A Systematic Scoping Review of Poststroke Engagement in Physical Activity

Abstract

Objective

To synthesize knowledge about physical activity engagement in community living stroke survivors.

Data Sources

Six electronic databases (Scopus, PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, Cumulative Index to Nursing and Allied Health Literature, PsycINFO/PsycARTICLES, and Web of Science) were searched between April 2023 and March 2024. Search parameters were set from database inception to 2024.

Study Selection

Studies were included if they were randomized controlled trials, (quasi)experimental, cross-sectional, or cohort/longitudinal; published in English; focused on measuring or promoting physical activity during rehabilitation or the postrehabilitation phase; and participants were at least 18 years old, living in the community, and had a clinical diagnosis of stroke.

Data Extraction

Data extracted included general study information; study design; participant characteristics; type of physical activity; program characteristics; type of support provided, type of physical activity engagement measure; and engagement outcomes and rates.

Data Synthesis

Twenty-one studies were included in this review; 6 promoted or measured physical activity as part of formal rehabilitation programs and 15 measured physical activity postrehabilitation. More than half of the studies reported low physical activity engagement. Eleven studies used objective measures of physical activity, whereas 8 used subjective measures and 2 used both. Only the 6 formal rehabilitation programs provided additional support for physical activity engagement, but the support was generally not underpinned by motivational theory or evidence. Walking and walk-related tasks were the most promoted type of physical activity poststroke.

Conclusions

We recommend that future interventions provide additional support for engagement that is underpinned by contemporary motivation theory. The type of support offered should differ depending on whether physical activity is part of a formal rehabilitation program or simply encouraged during postrehabilitation.

Stroke is the second leading cause of death and disability worldwide.¹ Although stroke mortality is declining, the number of people affected by stroke and stroke reoccurrence is growing,² leading to more people with disability and an escalating burden on rehabilitation services. As such, supporting effective rehabilitation and further recovery postrehabilitation after stroke constitutes a major societal and public health challenge,³ especially given the considerable comorbidities (eg, hypertension, diabetes)⁴ and socioeconomic consequences.⁵

Physical activity is an important part of stroke rehabilitation and ongoing recovery and refers to any movement that leads to energy expenditure,⁶ including recreational, leisure, social, and domestic activities.⁷ However, this definition is not yet established in research in stroke contexts.⁸ Physical activity is often used interchangeably with exercise, which is a subset of physical activity that is planned, structured, and repetitive and focuses on the improvement or maintenance of physical fitness.⁶ Approaches to enhancing structured, supervised exercise engagement are likely to be very different than unsupervised and unstructured physical activity. For this reason, the present review does not consider structured exercise.

Physical activity is important for stroke survivors because it can reduce cardiovascular risk factors and improve cognitive function, functional recovery, and quality of life.^9,10 It also helps prevent secondary strokes and developing other comorbid conditions such as depression, diabetes, and cardiac events.^9,11 Therefore, it is unsurprising that the National Clinical Guideline for Stroke¹² recommends 30 minutes of daily physical activity. Yet, stroke survivors typically only reach physical activity levels well below the recommended guidelines,¹¹ especially within community settings.¹³ The reasons for physical inactivity are complex and often predispose stroke survivors to a sedentary lifestyle.¹⁴ Therefore, a greater understanding of engagement of community living stroke survivors to physical activity is needed to promote improvement in physical activity behavior, rehabilitation success, and patient recovery.^15,16

Previous reviews on this topic have focused on the effects of interventions on physical activity levels and adherence,^17,18 and a recent review reported adherence rates to home programs and identified common strategies to support adherence.¹⁹ However, these reviews did not distinguish between structured exercise and physical activity. There are 2 reviews^8,20 that separated physical activity and exercise studies within stroke research, but they did not focus on stroke survivors’ engagement to physical activity but rather understanding the content and potential of the interventions. Additionally, previous reviews have excluded objective measures of physical activity (eg, accelerometer)²¹ or certain study designs.²² Other reviews have focused on physical activity adherence factors²³ or measurements.²² This is the first systematic scoping review, as far as we know that aimed to synthesize and summarize what is currently known about physical activity engagement in community living stroke survivors by solely focusing on unstructured physical activity. The research questions steering the current review were:

• 1.How is poststroke physical activity engagement measured?
• 2.How is poststroke physical activity engagement supported?
• 3.What are stroke survivors’ engagement rates in physical activity?
• 4.What type of poststroke physical activity is measured or promoted?

