1
Introduction
As the 21st century begins, the nation is facing increasing concern about health problems, particularly chronic diseases such as diabetes, and escalating health care costs. Our society is becoming more diverse in every conceivable way, and diversity is often, though not always, associated with health disparities. On the other hand, we have available increasingly robust theories about health behavior and exciting new opportunities for expanded health communication strategies through enhanced technologies. These offer the potential for interactive approaches that can be tailored more precisely to individual and small group needs.
Many broad social, economic, and political forces affect the health status of populations (e.g., Institute of Medicine, 2000), and these forces ultimately must be addressed. Yet health communication, in its various applications, offers a potentially important approach to a better informed and presumably healthier population by focusing on the behavioral aspects of risk factors, such as diet, smoking, alcohol use, sedentary lifestyle, and sexual behavior (McGinnis and Foege, 1993).
In 1999, the Institute of Medicine convened the Committee on Communication for Behavior Change in the 21st Century: Im-
proving the Health of Diverse Populations, in response to requests from the Board on Neuroscience and Behavioral Health and the Board on Health Sciences Policy to examine the potential of health communication strategies to improve public health behaviors, especially among demographically and socioculturally diverse populations. Health communication strategies, for purposes of this volume, are defined as approaches that seek to persuade or motivate people to change their behavior in order to improve their health. This volume is concerned primarily with communication strategies that are designed for larger groups or the public, rather than individual persuasion strategies such as doctor-patient communication. However, the committee recognized that health communication can be initiated by a variety of sources, including health care providers, campaign developers, and individuals seeking health information.
The salience and timeliness of examining how communication strategies relate to diversity within the U.S. population are emphasized by recent federal initiatives on disparities in health. Examples include the Eliminating Racial and Ethnic Disparities in Health campaign, Healthy People 2010, and the Minority Health and Health Disparities Research and Education Act of 2000. Although informed by the concerns about disparities raised by these initiatives, this volume adopts a considerably broader view of social and cultural differences that will be addressed briefly here and in detail in Chapter 7.
Substantial evidence exists on the relationship of diversity— mostly categorized by race/ethnicity, socioeconomic status, age, and gender—and health status (Centers for Disease Control and Prevention, 2001). The less socially or politically advantaged a population group is, the more compromised its health status. There is also reasonably good evidence that health communication campaigns can influence health behavior (e.g., Hornik, 2002). However, there is little evidence on the enhanced impact of health campaigns that are planned with special attention to addressing the needs of diverse audiences. This does not mean that health communication campaigns have not taken diversity into account.
Most campaigns are concerned about diversity and shape their message and placement based on age, gender, ethnicity, race, sexual preference, or other relevant personal variables. However, there is a striking lack of data available to answer essential questions about diversity—for example, does a health communication effort that takes extra care in understanding and responding to the needs of a diverse audience show effects that are greater than if such efforts are not made? In other words, it is not possible now to answer the question of whether added value, defined as better health outcomes, results from incorporating a focus on diversity compared to an intervention designed for a more general audience.
The challenge for those involved in the design, implementation, evaluation, and research concerning health communication is to influence behaviors with the greatest potential to significantly improve health outcomes across demographically and culturally diverse population groups. This volume addresses the challenge of improving health communication in a racially and culturally diverse society. Although targeting messages to specific audiences should increase effectiveness, this volume describes the difficulty and complexity of categorizing audiences in a meaningful way. Many of the “markers” that are used to describe groups are of little relevance and utility in crafting communication—or must be considered in a subtle way. Belief systems, religious and cultural values, life experiences, and group identity are all powerful filters through which information is received and processed. We need to know more about these in order to understand how to frame messages that will be understood and lead to action. A major goal of this volume is to clarify and evaluate the potential of health communication strategies to meet this challenge.
The committee that was formed to explore this fundamental question of the value of integrating diversity factors more scientifically into health communication approaches includes experts in anthropology, psychology, mental health, cancer prevention and control, health behavior change and theory, communication and the media, social marketing, and public health. This expertise was
supplemented by consultants in communication technology and ethics. The committee’s specific charges were as follows:
-
Review existing theory and research applications in health communication and health behavior change, especially as they relate to culturally diverse populations, and define research areas that would benefit from expanded or new research efforts;
-
Consider up to three specific examples of health communication interventions to evaluate whether and how those strategies affect culturally diverse groups; and
-
Recommend how health communication strategies may be designed and implemented to achieve sustained gains in public health across cultural groups.
