Abstract

Introduction

Systemic arterial hypertension (SAH) or high blood pressure is a chronic noncommunicable disease, frequently asymptomatic, considered a leading risk factor for cardiovascular diseases. SAH is defined as systematic systolic/diastolic levels greater than or equal to 140/90mmHg ¹1. Whelton PK, Carey RM, Aronow WS, Casey Jr. DE, Collins KJ, Dennison Himmelfarb C, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults. Executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71:1269-324..

The occurrence of SAH has often been linked to social inequalities ²2. Leng B, Jin YN, Li G, Chen L, Jin N. Socioeconomic status and hypertension: a meta-analysis. J Hypertens 2015; 33:221-9.^,33. Mills KT, Bundy JD, Kelly TN, Reed JE, Kearney PM, Reynolds K, et al. Global disparities of hypertension prevalence and control: a systematic analysis of population-based studies from 90 countries. Circulation 2016; 134:441-50.. Although prevalence of the disease tends to decrease in many wealthy countries, it has increased in low and middle-income nations ³3. Mills KT, Bundy JD, Kelly TN, Reed JE, Kearney PM, Reynolds K, et al. Global disparities of hypertension prevalence and control: a systematic analysis of population-based studies from 90 countries. Circulation 2016; 134:441-50.. SAH affects approximately 30% of the adult population in Brazil ⁴4. Picon RV, Fuchs FD, Moreira LB, Riegel G, Fuchs SC. Trends in prevalence of hypertension in Brazil: a systematic review with meta-analysis. PLoS One 2012; 7:e48255.. In this context, SAH has been associated with various indicators of socioeconomic status, such as occupation, income and education ²2. Leng B, Jin YN, Li G, Chen L, Jin N. Socioeconomic status and hypertension: a meta-analysis. J Hypertens 2015; 33:221-9.. Among these, education has shown a strong inverse association with SAH in Brazil ⁴4. Picon RV, Fuchs FD, Moreira LB, Riegel G, Fuchs SC. Trends in prevalence of hypertension in Brazil: a systematic review with meta-analysis. PLoS One 2012; 7:e48255.^,55. Cipullo JP, Martin JF, Ciorlia LA, Godoy MR, Cacao JC, Loureiro AA, et al. Hypertension prevalence and risk factors in a Brazilian urban population. Arq Bras Cardiol 2010; 94:519-26.^,66. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.. Most of the evidence for this association between schooling and SAH in Brazil has come from cross-sectional studies ⁵5. Cipullo JP, Martin JF, Ciorlia LA, Godoy MR, Cacao JC, Loureiro AA, et al. Hypertension prevalence and risk factors in a Brazilian urban population. Arq Bras Cardiol 2010; 94:519-26.^,66. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146..

Inequalities in SAH have also appeared in relation to color, ethnicity, or “race”. The disease is more prevalent among individuals classified as black or brown when compared to whites ⁶6. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.^,77. Barber S, Diez Roux AV, Cardoso L, Santos S, Toste V, James S, et al. At the intersection of place, race, and health in Brazil: residential segregation and cardio-metabolic risk factors in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Soc Sci Med 2018; 199:67-76.. The literature has suggested that ethnic and racial inequalities in SAH reflect unfavorable social conditions related to interpersonal discrimination and structural racism rather than biological differences between whites and blacks ⁶6. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.^,88. Kaufman JS, Cooper RS. Commentary: considerations for use of racial/ethnic classification in etiologic research. Am J Epidemiol 2001; 154:291-8..

The current article thus aims to analyze the relationship between intergenerational educational mobility (IEM) and SAH in a sample of Brazilian adults. We also aim to verify the role of interpersonal and racial discrimination as effect modifiers in the relationship between IEM and SAH. The goal is to contribute to the understanding of inequalities in SAH in low and middle-income countries and thus better orient public policies aimed at promoting the health of socially excluded or marginalized populations.

