Experiences of Racism and Risk of Preeclampsia in the Black Women’s Health Study

Abstract

Background:

Black women in the United States experience higher rates of preeclampsia incidence, severity, and case fatality than white women, which contributes to cardiovascular disease disparities. Differences in cardiometabolic risk factors do not fully explain this disparity, and there is growing consensus that racism may be a root cause. Few studies have evaluated associations between racism and preeclampsia among a cohort of exclusively Black women.

Methods:

We analyzed data from the Black Women’s Health Study, a follow-up study of approximately 59,000 Black women in the United States established in 1995. We included women who were nulliparous at time of enrollment and had their first birth from 1995 through 2009. Racism was measured as a total daily racism score (reported as daily exposure to interpersonal racism) and an institutional racism summary score (reported as ever treated unfairly due to race with respect to police, housing, or on-the-job). We used logistic regression models adjusted for demographic, reproductive, and health factors to estimate odds ratios (ORs) and 95% confidence intervals (CIs).

Results:

Women in the highest quartile of daily racism had a 50% higher odds of preeclampsia compared to the lowest quartile (aOR 1.50; 95% CI 1.13–1.99). Women reporting institutional racism in two domains had 47% higher odds compared to those reporting none (aOR 1.47; 95% CI 1.14–1.91).

Conclusions:

Greater experiences of daily and institutional racism were associated with an increased risk of preeclampsia, suggesting that racism contributes to the disproportionate burden of preeclampsia among Black women in the United States.

Graphical Abstract

INTRODUCTION

Preeclampsia is a multisystemic disorder of pregnancy characterized by maternal vascular dysfunction, placental abnormalities, and systemic inflammation.^1,2 It affects 2% to 8% of pregnancies in the United States (U.S.) and is a major cause of maternal morbidity and mortality, acute cardiovascular complications, and long-term cardiovascular disease.^3–8 Importantly, there are notable racial disparities in preeclampsia incidence, severity, and case fatality. Compared to white women, Black women have a 1.5 times higher incidence of severe preeclampsia, >3 times higher incidence of preeclampsia superimposed on chronic hypertension, and 2-3 times higher preeclampsia-related case fatality.^9–11 Elevated preeclampsia risk among Black women relative to other racial and ethnic groups has also been observed in multi-ethnic and international cohorts, indicating that these disparities are not limited to comparisons with White populations.^12,13 These disparities have remained persistent—and in some cases have worsened over time—as reflected in hypertension-related maternal mortality rates, which have been nearly four times higher among Black women than white women over the past forty years.¹⁴ Importantly, this excess risk among Black women extends to the long-term cardiovascular sequelae of preeclampsia such as chronic hypertension and cardiovascular disease.^15–17 Ultimately, these inequities fuel the dire state of Black maternal health in the U.S., where Black women are 2 to 4 more times likely to die as a complication of pregnancy than their white counterparts and face similar disparities related to severe maternal morbidity.^18–20 In this manuscript, we use the Centers for Disease Control and Prevention (CDC) definition of maternal health, referring to health and well-being during pregnancy, childbirth, and the postpartum period.²¹

While the exact causes of preeclampsia remain unknown, it is generally agreed that endothelial cell activation, intravascular inflammation, syncytiotrophoblast stress, and uteroplacental ischemia lead to clinical manifestations of hypertension, proteinuria, and other signs of end-organ damage after 20 weeks gestation.^1,2 Previous studies have also found that certain pregnancy-related factors (e.g., prior preeclampsia, nulliparity, multiple pregnancies), chronic diseases (e.g., increased body mass index (BMI), chronic hypertension, and diabetes mellitus) and social factors (e.g., low socioeconomic status, and psychosocial stressors) increase the risk of preeclampsia.^22–25

However, none of these aforementioned factors completely explain the persistent disparity Black women face. Experiences of racism (see Figure 1 for conceptual definitions), particularly interpersonal and institutional forms as operationalized in the present analysis, are dominant psychosocial stressors that contribute to racial health disparities,^26–28 including high rates of maternal morbidity and mortality among Black women in the U.S.^29–35

Figure 1.

Conceptual definitions of forms of racism referenced in this study. This figure presents conceptual definitions of multiple forms of racism to situate the study within the broader literature. In the present analysis, interpersonal racism was operationalized using the daily racism scale, and institutional racism was assessed based on self-reported experiences of unfair treatment across institutional domains (e.g., employment, housing, and policing). Structural and obstetric racism are included for conceptual context but were not directly measured. Definitions were informed by prior literature (references 23, 46–49).

