Abstract
Background. Health professionals assign diabetes patients “homework” in that they give them instructions on how to manage diabetes, recognizing that most diabetes care takes place in the home setting. We studied how homework is practiced and whether knowledge and behavioral practices related to diabetes self-management diffuse from patients to their housemates. Method. This mixed-methods study combined quantitative data from a household survey including 90 rural Ugandan households (50% had a member with type 2 diabetes [T2D]) with qualitative data from health facilities and interviews with 10 patients with T2D. Focus for data collection was knowledge and practices related to diabetes homework. A generalized mixed model was used to analyze quantitative data, while content analysis was used for qualitative data analysis. Results. Patients with T2D generally understood the diabetes homework assignments given by health professionals and carried out their homework with support from housemates. Although adherence to recommended diet was variable, housemates were likely to eat a healthier diet than if no patient with T2D lived in the household. Knowledge related to diabetes homework diffused from the patients to housemates and beyond to neighbors and family living elsewhere. Knowledge about primary prevention of T2D was almost absent among health staff, patients, and relatives. Conclusions. Homework practices related to T2D improve diabetes-related knowledge and may facilitate healthy eating in nondiabetic housemates. These findings suggest that having a chronic disease in the household provides an opportunity to improve health in the entire household and address the lack of knowledge about prevention of T2D.
Keywords
The prevalence of type 2 diabetes (T2D) in selected rural and urban areas in Uganda has already reached levels comparable to many European countries, with observations of up to 7.4% and 8.1% of adults above 35 years of age having the disease, respectively (Lasky, Becerra, Boto, Otim, & Ntambi, 2002; Mayega et al., 2013). This is a challenge to both the individual with T2D and the health system in a country already heavily burdened with communicable diseases (World Health Organization, 2008) and financial constraints. Furthermore, in contrast to the HIV/AIDS epidemic, no bilateral or multilateral aid (PEPFAR, 2014; The Global Fund, 2014; USAID, 2014), and few nongovernmental organizations have compensated for the lack of governmental programs, awareness campaigns, or prevention of future cases of T2D.
We recently demonstrated that members of diabetic households in rural Uganda had higher levels of diabetes-related knowledge and a healthier cardiometabolic profile than those from nondiabetic households (Nielsen et al., 2015), suggesting that knowledge and healthy behaviors related to diabetes self-management may diffuse from the patient to other household members. This may be an important unintended benefit for rural African settings, particularly given the lack of public awareness campaigns or governmental diabetes education and prevention programs, the low recognition of diabetes (Peck et al., 2014; Whyte, 2012), and limited access to Internet, television, and newspapers, and to some extent radio, for health education.
The potential for a shared home environment to serve as a driver for positive influences in chronic disease management and prevention is supported by studies from the United States showing that untreated spouses experienced weight loss (Gorin et al., 2008; White et al., 1991) and improved their diet (Shattuck, White, & Kristal, 1992) when sharing a household with a spouse enrolled in a lifestyle intervention program. Similarly, a study from the United States found a higher level of diabetes-related knowledge in relatives of patients with T2D as compared to individuals with no family history of the disease (Baptiste-Roberts et al., 2007). Other studies, however, have found a higher diabetes risk in relatives and spouses compared with individuals with no spousal or family history of T2D (Hemminki, Li, Sundquist, & Sundquist, 2010; Leong, Rahme, & Dasgupta, 2014; Wilson et al., 2007). Either way, patterns of eating and physical activity cluster among family members living in the same household (Pachucki, Jacques, & Christakis, 2011; Sonneville et al., 2012).
Optimal management of T2D requires a complex and demanding behavioral management regimen involving daily monitoring of symptoms and blood glucose levels, adherence to medication, and often significant changes in diet and physical activity level (Gonder-Frederick, Cox, & Ritterband, 2002). Due to the comprehensive regimen, most of the diabetes management takes places outside the health system and patients and their families are thus responsible for the vast majority of the diabetes care (Rintala, Jaatinen, Paavilainen, & Astedt-Kurki, 2013). To self-manage the disease, patients with diabetes (Nicolucci et al., 2013) and their family members (Kovacs Burns et al., 2013) are to some extent informed and educated by health workers.
