NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Panel on Race, Ethnicity, and Health in Later Life; Bulatao RA, Anderson NB, editors. Understanding Racial and Ethnic Differences in Health in Late Life: A Research Agenda. Washington (DC): National Academies Press (US); 2004.

Cover of Understanding Racial and Ethnic Differences in Health in Late Life

Understanding Racial and Ethnic Differences in Health in Late Life: A Research Agenda.

Show details

7Prejudice and Discrimination

Studies of race and health frequently invoke racism, prejudice, and discrimination as possible reasons for high levels of morbidity and mortality among black (Jackson et al., 1996; Krieger, 1999; Williams and Neighbors, 2001) and among other racial and ethnic minorities (e.g., Amaro et al., 1987; Salgado de Snyder, 1987). Definitions of these terms vary, and no definitions are universally accepted (Clark, 2004). For our purposes, we use these terms somewhat interchangeably as indicating negative attitudes toward or biased treatment of one group by another (Williams et al., 2003).

Various types of racism have been described (Jones, 1997): personal, which may be considered the same as prejudice (Allport, 1958); institutional, involving a set of environmental conditions, such as housing market conditions, that favors one group over another; and cultural, referring to shared beliefs about the superiority of one group over another. Racism also often involves control by one group over resources that another group wants or needs (Jones, 1997).

Discrimination refers to unequal treatment based on group membership. What actual perceptions, attitudes, or behaviors these constructs refer to depends on the context—the nature and timing of events, their frequency, severity, and duration, whether they are acute or chronic—and on how they are perceived and interpreted, whether intent is attributed, and how they may later be distorted in memory (Williams et al., 2003).

Prejudice, discrimination, and racism could affect health in several ways. First, discrimination could determine a group's living conditions and life chances, affecting such areas as education, employment, and housing. As we note above, low socioeconomic status is one of the most important predictors of adverse changes in health status (Anderson and Armstead, 1995; Williams, 1990; Williams and Collins, 1995), though the specific mechanisms by which low status compromises health have yet to be adequately elucidated (Anderson and Armstead, 1995; Clark et al., 1999). Similarly, all the mechanisms by which discrimination limits economic and social opportunities still need to be fully accounted for (Williams and Collins, 2001), but that it has historically had an effect on minority socioeconomic status is unquestioned.

Second, discrimination could lead to differences in access to and quality of health care (Blendon et al., 1989; Council on Ethical and Judicial Affairs, 1990; Institute of Medicine, 2002), a possibility we examine in Chapter 10. Third, the experience of specific incidents of unfair treatment on the basis of race or ethnicity may generate psychic distress and other changes in physiological processes that adversely affect health (Clark, 2004; Clark et al., 1999; Landrine and Klonoff, 1996; McNeilly et al., 1996). Fourth, some of the coping strategies that people use as they grapple with inequitable living conditions and a hostile psychosocial environment, such as internalizing negative stereotypes (White et al., 2000) or using drugs and alcohol (Jackson and Ramon, 2002), may also impair physical and psychological functioning (Clark, 2004).

We focus in the rest of this chapter on the third effect, with some reference to the fourth.

Early literature on black health, especially mental health, reflects a clear consensus that racism and discrimination have adverse effects (e.g., McCarthy and Yancey, 1971). That some degree of discrimination continues is clear: for example, audit studies continue to document discrimination in housing and employment (Fix and Struyk, 1993). However, there have been comparatively few attempts to explore empirically the health effects of such discrimination among blacks, whether on children, adolescents, or adults (Jackson et al., 1996; Landrine and Klonoff, 1996; Thompson, 1996; Utsey and Ponterotto, 1996). There have been even fewer empirical studies of any kind on other racial and ethnic groups (Williams et al., 2003). Researchers have continued to note that discrimination is an important factor in understanding black health status, and some suggest that it may account for particular patterns of association (Landrine and Klonoff, 1996). Fernando (1984) even proposed that racial discrimination does not just add to stress; it is an actual pathogen. Nevertheless, these constructs and arguments have received limited empirical attention (Harrell et al., 1998; Krieger, 1999), especially as they relate to the life course and aging.

