https://orcid.org/0000-0002-9809-9859
Figures
Abstract
Background
Although the benefits of regular physical activity (PA) after stroke are well established, many stroke survivors do not achieve recommended PA levels. To date, studies exploring determinants to PA have not used a behaviour change theory and focused on stroke survivors with physical disabilities. As a precursor to an intervention development study, we aimed to use the Theoretical Domains Framework (TDF) to identify factors influencing PA in stroke survivors with physical disabilities in Singapore.
Methods
Between November 2021 and January 2022, we conducted interviews with 19 community-dwelling stroke survivors with a weak arm and/or leg. An interview guide based on the TDF was developed. We analysed the data deductively by coding interview transcripts into the theoretical domains of the TDF, and then inductively by generating themes and belief statements. To identify relevant TDF domains, we prioritised the domains based on the frequencies of the belief statements, presence of conflicting belief statements and evidence of strong belief statements.
Results
Eight of the 14 TDF domains were relevant, and included environmental context and resources, knowledge, social influences, emotion, reinforcement, behavioural regulation, skills and beliefs about capabilities. The lack of access, suitable equipment and skilled help often limited PA participation at public fitness spaces such as parks, gyms and swimming pools (environmental context and resources). While a few stroke survivors expressed that they had the skills to engage in regular PA, most expressed not knowing how much and how hard to work, which exercises to do, which equipment to use and how to adapt exercises and equipment (knowledge and skills). This often left them feeling afraid to try new activities or venture out to new places for fear of the unknown or adverse events (e.g., falls) (emotion). For some, doing the activities in a group encourage them to get out and engage in PA (social influences).
Conclusions
In stroke survivors with physical disabilities, environmental context and resources had a significant influence on PA participation, and this often had a spill over effect into other domains. Our results inform a complex behaviour change intervention to improve PA after stroke, and has implications for intervention design for people with physical disabilities.
Citation: Kwah LK, Doshi K, De Silva DA, Ng WM, Thilarajah S (2024) What influences stroke survivors with physical disabilities to be physically active? A qualitative study informed by the Theoretical Domains Framework. PLoS ONE 19(3): e0292442. https://doi.org/10.1371/journal.pone.0292442
Editor: Nicola Diviani, Swiss Paraplegic Research, SWITZERLAND
Received: June 26, 2023; Accepted: September 20, 2023; Published: March 28, 2024
Copyright: © 2024 Kwah et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The underlying minimal data for this study can not be shared publicly because they contain potentially identifying or sensitive patient information. Data are restricted by a research collaboration agreement signed between five organizations involved in this project. Ethical approval was given by the Singapore Institute of Technology Institutional Review Board (SIT IRB). Data are available upon request from the Corresponding Author and from Mark Alexander, Secretariat (SIT IRB), via email (mark.alexander@singaporetech.edu.sg), for researchers who meet the criteria for access to confidential data.
Funding: The current study is funded by a Singapore Institute of Technology (SIT) Ignition grant. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: PA, Physical Activity; TDF, Theoretical Domains Framework; BCW, Behaviour Change Wheel; SNSA, Singapore National Stroke Association
Introduction
The benefits of regular physical activity (PA) after stroke are well known. Regular PA reduces the risk of another stroke and improves the physical, mental and overall well-being of people after stroke [1, 2]. Despite recommendations of PA levels for people with disabilities or chronic illnesses (e.g., 150–300 minutes per week of moderate intensity PA [3], 6500–8500 steps/day [4]), most stroke survivors do not meet these PA guidelines [5]. Worldwide, stroke survivors achieved an average of 5535 steps in the subacute phase and 4078 steps in the chronic phase [5]. These numbers are similar in Singapore, with stroke survivors clocking a median of 4870 steps in the subacute phase [6], highlighting the prevalent issue of physical inactivity after stroke.
Factors influencing PA after stroke such as physical ability, environmental issues (access, transport, cost), social networks and participation in meaningful activities other than “exercise” have been reported in the United States, United Kingdom, Canada and Australia [7, 8], though stroke survivors in Singapore might highlight different determinants influencing PA levels, considering the differences in culture, clinical practices and health care systems. For example, a barrier such as cost of gym membership (reported by stroke survivors in the United States [9]) might not be as frequently encountered by stroke survivors in Singapore as all Singapore citizens and permanent residents are accorded $100 credit to access public gyms and swimming pools for free [10]. Considering the unique aspects of the public health infrastructure in Singapore, it is possible that barriers to PA faced by stroke survivors in Singapore might differ from others in developed countries [7, 8]. A clear picture of the barriers and facilitators influencing PA after stroke in Singapore is needed to design interventions that have a high chance of working and can be adopted into practice successfully [11].
To design an intervention that has a high chance of working and adoption into practice, it is necessary to have a clear picture of the determinants influencing the targeted behaviour [11]. While studies exploring barriers and facilitators to PA after stroke have been published [7, 8, 12], few studies have used behaviour change theories or frameworks to inform their results [12, 13] and focused on stroke survivors with physical disabilities. This group of stroke survivors are likely to be the least active, considering that stroke survivors consistently walked less steps per day compared to other chronic conditions (e.g., Chronic Obstructive Pulmonary Diseases, diabetes) and adults with physical disabilities/mobility limitations were the least physically active compared to adults with hearing, visual or cognitive impairments and adults with no disabilities [14].
