Introduction

Black women have consistently shown higher rates of HIV amounting to 59% of new diagnoses in 2017 for women in the USA [1]. As a marginalized group, they also experience unique stressors such as microaggressions, racial discrimination, poverty, and violence/abuse that make them especially vulnerable to poor physical and psychological health [2,3,4]. They experience daily microaggression, a term coined by Pierce [5] to describe subtle insults, intentional or unintentional, made to belittle, or make negative assumptions about someone of a certain group based on race, culture, or gender. Microaggressions have been shown to impact Black women’s mental and physical health [4, 6, 7] and have been linked to depression and posttraumatic stress disorder symptoms among Black women living with HIV (BWLWH) [4, 8]. Similarly, higher self-silencing has been associated with higher depressive symptoms among women living with HIV [9]. The term self-silencing describes behaviors such as containing one’s emotion, thoughts, and desires in order to avoid conflict or damaging a relationship [6]. A core symptom of depression is suicidal ideation and attempt [10] and as such both microaggression and self-silencing may be related to suicidality among BWLWH, but this remains unexplored.

Black women experience microaggressions daily [11] and specifically GRMS, which are microaggressions on the basis of their intersectional identities as Black and women that are linked to stereotypes such as being angry, strong, and assertive [7]. The small body of empirical literature focusing on GRMS both among Black women in general and BWLWH has indicated links to depressive symptoms [4, 7, 12]. For instance, Williams and Lewis [12] found that GRMS was related to psychological distress and disengagement and depressive symptoms in a general sample of Black women [12]. We are aware of only three studies that have been published on GRMS among BWLWH [4, 8, 13]. The first study found that GRMS was associated to barriers to HIV care and predicted barriers to care above and beyond the contribution of race- and HIV-related discrimination [13]. Findings from the second study indicated that GRMS predicted higher PTSD symptoms and cognitions and made unique contributions to these symptoms above race- and HIV-related discrimination [4]. Similar findings were also reported in the third publication with GRMS predicting depression diagnosis above and beyond race- and HIV-related discrimination [8]. Suicidality and self-silencing are important issues to understand in relation to GRMS and no literature has examined these relationships among BWLWH.

The Theory of Silencing the Self (TSTS) first explained by Jack [6] describes that women silence their thoughts and feelings, by not communicating needs or directly expressing themselves to avoid disagreements and preserve the relationship [14, 15]. This behavior demonstrated in Black women may be reinforced by perceived gender norms and stereotypes such as the Strong Black Woman (SBW) schema [16] and self-sacrifice in relationships (i.e., prioritizing partners happiness more than their your own). In the attempt to perform traditional female roles and neglect their own feelings, self-silencing behavior is related to worsened depressive symptoms and feelings of anger and frustration overtime [6, 17]. This theory remains especially true for BWLWH given the vulnerability of their relationships that may be influenced by financial dependency, abuse, fear of being rejected, and other factors [3, 17]. Previous research has focused on self-silencing in relation to depression, medication adherence, resilience, and the SBW schema [6, 15,16,17,18,19]. However, none has focused on the relationship between self-silencing and suicidal ideation in BWLWH.

Quantitative research on suicide among African Americans has been limited [20] and some existing research has indicated lower rates of suicidality among African Americans compared to other racial groups (e.g., rate of 2.4 among African American women compared to 6.9 among White women) [21]. However, research indicates that people living with HIV may be more vulnerable to suicidal ideation (thoughts) and attempts when considering factors such as depression, substance abuse, and trauma of initial HIV diagnosis [10, 22]. Low-income BWLWH also experience risk factors for suicide attempt such as PTSD, hopelessness, relationship discord, and intimate partner violence that heighten their chances of suicidal behavior and ideation [22]. Research also indicates that living with HIV includes stressors unique to the virus such as loss of loved ones and peers due to HIV and rejection due to HIV-stigma that may also trigger feelings of hopelessness, lifetime suicidality, and depression [23].

Multiple sources have linked microaggression to suicidality through depressive symptoms but not in Black women [7, 12, 24]. Among a diverse sample of young adults (33% African American), researchers found that experiencing various types of racial microaggression can lead to an increase in depressive symptoms which subsequently may result in increased suicidality [24]. Researchers have also indicated a relationship between suicidality and self-silencing through depressive symptoms among depressed women that arise from lack of expression and self-condemnation in relationships [6, 25]. However, scholars have not yet investigated the association between microaggression and self-silencing as they relate directly to suicidality in BWLWH. Women of this community endure daily microaggressions and discrimination that may reinforce self-silencing in instances of feeling shame, invisible, or inferior [7, 11, 14]. This study aims to investigate among BWLWH how GRMS and silencing the self may be related to higher suicidal ideation and is guided by Kimberle Crenshaw’s Theory of Intersectionality (1989) [26], which highlighted that for Black women being Black and a woman within systems of marginalization creates an experience of discrimination that is not one-sided but compounded on the basis of gender and race.

