patients with non-vestibular dizziness. d Statistically significant difference within the same direction between VDZ and NVDZ.
vestibular pathology when controlling for age and gender. In contrast, Fritz el al. suggested the use of BW as a clinical tool to identify older adults with impaired mobility at risk for fall, given their identification that ‘fallers’ walked slower and had smaller stride lengths than ‘non-fallers’ during FW and BW [14]. They also reported fallers had a BW velocity less than 0.6 m/s and suggested using this measure to screen for fall risk [14]. In contrast, BW velocities in individuals we identified as fallers varied widely (range 0.14–1.1 m/s; mean 0.5 0.28 m/s), with no clear difference compared to non-fallers. Although both studies collected their falls data by asking participants about their history of falls (in the prior 6 months for Fritz study; in the prior 12 months for ours); the study populations were very different (older adults [14] versus individuals with dizziness and imbalance). Another explanation may be that individuals with dizziness already show gait aberrations that exist similarly for FW and BW. Therefore, any differences between fallers and non-fallers within a group of dizzy individuals may not be as apparent.
5. Conclusion Backward walking is different from forward walking within both healthy controls and patients with symptoms of dizziness. When comparing between groups and correcting for age and gender, only velocity, cadence, step time, and stride time were significantly affected by having dizziness; there were no differences in FW characteristics. BW appears to be a better biomarker than FW at identifying individuals with symptoms of dizziness, though it does not appear to characterize those who fall in this population. Acknowledgments Michael C. Schubert was funded in part by NASA through HRP grant NNX10AO19G. NASA did not have an active role in this study. Conflicts of interest: All authors disclose that there is no financial and personal relationships with other people or organizations that could inappropriately influence (bias) this work.
4.4. Limitations Although we controlled for age, it is important to recognize the difference in age between the younger healthy controls and older individuals. Our results in the healthy group are aligned with prior results and previous reports acknowledge the impact of age and gender in both FW and BW [17,18], although there was no difference in falls or age with our patient groups. Our sample size was relatively small and may not be representative of a general population, though we are confident in the pattern of differences illustrated between FW and BW in individuals with dizziness. Future studies of the specific gait abnormalities within various vestibular pathologies in a more homogeneous sample would be valuable.
References [2] Agrawal Y, Carey JP, Della Santina CC, Schubert MC, Minor LB. Disorders of balance and vestibular function in US adults: data from the National Health and Nutrition Examination Survey, 2001-2004. Arch Intern Med 2009;169:938–44. [3] Herdman SJ, Blatt P, Schubert MC, Tusa RJ. Falls in individuals with vestibular deficits. Am J Otol 2000;21:847–51. [4] Agrawal Y, Carey JP, Della Santina CC, Schubert MC, Minor LB. Diabetes, vestibular dysfunction, and falls: analyses from the National Health and Nutrition Examination Survey. Otol Neurotol 2010;31:1445–50. [5] Tinetti ME, Kumar C. The patient who falls: ‘‘It’s always a trade-off’’. JAMA 2010;303:258–66. [6] Winter DA. Human balance and posture control during standing walking. Gait Posture 1995;3:193–214.
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503 [7] Celebisoy N, Bayam E, Gulec F, Kose T, Akyurekli O. Balance in posterior and horizontal canal type benign paroxysmal positional vertigo before and after canalith repositioning maneuvers. Gait Posture 2009;29: 520–3. [8] Cohen HS, Sangi-Haghpeykar H. Walking speed and vestibular disorders in a path integration task. Gait Posture 2011;33:211–3. [9] Arthur JC, Kortte KB, Shelhamer M, Schubert MC. Linear path integration deficits in individuals with abnormal vestibular afference. Seeing Perceiving 2012;25:155–78. [10] Zangemeister WH, Bulgheroni MV, Pedotti A. Normal gait is differentially influenced by the otoliths. J Biomed Eng 1991;13:451–8. [11] Angunsri N, Ishikawa K, Yin M, Omi E, Shibata Y, Saito T, et al. Gait instability caused by vestibular disorders – analysis by tactile sensor. Auris Nasus Larynx 2011;38:462–8. [12] Prudham D, Evans JG. Factors associated with falls in the elderly: a community study. Age Ageing 1981;10:141–6. [13] Hackney ME, Earhart GM. Backward walking in Parkinson’s disease. Mov Disord 2009;24:218–23. [14] Fritz NE, Worstell AM, Kloos AD, Siles AB, White SE, Kegelmeyer DA. Backward walking measures are sensitive to age-related changes in mobility and balance. Gait Posture 2013 Apr;37(4):593–7. [15] Grasso R, Bianchi L, Lacquaniti F. Motor patterns for human gait: backward versus forward locomotion. J Neurophysiol 1998;80:1868–85. [16] Winter DA, Pluck N, Yang JF. Backward walking: a simple reversal of forward walking? J Mot Behav 1989;21:291–305. [17] Laufer Y. Age- and gender-related changes in the temporal-spatial characteristics of forwards and backwards gaits. Physiother Res Int 2003;8: 131–42.