In recent years, there has been an increasing demand to develop complex interventions to improve health [15, 16]. As health challenges become more complex, it is evident that health interventions targeted at changing human behaviour need to be multi-component, that is, involve a multitude of actions delivered by multiple individuals at multiple levels (e.g., policy makers at health systems level, health professionals at hospital level, patients at individual level) [16]. Although no single approach has been shown to be more effective than others, a framework/systematic approach for intervention development is often recommended to reduce research waste and increase the chances of producing a successful intervention [15]. In line with recommendations from the UK Medical Research Council to draw on theory when designing interventions [17, 18], we based our interview questions on the Theoretical Domains Framework (TDF) in order to map our findings back to the Behaviour Change Wheel (BCW) to consider the full range of intervention functions and select the most appropriate ones [11, 19]. Unlike other theoretical models of behaviour (e.g., the Theory of Planned Behaviour and the Health Belief Model) which might only explain some influences on behaviour [11], the TDF covers cognitive, affective, social and environmental influences on behaviour through synthesising 33 theories of behaviour and behaviour change categorised into 14 domains [19]. The BCW was also chosen due to its consideration of nine intervention functions and seven policy categories as elements to consider in the development of a complex intervention [11]. Using the TDF and BCW will therefore ensure a comprehensive and systematic process in identifying intervention functions that might be the most promising in encouraging uptake of PA by stroke survivors in Singapore. The process of identifying barriers and facilitators to PA prior to designing interventions will promote the design of interventions that are tailored to the specific context in which the behaviour occurs, and thereby increase the likelihood of behaviour change.
In this study, we aimed to use the TDF to explore barriers and facilitators influencing PA in stroke survivors with physical disabilities in Singapore.
Methods
Study design
We hereby report our study methods in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist [20]. We conducted a semi-structured interview study using the TDF as a guide to develop the interview questions [19]. The behaviour of interest was in staying or being more physically active after stroke, in line with PA recommendations for people with disabilities or chronic illnesses [3]. This study is part of a larger body of research work aimed at developing a complex intervention to improve PA after stroke in Singapore. The study has been granted ethical approval by the Singapore Institute of Technology Institutional Review Board (project number: 2021104).
Participant selection
Participants learnt about the study via a recruitment advertisement distributed via web-based methods (e.g., mailing list of stroke support organisations, social media, Whatsapp group chats), word-of-mouth and/or snowball sampling. We included participants who were aged 21 years or over, English-speaking, living in the community, had an affected arm and/or leg, and were interested in being more active. In accordance with the TDF guide, we estimated a minimum sample size of 10 participants but continued to recruit more in groups of three till data saturation (i.e., no new information on barriers and facilitators to PA) had been achieved [19]. Participants received a $20 voucher to reimburse them for their time and effort upon completing the interviews.
Setting
Interviews with stroke survivors were conducted face-to-face in their homes, or virtually (via Zoom) depending on the preferences of stroke survivors in light of COVID-19 restrictions. In some instances, family members and/or caregivers were around during the interviews but they did not participate in the interviews, and their presence did not influence the sharing of the stroke survivors. Interviews commenced in November 2021 and were completed in January 2022.
Data collection
Prior to the interview, stroke survivors completed a short face-to-face survey where baseline characteristics were obtained. Data included age, gender, ethnicity, employment status, years post-stroke, type of mobility aid, level of disability and perceived minutes of moderate intensity activity engaged in the past week. The level of disability was measured using the Modified Rankin Scale (mRS) that described the amount of assistance needed by stroke survivors to conduct their daily activities. To ensure standardised scoring of the mRS, we used the same structured interview that had previously shown increased agreement between raters [21, 22]. Moderate intensity activity was described as movements, or exercises that raise one’s breathing rate, or make one huff and puff [23]. The timeframe of one week was used as per recommendations in the World Health Organization [3] The interview guide (including questions and prompts) (see S1 File) was developed by the research team with reference to the TDF and BCW publications [11, 19, 24]. We used audio recordings to collect the data during the interviews. Field notes were documented after some of the interviews. Audio recordings were subsequently transcribed by an external transcription service.
Research team and reflexivity
Interviews were conducted by the first author (KLK) and a research assistant (AK). For some interviews, undergraduate physiotherapy students were present (only one at a time) as the project was offered as a Honours Thesis project. Both KLK and AK had more than 10 years of clinical experience in the area of stroke. KLK (female, age 35–45) is an academic teaching at the local University and has a PhD in the area of stroke, while AK (female, age 30–40) is a senior physiotherapist with first class Honours from a BSc (Physiotherapy) degree. None of the participants were known to AK prior, while a few of the participants were known to KLK due to her involvement as volunteer in the Singapore National Stroke Association (SNSA). We did not think this would cause any substantial bias to the interviews as AK conducted most of the interviews, and KLK was not involved in virtual volunteering activities during the COVID-19 period which spans the duration of the study. Both KLK and AK did not have first-hand experience with disability, only second-hand experience having cared for people with stroke and other neurological conditions. Prior to the interviews, a 2-hour training session was held between KLK, AK and ST. All had access to the interview materials including the interview guide, TDF and BCW publications. ST had experience with developing interview questions, conducting interviews and analysing interview data [25]. During the training session, the team discussed each of the interview question in-depth, the use of prompts, and the relationship of the question to the TDF domains. The first batch of three interviews were reviewed by KLK, AK and ST to ensure questions were asked in an open-ended way to gather relevant responses to each of the TDF domains.
Analysis and findings
After transcriptions of interviews, analysis of transcripts took a further 6 months and followed a two-stage process, using both deductive and inductive methods [19]. First, KLK and ST used a deductive approach to map four interview transcripts to the 14 domains of the TDF. Both KLK and ST open coded four transcripts (20% of all transcripts) line by line and met frequently to develop and revise the coding framework. Where there was discrepancy (e.g., whether a quote belongs in a TDF domain), both researchers discussed to reach consensus. In some cases, the participant’s quote was assigned to more than one domain [19]. If consensus was not reached, a third researcher (KD) was consulted. Results of these discussions were documented in a codebook to improve reliability of coding. We repeated the process till the Cohen’s Kappa (κ) between coders reached substantial agreement (more than 60%) [19]. KLK then coded the remaining transcripts using the codebook as reference. Second, KLK and ST used an inductive approach to generate themes and belief statements (i.e., a statement that provides detail about the perceived role of the domain in influencing the behaviour [26]) of the interview data within each domain of the TDF. To identify relevant TDF domains (i.e., domains that are likely to improve PA levels if targeted in interventions), we used the recommended three criteria: a) high frequency of the belief statements in domains, b) presence of conflicting belief statements, and c) evidence of strong beliefs that may affect the behaviour [19]. We regarded frequency as the number of different participants (out of 19) who expressed the belief statement rather than the number of times it was expressed [27, 28]. For the domains that did not meet all three prioritisation criteria, we deemed them as not relevant. We further categorised all domains (i.e., relevant and not relevant domains) according to level of importance: high importance (if all three criteria were met), moderate importance (if two out of three criteria were met), and low importance (if one or none of the three criteria were met) [27]. Relevant domains were equivalent to domains of high importance, while domains that were not relevant were equivalent to domains of moderate or low importance [27]. Microsoft Excel and Quirkos software were used to organise and analyse the data.
