Background
Perceived discrimination has been defined as an individual’s perception of being treatedunfairly by other people due to some personal attribute, such as race, ethnicity, age, gender, socioeconomic status, weight, sexual orientation, or other characteristics [
1,
2]. Unfair treatment usually involves discriminatory practices of dominant groups to maintain privileges they accrue through subordinating groups they oppress, which typically revolve around notions of innate superiority and inferiority, difference, or deviance [
3,
4].
A large and growing body of research suggests that self-reported experiences of discrimination are a form of psychological stress that has an adverse impact on both mental and physical health outcomes, across multiple population groups in a wider range of cultural and national contexts [
4‐
6]. Researchers have found associations between reports of discrimination and distress, clinically diagnosed mental disorders (e.g., generalized anxiety, posttraumatic stress disorder, depression), all-cause mortality and a variety of objective clinical disease outcomes such as preclinical endpoints - nighttime blood pressure and visceral fat - and silent indicators of premature aging - higher allostatic load and increased oxidative stress [
5,
6]. Regardless of effects on health, unjustly denying people fair treatment constrain possibilities for living fully expressed and create patterns of health inequities [
3,
4]. Thus, perceived discrimination’ studies may contribute to identify what drives population patterns of health and health inequities and to generate knowledge useful for guiding policies and actions to tackle inequities.
Research on discrimination and health focused on older adults has been scarce, comparatively with younger and middle-aged adults. In general, researchers suggest that older adults who internalize negative attitudes towards themselves are at increased risk for functional [
7] and cognitive decline [
8], life dissatisfaction [
1,
9], depression [
1], besides social withdrawal [
9,
10], reduction in cultural engagement and reluctance to visit health professionals [
10]. Among older adults who live in developing countries, literature is even scarcer. The studies usually represent major urban centers, focus primarily on racial discrimination and older adults just integrate larger samples composed by younger participants rather than constitute the target population under analysis [
11‐
13].
In Brazil, evidences from a large metropolitan region showed that around 9% of the respondents reported some type of discrimination, higher among black, women and lower among those aged 60 years and over [
11]. Additional associated factors included poorer health status and low social trust [
11]. Data from a larger survey of civil servants showed that among individuals aged 40 and older, black were more likely to report lifetime discrimination, particularly black men and for most race-by-gender groups, perceived discrimination increased with educational attainment [
13].
Brazil has rapidly transitioned from a low-income country, primarily rural in the mid-1950s, to one of the top ten economies in the world, with 84% of the population living in urban areas [
14]. The latest Brazilian census indicated that nearly 33 million people aged 50 and older lived in cities, and about 6 million were rural dwellers [
15]. The magnitude of rural Brazil is not negligible and these areas still play a fundamental role in keeping economics, social cohesion and environmental sustainability [
16,
17]. In its turn, urban dwellers may experience advantages and disadvantages [
18,
19], such as many adverse health outcomes and large health inequalities [
20]. It already became apparent that urban disadvantaged areas may have similar if not worse health outcomes than rural areas [
18,
21]. Some authors argue that discrimination would manifest strongly in areas with higher levels of social inequalities [
3]. If where people live matters, accurate measures of perceived discrimination might consider how the place of residence affects discriminatory experiences.
Using data from the Brazilian Longitudinal Study of Aging (ELSI-Brazil) baseline survey, we examined the cross-sectional association between perceived discrimination and urban/rural place of residence, among older adults. Our objective is also to identify the main individual characteristics related to higher reports of discriminatory experiences, fostering better understanding of the distribution, magnitude, and interrelationships among risk factors for exposure to discrimination. Considering discrimination may manifest strongly in areas with higher levels of social inequalities, we hypothesized that urban residence would be positively associated with perceived discrimination, independent of sociodemographic characteristics. Given the marginalized racial status of blacks in Brazil and the evidences from abovementioned studies, we also expected higher reports of discriminatory experiences within this group.
Discussion
In the first representative epidemiologic investigation of perceived discrimination among older adults in Brazil, we found that around 17% reported a discriminatory experience (any) in the previous year. Furthermore, we also found independent effects of place of residence on perceived discrimination, even after adjusting for socioeconomic status, health status and neighborhood social environment. Older adults living in urban areas reported 34% more discrimination than their rural counterparts. This finding supports our initial hypothesis that place of residence matters and urban residence is positively associated with perceived discrimination, independent of other characteristics. Moreover, our results also show that yellow and indigenous older adults were more likely to report discrimination, while blacks occupied the second worse-off position, followed by browns. It constitutes an unexpected finding, once we first hypothesized higher reports of discriminatory experiences among blacks.
