Influence of the use of cane on the gait cycle of individuals who are blind

Abstract

The aim of this study was to perform a biomechanical characterization of the gait cycle in individuals who are blind. Five individuals with ages between 16 and 19 years participated in this study. The task consisted of walks of 12m measured in two conditions: (1) with cane and (2) wihtout cane; a total of 20 walks in each condition were performed. During walks, participants were monitored with a Vicon 3D-motion capture system. Spatialtemporal, kinematic, kinetic, and dynamic parameters were recorded and compared between the two conditions. We observed an interaction between the condition and ankle angular measures (p = .003); the interaction was due to differences induced by condition in instants ‘opposite toe off’ (p = .045) and ‘opposite initial contact’ (p = .019). We also obtained a significant difference in the negative ankle-joint-power measures between conditions (p = .044). This study showed that the use of cane changes the gait pattern in individuals who are blind. The subtle changes in ankle behaviour when walking with a cane, compared with no cane, suggest better application of the force during the initial stages of support leading to a more comfortable gait. This type of assessment of gait may be important to improve mobility training and rehabilitation strategies.

Keywords

Biomechanics cane gait orientation and mobility impaired vision blind

Introduction

Persons who are blind often show differences in gait variables when compared to persons with normal vision (Hallemans, Beccu, Van Look, Ortibus, & Truijen, 2009; Moura e Castro, 1993; Rosen, 2010). Gait is known as a sequence and standardized reproduction of movements between the lower limbs interaction and the total body mass, allowing the individual to move from one point to another (Baker, 2013; Perry, 2005). The gait cycle is a set of events occurring during the time interval between the accomplishment of a determined contact of a foot and the repetition of the same type of contact with the same foot (Gabriel, 2001). For a representation of the cycle, see Supplementary Methods, Appendix Figure 1. Studies investigating gait in individuals with vision impairment (we use vision impairment when referring groups that can include blind or low vision subjects) found a significant heterogeneity in the gait cycle (Long & Giudice, 2010; Pereira, 1984; Rosen, 2010). In these studies, that has been attributed to the absence or reduction of vision leading to a limited amount of motor experiences. Gait difficulties can be so debilitating that some individuals with vision impairment avoid walking without a human guide when performing their daily activities because they are afraid of falls and bumping into objects (Mason & McCall, 2013; Moura e Castro, 1993). The use of mobility devices is an alternative to a human guide.

Mobility devices such as a cane, used by individuals with impaired vision, can interfere with the biomechanical aspects of gait (Geruschat & Smith, 2010). A cane provides the user with sensory information that helps him or her to detect what lies ahead and, in this way, will have a great influence on the motor response and consequently on its gait cycle (LaGrow, 2010). As such, a cane can help to increase confidence and eventually motivation to initiate independent and faster movements leading to a change in the biomechanical aspects of gait (Geruschat & Smith, 2010). At this point, it is important to define biomechanics as a complementary diagnostic tool used to study movements and forces associated with human motion (Winter, 1987). Movements and forces result from mechanical relations established and controlled by the individual. There is evidence in the literature that gait patterns can predict the risk of falls, reduction in quality of life, and eventually increased risk of mortality (Hallemans et al., 2009; Hallemans, Ortibus, Meire, & Aerts, 2010; Hallemans, Ortibus, Truijen, & Meire, 2011; Iosa, Fusco, Morone, & Paolucci, 2012; Ramsey, Blasch, Kita, & Johnson, 1999).

Some studies have analysed the gait cycle of persons with impaired vision in laboratories equipped with modern biomechanical methods that allow the analysis of three-dimensional aspects of the cycle (Hallemans et al., 2009; Hallemans et al., 2010; Hallemans et al., 2011; Iosa et al., 2012). The gait cycle of persons with impaired vision, not using a cane, is characterized by low speed, by short stride length, and with long double support periods (Hallemans et al., 2009; Hallemans et al., 2010; Nakamura, 1997). Other authors not only measured different parameters but also observed characteristics similar to immature gait patterns such as longer stride width, flexed knees, decreased ‘initial contact’ (stance sub-phase), and a large degree of out-toeing when compared with normal sighted individuals (Rosen, 2010). However, these studies did not clarify what happens to the gait cycle of the same person with impaired vision when using a cane compared with not using it. Better understanding of the impact of the cane in the gait cycle can help with the decisions that need to be taken by patients and therapists during vision rehabilitation.

