Shaping Environments for Reductions in Type 2 Diabetes Risk Behaviors: A Look at CVD and Cancer Interventions

Before reviewing how CVD and cancer intervention programs have addressed behavioral risk factors, we must first define what is meant by “environment.” How we conceptualize the environment will affect research and intervention methodologies. Webster’s New World Dictionary defines “environment” as “something that surrounds—surroundings; all the conditions, circumstances, and influences surrounding and affecting the development of an organism or group of organisms.” To make the concept of environment useful, consider the theoretical model of how our surroundings affect health behaviors presented in Figure 1.

The left center of the figure, labeled “health behavior decisions,” represents an individual who makes decisions about eating, exercise, and health maintenance behaviors in the context of day-to-day living. Material resources include money, a vehicle, a home, health insurance, and so forth. Psychological resources include current emotional states, such as depression, fear, anxiety, or worry, and coping strategies that can affect behavioral decisions.

The center box, labeled “proximal-distal continuum,” represents the individual’s interface and interaction with the environment. If behaviors are influenced by systems that operate within larger environments, those who aim to change environments may target their geophysical and sociocultural components.

The geophysical system represents a person’s physical surroundings, which can have both negative and positive impacts on health behaviors. For example, transportation systems affect levels of physical activity.1 Communications systems, including broadcast and print media, television, computers, and the Internet, disseminate information, art, and entertainment. However, research has linked time spent accessing such media to development of chronic diseases.14,15 Along with intense marketing of high-calorie, high-fat food choices, the agricultural and food distribution systems have contributed to the inverted food pyramid that emerges when food consumption among Americans is compared to recommended dietary guidelines. Manufacturing systems have evolved in a manner that has made work, in many cases, less physically demanding. In addition, increased income allows people to purchase more labor-saving devices.

Social systems are made up of people who, through interactions in families, schools, social and religious organizations, and workplaces, have considerable influence on one another’s health-related behaviors and perceptions. Cultural systems, in general, refer to the language, ideas, beliefs, and behavioral norms that go with belonging to a particular social group. Cultural norms, especially those based on ethnicity and region, can have a direct effect on food choices and also affect the role that food plays in family and group life. The modern economic system, through wages and credit, allows individuals to gain access to resources such as transportation, housing, food, and health care. Differential access to such resources generally results in disparities in health outcomes.16 

To summarize, the American geophysical environment has created a physical environment that requires very little physical activity for most people, markets a food supply that promotes eating fast food, and gives people the time and money to achieve this lifestyle. Systems of the sociocultural environment can have a great impact on the ability, desire, and extent to which one indulges in these behaviors.

It should be obvious, however, that this method of characterizing environments by its parts is artificial since physical and social systems are inextricably intertwined. In the next section, we will highlight examples from CVD and cancer intervention projects that have endeavored to bring about changes in health outcomes.

Two major interventions sponsored by the National Heart, Blood, and Lung Institute set the stage for addressing CVD risk factors at the community level: the Stanford Five-City Project (SFCP) and the Minnesota Heart Health Project (MHHP). These studies were designed to produce changes in CVD risk, mortality, and morbidity. Descriptions of how each sought to influence environmental systems to affect changes may be instructive for diabetes health advocates.

The SFCP. The SFCP17 explored the extent to which a comprehensive program of health promotion through mass media would affect changes in CVD risk, morbidity, and mortality.

Two treatment and three control communities in northern California were included. Beginning in 1979, the central hypothesis driving this intervention was that influencing behaviors through media-driven education campaigns and community organization would 1) reduce the prevalence of CVD risk factors in people aged 12–74 years, and 2) this reduction in risk would lead to decreased CVD-related morbidity and mortality in people aged 30–74 years.

From 1980 to 1986, the two treatment communities received what was called a “multifactor risk reduction education program.” The program consisted of messages focusing on cholesterol reduction through dieting; blood pressure reduction through getting regular medical care; reduction of salt intake; weight maintenance through diet and exercise; adherence to medication regimens; and reducing cigarette smoking. Each adult in the intervention communities was exposed to ∼5 hours per year of these messages through classes, lectureships or workshops, television and radio, booklets and kits, and newspapers and newsletters.18 

Results indicated that positive changes (increased knowledge about CVD, blood pressure, smoking, and cholesterol), though modest, occurred in residents of the two treatment towns when compared to the three control towns.18 Cross-sectional surveys showed that residents of the two treatment towns were more knowledgeable about blood pressure and hypertension19 and exhibited positive changes related to cigarette smoking prevalence20 and cholesterol reduction.21 

The other arm of the SFCP was to examine morbidity and mortality rates due to coronary heart disease, myocardial infarction, and stroke in relation to findings from the multifactor risk reduction media education program. No significant differences were found in morbidity and mortality rates in the treatment and control towns when compared to controls.22 

The MHHP. The MHHP used seven strategies to communicate “heart-healthy” messages: 1) involvement of community leaders and organizations; 2) media education; 3) population-based risk factor screening and education; 4) adult education classes; 5) youth and parent education; 6) health professional education; and 7) community-wide risk factor education campaigns.23 Like the SFCP, there were no significant changes in overall CVD mortality and morbidity.24 There were, however, significant improvements related to reducing risk through exercise, smoking cessation programs, and nutrition.