Results
Participants
Twenty-four participants responded to the recruitment advertisement, though only 19 completed the interviews. Reasons for non-participation included declining to sign the consent form (n = 2), no weakness in arm/leg (n = 2), and not being able to communicate in English (n = 1). The characteristics of our participants are shown in Table 1. Of the 19 participants interviewed, most were male (53%), of Chinese ethnicity (89%) with a mean age of 55 years (range 42 to 68 years). All were in the sub-acute to chronic phase post-stroke (median 9 years; IQR 4 to 13 years post-stroke). More than half of our participants (58%) had slight disability (score of 2 on Modified Rankin Scale), and more than half (74%) frequently used a mobility aid. Only a small number of participants (16%) reported having met 150 minutes of moderate intensity PA per week. Interviews lasted typically for 30 to 60 minutes. Findings of the transcripts were summarised and returned to participants for feedback in a separate co-design workshop. We recruited till data saturation was reached, and determined this by reviewing transcripts concurrently in batches of three [29]. In the last three interviews, all relevant quotes were coded to theoretical domains of TDF that already had quotes from prior interviews. As no new information was gained, we stopped at 19 participants.
Inter-rater reliability
Inter-rater reliability between two raters (KLK and ST) was assessed from four interviews, and reached agreement of 76% (95% CI 69 to 83%) (considered “substantial agreement” [30]) after four rounds of coding and discussion.
Relevant TDF domains
In total, 497 quotes from 19 interviews generated 54 belief statements, coded into 14 domains of the TDF. Only 355 quotes were regarded in the frequency count as we ensured the number reflected the number of different participants who expressed the belief statement rather than the number of times it was expressed. Eight relevant TDF domains were identified (Fig 1). Relevant domains were likely to influence behaviour and considered of high importance (Table 2). These domains included environmental context and resources, knowledge, social influences, emotion, reinforcement, behavioural regulation, skills and beliefs about capabilities (Table 3).
Environmental context and resources.
Many factors influenced stroke survivors’ participation in regular PA, of which environmental context and resources played a significant part. Most stroke survivors reported issues with access and use of public spaces, and these issues included lack of ramp access into swimming pools, wheelchair access into public spaces, inability to use accessible toilets (either due to toilets not working or being used by abled people) and the lack of accessible parking lots. Exercise equipment at the outdoor fitness zones and gyms were often deemed unsuitable, with most stroke survivors reporting difficulty in positioning their affected arm/leg onto the machines and getting onto the seats of the machines which are either too small or were not adjustable to allow use of the machines. In these instances, physical assistance (both skilled and unskilled) were deemed useful (e.g., having someone to assist them into the pool, or to guide them on the use of the machines). For the few who reported having minimal issues with access and use of space and use of equipment, they visited the swimming pool or the gyms in groups (sometimes supported by volunteers) and also learnt to rely on adaptive equipment (e.g., ankle-foot orthosis) or adapt everyday objects (e.g., using milk bottles as weights) to allow them to engage in more PA. Unsurprisingly, the COVID-19 pandemic made it even harder for stroke survivors to engage in PA due to the cancellation of group exercise classes and social outings. Although virtual activities replaced the physical activities, it was not possible to replicate fully the atmosphere and physical connections that one made with others in a face-to-face setting. There were also limited gym bookings available due to capped group sizes, and the risk of infection deterred some stroke survivors from heading out of their homes. Other PA barriers related to environmental context and resources also included the lack of finances, competing priorities/commitments and poor weather though these were much smaller obstacles compared to the access and use of public spaces, and use of equipment in public spaces.
“[…] Normal swimming pool stroke people not quite suitable, because even it’s just a narrow pool you can’t go down to the pool because the steps beside are not suitable to go down.” (P6)
“…it’s quite difficult for person with disability to join [outdoor fitness zones] […] I can’t find even one [outdoor fitness equipment] I can sit down and start doing. … The wheel… I can’t sit on the seat because it’s not very accessible to me. Because I have to get down. Then I have to shift myself in right? So when I grab hold of the wheel, there’s nothing to tie the leg… Then those who are doing the hand…sometimes too tight, and I can’t put the hand on the wheel.” (P4)
“…because of Covid, we cannot go [out] most of the things, you know, those organised outings… we cannot meet other people or friends or outings.” (P2)
Knowledge.
While some stroke survivors had knowledge of the PA guidelines in terms of duration and number of steps to achieve for health benefits (e.g., 30 minutes or 10,000 steps daily), none mentioned the intensity of PA needing to be of moderate intensity. Thus there appeared to be a greater awareness of the recommended duration or quantity of PA (i.e., how much one should be working) than the intensity of PA (i.e., how hard one should be working). Almost all stroke survivors knew about the benefits of PA and consequences of physical inactivity. The knowledge domain was closely linked to the skills domain. Stroke survivors highlighted the lack of information on appropriate PA opportunities, and not knowing what exercises and machines are appropriate for them to use as factors influencing their engagement in PA.
