- Email: [email protected]
Local dynamic stability and gait variability during attentional tasks in young adults
Gait & Posture 55 (2017) 105–108
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Local dynamic stability and gait variability during attentional tasks in young adults
MARK
⁎
Rina Márcia Magnania, , Georgia Cristina Lehnena, Fábio Barbosa Rodriguesa, Gustavo Souto de Sá e Souzaa, Adriano de Oliveira Andradeb, Marcus Fraga Vieiraa a b
Universidade Federal de Goiás, Bioengineering and Biomechanics Laboratory, Goiânia, Goiás, Brazil Universidade Federal de Uberlândia, Biomedical Engineering Laboratory, Uberlândia, Minas Gerais, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords: Human gait Cell phone Dual-task Kinematics Local dynamic stability Entropy
Cell phone use while walking may be a cognitive distraction and reduce visual and motor attention. Thus, the aim of this study was to verify the effects of attentional dual-tasks while using a cell phone in different conditions. Stability, regularity, and linear variability of trunk kinematics, and gait spatiotemporal parameters in young adults were measured. Twenty young subjects of both genders were asked to walk on a treadmill for 4 min under the following conditions: (a) looking forward at a fixed target 2.5 m away (walking); (b) talking on a cell phone with unilateral handling (talking); (c) texting messages on a cell phone with unilateral handling (texting); and (d) looking forward at the aforementioned target while listening to music without handling the phone (listening). Local dynamic stability measured in terms of the largest Lyapunov exponent decreased while handling a cell phone (talking and texting). Gait variability and regularity increased when talking on a cell phone, but no variable changed in the listening condition. Under all dual-task conditions, there were significant increases in stride width and its variability. We conclude that young adults who use a cell phone when walking adapt their gait pattern conservatively, which can be because of increased attentional demand during cell phone use.
1. Introduction Attention and executive functions from cognitive areas are active during gait motor control. Performing gait in conjunction with another task, such as talking or typing on a cell phone, requires cognitive, neuromotor, physical, and memory skills; in addition, there is a competition for visual attention between the two tasks [1–4]. Thus, the dual-task paradigm has been used to evaluate the role of concurrent attentional demand in the motor control of human gait; in this setting, increased risk of falling, kinematic variability, and gait instability were observed [5–7]. Studies showed that dual-tasks using the cell phone, including texting, reading, and playing logical games, have an impact on the locomotion motor ability [1,2,8,9]. However, such dual-tasks are not executed spontaneously. To date, dual-tasks routinely practiced by young people using the cell phone have not been investigated, including unilateral handling texting and talking, and listening to music. According to TeleGeography, in 2013, 77% of the people worldwide used cell phone text messaging as a communication method. In 2015, there were 7.1 billion active devices, and there were 7.3 billion people [10]. When using a cell phone, individuals need to focus on a small
⁎
Corresponding author. E-mail address: [email protected] (R.M. Magnani).
http://dx.doi.org/10.1016/j.gaitpost.2017.04.019 Received 27 June 2016; Received in revised form 12 April 2017; Accepted 13 April 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
portable screen, which requires increased levels of manual dexterity, head and neck flexion, and concentration, all of which leads to reduced visual information input from the individual’s surroundings, increased working memory use, and executive control requirement [9,11]. Furthermore, several studies have shown the dangers of concurrent cell phone use while driving or walking, which may lead to an accident [1] or even death [2,8,11]. To maintain stability, executive and attention functions alter gait patterns during dual-task walking, as reported in young adults walking on a treadmill while using a cell phone [9]. In addition, when walking overground, the dual-task paradigm using cell phone increased the variability in spatiotemporal gait parameters, which has been related to decreased walking speed [2,8,10,12,13]. Schabrun et al. [10] also found increased trunk variability when individuals walked on a treadmill with a constant speed that was equal to normal the over-theground speed [10]. Similarly, Kao et al. [1] found a significantly greater trunk variability in treadmill walking during dual-task while using cell phone. Although studies evaluating regularity while walking and dualtask are scarce, one study was an examination of the effect of cell phone texting on the postural stability of young adults; the results from this
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
questions pertained to general and personal knowledge with the purpose of causing cognitive distractions through informal conversation.
study indicated that gait regularity did not change [14]. Gait variability is influenced by the ability to optimally control gait from one stride to the next [15]. Variability is quantified using linear or nonlinear measures; because the variations in human movement are distinguishable from noise, they have a deterministic origin, being neither random nor independent. Variability can be quantified using linear measures, such as magnitude variability (e.g., the average standard deviation along strides) [16], or nonlinear measures, such as entropy that quantifies the structure of the temporal variability or regularity of a time series [17]. Similarly, stability can be quantified using measures derived from nonlinear dynamics, such as the largest Lyapunov exponent (λs), which is a nonlinear measure of local dynamic stability (LDS) [5,18,19]. Variability refers to the motor system’s ability to perform in a wide variety of tasks and environmental constraints; stability refers to the dynamic ability to recover from an external perturbation, including dual-task. Thus, linear (trunk and spatiotemporal parameters variability) and nonlinear variables (entropy and stability) were computed in the present study. Since trunk kinematic data are the most sensitive to differences between different groups [20] and because maintaining stability of the upper body is critical for human locomotion [21], we computed gait features from trunk kinematics. Therefore, the objective of the present study was to verify the effects of attentional dual-tasks using a cell phone on gait LDS, regularity, linear variability, and spatiotemporal parameters in young adults. We hypothesized that LDS decreases under the proposed dual-task conditions while linear variability increases correspondingly.
