Double Impact: A Dyadic Discrimination Model for Poor, Minority, and Pregnant Couples

Abstract

Frequent daily discrimination compounds the negative health impacts of those with multiple marginalized identities, including pregnant mothers and their children. We used a dyadic, moderated, mediated model of 296 young, expectant, poor, urban, primarily minority couples. In this study, we explored if a multiple pathway discrimination model explained the relationship between multiple marginalized identities and health (depression and stress). We also examined if a mediated (discrimination moderated by gender) model, within a minority-stress and intersectional framework explained the relationship with depression and stress for couples. We observed that frequent daily discrimination was associated with negative health outcomes (depression and stress). Women reported significantly more depression than men. Frequent daily discrimination mediated the relationship between multiple marginalized identities and depression and stress and having a partner with multiple marginations increased one’s personal depression and stress. Our observations suggest that discrimination’s impact on health is experienced during pregnancy and the more marginalized identities one carries, the more impact it may have. Further, having a partner with multiple marginalized identities also impacts the depression and stress reported by women. Inventions to address depression and stress outcomes may be strengthened by considering multiple marginalized identities and include couples.

Introduction

Approximately 70% of US adults from minority identities report experiencing racial/ethnic discrimination, with 61% reporting frequent daily discrimination [1]. Ethnic/racial discrimination is characterized by differential treatment and assumptions toward ethnically/racially marginalized individuals because of their ethnicity, race, or associated personal characteristics [2]. Ethnic/racial discrimination can either be enacted by an individual and/or through institutional policies that are often subtle, seemingly innocuous, and pervasive [3, 4]. People who experience ethnic/racial discrimination have higher rates of physical and psychological symptoms, such as depression and stress [5–7]. The negative outcomes associated with ethnic/racial discrimination can be exacerbated by the presence of other forms of discrimination (e.g., gender, income-related discrimination). The negative outcomes associated with multiple forms of discrimination (because of multiple marginalized identities) have been shown to worsen depression symptoms [8, 9].

The effects of discrimination, whether based on one (i.e., ethnic/racial) or multiple (e.g., race, age, ethnic/racial) marginalized characteristics, appear to also persist during some of life’s important transitions such as pregnancy and parenthood [10]. Pregnant minority women’s report of discrimination experiences has been shown to be positively associated with depression, perceived stress, and low infant birth weight [9–12]. Antenatal depression and chronic stress during pregnancy is linked to an overproduction of cortisol, increased blood pressure, constricted blood flow to the fetus, and increased risk for preterm delivery [13, 14].

Research shows Black women experience higher rates of early-term births (1.5 times of white and 1.4 times of Hispanic women), low weight births (2 times of white and Hispanic women), and infant mortality (2.5 times of white and 2 times of Hispanic women) [15, 16]. One possible explanation for this difference in birth outcomes for Black women is frequent daily discrimination. Young, urban women of color experiencing frequent daily discrimination report higher rates of depression and stress [11, 17]. Research has shown that depression mediated the relationship between frequent daily discrimination and lower birthweight for young, urban women of color [17]. Further, frequent daily discrimination increased the amount of stress reported by young, urban women of color, and their partners [11].

Extant literature examining the relationship between discrimination and health has focused primarily on one type of discrimination. Examining how multiple marginalized identities (e.g., poverty, race), which often co-occur, can lead to more frequent daily discrimination, and exacerbate negative health outcomes, is warranted [18, 19]. Studying one type of discrimination’s impact on health has resulted in mixed research findings. For example, research on the effects of discrimination and birth outcomes, has observed that the infant mortality rate is higher for educated Black women than less educated White women [13]. Conversely, Black males report experiencing more frequent daily discrimination than Black women, despite Black women having dual minority status [20, 21]. Researchers have attributed these mixed findings to the sometimes uniquely experienced or additive nature of discrimination. Further, few researchers have considered how multiple factors used to explain the experience of discrimination combine and exacerbate poor health. This failure needs academic attention through the examination of a multidimensional discrimination model and its impact on health [22]. There are several theories that may help to explicate how discrimination and the multiple ways it is elicited negatively impact health. Included are minority stress and intersectionality theories. Minority stress theory [23] posits that being a minority in a marginalizing society causes psychological stress. Minority stress is often experienced through discrimination and slights, the expectation of discrimination, and the internalization of discrimination. This results in deteriorating coping processes in response to these encounters [24]. Stress is most heightened during the conflict between the norms of the dominant culture and those of the marginalized group [25]. Building on the premise of minority stress theory, intersectionality theory underscores the potentially compounding impacts of having multi-stigmatizing identities (e.g., being Black and a woman) [26]. As a result, individuals with multi-stigmatizing minority identities may experience double or triple jeopardy and multiple ways health is undermined [26, 27].