“How long? I’m not, I don’t really have analysed that, so. I don’t really know how long. I will say maybe at least half an hour to an hour minimum. In a day, yeah. How hard? No, I don’t go by that kind of gauge. I usually set my own target and go, so it’s like if I have to do like 10 set of the exercise per day. Then I finish all my 10 set. Yeah, then is like I finished up my 10 set, I finished up. Then if I can do more that day then I’ll do another 10 set maybe in the afternoon or in the evening.” (P9)
“I have been asking around how do I sign up for swimming lesson but nobody seems to know about it. They say that now with Covid everything is stopped. Because I used to be a swimmer and I love to swim so I’m hoping to see whether I can use back some function. So before the stroke, [I was] swimming quite a bit. And then ever since after the stroke, [I] haven’t really swum. Cannot swim at all because my left side is very weak.” (P16)
Skills.
As the quotes on exercises and machines were related to their ability to carry out a physical task that would improve with practice, most of the quotes were coded in the skills domain. Some of these skills were related to pushing into intensive exercise, progression and adaptation of exercises, problem-solving in new environments, and using machines in gyms and outdoor fitness zones. Without the knowledge and skills, stroke survivors shared that they were either still doing the same exercises/activities previously prescribed by their physiotherapists in the past, intentionally limiting activities in new environments (e.g., crossing the roads, visiting busy malls, getting on escalators), or not using any of the equipment such as treadmills, cycle machines and leg press in gyms and outdoor fitness zones.
“I have not come to the extent that I really feel the exercise is hard for me because when I do all these exercise, normally I have a qualified physiotherapist PT or whoever around, and normally they are always very protective, a little bit they said oh, you should be tired, you should stop. Even I know that I can go further, but normally to safeguard they say okay stop now, I think you’re tired.” (P6)
“[…] Like gyms wise, you can’t do that much because one hand cannot lift…I already tried […] the thing is our so called disabled arm, we can’t do much. People like us, even now you put on those like strap to the hand bicycle, no use, we’ll trip… . You know the treadmill when you run, the more the tread is there the speed is there. But if you don’t want to move, if move too fast cannot. So I hope, but it has to be much much slower or there is a limit so that even if we push stop, we won’t fall down. Because once you step out the treadmill once you cannot pace with the starting, you will trip.” (P7)
Social influences.
Stroke survivors stated both positive and negative influences of people around them in influencing their PA, though there were more positive than negative influences. Most stroke survivors shared on the physical, emotional and mental help received from family, friends/peers, colleagues and strangers which have motivated them to engage in more PA, be it walking outdoors, visiting the gym, swimming, returning to work activities, and attending social events/outings organised by SNSA. Negative influences were few, and were associated with nagging, lack of time to assist and lack of motivation to exercise on the part of the family and friends. Here, it is interesting to note that although some stroke survivors enjoyed the company of their family and friends during PA, they also expressed a preference to be alone at times so that they can focus on exercising (as they tend to talk more than exercise more in the presence of company).
“Of course in the group, besides exercise physically, you actually have a conversation. You actually in a ommunication with other group, so that helps in so called that interaction with others. There’s also kind of social interactive. That’s why it’s help not only physical, buy also mentally. We start to encourage one another. This is also more important, quite important.” (P6)
“My [family member] will stop me from [being] more active. Yeah because whenever I want to go down to … our new mall here has a [name of shop]. So I wanted to go there and see. She just say because of the pandemic, why, she will ask me why I want to go out because wherever I want to go is a bit troublesome because need to wear the bra. Yes. Then my shoes, I need to wear my ankle support to go down […] Stops me sometimes.” (P8)
Emotions.
Both positive and negative emotions were reported to influence PA. Stroke survivors reported gaining pleasure from socialising in a group, getting out and about in nature, and experiencing the endorphins from exercise. They also reported an increase in their confidence levels when they became more independent in their daily activities, or learnt to do a new activity in an adapted manner without requiring as much help as before. Where negative emotions were reported, these included the fear of falling or other adverse events, fear of the unknown and having doubts/worries on what to expect, particularly in new environments. Some stroke survivors also expressed discomfort/helplessness/frustration when they considered the complications limiting their PA (e.g., pain, spasticity and bladder issues).
“To me is like try to get back your life. Like you are still capable of doing things. By doing things and also adapting. Using ways to adapt, you find well, you can do it. then satisfaction.” (P18)
“No, it’s not very hard, but if my inversion problem was not there then probably I could have tried even swimming or Aqua therapy and all, but I can’t go down without my AFO like if I have to get into water, then I have to remove the AFO. I’m scared that I’ll slip and fall.” (P1)
Reinforcement.
On what might positively or negatively reinforce PA, the greatest determinant was the effects gained with PA. Stroke survivors reported that the benefits of PA (i.e., opportunities to improve and to be better physically, mentally and emotionally) were what kept them going. Some reported a sense of achievement at seeing themselves improve (e.g., being able to walk independently with no aid), while others shared that the consequences of physical inactivity (e.g., weight gain, weakness, falls, diabetes) were what prevented them from stopping their PA. Opportunities to connect with others in a group were also reported by stroke survivors to be positive reinforcement to engage in more PA. Vouchers and financial incentives received mixed responses. While some stroke survivors welcomed financial incentives (provided they could put on the PA tracking devices with ease or could walk up to the recommended number of steps), some were ambivalent or did not find the financial incentives appealing, particularly when compared to the opportunities to improve and meet others for social connections.
“I think it’s more intrinsic. If I’m active I’m able bodied, I can walk better, I think that’s the motivation for me. Cannot walk then I’m stuck in the house, that kind of thing, that’s very bad. That is my motivating point, to push myself such that I will walk for tomorrow.” (P15)
‘But then it [voucher/financial incentive] doesn’t appeal to me. I don’t know why, but it appealed to my sister there. I don’t know because like her age 60 over years old maybe she don’t work so she got more time doing all these collect points and walk.” (P4)
Behavioural regulation.