Educational mobility, discrimination, and hypertension in Brazil

Life-course epidemiology is the study of long-term biological, behavioral, and psychosocial processes that link health risk in adults to physical or social exposures during intrauterine life, childhood, adolescence, and early adulthood or events transmitted across generations ⁹9. Ben-Shlomo Y, Cooper R, Kuh D. The last two decades of life course epidemiology, and its relevance for research on ageing. Int J Epidemiol 2016; 45:973-88..

The social mobility model is important here, considering that individuals’ health can be influenced by variation in their life-course social status, including upward or downward social mobility as well as persistence in the same position for indeterminate periods. Such variation can occur both within the same generation (intragenerational) and across different generations (intergenerational) ¹⁰10. Hallqvist J, Lynch J, Bartley M, Lang T, Blane D. Can we disentangle life course processes of accumulation, critical period and social mobility? An analysis of disadvantaged socio-economic positions and myocardial infarction in the Stockholm Heart Epidemiology Program. Soc Sci Med 2004; 58:1555-62..

Social inequalities persist in Brazil, as in other low and middle-income countries ¹¹11. Leubolt B. Social policies and redistribution in Brazil. Geneva: International Labour Office/Global Labour University; 2014.. Analyses of intergenerational mobility suggest a tendency for individuals to remain in certain socioeconomic strata, especially at the low and high social extremes ¹²12. Torche F. Intergenerational mobility and inequality: the Latin American case. Annu Rev Sociol 2014; 40:619-42.. Besides, even in the presence of upward mobility, the “leap” by underprivileged individuals in relation to their parents proves insufficient to guarantee material equality with the better-off portion of the population ¹³13. Mahlmeister R, Ferreira SG, Veloso F, Menezes Filho N, Komatsu BK. Revisitando a mobilidade intergeracional de educação no Brasil. São Paulo: Insper - Instituto de Ensino e Pesquisa; 2017..

Despite the inverse association between schooling and SAH ⁵5. Cipullo JP, Martin JF, Ciorlia LA, Godoy MR, Cacao JC, Loureiro AA, et al. Hypertension prevalence and risk factors in a Brazilian urban population. Arq Bras Cardiol 2010; 94:519-26.^,66. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146., little is known about the relationship between this health outcome and IEM in Brazil. Only one Brazilian study was found, but it was limited to an analysis of intragenerational social mobility and SAH, and no statistically significant association was observed ¹⁴14. Guimarães JMN, Griep RH, Clarke PJ, Fonseca MJM, Barreto SM, Giatti L, et al. Intra-generational social mobility and changes in blood pressure: longitudinal analysis from the ELSA-Brasil Study. Am J Hypertens 2018; 31:672-8..

The pattern of educational levels across generations in Brazil also exhibits inequalities according to color, ethnicity, or “race”. The likelihood of black and brown Brazilians remaining at low levels of education is greater than for whites ¹³13. Mahlmeister R, Ferreira SG, Veloso F, Menezes Filho N, Komatsu BK. Revisitando a mobilidade intergeracional de educação no Brasil. São Paulo: Insper - Instituto de Ensino e Pesquisa; 2017.^,1515. Marteleto LJ. Educational inequality by race in Brazil, 1982-2007: structural changes and shifts in racial classification. Demography 2012; 49:337-58.. According to various authors, this “racial” inequality in education reflects interpersonal discrimination and structural racism, part of the country’s historical trajectory based on slavery and the power relations established during Brazil’s colonial period ¹⁶16. Hasenbalg C. Discriminação e desigualdades raciais no Brasil. 2ª Ed. Rio de Janeiro: Editora UFMG; 2005..

Considering that Brazil is a country with intense and persistent social inequalities and with a significant black and brown population, where SAH is a public health problem, more studies on the link between discrimination and the socioeconomic disadvantages that affect the population’s blood pressure are extremely important.

