Beyond biology: Social and geographic determinants of hypertension in rural Alabama communities

TL;DR

This study explores how race, gender, age, and geography affect blood pressure in rural Alabama, revealing patterns that could help reduce health disparities.

Contribution

The study identifies distinct demographic and geographic patterns in blood pressure that highlight the role of social determinants in cardiovascular disparities.

Findings

Age strongly predicts systolic blood pressure, with steeper increases among Black men compared to Black women and White men.

Diastolic blood pressure is higher among Black adults and varies by geographic location, indicating local social determinants.

Geographic variation in blood pressure suggests the need for place-based interventions to address cardiovascular disparities.

Abstract

Hypertension remains a leading modifiable risk factor for cardiovascular morbidity and mortality in the United States. Despite advances in detection and treatment, disparities in hypertension prevalence, control, and outcomes persist across racial, gender, and geographic lines, particularly in the rural Deep South. To examine the associations between race, gender, age, and geography on systolic and diastolic blood pressure among adult patients in West Alabama, and to identify patterns that may inform equitable, place-based interventions. De-identified health records from 3,462 adult patients across nine West Alabama counties were analyzed. Systolic and diastolic blood pressure were modeled as dependent variables using multivariate analyses including race, gender, age, and zip code. Interaction terms were examined to assess moderating effects between demographic and geographic variables. Systolic blood pressure (SBP) was most strongly predicted by age (p < 0.0001), with steeper increases among men. The steeper increase for men with age was driven by Black men – Black men showed a steeper SBP increase with age than Black women while White women showed a steeper increase than White men. Diastolic blood pressure (DBP) was independently associated with race, gender, and zip code (p < 0.05), with higher DBP among Black adults across geographic areas. Geographic variation persisted for both SBP and DBP, suggesting the influence of local social determinants of health. Distinct demographic and geographic patterns in blood pressure highlight the intersection of biological and structural factors driving cardiovascular disparities in rural Alabama. Addressing these inequities will require multilevel strategies integrating clinical care, community resources, and place-based policy interventions.

Figures

Linear Prediction of Systolic Blood Pressure Across Age by Gender.

Linear Prediction of Systolic Blood Pressure Across Age by Gender for Race = Black.

Linear Prediction of Systolic Blood Pressure Across Age by Gender for Race = White.

Full text

Introduction

Hypertension remains one of the leading modifiable risk factors for cardiovascular morbidity and mortality in the United States. Despite overall advances in detection and management, disparities in the prevalence of hypertension, control, and outcomes remain across racial, gender, and geographic lines [1,2]. African American adults in particular experience earlier onset, higher prevalence, and lower rates of blood pressure control compared to White adults, contributing to disproportionately high rates of stroke, kidney disease, and cardiovascular mortality [3,4]. Additionally, rural Americans have higher odds of having hypertension than urban dwellers [5,6]. These inequities persist even when adjusting for income, education, and access to care [7], underscoring the potential role of structural racism in shaping cardiovascular outcomes [8,9].

Gender also plays an important role in the epidemiology of hypertension [10,11]. Men are more likely to develop elevated blood pressure earlier in life, whereas women often exhibit steeper increases in both systolic and diastolic pressures after midlife, coinciding with menopause and changes in vascular physiology [10]. These trajectories result in a narrowing of the gender gap in later life [2], with older women facing particularly high risks of hypertension-related cardiovascular complications. Gender disparities are further shaped by social roles, caregiving responsibilities, and health care access, which influence adherence to treatment and engagement in preventive care [12,13].

Geography and community context further modify hypertension risk and outcomes. Rural populations experience higher rates of hypertension and cardiovascular disease, driven by social determinants of health (SDoH) including provider shortages, limited access to specialty care, longer travel distances, and economic disadvantage [7,14,15]. In Alabama, particularly in the Black Belt region, these barriers intersect with persistent poverty, underfunded health infrastructure, and racial inequities in the built environment (housing, food access, water systems), producing some of the highest cardiovascular mortality rates in the nation [16]. Geographic variation in blood pressure outcomes suggests that neighborhood- and zip code–level factors are as critical as individual clinical predictors [17].

