Understanding the cultural meanings of stroke in the Ghanaian setting: A qualitative study exploring the perspectives of local community residents

Background: Stroke has undergone different medical constructions over the years. While the medical profession posits that disease is a biological condition, universal and unchanging, social constructionists perceive illness as the social meaning of the biological condition. Even though the medical notion of stroke is monolithic and sometimes contradicts the representations by local community residents, little attention has been paid to understanding the cultural meanings of stroke. This study explores the cultural meanings of stroke in five different cultural settings across Ghana. Methods: 30 focus group discussions (FGDs) were conducted with local community members in five communities (Ga Mashie, Tafo, Gyegyeano, Chanshegu and Agorve) located in five regions in Ghana. The FGDs were conducted in Ga, Twi, Fante, Ewe and Dagbani, and were transcribed verbatim into English. The transcripts were analysed thematically. Results: The local words used for stroke in all the five cultural settings focused on physical disability associated with stroke after its onset, and this formed the dominant source of fear about the condition. Participants mentioned that spiritual and left-side stroke have the most debilitating impact on the sufferer. Although there was a general consensus that anyone can be at risk of stroke, there was a gender dynamics in the explanation of risk relativity. Participants believed that stroke can be cured through early detection and treatment, use of herbal medicines, and availability of financial resources. Compared to other disabling conditions, the community residents perceived stroke to be more severe due to the multifaceted disabilities associated with the condition. Conclusions: This study showed that the social meanings of stroke in the five communities are multifaceted, and reflected co-existence of biomedical and cultural frameworks. The findings showed the need to pay good attention to the sociocultural context when developing interventions strategies on stroke prevention and control in Ghana.


Introduction
Over the years, stroke has undergone different medical constructions globally, and these influence perceptions and practice on management and prognosis. From the time of Hippocrates to the first half of the twentieth century, medical science mostly referred to stroke as "apoplexy", meaning any sudden condition that began with loss of consciousness especially one in which the sufferer died within a matter of seconds after losing consciousness (Cooke, 1824, cited in Pound et al., 1997. Hence, the global medical discourse was that apoplexy was a sudden acute event, with no hope of survival. For instance, Osler (1909), in his book on "principles and practice of medicine" made a suggestion that when an apoplectic sufferer is completely paralysed, the friends should be told at the onset that there is a slim chance of recovery, and in cases where the hemiplegia persists for more than three months and muscle contractures had developed, 'it is the duty of the physician to explain to the patient, or to his friends, that the condition is past relief, that medicines and electricity will do no good, and that there is no possible hope of cure ' (Pound et al., 1997). Even though there was pessimism about the prognosis of apoplexy, most especially during the 18 th century, physicians hoped for a future in medical advancement with prospects of better outcomes.
Towards the end of the 18 th century, there was a shift in the conceptualisation of apoplexy, and attention was paid to the characteristics that people with the illness shared, rather than focus on individual characteristics. In addition, even though there had been a distinction between apoplexies caused by obstruction and the ones caused by haemorrhage prior to this period, there was a new explanation that both conditions were caused by degeneration of the arterial wall (Price, 1966). As a result, the term 'cerebrovascular disease' emerged in medical discussions and the use of the word 'apoplexy' faded away gradually (Pound et al., 1997). Based on the medical construction of cerebrovascular disease, the emphasis on the person experiencing the condition began to fade away and 'patients became less and less visible in medical texts ' (Pound et al., 1997). The focus of the physicians shifted to the inside of the body of the sufferer, and the disease became more separate from the person in whose body it resided. During the 1950s, new treatment procedures such as angiography, cerebrovascular surgery and anticoagulants were developed for cerebrovascular disease and this formed the basis for a gleam of hope for better prognosis.
In spite of the fact that new treatment strategies for stroke were introduced during the 1950s, studies in North America showed that culturally informed ideas about aging shaped the provision of stroke care and the possibilities and perceived value of intervening. For instance, up to the 1970s, stroke was equated with 'nothing more to do' stage of a terminal illness and connoted hopelessness in a General Hospital in North America. According to Hoffman, 'nothing can be done' paralleled the concept of 'nothing more to do' when it came to stroke management and this was due to low education and training on stroke management as well as institutional policies and professional values that minimized the benefits that health professionals could obtain through the care of stroke patients (Hoffman, 1974). This, however, influenced staff attitudes towards stroke patients and they either withdrew from participation in care of these patients or denied the complications of stroke by managing patients in a way that concealed from them the chronicity of their disability. They let the patients believed that they could recover completely if only they tried enough because the responsibility for getting well lied with them (i.e. patients) and not physicians. Hoffman explained that belief in the possibility of complete elimination of disability delayed patients from making 'satisfactory long-term adjustment to their condition' (Hoffman, 1974). As a result, some of the rehabilitation strategies were unsuccessful because patients had the perception that they could get around their disability since rehabilitation could only enable them to live in spite of disability, and not without it. Nevertheless, research shows that stroke rehabilitation do not promote actual physiological recovery, but only enables patients to fully use whatever function remains poststroke, to gain strength and endurance (Kaufman & Becker, 1991).
Although the word 'stroke' firstly occurred in English literature in the late 16th century (1599) (Pound et al., 1997), it became widely used from the 1950s. The term 'stroke' was used to capture the sudden nature of an acute event, and was adopted at the time when the initial surgical or drug treatment for the condition was dashed. In 1970, a new Journal called "STROKE" was developed to synthesise evidence on stroke in a way that would help physicians across the world get access to updated information on stroke (Millikan & Clarke, 1970); this journal probably increased the use of the word "stroke". Through medical advancement, stroke is now seen as a condition of vascular origin but with hope of rehabilitation through physiotherapy and occupational therapy (Peszczynski et al., 1972;Walker et al., 2013). The World Health Organization therefore define stroke as "rapidly developing clinical signs of focal (or global) disturbance of cerebral function, with symptoms lasting 24 hours or longer or leading to death, with no apparent cause other than of vascular origin" (Truelsen et al., 2000).
Even though the medical construction of an illness is important for management and prognosis, perceptions of local community

