Abstract
Despite improved hypertension (HTN) awareness and treatment, racial disparities in HTN prevalence persist. An understanding of the biopsychosocial determinants of HTN is necessary to address racial disparities in the prevalence of HTN. This review examines the evidence directly and indirectly linking multiple levels of racism to HTN.
Published empirical research in EBSCO databases investigating the relationships of three levels of racism (individual/interpersonal, internalized, and institutional racism) to HTN was reviewed.
Direct evidence linking individual/interpersonal racism to HTN diagnosis is weak. However, the relationship of individual/ interpersonal racism to ambulatory blood pressure (ABP) is more consistent, with all published studies reporting a positive relationship of interpersonal racism to ABP. There is no direct evidence linking internalized racism to BP. Population-based studies provide some evidence linking institutional racism, in the forms of residential racial segregation (RRS) and incarceration, to HTN incidence. Racism shows associations to stress exposure and reactivity as well as associations to established HTN-related risk factors including obesity, low levels of physical activity and alcohol use. The effects vary by level of racism.
Overall the findings suggest that racism may increase risk for HTN; these effects emerge more clearly for institutional racism than for individual level racism. All levels of racism may influence the prevalence of HTN via stress exposure and reactivity and by fostering conditions that undermine health behaviors, raising the barriers to lifestyle change.
American Journal of Hypertension, advance online publication 17 February 2011; doi:10.1038/ajh.2011.9
Racial disparities in hypertension (HTN) continue to be a pressing problem in the United States. There is consistent evidence that black Americans are more likely than white Americans to develop HTN. Prevalence rates for black adults range from 30.6 to 40.5%; whereas the rates for white range from 24.4 to 29%.1–5 There is also evidence of racial disparities in blood pressure (BP) control (control rates: blacks 44.1–65.2%; whites 55.6–86.3%),5–9 although not all studies have found race differences.10,11
These disparities exist despite the fact that black Americans are more likely to be aware of their HTN12 and in some cases are more likely to receive treatment for HTN than are white Americans.7,8 Even when black Americans are as or more adherent to antihypertensive treatment than are whites6,13 disparities in BP control are manifest. To address the high prevalence of HTN among black Americans, it may be useful to identify other variables, including different psychosocial stressors, that might serve as potential individual-level and environmental risk factors that disproportionately affect black Americans, and to understand the ways in which these variables may operate to increase HTN prevalence.
Racism has been hypothesized to serve as a psychosocial stressor contributing to the excess rates of HTN among black Americans.14–18 The goal of this review is to provide a detailed evaluation of the evidence linking individual/interpersonal, internalized, and institutional racism to HTN and to known risk factors for HTN, including obesity, fitness, and alcohol use, as well as psychosocial stress. We hope to provide an evidence base that can inform further examination of the role of racism in the development and course of HTN.
We specifically investigate the effects of racism on black Americans, because the majority of published research on the relation of racism to HTN has focused on black Americans. It is important to note, though, there are also significant disparities for other ethnic groups.3,6,19,20
Constructs and Definitions
Most broadly, racism has been defined as “the beliefs, attitudes, institutional arrangements, and acts that tend to denigrate individuals or groups because of phenotypic characteristics or ethnic group affiliation.”16 Racism or ethnic discrimination can be considered as a form of social ostracism. Phenotypic or cultural characteristics are used to render individuals outcasts, making them targets of social exclusion, unfair treatment, and harassment; and consequently, either directly or indirectly, depriving them of social and economic opportunities and threatening personal safety.21 Detailed reviews concerning the conceptualization and measurement of racism are available elsewhere.16,18,22
Racism can occur on multiple levels: individual/interpersonal, internalized, and institutional.22,23 Individual-level racism includes episodes of race-based maltreatment that are perpetrated by individuals and targeted at other individuals.17,22 In the context of an interpersonal exchange, these exchanges are considered interpersonal racism, which has been defined as “directly perceived discriminatory interactions between individuals whether in their institutional roles or as public and private individuals.”22 Individual-level racism is typically assessed with self-report surveys inquiring about exposure to acts perceived as discriminatory, unfair, or disrespectful (i.e., refs. 24,25,27). Self-report surveys assess the subset of experiences of ethnicity-related maltreatment that are directly perceived by the target and are generally labeled perceived racism or ethnic discrimination.