2.3. Data analysis Before data analysis, except for calculating the short-term largest Lyapunov exponent (λs) and sample entropy (SEn), kinematic data were low-pass filtered using a fourth-order, zero-lag Butterworth filter with a cut-off frequency of 6 Hz. All parameters were calculated for 150 strides. The first and last 15 s of each trial were discarded [19]. A customized MATLAB code was used for data analysis. The LDS was assessed by λs calculated using Rosenstein’s algorithm [25]. Briefly, the mediolateral (ML), anteroposterior (AP), and vertical (V) T2 marker velocities were calculated from that marker data by using the three-points method [26]. Next, the velocity signal was timenormalized to 15,000 samples by preserving differences in stride time between strides [23]. A high-dimension attractor was constructed using the normalized ML, AP, and V T2 marker velocities and their delayed copies. A delay of 10 samples was used based on the mean value of the minimum of the mutual information function across all data, and a dimensionality of 5 was found to be sufficient based on the results of a global false nearest-neighbor analysis. For each point in the state-space, the nearest neighbor was found, with a minimal distance of a mean period corresponding to one step, and the Euclidean distance between these points was tracked for 10 strides. Then, a time vs log of the Euclidian distance curve was calculated for all neighboring points. Next, a divergence curve was calculated as the mean of those timedistance curves. The short-term λs was calculated as the slope of the linear fit of the first 50 samples (the time needed for one step) in the divergence curve. Thus, λs indicates the relative rate of divergence over one step, resulting from a small perturbation in the initial conditions. This method assumes that motor control ensures a dynamically stable gait if the divergence remains low between trajectories in a reconstructed state space that reflects gait dynamics [23]. Sample entropy (SEn) was calculated to quantify the degree of regularity of temporal variations of ML, AP, and V trunk velocities. SEn is the negative natural logarithm of the conditional probability of two m-dimensional delayed vectors that are close within a tolerance r, remain close in the (m + 1) dimensional state space without allowing self-matches [24,25]. The parameter values of m = 2 and r = 0.2 were selected. SEn reflects the likelihood that similar patterns of observations will not be followed by additional, similar observations. A time series containing numerous repetitive patterns – one that is predictable – has a relatively small SEn whereas a less predictable process has a higher SEn. To compute gait cycle variability (VAR), trunk acceleration during each stride was time-normalized (0–100%). At each of the 101 normalized time points, the SD of ML trunk acceleration over 150 strides was calculated. Next, the average SD of these 101 SDs was calculated [16]. Gait spatiotemporal parameters and their SDs in the trial were also used to assess changes in gait linear variability and pattern. Step width (SW) was determined as the ML distance between two consecutive heelstrikes, and the average step length (SL) was calculated from the average treadmill speed and the average step frequency ratio [26]. Step frequency (SF) was determined as the inverse of the average duration between two consecutive heel-strikes between limbs (i.e., left followed by right, or right followed by left), which were detected as the zerocross of the heel markers’ AP velocity [27].
2. Methods 2.1. Subjects Twenty healthy young subjects (10 males and 10 females, 24.5 ± 3.3 years old, 69.0 ± 13.7 kg, and 1.62 ± 36.7 m) were recruited into the study after obtaining written informed consent that was approved by the local research ethics committee. The inclusion criteria were an age of between 18 and 30 years and cell phone use. The exclusion criteria were any cognitive or musculoskeletal disease or impairment and disabling surgery in the last 12 months. 2.2. Equipment and procedures For gait assessment, kinematic analysis was performed using a 3D motion capture system comprising of 10 infrared cameras operating at 100 Hz (Vicon Nexus, Oxford Metrics, Oxford, UK). Nine reflective markers were attached to the lateral malleoli, heels, and heads of the second and the fifth metatarsals (bilaterally) and the spinous process of the second thoracic vertebrae (T2). Our stability analysis was focused on trunk motion because maintaining stability of the upper body is critical for human locomotion [22]. To isolate the effect of the attentional task, the same average speed found in the previous pilot study (4 km/h) was adopted for all subjects under the following conditions: (a) looking forward at a fixed target 2.5 m away (walking); (b) talking on a cell phone with unilateral handling (talking); (c) texting messages on a cell phone with unilateral handling (texting); and (d) looking forward at the aforementioned target while listening to music without handling the phone (listening). Each experimental condition was tested randomly in two 4-min trials with a 2-min rest periods between trials. The first trial was used for protocol familiarization, and only the results of the second trial were analyzed. The subjects used their own cell phones because they were familiar with the devices. An experimenter in another room asked questions and sent messages to the subjects during the corresponding trials, and no special instructions were provided about language, orthographic correction, correct answers, and a needed number of typed words. The
2.4. Statistical analysis Data that conformed to a normal distribution (Shapiro-Wilk test, p > 0.05) were analyzed using a repeated measure analysis of variance (ANOVA) with paired comparisons as post-hoc tests. Data that did not conform to a normal distribution were analyzed with the 106
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
Table 1 Non-linear and linear variability and percentage change of each variable.
ML λs AP λs V λs SEn VAR (m/s2) SL (cm) SF (Hz) SW (cm) SLSD (cm) SFSD (Hz) SWSD (cm)
Walking
Talking
%
Texting
%
Listening
%
p-value
1.53 ± 0.25a,b 1.70 ± 0.51f 1.40 ± 0.27 0.36 ± 0.07h 0.23 ± 0.11 60.60 ± 3.30 1.66 ± 0.13 6.65 ± 2.91 1.48 ± 0.30 0.04 ± 0.01 2.20 ± 0.40
1.71 ± 0.26a,c,d 1.76 ± 0.50 1.27 ± 0.28 0.31 ± 0.08h,i 0.24 ± 0.14k 60.4 ± 3.30 1.70 ± 0.09 7.92 ± 2.89 1.47 ± 0.29 0.04 ± 0.01 2.28 ± 0.37
12 4 −9 −14 4 0 2 19 −1 0 4
1.81 ± 0.21b,c,e 1.92 ± 0.32f,g 1.29 ± 0.34 0.34 ± 0.05 j 0.18 ± 0.04 k 60.4 ± 3.20 1.69 ± 0.07 7.61 ± 2.89 1.44 ± 0.25 0.04 ± 0.01 2.33 ± 0.28l
18 13 -8 -6 −22 0 2 14 −3 0 6
1.53 ± 0.20d,e 1.71 ± 0.4g 1.26 ± 0.30 0.37 ± 0.07i,j 0.21 ± 0.15 60.9 ± 2.90 1.70 ± 0.08 6.90 ± 2.97 1.47 ± 0.29 0.04 ± 0.01 2.12 ± 0.25l
0 1 −10 3 −9 0 2 4 −1 0 −4
< 0.001* 0.001* 0.177** < 0.001* 0.016* 0.329** 0.400** 0.050** 0.821** 0.392* 0.004*
a–l: pairwise comparisons at p < 0.05; ML: mediolateral; AP: anteroposterior; V: vertical; λs: short-term largest Lyapunov exponent; SEn: sample entropy; VAR: gait cycle standard deviation variability; SLSD: step length standard deviation; SFSD: step frequency standard deviation; SWSD: stride width standard deviation. %: change in relation to walking condition – a positive value means an increase, while a negative one means a decrease. * Friedman Test. ** Repeated Measures ANOVA.
