Sex differences in blood pressure hemodynamics and white matter hyperintensity volume across racial ethnic groups—HABS‐HD

Figures

• 

  FIGURE 1. Sex differences in blood pressure across race and ethnic groups. Boxplots with overlaid jittered data points display systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP), and mean arterial pressure (MAP) by sex (male, female) within each race and ethnic group: non‐Hispanic White (NHW), non‐Hispanic Black (NHB), and Hispanic. Female participants are shown in firebrick and male participants in dark gray. Individual points represent participants, and boxplots summarize group‐level distributions. Corresponding sex differences were quantified using analysis of covariance models adjusted for age and education attainment.

• 

  FIGURE 2. Association of blood pressure metrics with white matter hyperintensities (WMHs) by race and ethnicity and sex. Rows denote race and ethnic strata and columns denote blood‐pressure (BP) metrics. A, non‐Hispanic White. B, non‐Hispanic Black. C, Hispanic. Columns (left → right) denote: systolic BP (SBP), diastolic BP (DBP), pulse pressure (PP), and mean arterial pressure (MAP). Within each panel, semi‐transparent points (jittered for visibility) show individual participants. Lines are sex‐specific predicted values from linear models with a BP × sex interaction; shaded ribbons are 95% confidence intervals around those predictions. BP predictors are scaled per 10 mmHg, but x axes are plotted in mmHg (i.e., BP10×10). The outcome is intracranial volume (ICV)–adjusted log WMH residuals. All models are adjusted for age, years of education, magnetic resonance imaging scanner, diabetes, dyslipidemia, obesity, and tobacco dependence. Color/linetype encodes sex (male vs. female) and is consistent with each panel's legend.

• 

  FIGURE 3. Associations of systolic blood pressure, diastolic blood pressure, pulse pressure, and mean arterial pressure with white matter hyperintensity (WMH) volume stratified by antihypertensive use and race ethnicity. Forest plots show the standardized β coefficients (95% confidence interval [CI]) from linear regression models evaluating the associations between blood pressure metrics and WMH volume among (A) non‐Hispanic White, (B) non‐Hispanic Black, and (C) Hispanic participants. Each panel displays results stratified by antihypertensive medication use (left: no antihypertensives; right: on antihypertensives). For each blood pressure (BP) metric—systolic BP, diastolic BP, pulse pressure, and mean arterial pressure—the plot displays the BP main effect (black) and the BP × sex interaction effect (colored: red for non‐Hispanic White, purple for non‐Hispanic Black, lavender for Hispanic participants). All models were adjusted for age, education, magnetic resonance imaging scanner type, diabetes, dyslipidemia, obesity, and tobacco use. The vertical dashed line represents the null value (β = 0). Positive estimates indicate higher WMH burden with increasing BP metric values; interaction…

• 

  FIGURE 4. Joint pulse pressure–diastolic blood pressure distributions and associations with WMH volume across race and ethnic groups. A, Sankey diagrams illustrating the distribution of participants across combined pulse pressure (PP) and diastolic blood pressure (DBP) categories among non‐Hispanic White (NHW), non‐Hispanic Black (NHB), and Hispanic participants. PP was categorized as narrow (< 40 mmHg), normal (40–60 mmHg), or widened (> 60 mmHg), and DBP as low (< 60 mmHg), normal (60–80 mmHg), or high (> 80 mmHg). Flow widths represent the number of individuals in each PP–DBP combination, with node sizes and embedded numeric counts indicating category frequencies. B, Adjusted associations between sex (female vs. male), PP categories (reference = normal PP), and their interaction in relation to log‐transformed, intracranial volume–normalized white matter hyperintensity (WMH) volume. Results are shown separately for NHW, NHB, and Hispanic participants. Estimates reflect regression coefficients with 95% confidence intervals (CIs) from models adjusted for age, education, magnetic resonance imaging scanner type, diabetes, dyslipidemia, obesity, and tobacco use. Positive values indicate…