Methods

Participants

We analyzed data from 1,720 participants in the EpiFloripa Adult Study (ages 20-59 years at baseline). The study began in 2009, with the first follow-up wave in 2012 (n = 1,222) and the third in 2014 (n = 862). In 2009 and 2012, the interviews were held at the participants’ homes, in the urban area of Florianópolis, capital of the state of Santa Catarina in southern Brazil. Sampling was done in two-stage clusters, with the first involving selection of census tracts according to head-of-household’s income deciles, constituting the primary analytical units. The second stage was the selection of households for the home visits. Pregnant women in the second or third trimester and women with infants six months of age or younger did not have their blood pressure measured or anthropometric measures taken. In the last wave, participants went to a center organized for the data collection and measurements. In all three waves, blood pressure was measured two to three times, using a digital pulse sphygmomanometer. The questionnaire applied in 2009 included questions on the participants’ schooling, prior diagnosis of SAH, and use of antihypertensive medication. In the 2012 wave, data were collected on parents’ education, besides information on experiences with discrimination at the individual level. In 2014, data were collected again on the participants’ schooling.

The Institutional Review Board of the Federal University of Santa Catarina approved the study under protocols 351/08 of December 15, 2008, and 1772/11 of February 28, 2011. Participants signed a free and informed consent form after receiving an explanation on the study’s objectives. Details on the baseline methodological procedures of the EpiFloripa Adult Study have been described in a previous publication ¹⁷17. Boing AC, Peres KG, Boing AF, Hallal PC, Silva NN, Peres MA. EpiFloripa Health Survey: the methodological and operational aspects behind the scenes. Rev Bras Epidemiol 2014; 17:147-62..

Dependent variable

Individuals were classified as hypertense with systolic/diastolic blood pressure greater than or equal to 140/90mmHg ¹1. Whelton PK, Carey RM, Aronow WS, Casey Jr. DE, Collins KJ, Dennison Himmelfarb C, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults. Executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71:1269-324. and/or when they reported a medical diagnosis of SAH and use of antihypertensive medication. Considering the participants’ follow-up over time at three different data collection moments, the dependent variable was defined as the presence of hypertension (yes/no) on at least one of the three occasions. The specific analysis of incident cases of hypertension is the object of another study under way and will be published in due course.

Exposures

Years of schooling were used as an indicator of socioeconomic status both for the participants (current socioeconomic status) and for their parents (parents’ socioeconomic status during the participants’ childhood, hereinafter “childhood socioeconomic status”). Education is a widely used indicator for socioeconomic status in epidemiological studies, since it is generally available for both sexes; it is also a quantifiable measure and especially is more stable over time when compared to occupation and income ¹⁸18. Krieger N, Williams DR, Moss NE. Measuring social class in US public health research: concepts, methodologies, and guidelines. Annu Rev Public Health 1997; 18:341-78..

IEM was calculated as the variation between current socioeconomic status and childhood socioeconomic status. Three types of IEM were calculated: paternal education, maternal education, and family education. Family IEM was created on the basis of greater mobility between paternal and maternal IEM. Thus, when paternal mobility was low and maternal mobility was downward, the resulting family IEM variable consisted of maternal mobility. When paternal mobility was upward and maternal mobility was downward, the family IEM reflected paternal mobility, and so on.

The categories for the IEM variable were thus as follows: low (for participants with low schooling and whose parents likewise had low schooling), high (for individuals and parents with high schooling), upward (for participants with high schooling, as opposed to their parents), and downward (for individuals in this study with less schooling than their own parents). Categorization of socioeconomic status variables was based on the division of educational levels in Brazil: primary schooling (0-9 years), secondary (10-12 years), and university (> 12 years) ¹⁹19. Saiba como é a divisão do sistema de educação brasileiro. Agência Brasil 2014. http://www.brasil.gov.br/educacao/2014/05/saiba-como-e-a-divisao-do-sistema-de-educaçãoi-brasileiro/view (acessado em 30/Jul/2014).
http://www.brasil.gov.br/educacao/2014/0... . This division was adopted in Brazil in 1996. Previously, secondary school ended with 11 years of schooling. In order to reduce classification errors, individuals under 30 years of age, i.e., who were 16 years or younger in 1996, were classified according to the new division, while those 30 years or older, or 17 years or older in 1996, were classified according to the previous division.

