Abstract
Despite growing interest in the social determinants of hypertension, nationally representative studies examining intra-urban effects of individual and contextual socioeconomic conditions within unequal urban areas, such as those in Latin America, remain scarce. This study describes gender disparities in the association of individual and contextual socioeconomic conditions with hypertension among older adults residing in urban areas of Brazil. We analyzed data from 6,767 participants from the baseline (2015–2016) of the Brazilian Longitudinal Study of Aging (ELSI-Brazil), a nationally representative cohort of community-dwelling adults aged ≥ 50years. Analyses were conducted as part of the ELSI-Urbe project, which integrates individual-level ELSI-Brazil data with contextual urban information. Hypertension was self-reported. The measure of socioeconomic condition was education, and the contextual measure was the Brazilian Deprivation Index (IBP; acronym in Portuguese) of the census tract of the participant’s residence. Multilevel logistic regression models (individuals and census tracts), adjusted for age and stratified by gender, were used. The prevalence of hypertension by education and the IBP differed between men and women. In women, higher education (≥9 years versus ≤4 years of schooling) was associated with a lower chance of hypertension (odds ratio [OR] = 0.62; 95% confidence interval [CI] = 0.52–0.74), and residing in census tracts with higher deprivation was associated with a higher chance of hypertension (OR per standard deviation [SD] = 1.04; 95%CI = 1.01–1.09). In men, there was no significant association between education or IBP and hypertension. Our findings suggest that public policies in middle- and low-income countries, such as Brazil, should be gender-sensitive context-specific to effectively address the burden of this disease.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11524-025-00998-5.
Keywords: Neighborhoods, Socioeconomic factors, Hypertension, Multilevel Analysis
Introduction
The prevalence of hypertension, the leading cause of cardiovascular diseases and premature death [1], is decreasing in high-income countries but increasing in low- and middle-income countries [2]. In these regions, beyond the rapid population aging, urbanization has been recognized as an important determinant of the growing burden of hypertension and its risk factors [3]. Urban living exposes individuals to higher pollution levels (air, water, light, and noise), increased stress, and reduced access to green spaces [3]. Moreover, urban environments contribute to nutritional transition, decreased physical activity, and higher consumption of tobacco and alcohol [4]. Increased exposure to these factors is not evenly distributed across urban populations and tends to be concentrated among the most socioeconomically disadvantaged groups [3].
Among urban residents, individual socioeconomic conditions are strongly associated with hypertension in high-income countries, particularly among women [5, 6]. In low- and middle-income countries, however, studies are limited and suggest the presence of significant heterogeneities in the social patterning of hypertension [7–9]. Socioeconomic characteristics at the neighborhood level — that is, the local context (e.g., census tracts) — are also linked to hypertension [9–13]. A previous study demonstrated that residing in a neighborhood with a lower socioeconomic level was associated with an increase in blood pressure over time and a higher incidence of hypertension over a 9.7-year follow-up period [11]. Despite this evidence, studies investigating neighborhood-level effects in low- and middle-income countries remain scarce and are predominantly based on local samples, often restricted to a single city [14, 15]. Moreover, nationally representative studies have examined the urban environment at broader geographic scales, such as entire cities or larger spatial units like the Brazilian Federative Units [8, 16].
Individual and neighborhood socioeconomic conditions can contribute to the development of hypertension through multiple processes. At the individual level, socioeconomic conditions can influence health behaviors and stress-related processes [6]. Neighborhood socioeconomic conditions, in turn, encompass environmental characteristics such as the availability of physical spaces for purchasing healthy foods, safe places for physical activity, transportation services, and access to healthcare services [12, 13]. Additionally, residing in a neighborhood with greater social vulnerability, characterized by poor sanitation, unhealthy housing, high crime rates, violence, and inadequate investment in human capital, can elevate chronic stress levels, thereby contributing to the development of hypertension [12]. Additionally, research indicates that older individuals [10] and women [17] are more susceptible to the characteristics of the environment where they live.
Brazil, the most populous country in Latin America and a middle-income nation, has undergone rapid urbanization and population aging in recent years. Currently, over 60% of the population lives in urban areas, and 15.6% are classified as older adults [18]. This demographic transition has been accompanied by an epidemiological transition characterized by an increasing prevalence of non-communicable chronic diseases and their risk factors, such as hypertension, which are more common in this age group [19]. This study contributes by using data from a nationally representative survey of Brazilians aged 50 years and over to assess the association between neighborhood characteristics and hypertension. It also addresses a gap in the literature by incorporating intra-urban data, enabling the identification of more vulnerable groups and supporting the planning of more effective public health interventions.
The goal of this study was to investigate the association of individual and contextual socioeconomic conditions with hypertension by gender, among older adults residing in urban areas of Brazil. We hypothesized that poor individual and contextual socioeconomic conditions would be associated with a higher prevalence of hypertension. In addition, we hypothesized that associations of individual-level education and contextual socioeconomic conditions would be stronger among women than men.