The tasks or assignments of a disease treatment regimen patients and their families are expected to carry out in the home setting are conceptualized by Mattingly, Gron, and Meinert (2011) as “chronic homework.” Such tasks include taking (and giving) medicine, monitoring symptoms, exercising, and following a specific diet. The homework is assigned by experts in the clinic, but carried out by the patient and the family at home. Thus, outside the health system, homework changes its status from a set of distinct technical tasks (as the professional health staff envisions) into an extensive social project that generates new and unanticipated demands and relationships among family members (Mattingly et al., 2011). In this study, we adopt the concept of homework, and our interests in the dynamics of the “social project” focus on homework’s consequences for knowledge and health practices in the domestic setting where the homework is undertaken. Thus, in this study, homework is operationalized as the tasks related to management of T2D about which patients are educated in the professional health sector and which have to be carried out in the home setting.
Essential components of the homework related to T2D such as balanced diet and increased physical activity (Gonder-Frederick et al., 2002) can also prevent T2D (Knowler et al., 2009; Lindstrom et al., 2006; Pan et al., 1997; Ramachandran et al., 2006). Yet there has been little research on how family members may learn and benefit from participating in the homework and from supporting a chronically ill family member.
To determine whether the homework assigned to patients with T2D facilitates diabetes-knowledge flow from the health system to the patients and in turn to the others in the household, we conducted a mixed-methods study in southwestern Uganda, comprising a household survey combined with qualitative methods. In addition, we explored if there is a relationship between the practices of homework and our previous findings with respect to a higher level of diabetes-related knowledge and a healthier cardiometabolic risk factor profile in diabetic households.
Method
From November 2011 to April 2013, fieldwork was carried out at health facilities providing diabetes treatment and in 90 rural households (sampling is described elsewhere; Nielsen et al., 2015) in Kasese District in southwestern Uganda. To answer the research questions, the study adopted a mixed-methods approach with a primarily explanatory design, which is a design where potential relationships are assessed with quantitative data, but explained by qualitative data (Creswell & Plano Clark, 2011). Thus, in the explanatory design with quantitative data we aimed at determining the level and areas of diabetes-related knowledge, while qualitative data were collected to explain potential mechanisms of knowledge flow and to clarify and add texture to the quantitative data. Furthermore, the quantitative data should test the generalizability of the qualitative data. The mixed-methods approach was also used as an exploratory design where qualitative data were used to develop the quantitative tools (structured questionnaires) used in the study (Creswell & Plano Clark, 2011). All data were collected by the first author in English assisted by local research assistants who were fluent in the local language, Lukonzo, and English.
Study Area
Kasese District was chosen because Kagando Hospital, a private not-for-profit hospital, has been offering diabetes treatment since 1965 and had diabetes patient files dating back to 2004. In addition, the hospital offers a weekly diabetes clinic, where the hospital provides patients with T2D with “homework.”
In 2012, the population of Kasese District was estimated at 747,800. The majority of the inhabitants belong to the Bakonzo ethnic group, who traditionally inhabited the Rwenzori Mountains on the borderlands of Uganda and the Democratic Republic of Congo. About 48% of the population in Kasese District lives below the poverty line, and 67.6% are literate (Kasese District Local Government & Uganda Bureau of Statistics, 2012). The population is young (57.5% under 18 years of age; Kasese District Local Government & Uganda Bureau of Statistics, 2012), the fertility rate is high (7.4%), and polygamy is common (Reno, Twinamasiko, & Mugisa, 2012). The majority of the inhabitants in Kasese District live in rural areas (75.3%), and approximately 80% depend on small scale subsistence farming with cassava, sweet potatoes, matoke (plantain), and cash crops such as coffee as the main crops (Reno et al., 2012). The majority of people live in houses made of mud or unburned bricks with iron sheet roofs and with no electricity or piped water. Land is inherited patrilineally, and due to the high population growth, the size of land holdings is decreasing rapidly (Kasese District Local Government & Uganda Bureau of Statistics, 2012).
Data Collection
This study included data from three different sources (see Table 1 for an overview of data sources):
Observational data included the following: (a) general observations made over the course of 14 months in the study area, including informal conversations with patients at the diabetes clinic and staff at Kagando Hospital. In addition, 12 health facilities, 3 pharmacies, 10 drug shops, and 2 private clinics were visited, yielding general information about health-seeking patterns and possibilities for patients with T2D in the district; and (b) observations at the diabetes clinic at Kagando Hospital to investigate the content of the diabetes education at the hospital, and thus, the specific task and instructions of the homework assigned to the patients at the weekly diabetes clinic at the Hospital.
