Racism, shame, and stress reactivity among young black women

Abstract

Black women experience disproportional rates of cardiovascular disease (CVD) warranting further exploration of CVD risk factors. Growing evidence suggests acute stress reactivity studies may elucidate the mechanisms driving psychosocial correlates of CVD risks. Race-related stress has been identified as a CVD risk factor among Black women though recent evidence suggests emotions may facilitate these risks. Black women may be vulnerable to shame related to frequent racist experiences. Yet, no study to date has examined racism, shame, and stress reactivity in this population. The current study utilized mixed linear models to test for time and group effects of racism and shame on stress reactivity (e.g., cortisol and C-reactive protein [CRP]) among 34 Black women who completed the Trier Social Stress Test. Tests for two-way interactions (i.e., shame by racism) were also performed. Significant time and group effects were observed for shame and racism on stress reactivity. Black women who experienced greater lifetime racism, stress appraised, but lower racism during the past year, exhibited greater CRP responses. Black women who experienced high levels of shame and racism during the past year and their lifetime demonstrated greater cortisol reactivity. These results prompt further research on racism and shame as CVD risk factors among Black women.

1 |. INTRODUCTION

Cardiovascular diseases (CVD), including coronary heart disease, heart failure, and hypertension, are the leading cause of death among Black women 18 years and older in the United States (Virani et al., 2021). Recent trends suggest Black women continue to exhibit excess CVD morbidity despite ongoing efforts to reduce CVD risk factors within the population (Cunningham et al., 2017; Kalinowski et al., 2019; Virani et al., 2021). Moreover, Black women develop CVD earlier in life with prominent CVD disparities observed among Black women aged 35–54. These trends warrant further exploration of risk factors contributing to CVD risk among Black women.

A breadth of studies have linked racism, both personally and institutionally mediated, to CVD risk (Brondolo et al., 2003; Lewis et al., 2014; Sims et al., 2012). For example, there is compelling evidence that racism increases the risk of hypertension, poor cardiovascular health, and health risk behaviours (Calvin et al., 2003; Panza et al., 2019; Sims et al., 2012). Yet, the underlying mechanisms that link racism to physical outcomes are not clearly understood. One prominent explanation points to the long-term physiological consequences of racism-related stress. Racism-related stress contributes to physiological wear and tear through the activation of the hypothalamic-pituitary-adrenal (HPA) axis, sympathetic-adrenal-medullary (SAM) axis, and immune systems (Geronimus, 1991; McEwen, 1998). Allostasis, or the regulation of these stress response systems, prepares the body to meet internal and external demands to achieve optimal physiological function during stress (McEwen, 1998; Sterling, 2012; Sterling & Eyer, 1988). Regulation of these systems is predicated on past experiences which enables the brain to predict the demand necessary to maintain physiological balance amidst future stressors. These adaptive coordinated stress responses are designed to optimize performance with minimal physiological costs (Sterling, 2012).

Racism-related stress has been theorized as a chronic stressor due to the systemic nature of racism that can manifest via interpersonal interactions, structural challenges, and institutional policies (Brody et al., 2014; Geronimus, 1991). Chronic stress promotes sustained demand that alters physiological “set-points” to provide levels of response that satisfy the most common level of demand. For example, persistent immune activity can promote sustained cytokine levels in response to anticipated demand increasing the risk of inflammatory disorders. Progressive physiological wear and tear from chronic stress-induced arousal is considered allostatic load (Cunningham et al., 2017; Gianaros & Jennings, 2018; Kalinowski et al., 2019; McEwen & Seeman, 1999; Schulkin & Sterling, 2019). Allostatic load has been theoretically and empirically linked to CVD risks (Borrell et al., 2020; Gianaros & Jennings, 2018; Seeman et al., 1997). Moreover, Black women have exhibited allostatic load at younger ages compared to women of racial minority groups (Geronimus et al., 2006, 2010). This phenomenon has been explained, in part, by the disproportional burden of race-related stress among Black women (Chyu & Upchurch, 2011; Geronimus et al., 2006).