The outcomes of the present review can provide a more complete understanding of stroke survivors’ physical activity and inform the design and delivery of interventions to improve stroke survivors’ engagement in physical activity in rehabilitation and postrehabilitation phases. Focusing on physical activity, and not structured exercise, will offer greater specificity in any recommendations to increase adherence and engagement in future rehabilitation work. This is important because most stroke survivors are relatively inactive and adherence to physical activity within the stroke population is challenging.

Methods

This systematic scoping review was conducted using guidance on how to conduct a systematic scoping review²⁴ and the Preferred Reporting Items for Systematic reviews and Meta-Analyses protocols extension for Scoping Reviews²⁵ The review protocol can be accessed at the Open Science Framework (https://osf.io/tge3d).

Eligibility criteria

Study designs were limited to randomized controlled trials (RCTs), (quasi)experimental, cross-sectional, or cohort/longitudinal. The current review excluded articles if they were review papers, case reports, commentaries or other opinion pieces, book chapters, qualitative studies, or protocols. The inclusion criteria for studies were as follows: (1) published in English; (2) focus on measuring or promoting physical activity with community living (ie, discharged from the hospital and/or stroke ward) stroke survivors; (3) include participants who were 18 years old or older; and (4) were in any poststroke phase and had a clinical diagnosis of stroke. Studies with a mixed population (ie, stroke survivors and those with other neurologic conditions) were included if the subanalysis focused on stroke. Articles were excluded if the physical activity was, in fact, exercise (ie, planned, structured, and repetitive).⁶

Search strategy

Searches were carried out between April 2023 and March 2024 in 6 electronic databases (Scopus, PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, Cumulative Index to Nursing and Allied Health Literature, PsycINFO/PsycARTICLES, and Web of Science). The date range parameters for publication were the maximum permitted by each database (ie, we searched from database inception to 2024). The reference lists of articles that met the inclusion criteria were manually searched to identify additional articles. Forward citation checks of included studies were also performed for the same reason. All searches were re-run prior to the final analyses. The search strategy (table 1) consisted of 3 concepts: stroke, rehabilitation, and physical activity. For each concept, multiple keywords were developed. During the search, the keywords for each concept were combined with the Boolean operator “OR” and each concept was combined with the other concepts using the Boolean operator “AND.” The search results were then imported onto Covidence,²⁶ an online systematic review management tool widely used within health care research.²⁷ Once duplicates were removed, titles and/or abstracts of articles were screened by the first author to identify studies that potentially meet the inclusion criteria. A university library or the authors of published abstracts were contacted to request the full-text versions of studies where relevant. The full-text of potentially eligible studies was retrieved and independently assessed for eligibility by the reviewer. Any concerns over eligibility were resolved through discussion with the last author, who made the final decision on inclusion. The PRISMA flow chart summarizing the outcomes of this process can be seen in figure 1.

Table 1.

Search strategy (the following concepts were used in combination [“AND”]).

Search Strategy
Key concept 1: Stroke
“stroke” OR “cerebrovascular accident” OR “CVA” OR “cerebrovascular event” OR “cerebrovascular disorder” OR “cerebrovascular insult” OR “cerebrovascular trauma” OR “apoplexy” OR “severe stroke” OR “post-stroke” OR “stroke disabilit*” OR “intracerebral haemorrhage” OR “ICH” OR “ischemic stroke” OR “cerebral infarction” OR “brain infarction” OR “intracranial haemorrhage” OR “intracerebral hemorrhage” OR “intracranial hemorrhage” Key concept 2: Rehabilitation “rehab*” OR “recovery” OR “aftercare” OR “therap*” Key concept 3: Physical activity “physical activity” OR “exercise” OR “movement” OR “physical therapy” OR “physical conditioning” OR “daily activit*” OR “leisure activit*”

Fig 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analysis Scoping Review (PRISMA-ScR) flow diagram.

