Contents lists available atScienceDirect
Gait & Posture
journal homepage:www.elsevier.com/locate/gaitpost
Comparison of upper limb burn injury versus simulated pathology in terms
of gait and footprint parameters
Özden Özkal
a,*
, Melek Merve Erdem
b, Kemal K
ısmet
c, Semra Topuz
a aHacettepe University, Faculty of Physical Therapy and Rehabilitation, Ankara, TurkeybRecep Tayyip Erdoğan University, School of Güneysu Physical Therapy and Rehabilitation, Rize, Turkey cSelçuk University, Faculty of Nursing, Konya, Turkey
A R T I C L E I N F O Keywords: Arm swing Burn injury Gait analysis Plantar pressure Spatiotemporal characteristics A B S T R A C T
Background: Little is known about whether a simulated upper limb condition reflects a real (burn-injury) upper limb pathology in terms of gait/footprint parameters.
Research question: The main aim of this study was to investigate the differences in these parameters between two conditions (real-simulation).
Methods: The study included burn patients (n = 30) and a control group of 30 healthy subjects. Gait and footprint parameters were evaluated using the GAITRite electronic walkway. Kinesiophobia and pain were as-sessed with the Tampa Kinesiophobia Scale and Visual Analog Scale, respectively. Gait evaluation of the control group was performed randomly in two conditions:1. Normal arm swing (control group) 2.Elbowflexed at 90° with a bandage (simulated group).
Results: Step and stride length in the burn group were significantly shorter than in the other groups (p < 0.05). Stance phase was significantly higher while swing phase, velocity and cadence were lower in the burn group (p < 0.05). Peak time in the midfoot for both sides were significantly higher in the burn group (p < 0.05). Peak time in the hindfoot for the affected side was significantly lower while peak time in the hindfoot for the intact side was significantly higher in the burn group compared to the simulated group (p < 0.05). There were significant correlations between pain, kinesiophobia and velocity, and cadence in the burn group (p < 0.05). Significance: Compared to the other groups, patients with burn injury have different gait/footprint parameters due to increased pain and kinesiophobia. To determine the effects of upper limb injury and arm swing on gait parameters, a real pathology should be considered rather than a simulated pathology.
1. Introduction
Upper limb burn injury (ULBI) often leads to significant dysfunction [1]. Severe pain, limited range of motion in the joint and kinesiophobia may affect upper limb functionality and activities of daily living in the acute term of injury [2,3]. Increased pain, and limitations in range of motion and function of the upper limb may cause deterioration in re-ciprocal arm swing which plays a substantial role in gait pattern by contributing to the co-ordination of the trunk, pelvis and leg move-ments, reducing energy expenditure and improving gait stability and balance [4–7].
Previous studies have examined the effects of arm swing on gait parameters and stability in real pathological diseases such as cerebral palsy, stroke, Parkinson’s disease and transhumeral amputees [8–11]. Those studies have reported that disruptions in reciprocal arm
movements during walking lead to significant alterations in gait char-acteristics, such as decreased cadence, gait velocity and step and stride length asymmetry [8–11]. It was reported that the amount of arm swing was associated with gait speed in Parkinson’s disease [8]. In addition, it has been shown that a decrease in step length of contralateral lower extremity was correlated with a decrease in amount of arm swing on the amputated side in unilateral transhumeral amputees [11].
With the exception of these diseases, there are few studies that have investigated simulated pathologies focused on the restriction of arm motion on gait parameters in healthy individuals [6–12]. It has been shown that the prevention of arm movement with casting has no effect on gait velocity or cadence, although there is a significant reduction in the step time and length of the casted limb compared to the non-casted limb [12]. A similar study conducted on simulated elbow contracture demonstrated that single support time, stride length and gait velocity
https://doi.org/10.1016/j.gaitpost.2019.10.027
Received 26 August 2019; Received in revised form 5 October 2019; Accepted 17 October 2019
⁎Corresponding author at: Hacettepe University, Faculty of Physical Therapy and Rehabilitation, Samanpazari, Ankara, 06100, Turkey. E-mail addresses:ozdenozkal@gmail.com,ozdenozkal@hacettepe.edu.tr(Ö. Özkal).