To motivate exercise behavior, the MHHP conducted annual exercise campaigns in two relatively small intervention communities from 1982 to 1989.25 The program, called “The Shape-Up Challenge,” was designed as annual month-long worksite exercise competitions. The study involved 17,626 employees from 119 participating companies. During the month-long competitions, employees recorded minutes spent daily in aerobic activities. Incentives were used for both intra-group cooperation and inter-group competition. The results showed that smaller companies had significantly greater participation rates than larger ones and that women were more likely to participate than men.25 

Another notable undertaking of the MHHP was its “Quit Smoking” contests. Recruitment methods included community-wide direct mail campaigns, newspapers, contest flyers, and school presentations. Smokers were eligible for monthly and grand prizes for enrolling during the designated period and abstaining from smoking for at least 1 month. Approximately 70% (918 people) of those who returned initial interest cards also pledged to quit smoking. Of those, 16.7% reported abstinence for 1 month compared to 9.2% of those who did not make the pledge. This contest was especially successful in recruiting individuals with less than a high school education.26 

A nutrition component of the MHHP focused on long-term outcomes of a school and community-based intervention.27 Sixth-graders in intervention and control groups reported two key aspects of their eating behaviors: knowledge and food preferences and food-salting behaviors. Children were surveyed annually for 7 years beginning in 1983. For girls in the intervention community, these two key variables were more favorable and significantly higher during the follow-up period when compared to their male counterparts. When compared to boys in the control community, boys in the intervention community were more knowledgeable about healthy choices and were less likely to add salt to foods.27 

Environments and impact of the SFCP and MHHP. The SFCP and MHHP serve as models for involving and engaging different aspects of the environment in addressing nutrition, physical activity, and cigarette smoking. In addition, results from each program highlight the importance of defining the target audience in order to have a greater effect on outcomes.

Many of the unhealthy lifestyle messages that individuals receive come through the media. However, the SFCP utilized this influential component of the treatment communities’ physical environment to disseminate its educational messages. To increase the effectiveness of message delivery, outreach to the community was augmented by involving other important entities embedded in sociocultural systems, such as local health organizations, voluntary associations, educational institutions, and school health programs.17 It was, therefore, a complex interplay of many systems within both the physical and sociocultural environments that contributed to the changes in CVD risk behaviors.

Like the SFCP, the MHHP involved many different environmental systems in the target communities to promote a reduction in CVD risk behaviors. Efforts to increase physical activity took place in worksites.25 This approach targeted participants’ physical environments by incorporating positive health behaviors in places where they spend the majority of their time during the day. This approach also played on elements of the sociocultural environment by providing incentives for intra-group and inter-group interactions among the employees. Though results showed that workers in smaller companies were more likely to participate than those in larger ones, the key finding was that individuals’ behaviors changed when vehicles for participating in a healthy behavior were made readily available.

In the “Quit Smoking” contests, the MHHP used primarily print media for recruitment and was successful in getting people to return pledge cards agreeing to stop smoking. Both people who agreed to quit and those who did not agree to quit reported abstinence for at least 1 month. The MHHP also targeted the sociocultural environment of school-aged children in the nutrition component of the study. These strategies proved to be a viable and seemingly sustainable method of influencing eating behaviors of young children.27 

The SFCP and the MHHP demonstrated the importance of designing interventions that have both specific targets and sufficient reach into their communities. The SFCP resulted in only modest changes in knowledge about CVD risk factors,19 positive changes in smoking prevalence,20 and cholesterol reduction21 among residents of the intervention cities when compared to controls. Though several parameters may have contributed to this trend, it is likely that designing different methods of disseminating education for specific subsets of the population might produce more emphatic results.

With the exception of one treatment and one control city in which 20% of the residents were Mexican Americans, the majority of the residents in the study were white and non-Hispanic. As part of the SFCP, Ribisl et al.28 found that Hispanic and white men with lower educational status had more CVD risk factors and received fewer educational messages than Hispanic and white men with higher educational attainment. These findings reveal that even when people are exposed to health-enhancing information in their physical environments, other factors, such as economics, may be greater determinants of risk behaviors.