The aim of this study was to understand the impact of using a cane on gait variables of individuals who are blind. We investigated this by comparing biomechanical variables when individuals who are blind walked with or without a cane.

Methods

Participants

All participants had vision impairment from birth; at the time of the study, vision was reduced to light perception and therefore can be classified as blind. Exclusion criteria included existence of psychiatric conditions, physical disabilities, and communication difficulties. Information recorded included weight and height, length of both legs, and time as an independent cane user. The leg length was measured from the anterior superior iliac spine to the medial malleolus, passing through the knee joint. These results are summarized in Table 1; all participants were right-hand dominant and the cane was held in the dominant hand. All five participants (three were females) attended or had finished high school at the time of the study; their school is a reference school for education of visually impaired students in Braga, Portugal. The study conformed to the tenets of the Declaration of Helsinki. Participants or their legal guardians signed the informed consent form after receiving oral information about the purpose and agreeing to take part in the study.

Table 1.

Demographic and anthropometric information of the five participants.

	Age (years)	Weight (N)	Height (mm)	Left leg length (mm)	Right leg length (mm)	Cane user (years)
M	18	652.1	1633	872	872	4
SD	1.4	150.0	91	48	50	1.6
Min	16	413.8	1559	825	820	2
Max	19	840.4	1763	945	950	6

M = mean; SD = standard deviation; Min = minimum; Max = maximum, N= Newton. All participants were right-dominant, and therefore results for the right limbs can be interpreted as results for the dominant limb and for the left as results for the non-dominant limb.

Equipment for biomechanical assessment

The biomechanical assessment of gait was performed at MovLab – the Motion Capture Lab in Universidade Lusófona de Humanidades e Tecnologias – in Lisbon.

Motion capture was supported by the Vicon 3D-motion capture MX system. This system was based on 10 MX cameras (8 × 1.3 GB; 2 × 2.0 GB) connected to the MX Ultranet controlling hardware unit which was used to track the motion of 41 spherical reflective markers (9.5 mm in diameter) placed on anatomical standardized landmarks of the participants represented in Figure 1 (Model Plug-in-Gait – Full Body). Anthropometric data were measured with an SECA 764 balance and by anthropometric instruments from Siber Hegner. Kinematic and kinetic data were recorded at 200 Hz. A force plate (AMTI-OR6) was mounted on the floor of the capture volume recording the ground reaction force data at 1000 Hz. The force plate was connected to a strain gauge amplifier (AMTI MSA-6 MiniAmp) and to the Vicon^® MX Control in order to synchronize the Vicon MX Ultranet. A host computer ran the data acquisition and processing software Nexus 1.7 from Vicon. The software supported data acquisition and kinematic, kinetic, and dynamic data processing.

Figure 1.

Plug-in-gait marker placement (‘Vicon’s plug-in-gait marker placement’) and an example of the subject viewer with a Software Vicon.

Procedure for data collection

Spatialtemporal, kinematic, kinetic, and dynamic information of gait pattern was recorded with and without a cane. Each participant used his own cane that is normalized according to the height of the user. Walks were performed within a rectangle in which the force platform was located (see Figure 2). A full recognition of space was performed before the measurements to ensure that participants were confident and informed about the direction of motion during the measurements. Carpets, marked in Figure 2, helped with tactile discrimination on the path and provided information about direction of motion.

Figure 2.

