Central Hemodynamics in African American Women: Examining the Role of Superwoman Schema Endorsement

Abstract

Background

African American women bear a disproportionate burden of cardiovascular diseases, potentially due to altered central hemodynamics. Racism and sexism often lead to African American women taking on numerous caretaking roles and overall increases their use of the Strong Black Woman (ie, Superwoman) mindset, which may have negative health consequences. We hypothesized that endorsing the Superwoman role and its Obligation to Help Others dimension would be associated with a deleterious central hemodynamics profile in African American women.

Methods and Results

Using cross‐sectional data, we examined central systolic blood pressure (mm Hg; n=408), augmentation index (percentage, adjusted for height and heart rate; n=408), and pulse wave velocity (m/s; n=368) in African American women aged 30 to 46 years. The Giscombe Superwoman Schema (SWS) questionnaire assessed endorsement of Overall SWS (range, 0–105) and SWS–Obligation to Help Others (range, 0–3). Multiple linear regression modeled associations between Overall SWS (10‐unit increments) and SWS–Obligation to Help Others (1‐unit increments) and central hemodynamics while adjusting for pertinent sociodemographic, clinical, and psychosocial factors. In fully adjusted models, central systolic blood pressure was significantly associated with Overall SWS (β=0.83 [95% CI, 0.19–1.47]) and SWS–Obligation to Help Others (β=2.03 [95% CI, 0.39–3.67]). Augmentation index was associated with Overall SWS (β=0.66 [95% CI, 0.02–1.30]) and SWS–Obligation to Help Others (β=2.21 [95% CI, 0.58–3.84]). Significant associations were not observed between pulse wave velocity and SWS.

Conclusions

Greater endorsement of the Superwoman role and prioritizing caregiving over self‐care were associated with higher central systolic blood pressure and augmentation index, which may contribute to adverse cardiovascular health among African American women.

Nonstandard Abbreviations and Acronyms

MUSE

Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health

SWS

Superwoman Schema

SWS‐Help

SWS—Obligation to Help Others

Clinical Perspective.

What Is New?

• We investigated whether endorsement of the Superwoman Schema, a sociocultural framework used by African American women amid historical and persistent race‐ and gender‐based oppression, was associated with a detrimental central hemodynamics profile.

• African American women who prioritized caregiving over self‐care and overall felt more obligated to fulfill the Superwoman role in society exhibited higher central systolic blood pressure and augmentation index, which may partially explain the increased risk of cardiovascular disease and stroke in this population.

What Are the Clinical Implications?

• Researchers, clinicians, and policymakers aiming to reduce cardiovascular disease and stroke risk in African American women should strongly consider and incorporate their unique lived experiences, including, but not limited to, Superwoman Schema endorsement.

African American women have the highest prevalence of hypertension and continue to experience the highest rates of cardiovascular disease (CVD) and stroke relative to women from other racial and ethnic groups. ¹ Early detection and tracking of subclinical CVD risk markers can serve as critical tools for understanding the development of hypertension, CVD, and stroke and aiding in their prevention. ² , ³ Central systolic blood pressure, augmentation index, and pulse wave velocity are independent predictors of cardiovascular events and death and thus have emerged as promising CVD risk markers. ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ Central systolic blood pressure has been posited as a better predictor of CVD and stroke than peripheral blood pressure because it more closely reflects the pressure exerted on target organs (heart, brain, and kidneys). ⁹ , ¹⁰ Further, greater pulse wave reflections from the periphery (quantified as augmentation index) and increased arterial stiffness (quantified as pulse wave velocity) are known contributors to elevated central systolic blood pressure. ² Prior studies indicate that central systolic blood pressure, augmentation index, and pulse wave velocity are all elevated in African American relative to White men and women. ¹¹ , ¹² Several studies have investigated the impact of traditional CVD risk factors (eg, smoking ¹³ ) and interventions (eg, exercise ¹⁴ , ¹⁵ ) on the aforementioned central hemodynamic parameters in African American men and women combined. However, limited studies ¹⁶ have investigated the impact of psychosocial factors on central hemodynamics specifically among African American women, who may face unique psychosocial stressors due to their dual minority status in society. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ Importantly, psychosocial stress contributes to biological responses (allostasis), which, over time, cause neurohormonal imbalances, including augmented systemic inflammation and oxidative stress (allostatic overload). ²¹ , ²² These neurohormonal imbalances, in turn, can lead to altered central hemodynamics including increased arterial stiffness ²³ , ²⁴ ; however, the precise psychophysiological pathways remain incompletely understood. ²⁰