In the MHHP, females were more likely to participate in the worksite exercise programs,25 and the “Quit Smoking” contests were especially successful in recruiting individuals with less than a high school education.26 This demonstrates the need to tailor interventions based on factors such as sex, age, and socioeconomic status.

From the SFCP and MHHP, we learned that multiple environments can be modified in order to reduce risk behaviors. These studies were also helpful in understanding that crucial to the success of such interventions is defining communities in a manner that is sensitive to factors such as sex, ethnicity, age, and socioeconomic status. We will now describe, briefly, an ongoing demonstration project that provides insight into how to change multiple environments in a target population.

Much of what we have learned about changing environments for better health outcomes has come from our work on the Nashville REACH (Racial and Ethnic Approaches to Community Health) 2010 initiative. This demonstration project, funded by the Centers for Disease Control and Prevention, is part of a larger effort to address a broad range of health disparities among different ethnic groups in communities across the United States.

The goal of Nashville REACH 2010, specifically, is to reduce diabetes and CVD-related disparities among African Americans living in North Nashville by the year 2010. The North Nashville community was targeted based on data indicating that African Americans living there have higher age-adjusted death rates (deaths per 100,000 people) and premature death rates (years of potential life lost per 100,000 people) because of CVD and diabetes than do whites in the same county. (The age-adjusted and premature death rates for North Nashville and Davidson County appearing in this proposal were calculated by Dr. Jesse Huang, chief epidemiologist of the Metropolitan Health Department of Nashville and Davidson County, using a file of 1998 death records for Davidson County residents supplied by the Tennessee Department of Health. The age-adjusted rates were adjusted to the 1940 standard population using 1998 census population estimates obtained from Claritas, Inc., Arlington, Va. The premature death rates were calculated using <75 years as the definition of prematurity.)

A logic model illustrating Nashville REACH 2010 goals, plans, and anticipated outcomes is presented in Figure 2. On the left side of the logic model, the problem is stated. Next, proceeding to the right, planning input from various sources is denoted.

One of the most important features of Nashville REACH 2010 is the active involvement of the community members and coalition members representing several academic institutions, the local health department, faith-based organizations, and others under the direction of a community health center. When presented with the disparity statistics, members of the community action plan committee proposed a systematic plan to facilitate change.

The model illustrates the development of four strategy teams focusing on behavioral and environmental risk factors and engaging a wide range of community resources to accomplish their goals. It then integrates more specific goals, namely 1) creating readiness to change, 2) attending to environmental supports and barriers, and 3) introducing behavioral supports.

Efforts to increase the community’s readiness to change have included presentations to various community, civic, and grassroots organizations and dissemination of the message through print media. In addition, strategy team members have been featured on local radio and television networks.

Interventions designed to attend to environmental supports and barriers include promoting change in infrastructure of schools, worksites, churches, and neighborhoods. Toward this end, a number of churches have begun to work actively with REACH to promote low-fat eating and offer exercise classes. Local restaurant owners have been challenged to add healthier items to their menus and to become advocates and models for change among other restaurant owners. Moreover, a manual was developed to guide community organizations in planning and implementing health screenings, with an emphasis on follow-up. Team members have also developed community audits called “Walk Abouts,” through which community members systematically document both favorable and unfavorable aspects of the environment and make concerns known to the appropriate community leaders. Additionally, clinic hours have been extended for a community health center serving the target community.

Another central goal in implementing REACH interventions was to introduce behavioral supports for changes in risk behaviors. One way REACH proposes to do this is by launching a labeling campaign that will enable residents to identify the healthier food choices on restaurant menus and in grocery stores. The theory behind this is that if healthier foods sell, then more will be ordered, labeled as healthy, and purchased subsequently.

The aforementioned logic model strategies highlight only a few examples in which Nashville REACH 2010 is attempting to make environmental changes. The model concludes on the right with strategies leading to a reduction in risk behaviors. The enclosed loop demonstrates the key role of multiple aspects of the environment in addressing the problem and seeing goals to fruition.

Morbidity and mortality are often related to smoking behaviors in the case of lung cancer and underutilization of preventive care services in breast, cervical, and prostrate cancers. Two community-based studies addressing these risk behaviors are summarized in the following sections.

