THE Institute of Medicine report Unequal Treatment1 outlines a disturbing picture of health disparities in contemporary United States. Persuasive evidence presented in that report and elsewhere links race/ethnicity to worse health outcomes for individuals who are members of historically underserved groups. Disparate survival rates for underserved segments of the US society have been documented with respect to heart disease, cerebrovascular disease, malignant neoplasms, diabetes mellitus, and other conditions. Economic security, access to health insurance, and adequate resources for personal and family health are key factors in an individual's health and well-being. Even controlling for these factors, perceived discrimination adds a further component and has been associated with such negative health outcomes as hypertension, cardiac disease, major depression, early initiation of substance abuse, and low birth weight. On the positive side, self-care and environmental enhancement can contribute substantially to the amelioration of these conditions. Taken together, the current body of literature does not identify a single, or even multiple, variable(s) that can fully explain health disparities.

In her widely disseminated book Nickel and Dimed: On (Not) Getting By in America, Barbara Ehrenreich2 set out to describe the plight of women working at low-wage jobs. It is important to point out that Ehrenreich's report was journalistic, not a research report. Indeed, her book was intended for popular, not scientific, consumption. Scientists typically strive to present objective, unbiased data and avoid personalizing their research in this way. On the other hand, a report such as Nickel and Dimed can help scientists follow the wise advice of an articulate Community Advisory Board member who noted, "Researchers have got to learn that they are dealing with human beings NOT subjects if they expect to do good research!!" (Letitia J. Byrd, Member University of Michigan Minority Health Research Program Community Advisory Board, February 16, 2006) Ehrenreich's account of what it was like to work and to attempt to cover living expenses as an employee in various minimum-wage jobs adds valuable nuances to scientific reports about the situation of low-income women. Her qualitatively thick examples enable scientists who seek to understand health disparities to put a human face on the anonymous statistics. These insights may enhance scientific capacity to better understand the context that sustains health disparities and to develop interventions to address effectively the unacceptable health disparities that are the reality in the United States today.

This article examines current circumstances in the United States that increase health disparities. Minority women are particularly vulnerable because of environmental stress and poor nutrition. Furthermore, absent or inadequate healthcare coverage deters the use of preventive healthcare practices. Organizing themes are as follows: underserved, which describes factors that contribute to health disparities and examines the consequences particularly for underserved minority groups; understudied, which examines potentially fruitful, but as yet insufficiently investigated, avenues of research needed to better understand the basis for health disparities; and underestimated, which argues that without full-blown researcher-community partnerships, interventions to address health disparities will be flawed and ineffective.

UNDERSERVED: FACTORS THAT CONTRIBUTE TO DISPARATE HEALTH OUTCOMES

In her report of life on minimum wage, Ehrenreich aimed to document what it was like to try to stretch a low-wage salary to pay for basics like housing, food, and clothing. She lived in circumstances that many of us never confront. Ehrenreich has a PhD in sociology, has written for the popular press for many years, and obviously is not living in poverty in her real life. For the scientist who aims to understand the insidious nature of health disparities, an important fact is that Ehrenreich's life below the poverty line had clear health consequences that her education alone did not equip her to surmount.

On average, Ehrenreich was able to secure employment for which she earned $6 to $7 an hour, a wage that yielded a monthly income of about $1,200. Sadly, this is typical for many women who are more likely to be low paid than are male workers: in comparison to 41% of the male workforce, 59% of women work in jobs that pay below the US official poverty level.3

For low-wage earners it is difficult to ensure adequate nutrition. For example, Ehrenreich recounts that despite her awareness of the importance of nutrition, she was compelled to obtain her meals at fast-food places or at convenience stores. This circumstance stemmed from the fact that the living accommodations that she could afford were in poor areas that rarely had supermarkets close by. As a result, she ate disproportionate amounts of canned food and noodles, and very little fresh produce.

Women in low-wage jobs also encounter health hazards that are not consistently recognized. In a 1994 report on occupational hazards for women, the World Health Organisation pointed to psychological and ergonomic stressors encountered by piece workers or workers on an assembly line, exposure to toxic substances in the microelectronics industry, and repetitive strain injury among housekeepers and health workers.4 For example, as a domestic worker Ehrenreich was required to strap a 40-lb vacuum cleaner on her back and to move rapidly through each home, cleaning in high places as well as carpets, and stooping to vacuum under furniture. She describes living in a world of pain that required management with regular doses of over-the-counter pain medications. Work conditions such as these understandably can lead to chronically poor health.