0966-6362/ © 2019 Elsevier B.V. All rights reserved.
decreased while double support time increased in simulated elbow conditions compared to the control condition [6].
In terms of prevention of arm swing, although gait parameters of simulated pathologies have been shown to be different from healthy individuals, it is not clear whether a simulated upper limb condition reflects an unilateral ULBI in terms of gait and footprint parameters. Because, most of the studies evaluating the effect of arm swing on gait were performed in real pathologies such as parkinson, stroke or cere-bral palsy in which the lower extremities are generally affected as well as the upper extremity. To determine the effect of arm swing, only one study was conducted on isolated unilateral transhumeral amputees [11]. However the evaluation of an isolated and unilateral ULBI and a simulated upper limb pathology together has not been studied to date. Therefore, the aim of this study was to determine the differences in terms of gait and footprint parameters between a simulated upper limb pathology and a unilateral ULBI and to investigate the effects of acute unilateral ULBI on these parameters.
2. Methods
This case-control study included a total of 60 individuals aged 20–48 years, comprising 30 patients with unilateral ULBI, and 30 gender-matched healthy subjects. Patients were excluded if they had any additional diseases that could affect gait parameters or mask the effect of the burn injury. Exclusion criteria for both groups included: (1) having a burn injury in any area of the body other than the upper ex-tremity, (2) orthopedic, musculoskeletal, neurological problems or previous lower extremity surgical interventions that may affect the mobility and gait. The procedure of the study was explained to the participants and those who agreed to participate voluntarily provided written informed consent and were included in the study. Approval for the study was granted by the Institutional Research Ethics Commitee. 2.1. Intervention
For the burn injury patients, the elbow was fixed to the anterior trunk at 90 degreesflexion during walking because of pain. This pattern was the natural and antalgic gait which was generally used in their daily life because of pain and kinesiophobia in the acute period of the burn injury. Therefore, the gait of the burn injury patients was eval-uated using the position defined above with their burn dressing. To understand the differences of gait parameters between simulated and real upper limb injury, the dominant or non-dominant upper ex-tremities of the healthy individuals were restricted in accordance with the pre-determined restriction protocol applied to the dominant or non-dominant side of the upper extremities of the individuals matched to the burn group. To be able to mimic the antalgic gait patterns of the burned individuals, the upper extremity of the healthy individuals was fixed to the anterior trunk at 90° elbow flexion with an elastic bandage, as in the patients with ULBI. Gait analysis was then applied randomly to the healthy participants in two different conditions: 1) Normal arm swing (control group) 2) Elbowflexed at 90° in the arm restricted po-sition (simulated group).
2.2. Gait analysis
The GAITRite (CIR System INC. Clifton, NJ, USA, 07012) electronic walkway was used to evaluate the spatio-temporal and footprint para-meters during walking [13]. The gait parameters of the individuals in the burn group were assessed within 3 days of the burn injury. Gait evaluation of patients with ULBI was performed before burn dressing. Participants walked barefoot at self-selected pace. Gait analysis was started after a trial to allow the participants to become familiar with the test. Gait velocity, cadence and bilateral gait parameters of step length, stride length, percentage of stance and swing were used for analysis. In addition spatio-temporal parameters, and footprint variables
demonstrating plantar pressure distribution were analysed during walking. The GAITRite system defines the footprint as 12 trapezoids with 6 on the lateral and 6 on the medial side. To simplify the inter-pretation of results, the 12 trapezoids are grouped into three areas: hindfoot, midfoot and forefoot. The peak time and peak pressure values are obtained for the 3 areas of the footprint. The GAITRite system ex-presses the plantar pressure value as a percentage of the maximum pressure [14].
2.3. Kinesiophobia and pain assessment
To assess the fear of movement, the Turkish version of the Tampa Scale for Kinesiophobia (TSK) was used [15]. The TSK inclues 17 items which evaluate the fear of re-injury due to pain or movement. The TSK is rated on a 4-point Likert scale ranging from 17 to 68, with higher scores indicating a high level of kinesiophobia [15,16]. The TSK has been shown to be a reliable tool for the measurement of kinesiophobia in patients with ULBI [17].