The results of this effort are intended to inform policy makers, researchers, funders/sponsors, advocacy organizations, practitioners, and others in developing, implementing, evaluating, and conducting research on communication strategies to improve the public’s health.
CONTEXT
Health and Diversity
The importance of understanding diversity for public health is clearly evident. Yet current categories, which use arbitrary groupings of “race” and ethnicity, are not useful enough and do not contribute to necessary knowledge on the relationship between diversity and health behavior. New ways of understanding diversity will be required to develop more appropriate, and presumably more successful, approaches for health communication efforts. Yet jettisoning current categories of diversity as a means for identifying need and targeting programs will likely be controversial. Many political organizations, constituency groups, the U.S. Congress, and major administrative agencies of the government are organized around current definitions and categories. If the definition of di-
versity is changed, then the basis for funding decisions will change. This uncertainty could result in considerable argument, which happened during efforts to change the census categories for racial/ ethnic identification. Making diversity scientifically and programmatically meaningful undoubtedly will cause political conflict.
Categories for Describing Diversity
Discussions of disparities and health outcome differences across cultural groups contrast broad ethnic or racial categories that, although they have some utility for simplifying and focusing public discussion, relate little, if at all, to variation in socioculturally based beliefs and behaviors that have specific implications for health outcomes. The widely used broad ethnic categories, in particular, are simply not valid indicators of meaningful variation in health-related beliefs and behaviors. The most useful information for framing this volume—data that provide a much deeper and more sophisticated understanding of how specific beliefs and behaviors and health status covary across the U.S. population and of how health behavior is shaped by sociocultural processes—is not available, and one of the committee’s recommendations is to develop this exact information. For example, how individuals come to think about and make sense of the world is mediated, to a large extent, by their social group identification and participation and by their life experiences. The term “cultural processes” is used to refer to these socially grounded ways of learning that contribute to how an individual thinks, feels, and acts.
Part of the challenge in addressing health behavior change for diverse populations is determining the circumstances under which diversity matters. The categorical comparisons that are available, and that are used throughout this volume, emphasize the importance of attending to cultural differences, even if presenting information in this way risks misguiding the reader into thinking these categories are relevant to the committee’s charge.
Economic status and social class are considered dimensions of diversity for this volume because there are differences in health
outcomes, and presumably in health risks, for these groupings (Lynch and Kaplan 2000; House and Williams, 2000). Although economic status and social class can be considered separately, they are most often grouped. Socioeconomic status (SES) commonly includes education, occupation, and income, but other factors such as net worth may be considered. The measurement of socioeconomic status has been discussed at length in the literature, and is clearly summarized in Lynch and Kaplan (2000). Similarly, the relationship between socioeconomic status and health has been discussed in detail in a recent Institute of Medicine report, Promoting Health (2000).
In discussing the relationship between SES and race/ethnicity, House and Williams (2000:97) note: “Although not useful as biological markers, current racial/ethnic categories capture an important part of the inequality and injustice in American society. . . . There are important power and status differences between groups.” In fact, as they document, these power and status differences relate directly to the health disparities through factors such as environmental conditions and access to preventative and curative health care. They also note that SES alone does not explain health disparities. Furthermore, caution must be applied to the measurement of SES, and to assumptions about relationships between poverty and culture. As just mentioned, poor health outcomes are not confined to people with low SES, nor does low SES mean inevitably poorer outcomes.
Age and gender also show differences in health outcomes. These are partly because of biological factors, but also because of other kinds of factors underlying health disparities such as access to care, education, and environmental contexts, as well as health beliefs and behaviors. Both age and gender affect how people are treated in U.S. society. The very old and the very young are likely to have fewer rights and be subjected to more discrimination, as are women, regardless of ethnicity.
The committee members acknowledge that sexual orientation is a relevant and important parameter of diversity for the develop-
ment, implementation, and evaluation of effective health communication interventions. Indeed, significant differences have been reported on various health measures for women and men, depending on sexual orientation (Institute of Medicine, 1999a; Dean et al., 2000). Many of the differences in health outcomes are related to differential risk behaviors, such as higher risk of sexually transmitted diseases and AIDS from unsafe sexual behavior in gay men, higher rates of poor diet and being overweight among lesbian women, and higher rates of smoking, alcohol use, and substance abuse among lesbian, gay, and bisexual persons.