Internalized racism is defined as “the acceptance, by marginalized racial populations, of the negative societal beliefs and stereotypes about themselves.”28 The internalization of negative stereotypes about ones' own group may develop in response to repeated exposure to ethnicity-based maltreatment, as a function of cultural communications of attitudes toward stigmatized groups, and from familial or other socialization processes, as well as other mechanisms.29,30 In studies of BP among black individuals, internalized racism has been assessed with a self-report scale (i.e., Nadanolitization scale)31 that measure the degree of agreement with typical stereotypes about the black individuals.
Institutional racism refers to specific policies and/or procedures of institutions (i.e., government, business, schools, churches, etc.) which consistently result in unequal treatment or outcomes for particular groups, even though other nonrace-related factors may also be associated with the disparate outcomes.32,33 Policies resulting in unequal treatment can be considered as a form of racism, despite the absence of evidence of deliberate racial prejudice on the part of the policy-makers. This is the case when majority-group policy-makers are less aware of or responsive to the consequences of these policies for minority stakeholders.34 In general, research on the relationship of institutional racism to HTN has focused on the relationship of BP to the tangible outcomes of these policies, including access to education or health care, residential segregation, incarceration, among other outcomes.35,36 Two outcomes that have been specifically studied in relation to HTN include residential racial segregation (RRS) and incarceration.
Residential segregation refers to “the degree to which groups of people categorized on a variety of scales (race, ethnicity, income) occupy different space within urban areas.”37 We focus on race-based residential segregation (RRS), which is likely to be a function of a number of both historical and current actions on the part of institutions (i.e., real estate developers, lending organizations, employers) as well as the actions of individuals within neighborhoods.37 RRS also serves as a proxy for the extent to which black individuals are ostracized by other groups.38 Across all income groups, blacks tend to live in more racially segregated areas than do whites, but RRS is most pronounced among individuals with low levels of income and education.35 Strategies for conceptualizing and quantifying RRS have been well reviewed elsewhere.37,39,40 Examples of measures include the index of dissimilarity and the proportion of black residents in a given area, a measure used in most studies of HTN despite some limitations to its interpretability.37,41
Rates of incarceration in the criminal justice system can also be regarded as an index of institutional racism.42,43 In comparison to whites, most evidence suggests that black Americans are more likely to be incarcerated, even when controlling for a wide range of case and jurisdiction-related variable.42 These differences have developed in part, because of stereotypes about the propensity of black Americans to be violent, as well as legal and policing policies and practices.42,43
This review extends our prior work and examines studies of adults linking each level of racism to HTN diagnosis or to BP levels (with BP levels serving as a proxy for a documented diagnosis of HTN).14,44 To obtain all relevant studies, we searched all EBSCO-host-related databases, including MEDLINE and Psych Info using the terms: racism, racial discrimination, ethnic discrimination, institutional racism, internalized racism, self-stereotyping, residential segregation, racial segregation, racial residential segregation, and incarceration combined with BP, cardiovascular response, reactivity, HTN, and health. All papers were searched for any additional relevant references. Papers available through August 2010 were included. Table 1 includes the details of all reviewed studies for each level of racism.
To further understand the mechanisms through which racism may affect HTN, we also investigate the relationship of racism to obesity, low levels of fitness, and excess alcohol consumption. Each has been documented to be associated with increased HTN prevalence.20,45 Reductions in these risk factors have been associated with improvements in BP, and they are frequent targets of physician recommendations.45
We include data on psychosocial stress as a risk factor, although the relationship is not as well documented or accepted as lifestyle-related risk factors. Events and conditions are perceived as stressful when they are appraised as salient and threatening, and present demands for coping that are perceived to exceed the individual's resources.46 Both systematic and conceptual reviews suggest that chronic, but not acute stressors are more likely to be associated with increased risk for HTN.47,48
All levels of racism can result in acute stress exposure, but racism is widely regarded as a chronic stressor.16,21,49 Interpersonal racism takes the form of discrete events, including both overt and covert episodes of race-related maltreatment. These acute events can become chronic stressors if they occur frequently and/or if the experience has persistent negative effects. For example, the acute effects of race-related maltreatment may be maintained if the targeted individual experiences constraints on his or her ability to resolve the situation or cope with its aftermath.21,46 Institutional racism is associated with conditions (e.g., residential segregation, incarceration) that present additional obstacles or sustained demands that can act as chronic stressors.