The severity of pain of ULBI patients, and the feelings of discomfort due to the restriction of the control group were recorded using a Visual Analogue Scale (VAS) ranging from 0 (no pain / no discomfort) to 10 (extreme pain / extreme discomfort). The VAS is a proven reliable tool for the measurement of pain in the acute term of injury [18]. The mean score of pain intensity during the day was recorded in patients with ULBI. The pain duration of ULBI patients was questioned. Patients in-dicated that their pain persisted throughout the day. When the factors that lead to increase in pain was questioned, patients stated that their pain increased after burn dressing. Therefore gait evaluation of patients was performed before burn dressing to eliminate negative effects of pain.
2.4. Data analysis
Bilateral gait parameters were recorded as“affected” or “intact” sides according to the burned/restriction side of the upper limb. It was accepted that the lower limbs were matched with the ipsilateral upper limbs while recording the bilateral gait and footprint parameters; right-side gait parameters were recorded as affected side for right-side uni-lateral ULBI.
In the comparison of bilateral gait parameters between all the groups, the burned side of the patients with ULBI was compared to the restricted side of the simulated group and the non-dominant side of the control group. The intact side of the burned patients was compared to the non-restricted side of the simulated group and the dominant side of the control group.
Statistical analysis was performed using IBM SPSS 21.0 software. Normality tests (Kolmogorov–Smirnov, and Shapiro–Wilk) and histo-gram were used to determine conformity of the variables to normal distribution. The Student’s t-test or Mann-Whitney U test was used in the comparison of physical and clinical characteristics. The Paired Samples t-test or the Wilcoxon Signed Rank test was used for compar-isons between the affected and intact sides.
The Kruskal–Wallis test or One-way ANOVA was used to compare gait and footprint parameters between the three groups. If there was a significant difference between the 3 groups, the Tukey or Tamhane’s T2 correction according to the test of homogeneity of variances for One-way ANOVA or pairwise comparisons with adjusted significance levels for Kruskal-Wallis test, were used to determine significant pairwise differences for each analysis. Spearman or Pearson correlation coeffi-cient was used for correlation analysis between the gait parameters and pain/discomfort, kinesiophobia. A value of p < 0.05 was accepted as statistically significant.
3. Results
participants are presented inTable 1. The groups were similar in re-spect of age, gender, height and weight (p˃ 0.05). The subjects with ULBI had higher pain and kinesiophobia values than those of the si-mulated group (p < 0.05).
When the all groups were compared in terms of gait parameters, there were significant differences in step length, stride length, swing and stance for the two sides (affected-intact) (p < 0.05). Pairwise comparisons showed that step and stride length of both sides (a ffected-intact) in the burn group were significantly shorter than those of the control and simulated groups (p < 0.001). Swing phase was sig-nificantly lower for both sides (affected-intact) while stance phase was significantly higher for the affected side in the burn group compared to the control and simulated groups (p < 0.05). Stance phase for the in-tact side was significantly higher in the burn group than in the control group (p < 0.05). ULBI patients had a slower gait velocity than both the control and simulated groups (p < 0.001). Cadence was sig-nificantly lower in the burn group than in the control group (p = 0.027). Peak time for forefoot (both affected and intact sides) and peak pressure for all sides and regions (forefoot-midfoot-hindfoot) were found to be similar in all the groups (p˃ 0.05). Peak time for affected and intact midfoot were significantly higher in the burn group com-pared to the other groups (p < 0.05). Peak time for affected hindfoot was significantly lower while the intact hindfoot was significantly higher in the burn group than in the simulated group (p < 0.05) (Table 2).
In the comparisons of the bilateral gait and footprint parameters between the affected and intact sides for all groups, there was no sig-nificant difference for all parameters (p ˃ 0.05) (Table 3). According to the correlation analysis between velocity/cadence and pain/kinesio-phobia, there were significant correlations between pain and velocity, and cadence in the burn group (p < 0.05) (Table 4).