Methods
Data Source
This study used data from the ELSI-Urbe project, "The influence of the physical and social environment on the health of older Brazilian adults: a longitudinal multimethod population-based study,” coordinated by the Belo Horizonte Urban Health Observatory of the Universidade Federal de Minas Gerais (OSUBH-UFMG). The ELSI-Urbe project investigates how the physical and social characteristics of neighborhoods influence the health and well-being of older adults over time. It is part of the broader Brazilian Longitudinal Study of Aging (ELSI-Brazil), which examines the dynamics of population aging in Brazil, its determinants, and the demand for social care and health services in this population.
The ELSI-Brazil is a population-based longitudinal study, representative of the non-institutionalized Brazilians aged 50 and over residing in 70 municipalities across all five major regions of the country [20]. The sampling design of ELSI-Brazil employed utilized a multistage stratified approach, which combined stratification of primary sampling units (municipalities), census tracts, and households. Data were collected using a structured questionnaire divided into three main modules: (1) household, (2) individual, and (3) physical measurements (including anthropometry, blood pressure, and physical functioning). The first wave was conducted in 2015–2016, with participation from 9,412 individuals, and the second wave took place in 2019–2021, with 9,949 participants. Detailed information regarding the study design and recruitment methods has been published elsewhere [20]. The ELSI-Brazil was approved by the Ethics Committee of the Oswaldo Cruz Foundation – Minas Gerais (CAAE: 34,649,814.3.0000.5091); all participants or their legal guardians have signed an informed consent form.
The ELSI-Urbe developed a multilevel data structure to integrate individual-level data collected in ELSI-Brazil with contextual data obtained from open-access sources and aggregated across different geographic units of residence (regions, municipalities, neighborhoods, and census tracts). At the individual level, only participants residing in urban areas were included. At the contextual level, the smallest geographic unit used for aggregation was the census tract in which the participants lived.
Each participant was assigned a common geocode to link individual and contextual data. This geocode was used exclusively to construct the multilevel database and was deleted after processing to ensure confidentiality and compliance with data protection guidelines. Non-public geographic identifiers were stored in a secure environment with access restricted to authorized researchers.
Study Sample
For this study, the ELSI-Urbe sample was structured using a two-level multilevel design: individuals, obtained from the first wave of ELSI-Brazil, nested within their residential census tracts. Thus, the eligible population consisted of participants residing in urban areas with available census tract information (n = 7,751 individuals nested within 609 census tracts).
Our analyses were restricted to eligible individuals who completed the interview independently — without the assistance of a proxy respondent — and who had complete data on both the outcome and the exposure variables at the individual and contextual levels. Further details on the inclusion and exclusion criteria are provided in Supplementary Fig. 1.
Dependent Variable
Hypertension
Participants were defined as having hypertension if they reported that a physician had told them that they had hypertension. The information was obtained through the following question: “Has a doctor ever told you that you have hypertension (high blood pressure)?” (“no”; “yes”; “yes, only during pregnancy”; “didn’t know/didn’t answer”). Participants who answered “yes, only during pregnancy” or “didn’t know/didn’t answer” were excluded from the analysis.
Independent Variables
Individual-Level Education
Education was used as a proxy for individual socioeconomic conditions due to its wide availability for both sexes and stability over time [21]. Beyond its impact on labor market access and income, education is linked to health behaviors [21] and stands out as the strongest predictor of hypertension compared to income or occupation [5]. This variable was obtained from the question: “What was the highest grade in school that you completed?”. The question had the following answer options: 1) never studied; 2) 1 st grade (Elementary School); 3) 2nd grade (Elementary School); 4) 3rd grade (Elementary School); 5) 4th grade (Elementary School); 6) 5th grade (Elementary School); 7) 6th grade (Middle School); 8) 7th grade (Middle School); 9) 8th grade (Middle School); 10) 9th grade (High School); 11) 10th grade (High School); 12) 11/12th grade (High School); 13) school for adults; 14) some college (incomplete); 15) complete college; 16) specialization/medical residency; 17) Master’s degree; 18) Doctoral degree (Ph.D.). In Brazil, most of the older population has a low level of education [22]. For this study, education was categorized as follows: ≤ 4 years; 5 to 8 years; ≥ 9 years. These categories better capture educational inequalities within this cohort, as demonstrated in a study that used the same categorization [16].
Brazilian Deprivation Index (IBP)
The Brazilian Deprivation Index (IBP, acronym in Portuguese), an indicator developed using data from the 2010 Demographic Census [23], was used as a continuous variable to characterize the contextual socioeconomic condition at the census tract level. The IBP is based on three components: (1) the percentage of households with a per capita income ≤ 1/2 minimum wage; (2) the percentage of individuals over 7 years of age who are illiterate; and (3) the average percentage of individuals with inadequate access to basic sanitation services — including sewage, water supply, garbage collection — as well as the absence of a bathroom, bathtub, or shower [23]. A deprivation score was created by summing the standardized z-scores of these three indicators, weighted by the population of each census tract. It was designed to measure social inequalities at the smallest geographic units in Brazil. Higher IBP scores indicate greater deprivation, whereas lower scores represent less deprived areas [23].