An additional distinction of clinical importance lies in the separate contributions of systolic and diastolic blood pressure. Systolic hypertension, which increases with arterial stiffness and age, is a stronger predictor of cardiovascular events in older adults, whereas diastolic hypertension has been linked to stroke and cardiovascular risk among younger populations [18]. Research suggests that systolic pressure is more strongly associated with age and gender, while diastolic disparities often reflect racial differences and exposure to environmental and psychosocial stressors [3]. This makes simultaneous examination of systolic and diastolic outcomes essential for identifying nuanced disparities across populations.

Although many studies have documented hypertension disparities, fewer have examined how systolic and diastolic pressures vary simultaneously across race, gender, and geography in large, real-world datasets that capture diverse patients across health systems. Health information exchanges (HIEs) provide a unique opportunity to evaluate blood pressure outcomes at scale, integrating data from multiple providers and payers to reflect community-level patterns. Leveraging HIE data allows for a more comprehensive understanding of how demographic and geographic variables interact to shape cardiovascular risk, particularly in underserved rural areas.

This study focused on counties in West Alabama. Alabama has consistently been found to have a high prevalence of hypertension and a low control rate [19,20]. Data from the Alabama state health information exchange was used to examine differences in systolic and diastolic blood pressure by race, gender, age, and residential zip code in nine West Alabama counties. By distinguishing between systolic and diastolic pressures and incorporating geographic variation, the analysis provides a nuanced perspective on how biological, social, and structural factors converge to produce disparities in hypertension outcomes among adults in West Alabama.

Methods

Study design and data source

This study is part of a larger project designed to better understand and address factors that may underlie the high rate of hypertension in West Alabama, specifically 9 West Alabama counties (see Table 1). A retrospective analysis of blood pressure data from Alabama’s Health Information Exchange (HIE) from these nine counties from April 1, 2025 to May 9, 2025 was conducted. The authors obtained minimally identifiable protected health information, limited to patient ZIP codes. All data were transferred to a secure server and stored in a HIPAA-compliant environment, with access restricted to research team members approved by the Institutional Review Board (IRB). The sample included single clinic visits of 3,462 adult participants with available demographic and clinical information, including race, gender, age, residential zip code, and blood pressure values. All analyses were conducted using SAS 9.4, with significance defined at alpha level 0.05.

Table 1: Demographics of Target Area.

Ethics statement

The study was approved by the University of Alabama IRB (# 23-08-6857). A waiver of written informed consent was obtained from the IRB.

Blood pressure classification

Blood pressure stage was defined using established clinical cut points as suggested by the American Heart Association:

Normal-elevated blood pressure: systolic <130 mmHg and diastolic <80 mmHgStage 1 hypertension: systolic 130–139 mmHg or diastolic 80–89 mmHgStage 2 hypertension: systolic 140–180 mmHg or diastolic 90–120 mmHgHypertensive crisis: systolic >180 mmHg or diastolic >120 mmHg

Descriptive analyses

Frequencies of categorical variables (race, gender, blood pressure stage, and zip code) were summarized. Frequencies of blood pressure stages were examined overall and stratified by race and gender. Continuous variables (age, systolic blood pressure, and diastolic blood pressure) were summarized and descriptive statistics including sample size, mean, standard deviation, median, minimum, and maximum were produced. Group-specific means were calculated separately by race and by gender.

Normality assessment

The distributions of age, systolic blood pressure, and diastolic blood pressure were evaluated, stratified by race and gender. Tests of normality indicated that continuous outcomes were not normally distributed. Although analysis of variance (ANOVA) assumes normally distributed residuals, ANOVA is generally robust to moderate violations of this assumption in the presence of large sample sizes and reasonably balanced groups. Given the total sample size (n = 3,462) and analytic objectives, ANOVA was retained as the primary inferential method.