In-depth interviews with (a) the diabetes nurse in charge of the clinic to investigate the content of the diabetes education at Kagando hospital and (b) with 10 patients with T2D in their homes to explore practices of homework and family support for it in the home setting. For the patient interviews, an interview guide with open-ended questions was developed based on observations at the diabetes clinics and informal conversations with diabetes patients and other people. The interview guide was organized into five domains: (a) life history, (b) diagnosis of and life with diabetes, (c) treatment of diabetes, (d) knowledge and self-care of diabetes, and (e) support in managing life with diabetes. The 10 patients were selected from the household survey in which the patients with T2D had been diagnosed with diabetes for a minimum of 2 years, were 40 years of age or older at the time of diagnosis, and had attended at least two diabetes consultations at Kagando Hospital (Nielsen et al., 2015). From this cohort the 10 patients were purposefully selected to represent both genders and different levels of control of T2D (defined by glycosylated hemoglobin [HbA1c] according to Ali et al., 2013).
Quantitative data from a household survey of 90 households on (a) demography and socioeconomic factors and (b) diabetes-related knowledge in individuals without known diabetes were collected using a structured questionnaire. The development of the questionnaire was based on observations at the diabetes clinic at Kagando Hospital and should thus reflect the content of the diabetes education. These data were collected as part of the household survey, described in detail elsewhere (Nielsen et al., 2015). In short, 90 households were surveyed, of which 45 had a resident with diagnosed T2D; 392 nondiabetic individuals ≥13 years of age, who shared food on a daily basis and had lived in the household for at least 3 months prior to the study, were included. Socioeconomic status (SES) and educational level were higher in diabetic households than nondiabetic households (p = .001 and p = .003, respectively), but no differences were seen in relation to sex or age (p = .268 and p = .923, respectively).
Overview of Data Sources, Themes, and Key Findings.
Data Analysis
Qualitative Data
Ongoing data analysis (Pope, Ziebland, & Mays, 2000) was used for the exploratory design phase where qualitative data were used to develop the quantitative questionnaires and interview guide for in-depth interviews.
All qualitative data analyses followed the principles of content analysis (Graneheim & Lundman, 2004), reading through the interviews and manually organizing and coding data into analytical units, meaning units, and themes. We identified and coded for four themes concerning households, education, homework, and knowledge. During the analyses the authors often returned to the full interviews to ensure that the subinterpretations were in agreement with the overall context of the interviews. Patterns of the data are reported and direct quotations are retained to provide examples of people’s responses.
Quantitative Data
We previously showed that household members to a patient with T2D have a higher level of diabetes knowledge as compared to individuals not having T2D in their household (Nielsen et al., 2015). Thus, to examine underlying patterns of diabetes-related knowledge in individuals without known diabetes (160 and 232 individuals from diabetic and nondiabetic households, respectively), these individuals were asked whether 10 statements were true or false. These statements were chosen to represent different aspects of diabetes knowledge: perceptions of the chronicity of T2D, the importance of self-management, and knowledge related to homework. The answers were coded as incorrect (0) or correct (1) based on the content of the diabetes education at Kagando Hospital. A generalized mixed model was used to analyze the binary outcome variable—knowledge—as a function of living in a diabetic household including sex, age, SES, and education as covariates. Household was used as a random effect to account for within-household clustering. In case of missing values, the missing answer to a specific question was excluded from the analysis, whereas valid answers to other questions from the same individual were included in the subsequent analyses.
Ethics
The aim of the project was introduced both verbally and in writing to the participants. A written informed consent form was read aloud in the local language Lukonjo, and participants agreeing to participate either signed or put their thumb print on the form. The participants were carefully informed that declining participation would not affect their current and future treatment options at Kagando Hospital. The study protocol was approved by The Uganda National Council of Science and Technology, Makerere University School of Medicine Research & Ethics Committee, St. Raphael of St. Francis Hospital Nsambya and Kagando Hospital. In publications, patients have been given pseudonyms to preserve confidentiality.
Results
Results are presented according to four themes supported by quantitative and qualitative data: (a) Patients and their households: Characteristics, relationships, and responsibilities; (b) Diabetes education and homework assignments; (c) Diabetes-related knowledge in nondiabetic individuals; and (d) Involvement in homework and flow of diabetes-related knowledge. The last theme includes three subthemes: “Food and exercise practices in the diabetic households”; “Blood glucose, medication, and clinic attendance”; and “Perceptions and knowledge about diabetes risk factors and prevention.” Table 1 provides an overview and shows from which source the data on each theme were obtained.