Several researchers have argued that as a product of long‐standing structural racism, ²⁵ African American women are more likely to find themselves in roles that require them to expend substantial physical and psychological resources caring for others. ²⁶ For example, in part because African American men are disproportionately incarcerated, African American women have the highest prevalence of single motherhood and are therefore more likely to take on a greater burden of psychological and financial stress that accompanies raising children by themselves. ²⁷ , ²⁸ , ²⁹ This strong obligation to care for and help others in numerous contexts is 1 dimension of the Superwoman Schema (SWS), a psychological framework that many African American women adopt to help manage the accompanying expectations, stereotypes, and stressors associated with their dual minority status. ³⁰ , ³¹ Although SWS encompasses dimensions that are relatively masculine or gender neutral, such as suppressing emotions, resisting vulnerability, manifesting strength, and an intense motivation to succeed, SWS–Obligation to Help Others (SWS‐Help) stands out as having notably feminine characteristics. ³⁰ , ³² While some aspects of SWS are thought to be protective for physical and mental health, SWS‐Help appears to be detrimental. ³⁰ For example, greater endorsement of SWS‐Help has been directly linked to depression ³³ and poor sleep quality ³⁰ among African American women. Thus, emerging evidence suggests that SWS‐Help may be particularly relevant (and detrimental) to behavioral and mental health among African American women; however, more research is needed to understand whether SWS‐Help endorsement has a concomitant impact on cardiovascular health and function.

Research on SWS and physical health is in its early stages; thus, few studies ³⁴ , ³⁵ have investigated the impact of SWS endorsement, and specifically SWS‐Help, on CVD risk markers. Importantly, understanding the relationship between salient psychosocial factors and subclinical CVD risk markers will be critical for mitigating cardiovascular health disparities among African American women. ¹⁹ , ³⁶ , ³⁷ Accordingly, we tested the hypothesis that Overall SWS and SWS‐Help endorsement would be positively associated with central systolic blood pressure, augmentation index, and pulse wave velocity in African American women.

Methods

Participants

We analyzed baseline data from 422 African American women from the MUSE (Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health) cohort. Participants were recruited from December 2016 to March 2019 via Metro Atlanta residential and voter registration lists. Inclusion criteria were self‐identifying as a Black or African American woman, age 30 to 45 years at the time of enrollment, and being premenopausal with at least 1 ovary. Exclusion criteria were history of clinical CVD, currently pregnant or lactating, clinical conditions known to influence CVD (eg, kidney disease, HIV/AIDS), current treatment for mental health conditions, shift work, illicit drug use, or alcohol abuse. By design, 50% of the participants reported an income above the median Georgia income of US$50 000, while the remaining 50% reported a personal income below Georgia's median income. Lastly, we utilize the term “African American” throughout because “Black” refers to race alone and includes individuals who identify as African (eg, immigrants from the African continent) or Afro‐Caribbean (individuals from the West Indies), as well as those who identify as African American. ³⁸ , ³⁹ These 3 groups of women have different risk profiles and distinct histories in the United States. ⁴⁰ , ⁴¹ Thus, it is unclear whether the construct of Superwoman Schema would operate similarly across every group that identifies as “Black.” Furthermore, we prioritized the term African American over Black American to be consistent with MUSE study materials and the foundational SWS papers. ³⁰ , ³²

Study Design

The participants completed an in‐person visit at Emory University, which included self‐administered surveys and face‐to‐face interviews to assess the exposures of interest as well as a number of covariates, as described below. Physical measurements, including resting heart rate and blood pressure, height, weight, and central hemodynamic measurements, were also obtained during the in‐person visit. Clinical measurements were taken in a climate‐controlled room (20–24 °C), and participants were asked to fast for at least 4 hours before arrival.