503
[18] Laufer Y. Effect of age on characteristics of forward and backward gait at preferred and accelerated walking speed. J Gerontol A Biol Sci Med Sci 2005;60:627–32. [19] Whitney SL, Hudak MT, Marchetti GF. The dynamic gait index relates to selfreported fall history in individuals with vestibular dysfunction. J Vestib Res 2000;10:99–105. [20] Bilney B, Morris M, Webster K. Concurrent related validity of the GAITRite walkway system for quantification of the spatial and temporal parameters of gait. Gait Posture 2003;17:68–74. [21] Schniepp R, Wuehr M, Neuhaeusser M, Kamenova M, Dimitriadis K, Klopstock T, et al. Locomotion speed determines gait variability in cerebellar ataxia and vestibular failure. Mov Disord 2012;27:125–31. [22] French B, Thomas LH, Leathley MJ, Sutton CJ, McAdam J, Forster A, et al. Repetitive task training for improving functional ability after stroke. Cochrane Database Syst Rev )2007;(October (4)). CD006073. [23] Hicheur H, Terekhov AV, Berthoz A. Intersegmental coordination during human locomotion: does planar covariation of elevation angles reflect central constraints? J Neurophysiol 2006;96:1406–19. [24] Flynn TW, Connery SM, Smutok MA, Zeballos RJ, Weisman IM. Comparison of cardiopulmonary responses to forward and backward walking and running. Med Sci Sports Exerc 1994;26:89–94. [25] Hooper TL, Dunn DM, Props JE, Bruce BA, Sawyer SF, Daniel JA. The effects of graded forward and backward walking on heart rate and oxygen consumption. J Orthop Sports Phys Ther 2004;34:65–71. [26] Perring S, Summers T. Laboratory-free measurement of gait rhythmicity in the assessment of the degree of impairment and the effectiveness of rehabilitation in individuals with vertigo resulting from vestibular hypofunction. Physiol Meas 2007;28:697–705.
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Forward and backward locomotion in individuals with dizziness Marcela Davalos-Bichara a, Maria Geraldine Zuniga a, Yuri Agrawal a, John P. Carey a, Michael C. Schubert a,b,* a b
Department of Otolaryngology, Johns Hopkins University School of Medicine, Baltimore, MD, United States Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, Baltimore, MD, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 August 2013 Received in revised form 13 May 2014 Accepted 16 June 2014
The vestibular system plays an important role in locomotion. Individuals with vestibular pathology present with gait abnormalities, which may increase their fall frequency. Backward walking (BW) has been suggested as a predictor of falls in other patient populations; however it has not been studied in individuals with dizziness. Our aims were: (1) to investigate the differences in forward walking (FW) and BW both between and within 3 groups: Healthy controls, individuals with dizziness and vestibular pathology, and individuals with dizziness without vestibular pathology, (2) describe differences in FW and BW between individuals that have fallen and those that have not. We studied 28 healthy controls (mean 53.8 17 years), 21 individuals with pathophysiology of the vestibular system (mean 68.5 13 years), and 18 individuals without a vestibular cause for their dizziness (mean 67.4 17 years). Subjects performed 2 FW and 2 BW trials over the GAITRite walkway. Data on history of falls in the preceding year were collected. We found BW was different to FW within each group. When comparing between groups and correcting for age and gender, only BW velocity (beta = 11.390, p = 0.019), cadence (beta = 8.471, p = 0.021), step time (beta = 0.067, p = 0.007) and stride time (beta = 0.137, p = 0.005) were significantly affected by having dizziness, with no differences in FW characteristics. There were no differences between FW and BW between fallers and non-fallers. BW appears to be a better biomarker than FW for identifying individuals with symptoms of dizziness; though it does not appear to characterize those who fall. ß 2014 Elsevier B.V. All rights reserved.