Amendments from Version 1
This version addresses all major and minor criticisms of the reviewers. In the introduction section, in addition to the historical overview presented earlier, I have referred to some studies from North America which illustrate how culturally informed ideas about ageing shape the provision of stroke care and the possibilities and perceived value of intervening. In the methodology section, I have provided information about the burden of stroke in Ghana and the current provision of care in the country. Further, I reorganised the results and discussion sections to provide more coherence to the paper and to tell the stories that are emerging from the data. Also, sentences which demonstrate the judgement that medical thinking is superior have been reworded to show that social and cultural meanings are equally valid and important in stroke management and control. .

REVISED
residents are equally important, albeit often different from the medical perception. Sociologists believe that there are complex relationships between the medical and cultural notion of disease. Social constructionists particularly posit that individuals and groups produce their own conceptions of reality, and that knowledge itself is the product of social dynamics. For the medical profession, disease is a biological condition, universal and unchanging; however, social constructionists perceive illness as the social meaning of the 'biological' condition. That is, diseases are not just biological conditions, they have social meanings (Sontag, 1978) and these exist for social rather than pure biological reasons (Conrad & Barker, 2010).
The social meanings of illnesses shape how people make sense of and respond to an illness. Ervin Goffman argued that some illnesses are stigmatised and considered disabilities while others are not (Goffman, 2009). Despite the fact that there is nothing inherent in an illness that makes it stigmatising, the social meanings acquired through social interactions makes a condition stigmatised or considered as disabilities unlike others (Goffman, 2009). The cultural meanings of illness are therefore important because they have an impact on the way the illness is experienced, how the illness is depicted, the social response to the illness and what policies are created concerning the illness. The cultural meanings also shed some light on the rationale for noncompliance which is relevant for developing contextcentered strategies to improve effective implementation of medical treatment (Conrad & Barker, 2010).
While the medical notion of stroke has already been established, the cultural meanings of stroke have received little attention. Allan Young particularly argued that there is usually a difference between diagnosis of an illness as part of the biomedical model and illness experience as part of the sociological or anthropological understandings and experience of such illness (Young, 1976). In British Colonial Africa, despite the fact that biomedical knowledge played an important role in the wider creation of knowledge of 'the African', many biomedical theories and interventions failed because no consideration was given to the social and political context (Vaughan, 1991). A few existing studies in some parts of Africa showed that people draw on both medical and supernatural frameworks of understanding to explain stroke and its associated disruptions to individual biographies (Legg & Penn, 2013;Mshana et al., 2008a;Mshana, 2008b;Penn, 2007;Penn et al., 1992;Sanuade, 2016), and communities regard both the clinical and social diagnostics as important in stroke management (Hundt et al., 2004). The aim of this study was to explore the cultural meanings of stroke in five communities across Ghana. Specifically, this study explored perceptions on stroke definition, types, cure, 'group at risk of', and general community ideas. Findings from this study can help to understand how stroke is situated within the Ghanaian cultural context, and this is important for developing intervention strategies on prevention and management.

Design
This was a cross-sectional qualitative study. Data were gathered through focus group discussions (FGDs) with local community residents and none of these participants had stroke at the time of the data collection. Six FGDs were conducted in each of the study sites and participants were segmented by age and sex ( Figure 1). A total of 30 FGDs were conducted in five communities across Ghana through a purposive sampling technique. Participants' profiles are presented in Table 1. The number of participants in each of the group discussions ranged from six to ten. Research has shown that this range is adequate to facilitate informative conversation and for the group discussion not to become disorderly or fragmented (Krueger, 1994;Nyumba et al., 2018). The Standards for Reporting Qualitative Research (SRQR) guidelines were used in this study (O'Brien et al., 2014).