Other Variables
Age and gender were also included in this study. Age (in years) was used as an adjustment variable due to its direct association with the prevalence of hypertension, particularly in older populations [19]. Gender, defined as sex assigned at birth and categorized as men or women, was considered a potential moderator and used as a stratification variable, based on evidence indicating biological, behavioral, and social differences in the prevalence and determinants of hypertension between men and women [17, 24, 25].
Statistical Analysis
All analyses were stratified by gender. Initially, descriptive statistics were performed using frequency distributions (%), means, and standard deviations (SD) according to the hypertension status. Posteriorly, age-adjusted hypertension prevalence estimates, and their respective 95% confidence intervals (95% CI) were calculated according to individual education categories (≤ 4, 5–8, ≥ 9 years of education) and the tertiles of the IBP of the census tract (the 1 st tertile corresponds to the stratum with the least deprivation and the 3rd tertile to the stratum with the greatest deprivation). In this analysis, the IBP was categorized into tertiles using the subsample of census tracts included in the study. Adjusted prevalence estimates were calculated using a binary logistic regression model.
Bivariate and adjusted associations between the outcome and independent variables were analyzed using two-level multilevel logistic regression (individuals and census tracts), including a random effect for the census tract [26]. First, we adjusted a model without independent variables but with a random effect at the census tract level to assess the contribution of variability between census tracts to hypertension (Model 1). Model 2 included individual education and Model 3 included individual education and the IBP. Models 2 and 3 were adjusted for age (in years). Odds ratios (OR) and their respective 95% CI were estimated.
A sensitivity analysis was conducted using measured blood pressure to assess the robustness of the findings. Hypertension was defined as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg and/or the use of antihypertensive drugs. Additionally, a sensitivity analysis was performed considering only census tracts with at least three eligible older adults, as some census tracts had a small number of older adults in the sample. All estimates accounted for sampling weights and the complex sample design. The survey command (svy) was used for descriptive analysis. Analyses were performed using the statistical software STATA version 17, with a significance level of 5%.
Results
From the total of 7,751 eligible participants for the present analysis, 6,767 completed the interview independently, without the assistance of a proxy respondent, and provided complete information for all variables of interest. Therefore, the final sample comprised 6,767 participants distributed across 605 census tracts. Details on exclusions are presented in Supplementary Figure S1. The median number of participants per census tract was 11.2 (minimum = 1; maximum = 24), with 2.9% of census tracts having up to three participants (n = 18) (Table 1).
Table 1.
Distribution of participants by Brazilian census tracts
| Total 1 | Women 1 | Men 1 | |
|---|---|---|---|
| Number of participants | 6,767 | 3,876 | 2,891 |
| Number of census tracts | 605 | 553 | 590 |
| Participants per census tract | |||
| Mean (Minimum – Maximum) | 11.2 (1—24) | 7.0 (1—16) | 4.9 (1—14) |
| ≤ 3 participants per census tract, n (%) | 18 (2.9) | 60 (10.8) | 104 (17.6) |
1 Unweighted
Table 2 displays the descriptive characteristics and bivariate analysis according to the hypertension status for the total sample and stratified by gender. The mean age of participants was 61.6 years (SD = 9.1), with 54.4% being women, and 44.3% having between 0 and 4 years of education. The mean age was 62.0 years (SD = 9.5) for women and 61.1 years (SD = 8.6) for men. Regarding individual education, 45.6% of women and 42.7% of men had up to 4 years of education. The overall prevalence of hypertension was 51.4% (95% CI = 49.3–53.6%) (data not shown in table), with a significantly higher proportion among women (57.8%; 95% CI = 54.4–61.1%) compared to men (42.2%; 95% CI = 38.9–45.6%).
Table 2.