Some authors contend that social processes involved in the transition from primary to secondary school have less impact than the transition from secondary school to university, when individuals receive a greater load of responsibilities, often starting life away from their parents and dealing with new situations and health risks ²⁰20. Fromme K, Corbin WR, Kruse MI. Behavioral risks during the transition from high school to college. Dev Psychol 2008; 44:1497-504.. Primary and secondary schooling were thus condensed into one socioeconomic status category, “low”. University education was defined as the “high” category of socioeconomic status.

As for childhood socioeconomic status, the study took into account the improvement in education in Brazil in recent decades ²¹21. Organisation for Economic Co-operation and Development. Education at a glance 2011: OECD Indicators. Paris: OECD Publishing; 2011.. Thus, childhood socioeconomic status was categorized according to the parents’ median years of schooling in relation to the study participant’s age. Parents with schooling below the median for the participants’ age were classified as having low education. Parents with schooling above the median for the participants’ age were classified as having high education.

Confounding and interaction variables

The confounding variables sex, age (20-29, 30-39, 40-49, ≥ 50 years), and total household income (quintiles) were tested in the procedure for definition of the final models, since the literature suggests a frequent association between them and the target variables ²2. Leng B, Jin YN, Li G, Chen L, Jin N. Socioeconomic status and hypertension: a meta-analysis. J Hypertens 2015; 33:221-9.^,2222. Janicki-Deverts D, Cohen S, Matthews KA, Jacobs DR Jr. Sex differences in the association of childhood socioeconomic status with adult blood pressure change: the CARDIA study. Psychosom Med 2012; 74:728-35.. The analyses were also adjusted for the effect of the time transpired between the waves. Interpersonal discrimination was measured with the Everyday Discrimination Scale (EDS), which allows analyzing unfair treatments over the course of life in different domains, as well as the reasons for discrimination and the study participant’s classification of a given event as discriminatory or non-discriminatory. The EDS was drafted to assess the effects of discrimination on health outcomes and behaviors ²³23. Bastos JL, Faerstein E, Celeste RK, Barros AJ. Explicit discrimination and health: development and psychometric properties of an assessment instrument. Rev Saúde Pública 2012; 46:269-78.. Skin color/“race” was assigned according to visual inspection by the interviewer, because as some authors suggest ²⁴24. La Veist TA. Beyond dummy variables and sample selection: what health services researchers ought to know about race as a variable. Health Serv Res 1994; 29:1-16., this reflects how participants are “perceived” by others. This process can also be called physiognomy and refers to a judgment based on appearance, which can determine the degree of exposure to health risks ²⁴24. La Veist TA. Beyond dummy variables and sample selection: what health services researchers ought to know about race as a variable. Health Serv Res 1994; 29:1-16..