Recent studies have focussed on acute stress reactivity to understand how race-related stress contributes to allostatic load and subsequent CVD risk. Acute stress responses are instrumental in the development of cardiovascular risk status (Chida & Steptoe, 2010). Stress reactivity studies among Black populations have demonstrated that racism has been associated with greater blood pressure reactivity (Clark & Adams, 2004), increased inflammatory activity (Lewis et al., 2010; Lucas et al., 2017), and blunted cortisol responses (Richman & Jonassaint, 2008; Urizar et al., 2019). Emotional regions in the brain have been shown to promote the coordination of these physiological stress responses, including activation of the HPA axis and sympathetic nervous system, to accommodate acute stressors (Panksepp, 1998). Several studies have specifically examined the mediating role of negative emotion in the relationship between racism and acute stress reactivity. Cheadle et al. (2020) examined racism-related experiences, negative emotion, and electrodermal activity via changes in sweat excretions to detect sympathetic nervous system arousal in a diverse sample of young adults. The authors found that racism-related experiences predicted both negative emotions and increased sympathetic nervous system activity (Cheadle et al., 2020). A second study by Hittner and Adam (2020) differentiated negative emotion versus negative social evaluative emotion (e.g., embarrassment, rejection, shame) to examine how emotions mediate the association between discrimination and diurnal cortisol. This study revealed negative social evaluative emotions relative to negative emotions were a stronger mediator of the associations between discrimination and diurnal cortisol (Hittner & Adam, 2020). Taken together, these findings unveiled understudied emotional mechanisms by which Black women may physiologically embody racism.

Frequent experiences of racism have been shown to predict internalized shame in Black populations (Johnson, 2020). Shame is an emotional response to perceived threats to social esteem, social status, and acceptance (Gruenewald et al., 2004; Lewis, 1971). Perceived threats occur in an array of social situations often stemming from circumstances where individuals feel unworthy of acceptance (Dickerson, Gruenewald, et al., 2004). The negative consequences of such threats are exacerbated when these experiences are perceived as uncontrollable and/or associated with unwanted identities within social spaces (Dickerson, Kemeny, et al., 2004; Ferguson & Eyre, 2000; Gruenewald et al., 2004).

Shame may provide an important focal point of study as theorists have posited shame as a facilitator of other emotions such as anger or sadness (Ferguson & Eyre, 2000; Hejdenberg & Andrews, 2011; Lewis, 1971). For example, Ferguson and Eyre (2000) revealed unwanted identities elicited shame which, in turn, instigated a strong anger response. Black women may be vulnerable to shame via social threats stemming from racialized stereotypes that depict Black women as, for example, angry or immoral (Harris-Perry, 2011). These stereotypes ascribe unwanted identities and attributes that are often compounded by societal regulations of Black women’s hair, sexuality, and bodies increasing the frequency of shame experiences (Ashley, 2014; Ellis-Hervey et al., 2016; Harris-Perry, 2011). These experiences are systemic for Black women as these social threats are rooted in societal beliefs and enforced through interpersonal interactions, mass media, and institutional practices (Harris-Perry, 2011; Johnson, 2020). Long term exposure to shame experiences can promote internalized shame contributing to feelings of inferiority and inadequacy (Cook, 1988; Rybak & Brown, 1996).

Shame has been strongly associated with an increased biological stress response, demonstrated by increased levels of cortisol and inflammatory markers in stress reactivity studies though these studies are limited among Black populations (Dickerson, Kemeny, et al., 2004; Kemeny et al., 2004). A recent study by Johnson and Urizar (2021) examined blood pressure responses to social stress among young Black women who reported high versus low shame. This study revealed that Black women who reported high shame exhibited significantly prolonged diastolic blood pressure reactivity and poor diastolic blood pressure recovery compared to Black women with low shame (Johnson & Urizar, 2021). This suggests shame may be associated with prolonged SNS activity responses to stress in this population. To date, little is known about the collective impact of race-related stress and shame on stress reactivity among Black women.

C-reactive protein (CRP) and cortisol are prominently investigated biomarkers in the investigation of allostatic load. Moreover, cortisol, an indicator of HPA axis activity, and C-reactive protein (CRP), an inflammatory marker, are two important biological indicators of CVD risk (Clearfield, 2005; Danesh et al., 2004; Kirschbaum & Hellhammer, 1989). A large number of stress reactivity studies have included cortisol as a measure of stress reactivity (Gruenewald et al., 2004; Kirschbaum & Hellhammer, 1989; Knight et al., 2021; Urizar et al., 2019). Stress reactivity studies among Black women have shown blunted cortisol responses to stress suggesting evidence of physiological wearing (Chong et al., 2008; Richman & Jonassaint, 2008; Urizar et al., 2019). Chronic inflammation has been hypothesized as one mechanism for diminished HPA axis responses such that decreased levels of cortisol reactivity are associated with increased inflammatory activity (Knight et al., 2021; Nijm & Jonasson, 2009).