Those with multiple marginalized identities report more daily discrimination than those who experienced no discrimination [28] Grollman’s [18, 28] studies of older adult males and youths found everyday discrimination mediated the relationship between multiple marginalized identities and health. Seng’s [19] study found women with multiple marginalized identities who reported frequent discrimination during pregnancy experienced significantly lower quality of life and more PTSD symptoms. Further, frequent daily discrimination in this study mediated the relationship between marginalized identities and PTSD and quality of life.

Male partners also impact the life of mothers during pregnancy. Black women experiencing low partner support during pregnancy have a sevenfold higher odds of depression than non-Hispanic white women [29]. Research also suggests that the mental health sequelae observed during pregnancy is reciprocal with 13% of fathers’ reported depressive symptoms being correlated with their pregnant partners’ reported depressive symptoms [30, 31].

The current study seeks to explore these complex relationships by examining an additive, interactive, and multidimensional model of multiple marginalized identities on the mental health of pregnant women and their partners. We considered the impact of social factors like partner’s experiences of discrimination and marginalization on antenatal health. By adding partner effects, we hope to build on extant literature that demonstrates that there are more pathways through which multidimensional discrimination can impact antenatal health [22]. We examined a moderated, mediated discrimination model within a dyadic framework for pregnant couples (see Fig. 1) on depression and stress. Our model expounds on the prior research to bring in partners and looks at the individual differences that men and women experience by including gender as a moderator while adding partners. Our research tests three models: (hypothesis/aim 1) actor effect mediation model. We hypothesize that the effects of an individual’s multiple marginalized identities on their mental health (depression, stress) will be mediated by their frequent daily discrimination experiences: (hypothesis/aim 2) partner effect mediation model. We hypothesize that having a partner with multiple marginalized identities will be mediated by the frequency of daily discrimination experienced: (hypothesis/aim 3) moderation. Gender will moderate the paths tested in aims 1 and 2. We hypothesize there will be moderation by gender on the actor and partner mediation models in hypothesis one and two with the effects possibly being greater for women compared to men.

Fig. 1.

Discrimination conceptual model

Methods

Sample

We used cross-sectional data from a longitudinal study of 296 young, expectant couples who visited four Connecticut university-affiliated hospitals between July 2007 and February 2011. After obtaining the necessary permissions, securing the interest of both partners, and screening for eligibility, we conducted baseline interviews. The inclusion criteria were (a) second or third-trimester pregnancy at the time of the baseline interview; (b) women aged 14–21 and men aged 14 + at the time of the interview; (c) both romantically involved; (d) both biological parents of the unborn baby; (e) both agree to participate, and (f) both speak English or Spanish. To ease the completion of surveys for those with low reading skills, we used computer-assisted self-interviews in which the questions were read [32]. Participation was voluntary and did not affect health care or social services received. Participants received $25 for each interview.

Measures

Outcomes

Stress was measured using Cohen and Williamson’s [33] 10-item Perceived Stress Scale. This measures individuals’ appraised level of stress, external demands, and anxiety during the past month. Items included direct questions about being “nervous and stressed” to “able to control irritations in your life” on a 5-point Likert scale from never (0) to very often (5). Higher scores indicated more stress and anxiety symptoms. Cronbach alpha for the sample was good for both females (alpha = 0.787) and males (alpha = 0.734).

