Introduction

In the United States (US), first- and second-generation immigrant individuals, the majority of whom are racially/ethnically minoritized (REM), represent approximately 14% and 12% of the population, respectively [1, 2]. Research conducted over the last 15 years demonstrates the lifetime prevalence of depression and anxiety disorders in this population to be approximately 15% and 10%, respectively [3, 4]. While REM immigrant populations have lower rates of mental health disorders relative to non-immigrants (both non- and REM) [3], they face similar racial/ethnic disparities in access to quality mental health treatment to those of non-immigrant US REM groups [5]. REM immigrant individuals also experience racial/ethnic discrimination like that of non-immigrant REM individuals [6], and since 2015, anti-immigrant sentiment specifically has experienced an upward trend [7], recently exacerbated by the COVID-19 pandemic [8]. Racial/ethnic discrimination is correlated with depression/anxiety symptomology [9,10, 10 11] that in extreme cases lead to clinical disorders [12]. Therefore, it is essential to study the mechanisms explaining the connection between discrimination and mental health among REM immigrant individuals to effectively tailor mental health services to their needs. One often-examined mechanism is cultural identity and more specifically how cultural identity is shaped by experiencing discrimination [13].

TheoreticalFramework: Ethnic/National Identity and Rejection (Dis)Identification Models

REM immigrant individuals exist within a multicultural environment and tend to be bicultural, with ethnic (heritage) and national (host) cultural identities [14]. Ethnic identity is one’s identification with, sense of belonging to, and attitudes toward an ethnic group [15]. Feelings of belongingness/identification with the host society are captured by national (host) identity [16], which for REM immigrant people exists on a separate continuum from their ethnic identity [17]. As minoritized individuals, REM immigrants’ experiences in the host country and their cultural identities [18] are highly influenced by their experiences with majority group members [13]. The rejection-identification and rejection-disidentification (RIM/RDIM) models posit that racial/ethnic discrimination leads to increased minority group and decreased superordinate group (i.e., host national) identification, respectively, among minoritized group members [19, 20]. According to RIM/RDIM, discrimination is a form of social exclusion that can deprive individuals of their need to belong. To buffer these negative outcomes associated with experiencing discrimination, individuals identify more with the in-group and less with the broader out-group [21, 22]. In support of RIM, Armenta and Hunt [23] found that ethnic discrimination related to improved psychological wellbeing through ethnic identity among Hispanic adolescents. The findings for RDIM are equivocal but disidentification from the host society seems to be associated with poorer mental health [24, 25]. For example, ethnic discrimination related to feelings of depression via weaker host national identity in a sample of immigrant adolescents [26]. Furthermore, racial/ethnic discrimination is associated with having a stronger ethnic than host national identity [27].

To our knowledge, no studies have examined whether both processes together explain the relation between racial/ethnic discrimination and mental health. That is, as an individual experiences more racial/ethnic discrimination and their ethnic identity becomes stronger while their national (host) identity becomes weaker, what is the role of this identity discordance in mental health? Does this cultural identity discordance reduce bicultural identity harmony—a factor associated with higher depression/anxiety symptoms among REM immigrant individuals [28, 29]?

Bicultural Identity Harmony

For immigrant individuals, the interplay between ethnic and national (host) identities, regardless of their strength, is captured by bicultural identity integration (i.e., BII). One aspect of BII is bicultural identity harmony, or the degree of perceived compatibility between the two identities [30]. Individuals who perceive harmony between their cultural identities usually have fewer depression/anxiety symptoms than individuals who perceive their identities as conflicting [28, 29, 31]. Studies suggest that stronger ethnic and national identities relate to greater bicultural identity harmony, and the latter may explain why stronger ethnic and national identities are associated with fewer depression/anxiety symptoms [28, 29, 32, 33]. Huynh and colleagues [28] found that ethnic identity related to higher bicultural identity harmony, which in turn related to fewer depression symptoms in REM immigrant college students. Other studies also reveal a negative relation between racial/ethnic discrimination and bicultural identity harmony [34, 35] that may help explain why racial/ethnic discrimination is associated with higher depression symptoms [28].