4. Discussion
The aim of the present study was to determine the differences in gait and footprint parameters between an acute-unilateral ULBI and a si-mulated upper limb pathological condition. It was demonstrated that subjects with ULBI had lower step and stride length, velocity and ca-dence than both the control and simulated groups. In addition, stance time was increased while swing phase was decreased in the burn group
compared to the other groups. It was also observed that patients with ULBI had different footprint parameters in terms of peak time in both midfoot and hindfoot from those of the control and simulated groups. Thesefindings showed that simulated upper limb pathology does not reflect exactly a unilateral ULBI in terms of gait and footprint para-meters. To the best of our knowledge, this is thefirst report showing the effects of a simulated and real upper limb pathological condition on gait by matching of participants.
Greater declines in step and stride length for both affected and in-tact sides, velocity, and cadence were noted in patients with ULBI, which was consistent with thefindings of previous studies that have shown that pathologies including the upper limb cause a less stable and slower gait pattern [8–11]. It is known that step length is associated with gait velocity and the increase in step length is compensated by the increase in arm swing amplitude [19]. The decline in step and stride length in the burn group may have resulted from the lack of arm swing and decreased velocity. Thefindings of the present study also demon-strated that increased pain severity and kinesiophobia level caused a decrease in gait velocity and cadence in patients with ULBI. Recent studies have generally reported that the distruption or prevention of arm-swing leads to decreased velocity in real or simulated pathological conditions [6–12], although increased pain and kinesiophobia are other important factors affecting gait velocity in an acute burn injury. In addition to changes in those gait parameters, the increase in step phase and reduced swing phase times were also observed in patients with burn injury. These changes in swing and stance phase could be attrib-uted to decreased velocity and cadence in addition to the determinant of a less stable walking pattern. It has been reported that increased walking speed is an indicator of a more stable gait pattern when com-pared to slow walking due to a decrease in the mediolateral center of gravity movements [20].
Contrary to most previous studies [6,7] which focused on the pre-vention of arm movement in healthy subjects, the present study de-monstrated that the restriction of arm motion has no significant effect on step and stride length, swing, stance, velocity and cadence. Only one similar simulated study reported that upper limb casting does not lead to a change in gait velocity while significantly reducing step length and single support time on the casted side when compared to the non-casted side [12]. Those simulated studies indicated that the additional weight of the orthosis or casting used to restrict the arm swing may have Table 1
Physical, Clinical and Burn Injury Characteristics of Participants.
Group 1 (ULBI) (n = 30)
Group 3 (Simulated) (n = 30)
p
Physical and Clinical Characteristics Age (year) 32.96 ± 10.21 29.50 ± 5.43 0.287
Gender (F/M) 26/4 26/4 1.00
Height (cm) 163.70 ± 6.70 165.73 ± 5.33 0.199
Weight (kg) 66.43 ± 12.09 62.26 ± 6.75 0.106
Pain /Discomfort 5.01 ± 2.41 2.75 ± 1.97 < 0.001* Kinesiophobia 43.20 ± 6.67 34.16 ± 7.24 < 0.001*
Burn Injury Characteristics Affected Extremity (D/ND) 17/13 17/13 1.00
TBSA (%) 2.86 ± 1.47 – –
Depth of Burn Injury 2nd degree deep and superficial – – Cause of Burn Injury
(Scald /Contact) 27/3 – – Burn localization n(%) Hand 8(26.64%) – – Hand + Wrist 9(29.97%) Hand + Wrist + Forearm 6(19.98%) Forearm 4(13.32%) Forearm + upper arm + elbow 2(6.66%) Shoulder + upper arm + forearm 1(3.33%) Affected Joint
n(%)
Wrist 15(49.95%) – –
Elbow 2(6.66%)
Shoulder 1(3.33%)
Data are given as mean ± standard deviation or ratio. ULBI = upper limb burn injury; TBSA = total burn surface area; F = female; M = male; D = dominant; ND = non-dominant; n = number of participants *p < 0.05.
affected the gait parameters [6,12]. The similarity in term of gait parameters between the control and simulated groups in the present study could have resulted from the lack of additional weight of ban-dages or the lack of time spent with elastic banban-dages. Even though the arm motion was prevented similar to the real pathology, it is not pos-sible to simulate the pain and kinesiophobia experienced by patients with ULBI.