Most stroke survivors shared that they monitored or kept track of how much they engaged in PA using a mix of gadgets (e.g., Google Fit app on mobile phone, Garmin tracker on watch) and simple measures (e.g., duration and reps of an activity). However, in terms of how hard they worked during PA, most do not monitor this via objective measures but relied on subjective “feel” (i.e., stopping when they felt tired/pain or did not feel well). These findings were interesting to note and corroborated with earlier findings (in the Knowledge domain) regarding the lack of awareness of the recommended intensity of PA (i.e., how hard one should be working).
“I have Google Fit app on my phone, on Steps app which counts my number of steps and how much I have walked.” (P1)
"I just go with the day’s flow so if I that day don’t feel too good or feel very tired. I don’t force myself to do it…” (P9)
Beliefs about capabilities.
Most stroke survivors reported having reduced confidence in engaging in more PA in terms of adapting exercises and using equipment. These beliefs were often associated with the presence of complications or other health challenges, the need for physical assistance during activities, and environmental barriers perceived to be out of their control. For the small number who expressed confidence in engaging in more PA, they did not require physical assistance to walk and were able to engage in PA on their own.
“Yeah, vigorous exercise. I don’t think I can really ever do it, the only thing I can do is walk, walk and walk. There’s nothing much I can do.” (P12)
“I am not confident of using the treadmill. Prior to my stroke, I used to run…But now, I’m a little underconfident…Yes. I wish I could run again.” (P1)
Not relevant TDF domains
Six TDF domains were identified as not relevant (Fig 2). These domains were unlikely to influence behaviour and were considered of moderate and low importance (Table 2). These domains were beliefs about consequences, goals, intentions, optimism, social/professional role and identity, and memory, attention and decision processes (Table 4). Although domains such as goals, intentions and optimism had conflicting belief statements, they were considered of moderate importance due to low frequency and/or weak strength of belief statements. Conversely, beliefs about consequences had no conflicting belief statements but was high in frequency and strong in strength. The remaining domains of social/professional role and identity, and memory, attention and decision processes had no conflicting belief statements, and were weak in strength, and low in frequency and weak in strength respectively (Table 2).
Optimism.
For a few stroke survivors, they were optimistic in finding solutions to their problems, often in part due to past experiences and strategies in adapting their activities, acceptance of the stroke recovery process, and having options and other networks to tap on (e.g., church groups). For others, they were less optimistic as the problems centred around their post-stroke complications (e.g., spasticity and not getting the right advice) and the COVID-19 pandemic, problems that were deemed to be beyond their control.
“I think the chances [of being physically active] is probably less than 20%, unless I find some kind of alternative solution.” (P3)
Intentions.
While a few stroke survivors shared their intentions to engage in a new PA (with some clearly stating their goals and future plans), a few also reported feeling lazy and having reduced motivation to engage in more PA in light of work commitments and desire for more company during PA.
“So I’m going to find ways and where is it and where is the right time and right place to do the thing…As I said, my next target is I wanted to ride on a bike and rock climbing. So these are things that I want to do. Need to plan who will be going with you. The equipment that you need to get. Just like if you want to ride on a bike, what type of bike that is suitable for us? Which bike should I buy? What are the things that I should do first before I grab on to actual bike, all these you need to plan. Because before you want to do, you gotta have the things to do it.” (P5)
Goals.
Most stroke survivors had goals which they wanted to achieve, be it trying new things (e.g., aqua therapy, running, cycling) or returning to their pre-morbid level of function (e.g., hiking, travelling). Few stroke survivors had clear action plans related to their PA goals. Only one individual had reported having no specific goals to work towards.
“Yeah, for sure. I mean I want to be working walking better.. Of course, running it will be a optimal goal, but I want to be realistic also. So my partner she loves to go hiking. So the thing is, hiking is a challenge for me because it’s not even grounds. Up slope uneven grounds, up slope, down slope slope, so yeah. I’ve been to some of the nature reserves is actually not easy, but if since as long as there’s a real thing. railing for safety. So you can hold on to it. I think it’s okay.” (P3)
Beliefs about consequences.
All stroke survivors had accurate beliefs about the benefits of physical activity and consequences of physical inactivity. These included the effects of PA such as reducing the risk of stroke, reducing risk factors for stroke (e.g., blood pressure, cholesterol, blood sugar levels), managing complications (e.g., spasticity), reducing weight gain, reducing low mood, improving mental stimulation, preventing muscle atrophy and improving mobility.
“Like for example, the brain becomes slower and then there will be heart problem. And then hypertension which already I have. Also we’re trying to move to get it going and most important is I realise that by not moving it can have DVT which I’m having. And maybe another stroke.” (P16)
Social/professional role and identity.
Stroke survivors reported that they were personally responsible for their own health, and saw being physically active as something they should be accountable for. They did not expect others to assume responsibility for them. One stroke survivor shared that being physically active helped to reduce the stigma of being a stroke survivor as she learnt to do more things on her own.
“I think it’s my responsibility. Nobody is responsible for our improvement if we don’t want to take, take into responsibility at the end of the day is we ourselves suffer.” (P9)
Memory, attention and decision processes.
Few stroke survivors commented that they seldom forgot to engage in PA as these activities were now part of their routine everyday lives.
“But generally because you go through that motion, Hmm take that moments of needs so it’s sort like it’s around that there about. In a way, routine […]” (P18)
Discussion
Our study is the first to report determinants to PA amongst stroke survivors with physical disabilities. Using a theoretical model of behaviour, we have identified eight relevant domains that are likely to improve PA levels after stroke if targeted in future interventions. These domains include environmental context and resources, knowledge, social influences, emotion, reinforcement, behavioural regulation, skills and beliefs about capabilities.