While the experiences that Ehrenreich described were riveting, the reader was always aware that the author was in the situation by choice and would be returning to a comfortable middle-class life. However, the impact of poverty on healthcare is an ever-present reality for those who live below the poverty line. The November 6, 2005, issue of the New York Times reported just such a situation as experienced by a real-life woman.

Ms Jamie McFarland (not her real name) lived in Yonkers, New York, and had a job as a security guard several miles away. She was paid $9 an hour, or $1,548 a month, and on this income supported herself and her 2 school-aged children. She took the bus to work each day and was managing, until November 2005 when the New York City Bus Driver Union went on strike. When Ms McFarland's bus transportation to her job was no longer available, she had to rely on taxi cabs, which increased her daily transportation cost by a factor of nearly 10. After several weeks of this, she found that the money she had left barely covered the rent for her apartment. She and her young son and daughter subsisted for weeks on rice and beans because she could buy these items cheaply in bulk. In describing her experience, she noted, "We were down to peanut butter and jelly sandwiches for breakfast, lunch, and dinner because I couldn't afford to buy food." It is not difficult to imagine the crisis that someone in Ms McFarland's situation would encounter, should her daughter develop a serious health condition. It is also disturbingly clear that there were no funds in her tight budget for any preventive healthcare. At her age she should be having annual Pap tests and mammograms, but it is highly unlikely that she does. As a result, she carries a far greater risk of being diagnosed with advanced, even untreatable, cancer of the cervix or breast.

It is well documented in an Institute of Medicine1 report that individuals who are African American, Hispanic, or American Indian often receive inadequate healthcare and have morbidity and mortality outcomes that reflect this discrepancy. The report indicates that healthcare disparities are consistent and extensive across a wide range of medical conditions and healthcare services and that they are unquestionably associated with worse health outcomes. The risk of not having health insurance is much higher in the United States for Hispanics, American Indians, African Americans, or Asian Americans (particularly South Pacific Islanders) in comparison to non-Hispanic Whites.5 Without healthcare insurance the cost of even maintenance care for a chronic disease such as diabetes mellitus or hypertension is daunting. And preventive healthcare is a remote luxury.

UNDERSTUDIED: KNOWLEDGE NEEDED TO INFORM THE HEALTH DISPARITIES RESEARCH AGENDA

Researchers attempting to better understand and to develop interventions to address health disparities are confronted with gaps where insufficient knowledge blocks progress. Understudied areas include the impact of institutional and internalized racism, the challenge to low-wage workers as they strive to combine work and family roles, and the contribution of a residence in a disadvantaged neighborhood.

Institutional/internalized racism

Despite the substantial role that income plays in disparate health outcomes, this factor does not explain all of the variance, given that disparities in healthcare occur by race/ethnicity across all levels of income. For example, Williams reported marked racial/ethnic differences among women for the health indicators of hypertension and overweight.6 White women had lower levels of both hypertension and overweight than their African American and Mexican American counterparts. Rates of hypertension were about 1.8 times as high for African American women and increased for American in comparison to those for White women, and both African American and Mexican American women were more than 1.5 times as likely to be overweight as White women. Several patterns can be seen: Income is strongly linked to hypertension for African American and White women and to overweight for White and Mexican American women. Overall, women with lower levels of income have worse health than their economically favored counterparts. However, income was unrelated to hypertension for Mexican American women and was not strongly associated with overweight for African American women. Moreover, racial differences persist at every level of socioeconomic status, emphasizing that race is more powerful than socioeconomic status. Williams6 concluded that this pattern may reflect complex interactions between racial/ethnic status and migration history or culture, long-term effects of exposure to social and economic adversity during childhood, independent contributions of institutional and individual discrimination, or the noncomparability of socioeconomic status indicators across racial/ethnic populations. Because women in the United States (particularly women of color and their children) are more likely to live below the poverty line, they are in triple jeopardy (gender, race/ethnicity, and decreased economic power). Women's location in the structure of society places them where they have diminished access to power, privilege, and resources. And, poor women bear the additional burden of being more likely to find themselves in abusive relationships, a factor that carries with it increased risk for major depression and for substance abuse.7,8