When the differences in gait parameters between the affected and intact sides within the groups were examined, all the groups were ob-served to have symmetric gait pattern. These results are consistent with a previous study that investigated the effects of simulated elbow con-tracture on gait [6]. However, gait asymmetry may be affected in chronic pathological conditions that involve the upper limb [8–11] Therefore, it is possible to say that a simulated pathology is more si-milar in terms of gait symmetry to an acute ULBI.
Studies conducted on real pathologies or simulated patologies have generally focused on spatio-temporal parameters of gait. However, evaluation of gait and footprint parameters together has not been in-vestigated to date. The present study results pertaining to footprint parameters indicated that peak time in both the affected and intact midfoot is increased in the patients with ULBI compared to the control and simulated groups. It is known that increased peak time is correlated with lower gait velocity and increased stance time [21]. Similarly, de-creased velocity and cadence or inde-creased stance phase may lead to increased peak time in the midfoot in patients with burn injury. How-ever, this increase was observed in only the midfoot areas. Therefore, another possible reason for this increase in peak time may be associated with a new gait stabilization strategy which has developed following the burn injury to obtain a more stable gait pattern or to maintain functional balance during walking because of increased pain and ki-nesiophobia. It was also found that the peak time in the affected hindfoot is decreased while it is increased in the intact hindfoot. Re-duced peak time in the affected hindfoot may have developed as a result Table 2
Comparison of gait and footprint parameters between the groups.
Group 1 (ULBI)
Group 2 (Control)
Group 3 (Simulated) Group 1-2-3 Group 1-2 Group 1-3 Group 2-3
Mean ± SD Mean ± SD Mean ± SD p p p p
Affected Side Step Length (cm) 52.76 ± 6.15 63.64 ± 6.17 62.75 ± 8.90 < 0.001* < 0.001* < 0.001* 1.000 Stride Length (cm) 106.76 ± 12.52 128.14 ± 11.41 125.40 ± 2.30 < 0.001* < 0.001* < 0.001* 1.000 Swing (%) 35.82 ± 1.83 37.46 ± 2.20 36.16 ± 4.79 0.001* 0.001* 0.039* 0.733 Stance (%) 64.16 ± 1.84 62.55 ± 2.20 63.84 ± 4.78 0.001* 0.001* 0.046* 0.724
Intact Side Step Length (cm) 53.26 ± 6.56 63.89 ± 5.30 62.52 ± 6.04 < 0.001* < 0.001* < 0.001* 1.000 Stride Length (cm) 107.46 ± 11.68 128.09 ± 11.51 127.28 ± 16.64 < 0.001* < 0.001* < 0.001* 1.000 Swing (%) 35.65 ± 2.43 37.18 ± 2.31 37.15 ± 1.17 0.001* 0.002* 0.017* 1.000 Stance (%) 64.00 ± 2.06 62.48 ± 1.25 62.85 ± 1.16 0.003* 0.003* 0.051 1.000 Affected Forefoot peak time (s) 1.36 ± 1.82 1.35 ± 0.53 1.50 ± 0.55 0.250 peak pressure (%) 46.81 ± 27.31 38.18 ± 4.72 39.62 ± 4.52 0.321 Intact Forefoot peak time (s) 1.51 ± 2.34 1.36 ± 0.55 1.49 ± 0.58 0.295 peak pressure (%) 44.42 ± 23.20 37.92 ± 4.10 38.04 ± 4.48 0.760 Affected Midfoot peak time (s) 2.68 ± 1.64 1.42 ± 0.28 1.53 ± 0.48 < 0.001* < 0.001* 0.002* 1.000 peak pressure (%) 24.25 ± 14.94 22.14 ± 2.89 21.83 ± 3.49 0.877 Intact Midfoot peak time (s) 2.67 ± 3.33 1.38 ± 0.22 1.42 ± 0.23 < 0.001* 0.001* 0.003* 1.000 peak pressure (%) 21.86 ± 10.74 21.44 ± 3.64 22.35 ± 3.40 0.448 Affected Hindfoot peak time (s) 0.22 ± 1.88 2.16 ± 0.44 2.16 ± 0.60 < 0.001* < 0.001* < 0.001* 1.000 peak pressure (%) 38.23 ± 27.32 39.73 ± 4.80 38.67 ± 4.57 0.678 Intact Hindfoot peak time (s) 2.10 ± 12.36 2.15 ± 0.41 2.08 ± 0.57 < 0.001* < 0.001* < 0.001* 1.000 peak pressure (%) 36.50 ± 22.17 40.71 ± 4.57 39.69 ± 4.78 0.471 Velocity (cm/s) 92.74 ± 17.56 118.06 ± 11.95 111.92 ± 14.09 < 0.001* < 0.001* < 0.001* 0.245 Cadence (step/min) 104.59 ± 11.48 111.12 ± 6.40 108.37 ± 6.99 0.025* 0.027* 0.341 0.314
* p < 0.05; ULBI = upper limb burn injury; SD = standard deviation. Table 3
Comparison gait and footprint variables between affected and intact sides within groups.
