|Home | About | Journals | Submit | Contact Us | Français|
As developing countries shift to increasing prevalence of cardiovascular risk factors and diseases (CVD), prevention efforts, both primary and secondary, become a public health priority. Designing effective methods requires a clear understanding of local beliefs and practices regarding health risks and behaviors.
A mixed gender and age team deployed a Rapid Assessment Protocol (participant observation; interviews) over three days. Interviews from 25 residents of Carriacou, Grenada included leaders and community members representing a range of demographic characteristics (gender, age, employment).
Residents expressed general uncertainty about their actual health. While acknowledging that certain conditions (e.g. diabetes, hypertension) were prevalent, heredity was viewed as being more strongly associated with CVD. Not being able to work or carry out one’s daily activities often drove health care seeking behavior (evaluation, care or initiating lifestyle changes). Health improvement activities when practiced were fragmented, not an overall lifestyle change. Physical activity was implicitly valued but not universally practiced; it declined with age and increasing work and other commitments.
While public health programs benefit from understanding community attitudes and beliefs, research to inform program development is often not undertaken or if undertaken not effectively utilized to make needed program modifications. Key to our conclusions was their perspective on health as illness oriented and reactive, strongly associated with heredity rather than preventive and associated with behavior change. A preventive focus informed by local practices is fundamental to designing effective and sustainable primary and secondary prevention programs and particularly useful in developing countries.
“We don’t worry except when you get sick” — middle aged woman
The epidemiologic transition  describes the shift in population mortality from childhood infectious diseases, nutrient deficiencies, and epidemics at all ages, to degenerative and lifestyle-related diseases at a later age. Many developing countries are undergoing a contemporary version of this transition wherein improved public health measures and medical care help individuals live longer, more productive lives. Concurrently these countries often experience changes in diet (e.g. fewer whole grains, higher fat intake) and reductions in physical activity that lead to higher prevalence of cardiovascular (CVD) risk factors and disease .
Chronic noncommunicable diseases were estimated to contribute to 59% of global mortality (~32 million deaths) and 43% of the global burden of disease (1998 estimates) . CVDs accounted for 31% of worldwide deaths and 10% of global disease burden (based on disability-adjusted life years). A dramatic shift will occur by 2025, with low and middle income countries accounting for over 80% of world CVD burden. While demographic shifts will drive much of this change, so too will the dramatic increases in urbanization (projected to exceed 40% by 2021). This trend and its associated lifestyles will contribute to higher rates of diabetes and other important cardiovascular risk factors .
Population based approaches to reduce CVD have had mixed success in developed and developing countries. Community education, incorporated into the Stanford Three- Community Study, demonstrated a reduced risk of CVD after 2 years among intervention communities relative to control communities . The Finnish North Karelia Project, a comprehensive community program to reduce CVD risk failed to demonstrate a significant difference between intervention and control communities [5, 6]. In the Minnesota Heart Health Program, selected intervention components were effective in reducing risk in assigned groups , but overall the program’s effects were only modest and not statistically significant. In the Pawtucket Heart Health Program, early significant differences between a community exposed to an educational, screening, and counseling program and a control community were not sustained . In the Stanford Five-City Project, after a mass-media educational campaign, the intervention community lowered its short-term relative risk but over time experienced no significant difference in CVD events compared to control cities. The authors concluded that, “[i]t is most likely that some influence affecting all cities, not the intervention, accounted for the observed change” (p. 322) .
Impact evaluations of similar programs in developing countries are limited. An intervention in Mauritius involving fiscal and legislative measures, the use of mass media, along with school, workplace, and community education, yielded significant 5-year reductions in smoking, hypertension, and hypercholesterolemia prevalence . Concurrently, diabetes and obesity prevalences, increased by 15% and 56%, respectively.
These studies, although community focused, did not employ research strategies that incorporated the community into the intervention’s design, which may explain the lack of sustained programmatic success while emphasizing the importance of ongoing efforts and the need to tailor programs to the target community. Understanding local norms, beliefs and attitudes is a critical first step in engaging the community and designing an effective sustainable intervention.
The World Heart Federation (WHF) initiated an epidemiologic study based in the three- island country of Grenada to examine effects of the epidemiologic transition on cardiovascular risk. In selecting this Western Hemisphere country, WHF considered several factors. First, Grenada’s geographic isolation made it amenable to public health measures focusing on the modification of local health practices. Second, the traditional Grenadian diet included liberal amounts of seafood, fresh fruit and vegetables, and thus preserving this diet would be expected to help prevent CVD. Third, Grenada was early in the epidemiologic transition, a stage where interventions that mitigate the transition’s negative health consequences would be more likely to have a positive effect. Lastly, its population of 95,000 would allow for a manageable project.