Exploration of CRP responses to acute stress has been limited. This is possibly explained by the requirement of stress studies to incorporate venipuncture to measure CRP in venous blood and its potential contribution to acute stress (Girgis et al., 1988; McDade et al., 2007). However, noninvasive measures such as salivary CRP have shown promise in understanding the role of CRP in acute stress. Several studies have shown that CRP is reactive to acute stress (Marsland et al., 2017; Steptoe et al., 2007). However, these studies show a late phase CRP response with increased levels occurring during stress recovery (Lucas et al., 2017, 2016; Marsland et al., 2017; Steptoe et al., 2007). Specifically, Lucas et al. (2016) and (2017) both explore CRP responses to acute stress among Black samples. These studies revealed CRP responses to acute stress were related to self-beliefs (e.g., racial identity) and world beliefs (inconsistency between beliefs and experiences) within these samples. Thus, there is evidence suggesting cortisol and CRP may be useful for elucidating mechanisms of adaptive physiological reactivity given frequent experiences of race-related stress and emotional states.

The current study sought to examine the effect of perceived racism and internalized shame on cortisol and CRP reactivity to social stress among young Black women. Therefore, this study hypothesised that Black women who report a higher frequency of racist events will exhibit blunted cortisol responses and heightened CRP reactivity when exposed to a laboratory stressor (Trier Social Stress Test [TSST]). It was further hypothesized that the effect of racism on cortisol and CRP reactivity would be moderated by internalized shame such that those reporting higher levels of shame will experience lower cortisol reactivity but greater CRP reactivity than those with low internalized shame. Findings from the study may further implicate shame as a key emotion that may facilitate CVD risk when Black women experience racism.

2 |. METHODS

2.1 |. Participants

The present study recruited 34 Black women using both an undergraduate psychology participant pool from a Southern California university and flyer distribution (print and electronic) within the local university area. Recruitment materials included information on the study purpose, study eligibility, and research protocol. To be eligible for the study, participants were required to be at least 18 years old and self-identify as African American/Black. Exclusion criteria were developed to account for factors (e.g., smoking status, pregnancy, mental state) that have been shown to reduce the validity of stress reactivity measurements (e.g., cortisol) (Linares et al., 2020). These criteria excluded individuals who smoked (marijuana/tobacco) within the past 30 days, on antidepressants, or were pregnant. Individuals were also excluded if they had a previous diagnosis of cardiovascular disease, diabetes, or mental illness (e.g., bipolar disorders, clinical depression).

2.2 |. Procedures

Eligible participants were provided an informed consent form before their study participation. Once informed consent was obtained, participants were directed to a questionnaire that assessed the frequency of racist experiences, shame, perceived stress, and self-esteem. Participants were compensated with a $10 gift card after completing the electronic questionnaire. After completion of the questionnaire (approximately 15 min), participants’ height and weight were collected. Participants were then asked to submit a saliva sample using the passive drool method (Salimetrics & SalivaBio, 2011). Participants were then escorted to a second room to complete the TSST, a two-hour laboratory highly standardized stress reactivity protocol (Kirschbaum et al., 1993) shown to reliably induce cortisol and CRP changes in ethnically diverse samples (Lucas et al., 2016; Urizar et al., 2019). The TSST protocol incorporated two stress reactivity tasks in front of two judges (i.e., research assistants) who videotaped their performance. These tasks, lasting a total of 15 min, consisted of public speaking simulated job interview and a mental arithmetic task (serial subtraction). Participants exited the room and were asked to provide a saliva sample immediately upon exiting. Four additional saliva samples were taken at 15, 30, 45, and 60 min after the TSST during a 60-min recovery period. Participants were debriefed during the recovery period and compensated with a $25 gift card. All study procedures were approved by the university’s Institutional Review Board (IRB) and all procedures were followed in accordance with the ethical standards of the IRB.

2.3 |. Measures

2.3.1 |. Demographic questionnaire

Participants completed a demographic questionnaire that assessed ethnicity (e.g., African American/Black, African, Afro-Latino), age, and sexual orientation among the sample.

2.4 |. Racism

The SRE is an 18-item self-report questionnaire that examines the frequency of racist discrimination in the past year, one’s entire life, and the stress appraised from racist events (Landrine & Klonoff, 1996). The SRE is a validated measure that assesses event based acute forms of racism. Moreover, this scale specifically assesses stress appraised from racist experiences. The authors believed that variations in stress appraised from racist experiences may contribute to varying stress responses, motivating the use of the SRE. The frequency of racist events was reported using a 6-point rating scale ranging from 1 (this has never happened to you) to 6 (this has happened almost all of the time).

Average scores for frequency of racist events during the past year, over one’s lifetime, and stress appraised for racist events were measured with higher scores representing greater experiences of racism and race-related stress. The SRE has shown moderate to high reliability in Black populations for past year racist events (α= 0.88–0.95) entire life racist events (α = 0.88–0.95), and stress appraisal (α= 0.93) (Johnson, 2020; Landrine & Klonoff, 1996).

