Walking capacity, performance, and confidence as predictors of falls in post-stroke individuals

ABSTRACT

This study aims to assess the correlation between walking capacity, performance, and confidence and the number of falls in post-stroke individuals, in addition to investigate which of these walking-related variables could predict falls in this population. In total, 95 post-stroke individuals were evaluated according to the number of falls in the last year and walking capacity (10-meter walk test), walking performance (ABILOCO-Brazil) and walking confidence (modified Gait Efficacy Scale). Pearson’’s correlation coefficient was used to assess the correlations between variables. Multiple linear regression was used to assess whether walking-related variables can predict falls in post-stroke individuals. No significant correlations were found between capacity (p=0.87) and walking performance (p=0.06) and number of falls. A significant, negative correlation, with moderate magnitude, was found only between walking confidence and falls (r=−0.43; p<0.01). In the regression analysis, only walking confidence remained in the model (p<0.01) as a predictor of the number of falls in post-stroke individuals, explaining 18% (R²=0.18) of this variable. Thus, this study results demonstrated that only walking confidence is directly related to the number of falls in post-stroke individuals, which is also the only variable that can be considered a predictor of this event in this population.

Keywords:
Gait; Accidental Falls; Stroke

RESUMO

O objetivo deste estudo foi avaliar a correlação entre capacidade, desempenho e confiança da marcha e número de quedas em indivíduos após sofrerem acidente vascular encefálico (AVE), além de investigar quais dessas variáveis relacionadas à marcha poderiam predizer as quedas nessa população. Foram avaliados 95 indivíduos pós-AVE, de acordo com número de quedas no último ano e capacidade de marcha (por meio de teste de caminhada de 10 metros), desempenho na marcha (pela aplicação do questionário ABILOCO-Brasil) e confiança na marcha (pela análise de modified gait efficacy scale). O coeficiente de correlação de Pearson foi utilizado para avaliar as correlações entre as variáveis. A regressão linear múltipla foi aplicada para avaliar se as variáveis relacionadas à marcha são capazes de predizer as quedas em indivíduos pós-AVE. Não foram encontradas correlações significativas entre capacidade (p=0,87) e performance na marcha (p=0,06) e número de quedas. Correlação significativa, negativa, de magnitude moderada, foi encontrada somente entre confiança na marcha e quedas (r=−0,43; p<0,01). Na análise de regressão, apenas a confiança na marcha permaneceu no modelo (p<0,01) como preditora do número de quedas em indivíduos pós-AVE, sendo capaz, sozinha, de explicar 18% (R²=0,18) dessa variável. Assim, os achados do estudo demonstraram que apenas a confiança na marcha está diretamente relacionada com o número de quedas em indivíduos pós-AVE, sendo essa também a única variável que pode ser considerada preditora desse evento nessa população.

Descritores:
Marcha; Acidente por Quedas; Acidente Vascular Cerebral

RESUMEN

El objetivo de este estudio fue evaluar la correlación entre la capacidad de la marcha, rendimiento y confianza con el número de caídas en personas después del accidente cerebrovascular (ACV), además de identificar las variables relacionadas con la marcha que podrían predecir caídas en esta población. Se evaluó a un total de 95 personas pos-ACV de acuerdo al número de caídas en el último año y la capacidad de la marcha (por medio de la prueba de caminata de 10 metros), el rendimiento de la marcha (aplicando el cuestionario ABILOCO-Brasil) y la confianza en la marcha (por el análisis de modified gait efficacy scale). Para evaluar las correlaciones entre variables, se utilizó el coeficiente de correlación de Pearson. Para determinar si las variables relacionadas con la marcha pueden predecir caídas en personas pos-ACV, se aplicó regresión lineal múltiple. No se encontraron correlaciones significativas entre la capacidad (p=0,87) y rendimiento de la marcha (p=0,06) con el número de caídas. Se encontró únicamente una correlación negativa significativa de magnitud moderada entre la confianza en la marcha y las caídas (r=−0,43; p<0,01). En el análisis de regresión, solamente la confianza en la marcha permaneció en el modelo (p<0,01) como la predictora del número de caídas en personas pos-ACV pudiendo explicar, por sí sola, el 18% (R²=0,18) de esta variable. Los hallazgos del estudio demostraron que la confianza en la marcha fue la única variable que estuvo directamente relacionada con el número de caídas en personas pos-ACV, además de ser la única predictora de caídas en esta población.