A recent literature review categorized more than 60 studies that examined different health outcomes that could be linked to discrimination.9 The studies were categorized according to those that demonstrated positive, negative, conditional, or no association between individuals' experience with perceived discrimination and such markers of health as mental health, physical health, and health behavior. Most of the study results demonstrated a positive association. That is, individuals who perceived higher levels of racial or ethnic discrimination were likely to have more psychological distress, major depression, hypertension, mortality, or to smoke or use excess alcohol. Although some studies were inconclusive and showed no association, no studies were found that demonstrated a negative association. If that had been true, the higher the perceived discrimination the lower the levels of hypertension, depression, or other condition. Such a relationship was never observed. Conditional associations pointed to avenues of study that could add a more nuanced understanding of how discrimination can affect health. For example, perceptions of discrimination accounted for elevated levels of hypertension in some reviewed studies, while in others the effect was dependent on such factors as coping style, sex, social class, or ethnicity. With respect to other cardiovascular outcomes, it is noteworthy that a positive association between chronic daily discrimination and the development of atherosclerotic disease (in the form of intima media thickness) has been demonstrated in the carotid artery among African American women, but not among White women.10

The 2004 Annual Review of Nursing Research, focused on eliminating health disparities among racial and ethnic minorities in the United States, reviewed more than 40 articles that provide evidence for structural and racial barriers to healthcare.11 In addition to lower rates of health insurance and lower socioeconomic status, other barriers to healthcare included lower rates of health literacy, lower fluency in English, certain cultural and spiritual beliefs, and a lack of confidence in healthcare providers. The perception is strong among minority groups that the (predominantly White) healthcare system is uncaring. Qualitative investigations with women who were primarily African American found women reluctant to use health services because of perceived racism and bias, long periods of waiting time to see the healthcare provider, and the negative manner with which they were approached and addressed.12,13 Clearly, perceived discrimination must be investigated more intensively and with consideration of the potential moderating effects of coping style and environmental factors.

Conflicting family and work demands

The lack of healthcare insurance is a major explanatory variable in health disparities. Hence the capacity of a woman to secure and retain a job that provides this benefit is of great importance. However, the challenge to combine work and family roles is more daunting for women in low-wage jobs, as Susan Lambert,14 who has studied these special challenges, points out. Ensuring that non-school-age children will be cared for involves substantial juggling and often entails reliance on other women whose circumstances are similarly challenging. Recommended research to better understand the nature of these issues includes investigation of family well-being as women move off welfare, particularly examination of different outcomes associated with various welfare-to-work programs; special needs and circumstances of lower-wage workers and their families, particularly small differences in job design that may enhance the balance of work and family roles; and investigation of programs to more effectively recruit and retain lower-wage workers, such as mentoring programs, increased flexibility of job requirements, and increased opportunities for advancement. Other scientists argue for a better fit between social resources and the reality of low-wage women's lives; they suggest expanded hours and days of the week when women might interface with agency representatives, incorporating modes beyond face-to-face such as voice mail or online communication, and expanded services such as child care, housing assistance, job search counseling, and clothes-for-work banks.15

Residing in a disadvantaged neighborhood

The simple fact of living in a disadvantaged neighborhood has been shown to have direct effects on residents' health. With respect to the risk for coronary heart disease in women, for example, the residents of disadvantaged neighborhoods have been shown to have increasing risk as socioenvironmental characteristics become less favorable.16 Even controlling for individual-level variables, women in less-well-off neighborhoods were at greater risk for smoking, elevated systolic blood pressure, and higher serum cholesterol levels-findings that suggest neighborhood social context is a factor to be considered in linking social structure to coronary heart disease.17,18 While this linkage is far from simple, some aspects seem straightforward. For example, the spatial arrangement of a disadvantaged neighborhood may not effectively support recommended levels of physical activity, a recognized factor in reducing the risk of coronary heart disease. And when safe, accessible places to walk or work out are not available, as is true of most inner-city neighborhoods, it is not surprising that the rate of exercise is low among inner-city dwellers compared to individuals who live in areas that are safer. More study is needed to disentangle the role of the less-well-off neighborhood per se and the absence of specific amenities such as safe parks for walking in order to better understand health disparities.