Keywords: Backward walking Locomotion Fall risk Vestibular dysfunction GAITRite
1. Introduction The vestibular system contributes significantly to locomotion through reflex mediation of gait and gaze stability. It is also critical for spatial orientation and navigation. Individuals with vestibular dysfunction present with gait abnormalities and those with bilateral hypofunction report 50% more falls than healthy individuals [2–4]. The high morbidity and mortality associated with falls cost the US health care system 30 billion dollars in 2010 [5]. Forward locomotion requires an accurate coordination of information from the vestibular, visual, somatosensory, and musculoskeletal systems [6]. Sensory and motor processing mediated through the vestibulospinal and vestibulo-ocular reflexes perform critical roles in locomotion [7–11]. This is
* Corresponding author at: Johns Hopkins Outpatient Center, 601N. Caroline St. Suite 6245, Baltimore, MD 21287, United States. Tel.: +1 410 955 7381; fax: +1 410 614 7222. E-mail addresses: [email protected] (M. Davalos-Bichara), [email protected] (M.G. Zuniga), [email protected] (Y. Agrawal), [email protected] (J.P. Carey), [email protected] (M.C. Schubert). http://dx.doi.org/10.1016/j.gaitpost.2014.06.008 0966-6362/ß 2014 Elsevier B.V. All rights reserved.
evidenced by a slower gait speed in benign paroxysmal positional vertigo (BPPV) [7], poor path integration in vestibular hypofunction [8,9], and increased variability in stance and swing phase characteristics in vestibular schwannoma as well as vestibular neuritis [11]. These gait aberrations may explain the increased fall risk in individuals with vestibular dysfunction, estimated to occur during locomotion nearly fifty percent of the time [12]. Although backward walking (BW) occurs much less frequently than forward walking (FW), it is still necessary for independence in daily life (e.g. removing an item from the oven), and many falls occur while moving in this direction [13,14]. The characteristics of BW have been described simply as reversal of FW [15,16]. Evidence to support this descriptor include a correlation between speed, gait cycle, and step length in both directions [15]. However, some notable differences are reported between the mechanical and temporospatial properties of FW relative to BW [15,16]. For example, compared with FW, BW is characterized by having a slower gait speed, reduced stride length, reduced swing phase, and an increase in time spent in double support [17,18]. Some argue that these differences may be explained by the anatomical constraints of the lower limbs (antero-posterior asymmetry and their multi-jointed nature) [16], while others
500
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
explain it as an inability of the central nervous system to maintain its behavioral goal of conserving the pattern of agonist and antagonist muscle action at the joints [15]. Age too appears to affect BW uniquely relative to FW, as older adults exhibit a larger magnitude of change in temporospatial characteristics during BW. Compared to younger individuals, older adults show greater decrements in gait speed, stride length, and percent of time in swing phase with a greater increment in percent of time spent in stance and double support [17,18]. BW has also has been found to have an increased variability in double support, stride length, step time and swing time when compared to younger adults [14]. Individuals with impaired mobility have even greater disparity between FW and BW. Older adults with impaired mobility and who have fallen, have slower gait speed, shorter stride length, increased time spent in double support, a wider base of support, and an increase in step-time variability in BW when compared to FW. This does not exist in older adults who have not fallen [14]. As a result, the authors suggested BW may be a clinical tool to identify risk of falling in older adults with impaired mobility [14]. In individuals with Parkinson’s disease, BW is characterized by slower gait speed, shorter strides, smaller percent of time spent in swing phase, and larger percent of time spent in double support and stance phases, compared with healthy controls [13]. BW has not yet been studied in individuals with dizziness and balance disorders. The aim of this study was to investigate the temporospatial characteristics of FW and BW in individuals with complaints of dizziness and imbalance compared to healthy controls. Additionally, we were interested in the specific role vestibular pathology may have on any differences between FW and BW; thus we subdivided the individuals into two groups – those with and without a verifiable vestibular pathology but each still reporting dizziness. We defined dizziness to include symptoms of vertigo, imbalance, lightheadedness, wooziness, and/or any headmotion induced sense of malaise. We further sought to identify differences in gait parameters during FW and BW between those individuals that have fallen and those that have not. We hypothesized that individuals with dizziness and imbalance symptoms and a history of falls would have temporospatial gait characteristics indicative of motor impairment during FW and BW with an increased variability compared to those without dizziness or a history of falls. 