RACIAL DISCRIMINATION AND HEALTH IN :EVIDENCE FROM A POPULATION-BASED SURVEY

Objective: To investigate the association be- Ana Luiza Braz Pava˜o, MD; George Basil Ploubidis, PhD; tween racial discrimination and three health Guilherme Werneck, PhD; Moˆnica Rodrigues Campos, PhD outcomes: self-rated health (SRH), physical mor- bidity and depression, in the Brazilian population. INTRODUCTION more recently, the introduction of Design: Cross-sectional study based on data quotas to permit the admission of obtained from a representative national survey Racism, the belief that some races African descendants to some public carried out in 2008: The Research for Social Dimension of Inequalities. are superior to others, is used to devise universities. Many efforts have been and justify actions that create inequality made to minimize the effects of a Participants: 3,863 household heads who between racial groups.1 Perceived racial history of racism and racial discrimina- classified themselves as Blacks or Mullatoes discrimination is one aspect of racism tion in Brazil, however, these effects still regarding their race and answered the entire that is increasingly receiving empirical remain in Brazilian society and they research questionnaire. attention as a class of stressors that could may affect the health of people of 2 Main Outcome Measures: Racial discrimina- have negative consequences for health. African ancestry. tion was measured through a scale of 9 domains According to Krieger, there are five There is growing evidence that racial based on a previously validated instrument and main key pathways through which discrimination is an important risk classified into two categories: no discrimination racism can harm health: economic and factor for health.10 In a systematic and any experience of racial discrimination. SRH social deprivation; toxic substances and review of 138 studies on self-reported was based on the question from the SF-36 quality of life instrument. Physical morbidity and depres- hazardous conditions; socially inflicted racism and health, mainly developed in sion were obtained from a list of chronic diseases trauma (mental, physical, and sexual); the United States and Europe, the from the questionnaire. Regression analysis was targeted marketing of commodities such authors found an association between carried out for the three health outcomes as junk food and psychoactive substanc- self-reported racism and ill health in controlling for socioeconomic, demographic, es; and inadequate or degrading medical oppressed racial groups.11 Ameta- health behavior variables, and body mass index. care.3 analysis on perceived discrimination Results: Racial discrimination was negatively Brazil, among on the and health also showed a significant associated with health for all evaluated out- American continent, has the largest negative effect of racial discrimination comes. Any experience of racial discrimination population of individuals with African on both physical and mental health was associated with 1.37 more chance of outcomes and revealed that discrimina- having worse SRH, 1.55 more chance of ancestry. In a total population of 190 tion was associated with higher levels of having more physical morbidities and 1.77 million, 7.4% were classified as Black 4 more chance of having depression, even after and 42.3% as Mulattoes. Miscegena- stress response and also with unhealthy 12 controlling for confounders. tion is intrinsically connected with behaviors. In Brazil, few studies have Brazilian history and, although now evaluated racial inequalities in health, Conclusion: An impact of racial discrimination and the studies that evaluated the on the health of the Brazilian population was considered a sign of racial tolerance, found, regardless of the health indicator used, individuals with Black skin color have impact of racial discrimination on which revealed that depression was the health poorer social and health indicators than health are scarce. Still, the existing outcome with the most pronounced associa- Whites.5–9 This situation of worse social evidence suggests that Blacks and Mul- tion. (Ethn Dis. 2012;22[3]:353–359) and health status is an artifact of latoes have worse health status than Whites.8,9,13,14 Key Words: Cross-sectional Studies, Racial centuries of slavery, which lasted until According to Chor et al,8 Blacks have Discrimination, Health Status Indicators, Pop- the 19th century. After that, many ulation, Brazil advances were achieved: the end of more premature deaths and twice the slavery in 1888; the first concepts of chance of death due to physical attacks than Whites. Another study evaluated From Department of Epidemiology, racism that appeared in Brazilian law in Social Medicine Institute, State University 1951; racism viewed as a