Data extraction

Data extracted (table 2) included general study information (authors, year of publication); study design; participant characteristics (sample size, age, sex, type of stroke); type of physical activity; program characteristics (eg, duration, frequency); type of physical activity engagement measure, type of support provided (if applicable); and engagement outcomes and rates (if applicable).

Table 2.

Data extraction and study characteristics.

Author, Year of Publication	Study Design	Participant Characteristics	Type of Physical Activity	Program Characteristics	Physical Activity Engagement Measure	Physical Activity Engagement Support	Physical Activity Engagement Trend and Rates
Cha et al37 (2022)	Pre–post clinical trial	Sample size: 139 Age: 60-65 Sex: 96 males and 43 females Type of stroke: 123 ischemic, 16 hemorrhagic	Vigorous or moderate-intensity activity and walking.	Monthly telephone assessment using the IPAQ for 3 mo.	IPAQ	Educational program	Significantly higher proportion of participants in the intervention group (71.6%) were physically active compared to those in the control group (34.7%). There was a 60% increase in participants who became physically active.
Fini et al11 (2021)	Single-center, prospective longitudinal observation study	Sample size: 79 Age: mean ∼65 Sex: 54 males and 25 females Type of stroke: 54 infarct/ischemic, 25 hemorrhagic	Physical activity operationalized as total steps and duration of moderate-to-vigorous physical activity	Accelerometer worn for 7 d at each of the 4 assessment times (baseline, at 6, 12 and 24 mo).	SWAB	No support	Daily step count was low at all time points. Most of the sample achieved physical activity recommended guidelines but this declined from 72% to 65% over a 2y period.
de Graaf et al55 (2023)	Prospective cohort study	Sample size: 200 Age: mean ∼67.8 ± 11.2 Sex: 128 males and 72 females Type of stroke: 183 ischemic stroke, 17 not specified	Physical activity operationalized as total walking bouts and duration of moderate-to-vigorous physical activity	Accelerometer (Activ8, a 3-axial accelerometer) worn for 2 consecutive wk after 1 y after hospital discharge	Activ8, a 3-axial accelerometer	No support	Physical activity improved in the first 6 mo after hospital discharge, but there was a decline in physical activity participation between 6 mo and 1 y
Guediri et al38 (2022)	Comparative prospective, parallel randomized controlled trial	Sample size: 47 Age: mean ∼60 Sex: 9 males and 38 females Type of stroke: 14 ischemic, 33 hemorrhagic	Physical activity operationalized as total steps	Experimental group: Educational phase, and second phase: weekly telephone calls and a home visit every 3 wk for 6 mo. Control group: follow-up at 1 and 6 mo.	Accelerometer (SenseWear armband, BodyMedia)	Educational program	Nonsignificant increase in physical activity from 4223 steps after the first month to 5104 steps at 6 mo.
Askim et al39 (2018)	Multicenter, pragmatic, single-blinded, randomized controlled trial	Sample size: 380 Age: mean ∼72 Sex: 231 males and 149 females Type of stroke: 346 infarct/ischemic, 34 hemorrhagic	Physical activity included housework, walking, and gardening.	LAST program-Intervention group aim: 30 min of physical activity daily. Plus, monthly meeting with coach for 18 mo.	Self-reported training diaries	Individualized coaching program	Increasing numbers of compliers to 20-30 min of daily physical activity during follow-ups, and participants receiving regular individualized coaching for 18-mo were more active than participants who received standard care.
Jilimin and Mazlan50 (2023)	Cross-sectional study	Sample size: 45 Age: mean ∼54.3±9.5 Sex: 27 males and 18 females Type of stroke: 34 ischemic, 11 hemorrhagic	Vigorous or moderate-intensity activity and walking.	Completed the IPAQ	IPAQ	No support	Lower physical activity levels compared to comparable data with healthy participants. Nevertheless, around 75% of participants achieved the required activity time as recommended in stroke guidelines.
Joseph et al47 (2018)	Cross-sectional study	Sample size: 45 Age: mean ∼58 Sex: 23 males and 22 females Type of stroke: not reported	Physical activity operationalized as the total number of steps and the total activity counts	Accelerometer worn for 5 consecutive days	Accelerometer (Actigraph GT3Xþ, Pensacola, FL)	No support	85% of participants did not achieve 22 min/d of physical activity, thus not meeting the recommended guidelines.
Kanai et al48 (2019)	Cross-sectional study	Sample size: 61 Age: mean ∼67 Sex: 46 males and 15 females Type of stroke: 48 ischemic, 13 hemorrhagic	Physical activity operationalized as total steps number of steps and duration of moderate-to-vigorous physical activity	Fitbit One 3-dimensional accelerometer for 7d	Fitbit One 3-dimensional accelerometer	No support	Participants achieved 13.9 min/d of physical activity, thus not achieving the recommended guidelines.
Lau et al28 (2022)	Prospective cohort study	Sample size: 40 Age: mean ∼53 Sex: 23 males and 17 females Type of stroke: 28 ischemic, 12 hemorrhagic	Measured free-living physical activities in 4 domains: instrumental activities of daily living (shopping), high-physical-demand leisure activities (gardening), low-physical-demand leisure activities (watching TV), and social activities (visiting family or friends).	Completed PIEL Survey App and wore The activPAL accelerometer for 7d	PIEL Survey App and The activPAL accelerometer	Contacted participants to encourage and support physical activity engagement	Participants reported meeting recommendations by averaging 45 min/d of physical activity.