Theme 1. Patients and Their Households: Characteristics, Relationships, and Responsibilities
The characteristics of the 10 patients participating in the qualitative interviews and their households are described in Table 2. In short, the 10 patients had lived with diabetes for 3 to 31 years (median: 11 years), and only 3 had a level of HbA1c or complications corresponding to good control of their diabetes. None of the patients could measure their blood glucose levels at home and would only have this measure conducted when visiting a health facility. From the patient files at Kagando Hospital (dating back to 2004), the observed median number of diabetes consultations attended at Kagando Hospital was 25 (range: 3-64); patients with a higher number of consultations were also the patients who have lived longest with the diabetes diagnosis. The mentioned patient characteristics did not differ statistically between the 10 patients taking part in the qualitative interviews and the overall patient group measured in the household survey (p > .05 for all tests, data not shown).
Characteristics of the 10 Patients With Type 2 Diabetes and Their Households.
In line with the inclusion criteria, the 90 households in the survey had at least two generations living together—parents and children or grandchildren and grandparents; some included three generations. More women than men lived in the households (57.2% vs. 42.8%), which was primarily due to especially younger men studying or working in towns. Polygamy was common and one of the men with T2D had 10 wives. In general remarrying was frequent, but women who remarried did so when they were still in their reproductive age. Women did the cooking in 88 of the 90 households, and the cooking duty was most often given to a younger woman who had finished school.
Theme 2. Diabetes Education and Homework Assignments
From observations at Kagando Hospital diabetes clinic and the interview with the diabetes nurse, we learned that the educational component at the weekly diabetes clinic was focused on homework-related tasks like nutritional instructions, advices related to exercise, smoking cessation, awareness of symptoms of hypo- and hyperglycemia, and intake of medicine. The diabetes nurse illustrated his instructions simply, for example, by using his palm to show patients carbohydrate and fruit amounts and the tip of the little finger to show recommended salt usage during cooking. Some foods were totally banned, such as sugar, sodas, packaged foods such as biscuits, and in some cases salt (if the patient also had hypertension), while others—cooking oil, meat, and salt—should be limited. In relation to exercise, the patients were encouraged to walk instead of taking a boda-boda (local motorbike taxi) and told to do “digging” (gardening/farming) and household chores daily. In terms of adherence to treatment, the patients were told always to take the medicine at the same time of the day and always to eat immediately after.
All 10 patients participating in the qualitative interviews reported that they had received their first diabetes education at the time of diagnosis. In relation to homework, all the patients recalled being told about “restricted food,” which was food they could not eat at all or food they could only eat in very limited quantities. Of food to avoid, sugar, sugary foods, cooking oil, and fried food were mentioned by the majority of the patients. Some of the patients had also been advised to avoid starchy foods. Two of the starchy foods that were restricted were staples of the local diet: matoke (plantain bananas) and ubunde (cassava with millet); however, it was recommended that ubunde could be prepared using primarily millet.
Patients reported that they were taught to take care and not to cut themselves when digging or fetching firewood, since wounds are difficult to heal for people with diabetes. In terms of exercise, the majority of the patients said they were instructed to walk or “dig” every day: “We should avoid hard work from morning to night, but we should work at least one or two hours per day. But not before taking drugs” (Suzan).
No major differences were found between the lessons observed at the diabetes clinic, the interview with the diabetes nurse, and in the 10 patients’ recall of the homework assignments. However, we noticed a disagreement regarding fruit, where the diabetes nurse told the patients to limit those food items, while some patients recalled that they could not eat fruit at all.
Theme 3. Diabetes-Related Knowledge in Nondiabetic Individuals
The results from the household survey showed that overall, members from diabetic households had a higher knowledge than members from nondiabetic households; 73.9% (95% confidence interval [CI] [68.7, 79.2]) of the diabetic household members assessed the knowledge statements correctly versus 66.8 (95% CI [57.9, 75.7]) of members from a nondiabetic household (see Table 3). More specifically, members of diabetic households were more likely to know the correct answer on 4 of the 10 statements (p < .05). Specifically, knowledge differed for questions related to “comorbidities of diabetes,” “how to handle low blood sugar,” “if sugary foods affect blood sugar,” and “if exercise is good for a patient with T2D.” After additional adjustment for SES and education, the difference found in relation to “how to handle low blood sugar” was not statistically significant (p = .207). In none of the other questions did additional adjustment for SES and education change the findings.