The attentional demand under the texting and talking conditions in our study could have been challenging enough to cause significant decreases in ML and AP stability, which are attention- and physicaldemanding tasks. Although Dingwell et al. [5] found slight changes in LDS during a Stroop test, these changes were not remarkable because the dual-task used may not have been challenging enough. A Stroop test imposes only a cognitive demand without an associate motor demand. Some studies indicated that pure cognitive tasks (mathematical calculation game) did not affect margin of stability as much as associating the same cognitive task with cell phone bimanual handling [1,9]. Regarding visual demands, it has been well-established that the human visual system provides information about the environment and plays an important role during locomotion in terms of maintaining stability [28], inducing a more cautious gait pattern [9]. During cell phone use, head motion is reduced to keep the gaze fixed on the phone, which may alter vestibular and visual sensory signals for controlling balance, a situation associated with a high risk of falling [1,9,10]. This can explain the decreased LDS in the talking and texting conditions compared to the walking condition. Concerning motor demands, cell phone use intrinsically alters the upper limb movements during gait. Arm swing assists in balance recovery after postural disturbance, helps stabilize the body, enhances lateral balance, and reduces the metabolic cost of walking [1,9,10]. Cell phone manipulation leads to reduced arm swinging and altered head orientation, modifying walking performance in addition to being a cognitive distraction. This situation might reduce walking stability or induce the adoption of energetically costly stabilization strategies, such as increased trunk muscle activation or step width adjustment [1]. In the present study, VAR increased only under the talking condition, while SW and SWSD increased significantly under the texting and the talking conditions. A possible explanation for these results is that such conditions involve cognitive and physical resources, gross and fine motor function integration, and near and far vision [29], forcing the subjects to adopt a conservative gait with increased SW. The increased gait regularity observed in the talking condition compared to the walking condition can be interpreted as an ineffective balance control, similar to that observed in the dual-task postural control study [30]. While properly answering the questions being asked by the experimenter, the subjects appeared to be more focused on gait than in the walking condition, and in some situations, it has been reported that increased attentional focus may be detrimental for balance control [30]. The texting condition presented the lowest LDS (greatest λs) and SEn, probably because of the reduction of the visual field because in this condition, the gaze is directed at the cell phone and cell phone manipulation decreases arm swing. By contrast, under the listening
Friedman's test, with a Wilcoxon-signed rank test as the post-hoc. Bonferroni corrections were applied to all p-values from the post-hoc analyses. A statistical analysis was performed with SPSS software (version 23, SPSS Inc., Chicago, IL), and the significance level was set to p < 0.05. 3. Results LDS, sample entropy, gait linear variability, and spatiotemporal parameters and their standard deviations are summarized in Table 1. The ML and AP λs showed significant differences, with increased values under the talking and the texting conditions compared to the walking trial. The talking condition exhibited decreased SEn and increased VAR, denoting increased regularity and increased variability, respectively. 4. Discussion With young adult participants, the present study verified the effects of attentional dual-tasks on gait LDS, regularity, and linear variability by using a cell phone in different conditions. Our initial hypothesis could be partially proved because we observed a decrease in LDS in both the ML and the AP directions and an increase in gait linear variability, denoted by increased VAR and SWSD in both the talking and the texting conditions compared to the walking condition. The same results, however, were not found for the listening condition. Unexpectedly, gait regularity increased in the talking condition. Aiming to cover the knowledge gap, our concern was reproducing routine situations with the use of a cell phone while walking. The findings showed that the situations related to the evaluated conditions are of great concern regarding trunk stability and gait variability. These findings take into account the decision to impose the same speed for all conditions, which was motivated by the inability to know if the observed gait pattern changes would have been caused by the attentional challenge or by a change in walking speed because the subjects tended to reduce their speed in such conditions in the pilot study. We interpreted this decrease in the walking speed as an adaptation to maintain gait stability, and this would not be easy to isolate from the dual-task’s effect. Several cognitive and motor aspects could explain the findings of the present study. Schabrun et al. [10] observed changes in gait pattern because of cell phone use. They attributed the changes to three major factors: cognitive distraction, reduction in visual attention toward the environment, and changes in physical demands, such as reduced arm swinging and altered head orientation. These factors may also influence trunk stability and gait variability. 107
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
parameters during ramp ascent and descent, Phys. Ther. Rehabil. Sci. (2014) 43–48. [9] J.R. Marone, P.B. Patel, C.P. Hurt, M.D. Grabiner, Frontal plane margin of stability is increased during texting while walking, Gait Posture 40 (2014) 243–246, http:// dx.doi.org/10.1016/j.gaitpost.2014.04.188. [10] S.M. Schabrun, W. Den Hoorn, A. Moorcroft, C. Greenland, P.W. Hodges, Texting and walking: strategies for postural control and implications for safety, PLoS One 9 (2014), http://dx.doi.org/10.1371/journal.pone.0084312. [11] J.M. Aldridge, The Effect of Concurrent Cell Phone Use and Walking on Gait Characteristics, University of Nevada, Las Vegas, 2009. [12] Z. Hazar, L. Eker, İ. Yüksel, Demographic and anthropometric factors predicting the grip strength and endurance in adolescent tennis players, Turkiye Klin. J. Sport Sci. 7 (2015) 24–28, http://dx.doi.org/10.5336/sportsci.2014-41598. [13] S. Demura, M. Uchiyama, Influence of cell phone email usage on gait characteristics, Eur. J. Sport Sci. 9 (2009) 303–309. [14] N.R. Nurwulan, B.C. Jiang, H. Iridiastadi, Posture and texting: effect on balance in young adults, PLoS One (2015) 1–10, http://dx.doi.org/10.1371/journal.pone. 0134230. [15] J.M. Hausdorff, D.A. Rios, H.K. Edelberg, Gait variability and fall risk in community-living older adults: a 1-year prospective study, Arch. Phys. Med. Rehabil. 82 (2001) 1050–1056, http://dx.doi.org/10.1053/apmr.2001.24893. [16] J.B. Dingwell, J.P. Cusumano, P.R. Cavanagh, D. Sternad, Local dynamic stability versus kinematic variability of continuous overground and treadmill walking, J. Biomech. Eng. 123 (2001) 27–32, http://dx.doi.org/10.1115/1.1336798. [17] M. Costa, C.-K. Peng, A.L. Goldberger, J.M. Hausdorr, Multiscale entropy analysis of human gait dynamics, Physica A 330 (2003) 53–60, http://dx.doi.org/10.1016/j. physa.2003.08.022. [18] K.S. van Schooten, L.H. Sloot, S.M. Bruijn, H. Kingma, O.G. Meijer, M. Pijnappels, J.H. Van Dieen, Sensitivity of trunk variability and stability measures to balance impairments induced by galvanic vestibular stimulation during gait, Gait Posture 33 (2011) 656–660, http://dx.doi.org/10.1016/j.gaitpost.2011.02.017. [19] L. Hak, H. Houdijk, P.J. Beek, J.H. Van Dieën, Steps to take to enhance gait stability: the effect of stride frequency, stride length, and walking speed on local dynamic stability and margins of stability, PLoS One 8 (2013) 1–7, http://dx.doi.org/10. 