Unlike the other defining constructs of diversity in this volume, however, sexual orientation is not routinely used as a demographic grouping to assess or report differences in health behaviors or health outcomes. In fact, the collection of reliable information about gay and lesbian populations has been constrained by political obstacles and methodological challenges (Dean et al., 2000; Institute of Medicine, 1999a).
Similarly, other dimensions of diversity are highly relevant to communication, but we cannot focus on them in an examination of existing studies. For example, the issue of an individual’s primary language is minimally addressed in this volume, although it is clearly an important topic. For the decade starting in 1990, the “most recent wave of immigration is composed largely of non-European, non-English speaking ‘people of color’ arriving in unprecedented numbers from Asia, the Caribbean, and Latin America” (Suárez-Orozco, 2000:5). Significant numbers of U.S. residents also communicate primarily using languages other than standard English. Literacy and education are additional dimensions of diversity that are not closely examined. The lack of attention to such factors should not be viewed as a measure of their importance, but rather reflect constraints in terms of what reasonably could be examined within the scope of the committee’s time frame and available data.
The U.S. Population and Categorical Demographic Diversity
The U.S. population now has more than 281 million residents. The number of persons in the United States has tripled over the past century among persons under age 65, and increased by a factor of 11 among those age 65 and over (U.S. Census Bureau, 2000). Ethnic diversity has increased among all age categories in the population, particularly among younger populations. By the year 2050, the percentage of non-Hispanic whites in the United States is expected to decrease to 53 percent overall. As many as 56 percent of adolescents are projected to represent ethnic minority groups; that is, groups other than non-Hispanic whites or EuroAmericans (National Center for Health Statistics, 2000; Brown et al., 1996). More people are completing formal education, poverty rates have reached their lowest levels since 1979, and real median household income has reached a record high, increasing for all types of households in all regions of the United States in 1998 (U.S. Census Bureau, 1998, 2000; Dalaker and Proctor, 2000).
All of these changes in demographics suggest that the cultural landscape of the U.S. population is changing toward a larger, older, more ethnically diverse, and more educated population. For example, Hispanics constitute a relatively younger population lagging in education and income, while whites are older and have higher accumulated wealth. Among populations age 15 and older, non-Hispanic whites (83 percent) were most likely to have at least a high school education in 2000, followed by Asian Americans/ Pacific Islanders (78 percent), non-Hispanic African-Americans (72 percent), and Hispanics (53 percent) (Bennett and Martin, 1995).
The income gap between rich and poor remains considerable (U.S. Census Bureau, 1998). In 2000, Asian Americans/ Pacific Islanders had the highest median income ($55,500) and Hispanics and African-Americans had the lowest ($33,400 and $30,400, respectively). The poverty rate across all groups is approximately 11 percent, with African-Americans and Hispanics well above the mean and non-Hispanic whites well below the mean (U.S. Census Bureau, 2001).
In 1995, approximately 49 million people (20 percent) in the United States lived in a household that had at least one difficulty in meeting basic needs, such as the inability to meet essential living expenses (e.g., pay utility bills, mortgage, or rent), buy food, or seek medical or dental care when needed (Bauman, 1995). Approximately 5 percent of American households reported that members sometimes did not have enough food to eat, while nearly 20 percent reported either not having enough food or the kind of food they wanted to eat (Bauman, 1995). Income, ethnicity, age, gender of householder, health insurance coverage, and Hispanic origin were among the major risk factors for not having enough food to eat.
The burden of death and disease frequently falls most heavily on some ethnic groups and on those with lower levels of education and income. The age-adjusted rates for the leading causes of death among men and women in the three largest ethnic groups (non-Hispanic whites, non-Hispanic Blacks, and Hispanics) are shown in Tables 1-1 and 1-2 (Centers for Disease Control and Prevention, 2001).1 These tables show that the highest rates among all groups are for heart disease and cancer. However, significant variations exist. Non-Hispanic white and Black males and females have significantly higher rates for these two diseases than Hispanics. For the most part, Hispanics have significantly lower rates across all leading causes of death. Male non-Hispanic Blacks have the highest rates of death for several diseases, including heart disease, cancer, cerebrovascular disease, and diabetes. Furthermore, the disparities among these groups increase with age; for example, the difference in mortality rate between Black and white males is three times greater at age 65 than at age 45.