Depression was measured using Radloff’s [34] 20-item Center for Epidemiologic Studies Depression Scale. The scale items are designed to measure symptoms associated with depression. Study participants responded to 15-adapted questions on depression symptoms on a 5-point Likert scale ranging from “rarely or none of the time (less than 1 day)” to “most or all of the time (5–7 days). Higher scores indicated more depression symptoms over the past week. Five somatic items were removed because they overlap with common symptoms of pregnancy. Cronbach alpha for our working sample was good for females (alpha = 0.841) and males (alpha = 0.795).

Mediator

Frequent daily discrimination was assessed using an adapted version of Essed’s [3] 20–item 6-point Likert Daily Life Experiences Scale. Introduced in 1991, Essed’s scale measures experiences of discrimination by assessing experiences such as speaking for the entire group, being observed or followed, etc. The Daily Life Experiences Scale range is 0 = “never,” 1 = “less than once a year,” 2 = “a few times a year,” 3 = “about once a month,” 4 = a “few times a month,” and 5 = “once a week or more.” A mean frequent discrimination of 1 means, on average, a person experiences frequent discrimination is less than once a year. We summed the total number of frequent daily discrimination then divided by the total number of items to obtain the average frequent daily discrimination score for each participant. Since each scale item ranged from 0 to 5, the average frequent daily discrimination, ranged from 0 to 5. Cronbach alpha for the sample was very good for females (alpha = 0.924) and males (alpha = 0.939).

Predictor

Multiple marginalized identities

After identifying the frequency of discrimination, participants were asked to ascribe reasons for their daily experiences of discrimination on the Daily Life Experiences Scale. Reasons for discrimination were selected from nine discrimination categories (i.e., race, ethnicity, gender, age, income, language, physical appearance, sexual orientation, other). “Other” was a write-in box that was excluded from the analyses. Participants were allowed to indicate one or more of the discrimination categories that they felt were the reason for the experience of discrimination. The number of multiple marginalized identities was coded as zero, one, or two or more. Study participants with two or more marginalized identities were our reference group and were compared to those having zero or one marginalized identity. This approach strengthened our multiple marginalization conclusions in our discrimination model.

Control variables

We controlled for the variables age, gender, race, education level, employment status, and personal income because they have all been shown to be related to discrimination, stress, and depression [1, 35].

Statistical analyses

We used SPSS 22 to examine demographic information and the types of discrimination experienced. We used profile analysis to understand the nature of discrimination across genders [36]. Gender and frequent daily discrimination items functioned as within-subject variables (within couples). We examined how discrimination experiences impacted depression and stress using the actor-partner interdependence model (APIM). This approach adjusts for the correlated nature of dyadic data [37]. APIM was estimated using generalized estimated equations. Actor effects are an individual’s predictors relating to the individual’s outcome. Partner effects are partners’ predictors relating to the individual’s outcome [38]. The potential study sample size was 592; however, our working sample was n = 578 because of missing data.

To examine the mediational relationship between frequent daily discrimination, multiple marginalized identities, depression, and stress, we used Monte Carlo simulated confidence intervals of the product of the indirect paths using RMediation [39]. Each indirect path was moderated by gender. The indirect paths were from multiple marginalized identities to frequent daily discrimination and from frequent daily discrimination to stress/depression. If the interaction with gender was significant on the indirect paths, simple effects analysis was conducted to describe the nature of the interaction (i.e., whether women were affected more than men or men affected more than women); if the interaction with gender was not significant, the main effects, represented as the indirect effects across both genders, were reported.

Results

The majority (80%) of the study participants were African American or Hispanic. Women’s age range was 15–21 years (M = 18.71; SD = 1.63). Men’s age range was 14–40 (M = 21.33; SD = 4.06). Average personal income for women was $5835 (SD = $7447) and $10,870 for men (SD = $11,858). Women’s average years of education was 11.75 years (SD = 1.82) and men’s average years of education was 11.84 (SD = 1.89). Twelve years of schooling equals to meeting and completing the requirements for high school graduation. Eight percent (8%) of women worked full time, 21% worked part time, and 71% were unemployed. Thirty-one percent (31%) of the men worked full time, 29% worked part time, and 40% were unemployed (see Table 1). Men (M = 1.022, SD = 1.01) reported significantly more frequent discrimination than women (M = 0.846, SD = 0.809, paired t(289) = 2.436, p = 0.015).