Methods

Participants

One hundred and nineteen BWLWH in the South Florida region were recruited to participate in a clinical intervention study between October 2017 and May 2018. Participants were recruited through flyers, posters distributed in different community locations such as health clinics, hospitals, and community centers. Potential participants who contacted the lab were screened to determine their eligibility for the study. The eligibility requirements included (1) ≥ 18 years of age, (2) cisgender female, (3) identify as Black and/or African American, (4) history of abuse/trauma (i.e., responding “yes” to “During your lifetime have you experienced trauma or abuse?”), (5) prescribed antiretroviral therapy (ART) for at least the last 2 months, (6) English speaking, and (7) possibility of low ART adherence, detectable viral load within the past year, and/or missed HIV-related medical visits within the past year.

Participants who were deemed eligible then came in for two baseline assessments over the span of 2 weeks. During the first baseline assessment, the participants signed informed consent documents, followed by a self-report survey battery, and lastly a semi-structured clinical interview administered by a trained clinician/staff. Participants who completed both the first and second baseline assessment received a total of US $50 (US $25 given during the first visit and the second visit). All procedures done in the study were approved by the University of Miami’s Institutional Review Board.

Measures

Self-Report Sociodemographic Survey

The survey included questions on age, their country of birth, living situation, employment status, religious affiliation, relationship status, sexual orientation, annual income, and HIV viral load status (undetectable vs detectable).

Silencing the Self [27]

The Silencing the Self-Scale consists of four subscales (Externalized Self-perception, Care as Self-sacrifice, Silencing the Self, The Divided Self) to understand how women navigate self-suppression of their thoughts, emotions, and actions to avoid problems in important relationships. The scale consists of 31 items and sample items include “Considering my needs to be as important as those of the people I love is selfish” and “Caring means choosing to do what the other person wants, even when I want to do something different.” The items were rated by participants using a 5-point Likert scale ranging from strongly disagree to strongly agree and total scores ranged from 31 to 155. The scale has shown excellent test-retest reliability and internal consistency with alpha scores ranging from (.86 to .94) [27].

Mini International Neuropsychiatric Interview (MINI; [28])

The MINI [29, 30] was used to assess suicidality in BWLWH. The MINI is a shortened diagnostic assessment to determine DSM-5 and ICD-10 psychiatric disorders such as major depression. The suicidality module (B) focuses on frequency (occasional, often, and very often) of the ideation, and intensity (mild, moderate, severe). The interview also focuses on what may be triggering suicidal ideation, and what actions participants have taken in the past attempts. The MINI has been used in a previous study with BWLWH [8] and in other studies has also shown great reliability and sensitivity for detecting disorders [28].

Gendered Racial Microaggressions Scale-Black Women (GRMS-BW; [31])

The GRMS-BW captures microaggression experienced by Black women by measuring frequency of the occurrences (how often they happen) and stress appraisal (how stressful the event was). The scale includes four subscales: Assumptions of Beauty and Sexual Objectification subscale, Silenced and Marginalized subscale, Strong Black Woman subscale, and Angry Black Woman subscale. Some of the experiences listed on the scale include items such as “Someone has tried to “put me in my place,” “Assumed I did not have much to contribute to the conversation,” “Someone told me to calm down,” and “Negative comments about my hair when natural.” The frequency scale ranges from 0 to 5 (0 = never, 1 = less than once a year …. 5 = once a week or more), and stress appraisal of the event was rated on a scale from 0 to 5 (0 = never happened, 1 = not at all stressful ... 5 = extremely stressful). Lastly, Cronbach’s alpha scores (appraisal α = 0.93 and frequency α = 0.92) have shown good validity and reliability in measuring these experiences [31].

Analyses

Statistical analyses were done using SPSS 25. The analysis included all one hundred and nineteen participants in the study. Multivariable logistic regressions were done to assess if microaggression (appraisal and frequency) and self-silencing predicted suicidality separately and if both factors predicted this outcome together. The first set of analyses included covariates of age and education, predictors (microaggression and/or self-silencing), and outcomes of current suicidality, lifetime suicidality, and suicidality severity. The second set of analyses included age and education (block 1), major depression (block 2), microaggression or self-silencing (block 3), and outcomes of suicidality.