The study context
Ghana is a multi-ethnic, multi-religious and multi-cultural country with ten administrative regions. Its current population, which is estimated at about 25 million (Ghana Statistical Service, 2012), is a mix of large and small ethnic groups. Some of the major groups in the country include the Asante, Fante, Ga, Ewe and Dagomba. Even though people from various ethnic groups are dispersed across the country due to easy geographical and social mobility, many of these people still retain their cultural identities (Asante & Gyimah-Boadi, 2004).  The Burden of Disease Study conducted in 2016 showed that stroke is the eighth leading cause of death in Ghana and a major contributor to the Disability Adjusted Life Years Lost (DALYs) (Ghana Ministry of Health, 2016). Evidence also suggests increase in stroke prevalence and incidence in the country. Despite the increase in stroke morbidity and mortality (Agyemang et al., 2012), there is limited evidence-based stroke services and low priority for stroke care in Ghana. The health system landscape in the country is such that healthcare delivery is provided by the formal medical healthcare system as well as informal non-medical healthcare such as faith-based and ethno-medical systems. Out of these healthcare delivery systems, the formal medical healthcare system is seen as the only 'legitimate' healthcare system set up for stroke management and control. Even though there are tertiary teaching hospitals, regional hospitals, district hospitals and sub-district health centres in the country, the district hospitals and subdistrict health centres often have limited clinical capacity for stroke care. Also, the referral hospitals which are supposed to be well equipped for stroke management and control often lack adequate capacity for acute stroke care. For instance, there is only one stroke unit in the country. Out of the 11 major referral hospitals in the country, 7 hospitals have functional computed tomographic (CT) scan and only one hospital have functional magnetic resonance imaging (MRI) scan services (Baatiema et al., 2017). In addition, none of these hospitals has an occupational or a speech therapist to assist in providing acute stroke care (Baatiema et al., 2017).

Study communities
Data were collected in five communities (Ga Mashie, Tafo, Gyegyeano, Chanshegu and Agorve) located in five regions in Ghana (Greater Accra, Ashanti, Central, Northern and Volta regions, respectively) ( Figure 2). Participants were selected from these communities in order to capture nuances in stroke perceptions from the major ethnic groups in Ghana.
Ga Mashie is located on the Atlantic Coast of the Greater Accra Region of Ghana. It is home to the Ga people who speak the Ga language. The area is referred to as 'Old Accra' because it is the place where the original Ga ethnic group first settled; this makes Ga Mashie the oldest community in Accra (Mahama et al., 2011). Ga Mashie, constituting James Town and Ussher Town, covers an area of 100 hectares along the southwest coast of Accra. The population of Ga Mashie has grown rapidly over the years and it is currently one of the most densely populated communities in Accra (GSS, 2012). It is characterised by low income, low levels of education, inadequate access to improved sanitary facilities, low access to healthcare services, high population density and double burden of communicable and non-communicable diseases (NCDs) (Awuah et al., 2014;Mahama et al., 2011). Furthermore, there have been several NCD studies and interventions in these communities (Awuah et al., 2014;de-Graft Aikins et al., 2014;Sanuade, 2016) and this indicates that the residents may have more NCD knowledge compared to the other communities. Additionally, there was previous contact with Ga Mashie residents and this made recruitment easier compared to the rest of the communities.
Tafo is a town in Kumasi Metropolitan District in the Ashanti Region of Ghana. The largest ethnic group in this city is the Asante, a sub-group of the larger Akan ethnic group, and the population speaks the Twi language. Although there is ethnic and cultural diversity in this community, the population is close-knit due to the presence of a strong traditional administrative set-up that fosters cohesion among the diverse ethnic groups. The community is densely populated, has low level of education, and low socio-economic status, and poor sanitation Gyegyeano is located in Cape Coast in the Central region of Ghana. It was founded by the people of Oguaa. It is one of the most historic cities in Ghana and has had a lot of interactions with the Europeans (the British, the Portuguese, the Swedish, the Danish and the Dutch). It is home to the Fante people, a sub-group of the larger Akan ethnic group who speak Fante. The city has some of the Ghana's best secondary and Technical schools and so attracts people from across the country.
Chanshegu is located in Tamale, which is one of the six Metropolitan Assemblies in Ghana and the only Metropolis in the three Northern regions (i.e. Upper East, Upper West and Northern regions). Chanshegu is occupied by an ethnic group called the Dagombas; they are the largest ethnic groups in Tamale and speak Dagbani. This community has a traditional administrative head who is the leader of the community. Although the community is sparsely populated, the people are close-knit and are predominantly Muslims. It is characterised by low income, low level of education, poor sanitation, and low access to healthcare services.
Agorve is located in Keta, a town in the Volta Region of Ghana. It is the capital of the Keta Municipal District. The area was an important trading post between the 14th and the late 20th centuries. It was settled by the Anlo Ewe, a sub-group of the Ewe people who, in the 17th century, migrated to the area from Ketu, in Benin and speak 'Anlogbe', a dialect of the Ewe language. They are a patrilineal society governed by a hierarchical, centralized authority. . The aims of the study were explained to the participants and written consent was obtained from all of them. Permission to record the group discussions was obtained from the study participants prior to the start of the group discussions. Confidentiality was strictly observed. At the end of the data collection, the recordings were transferred from audio recorders to a computer and were deleted after the extraction. Currently, they are being stored on a secure laptop and will be deleted on October 16 2018.