Although a small number of studies have examined structural racism in relation to preeclampsia,^36,37 few have simultaneously evaluated interpersonal and institutional racism in relation to preeclampsia risk within a prospective cohort of Black women. Moreover, much of the existing literature relies on White comparator groups, which may obscure mechanisms through which racism operates within Black populations. ^38,39Race-conscious study designs that center Black women without defaulting to White referents may provide a more precise understanding of racism-related risk pathways. ^38,39

Our study addresses key conceptual (e.g., multidimensional exposures), methodological (e.g., reliance on white comparator groups), and empirical (e.g., limited prospective data focused on preeclampsia) gaps in the literature by examining the association between interpersonal and institutional racism and risk of preeclampsia among a large, socioeconomically diverse cohort of Black women in the Black Women’s Health Study (BWHS).

METHODS

Because of the sensitive nature of the data collected for this study, requests to access the dataset from qualified researchers trained in human subject confidentiality protocols may be sent to Black Women’s Health Study (BWHS) at Boston University at bwhs@bu.edu.

Conceptual Framework

We drew on several established frameworks to conceptualize the relationship between racism and adverse maternal health outcomes, including preeclampsia.^31,40–42 These frameworks identify both structural mechanisms (e.g., systemic and institutional racism) and individual-level mechanisms (e.g., interpersonal and internalized racism) as contributors to chronic stress, inflammation, and comorbid conditions. A visual depiction of our final adapted conceptual model is provided in the Supplement (Figure S1).

Study Population

We studied a subset of participants in the BWHS, a prospective follow-up study of “self-identified US Black women”. ^43,44 In this study, “self-identified U.S. Black women” refers to participants who self-reported their race as Black or African American and resided in the United States at enrollment, consistent with standard practice in U.S. epidemiologic research. Self-identification reflects individuals lived racialized experiences within the U.S. sociocultural and structural context, rather than externally assigned categories. This approach aligns with best practices in prior large-scale cohort studies, including the Black Women’s Health Study, and with federal guidance on the collection of race and ethnicity data, and is widely used to capture the social dimensions of race relevant to health outcomes.^45,46

In 1995, the BWHS enrolled approximately 59,000 women ages 21 through 69 who completed a 16-page health questionnaire. Participants complete mail and web-based questionnaires biennially to update health information on exposures and to ascertain incident disease. Completing and returning the questionnaires implies consent to participate. Follow-up is complete for >85% of potential person-years of the original cohort.^47,48 The Boston University Medical Campus Institutional Review (IRB) approved the original study protocol and the Johns Hopkins School of Medicine IRB approved this analysis.

We restricted the analytic cohort to nulliparous women who had their first pregnancy during the study period (1995 through 2009). We excluded women who remained nulliparous throughout follow-up (and thus were not at risk of developing the outcome, preeclampsia) or whose parity status was unknown ( and thus nulliparity at baseline could not be verified). We also excluded women who did not respond to questions on racism. A flowchart detailing the stages of sample selection can be found in Figure 2.

Figure 2.

Flowchart of sample selection, BWHS 1995 – 2009.

Assessment of Preeclampsia (Outcome)

In 1999 and 2009, the BWHS asked all women about any prior diagnosis of preeclampsia/toxemia. Detailed pregnancy outcomes were also queried on biennial questionnaires for women with births between 1995 and 2003. Specifically, women were asked if their baby had been born at least three weeks early and whether the preterm birth was because “labor was induced or had c-section because blood pressure was too high (preeclampsia/toxemia).”

Assessment of Self-reported Racism (Exposure)

Guided by conceptual definitions of racism described in Figure 1, we operationalized interpersonal and institutional forms of racism using validated self-report measures available in the BWHS; structural and obstetric racism were not directly measured in this study and are discussed conceptually in relation to our findings.

We evaluated interpersonal and institutional experiences of racism using a daily racism scale and an assessment of racism within specific institutional sectors, respectively. The 1997 BWHS questionnaire addressed experiences of daily racism adapted from Williams et al.⁴⁹ Daily racism was assessed with the question, “In your day-to-day life, how often have any of the following things happened to you?” followed by five specific scenarios: “You received poorer service than other people in restaurants or stores,” “People act as if they think you are not intelligent,” “People act as if they are afraid of you,” “People act as if they think you are dishonest,” and “People act as if they are better than you.” Response choices were “never,” “a few times a year,” “once a month,” “once a week,” and “almost every day,” coded as 1 to 5. A total daily racism score was averaged, rank ordered and categorized into quartiles (Quartile 1 lowest; Quartile 4 highest). Because tied values prevented exact quartile cut-points, we used approximate quartiles based on the rank-ordered distribution; however, for consistency, these groups are referred to as quartiles throughout the manuscript and tables.