crime by the health inequalities based on sex and of (ALBP) and Department of Federal Constitution of 1988; and, ethnicity and found that Black males, Population Studies, Faculty of Epidemiology White females and Black females had a and Population Health, University of Lon- higher chance of having fair or poor don (GP) and Department of Epidemiology, health status compared to White males.13 Social Medicine Institute, State University of Address correspondence to Ana Luiza Rio de Janeiro (GW) and Department of Braz Pava˜o, MD; Av. Lu´cio Costa, nu 3360, A longitudinal study of Brazilian workers Social Sciences, National School of Public aptu 303 / 01; 22630-010; (+5521) 2496- showed that Blacks and Mulattoes who Health, Oswaldo Cruz Foundation (MRC). 9111; [email protected] experienced racial discrimination had

Ethnicity & Disease, Volume 22, Summer 2012 353 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al

50% more chance of having high blood a system of sociological indicators to questionnaire - Short-Form 36, ‘‘In pressure than those in the same racial evaluate the process of inequality and general, would you say your health is: group who did not experience that social mobility in Brazil and also to excellent, very good, good, fair or condition.14 The aim of our study was assess ’ life conditions. The poor?’’ Physical morbidity was assessed to analyze the impact of the racial study was coordinated by the University through a list of chronic diseases, from discrimination experience on the health Institute of Research of Rio de Janeiro which the interviewee had to say if any of Brazilian Blacks and Mullatoes, by (IUPERJ) and included the collabora- physician or health professional has ever estimating the association between racial tion of several others Brazilian academic told him/her that he/she had had the discrimination and three health out- and research institutions. condition (yes/no answer), based on the comes: self-rated health, physical mor- The survey population was obtained question, ‘‘For each disease I will read bidity and depression. We controlled the through stratified multi-stage random now, I would like you to say if any analysis for other health risk factors (ie, sampling and the unit of analysis was physician or other health professional socioeconomic, demographic, health be- the residence. National participants has ever told you that you have it.’’ The havior variables, and body mass index) to were classified into 6 domains according list of chronic diseases included 16 isolate the effect of racial discrimination to the geographic region of origin and morbid conditions: arthritis, cancer, on health. whether it was an urban or a rural area. cardiac disease, diabetes, hypertension, A sample stratum with the 10% richest tuberculosis, spinal disease, cirrhosis, clusters was created to improve reliabil- gout disease, tendonitis, repetitive stress METHODS ity of inequality indicators. Participants injury, chronic bronchitis, depression, were interviewed by a self-assessed chronic renal disease, asthma, hyper- This is a cross-sectional study based questionnaire and their anthropometric cholesterolemia. The original list in- on the Brazilian national survey carried measures were collected. cluded one mental health condition out in 2008 (Research for Social Data collection occurred between (depression), but we decided to exclude Dimension of Inequalities) which con- July, 2008 and December, 2008. The this information to focus on physical sisted of 8,048 households randomly questionnaire comprised several modules: chronic conditions. sampled from all five Brazilian geo- house features; household head’s and The exposure variable was experi- graphic regions (North, Northeast, spouse’s characteristics; parents’, siblings’ ence of racial discrimination. It was Mid-West, Southeast and South), com- and friends’ characteristics; economic assessed through the question, ‘‘Have prising 12,324 individuals (household activities and job; health, anthropometry, you ever felt discriminated, or unable to head and spouse) aged $20 years. The fecundity; life-related conditions; percep- do something, or has anyone ever main purpose of our study was to create tions of justice; and experience of bothered you or made you feel inferior discrimination. The modules were an- because of your color?’’ The interview- swered both by the family’s head and the ees had to answer this question in 9 spouse, except in the case of the modules domains: at school, getting a job, The aim of our study was to related to life-related conditions; percep- getting housing, at work, on the street analyze the impact of the tions of justice and experience of dis- or in a public setting, getting medical crimination, which were answered exclu- care, at a restaurant, at a bank opening racial discrimination sively by the household head. In one of an account or getting a loan, or by the experience on the health of the modules, household head’s and police or justice. This instrument of spouse’s characteristics, participants were self-reported racial discrimination, Brazilian Blacks and asked how they would classify themselves called experiences of discrimination Mullatoes, by estimating the in relation to their skin color or race. The (EOD), is based on previous work by final sample was composed of 3,863 Krieger.15 We used the revised version, association between racial household heads who had complete data which asked about the frequency of on the questions related to racial discrim- occurrence of racial discrimination.16 In discrimination and three ination and who classified themselves as this version, interviewees also had to health outcomes: self-rated Blacks or Mullatoes. answer how often this happened (never, Three different health outcomes once, 2 or 3 times, 4 times or more). health, physical morbidity were evaluated: self-rated health, phys- Then, an index variable was created and depression. ical morbidity and depression. Self- comprising the sum score of answers rated health was obtained from the to these questions (0 to 45). Based Portuguese version of the quality of life on previous research, the index was

354 Ethnicity & Disease, Volume 22, Summer 2012 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al subsequently recoded to a dichotomous Table 1. Descriptive analysis for sociodemographic characteristics of 3,863 variable: no discrimination vs any Brazilian Blacks and Mullatoes. Data from Research for Social Dimension of 10 experience of racial discrimination. Inequalities, 2008 The psychometric properties of the self-reported racial discrimination in- n % strument had been previously evaluated Sex male 2,478 65.6 and indicated that EOD could be female 1,385 34.4 Age, years 20 to 39 1,244 31.3 validly employed with working class 40 to 64 1,894 49.9 16,17 African Americans. $65 725 18.8 The analysis also included several Race Mulatto 2,986 77.2 Black 877 22.8 covariates that may affect health and a Income #0.5 MW 1,780 48.0 could be acting as potential confounders: .0.5 and #1 MW 1,068 32.3 sex, age, race, income, , smok- .1 and #1.5 MW 281 7.9 ing habits, alcohol consumption, physi- .1.5 and #2 MW 171 5.1 .2 MW 227 6.7 cal activity, and body mass index (BMI). Education none 733 21.0 Age was classified into 3 categories: 20– 1 to 4 years 1,101 31.9 39 years, 40–64 years and $65 years. 5 to 8 years 792 20.7 Race classification was divided into secondary school 746 20.0 college or more 217 6.3 Blacks or Mulattoes. Income classifica- Total 3,863 100.0 tion was based on the Brazilian monthly a Brazilian minimum wage. minimum wage (MW) of May, 2010 of 510.00 Brazilian reais (approximately US $300.00) and has five categories: ling for the confounders. For self-rated healthy, followed by 34.8% considered #0.5 MW, .0.5 and # 1 MW, .1 and health and physical morbidity, ordinal in overweight (Table 2). #1.5 MW; .1.5 and #2 MW, and .2 logistic regression was used, and, for Descriptive analysis for the health MW. Education was categorized into depression, binary regression analysis outcomes showed that 36.3% of the five strata: none, 1 to 4 years, 5 to 8 years, was applied. Results are presented as population considered their health secondary school, and college or more. adjusted odds ratios with 95% confi- good, followed by 31.7% who consid- Smoking habits were classified into three dence intervals. Data were analyzed ered their health just fair. Regarding categories: never smoked, smoked but using Stata/IC software version 11 physical morbidity, the majority report- stopped smoking, and smokes. The (Stata Corporation, College Station, ed having no diseases (40.7%), whereas categories for alcohol consumption (re- USA). 26.4% indicated the presence of one garding consumption in the last year) physical chronic morbidity. Finally, were: none, 1 time/month, 1–4 times/ depression was present in 9.2% of the month, up to 3 times/week, from 4 RESULTS interviewees. In relation to the exposure times/week to daily. For physical activity, variable, no experience of racial discrim- the classification was: sitting most of the The majority of the population was ination was referred in 77.7% of cases day, not walking very much during the male (65.6%) while the most prevalent and any level of experience of racial day, walks a lot during the day but age was 40 to 64 years (49.9%). discrimination was found in 22.4% of without carrying things, carrying light Regarding race, 77.2% classified them- household heads (Table 3). loads or doing light exercises regularly, selves as Mullatoes and 22.8% as Blacks. We observed an association between and carrying heavy loads or exercising The most prevalent income level was racial discrimination and self-rated heavily. The categories for BMI were: the lowest one (#.5 MW) (48.0%) and health after statistically controlling for low weight, healthy, overweight and the most frequent level of education was several confounders. We found that for obese. 