Data analysis
The analysis was performed after all the group discussions were conducted, transcribed and quality checked. All interviews were transcribed verbatim from Ga, Twi, Fante, Ewe and Dagbani into English language by a team of transcribers with Ga, Twi, Fante, Ewe, Dagbani and English language competence. Data were checked and cleaned, stored as Microsoft Word files (using MS-Word, 2010) and exported to ATLAS TI7. Thereafter, data were analysed using thematic approach (Attride-Stirling, 2001). The first stage of analysis involved reading all the transcripts to identify emerging codes and themes. A mix of deductive and inductive codes was employed for the analytical framework. The deductive codes were derived from previous studies on stroke reviews (Karenberg, 1994;Morris, 2011;Pound et al., 1997), and conceptual ideas were drawn from the social construction of illness theory (Conrad & Barker, 2010). Several discussions were held with a colleague, an expert in African history, so as to develop an elaborate codebook that captures social meanings of illness in an African setting. The second stage of the analysis involved identifying the linkages between codes, themes, and appropriate respondent quotes, and existing research. The interpretive process was shaped by the social construction of illness framework to highlight the cultural meanings and description of stroke by the five communities.

Results
The findings are grouped under seven themes: 1) local words for stroke and their meanings; 2) community ideas on stroke; 3) typology of stroke; 4) perceptions on cure of stroke; 5) groups at risk of stroke ; 6) most fearful thing(s) about stroke, and; 7) stroke and other disabling conditions.

Local words for stroke and their meanings
Participants were asked to provide the local words used for stroke, and what these meant. Their responses are summarised in Table 2. The medical English word 'stroke' was used during the group discussions and participants were asked to provide the local words (Box 1). Also, many of the community members were familiar with the word 'stroke' and could recognise who a stroke survivor is in line with medical descriptions. In the five communities, stroke was described as: 1) a debilitating or paralytic illness; 2) one leg one arm; 3) a ghost illness, and; 4) a shivering illness. All the participants referred to stroke as a 'paralytic or debilitating illness' because it is an illness which breaks the sufferer down by causing loss of body functions and incapacitation. Stroke was also described as 'one leg one arm' among the Ga (in Ga Mashie) and Ewe (in Agorve) because after its onset, the sufferer either experiences weakness at the right or the left side (i.e. arm and leg). In addition, Ga Mashie participants referred to stroke as a ghost illness "sisa hela"; they mentioned that this term was developed in the 1960s when community knowledge on causes and management of stroke was low. Sisa hela (ghost illness) is generally used among Ga to refer to any illness that people don't know how it came about and curing it is difficult. Finally, Ga Mashie participants referred to stroke as " kporkpormorhela" which literally means epilepsy in Ga; this is because the onset of stroke triggers uncontrollable shaking similar to epilepsy.

Community ideas about stroke
Participants were asked to state the kinds of ideas or beliefs that people in their communities have about stroke. Four different ideas emerged from their narratives. In Ga Mashie, stroke was referred to as a ghost illness due to lack of knowledge on causes and management. It was also a belief in the community that stroke occur as a result of being cursed. Those from Chanshegu mentioned several pluralistic ideas about stroke. First, there was the idea that stroke affects only those who are strong, and these are mostly men. They mentioned that some people in their community also have the idea that stroke can affect anyone irrespective of whether they practice good lifestyles or not. There was also the belief that stroke can only attack someone in the midst of other people. That is, stroke does not affect the sufferer when he/she is alone. In addition, some of the community members believed that stroke is not a hospital illness as it can only be managed with local herbs. Finally, stroke was seen as the handiwork of God in this community. That is, stroke occurrence is neither due to the weakness of the sufferer nor his or her lifestyles, but it is a divine work of God, and this is why the place for stroke management is not the hospital.

Typology of stroke
When participants were asked to describe stroke in their local contexts, typology of stroke emerged as a thematic area of discussion. This typology broadly focused on stroke causation and characteristics. With respect to causation, participants from Agorve and Gyegyeano differentiated between natural and spiritual stroke, although there were contradictions in their explanations. They mentioned that natural stroke is caused by hypertension while spiritual stroke is caused by witchcraft activities, sorcery and curses. There was the belief among those from Agorve that while natural stroke is difficult to manage, spiritual stroke can be easily cured. They particularly mentioned that: . On the contrary, those from Gyegyeano mentioned that while natural stroke is easy to manage, spiritual stroke is difficult to manage because no matter the amount of medicine taken, the stroke will eventually kill the sufferer. This was evident in the narratives of the adult male group.
…there are some orthodox medicine as well as traditional medicine for natural stroke. It's the spiritually caused stroke which is dangerous. No matter the kind of medicine you take, the stroke will still kill you [Adult male, Gyegyeano] Regarding stroke characteristics, participants from Ga Mashie and Gyegyeano explained the difference between mild and severe stroke. They mentioned that while mild stroke affects some parts of the body, severe stroke affects all parts of the body and makes the sufferer bedridden. One of the participants from the young women group in Gyegyeano said that: Although it was mentioned that anyone is at risk of stroke, participants were of the opinion that stroke cases are higher among the elderly. There was however the notion that even though the elderly are more affected by stroke, the prevalence is low among those who are 80 years and above. In addition, even among the elderly, participants mentioned that men are more at risk of stroke than women because of: frequent thinking/worrying due to marital breakdown and joblessness; indulgence in indiscriminate sex such as having multiple sexual partners and having sex while standing, and; the notion that stroke affect those who are strong and men are usually the strong ones. Further, the young women group in Gyegyeano mentioned that the reason why women are less likely to get stroke is because their menses act as a protective factor.

I learnt that when women go through their menses, all the waste products in the blood is expelled and that's why they aren't likely to get stroke as men [young women, Agorve]
On the contrary, the elderly women group in Cape Coast mentioned that stroke is common among women because women experience a lot of stress and worrying compared to men.