The Current Study

While research documents associations between racial/ethnic discrimination, ethnic/national identity, bicultural identity harmony, and symptoms of depression/anxiety, no study has examined these factors together. Such a study is necessary as it can illuminate how the simultaneous divergent shifts in ethnic/national identities associated with racial/ethnic discrimination relate to mental health among REM immigrant individuals. Supported by previous studies [32, 33, 35, 36], our conceptual framework (Fig. 1) proposes that the relation between racial/ethnic discrimination and depression/anxiety symptoms is mediated sequentially by ethnic/national identity discordance and bicultural identity harmony. Further, in accordance with RIM/RDIM [27], we hypothesize that the path from racial/ethnic discrimination to cultural identity discordance will only be significant among those who report stronger ethnic than national identity. Lastly, we hypothesize negative relations between racial/ethnic discrimination and bicultural identity harmony and cultural identity discordance and depression/anxiety symptomatology.

Fig. 1
figure 1

Conceptual Model for the Relations Between Racial/ethnic Disciscrimination, Cultural Identity Discordance, Bicultural Identity Harmony, and Symptoms of Depression/Anxiety. Note: +/- signs represent hypothesized directions of associations between pairs of variables

Full size image

Method

Participants

Data for this secondary data analysis originated from a larger study on psychosis risk conducted at an urban and racially/ethnically diverse university in the Northeast. The inclusion criteria were: being between 18 and 29 years old, identifying as Black/African American and/or being born abroad or having at least one parent born abroad. Data were collected between September 2014 and May 2018 and 1053 individuals participated. Only participants who were first- or second-generation immigrant and REM were included herein (N = 877). All participants filled out self-report measures on a laboratory computer for course credit. The university’s IRB approved the study. Two participants had missing data and were excluded from analysis.

Measures

Racial/Ethnic Discrimination

 Participants completed the Experience of Discrimination scale (EOD; [37]), which assessed whether respondents ever experienced discrimination due to their race/ethnicity/color in nine domains (e.g., Work). For each domain endorsed, participants indicated the frequency using a 3-point scale: 1 (once), 2 (two to three times), 3 (four or more times). The scale has demonstrated good reliability/validity in REM samples [37], and the Cronbach’s alpha for the current sample was 0.66. The frequencies from each domain were aggregated into a score (0–27), with higher values indicating more lifetime experiences.

Bicultural Identity Harmony

We used the 10-item bicultural identity harmony subscale of the Bicultural Identity Integration Scale-Version 2 (BIIS-2; [28]). Respondents rated each item using a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree). A sample item is “I rarely feel conflicted about being bicultural.” Responses were averaged into a scale score, with higher scores indicating more bicultural identity harmony. The scale has previously shown convergent/discriminant validity with other measures of cultural identity and acculturation [28]. The Cronbach’s alpha estimate for the current sample was 0.82.

Ethnic Identity

We used the Multigroup Ethnic Identity Measure-Revised (MEIM-R; [15]), a 6-item questionnaire, to assess identification with an ethnic group. Participants rated each item on a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree). A sample item is “I have a strong sense of belonging to my own ethnic group.” We averaged responses into a scale score, with higher scores indicating a stronger ethnic identity. The scale has demonstrated good internal consistency in research with diverse student populations [15, 38]. The Cronbach’s alpha estimate for the current sample was 0.86.

American Identity

We used the American Identity Measure (AIM; [39]), a 12-item questionnaire, to assess identification with the US. Respondents rated each item using a Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). A sample item is “I have a lot of pride in the United States.” We averaged the responses, with higher scores corresponding to stronger American identity. In prior research, the scale demonstrated good internal consistency and associations with other measures of American identification/culture [39]. For this sample, the Cronbach’s alpha estimate was 0.87.