Table 1-3 shows the age-adjusted death rates for four major causes of death from cancer for whites, African-Americans, Hispanics, Asian Americans/Pacific Islanders, and Native American
TABLE 1-1 Age-Adjusted Rates for 10 Leading Causes of Death by Ethnic Group—Males (rate per 100,000)
|
|
Non-Hispanic |
Non-Hispanic |
|
Disease |
All |
White |
Black |
Hispanic |
Heart |
328.1 |
329.5 |
344.3 |
212.7 |
All cancer |
251.6 |
251.4 |
350.1 |
151.4 |
Cerebrovascular diseases |
62.4 |
60.5 |
89.7 |
44.6 |
Chronic obstructive pulmonary disease |
58.1 |
61.3 |
51.4 |
27.3 |
Unintentional accidents |
50.6 |
49.1 |
64.2 |
47.2 |
Pneumonia and influenza |
28.0 |
28.0 |
33.0 |
18.6 |
Diabetes mellitus |
27.7 |
25.0 |
50.1 |
34.5 |
Suicide |
18.2 |
20.3 |
10.8 |
10.7 |
Kidney infections |
16.2 |
14.8 |
33.8 |
12.9 |
Chronic liver disease and cirrhosis |
13.7 |
12.7 |
15.6 |
23.0 |
SOURCE: Centers for Disease Control and Prevention (2001). |
TABLE 1-2 Age-Adjusted Rates for 10 Leading Causes of Death by Ethnic Group—Females (rate per 100,000)
|
|
Non-Hispanic |
Non-Hispanic |
|
Disease |
All |
White |
Black |
Hispanic |
Heart |
220.9 |
218.1 |
297.0 |
146.5 |
All cancer |
169.9 |
172.1 |
205.6 |
101.4 |
Cerebrovascular diseases |
60.5 |
59.6 |
80.0 |
36.6 |
Chronic obstructive pulmonary disease |
38.2 |
41.5 |
24.5 |
15.3 |
Diabetes mellitus |
23.3 |
19.5 |
51.7 |
32.6 |
Unintentional accidents |
22.7 |
23.1 |
24.4 |
15.5 |
Pneumonia and influenza |
20.8 |
21.1 |
21.7 |
13.5 |
Alzheimer’s disease |
17.6 |
18.8 |
12.4 |
8.4 |
Kidney infections |
11.2 |
9.7 |
26.6 |
8.8 |
Septicemia |
10.5 |
9.5 |
23.0 |
6.8 |
SOURCE: Centers for Disease Control and Prevention (2001). |
TABLE 1-3 Age-Adjusted Rates for Four Major Causes of Death from Cancer by Ethnic Groupa (rate per 100,000)b
Cancer |
White |
African-American |
Asian American/Pacific Islander |
Native American |
Hispanicc |
Lung and bronchus |
|
||||
Males |
69.5 |
99.5 |
34.2 |
40.9 |
31.6 |
Females |
34.0 |
33.0 |
14.9 |
19.8 |
11.0 |
Colon and rectum |
|
||||
Males |
21.3 |
27.7 |
13.1 |
11.6 |
13.1 |
Females |
14.3 |
19.9 |
8.9 |
8.9 |
8.3 |
Prostate (males) |
23.3 |
54.1 |
10.4 |
14.2 |
16.2 |
Breast (females) |
25.3 |
31.4 |
11.2 |
12.1 |
15.1 |
aData are from the American Cancer Society, Surveillance Research (2001). SOURCE: Surveillance, Epidemiology, and End Results Program, Division of Cancer Control and Population Science, National Cancer Institute (2000). Mortality derived from data originating from the National Center for Health Statistics, Centers for Disease Control and Prevention (2000). bPer 100,000, age adjusted to the 1970 U.S. standard population. cHispanic is not mutually exclusive from white, African-American, Asian American/ Pacific Islander, and Native American. |
for both males and females based on data from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute (Centers for Disease Control and Prevention, 2000). Black males have the highest death rates for lung, colon, and prostate cancer; Black women have the highest death rates for colon and breast cancer. Asian Americans/Pacific Islanders have the lowest rates of breast and prostate cancer; Hispanics have the lowest rates of lung cancer. Several factors may contribute to these disparities, including health behaviors, access to and availability of prevention and treatment services, patterns of service utilization, environmental and occupational risks, community support and cohesion, differences in insurance coverage, and underlying biological risk factors.