Gait and Footprint Characteristics Group 1 (ULBI) Group 2 (Control) Group 3 (Simulated) p p p Step Length (cm) 0.437 0.607 0.835 Stride Length (cm) 0.313 0.896 0.432 Swing (%) 0.733 0.299 0.550 Stance (%) 0.660 0.440 0.557 Forefoot peak time (s) 0.673 0.789 0.139 Forefoot peak pressure
(%)
0.434 0.753 0.060
Midfoot peak time (s) 0.658 0.960 0.082 Midfoot peak pressure
(%)
0.223 0.278 0.461
Hindfoot peak time (s) 0.704 0.365 0.861 Hindfoot peak pressure
(%)
0.835 0.293 0.365
ULBI = upper limb burn injury.
Table 4
Correlations between pain/discomfort, kinesiophobia and gait characteristics.
Gait Characteristics Group 1 (ULBI)
Group 3 (Simulated)
Pain Kinesiophobia Discomfort Kinesiophobia
Velocity (cm/s) −0.791* −0.790* 0.353 −0.030 Cadence (step/min) −0.639* −0.627* 0.342 −0.039
of patients avoiding weight-bearing on the affected limb beacuse of increased pain and kinesiophobia. In addition, both peak time in the affected side for the midfoot and peak time in the intact side for the hindfoot may be increased to compensate for the decrease in peak time in the affected side for the hindfoot in patients with ULBI.
One of the limitations of the current study can be said to be the lack of long-term follow-up of the patients with ULBI. The patients with ULBI may distrupt gait pattern/mechanics as contractures occur, or vice versa, normalize the gait pattern by compensating over time as the pain decreases. Therefore, it is not possible to generalize thesefindings of the gait pattern of patients with ULBI. Another limitation could be con-sidered to be the short duration of elastic bandage application in the simulated group. Future studies should investigate whether simulated pathological conditions which have been designed for lower limbs re-flect a real pathology.
5. Conclusions
The results of this study showed that patients with ULBI have dif-ferent gait and footprint parameters from those of the simulated group because of increased pain and kinesiophobia. These results are clini-cally important as the real and simulated pathological conditions were evaluated together and the effects of an acute ULBI on these parameters were demonstrated. Thefindings of the present study might provide a rationale of why a real pathology should be evaluated rather than a simulated pathology to investigate the effect of arm-swing on gait. Funding
There is nofinancial support. Ethical approval
Hacettepe University Non-interventional Clinical Research Ethics Board (Decision number: GO 18/785-34).
Declaration of Competing Interest
The authors declare no conflicts of interest. References
[1] A. Wu, D.W. Edgar, F.M. Wood, The QuickDASH is an appropriate tool for
measuring the quality of recovery after upper limb burn injury, Burns 33 (2007) 843–849.