Statistical analyses

Multilevel analyses were performed with Stata 14 software (https://www.stata.com), using generalized linear latent and mixed models (GLLAMM) to obtain the random intercept and coefficient with dichotomous outcome (with/without SAH). The final multilevel models were defined by the forward strategy (p < 0.05), inserting each of the confounding variables (sex, age, income, and time between study waves) and interaction variables (interpersonal discrimination and skin color/race) sequentially, comparing the nested models using the maximum likelihood test. Analysis of the GLLAMM with and without a random gradient was also performed. Interaction between discrimination and IEM was tested in stratified analysis by skin color/race, creating a new variable resulting from the multiplication of interpersonal discrimination (dichotomous) and IEM (polytomous). These interaction models considered the role of discrimination on racial inequalities in education ¹³13. Mahlmeister R, Ferreira SG, Veloso F, Menezes Filho N, Komatsu BK. Revisitando a mobilidade intergeracional de educação no Brasil. São Paulo: Insper - Instituto de Ensino e Pesquisa; 2017. and their connection to blood pressure in adults ⁶6. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.^,77. Barber S, Diez Roux AV, Cardoso L, Santos S, Toste V, James S, et al. At the intersection of place, race, and health in Brazil: residential segregation and cardio-metabolic risk factors in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Soc Sci Med 2018; 199:67-76.. The test hypothesis took into account that the effect of IEM could be intensified in situations of interpersonal discrimination, especially for black and brown Brazilians. All the statistical analysis took the study’s sampling structure and weights into account.

Results

The study sample was mostly female, under 40 years of age, in the fourth quintile of total household income, and with low current and childhood socieconomic status. The losses were due mainly to participants having moved to other cities, not being found at home, or having refused to remain in the study. Table 1 shows the participants’ distribution according to baseline sociodemographic characteristics (2009), including the report of lifetime experiences with discrimination (2012) and prevalence of SAH. Higher prevalence of SAH was seen in men, individuals over 50 years of age, those classified as black or brown, in the third income quintile, with low childhood and current SES, and with downward and low IEM. Prevalence of SAH was similar between individuals reporting discrimination and those not reporting it.

Comparisons of models with and without random coefficients did not provide additional information with the gradient’s inclusion, so we opted for models with the random intercept only. Three final models were selected, hierarchically testing the life-course effect of IEM on SAH. The multilevel analysis with random effects showed that the influence of IEM was maintained even after adjusting for confounding, as shown in Table 2. Constantly low socioeconomic status across generations (low IEM) was considered the reference category. Downward IEM increased the odds of SAH. This pattern was observed in models 1, 2, and model 3. In the opposite direction, upward IEM reduced the odds of SAH, both when paternal mobility was tested (model 1), as well as maternal mobility (model 2) and family mobility (model 3). Persistently high socioeconomic status across generations (high IEM) significantly reduced the odds of SAH in all the models. In Table 2, considering the influence of random effects, not measured by the study variables, intraclass correlations (ICCs) for the random intercepts in the three models (0.91) indicated high propensity for the participants to remain in their initial socioeconomic status throughout the study, due to individual characteristics. Age, sex, and follow-up time all showed high influence on the odds of SAH.

In the analysis of the interaction between discrimination and IEM, the group that did not report discrimination was considered the reference category. As shown in Table 3, the inverse relationship between IEM and presence of SAH was maintained in individuals that reported discrimination when compared to those with no such report. In practically all of the interaction models tested, the odds of SAH increased in low and downward IEM among individuals that reported discrimination compared to those not reporting it, except in the paternal model’s downward category. In the three models, individuals reporting discrimination showed lower odds of SAH in upward and high IEM, and the results in the high category were statistically significant.

However, analyzing the interaction between IEM and discrimination graphically, the above-mentioned inverse tendency practically disappeared. There was a reduction in the odds of SAH in nearly all the IEM categories when discrimination was reported, and this was repeated in all three models tested (Figures 1a, 1b, and 1c). The only exceptions were in the downward category in the maternal model (Figure 1a) and family model (Figure 1c).

Stratifying the analyses by skin color/ “race”, no effect modification was seen in the maternal model when comparing black or brown individuals to whites (Figure 2). However, in the paternal model (Figure 3) and family model (Figure 4), the odds of SAH differed between black/brown individuals and whites in low and downward IEM. Meanwhile, in the paternal model, the odds of SAH increased among black or brown individuals who reported discrimination in low and downward IEM, while the odds decreased in white individuals in the same situation. In the family model, the odds of SAH did not vary in the low category and increased in the downward category among black or brown individuals reporting discrimination, compared to those not reporting it, but decreased in whites in the same situation in the low and downward categories. In upward and high IEM, the odds of SAH were lower among those who reported discrimination in all the models and between black or brown individuals and whites.