2. Methods Twenty-eight healthy subjects aged 23–81 years (mean 53.8 17 years), 21 individuals with vestibular dizziness (VDZ), aged 36–89 years (mean 68.5 13 years), and 18 individuals with non-vestibular dizziness (NVDZ), aged 36–94 years (mean 67.4 17 years), were included in the study. All individuals were recruited from the outpatient otolaryngology clinic at our institution. VDZ individuals included those diagnosed with a peripheral vestibular lesion by history and physical exam: Me´nie`re’s disease based on clinical history and audiologic testing; BPPV based on positive Dix-Hallpike in the case of vertical canalithiasis or roll maneuver in the case of horizontal canalithiasis; vestibular hypofunction based on abnormal headimpulse test; and superior semicircular canal dehiscence syndrome based on abnormal CT scan. NVDZ individuals included those with complaints of dizziness but with normal vestibular function based on clinical and vestibular function testing. All study participants gave informed consent. This study was approved by the Institutional Review Board. Both patient groups were screened for cognitive impairment using the Mini-Mental State Examination (MMSE). A threshold score of >24 was used as inclusion criteria. Subjects were also asked for data on history of falls and completed the
Dizziness Handicap Inventory. Each subject was assessed using the Dynamic Gait Index (DGI) (Table 1). The DGI is a validated behavioral measure of fall risk in vestibular hypofunction [19]. The temporospatial characteristics of gait were measured using the GAITRiteTM electronic walkway (CIR Systems, Inc.). The GaitRiteTM walkway contains 13,824 sensors encapsulated in a roll up carpet to produce an active area 2-feet wide and 22-feet long. Footfall information was recorded only on the central region (4.88 m) of the entire 6.7-m long walkway. The GAITRiteTM walkway is a validated and reliable tool for gait analysis comparable with video analysis techniques [20]. Participants were asked to perform two FW and two BW trials. They were instructed to walk at their self-selected speed, look straight ahead (they were not allowed to turn their head during BW), and let their arms swing naturally by their sides. Subjects started and finished the walk 1 meter before and after the GAITRiteTM walkway to prevent premature acceleration and deceleration of speed while walking on/off the GAITRiteTM. Safety during the walking trials was ensured by having an investigator walk beside the subject so as not to lead them. The patient subjects were recruited directly from the clinic and were therefore experiencing symptoms during the trials. 2.1. Statistical analysis Only valid trials were analyzed. We considered a trial valid when it included clear footsteps and in which the participant did not step off the mat or stop walking before the 1 m area beyond the end of the mat. Approximately 15% of all trials were discarded as invalid on this basis. Mean and one standard deviation (SD) error bars were calculated for the GAITRite walkway data, captured during both FW and BW. There was no statistically difference between left foot and right foot data, therefore data from both sides were pooled to increase the number of events [21]. All variables were tested for normality using the Kolmogorov–Smirnov test. FW and BW means were compared using paired-samples Student’s ttests and one-way repeated-measures ANOVA with Bonferroni post hoc tests to control for multiple comparisons. Multiple logistic regressions, correcting for age and gender, were used to further assess how the symptom of dizziness impacts gait parameters for both FW and BW. Only the VDZ and NVDZ groups were included in the falls analysis. Independent sample t-tests were used to assess the difference in temporospatial characteristics between fallers and non-fallers for both FW and BW. Levene’s test of equality of variances was used to assess if different variances should be assumed or not for each comparison. Statistical analyses were
Table 1 Characteristics of the 3 groups participating in the study.
N Male (n (%)) Age in years MMSE VADL DHI DGI History of falls (n (%))
HS mean (SD)
VDZ mean (SD)
NVDZ mean (SD)
28 12 (43) 53.8 (17)
21 11 (50) 68.5 (13) 28.9 (3.1) 46.9 (20.6) 30.9 (20.1) 15.3 (4.2) 9 (43)
18 11 (61) 67.4 (17) 29.2 (1.7) 41.5 (24.1) 30 (25.6) 17.5 (5.5) 8 (44)
p-value
0.48 0.003* 0.78 0.51 0.37 0.19 1.00
HS: Healthy Subjects; VDZ: Vestibular dizziness; NVDZ: Non-vestibular dizziness; MMSE: Mini-Mental State Examination; VADL: Vestibular Activities of Daily Living; DHI: Dizziness Handicap Inventory; DGI: Dynamic Gait Index. * Statistically significant between the 3 groups (p < 0.05).