2.5 |. Shame

The ISS is a 30 item self-report questionnaire in which 24 questions comprise a scale to measure the extent to which shame has been internalized as a consequence of long term shame exposure (Cook, 1988). Examples of questions used on the ISS include “have an overpowering dread that my faults will be revealed in front of others” and “I feel somehow left out”. Participants rated their level of agreement using a 5-point rating scale (1 = Never to 5 = Almost Always). The ISS has shown high reliability among Black populations (α = 0.94 −0.96) and correlation with perceived racist events (Johnson, 2020).

2.6 |. Salivary cortisol and CRP (stress reactivity)

Six total salivary samples were collected throughout the study, including one sample before (baseline) and immediately after the TSST and 15, 30, 45, and 60 min after the TSST. Participants provided these samples using a passive drool method by allowing saliva to pool in their mouths. Participants were then asked to hold a saliva collection aid (similar to a straw) inside a collection tube using both hands (Salimetrics & SalivaBio, 2011). They were instructed to tilt their head forward and carefully spit into the collection tube. They were asked to repeat this procedure until they provided a minimum of 1.8 ml of saliva. Salivary cortisol samples were analysed via ELISA immunoassay (ELISA, IBL-America, Minneapolis, MN). Intra- (4.6%) and inter-assay (6.0%) variability were both under 10%. Cortisol values were then logarithmically transformed (base 10, converted from μg/dl to nmol/l) given that salivary cortisol values are typically skewed, with higher values reflecting greater cortisol levels. Salivary CRP was analysed via high sensitivity enzyme immunoassay (Salimetrics Salivary C-reactive Protein Assay Kit, Cat. No. 1–2102) in which Intra-(3.2%) and inter-assay (2.6%) variability were also under 10%. Like cortisol, CRP was logarithmically transformed (base 10) as CRP exhibited a positive skew.

2.7 |. Covariates

2.7.1 |. Body mass index

BMI has been shown to influence both cortisol and CRP (de Punder et al., 2019). Therefore, BMI was included as a covariate in the current study. Bodyweight and standing height without shoes using a Dectecto Physicians’ Scale was assessed for each participant. BMI was calculated as weight (kg) divided by height squared (m²).

2.8 |. Self-esteem

Self-esteem was included in a covariate in this study due to previous literature suggesting a significant association with racism and shame in Black populations (Johnson, 2020). Self-esteem was measured using six questions from the ISS’s self-esteem subscale (1 = Never to 5 = Almost Always). Scores above 18 indicate higher self-esteem (Cook, 1988). The measure has demonstrated good reliability among Black samples (α = 0.82).

2.9 |. Perceived stress scale (PSS)

PSS was included as a covariate to account for the potential impact of perceived stress on stress reactivity (Johnson & Urizar, 2021). Participants perceived stress at the time of the study was measured using the 14 questions self-report PSS (Cohen et al., 1983). The PSS measures participants’ degree of stressfulness from daily life situations by rating the frequency of experiencing seven positively worded and seven negatively worded stress items over the past 4 weeks (0 = never, 1 = almost never, 2 = sometimes, 3 = fairly often, 4 = very often). Positively worded stress questions were reversed coded such that lower scores indicated higher perceived stress. This scale demonstrated very good reliability among Black populations (α = 0.95) (Johnson & Urizar, 2021; Mendez et al., 2013).

2.10 |. Statistical analysis

Independent samples t-tests and Pearson’s Chi-squared were performed to examine between group differences (low vs. high) for past year racism, entire life racism, stress appraised from racism, and shame by participants characteristics for continuous and categorical variables, respectively. A mixed effect linear model was used to test for possible time and group effects of shame and racism (past year, entire life, stress appraisal) on salivary cortisol and CRP patterns over the six study time points (baseline and 1, 15, 30, 45, and 60 min post-TSST). This approach accounts for intra-individual variation of baseline cortisol, missing data, unequally spaced data points, non–independence of repeated measures data, and is a more appropriate test for handling data that may violate normality assumptions (Hruschka et al., 2005; Urizar et al., 2019). This analytic approach also allows for the examination of within- and between-person variation to the TSST and accounts for autocorrelation of observations across time. Finally, it tests for linear, quadratic, and cubic cortisol patterns, which are particularly important to examine when there are individual differences in participants’ cortisol peak reactivity and recovery following the TSST.

The mixed model was estimated by maximum likelihood using SAS PROC MIXED (SAS Institute, Cary, NC, USA). Simple effect analyses testing average differences across groups and covariate-adjusted analyses that additionally controlled for BMI, self-esteem, and perceived stress were conducted. The effect sizes for these time and group effects are presented as partial eta squared (ηp2), as is recommended for mixed models (Baguley, 2009). A second, exploratory mixed effect linear model was performed to test for significant two-way interactions (i.e., shame by racism) that could influence shame effects on salivary cortisol and CRP. Main effects and interaction terms of p < 0.05 were considered to be statistically significant. The simple effect slopes of continuous variables (e.g., shame) were then used to illustrate significant moderation effects by racism on crude cortisol and CRP values. Additionally, a dichotomized version of the shame and racism (past year, entire life, stress appraisal) variables was used via median split to better illustrate group differences (i.e., low vs. high shame and racism) on cortisol and CRP responses. The least-squares means method was used to compare group means for all significant effects.