Similar to other studies in high income countries [7, 8, 12], the broad domains of environmental context and resources, social influences and beliefs about capabilities were highlighted in our cohort as key factors influencing PA after stroke. There was congruence between our data and that of other studies, as stroke survivors often reported having stroke-related impairments which affected their confidence in engaging in PA [8, 12], and the importance of a supportive network of friends and family to provide the psychological and social support to engage in PA [7, 8, 12]. Although the domains were broadly similar to other studies, there were two notable differences within the environmental context and resources domain.
First, the lack of transport and high cost of exercise programs were often featured under environmental factors influencing PA in the United States [9, 31] and United Kingdom [12]. In our study, stroke survivors reported greater concerns with access and use of public spaces for PA. In Singapore, lack of transport and high cost of exercise programs were likely less of an issue due to the widespread availability of public spaces to exercise. For example, public gyms are situated in most neighbourhoods [32], outdoor fitness zones are built near public housing flats [33] and the park connector network linking major parks and nature areas across Singapore offers a variety of routes and trails for walking and cycling [34]. The national initiative of according all Singapore citizens and permanent residents $100 credit to access public gyms and swimming pools [10] had also been useful, though it was apparent that stroke survivors used the $100 credit but were not able to access and use the facilities and equipment optimally (e.g., lack of ramp access in swimming pools limited easy access in and out of the swimming pools). Despite the rise of inclusive gyms and swimming pools in Singapore [35, 36], it appears that further solutions are needed for successful uptake. Some possible solutions offered by stroke survivors during the subsequent co-design workshops included raising awareness of inclusive sport facilities, running group exercises classes, and providing physical assistance and adaptive equipment (e.g., arm and leg straps) during the classes. These solutions are likely to have implications on intervention design for other people with physical disabilities.
Second, the impact of COVID-19 on stroke survivors’ PA levels was evident in our study as it was conducted during the pandemic while most studies looking at factors influencing PA were conducted before the pandemic [7, 8, 12]. Other than the COVID-19 restrictions which led to the cancellation of face-to-face exercise classes and social outings, and difficulty in securing gym bookings, stroke survivors found it hard adjusting to virtual activities and were more likely to stay at home in order to reduce the risk of infection. These factors highlight the need to consider country-specific factors in the current climate in order to develop targeted solutions for stroke survivors with physical disabilities to engage in PA after stroke.
Existing strategies and national initiatives targeted at improving PA in the general population were not accessible to stroke survivors with physical disabilities. In Singapore, there is a successful public health campaign termed the National Steps Challenge which uses trackers, rewards and mass media campaigns to engage the public in more PA [37]. To date, the campaign has registered 696 907 participants, engaged 60% of them in active participation for a median of 74 (IQR, 14–149) days, and increased mean daily step count by about 1579 (95% CI, 1564–1594) steps during the main challenge phase and 934 (95% CI, 916–952) steps during the maintenance phase. While some stroke survivors in our study could recall the importance of achieving 10,000 steps per day for health benefits (a possible outcome achieved via the public health campaign), some felt that the goal was unattainable and therefore did not partake in the campaign. When asked about incentives or reinforcements to stay active, the greatest determinant was the physical, mental and emotional benefits gained with PA, while vouchers and financial incentives received mixed responses. A previous nationwide survey in Singapore had identified the top three barriers to PA to be lack of time (65.3%), fatigue (64.7%) and pollution (56.1%) (n = 2867 participants sampled from a national administrative database of all residents in Singapore) [38]. These barriers differed greatly from the barriers to PA highlighted by stroke survivors in our study. Hence, it is likely that we need new strategies beyond the existing strategies and national initiatives to target stroke-specific barriers and improve PA after stroke in Singapore.
Using the recommended prioritisation criteria [19], we had identified eight TDF domains to be relevant (i.e., domains most likely to change behaviour if targeted in future interventions). To ensure we did not miss out any TDF domains that might have a significant influence on changing behaviour, we further rated the importance of each TDF domain as high, moderate or low importance depending on the number of criteria met [27], and discussed each domain and its potential influence on behaviour based on existing literature and data from our study. Of the remaining TDF domains of moderate and low importance (Table 2), we highlighted the intentions domain to be of potential influence due to the proven efficacy of implementation intention interventions in improving PA [39–41], presence of conflicting belief statements and the strength of belief statements from our data. According to the Theory of Planned Behaviour, intention refers to an individual’s willingness to act and predicts behaviour [42]. Although there remains an intention-behaviour gap (i.e., not all individuals with strong intention initiate the behaviour), implementation of intention interventions including action planning and coping planning are often incorporated to improve PA [39]. We are uncertain why the intentions domain was not relevant in our study, but the lack of motivation is not consistently reported in other studies on barriers to PA in the stroke population [8]. It is possible that in light of other problems such as stroke-related impairments, environmental barriers, low self-efficacy resulting in low engagement in PA, the lack of motivation becomes a smaller determinant to PA behaviour.
Strengths and limitations
The strengths of our study include the use of a comprehensive theoretical model of behaviour [11, 19, 24], and the inclusion of stroke survivors with physical disabilities. Current studies often include stroke survivors with no physical disabilities [7, 8, 12], and most have not used behaviour change theories to underpin their results [7, 8]. Despite the strengths, our study had several limitations. First, our study focused predominantly on the viewpoints of stroke survivors and did not capture the viewpoints of caregivers. Considering the pivotal role that caregivers play in the recovery journey and their unique experiences of healthcare services [43], it is worthwhile to explore determinants to PA from the caregiver’s viewpoint in order to target barriers which caregivers might foresee when supporting stroke survivors in their PA journey. Second, barriers and facilitators reported by stroke survivors were perceived, and might not necessarily reflect the actual barriers and facilitators to PA experienced in real life. Third, all interviews were conducted in English. Although English remains the dominant language spoken at home in Singapore [44], at least half of Singaporeans spoke other languages including Mandarin, Chinese dialects, Malay, Tamil and others. Future studies should consider interviewing caregivers and including other methods of data collection such as observations in stroke survivors’ homes to provide a richer picture of stroke survivors’ PA levels and further contextualise interview data [45]. Resources should also be allocated to assist with the conduct, analyses and translation of interview data in different languages to reflect viewpoints of all stroke survivors in Singapore.