Institutional racism was assessed with the question, “Have you ever been treated unfairly due to your race in any of the following circumstances? (1) Job, (2) Housing, (3) Police.” The institutional racism score was a sum of positive responses (0 to 3). Responses were summarized across items to capture cumulative exposure to racism, consistent with prior applications of validated discrimination scales. This approach improves measurement reliability and reflects the chronic nature of racism exposure demonstrated in prior Black Women’s Health Study analyses. ^50,51 Item-level stratified analyses were not pursued due to limited power and small cell sizes for individual items. Descriptive frequencies of each institutional racism domain (job, housing, police) are provided in the Supplemental Material (Table S5).

Covariate Assessment

We considered several variables potentially associated with racism and preeclampsia for inclusion in multivariable models, most of which were queried on the 1995 (baseline) or 1997 questionnaires and updated throughout follow-up. We further classified these variables as confounders or mechanistic variables based on associations with the exposure and outcome identified in prior studies.^23,50,51 See Figure S2.

In determining which covariates to include in the final models, we evaluated collinearity among cardiometabolic covariates and considered their potential role as intermediates in the relationship between racism and preeclampsia. Because hypertension, diabetes, and hypercholesterolemia may lie on the causal pathway between racism and preeclampsia and were collinear with body mass index, these variables were not included in the primary multivariable models but were examined in sensitivity analyses (Model 4; Tables S6–S7).

Statistical Analyses

We generated summary statistics to characterize participants overall and by preeclampsia status. We characterized missingness and conducted multiple imputation for missing covariates (BMI, education, health insurance, alcohol consumption, age at first birth). Missingness was examined under the recognized mechanisms of Missing Completely at Random (MCAR) and Missing at Random (MAR).⁵² We used Little’s test⁵³ to assess for MCAR and found no strong violations. We also compared regression estimates from models with and without missing covariates (Models 3 and 1, respectively) to evaluate potential bias. Based on these findings, we conducted Multiple Imputation by Chained Equations (MICE)⁵⁴ to impute missing values.

We conducted unadjusted and adjusted logistic regression analyses with the imputed data set to calculate Odds Ratios (ORs) along with 95% confidence intervals (CIs) of daily and institutional racism associated with preeclampsia.

The approach to modeling was as follows: Model 1 was unadjusted; Model 2 adjusted for age at first birth, education, health insurance, and total gravidity; Model 3 additionally adjusted for BMI, smoking, and alcohol consumption. BMI (measured continuously) was included in the final model as a potential mechanistic variable in lieu of hypertension, type 2 diabetes, and high cholesterol, given the high collinearity among these variables. We additionally fit an alternative model that adjusted for these cardiometabolic conditions (Model 4; Supplemental Tables S5–S6). Statistical significance was defined as a 95% confidence interval (CI) that did not cross 1.0. We conducted the analyses using R Statistical Software (version 4.2.2; R Core Team 2022).

We conducted sensitivity analyses to check the robustness of the results. In the first, we limited the study population to women with only one birth during the study period—testing whether the model results were similar to the full study population. In the second, we re-ran the models using the raw, non-imputed dataset, evaluating whether the ORs differed when restricted to the population with complete data in the main models.

RESULTS

A total of 5,181 participants were included in the analysis. Baseline characteristics according to preeclampsia status are shown in Table 1. The prevalence of preeclampsia in the study was 9.3%. Compared to those without preeclampsia, women with preeclampsia had higher total daily racism scores (2.30 vs. 2.20) and were more likely to experience institutional racism across two domains (25.5% vs 21.1%). They were also more likely to have a BMI >30 kg/m² (27.1% vs. 19.4%), diabetes (9.1% vs. 5.8%), high cholesterol (22.2% vs. 17.2%), and hypertension (38.1% vs. 18.6%). No differences were observed in age at first birth, smoking, or alcohol consumption. Women with preeclampsia were slightly more likely to have health insurance (96.3% vs. 92.6%).

Table 1.