Measurements

Superwoman Schema

To quantify SWS endorsement, we used the validated 35‐item Giscombe Superwoman Schema Questionnaire. ³⁰ , ³² Participants were asked to indicate the extent to which various SWS statements are true for them, with response options ranging from “not true at all” (coded 0) to “this is true for me all the time” (coded 3). The Giscombe Superwoman Schema Questionnaire contains the following 5 subscales: (1) Obligation to Help Others (SWS‐Help; 9 items; eg, “When others ask for my help, I say yes when I should say no.”); (2) Intense Motivation to Succeed (6 items; eg, “I accomplish my goals with limited resources”); (3) Resistance to Being Vulnerable (7 items; eg, “It's hard for me to accept help from others.”); (4) Obligation to Suppress Emotions (7 items; eg, “I display my emotions in privacy.”); and (5) Obligation to Present an Image of Strength (6 items; eg, “I have to be strong.”). ³⁰ Using the questionnaire instructions provided directly by Woods‐Giscombe, ³⁰ , ³² we calculated an average score for each subscale (range, 0–3), as the average score allows us to compare across subscales with differing numbers of items, as well as an Overall SWS score by summing the responses from all 35 items (range, 0–105). Higher scores indicate greater endorsement of the various SWS constructs. As noted previously, the current analysis focused on Overall SWS and SWS‐Help, with Cronbach's α of 0.95 and 0.89, respectively. ³⁰ , ³² The test–retest reliability for the 5 SWS subscales over 4‐ to 8‐week ranges from r=0.46 to r=0.89 (P<0.05 for all), and SWS‐Help, specifically, has a test–retest reliability of r=0.89 (P<0.05). ³⁰

Central Hemodynamics

In accordance with established practices for quantifying arterial stiffness and central hemodynamics in humans, ⁸ , ³⁸ each participant underwent pulse wave analysis and pulse wave velocity assessments via the SphygmoCor XCEL system (AtCor, Naperville, IL). Participants rested supine for ≈15 minutes before the measurements, which were completed on the right side of the body. For pulse wave analysis, a cuff was placed around the upper arm to obtain arterial pulsations from the brachial artery. The pressure waveform obtained at the brachial artery was then subjected to an online generalized transfer function, thereby allowing the synthesis of a central (aortic) blood pressure waveform. ³⁹ Outputs from this procedure were central systolic blood pressure and the augmentation index. For the carotid‐to‐femoral pulse wave velocity assessment, the distances between the following sites were obtained: palpated common carotid artery pulsation to the suprasternal notch; suprasternal notch to the top of a blood pressure cuff placed on the upper thigh; and the palpated common femoral artery pulsation to the top of the thigh cuff. Femoral artery pulsations were detected by a thigh cuff, while simultaneous common carotid artery pulsations were detected via applanation tonometry. The SphygmoCor XCEL unit used the distance measurements along with the detected pulsations to calculate pulse transit time and thus pulse wave velocity. All pulse wave analysis and pulse wave velocity measurements were completed once, and repeat measurements were taken only when necessary to meet quality control standards.

Covariates

Covariates were chosen on the basis of prior studies ⁴⁰ , ⁴¹ to ensure that the findings were not confounded by sociodemographic and clinical factors traditionally associated with CVD outcomes. We assessed the following sociodemographic factors via self‐report: age, annual family income (<$35 000; $35 000–$49 999; $50 000–$74 999; and >$75000), and education (years). Body mass index, in kg/m², was calculated from height and weight measurements acquired from a stadiometer and scale. Current smoking, presence of diabetes, and antihypertensive medication use were determined via self‐report. Resting peripheral blood pressure was obtained after 15 minutes of rest in the seated position (feet flat on the floor, arm supported at heart level) via an automated sphygmomanometer, and an average of at least 2 measurements was used. Resting heart rate was obtained during the SphygmoCor assessment. Severity of depressive symptoms was assessed via self‐report using the Beck Depression Inventory (range, 0–60; higher scores indicate greater symptom severity), ⁴² excluding 1 sleep‐related question as in prior studies of a similar nature. ⁴³ , ⁴⁴ To ensure that the results were uniquely attributable to the SWS as opposed to general caregiving or lack of social support, we measured caregiving and social support via the following questions that have been used previously in large‐scale survey studies of African American adults ⁴⁵ , ⁴⁶ , ⁴⁷ : (1) How often do you help out people in your family?; (2) How often do you help out your friends?; (3) How often do people in your family help you out?; and (4) How often do your friends help you out? Responses ranged from 1 to 4 (never to very often) for question 1 and 1 to 5 (never needed/never to very often) for questions 2 to 4. ⁴⁵ , ⁴⁶ , ⁴⁷

