Thilarajah S1,2, Bower K3, Pua YH1, Williams G3,4, Tan D1,5, Koh G6, Clark R7
1Singapore General Hospital, Singapore, Singapore, 2University of the Sunshine Coast, Queensland, Australia, 3The University of Melbourne, Victoria, Australia, 4Epworth Healthcare, Victoria, Australia, 5Singapore Institute of Technology, Singapore, Singapore, 6National University of Singapore, Singapore, Singapore, 7University of Sunshine Coast, Queensland, Australia
Background: Stroke survivors are not meeting recommended levels of physical activity. There is consensus that the key modifiable prognostic factors that contribute to post-stroke physical activity levels need to be identified so that new interventions can be developed to target these factors.
Purpose: This exploratory prospective cohort study investigated the modifiable prognostic factors at discharge from inpatient rehabilitation that contribute to physical activity levels at three months post-discharge.
Methods: Participants with mild disability were recruited from an inpatient rehabilitation unit in Singapore. The candidate factors (i.e., measures of gait speed, balance, strength, mood and motivation) were measured within one week prior to discharge from inpatient rehabilitation and repeated at three months following discharge. Outcomes measuring different aspect of physical activity were also completed at this timepoint. An ankle-worn accelerometer was used to measure walking-related physical activity (steps/day), the Activity Card Sort (ACS) as a measure of physical activity participation, and the International Physical Activity Questionnaire Short 7 days (IPAQ-S7) to evaluate the intensity of physical activity. Separate multivariable linear regression models or proportional odds regression models were used to evaluate the associations of the candidate variables with the physical activity outcomes, adjusting for age, gender, stroke severity, and time since stroke. To avoid assuming linearity, all continuous predictors were modelled as restricted cubic splines, unless there was insufficient evidence against the linearity (null) assumption (P>.20).
Results: Sixty-six individuals consented to participate in the study, with 64 completing baseline assessment (median time post stroke: 24 days) and 55 completing the follow-up assessment (mean age: 59 years, median time post stroke: 117days). Physical function measures (i.e., gait speed and step test) were found to be significant modifiable factors contributing to physical activity participation (i.e., ACS), walking related physical activity (i.e., steps/day) and intensity of physical activity (i.e., IPAQ-S7). Strength of the affected dorsiflexor muscles contributed to physical activity participation and intensity. Higher intrinsic and introjected regulation, measured using the Behavioural Regulation in Exercise Questionnaire-2, contributed to higher physical activity participation. Anxiety (Hospital Anxiety and Depression Scale) demonstrated a non-linear relationship with physical activity participation. A steep rise in physical activity participation with increasing anxiety levels was observed up to an approximate threshold score of 4-5, above which there was no appreciable association between anxiety and activity.
Conclusion(s): Gait speed and balance are modifiable factors that contribute to all aspects of post-stroke physical activity at discharge from inpatient rehabilitation. Anxiety demonstrated a non-linear association with participation in high demand leisure activities at discharge. The results demonstrated that better physical function contributed to future increased physical activity.
Implications: Physical activity levels can be influenced by multiple factors such as physical function and psychosocial factors. The factors that contribute to physical activity levels may also differ depending on the stage of stroke recovery. The findings of this study stress the importance of optimising gait and balance early after stroke. However, motivation and mood were also significant modifiable factors and may potentially be targeted in interventions to improve physical activity.
Keywords: Physical activity, stroke, predictor
Funding acknowledgements: Australian Government RTP Scholarship (Author ST)
NHMRC Career Development Fellowship (Author RC)
Singapore General Hospital Research Grant (Author DT)
Purpose: This exploratory prospective cohort study investigated the modifiable prognostic factors at discharge from inpatient rehabilitation that contribute to physical activity levels at three months post-discharge.
Methods: Participants with mild disability were recruited from an inpatient rehabilitation unit in Singapore. The candidate factors (i.e., measures of gait speed, balance, strength, mood and motivation) were measured within one week prior to discharge from inpatient rehabilitation and repeated at three months following discharge. Outcomes measuring different aspect of physical activity were also completed at this timepoint. An ankle-worn accelerometer was used to measure walking-related physical activity (steps/day), the Activity Card Sort (ACS) as a measure of physical activity participation, and the International Physical Activity Questionnaire Short 7 days (IPAQ-S7) to evaluate the intensity of physical activity. Separate multivariable linear regression models or proportional odds regression models were used to evaluate the associations of the candidate variables with the physical activity outcomes, adjusting for age, gender, stroke severity, and time since stroke. To avoid assuming linearity, all continuous predictors were modelled as restricted cubic splines, unless there was insufficient evidence against the linearity (null) assumption (P>.20).
Results: Sixty-six individuals consented to participate in the study, with 64 completing baseline assessment (median time post stroke: 24 days) and 55 completing the follow-up assessment (mean age: 59 years, median time post stroke: 117days). Physical function measures (i.e., gait speed and step test) were found to be significant modifiable factors contributing to physical activity participation (i.e., ACS), walking related physical activity (i.e., steps/day) and intensity of physical activity (i.e., IPAQ-S7). Strength of the affected dorsiflexor muscles contributed to physical activity participation and intensity. Higher intrinsic and introjected regulation, measured using the Behavioural Regulation in Exercise Questionnaire-2, contributed to higher physical activity participation. Anxiety (Hospital Anxiety and Depression Scale) demonstrated a non-linear relationship with physical activity participation. A steep rise in physical activity participation with increasing anxiety levels was observed up to an approximate threshold score of 4-5, above which there was no appreciable association between anxiety and activity.
Conclusion(s): Gait speed and balance are modifiable factors that contribute to all aspects of post-stroke physical activity at discharge from inpatient rehabilitation. Anxiety demonstrated a non-linear association with participation in high demand leisure activities at discharge. The results demonstrated that better physical function contributed to future increased physical activity.
Implications: Physical activity levels can be influenced by multiple factors such as physical function and psychosocial factors. The factors that contribute to physical activity levels may also differ depending on the stage of stroke recovery. The findings of this study stress the importance of optimising gait and balance early after stroke. However, motivation and mood were also significant modifiable factors and may potentially be targeted in interventions to improve physical activity.
Keywords: Physical activity, stroke, predictor
Funding acknowledgements: Australian Government RTP Scholarship (Author ST)
NHMRC Career Development Fellowship (Author RC)
Singapore General Hospital Research Grant (Author DT)
Topic: Neurology
Ethics approval required: Yes
Institution: SingHealth Pte Ltd
Ethics committee: SingHealth Research Ethics Committee
Ethics number: 2015/2010
All authors, affiliations and abstracts have been published as submitted.