Schematic representation showing the walking path and the approximate location of the force platform. The force platform (or force plate) is close to the start of the path because, according to the experience of our laboratory, this leads to a better constancy (reduces the number of invalid trials) of the measurements for the support values (Gabriel, Monteiro, Moreira, Faria, & Abrantes, 2010; L. Santos & Abrantes, 2010).

Participants were wearing appropriate clothing – bathing or bikini shorts – barefoot and had passive retroreflective marks glued with antiallergenic adhesive tape on the body, according to the method described for the biomechanical assessment (plug-in-gait – full body). Each participant performed 40 walks under two conditions: 20 walks with a cane and 20 walks without a cane. In each condition, 10 walks were performed with the left foot making contact with the force platform and 10 with the right foot. The first foot in contact with the ground was chosen by the participant. Participants were instructed to walk at their normal (preferred) walking speed. Each measurement took approximately 30 s (including preparation and walking). Measurements were performed alternately with/without the cane, walks in which the measurements in the force platform failed were repeated.

Variables analysed

Eighteen gait variables divided into four categories were studied. The distribution by category is spatiotemporal variables – (1) step width, (2) step length, (3) stride length, (4) centre of gravity’s velocity, (5) cadence, (6) speed, (7) stance time, (8) swing time, (9) double support time; kinematic variables – (10) pelvic tilt angle, (11) hip angle, (12) knee angle, (13) maximum value of knee flexion, (14) ankle angle; kinetic variables – (15) ground reaction force anteroposterior dynamic; (16) vertical component; dynamic variables – (17) maximum value of positive and (18) negative ankle joint power (AJP).

The description of the variables is given in Figures 3 and 4 and Supplementary Appendix (see also D. Santos, 2012) and the results are given in Tables 2 to 4.When applicable, variables were measured and compared for four instants of the gait cycle.

Figure 3.

(a–d) Representation of the four instants of the gait cycle described and analysed in this study. (e) Lateral perspective of the gait cycle for the right leg. IC: ‘initial contact’: right foot hits the ground; oTO: ‘opposite toe off’: contralateral foot (left foot) leaves the ground; oIC: ‘opposite initial contact’: contralateral foot (left foot) hits the ground; TO: ‘toe off’: right foot leaves the ground. (f) Interior perspective of the gait cycle for the right leg.

Figure 4.

Values of the ankle angle for four instants of the gait cycle (IC: initial contact, oTO: opposite toe off, oIC: opposite initial contact, TO: toe off). The two conditions tested were cane (C: thick line) and no cane (NC: dotted line); the error bars represent one standard deviation of the mean and are plotted above the line for C and below the line for NC. The expected curve (thin line) has been adapted from The Pathokinesiology Service and the Physical Therapy Department (1996).

Table 2.

Summary of the spatiotemporal variables for all participants and trials divided by condition (C: cane; NC: no cane).

Spatiotemporal variables	Condition		Mean difference	p value
Spatiotemporal variables	NC (M ± SD)	C (M ± SD)	Mean difference	p value
Step width (m)	0.096 ± 0.056	0.086 ± 0.060	–0.010	.230
Step length (m)	0.629 ± 0.130	0.642 ± 0.106	0.013	.430
Stride length (m)	1.257 ± 0.215	1.266 ± 0.185	0.009	.748
CoG velocity
IC (m/s)	1.111 ± 0.351	1.197 ± 0.996	0.086	.417
oTO (m/s)	1.286 ± 0.976	1.186 ± 0.331	–0.100	.333
oIC (m/s)	1.185 ± 0.338	1.161 ± 0.299	–0.024	.596
TO (m/s)	1.261 ± 0.341	1.239 ± 0.300	–0.022	.636
Cadence (stride/min)	54.77 ± 7.159	53.51 ± 6.235	–1.260	.186
Speed (m/s)	1.140 ± 0.278	1.160 ± 0.322	0.025	.563
Stance time (s)	0.676 ± 0.116	0.687 ± 0.170	0.011	.481
Swing time (s)	0.438 ± 0.046	0.449 ± 0.047	0.011	.079
Double support time (s)	1.257 ± 0.215	1.266 ± 0.185	0.009	.748

CoG: centre of gravity; IC: initial contact; oTO: opposite toe off; oIC: opposite initial contact; TO: toe off.