Characteristics of study participants and bivariate analysis1 stratified by gender
| Characteristics | Total Sample | Hypertension | ||
|---|---|---|---|---|
| Yes | No | OR (95% CI) | ||
| TOTAL SAMPLE | ||||
| Individual | ||||
| Age, mean (SD) | 61.6 (9.1) | 63.3 (9.4) | 59.7 (8.4) | 1.04 (1.03–1.05)*** |
| Gender, % | ||||
| Men | 45.6 | 42.2 | 49.0 | 1.00 |
| Women | 54.4 | 57.8 | 51.0 | 1.28 (1.15–1.43)*** |
| Individual Education, % | ||||
| ≤ 4 years | 44.3 | 49.7 | 38.5 | 1.00 |
| 5 to 8 years | 23.3 | 22.6 | 24.0 | 0.74 (0.65–0.85)*** |
| ≥ 9 years | 32.4 | 27.7 | 37.5 | 0.58 (0.51–0.66)*** |
| Contextual | ||||
| Brazilian Deprivation Index, mean (SD) | - 1.21 (1.65) | - 1.16 (1.68) | - 1.25 (1.62) | 1.01 (0.98–1.04) |
| n (unweighted) | 6,767 | 3,581 | 3,186 | |
| WOMEN | ||||
| Individual | ||||
| Age, mean (SD) | 62.0 (9.5) | 63.9 (9.9) | 59.8 (8.5) | 1.05 (1.04–1.06)*** |
| Individual Education, % | ||||
| ≤ 4 years | 45.6 | 52.4 | 37.4 | 1.00 |
| 5 to 8 years | 22.5 | 22.1 | 23.1 | 0.71 (0.59–0.85)*** |
| ≥ 9 years | 31.9 | 25.5 | 39.5 | 0.48 (0.41–0.56)*** |
| Contextual | ||||
| Brazilian Deprivation Index, mean (SD) | - 1.20 (1.67) | - 1.10 (1.73) | - 1.31 (1.58) | 1.06 (1.02–1.10)** |
| n (unweighted) | 3,876 | 2,190 | 1,686 | |
| MEN | ||||
| Individual | ||||
| Age, mean (SD) | 61.1 (8.6) | 62.6 (8.5) | 59.7 (8.4) | 1.03 (1.02–1.04)*** |
| Individual Education, % | ||||
| ≤ 4 years | 42.7 | 46.1 | 39.6 | 1.00 |
| 5 to 8 years | 24.2 | 23.4 | 25.0 | 0.82 (0.67–1.01) |
| ≥ 9 years | 33.1 | 30.5 | 35.4 | 0.76 (0.62–0.92)** |
| Contextual | ||||
| Brazilian Deprivation Index, mean (SD) | - 1.21 (1.62) | - 1.23 (1.59) | - 1.19 (1.64) | 0.96 (0.92–1.01) |
| n (unweighted) | 2,891 | 1,391 | 1,500 | |
SD: standard deviation. OR: odds ratio. CI: confidence interval. Significance level: ** < 0.01, *** < 0.001
1The bivariate model between individual and contextual characteristics and arterial hypertension was carried out using multilevel analysis with a random effect at two levels: individual nested in the census tract
All estimates considered the sampling weights and the complex sample design
The bivariate analysis indicated that women had 28% higher odds of reporting hypertension compared to men. For both genders, age showed a positive association with hypertension. Among women, higher education levels were associated with lower odds of hypertension, while residing in census tracts with higher deprivation was associated with higher odds of this disease. Among men, those with 9 years or more of education exhibited lower odds of hypertension (Table 2).
Figure 1 illustrates the age-adjusted prevalence of hypertension for the total sample and stratified by gender, according to individual education categories and tertiles of the IBP of the census tract. The prevalence of hypertension showed a statistically significant dose–response gradient concerning individual education only among women, with a higher prevalence observed among those with lower education levels (≤ 4 years) (Fig. 1A). Similarly, a significant dose–response gradient in the prevalence of hypertension was observed across the tertiles of the IBP, also only among women, with a higher prevalence among those residing in census tracts with higher deprivation (3rd IBP tertile) (Fig. 1B).
Fig. 1.
Prevalence of arterial hypertension adjusted1 by age according to: a) individual education and b) Brazilian Deprivation Index (IBP) of the census tract (in tertiles). 1 Age adjustment was conducted using the age distribution of the sample.
Table 3 presents the fixed effects and the variance of the random intercept for the models stratified by gender. The higher variance observed among men suggests higher neighborhood-level heterogeneity in hypertension compared to women (see variance component in Table 3). Regarding the fixed parameters, model 3 results show that the odds of hypertension were 38% lower among women with 9 years or more of education compared to those with 4 years or fewer of education. Additionally, IBP was positively associated with hypertension among women, indicating that higher deprivation of the census tract was associated with higher odds of women reporting hypertension. Among men, although no associations reached statistical significance, a similar inverse gradient was observed for education. IBP was not associated with hypertension.
Table 3.
Adjusted associations between individual education, Brazilian Deprivation Index (IBP) of the census tract, and arterial hypertension, stratified by gender, for older Brazilian adults residing in urban areas
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| WOMEN | |||
| Individual Education | |||
| ≤ 4 years | 1.00 | 1.00 | |
| 5 to 8 years | 0.85 (0.71–1.02) | 0.88 (0.73–1.06) | |
| ≥ 9 years | 0.59 (0.50–0.70)** | 0.62 (0.52–0.74)** | |
| Brazilian Deprivation Index | – | 1.04 (1.01–1.09)* | |
| Random parameter | |||
| Variance (standard error) | 0.0249 (0.0409) | 0.0246 (0.0424) | 0.0189 (0.0417) |
| MEN | |||
| Individual Education | |||
| ≤ 4 years | 1.00 | 1.00 | |
| 5 to 8 years | 0.95 (0.77–1.18) | 0.92 (0.75–1.14) | |
| ≥ 9 years | 0.89 (0.73–1.09) | 0.84 (0.68–1.04) | |
| Brazilian Deprivation Index | – | 0.95 (0.90–1.00) | |
| Random parameter | |||
| Variance (standard error) | 0.2155 (0.0743) | 0.1465 (0.0702) | 0.1413 (0.0688) |
OR: odds ratio. CI: confidence interval. Significance level: * < 0.05, ** < 0.01
The Brazilian Deprivation Index of the census tract was standardized to a mean of 0 and a standard deviation (SD) of 1
Model 1: Null model (random effect at the census tract level)
Model 2: Education, adjusted for age (random effect at the census tract level)
Model 3: Education and Brazilian Deprivation Index, adjusted for age (random effect at the census tract level)
All estimates considered the sampling weights and the complex sample design
The sensitivity analyses using measured blood pressure to define hypertension yielded results similar to those of the main analysis for both women and men, maintaining consistency in directionality and association patterns for both exposures in the full model. The results for the random parameters were also consistent with the main analysis (Supplementary Table 1). Similarly, multilevel models that included only census tracts with at least three eligible older adults showed results comparable to those observed in the main analysis (data not shown).