Inferential analyses

Two separate general linear models (GLM) were estimated to evaluate predictors of systolic and diastolic blood pressure. Independent variables included race, gender, age (continuous), and patient residential zip code. Zip code was specified as a categorical control variable to account for geographic variation. The following interaction terms were included in both models: race × gender, race × age, race × zip code, gender × age, and gender × zip code. For interactions involving continuous age, slope analyses were performed to evaluate whether the relationship between age and blood pressure differed across demographic subgroups. Specifically, estimates were computed for separate regression slopes of age on systolic blood pressure and diastolic blood pressure within levels of gender and within racial groups. These slope estimates allowed assessment of whether systolic and diastolic disparities disproportionately varied by gender or race across the age span. For interactions involving geographic variation (by zip code), least squares means (LS-means) estimates were generated. Specifically, LS-means were used to evaluate differences in systolic blood pressure across zip codes within subgroups defined by race and gender, and to evaluate differences in diastolic blood pressure across counties within racial groups. Pairwise differences of LS-means were requested to further explore significant interaction effects.

Multi-factor ANOVA of Systolic and diastolic blood pressure

Separate multi-factor ANOVA models were created to determine effects of gender, age, and zip code on systolic and diastolic blood pressure and whether those effects are independent or dependent on one (or more) of the other factor(s).

Results

Blood pressure summary

There are 3,462 individuals included in the analysis data; of this total 2131 are African American (62%), 1331 are Caucasian (38%), 895 are male (26%), and 2567 are female (74%). Normal blood pressure has 1142 (33%), Stage 1 hypertension has the most individuals with 1197 (35%), Stage 2 hypertension has 1044 (30%), and hypertensive crisis has the least with 79 (2%). Table 2 illustrates the overall demographic summary statistics for race, gender, blood pressure stage, and blood pressure (systolic and diastolic).

Table 2: Summary Table of Age, Systolic and Diastolic Blood Pressure, and Blood Pressure Stage Overall and by Gender and Race.

Systolic Blood Pressure (SBP)

The results indicate statistically significant effects of race, age, and geographic area (zip code) on SBP. Age and race emerged as the strongest predictors (both p < 0.0001), while zip code also showed significant geographic variation (p = 0.0328). Moreover, racial differences in blood pressure varied by zip code (p = 0.0294), and the relationship between age and SBP differed between males and females (p = 0.0244). In addition, gender differences in blood pressure depended on geographic area (p < 0.0001). A complete summary of model estimates is presented in Table 3.

Table 3: Type III SS Summary of Multi-Factor ANOVA of Systolic Blood Pressure.

Systolic blood pressure increased markedly with age, though the rate of increase differed by gender. Fig 1 presents predicted SBP as a function of age, stratified by gender. Both men and women showed significant positive associations between age and SBP, with steady increases across the lifespan. The estimated slope of age was 0.36 mmHg per year for men and 0.30 mmHg per year for women, indicating a steeper increase among men. At younger ages, men had slightly higher predicted SBP, and this difference widened with advancing age. The 95% confidence bands around each regression line demonstrate that the age effect on SBP is robust in both groups, with partial overlap between genders. Overall, these findings confirm that systolic blood pressure rises with age for both sexes, however, unlike the previous literature the rate of increase is greater in males.

Linear Prediction of Systolic Blood Pressure Across Age by Gender.

Race was also a significant predictor, and the interaction between race, age, and gender was further examined. Figs 2 and 3 illustrate the age-by-gender slopes for Black and White participants, respectively. Among Black participants, the estimated slope for age was β = 0.46 for males and β = 0.34 for females. At younger ages, Black females had higher predicted SBP than Black males; however, the steeper male slope led to convergence and eventual crossover around age 45, with males surpassing females thereafter, contradicting previously reported gender differences. Among White participants, the estimated slope for age was β = 0.22 for males and β = 0.26 for females. White males exhibited higher SBP at younger ages, but because White females’ slope was steeper, their predicted SBP converged with and surpassed that of males around age 90. This steeper slope for White females aligns with prior literature indicating that women have lower SBP at younger ages but experience a more rapid increase with aging.

Linear Prediction of Systolic Blood Pressure Across Age by Gender for Race = Black.

Linear Prediction of Systolic Blood Pressure Across Age by Gender for Race = White.