Table 1.

Discrimination demographics

Demographics
	Men	Women	p
Race/ethnicity			0.045
Black	144(49)	117(40)
Latino	108(36)	117(40)
White	44(15)	62(20)
Employment			.000*
Full-time	92(31)	23(8)
Part-time	86(29)	61(21)
Unemployed	116(40)	212(71)
Personal income	10,870 ± 11,858	5,835 ± 7,448	.000*
Years of education	11.844 ± 1.89	11.745 ± 1.82	0.456
Daily discrimination	1.022 ± 1.01	.8460 ± .809	.015*
Age	21.33 ± 4.056	18.71 ± 1.630	.000*
Depressive symptomology	8.89 ± 6.617	10.526 ± 7.783	.002*
Perceived stress	15.45 ± 6.312	16.719 ± 6.244	.004*
Multiple marginalized identities			0.985
Zero	70 ( 24)	71 (24)
One	97 (34)	101(34)
Two or more	123 (42)	124 (42)
Frequent daily discrimination			p
Multiple Marginalized Identities	Zero	Two +
Men	.144 ± .361	1.35 ± .943	< .001*
Women	.1204 ± .320	1.207 ± .806	< .001*
Frequent Daily Discrimination			p
Multiple Marginalized Identities	One	Two +
Men	1.234 ± 1.037	1.35 ± .943	0.7991
Women	.9032 ± .732	1.207 ± .806	0.004*

N(%) mean ± standard deviation

N = 290 Men, N = 296, women dyads = 290

^*Significant

Race was described as the primary reason for discrimination (men 50.0%, women 41.6%), followed by age (men 31.8%, women 44.6%). For women, gender was third (27.0). For men, physical appearance (26.4%) was third. Multiple marginalized identities (2 +) were reported by 42% of our sample, 34% reported one marginalized identity, and 24% reported no marginalized identities (see Table 1). Race and age (26%) were the top two multiple marginalization across gender and physical appearance, race, and age (30%) were the top three multiple marginalized identities across both genders.

Profile analysis

Men and women differed significantly on frequent daily discrimination, (F(13.469, 3879.04) = 42.514, p = 0.000). Being stared at (M = 1.684, SD = 0.081) and overhearing an offensive joke or comment (M = 1.4, SD = 0.081) were the most significantly reported frequent daily discriminations reported by both genders. However, we observed significant differences in the type of frequent daily discrimination experienced by gender F(13.881, 3997.69) = 3.548, p = 0.000; Fig. 2). Significantly higher for men than women: being accused or treated suspiciously (M_men = 1.08, M_women = 0.66; J = 0.422, p = 0.000), someone being afraid or intimidated by them (M_men = 1.25, M_women = 0.71; J = 0.533, p = 0.000), being observed or followed (M_men = 0.96, M_women = 0.55; J = 0.405, p = 0.000), being avoided (M_men = 0.65, M_women = 0.46; J = 0.190, p = 0.038), being mistaken for someone who serves (M_men = 0.55, M_women = 0.28; J = 0.266, p = 0.001), being taunted (M_men = 0.84, M_women = 0.60; J = 0.242, p = 0.016), and being asked to speak for or represent their entire group (M_men = 0.99, M_women = 0.72; J = 0.27, p = 0.013). Being treated rudely or disrespectfully had higher scores for women than men but the mean differences were not significant (M_men = 1.03, M_women = 1.12; J = -0.087, p = 0.437).

Fig. 2.