Discussion
The present study examined the association between socioeconomic conditions at different levels (individual and neighborhood) and hypertension in a nationally representative sample of older Brazilian adults living in urban areas. Our findings revealed a high prevalence of hypertension, particularly among women, aligning with previous literature [27]. Additionally, we observed an inverse association between individual education and hypertension among women. Furthermore, residing in census tracts with higher levels of deprivation was associated with higher odds of hypertension in women. Among men, although no statistically significant associations were observed, a similar inverse gradient was noted for education, suggesting a potential socioeconomic pattern that may be relevant and merits further investigation.
Research conducted in high-income countries shows a higher prevalence of hypertension among individuals with lower levels of education compared to those with higher education, with this association being more consistent and stronger among women [5]. Data from low- and middle-income countries, including those from Latin America, show variability in educational gradients [8, 9, 14]. For instance, a recent study analyzing data from eight Latin American countries (Argentina, Brazil, Chile, Colombia, El Salvador, Guatemala, Mexico, and Peru) found that higher education was associated with lower odds of hypertension in women and higher odds in men [8]. Studies examining the association between education and hypertension among urban residents (aged 18 years and over) in Argentina [9] and Brazil [14] also observed an inverse association in women but no association in men. Thus, our findings regarding the association between education and hypertension align with those reported among women in high-income countries and Latin America, and they support previous Latin American studies that reported no association among men.
While the reasons for observed gender differences remain unclear, several explanations linking higher education to a decreased risk of hypertension in women warrant consideration. Education may provide greater benefits for women and compensate for potential disparities in access to other resources such as income, power, and authority [19]. Furthermore, the distribution of risk factors and access to healthcare based on socioeconomic position may differ between men and women [24]. The chronic stress stemming from work and home environments, such as informal employment, single motherhood, role overload, and experiences of violence, may also affect men and women differently [29, 30].
The present study also found that census tract deprivation, as measured by the IBP, is associated with hypertension, but the patterns vary by gender. Among women, living in census tracts with higher deprivation was significantly associated with higher odds of hypertension, whereas no significant association was observed among men. These findings are consistent with those reported in a recent study conducted in Argentina among adults aged 18 years and older [9]. There are several plausible pathways through which census tract and/or neighborhood socioeconomic deprivation may contribute to the development of hypertension. These include aspects of the neighborhood’s-built environment, such as the availability of processed food outlets, physical activity facilities, healthcare access, transportation options, and exposure to environmental pollutants [12, 13].
Differential psychosocial stress may explain the gender disparities found in our study regarding the relationship between census tract deprivation and hypertension. Individuals living in areas of higher socioeconomic deprivation report greater daily difficulties and higher exposure to chronic social stressors in terms of both frequency and severity [31]. These experiences can lead to biological wear and tear, resulting in early health deterioration, and chronic hyperactivity or hypoactivity of allostatic systems [32], contributing to the development of hypertension [33]. Therefore, being a woman and residing in census tracts with higher socioeconomic deprivation may indicate greater distress and stress, given that women are more likely than men to experience chronic stress situations [29, 30]. Additionally, sex hormones that regulate the hypothalamic–pituitary–adrenal axis response to psychological stress act differently between men and women [25].
Questionnaires containing self-reported measurements have been widely used in population-based surveys [19] and other epidemiological studies [8, 9, 14], particularly in low- and middle-income countries, due to their cost-effectiveness and ease of implementation. Previous studies comparing self-reported and measured data have demonstrated the reliability of self-reported hypertension as a valid method for population-level estimation [34]. In the present study, sensitivity analysis using objective blood pressure measurements yielded similar results to those obtained in the main analysis based on self-reported hypertension for both women and men, reinforcing the findings of the previous validation study [34].