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
completed using SPSS (version 18, Chicago, IL). Alpha was set at p < 0.05 to define significance for all analyses.
501
3.4. Backward walking between groups
Our healthy control subjects were significantly younger than both patient groups (p < 0.05) but there was no difference in age between the individuals with and without a vestibular cause for their dizziness (p > 0.05, Table 1). An example of the recordings obtained from one individual in each group walking in both directions is shown in Fig. 1. The mean and SD for each gait variable for FW and BW are shown in Table 2.
Compared with the healthy control subjects, both patient groups had a slower gait velocity, lower cadence, slower stride velocity, reduced step and stride lengths, longer step and stride times, a larger base of support, and spent more time in double support (p < 0.05). We also found a greater variability (SD) for single support time, swing time, and stance time during BW (p < 0.05) (Table 2). However, after correcting for age and gender, the only variables that were significantly different between healthy and dizzy subjects were velocity (beta = 11.390, p = 0.019), cadence (beta = 8.471, p = 0.021), step time (beta = 0.067, p = 0.007) and stride time (beta = 0.137, p = 0.005).
3.1. History of falls
4. Discussion
3. Results
Only the patient groups were assessed for differences in FW and BW characteristics between individuals that reported a fall in the previous year and those that did not. The number of falls between the patient groups was similar; 43% of individuals with a vestibular diagnosis and 44% of those without a known vestibular cause for their dizziness. The ages of individuals who reported a fall were not different than the ages of the non-fallers (mean 71.3 15.1 vs. 65.5 14.4, respectively) (p > 0.05). In addition, there were no differences between fallers and non-fallers in any of the variables tested for FW or BW (p > 0.05). 3.2. Forward and backward walking within each group Healthy Control Subjects: The healthy control subjects had a faster velocity, cadence, stride velocity, larger step length, larger stride length, smaller base of support, less time spent in double support and in stance, and shorter step and stride times in FW than in BW (p < 0.05). There was an increased variability (i.e. greater SD) in every gait variable tested during BW when compared to FW within this group. Vestibular Dizziness: Individuals with a vestibular cause for their dizziness had a faster velocity, higher cadence, faster stride velocity, larger step length and stride length, smaller base of support, and less time spent in double support in FW than in BW (p < 0.05). There was no statistical difference between FW and BW in either step time or stride time (p > 0.05; Table 2, gray-shaded area). This patient group showed increased gait variability in BW when compared to FW for all parameters tested (p < 0.05) except for step length and stride velocity (p > 0.05; Table 2, grayshaded area). Non-Vestibular Dizziness: Individuals with dizziness but no vestibular pathology had a faster velocity, higher cadence, faster stride velocity, larger step length and stride length, smaller base of support, spent less time in double support, and had reduced step and stride times during FW compared with BW (p < 0.05). The patient group showed an increased gait variability in BW when compared to FW for all of the parameters tested (p < 0.05), except for step length, base of support, and stride velocity (p > 0.05; Table 2, gray-shaded area). 3.3. Forward walking between groups Although we found some differences between the groups in FW (Table 2), these were no longer significant after correcting for age and gender (p > 0.05).
[(Fig._1)TD$IG]
4.1. Effect of dizziness on backward walking An important finding of this study is that amongst a heterogeneous population of individuals with complaints of dizziness and imbalance (with or without identifiable vestibular pathology), only BW was found to demonstrate detectable differences in gait characteristics in individuals compared to healthy controls when correcting for age and gender. Of the variables measured during BW, our data suggest that subjects with dizziness have significantly lower velocity, lower cadence, reduced step time, and reduced stride time compared to healthy controls. We believe these gait performance data are related to having dizziness, an impairment that likely costs the individual greater energy consumption (reduced efficiency) when performing less common tasks. Motor learning occurs through repeated exposure to correction of error [22]. Backward walking is a less common direction of locomotion; thereby less time (less exposure) is spent learning how to do it efficiently. Compounding this reduced exposure is the symptom of dizziness that the brain considers to be an impairment, due to the aberrant vestibular (i.e. reflexive, spatial) information. In this scenario, we propose dizziness has a greater impact (negative) during performance of a less common motor task. There is indirect evidence to support this theory in the amount of energy required for certain tasks. It has been suggested that BW has higher energy expenditure than FW based on higher mean EMG activity [15,23]. Additionally during BW there are greater cardiorespiratory demands with higher oxygen consumption and increased heart rate [24,25]. These data suggest that BW might be a biomarker for individuals with symptoms of dizziness. Our data further suggest it is the subjective perception of dizziness and not the objective evidence of vestibular pathophysiology that is a greater contributor to abnormal gait while walking backwards. 4.2. Effect of dizziness on forward walking Prior studies have investigated FW gait in individuals with vestibular pathology and have shown differences relative to healthy controls [7–9,11,26]. Although our data show a similar trend, we did not detect a difference in FW between healthy control and patient subjects. Our patient population included a diverse array of vestibular and non-vestibular pathologies and may explain the apparent discrepancy; previous studies have described such differences in more homogeneous populations. Previous studies have shown an increased variability of time during stance, swing and stride during FW in individuals with vestibular pathologies [23,26]. We corroborate this finding and additionally report a similar increased variability during BW in individuals with dizziness. 4.3. Fall risk
Fig. 1. Illustrated is GAITRite walkway data capture of Forward (FW) and Backward Walking (BW) from one subject each from the three groups; Healthy Control, Vestibular Dizziness, and Non-Vestibular dizziness.