1371/journal.pone.0082842. [20] H.G. Kang, J.B. Dingwell, Dynamic stability of superior vs. inferior segments during walking in young and older adults, Gait Posture 30 (2009) 260–263, http://dx.doi. org/10.1016/j.gaitpost.2009.05.003. [21] F. Prince, D. Winter, P. Stergiou, S. Walt, Anticipatory control of upper body balance during human locomotion, Gait Posture 2 (1994) 19–25, http://dx.doi.org/ 10.1016/0966-6362(94)90013-2. [22] P.M.M. Young, J.B. Dingwell, Voluntarily changing step length or step width affects dynamic stability of human walking, Gait Posture 35 (2012) 472–477, http://dx. doi.org/10.1016/j.pestbp.2011.02.012.Investigations. [23] S.M. Bruijn, O.G. Meijer, P.J. Beek, J.H. van Dieën, Assessing the stability of human locomotion: a review of current measures, J. R. Soc. Interfaces 10 (2013) 20120999, http://dx.doi.org/10.1098/rsif.2012.0999. [24] S. Ramdani, B. Seigle, J. Lagarde, F. Bouchara, P.L. Bernard, On the use of sample entropy to analyze human postural sway data, Med. Eng. Phys. 31 (2009) 1023–1031, http://dx.doi.org/10.1016/j.medengphy.2009.06.004. [25] J.S. Richman, J.R. Moorman, Physiological time-series analysis using approximate entropy and sample entropy, Am. J. Physiol. Heart Circ. Physiol. (2000) 2039–2049. [26] P. Terrier, F. Reynard, Effect of age on the variability and stability of gait: a crosssectional treadmill study in healthy individuals between 20 and 69 years of age, Gait Posture 41 (2015) 170–174, http://dx.doi.org/10.1016/j.gaitpost.2014.09. 024. [27] A.D. Patel, J.R. Iversen, Y. Chen, B.H. Repp, The influence of metricality and modality on synchronization with a beat, Exp. Brain Res. 163 (2005) 226–238, http://dx.doi.org/10.1007/s00221-004-2159-8. [28] A. Hallemans, E. Ortibus, F. Meire, P. Aerts, Low vision affects dynamic stability of gait, Gait Posture 32 (2010) 547–551, http://dx.doi.org/10.1016/j.gaitpost.2010. 07.018. [29] V. Agostini, F. Lo Fermo, G. Massazza, M. Knaflitz, Does texting while walking really affect gait in young adults? J. Neuroeng. Rehabil. 12 (2015) 86, http://dx. doi.org/10.1186/s12984-015-0079-4. [30] S.F. Donker, M. Roerdink, A.J. Greven, P.J. Beek, Regularity of center-of-pressure trajectories depends on the amount of attention invested in postural control, Exp. Brain Res. 181 (2007) 1–11, http://dx.doi.org/10.1007/s00221-007-0905-4.
condition, there were no significant differences in LDS SEn, SWSD, and VAR when compared to the walking condition, possibly because this condition did not impose any restriction to the visual field, arm swing, or attentional focus. This may be related to music having a remarkable ability to drive rhythmic, metrically organized motor behavior [27]. Future studies should address the limitations of this study and other tasks, such as texting with both hands, increased cognitive demand, and should specifically look at the influence of reduction in visual field and other changes in physical demands in gait parameters. 5. Conclusion Based on the collected data, we observed that attentional dual-tasks, such as cell phone use while walking, alter gait features. This was assessed in a young adult population using LDS, gait regularity, and linear variability. The results indicated that cell phone use may compromise walking stability, which in turn may increase the risk of falling. Attentional dual-tasks that reduce the visual field and upper limb movement, such as the texting and talking conditions assessed in the present study, increase gait linear variability; although, the gait regularity increase while talking on a cell phone. Future studies could focus on strategies to develop safety guidelines for pedestrians using cell phones. Conflict of interest There are no conflict of interest to declare. Acknowledgements The authors are grateful to government agencies CAPES, CNPq, FAPEG and FAPEMIG. References [1] P.C. Kao, C.I. Higginson, K. Seymour, M. Kamerdze, J.S. Higginson, Walking stability during cell phone use in healthy adults, Gait Posture 41 (2015) 947–953, http://dx.doi.org/10.1016/j.gaitpost.2015.03.347. [2] E.M. Lamberg, L.M. Muratori, Cell phones change the way we walk, Gait Posture 35 (2012) 688–690, http://dx.doi.org/10.1016/j.gaitpost.2011.12.005. [3] S. Schaefer, D. Jagenow, J. Verrel, U. Lindenberger, The influence of cognitive load and walking speed on gait regularity in children and young adults, Gait Posture 41 (2015) 258–262, http://dx.doi.org/10.1016/j.gaitpost.2014.10.013. [4] T. Szturm, P. Maharjan, J.J. Marotta, B. Shay, S. Shrestha, V. Sakhalkar, The interacting effect of cognitive and motor task demands on performance of gait, balance and cognition in young adults, Gait Posture 38 (2013) 596–602, http://dx. doi.org/10.1016/j.gaitpost.2013.02.004. [5] J.B. Dingwell, R.T. Robb, K.L. Troy, M.D. Grabiner, Effects of an attention demanding task on dynamic stability during treadmill walking, J. Neuroeng. Rehabil. 5 (2008) 12, http://dx.doi.org/10.1186/1743-0003-5-12. [6] N. Polskaia, The Impact of Auditory and Visual Cognitive Tasks on Postural Control in Young Adults 128 Univ. Ottawa, 2015, http://dx.doi.org/10.1017/ CBO9781107415324.004 Thesis. [7] X. Qu, Age-related cognitive task effects on gait characteristics: do different working memory components make a difference? J. Neuroeng. Rehabil. 11 (2014) 149, http://dx.doi.org/10.1186/1743-0003-11-149. [8] H. Kim, J. Park, J. Cha, C. Song, Influence of mobile phone texting on gait
108
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Local dynamic stability and gait variability during attentional tasks in young adults
MARK
⁎
Rina Márcia Magnania, , Georgia Cristina Lehnena, Fábio Barbosa Rodriguesa, Gustavo Souto de Sá e Souzaa, Adriano de Oliveira Andradeb, Marcus Fraga Vieiraa a b
Universidade Federal de Goiás, Bioengineering and Biomechanics Laboratory, Goiânia, Goiás, Brazil Universidade Federal de Uberlândia, Biomedical Engineering Laboratory, Uberlândia, Minas Gerais, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords: Human gait Cell phone Dual-task Kinematics Local dynamic stability Entropy
Cell phone use while walking may be a cognitive distraction and reduce visual and motor attention. Thus, the aim of this study was to verify the effects of attentional dual-tasks while using a cell phone in different conditions. Stability, regularity, and linear variability of trunk kinematics, and gait spatiotemporal parameters in young adults were measured. Twenty young subjects of both genders were asked to walk on a treadmill for 4 min under the following conditions: (a) looking forward at a fixed target 2.5 m away (walking); (b) talking on a cell phone with unilateral handling (talking); (c) texting messages on a cell phone with unilateral handling (texting); and (d) looking forward at the aforementioned target while listening to music without handling the phone (listening). Local dynamic stability measured in terms of the largest Lyapunov exponent decreased while handling a cell phone (talking and texting). Gait variability and regularity increased when talking on a cell phone, but no variable changed in the listening condition. Under all dual-task conditions, there were significant increases in stride width and its variability. We conclude that young adults who use a cell phone when walking adapt their gait pattern conservatively, which can be because of increased attentional demand during cell phone use.