This study presents some limitations. First, we acknowledge that the IBP, constructed using data from the 2010 Demographic Census, may not reflect changes in the distribution of deprivation over the past decade. To minimize the temporal mismatch between contextual and individual data, we used baseline data from ELSI-Brazil (2015–2016), collected shortly after the census. Second, we recognize that other risk factors for hypertension, such as physical inactivity, unhealthy dietary habits, and obesity, were not included in our analyses. However, these factors are likely mediators of the associations investigated [12, 35], and we aimed to estimate the overall effect and not the direct association between exposure and outcome. Therefore, residual confounding due to unmeasured or unknown factors cannot be entirely ruled out. Third, relates to the interpretation of variance estimates in multilevel models. In our data, we observed substantial sampling-related heterogeneity, particularly among men. As shown in Table 1, the average number of participants per census tract was lower for men (mean = 4.9) than for women (mean = 7.0). Additionally, the proportion of census tracts with three or fewer individuals was considerably higher among men (17.6%) compared to women (10.8%). These small and unbalanced cluster sizes compromise the precision of the estimated random intercept variance, which requires careful interpretation under such conditions [26].
It is essential to note that the ELSI-Brazil study was not initially designed for multilevel analysis, and its sampling framework was not based on clusters such as census tracts. Nevertheless, it enabled the first study in Brazil to simultaneously examine the association of individual and neighborhood socioeconomic conditions with hypertension from a gender perspective, using census tracts and a nationally representative sample. The multilevel analysis allows for distinguishing the effects of the environment from individual characteristics [26], contributing to more effective public policies aimed at reducing health inequalities. Additionally, using a representative sample enhances the generalizability of the findings, making the conclusions applicable to different urban contexts across the country. Furthermore, no previous study conducted in Brazil has investigated the relationship between census tract socioeconomic deprivation and hypertension with census tract socioeconomic deprivation on an urban scale from a gender perspective.
In conclusion, our study reinforces evidence that individual and contextual socioeconomic conditions are associated with hypertension and highlights variations by gender. Lower individual education and higher IBP of the census tract (indicative of higher deprivation) were linked to increased odds of hypertension among women. In contrast, such associations were not statistically significant among men. These findings underscore the importance of addressing social and gender disparities in health planning for controlling hypertension among older Brazilian adults. Therefore, implementing and intensifying public policies aimed at addressing social determinants, such as poverty alleviation and improving access to quality education among socioeconomically disadvantaged groups, is crucial. Furthermore, considering intra-urban disparities, there is a need for more targeted actions to support vulnerable groups, including enhancing access to essential services, promoting conditional cash transfers, and implementing urban interventions that foster healthy environments in both living and working contexts, particularly for women. Finally, this study highlights the value of incorporating multilevel perspectives to better understand the contextual social determinants of health. Future research would benefit from sampling designs that are specifically structured to support multilevel analysis, enabling more robust and reliable estimates of contextual effects and advancing evidence to inform equity-oriented urban health policies.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The authors would like to thank the participants and professionals who collaborated with ELSI-Brazil. They also wish to thank the Brazilian National Council for Scientific and Technological Development (CNPq) for the productivity research grant awarded to authors WTC, AALF and MFLC. DMC acknowledges the doctoral scholarship provided by the Coordination for the Improvement of Higher Education Personnel (CAPES).
Abbreviations
- ELSI-Brazil
Brazilian Longitudinal Study of Aging
- OSUBH-UFMG
Belo Horizonte Urban Health Observatory of the Universidade Federal de Minas Gerais
- IBP
Brazilian Deprivation Index
- SD
Standard Deviation
- CI
Confidence Interval
- OR
Odds Ratio
Author Contributions
DMC, ACSA, and WTC conceived the study. DMC and ACSA performed the statistical analyses. DMC drafted the first version of the manuscript. BSM, LSB and MFLC participated in or supported data collection. All authors participated in the interpretation of results and approved the final version of the manuscript.
Funding
ELSI-Brazil was funded by the Ministry of Health: DECIT/SCTIE – Department of Science and Technology of the Secretariat of Science, Technology and Strategic Inputs (Processes: 404965/2012–1 and TED 28/2017); COPID/DECIV/SAPS—Coordination of Health of the Elderly in Primary Care, Department of Life Cycles of the Secretariat of Primary Health Care (Processes: 20836, 22566, 23700, 25560, 25552 and 27510). ELSI-Urbe was funded by FAPEMIG-Foundation for Research Support of the State of Minas Gerais (project APQ-02425–21). The author DMC was also supported by the Coordination for the Improvement of Higher Education Personnel (CAPES) through a doctoral scholarship. The authors AALF and WTC are supported by the Brazilian National Council for Scientific and Technological Development (CNPq) through a research productivity scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or manuscript preparation.
Data Availability
The data that support the findings of this study are available on the ELSI-Brazil homepage after registration (https://elsi.cpqrr.fiocruz.br/en/home-english/en-data-access/).
Declarations
Ethics Approval
The ELSI-Brazil was approved by the Ethics Committee of the Oswaldo Cruz Foundation – Minas Gerais (CAAE: 34649814.3.0000.5091). All participants or legal guardians have signed an informed consent form.