Along these lines, Balfour and Kaplan19 linked preservation of functional status in the elderly to such neighborhood characteristics as traffic, noise, and inadequate lighting. Importantly, these investigators used a longitudinal design that enabled them to show a neighborhood effect that persisted after taking baseline functional status into account. A critical review of 25 studies examining the influence of neighborhood characteristics on health determined that all but 2 studies reported a significant association between at least 1 measure of social environment and a health outcome after adjusting for individual socioeconomic status.20

Investigations such as those outlined above are needed to disentangle the role of personal choice versus the impact of neighborhood. Considering the factors that support the health behavior of accessing preventive services, one might ask how much of a decision to not access such services is due to personal disinterest versus the burden of, say, transportation that accessing the services imposes? I worked as a nurse practitioner in inner-city Detroit for several years. Breast and cervical cancer screening was provided to uninsured women free of charge, so the economic barrier to this preventive service was removed. Some of my patients traveled for more than 2 hours via bus to get to their screening appointments-they had to transfer several times, each time requiring a long interim wait for the next bus. I am sure that many of their counterparts, especially those who had to travel with several children, elected to skip the free services. Electing not to access these services seems to be more logically attributable to environment than to personal choice. Just getting to a healthcare appointment for women in such circumstances can be considered a source of environmental stress-especially if the woman must take unpaid time off from work. On the other hand, investigations into what factors enabled significant numbers of women to schedule and to keep preventive screening appointments would provide valuable data about current strengths in the patient base that might be built into service models.

Hogue and Bremner21 present a model that emphasizes consideration of the impact of environmental stress, which broadly encompasses such factors as institutional/ internalized racism, conflicting family and work demands, and residence in a disadvantaged neighborhood. Stress age includes genetics and environment, personality, and other factors. A consistent level of environmental stress in the form of daily or weekly exposure to institutional racism, for example, could well take its toll on health status, increasing the stress age of an individual. This factor must be taken into consideration when one examines such negative health outcomes as the higher prevalence of hypertension and of low-birth-weight infants in the African American community. Factors that might moderate these negative outcomes are such situational factors as personal resilience or economic resources. All of these factors likely play a role in disparate health outcomes for underserved groups. We know the disparities exist, but there are great gaps in our understanding of the reasons why. Research is needed to explicate the mechanism(s) that could guide our efforts to develop interventions to reduce and eventually eliminate health disparities.

UNDERESTIMATED: COMMUNITY PARTNERS ARE ESSENTIAL TO THE DEVELOPMENT OF EFFECTIVE INTERVENTIONS TO ADDRESS HEALTH DISPARITIES

Many factors account for health disparities that must be understood by scientists before the problem of disparities can be adequately addressed. A major factor that mandates vigorous investigation is the impact of socially determined exposure. To advance our understanding in this area, scientists must make marked changes in the way research is done, specifically researcher-community partnerships.

To pose and answer the most relevant clinical research questions with respect to health disparities, the community must be an integral part of the research process, from conceptualization of the research question to translation of research results. This was affirmed by the recent report examining the Health Disparities Research Plan of the NIH: Unfinished Business, which recognizes the essential role of community input into well-designed clinical research and recommends the inclusion of community-based participatory research as an objective.22 This underlines the growing recognition of the role that an engaged community must play in the development, conduct, and interpretation of effective clinical and translational research. However, this position is not a comfortable one for many scientists whose training has emphasized the development of specialized expertise. It may be difficult for such an expert to acknowledge that members of underserved communities have critical expertise about priority community concerns and essential insights into how a particular intervention can be best adopted and maintained. As a result, many researcher-community partnerships have been marred by mistrust and ineffective, disjunctive collaborations.