Racism and HTN: Examining the Associations of Interpersonal, Internalized, and Institutionalized Racism to HTN Diagnosis or BP Levels and HTN Related Risk Factors
Individual/interpersonal racism
The bulk of the research on racism and HTN has investigated the effects of individual-level or interpersonal racism.15,17 Most studies employed within-group designs to investigate the degree to which the intensity of exposure to racism affects risk for HTN within black individuals. In our prior review,14 we indicated limited direct relationships of racism to HTN diagnosis. The subsequent publications support this conclusion.
To date there have been 12 observational studies (described in 13 papers) which included black adults and which examined the relationship between self-reported exposure to interpersonal racism and resting BP level (e.g., a mean of two or three readings taken under standardized conditions)47–55 or self-reported or physician-diagnosed hypertensive status.48,56–58 Seven studies did not find a direct relationship between perceived racism and BP when the investigators examined the sample as a whole.47,50,52–54,56,57 Two studies have found a negative relationship either among older participants54 or among the participant group as a whole.59 There are two studies that report a U-shaped relationship of racism to HTN, in which, depending on participants' race, gender, and social class, there were elevated BP levels in those experiencing high levels of racism or no racism vs. moderate levels of racism.48,51 There are only two studies that reported a positive relationship between self-reported racism and either BP level or self-reported diagnosis of HTN either in the group overall49 or in one subgroup (i.e., non-US born women).58 However, one of these studies included a small sample (n = 89), only 18% of whom were black.49
In contrast, the data from ambulatory BP (ABP) monitoring studies are more consistent. ABP, and in particular nocturnal ABP, is regarded as a more reliable predictor of target organ damage than are clinic measures.60 Ambulatory monitoring also captures BP reactivity to daily events. The six studies of adults all reported positive relationships between perceived racism/discrimination and either daytime ABP,61 nighttime ABP or BP dipping,62–65 or both.66
There are substantial variations in the quality of these studies of HTN diagnosis, clinic BP or ABP. Only four of the studies employed population-based or randomly selected samples.55–57,59 As shown in Table 1, some studies employed measures with a small number of items inquiring about discrimination in general or experiences of discrimination in specific venues (i.e., such as work or medical care),48,52,55,56,59 and very little psychometric information was provided about these measures. Others studies included measures that have been subjected to extensive psychometric testing (e.g., Perceived Racism Scale, Everyday Discrimination, or the Perceived Ethnic Discrimination Questionnaire-Community Version).47,49,50,54 The studies of ABP (vs. those of BP level or HTN diagnosis) were more likely to include measures with known and good psychometric properties. However, it is important to note that neither the more limited scales, nor those with good psychometric properties yielded positive effects in studies of clinic BP or HTN status. In contrast, the same scales (i.e., the Perceived Racism Scale, Everyday Discrimination/Unfair treatment) were associated with ABP, even in studies with much smaller samples.61–66
As is the case with all self-report measures, scores on measures of perceived racism may contain some error. The scales measuring perceived racism cannot distinguish between the target's perceptions of racial bias in cases in which these perceptions are accurate (i.e., the perpetrators' actions were motivated by racial bias) vs. those in which the target's perceptions are a function of misperceptions or misattributions to discrimination. To attempt to control for intrapersonal factors such as hostility or neuroticism that might influence the perceptions of racism (and potentially HTN), but may develop from nonracism-related factors (e.g., temperament, family functioning, etc.), some investigators have included measures of personality characteristics as covariates.62,65 Three ABP studies in which measures of negative-affect related traits (e.g., hostility or neuroticism) were included as covariates find that the effects of perceived racism on ABP remain robust and significant.62,64,65
Measures of perceived racism which inquire about discrimination in a variety of venues could elicit answers reflecting perceptions of institutional racism (i.e., perceptions of being mistreated as a function of institutional policies) rather than experiences of interpersonal maltreatment. Racism may also affect an individual's access to economic and social resources, and in turn affect HTN risk through deprivation. Consequently, most investigators included measures of individual level or neighborhood socioeconomic status (SES) as a partial control for the effects of these environmental or institutional variables. In studies in which the effects of SES were explicitly evaluated,67 the inclusion of SES as a control variable did not eliminate the effects of perceived racism/discrimination on ABP.62 However, some studies suggest that SES moderates the effects of racism on BP, although the direction of effects is not consistent and additional work is needed.48,55,57
Individual-level racism may also have health effects in circumstances in which the targeted individual is unaware of the exposure.68,69 Some investigators have advocated the use of measures of unfair treatment or discriminatory behavior (e.g., the Everyday Unfair Treatment Scale68,69) which assess exposure to interpersonal experiences that are likely to be a function of racial discrimination, without requiring participants to attribute the maltreatment to racial bias. These scales can be considered as a measure of the construct “everyday unfair treatment” rather than racial discrimination per se, because individuals can perceive themselves as targeted for unfair treatment for many reasons (i.e., including their social class or gender). Some investigators have included additional questions about the attributions for the maltreatment; however, none of the studies of HTN or ABP in adults included these items.