Lin et al53 (2022)	Cross-sectional descriptive study	Sample size: 208 Age: mean ∼70.25±9.08 Sex: 125 males and 83 females Type of stroke: 168 ischemic, 33 hemorrhagic, and 7 not specified	Walking or going up and down stairs	Completed the PAEAS	PAEAS	No support	The mean adherence rate of stroke survivors to physical activity recommendations was 62%.
Lin et al56 (2023)	Cross-sectional study	Sample size: 596 Age: mean ∼65.70±10.76 Sex: 347 males and 249 females Type of stroke: 421 ischemic, 175 hemorrhagic	Walking or going up and down stairs	Completed the PAEAS	PAEAS	No support	Participants who were < 1-y poststroke reported slightly higher physical activity adherence compared to participants who were 1-10 y poststroke.
Miller et al45 (2021)	Cross-sectional study	Sample size: 249 Age: mean ∼63 Sex: 120 males and 129 females Type of stroke: not reported	Physical activity operationalized as total steps	FitBit One or FitBit Zip device worn for a minimum of 3 d	FitBit One or FitBit Zip device	No support	The participants achieved an average step count of 4353/d, thus not meeting the recommended guidelines.
Miller et al46 (2022)	Cross-sectional study	Sample size: 282 Age: mean ∼63 Sex: 146 males and 136 females Type of stroke: not reported	Physical activity operationalized as total steps	FitBit One or FitBit Zip device worn for a minimum of 3 d	FitBit One or Zip device	No support	The participants achieved an average step count of 4563/d, thus not meeting the recommended guidelines.
Rand et al54 (2009)	Cross-sectional study	Sample size: 40 Age: mean ∼66-67 Sex: 13 males and 27 females Type of stroke: 37 ischemic, 3 hemorrhagic	Physical activity operationalized as total activity counts and associated energy expenditure.	Accelerometer worn for 3 consecutive days.	Accelerometer (Actical Mini-Mitter Co)	No support	58% of participants did not meet recommended levels of physical activity based on activity counts and associated energy expenditure.
Robinson et al29 (2011)	Cross-sectional study	Sample size: 50 Age: mean ∼65 Sex: 22 males and 23 females Type of stroke: left side effected (25), right side effected (23), bilateral (2)	Walking-related activities (eg, walk to the restaurant)	Completed TAL and wore pedometer for 7 d	TAL and pedometer (VKRFitness Twin Step)	No support	The participants achieved an average step count of 2540/d, thus not meeting the recommended guidelines.
Robinson et al57 (2013)	Cross-sectional study	Sample size: 30 Age: mean ∼68-69 Sex: 16 males and 14 females Type of stroke: left side effected (40%), right side effected (∼53%), bilateral (∼7%)	Walking-related activities (eg, walk to the restaurant)	TAL was repeated once a week for 3 subsequent weeks.	TAL	No support	Lower levels of physical activity in stroke survivors compared to comparable data with healthy participants.
Roos et al51 (2012)	Cross-sectional study	Sample size: 54 Age: mean ∼64 Sex: not reported Type of stroke: not reported	Physical activity operationalized as total steps and walking bouts.	SAM worn for 3 consecutive days	SAM	No support	Lower levels of physical activity in stroke survivors compared to comparable data with healthy participants.
Walter et al40 (2015)	Multiple-participant, 2-phase single-subject randomized controlled trial	Sample size: 5 Age: mean∼ 81 Sex: 2 males and 3 females Type of stroke: 2 left side effected, 3 right side effected	Walking indoors or outdoors, household activities.	Three home visits in 3 mo by the research physiotherapist.	Physical activity engagement diary	Motivational support (motivational interviewing) and help sheets.	All participants maintained engagement in physical activity for the duration of study.
Wondergem et al49 (2019)	Cross-sectional study	Sample size: 190 Age: mean ∼68 Sex: 123 males and 67 females Type of stroke: 174 infarct/ischemic, 16 not specified	Physical activity operationalized as total duration of moderate-to-vigorous physical activity.	Accelerometer worn for at least 7 d of at least 10 h/d for valid data.	Activ8, a 3-axial accelerometer	No support	Approximately 77% of participants reported less than 30 min of physical activity (not meeting the recommendations), but 23% of participants reported around 40 min of physical activity.
Wu et al42 (2023)	Quasi-experimental study	Sample size: 26 Age: mean range ∼33-67 Sex: 12 males and 14 females Type of stroke: not reported	Physical activity operationalized as total steps	Uploaded data of walking to the health management platform for 90 d. The experimental group received affirmative encouragements.	WowGoHealth App and wearable device (Gordon smart Bluetooth heart rate bracelet GSH417-M2)	Affirmative encouragements	A significant increase in the number of daily steps of participants receiving a 12-wk health management platform intervention from 4201 to 6813 steps.
Yi et al52 (2015)	Cross-sectional study	Sample size: 60 Age: mean ∼68 Sex: 37 males and 23 females Type of stroke: 43 ischemic, 17 hemorrhagic	Leisure activities (eg, gardening, household activities, social and outdoor activities)	Completed FSE questionnaire.	FSE	No support	Lower physical activity levels compared to physical activity prior to suffering a stroke.