Diastolic Blood Pressure (DBP)

The results indicate statistically significant effects of race, gender, age, and geographic area (zip code) on diastolic blood pressure. Race, age, and gender were strong, independent predictors of DBP after adjustment for all other variables (p < 0.0001). Geographic area was also a significant factor, with notable differences across zip codes (p = 0.0127). Furthermore, racial differences in DBP varied by geographic area, as indicated by a significant race × zip code interaction (p = 0.0048). A full summary of the model estimates is provided in Table 4.

Table 4: Type III SS Summary of Multi-Factor ANOVA of Diastolic Blood Pressure.

Diastolic blood pressure was significantly influenced by race, gender, age, and geographic location. Race remained an independent predictor, and the magnitude of racial differences varied across zip codes. Geographic location itself contributed meaningfully to variability in DBP, underscoring the importance of contextual environmental factors.

Follow-up (slicing) analyses revealed that DBP varied significantly across zip codes for both Black (p < 0.0001) and White (p = 0.0342) participants. To further characterize these differences, adjusted mean DBP values were compared across zip codes. Among Black participants, the adjusted means ranged from the high 60s to high 90s mmHg, with some zip codes exhibiting markedly elevated estimates. Wider confidence intervals in sparsely populated zip codes suggested greater uncertainty in those areas. Among White participants, adjusted mean DBP values ranged from the low 60s to high 80s mmHg, consistently lower than those observed among Black participants across all geographic areas.

While many pairwise differences between zip codes were not statistically significant, the overall race × zip code interaction confirmed significant variability in diastolic blood pressure across geographic areas for both racial groups. These findings provide evidence of both intra- and inter-racial variability in DBP that is strongly influenced by local geographic context.

Discussion

This analysis of blood pressure data from more than 3,462 patients in West Alabama reveals meaningful disparities in blood pressure outcomes across race, gender, age, and geography. Importantly, the findings underscore distinct patterns for systolic versus diastolic blood pressure, suggesting that these measures capture different physiological mechanisms and risk profiles. Across all age groups, Black participants exhibited higher mean SBP and DBP compared to White participants. SBP exhibited significant gender differences, whereas DBP was shaped by both race and gender effects.

Systolic blood pressure was most strongly predicted by age, consistent with extensive evidence that systolic hypertension becomes increasingly prevalent with arterial stiffening and vascular aging [21]. Among White participants, SBP followed the previously reported pattern, with women exhibiting greater increases with age than men. This has important clinical implications given that elevated systolic pressure is a well-established predictor of myocardial infarction, stroke, and overall cardiovascular morbidity and mortality in older adults [22]. Moreover, Ji and colleagues [23] demonstrated that women experience increased cardiovascular risk even at lower systolic thresholds than men, underscoring the importance sex-specific blood pressure monitoring and intervention strategies.

Contrary to prior work indicating that women experience sharper increases in SBP with age [10], the present data suggest a steeper trajectory among men overall that is driven by Black patients. One explanation for the previous finding is that women show a steeper age-related increase in SBP due to hormonal changes [10]. Indeed, Black women do experience increases in SBP with age, and their slope is steeper than that of White women, corroborating Richardson and Brown’s [24] findings that race and gender may have a multiplicative effect on hypertension risk. However, Black patients demonstrated a markedly different trajectory than White patients: younger Black women had higher SBP than their male counterparts, but Black men showed a steeper rise in SBP with advancing age. Prior research has linked low socioeconomic position in both childhood and adulthood to higher blood pressure [25], particularly among Black men [26]. In a previous study examining perceived stress and depression among Black residents of the Alabama Black Belt, we found that men reported increasing stress, anxiety, and depression with age, whereas women did not [27]. This pattern suggests that psychosocial stressors may contribute disproportionately to age-related increases in SBP among Black men and could help explain the observed gender differences in hypertension risk.

By contrast, diastolic blood pressure was independently predicted by race, with Black adults exhibiting significantly higher DBP values even after controlling for age, gender, and geographic area. This pattern echoes prior findings from Carnethon et al. [3], who documented persistent racial disparities in hypertension despite adjustments for socioeconomic and clinical covariates. The independent association between race and DBP may reflect the cumulative physiological consequences of chronic psychosocial stress, structural racism, and adverse neighborhood conditions - factors that have been shown to increase vascular resistance and elevate diastolic pressure [28,29]. McEvoy et al. [30] also reported that isolated diastolic hypertension remains a significant predictor of cardiovascular disease in younger adults, underscoring the long-term implications of early disparities.