Discrimination profiles by gender

Aim 1 (Actor Effect Mediation Model)

Study participants with multiple marginalized identities experienced more daily discrimination than those with none (B =  − 1.139, p = 0.000) and one marginalized identity (B =  − 0.206, p = 0.014). There were significant positive associations between personal frequent daily discrimination and depression (B = 2.488, p = 0.000) and stress (B = 2.443, p = 0.000; Table 2).

Table 2.

Discrimination mediation table on depression and stress

Multiple marginalizated identities and frequent daily discrimination on depression and stress, mediation model
Variable	Frequent daily discriminationa			Depression			Stress
	B	SE (B)	p	B	SE (B)	p	B	SE (B)	p
Covariates
Intercept	1.678	0.338	0.000	0.785	3.0598	0.797	8.415	2.331	0.000
Race/ethnicitya
Black	Reference			Reference
Latino	− 0.033	0.0731	0.656	0.244	0.6285	0.698	− 0.011	0.568	0.985
White	0.220	0.0888	0.013*	− 0.783	0.868	0.367	− 0.562	0.7434	0.449
Employmenta
Full-time	Reference			Reference			Reference
Part-time	0.01	0.1067	0.922	− 0.67	0.8117	0.409	− 1.042	0.7404	0.159
Unemployed	0.074	0.1108	0.504	− 0.698	0.8383	0.405	− 0.247	0.6941	0.722
Gendera
Male	Reference			Reference			Reference
Female	− 0.306	0.0805	0.000*	2.644	0.5936	.000*	2.11	0.4785	0.000*
Personal incomea	− 0.028	0.0501	0.572	− 0.499	0.3219	0.121	− 0.363	0.2551	0.155
Years of educationa	0.020	0.0243	0.401	0.031	0.1643	0.849	− 0.046	0.1203	0.701
Agea	− 0.027	0.0133	0.044*	0.168	0.1028	0.103	0.216	0.0806	0.007*
0 vs.2 + multiple Marginalized identitiesa	− 1.139	0.0658	0.000*	1.512	0.8172	0.064	− 0.054	0.7546	0.943
1 vs.2 + multiple marginalized identitiesa	− 0.206	0.0836	0.014*	0.551	0.6416	0.390	− 0.875	0.5595	0.118
0 vs.2 + multiple marginalized identitiesp	− 0.160	0.0831	0.054*	2.25	0.815	0.006*	1.486	0.7214	0.039*
1 vs.2 + multiple marginalized identitiesp	0.032	0.0793	0.691	1.095	0.6574	0.096	0.707	0.5776	0.221
Average frequent daily discriminationp				0.667	0.3878	0.085	0.829	0.3334	0.013*
Indirects (b)
Average frequent daily discriminationa				2.488	0.3864	0.000*	2.443	0.3076	0.000*
				B	SE (B)	CI	B	SE (B)	CI
Total indirect effect (ab)
0 vs.2 + multiple marginalized identitiesa				− 2.816	0.347	[− 3.47, − 2.106]	− 2.768	0.265	[− 3.273, − 2.231]
1 vs.2 + multiple marginalized identitiesa				− 0.490	0.167	[− 0.789, − 0.125]	− 0.485	0.169	[− 0.786, − 0.12]
0 vs.2 + multiple marginalized identitiesp				− 0.376	0.174	[− 0.679, 0.009]	− 0.373	0.175	[− 0.680, 0.008]
1 vs.2 + multiple marginalized identitiesp				0.101	0.210	[− 0.244, 0.571]	0.095	0.203	[− 0.251, 0.541]

a actor effect, p partner effect

Frequent daily discrimination mediated the relationship between those with two or more marginalized identities versus none and depression (B =  − 2.816, CI − 3.470, − 2.106]) and stress (B =  − 2.768, CI [− 3.273, − 2.231]). Frequent daily discrimination mediated the relationship between those with two or more marginalized identities versus one and depression (B =  − 0.490, CI [− 0.789,-0.125]) and stress (B =  − 0.485, CI [− 0.786, − 0.120]).