Palabras clave:
Marcha; Accidentes por Caídas; Accidente Cerebrovascular

INTRODUCTION

A cerebrovascular accident (CVA) is defined as an acute neurological dysfunction with rapid onset of symptoms-which has a vascular origin and may vary according to the affected region of the brain¹. With high mortality, the CVA is responsible for producing significant physical, psychological, cognitive, and social deficits in affected individuals, resulting in high morbidity and disability rates, especially for older adults². CVA impairs the body’s structures and functions and, as a result, the individual may have difficulties to perform basic activities of daily living³, generating different public health problems, such as falls⁴. Consequences of falls are serious for the health of these individuals, including fractures, major injuries, pain, fear of falling or post-fall syndrome, institutionalization and even death⁵.

Gait is a factor directly associated to falls, since if an individual with a CVA has a reduction in gait speed, this can be considered a causal factor for falls⁶. According to the International Classification of Functioning, Disability and Health (ICF), the concept of locomotion or gait is the ability to move while standing on a surface, so that one foot is always on the ground, as when walking, walking slowly, walking forwards, backwards or sideways⁷. There are two significant qualifiers for gait assessment, which are capacity and performance⁷. Capacity describes the ability of an individual to perform a task or an action and, to assess it, it is necessary to have a “standardized” environment to neutralize the variable influence of different environments⁷. On the other hand, performance describes what the individuals do in their usual living environment, that is, the “involvement in a life situation” or “the lived experience” of people in the context in which they live⁷. Both should be considered during gait assessment, as they provide additional information about the patient’s condition.

Another significant factor that must be evaluated in the gait of individuals with CVA, and which may be directly related to falls, is trust, since individuals who limit their activities because of fear of falling, particularly increase the risk of becoming fallers⁸. Confidence in gait can decrease after a stroke, due to a set of limitations that restrict mobility and the ability to perform daily life tasks. Thus, individuals with low self-confidence tend to focus more on their limitations and deficiencies⁹, changing their perception of the risk of falling and increasing its incidence.

There are previous studies investigating the relationship between gait and falls in post-stroke individuals. Faria et al.¹⁰ aimed to compare hemiparetic patients with and without a history of falls in the last six months, according to the components of the ICF and, specifically for gait speed, no significant difference was found between fallers and non-fallers. Another study, with 23 post-stroke individuals, aimed to verify the relationship between gait speed and self-efficacy for falls, finding no significant correlation between both measures¹¹. The study by Harris et al.¹² investigated the relationship between mobility and falls in 99 post-stroke individuals, also without finding mobility as an explanation for the risk of falls in this population. Although studies suggest that gait is not related to falls, all studies evaluated such activity through its ability, and no previous studies were found that investigated gait as a predictor of falls, including measures of performance and confidence.

Such results could help professionals to determine if there is a correlation and, if so, which factor contributes most to the reduction in the number of falls in this population, prioritizing their treatment during the rehabilitation process. Thus, the aim of this study was to evaluate the correlation between gait capacity, performance and confidence and the number of falls in post-stroke individuals, in addition to investigating which of these variables related to gait could predict falls in this population.

METHODOLOGY

Study design

This is an observational, cross-sectional study.

Participants

The recruitment period for participants was from August 2018 to March 2020, selected from members of the community in general. Inclusion criteria were: (1) individuals of both sexes, with no age restriction; (2) patients diagnosed with stroke and more than six months post-injury; and (3) being able to walk and answer questionnaires. Participants who had any other neurological or musculoskeletal conditions that could interfere with the tests were excluded.

Procedures

After the initial collection of the participants’ demographic and clinical data, the inclusion criteria and signature of the informed consent form were analyzed. Subsequently, participants were asked the number of falls in the last year, and measures of gait ability, performance and confidence were collected. All collections were performed at the Laboratory of Neurology at the Universidade Federal de Minas Gerais.

Gait ability

The ability to walk was assessed using the 10-meter walk test¹³. The test was performed on a flat surface, and the time needed to travel the central 10 meters was recorded in seconds using a digital stopwatch and used to calculate the individual’s usual speed, reported in m/s. Standardized instructions for performing the test were always provided by the same examiner¹⁴.

Gait performance

To measure gait performance, the ABILOCO-Brasil¹⁵ questionnaire was used. The questionnaire contains 13 items about walking in different everyday situations, and is conducted in the form of an interview, at which time participants were asked to rate their perceptions, according to the answers “Impossible” or “Possible” (impossible=0, possible=1)^15),(16. After the end of the interview, the answers were submitted to online analysis on the website www.rehab-scales.org, which converted the raw scores into a linear measure (logits)^15),(16.