Consent for Publication
Not applicable.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.GBD Risk Factor Collaborators. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):1223–49. 10.1016/S0140-6736(20)30752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: a pooled analysis of 1201 population-representative studies with 104 million participants. Lancet. 2021;398(10304):957–980. 10.1016/S0140-6736(21)01330-1 [DOI] [PMC free article] [PubMed]
- 3.Schutte AE, Venkateshmurthy NS, Mohan S, Prabhakaran D. Hypertension in low- and middle-income countries. Circ Res. 2021;128(7):808–26. 10.1161/CIRCRESAHA.120.318729. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Stringhini S, Bovet P. Socioeconomic status and risk factors for non-communicable diseases in low-income and lower-middle-income countries. Lancet Glob Health. 2017;5(3):e230–1. 10.1016/S2214-109X(17)30053-3. [DOI] [PubMed] [Google Scholar]
- 5.Leng B, Jin Y, Li G, Chen L, Jin N. Socioeconomic status and hypertension: a meta-analysis. J Hypertens. 2015;33(2):221–9. 10.1097/HJH.0000000000000428. [DOI] [PubMed] [Google Scholar]
- 6.Nakagomi A, Yasufuku Y, Ueno T, Kondo K. Social determinants of hypertension in high-income countries: a narrative literature review and future directions. Hypertens Res. 2022;45(10):1575–81. 10.1038/s41440-022-01072-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kirschbaum TK, Sudharsanan N, Manne-Goehler J, De Neve JW, Lemp JM, Theilmann M, et al. The association of socioeconomic status with hypertension in 76 low- and middle-income countries. J Am Coll Cardiol. 2022;80(8):804–17. 10.1016/j.jacc.2022.06.012. [DOI] [PubMed] [Google Scholar]
- 8.Coelho DM, Andrade ACS, Silva UM, Lazo M, Slesinski SC, Quistberg A, et al. Gender differences in the association of individual and contextual socioeconomic status with hypertension in 230 Latin American cities from the SALURBAL study: a multilevel analysis. BMC Public Health. 2023;23:1532. 10.1186/s12889-023-16400-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tumas N, López SR, Bilal U, Ortigoza F, Roux AVD. Urban social determinants of non-communicable diseases risk factors in Argentina. Health Place. 2021;77: 102611. 10.1016/j.healthplace.2021.102611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Buys DR, Howard VJ, McClure LA, Buys KC, Sawyer P, Allman RM, et al. Association between neighborhood disadvantage and hypertension prevalence, awareness, treatment, and control in older adults: results from the University of Alabama at Birmingham Study of Aging. Am J Public Health. 2015;105(6):1181–8. 10.2105/AJPH.2014.302452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Claudel SE, Adu-Brimpong J, Banks A, Ayers C, Albert MA, Das SR, et al. Association between neighborhood-level socioeconomic deprivation and incident hypertension: a longitudinal analysis of data from the Dallas Heart Study. Am Heart J. 2018;204:109–18. 10.1016/j.ahj.2018.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chaix B, Bean K, Leal C, Thomas F, Havard S, Evans D, et al. Individual/neighborhood social factors and blood pressure in the RECORD Cohort Study: which risk factors explain the associations? Hypertension. 2010;55(3):769–75. 10.1161/HYPERTENSIONAHA.109.141234. [DOI] [PubMed] [Google Scholar]
- 13.Lê-Scherban F, Ballester L, Castro JC, Cohen S, Melly S, Moore K, et al. Identifying neighborhood characteristics associated with diabetes and hypertension control in an urban African American population using geo-linked electronic health records. Prev Med Rep. 2019;15: 100953. 10.1016/j.pmedr.2019.100953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Alves RFS, Faerstein E. Educational inequalities in hypertension: complex patterns in intersections with gender and race in Brazil. Int J Equity Health. 2016;15(1):146. 10.1186/s12939-016-0440-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wagner KJP, Boing AF, Subramanian SV, Bastos JL, André C, d’Orsi E. Effects of neighborhood socioeconomic status on blood pressure in older adults. Rev Saude Publica. 2016;50:78. 10.1590/S01518-8787.2016050006373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bento IC, Mambrini JVDM, Peixoto SV. Contextual and individual factors associated with arterial hypertension among Brazilian older adults (National Health Survey–2013). Rev Bras Epidemiol. 2020;23: e200078. 10.1590/1980-549720200078. [DOI] [PubMed] [Google Scholar]
- 17.Gustafsson PE, Hammarström A, San SM. Cumulative contextual and individual disadvantages over the life course and adult functional somatic symptoms in Sweden. Eur J Public Health. 2015;25(4):592–7. 10.1093/eurpub/cku189. [DOI] [PubMed] [Google Scholar]