3 |. RESULTS

3.1 |. Sample characteristics

The average age of participants in this study was 19 years old. Most participants self-identified as Black or African American and heterosexual (Table 1). No differences were observed between women in the low versus high groups for past year racism, entire life racism, stress appraisal, or shame on demographic characteristics. However, those with higher perceived racism in the past year reported significantly higher entire life racism (M = 2.05, SD = 0.76 vs. M = 3.20, SD = 1.05), stress appraisal (M = 2.61, SD = 1.07 vs. M = 3.90, SD = 1.13), and shame (M = 31.71, SD = 18.83 vs. M = 40.24, SD = 16.63). As expected, the higher stress appraisal group had higher perceived stress (M = 29.59, SD = 4.67 vs. M = 31.69, SD = 5.91). Lastly those in the high shame group had significantly higher stress appraised from racism (M = 2.84, SD = 1.10 vs. M = 3.68, SD = 1.31).

TABLE 1.

Demographic characteristics of study sample by racism and shame level (low vs. high)

	Past year racism		Entire life racism		Stress appraisal		Shame
	Low	High	Low	High	Low	High	Low	High
Demographics
Age [M (SD)]	19.07 (1.34)	19.76 (1.20)	19.11 (1.41)	19.86 (1.03)	19.13 (1.30)	19.71 (1.26)	19.60 (1.30)	19.23 (1.31)
Black/African American [n (%)]	15 (93.75%)	12 (66.66%)	16 (88.88%)	11 (73.33%)	13 (81.25%)	14 (82.36%)	14 (82.36%)	13 (81.25%)
African [n (%)]	1 (6.25%)	3 (16.67%)	1 (5.56%)	2 (13.33%)	1 (6.25%)	2 (11.76%)	1 (5.88%)	2 (12.50%)
Afro Latino [n (%)]		1 (5.56%)		1 (6.67%)	1 (6.25%)		1 (5.88%)
Biracial [n (%)]		2 (11.11%)	1 (5.56%)	1 (6.67%)	1 (6.25%)	1 (5.88%)	1 (5.88%)	1 (6.25%)
Heterosexual [n (%)]	13 (81.25%)	17 (100%)	16 (88.90%)	14 (93.33%)	14 (87.50%)	16 (94.12%)	15 (88.24%)	15 (93.75%)
Homosexual [n (%)]	1 (6.25%)		1 (5.55%)	0	1 (6.25%)		1 (5.88%)	0
Bisexual [n (%)]	2 (12.50%)		1 (5.55%)	1 (6.67%)	1 (6.25%)	1 (5.88%)	1 (5.88%)	1 (6.25%)
Study variables
Past year racism [M (SD)	1.50 (0.32)	2.67 (0.62)	1.72 (0.45)	2.55 (0.83)**	1.76 (0.53)	2.41 (0.83)	1.94 (0.69)	2.23 (0.82)
Entire life racist [M (SD)]	2.05 (0.76)	3.20 (1.05)***	1.92 (0.47)	3.51 (0.89)**	1.97 (0.56)	3.28 (1.01)**	2.36 (0.76)	2.89 (1.23)
Stress appraisal [M (SD)]	2.62 (1.07)	3.90 (1.13)***	2.59 (1.08)	4.11 (0.94)**	2.27 (0.66)	4.26 (0.87)***	2.84 (1.10)	3.68 (1.31)*
Shame [M (SD)]	31.71 (18.83)	40.24 (16.63)***	32.47 (4.48)	40.41 (15.44)	29.18 (17.56)	42.76 (16.21)***	20.41 (8.38)	51.47 (9.43)**
Perceived stress [M (SD)]	29.06 (4.92)	32.25 (5.39)	29.68 (1.08)	31.85 (6.03)	29.59 (4.67)	31.69 (5.91)**	29.65 (4.85)	31.63 (5.77)
Self-esteem [M (SD)]	3.27 (0.72)	3.10 (0.88)	3.18 (0.71)	3.18 (0.93)	3.18 (0.71)	3.19 (0.91)	3.15 (0.79)	3.22 (0.83)
Body Mass index [M (SD)]	23.40 (3.29)	24.26 (6.13)	24.47 (5.05)	23.03 (4.66)	24.78 (5.81)	22.88 (3.63)	24.09 (4.22)	23.57 (5.55)

Note: Pearson’s x² and independent samples t-tests were conducted to test for between group differences among categorical and continuous variables, respectively.