[2] M. Willebrand, G. Andersson, M. Kildal, B. Gerdin, L. Ekselius, Injury-related fear-avoidance, neuroticism and burn-specific health, Burns 32 (2006) 408–415. [3] K. Voon, I. Silberstein, A. Eranki, M. Phillips, F.M. Wood, D.W. Edgar, Xbox
KinectTMbased rehabilitation as a feasible adjunct for minor upper limb burns
re-habilitation: a pilot RCT, Burns 42 (2016) 1797–1804.
[4] Z. Yizhar, S. Boulos, O. Inbar, E. Carmeli, The effect of restricted arm swing on energy expenditure in healthy men, Int. J. Rehabil. Res. 32 (2009) 115–123. [5] S.M. Bruijn, O.G. Meijer, J.H. van Dieën, I. Kingma, C.J. Lamoth, Coordination of
leg swing, thorax rotations, and pelvis rotations during gait: the organisation of total body angular momentum, Gait Posture 27 (2008) 455–462.
[6] S.K. Trehan, A.L. Wolff, M. Gibbons, H.J. Hillstrom, A. Daluiski, The effect of si-mulated elbow contracture on temporal and distance gait parameters, Gait Posture 41 (2015) 791–794.
[7] M.P. Ford, R.C. Wagenaar, K.M. Newell, Arm constraint and walking in healthy adults, Gait Posture 26 (2007) 135–141.
[8] M.D. Lewek, R. Poole, J. Johnson, O. Halawa, X. Huang, Arm swing magnitude and asymmetry during gait in the early stages of Parkinson’s disease, Gait Posture 31 (2010) 256–260.
[9] J.L. Stephenson, S.J. De Serres, A. Lamontagne, The effect of arm movements on the lower limb during gait after a stroke, Gait Posture 31 (2010) 109–115. [10] T. Delabastita, K. Desloovere, P. Meyns, Restricted arm swing affects gait stability
and increased walking speed alters trunk movements in children with cerebral palsy, Front. Hum. Neurosci. 15 (2016) 354.
[11] S. Topuz, E. Kirdi, A.I. Yalcin, O. Ulger, H. Keklicek, G. Sener, Effects of arm swing on spatiotemporal characteristics of gait in unilateral transhumeral amputees, Gait Posture 68 (2019) 95–100.
[12] D. Dreyfuss, A. Elbaz, A. Mor, G. Segal, E. Calif, The effect of upper limb casting on gait pattern, Int. J. Rehabil. Res. 39 (2016) 176–180.
[13] K.E. Webster, J.E. Wittwer, J.A. Feller, Validity of the GAITRite® walkway system for the measurement of averaged and individual step parameters of gait, Gait Posture 22 (2005) 317–321.
[14] The Gaitrite Electronic Walkway: Measurements and Definitions, CIR System Inc., Sparta, NJ, 2006.
[15] O.T. Yilmaz, Y. Yakut, F. Uygur, N. Ulug, Turkish version of the Tampa Scale for Kinesiophobia and its test-retest reliability, Turk. J. Physiother. Rehabil. 22 (2011) 44–49.
[16] J.W.S. Vlaeyen, S.J. Linton, Fear-avoidance and its consequences in chronic mus-culoskeletal pain: a state of the art, Pain 85 (2000) 317–332.
[17] M.I. Sgroi, M. Willebrand, L. Ekselius, B. Gerdin, G. Andersson, Fear-avoidance in recovered burn patients: association with psychological and somatic symptoms, J. Health Psychol. 10 (2005) 491–502.
[18] P.E. Bijur, W. Silver, E.J. Gallagher, Reliability of the visual analog scale for mea-surement of acute pain, Acad. Emerg. Med. 8 (2001) 1153–1157.
[19] J. Romkes, K. Bracht-Schweizer, The effects of walking speed on upper body ki-nematics during gait in healthy subjects, Gait Posture 54 (2017) 304–310. [20] S.M. Bruijn, J.H. van Dieën, O.G. Meijer, P.J. Beek, Is slow walking more stable? J.
Biomech. 42 (2009) 1506–1512.
[21] E.B. Titianova, P.S. Mateev, I.M. Tarkka, Footprint analysis of gait using a pressure sensor system, J. Electromyogr. Kinesiol. 14 (2004) 275–281.