No direct effects were observed for study participants with two or more multiple marginalized identities versus none and depression (B = 1.512, p = 0.064) or stress (B =  − 0.054, p = 0.943). No direct effects were observed for study participants with two or more marginalized identities versus one and depression (B = 0.551, p = 0.390) and stress (B =  − 0.875, p = 0.118).

Aim 2 (Partner Effect Mediation Model)

Study participants whose partners had two or more marginalized identities reported marginally significantly higher personal frequent daily discrimination than those with partners with no marginalized identities (B =  − 0.160, p = 0.054). There was no indirect effect for participants whose partners had two or more marginalized identities verses one on personal frequent daily discrimination (B = 0.032, p = 0.691).

Partner’s frequent daily discrimination did not predict actor’s depression (B = 0.667, p = 0.085). However, the greater the partner’s frequent daily discrimination experiences, the greater the person’s perceived stress (B = 0.829, p = 0.013).

Frequent daily discrimination did not mediate the relationship between partner’s multiple marginalized identities and actor’s reported depression and stress symptoms (Table 2). However, there was a direct relationship between having a partner with two or more marginalized identities versus no marginalized identities and depression (B = 2.25, p = 0.006) and stress (B = 1.486, p = 0.039).

Aim 3 (Moderation)

Gender moderated the actor mediation model for depression but not stress. The indirect effect between those with two or more marginalized identities versus none and depression through frequent daily discrimination was stronger for women (B =  − 4.337, CI [− 5.349, − 3.238]) than men (B =  − 1.947, CI [− 2.723, − 1.086]). Similarly, the indirect effect between those with two or more versus one marginalized identities and depression through frequent daily discrimination was nearly two times greater for women (B =  − 0.754, CI [− 1.215, − 0.193]) than men (B =  − 0.331, CI [− 0.545, − 0.094]). Gender did not moderate the mediated relationship between multiple marginalized identities and stress (B = 0.936, p = 0.08).

For the partner mediation model (aim 2), gender did not moderate the frequent daily discrimination’s mediated relationship between partner’s marginalized identities and depression(B = 0.861, p = 0.183) or stress (B = . − 0.555, p = 0.304).

Discussion

This research used minority stress and intersectionality theories as frameworks to guide our examination of multiple pathways that discrimination affects mental health for young couples having a baby. In our model, we examined how multiple marginalized identities, through frequent daily discrimination, impacted reported depression and stress by expecting couples. Our study supported the growing literature showing the importance of an intersectional approach (e.g., having multiple marginalized identities on health and well-being) [18, 19, 21, 22]. Prior research examining the relationship between marginalized identities and negative health outcomes has produced mixed results [18, 19, 22].While 42% of our sample reported having multiple marginalized identities, having multiple marginalized identities did not have a direct impact on the mental health of expectant parents. However, there were indirect, mediated associations between multiple marginalized identities and higher levels of depression and stress. Individuals with multiple marginalized identities reported experiencing more discrimination, which was associated with higher levels of depression and stress. These observations support our hypotheses that a complex discrimination model with “multiple pathways” may better describe the relationship with negative health outcomes and should be the focus of future research [22]. Our study also found some gender effects. Though men experience more frequent daily discrimination than women, there were no significant gender differences between having multiple marginalized identities on the frequency of daily discrimination. Thus, having multiple marginalized identities increased frequent daily discrimination regardless of gender. The effects of discrimination were corrosive and detrimental for both genders.

Additionally, there were no differences between women and men in the levels of stress reported due to frequent daily discrimination. Thus, men and women reported the same amount of stress due to frequent daily discrimination. We found that women, when compared to men, who experienced more frequent daily discrimination reported significantly more depression. This observation is important as we consider the gendered impacts of discrimination and its potentially negative effects on mental health for the mother and her unborn child. Depression is linked to premature birth, low birth rates, and is the strongest predictor of postnatal depression [40].