Participant Characteristics Overall and by Preeclampsia Status, Black Women’s Health Study (BWHS), 1997 (N = 5181). ^*

Characteristics	Overall (N = 5181)	No Preeclampsia (N = 4699)	Preeclampsia (N = 482)
Total daily racism score
Mean (SD)	2.21 (0.77)	2.20 (0.77)	2.30 (0.78)
Median [Q1-Q3]	2.00 [1.60-2.60]	2.00 [1.60-2.60]	2.00 [1.80-2.60]
Total daily racism score, quartiles (Q), n (%)
Q1: 0 to 25th percentile	1401 (27.04)	1301 (27.69)	100 (20.75)
Q2: >25th to 50th percentile	1509 (29.13)	1358 (28.90)	151 (31.33)
Q3: >50th to 75th percentile	1126 (21.73)	1015 (21.60)	111 (23.03)
Q4: >75th to 100th percentile	1145 (22.10)	1025 (21.81)	120 (24.90)
Institutional racism, n (%)
No positive responses	1766 (34.09)	1629 (34.67)	137 (28.42)
One positive response	1892 (36.52)	1708 (36.35)	184 (38.17)
Two positive responses	1116 (21.54)	993 (21.13)	123 (25.52)
Three positive responses	407 (7.86)	369 (7.85)	38 (7.88)
Age at first birth, year
Mean (SD)	29.46 (4.25)	29.47 (4.25)	29.35 (4.21)
Educational attainment, n (%)
≤12 years	321 (6.20)	293 (6.24)	28 (5.81)
13-15 years	1611 (31.09)	1455 (30.96)	156 (32.37)
16 years	1888 (36.44)	1709 (36.37)	179 (37.14)
≥17 years	1361 (26.27)	1242 (26.43)	119 (24.69)
Health insurance, yes, n (%)	4815 (92.94)	4351 (92.59)	464 (96.27)
Smoking history, n (%)
Never smoker	4429 (85.49)	4009 (85.32)	420 (87.14)
Current smoker	335 (6.47)	307 (6.53)	28 (5.81)
Past smoker	417 (8.05)	383 (8.15)	34 (7.05)
Alcohol consumption, n (%)
Not current (<1 drink/week)	3970 (76.63)	3593 (76.46)	377 (78.22)
Current, 1-6 drinks/week	1134 (21.89)	1038 (22.09)	96 (19.92)
Current, ≥7 drinks/week	77 (1.49)	68 (1.45)	9 (1.87)
Body mass index, kg/m2
Median [Q1-Q3]	24.69 (21.92-28.73)	24.62 [21.92-28.35]	25.57 [22.59-30.61]
Body mass index, kg/m2, n (%)
Normal <25	2733 (52.75)	2514 (53.50)	219 (45.44)
Overweight 25-29.9	1404 (27.10)	1272 (27.07)	132 (27.39)
Obese ≥30	1044 (20.15)	913 (19.43)	131 (27.18)
Total gravidity, ≥2, n (%)	3168 (61.15)	2828 (60.18)	340 (70.54)
History of hypertension, ever, n (%)	1059 (20.44)	875 (18.62)	184 (38.17)
History of type 2 diabetes, ever, n (%)	320 (6.18)	276 (5.87)	44 (9.13)
History of high cholesterol, ever, n (%)	919 (17.74)	812 (17.28)	107 (22.20)

^*Bold values denote statistical significance defined as a p-value < 0.05

Daily Racism

Odds ratios and 95% CIs for the association between daily racism and preeclampsia are shown in Table 2. In the unadjusted model (Model 1), women in each increasing quartile of exposure to daily racism had significantly higher odds of preeclampsia compared to those in the lowest quartile. For example, women in the highest quartile had 52% greater odds of preeclampsia (OR 1.52; 95% CI 1.15, 2.01). After adjusting for age at first birth, education, health insurance, and gravidity (Model 2), the association remained significant (adjusted OR [aOR] 1.54; 95% CI 1.17, 2.04), and was only slightly attenuated with further adjustment for BMI, smoking, and alcohol consumption (Model 3) (aOR 1.50; 95% CI 1.13, 1.99).

Table 2.