Abbreviations: FSE, Functional Status Examination; IPAQ, International Physical Activity Questionnaire; PAEAS, Physical Activity and Exercise Adherence Scale; PIEL, Everyday Life; SAM, StepWatch Activity Monitor; SWAB, SenseWear MF Armband; TAL, Trip Activity Log.

Data synthesis

Data were analyzed using a narrative synthesis of the key findings from the included studies. After a description of the study and sample characteristics, the synthesis was structured around the 4 research questions of the review (ie, engagement measurements, the type of support, physical activity engagement, and the type of physical activity).

Results

Search results

The initial search yielded 17,308 articles that were imported into Covidence,²⁶ and after removing the 9230 duplicates, 8078 titles and abstracts were reviewed for inclusion. Among those, 188 articles progressed to the full-text review, and 17 studies met the inclusion criteria. During the full-text review, several studies were excluded because they contained secondary analysis following on from original studies and/or had study outcomes that were irrelevant to the present review. Checking reference lists of the included articles and relevant systematic and scoping reviews resulted in 4 additional studies that met the inclusion criteria. In total, 21 studies fulfilled the predefined inclusion criteria.

Study characteristics

The characteristics of the included studies are summarized in table 2. The reports were published between 2009 and 2023 and conducted in 13 countries, with most conducted in the United States and the Netherlands. Thirteen studies were cross-sectional, 4 were RCTs, 3 were cohort/longitudinal, and 1 employed a quasi-experimental design. The length of the longitudinal/cohort and RCT studies ranged from 7 days to 24 months.