1 to 4 years (31.9%) (Table 1). About individuals with any experience of racial half of the population said they had discrimination, the chance of having a Statistics never smoked (49.5%) or consumed worse perception about their health We developed three multivariable alcoholic beverages (54.1%), while the increased about 1.37 times (P5.002). logistic regression models to assess the majority of the interviewees said they Furthermore, being female, increased effect of racial discrimination on each usually walk during the day but without age, smoking habits, and being in health outcome: self-rated health, phys- carrying things (44.9%). Regarding overweight or obese raised the chance ical morbidity and depression, control- BMI values, 44.5% were classified as of having worse self-rated health. On

Ethnicity & Disease, Volume 22, Summer 2012 355 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al

with higher chances of having a worse Table 2. Descriptive analysis for health behavior and body mass index (BMI) of 3,863 Brazilian Blacks and Mullatoes. Data from Research for Social Dimension of physical health outcome. Individuals Inequalities, 2008 who stopped smoking had 1.43 times more chance of having more physical n % morbidities. Factors associated with a Smoking never 1,917 49.5 decrease in physical chronic morbidity stopped smoking 1,113 28.0 levels were: moderate level of alcohol smokes 833 22.5 Alcohol consumption none 2,086 54.1 consumption (1–4 times/month), and 1 time/month 733 17.6 higher levels of physical activity (walks 1–4 times/month 628 16.3 but without carrying, carrying light up to 3 times/week 226 6.0 loads, and heavy work) with a dose- from 4 times/week to daily 190 6.0 Physical activity sitting most part of the day 667 17.8 response gradient (Table 4). not walking very much 424 11.6 We observed a positive association walk but without carrying 1,774 44.9 between the experience of racial dis- carrying light loads 594 15.1 heavy work 367 10.7 crimination and depression. For any Body Mass Index low weight 132 3.4 experience of racial discrimination, the healthy 1,689 44.5 chance of having depression was in- overweight 1,343 34.8 creased 1.77 times (P5.001). In this obese 636 17.3 Total 3,863 100.0 model, factors that also showed an increased chance of having depression were: being female, income (only for the stratum: .1 and #1.5 MW), smoking the other hand, factors such as belong- ling for several confounders. We found habits (smokes), higher level of alcohol ingtohigherincomelevels,higher that any experience of racial discrimi- consumption (from 4 times/week to educational levels, moderate levels of nation was strongly associated with daily), and being obese. No differences alcohol consumption (from 1 time/ physical morbidity, compared to no were found for age. Factors that de- month to 3 times/week), and higher discrimination. The presence of some creased the chance of depression were: levels of physical activity reduced the degree of racial discrimination increased higher levels of education (for secondary chance of having a worse perception of the chance of having physical morbidity school, and college or more), and health (Table 4). 1.55 times (P,.001). Besides, as for physical activity (walks but without In the second model, we evaluated self-rated health, being female, increased carrying and carrying light loads). the association between racial discrimi- age, smoking habits, and being in Regarding race, for the three health nation and physical morbidity, control- overweight or obese were associated outcomes measures, no significant dif- ferences were found (Table 4). Table 3. Descriptive analysis for the health outcomes and the exposure variable. Data from Research for Social Dimension of Inequalities, Brazil, 2008 DISCUSSION n % Exposure variable Racial discrimination was strongly Racial discrimination no discrimination 2,994 77.7 associated with the three health out- any experience 869 22.4 Health outcomes comes analyzed in this study: self-rated Self-rated health excellent 404 10.3 very good 458 12.1 good 1,385 36.3 fair 1,251 31.7 We found that the experience poor 365 9.6 Physical morbidity healthy 1,579 40.7 of any level of racial 1 morbidity 991 26.4 2 morbidities 597 15.7 discrimination was associated 3 or more morbidities 696 17.3 Depression yes 371 9.2 with a 1.4 increase in chance no 3,492 90.4 Total 3,863 100.0 of having perceived ill health.