Crude and Adjusted Odds Ratios (OR, aOR)^* for the Association Between Preeclampsia and Daily Racism, Black Women’s Health Study (BWHS), 1995-2009 (N=5181). ^†

	Model 1	Model 2	Model 3
Characteristics	OR (95% CI)	aOR (95% CI)	aOR (95% CI)
Total daily racism score quartile(Q) ‡,
Q1: 0 to 25th percentile	Ref (1.00)	Ref (1.00)	Ref (1.00)
Q2: >25th to 50th percentile	1.45 (1.11 - 1.88)	1.44 (1.10 - 1.87)	1.42 (1.09 - 1.86)
Q3: >50th to 75th percentile	1.42 (1.07 - 1.89)	1.42 (1.07 - 1.89)	1.40 (1.05 - 1.87)
Q4: >75th to 100th percentile	1.52 (1.15 - 2.01)	1.54 (1.17 - 2.04)	1.50 (1.13 - 1.99)

Abbreviations: OR = Odds ratio; CI = Confidence Interval

^*Adjusted models: Model 1: unadjusted; Model 2: adjusted for age at first birth, education, health insurance, and total gravidity; Model 3: adjusted for variables in Model 2 and alcohol consumption, smoking, and BMI.

^†Bold value denoted 95% confidence intervals did not include 1.00.

^‡Daily racism was measured in quartiles of the averaged responses to the following five questions: (1) “You received poorer service than other people in restaurants or stores,” (2) “People act as if they think you are not intelligent,” (3) “People act as if they are afraid of you,” (4) “People act as if they think you are dishonest,” and (5) “People act as if they are better than you.” Response items were “never,” “a few times a year,” “once a month,” “once a week,” and “almost every day,” and coded as scores 1 to 5.

Institutional Racism

Odds ratios and 95% confidence intervals (CIs) for the association between institutional racism and preeclampsia are shown in Table 3. Compared to women who reported no experiences of institutional racism, those with two positive responses had 47% greater odds of preeclampsia in unadjusted analyses (Model 1: OR 1.47; 95% CI 1.14, 1.90). The association remained consistent after adjusting for age at first birth, education, health insurance, and gravidity (Model 2: aOR 1.47; 95% CI 1.13, 1.90) and was nearly unchanged after additional adjustment for BMI, smoking, and alcohol consumption (Model 3: aOR 1.47; 95% CI 1.14, 1.91). Women with one reported experience of institutional racism also had significantly higher odds of preeclampsia across models (Model 3: aOR 1.27; 95% CI 1.01, 1.60). However, the association for women reporting all three domains of institutional racism was not statistically significant (Model 3: aOR 1.28; 95% CI 0.88, 1.88).

Table 3.

Crude and Adjusted Odds Ratios (ORs)^* for Association Between Preeclampsia and Institutional Racism, Black Women’s Health Study (BWHS), 1995-2009 (N=5181) ^†

	Model 1	Model 2	Model 3
Characteristics	OR (95% CI)	aOR (95% CI)	aOR (95% CI)
Institutional racism‡,
No positive responses	Ref (1.00)	Ref (1.00)	Ref (1.00)
One positive response	1.28 (1.02 - 1.62)	1.29 (1.02 - 1.62)	1.27 (1.01 - 1.60)
Two positive responses	1.47 (1.14 - 1.90)	1.47 (1.13 - 1.90)	1.47 (1.14 - 1.91)
Three positive responses	1.22 (0.84 - 1.78)	1.28 (0.87 - 1.87)	1.28 (0.88 - 1.88)

Abbreviations: OR = Odds ratio; CI = Confidence Interval

^*Adjusted models: Model 1: unadjusted; Model 2: adjusted for age at first birth, education, health insurance, and total gravidity; Model 3: adjusted for variables in Model 2 and alcohol consumption, smoking, and BMI

^†Bold value denoted 95% confidence intervals did not include 1.00.

^‡Institutional racism was measured as the sum of positive (yes) responses to the following three questions: “Have you ever been treated unfairly due to your race in any of the following circumstances (yes/no)? (1) Job, (2) Housing, (3) Police.” The summed values were coded as none (no positive responses), 1 (one positive response), 2 (two positive responses), and 3 (all positive responses).

Sensitivity Analyses

In sensitivity analyses, we limited the study population to women with only one birth during the study period (n=2720) to test whether the final model results were similar to the full study analytic population. The final model results were similar, with slightly stronger associations for interpersonal racism. In the fully adjusted model, women in the highest compared to the lowest quartile of daily racism had 96% increased odds of preeclampsia (aOR 1.96; 95% CI 1.32, 2.92) (Table S1). Estimates for institutional racism were directionally consistent in the restricted sample of women with one birth, though not statistically significant (Table S2). Results remained robust using raw, non-imputed data, with similar effect sizes and statistical significance (Tables S3–S4).