The evidence that the experience of discrimination affects health outcomes is therefore spotty. The majority of reports that have looked at this issue do document an association between the experience of unequal treatment and a variety of health outcomes, including psychological distress, blood pressure, and mental health functioning (Harrell et al., 2003). But prospective studies of the long-term effects of chronic discrimination have not been conducted.


A recent review (Williams et al., 2003) identifies 53 separate community-based epidemiological studies of the association of experiences and perceptions of discrimination with health outcomes. Most of the U.S. studies have involved blacks, but there are some studies of other minorities, and studies have been conducted with immigrant groups in such other countries as Canada, England, and the Netherlands. The majority of the studies find that the self-reported experience of discrimination has an unfavorable effect, producing psychological distress, reduced psychological well-being, lowered self-esteem, impaired mental health, and even definable psychiatric disorders.

These correlational studies have also commonly examined self-reported overall health. More than 70 percent of the studies report poorer health among those who report discrimination. The studies have shown somewhat more variable relationships of discrimination to more specific health indicators. Blood pressure, an important health status measure, has sometimes been positively associated with discrimination, but sometimes has had no association or even a negative association. Cigarette smoking and alcohol use have also been linked to discrimination. Some studies attempt to show that perceptions of discrimination, net of socioeconomic factors, account for racially related health differences.

Systematic investigation of the role of discrimination in health over the life course is rare (Williams and Neighbors, 2001). One longitudinal panel study did find that reports of discrimination that were related to poorer health in the first year were still linked, 13 years later, to poorer mental health, though by that time they were related, somewhat surprisingly, to better self-reported physical health (Jackson et al., 1996). Another study that used reports of the experience of chronic discrimination found that these were related to subclinical carotid artery disease for black but not white premenopausal women (Troxel et al., 2003).


Harrell et al. (2003) recently reviewed 13 experimental studies of discrimination and health. Some studies focused on using analogs to racially charged stimuli in the laboratory and examining physiological reactions. For the most part, these studies show that such stimuli increase physiological arousal. What is unclear is how this response differs from arousal due to other stressors, such as those that would provoke anger.

A second set of studies tested the significance of past sensitization to racist stressors, with individuals who previously experienced discrimination assigned to various experimental conditions. Harrell et al. (2003) report, for instance, that individuals who embrace basic American values tend to be more reactive to racist material than other people.

Finally, a series of studies investigated whether physiological response is moderated by cultural affinities or personality factors, such as “John Henryism,” a dispositional orientation that leads individuals to work hard in the face of impossible barriers (James et al., 1984). These studies have shown mixed results. Harrell et al. (2003) argue that they show the need for more and better studies of basic physiological processes, particularly on cholinergic pathways that link anxiety and stress to cardiovascular reactions. To explain these linkages, the authors propose new models of allostasis and allodynamism that define physiological set points and the mechanisms that govern them. The argument is that both external stressors, such as the experience of discrimination, and internal processes alter these physiological set points, which has health implications. Harrell et al. (2003) suggest that studies in this area might use pharmacological blocks and brain imaging.


These correlational and experimental studies suggest that the subjective experience of bias and unequal treatment could affect particular health outcomes. However, the evidence is uneven and inconclusive, as almost every individual study has substantial inadequacies. Across the variety of studies, the definition and measurement of the factors of prejudice, racism, discrimination, and resulting unequal treatment are still relatively crude. Another problem is uncertain delineation of physiological pathways that serve as conduits for the effects of such factors on health. In addition, the conduits undoubtedly are affected by a host of contextual factors, such as socioeconomic status, individual host resistance factors, and coping styles and responses, as well as varying by age and possibly period and cohort.