All studies of HTN and clinic BP employed measures directly referring to race. All ABP studies included measures assessing experiences of unfair or discriminatory treatment in everyday life (i.e., Everyday Unfair Treatment, Perceived Racism Scale, and Perceived Ethnic Discrimination Questionnaire-Community Version). Four of these studies included measures which explicitly refer to race as a cause for the unfair treatment (i.e., Perceived Racism Scale and Perceived Ethnic Discrimination Questionnaire-Community Version),61–63,66 whereas two other studies included measures of unfair treatment that did not explicitly refer to race.64,65 Associations of unfair treatment/discrimination to ABP among blacks were found using either type of measure of unfair treatment.
Interpersonal racism and risk factors for HTN. Although two recent studies reported no concurrent relationship of racism to body mass index,70,71 another prospective investigation reported that increases in interpersonal racism were positively associated with weight gain over a period of 8 years.72 To our knowledge there have been no studies of the relationship of individual-level racism and the intake of specific nutrients such as sodium or potassium. Perceived racism has been associated with greater risk for any level of alcohol use (but not binge or heavy drinking) among black Americans.73 Prospective studies also indicate a relationship of perceived discrimination to increases in alcohol use, partially mediated by discrimination-related changes in psychological distress.74 The one study specifically examining physical activity in a large population-based sample, did not find a relationship with racism.75
In contrast, there is substantial, clear and consistent evidence that individual-level racism is associated with indices of psychological distress (e.g., negative affect, anger, depression, and anxiety), as well as personality characteristics (e.g., hostility, trait negative mood) that increase the experience of distress.15,17,76,77 Racism may also influence cardiovascular responses to stress exposure. BP reactivity to stress has been identified as an independent predictor of the development of HTN.78,79 Several studies report that perceived individual-level racism predicts the magnitude of BP reactivity to laboratory-induced stressors.80–85 However, others studies found no direct relationship of individual-level racism to BP reactivity or recovery, and report that the effects of racism emerged only when moderated by other characteristics (e.g., support or hostility).80,82,86
Interpersonal racism is more consistently related to perceived stress and negative emotions than to lifestyle-related factors, including physical activity or obesity. It is worth noting that some,50,57 although not all,54 studies of racism and HTN found that stress reactions to racism were associated with HTN diagnosis or BP level, even when exposure to race-based maltreatment was not. However, there are still very limited data.