Statistical Analysis

Descriptive statistics were used to depict participant characteristics. We used simple and multiple linear regression to examine the association between 10‐unit increments of Overall SWS endorsement (10 units were chosen because of the wide range of possible scores on the 0–105‐point scale) and 1‐unit increments of SWS‐Help endorsement (range, 0–3) and central hemodynamic parameters (central systolic blood pressure, augmentation index, and pulse wave velocity). Six models were constructed for each exposure‐outcome analysis. Model 1 was the base age‐adjusted model. Model 2 additionally adjusted for socioeconomic factors (income and education) and, for augmentation index models only, heart rate and height. ⁴⁸ Model 3 includes all factors in Model 2 plus clinical CVD risk factors (peripheral systolic blood pressure and diabetes) and antihypertensive medication use. Model 4 adds body mass index alone. Model 5 additionally adjusts for smoking as a behavioral risk factor. Model 6 is the fully adjusted model; it includes all factors in model 5 plus psychosocial factors (severity of depressive symptoms, caregiving, and social support). We additionally conducted exploratory regression analyses to examine associations among the 4 remaining SWS subscales and central hemodynamic variables. Fully adjusted models for central systolic blood pressure and augmentation index had an analytic sample of 386 due to 36 participants missing data on SWS, central hemodynamics, or any covariates, representing <10% missingness. For pulse wave velocity, the smaller sample size of 347 in the fully adjusted model is attributable to some participants having a thigh circumference exceeding that of the thigh cuff, rendering such measurements unattainable. To better understand associations among central systolic blood pressure, peripheral systolic blood pressure, SWS, and covariates traditionally associated with CVD and blood pressure, we conducted exploratory analyses using multiple linear regression and Pearson/Spearman correlations. All data were processed using SAS version 9.4 (SAS Institute; Cary, NC) and GraphPad Prism version 9.5.1 (GraphPad Software LLC, San Diego, CA). The level for statistical significance was set a priori at α=0.05.

Ethical Approval

Ethical approval was obtained from the Emory University Institutional Review Board (#IRB00085245), and the data supporting the findings of this study will be made available upon participant consent. All participants were informed of the procedures and risks before participating. Verbal and written informed consent were obtained from each participant.

Results

Participant Characteristics

Participant characteristics are provided in Table 1. The average age of the participants was 38±4 years (range, 30–46 years), and a wide range of sociodemographic backgrounds were represented. Seventeen percent of the participants reported antihypertensive medication use, 10% smoked tobacco, and 4% had diabetes. The average resting heart rate and peripheral blood pressure values were within accepted normal limits, but substantial variability in both was noted. Scores on the Beck Depression Inventory ranged from 0 to 39, averaging 6.0±6.9. Overall SWS and SWS‐Help endorsement scores were 67±18 and 1.71±0.68, respectively, and were generally similar to those observed in prior studies. ³⁰ , ³¹ , ⁴⁹ , ⁵⁰ Central systolic blood pressure and augmentation index measurements were successfully obtained in 408 participants. The average central systolic blood pressure was 117±15 mm Hg, which is 6 mm Hg higher than the age–sex reference value. ⁹ The average augmentation index of 21%±11% fell within the typical range for the age and sex of the participants. ⁴⁸ Pulse wave velocity measurements were successfully obtained in 368 participants. The smaller number for these measurements was attributable to the aforementioned cuff circumference limitation. The average pulse wave velocity was 6.2±1.5 m/s, which falls within the normal range for this population. ⁵¹ In general, Overall SWS and SWS‐Help endorsement were in line with previous reports, ⁴⁹ , ⁵² and the remaining SWS subscale results are provided in Table S1.

Table 1.

Participant Characteristics

Characteristic	Mean±SD or %	No.
Sociodemographic factors
Age, y	38±4	422
Income
<US$35 000	25	412
US$35 000–US$49 999	21	412
US$50 000–US$74 999	23	412
≥US$75 000	31	412
Education, y	15±2	416
Health risk factors and medication use
Resting heart rate, bpm	65±10	411
Resting SBP, mm Hg	119±15	421
Resting DBP, mm Hg	81±12	421
Body mass index, kg/m2	32.8±8.3	421
Current cigarette smoker	10	421
Diabetes	4	422
Beck Depression Inventory (score; range, 0–60)	6.0±6.9	420
Antihypertensives	17	422
General caregiving and social support
Caring for family (score; range, 1–4)	3.4±0.8	422
Caring for friends (score; range, 1–5)	3.8±1.0	419
Support from family (score; range, 1–5)	3.8±1.0	420
Support from friends (score; range, 1–5)	3.5±1.0	420
Superwoman Schema
Overall SWS endorsement (score; range, 0–105)	67±18	419
SWS‐Help endorsement (score; range, 0–3)	1.71±0.68	419
Central hemodynamics
Central SBP, mm Hg	117±15	408
Augmentation index, %	21±11	408
Pulse wave velocity, m/s	6.2±1.5	368

bpm indicates beats per minute; DBP, diastolic blood pressure; SBP, systolic blood pressure; SWS, Superwoman Schema; and SWS‐Help, SWS–Obligation to Help Others.