Cultural Identity Discordance and Type of Discordance

 We computed a cultural identity discordance score by taking the absolute value of the difference between each participant’s ethnic (i.e., MEIM-R) and American identity (i.e., AIM) scores. To capture cultural identity discordance type, we created a three-level variable comparing each participant’s ethnic/national identity scores (Table 1): level 1 included participants with higher ethnic than American identity; level 2 those with higher American than ethnic identity; and level 3 those who scored equally on both.

Table 1 Sample Demographic Characteristics (N = 877)
Full size table

Depression Symptoms

The 10-item Center for Epidemiologic Studies-Depression Scale (CES-D; [40]) assessed depression symptoms experienced in the previous week. Respondents rated items using a four-point Likert-type scale from 0 (rarely or none of the time) to 3 (all of the time). A sample item is “I felt that everything I did was an effort.” Scores range from 0 to 30, with a score of 10 or greater indicating potential clinical depression [40]. Previous research has demonstrated the validity of this scale among ethnically diverse populations [41]. In the current sample, the Cronbach’s alpha estimate was 0.81.

Anxiety Symptoms

 The short version of the State-Trait Anxiety Inventory-Trait Form-Anxiety Subscale (STAI-Trait; [42]), a 7-item questionnaire, measured current (i.e., how one “generally” feels) anxiety symptoms. A sample item from the scale is “I feel nervous and restless.” Items were presented in Likert scale, from 1 (almost never) to 4 (almost always). Aggregate scores range from 4 to 28, with scores of 16 or greater indicating potential clinical anxiety [43]. The measure demonstrated good reliability in studies with ethnically/racial diverse samples [44], the Cronbach’s alpha estimate here was 0.85.

Sociodemographic Characteristics

 Participants provided their age (years), sex at birth (male/female), family income before taxes, race/ethnicity, and immigrant status. We assessed race/ethnicity by asking respondents to select a race/ethnicity category from a list and then grouped the responses into a four-category variable: (1) Black, (2) Hispanic, (3) Asian, and (4) other. To determine immigrant status, we used participants’ and their parents’ places of birth. We classified participants born outside of the US as first-generation and those born in the US, but with at least one foreign-born parent, as second-generation immigrants.

Data Analysis

We examined bivariate relations among continuous variables using Pearson correlations. To test our conceptual model, we estimated two moderated-mediation models, one for each outcome. We obtained indirect effects using 10,000 bootstrapped replications, yielding 95% bias corrected confidence intervals. We assessed the extent to which the indirect effect via cultural identity discordance varied across discordance type (e.g., higher ethnic than American identity) with moderated mediation models. The models controlled for age, family income, sex, generation, and race/ethnicity, with male, first-generation, and Asian as reference categories. Prior to conducting all analyses, we confirmed that no variable had excessive skewness or kurtosis [45]. We verified linearity in the relation between discrimination and anxiety/depression symptoms using curve estimation. We ruled out multicollinearity because variance inflation factors were less than 2. To account for any bias due to violations of equality of variance assumptions, we computed standard errors using a heteroscedastic consistent estimator [46]. We conducted all analyses in SPSS 27, using the PROCESS macro (Model 83) to test moderated-mediation [47].

Results

Sample Characteristics and Bivariate Associations

Tables 1 and 2 present sample demographic characteristics and bivariate correlations for study variables, respectively. The sample was 62.1% female, average age was 19.86 (SD = 2.40), and 54.5% reported annual family income of less than $35,000. The racial/ethnic background of the participants was 19.4% Black, 35.2% Hispanic, 35.2% Asian, 10.2% other. The latter included multi-racial individuals. On average, participants reported 3–4 experiences of racial/ethnic discrimination, with 27.6% reporting none. Most of the sample reported stronger ethnic than American identity. The sample average scores were below the cut-offs for clinical levels of depression or anxiety [40, 43]. Discrimination was positively related to depression/anxiety symptoms and cultural identity discordance, and negatively related to bicultural identity harmony. Bicultural identity harmony was positively related to American identity and negatively to depression/anxiety symptoms.