“Quit Today.” Because cigarette smoking is the leading preventable cause of premature death in the United States, several studies have been designed specifically to encourage and aid people in efforts to stop smoking. One study, “Quit Today,”29 was prompted by the finding that African Americans have made little use of the Cancer Information Service (CIS) smoking cessation resources. Established in 1975, one of the goals of the CIS is to conduct community outreach activities to reach underserved populations.30 

The aim of the “Quit Today” study was to increase the utilization of CIS resources among African Americans by using targeted communications (primarily radio). Fourteen communities served by four CIS regional offices participated in the study. Development of the intervention was guided by examination of previous studies that illuminated quitting motives and barriers to quitting among African Americans. In addition, focus groups were conducted to identify specific reasons why African Americans underutilized the CIS.

After analysis of these findings, storyboards and audiotapes were developed and reviewed by members of the target audience in a second round of focus groups. Six radio spots were produced—two each for African-American contemporary programming, gospel, and jazz. A seventh radio spot was taken from the audio portion of the single television ad that aired. For each experimental community, 10 weeks of advertising took place—6 weeks in the fall and 4 weeks in the spring.

Results from the campaign showed that 709 smoking-related calls came from either the experimental or control communities. Of those calls, 565 came from African Americans. Calls from African Americans in the experimental communities were approximately 80 times greater than those in the control community. Moreover, the average number of calls to the CIS increased from 1.9 calls per week before the intervention to 86 and 40 per week during the fall and spring waves of the intervention, respectively.

Most African-American smokers stated that they heard about “Quit Today” on the radio. The second largest venue was television. African-American callers tended to be between the ages of 30 and 39 years and female. Approximately 20% of the African-American callers had not completed high school, and a much higher proportion had completed high school or some college when compared to other ethnic groups.

“Por La Vida.” Many of the efforts to address breast and cervical cancer place special emphasis on screening. One study provides an example of a program tailored to women of Latin-American decent. “Por La Vida”31 was designed to evaluate the impact of a lay worker–mediated program to increase cancer screening among low-income Latinas.

Lay workers were identified members of the Latino community, which viewed them as trusted natural helpers. The 36 recruited workers were trained to conduct weekly educational sessions. Workers then invited women they knew to participate in 12 small-group sessions where educational information related to breast and cervical cancer was presented.

Pre- and post-test results showed that breast self-exams increased among the intervention group. There were no significant differences, however, in professional breast exams between intervention and control groups. Women over 40 years of age in the intervention group reported an increase in mammography use. Additionally, women in the intervention and control groups reported increased Pap smear use. The increase was higher for the intervention group, however.

The Witness Project. The Witness Project32 was designed to increase mammography use among a population of underserved African-American women in rural Arkansas. The intervention took place in two intervention counties; two counties served as the control.

The Witness Project team was made up of African-American women who were breast- or cervical-cancer survivors. They delivered messages about breast health primarily in conjunction with other church-related activities. Beginning with prayer and personal statements of faith, the cancer survivors, referred to as Witness role models, shared their personal experiences, placing special emphasis on the importance of early detection and treatment. Participants were taught breast self-examination and also provided free and reduced-cost mammograms. Women in the control counties did not receive any breast education through the Witness project and only participated in telephone surveys.

Trained African-American women interviewers collected baseline and 6-month follow-up data from women in the intervention and control counties. The results showed that Witness Project participants significantly increased their practice of breast self-examination and mammography compared with women in the control counties.

Research on community-level interventions designed to reduce risks for CVD and cancer suggests several practical implications for diabetes health advocates.

1. Environmental influences on health behaviors are pervasive, complex, and very strong. Any single intervention program will, therefore, likely have only modest effects. Such effects are significant, however, when in combination with other health promotion activities that affect a community.

2. Members of a community are willing and able to make lifestyle changes when vehicles for participating in healthy behaviors and risk reduction behaviors are made convenient.

3. The most effective interventions are tailored to address multiple specific environmental factors associated with target communities. Successes in one community may not generalize to others if any of their environmental factors differ.

4. Environmental interventions must be sustained over time in order to achieve lasting effects. Limited resources might therefore be best allocated for modest interventions that make changes (e.g., identification of “heart-healthy” items or restaurant menus) that can affect many community members.

5. Diabetes educators can help patients directly by making them aware and promoting use of community programs consistent with their health goals. They can help indirectly by sponsoring and advocating environmental change programs.

6. The models presented in Figures 1 and 2 suggest a wide variety of important targets for community-level health promotion and risk reduction. Understanding the forces for and against change help health activists invest their resources wisely.

7. Research on environmental factors that influence CVD, cancer, and diabetes risks is in its infancy. Much more needs to be done. Diabetes educators, diabetes organizations, and diabetes-related funding agencies can help by including such research among their highest personal and corporate priorities.

This project was supported by National Institutes of Health (NIH) Grant P60 DK20593, NIH Training Grant DK 07061, and CDC Grant U50/CC4417280-01.