In a workshop sponsored by the NIH Council of Public Representatives, the Council specified that to improve and enhance the state of clinical research in this country it is "essential to build trust and relationships among all stakeholders."23 As a result of a growing awareness of healthcare disparities, many scientists now strive to obtain diverse sample representation. However, scientists must recognize that the issue is much broader than simply enhancing recruitment of a diverse sample. Rather, the community must be seen as an essential component at each phase of the clinical and translational research process. In fact, clinical and translational research that is conceptualized, implemented, and interpreted in a scientific vacuum, without engaged input from potential participants, is limited in its capacity to yield a valid test of the intervention.24 The absence of collaborative, trusting relationships with the community is of particular concern when one considers the possibility that this lack deters scientific capacity to address effectively healthcare disparities.25

Distrust of science and of public institutions, particularly within communities of color, hinders the development of productive, collaborative clinical and translational research relationships.26,27 Authoritative knowledge held by members of underserved groups has substantial implications for the development of meaningful research questions and selection of the most effective and ethical research methods.28,29 For example, an intervention study of physical activity to reduce obesity in the Latino community will not succeed if the research design fails to take environmental safety into account. Community input into the design phase of the project would avoid such an omission.

Three barriers hinder the development of collaborative, trusting community-university research partnerships:

1. Lack of stable infrastructure: Historically, many communities that have participated in clinical research studies have experienced resource withdrawal when an individual project ends.27 Although principal investigators decry "helicopter research" that swoops in to conduct an investigation and then departs without a trace, the reality is that when a funded project is completed, the research resources typically do vanish. This occurs whether the community is the inner city, rural America, or a service agency. Such discontinuity understandably fuels mistrust about subsequent research project participation. It also impedes researcher access to engaged community members who could make valuable recommendations in the early phases of the conceptualization of a research project.30

2. Lack of knowledge or cultural sensitivity (or both) on the part of both investigators and community members: Typically, scientists are not trained to value and incorporate community expertise into the full continuum of the research process.31 For example, funding streams have not supported the time and process to incorporate community expertise for project conceptualization and design. Moreover, community members may not have the skills and understanding of scientific precepts that enable them to fully contribute to project success. Cultural insensitivity is most problematic when it is on the part of the investigator and impedes the investigation of factors causing health disparities.32 However, insensitivity to the research culture on the part of community members, for example, random assignment, also negatively affects community trust.33 These negative events can be avoided with a community-based "collaborative process that equitably involves all partners in the research process and recognizes the unique strengths each brings."34^(p177)

3. Inadequate study design: Despite mistrust, racial/ethnic minority individuals are as willing to participate in health research as are their nonminority counterparts when the aims of the research and potential benefits to health are made clear to them and when barriers to participation are reduced.35 The onus is on the scientific community to communicate adequately the research aims and benefits to underserved groups and to reduce factors that deter their participation. This charge is best handled by deliberately inviting the expertise of community members to inform the research methods and supporting the engagement of community members in the process. For example, underrepresentation of minority groups can result from limitations imposed by cultural insensitivity on the part of the investigator.32,36 Such limitations include inadequate communication of the invitation to members of underrepresented groups regarding participation/eligibility criteria; use of data collection sites that are less accessible to minority groups; failure to provide written and verbal communication in grade-appropriate text or languages other than English; insufficient consideration of services and remuneration, such as child care and transportation reimbursement; and the lack of minority group knowledge about the importance of clinical research in improving the public health. Inattention to these issues can substantively reduce minority participation. Input from community members with decision-making authority and faculty with expertise in community-based research can alert scientists to the need to budget for and incorporate elements into their studies that will greatly enhance minority participation and retention and strengthen the validity of findings.

Moreover, translational research requires a 2-way transfer of knowledge: in one direction, efficacious scientific discoveries are applied in effective clinical interventions; in the other direction, clinical insights and knowledge of unmet community health needs flow to researchers to inform new hypotheses. Without infrastructure to support community participation and a bidirectional knowledge transfer, sound clinical and translational research is seriously compromised. Hence, the relationship between the scientific community and the public must be collaborative and grounded in mutual respect.37,38

To accomplish the recommended researcher-community partnership, scientists must learn how to work closely with community partners and to apply principles of effective community engagement in order to overcome these barriers. Scientists must honestly appraise the benefit that various research projects have contributed to the community and ask, How sustainable are our interventions once the period of funding ends? What tangible and long-lasting outcomes can we identify? After the project is over, what economic benefit do we leave for our community partners?

Many scientists have an honest desire to benefit the health of the community, but if an audit of the tangible outcomes of various community-based projects were conducted, it would most likely reveal a series of published articles (with grateful acknowledgment of funding from the National Institutes of Health or some other sponsor, because we know how critical that is to future funding), limited evidence of community partner input into the research questions to be asked, and limited evidence of community partner involvement at every step of the research process.