Regarding perceived discrimination prevalence, our findings differs from both international and national literature. Brazilian older adults reported a lower prevalence of perceived any discrimination when compared to their English (39.3%) and Americans counterparts (over 60%) [
2,
27]. These differences might be partially attributed to variability in the way discrimination was measured, along with possible cultural differences. Both above-mentioned studies assessed discrimination more broadly, based on the frequency with which five everyday discriminatory events had happened. Even though these scales are quite comparable to ELSI’s scale, once they measure if individuals were treated disrespectfully or received poorer quality services in day-to-day life situations, they keep slight differences. In addition, some authors argue that the persistent inequality in Brazil (Gini coefficient = 0.51 in 2015) might contribute to some individuals interpret discriminatory experiences as normal situations and do not report them [
3,
28]. Social inequalities in Brazil are still remarkable and higher than United States (Gini coefficient = 0.41 in 2016) and England (Gini coefficient = 0,33 in 2015) [
28]. Considering national data, our results showed an 8% higher prevalence of perceived discrimination [
11]. However, this previous study also included young and middle adults, all from a specific metropolitan region in Brazil, which might contribute to justify the differences found.
Regarding the place of residence, rural dwellers were less likely to acknowledge discrimination. It is possible that older adults living in rural areas encounter less discriminatory experiences than their counterparts who live in cities. In this case, rural dwellers would report lower levels of perceived discrimination and urban living might be understood as an exposure factor to discriminatory experiences. The core aspects and cultural values that define the rural ways of life in Brazil might contribute to these findings. Characteristics such as smaller populations and low demographic density, marked by less degree of social differentiation, stratification and complexity, where closer and solidarity relationships among neighbors prevail [
16], might favor low levels of discrimination occurrence. An alternative explanation for the higher levels of discrimination found in urban areas is that urban older adults are more aware of discrimination and therefore more readily to report it, or are more likely to label an experience as due to discrimination. Thus, the place of residence affects the tendency to perceive and/or report discriminatory treatment. An additional aspect of interesting regarding urban-rural differences is that no differences on health status were found. Even though urban-living in disadvantaged areas have similar if not worse health outcomes than rural areas, [
18,
21] rural living in Brazil is often related to important restrictions on access to goods, services and opportunities [
16,
29]. Further studies are necessary to better investigate this finding among older Brazilian. Additionally, considering the nested nature of our data, multilevel analysis might help us to improve our understanding on how the place of residence contributes to perceived discrimination.
Adjustment for socioeconomic status, health status and neighborhood social environment in the current study slightly reduced the relation of place of residence to perceived discrimination, suggesting that a part of this relation may be mediated by these factors. On age, as it increased, participants reported lower levels of discrimination and a significant negative association was observed for older old. Our findings are consistent with other studies [
6,
11] and might be explained by selective survival. Also, it is important to consider an effect of aging, such that experiences and/or perceptions of discrimination may change across life course, suggesting the importance of considering the specific cohort and period context in which discrimination was measured [
6,
13].
Regarding racial disparities, our study demonstrated that reports of discriminatory experiences were respectively 57, 49 and 20% higher among yellow/indigenous, blacks and browns, in comparison to whites. As yellows and indigenous constitute a minority group, also in numeric terms, researchers usually exclude these individuals from analysis due to the small sample size. To our knowledge, it was the first time a representative study investigated discrimination among yellow and indigenous older adults in Brazil. The higher levels of perceived discrimination we found among indigenous and blacks are consistent with the marks that slavery left in Brazil, regarding the social position of these individuals throughout successive generations. Aspects such as greater exposure to early mortality along with social disorganization and migration to cities, living in poor conditions, certainly have contributed to perpetuate the socioeconomic marginality of these groups [
30]. Lifetime socioeconomic differences across successive generations have been identified as the main cause of racial inequality in health [
4,
30]. It has been suggested that perceived discrimination and its impact on health play a central role in the origin of these inequalities [
4,
30]. Some authors also argue that racial discrimination in Brazil is mostly notable in “hard domains”, such as job market, affective-sexual relationships, and the interactions with the police [
3]. Under this perspective, discrimination in “hard domains” would occur as an expression of dominance and oppression, viewed as a struggle for power and privilege [
4]. In other contexts, termed “soft domains”, race would tend to seem irrelevant for social relations and interpersonal contact. These domains include both leisure places and activities, such as pubs, having a talk with the neighbor, samba, carnival, and locals of spiritual or religious manifestations [
3]. It might explain our findings of racial disparities on perceived discrimination in seeking health care, whereas health care settings as “hard domains”. However, we found significant racial disparities on perceived discrimination in social gatherings – understood as a “soft domain”, and no disparities in the work place - a “hard domain”. It seems plausible that the prevalence of perceived discrimination in the work place may decrease and no disparities among racial groups be observed, as individuals get older and leave the job market. From now, it is necessary to investigate the relationship between racial disparities and specific domains of perceived discrimination among older adults along with how disparities affect health outcomes. A deeper understanding on these topics did not constitute our objectives in this study.