SWS Endorsement and Central Systolic Blood Pressure

As shown in Table 2, in the age‐adjusted models, central systolic blood pressure was significantly associated with Overall SWS (β=1.01 [95% CI, 0.17–1.86]) and SWS‐Help (β=3.52 [95% CI, 1.37–5.67]). These associations were modeled continuously, but are illustrated in Figure [A] (Overall SWS) and Figure [B] (SWS‐Help) on the basis of the following established SWS endorsement scoring categories ³⁰ , ³² : Overall SWS: low (0–35), moderate (36–70), and high (71–105); SWS‐Help: low (0–9), moderate (10–18), and high (19–27). These cut points were used for descriptive purposes only to characterize the nature of the association. Both Figure [A] and [B] demonstrate an approximate dose–response association, where higher reports of Overall SWS and SWS‐Help were associated with greater central systolic blood pressure. In the fully adjusted models, central systolic blood pressure remained significantly associated with both Overall SWS (β=0.83 [95% CI, 0.19–1.47]) and SWS‐Help (β=2.03 [95% CI, 0.39–3.67]) (Table 2).

Table 2.

SWS Endorsement and Central Systolic Blood Pressure in MUSE Study Participants

	β	95% CI	P value	No.
Overall SWS
Model 1	1.01*	0.17–1.86	0.019	405
Model 2	0.91*	0.04–1.77	0.040	392
Model 3	0.85*	0.21–1.49	0.010	392
Model 4	0.73*	0.16–1.30	0.012	391
Model 5	0.67*	0.10–1.25	0.022	390
Model 6	0.83*	0.19–1.47	0.011	386
SWS‐Help
Model 1	3.52†	1.37–5.67	0.001	405
Model 2	3.14†	0.93–5.35	0.006	392
Model 3	2.65†	0.99–4.30	0.002	392
Model 4	1.78*	0.29–3.26	0.019	391
Model 5	1.69*	0.20–3.18	0.026	390
Model 6	2.03*	0.39–3.67	0.015	386

The β coefficient represents the change in central systolic blood pressure (mm Hg) associated with a 10‐unit (Overall SWS) or 1‐unit (SWS‐Help) increase in SWS endorsement score. Model 1: adjusted for age. Model 2: adjusted for model 1+education and income. Model 3: adjusted for model 2+peripheral systolic blood pressure, diabetes, and antihypertensive medications. Model 4: adjusted for model 3+body mass index. Model 5: adjusted for model 4+smoking. Model 6: adjusted for model 5+severity of depressive symptoms, caregiving, and social support. MUSE indicates, Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health Study; SWS, Superwoman Schema; and SWS‐Help, SWS–Obligation to Help Others.

^* P<0.05.

^† P<0.01.

Figure 1. SWS endorsement, central systolic blood pressure, and augmentation index in MUSE study participants.

A, Overall SWS endorsement and central systolic blood pressure. B, SWS‐Help endorsement and central systolic blood pressure. C, Overall SWS endorsement and augmentation index. D, SWS‐Help endorsement and augmentation index. Data are presented as age‐adjusted means ±95% CI for illustrative purposes only as a complement to the primary regression results in Tables 2 and 3. N=405 for all models, which were analyzed via multiple linear regression and pairwise comparisons. *P<0.05. **P<0.01. MUSE indicates Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health Study; SWS, Superwoman Schema; and SWS‐Help, SWS–Obligation to Help Others.

SWS Endorsement and Augmentation Index

In age‐adjusted models, augmentation index was associated with Overall SWS (β=0.58 [95% CI, 0.00–1.16]) and SWS‐Help (β=2.02 [95% CI, 0.54–3.50]) (Table 3). Figure [C] (Overall SWS) and Figure [D] (SWS‐Help) illustrate these findings using the categorical SWS endorsement levels and depict an approximate linear association between SWS endorsement and augmentation index. In fully adjusted models, augmentation index was associated with Overall SWS (β=0.66 [95% CI, 0.02–1.30]) and SWS‐Help (β=2.21 [95% CI, 0.58–3.84]) (Table 3).

Table 3.