Research into the effects of prejudice and discrimination on health differences requires some systematization. Such constructs as prejudice, discrimination, and racism have shifting definitions across studies and are often poorly operationalized. The confusion from continual redefinition makes it difficult for studies to build on one another. Measurement is also a problem, particularly the determination of discrimination from self-reports, which is the usual practice in nonexperimental studies. Response biases are possible in such data and may not be independent of response biases in self-reported health status (Williams et al., 2003). Biases could even affect longitudinal studies, when prior experiences are reinterpreted in the light of subsequent events, though some closed-cohort longitudinal studies suggest this is not a critical issue (Jackson et al., 1996).

Methodological problems go beyond measurement, however, and require better study design (Krieger, 1999; Williams et al., 2003). Longitudinal studies are clearly superior to correlational studies (and avoid the methodological and ethical issues involved in discriminatory treatment of experimental subjects), but they also have limitations, which they generally share with other studies of the effects of stress. Selection processes, memory distortions, and period events with broad effects on cohorts can all complicate the design and interpretation of results. Dealing with all such issues in an efficient design would be the goal, but it is not easy to achieve.

Research Need 10: Determine the lifetime effects of prejudice and discrimination on health using longitudinal data and a framework that centers on stress and its effects.

Stressful events and experiences have been reliably linked to heath outcomes, as we discuss in the next chapter. However, what roots stress may have in prejudice and discrimination (Myers and Hwang, 2004; Pearlin, 1989) require better delineation. There is a need to distinguish among traumatic events and between macro- and microstressors (Williams et al., 2003), and the relationships may be complicated. Discriminatory experiences may combine with other life stressors that affect health. But stress resulting from discrimination may be less easy to deal with through normal coping responses than stress from other sources, and different groups may have generally different ways of dealing with stress. For instance, active and passive coping responses work as well for blacks in response to normal life stressors (low income, negative life events, deaths of relatives and friends, etc.) as for other groups, but blacks are reported to have relatively few effective coping responses to poor treatment due to racial prejudice (Jackson et al., 2003).

The effects of discrimination on the experience of stress and health outcomes may involve lags and host resistance factors and may change over the life course, influenced by personality and other life experiences, such as resource acquisition, exposure, and support processes. Effects related to aging have to be seen in the context of period and cohort variation. Experiences of discrimination may be tied to particular periods or significant historical events (such as the 1960s civil rights movement). And birth cohorts each have their own history, possibly reacting to events differently because of the stage in the life course at which the events are experienced. A framework that combines aging, period, and cohort factors is therefore needed to understand how early experiences may lead to a cascade of subsequent health-relevant events and how experiences may have different effects over the life course. Such a framework is also needed to put scientific observations in context, since these observations necessarily pertain to particular periods and may be of limited relevance to individuals late in the life course. Models for the complex biopsychosocial processes involved in stress reactions to the experience of discrimination also require development (Clark et al., 1999; Harrell et al., 2003; Williams et al., 2003), as we discuss further in the next chapter.

Research Need 11: Evaluate the effects of prejudice and discrimination on the health of minorities other than blacks.

Other racial and ethnic groups, such as American Indians and Alaska Natives, have been subject to prejudice over long periods. Immigrants have also been discriminated against, though as they assimilate and new immigrants enter, the targets shift. Arab Americans and Muslims are the latest to feel targeted. Yet indicators for the health of older adults in these groups are more favorable than indicators for blacks—and indicators actually deteriorate for immigrants as they assimilate and prejudice presumably declines. Does prejudice have effects on health in these groups, but are the effects counterbalanced by other factors, such as immigrant selectivity or better socioeconomic status? Or is prejudice against these groups weaker or less pervasive, of a different quality, or for some reason less consequential for health than among blacks? The answers could have implications not only for these racial and ethnic groups, but also for understanding the mechanisms that link prejudice and health for any group.

Copyright © 2004, National Academies.
Bookshelf ID: NBK24680


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (790K)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...