Comparison of Diabetes-Homework Related Knowledge in Individuals Without Known Diabetes From Diabetic and Nondiabetic Households.
Note. OR = odds ratio; CI = confidence interval; T2D = type 2 diabetes.
Age and sex adjusted OR of answering correctly according to household status. bStatistically significant at .05 level. cAdjusted for sex and age. dData missing for one individual from a diabetic household. eData missing for one individual from a diabetic household.
Theme 4. Involvement in Homework and Flow of Diabetes-Related Knowledge
From the in-depth interviews with the patients, two sources of diabetes-related knowledge for household members were documented: (a) via the hospital when the patient with T2D was admitted and (b) from the patient himself or herself. No family members had received diabetes-related education targeting family members. This information was consistent with observations at the diabetes clinic, where family members were not observed to be educated about diabetes or to accompany the patient to the clinic—unless the patient was very ill.
All patients reported that the best informed household member was a female. For male patients it was their wife, whom they had educated or who accompanied them at the time of diagnosis and/or admission, while for female patients it was primarily a daughter, granddaughter, or daughter-in-law. However, the patients also talked about their diabetes with the other members of the household.
In relation to support for the homework, the same gender pattern appeared; the majority of family members mentioned were wives, daughters, granddaughters, and daughters-in-law. Nevertheless, two male patients also told that their sons were involved in accompanying them to the diabetes clinic or in more social support. Otherwise, male family members were mostly mentioned in relation to financial support.
Patients of both genders reported that they also educated family members not living in the same household, and some patients were accompanied to the hospital by their children living elsewhere.
Patients at the diabetes clinic confirmed that the diabetes-related education of family members relied on the patient. To take on this role, some requested tools like the posters used in the clinic to educate their family members.
In relation to flow of knowledge, 5 of the 10 patients also mentioned neighbors and friends as people they would talk to about diabetes: “Sometimes some neighbors visit me and ask me about sukali [diabetes] and I teach them only to take little sugar, and use less cooking oil” (Josephine). One patient had family members living outside his household bringing him special diabetes foods (green leafy vegetables and millet), which indicates that the knowledge about food diffused to family members beyond the household.
Subtheme 4.1. Food and Exercise Practices in the Diabetic Households
None of the 10 patients participating in the qualitative interviews did the cooking. Yet they had all educated a female household member to carry out the homework instructions when preparing diabetic food. Often the cooking of diabetic food was integrated in the “normal” cooking procedures by preparing all food without or with very little cooking oil, salt, and sugar. A portion would then be taken out—the diabetic food for the patient—and the mentioned ingredients were added to the rest of the food to be eaten by the household members without diabetes.
Homework was adjusted to daily life practices in different ways, mostly in relation to “prohibited food,” as the patients ate the food despite knowing it was banned. One reason for doing so was financial: “We should make ubunde with millet, but [millet] is expensive, and I only eat plain ubunde [made of homegrown cassava]” (Godfrey). Other reasons were the preferences or needs of other household members: For the family I allow a little cooking oil but for salt I’m strict. They prepare the sauce with very little and the others can add more if they like. For the happiness of the family, I allow them to prepare sauce with cooking oil. I don’t make them prepare sauce separately for me. (Tom) My wife is sick and will have her uterus removed, and I will not give her a hard time, so I take the ubunde without millet [only made of cassava], even though I’m not allowed to take it without. (Moses)
Most of the patients interviewed reported practicing exercise daily. However, from the patients’ narratives, exercise received less attention than diet and none of the patients shared stories about family members taking part in their exercise.