Participant characteristics

The participants were mostly recruited from hospital databases, with fewer participants recruited through rehabilitation services (eg, stroke clubs) and health centers. Across all studies, 2826 adult community living stroke survivors participated (mean= ∼135 participants per study), of which ∼57% were men (1 study did not report sex). The age of participants ranged from 25 to 90 years, with most participants 60-70 years. Participants had experienced their stroke between 0 and ∼40 years prior to the study start. The target population were individuals with ischemic stroke (or cerebral infarction; n=1673), hemorrhagic stroke (n=372); and non-specified types of strokes (n=781). Eight studies did not report information on the type of stroke. Only 2 studies reported including participants with aphasia and, in almost all the studies, participants ranged from showing no symptoms to moderate disability. Table 2 reports the participant characteristics from each study.

Measures of physical activity engagement

Thirteen studies used objective measures of physical activity (eg, accelerometers); 2 of these studies^28,29 also employed subjective measures of physical activity. Where objective measures were applied, physical activity was commonly operationalized as total step counts. Subjective measures of physical activity included physical activity logs (eg, Trip Activity Log),³⁰ questionnaires, or diaries (sometimes referred to as “training diaries”) to assess physical activity engagement. Self-report measures included the Participation in Everyday Life survey³¹; Physical Activity and Exercise Adherence Scale³²; the International Physical Activity Questionnaire^33,34; the Functional Status Examination³⁵; and the Saltin-Grimby Physical Activity Level Scale.³⁶

Physical activity engagement support

Within the review, 6 studies focused on physical activity during rehabilitation programs, whereas the other 15 examined physical activity during stroke survivors’ postrehabilitation phase. The 6 formal programs provided additional support to their participants to engage in physical activity. Two of these studies^37,38 implemented an educational phase/program that included consultation and explanation of the general benefits of physical activity and rehabilitation, recommendations on activities, and encouragement to perform physical activity as part of rehabilitation. One study provided an individualized coaching program,³⁹ and 1 included motivational support⁴⁰ using face-to-face motivational interviewing⁴¹ during 2 visits and face-to-face help sheets during 1 visit that helped with participants’ action planning. Wu et al⁴² provided affirmative encouragements on a health management platform to motivate participants in the experimental group, and Lau et al²⁸ contacted participants via phone to check-in and encourage physical activity engagement. Generally, the additional support in these studies was not underpinned by contemporary motivation theory or evidence.

Stroke survivors’ engagement trend and rates to physical activity

Eleven cross-sectional studies reported generally lower-level physical activity engagement than recommended guidelines (ie, 150 min/wk in moderate-to-vigorous physical activity⁴³; between 6000 and 8500 steps/d for populations with disability⁴⁴). Where step counts were measured, the typically reported average daily step count was around 4000 steps (eg, Miller et al^45,46), whereas studies that measured minutes of physical activity showed varying results (ie, Joseph et al,⁴⁷ Kanai et al,⁴⁸ and Wondergem et al⁴⁹).

Other studies reported lower physical activity in stroke survivors compared to comparable data with healthy participants (eg, Jilimin and Mazlan,⁵⁰ Roos et al⁵¹) or physical activity prior to suffering a stroke (eg, Yi et al⁵²). Lin et al⁵³ used the Physical Activity and Exercise Adherence Scale and reported a 62% mean adherence rate of stroke survivors to physical activity recommendations. Rand et al⁵⁴ noted that 58% of their participants did not meet recommended levels of physical activity based on activity counts and associated energy expenditure. Overall, studies using objective measures showed relatively lower rates of physical activity compared to self-reported physical activity.

Two longitudinal studies reported declining levels of physical activity over time, albeit in different ways. Fini et al¹¹ reported the percentage of participants who met recommended guidelines (ie, min of physical activity/d) declined from 72% to 65% over a 2-year period. de Graaf et al⁵⁵ noted that physical activity improved in the first 6 months after hospital discharge, but there was a decline in physical activity participation between 6 months and 1 year. A cross-sectional comparison study⁵⁶ found that participants who were less than 1 year poststroke reported slightly higher physical activity engagement compared to participants who were 1-10 years poststroke.