Stroke is common among women because women worry a lot more than men.. in our daily activities, we carry too much heavy loads. When it gets to a point in time that practice brings sicknesses upon us [Elderly women, Agorve]
Most fearful things about stroke The dominant theme on what participants fear most about stroke was the physical disability associated with the illness. All the groups mentioned that they are afraid of post-stroke physical disabilities such as mouth distortion, speech impairment and paralysis which confines the sufferer to one place and incapacitates such individual not to engage in any activity. The adult women group in Tafo mentioned that their fear of stroke is that it breeds poor interpersonal communication by making it difficult for the sufferer to communicate their needs to others.
Further, the sudden change in the biography of the sufferer constitutes the fear for some of the participants. Particularly, they mentioned that they are frightened when they see a stroke survivor and this is because the sufferer may be moving about normally at one moment and suddenly, the person has a stroke and the entire body functions change. One of the members of the adult women group in Gyegyeano said that:

Yes, when I see a stroke patient I become frightened. This is because some few days ago the stroke patient was moving about normally now he's paralyzed in some parts of the body. The sudden change really scares me [Adult women, Gyegyeano]
In addition, participants said that they fear the psychological disability associated with stroke and this is because after the stroke onset, the sufferer engages in frequent worrying and thinking due to changes in the physical body and life trajectory. Inability to cater for family due to stroke complications featured in all the group discussions in Chanshegu. These participants mentioned that the onset of stroke makes it difficult for the sufferer to take care of his/her family because of restrictions in activities of daily living. As a result, there is a serious effect on the economic circumstances of the entire household and children's education and general well-being are affected.
Some of the participants from Agorve mentioned that what they fear most about stroke is the fact that it brings about changes in dietary patterns/practices. They mentioned that a stroke survivor may no longer have the opportunity to indulge in some foods which he/she likes. For instance, the elderly male group said that: ..when you get stroke, there are certain foods you

can't eat anymore. So that is what frightens us about it [Elderly men, Agorve]
Furthermore, the young men group in Chanshegu said that one of the things they fear most about stroke is the fact that it can affect anyone. Finally, some of the participants mentioned that the fact that stroke eventually causes death makes them afraid. The belief for some people was that no matter how much an individual manages the condition, the person will eventually die as a result of the condition.
My prayer at all times has always been against stroke. If I am sitting and it comes to mind, I always feel like crying. Ever since I grew to a certain age if even some is cured of stroke, by and by the person will not survive for long. No matter how cured the person will be it will get to a point that you will see that the person has died as a result of that same illness. So for this illness, we are just praying to God that wherever it's cure will come from to help us. It will be good because if you hear that someone has been treated of stroke and is cured, by and by the same ailment will be back again [Adult men, Chanshegu] Stroke and other disabling conditions Participants were asked to explain if there is a difference between people living with stroke with post-stroke disabilities and people living with other forms of disabling conditions. Participants differentiated between post-stroke disabilities and other disabling conditions such as mental disability (e.g. psychosis) and physical disabilities (e.g. impaired-eye and other physical deformities). The first category of people believed that poststroke disability is better than mental disability. For instance, with respect to stroke-induced disability and psychosis, participants mentioned that although the speech of the stroke sufferer may be impaired, he/she will still be able to hold meaningful conversation with others; however, a psychotic person finds it difficult to hold logical communication with others.
Someone with psychosis will speak alright but it won't make sense. You will hear everything he says but his speech The second category believed that stroke is worse than other disabling conditions because of its multifaceted disabilities, and this formed the dominant theme. The narratives of the participants suggest that for someone with physical disabling condition, he/she may only have just that particular disability to deal with; however, stroke is usually associated with multiple disabilities which may be physical, mental, social, and cognitive. Another reason given was that people with physical disabling condition are better than those living with stroke because they are created by God and so they are not sick.