Negative mean differences indicate a reduced value during walks with a cane compared to walks without a cane. Differences were considered statistically significant for p < .05.

The four instants of the gait cycle are shown in Figure 3 and are defined as follows: ‘initial contact’ (IC, Figure 3(a)) which corresponds to instant of 0% of cycle gait when the foot makes contact with the ground, ‘opposite toe off’ (oTO, Figure 3(b)) which corresponds to instant of 10% of cycle gait when the contralateral foot leaves the ground, ‘opposite initial contact’ (oIC, Figure 3(c)) which corresponds to instant of 50% of cycle gait when the contralateral foot hits the ground, and ‘toe off’ (TO, Figure 3(d)) which corresponds to instant of 60% of cycle gait when the foot leaves the ground.

Statistical procedures

The mean value of both feet for each run and variable was used for statistical analysis because there were no statistically significant differences between right and left feet. Repeated measures were analysed with linear mixed models (LMMs; SPSS, v22). LMMs are one of the recommended methods to analyse repeated measures in small samples; a detailed explanation with mathematical details can be found in the literature (Muth et al., 2015; Skene & Kenward, 2010). For each outcome measure, we used 40 measurements of each participant giving a total of 200 measurements. For LMMs, participants were defined as ‘random factor’ and ‘condition’ as fixed factor. Two conditions were defined with Condition 1 = cane (C) and Condition 2 = no cane (NC). According to the outcome, other fixed factors were defined. For example, to analyse all but maximum knee angle in the kinematic variables (shown in Table 3), the ‘instant of the gait cycle’ with four levels was defined as fixed factor.

Table 3.

Summary of the kinematic variables for all participants and trials divided by condition (C: cane; NC: no cane). Differences that reached statistical significance are in bold font.

Kinematic variables	Condition		Mean difference	p value
Kinematic variables	NC (M ± SD)	C (M ± SD)	Mean difference	p value
Pelvic tilt angle
IC (°)	11.3 ± 4.0	11.2 ± 3.5	–0.2	.761
oTO (°)	11.1 ± 3.7	10.7 ± 3.3	–0.4	.375
oIC (°)	10.7 ± 4.4	10.6 ± 4.0	–0.1	.861
TO (°)	10.2 ± 4.0	9.8 ± 3.5	–0.4	.440
Hip angle
IC (°)	34.6 ± 5.1	34.8 ± 5.6	0.2	.812
oTO (°)	32.2 ± 5.5	32.7 ± 6.4	0.5	.590
oIC (°)	–8.4 ± 4.6	–8.6 ± 3.8	–0.2	.708
TO (°)	–3.1 ± 4.7	–3.4 ± 4.8	–0.3	.667
Knee angle
IC (°)	6.7 ± 5.8	4.4 ± 6.7	–2.3	.247
oTO (°)	14.8 ± 4.3	14.2 ± 14.8	–0.6	.748
oIC (°)	10.7 ± 5.8	9.3 ± 10.7	–1.4	.480
TO (°)	32.2 ± 4.8	30.7 ± 32.2	–1.5	.439
Max. knee flexion (°)	56.4 ± 5.8	55.2 ± 18.6	–1.2	.546
Ankle angle
IC (°)	3.8 ± 4.7	2.8 ± 4.1	–1.0	.098
oTO (°)	0.8 ± 4.2	–0.4 ± 3.8	–1.1*	.045*
oIC (°)	14.1 ± 8.3	16.6 ± 6.5	2.5*	.019*
TO (°)	–6.8 ± 10.6	–4.2 ± 8.8	2.6	.063

IC: initial contact; oTO: opposite toe off; oIC: opposite initial contact; TO: toe off; Max. knee flexion: maximum value of knee flexion.

Negative mean differences indicate a reduced value during walks with a cane compared to walks without a cane.