We did not find any differences in BW or FW in those who fell between individuals with (VDZ) and those without (NVDZ)
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503
502
Table 2 Spatiotemporal characteristics during forward and backward walking for the 3 groups expressed as mean and 1 Standard Deviation. Healthy subjects
Velocity (cm/s) Cadence (steps/min) Step Time Stride Length (cm) Step Length (cm) Stride time Base of Support (cm) Single support phase (%) Double support phase (%) Swing time (%) Stance time (%) Stride Velocity Step Time SD Stride Length SD Stride Time SD Step Length SD Base of support SD Stride Velocity SD Single support time SD Double support time SD Swing Time SD Stance time SD
Forward mean (SD)
Backward mean (SD)
118.96 (19.47)a,b,c 106.81 (9.83)a,b 0.57 (0.05)a,b 133.74 (17.44)a,b,c 66.75 (8.86)a,b,c 1.13 (0.11)a,b 8.31 (2.97)a,c 35.24 (2.23)a,c 29.67 (4.35)a,c 35.23 (2.19)a,c 64.77 (2.19)a,c 119.33 (19.19)a,b,c 0.01 (0.01)a,c 2.29 (1.88)a,c 0.02 (0.02)a,c 1.67 (1.09)a,b,c 1.68 (1.25)a,b 2.97 (2.59)a 0.01 (0.16)a,b 0.27 (0.02)a 0.014 (0.016)a,b,c 0.016 (0.018)a,c
81.48 101.97 0.60 96.00 47.86 1.19 18.90 33.82 32.80 33.83 66.17 82.08 0.03 6.04 0.04 3.82 2.25 5.49 0.21 0.03 0.02 0.03
(25.59)b,c (13.24)b (0.076)b (26.96)b,c (13.49)b,c (0.15)b,c (3.62)b (3.89)b,c (7.57)b,c (3.87)b,c (3.88)b,c (25.85)b (0.01)c (3.01) (0.02)c (1.70) (1.12) (2.93) (0.11)b,c (0.03)c (0.12)b,c (0.02)b,c
Vestibular dizziness (VDZ)
Non-vestibular dizziness (NVDZ)
Forward mean (SD)
Backward mean (SD)
Forward mean (SD)
98.34 (19.63)a 100.36 (10.01)a 0.60 (0.06) 117.86 (18.17)a 58.66 (9.18)a 1.20 (0.12) 9.35 (4.37)a 34.08 (2.29)a,d 32.03 (4.14)a,d 34.07 (2.19)a,d 65.94 (2.19)a,d 99.14 (19.61)a 0.04 (0.02)a 3.30 (2.02)a 0.04 (0.25)a 2.47 (1.64) 2.62 (1.85)a 3.67 (1.95) 0.02 (0.02)a 0.02 (0.02)a 0.02 (0.01)a 0.025 (0.02)a
55.01 (19.63) 92.32 (12.71) 0.66 (0.11) 70.98 (21.75) 35.33 (11.09) 1.33 (0.22) 21.98 (5.25)d 29.76 (4.44) 40.46 (8.54) 29.76 (4.44) 70.23 (4.45) 55.41 (19.87) 0.03 (0.02) 6.23 (3.18) 0.59 (0.03)d 4.28 (1.72) 2.46 (1.20) 4.90 (2.96) 0.33 (0.18) 0.05 (0.033)d 0.033 (0.02) 0.025 (0.04)
98.59 103.23 0.59 112.78 56.24 1.17 10.42 32.29 35.26 32.29 67.71 99.08 0.06 3.74 0.05 2.74 1.99 3.55 0.02 0.43 0.02 0.04
(37.82)a (11.63)a (0.074)a (39.47)a (19.75)a (0.14)a (2.93)a (4.62)a (8.88)a (4.59)a (4.59)a (38.05)a (0.06)a (2.90) (0.08)a (1.82) (1.33) (2.06) (0.02)a (0.08)a (0.02)a (0.08)a
Backward mean (SD) 46.92 (30. 82) 97.01 (22.43) 0.65 (0.144) 63.02 (33.77) 28.87 (19. 60) 1.31 (0.29) 19.40 (4.52) 27.17 (6.80) 46.91 (14.09) 27.15 (6.71) 27.15 (6.71) 60.03 (66.54) 0.03 (0.06) 9.53 (20.96) 0.09 (0.09) 8.98 (28.53) 2.03 (1.09) 29.58 (148.52) 0.041 (0.03) 0.094 (0.12) 0.04 (0.03) 0.08 (0.08)