- 18.Instituto Brasileiro de Geografia e Estatística (IBGE). Ministério do Planejamento e Orçamento. Censo Demográfico 2022. População por idade e sexo: pessoas de 60 anos ou mais de idade. Resultados do universo. Brasil, Grandes Regiões e Unidades da Federação. Rio de Janeiro: IBGE; 2023. Accessed April 1, 2025. https://biblioteca.ibge.gov.br/visualizacao/livros/liv102038.pdf
- 19.Brasil. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância de Doenças e Agravos não Transmissíveis e Promoção da Saúde. Vigitel Brasil 2017: vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico: estimativas sobre frequência e distribuição sociodemográfica de fatores de risco e proteção para doenças crônicas nas capitais dos 26 estados brasileiros e no Distrito Federal em 2017. Brasília: Ministério da Saúde; 2019. Accessed April 1, 2025. https://bvsms.saude.gov.br/bvs/publicacoes/vigitel_brasil_2017_vigilancia_fatores_risco_1ed_rev.pdf
- 20.Lima-Costa MF, Andrade FB, Souza PRB Jr, Neri AL, Duarte YAO, Castro-Costa E, et al. The Brazilian Longitudinal Study of Aging (ELSI-BRAZIL): objectives and design. Am J Epidemiol. 2018;187(7):1345–53. 10.1093/aje/kwx387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey SG. Indicators of socioeconomic position (part 2). J Epidemiol Community Health. 2006;60(2):95–101. 10.1136/jech.2004.028092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Instituto Brasileiro de Geografia e Estatística (IBGE). Síntese de Indicadores Sociais: uma análise das condições de vida da população brasileira: 2024. Rio de Janeiro: IBGE; 2024. Accessed April 1, 2025. https://biblioteca.ibge.gov.br/visualizacao/livros/liv102144.pdf
- 23.Allik M, Ramos D, Agranonik M, Pinto Júnior EP, Ichihara MY, Barreto ML, et al. Developing a Small-Area Deprivation Measure for Brazil. Glasgow: University of Glasgow; 2020. 10.36399/gla.pubs.215898
- 24.Levorato CD, Mello LM, Silva AS, Nunes AA. Factors associated with the demand for health services from a gender-relational perspective. Cien Saude Colet. 2014;19(4):1263–74. 10.1590/1413-81232014194.01252013. [DOI] [PubMed] [Google Scholar]
- 25.Pasquali R. The hypothalamic-pituitary-adrenal axis and sex hormones in chronic stress and obesity: pathophysiological and clinical aspects. Ann N Y Acad Sci. 2012;1264:20–35. 10.1111/j.1749-6632.2012.06569.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Merlo J, Chaix B, Ohlsson H, Beckman A, Johnell K, Hjerpe P, et al. A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. J Epidemiol Community Health. 2006;60(4):290–7. 10.1136/jech.2004.029454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Malta DC, Bernal RTI, Prates EJS, Vasconcelos NM, Gomes CS, Stopa SR, et al. Self-reported arterial hypertension, use of health services and guidelines for care in Brazilian population: national Health Survey, 2019. Epidemiol Serv Saude. 2022;31(spe1): e2021369. 10.1590/SS2237-9622202200012.especial. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ross CE, Mirowsky J. Gender and the health benefits of education. Sociol Q. 2010;51(1):1–19. 10.1111/j.1533-8525.2009.01164.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kandasamy S, Anand SS. Cardiovascular disease among women from vulnerable populations: a review. Can J Cardiol. 2018;34(4):450–7. 10.1016/j.cjca.2018.01.015. [DOI] [PubMed] [Google Scholar]
- 30.Vega AP. Pobreza feminina desvendando suas raízes. TEL Temp Esp Linguagem. 2019;10(2):177–200. 10.5935/2177-6644.20190025. [Google Scholar]
- 31.Bak Bak CK, Andersen PT, Bacher I, Bancila DD. The association between socio-demographic characteristics and perceived stress among residents in a deprived neighbourhood in Denmark. Eur J Public Health. 2012;22(6):787–92. 10.1093/eurpub/ckr135. [DOI] [PubMed] [Google Scholar]
- 32.McEwen BS. Protective and damaging effects of stress mediators. N Engl J Med. 1998;338(3):171–9. 10.1056/NEJM199801153380307. [DOI] [PubMed] [Google Scholar]
- 33.Keita AD, Judd SE, Howard VJ, Cushman M, Safford MM, McClure LA, et al. Associations of neighborhood area level deprivation with the metabolic syndrome and inflammation among middle- and older-age adults. BMC Public Health. 2014;14:1319. 10.1186/1471-2458-14-1319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lima-Costa MF, Peixoto SV, Firmo JOA. Validity of self-reported hypertension and its determinants (the Bambuí study). Rev Saude Publica. 2004;38(5):637–642. Accessed April 1, 2025. http://www.scielo.br/pdf/rsp/v38n5/21750.pdf [DOI] [PubMed]
- 35.Nejatinamini S, Campbell DJ, Godley J, Minaker LM, Sajobi TT, McCormack GR. The contribution of modifiable risk factors to socioeconomic inequities in cardiovascular disease morbidity and mortality: a nationally representative population-based cohort study. Prev Med. 2023;171: 107497. 10.1016/j.ypmed.2023.107497. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available on the ELSI-Brazil homepage after registration (https://elsi.cpqrr.fiocruz.br/en/home-english/en-data-access/).