(*) Differences statistically significant for p < .05.

Results

Spatiotemporal results

A summary of the values obtained for the nine spatiotemporal variables analysed is provided in Table 2. Descriptive statistics results about the within-participant variability are available in Supplementary results. For spatiotemporal variables, the mean differences (mDif) between condition C (higher values for C) and NC were as follows: mDif = 0.013 m for step length, mDif = 0.009 m for stride length, mDif = 0.025 m/s for speed, mDif = 0.011 s for stance time, mDif = 0.011 s for swing time, and mDif = 0.009 s for double support time. Lower values were obtained for C when compared with NC for step width: mDif = 0.010 m and cadence: mDif = 1.3 stride/min. The trend observed when using a cane was that participants moved faster (increased speed, step length, stride length, and reduced cadence) and had better stability (reduced step width). Spatiotemporal values in both conditions assessed seem to differ from the expected values obtained in normal sighted persons (Neumann, 2002).

Kinematic results

Kinematic results are summarized in Table 3 and show an interaction between conditions (C and NC) and instants of the gait cycle for the ankle angle. The interaction is shown in Figure 4 and was due to the ankle angles obtained for instants oTO (p = .045) and oIC (p = .019). No other statistically significant differences between the two conditions were observed for the remaining variables. However, compared with the expected curve, there was a shift of the curve that shows a tendency towards an exaggerated ankle dorsiflexion at all instants of the gait cycle.

Kinetic and dynamic results

Kinetic and dynamic values are summarized in Table 4. For dynamic variables, the maximum value of negative AJP was different for the two conditions (mDif = –0.190 W/kg, p = .044). In this parameter, the blind participants using the cane show statistically significant higher values.

Table 4.

Summary of the kinetic and dynamic variables for all participants and trials divided by condition (C: cane; NC: no cane). Differences that reached statistical significance are in bold font.

Kinetic/dynamic variables	Condition		Mean difference	p value
Kinetic/dynamic variables	NC (M ± SD)	C (M ± SD)	Mean difference	p value
Kinetic variables
GRF VERT
oTO (N/kg)	533.81 ± 186.16	550.83 ± 188.62	17.02	.521
oIC (N/kg)	593.47 ± 153.74	611.92 ± 135.47	18.45	.369
GRF AP
oTO (N/kg)	–102.52 ± 57.84	–110.71 ± 64.38	–8.19	.345
oIC (N/kg)	123.85 ± 67.32	135.53 ± 63.15	11.69	.207
Dynamic variables
Max. positive AJP (W/kg)	3.49 ± 1.64	3.55 ± 1.54	0.06	.788
Max. negative AJP (W/kg)	–0.74 ± 0.83	–0.93 ± 0.44	–0.19*	.044*

oTO: opposite toe off; oIC: opposite initial contact; GRF VERT/AP: ground reaction force vertical/anteroposterior component; Max. positive/negative AJP: maximum value of positive/negative ankle joint power.

Negative mean differences indicate a reduced value during walks with a cane compared to walks without a cane.

(*) Differences statistically significant for p < .05.

Discussion and conclusion

This study was motivated by the current limited knowledge regarding gait patterns of blind people while using a cane. The specific objective of the study was to provide evidence about the influence of the use of cane in biomechanical parameters and, ultimately, gather evidence that the use of cane leads to better control of gait. We found differences between the two conditions studied, cane and no cane, for two of the variables analysed: ankle angle and maximum value of negative AJP.

Our study shows that the amplitude of ankle movement increases when using a cane compared with no cane, which is indicative of increased stability and is associated with reduced plantar contact of the foot and improvement in shock absorption. That is, when participants walked with cane, the necessity of sensory exploration reduced and the foot was used less to provide haptic feedback information about the ground. This suggests a shift in strategy from a cautious gait towards a more relaxed gait.