356 Ethnicity & Disease, Volume 22, Summer 2012 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al

Table 4. Logistic regression analysis of the association between racial discrimination and the three health outcomes: self-rated health, physical morbidity and depression. Research for Social Dimension of Inequalities (PDSD), Brazil, 2008

Self-rated health Physical morbidity Depression OR CI 95% OR CI 95% OR CI 95% Discrimination no discrimination 1.00 1.00 1.00 any experience 1.37a 1.12–1.68 1.55b 1.25–1.91 1.77a 1.25–2.50 Sex male 1.00 1.00 1.00 female 1.55b 1.29–1.87 1.99b 1.65–2.40 3.65b 2.55–5.23 Age, years 20 to 39 1.00 1.00 1.00 40 to 64 2.01b 1.64–2.45 2.80b 2.25–3.48 1.18 .78–1.77 $65 3.09b 2.30–4.16 6.18b 4.59–8.31 .84 .48–1.46 Race Mullato 1.00 1.00 1.00 Black 1.15 .95–1.39 1.10 .90–1.34 1.35 .91–2.01 Income #0.5 MWc 1.00 1.00 1.00 .0.5 and #1 MW .73a .61–.89 1.13 .93–1.37 1.19 .81–1.74 .1 and #1.5 MW .71a .52–.96 1.15 .83–1.58 1.91a 1.07–3.41 .1.5 and #2 MW .43b .28–.65 1.30 .86–1.97 1.67 .79–3.60 .2 MW .63a .43–.91 1.16 .79–1.71 1.85 .91–3.77 Education none 1.00 1.00 1.00 1 to 4 years .73a .58–.92 1.02 .81–1.27 .83 .55–1.25 5 to 8 years .54b .41–.71 1.08 .82–1.41 .84 .52–1.35 secondary school .48b .36–.65 .84 .63–1.12 .52a .29–.94 college or more .28b .19–.41 .76 .48–1.21 .26a .10–.66 Smoking never 1.00 1.00 1.00 stopped smoking 1.40a 1.14–1.71 1.43b 1.17–1.75 1.14 .78–1.69 smokes 1.45a 1.18–1.80 1.31a 1.05–1.62 1.61a 1.10–2.36 Alcohol consumption none 1.00 1.00 1.00 1 time/month .80 .64–1.01 .83 .66–1.04 .98 .62–1.54 1–4 times/month .70a .55–.89 .75a .58–.97 1.36 .83–2.22 up to 3 times/week .62a .45–.85 .83 .59–1.17 1.11 .48–2.55 from 4 times/week to daily .85 .58–1.24 .87 .58–1.31 2.94a 1.52–5.69 Physical activity sitting most part of the day 1.00 1.00 1.00 not walking very much 1.01 .73–1.40 .89 .65–1.20 .71 .42–1.18 walk but without carrying .69a .54–.87 .77a .62–.97 .58a .39–.84 carrying light loads .52b .39–.69 .58b .43–.78 .39a .21–.72 heavy work .45b .32–.63 .43b .30–.63 .64 .32–1.28 Body mass index healthy 1.00 1.00 1.00 low weight 1.10 .68–1.79 .82 .56–1.22 1.41 .63–3.15 overweight 1.22a 1.01–1.47 1.41b 1.17–1.70 1.16 .81–1.67 obese 1.79b 1.42–2.26 2.59b 2.04–3.28 1.53a 1.01–2.32

a P,.05. b P,.001. c Brazilian minimum wage. health, physical morbidity and depres- studies, two of them systematic reviews, reduced odds of excellent to good sion, even after the adjustment for have shown the association between health.10 socioeconomic, demographic, health perceived racial discrimination and poor With respect to physical morbidity, behavior variables, and BMI. Self-rated self-reported health.11,12,19 In a study racial discrimination increased the health is a subjective indicator consid- carried out with 1,722 African Amer- chance of having chronic diseases (eg, ered to be the individual perception of ican young adults, the authors also arthritis, heart disease, cancer, tubercu- health that combines both physical and evaluated the association between per- losis, diabetes, and hypertension) about emotional components, including sense ceived racial discrimination based on 1.6 times. This result is consistent with of wellbeing and satisfaction with life.18 seven domains (at school, getting a a longitudinal study carried out in a We found that the experience of any job, getting housing, at work, at home, Brazilian sample of workers, which level of racial discrimination was asso- getting medical care, on the street or found 50% more chance of hyperten- ciated with a 1.4 increase in chance of in a public setting) and self-rated sion among non-Whites who experi- having perceived ill health. This result is health, and found that self-perceived enced racial discrimination.14 The au- consistent with previous findings. Many discrimination was associated with a thors argued that the effect of racial