1. Introduction Attention and executive functions from cognitive areas are active during gait motor control. Performing gait in conjunction with another task, such as talking or typing on a cell phone, requires cognitive, neuromotor, physical, and memory skills; in addition, there is a competition for visual attention between the two tasks [1–4]. Thus, the dual-task paradigm has been used to evaluate the role of concurrent attentional demand in the motor control of human gait; in this setting, increased risk of falling, kinematic variability, and gait instability were observed [5–7]. Studies showed that dual-tasks using the cell phone, including texting, reading, and playing logical games, have an impact on the locomotion motor ability [1,2,8,9]. However, such dual-tasks are not executed spontaneously. To date, dual-tasks routinely practiced by young people using the cell phone have not been investigated, including unilateral handling texting and talking, and listening to music. According to TeleGeography, in 2013, 77% of the people worldwide used cell phone text messaging as a communication method. In 2015, there were 7.1 billion active devices, and there were 7.3 billion people [10]. When using a cell phone, individuals need to focus on a small
⁎
Corresponding author. E-mail address: [email protected] (R.M. Magnani).
http://dx.doi.org/10.1016/j.gaitpost.2017.04.019 Received 27 June 2016; Received in revised form 12 April 2017; Accepted 13 April 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
portable screen, which requires increased levels of manual dexterity, head and neck flexion, and concentration, all of which leads to reduced visual information input from the individual’s surroundings, increased working memory use, and executive control requirement [9,11]. Furthermore, several studies have shown the dangers of concurrent cell phone use while driving or walking, which may lead to an accident [1] or even death [2,8,11]. To maintain stability, executive and attention functions alter gait patterns during dual-task walking, as reported in young adults walking on a treadmill while using a cell phone [9]. In addition, when walking overground, the dual-task paradigm using cell phone increased the variability in spatiotemporal gait parameters, which has been related to decreased walking speed [2,8,10,12,13]. Schabrun et al. [10] also found increased trunk variability when individuals walked on a treadmill with a constant speed that was equal to normal the over-theground speed [10]. Similarly, Kao et al. [1] found a significantly greater trunk variability in treadmill walking during dual-task while using cell phone. Although studies evaluating regularity while walking and dualtask are scarce, one study was an examination of the effect of cell phone texting on the postural stability of young adults; the results from this
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
questions pertained to general and personal knowledge with the purpose of causing cognitive distractions through informal conversation.
study indicated that gait regularity did not change [14]. Gait variability is influenced by the ability to optimally control gait from one stride to the next [15]. Variability is quantified using linear or nonlinear measures; because the variations in human movement are distinguishable from noise, they have a deterministic origin, being neither random nor independent. Variability can be quantified using linear measures, such as magnitude variability (e.g., the average standard deviation along strides) [16], or nonlinear measures, such as entropy that quantifies the structure of the temporal variability or regularity of a time series [17]. Similarly, stability can be quantified using measures derived from nonlinear dynamics, such as the largest Lyapunov exponent (λs), which is a nonlinear measure of local dynamic stability (LDS) [5,18,19]. Variability refers to the motor system’s ability to perform in a wide variety of tasks and environmental constraints; stability refers to the dynamic ability to recover from an external perturbation, including dual-task. Thus, linear (trunk and spatiotemporal parameters variability) and nonlinear variables (entropy and stability) were computed in the present study. Since trunk kinematic data are the most sensitive to differences between different groups [20] and because maintaining stability of the upper body is critical for human locomotion [21], we computed gait features from trunk kinematics. Therefore, the objective of the present study was to verify the effects of attentional dual-tasks using a cell phone on gait LDS, regularity, linear variability, and spatiotemporal parameters in young adults. We hypothesized that LDS decreases under the proposed dual-task conditions while linear variability increases correspondingly.