Three RCTs, 1 cohort, and 1 quasi-experimental study reported increased or relatively high physical activity behavior (ie, achieving physical activity recommendations) over time. These studies mostly used subjective measures, and explicitly promoted and supported physical activity as part of their rehabilitation program. For example, the Life After STroke³⁹ program reported increasing numbers of compliers to 20-30 minutes of daily physical activity during follow-ups, and that stroke survivors who received regular individualized coaching for 18 months were more active than participants who received standard care. Moreover, a study⁴² using objective measures displayed a significant increase in the number of daily steps of participants receiving a 12-week health management platform intervention from 4201 to 6813 steps. In one cohort study,²⁸ participants reported meeting recommendations by averaging 45 min/d of physical activity. Increases in physical activity were also more common in studies that provided motivational support to participants (eg, educational programs, motivational training, or individualized coaching). Cha et al³⁷ provided educational support to the intervention group and reported that a significantly higher proportion of these participants (71.6%) were physically active (eg, walking for at least 30min/d) compared to those in the control group (34.7%). At the end of the intervention, there was a 60% increase in participants who were categorized as physically active. Guediri et al³⁸ provided additional motivational support to participants but reported a nonsignificant increase in physical activity (from 4223 steps after the first month to 5104 steps at 6mo).

Type of physical activity being measured or promoted

Six studies focused on walking and walking-based tasks (eg, walking to a bank or restaurant), and 4 measured and/or promoted general everyday free-living physical activities that included gardening and household/housework activities, leisure activities, volunteering or work activities, and physically demanding social activities (eg, assisting others with activities of daily living). Two studies^29,57 only collected data on the number of trips and walking-related activities that happened outside of participants’ homes and property. Studies using objective measures were not able to specify the type of physical activities that participants carried out but rather operationalized physical activity as total step counts and/or duration of physical activity.

Discussion

The present review aimed to establish how physical activity engagement is measured; how physical activity engagement is supported; the typical levels of stroke survivors’ physical activity engagement; and what physical activity is being measured or promoted. The insights from the current review can offer potential directions and strategies for rehabilitation interventions to improve long-term physical activity engagement in stroke survivors. Meeting these aims is important because physical activity levels are typically low and engagement in physical activity is challenging for stroke survivors.

Measures of physical activity engagement

More than half of the studies used objective measures (eg, accelerometery) to assess physical activity engagement within stroke survivors, the most common of which was step counts measured via accelerometery. The most frequently used subjective measures were physical activity questionnaires. The choice of physical activity measure has significant implications. Decreasing trends and relatively lower rates of engagement were typically observed in studies that used objective measures of engagement in physical activity which is consistent with trends outside the rehabilitation context.⁵⁸ On the other hand, studies that promoted physical activity as part of a formal rehabilitation program and used subjective measures typically showed improved physical activity behavior or relatively higher engagement in physical activity. Because both subjective and objective measurements of physical activity engagement are valuable and offer different information, it would be advantageous if physical activity engagement was measured using both methods in future research to fully capture and understand physical activity of stroke survivors. However, it is appreciated that this requires extra resources and adds burden to participants.

Stroke survivors’ engagement trend and rates of physical activity

The reviewed studies, particularly those that examined physical activity postrehabilitation, generally reported that stroke survivors’ physical activity levels are below guidelines. Thus, physical activity recommendations seem to be overly difficult for stroke survivors living in the community to achieve.⁴⁸ This could be because of the lack of detail around how to support the facilitation and application of the given guidelines.^59,60 Because physical activity is crucial for stroke survivors,⁹ appropriate implementation, reflection, and evaluation of these guidelines might be necessary to improve long-term engagement in physical activity.^60,61

In contrast, the findings of the longitudinal/cohort studies and RCTs typically showed an increase in physical activity engagement levels that commonly exceeded recommended levels, especially at the beginning of the programs. Most of these studies also provided additional motivational support for stroke survivors to engage in physical activity, but few examined physical activities after the support had ended. It would be interesting to examine if physical activity remained high when additional support ended because previous research has highlighted phases of rehabilitation with different levels of adherence.⁶² Each phase likely requires different motives for physical activity; therefore, future interventions should be adapted accordingly.