Results for the dynamic variables in Table 4, maximum value of positive/negative AJP, reinforce the idea of a more relaxed gait with cane. Joint power reflects the net rate of generating or absorbing energy by all muscles and other connective tissues crossing a joint. A positive joint-power value (AJP) indicates power generation, which reflects a concentric muscle activation and a release of energy from the previously elongated connective tissues, whereas a negative value indicates power absorption, which reflects an eccentric muscle activation and the passive stretching of the connective tissues (Neumann, 2002; Winter, 1987). In our case, the range (the difference between maximum and minimum AJP) increased with the cane mostly due to a significant increase in the negative AJP. Thus, our results with cane show higher values of absorption, which demonstrate a more efficient absorption of energy by the antagonistic muscles, giving the inferior limb more stability and better shock absorption. As other authors have shown in individuals with normal vision, higher absolute values of maximum negative AJP are associated with comfortable walking and absolute values for this variable tend to reduce when adults with gait problems are forced to walk quickly (Graf, Judge, Ounpuu, & Thelen, 2005).

The methodology used in this study is a major strength, the plug-in-gait model of Vicon represents the entire body allowing a detailed knowledge of the integrated motor behaviour in three dimensions. We recognize some limitations, namely, the small sample size and the lack of a control group. These limitations will be taken in consideration in future studies.

To conclude, in this study, we showed that biomechanics analysis can be successfully used to characterize gait patterns in individuals who are blind. This precise and objective technique captured subtle changes in ankle behaviour when participants walked with the cane (when compared with no cane). The observed changes indicate that individuals would be able to stop the movement due to better application of the force during the initial stages of support. This type of assessment of gait may be important in improving mobility training and rehabilitation strategies.

Supplemental Material

3Appendix_-_Supplementary_methods – Supplemental material for Influence of the use of cane on the gait cycle of individuals who are blind

Supplemental material, 3Appendix_-_Supplementary_methods for Influence of the use of cane on the gait cycle of individuals who are blind by Diana Santos, João MCS Abrantes, Peter Lewis and António Filipe Macedo in The British Journal of Visual Impairment

Supplemental Material

4Appendix_-_Supplementary_results – Supplemental material for Influence of the use of cane on the gait cycle of individuals who are blind

Supplemental material, 4Appendix_-_Supplementary_results for Influence of the use of cane on the gait cycle of individuals who are blind by Diana Santos, João MCS Abrantes, Peter Lewis and António Filipe Macedo in The British Journal of Visual Impairment

Footnotes

Acknowledgements

We are highly grateful to Dr Ivo Fialho Roupa for providing the expertise for using the Vicon system. His dedication made possible the collection, processing, and analysis of the data.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Diana Santos

António Filipe Macedo

References

Baker

(2013). Measuring walking: A handbook of clinical gait analysis. London, England: Mac Keith Press.

Gabriel

(2001). Biomechanics of hiking (pp. 1–85). Vila Real, Portugal: Universidade de Trás-os-Montes e Alto Douro.

Gabriel

Monteiro

Moreira

Faria

Abrantes

(2010). Assessment based on plantar pressure: Emphasis on its use in symptom-free postmenopausal women. In Black

(Ed.), Overweightness and walking (pp. 59–73). New York: Nova Science Publishers.

Geruschat

Smith

(2010). Improving the use of low vision for orientation and mobility. In Wiener

Welsh

lasch

(Eds.), Foundations of orientation and mobility (Vol. 2, pp. 54–90). Sewickley, PA: American Foundation for the Blind Press.

Graf

Judge

Ounpuu

Thelen

(2005). The effect of walking speed on lower-extremity joint powers among elderly adults who exhibit low physical performance. Archives of Physical Medicine and Rehabilitation, 86, 2177–2183.

Hallemans

Beccu

Van Look

Ortibus

Truijen

(2009). Visual deprivation leads to gait adaptations that are age and context-specific: II. Kinematic parameters. Gait & Posture, 30, 307–311.

Hallemans

Ortibus

Meire

Aerts

(2010). Low vision affects dynamic stability of gait. Gait & Posture, 32, 547–551.