2.3. Data analysis Before data analysis, except for calculating the short-term largest Lyapunov exponent (λs) and sample entropy (SEn), kinematic data were low-pass filtered using a fourth-order, zero-lag Butterworth filter with a cut-off frequency of 6 Hz. All parameters were calculated for 150 strides. The first and last 15 s of each trial were discarded [19]. A customized MATLAB code was used for data analysis. The LDS was assessed by λs calculated using Rosenstein’s algorithm [25]. Briefly, the mediolateral (ML), anteroposterior (AP), and vertical (V) T2 marker velocities were calculated from that marker data by using the three-points method [26]. Next, the velocity signal was timenormalized to 15,000 samples by preserving differences in stride time between strides [23]. A high-dimension attractor was constructed using the normalized ML, AP, and V T2 marker velocities and their delayed copies. A delay of 10 samples was used based on the mean value of the minimum of the mutual information function across all data, and a dimensionality of 5 was found to be sufficient based on the results of a global false nearest-neighbor analysis. For each point in the state-space, the nearest neighbor was found, with a minimal distance of a mean period corresponding to one step, and the Euclidean distance between these points was tracked for 10 strides. Then, a time vs log of the Euclidian distance curve was calculated for all neighboring points. Next, a divergence curve was calculated as the mean of those timedistance curves. The short-term λs was calculated as the slope of the linear fit of the first 50 samples (the time needed for one step) in the divergence curve. Thus, λs indicates the relative rate of divergence over one step, resulting from a small perturbation in the initial conditions. This method assumes that motor control ensures a dynamically stable gait if the divergence remains low between trajectories in a reconstructed state space that reflects gait dynamics [23]. Sample entropy (SEn) was calculated to quantify the degree of regularity of temporal variations of ML, AP, and V trunk velocities. SEn is the negative natural logarithm of the conditional probability of two m-dimensional delayed vectors that are close within a tolerance r, remain close in the (m + 1) dimensional state space without allowing self-matches [24,25]. The parameter values of m = 2 and r = 0.2 were selected. SEn reflects the likelihood that similar patterns of observations will not be followed by additional, similar observations. A time series containing numerous repetitive patterns – one that is predictable – has a relatively small SEn whereas a less predictable process has a higher SEn. To compute gait cycle variability (VAR), trunk acceleration during each stride was time-normalized (0–100%). At each of the 101 normalized time points, the SD of ML trunk acceleration over 150 strides was calculated. Next, the average SD of these 101 SDs was calculated [16]. Gait spatiotemporal parameters and their SDs in the trial were also used to assess changes in gait linear variability and pattern. Step width (SW) was determined as the ML distance between two consecutive heelstrikes, and the average step length (SL) was calculated from the average treadmill speed and the average step frequency ratio [26]. Step frequency (SF) was determined as the inverse of the average duration between two consecutive heel-strikes between limbs (i.e., left followed by right, or right followed by left), which were detected as the zerocross of the heel markers’ AP velocity [27].
2. Methods 2.1. Subjects Twenty healthy young subjects (10 males and 10 females, 24.5 ± 3.3 years old, 69.0 ± 13.7 kg, and 1.62 ± 36.7 m) were recruited into the study after obtaining written informed consent that was approved by the local research ethics committee. The inclusion criteria were an age of between 18 and 30 years and cell phone use. The exclusion criteria were any cognitive or musculoskeletal disease or impairment and disabling surgery in the last 12 months. 2.2. Equipment and procedures For gait assessment, kinematic analysis was performed using a 3D motion capture system comprising of 10 infrared cameras operating at 100 Hz (Vicon Nexus, Oxford Metrics, Oxford, UK). Nine reflective markers were attached to the lateral malleoli, heels, and heads of the second and the fifth metatarsals (bilaterally) and the spinous process of the second thoracic vertebrae (T2). Our stability analysis was focused on trunk motion because maintaining stability of the upper body is critical for human locomotion [22]. To isolate the effect of the attentional task, the same average speed found in the previous pilot study (4 km/h) was adopted for all subjects under the following conditions: (a) looking forward at a fixed target 2.5 m away (walking); (b) talking on a cell phone with unilateral handling (talking); (c) texting messages on a cell phone with unilateral handling (texting); and (d) looking forward at the aforementioned target while listening to music without handling the phone (listening). Each experimental condition was tested randomly in two 4-min trials with a 2-min rest periods between trials. The first trial was used for protocol familiarization, and only the results of the second trial were analyzed. The subjects used their own cell phones because they were familiar with the devices. An experimenter in another room asked questions and sent messages to the subjects during the corresponding trials, and no special instructions were provided about language, orthographic correction, correct answers, and a needed number of typed words. The
2.4. Statistical analysis Data that conformed to a normal distribution (Shapiro-Wilk test, p > 0.05) were analyzed using a repeated measure analysis of variance (ANOVA) with paired comparisons as post-hoc tests. Data that did not conform to a normal distribution were analyzed with the 106
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
Table 1 Non-linear and linear variability and percentage change of each variable.
ML λs AP λs V λs SEn VAR (m/s2) SL (cm) SF (Hz) SW (cm) SLSD (cm) SFSD (Hz) SWSD (cm)
Walking
Talking
%
Texting
%
Listening
%
p-value
1.53 ± 0.25a,b 1.70 ± 0.51f 1.40 ± 0.27 0.36 ± 0.07h 0.23 ± 0.11 60.60 ± 3.30 1.66 ± 0.13 6.65 ± 2.91 1.48 ± 0.30 0.04 ± 0.01 2.20 ± 0.40
1.71 ± 0.26a,c,d 1.76 ± 0.50 1.27 ± 0.28 0.31 ± 0.08h,i 0.24 ± 0.14k 60.4 ± 3.30 1.70 ± 0.09 7.92 ± 2.89 1.47 ± 0.29 0.04 ± 0.01 2.28 ± 0.37
12 4 −9 −14 4 0 2 19 −1 0 4
1.81 ± 0.21b,c,e 1.92 ± 0.32f,g 1.29 ± 0.34 0.34 ± 0.05 j 0.18 ± 0.04 k 60.4 ± 3.20 1.69 ± 0.07 7.61 ± 2.89 1.44 ± 0.25 0.04 ± 0.01 2.33 ± 0.28l
18 13 -8 -6 −22 0 2 14 −3 0 6
1.53 ± 0.20d,e 1.71 ± 0.4g 1.26 ± 0.30 0.37 ± 0.07i,j 0.21 ± 0.15 60.9 ± 2.90 1.70 ± 0.08 6.90 ± 2.97 1.47 ± 0.29 0.04 ± 0.01 2.12 ± 0.25l
0 1 −10 3 −9 0 2 4 −1 0 −4
< 0.001* 0.001* 0.177** < 0.001* 0.016* 0.329** 0.400** 0.050** 0.821** 0.392* 0.004*
a–l: pairwise comparisons at p < 0.05; ML: mediolateral; AP: anteroposterior; V: vertical; λs: short-term largest Lyapunov exponent; SEn: sample entropy; VAR: gait cycle standard deviation variability; SLSD: step length standard deviation; SFSD: step frequency standard deviation; SWSD: stride width standard deviation. %: change in relation to walking condition – a positive value means an increase, while a negative one means a decrease. * Friedman Test. ** Repeated Measures ANOVA.