Discussion

The current article analyzed the influence of IEM on the occurrence of SAH in adults in a city in southern Brazil. The results of the multilevel longitudinal analysis with random effects showed that persistence in high socioeconomic status (high IEM) significantly reduced the odds of SAH in all the models. Interaction between interpersonal discrimination and IEM was also tested in an analysis stratified by skin color/“race”, showing that the odds of SAH increased in the downward category of IEM among black or brown individuals who reported discrimination, compared to those not reporting it.

The high prevalence of SAH in the current study (44.9%) is close to the prevalence observed by other authors in Latin America and the Caribbean (39.1%, 95%CI: 33.1; 45.2) ²⁵25. Sarki AM, Nduka CU, Stranges S, Kandala NB, Uthman OA. Prevalence of hypertension in low- and middle-income countries: a systematic review and meta-analysis. Medicine (Baltimore) 2015; 94:e1959.. However, it exceeds the prevalence rates reported in other Brazilian studies, such as the National Health Survey of 2013 (22.9% in the South of Brazil) ²⁶26. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional de Saúde 2013. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2014. and the Vigitel Brazil 2017 Supplementary Health survey (21,5% in Florianópolis, Santa Catarina State) ²⁷27. Agência Nacional de Saúde Suplementar; Ministério da Saúde. Vigitel Brasil 2017 Saúde Suplementar: vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico [recurso eletrônico]. Brasília: Ministério da Saúde; 2018.. The difference is that the data from the two latter studies only refer to individuals who reported a medical diagnosis of SAH, while the current study included the use of antihypertensive medication (6% prevalence at baseline) and the mean of two measurements of systolic/diastolic blood pressure, which may have contributed to the higher observed prevalence.

In this study, hypertension was more prevalent in black and brown individuals (51.3%) than in whites (44.1%), corroborating previous studies ⁶6. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.. In racially unequal countries like Brazil, color/ “race” is an indicator of social class and structural racism, besides acting as a target for interpersonal discrimination, affecting health unfavorably ²⁸28. Paradies Y, Ben J, Denson N, Elias A, Priest N, Pieterse A, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One 2015; 10:e0138511.. A study in Cuba, a country with less racial inequality than Brazil, found no difference in SAH prevalence between blacks and whites ²⁹29. Ordunez P, Kaufman JS, Benet M, Morejon A, Silva LC, Shoham DA, et al. Blacks and Whites in the Cuba have equal prevalence of hypertension: confirmation from a new population survey. BMC Public Health 2013; 13:169..

No relevant differences were observed in the prevalence of SAH between total household income quintiles, between childhood educational levels, or between individuals with or without experiences of discrimination. Although the prevalence of SAH was similar in these groups, studies have been performed to demonstrate that life-course experience with discrimination ⁶6. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.^,2828. Paradies Y, Ben J, Denson N, Elias A, Priest N, Pieterse A, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One 2015; 10:e0138511., material deprivation since childhood ³⁰30. Bowen ME. Coronary heart disease from a life-course approach: findings from the health and retirement study, 1998-2004. J Aging Health 2010; 22:219-41., and parents’ low schooling ³¹31. Rodríguez López S, Bensenor IM, Giatti L, Molina MDC, Lotufo PA. Association between maternal education and blood pressure: mediation evidence through height components in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Ann Hum Biol 2017; 44:243-25 can influence blood pressure and health in adulthood. In the study sample, SAH was more prevalent in individuals with low schooling, corroborating observations from previous studies ²2. Leng B, Jin YN, Li G, Chen L, Jin N. Socioeconomic status and hypertension: a meta-analysis. J Hypertens 2015; 33:221-9.^,55. Cipullo JP, Martin JF, Ciorlia LA, Godoy MR, Cacao JC, Loureiro AA, et al. Hypertension prevalence and risk factors in a Brazilian urban population. Arq Bras Cardiol 2010; 94:519-26.^,66. Alves RF, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health 2016; 15:146.. Individuals with downward IEM also displayed high prevalence of SAH, followed by those who maintained low educational level throughout life. A longitudinal study of twins in Sweden ³²32. Högberg L, Cnattingius S, Lundholm C, Sparen P, Iliadou AN. Intergenerational social mobility and the risk of hypertension. J Epidemiol Community Health 2012; 66:e9. and a cross-sectional study in the USA with black men ³³33. James SA, Van Hoewyk J, Belli RF, Strogatz DS, Williams DR, Raghunathan TE. Life-course socioeconomic position and hypertension in African American men: the Pitt County Study. Am J Public Health 2006; 96:812-7. found that prevalence of SAH was higher in individuals that remained in low socioeconomic status throughout life.