Geographic variation across zip codes further highlights the powerful influence of social determinants of health (SDoH). Both systolic and diastolic pressures varied significantly by location, suggesting that contextual factors such as socioeconomic deprivation, limited health care infrastructure, food insecurity, and transportation barriers play a key role in shaping cardiovascular risk [31,32]. These findings align with previous work demonstrating that neighborhood-level socioeconomic disadvantage is associated with poorer blood pressure control and worse cardiovascular outcomes [17]. In West Alabama, where rural communities continue to experience underinvestment in health services and high poverty rates [33], the geographic disparities observed here underscore the urgency of place-based strategies for hypertension control.

Combined census data from the nine West Alabama counties included in this analysis—Bibb, Fayette, Greene, Hale, Lamar, Perry, Pickens, Sumter, and Tuscaloosa—provide important socioeconomic context. Across these counties, the median household income is approximately $50,000, with a per capita income of$24,000. The overall poverty rate is 19%, and roughly 22% of residents rely on the Supplemental Nutrition Assistance Program (SNAP). A GINI index (a summary measure of income inequality) of 0.42 reflects moderate income inequality, while labor force participation (57%) and unemployment (4.8%) further underscore the region’s economic challenges. Considerable variation exists by zip code: for example, areas with the lowest prevalence of stage 1 and 2 hypertension were located in Tuscaloosa County (e.g., zip code 35475, median income $104,852), whereas those with the highest hypertension prevalence had substantially lower median incomes (e.g., zip code 36744 in Perry County,$33,983). These socioeconomic disparities highlight the broader context of structural disadvantage that likely contributes to elevated cardiovascular risk across the region.

Limitations

Several limitations should be acknowledged. Socioeconomic and behavioral factors (e.g., diet, weight, physical activity, medication adherence) were not uniformly available, limiting the ability to fully adjust for all risk factors. Additionally, geographic patterns were analyzed at the zip code level, which may obscure important within-community variation. The analysis focused on only two races. There is a growing Hispanic population that should be included in future studies. Finally, the data were derived from a health information exchange (HIE) system, which may be subject to variability in measurement practices and incomplete data capture across sites. Although the majority of federally qualified health centers in the state are connected to the HIE, participation is voluntary and not all private or smaller rural clinics contribute data. Consequently, the sample may be biased toward lower-income patients who more commonly receive care in safety-net settings. The relatively small sample size likely reflects several structural factors. Alabama does not mandate HIE participation for healthcare providers, and adoption remains limited among smaller rural clinics. In addition, with the exception of Tuscaloosa County, the included counties are sparsely populated and characterized by limited healthcare access, further constraining data availability. Finally, the HIE did not consistently capture information on anti-hypertensive medication use, limiting the ability to fully account for treatment effects.

Conclusions

aken together, these findings suggest that strategies to reduce hypertension disparities must address both biological aging processes and the structural inequities that shape exposure to risk. Clinically, systolic blood pressure appears most sensitive to age and gender, whereas diastolic pressure disproportionately affects Black adults across the lifespan. We emphasize that observed differences between Black and White adults are not hypothesized to reflect biological differences. Rather, we posit that these disparities are driven by differential exposure to chronic and cumulative toxic stress, shaped by social and environmental conditions. Public health interventions should prioritize rural and underserved areas, where geographic variation in blood pressure reflects broader inequities in access to care, housing stability, transportation, and access to foods and beverages that promote health.

The use of HIE data in this study demonstrates the value of population-level surveillance for identifying nuanced disparities that might be missed in smaller clinical samples. By integrating race, gender, and geography, the analysis highlights the need for multilevel interventions - from clinical management and prevention programs to structural and community-based efforts targeting the social determinants of cardiovascular health. Strengthening rural health infrastructure, expanding access to preventive care, and empowering community-based organizations in West Alabama represent critical next steps toward reducing the disproportionate cardiovascular burden experienced by Black communities and by women in later life.