Discussion
This study explored the cultural meanings of stroke in five communities in Ghana. All the words used to capture stroke in the five cultural settings centered on the physical disabilities associated with stroke after its onset, with no mention of other disabilities such as cognitive dysfunction, behavioural disability, psychological disability, etc. In addition, the dominant theme on what participants fear most about stroke was the post-stroke physical disability. This is similar to the study by Hundt et al., 2004 where the local words used by Shangaan Tsongan Mozambican refugees and South Africans in the Agincourt sub-district of Limpopo Province for the medical term 'stroke centered on weakness in one side of the body or paralysis. The present findings suggest that the physical impact of stroke is more apparent and devastating in these communities. This is not surprising because even for stroke survivors, the physical impact of the illness constitutes the most important source of worry to them (Sanuade, 2016). Stroke particularly brings about multifaceted disruptions which affect the life trajectories of the sufferers (Bury, 1982;Becker, 1993;Kaufman, 1988;Owolabi & Ogunniyi, 2009;Sampane-Donkor et al., 2014). The implication of this is that interventions that address the physical impact of stroke may be easily welcomed in these communities. Also, this finding suggests the need to provide more education on the multifaceted impact of stroke in these communities.
Regarding stroke typology, participants had multiple ideas about stroke causation. Their narratives also suggest that adequate treatment of stroke depends on what the root cause of the illness is. While the 'naturally caused' strokes can be treated by physicians, 'spiritually caused' strokes can be treated either through faith or traditional healing. Furthermore, there was a general notion among the participants that stroke can be cured through early detection and treatment, use of herbal medicines, and availability of financial resources. The Ghana Health Service can capitalize on this by equipping local hospitals in such a way that early diagnosis could be made and patients could be stabilized before being referred to district or tertiary hospitals for further treatments. Previous research in Ghana showed that stroke has a huge financial impact on the sufferers and their families (Sanuade, 2016); it is important that stroke treatment is covered in the National Health Insurance Scheme (NHIS) as this may improve the biomedical management of the condition. This can also be facilitated through the provision of NCD products such as laboratory tests and NCD care products at different hospitals (e.g. blood glucose meter, BP monitors) (de-Graft Aikins et al., 2014). Regarding the use of herbal medicines as a treatment plan to cure stroke, the present findings reinforce the fact that the use of herbal medicine is part-and-parcel of the health seeking behaviour of Ghanaians (Awuah et al., 2014;Okyerefo & Fiaveh, 2017), and will remain so for a long time. In the meantime, pharmacology tests to showcase 'safe' herbal medicines for stroke management needs to be a continued pursuit by the government.
The findings also suggest that participants' perceptions on stroke causation and management are a mix of biomedical and sociocultural/spiritual factors. However, primary intervention and management strategies in Ghana mostly focus on the biomedical explanations of illness and do not take the socio-cultural explanations into consideration. This partly explains why many of these interventions have been unsuccessful. An important way that may help to bridge the gap between the medical construction of stroke and the social representations of the illness in Ghana is making use of the community health workers, through the taskshifting framework. Since community health workers have been trained to provide care in dominant community health problems, they can equally be trained on stroke so as to increase their knowledge, confidence and ability to provide information, care, and support for local community residents (de-Graft Aikins et al., 2014). This may consequently enhance primary prevention initiatives, acute care and rehabilitation and longer term social care for stroke in these communities.
Participants also believed that there are gradients in stroke severity. Generally, there was the notion that spiritual and leftsided strokes are more severe than natural and right-sided strokes, respectively. The findings suggest a high level of spiritual explanation of disease conceptualisations in these communities and this is because the activities of witchcraft, curse, and ghosts featured prominently in how stroke was generally perceived. Previous anthropological accounts in Ghana and other African countries showed that these activities are central to how an illness is perceived (Ashforth, 2003;Evans-Pritchard, 1976;Field, 1937;Hundt et al., 2004;Moore & Sander, 2003). The findings also showed that these beliefs are still part of the social fabric of Ghanaians and reinforces Allan Young's idea that there is a difference between biomedical diagnosis and sociological understandings and experience of an illness (Young, 1976). Further, the findings underpin what these previous studies showed that illness causation and experience in an African setting is a mix of biological, psychological, social and cultural factors. It is important that these contexts are considered when developing intervention strategies on stroke in these communities. Based on medical perspective, stroke has two main types namely haemorrhagic and ischemic stroke, and their treatment strategies are radically different. None of the participants in this study differentiated between these two types of stroke; perhaps, they had no idea how they are different in terms of aetiology and management. It may be easier for local community residents to have more knowledge of this if the stroke types are classified as two separate diseases instead of one disease with two types. This is because previous engagements with some Ghanaian communities suggest that community residents think in terms of different disease conditions rather than different types of the same condition. They sometimes find it confusing to distinguish between types of a particular illness. Hence, in order to simplify identification of stroke types by 'non-health professionals', the possibility of classifying them as two separate conditions should be looked into by the medical community.
With respect to the groups at risk of stroke, the findings showed that there were age and gender dynamics. Regarding age, many had the notion that older people have higher risk of stroke, and this is consistent with what literature has shown (Feigin et al., 2017;Zhang et al., 2012). Research showed that stroke incidence is higher among the elderly because they are more likely to have higher prevalence of classic vascular risk factors such as ischemic heart diseases, chronic heart failure, and atrial fibrillation, which build-up over time (Yao et al., 2012). Concerning the gender dynamics on risk relativity, while the male groups mentioned that men are more at risk of stroke, the female groups stated that women have higher risk. This suggests that both gender perceived themselves as 'victims of risk'. Research has shown that there are no inherent biological characteristics that predispose a particular gender to stroke; however, relativity of risk is dependent on different exposure to stroke risk factors, medical treatment, and therapeutic interventions (Morović et al., 2012). Interventions that emphasise this information need to be promoted in these communities. An unexpected theme which emerged during the group discussions with young women was the mention of menstruation as a protective factor for stroke. This contradicts the findings from a study of 1,412 postmenopausal Japanese women (Murakami et al., 2016). The study showed that during the 12 year follow-up (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010), menstrual factor such as early menarche was associated with incidence of stroke and cerebral infarction (Murakami et al., 2016).
Since Murakami et al. (2016)'s findings cannot be extrapolated to Ghanaian setting, this study recommends the need for future studies to explore the association between menstrual factors and stroke, especially in the five communities.
Generally, participants mentioned that stroke is more severe than other disabling conditions due to the multifaceted disabilities associated with the illness. This corroborates what other research has shown, that stroke is a major cause of complex disability (Adamson et al., 2004;Lai et al., 2003). Compared with other diseases, stroke is a major health problem that can cause multiple or concurrent disabilities in an individual (Lai et al., 2003) which can influence all dimensions of life, including the simplest self-care tasks (Adamson et al., 2004). In addition, participants perceived stroke to be more severe than other disabling conditions because while people with other mental or physical disabling conditions are not considered as sick because they are God's creations, those with stroke are perceived as sick. This belief reinforces the constructionists' idea that 'there is nothing inherent about a condition that makes it stigmatizing; rather, it is the social response to the condition and some of its manifestations, or the type of individuals who suffer from it, that make a condition stigmatised', This also supports the social model of disability that disability is determined by the way society is organised and cannot be reduced to a mere biological problem located in an individual's body (Barnes & Mercer, 2010).