SWS Endorsement and Augmentation Index in MUSE Study Participants

	β	95% CI	P value	No.
Overall SWS
Model 1	0.58*	0.00 to 1.16	0.049	405
Model 2	0.58	−0.03 to 1.18	0.061	392
Model 3	0.62*	0.03 to 1.21	0.038	392
Model 4	0.60*	0.03 to 1.18	0.040	391
Model 5	0.52	−0.06 to 1.10	0.077	390
Model 6	0.66*	0.02 to 1.30	0.045	386
SWS‐Help
Model 1	2.02†	0.54 to 3.50	0.008	405
Model 2	2.07†	0.52 to 3.61	0.009	392
Model 3	2.08†	0.57 to 3.59	0.007	392
Model 4	1.79*	0.30 to 3.27	0.018	391
Model 5	1.66*	0.18 to 3.14	0.028	390
Model 6	2.21†	0.58 to 3.84	0.008	386

The β coefficient represents the change in augmentation index (%) associated with a 10‐unit (Overall SWS) or 1‐unit (SWS‐Help) increase in SWS endorsement score. Model 1: adjusted for age. Model 2: adjusted for model 1+education, income, heart rate, and height. Model 3: adjusted for model 2+peripheral systolic blood pressure, diabetes, and antihypertensive medications. Model 4: adjusted for model 3+body mass index. Model 5: adjusted for model 4+smoking. Model 6: adjusted for model 5+severity of depressive symptoms, caregiving, and social support. MUSE indicates Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health Study; SWS, Superwoman Schema; and SWS‐Help, SWS–Obligation to Help Others.

^* P<0.05.

^† P<0.01.

SWS Endorsement and Pulse Wave Velocity

Overall SWS and SWS‐Help endorsement were not associated with pulse wave velocity (Table 4). In the age‐adjusted model, Overall SWS tended to be associated with pulse wave velocity (β=0.08 [95% CI, −0.01 to 0.17]); however, this result and the remaining SWS‐pulse wave velocity estimates were not statistically significant following adjustment for sociodemographic and clinical CVD risk factors.

Table 4.

SWS Endorsement and Pulse Wave Velocity in MUSE Study Participants

	β	95% CI	P value	No.
Overall SWS
Model 1	0.08	−0.01 to 0.17	0.084	365
Model 2	0.06	−0.03 to 0.15	0.215	353
Model 3	0.04	−0.05 to 0.12	0.409	353
Model 4	0.03	−0.05 to 0.12	0.401	352
Model 5	0.04	−0.05 to 0.12	0.403	351
Model 6	0.01	−0.08 to 0.10	0.851	347
SWS‐Help
Model 1	0.10	−0.13 to 0.33	0.377	365
Model 2	0.04	−0.19 to 0.28	0.714	353
Model 3	−0.02	−0.24 to 0.20	0.862	353
Model 4	−0.05	−0.26 to 0.16	0.631	352
Model 5	−0.05	−0.27 to 0.16	0.619	351
Model 6	−0.13	−0.37 to 0.10	0.267	347

The β coefficient represents the change in pulse wave velocity (m/s) associated with a 10‐unit (Overall SWS) or 1‐unit (SWS‐Help) increase in SWS endorsement score. Model 1: adjusted for age. Model 2: adjusted for model 1+education and income. Model 3: adjusted for model 2+peripheral systolic blood pressure, diabetes, and antihypertensive medications. Model 4: adjusted for model 3+body mass index. Model 5: adjusted for model 4+smoking. Model 6: adjusted for model 5+severity of depressive symptoms, caregiving, and social support. MUSE indicates Mechanisms Underlying the Impact of Stress and Emotions on African American Women's Health Study; SWS, Superwoman Schema; and SWS‐Help, SWS–Obligation to Help Others.

Exploratory Analysis

The results from the exploratory subscale analyses are provided in Tables S2 through S4. In general, these results were mixed and inconclusive. For example, central systolic blood pressure was associated with SWS–Obligation to Suppress Emotions but none of the remaining subscales (Table S2). Augmentation index was not associated with any of the subscales in the fully adjusted exploratory analyses (Table S3). Additionally, although pulse wave velocity was associated with SWS–Intense Motivation to Succeed in the demographic‐adjusted model, the association was no longer significant in the fully adjusted model, and pulse wave velocity was not associated with any other SWS subscales in any of the models (Table S4).

There was a strong correlation between central and peripheral systolic blood pressure (r=0.97; P<0.0001). In demographic‐adjusted models, associations between SWS‐Help and peripheral systolic blood pressure are similar, albeit slightly attenuated relative to the central systolic blood pressure findings (β=2.91 [95% CI, 0.56–5.27] versus β=3.14 [95% CI, 0.93–5.35]). The correlations between 7 of 8 covariates and peripheral systolic blood pressure tended to be lower than those of central systolic blood pressure (Table 5).

Table 5.