Subtheme 4.2. Blood Glucose, Medication, and Clinic Attendance
None of the patients had access to a glucose meter at home and only knew their blood glucose level by attending the diabetes clinic. The individual blood glucose values were not discussed in the educational session at the diabetes clinic, but symptoms of high and low blood glucose were often themes in the education. During the individual consultation, the health professional reviewed blood glucose values, asked the patient questions, and then prescribed medicine. We never observed a health professional discussing the values with the patient. Nevertheless, the majority of the patients knew the blood glucose numbers, and these values seemed to be of great importance to the patients: At the hospital they don’t tell the numbers—just write them. If they [the nurses] write large I can read them myself, but if they write too small, I have to ask my brother’s granddaughter or the neighbor’s child to read them because I do not see well. (Tom) The blood sugar should be between 4 and 7—but it should not be lower. (Jimmi, who was proud of his measurement of 4.1)
The importance of the blood glucose values was not limited to the patients, as two of the wives interviewed mentioned the blood glucose level as important to them: When Nelson returns from the hospital, I always ask what the number was. I ask before he even sits down and he then just gives me the paper with the number. If the value is 8, I get worried. (Wife of patient with T2D) When Godfrey comes back from diabetes clinic, I always ask him about the numbers. Godfrey tells me if the number is high or normal, but never the value. . . . These days it always goes up and I become sad. I never ask him about the numbers for pressure [blood pressure]. I’m only interested in the diabetes. (Wife of patient with T2D)
Another wife did not know the values of a high blood glucose level, but she knew the symptoms of her husband’s diabetes if he had not been taking his medicine: “He over-urinates, has poor vision, puffiness in the face and overeats.”
In relation to intake of medicine and adhering to the medicinal part of the treatment regime, all 10 patients said they were responsible for remembering this part of the treatment. However, one patient (Suzan) reported that despite herself being in charge of taking medicine, her granddaughter sometimes helped her: “When I’m in bed and my granddaughter has prepared food, she wakes me up and says, ‘Food is ready, take your medicine.’”
Regarding appointments at the diabetes clinic, only one patient (Suzan) indicated that the family kept track of her appointments, but in most households, the family members knew when the patients had been to a diabetes clinic.
Subtheme 4.3. Perceptions and Knowledge About Diabetes Risk Factors and Prevention
The majority of the patients told similar stories of how their family reacted when they realized that the patient had diabetes; they thought a diagnosis with diabetes was the same as a “death sentence”: “When I heard about the diabetes I felt anxiety and thought Peter was going to die and I would be a widow soon” (wife of patient with T2D). However, seeing the patients living with diabetes for many years seemed to change this perception of diabetes from a deadly disease to a disease that could be managed and lived with: “My children thought I was going to die, but now that I have spent many years with them, they don’t think I’m going to die” (Josephine).
Causes of diabetes were not taught extensively at the diabetes clinic, except for discussions about “Myth and Facts about Diabetes,” such as whether “diabetes is a curse.” There was no discussion of lifestyle-related risk factors and none of the patients mentioned such factors as causes of their diabetes. In terms of their diabetes, all the patients participating in the qualitative interviews referred to God. Nevertheless, despite the patients claiming that only God knew why they got diabetes, they did reflect on the causes of diabetes in other people: Diabetes may be caused by eating too much sugar, but I did not even take tea [in Uganda tea ideally contains a high amount of sugar]. (Nelson) It is rich people, who are the one to get it and the fat people, so I did not mind about it.(Margaret)
However, later in the interview Margaret said that “before I fell sick, I had gained some weight.
One family member (Nelson’s wife) had also reflected about the causes of diabetes: “We live in the same family and we eat the same—why do we not all get diabetes?”
The connection between causes and prevention of T2D was rarely mentioned by nurses at the hospital, patients and their families, or by other people in the study area. People were unsure about the link between prevention, lifestyle, and causes of the disease, as was illustrated by an observation in a pharmacy in the district capital, Kasese Town. When the pharmacist heard that we were conducting a research project about diabetes, she told us that two of her siblings had died from the disease and both of her parents lived with diabetes. She and her family asked themselves, “What is causing the diabetes? Is it what we eat? We eat cassava, cassava roots for lunch and bread made out of cassava flour for dinner. It is lots of carbohydrates.” She was obviously very afraid of getting diabetes: “I check myself more often [measuring the blood glucose level] because I’m a health worker. . . . Others in the family don’t check. Even malaria makes me scared [that I have diabetes].” Despite the fact that Janet was scared and she knew about diet and exercise advice for patients with diabetes, she had not linked the diet and exercise instructions for management of diabetes with prevention of the same disease: “It [the diet and exercise advice] is for that one. The one who is sick [the patient with diabetes] is supposed to do it alone. Me, I just forget. I don’t change. I don’t follow their diet.”
Discussion
This mixed-methods examination of the flow of diabetes knowledge from the health care system in rural Uganda to patients and to households yielded three key findings. First, patients with T2D registered at a diabetes clinic generally understood the educational messages and carried out their diabetes homework. Second, patients with T2D disseminate their knowledge to the other members in the household and beyond, which was found both in the qualitative and quantitative data. Thereby patients with T2D may be an unexploited opportunity to increase awareness and knowledge about T2D. Third, in contrast to the high level of knowledge related to self-management of T2D, knowledge about T2D prevention was almost completely absent in the study area.