The research reported here was the first part of this multi-stage community-based research project to study heart health in each of the three islands of Grenada. Given the pre-existing differences among the three islands (e.g. infrastructure, industry, access to health-related resources) generalizability of findings from one island to another was not assumed. On each island a similar methodology would be followed: formative research (using qualitative methods); CVD and risk factor surveillance (epidemiologic survey); and programmatic interventions. The latter would be designed to moderate the development (primary prevention) or progression (secondary prevention) of CVD and its risk factors. The formative findings would inform the surveillance methods and processes and the program interventions to assure alignment with the local environment and the community’s beliefs and practices. We report findings from the formative research conducted on one of the islands, Carriacou.
This project employed Rapid Assessment Procedure (RAP), an anthropologically based method, used internationally to assess health-related attitudes and behavior and health systems [11-16]. Using mixed-methods, this focused, systematic approach provides for time-efficient formative research  Given its abbreviated timeframe, RAP does not replace comprehensive ethnography; rather its focused, qualitative approach provides key insights about a particular issue or situation within a given community or sub-group (e.g. how different groups perceive their health, understand causes, view solutions).
Grenada includes two smaller islands (Carriacou and Petit Martinique) in addition to its main island (Grenada). The research reported here describes our findings on Carriacou. With a population between 5000 – 8000, Carriacou has one small urban area (Hillsborough) with ferry service, a small airport and approximately 10 geographically distributed parishes (districts) in both inland and coastal areas. A hilly island of 13 square miles with winding, mostly paved roads connecting communities, Carriacou has a hospital and main clinic situated inland on one of the island’s peaks and three additional clinics located in the north, central and southern districts. Carriacou has several public primary and two secondary schools. Employment ranges from farming and fishing to shipping and construction as well as service industries (government, education, retail, health). Residents rely primarily on public transportation. Car ownership is the exception. Over 12 different religions are practiced on Carriacou.
RAP on Carriacou utilized anthropological techniques, primarily in-depth interviewing coupled with direct and participant-observation lasting approximately three days. The semi-structured interview guide included eight key domains considered to be relevant to understanding cardiovascular health, health behaviors and health service utilization. For parsimony, only key data from selected domains are reported here (Table 1). In additional to US and Grenadian-based institutional review board approvals, the project’s local community advisory board reviewed the methods and facilitated access to key community members.
During the three days of intensive data collection, daily team briefings reviewed progress and included an assessment of interviewees’ representativeness across socio-demographic sub-groups and adequacy of data within each domain. Commonalities and differences across the data provided the basis for preliminary discussions of potential themes. Preliminary themes and recommendations emerged after iterative review of notes and were finalized following review and feedback from the local advisory board. The latter served as a form of respondent verification, reaffirming that our interpretation of the formative data was consistent with community beliefs and practices.
A mixed-gendered team consisting of three University of Rochester (UR) representatives and two residents of Carriacou, Grenada was led by a UR faculty member. Team members completed a joint four-hour RAP training workshop (discussion, role plays of protocols and interview guide, ethical principles, cultural competency, purposive sampling strategies, and analytic matrices).
The 5-member team completed a windshield tour of the island and participant observations in retail and eating establishments, public venues (town squares, ferry landing site, beaches) and religious events. The team identified potential key informants (adults over age 18) based on their position, availability, representativeness (e.g., age, gender) or referral from a prior informant. When needed, approval from supervisors was obtained prior to approaching an employee during work. All interviewees received a small informational card and an explanation of the project including the verbal consent language. No one refused. Several respondents, who initially had only a few minutes to talk, spent 60-75 minutes being interviewed. Most interviews lasted 30 minutes. Written notes were taken omitting identifiers other than general information about age category, gender and nature of employment (Table 2). Of the 25 key informants interviewed, over half occupied leadership or authority positions in healthcare, education, government or religion . Three interviewees were members of the project’s local advisory board.
Interviewees represented men and women ranging in age from early 20s into their 80s and included individuals from higher and lower socioeconomic status. Drawn from different communities, they included parents (in single and two-parent households), retirees, unmarried and widowed individuals and post-migration returnees. Employed participants held positions with regular weekday hours or were classified as self-employed (sometimes in several different lines of work). Given the absence of census or other demographic data about Carriacou the representativeness of the respondents cannot be determined.
Beliefs regarding what constitutes health varied, ranging from a simple “absence of disease” definition to a more inclusive framework (social and economic factors). See Table 3 for specific respondent statements.
Health problems were thought to be primarily hypertension (referred to as “high blood pressure” or simply “pressure”) and diabetes (referred to as “sugar”). Other problems were rarely mentioned (e.g. cancer, asthma, STDs, HIV).