Ethnicity & Disease, Volume 22, Summer 2012 357 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al discrimination on the incidence of mental health status, measured using ities in health have received little hypertension was probably mediated the Mental Component Summary investigation in the Brazilian popula- through chronic stress, which directly (MCS12) of the Medical Outcomes tion.8 The objective of our study was to contributes to disease occurrence and to Study Short Form 12, among Black help evaluate the impact of racial health behavioral attitudes towards hy- and Latino immigrants in New Hamp- discrimination on health, especially in pertension (for example, alcohol con- shire, United States.23 Self-reported a developing like Brazil, where sumption, sedentariness, diet changes). racial discrimination was associated with the knowledge of this impact is still rare. Thereisalsoevidenceinliterature increased risk of mental disorders It is commonly believed that a high level showing that repeated surges of blood among Asian Americans and this rela- of tolerance concerning racial differenc- pressure, potentially from racial dis- tionship was not explained by social es is the norm in Brazil in the concept of crimination, can lead to physical dam- desirability, physical health, other stress- racial democracy.7,9 However, our find- age to blood vessels and, ultimately, ors, and sociodemographic factors.24 In ings, in accordance with other studies atherosclerosis.20 Corroborating this ev- Brazil, a study carried out among carried out in national and foreign idence, a recent study has shown that adolescents from an urban area found settings, revealed that racial discrimina- exposure to discrimination, even after the prevalence of major depression tion produces a considerable effect on controlling for age, race, socioeconomic among those reporting racial discrimi- physical, mental, and perceived health status, health status and health behavior, nation 10.4% higher than the preva- of the Brazilian population of Blacks predicted higher circulating levels of E- lence among adolescents that did not and Mullatoes, despite the adjustment selectin in men, which is a marker of report any experience of racial discrim- for several confounders. endothelial dysfunction associated with ination and this result persisted after the The most important limitation of atherosclerosis and cardiovascular dis- adjustment for age, sex, socioeconomic our study was the cross-sectional nature ease risk.21 status, skin color and self-esteem.9 of the design, which precludes the Depression was the health outcome There is biological evidence that achievement of temporality between which had the strongest association with supports the occurrence of physiological exposure and outcome (the guarantee racial discrimination in this study. The stressful reactions when individuals are that the exposure precedes the outcome chance of having depression after any exposed to situations of racism and in time) impairing inferences about 30 experience of racial discrimination in- discrimination.25 It may be that either cause and effect. Another limitation creased 77% compared to no experience direct pathopsychological effects on the is the self reported nature of our data. of racial discrimination. This is consis- brain or indirect neurophysiological Among the strengths of this study were tent with many studies. Paradies carried changes wrought through other body the nationally representative popula- out a systematic review on self-reported systems (or a combination of both) tion-based sample and the availability racism and health and found the most mediate the association between self- of many data related to the sample, consistent findings for negative mental reported racism and poor mental health which made possible the measurement health outcomes (for example, psycho- outcomes.11 