Gait confidence

Gait confidence was assessed using the modified gait efficacy scale (mGES), a 10-item scale that assesses the individual’s perception of their level of gait confidence in challenging circumstances^17),(18. Items are rated individually on a 10-point Likert scale, in which a score of 1 represents “no confidence at all” and a score of 10 represents “complete confidence”, thus making scores from 10 to 100. The higher the score achieved by the individual, the greater the confidence in their mobility ability¹⁸.

Sample calculation

The sample size calculation was performed based on regression analysis, which considered the inclusion of three independent variables (ability, performance and confidence in walking). According to the formula n=30+10k, where k is the number of possible predictor variables to be included¹⁹, the minimum sample to be considered in the study is 60 individuals.

Statistical analysis

Descriptive statistics and normality tests (Kolmogorov-Smirnov) were performed for all variables. Pearson’s correlation coefficient was used to assess correlations between gait measures (ability, performance and confidence) and the number of falls. The strength of the correlations was classified as low (r<0.30), moderate (0.30≤r≤0.50) and high (r>0.50)²⁰. Multiple linear regression was used to assess which gait variable (ability, performance and confidence) best predicts falls. All analyses were performed using SPSS 23.0 statistical software with a significance level of 5%.

RESULTS

Out of 714 individuals, 619 could not be contacted, refused to participate, or did not meet the inclusion criteria. Thus, 95 individuals participated in the study, 38 men (40%). The mean age of participants was 67 years (SD=13), with a predominance of ischemic stroke (76%), and a mean onset time of 30 months (SD=27). Table 1 shows the detailed characteristics of the participants.

Table 1
Characteristics of the participants (n=95)

Age, years, mean (SD)	67 (13)
Sex, man, number (%)	38 (40)
CVA type, number (%)
Ischemic	72 (76)
Haemorrhagic	16 (17)
Unknown	7 (7)
Affected side, number (%)
Right	52 (55)
Left	43 (45)
Number of episodes, number (%)
1	63 (66)
>1	32 (34)
Onset time, months, average (SD)	30 (27)
Walking speed, m/s, average (SD)	0.79 (0.27)
Gait performance, ABILOCO (logits), average (SD)	2.95 (2.31)
Gait confidence, mGES (0-100), average (SD)	69 (22)
Number of falls, mean (SD)	1.15 (3.62)

• SD: standard deviation; CVA: stroke; mGES: modified gait efficacy scale.

Significant correlations between capacity (p=0.87) and gait performance (p=0.06) and number of falls were not found. Significative, negative, of moderate magnitude correlation was only found in gait confidence and falls (r=−0.43; p<0.01). In the regression analysis, only gait confidence remained in the model (p<0.01) as a predictor of the number of falls in post-stroke individuals, explaining 18% (R²=0.18) of this variable (Table 2).

Table 2
Regression analysis of possible gait-related variables as predictors of the number of falls (n=95)

Model	B (SD)	B	R2 adjusted	SEE
Model 1
Constant	5.47 (1.02)	-	-	-
Gait confidence	−0.07 (0.01)	−0.43	0.18	3.29

• SD: standard deviation; SEE: standard error of estimate.

DISCUSSION

This study aimed to evaluate the relationship between gait ability, performance, and confidence and the number of falls in post-stroke individuals, in addition to investigating which of these three gait-related variables could predict falls in this population. The results showed that there was a significant correlation only with gait confidence, which is also the only variable that can be considered a predictor of the number of falls in post-stroke individuals.

Similar to previous findings in the literature, this study also did not find a significant correlation between gait ability, assessed by gait speed, as well as previous studies and falls^10)-(12. Gait speed measurements are often used to measure the capacity of post-stroke individuals¹⁴. However, walking, especially in the community, requires gait adaptation in adverse conditions, with attentional demands and performing additional tasks, for example, carrying a bag²¹, components not included in the gait speed test or in ABILOCO-Brasil, performance measure which also did not show a significant correlation with falls. The recovery of walking after a stroke, directly related to performing such activity without the presence or with the fewest possible number of falls, is complex, and requires not only gait speed, but mainly the ability to perform more complex walking tasks in diverse environments²¹, since capacity is not measured in an individual’s real-life environment, but in a standardized environment. In addition, although ABILOCO assesses performance in different activities in the patient’s real-life environment, this instrument assesses the perception regarding the possibility of performing or not certain tasks¹⁶, but does not estimate their perception regarding the level of difficulty or confidence of the patients in the suggested activities.