^*p < 0.05

^**p < 0.01

^***p < 0.01.

3.2 |. Cortisol response to the TSST

Mixed effect linear model analyses of change revealed that the TSST elicited a significant biological cortisol response among participants (order 5 pattern), adjusting for BMI, self-esteem, and perceived stress. The greatest cortisol reactivity (increase in cortisol) occurred from timepoint 1 (baseline, M = 4.83 nmol/L) to timepoint 3 (15 minutes post TSST, M = 5.28 nmol/L). The greatest cortisol recovery (decrease in cortisol) was observed from timepoint 3 (15 minutes post TSST, M = 5.28 nmol/L) to timepoint 6 [60 min post TSST, M = 2.95 nmol/L; F(1, 33) = 21.53, p < 0.001, ηp²= 0.33; see Figure 1].

FIGURE 1.

Changes in salivary cortisol following a laboratory stressor

3.3 |. Main effects of shame and racism on cortisol responses

Shame.

A mixed effect linear model analysis of change determined that there was no significant main effect of shame on cortisol responses to the TSST, controlling for BMI, self-esteem, and perceived stress [F(1, 33) = 0.84, p = 0.367, ηp² = 0.03].

Racism.

A mixed effect linear model analysis of change determined that there was a significant main effect for past year racism on cortisol responses to the TSST, controlling for BMI, self-esteem, and perceived stress [cubic pattern; F(1, 32) = 5.19, p = 0.031, ηp² = 0.16]. Specifically, Black women who reported experiencing a higher degree of racism during the past year had a heightened cortisol response to the TSST compared to those who reported experiencing less racism in the past year (see Figure 2a). Results also demonstrated a significant main effect of entire life racism on cortisol responses, controlling for BMI, self-esteem, and perceived stress [cubic pattern; F(1, 32) = 6.34, p = 0.018, ηp² = 0.19]. Specifically, Black women who reported experiencing a higher degree of racism in their lifetime had a heightened cortisol response to the TSST (but lower cortisol levels overall) compared to those who reported experiencing less racism in their lifetime (see Figure 2b). There was no significant main effect of stress appraisal racism on cortisol responses to the TSST [F(1, 32) = 1.88, p = 0.182, ηp² = 0.07].

FIGURE 2.

Changes in salivary cortisol following a laboratory stressor by racism type (past year & entire life). (a) Cortisol patterns by racism in the past year (low vs. high). (b) Cortisol pattern by entire life racism (low vs. high)

3.4 |. Interaction effects of shame by racism on cortisol responses

A mixed effect linear model analysis of change showed that there was a significant three-way interaction between shame and racism [Time × Shame × Racism (Past Year, Entire Life, Stress Appraisal)] on cortisol responses to the TSST controlling for BMI, self-esteem, and perceived stress. The dichotomized shame and racism variables (i.e., median split) were used to plot these differences (Figure 3a–f). Specifically, among Black women who experienced a higher degree of racism during the past year, those who also experienced high levels of shame demonstrated a heightened cortisol response to the TSST compared to those experiencing low levels of shame [quadratic pattern; F(1, 32) = 8.87, p = 0.006, ηp² = 0.26; see Figure 3b]. Similarly, among Black women who experienced a higher degree of racism in their lifetime, those who also experienced high levels of shame demonstrated a heightened cortisol response to the TSST compared to those experiencing low levels of shame [order 5 pattern; F(1, 32) = 10.10, p = 0.004, ηp²= 0.29; see Figure 3d]. Finally, among Black women who experienced higher stress appraisal from racism, those who also experienced high levels of shame demonstrated a heightened cortisol response to the TSST compared to those experiencing low levels of shame [order 5 pattern; F(1, 32) = 8.76, p = 0.007, ηp² = 0.26; see Figure 3f].

FIGURE 3.

Changes in salivary cortisol following a laboratory stressor by shame and racism group. (a) Cortisol patterns for low shame group by past year racism. (b) Cortisol patterns for high shame group by past year racism. (c) Cortisol patterns for low shame group by entire life racism. (d) Cortisol patterns for high shame group by entire life racism. (e) Cortisol patterns for low shame group by stress appraisal racism. (f) Cortisol patterns for high shame group by stress appraisal racism

3.5 |. CRP response to the TSST

Mixed effect linear model analyses of change revealed that the TSST elicited a significant decline in biological CRP levels among participants (order 5 pattern), adjusting for BMI, self-esteem, and perceived stress. The greatest decrease in CRP was observed from timepoint 1 (baseline, M = 768.12 pg/ml) to timepoint 6 [60 min post TSST, M = 481.52 nmol/L; F(1, 33) = 19.70, p < 0.001, ηp²= 0.10; see Figure 4].