The attentional demand under the texting and talking conditions in our study could have been challenging enough to cause significant decreases in ML and AP stability, which are attention- and physicaldemanding tasks. Although Dingwell et al. [5] found slight changes in LDS during a Stroop test, these changes were not remarkable because the dual-task used may not have been challenging enough. A Stroop test imposes only a cognitive demand without an associate motor demand. Some studies indicated that pure cognitive tasks (mathematical calculation game) did not affect margin of stability as much as associating the same cognitive task with cell phone bimanual handling [1,9]. Regarding visual demands, it has been well-established that the human visual system provides information about the environment and plays an important role during locomotion in terms of maintaining stability [28], inducing a more cautious gait pattern [9]. During cell phone use, head motion is reduced to keep the gaze fixed on the phone, which may alter vestibular and visual sensory signals for controlling balance, a situation associated with a high risk of falling [1,9,10]. This can explain the decreased LDS in the talking and texting conditions compared to the walking condition. Concerning motor demands, cell phone use intrinsically alters the upper limb movements during gait. Arm swing assists in balance recovery after postural disturbance, helps stabilize the body, enhances lateral balance, and reduces the metabolic cost of walking [1,9,10]. Cell phone manipulation leads to reduced arm swinging and altered head orientation, modifying walking performance in addition to being a cognitive distraction. This situation might reduce walking stability or induce the adoption of energetically costly stabilization strategies, such as increased trunk muscle activation or step width adjustment [1]. In the present study, VAR increased only under the talking condition, while SW and SWSD increased significantly under the texting and the talking conditions. A possible explanation for these results is that such conditions involve cognitive and physical resources, gross and fine motor function integration, and near and far vision [29], forcing the subjects to adopt a conservative gait with increased SW. The increased gait regularity observed in the talking condition compared to the walking condition can be interpreted as an ineffective balance control, similar to that observed in the dual-task postural control study [30]. While properly answering the questions being asked by the experimenter, the subjects appeared to be more focused on gait than in the walking condition, and in some situations, it has been reported that increased attentional focus may be detrimental for balance control [30]. The texting condition presented the lowest LDS (greatest λs) and SEn, probably because of the reduction of the visual field because in this condition, the gaze is directed at the cell phone and cell phone manipulation decreases arm swing. By contrast, under the listening
Friedman's test, with a Wilcoxon-signed rank test as the post-hoc. Bonferroni corrections were applied to all p-values from the post-hoc analyses. A statistical analysis was performed with SPSS software (version 23, SPSS Inc., Chicago, IL), and the significance level was set to p < 0.05. 3. Results LDS, sample entropy, gait linear variability, and spatiotemporal parameters and their standard deviations are summarized in Table 1. The ML and AP λs showed significant differences, with increased values under the talking and the texting conditions compared to the walking trial. The talking condition exhibited decreased SEn and increased VAR, denoting increased regularity and increased variability, respectively. 4. Discussion With young adult participants, the present study verified the effects of attentional dual-tasks on gait LDS, regularity, and linear variability by using a cell phone in different conditions. Our initial hypothesis could be partially proved because we observed a decrease in LDS in both the ML and the AP directions and an increase in gait linear variability, denoted by increased VAR and SWSD in both the talking and the texting conditions compared to the walking condition. The same results, however, were not found for the listening condition. Unexpectedly, gait regularity increased in the talking condition. Aiming to cover the knowledge gap, our concern was reproducing routine situations with the use of a cell phone while walking. The findings showed that the situations related to the evaluated conditions are of great concern regarding trunk stability and gait variability. These findings take into account the decision to impose the same speed for all conditions, which was motivated by the inability to know if the observed gait pattern changes would have been caused by the attentional challenge or by a change in walking speed because the subjects tended to reduce their speed in such conditions in the pilot study. We interpreted this decrease in the walking speed as an adaptation to maintain gait stability, and this would not be easy to isolate from the dual-task’s effect. Several cognitive and motor aspects could explain the findings of the present study. Schabrun et al. [10] observed changes in gait pattern because of cell phone use. They attributed the changes to three major factors: cognitive distraction, reduction in visual attention toward the environment, and changes in physical demands, such as reduced arm swinging and altered head orientation. These factors may also influence trunk stability and gait variability. 107
Gait & Posture 55 (2017) 105–108
R.M. Magnani et al.
parameters during ramp ascent and descent, Phys. Ther. Rehabil. Sci. (2014) 43–48. [9] J.R. Marone, P.B. Patel, C.P. Hurt, M.D. Grabiner, Frontal plane margin of stability is increased during texting while walking, Gait Posture 40 (2014) 243–246, http:// dx.doi.org/10.1016/j.gaitpost.2014.04.188. [10] S.M. Schabrun, W. Den Hoorn, A. Moorcroft, C. Greenland, P.W. Hodges, Texting and walking: strategies for postural control and implications for safety, PLoS One 9 (2014), http://dx.doi.org/10.1371/journal.pone.0084312. [11] J.M. Aldridge, The Effect of Concurrent Cell Phone Use and Walking on Gait Characteristics, University of Nevada, Las Vegas, 2009. [12] Z. Hazar, L. Eker, İ. Yüksel, Demographic and anthropometric factors predicting the grip strength and endurance in adolescent tennis players, Turkiye Klin. J. Sport Sci. 7 (2015) 24–28, http://dx.doi.org/10.5336/sportsci.2014-41598. [13] S. Demura, M. Uchiyama, Influence of cell phone email usage on gait characteristics, Eur. J. Sport Sci. 9 (2009) 303–309. [14] N.R. Nurwulan, B.C. Jiang, H. Iridiastadi, Posture and texting: effect on balance in young adults, PLoS One (2015) 1–10, http://dx.doi.org/10.1371/journal.pone. 0134230. [15] J.M. Hausdorff, D.A. Rios, H.K. Edelberg, Gait variability and fall risk in community-living older adults: a 1-year prospective study, Arch. Phys. Med. Rehabil. 