DISCUSSION

In this cohort of Black women, those in the highest quartile of daily racism faced an approximately 1.5 times higher odds of preeclampsia than those in the lower quartile, with somewhat stronger associations noted for women who had only one birth during follow-up. Risk increased across quartiles, with the highest quartile showing a statistically significant association. In addition, women who reported institutional racism in two domains had 47% higher odds of preeclampsia compared to those reporting none. Our findings help to shed light on the potential role of racism in the persistent disparities in Black maternal health in the US and add to the growing body of evidence demonstrating the detrimental effects of racism on women’s health.

Our findings build on a growing body of research linking racism—both structural and interpersonal—to hypertensive disorders across the reproductive life course. In a single hospital study in Chicago, Mayne et al used electronic health record data representing 4,748 Black births to evaluate the relationship between segregation and hypertensive disorders of pregnancy (HDP). The authors found that residential segregation was associated with greater odds of HDP among Black women living in high-poverty neighborhoods.³⁶ Francis et al. analyzed intergenerationally linked birth records from over 45,000 Black mothers and their infants in California to examine how exposure to racial and economic segregation across the life course influences HDP risk. Using the Index of Concentration at the Extremes, they found that living in deprived neighborhoods during either early childhood or adulthood was associated with increased odds of HDP, with the highest risk observed among women who lived in deprived neighborhoods at both time points.³⁷ Both studies utilized geospatial proxies of structural racism. In contrast, our study captured self-reported experiences of daily interpersonal and institutional racism, offering a more direct lens on lived racism-related stress. Each approach has strengths and limitations, with geospatial measures offering structural insight and self-report measures reflecting individual perception and burden.

Recently, Janevic et al. combined these approaches in a prospective study of a multiracial cohort (N = 373), reporting that both gendered racial microaggressions—a form of personally mediated racism—and structural racism, measured using the multidomain Structural Racism Effect Index (SREI; e.g., across education, income, housing, criminal justice indices), were associated with elevated postpartum blood pressure. Their study represents a robust model of how structural and interpersonal exposures can jointly shape maternal cardiovascular outcomes.⁵⁵.While their focus was on the postpartum period, our findings complement theirs by identifying associations between racism and hypertensive risk during pregnancy, reinforcing the importance of assessing racism across the life course.

Racism is increasingly recognized as a chronic psychosocial stressor that can disrupt neuroendocrine, vascular, and immune pathways during pregnancy. Prior studies have linked psychosocial stress to preeclampsia and other adverse outcomes, but definitions of “stress” were broad and heterogeneous, including depression, anxiety, financial concerns, and global phenomena (e.g., COVID-19 pandemic and catastrophic weather).^56,57 Conceptual models such as the weathering hypothesis posit that chronic exposure to racism-related stress contributes to cumulative physiologic wear over time through sustained activation of neuroendocrine and inflammatory pathways.^58–60This framework, often operationalized health deterioration and adverse cardiometabolic and reproductive outcomes among Black women. ^59,61–63 Our findings are consistent with this broader conceptual literature suggesting that racism-related stress exposures may be relevant to hypertensive risk during pregnancy. While racism is considered a form of psychosocial stress, to date the literature on psychosocial stress and adverse pregnancy outcomes has mostly excluded racism.

The dearth of research in this area is even more troubling considering consistent evidence that racial disparities exist in the experience of psychosocial stressors before, during, and after pregnancy and that non-Hispanic Black women are more likely to have greater psychosocial stressors than their white counterparts.^58,64–66

Our study adds to a limited body of work that examines racism more directly. Earlier research has shown associations between racism exposure and maternal blood pressure⁶⁷ or has evaluated chronic stress through biomarkers such as allostatic load, which have been linked to preeclampsia risk .^61,64 These pathways are consistent with the weathering hypothesis, which describes the cumulative physiologic toll of racism-related stress, particularly for Black women.

Cardiovascular Health Implications

Numerous studies have explored the connections between racism and various cardiometabolic risk factors, such as racism’s correlation with hypertension,^68,69 obesity,⁷⁰ and diabetes.⁴⁸ Our findings combined with established findings on preeclampsia independently predicting cardiovascular disease,⁶ suggest that experiences of racism may be associated with cardiometabolic risk factors that are in turn linked to increased cardiovascular mortality. These patterns are consistent with conceptual models in which racism may operate upstream of cardiometabolic risk factors that contribute to cardiovascular mortality, although causal pathways cannot be established in this observational study.