FIGURE 4.

Changes in C-Reactive Protein (CRP) following a laboratory stressor

3.6 |. Main effects of shame and racism on CRP responses

Shame.

A mixed effect linear model analysis of change determined that there was a significant main effect for shame on CRP responses [cubic pattern; F(1, 32) = 6.10, p = 0.019, ηp² = 0.10]. Specifically, Black women who reported experiencing a higher degree of shame showed more blunted CRP responses to the TSST (i.e., lower CRP levels throughout) compared to those who reported experiencing less shame (see Figure 5).

FIGURE 5.

CRP patterns by Shame Group (low vs. high)

Racism.

There was no significant main effect of past year racism on CRP responses to the TSST [F(1, 33) = 3.03, p = 0.094, ηp²= 0.11]. However, there was a significant main effect of entire life racism on CRP responses, controlling for BMI, self-esteem, and perceived stress [linear pattern; F(1, 32) = 4.65, p = 0.041, ηp² = 0.16]. Specifically, Black women who reported experiencing a higher degree of racism in their lifetime showed significantly higher CRP responses to the TSST compared to those who reported experiencing less racism in their lifetime (see Figure 6a). Finally, results demonstrated that there was a significant main effect of stress appraisal racism on CRP responses, controlling for BMI, self-esteem, and perceived stress [linear pattern; F(1, 32) = 14.33, p = 0.001, ηp² = 0.36]. Specifically, Black women who reported experiencing a higher stress appraisal from racism showed a blunted CRP response immediately after the TSST compared to those who reported experiencing lower stress appraisal from racism (see Figure 6b).

FIGURE 6.

Changes in salivary C-Reactive Protein (CRP) following a stressor by racism type (entire life & stress appraisal). (a) CRP patterns by entire life racism (low vs. high). (b) CRP patterns by stress appraisal racism (low vs. high)

3.7 |. Interaction effects of shame by racism on CRP responses

A mixed effect linear model analysis of change showed that there was a significant three-way interaction between shame and racism [Time × Shame × Racism (Past Year, Entire Life, Stress Appraisal)] on CRP responses to the TSST controlling for BMI, self-esteem, and perceived stress. Specifically, among Black women who perceived a lower degree of racism during the past year, those who also experienced high levels of shame showed increased CRP responses, particularly at 45 min Post-TSST compared to those experiencing low levels of shame [cubic pattern; F(1, 32) = 6.10, p = 0.019, ηp²= 0.10; see Figure 7a,b]. In contrast, among Black women who experienced a higher degree of racism in their lifetime, those who also experienced high levels of shame demonstrated greater CRP reactivity to the TSST (15- and 45-min Post-TSST) compared to those experiencing low levels of shame [order 4 pattern; F(1, 32) = 4.49, p = 0.042, ηp² = 0.11; see Figure 7c,d]. Similarly, among Black women who experienced higher stress appraisal from racism, those who also experienced high levels of shame had greater CRP responses to the TSST (15- and 45-min Post-TSST) compared to those experiencing low levels of shame [order 5 pattern; F(1, 32) = 6.75, p = 0.014, ηp² = 0.11; see Figure 7e,f].

FIGURE 7.

Changes in salivary C-reactive protein (CRP) following a laboratory stressor by shame and racism group. (a) CRP patients for low shame group by past year racism. (b) CRP patients for high shame group by past year racism. (c) CRP patients for low shame group by entire life racism. (d) CRP patients for high shame group by entire life racism. (e) CRP patients for low shame group by stress appraisal racism. (f) CRP patients for high shame group by stress appraisal racism

4 |. DISCUSSION

This study sought to examine the relationship between racism, shame, and stress reactivity via cortisol and CRP among young Black women. Consistent with previous literature, Black women in this study exhibited blunted cortisol responses to stress suggesting early HPA axis dysfunction within this sample (Chong et al., 2008; Urizar et al., 2019). Black women experienced minimal cortisol reactivity demonstrating a mean increase of only 0.45 nmol/L 15 minutes after stress. In fact, cortisol levels significantly decreased 1.88 nmol/L below baseline during stress recovery. Previous research suggested cortisol and CRP response would be inversely related such that decreased cortisol levels would coincide with heightened CRP reactivity (Knight et al., 2021; Nijm & Jonasson, 2009; Richman & Jonassaint, 2008). However, we also observed blunted CRP responses that indicated a significant decrease in mean CRP levels of 286.6 pg/ml from baseline to the end of the stress recovery phase.

Contrary to our hypothesis, our findings suggest Black women who reported more frequent experiences of racism over the past year and entire life racism exhibited greater cortisol reactivity. Cortisol reactivity was even greater among Black women reporting more frequent experiences of racism over their lifetime. Black women continued to show blunted CRP responses to stress when accounting for racism. However, women with greater experiences of racism over their lifetime racism exhibited significantly higher CRP than women with less frequent racist experiences suggesting that women in this group may be subject to persistent inflammatory activity.