Table 2 Pearson correlations
Full size table

Moderated Mediation Analyses

Standardized regression estimates for the moderated mediation models are depicted in Fig. 2 for depression and Fig. 3 for anxiety symptoms. Racial/ethnic discrimination was positively associated with cultural identity discordance. Cultural identity discordance was negatively associated with bicultural identity harmony and bicultural identity harmony was negatively associated with depression/anxiety symptoms. Discrimination was positively associated with both depression and anxiety symptoms. As shown, the magnitude of these associations was small.

Fig. 2
figure 2

 Regression coffiecients for regression model with depression symptoms as the outcome variable. Unstandardized coefficients (B) are given first followed by standard errors (SE) in parentheses and standardized coefficients (β) in brackets; ** p < 0.01; * p < 0.05; Interaction 1; Racial/Ethnic Discrimination × Contrast 1[Cultural Identity Discordance Type values 0 (Ethnic Identity > American Identity) and 2 (Ethnic Identity = American Identity) are compared with value 1 (Ethnic Identity < American Identity)]; Interaction 2; Racial/Ethnic Discrimination × Contrast 2 [Cultural Identity Discordance Type values 0 (Ethnic Identity > American Identity) and 1 (Ethnic Identity < American Identity) are compared with value 2 (Ethnic Identity = American Identity)]

Full size image
Fig. 3
figure 3

Regression coffiecients for regression model with anxiety symptoms as the outcome variable. Unstandardized coefficients (B) are given first followed by standard errors (SE) in parentheses and standardized coefficients (β) in brackets; ** p < 0.01; * p < 0.05; Interaction 1; Racial/Ethnic Discrimination x Contrast 1[Cultural Identity Discordance Type values 0 (Ethnic Identity > American Identity) and 2 (Ethnic Identity = American Identity) are compared with value 1 (Ethnic Identity < American Identity)]; Interaction 2; Racial/Ethnic Discrimination x Contrast 2 [Cultural Identity Discordance Type values 0 (Ethnic Identity > American Identity) and 1 (Ethnic Identity < American Identity) are compared with value 2 (Ethnic Identity = American Identity)]

Full size image

Table 3. presents indirect effects and indices of moderated mediation for each outcome. According to indices of moderated mediation, the relation between discrimination and cultural identity discordance was only significant for individuals with a stronger ethnic than American identity (Fig. 4). The serial mediation effect via cultural identity discordance and bicultural identity harmony was significant and accounted for 1.3% of the total effect on depression and 1.7% on anxiety. Discrimination was also indirectly associated with depression and anxiety symptoms through bicultural identity harmony, which mediated 9.9% and 12.7% of the total effects, respectively.

Table 3 Indirect effects and indices of moderated mediation for regressions predicting depression and anxiety symptoms
Full size table
Fig. 4
figure 4

Cultural identity discordance as a function of racial/ethnic discrimination and cultural identity discordance type. Ethnic identity = Amerian identity cultural identitiy discordance type is omitted as it is redudandant

Full size image

Discussion

We tested a theoretical model that proposes cultural identity discordance as a mechanism that partly explains the connection between racial/ethnic discrimination and increased depression and anxiety symptoms among US first- and second-generation REM immigrant individuals. Our findings indicate that racial/ethnic discrimination was positively associated with cultural identity discordance, but only for individuals who had a stronger ethnic than American (i.e., national) identity. Additionally, only ethnic (but not American) identity correlated with racial/ethnic discrimination. Combined, these findings suggest that consistent with RIM, racial/ethnic discrimination relates to higher ethnic identity, but contrary to RDIM, it may not relate to lower host national identity. Thus, the discordance between ethnic and host national identities associated with racial/ethnic discrimination may have stemmed from a burgeoning ethnic and static national identities. This discordance appears to be important as it was negatively associated with bicultural identity harmony, indicating that as ethnic and national identities diverge, the more conflict one feels among them. These findings may suggest that as host national identity strengthens and mirrors the strength of ethnic identity, regardless of whether these are high or low, the two identities are perceived as more harmonious.