Regarding gait confidence, the only variable that showed a significant correlation and the only predictor found in the study, this is an ability that can precede the actual ability and influence the performance of daily activities²². In post-stroke individuals, even if they manage to increase gait speed, it may be that this improvement is not transferred to the home environment or to community walking, due to low confidence¹⁸. Individuals with low gait confidence also demonstrate lower performance¹⁸. The loss of confidence in gait generates fear of falling and, as a result, individuals restrict their daily activities, generating a cycle that starts with fear and results in restriction, which consequently leads to decreased muscle strength and balance deficit, generating even more disabilities and greater risk of falls²³.

The main positive point of this study is the realization of an unprecedented research, which fills an important gap in the literature, with the use of low-cost, easy-to-handle, quick-to-apply instruments, with appropriate measurement properties and that can be used in the home environment. However, the use of a convenience sample is the main limitation of this article. Although this sample was large and drawn from many different locations, it was not randomly selected and therefore may not be fully representative of the general population.

CONCLUSION

The study findings showed that only gait confidence is directly related to the number of falls in post-stroke individuals, which is also the only variable that can be considered a predictor of this event in this population. Thus, rehabilitation, which aims to reduce the number of falls in post-stroke individuals, should focus on a plan that includes, mainly, training capable of increasing confidence in this activity, since gains only in capacity and performance possibly will not be able to significantly improve the number of falls in these individuals.