The most notable differences for cortisol and CRP reactivity were observed between Black women in low versus high shame groups. We found that the effect of racism on cortisol and CRP reactivity was significantly moderated by internalized shame. Surprisingly, Black women with a higher frequency of racism (i.e., past year and lifetime) with increased shame exhibited greater cortisol reactivity at 15 min after stress compared to women with low shame who exhibited minimal cortisol reactivity. This is consistent with other studies linking shame and greater cortisol reactivity (Dickerson, Gruenewald, et al., 2004; Gruenewald et al., 2004). Given this finding, it was believed that conditions contributing to increased cortisol reactivity (i.e., high racism and high shame) should also contribute to decreased CRP. However, we observed a similar pattern among Black women who experienced a high frequency of racism over their lifetime and stress appraised from racism. Coupled with greater shame, Black women who experienced greater lifetime racism and racism-related stress exhibited significant increases in CRP levels at 15 min post-TSST relative to women with low shame. However, CRP stress reactivity response patterns deviated from cortisol at 45 min Post-TSST. CRP responses peaked 45 min after social stress suggesting that while cortisol reactivity began to diminish at 15 min, CRP reactivity continued for 30 additional minutes. These findings align with research indicating acute stress elicits late phase CRP responses during stress recovery (Lucas et al., 2017, 2016; Marsland et al., 2017; Steptoe et al., 2007). Moreover, it also supports research suggesting immune responses to stress (i.e., C-reactive protein) may have an inhibitory effect on cortisol (Knight et al., 2021; Nijm & Jonasson, 2009). Surprisingly, Black women with high shame but low perceived racism in the past year elicited C-reactive protein response patterns similar to those with greater racism over the lifetime and stress appraised from racism. Thus, shame, rather than racism, may be a more significant facilitator of CRP reactivity among Black women with lower perceived racism in the past year. The relationship between recent experiences of racism and shame should be evaluated in future stress reactivity to better explain this finding.

4.1 |. Limitations

These findings should be viewed with caution in light of several limitations. The study derived from a relatively small sample size (n = 34) of young Black women in Southern California thus limiting the generalisability of our findings. The current study findings are also exploratory in nature due to the limited sample size despite yielding medium to large effects (ηp² = 0.07–0.36). However, the reliability of the estimates can vary due to the sample size. Lastly, it is possible that some surveys that measured, for example, racism, may have increased stress levels in Time 1 to stress levels which could possibly explain the lack of cortisol and cortisol reactivity to TSST. However, we believe this impact on these surveys is minimal given previous studies that show blunted cortisol among Black women, late phase CRP responses to acute stress, and our findings indicating cortisol and CRP responses to acute stress (TSST) differ by high versus low shame. Other primary stress biomarkers (e.g., DHEAS, norepinephrine, epinephrine) can provide additional information regarding the acute stress response. This warrants further exploration of racism, shame and stress reactivity in studies including these biomarkers in larger sample sizes. Nonetheless, this study is one of the first to examine cortisol and CRP responses to a laboratory social stressor by racism and shame. This provides new insight into how specific psychosocial factors can contribute to how one physiologically responds to social stress.

5 |. CONCLUSIONS AND IMPLICATIONS

Our findings suggest that internalized shame among Black women may play an important role in adaptive physiological responses to race-related stress. Increased shame among Black women experiencing racism and racism-related stress did not elicit similar cortisol and CRP responses to stress. Heightened cortisol reactivity was observed among women experiencing more frequent racist experiences and high shame for 15 minutes after stress. Alternatively, CRP responses peaked approximately 45 minutes after laboratory stressors for women who experience high shame coupled with high lifetime racism and racism-related stress. Admittingly, these findings are complex, somewhat supporting the inverse relationship often observed between cortisol and CRP. However, the complexity of these findings necessitates further exploration of physiological adaptive responses to specific psychosocial stressors. In addition, this study provides valuable insight into the role of emotions, shame specifically, in influencing adaptive physiological responses to social stress among young Black women. Both cortisol and C-reactive protein responses to social stress, not accounting for shame, elicited minimal responses to a laboratory stressor. Reactivity was only unmasked when considering the role of shame among Black women in this study. This suggests shame may play a key role in facilitating the embodiment of racism. Further exploration of shame, racism, and CVD risk in larger studies is warranted. Moreover, recent studies suggested effective emotion regulation may improve cardiovascular health. This current study supports the exploration of interventions that incorporate emotion regulation strategies targeting racism related shame as a useful strategy for reducing CVD risk among Black women.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available upon a reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.