Physical activity behaviour after stroke : a COM-B informed analysis of activity accumulation and post-stroke determinants

Physical activity is a key contributor to recovery after stroke, underpinning functional improvement, enhancing health-related quality of life, and reducing the risk of recurrent events. Yet despite robust evidence and clear national and international guidelines, people with stroke remain insufficiently active and highly sedentary. The impact of stroke is multi-faceted, affecting physical, cognitive and emotional domains that shape engagement in physical activity; however, these factors are rarely integrated into intervention design. In addition, current approaches to measuring and designing physical activity interventions focus on total volume and intensity, which may not adequately capture the detail of how activity is accumulated in daily life. Developing a more comprehensive understanding of activity patterns, and the factors influencing engagement, is therefore key to inform effective and individualised strategies to support participation in this population. This thesis aimed to advance understanding of physical activity behaviour in people with stroke guided by the COM-B model of behaviour change. It established a novel approach to activity analysis that captures not only volume and intensity, but also how frequently activity is undertaken and how long it is sustained, providing a more detailed and clinically meaningful description of activity patterns, generating insights into physical activity behaviour that are not captured by conventional measures. The first study, an overview of reviews, synthesised evidence from systematic reviews of interventions to improve physical activity after stroke and categorised intervention components using the COM-B model. Synthesis of 20 reviews demonstrated substantial heterogeneity in intervention design; however, it revealed that interventions incorporating at least two components of the COM-B - capability, opportunity and/ or motivation - were more likely to be effective. These findings support the importance of multi-component, theory-informed approaches to behaviour change. The second and third studies, used physical activity data from a prospective cohort study, OX-CHRONIC (n=105), to describe physical activity across multiple dimensions: amount, intensity, frequency and duration; and examine associations with post-stroke factors. Cross-sectional analysis (n=83) demonstrated low physical activity (114 ± 82 minutes), and substantial within- and between- subject variability in all activity dimensions. Activity was predominantly accumulated in frequent (n=35 ± 19.5) short bouts of less than five minutes (2.9 ± 1.3 minutes), although mean maximum bout length was substantially longer (27.7 ± 24.5 minutes). Distinct associations were observed between post-stroke factors and activity dimensions; whilst bout frequency decreased in relation to greater fatigue (B=-0.28, p=0.045) and there were significantly fewer bouts in those with clinical symptoms of apathy (t(78)=2.07, p=.042); bout duration was shorter in those with reduced cognition (B=0.10, p=.002) and significantly shorter in those with clinical signs of depression (t(77)=2.77, p=.01). Longitudinal analysis (n=62) demonstrated an increase in overall physical activity over one year (27.42 ±74.19 minutes), driven primarily by changes in bout frequency (n=5.99 ± 13.88), whilst bout duration remained unchanged. Associations between activity dimensions and post-stroke factors were largely stable; however worsening cognition (B= -.04, p = .03) and increasing apathy (B= 0.07, p=.03) were associated with shorter bout duration over time. This thesis advances understanding of physical activity after stroke by demonstrating the importance of how activity is accumulated, rather than focusing solely on total volume. The overall increase in bout frequency over time, and stability of bout duration suggests frequency as the more modifiable dimension to target sustainable change in activity behaviour; however, difference in daily mean and maximum bout duration indicates that individuals may have capacity to target bout length. Differential and stable associations between post-stroke factors and specific activity dimensions highlight opportunities for more targeted and individualised intervention strategies based on post-stroke presentation and individual activity behaviour. Collectively, these findings support a shift in how physical activity is conceptualised, measured, and supported in people with stroke, with implications for future research and clinical practice in development of theory-informed, personalised approaches to behaviour change.

Name dc:type.qualificationname

Ph.D

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

Oxford Brookes University

Author dc:creator

• Paterson, Sarah

Advisors dc:contributor.advisor

• Collett, Johnny
• Dawes, Helen

DOI dc:identifier.doi

https://doi.org/10.24384/r7tt-9130

OAI record oai:identifier

https://radar.brookes.ac.uk/radar/oai?verb=GetRecord&metadataPrefix=uketd_dc&identifier=tle%3A0d39ef4b-d1e9-4ebf-920d-1fe241896d7b%3Ad6bd9758-527a-46cd-bfe2-c433766e8fca%3A1