Frequent daily discrimination mediated or explained more of the relationship between multiple marginalized identities and depression for women than men. Having multiple marginalized identities increased reported frequent daily discrimination, which was negatively associated with women’s depression when compared to men. Although gender moderated the mediated relationship for depression, gender did not moderate the mediated relationship for stress. These observations may be explained by extant literature that suggests that women’s depression may be the result of their internalizing and ruminating responses to frequent discrimination [41]. Teaching cognitive skills to manage these stressors may be important in reducing the negative antenatal health outcomes experienced by pregnant women of color during and following childbirth.

Finally, having a partner who experienced frequent daily discrimination increased stress for both genders. These observations may be the result of the partners’ stress being transferrable, either through an empathetic reaction, and or direct transference of stress via social interaction [42]. This observation also calls for research that provides both individuals involved in a pregnancy with strategies to manage the stress associated with discrimination and tools to minimize the impacts, personally and in the relationship.

Although we did not observe a direct effect for personal multiple marginalized identities on depression or stress, we did find a direct effect between partners’ multiple marginalized identities and depression and stress. This suggests that having a partner with multiple marginalized identities may affect one’s depression and stress for both genders, which can have negative implications for the pregnant female’s health and speaks to the corrosive nature of the experience.

Our findings suggest a couples-based antenatal intervention that targets individuals with multiply marginalized identities may improve the mothers and their partner’s mental health. Successful interventions that reduce stress like yoga or creative arts programs have traditionally focused on women [43]. Given the effectiveness of these strategies, there are questions about the increased potency of such approaches if tailored to the couple during the antenatal period and calls for other gendered support strategies. Using dancing, storytelling, and yoga, with education, may help couples cope with discrimination and instill life and relationship skills that may extend beyond the antenatal period. These strategies may be incorporated into recent successful couple-based interventions that focus on education through emotional self-management, communication, and mutual support strategies [44].

A major strength of our study is its large, young, and primarily, minority sample. Our observations may have implications for other populations (e.g., those who live in a rural setting and interracial households). Observations from this study may also have applicability for other ethnically and racially minority groups with additional markers of marginalization (e.g., transgender women and men). In this study, roughly 28% of our participants, across all genders, selected other as the reason for their marginalization. Future studies should include a broader list of marginalized identities (e.g., ableism, cultural discrimination).

While this research has several strengths, there are several limitations. This research relied on study subject recall. In this recall, participants might have underestimated their discrimination experiences. Another limitation is that we did not have the sample size to explore specific marginalized identities (e.g., Black woman). Future studies could examine differences in the results by specific multiple marginalized identities. Finally, the study used cross-sectional data and does not provide any evidence of causal pathways. Future studies should explore longitudinal examination of the model. Further, one could test if the model is invariant over time.

The observations from this study lend support to minority stress and intersectionality theories. Future studies are needed that build on these observations. For example, studies could consider other moderators like social support and neighborhood problems/cohesion, factors that have been shown to impact negatively mental health outcomes in young, expectant couples [11]. Future studies could also consider how one’s partners’ experience of discrimination is impacted by pathways like social support, coping, and relationship functioning. We strongly suggest that future studies be conducted using a diverse and multiply marginalized population(s) to confirm whether the model is invariant across time and populations.

Other suggested mediation pathways for future studies could include stigma and racial identity development. Frequent daily discrimination could be redefined as overt versus subtle discrimination to see if the type of discrimination experienced have a different impact on health for multiple marginalized individuals.

Lastly, the psychosocial, cross-over stress of frequent daily discrimination on personal depression and levels of stress could be the result of vicarious racism. Vicarious racism is defined as “the secondhand exposure to racial discrimination and/or prejudice directed at another individual” [45]. For a primarily non-minority, older population, Wofford [46] argued that vicarious racism explained the association between partners’ everyday discrimination and negative health outcomes, depression, and relationship strain. The effects were mediated by relationship strain. However, vicarious racism requires the secondhand exposed victim to be cognizant of the discrimination. Neither the Wofford [46]study nor our study asked about knowledge and disclosure of partner’s negative discrimination experiences. Future studies should include the cognitive measures of partner’s negative discrimination experiences as a potential moderator on personal health.

Declarations

The first author was neither involved in the recruitment of study participants nor in obtaining their informed consent.