R. Barclay1, D. Jiang2, Y. Liu2, J. Ripat3, R. Tate2, S. Webber1
1College of Rehabilitation Sciences, Rady Faculty of Health Sciences, University of Manitoba, Physical Therapy, Winnipeg, Canada, 2Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Community Health Sciences, Winnipeg, Canada, 3College of Rehabilitation Sciences, Rady Faculty of Health Sciences, University of Manitoba, Occupational Therapy, Winnipeg, Canada
Background: Community ambulation is important for the health and well-being of adults as they age. It includes aspects such as walking for shopping, accessing public transportation or getting to social activities; as well as walking for leisure or exercise. Despite health benefits of community ambulation, approximately 34% of Canadian adults aged 45 and older walk outside of their home only twice per week or less. A model of community ambulation after stroke was developed by Barclay et al in 2015. However, this has not been empirically verified among older adults.
Purpose: The objective of this study was to test an existing model of community ambulation post stroke and to validate the model for older adults.
Methods: Ethics approval was received from a university health research ethics board. The Canadian Longitudinal Study of Aging (CLSA) is a rich dataset of self-report and physical variables of participants aged 45-85. Secondary data analysis of the comprehensive baseline and maintaining contact questionnaire data from the CLSA was completed. Variables included self-report questionnaires and physical assessments of gait speed, timed up and go (TUG), five times sit to stand (STS) representing leg strength, and single leg balance. Structural Equation Modeling (SEM) was used to test the pathways for community ambulation, based on the 2015 model. Multi-group SEM was also conducted to examine sex differences. A measurement model was developed using the following latent variables [and associated measured variables]: Ambulation [frequency of walking outside in last week, number of hours of walking per day, ability to walk with help, life-space index - frequency and aids used]; Health Perception [general health perception, pain frequency, pain intensity]; Gait speed [comfortable gait speed, TUG, STS, balance]; Environment [afraid to walk alone in neighborhood, rural or urban setting]. Variables of depression, falls, age and sex were added to the structural model.
Results: Of the 30,097 participants in the comprehensive dataset, the mean age was 63.0 (SD 10.3) and 15,320 (50.9%) were female. The final model had a close fit with RMSEA (90% CI) = 0.029 (0.029, 0.030), CFI = 0.931. Results demonstrated that health perception, gait speed and environment all have a positive impact on ambulation. Gait speed has both direct and indirect impact (through number of falls) on ambulation. Multi-group SEM results reveal a similar measurement model and pathways for males and females.
Conclusion(s): The community ambulation model post stroke was validated with adults aged 45-85. The model appears consistent for both males and females. Future research will explore models for adults aged 65 and over, and people with osteoarthritis.
Implications: Identifying factors and associations of community ambulation in adults aged 45-85 will assist physiotherapists in determining community ambulation goals in collaboration with clients, and assist in focusing assessment and intervention. It is important to understand factors that are barriers or facilitators to community ambulation in adults aged 45-85 to enable optimization of prevention and rehabilitation programs. Results may be beneficial in policy planning for prevention of mobility decline as individuals age.
Funding, acknowledgements: Canadian Institutes of Health Research Catalyst Grant: Analysis of CLSA Data
Keywords: Community ambulation, older adult, structural equation modeling
Topic: Older people
Did this work require ethics approval? Yes
Institution: University of Manitoba
Committee: Health Research Ethics Board
Ethics number: HS21345 (H2017:406)
All authors, affiliations and abstracts have been published as submitted.
