Living with consequences of stroke and risk factors for unhealthy diet- experiences among stroke survivors and caregivers in Nairobi, Kenya

A cross sectional study using qualitative and quantitative methods involving two rounds of data collection.

In November 2018, the Stroke Association of Kenya (SAK) invited their members (i.e., persons who had had a stroke, caregivers, spouses, parents, children, siblings and/or friends) to participate in a half day seminar about stroke in a location in Nairobi organized by SAK. The invitation also included information about the possibility of participating in a study about living in Kenya after suffering a stroke. The inclusion criteria for the participants in the study were: 1) stroke diagnosis 2) no psychiatric diagnosis, 3) ability to comprehend and reply to instructions in English or in the local language.

After the presentation within the seminar, all people with stroke living in the community that attended the seminar were provided written and verbal information about the study and agreed to participate (n = 30) in the first round of data collection. All the participants were members of the SAK and participated voluntarily.

In the first round of data collection persons with stroke (n = 30) were interviewed individually by students from Kenya Medical Training College (KMTC) in Nairobi who had been trained for the purpose by the authors SG and LvK. During the interviews, a self-reported questionnaire developed by the authors (LvK, GE, SG) and previously used in Uganda was used to collect the sociodemographic data including age, sex, marital status, handedness, number of children, work status, level of education and housing. Also information on type of stroke, affected side of the body, medication for hypertension and rehabilitation received was asked for during the interview by the person with stroke which also was confirmed by the present significant others (see Supplementary material 1). The Barthel Index (BI) was used to establish the level of dependence in activities of daily living (ADL) [13]. The persons ‘perceptions of the impact of stroke was collected using the Stroke Impact Scale (SIS) 3.0 Uganda version [14]. The scale includes eight domains of functioning: Strength, Hand function, ADL/Instrumental activities of daily living (IADLs), Mobility, Communication, Emotion, Memory and Thinking, and Participation. The SIS aggregated scores range from 0 (indicating more problems in everyday life) to 100 (indicating fewer problems). The SIS 3.0 Uganda version also includes one item to assess the participant’s perceived overall stroke recovery to be reported on a visual analogue scale ranging from 0 = no recovery to 100 = full recovery.

For the second round of data collection conducted in February – March 2019, the first author (ME) contacted the participants who had provided their telephone numbers during the first round of data collection. They were additionally informed about the study and asked if they were willing to participate in an interview that would take approximately 60–90 min. Nine persons who had had a stroke agreed to participate and the interviews took place in their homes or at the campus of KMTC. Six of the nine participants were accompanied by their caregivers that also participated in the interviews. Therefore, both verbal and written consent was sought from both the person with stroke and their accompanying caregivers. A semi-structured interview was conducted by the first author that focused on the experience and nutritional aspects of living with the consequences of stroke in Kenya (see Supplementary material 2) as well as anthropometric measurements and data on type and frequency of food consumption were collected. Those persons travelling to KMTC received reimbursement for their travelling expenses.

Anthropometric measurements of height, weight, and waist circumference (WC) of the person with stroke were collected. Height was measured using a steel tape attached to a flat wall. The person with stroke was asked to stand to the best of their ability on a flat surface against the wall. The measurement was taken at a point perpendicular to the head, indicated by a wooden block, to the nearest 0.1 cm. Weight was measured to the nearest 0.1 kg using the BMI Beurer SR BF2 scale and BMI indicating normal weight of 18.5–24.9, overweight of 25–29.9 and obese weight over 30 [15]. WC was measured to the nearest 0.1 cm using a tape measure. The risk of developing cardiovascular disease (CVD) increases with a WC of over 88 cm in women and 102 cm in men [15]. Social and demographic data and data on risk factors for stroke and eating habits were collected. To obtain information and guide the qualitative interview on the type and frequency of food consumed, a food frequency questionnaire (FFQ) from the Kenya National Clinical Nutrition and Dietetics Reference Manual was used [16]. In order to capture the participants’ experiences and perspectives, an interview guide that provided guidance on questions regarding the pathway from stroke onset to the time of the interview, the consequences of stroke in everyday life activities, nutritional status and eating habits was used. The interviews were digitally recorded and transcribed verbatim.