A paradigm shift is needed to generate research that makes more of a direct contribution to the public's health. This transition begins with a reconceptualization of the role of community partners. Currently, positive community relationships are viewed as a means to an end that will ensure that the community has sufficient regard for the project so that community leaders will encourage other community members to enroll. Scientists tend to be satisfied if the sample demographics reflect diversity. The notion of community engagement must be expanded so that research projects embody the following principles of effective community-based participatory research34 in the research plan:

* Facilitation of collaborative, equitable partnerships in all phases of clinical and translational research;

* Involvement of empowering and power-sharing processes;

* Utilization of a co-learning process;

* Development of systems that support local capacity building;

* Achievement of a balance between research and action;

* Establishment of an ongoing presence that is not project dependent.

A model research project that has effectively incorporated these principles is the East Side Village Health Worker Partnership in Detroit.39 Following the principles of community-based participatory research, these scientists involved members of the community from the project's inception. They accomplished this by establishing a community advisory board through which community members participated in discussions of the research questions, research design, data collection and analysis, and interpretation of results. As the project evolved, community advisory board members brought invaluable insight with regard to the importance of placing priorities on the community concerns of policing and safety, economic development, support for parents, and health services, with emphasis on cardiovascular disease and diabetes.

Examples of how the additional community-based participatory research principles might be applied are outlined here. Use of empowering and power-sharing processes include structuring community advisory boards to ensure equal or overrepresentation of community members (as opposed to 1 or 2 "token" community members), designating a member of the community to serve as the chair of the community advisory board (and compensating the individual for his or her time and effort), and requiring researchers who aim to conduct their research in a community-based agency to present the proposed research to the board for approval. Co-learning processes involve mutual examination of culturally sensitive areas where both the researcher and the community member have opportunities to gain insight. Local capacity building can be accomplished by hiring, training, and paying community members to assist with the conduct of the research; personnel directories may facilitate additional job opportunities on other research projects for skilled team members. Attention to the balance between research and action will ensure that the results of the research are returned to the community; members of the research team with relevant health promotion expertise can also make their services available to community groups seeking presentations on health topics. Sustaining an ongoing presence is a challenge for an individual principal investigator, whose resources end with the funding period, but can be attained within the mission of an institution. For example, the University of Michigan recently established the Detroit Center in the heart of Detroit. In the context of advancing research within an underserved community, this building provides space to researchers for the conduct of interviews or other data collection, for the implementation of selected protocols, and for public meetings to disseminate research results.

After more than 5 years' presence on Detroit's East Side, the East Side Village Health Worker Partnership has generated tangible outcomes. Various village health workers have been elected to citizens groups such as the Community Relations Board of the local police station. They have formed the Economic Development Action Group, which supports job fairs and microenterprises such as hair design shops. Social support networks have been strengthened by such groups as the Sisters Roundtable Action Group. And programs are in place that promote healthy eating-minimarkets now stock fresh produce at affordable prices, and the Wisdom Strut program offers elders a safe venue for exercise.

Nursing, in partnership with other healthcare disciplines, has great capacity to change the face of healthcare disparities. We can change the way research is being done by doing what nurses do best-that is, respecting the dignity of each person and valuing the individual strength that each person brings whether it is in a healthcare or research encounter. I believe that we must do this, not simply because it is the right thing to do although it certainly is, but because the integrity of our science and of our clinical programs depends upon it. Indeed, if we do not include the voice of the recipient who will ultimately incorporate our intervention into her life and the provider who must deliver it, we cannot be sure that we have developed a feasible intervention. As nurses, this comes as no surprise. Patient-centered care is our creed, and we know that the patient is the major stakeholder in her care. Similarly, the research participant must be viewed through this lens.

In summary, most of us came to nursing with the dream of providing a service that would benefit the health of people. When we value, intentionally seek, and incorporate input from the individual or group whose health we aim to benefit with our research, we increase our chances of developing a feasible, cost-effective intervention. By partnering with diverse members of the community, we move closer to our goal of reducing healthcare disparities and making a tangible impact on the health of the people.

REFERENCES