Internalized racism
To our knowledge, there is only one study directly assessing the effects of internalized racism, assessed with a modified version of the Nadanolitization scale on BP in Afro-Caribbean women. The authors did not find a direct relationship of internalized racism to resting BP.87
The measure of internalized racism (i.e., the Nadanolitization scale) assesses the belief that members of one's group have characteristics that correspond to common stereotypes about the group. This may or may not be related to self-stereotyping (i.e., the degree to which the individual has incorporated these stereotypes into his or her self-concept). New research is employing methods from cognitive psychology, including variations on the Implicit Association Test, to assess nonconscious self-stereotyping.88
Internalized racism and HTN risk factors. Internalized racism does not show a relationship with body mass index,87 but is more closely associated with abdominal obesity, with three of four studies reporting a significant positive relationship.89–91 One study suggests that internalized racism is associated with perceived stress among black women.87
Institutional racism
The data linking racial residential segregation to HTN is mixed. RRS has been associated with greater risk for HTN among both black and white mothers, such that mothers living in more racially segregated areas (e.g., areas in which there were high percentages of black individuals and in which residents were less likely to interact with others who were not black) were more likely to report having chronic HTN and to be at risk for pregnancy-related HTN, controlling for neighborhood poverty and other factors.92 Another study reported that the percentage of black individuals living in an area was correlated with the percentage of individuals with HTN, but did not control for other facets of the neighborhood in which black individuals predominate, including higher density and lower cost of housing, variables independently associated with HTN.93
In contrast, two population-based studies of black adults did not find any relationship of the proportion of blacks living in the neighborhood to prevalence of HTN among black adults94 controlling for other neighborhood risks.95 Fang et al. reports that for black adults, there was no effect of residential area on HTN-related mortality.96 Finally, a large population-based study of black women from communities varying in size, segregation, and other factors found that racial segregation, as assessed by an index evaluating the likelihood that individuals would interact with others of another ethnicity or race, was negatively associated with a measure of 10-year risk for coronary heart disease in which HTN was one factor used to comprise the measure.64
To our knowledge, there is only one study specifically examining the link between incarceration and HTN.97 The investigators report that a history of incarceration was associated with HTN prevalence and new incidences of HTN because incarceration across both black and white adults drawn from a national sample of young adults drawn from the Coronary Artery Risk Development in Young Adults (CARDIA) study. The effects were strongest for black men, the group most likely to have been incarcerated, but interactions of race and incarceration were not significant.
There is substantial evidence that black Americans live in more disadvantaged communities than other groups.34 The limited available evidence suggests that neighborhood disadvantage may mediate the relationship of RRS to HTN.40 Low levels of neighborhood economic resources, including housing quality and affluence have been associated with increased prevalence of HTN,98,99 as have perceptions of social stress in the community, including crime, perceptions of safety, marital instability, and crowding.100–103 There may also be additional environmental factors influencing racial disparities in HTN, given the wide geographic disparities in rates of HTN among both blacks Americans.104 Efforts to intervene to reduce HTN will require an understanding of the specific circumstances or deprivations that are most closely associated with HTN and which mediate the effects of RRS on HTN.40
Institutional racism and HTN risk factors. Data from most,105–108 but not all109 studies suggest that living in neighborhoods with higher levels of RRS is associated with a higher prevalence of obesity. The data on the association of neighborhood affluence to obesity is clear: rates of obesity are higher in neighborhoods with low vs. high SES.37,110
To our knowledge, there is no direct evidence that RRS is independently linked to higher rates of alcohol abuse or dependence. However, there is evidence that economically disadvantaged neighborhoods and those with higher levels of neighborhood stress are associated with a higher rate of alcoholism.111–113 The available data on RRS suggest that individuals living in more segregated communities are less likely to be physically active.114 RRS has been associated with both objective indices of stress (e.g., crime),115 subjective reports of neighborhood stress,116 and fewer community resources for stress reduction (e.g., parks, recreational facilities, etc.).117
A portion of these neighborhood effects on HTN risk factors may be a function of the barriers to obtaining healthy foods and accessing recreational facilities, combined with greater access to liquor stores.110,118,119,120,121 In one experimental study in which very low income individuals from low income neighborhoods were randomly assigned to live in new, higher income neighborhoods revealed decreases in obesity (but not HTN) over a 5-year period.122 Similarly, in the Yonkers project, low income minority families who were randomly assigned to be able to move to middle class neighborhoods reported less alcohol abuse than did families unable to move.123
Summary
Black individuals remain at higher risk for the development of HTN than do white individuals, despite improvements in awareness and treatment. There is evidence that racism appears to affect risk for HTN, but the effects are complex. Among black Americans, interpersonal racism is associated with ABP,61,62,63,66,124 and in particular nocturnal BP, although it does not appear to be reliably associated with resting measures of BP or HTN diagnosis.47,50,52–54,56,57,59 There is mixed evidence linking RRS, an index of institutional racism, to HTN prevalence and BP levels,92,93 and emerging evidence that prison incarceration is associated with HTN prevalence.97 It is not clear if the effects of RRS are attributable to the degree of racial isolation or the degree of deprivation associated with the neighborhood, as neighborhood SES is inversely associated with HTN incidence.98,99 There is no evidence directly linking internalized racism to BP, but there have been very few studies.