Conclusions
In stroke survivors with physical disabilities, environmental context and resources had a significant influence on PA participation, and this often had a spill over effect into other domains. While some of our findings were similar to findings in other high-income countries, some factors were unique to the PA ecosystem in Singapore. Our results inform the development of a complex behaviour change intervention targeted at improving PA in stroke survivors with physical disabilities in Singapore.
Supporting information
S2 File. Consolidated criteria for reporting qualitative studies (COREQ): 32-item checklist.
https://doi.org/10.1371/journal.pone.0292442.s002
(DOCX)
Acknowledgments
We are grateful for the support rendered by the stroke survivors and staff at the Singapore National Stroke Association. We also thank Abigail Koh for conducting most of the interviews, and Elaine Lum, Duncan Smith and Lou Atkins for providing insights during the data analysis phase. Without the assistance of these individuals, we would not have been able to complete the study.
References
- 1. Fini NA, Bernhardt J, Said CM, Billinger SA. How to address physical activity participation after stroke in research and clinical practice. Stroke. 2021;52(6):e274–e7. pmid:33951930
- 2. van Wijck F, Bernhardt J, Billinger SA, Bird ML, Eng J, English C, et al. Improving life after stroke needs global efforts to implement evidence-based physical activity pathways. International Journal of Stroke. 2019:1747493019840930. pmid:30975042
- 3.
WHO. Physical activity: World Health Organization; 2020 [Available from: https://www.who.int/news-room/fact-sheets/detail/physical-activity.
- 4. Tudor-Locke C, Craig CL, Aoyagi Y, Bell RC, Croteau KA, De Bourdeaudhuij I, et al. How many steps/day are enough? For older adults and special populations. International Journal of Behavioral Nutrition and Physical Activity. 2011;8:80. pmid:21798044
- 5. Fini NA, Holland AE, Keating J, Simek J, Bernhardt J. How physically active are people following stroke? Systematic review and quantitative synthesis. Physical Therapy. 2017;97(7):707–17. pmid:28444348
- 6. Thilarajah S, Bower KJ, Pua YH, Tan D, Williams G, Larik A, et al. Modifiable factors associated with poststroke physical activity at discharge from rehabilitation: Prospective cohort study. Physical Therapy. 2020;100(5):818–28. pmid:31995190
- 7. Espernberger KR, Fini NA, Peiris CL. Personal and social factors that influence physical activity levels in community-dwelling stroke survivors: A systematic review of qualitative literature. Clinical Rehabilitation. 2021;35(7):1044–55. pmid:33586479
- 8. Nicholson S, Sniehotta FF, van Wijck F, Greig CA, Johnston M, McMurdo ME, et al. A systematic review of perceived barriers and motivators to physical activity after stroke. International Journal of Stroke. 2013;8(5):357–64. pmid:22974010
- 9. Rimmer JH, Wang E, Smith D. Barriers associated with exercise and community access for individuals with stroke. Journal of Rehabilitation Research & Development. 2008;45(2):315–22. pmid:18566948
- 10.
SportSingapore. Active SG—Membership Singapore: Singapore Sports Council; 2020 [Available from: https://www.myactivesg.com/about-activesg/membership.
- 11. Michie S, van Stralen MM, West R. The Behaviour Change Wheel: a new method for characterising and designing behaviour change interventions. Implementation Science. 2011;6:42. pmid:21513547
- 12. Nicholson SL, Donaghy M, Johnston M, Sniehotta FF, van Wijck F, Johnston D, et al. A qualitative theory guided analysis of stroke survivors’ perceived barriers and facilitators to physical activity. Disability and Rehabilitation. 2014;36(22):1857–68. pmid:24377329
- 13. Bailey R. Examining daily physical activity in community-dwelling adults with stroke using social cognitive theory: an exploratory, qualitative study. Disability and Rehabilitation. 2020;42(18):2631–9. pmid:30973766
- 14. Carroll DD, Courtney-Long EA, Stevens AC, Sloan ML, Lullo C, Visser SN, et al. Vital signs: Disability and physical activity—United States, 2009–2012. MMWR Morbidity and Mortality Weekly Report. 2014;63(18):407–13. pmid:24807240
- 15. O’Cathain A, Croot L, Duncan E, Rousseau N, Sworn K, Turner KM, et al. Guidance on how to develop complex interventions to improve health and healthcare. BMJ Open. 2019;9(8):e029954. pmid:31420394
- 16. O’Cathain A, Croot L, Sworn K, Duncan E, Rousseau N, Turner K, et al. Taxonomy of approaches to developing interventions to improve health: a systematic methods overview. Pilot and Feasibility Studies. 2019;5:41. pmid:30923626
- 17. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ. 2008;337:a1655. pmid:18824488
- 18. Skivington K, Matthews L, Simpson SA, Craig P, Baird J, Blazeby JM, et al. A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ. 2021;374:n2061. pmid:34593508
- 19. Atkins L, Francis J, Islam R, O’Connor D, Patey A, Ivers N, et al. A guide to using the Theoretical Domains Framework of behaviour change to investigate implementation problems. Implementation Science. 2017;12(1):77. pmid:28637486
- 20. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. International Journal for Quality in Health Care. 2007;19(6):349–57. pmid:17872937
- 21. Wilson JT, Hareendran A, Grant M, Baird T, Schulz UG, Muir KW, et al. Improving the assessment of outcomes in stroke: use of a structured interview to assign grades on the modified Rankin Scale. Stroke. 2002;33(9):2243–6. pmid:12215594
- 22. Wilson JT, Hareendran A, Hendry A, Potter J, Bone I, Muir KW. Reliability of the modified Rankin Scale across multiple raters: benefits of a structured interview. Stroke. 2005;36(4):777–81. pmid:15718510
- 23. Billinger SA, Arena R, Bernhardt J, Eng JJ, Franklin BA, Johnson CM, et al. Physical activity and exercise recommendations for stroke survivors: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2014. pmid:24846875
- 24.