Regarding this mechanism, of different health outcomes, as well as logical/psychiatric/emotional distress some investigations have revealed that potential confounding factors. The and depression/depressive symptoms).11 coping strategies and other processes of occurrence of racial discrimination and According to Borrell, mental health has psychological compensation moderate the presence of racial inequalities affect been the outcome most commonly the relationships between discrimina- the capacity for integration of Blacks examined when studying racial discrim- tion and physiological variables.26 For into Brazilian society and jeopardize the 10 proposal to build a democratic society ination effects and most of them example, studies have shown that ethnic with equal opportunities for all.7 There- report an association between perceived identity (the strength of identification fore, the results of this study highlight discrimination and worse mental health with an ethnic group, having a sense of theimportanceofcreatingpolitical regardless of the indicator used. A study ethnic pride, involvement in ethnic actions towards eradicating racism in carried out in South Africa showed that practices, and cultural commitment to order to minimize racial disparities in perceived chronic racial and non-racial one’s racial/ethnic group) is a coping health. discrimination acts independently of resource for racial/ethnic minorities, demographic factors, other stressors, buffers the stress of racial/ethnic dis- psychological factors and multiple so- crimination, and may protect mental ACKNOWLEDGMENTS cioeconomic status indicators to ad- health.27,28 22 This project was supported by Pava˜o ALB, versely affect mental health. Another The understanding of racism is PhD scholarship from National Council for study found that racial discrimination crucialtotheknowledgeofracial Scientific and Technological Development was an important predictor of poor inequalities in health.29 Racial inequal- (CNPq), grant number: 140191/20097.

358 Ethnicity & Disease, Volume 22, Summer 2012 DISCRIMINATION AND HEALTH IN BRAZIL - Pava˜o et al

REFERENCES 12. Pascoe EA, Richman SL. Perceived discrimi- 23. Gee GC, Ryan A, Laflamme DJ, Holt J. Self- 1. Bhopal R. Spectre of racism in health and nation and health: a meta-analytic review. reported discrimination and mental health health care: lessons from history and the Psychol Bull. 2009;135(4):531–554. status among African descendants, Mexican United States. BMJ. 1998;316(7149): 13. Barata RB, Almeida MF, Montero CV, Silva Americans, and other Latinos in the New 1970–1973. ZP. Health inequalities based on ethnicity in Hampshire REACH 2010 initiative: the added 2. Williams DR, Mohammed SA. Discrimina- individuals aged 15 to 64, Brazil, 1998. Cad dimension of immigration. Am J Public tion and racial disparities in health: evidence Sau´de Pu´blica. 2007;23(2):305–313. Health. 2006;96(10):1821–1828. and needed research. J Behav Med. 2009; 14. Faerstein E, Chor D, Werneck G, Lopes C, 24. Gee GC, Spencer M, Chen J, Yip T, Takeuchi 32(1):20–47. Lynch J, Kaplan G. Race and perceived racism, DT. The association between self-reported racial 3. Krieger N. Does racism harm health? Did education, and hypertension among Brazilian discrimination and 12-month DSM-IV mental child abuse exist before 1962? On explicit civil servants. Am J Epidemiol. 2004;159:35. disorders among Asian Americans nationwide. questions, critical science, and current contro- 15. Krieger N. Racial and gender discrimination: Soc Sci Med. 2007;64(10):1984–996. versies: an ecosocial perspective. Am J Public risk factors for high blood pressure? Soc Sci 25. Harrell JP, Hall S, Taliaferro J. Physiological Health. 2003;93(2):194–199. Med. 1990;30(12):1273–1281. responses to racism and discrimination: an 16. Krieger N, Smith K, Naishadham D, Hartman assessment of the evidence. Am J Public Health. 