To describe the sample regarding demographic and clinical characteristics, functioning, anthropometric measures and perceived impact of stroke descriptive statistics were used. An algorithm [17] was used to produce the SIS aggregated scores in each domain. The qualitative data were analysed by the first author using a constructivist grounded theory (GT) approach [18]. GT uses an inductive approach to data collection that is non-linear and should be seen as flexible guidelines. Accordingly, the analysis was carried out in two phases and, in the first phase, initial coding was conducted using code labels very close to the text. The second phase used a systematic constant comparative method including moving back and forth in the text and to the emerging codes to analyse and continue categorising. Two additional interviews with new participants were added to reach saturation in the categories. Memos describing the findings and reflections from the interviews and emerging categories were written throughout the process. The analysis resulted in one core category and five sub-categories. The qualitative and quantitative findings of this study were then compiled and synthesised, starting with a structure based on the qualitative categories that emerged using GT constructivist analysis [18] and supported by the quantitative data. Throughout the data analysis process, the last author (SG) repeatedly asked important questions to enhance the credibility of the results. To improve the resonance and as a triangulation of the findings [18], the different categories and subcategories were discussed between the first author (ME) and SG. As a final step in the analysis, these results were discussed with all authors, where the third (AB) and the fourth author (JA) represented the Kenyan context of cultural competence, as a form of informal triangulation [19].

The socio demographic data and the clinical characteristics are presented in Table 1.

The impact of stroke assessed by use of SIS 3.0, Uganda version is presented in Table 2.

The demographics for the nine participants participating in second data collection, including anthropometric measurements, are presented in Table 1, right column. The socio-economic status was mixed, ranging from slum-dwelling participants to owning several properties. However, most of the participants had a low socio-economic status (SES). The persons with stroke all stated that they had been seriously affected by their stroke and had experienced lengthy hospital stays ranging from two weeks to several months. The participants and their households, all followed the Kenyan staple diet according to the FFQ with the majority of participants consuming white maize at least twice a day.

The mean BMI fell in the overweight range and the lowest BMI was 24. The WC of all five women and two men fell within the increased risk of developing CVD [15].

The self-rated data from the nine participants with stroke of which four were males showed that regarding information provided from health professionals, five of the participants had received information on stroke, three of them had receive information on risk factors for stroke, two of them had received information on possible beneficial effects of diet while seven of them had been provided information from the SAK, relatives, friends or from a nutrition student. Eight of the participants had experienced difficulties chewing or swallowing food or drinks at onset and three of them still experienced dysphagia at time of the interview.

They reported that six caregivers were the main facilitator of their healthy diet. In seven out of nine cases, the person with stroke was fully dependent on their caregiver to manage everyday tasks. Caregivers assisted with dressing, walking, taking a bath, going to the bathroom, cooking, feeding and drinking. The caregiver was also the main facilitator of maintaining social relations, procuring information on diet and care, managing the household economy including the food budget, doctors’ appointments, medication, rehabilitation services, as well as school fees for children. The role of the caregiver differed depending on the participant’s gender. The spouse of the male participants was the primary caregiver. None of the spouses of the women were primary caregivers.

After compilation and analysis of the qualitative and quantitative data from the participants who have had a stroke (n = 9) and caregivers (n = 6) one core category, and five sub-categories emerged. The core category was: The caregiver as the main definer of health and enabler of a healthy diet among persons who have had a stroke. The sub-categories were: (1) The perspectives on health as central in the recovery process, (2) The caregiver as the collector of information on healthy diet, (3) The caregiver as the manager of the person’s with strokes limitations in eating (4) The caregiver as the manager of the underlying modifiable risk factors, and (5) Need of compromised healthy diet due to loss of income.

The persons with stroke stated that they had difficulty feeding themselves, either due to apparent dysphagia, upper extremity paresis or mood changes, for example, depression. A clinical assessment of dysphagia appeared to be lacking even though it is recommended for all stroke patients. Also, dietary modifications have been standardised [10, 18]. In a large review on nutrition in stroke patients, Corrigan et al. found that early detection and treatment of malnutrition – common in Kenya and often a finding in persons with stroke – can affect the person’s ability to participate in rehabilitation and recovery [10]. The apparent lack of clinical assessment of dysphagia in Kenya has been studied and a study assessing the knowledge on stroke care in healthcare professionals in SSA was conducted at one of Kenya’s largest referral hospitals [20]. The study was the first of its kind and the authors conducted a cross-sectional survey of 300 healthcare professionals, of which 199 responded. The findings showed a wide range in knowledge on stroke care and only 25% of responders reported performing dysphagia tests while 30% reported never administering tests [20]. 50% of responders reported documenting a lipid profile of the patients and the same number attended diabetes screening. However, our study had too few participants to determine whether the results correspond with the findings of Lin et al. The experiences of the participants in this study with healthcare professionals were similar to the results of the survey [20]. The authors described the positions held by the responders to the survey and highlighted how stroke care was administered by a single neurologist.