Physical activity engagement support

Some studies focused on physical activity as part of rehabilitation programs, whereas most of them more broadly examined physical activity postrehabilitation. This is important as support should differ depending on whether programs aim to motivate physical activity beyond the program itself (ie, postrehabilitation) or simply during the program. For example, when the aim is to encourage physical activity without the structure of a rehabilitation program, it would be beneficial if stroke survivors were provided with self-management skills to effectively manage their physical activity and condition.⁶³ Self-management for stroke survivors could help to create habits by introducing environmental regularity (ie, physical activity takes place at a regular location)⁶⁴ or using monitors and reminders (eg, digital watches and phone apps that encourage physical activity behaviors).⁶⁵ In contrast, formal rehabilitation programs could increase health care professional’s knowledge and skills by providing developmental training on how to best apply motivational support and deliver physical activity rehabilitation programs.⁶⁶ This support could focus on facilitating physical activity engagement beyond the rehabilitation phase to continue the recovery from stroke. In addition, physical activity levels in the postrehabilitation stage should not be compared to those in formal rehabilitation programs because the latter has more provision and support.

Previous stroke research has underscored the importance of extra support to enhance physical activity.^67,68 Nonetheless, the additional support identified in interventions included in this review tended not to be underpinned by contemporary motivation theory or evidence. Health care professionals consider motivation to engage in rehabilitation to be a key factor of rehabilitation outcome⁶⁹ because motivated stroke survivors tend to be more engaged and involved in rehabilitation interventions.⁷⁰ Therefore, physical activity support should be underpinned by up-to-date understanding of motivation to increase the effect and effectiveness of physical activity interventions⁶⁵ and assess the relationship between motivation and rehabilitation adherence or engagement.

Type of physical activity being measured or promoted

Walking and walk-related tasks were the most frequently reported and promoted types of physical activity; however, the studies that differentiated between modes of physical activity were relatively few. This emphasis on walking is likely because it is considered a more acceptable and suitable alternative for stroke survivors to engage in compared to more demanding and challenging physical activity or exercise.⁷¹ Walking can be done continuously in community settings, whereas stroke survivors might find it hard to implement high-intensity exercise into their daily lives and tasks.⁷² Walking is cost-effective, generally, no additional equipment is needed, and depending on stroke severity, stroke survivors can perform it on their own with or without walking aids.⁷³ Compared to walking, less is known about other free-living activities (eg, gardening or household/housework activities); therefore, future research could explore different types of free-living physical activities that stroke survivors engage in to help determine their preferences and possible drivers for physical activity.⁷⁴

Study limitations and future research directions

As in all reviews, it is possible that relevant articles were missed because the gray literature was not searched and we only included published full-text articles in English which may have led to an incomplete synthesis of data in this area. Because the aim of this review was to look at a wide breadth of literature, studies were included regardless of their quality. Although qualitative research can provide valuable contextual information⁷⁵ and broader understanding of how physical activity engagement and interventions are experienced,⁷⁶ we included only quantitative studies because we wanted to provide an overview of quantitative engagement measures and rates of physical activity.

Future research can also target different participant characteristics that were typically found within the current review, with most participants being relatively physically able (eg, mild motor impairment). In the future, researchers need to consider less able and nonambulant stroke survivors and examine alternative physical activities, such as dance techniques that can be performed sitting⁷⁷ or assisted walking training.⁷⁸ Likewise, more studies should focus on stroke survivors with aphasia because it affects approximately 35% of people with stroke.⁷⁹ Most stroke survivors across all included studies were between 60 and 70 years of age. Physical activity interventions and programs for older stroke survivors might be designed differently as opposed to younger cohorts.⁸⁰ Therefore, future programs should focus on including stroke survivors under the age of 60 years.⁸¹

Conclusions

The current review presents a comprehensive summary and understanding of community living stroke survivors’ engagement in physical activity. A specific definition of physical activity was applied because of the absence of clear and established definitions for physical activity in stroke research.⁸ This precision is important because physical activity has very different motivational and adherence challenges compared to structured exercise and sport. Some studies focused on physical activity during rehabilitation, while most measured or promoted physical activity postrehabilitation. This difference is important because the most optimal methods to support physical activity in these contexts will likely differ. Typically, lower physical activity engagement rates were observed in studies using objective measures of physical activity compared to those using subjective measures. Additional support to help adhere to physical activity was generally effective in the short term, but future interventions that provide additional support should be underpinned by contemporary motivation theory.

Disclosure

The investigators have no financial or nonfinancial disclosures to make in relation to this project.