REFERÊNCIAS

• ¹
  Silva SM, Corrêa JCF, Pereira GS, Corrêa FI. Social participation following a stroke: an assessment in accordance with the international classification of functioning, disability and health. Disabil Rehabil. 2017;41(8):879-86. doi: 10.1080/09638288.2017.1413428.
  » https://doi.org/10.1080/09638288.2017.1413428
• ²
  Teixeira-Salmela LF, Oliveira ESG, Santana EGS, Resende GP. Fortalecimento muscular e condicionamento físico em hemiplégicos. Acta Fisiatrica. 2000;7(3):108-18. doi: 10.5935/0104-7795.20000001.
  » https://doi.org/10.5935/0104-7795.20000001
• ³
  Elloker T, Rhoda A, Arowoiya A, Lawal IU. Factors predicting community participation in patients living with stroke, in the Western Cape, South Africa. Disab Rehabil. 2019;41(22):2640-7. doi: 10.1080/09638288.2018.1473509.
  » https://doi.org/10.1080/09638288.2018.1473509
• ⁴
  Geyh S, Cieza A, Schouten J, Dickson H, Frommelt P, Omar Z, et al. ICF core sets for stroke. J Rehabil Med. 2004;(44 Suppl):135-41. doi: 10.1080/16501960410016776.
  » https://doi.org/10.1080/16501960410016776
• ⁵
  Marchetti GF, Whitney SL. Older adults and balance dysfunction. Neurol Clin. 2005;23(3):785-805. doi: 10.1016/j.ncl.2005.01.009.
  » https://doi.org/10.1016/j.ncl.2005.01.009
• ⁶
  Gama ZAS, Conesa AG, Ferreira MS. Epidemiología de caídas de ancianos en España: una revisión sistemática. Rev Esp Salud Publica. 2008;82(1):43-56.
• ⁷
  Organização Mundial da Saúde; Organização Panamericana de Saúde. CIF: classificação internacional de funcionalidade, incapacidade e saúde. São Paulo: Edusp; 2003.
• ⁸
  Friedman SM, Munoz B, West SK, Rubin GS, Fried LP. Falls and fear of falling: which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention. J Am Geriatr Soc. 2002;50(8):1329-35. doi: 10.1046/j.1532-5415.2002.50352.x.
  » https://doi.org/10.1046/j.1532-5415.2002.50352.x
• ⁹
  Lopes KT, Costa DF, Santos LF, Castro DP, Bastone AC. Prevalence of fear of falling among a population of older adults and its correlation with mobility, dynamic balance, risk and history of falls. Braz J Phys Ther. 2009;13(3):223-9. doi: 10.1590/S1413-35552009005000026.
  » https://doi.org/10.1590/S1413-35552009005000026
• ¹⁰
  Faria CDCM, Saliba VA, Teixeira-Salmela LF, Nadeau S. Comparison between post-stroke hemiparetic subjects with and without history of falls on the basis of the international classification of functioning, disability and health. Fisioter Pesqui. 2010;17(3):242-7. doi: 10.1590/S1809-29502010000300010.
  » https://doi.org/10.1590/S1809-29502010000300010
• ¹¹
  Ricci NA, Ferrarias GP, Molina KI, Dib PM, Alouche SR. Velocidade de marcha e autoeficácia em quedas em indivíduos com hemiparesia após acidente vascular encefálico. Fisioter Pesqui. 2015;22(2):191-6. doi: 10.590/1809-2950/14484522022015.
  » https://doi.org/10.590/1809-2950/14484522022015
• ¹²
  Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL. Relationship of balance and mobility to fall incidence in people with chronic stroke. Phys Ther. 2005;85(2):150-8.
• ¹³
  Salbach NM, Mayo NE, Higgins J, Ahmed S, Finch LE, Richards CL. Responsiveness and predictability of gait speed and other disability measures in acute stroke. Arch Phys Med Rehabil. 2001;82(9):1204-12. doi: 10.1053/apmr.2001.24907.
  » https://doi.org/10.1053/apmr.2001.24907
• ¹⁴
  Nascimento LR, Caetano LCG, Freitas DCMA, Morais TM, Polese JC, Teixeira-Salmela LF. Different instructions during the ten-meter walking test determined significant increases in maximum gait speed in individuals with chronic hemiparesis. Braz J Phys Ther. 2012;16(2):122-7. doi: 10.1590/S1413-35552012005000008.
  » https://doi.org/10.1590/S1413-35552012005000008
• ¹⁵
  Avelino PR, Faria-Fortini I, Basílio ML, Menezes KKP, Magalhães LC, Teixeira-Salmela LF. Adaptação transcultural do ABILOCO: uma medida de habilidade de locomoção, específica para indivíduos pós-acidente vascular encefálico. Acta Fisiatrica. 2016;23(4):161-5. doi: 10.5935/0104-7795.20160031.
  » https://doi.org/10.5935/0104-7795.20160031
• ¹⁶
  Caty GD, Arnould C, Stoquart GG, Thonnard JL, Lejeune TM. ABILOCO: a rasch-built 13-item questionnaire to assess locomotion ability in stroke patients. Arch Phys Med Rehabil. 2008;89(2):284-90. doi: 10.1016/j.apmr.2007.08.155.
  » https://doi.org/10.1016/j.apmr.2007.08.155
• ¹⁷
  Avelino PR, Menezes KKP, Nascimento LR, Faria-Fortini I, Faria, CDCM, Scianni AA, et al. Adaptação transcultural da Modified Gait Efficacy Scale para indivíduos pós-acidente vascular encefálico. Rev Ter Ocup. 2018;29(3):230-6. doi: 10.11606/issn.2238-6149.v29i3p230-236.
  » https://doi.org/10.11606/issn.2238-6149.v29i3p230-236
• ¹⁸
  Newell AM, Vanswearingen JM, Hile E, Brach JS. The modified gait efficacy scale: establishing the psychometric properties in older adults. Phys Ther. 2012;92(2):318-28. doi: 10.2522/ptj.20110053.
  » https://doi.org/10.2522/ptj.20110053
• ¹⁹
  Knapp TR, Campbell-Heider N. Numbers of observations and variables in multivariate analyses. West J Nurs Res. 1989;11(5):634-41.
• ²⁰
  Cohen J. Statistical power analysis for the behavioral sciences. 2a ed. Hillsdale: Lawrence Erlbaum Associates; 1988.
• ²¹
  Robinson CA, Shumway-Cook A, Matsuda PN, Ciol MA. Understanding physical factors associated with participation in community ambulation following stroke. Disabil Rehabil. 2011;33(12):1033-42. doi: 10.3109/09638288.2010.520803.
  » https://doi.org/10.3109/09638288.2010.520803
• ²²
  McAuley E, Mihalko SL, Rosengren KS. Self-efficacy and balance correlates of fear of falling in elderly. J Aging Phys Act. 1997;5(4):329-40.
• ²³
  Jorstad EC, Hauer K, Becker C, Lamb SE. Measuring the psychological outcomes of falling: a systematic review. J Am Geriatr Soc. 2005;53(3):501-10. doi: 10.1111/j.1532-5415.2005.53172.x.
  » https://doi.org/10.1111/j.1532-5415.2005.53172.x

• 4
  Financing source: nothing to declare
• 6
  Approved by the Research Ethics Committee: No. CAAE 65765817.3.0000.5149.

Publication Dates

• Publication in this collection
  11 Mar 2022
• Date of issue
  Oct-Dec 2021

History

• Received
  06 Apr 2021
• Accepted
  15 Nov 2021