The Role of Communication
Recognizing the potential of verbal and nonverbal messages to modify beliefs and behavior, the public health community has been working with communication and media experts for more than 30 years to develop mass media campaigns to change health behavior. At the same time, behavioral scientists working in public health have developed communication strategies, including materials for individuals, programs at school and at work, programs delivered through health care providers, and large-scale campaigns.
The design and development of effective health communication strategies has become increasingly important with the development of new, but untested, technologies that could have potential for disseminating health communication across diverse cultural groups. This potential power to affect public health through communication, however, is attended by concerns that not all groups benefit equally from communication or social innovation, as exemplified by the “digital divide” in access to and use of the Internet (Rogers, 1995). Communication technologies may fundamentally change who has access to health information, and therefore shift the distribution of health-behavioral risk factors and health status. Would this shift increase or reduce health disparities as defined by conventional cultural categories? Would groups that differ in health-related beliefs and behaviors respond differently to health communication strategies? Are there options for optimizing health improvement from health communication programs? The answers to these questions may depend on whether disparities are malleable to communication strategies, and whether these strategies can be designed adequately to be effective for persons with different culturally based beliefs and behaviors.
Many of the basic benefits of communication strategies may be obtained across diverse cultural groups through a generic strategy that capitalizes on their similarities in beliefs and behaviors. Alternatively, it may be that the communication intervention or delivery strategy must be targeted to particular groups to have similar effects across groups, or tailored to the beliefs and values of the
group members to optimize the impacts. In this regard, it is of particular policy interest to know whether targeting and tailoring communication interventions is especially useful in improving the health of populations with historical or current health or health care disadvantages. Thus, the question of the marginal benefit of special attention in communication interventions to subgroups may depend on the goals of interventions and, specifically, how much attention should be focused on vulnerable populations—a societal ethical question. Furthermore, whether communication strategies or their implementation must be modified to be effective for different cultural groups will be a function of the particular health problem and subpopulation.
In setting the scope, we purposely attempted to include public health strategies, such as campaigns, and medical practice strategies, such as pamphlets or Internet-based communication by providers to larger groups of patients.
Ethical Considerations
A number of ethical principles need to be considered in the development and implementation of health communication for diverse populations, including avoidance of harm, providing benefit, respecting an individual’s autonomy to make a rational choice, and treating groups and individuals justly and equitably. These principles are not always easy to achieve. First, implementing ethical principles can be complicated by the developers’ need to consider tradeoffs among efficiency, cost, and improving the health of the most in need versus a broader range of persons. Second, a communication strategy that incorporates ethical principles may function effectively and appropriately for one segment of the population, but not for another. Third, unintended consequences may occur even with the most well-intentioned and well-executed health communication interventions. Such outcomes for diverse populations include confusion about the meaning of a message, unwarranted anxiety resulting from implying individual culpability, or the stigmatizing of certain cultural practices.
One method for dealing with these kinds of concerns is to seek and maintain mutual collaboration with the community during all phases of the communication intervention process. This method will increase the likelihood that the messages and strategies of the intervention will unfold in a manner that is appropriate for the intended audience.
APPROACH AND SCOPE
Given the important role that communication campaigns are likely to play in the coming years, the field of public health communication should be strengthened and become more science based. This means supporting research that improves our understanding of the values and context that determine how information will be received, supporting evaluation of campaigns to measure their effectiveness, and paying more attention to the ethical and social concerns that must be appreciated and respected in the design of communication campaigns.
In our review and analysis, we draw on several disciplines. First, we rely on communication sciences and fields such as social marketing to define the essential components and issues in communication strategies and to review the contributions of technological advances to those strategies in the near future.
Second, we rely on theories of culture and cultural process to help define the scope of inquiry into culturally based beliefs and behaviors and cultural comparisons. We begin with a broad view of sociocultural and demographic diversity, considering ethnic groups, gender differences, and differences in factors such as socioeconomic status and sexual orientation. Awareness of culture and the concepts of culture as shared understanding within a community is essential to improving our understanding of the dynamic factors that influence the process of health communication. Many potential sources of shared understanding, such as ethnicity, training for a specific occupation, education, age, religion, language, gender, and generation, may provide a basis for social groupings within which cultural processes unfold and contribute to intra-
cultural variation because individuals participate in or are exposed to different cultural processes.