4. Brazilian Institute of Geography and Statistics. C, Barbeau E. Experiences of discrimination: 2003;93(2):243–248. Pesquisa Nacional por Amostra de Domicı´lios validity and reliability of a self-report measure 26. Kawachi I, Berkman L. Social ties and mental (PNAD), 2008. www.ibge.gov.br. November for population health research on racism and health. J Urban Health. 2001;78(3):458–467. 12, 2010. health. Soc Sci Med. 2005;61(7):1576–1596. 27. Mossakowski KN. Coping with perceived 5. Torres C. Descendientes de africanos en la 17. Bastos JL, Celeste RK, Faerstein E, Barros discrimination: does ethnic identity protect Region de las Ame´ricas y equidad en materia AJD. Racial discrimination and health: a mental health? J Health Soc Behav. 2003;44(3): de salud. Rev Panam Salud Publica. 2002; systematic review of scales with a focus on 318–331. 11(5–6):471–479. their psychometric properties. Soc Sci Med. 28. Yip T, Gee GC, Takeuchi DT. Racial 6. Travassos C, Williams DR. The concept and 2010;70(7):1091–1099. discrimination and psychological distress: the measurement of race and their relationship to 18. Szwarcwald CL, Souza-Ju´nior PRB, Esteves impact of ethnic identity and age among public health: a review focused on Brazil and MAP, Damacena GN, Viacava F. Socio- immigrants and United States-born Asian the United States. Cad Sau´de Pu´blica. 2004; demographic determinants of self-rated health adults. Dev Psychol. 2008;44(3):787–800. 20(3):660–678. in Brazil. Cad Sau´de Pu´blica. 2005; 29. Nazroo JY. The structuring of health inequal- 7. Heringer R. Desigualdades raciais no Brasil: 21(Suppl):S54–S64. ities in health: economic position, racial ´ sıntese de indicadores e desafios no campo das 19. Brondolo E, Hausmann LRM, Jhalani J, et al. discrimination and racism. Am J Public Health. ´ ´ ´ ´ polıticas publicas. Cad Saude Publica. 2002; Dimensions of perceived racism and self- 2003;93(2):277–284. 18(Suppl):57–65. reported health: examination of racial/ethnic 30. Hill AB. The environment and disease: 8. Chor D, Lima CRA. Aspectos epidemiolo´gicos differences and potential mediators. Ann Behav association or causation? Proc R Soc Med. das desigualdades raciais em sau´de no Brasil. Med. 2011;42(1):14–28. 1965;58:295–300. Cad Sau´de Pu´blica. 2005;21(5):1586–1594. 20. McEwen BS, Seeman T. Protective and 9. Santana V, Almeida-Filho N, Roberts R, damaging effects of mediators of stress. AUTHOR CONTRIBUTIONS Cooper SP. Skin colour, perception of racismo Elaborating and testing the concepts of Design and concept of study: Braz Pava˜o, and depression among adolescentes in urban allostasis and allostatic load. Ann N Y Acad Ploubidis, Werneck, Rodrigues Campos Brazil. Child Adolesc Ment Health. 2007; Sci. 1999;896:30–47. Acquisition of data: Braz Pava˜o, Rodrigues 12(3):125–131. 21. Friedman EM, Williams DR, Singer BH, Ryff Campos 10. Borrell LN, Kiefe CI, Williams DR, Diez- CD. Chronic discrimination predicts higher Data analysis and interpretation: Braz Pava˜o, Roux AV, Gordon-Larsen P. Self-reported circulating levels of E-selectin in a national Ploubidis, Werneck health, perceived racial discrimination, and sample: the MIDUS study. Brain Behav Manuscript draft: Braz Pava˜o, Ploubidis, skin color in African Americans in the Immun. 2009;23(5):684–692. Werneck CARDIA study. Soc Sci Med. 2006;63(6): 22. Williams DR, Gonzalez HM, Williams S, Statistical expertise: Braz Pava˜o, Ploubidis, 1415–1427. Mohammed SA, Moomal H, Stein DJ. Werneck, Rodrigues Campos 11. Paradies Y. A systematic review of empirical Perceived discrimination, race and health Administrative: Braz Pava˜o, Rodrigues Cam- research on self-reported racism and health. in South Africa. Soc Sci Med. 2008;67(3): pos Int J Epidemiol. 2006;35(4):888–901. 441–452. Supervision: Braz Pava˜o, Ploubidis, Werneck

Ethnicity & Disease, Volume 22, Summer 2012 359