Correlations Between Central or Peripheral Systolic Blood Pressure and Key Study Covariates

	Central systolic blood pressure r or r s (P value)	Peripheral systolic blood pressure r or r s (P value)
Age, y*	0.142 (0.004)	0.140 (0.005)†
Education*	−0.094 (0.059)	−0.094 (0.061)†
Income‡	−0.055 (0.272)	−0.047 (0.349)†
Body mass index*	0.472 (<0.001)	0.448 (<0.001)†
Smoking‡	0.079 (0.110)	0.057 (0.248)†
Antihypertensives‡	0.344 (<0.001)	0.335 (<0.001)†
Diabetes‡	0.139 (0.005)	0.150 (0.002)
Depressive symptoms*	0.088 (0.076)	0.083 (0.096)†

r indicates Pearson correlation coefficient; and r _s, Spearman correlation coefficient.

^*Indicates use of Pearson correlation.

^† Values indicate instances where the correlation was lower for peripheral relative to central systolic blood pressure.

^‡ Indicates use of Spearman correlation.

Discussion

In this study, we found that, among African American women, greater endorsement of the SWS psychological framework and its Obligation to Help Others dimension were significantly associated with higher central systolic blood pressure and aortic augmentation index but not aortic pulse wave velocity. Accordingly, our data support the idea that African American women who feel a strong obligation to maintain the Superwoman role and care for others before themselves may be more likely to have a deleterious central hemodynamics profile, thereby increasing their risk of developing overt CVD.

Our strongest and most consistent findings were for central systolic blood pressure. Both Overall SWS and SWS‐Help endorsement were associated with elevated central systolic blood pressure, even after controlling for traditional risk factors, including peripheral blood pressure. In fully adjusted models, each increase in the level of SWS‐Help endorsement was associated with a 2‐mm Hg increase in central systolic blood pressure, a value large enough to move someone from having a normal, healthy central systolic blood pressure (<112 mm Hg) to having a value associated with an increased risk of major adverse cardiovascular events (>112 mm Hg). ⁵³ This is physiologically significant because central blood pressure is believed to most closely reflect the blood pressure impacting target organs, such as the brain, heart, and kidneys. ⁹ Indeed, our exploratory analyses suggest that (1) there may be a stronger association between central (relative to peripheral) systolic blood pressure and SWS and (2) relative to peripheral systolic blood pressure, central systolic blood pressure exhibits stronger associations with traditional CVD risk factors (eg, body mass index, smoking). Consequently, central systolic blood pressure could be an important early marker of changes in cardiovascular health in African American women, particularly those more inclined to take on the Superwoman role and prioritize caring for others over self‐care. In 1 of the few studies to examine associations between SWS endorsement and cardiovascular health in African American women, Perez et al ³⁵ found that African American women who reported greater SWS‐Help endorsement were more likely to have elevated or high blood pressure, which coincides with our findings.

Findings linking Overall SWS and SWS‐Help to the augmentation index were less robust, albeit significant for most models. In the fully adjusted model, each increase in the level of SWS‐Help endorsement was associated with a 2% increase in augmentation index, a value large enough to increase one's risk of premature coronary artery disease. ³ We did not observe significant associations for SWS and pulse wave velocity. This may be due to the age of the women in our cohort. While it has been demonstrated that the augmentation index increases linearly with age early in life and then plateaus in late life, the opposite is true for pulse wave velocity, indicating that the augmentation index may be a more sensitive marker of arterial stiffening in younger individuals. ⁵⁴ Therefore, given that the women in our cohort were primarily in their late 30s, it is possible that greater endorsement of Overall SWS and SWS‐Help may promote telltale signs of early vascular aging (ie, increased augmentation index) in this population.

Another study that investigated SWS and cardiovascular health found that brachial artery flow–mediated dilation in young (≈20 years), otherwise healthy Black women was associated with SWS–Intense Motivation to Succeed and SWS–Resistance to Being Vulnerable. ³⁴ The exploratory analyses in the present study did not demonstrate associations between either of these subscales and central hemodynamics. The differential findings between the specific SWS dimensions and cardiovascular parameters between these 2 studies could be attributable to the ≈20‐year age difference between the 2 cohorts (ie, SWS endorsement and its potential effects on health may vary with age), but this remains to be tested directly. Further, as shown in Tables S2 through S4, we did not observe any consistent associations between central hemodynamics and the 4 remaining subscales. Additional studies, using hypothesis‐driven approaches, could be used to elucidate whether these SWS dimensions might be impacting cardiovascular health and function.