Notably, cardiovascular-related disease ranks as the primary cause of maternal mortality among Black women. Black women also have the highest percentage of mortality related to hypertensive disorders of pregnancy compared to other racial groups.⁷¹ From a life course perspective, unaddressed experiences of racism may amplify this risk, perpetuating disproportionate mortality rates for Black women.

Recognizing these challenges, the Association of Black Cardiologists, the American College of Cardiology, and the American Heart Association issued a joint statement underscoring the crucial role of cardiologists in addressing the multifactorial drivers of Black maternal health disparities (including the role of racism).⁷²

Clinical Management Strategies

To improve pregnancy outcomes and patient-provider relationships, it is imperative for healthcare providers and institutions to recognize and address racism stress in the clinical encounter. Utilizing Patient Reported Experience Measures designed explicitly for Black mothers, such as the recently validated PREM-OB scale, and structured EHR screening for structural racism but also gendered racism and other intersectional experiences could help providers better understand and address the lived experiences that shape maternal health. ^73–75 Implicit bias training, while important, must be paired with institutional accountability and strategies such as upstander training,^75,76incident reporting mechanisms, and inclusion of racism as a contributing factor in maternal morbidity and mortality reviews.⁴²

Finally, storytelling is a powerful mechanism to humanize the experiences of Black women and call attention to the ongoing challenge of Black Maternal Health Care in the United States. In 2021, the American association of Black Cardiologists launched: “We Are the Faces of Black Maternal Health^™”⁷⁷ campaign, highlighting personal and professional experiences with Black maternal morbidity and mortality.⁷⁸ Hospitals should consider creating mechanisms for storytelling and reflection within their institutions to augment traditional bias mitigation training and contextualize and humanize the review of adverse maternal health outcomes.

Research Implications

The current study underscores the importance of comprehensively studying racism and diversifying the focus of maternal health beyond comparisons between Black and white women. Future research should continue to examine racism as a multidimensional exposure, incorporating both self-reported and structural measures. Studies that focus on the experiences of diverse groups of Black women, including those of higher socioeconomic status, are essential to understanding the full spectrum of racism’s impact. Standardized tools to capture structural, interpersonal, and gendered racism will help advance the field and guide intervention design.⁷³

Strengths and Limitations

Certain limitations of the study should be acknowledged. First, preeclampsia was self-reported as a clinician-diagnosed condition on BWHS questionnaires which did not capture the full clinical diagnostic criteria or timing relative to pregnancy (i.e. postpartum preeclampsia). Thus, this outcome may be subject to misclassification, including potential misclassification with other hypertensive disorders of pregnancy. . Prior studies suggest reasonable recall accuracy—especially for early-onset cases (sensitivity: 83–92%, specificity: 100%).^79,80 However, late-onset preeclampsia may be underreported (late-onset preeclampsia had a sensitivity range of 72.6−78.0% and a specificity range of 96.0−98.6%, whereas early-onset preeclampsia had a sensitivity range of 83.0−92.0% and a specificity of 100%.)^78,79 Any resulting misclassification is likely non-differential with respect to racism exposure and would be expected to bias associations toward the null. Additionally, because preeclampsia was ascertained through participant follow-up questionnaires, we were unable to collect outcome data from women who died prior to reporting; however, preeclampsia-related mortality in the cohort during the analytic window was exceedingly rare (one death), making meaningful bias from this mechanism unlikely.

Second, some clinically relevant covariates were not available in the BWHS pregnancy questionnaires and therefore could not be included in adjustment models. In particular, chronic kidney disease (CKD) status was not collected and could not be accounted for. Because underlying kidney disease may be associated with baseline proteinuria predating pregnancy, some outcome misclassification is possible, particularly given evidence that women with CKD have up to 10 times higher risk of developing preeclampsia than women without CKD, and that risk is disproportionately higher among Black women.^81,82 However, CKD complicates an estimated 0.1–3% of pregnancies, suggesting that the magnitude of potential unmeasured confounding is likely limited.⁸³ Additionally, partner- and family-level characteristics were not captured in the BWHS racism instruments, which focus on individuals’ lived experiences of racism; thus, we were unable to examine the potential modifying or buffering effects of social relationships on the association between racism and preeclampsia.