In carrying out their homework, the patients reported the greatest difficulty with the eating assignments. This is consistent with studies from both sub-Saharan Africa (Hjelm & Nambozi, 2008; Kolling, Winkley, & von Deden, 2010) and high-income countries (Denham, Manoogian, & Schuster, 2007; Wen, Shepherd, & Parchman, 2004) finding that the diet part of diabetes self-management is the hardest part to integrate at home. Yet even though none of the patients fully conformed to the homework instructions about food, their household members were likely to eat a healthier diet than if no patient with T2D had lived in the household; due to the homework, they used less salt, sugar, and cooking oil, and more millet instead of cassava.
The existing studies of diet and diabetes self-management (Denham et al., 2007; Hjelm & Atwine, 2011; Kolling et al., 2010; Wen et al., 2004) did not examine potential dietary changes in the patients’ housemates. However, the assumption of potential positive effects from sharing a household with a person adhering to special dietary instructions is consistent with studies from the United States that found greater weight loss and decreased energy intake in untreated spouses of people participating in a lifestyle intervention (Gorin et al., 2008; Shattuck et al., 1992; White et al., 1991).
According to the patients’ narratives, the diffusion of diabetes-related knowledge happened through the family members’ involvement in the homework and social support. The knowledge diffusion had a gender bias, with women being best informed. The pattern of women being the main providers of informal health care in the home and men supporting primarily in financial matters is consistent with a Ugandan study of HIV/AIDS (Taylor, Seeley, & Kajura, 1996). In relation to diet, the quantitative data found that members of nondiabetic households had the same level of knowledge as members of diabetic households. The disparity between the knowledge related to food and the knowledge related to other types of diabetic homework may be explained by higher awareness of issues related to food for the people in the study setting. Thus, many of the participants had lived through times of food shortage; the population of Kasese District has been through two wars during the last 50 years during which people were displaced from their agricultural land (Kasese District Local Government & Uganda Bureau of Statistics, 2012) and Kagando Hospital previously had a ward for undernourished children. Also, the patient reporting to educate neighbors had told these people about dietary components related to the diabetes. In contrast, exercise was rarely verbalized by the study participants or patients unless probed by the investigators, and no patients mentioned family members supporting them in their exercise, which is in line with a study from Zimbabwe (Hjelm & Mufunda 2010), where family support was mentioned in related to diet, but not exercise. However, our findings contrast with a study from the United States where the patients reported family support in relation to exercise (Wen et al., 2004). Yet, from the quantitative data, we saw that more individuals from diabetic than nondiabetic households knew that exercise is good for patients with T2D. The lack of attention toward exercise despite a higher quantitative knowledge in this area may be explained by the participants in the present study having a relatively high level of physical activity as a natural integrated part of their daily life due to the mountainous setting and subsistence farming as the primary livelihood. Though the participants were aware of the importance of physical activity, none of them expressed concerns regarding lack of exercise. Another study from Uganda found that rural, less educated people were little concerned with exercise, while urban, younger, and more middle-class people were more aware of its health value (Whyte, 2014). In a study from Cameroon, urban citizens reported avoiding physical activity despite being aware of its importance as they perceived exertion like walking to be a sign of poverty (Kiawi et al., 2006).
Previous studies of diabetes-related knowledge in relatives to a person with T2D focused on risk factors for T2D and did not include knowledge related to the patient’s homework (Baptiste-Roberts et al., 2007; Cullen & Buzek, 2009). From a family member perspective, diabetes has been researched mainly in relation to the burden associated with living with a person with T2D both in high-income (Fisher, Chesla, Skaff, Mullan, & Kanter, 2002; Kovacs Burns et al., 2013) and low-income settings (Sabone, 2008). Strikingly, potential benefits of this situation for the healthy family members have received little attention. Thus, we only found one study, including 17 middle- and high-income countries, reporting a positive impact of living with a patient with diabetes on at least one dimension of life (Kovacs Burns et al., 2013). Although the study did not include African or other low-income counties, and did not specify which dimensions were most important, the result is consistent with the results of the present study and our recent survey (Nielsen et al., 2015).
Literacy varied between patients and households and within households. However, after adjusting for education and SES, quantitatively measured homework knowledge was still higher in diabetic than nondiabetic households. Furthermore, we observed no major differences between the 10 patients in the understanding of the homework despite a large range in years of education.