Respondents saw a strong association between family history and development of certain conditions, such as diabetes and high blood pressure. Awareness of family history shaped what an individual watched for (e.g. symptoms); however care seeking was mainly for treatment not prevention. Preventing health problems or illnesses was not mentioned by lay respondents nor were the words ‘risk’ or ‘prevention’.
Health seeking behavior seldom occurred except when they were sick. This pattern was consistent across respondents regardless of whether they defined themselves as healthy or not as healthy. Regular medical check-ups were more common among women. Overall, many respondents noted that a functional limitation, especially affecting ability to work, increased whether the individual would consider or seek health care or screening or make a behavior change (e.g. attempting to lose weight, exercising more). Taking time off of work to seek care was also mentioned as a barrier.
Among lay respondents, heart disease was considered an acute, not a chronic condition.
While respondents viewed “pressure” as a major health problem, not all considered it fatal. Family history was cited as the major cause of hypertension. Respondents with hypertension usually treated their condition, but long term were more likely to use ‘bush’ medicine (local term for herbal or alternative medicine). ‘Bush’, typically a tea or drink make from a local plant (e.g. Carailli, sugarbush) was viewed as a natural treatment. Self-treatment with “bush” typically started after a confirmed diagnosis and, sometimes after brief use of prescribed medication.
Estimates of diabetes prevalence were high (40 – 70%) especially for those over age 40. Respondents viewed the risk of diabetes as fixed, due to heredity and aging, rather than modifiable through lifestyle. The association between diabetes and heart disease was rarely mentioned
Respondents indicated that high cholesterol was not common on Carriacou noting that individuals were protected by genetic factors coupled with their low fat diet. As with diabetes lay respondents did not associate high cholesterol with heart disease. Reports of cholesterol testing were rare. As with hypertension medications, individuals on cholesterol-lowering drugs may only take them for a month or two; no one mentioned treatment that included lifestyle modification.
Despite the absence of an orientation to prevention, many respondents expressed a personal responsibility for their own health. Religious beliefs, although central to many individuals’ lives, did not appear to affect or limit their health care-seeking behaviors.
Many respondents cited gardening as their most common type of physical activity (vegetable gardens, tending domesticated animals) (Table 4). Exercise was viewed as a part of everyday activities or work, rather than a scheduled activity.
Community members reporting participation in some team-based or programmed physical activity (swimming, track, cricket, basketball) tended to be under age 35. Middle aged respondents commented on declining physical activity due to age, societal changes (e.g. increase in motorized vehicles; availability of other entertainment), and increasing time constraints.
The heat and humidity of the Caribbean climate were not mentioned as factors affecting exercise or physical activity. Both male and female respondents referred to walking (alone or in groups) as a common source of local physical activity however respondents rarely reported walking themselves and few individuals were observed walking.
Eating habits are affected by accessibility, availability, convenience (ease of preparation), cost, as well as taste. Some respondents noted an increasing awareness of the importance of optimal nutrition.
Respondents reported a daily diet including some form of protein and carbohydrates. Marine based protein sources were ubiquitous (e.g. “lambi” (conch), fish) and considered to be components of the “traditional diet”. Daily vegetable consumption was common and ranged from one to three times a day. Most daily consumption was from starchy (e.g. pigeon peas, corn) rather than green leafy vegetables.
Most vegetables (carrots, caliloo [taro greens], okra, cabbage and greens) were available for purchase and eaten fresh; frozen or canned vegetables were commonly available and added to stews. Due to preparation time, weekend meals generally included a greater variety of foods than weekday meals. Vegetables were less available and more costly since Hurricane Ivan (Fall 2004). While availability was improving, some respondents did not think hurricane-related cost increases would return to pre-Ivan levels.
Fruit consumption was also part the daily diet often in the form of juice (tins or freshly squeezed). Pharmacies’ stocks included vitamins and the grocery sold nutritional/energy drinks, but when typical diets were discussed, these drinks were rarely mentioned. A range of oils (olive, coconut, vegetable) was available for purchase. While respondents mentioned using olive oil when possible, they also noted its higher cost as a barrier. Coconut and vegetable oils were more commonly mentioned for cooking. Lunch was generally the largest meal of the day. Adults commonly take lunch to work or buy take-away food. Snacking between meals was not a common practice but recognized as contributing to weight gain.
Body shape or size were not viewed as related to health. Neither thinness nor overweight were considered a good or ideal body size — being “thick” or filled out was preferable, particularly for women. More women than men appeared to be overweight or obese on Carriacou. Overweight was not viewed as a sign of health or power. Respondents associated weight or body size with heredity. Similarly, respondents emphasized that losing weight meant that you were probably ill (e.g. HIV/AIDS or diabetes). Women were more likely to have attempted to lose weight, with little reported success. Several dieting methods were mentioned (eating less, different food types). Over the counter diet pills (available in pharmacies) were never mentioned.