According to our results, the more educated participants perceived more discrimination. Those with higher educational attainment may be more aware of discriminatory treatment or more likely to identify it as such, or may engage in activities and move through social environments with a more diverse set of individuals, placing them at greater exposure (out of place) [
13]. Unexpectedly, no association was observed with household wealth, which might potentially protect individuals from exposure to situations that give rise to discrimination and provide a greater sense of control or security [
31]. Some authors identified level of education and wealth as the most significant correlates of perceived discrimination, regardless of the discriminatory situation itself [
10]. Considering reports of everyday discrimination among older adults, findings are quite heterogeneous concerning educational level. Some studies showed that the participants who reported having experienced discrimination were more likely to be more educated [
10,
31] and less wealthy [
10,
27,
31]. However, there are no significant differences regarding educational attainment for other samples [
27,
31]. Additionally, a Brazilian regional study including young, middle and older adults, showed no association between perceived any discrimination and both educational attainment and household wealth [
11]. In agreement with this previous study, we also observed that participants in poor health status and low social trust have consistently higher prevalence of reporting discrimination. Although we cannot establish causal relationships linking discrimination and poorer health status, there is considerable evidence on this association [
4,
5]. There is also consistent evidence that social support may moderate the effects of discrimination on health. However, our study design does not allow us to investigate if people have low trust due to experiences of discrimination or, whether people who have less trust are more likely to interpret some actions as discriminatory or to report experiences of discrimination more often. Regarding the longitudinal perspective of ELSI-Brazil, we will be able to explore the complex relationships between discrimination, poor health and low social trust among older adults in a near future.
It is also noteworthy that about one in ten participants reported perceived discrimination in seeking health care. This finding is quite consistent with both international and national literature [
10,
12,
31]. Discrimination may be evident in how clinical staff communicate with older patients and in the quality of care they provide. It can affect physicians’ behaviors and their decision-making, contributing to disparities in health care. For patients with major chronic conditions, such as older adults, discriminatory experiences may reduce engagement with the medical care system, affecting health promotion and disease management [
32]. Whereas that most Brazilian older adults depends exclusively of the national health system, the
Sistema Único de Saúde or SUS, our finding shows a very concerning trending of the healthcare in the country. The SUS provides comprehensive and universal care through decentralized management and provision of health services that are free of charge at the point of delivery [
14]. Universality, comprehensiveness and equity constitute the cornerstones of the system, which includes that all citizens must be treated respectfully and appropriately by health care professionals, regardless any personal attribute. Future research may provide better understanding on perceived discrimination in healthcare settings among Brazilian older adults.
Our study is not without limitations. First, the true magnitude of rural Brazil has been discussed [
16,
17]. The current methodology employed by IBGE establishes by means of municipal laws the limits of urban areas and classifies as rural those areas located outside these boundaries. If characteristics such as size, density, diversity and complexity define urban areas [
18], it is noteworthy that many Brazilian municipalities currently classified as urban do not present these core aspects [
16,
17]. A new urban-rural classification, aligned to international community, has been proposed by IBGE and will be incorporated to the 2020 Population Census. Based on three criteria - population size, demographic density, and localization in relation to the main urban centers - this new typology found 76% of the Brazilian population as predominantly urban, which corresponds to 26% of the total municipalities [
17]. Regardless this potential limitation, our study took into account the official classification employed in the country. Second, our research used a generic measure of discrimination. Unfortunately, our data did not allow participants to attribute discrimination to one or more possible causes, such as gender, race, age, weight, physical disability, or other. Despite attribute’s relevance, many studies indicate that the experience of unfairness or mistreatment may be more important for health than what the mistreatment is attribute to [
6]. Moreover, some authors argue that focuses on a single attribution for discriminatory experiences may ignore that individuals often occupy more than one socially disadvantaged status and these status may interact to shape their experiences, which is termed intersectionality. We do believe that our multivariate analysis partially offset this limitation, as we controlled for characteristics typically related to discrimination, such as age, gender, and SES. Third, this study may be subjected to bias, once place of residence might play a role on either minimization or vigilance bias. Minimization bias occurs when individuals perceive less discrimination than actually exists. Subtle ways of discrimination related to ambiguous situations could lead to minimization bias. It also might occur when individuals who belong to disadvantaged groups internalize unfair treatment as a natural or normal phenomenon [
3,
6]. In our analysis, rural older adults might be subjects to this type of bias. On the other hand, their urban counterparts might perceive more discrimination than actually exists, which is known as vigilance bias. Fourth, our study captured the perspective of individuals who reported discrimination. Another way to investigate it could consider the perspective of the stigmatizing individuals, which is termed actual discrimination (e.g. attitudes from health professionals towards groups with specific personal attributes). Together, both perspectives would show whether and to what extent perceived and actual discrimination effectively corresponds. Since we measured perceived discrimination, we cannot draw conclusions about levels of actual discrimination.
Our study also has important strengths. Using a national source of high quality information on individual aged 50 years or more, to our knowledge we carried out the first national study on perceived discrimination among older adults. Besides, it included participants commonly excluded from analyses: yellows, indigenous and rural dwellers. Due to the design of ELSI-Brazil, we will be able to provide longitudinal information on reports of discrimination in a near future.