Study limitation
The major limitation of this study is that the findings may not be generalizable to all Ghanaian communities or ethnic groups. It is possible that people who belong to the same ethnic group but located in separate communities may perceive stroke differently. However, the multi-site approach adopted in this study revealed interesting variations in the social meanings of stroke across different geographical locations. These meanings can enhance development of context-specific interventions for stroke in other African countries.

Conclusion
This study showed that the cultural meanings of stroke in five communities across Ghana are multifaceted and sometimes contradict the medical notion of the illness. Most often, people do not draw on one of either the cultural meanings or medical meaning, but can draw on these two meanings simultaneously or in a sequential manner. Since individuals and groups produce their own conceptions of reality and because knowledge itself is the product of social dynamics (Conrad & Barker, 2010), it is important that intervention strategies, targeted at improving stroke prevention and control, need to pay critical attention to the cultural meanings of the illness. This is because culture is an important building block for how diseases are interpreted and managed (Airhihenbuwa & Iwelunmor, 2012). Hence, if medicine as a practice or policies focus exclusively on the natural history of stroke, it may be difficult to ensure adequate primary prevention strategies and stroke rehabilitation in Ghanaian communities.
Since there are usually ontological divides between patients and physicians, many of the people living with stroke live in cultural reality that is sometimes different from that of the physicians. Thus, it is important that stroke clinicians have a good understanding of the social and cultural landscapes of the communities in which they work so that they can, together with local communities, come up with context-specific interventions for stroke management and control.Therefore, studies which explore the cultural meanings of the different aspects of stroke (i.e. causes, prevention, and management) and the association between menstrual factors and stroke need more attention. This study also highlights the need to promote education on the multifaceted impact of stroke in Ghanaian communities. Finally, in addition to paying attention to the cultural meanings of stroke, other factors such as access to care, cost of health care, economic realities and constraints of poverty should be given much consideration in stroke prevention and control.

Data availability
The full data for this study are not provided because the audio and transcripts of the interviews contain identifiable and sensitive information making it impossible to protect participants' confidentiality. To apply for access to the data, please contact the author at o.sanuade@ucl.ac.uk. Data will only be shared if the following conditions are met. First, anyone who applies to access the data must maintain the confidentiality of the participants. Second, the identity of any individual or organization mentioned in or in relation to the data must not be disclosed. Third, the author of this article must be informed about any publication in which the data will be used. Finally, if the data is used in any other publication, it should be acknowledged that it is of secondary source, and appropriate citation must be provided.

Grant information
This study was supported by the Wellcome Trust [106534].
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

School of Population Health & Environmental Sciences, King's College London, London, UK
This paper is interesting and reports novel findings. It is relevant given the recognition of the significance of non-communicable disease in low and middle income countries. Apart from the historical overview presented in the background to the paper, there is relatively little research on how stroke is perceived from a social-cultural perspective. However, there is some research that could be referred to: for example, work from North America has illustrated how 'culture' -for example culturally informed ideas about agingshapes the provision of stroke care and delineates the possibilities and perceived value of intervening .
Consideration of this work might also help the author to review what appears to be a tendency in the paper to assume that while lay people may indeed draw on both cultural and scientific ideas to talk about stroke (page 12) stroke in scientific discourse is culture free.
Some studies have reported research into cultural constructions of disability in Ghana and I wondered if these were relevant. I think it is unlikely that cultural understandings of stroke will exist entirely separately from wider constructions of disability, ageing, as well as disease and misfortune in general.
The study engaged a large number of participants and a great deal of effort must have been made to recruit this large and diverse range of participants. It wasn't entirely clear to me how participants were identified and invited to take part. Were any participants themselves stroke survivors or was this intended to collect information from the general population?
I would have liked a bit more epidemiological and health service information about the relative significance of stroke, incidence, and current provision of care in Ghana.
It might be useful to more clearly articulate implications for primary prevention initiatives, acute care and rehabilitation and longer term social care.
Minor points What does it mean to say that 'Medicine therefore appealed to the lay model in order to put the onus of recovery on the patient' (page 3)?
Page 8: '… in as much as such individual engages in poor lifestyles.' -rephrase.
One of the causes of stroke identified is thinking/worrying (page 8); does this equate to 'stress'?
Page 11: 'Stroke particularly bring about…': 's' missing Page 11: I wasn't convinced by the suggestion that it would be easier for local community residents to have more knowledge of the different aetiologies of stroke if these were classified as two different types of stroke, rather than one disease with two types. Comment: Apart from the historical overview presented in the background to the paper, there is relatively little research on how stroke is perceived from a social-cultural perspective. However, there is some research that could be referred to: for example, work from North America has illustrated how 'culture' -for example culturally informed ideas about ageing -shapes the provision of stroke care and delineates the possibilities and perceived value of intervening . Consideration of this work might also help the author to review what appears to be a tendency in the paper to assume that while lay people may indeed draw on both cultural and scientific ideas to talk about stroke (page 12) stroke in scientific discourse is culture free.