Gray: No significant difference between FW and BW within the same group. In FW, there was no significant difference found between healthy and dizzy subjects (VDZ and NVDZ) for any variable after correcting for age and gender. In BW, after correcting for age and gender, the only variables that were significantly different between healthy and dizzy subjects (VDZ and NVDZ) were velocity, cadence, step time, and stride time. a Statistically significant difference between forward walking (FW) and backward walking (BW) within the same group (p < 0.05). b Statistically significant difference within the same direction between healthy subjects and both patients with dizziness (VDZ and NVDZ). c Statistically significant difference within the same direction between HS and patients with non-vestibular dizziness. d Statistically significant difference within the same direction between VDZ and NVDZ.
vestibular pathology when controlling for age and gender. In contrast, Fritz el al. suggested the use of BW as a clinical tool to identify older adults with impaired mobility at risk for fall, given their identification that ‘fallers’ walked slower and had smaller stride lengths than ‘non-fallers’ during FW and BW [14]. They also reported fallers had a BW velocity less than 0.6 m/s and suggested using this measure to screen for fall risk [14]. In contrast, BW velocities in individuals we identified as fallers varied widely (range 0.14–1.1 m/s; mean 0.5 0.28 m/s), with no clear difference compared to non-fallers. Although both studies collected their falls data by asking participants about their history of falls (in the prior 6 months for Fritz study; in the prior 12 months for ours); the study populations were very different (older adults [14] versus individuals with dizziness and imbalance). Another explanation may be that individuals with dizziness already show gait aberrations that exist similarly for FW and BW. Therefore, any differences between fallers and non-fallers within a group of dizzy individuals may not be as apparent.
5. Conclusion Backward walking is different from forward walking within both healthy controls and patients with symptoms of dizziness. When comparing between groups and correcting for age and gender, only velocity, cadence, step time, and stride time were significantly affected by having dizziness; there were no differences in FW characteristics. BW appears to be a better biomarker than FW at identifying individuals with symptoms of dizziness, though it does not appear to characterize those who fall in this population. Acknowledgments Michael C. Schubert was funded in part by NASA through HRP grant NNX10AO19G. NASA did not have an active role in this study. Conflicts of interest: All authors disclose that there is no financial and personal relationships with other people or organizations that could inappropriately influence (bias) this work.
4.4. Limitations Although we controlled for age, it is important to recognize the difference in age between the younger healthy controls and older individuals. Our results in the healthy group are aligned with prior results and previous reports acknowledge the impact of age and gender in both FW and BW [17,18], although there was no difference in falls or age with our patient groups. Our sample size was relatively small and may not be representative of a general population, though we are confident in the pattern of differences illustrated between FW and BW in individuals with dizziness. Future studies of the specific gait abnormalities within various vestibular pathologies in a more homogeneous sample would be valuable.