Results

One hundred and nineteen BWLWH participated, and Table 1 presents socio-demographic characteristics and descriptive statistics on study measures. In short, women’s average age was 49 (range = 22–67), 63% had an annual income below $12,000, 66% completed high school or above, 58% were on disability, and 67% were renting a home/apartment. Twenty-six percent of the women had lifetime suicidality, 10% had current suicidality, and 23% had low suicidality severity. Women reported on average a moderate level of self-silencing (mean = 81.92) and low to moderate microaggressions (frequency mean =24.91, appraisal mean = 25.34).

Table 1 Sociodemographics and clinical characteristics of BWLWH
Full size table

Multivariable Associations of Microaggressions and Suicidality

Higher microaggression appraisal significantly predicted higher likelihood of current suicidality (β = 2.80, p < .01) (see Table 2). There were no significant relationships between microaggression frequency and current suicidality (β = 1.5, p = .14), lifetime suicidality, or suicidal severity. Microaggression appraisal did not significantly predict lifetime suicidality (β = 1.5, p = .09) or suicidality severity. When major depression was included microaggression (appraisal and frequency) did not contribute uniquely to current suicidality (appraisal β = 1.5, p =.11), lifetime suicidality, or suicidality severity above and beyond depression.

Table 2 Hierarchical logistic regressions of suicidality by microaggressions and self-silencing
Full size table

Multivariable Associations of Self-silencing and Suicidality

Higher self-silencing significantly predicted higher likelihood of current suicidality (β = 1.05, p < .01). Similarly, higher self-silencing significantly predicted higher likelihood of lifetime suicidality (β = 1.03, p < .01). However, self-silencing did not significantly predict suicidality severity. Nonetheless, when major depression was included, self-silencing contributed uniquely to current suicidality (β = 1.03, p < .05) and lifetime suicidality (β = 1.03, p < .01) above the contributions of depression.

Multivariable Associations of Predictors of Microaggression and Self-silencing Entered Together and Outcome of Suicidality

Multivariable regressions were conducted to determine if microaggression or self-silencing made unique contributions (above other predictors) to suicidality. We entered covariates (education, age, and income) in block 1, three predictors together (microaggression appraisal, microaggression frequency, and self-silencing) in block 2, and outcomes of current suicidality, lifetime suicidality, and suicidal severity separately. Our findings indicated that self-silencing significantly and uniquely contributed to lifetime suicidality above and beyond the contributions of microaggression frequency and appraisal (β = 1.03, p <.05). However, self-silencing did not significantly predict current suicidality (β = 1.03, p =.06) above and beyond the contributions of microaggression. Microaggression appraisal and frequency did not contribute uniquely to current suicidality (appraisal β = 2.36, p = .06; frequency β = .80, p = .52) or lifetime suicidality (appraisal β = .59, p = .09; frequency β = 1.46, p = .22). Also, none of the three predictors (microaggression appraisal, microaggression frequency, and self-silencing) showed a unique contribution to suicidal severity. Nonetheless, when major depression was also included in this model, self-silencing still contributed uniquely to lifetime suicidality (β = 1.03, p < .05), but did not contribute uniquely to current suicidality (β = 1.03, p = .11).

Discussion

Our analyses indicated multiple significant relationships between microaggression, self-silencing, and suicidality among BWLWH. First, we found that higher microaggression appraisal significantly predicted higher likelihood of current suicidality. This finding is consistent with previous research among BWLWH that found that GRMS is significantly associated with depression and PTSD symptoms [4, 8], which are both associated with suicidality [32, 33]. For instance, past studies have indicated that lifetime suicide rates are associated with depression among people with HIV [10]. Our finding is also consistent with results in a study done by O’Keefe et al. [24] among a diverse sample of young adults (33.3% African American), which found that racial microaggressions were related to increased depression and suicidal ideation. The nonsignificant findings for microaggression frequency in association with current and lifetime suicidality may be due to our sample size and perhaps in a larger sample size we would see significant findings. Nonetheless, the fact that there was a significant finding for microaggression appraisal, but not frequency may suggest that appraisal is more predictive in terms of suicidality than frequency. Perhaps how distressed the microaggressions make the women feel is more relevant to suicidality than how often women experience the microaggressions.

Our findings for self-silencing among BWLWH indicated that it was significantly related to current and lifetime suicidality. These associations are supported by past studies among women with depression, BWLWH, and a large sample of women with HIV (75% Black women) that found relationships between self-silencing, depression, anger, and self-blame [6, 15, 17, 18]. Past research indicates that through self-silencing behaviors, individuals may suffer from built up frustration and lack of expression in their relationships, leading to depressive symptoms and a high risk of suicidality [6, 25]. This is also consistent with past findings indicating that relationship discord, hopelessness, social support, and other factors experienced by Black women put them at risk for suicide attempts [22].