82 (2001) 1050–1056, http://dx.doi.org/10.1053/apmr.2001.24893. [16] J.B. Dingwell, J.P. Cusumano, P.R. Cavanagh, D. Sternad, Local dynamic stability versus kinematic variability of continuous overground and treadmill walking, J. Biomech. Eng. 123 (2001) 27–32, http://dx.doi.org/10.1115/1.1336798. [17] M. Costa, C.-K. Peng, A.L. Goldberger, J.M. Hausdorr, Multiscale entropy analysis of human gait dynamics, Physica A 330 (2003) 53–60, http://dx.doi.org/10.1016/j. physa.2003.08.022. [18] K.S. van Schooten, L.H. Sloot, S.M. Bruijn, H. Kingma, O.G. Meijer, M. Pijnappels, J.H. Van Dieen, Sensitivity of trunk variability and stability measures to balance impairments induced by galvanic vestibular stimulation during gait, Gait Posture 33 (2011) 656–660, http://dx.doi.org/10.1016/j.gaitpost.2011.02.017. [19] L. Hak, H. Houdijk, P.J. Beek, J.H. Van Dieën, Steps to take to enhance gait stability: the effect of stride frequency, stride length, and walking speed on local dynamic stability and margins of stability, PLoS One 8 (2013) 1–7, http://dx.doi.org/10. 1371/journal.pone.0082842. [20] H.G. Kang, J.B. Dingwell, Dynamic stability of superior vs. inferior segments during walking in young and older adults, Gait Posture 30 (2009) 260–263, http://dx.doi. org/10.1016/j.gaitpost.2009.05.003. [21] F. Prince, D. Winter, P. Stergiou, S. Walt, Anticipatory control of upper body balance during human locomotion, Gait Posture 2 (1994) 19–25, http://dx.doi.org/ 10.1016/0966-6362(94)90013-2. [22] P.M.M. Young, J.B. Dingwell, Voluntarily changing step length or step width affects dynamic stability of human walking, Gait Posture 35 (2012) 472–477, http://dx. doi.org/10.1016/j.pestbp.2011.02.012.Investigations. [23] S.M. Bruijn, O.G. Meijer, P.J. Beek, J.H. van Dieën, Assessing the stability of human locomotion: a review of current measures, J. R. Soc. Interfaces 10 (2013) 20120999, http://dx.doi.org/10.1098/rsif.2012.0999. [24] S. Ramdani, B. Seigle, J. Lagarde, F. Bouchara, P.L. Bernard, On the use of sample entropy to analyze human postural sway data, Med. Eng. Phys. 31 (2009) 1023–1031, http://dx.doi.org/10.1016/j.medengphy.2009.06.004. [25] J.S. Richman, J.R. Moorman, Physiological time-series analysis using approximate entropy and sample entropy, Am. J. Physiol. Heart Circ. Physiol. (2000) 2039–2049. [26] P. Terrier, F. Reynard, Effect of age on the variability and stability of gait: a crosssectional treadmill study in healthy individuals between 20 and 69 years of age, Gait Posture 41 (2015) 170–174, http://dx.doi.org/10.1016/j.gaitpost.2014.09. 024. [27] A.D. Patel, J.R. Iversen, Y. Chen, B.H. Repp, The influence of metricality and modality on synchronization with a beat, Exp. Brain Res. 163 (2005) 226–238, http://dx.doi.org/10.1007/s00221-004-2159-8. [28] A. Hallemans, E. Ortibus, F. Meire, P. Aerts, Low vision affects dynamic stability of gait, Gait Posture 32 (2010) 547–551, http://dx.doi.org/10.1016/j.gaitpost.2010. 07.018. [29] V. Agostini, F. Lo Fermo, G. Massazza, M. Knaflitz, Does texting while walking really affect gait in young adults? J. Neuroeng. Rehabil. 12 (2015) 86, http://dx. doi.org/10.1186/s12984-015-0079-4. [30] S.F. Donker, M. Roerdink, A.J. Greven, P.J. Beek, Regularity of center-of-pressure trajectories depends on the amount of attention invested in postural control, Exp. Brain Res. 181 (2007) 1–11, http://dx.doi.org/10.1007/s00221-007-0905-4.
condition, there were no significant differences in LDS SEn, SWSD, and VAR when compared to the walking condition, possibly because this condition did not impose any restriction to the visual field, arm swing, or attentional focus. This may be related to music having a remarkable ability to drive rhythmic, metrically organized motor behavior [27]. Future studies should address the limitations of this study and other tasks, such as texting with both hands, increased cognitive demand, and should specifically look at the influence of reduction in visual field and other changes in physical demands in gait parameters. 5. Conclusion Based on the collected data, we observed that attentional dual-tasks, such as cell phone use while walking, alter gait features. This was assessed in a young adult population using LDS, gait regularity, and linear variability. The results indicated that cell phone use may compromise walking stability, which in turn may increase the risk of falling. Attentional dual-tasks that reduce the visual field and upper limb movement, such as the texting and talking conditions assessed in the present study, increase gait linear variability; although, the gait regularity increase while talking on a cell phone. Future studies could focus on strategies to develop safety guidelines for pedestrians using cell phones. Conflict of interest There are no conflict of interest to declare. Acknowledgements The authors are grateful to government agencies CAPES, CNPq, FAPEG and FAPEMIG. References [1] P.C. Kao, C.I. Higginson, K. Seymour, M. Kamerdze, J.S. Higginson, Walking stability during cell phone use in healthy adults, Gait Posture 41 (2015) 947–953, http://dx.doi.org/10.1016/j.gaitpost.2015.03.347. [2] E.M. Lamberg, L.M. Muratori, Cell phones change the way we walk, Gait Posture 35 (2012) 688–690, http://dx.doi.org/10.1016/j.gaitpost.2011.12.005. [3] S. Schaefer, D. Jagenow, J. Verrel, U. Lindenberger, The influence of cognitive load and walking speed on gait regularity in children and young adults, Gait Posture 41 (2015) 258–262, http://dx.doi.org/10.1016/j.gaitpost.2014.10.013. [4] T. Szturm, P. Maharjan, J.J. Marotta, B. Shay, S. Shrestha, V. Sakhalkar, The interacting effect of cognitive and motor task demands on performance of gait, balance and cognition in young adults, Gait Posture 38 (2013) 596–602, http://dx. doi.org/10.1016/j.gaitpost.2013.02.004. [5] J.B. Dingwell, R.T. Robb, K.L. Troy, M.D. Grabiner, Effects of an attention demanding task on dynamic stability during treadmill walking, J. Neuroeng. Rehabil. 5 (2008) 12, http://dx.doi.org/10.1186/1743-0003-5-12. [6] N. Polskaia, The Impact of Auditory and Visual Cognitive Tasks on Postural Control in Young Adults 128 Univ. Ottawa, 2015, http://dx.doi.org/10.1017/ CBO9781107415324.004 Thesis. [7] X. Qu, Age-related cognitive task effects on gait characteristics: do different working memory components make a difference? J. Neuroeng. Rehabil. 11 (2014) 149, http://dx.doi.org/10.1186/1743-0003-11-149. [8] H. Kim, J. Park, J. Cha, C. Song, Influence of mobile phone texting on gait
108