References [2] Agrawal Y, Carey JP, Della Santina CC, Schubert MC, Minor LB. Disorders of balance and vestibular function in US adults: data from the National Health and Nutrition Examination Survey, 2001-2004. Arch Intern Med 2009;169:938–44. [3] Herdman SJ, Blatt P, Schubert MC, Tusa RJ. Falls in individuals with vestibular deficits. Am J Otol 2000;21:847–51. [4] Agrawal Y, Carey JP, Della Santina CC, Schubert MC, Minor LB. Diabetes, vestibular dysfunction, and falls: analyses from the National Health and Nutrition Examination Survey. Otol Neurotol 2010;31:1445–50. [5] Tinetti ME, Kumar C. The patient who falls: ‘‘It’s always a trade-off’’. JAMA 2010;303:258–66. [6] Winter DA. Human balance and posture control during standing walking. Gait Posture 1995;3:193–214.
M. Davalos-Bichara et al. / Gait & Posture 40 (2014) 499–503 [7] Celebisoy N, Bayam E, Gulec F, Kose T, Akyurekli O. Balance in posterior and horizontal canal type benign paroxysmal positional vertigo before and after canalith repositioning maneuvers. Gait Posture 2009;29: 520–3. [8] Cohen HS, Sangi-Haghpeykar H. Walking speed and vestibular disorders in a path integration task. Gait Posture 2011;33:211–3. [9] Arthur JC, Kortte KB, Shelhamer M, Schubert MC. Linear path integration deficits in individuals with abnormal vestibular afference. Seeing Perceiving 2012;25:155–78. [10] Zangemeister WH, Bulgheroni MV, Pedotti A. Normal gait is differentially influenced by the otoliths. J Biomed Eng 1991;13:451–8. [11] Angunsri N, Ishikawa K, Yin M, Omi E, Shibata Y, Saito T, et al. Gait instability caused by vestibular disorders – analysis by tactile sensor. Auris Nasus Larynx 2011;38:462–8. [12] Prudham D, Evans JG. Factors associated with falls in the elderly: a community study. Age Ageing 1981;10:141–6. [13] Hackney ME, Earhart GM. Backward walking in Parkinson’s disease. Mov Disord 2009;24:218–23. [14] Fritz NE, Worstell AM, Kloos AD, Siles AB, White SE, Kegelmeyer DA. Backward walking measures are sensitive to age-related changes in mobility and balance. Gait Posture 2013 Apr;37(4):593–7. [15] Grasso R, Bianchi L, Lacquaniti F. Motor patterns for human gait: backward versus forward locomotion. J Neurophysiol 1998;80:1868–85. [16] Winter DA, Pluck N, Yang JF. Backward walking: a simple reversal of forward walking? J Mot Behav 1989;21:291–305. [17] Laufer Y. Age- and gender-related changes in the temporal-spatial characteristics of forwards and backwards gaits. Physiother Res Int 2003;8: 131–42.
503
[18] Laufer Y. Effect of age on characteristics of forward and backward gait at preferred and accelerated walking speed. J Gerontol A Biol Sci Med Sci 2005;60:627–32. [19] Whitney SL, Hudak MT, Marchetti GF. The dynamic gait index relates to selfreported fall history in individuals with vestibular dysfunction. J Vestib Res 2000;10:99–105. [20] Bilney B, Morris M, Webster K. Concurrent related validity of the GAITRite walkway system for quantification of the spatial and temporal parameters of gait. Gait Posture 2003;17:68–74. [21] Schniepp R, Wuehr M, Neuhaeusser M, Kamenova M, Dimitriadis K, Klopstock T, et al. Locomotion speed determines gait variability in cerebellar ataxia and vestibular failure. Mov Disord 2012;27:125–31. [22] French B, Thomas LH, Leathley MJ, Sutton CJ, McAdam J, Forster A, et al. Repetitive task training for improving functional ability after stroke. Cochrane Database Syst Rev )2007;(October (4)). CD006073. [23] Hicheur H, Terekhov AV, Berthoz A. Intersegmental coordination during human locomotion: does planar covariation of elevation angles reflect central constraints? J Neurophysiol 2006;96:1406–19. [24] Flynn TW, Connery SM, Smutok MA, Zeballos RJ, Weisman IM. Comparison of cardiopulmonary responses to forward and backward walking and running. Med Sci Sports Exerc 1994;26:89–94. [25] Hooper TL, Dunn DM, Props JE, Bruce BA, Sawyer SF, Daniel JA. The effects of graded forward and backward walking on heart rate and oxygen consumption. J Orthop Sports Phys Ther 2004;34:65–71. [26] Perring S, Summers T. Laboratory-free measurement of gait rhythmicity in the assessment of the degree of impairment and the effectiveness of rehabilitation in individuals with vertigo resulting from vestibular hypofunction. Physiol Meas 2007;28:697–705.