Eight participants with stroke had a BMI indicating overweight. All five women and two men had a WC indicating increased risk for CVD. Overweight is part of the double burden of malnutrition that plagues Kenya and is probably an outcome of various lifestyle factors [4]. In 2012, Mbochi et al. published a study aimed at determining the predictors of overweight and obesity in adult women (n = 365) in Nairobi, Kenya and showed that BMI increased with increasing household expenditure [21]. No such study of a male population has been found, possibly because obesity and overweight are more prevalent in women [4, 20]. Although this study is too small to confirm the findings of Mbochi et al., the WCs of all the women put them at risk of CVD. As overweight and obesity are intertwined with other risk factors such as diabetes and hypertension, tackling the increasing prevalence of overweight in the population is perhaps a smart starting point for managing NCDs in Kenya.

This albeit small study raises questions that will need to be further explored in order to develop and implement both the national and community-based interventions that are needed to prevent and manage stroke and a healthy diet in Kenya [4]. Kenya is currently one of seven African countries that have developed food-based dietary guidelines, but the lack of nutritional specialists and poor access to healthcare services makes the effect of these guidelines questionable [1, 5]. There is much to learn from how HIC manage NCDs such as stroke. However, there is a need for LMIC such as Kenya to devise their own systems [8]. Unlike many HICs in which these lifestyle changes have occurred over several decades, LMIC countries experience these changes rapidly. Kenya, which has tripled its population over the last 35 years, does not have an extensive healthcare system and knowledge of stroke in the population is low. Thus, developing a national prevention programme poses multiple political and economic challenges [8]. However, this study aims to fill the knowledge gap that will enable policymakers and HCPs to better tailor the secondary prevention of stroke and provide advice on diets and physical activity in a Kenyan and SSA setting.

This study has explored and described a small sample of people living with the consequences of stroke in Nairobi and showed that the structural issues hindering a healthy diet are generalisable to many aspects of health care in LMIC [1, 21]. The participants were all members of the SAK and, at the time of the interviews, had received some information on stroke and risk factors. It is to be assumed that large numbers of people suffer stroke in Kenya, particularly in rural areas, and that these people never get access to the same level of health care or information. Essential healthcare services remain beyond the reach of most Kenyans [3, 22]. If this study were to be repeated with participants who represented different geographical areas of Kenya and who were not members of the SAK, it is possible that eating habits and factors affecting a healthy diet would differ from the results of this study. However, it is not unreasonable to assume that this study population has given a brighter picture of eating habits and the knowledge that exists within the community of people living with the consequences of stroke and their households than what actually exists in a more representative group. The interviews were conducted with English-speaking persons in a country that has two official languages (English and Kiswahili) and 47 ethnic groups. Although the participants represented several different ethnic groups, if the study were to be repeated, it would be of interest to include a wider demographic. Diet varies between ethnic groups and if possible, it would be of great interest to study these variations in relation to the development of stroke and other NCDs.

There have been several ethical considerations for this study even though the participants with stroke and their caregivers participated voluntarily and on their own initiative. Firstly, as stroke caused dramatic changes to the life of the participants and their households, the subject itself is sensitive. Secondly, people living in poverty are often ashamed about their situation. During the interviews, every effort was made to be sensitive to such aspects. As the main point of focus of the interviews was on nutritional aspects, not to value anything else that was said during the interview during the interviews. The reason for this is two-fold: firstly, in order to ensure that the interviewees did not feel they were being judged about their diets, and secondly, to try to avoid the interviewees describing a diet that they believed to be healthy, instead of what they were actually eating. As all information on diet has been self-reported, both under- and overreporting is to be expected. Underreporting often occurs in relation to alcohol consumption which, in Kenya, is frowned on in particular, as well as meat consumption [21]. Many of the interviewees requested individually tailored dietary advice from the first author at the time of the interviews. These persons were encouraged to use some of the information given in the Kenyan Dietary Guidelines. As most of them did not have internet access, the guidelines were sent to the SAK to relay to its members.