Consistent with previous research [28, 29, 31], bicultural identity harmony was negatively associated with depression/anxiety symptoms. Importantly, this study shows that some of the conflict between one’s cultural identities may be associated with the discordance in them linked with experiences of racial/ethnic discrimination. However, this cultural identity discordance was only relevant to depression/anxiety symptoms via its association with bicultural identity harmony. Furthermore, beyond the serial mediated pathway proposed, racial/ethnic discrimination directly related to bicultural identity conflict and depression/anxiety symptoms, and consistent with previous research [28, 35], bicultural identity harmony mediated the relation between discrimination and depression/anxiety symptoms.

This study highlights the interplay between ethnic/national identities and bicultural identity harmony in potentially explaining the association between experiences of racial/ethnic discrimination and depression/anxiety symptoms among REM immigrants. While the current study further cements the utility of bicultural identity harmony to this relation, it also provides evidence for the need to examine ethnic and national identity discordance. In the current study, neither ethnic nor national identity were directly associated with both racial/ethnic discrimination and depression/anxiety symptoms. The discordance between these identities was associated with both discrimination (directly) and depression/anxiety symptoms (indirectly via bicultural identity harmony). Thus, researchers examining these factors should include ethnic and host national identity discordance as key variables in their immigrant mental health models, a factor missing from extant research. Similarly, mental health practitioners working with REM immigrant clients experiencing depression/anxiety symptoms may identify difficulties with bicultural identity negotiation, specifically the conflicting experience of shifting cultural alliances associated with racial/ethnic discrimination.

Limitations and Future Directions

The study relied on cross-sectional data, which precludes us from making any causal statements. However, several of the component parts of the model are supported by longitudinal research including time-one racial/ethnic discrimination predicting time-two ethnic [48] and national identities [20], and bicultural identity harmony mediating the relation between time-one ethnic/national identities and time-two mental health outcomes [33]. Thus, experience of discrimination is an important factor to consider when understanding ethnic/national identities, which in turn are important for bicultural identity harmony, and consequently mental health. Nevertheless, future research is necessary to fully validate the directionality of the models tested here.

Next, while our mediation models were significant, the effect sizes were small. Thus, other factors may further explain the connection between racial/ethnic discrimination and depression/anxiety symptoms in REM immigrants (e.g., citizenship/residency status not captured here). Notwithstanding, most of the current sample reported two or fewer experiences of discrimination. Thus, the small magnitude of the mediating effect may at least be partially attributed to the relative lack of discrimination experienced by the participants. The sample consisted of college students attending a minority serving institution. As the racial/ethnic discrimination scale used asks participants to report discrete experiences of discrimination rather than overall perceptions of the treatment experienced by respondent’s race/ethnic group, participants in this study may have had fewer such experiences due to being in a predominantly minority environment. Similarly, the frequency of discriminatory experiences is likely to increase over time. Therefore, our sample may have fewer experiences of discrimination relative to older individuals, particularly in domains not frequently relevant to college students (e.g., obtaining bank loans). It is plausible that the effects found herein may be more pronounced in REM immigrant populations whose environments might place them at higher risk of discrimination. Future research using more diverse, community-based samples is necessary to further verify the magnitude of the effects we found.

New Contribution to the Literature

The current study presents new insights into the role played by divergences in ethnic and national identities as a mechanism linking racial/ethnic discrimination to mental health among REM immigrant individuals. As REM immigrant individuals engage with their ethnic side to cope with discrimination (i.e., rejection identification), they may feel like this aspect of their identity is overshadowing the side of themselves that feels connected to the society in which they reside (i.e., cultural identity discordance). This imbalance may result in feeling conflicted about belonging both to a group that is mistreated and a society whose members carry out the mistreatment (i.e., bicultural identity conflict). Clinicians should assist their REM immigrant clients in navigating identity challenges to cope with depression/anxiety symptoms related to discrimination.