This buildup of frustration and other consequences of self-silencing may be the common ground between current and lifetime suicidality. We found that higher self-silencing significantly predicted a higher likelihood of lifetime suicidality. Gender norms such as self-silencing are enculturated in childhood and reinforced by society through women’s lifetime; therefore, BWLWH may self-silence to preserve harmony in their relationships to meet gender-based societal norms [6]. For BWLWH, given their multiple marginalized identities, there are additional ways in which they may have to engage in self-silencing behaviors to persevere outside of the context of gender roles in intimate relationships. For instance, self-silencing may occur in the context of race- and HIV-related stigma and discrimination.

Additional findings compared all three predictors to emphasize the relationships between microaggression appraisal and frequency, self-silencing, and suicidality overall. We found that self-silencing best predicted different types of suicidality overall, but specifically, it uniquely contributed to lifetime suicide attempt above and beyond the contributions of both microaggression frequency and appraisal. This may relate back to the fact that self-silencing represents a gender role that is taught in early childhood and maintained/reinforced by society across a women’s lifespan; therefore, self-silencing may be more predictive of lifetime suicidality than microaggression frequency and appraisal captured in the past year.

Once depression was factored into our various analyses, self-silencing continued to contribute uniquely to suicidality above and beyond depression, indicating that self-silencing may play an important role in suicidality among BWLWH independent of depression. However, microaggression did not uniquely contribute to suicidality above depression, suggesting that the relationship between microaggression and suicidality among BWLWH may be explained largely by depression. Nonetheless, these findings help to shed light on how both self-silencing and microaggressions relate to suicidality among BWLWH.

Although the findings in this study make important contributions to the existing literature, there are some limitations, the first of which is our sample. The data was collected from BWLWH with histories of trauma living in a city in the Southeastern region of the USA, which limits the generalizability of our finding. Second, the cross-sectional design of the study prevents drawing casual inferences. Third, the self-report measures utilized may be susceptible to social desirability bias. Fourth, although we had a moderate sample size of one hundred and nineteen women, a larger sample size may have equipped us with greater power to detect additional significant findings among our variables of interest. For instance, for the suicide severity variable 23% of women had low severity and only 5% had high severity, which may have limited our ability to detect significance in relation to this outcome. Even with these limitations, our findings remain important to understanding the overall impact of microaggression and self-silencing on suicidality in BWLWH. Consistent with past research, our findings regarding the impact microaggression and self-silencing may have on severity, current, and lifetime suicidality have several meaningful implications. Future research efforts to determine factors influencing suicidality in BWLWH should account for self-silencing and microaggression as possible contributing factors. In addition, due to differences in microaggression appraisal and frequency findings, future studies are needed to further understand why these differences occur and which is the better predictor of all three-suicidality outcomes (lifetime, current, and severity).

Beyond future studies, interventions are needed at the national, community, and individual levels to alleviate microaggressions experienced by BWLWH and self-silencing. At the national level anti-racist and anti-misogynoir policies are needed to create a context where BWLWH do not face every day gendered racial microaggressions or need to engage in self-silencing to “preserve” harmony. In addition, policies are needed that increase the availability and reach of culturally congruent mental health services for BWLWH women to access for prevention or intervention around suicidal ideation. Similarly, at the community level financial resources are need to support initiatives to empower and support BWLWH as they work to enact positive change around mental and physical health in their communities. At the individual level, interventions are needed that are informed by the voices of BWLWH (e.g., [34]) and aim to enhance coping, emotional well-being, and self-advocacy as BWLWH navigate their day to day lives and face gendered racial microaggressions and gender roles expectations to self-silence.

No previous study has directly investigated the associations between microaggression, self-silencing, and suicidality in BWLWH. The findings in our study indicate that self-silencing and microaggression appraisal predict current and/or lifetime suicidality for BWLWH. Specifically, microaggression appraisal predicted higher likelihood of current suicidality, and higher self-silencing predicted higher likelihood of current and lifetime suicidality. Future research in a larger sample should explore the impact of microaggression (frequency and appraisal) and self-silencing on suicidality severity. Lastly, intervention research should be aimed at not only finding the best predictor of suicidality in BWLWH, but also understanding protective factors and coping strategies that may enhance psychological well-being among BWLWH with current and/or lifetime suicidality.