Third, we focus on key social science theoretical contributions to communication interventions and their dissemination. Social cognitive and behavioral theories have brought an important conceptual and structural integrity to communication intervention design and evaluation. Rather than reviewing individual theories, we describe how health behavior is determined by underlying beliefs and how these beliefs are shaped by communication. We discuss the contributions of behavioral and communication theories to conceptualize the relevant beliefs and behaviors and to hypothesize which interventions will affect them. In addition, we call on theories of diffusion of innovations (1) to understand which communication-based interventions might be disseminated and what might happen to those interventions and their impacts in the process of dissemination, and (2) to understand better the process of dissemination itself.
Fourth, we broaden the context for our main questions by considering the equity of benefits of communication interventions across diverse groups and the role of communication in reducing the disparities among these groups. Furthermore, because communication interventions often use persuasive techniques, we must examine the ethical issues involved in attempting to persuade diverse populations that include vulnerable groups to change their behaviors in the interest of public health goals.
We believe exciting new opportunities are present to achieve the goal of maximizing health benefits of communication strategies for a wide range of socioculturally diverse groups. Pursuing this goal involves understanding basic principles of communication interventions, such as how messages are constructed for targeted audiences and how they are effectively conveyed. Messages can be made more coherent and powerful and can be evaluated more effectively by organizing communication interventions around well-constructed, extensively studied social science theories. However, doing so requires consideration of what the specific beliefs and behaviors are for populations of interest and how
sociocultural groups that differ in their beliefs and behaviors may respond to or be affected differently by communication interventions.
The committee’s strategy in preparing this volume was to use an interdisciplinary approach to examine the current state of health communication to diverse populations and to suggest fruitful approaches for developing improved interventions in the future. More specifically, by drawing on the relevant disciplines, we explore two questions: (1) Can we improve our ability to optimize public health in the context of diversity through communication? (2) Can we improve the science base in this area?
Conducting a comprehensive evaluation of health communication is a challenging task given the size and complexity of the activities in this area. To make the task more manageable, we developed an approach to analyze the field through a broad examination of three exemplars. One exemplar is national or large-scale regional public health campaigns. The second is communication to improve mammography or breast cancer screening rates. The third area is the management of and outcomes for diabetes mellitus. These three exemplars offer our analysis a focus on a communication method (public health campaign), a public health problem (mammography screening), and disease management that spans public health and medical practice concerns (diabetes mellitus). For all three cases, a prime object of intervention is changing beliefs, intentions to act, or behaviors themselves, such as modifying personal health practices, accessing health care, or adhering to treatment recommendations.
All three involve areas of public health and medicine. For national campaigns and screening mammography, there has been extensive development of health communication strategies for large groups of people; for diabetes mellitus, few systematic communication efforts have been implemented. This area offers an important opportunity for the future development and testing of communication interventions to change behavior. There are surely other health conditions or types of communication intervention strategies in which the benefits of communication strategies for
improving health across diverse groups differ from what we observe for these three areas. However, our selected cases represent major areas of national attention and provide the opportunity to examine different methodological, theoretical, and research approaches to the design and conduct of communication strategies to change health behavior. We note that communication strategies are sometimes only one part of multicomponent interventions (e.g., to improve health) that combine communication and resources or that intervene both at group and individual levels. In reviewing the literature, we attended to interventions that use communication strategies exclusively as well as to those that include such strategies as a major component.
ORGANIZATION OF THE VOLUME
The volume is intended to inform policy makers, researchers, funders/sponsors, advocacy organizations, practitioners, and others in providing and evaluating communication interventions to improve public health. It is organized into eight chapters. Chapter 2 provides a framework for the committee’s review and analysis by examining the contributions of behavior and communication theory to the design, development, implementation, and evaluation of health communication to diverse populations. Chapters 3 through 5 present the three exemplars—national campaigns, mammography, and diabetes mellitus. As noted, our review of the diabetes literature resulted in the conclusion that little systematic work has been done on communication campaigns to change behavior in this area. As a result, the committee presents a discussion of diabetes as a challenge for health communication developers in the future. Chapter 6 reviews the state of technology and assesses its potential. Chapter 7 discusses the problems associated with developing appropriate definitions of diversity and provides guidance for a new direction. Chapter 8 presents the committee’s findings and recommendations.