Hallemans

Ortibus

Truijen

Meire

(2011). Development of independent locomotion in children with a severe visual impairment. Research in Developmental Disabilities, 32, 2069–2074. doi:10.1016/j.ridd.2011.08.017

Iosa

Fusco

Morone

Paolucci

(2012). Effects of visual deprivation on gait dynamic stability. Scientific World Journal, 2012, 974560. doi:10.1100/2012/974560

10.

LaGrow

(2010). Improving perception for orientation and mobility. In Wiener

Welsh

Blasch

(Eds.), Foundations of orientation and mobility (Vol. 2, pp. 3–26). Sewickley, PA: American Foundation for the Blind Press.

11.

Long

R. G.

Giudice

N. A.

(2010). Establishing and maintaining orientation for mobility. In Wiener

Welsh

Blasch

(Eds.), Foundations of orientation and mobility (Vol. 1, pp. 45–62). Sewickley, PA: American Foundation for the Blind Press.

12.

Mason

McCall

(2013). Visual impairment: Access to education for children and young people. New York, NY: David Fulton Publishers.

13.

Moura e Castro

(1993). Study of the influence of aerobic endurance capacity on orientation and mobility of the blind (PhD Thesis). Faculdade de Desporto da Universidade do Porto, Porto, Portugal. Retrieved from http://hdl.handle.net/10216/10235l.

14.

Muth

Bales

K. L.

Hinde

Maninger

Mendoza

S. P.

Ferrer

(2015). Alternative models for small samples in psychological research. Educational and Psychological Measurement, 76, 64–87. doi:10.1177/0013164415580432

15.

Nakamura

(1997). Quantitative analysis of gait in the visually impaired. Disability and Rehabilitation, 5(19), 194–197.

16.

Neumann

(2002). Kinesiology of the musculoskeletal system. Foundations for Physical Rehabilitation. St. Louis, MO: Mosby.

17.

The Pathokinesiology Service and the Physical Therapy Department. (1996). Observational gait analysis. Downey, CA: Los Amigos Research and Educational Institute.

18.

Pereira

(1984). Balance and spatial concepts in visually impaired children. Ludens, 8(2), 5–7.

19.

Perry

(2005). Marcha normal. In Perry

(Ed.), Análise da Marcha (Vol. 1, pp. 1–149). São Paulo, Brazil: Manole.

20.

Plug-in Gait Reference Guide. (2016). Vicon Motion 24 Systems Limited.

21.

Ramsey

Blasch

Kita

Johnson

(1999). A biomechanical evaluation of visually impaired persons’ gait and long-cane mechanics. Journal of Rehabilitation Research & Development, 36, 323–232.

22.

Rosen

(2010). Kinesiology and sensorimotor functioning for students with low vision. In Wiener

Welsh

Blasch

(Eds.), Foundations of orientation and mobility (Vol. 1, pp. 138–172). Sewickley, PA: American Foundation for the Blind Press.

23.

Santos

(2012). Influence of the use of walking stick on the gait of young people with blindness - intraindividual variability of biomechanical parameters (Master Thesis). Escola Superior de Educação Jean Piaget de Arcozelo-Canelas, Vila Nova de Gaia. Retrieved from https://www.dropbox.com/s/u1m75f1pnzd6hpd/TESE_DIANA_SANTOS.pdf?dl=0

24.

Santos

Abrantes

(2010). Multiple sclerosis subjects plantar pressure: A new tool for postural instability diagnosis. In Natal

J. T. R.

(Ed.), 6th TMSi, International Conference on Technology and Medical Sciences (pp. 197–203). Abingdon, UK: CRC Press.

25.

Skene

Kenward

(2010). The analysis of very small samples of repeated measurements I: An adjusted sandwich estimator. Statistics in Medicine, 29, 2825–2837. doi:10.1002/sim.4073

26.

Winter

(1987). The biomechanics and motor control of human gait. Waterloo, Ontario, Canada: Waterloo Biomechanics.