Racism may influence the incidence of HTN by increasing the incidence of HTN-related risk factors. There is limited evidence that interpersonal racism is associated with the development of obesity.72 RRS is associated with higher levels of obesity93 and lower levels of fitness.114 Both interpersonal racism and neighborhood deprivation and stress have been linked to alcohol use, but more data are needed on the effects of internalized racism and RRS. All levels of racism are associated with perceived stress, and individual-level racism, in particular, is associated with distress21 and stress reactivity.125
Conclusions
Taken together, the evidence suggests that institutional and interpersonal racism are likely to contribute to the development of HTN, although multiple mechanisms and trajectories may be involved. Individual-level racism, and potentially internalized racism, may act in part by increasing the frequency, magnitude, duration, and psychophysiological effects of stress exposure. The harsh or impoverished environments that are a function of institutional racism may add additional stress and raise barriers to achieving a healthy lifestyle.
The relationship of perceived racism to BP emerges more clearly, when the measures inquire about episodes of interpersonal maltreatment vs. global judgments of exposure to discrimination. This may reflect problems with the reliability of global discrimination measures or the strategies for measuring BP. However, it is also possible that the findings reflect the aspects of individual-level racism (i.e., stressful interpersonal maltreatment) that are most closely associated with BP.
Exposure to race-related maltreatment has been shown to be positively related to increased rates of negative interpersonal interactions in general.73,77 If the effects of perceived racism on BP are mediated through exposure to daily interpersonal maltreatment, the effects of racism on BP may not be apparent during brief conditions involving rest (or neutral or positive interactions with medical personnel). Instead, the effects of racism may be more likely to emerge when BP is assessed during everyday events, including episodes of interpersonal conflict. This is consistent with the finding that perceived racism/discrimination is more closely related to ABP than to resting clinic BP. The importance of ongoing interpersonal conflict to BP is underscored by our recent report that the level of daily interpersonal harassment predicted masked HTN (i.e., clinic normotension plus elevated ABP) in a sample of black and Latino(a) adults.126 Further study of the effects of racism on psychobiological responses to interpersonal relationships is needed, as is research on coping strategies that might moderate or buffer these effects.
Institutional racism is associated with conditions including neighborhood poverty, segregation, and incarceration that provide limited access to health promoting resources and constraints on the development and/or deployment of health promoting coping strategies.110 This suggests that resources and coping may mediate the relationship of institutional racism to HTN. Yet, experimental data suggest that the ways in which these variables act as mediators is complex. For example, moving to a less impoverished area was associated with decreases in obesity and alcohol abuse, but was not associated with changes in hypertensive status.122,123
Research is needed to understand the specific community-level variables that affect HTN incidence. It is possible that multiple environmental factors (e.g., high levels of stress exposure plus limited access to healthy foods) must be present to trigger the onset of HTN. Alternatively, environmental variables may exert an effect on risk for HTN only in the presence of genetic vulnerability. Different levels of racism may interact to compound risk by impairing coping. For example, the negative mood states that are a persistent effect of exposure to interpersonal racism may undermine the motivation needed to overcome the effects of institutional racism, including environments with few resources for healthy living.62 Risk factors may operate differently over the course of development. The early life effects of racism, including the well-documented effects of racism on birth weight,127,128 may set the stage for increased vulnerability to the additional challenges presented by chronic interpersonal maltreatment or neighborhood poverty.
Despite progress, research on racism and HTN is still in its early stages. To more fully understand the relationship of racism to HTN, it will be important to identify the ways each level of racism acts as a stressor or as a barrier to health promotion. These continued efforts will be necessary to identify targets for prevention and intervention.
This publication was made possible by prior support to E.B. from grant R01HL68590 and ongoing support to G.O. from the following grants: P60MD003421; R01HL087301; and R01HL078566. The contents of this work are solely the responsibility of the authors and do not represent the official views of NIH.
Disclosure:
The authors declared no conflict of interest.