The fixed effects of the multilevel longitudinal analysis showed that only high IEM was capable of significantly reducing the odds of SAH when compared to individuals with low IEM. This raises concern in Brazil, considering the persistence of schooling levels at the social extremes, hindering changes in participants’ socioeconomic status in relation to their parents when childhood socioeconomic status is low ¹²12. Torche F. Intergenerational mobility and inequality: the Latin American case. Annu Rev Sociol 2014; 40:619-42.^,1313. Mahlmeister R, Ferreira SG, Veloso F, Menezes Filho N, Komatsu BK. Revisitando a mobilidade intergeracional de educação no Brasil. São Paulo: Insper - Instituto de Ensino e Pesquisa; 2017.. In addition, even upward intergenerational schooling in Brazil has not produced the same favorable effect of maintaining high socioeconomic status throughout life. Corroborating the current study’s results, despite the methodological and contextual differences, a cross-sectional study in African American adults found an association between constantly high lifetime socioeconomic status and SAH (OR = 7.27, 95%CI: 1.91; 27.51) in comparison to constantly low socioeconomic status ³³33. James SA, Van Hoewyk J, Belli RF, Strogatz DS, Williams DR, Raghunathan TE. Life-course socioeconomic position and hypertension in African American men: the Pitt County Study. Am J Public Health 2006; 96:812-7.. Some authors suggest that schooling can affect health through a system of cumulative disadvantages involving behaviors with an impact on health and prevention and treatment of diseases, as well as through stressful workplaces and the residential neighborhood’s socioeconomic context ³⁴34. Cutler DM, Lleras-Muney A. Understanding differences in health behaviors by education. J Health Econ 2010; 29:1-28.^,3535. Zimmerman E, Woolf SH. Understanding the relationship between education and health. Washington DC: Institute of Medicine; 2014. (Discussion Paper).^,3636. Kaiser P, Diez Roux AV, Mujahid M, Carnethon M, Bertoni A, Adar SD, et al. Neighborhood environments and incident hypertension in the multi-ethnic study of atherosclerosis. Am J Epidemiol 2016; 183:988-97..

In addition, according to the random effects, the ICC of the random intercept (0.91 in the three models) showed that a large share of the outcome’s variance in the paternal model (model 1), maternal model (model 2), and family model (model 3) can be explained by the participants’ individual characteristics, which are unchangeable throughout the study and are not measured in the tested models. For example, some authors suggest that maternal breastfeeding and birthweight can produce latent effects that have the power to influence blood pressure in adulthood ²²22. Janicki-Deverts D, Cohen S, Matthews KA, Jacobs DR Jr. Sex differences in the association of childhood socioeconomic status with adult blood pressure change: the CARDIA study. Psychosom Med 2012; 74:728-35.^,3131. Rodríguez López S, Bensenor IM, Giatti L, Molina MDC, Lotufo PA. Association between maternal education and blood pressure: mediation evidence through height components in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Ann Hum Biol 2017; 44:243-25. Besides, situations of psychosocial stress such as material deprivation in childhood, adolescence, and youth can act on blood pressure in adulthood ³⁰30. Bowen ME. Coronary heart disease from a life-course approach: findings from the health and retirement study, 1998-2004. J Aging Health 2010; 22:219-41.^,3737. Poulton R, Caspi A, Milne BJ, Thomson WM, Taylor A, Sears MR, et al. Association between children's experience of socioeconomic disadvantage and adult health: a life-course study. Lancet. 2002; 360:1640-5..