Michie SA L.; West R. The Behaviour Change Wheel—A guide to designing interventions. Great Britain: Silverback Publishing; 2014.
- 25. Lim SH, Ang SY, Ong HK, Lee TZY, Lee TXL, Luo EZ, et al. Promotion of mobility among hospitalised older adults: An exploratory study on perceptions of patients, carers and nurses. Geriatric Nursing. 2020;41(5):608–14. pmid:32268947
- 26. Francis JJ, Stockton C, Eccles MP, Johnston M, Cuthbertson BH, Grimshaw JM, et al. Evidence-based selection of theories for designing behaviour change interventions: using methods based on theoretical construct domains to understand clinicians’ blood transfusion behaviour. British Journal of Health Psychology. 2009;14(Pt 4):625–46. pmid:19159506
- 27. Smith D, Cartwright M, Dyson J, Hartin J, Aitken LM. Barriers and enablers of recognition and response to deteriorating patients in the acute hospital setting: A theory-driven interview study using the Theoretical Domains Framework. Journal of Advanced Nursing. 2021;77(6):2831–44. pmid:33739478
- 28. Patey AM, Islam R, Francis JJ, Bryson GL, Grimshaw JM, Canada PPT. Anesthesiologists’ and surgeons’ perceptions about routine pre-operative testing in low-risk patients: application of the Theoretical Domains Framework (TDF) to identify factors that influence physicians’ decisions to order pre-operative tests. Implementation Science. 2012;7:52. pmid:22682612
- 29. Francis JJ, Johnston M, Robertson C, Glidewell L, Entwistle V, Eccles MP, et al. What is an adequate sample size? Operationalising data saturation for theory-based interview studies. Psychology & Health. 2010;25(10):1229–45. pmid:20204937
- 30. Landis JR, Koch GG. Measurment of observer agreement for categorical data Biometrics. 1977;33(1):159–74.
- 31. Hammel J, Jones R, Gossett A, Morgan E. Examining barriers and supports to community living and participation after a stroke from a participatory action research approach. Topics in Stroke Rehabilitation. 2006;13(3):43–58. pmid:16987791
- 32.
SportSingapore. Active SG—Facilities Singapore: Singapore Sports Council; 2020 [Available from: https://www.myactivesg.com/facilities/gyms.
- 33. SportifyCities. Singapore—Fitness zones Singapore2017 [Available from: https://sportifycities.com/singapore-fitness-zones/.
- 34. NationalParksBoard. Recreational Connectivity—Park Connector Network Singapore: Singapore; 2022 [Available from: https://www.nparks.gov.sg/gardens-parks-and-nature/park-connector-network.
- 35.
SportSingapore. Inclusive sport facilities—Swimming pools: Singapore Sports Council; 2022 [Available from: https://sportcares.sportsingapore.gov.sg/persons-with-disabilities/inclusive-swimming-pools/.
- 36.
SportSingapore. Inclusive sport facilities—Gyms Singapore: Singapore Sports Council; 2022 [Available from: https://sportcares.sportsingapore.gov.sg/persons-with-disabilities/inclusive-gyms/.
- 37. Yao J, Tan CS, Chen C, Tan J, Lim N, Muller-Riemenschneider F. Bright spots, physical activity investments that work: National Steps Challenge, Singapore: a nationwide mHealth physical activity programme. British Journal of Sports Medicine. 2020;54(17):1047–8. pmid:31857340
- 38. Koh YS, Asharani PV, Devi F, Roystonn K, Wang P, Vaingankar JA, et al. A cross-sectional study on the perceived barriers to physical activity and their associations with domain-specific physical activity and sedentary behaviour. BMC Public Health. 2022;22(1):1051. pmid:35614402
- 39. Peng S, Othman AT, Yuan F, Liang J. The effectiveness of planning interventions for improving physical activity in the general population: A systematic review and meta-analysis of randomized controlled trials. International Journal of Environmental Research and Public Health. 2022;19(12). pmid:35742582
- 40. Silva M, Sao-Joao TM, Brizon VC, Franco DH, Mialhe FL. Impact of implementation intentions on physical activity practice in adults: A systematic review and meta-analysis of randomized clinical trials. PLoS One. 2018;13(11):e0206294. pmid:30427874
- 41. Lin H, Yu P, Yang M, Wu D, Wang Z, An J, et al. Making specific plan improves physical activity and healthy eating for community-dwelling patients with chronic conditions: A systematic review and meta-analysis. Frontiers in Public Health. 2022;10:721223. pmid:35664117
- 42. Ajzen I. The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes. 1991;50(2):179–211.
- 43. Pindus DM, Mullis R, Lim L, Wellwood I, Rundell AV, Abd Aziz NA, et al. Stroke survivors’ and informal caregivers’ experiences of primary care and community healthcare services—A systematic review and meta-ethnography. PLoS One. 2018;13(2):e0192533. pmid:29466383
- 44.
GovernmentofSingapore. Department of Statistics Singapore—Home language of residents aged 5 years and over dashboard Singapore: Government of Singapore; 2022 [Available from: https://www.singstat.gov.sg/find-data/search-by-theme/population/education-language-spoken-and-literacy/visualising-data/home-language-of-residents-aged-5-years-and-over-dashboard.
- 45. Hall J, Morton S, Fitzsimons CF, Hall JF, Corepal R, English C, et al. Factors influencing sedentary behaviours after stroke: findings from qualitative observations and interviews with stroke survivors and their caregivers. BMC Public Health. 2020;20(1):967. pmid:32560713