Smoking tobacco was atypical. More men than women reported smoking. With one exception, of the few smokers observed all were male. Use of “weed” (marijuana) was mentioned as much more common than tobacco, and primarily by young adult males. “Kids are smoking too much marijuana today, wasting their lives”(older man).
Some respondents spontaneously mentioned stress, although their descriptions varied, from being experienced constantly to more situational. The latter was viewed as something brought on by one’s own commitments, particularly employment. One younger man, a non-native, expressed surprise that residents talked about stress, “[h]ow can you be stressed here…look at the view…there is no traffic.” Respondents did not consistently connect stress with health. Over-the-counter dietary supplements were available at the local pharmacies for one’s “hectic life”. Some respondents reported using relaxation, sleep, or ‘bush’ treatments for stress related headaches related. No one mentioned religion or faith as a way to deal with stress.
While public health programs benefit from understanding community attitudes and beliefs, research to inform program development is often not undertaken or if undertaken not effectively utilized to make needed program modifications. Key to our conclusions were their perspectives on health as illness oriented and reactive, strongly associated with heredity rather than preventive or associated with one’s behavior. With the shift to greater morbidity and mortality from chronic diseases, a longer term, preventive focus becomes a fundamental priority when designing cardiovascular risk reduction interventions.
Specific study conclusions on the design of cardiovascular surveillance and public health programs were wide ranging. To facilitate utilization, each conclusion generated from the findings was linked to specific recommendations (Table 5) informing the next phases of work on Carriacou (e.g epidemiologic survey; cardiovascular health programs and intervention).
Many local beliefs, customs, and traditions are closely related to self-perception of health among Carriacou residents. There was a general uncertainty about one’s actual health. While the residents acknowledged that conditions such as diabetes and hypertension were prevalent, they did not see them as related to CVD. They held strong beliefs about the role of heredity in disease development while also recognizing their personal responsibility to take care of one’s self. To maximize effectiveness, health program planning must acknowledge and incorporate these beliefs. Additionally any health communication or other health related discussion as part of the program should incorporate information about the interaction among prevention activities, risk reduction, and heredity in disease development. For example, in Carriacou the respondents lacked understanding of the linkages between conditions viewed as hereditary, such as diabetes and high blood pressure and the value of dietary modification.
Functional limitations such not being able to work or carry out one’s daily activities often drove health care seeking behavior that could include evaluation, care or initiating lifestyle changes. Building on this existing motivator may engage some individuals, in prevention or risk reduction activities, who might otherwise not be interested. Health improvement activities (e.g. regular consumption of vitamins/herbs, healthier oils, weight loss attempts) while employed by some residents appeared fragmented, not a general health related lifestyle change. Associating these activities to an individual’s risk profile obtained during surveillance activities could begin to reduce this fragmentation. Similarly, physical activity was implicitly valued but not universally practiced. Building physical activity into daily routines at work and home was common and may be a more effective, sustainable strategy than promoting exercise. This may help overcome the acknowledged decline of exercise with age and increasing work and other commitments. Lastly local dietary habits were affected by changes following Hurricane Ivan. While noting the change, residents did not appreciate that these changes (shifting away from their traditional diet) may be being institutionalized into the local diet.
Programs or approaches successfully implemented in developed countries to improve public health may not serve developing countries well. Formative research can inform proposed interventions by establishing an understanding of community norms and attitudes and in turn building this understanding into program or intervention design. This study describes a useful methodology to understand local views about cardiovascular health and risks, deployable in a developing country and how to translate findings into meaningful program recommendations.
Without the invaluable assistance of the Carriacou Project Advisory Board and Carriacou team members, Trevlyn Cox and Trisha Noel this work would not have been possible.
This project was supported by funding from the World Heart Federation.
None of the authors of this manuscript have any conflicts of interest in this study.
Ann M. Dozier, Department of Community and Preventive Medicine Box 278969 University of Rochester Rochester, New York 14627 Ph: 585.758-7812 Fx: 585-424-1469.
Robert Block, Department of Community and Preventive Medicine Box 644 University of Rochester Rochester, New York 14642.
Deborah Levy, HVMA Division of Medicine Brigham & Women’s and Faulkner Hospital 1153 Centre Street Boston, MA 02130.
Timothy D. Dye, Axios International Department of Research and Evaluation 7 boulevard de la Madeleine 75001 Paris France.
Thomas A. Pearson, Department of Community and Preventive Medicine Box 644 University of Rochester Rochester, New York 14642.