Third, experiences of interpersonal racism and institutional racism are not independent of one another and instead have a bidirectional, if not interactive, relationship. However, our study was underpowered to investigate a joint effect of institutional and interpersonal racism. Similarly, our measures of personally mediated racism did not fully capture other intersecting experiences of marginalization, such as those related to gender or social position. The surveys administered to participants in the BWHS were conducted in an earlier era, thus some of the contemporary tools that capture these nuances, like gendered racism scales and patient-reported experiences of gendered racism reflecting obstetric racism (Figure 1),^75,84 were not available but should be incorporated into future assessments within Black study populations.

Additionally, our operationalization of institutional racism as a summed count of domains (job, housing, police) assumes equivalence across types of experiences. This approach may obscure heterogeneity in the impact of specific domains, as different forms of institutional racism may have distinct relationships with health outcomes. Although item-level analyses were limited by sample size, future studies should examine domain-specific effects to better understand these relationships.

In parallel with these limitations, the present study is strengthened by the use of a large, socioeconomically and geographically diverse prospective cohort of Black women with detailed covariate data and the ability to conduct multivariable-adjusted and sensitivity analyses.

CONCLUSIONS

In summary, experiences of interpersonal and institutional racism were independently associated with increased preeclampsia risk among Black women in a large, diverse national cohort. These findings highlight the role of racism in shaping persistent disparities in Black maternal health in the United States.

PERSPECTIVES

This study demonstrates that greater exposure to interpersonal and institutional racism is associated with higher odds of preeclampsia among Black women, highlighting racism as a clinically relevant social exposure that may contribute to observed disparities in hypertensive disorders of pregnancy. Although causal inferences cannot be drawn from these observational data, the findings are consistent with conceptual models that frame racism as a chronic psychosocial stressor with potential downstream effects on vascular, inflammatory, and neuroendocrine pathways during pregnancy. These results underscore the importance of situating preeclampsia risk within a broader social and structural context, alongside established biomedical risk factors.

Future research should prospectively examine multidimensional measures of racism—including structural and obstetric racism—and incorporate biomarkers of stress and vascular dysfunction to clarify potential mechanistic pathways. Longitudinal and interventional studies are needed to evaluate whether reducing racism exposure within healthcare and community contexts is associated with improvements in hypertensive pregnancy outcomes. Integrating rigorously measured social exposures into cardiovascular and obstetric research may strengthen efforts to more fully characterize risk, identify modifiable upstream contributors to inequities, and inform multilevel strategies to reduce disparities in maternal cardiovascular health.

NOVELTY AND RELEVANCE.

What is New?

• The research addresses the gap in knowledge regarding racism and preeclampsia risk specifically among Black women.

• Greater experiences of self-reported daily racism and institutional racism were associated with an increased risk of preeclampsia among Black women.

  • Specifically, Black women in the highest quartile of a daily racism score had 1.5 times higher odds of preeclampsia compared to women in the lowest quartile.
  • Similarly, Black women who experienced institutional racism from two domains had 1.47 times higher odds of preeclampsia compared to those with no experience of institutional racism.

• The findings suggest that racism is a contributor to the disproportionately high prevalence of preeclampsia among Black women in the United States

What is Relevant?

• Racial disparities in preeclampsia are not fully explained by traditional clinical risk factors; social exposures such as racism may be relevant to risk stratification and prevention frameworks.

• Centering Black women’s lived experiences of racism—rather than relying solely on race-based comparisons—may help clarify pathways contributing to inequities in hypertensive disorders of pregnancy.

• Incorporating social determinants into cardiovascular and obstetric research aligns with emerging calls to integrate structural drivers of health into hypertension science.

Clinical/Pathophysiological Implications

• Racism-related stress may represent a relevant psychosocial exposure to consider in patient-centered, trauma-informed obstetric care, while acknowledging that causal pathways remain to be established.

• Standardized, scalable tools (e.g., EHR-integrated patient-reported measures) could facilitate systematic assessment of racism-related experiences in clinical and research settings.

• Health systems may consider incorporating consideration of racism and discrimination into maternal morbidity and mortality reviews to inform quality improvement and equity-focused interventions.

Non-standard Abbreviations & Acronyms

aIRR

Adjusted Incidence Rate Ratio

aOR

Adjusted Odds Ratio

BEAD

Biostatistics, Epidemiology and Data Management

BHWS

Black Women’s Health Study

HDP

Hypertensive Disorders of Pregnancy

MAR

Missing at Random

MCAR

Missing Completely at Random

MICE

Multiple Imputation by Chained Equations

OR

Odds Ratio

SREI

Structural Racism Effect Index