Although the precise mechanisms linking greater Overall SWS and SWS‐Help endorsement to higher central systolic blood pressure and augmentation index in African American women remain to be investigated directly, allostatic overload and detrimental health behaviors likely contribute. SWS endorsement promotes stress embodiment and is often accompanied by detrimental stress‐related health behaviors (eg, overeating) and delays in seeking health care. ³⁰ Combined, the liabilities of endorsing the SWS framework may take a toll on mental and physical health. The constant psychological stress that accompanies caring for others (parents, children, coworkers, etc) in lieu of self‐care may essentially get “under the skin” and drive physiological changes such as increased sympathetic nervous system activity and inflammation. ²¹ , ⁵⁵ When these systems are upregulated for sustained periods (ie, allostatic overload), they can wreak havoc throughout the body, which may include damages to the heart and vasculature, ²¹ , ²² including altered central hemodynamics as our data would suggest. Additional studies will be required to further elucidate the psychological and biological mediators and moderators of the SWS–cardiovascular health relationships observed in the present study. Since structural gendered racism is most likely the primary source of African American women's propensity to adopt the SWS framework, it should continue to be the intervention point for health equity for this population. ¹⁷ , ³⁰ , ³² , ⁴⁹ , ⁵⁶ However, interim strategies for mitigating any negative effects of high SWS endorsement among African American women should also be pursued. ⁵⁷ As the relevant literature suggests, these interventions could come in the form of mindfulness‐based stress reduction ⁵⁸ , ⁵⁹ or self‐compassion, ⁶⁰ , ⁶¹ among others. ⁵⁸ , ⁵⁹ , ⁶² , ⁶³ , ⁶⁴

This study is strengthened by several factors. By examining the impact of SWS on cardiovascular health, we centered African American women's experience as opposed to using more general psychosocial stress measures (eg, interpersonal racism). Additionally, our most pertinent results remained significant when adjusting for measures of caregiving and social support. This may make the findings more applicable to African American women specifically and shines a light on their unique experiences, challenges, and coping strategies. Our work is further strengthened by the socioeconomic diversity of the cohort, which helps ensure that the findings are representative of African American women from diverse backgrounds. However, this study also has several limitations worth discussing. First, the study design was cross‐sectional. Future work will need to examine the observed relationships in a longitudinal fashion to bolster confidence regarding SWS–cardiovascular health associations. The lack of association between SWS parameters and pulse wave velocity could be attributable to the missing data due to the thigh cuff circumference limitations discussed previously. In other words, African American women with a larger thigh circumference and high SWS endorsement may have greater pulse wave velocity, but we could not collect these data. Alternative means (eg, methods that do not use cuffs) for readily assessing carotid‐to‐femoral pulse wave velocity in people with larger thigh circumferences should be pursued. The Giscombe Superwoman Schema Questionnaire was specifically developed for and validated in African American women and speaks to their unique lived experiences. ³⁰ , ³² , ⁵² Accordingly, it is unknown whether the Giscombe Superwoman Schema Questionnaire or another similar questionnaire/construct would produce findings similar to those of the present study among women from other races and ethnicities. Nevertheless, studies should be designed to specifically test which psychosocial stress factors/coping strategies might be beneficial or detrimental to cardiovascular health in women of various racial and ethnic backgrounds. Finally, our findings are only representative of African American women in the metropolitan Atlanta area, and future work in a similar vein should incorporate multisite trials to improve nationwide representativeness.

In summary, this is the first investigation of SWS endorsement and central hemodynamics in African American women, a group disproportionately affected by hypertension, CVD, and stroke. We found that greater Overall SWS and SWS‐Help endorsement were associated with higher central systolic blood pressure and augmentation index but not pulse wave velocity. Greater SWS endorsement, a product of structural and interpersonal gendered racism, ¹⁷ , ³⁰ , ³² , ⁴⁹ may contribute to increased central systolic blood pressure and pulse wave reflections, both of which are known to increase the risk of CVD and stroke. ³ , ⁴ , ⁶ , ⁸ , ¹⁰ Researchers, clinicians, and policymakers should work together to eliminate gendered racism in society and improve methods for managing psychosocial stress to reduce CVD and stroke risk in African American women.

Sources of Funding

The MUSE study is funded by National Institutes of Health Grants R01 HL130471 and R01 HL158141 to TTL. Dr Lewis received additional funding from K24 HL163696. Drs Martin and Fields are supported by National Institutes of Health T32 HL130025 to Dr Vaccarino. Dr Erving is partially supported by P2CHD042849 awarded to the Population Research Center at the University of Texas at Austin by the Eunice Kennedy Shriver National Institute of Child Health and Human Development.

Disclosures

None.

Supporting information

Tables S1–S4