The lack of knowledge of causes of T2D and effective prevention approaches was prominent; not a single participant connected the dietary and physical activity principles of homework and self-management of T2D with prevention of T2D. The lack of knowledge of causes of T2D is consistent with another Ugandan study (Hjelm & Nambozi, 2008), but contrasts with a study from Cameroon where the majority of the participants mentioned avoidance of sugar as a means to prevent diabetes (Kiawi et al., 2006). The participants in the Cameroon study all lived in urban areas where educational level and access to information about causes and prevention of T2D may be higher than in the present rural study setting. Additionally, in both Tanzania (Aspray et al., 2000) and Kenya (Christensen et al., 2009), the prevalence of diabetes is higher in urban than rural settings, which may also contribute to a higher diabetes awareness and knowledge in urban than rural settings.
The lack of knowledge about prevention of T2D may explain why no relationship was found between the level of risk factors for cardiometabolic diseases and the level of diabetes-related knowledge in the household survey (Nielsen et al., 2015). According to “health belief models,” individuals must perceive themselves susceptible to a condition with potentially serious consequences and believe a course of action is available to reduce their susceptibility in order to change behaviors (Nutbeam & Harris, 2004). In our study most people said that “diabetes runs in families,” and the general perception was that diabetes was a serious potentially life-threatening disease, but the required knowledge of the possible actions to prevent the disease was absent. However, health belief models are criticized for the assumption that improvement in knowledge will be followed by healthier practices; in a review of health interventions in low-income settings, changes in knowledge were rarely followed by a change in behaviors (Yoder, 1997). Instead, the way homework was practiced may be more important than knowledge per se in relation to the diabetes risk factor profile in the household members and thus explain the healthier cardiometabolic profile in diabetic households than nondiabetic households (Nielsen et al., 2015).
Limitations of the Study
The patients with T2D were selected from a private not-for-profit hospital and had attended a minimum of two consultations. Those who had been able to pay and make at least two visits may not be representative for Ugandan patients with T2D in general. Furthermore, median time of living with diabetes was 11 years, and as older studies point to a mortality of up to 40% in the first 5 years after diagnosis (Castle & Wicks, 1980; McLarty, Kinabo, & Swai, 1990), the patients in the present study must have followed their homework and treatment to a certain extent or be healthier or less vulnerable to complications than patients who did not survive the first years after diagnosis. Therefore, these patients may be more likely to educate family members than patients not following the instructions. In addition, diabetic households had higher SES and educational level than nondiabetic households, which could have affected the level of diabetes-related knowledge positively due to greater experience with pedagogical exercises.
Conclusion
To our knowledge, this is the first study examining how patients with T2D can be a potential source for the dissemination of awareness and knowledge about diabetes in a low-income setting. Knowledge from patients with T2D diffuses within their households and may reach beyond to nonresident family members, to neighbors, and to other local contacts. The findings suggest that having a chronic disease in the household should not only be seen as a burden for the household but also as a potential for learning through homework. Although our informants did not speak explicitly about preventing diabetes, those who shared a household with an experienced patient benefitted from healthier eating practices. Patients must be reconsidered as resources in the growing epidemic of cardiometabolic and related chronic diseases. Enhancing awareness about the common benefits of healthy lifestyle for both self-management and primary prevention may further enhance the benefits that families seem to have. These finding are highly relevant for an informed discussion on how to manage and even prevent the growing epidemic of chronic diseases like T2D in low-income countries.
Footnotes
Acknowledgements
The authors wish to thank the individuals who opened their homes and took time to participate in this study. The field assistants are acknowledged for helping with the data collection. We also thank the staff at Kagando Hospital for their hospitality and support to the research project.
Authors’ Note
The funding agencies had no any influence on the study design; data collection, analysis, or interpretation; or the writing of this article.
Declaration of Conflicting Interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article:
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Novo Nordisk Fonden, University of Copenhagen, Aase & Ejnar Danielsens Fond, Thorvald Madsens Fond, and Christian & Otilia Brorsons Rejselegat.
Supplement Issue Note
This article is part of a Health Education & Behavior supplement, “Noncommunicable Diseases in Africa and the Global South,” which was supported by SAGE Publishing, with additional support from the National Heart, Lung, and Blood Institute Contract No. HHSN268201500073P. The entire supplemental issue is available open access for one year at 
.