The analyses of interaction between interpersonal discrimination and IEM showed a reduction in the odds of SAH in practically all the categories of IEM with reported discrimination and in all the models. This effect can result from internalized oppression ³⁸38. Krieger N. Embodiment: a conceptual glossary for epidemiology. J Epidemiol Community Health 2005; 59:350-5., a perspective that affirms that individuals do not always report what they experience, even when somatic manifestations demonstrate the opposite (embodiment). Thus, individuals that did not report discrimination because they did not see themselves as victims of such acts, even though they had been, may have constituted the “non-discriminated” reference category, leading to this contradictory result.

The inconsistency of the results found in the literature relating interpersonal discrimination to SAH may also have been due to non-reporting or non-recognition of certain forms of treatment as discriminatory. The most consistent evidence relates discrimination to outpatient blood pressure, which can be considered an indicator of an immediate reaction to stress, not dependent on the individuals’ understanding and reporting of a discriminatory act ²⁸28. Paradies Y, Ben J, Denson N, Elias A, Priest N, Pieterse A, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One 2015; 10:e0138511.. Importantly, the reduction in the odds of SAH among individuals that reported discrimination does not mean a beneficial effect of discrimination on blood pressure. Neither does it legitimize social injustice or the acceptance of social inequalities in SAH and in health in a broader sense. However, it highlights the need to understand how discrimination is perceived and reported and how it impacts individuals’ health. Besides, the search for solutions to fight discrimination does not depend on the reason that triggers such discrimination and overrides the results of scientific studies.

We also observed differences in the pattern of odds of SAH in the paternal and maternal models among black or brown individuals with low IEM, and among whites in the maternal model with downward IEM, indicating that mobility measured on the basis of the father can have a different effect from mobility measured on the basis of the mother. Brazil has an important share of single-parent families headed by women (16.3%) ³⁹39. Pinheiro L, Fontoura NO, Querino AC, Almeida CR, Firmino CR, Silva FHE, et al. Retrato das desigualdades de gênero e raça. Brasília: Instituto de Pesquisa Econômica Aplicada; 2015., suggesting a relevant maternal influence on health behaviors and beliefs in adulthood. However, more studies are needed that analyze the problem from this perspective.

Conclusion

The current article’s results reinforce the importance of awareness-raising on SAH in adults in southern Brazil from a life-course perspective. The study found that across generations, maintaining the parents’ high level of schooling favored the reduction of SAH, although random variables also influenced the increase in blood pressure. Upward IEM did not significantly reduce the effects of unfavorable childhood socioeconomic status. This reiterates the need for long-term public policies aimed at continuing education in the country as well as improvement in the population’s health.

The results of the stratified analysis by skin color/“race” suggest that situations of discrimination can act jointly with unfavorable IEM, increasing the odds of SAH, especially among black and brown individuals. This result provides even more backing for the argument that hypertension and its social inequalities should be confronted with public policies prioritizing education, which have not only a direct effect on the problem, but also an indirect effect, reducing the manifestations of discrimination.

Acknowledgements

The authors wish to thank the study participants, the Brazilian Graduate Studies Coordinating Board (CAPES, Funding Code 001), and the Brazilian National Research Council (CNPq), Ministry of Science, Technology, Innovations, and Communications.

References

Publication Dates

History