There was a general consensus in all the group discussions that anyone can be at risk of stroke if such individual engages in poor lifestyles.
Comment: One of the causes of stroke identified is thinking/worrying (page 8); does this equate to 'stress'?
Response: This is different from stress because the women used the concept of stress . differently from worrying/thinking Comment: Page 11: 'Stroke particularly bring about…': 's' missing Response: This has been addressed.
Comment: Page 11: I wasn't convinced by the suggestion that it would be easier for local community residents to have more knowledge of the different aetiologies of stroke if these were classified as two different types of stroke, rather than one disease with two types.
Response: I acknowledge the fact that this idea is debatable. However, previous engagements with some Ghanaian communities suggest that community residents think in terms of different disease conditions rather than different types of the same condition. They sometimes find it confusing to distinguish between types of a particular illness. Hence, in order to simplify identification of stroke types by 'non-health professionals', I suggest that the possibility of classifying them as two separate conditions should be looked into by the medical community. This is an interesting qualitative study on the cultural meanings of stroke based on 30 focus groups in 5 areas of Ghana which used 13 interviewers and with the recorded interviews in five languages translated verbatim to English. The review of literature is adequate. I have a few comments -I will deal with the minor ones first and then the more substantive one.
p.5 line 4 closely knitted: I think should be close knit. It is clear that this group are Muslim in contrast to the other groups and this is not mentioned, neither is there any analysis that draws on this. Perhaps this is 1,2 1 2 p.5 line 4 closely knitted: I think should be close knit. It is clear that this group are Muslim in contrast to the other groups and this is not mentioned, neither is there any analysis that draws on this. Perhaps this is relevant? I would suggest that the article could be improved by reorganising the paper so that it follows the topic guide less in terms of order. The community views of stroke could be added to the first part of the results about thinking about stroke which then gives more of a coherent form to the paper.
I think the results do not make the most of the data and occasionally are marred by a judgement that medical thinking is superior. This sentence demonstrates a judgement that this view is held because people did not know about stroke. This is at varience with the idea that social and cultural meanings of stroke are valid and important.....
'In addition, the Ga Mashie participants referred to stroke as a ghost illness -a term developed "sisa hela" in the 1960s when community knowledge on causes and management of stroke was low.' I would suggest you remove the sentence which ends 'lay model in order to put the onus of recovery on the patient' as it makes no sense.
It would be useful to distinguish between diagnosis of disease as part of the biomedical model and experience of illness as part of the sociological or anthropological understandings and experience of conditions. Alan Young's work is relevant here. There is clearly a difference between natural and spiritual strokes in terms of ideas of causation. This I think from a close reading is similar to other studies of plural ideas about illness and treatment. I think it should not to be conflated with the right and the left or mild and severe. So I would suggest you move the natural and spiritual to the beginning of the section as this is about causation and then the characteristics follow. Also the treatment for the natural and spiritual may be different. This is a pluralistic way of thinking about stroke so that these ways of thinking and treating are often combined. There can be multiple ideas about causation -for the spiritual -The idea that it is a ghost illness is an important one as are the ideas about being cursed or of witchcraft. For the natural ideas for reasons that include diet, worrying, and hypertension are also important Views on Severity and treatment are something else. So I would suggest that the results are reordered with careful attention to the data and the story that is emerging.
The ideas for intervention are unformed but could be developed further but perhaps not in this paper.
These revisions will improve the paper which is interesting and should be published.

If applicable, is the statistical analysis and its interpretation appropriate?
Yes superiority of medical thinking over that of the local communities. Sentences which demonstrate that medical thinking is superior have been reworded. For instance, the above sentence was actually provided by the participants. The sentence has been reworded to show that this was an explanation given by the participants and not the author's interpretation/judgement. The new sentence now reads as: In addition, Ga Mashie participants referred to stroke as a ghost illness "sisa hela"; they mentioned that this term was developed in the 1960s when community knowledge on causes and management of stroke was low.
Comment: I would suggest you remove the sentence which ends 'lay model in order to put the onus of recovery on the patient' as it makes no sense.

Response: This sentence has been taken out.
Comment: It would be useful to distinguish between diagnosis of disease as part of the biomedical model and experience of illness as part of the sociological or anthropological understandings and experience of conditions. Alan Young's work is relevant here. :

Response This comment has been addressed. A statement which brings out the difference between the biomedical diagnosis and sociological understandings and experience of illnesses has been included in the discussion section.
Comment: There is clearly a difference between natural and spiritual strokes in terms of ideas of causation. This I think from a close reading is similar to other studies of plural ideas about illness and treatment. I think it should not to be conflated with the right and the left or mild and severe. So I would suggest you move the natural and spiritual to the beginning of the section as this is about causation and then the characteristics follow.