Types of ethnic–racial identity belongingness predicting mental health and suicide ideation: Exploring racial and gender differences among Latine‐White and Black‐White Multiracial adolescents Authors: Chelsea Derlan Williams, N. Keita Christophe, Annabelle L. Atkin, Kimberly Lazarus, McKenzie N. Green, Troy J. Keys, B. Andi Lee, Nilee Kerr, Diamond Y. Bravo, Arlenis Santana Published: 2026 ABSTRACT AbstractThe current study examined whether three types of ethnic–racial identity (ERI) belongingness (i.e., minoritized identity belongingness, White identity belongingness, and Multiracial identity belongingness) were associated with anxiety symptoms, depressive symptoms, and suicide ideation among Multiracial adolescents. We also explored race‐gender group differences between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. Participants were 583 adolescents 13–17 years of age (M = 15.25, SD = 1.38). Findings indicated that youth had relatively high belongingness to their minoritized, White, and Multiracial identities, but there were significant mean differences among the race‐gender groups. Further, minoritized identity belongingness was associated with less depressive symptoms and suicide ideation among all youth, and less anxiety symptoms for all youth except Latino‐White males. Interestingly, Multiracial identity belongingness was associated with less anxiety symptoms for only Latino‐White males but was not significant for the other three race‐gender groups. Multiracial identity belongingness was also associated with less depressive symptoms for all youth but not associated with suicide ideation for any youth. White identity belongingness was not associated with outcomes across youth. Findings highlight the importance of focusing on nuances in ERI and impacts on Multiracial youths' mental health and suicide ideation, while also considering racial and gender group differences. Multiracial individuals comprise 10.2% of the population in the United States and have increased from 9 million to 33.8 million individuals in the last 10 years (Jones et al., 2021). Also, 32.5% of Multiracial individuals in the United States are also under the age of 18, making the Multiracial population the youngest of any racial group (Rico et al., 2023). It is essential to focus on Multiracial adolescents' unique experiences and outcomes in order to understand normative development for this growing population in the United States. Multiracial youth face an increased risk of depression, anxiety, and suicidal ideation compared with monoracial youth (Bolland et al., 2007; Cheng & Lively, 2009; Garcia et al., 2019; Udry et al., 2003; Williams et al., in press). A systematic review of research on internalizing symptoms among Multiracial adolescents found that between 7.5% and 55.2% reported internalizing symptoms, suggesting that depression and anxiety rates may be higher in this population than in the general adolescent population (Henson‐García et al., 2025). Compared with both monoracial minority and White adolescents, Multiracial youth report higher levels of loneliness, suicidal ideation, and extreme sadness and hopelessness (Garcia et al., 2019). Given mental health disparities among Multiracial adolescents, it is important to focus on factors that may predict these outcomes; however, existing work has primarily centered on testing how one's self‐reported ethnic–racial label impacts mental health outcomes. Beyond solely ethnic–racial label, the attitudes and behaviors individuals have about their ethnic–racial backgrounds that compose their ethnic–racial identity may be more informative (ERI; Williams et al., 2020). The process of navigating multiple identities can create challenges related to belongingness and identity formation for Multiracial youth (Garcia et al., 2019; Cooney & Radina, 2000; Udry et al., 2003). However, a positive ERI has been found to predict better mental health outcomes among both monoracial (Rivas‐Drake et al., 2014) and Multiracial youth (Green et al., 2024), making it an important predictor to examine. Despite work on other components of ERI, limited research has focused on ERI belongingness (i.e., individuals' sense of attachment and connection to their ethnicity/race/culture; Lee et al., 2022) and its impacts on mental health. To address gaps, the current study tested whether three types of ERI belongingness (i.e., belongingness to one's minoritized identity, belongingness to one's White identity, and belongingness to one's Multiracial identity) were associated with anxiety symptoms, depressive symptoms, and suicide ideation. We also explored whether there were race‐gender group differences in associations between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. Adolescence as a critical period for ERI belongingness During adolescence, youth make meaning of their ethnic and racial identities in ways that are informed by advancing cognitive capacities characterized by abstract reasoning and social perspective‐taking, through which they can increasingly interpret racial and ethnic experiences within the context of broader societal hierarchies and historical narratives (Quintana, 1994). Through identity‐relevant experiences (Williams et al., 2020) and contextual influences (e.g., schools, families), adolescents have spaces to explore their identities, create deeper affiliations with others (e.g., peers, teachers, siblings), and develop loyalties, shared values, beliefs, and attitudes with ingroup members. Together, these opportunities for ERI development have implications for improving adolescents' psychosocial, academic, physical, and mental health (Rivas‐Drake et al., 2014). One particular facet of ERI that remains understudied is an individual's sense of belonging to their ethnicity/race/culture (Hunter et al., 2019; Neville et al., 2014). Belongingness has been identified as a “lost” component of ERI (Lee et al., 2022; Williams et al., 2020), given that it was included in earlier conceptualizations (e.g., Phinney, 1992), and mentioned as a necessary element of ERI development in most racial and ethnic identity theories, but has only recently begun to be studied and defined as its own construct. According to the lifespan model of ERI (Williams et al., 2020), there are five dimensions of ERI: ethnic–racial awareness, affiliation, attitudes, and behaviors. The affiliation dimension captures individuals' sense of membership in or belongingness to one or more ethnic–racial groups, and it is posited that belongingness is a component within this broader dimension. ERI scholars (Umaña‐Taylor et al., 2014) have also highlighted that ERI components reflect content (attitudes and/or beliefs about one's group), and process (mechanisms involved in exploring and maintaining ERI). The affiliation dimension of ERI, and belongingness specifically, reflects content of ERI. In a grounded theory study of ERI belonging, or racial–ethnic‐cultural belonging among students of color, it was defined as a sense of “home,” or a recognition and return to comfort and kinship with their own ERI group, which includes three interrelated dimensions of connection, acceptance, and authenticity (Lee et al., 2025). A subsequent measure of ERI belonging, or racial–ethnic‐cultural belonging, created by and for people of color (Lee & Neville, 2024), found that greater ERI belonging was positively associated with general belonging, ERI, and general well‐being, and negatively associated with cultural homelessness (i.e., feeling a lack of group membership and/or cultural home; Navarrete & Jenkins, 2011) and loneliness. To date, most of the literature on belongingness for adolescents has centered on place‐based contexts, such as classrooms (Boston & Warren, 2017; Goodenow, 1993; Hussain et al., 2018) and universities (Hurtado & Carter, 1997; Strayhorn, 2009). Thus, we have a good sense of the impacts of students' feelings of belonging to their classrooms or broader university settings, but have less clarity on their sense of belonging to their ethnicity/race/culture. By studying ERI belongingness, we can therefore expand upon place‐based understandings to social identity‐based contexts (Lee et al., 2025). Among Multiracial youth, belongingness may be particularly nuanced because they may have a sense of belonging (or not) with multiple identities. For Multiracial individuals, the experience of belonging differs from monoracial individuals due to unique challenges faced with navigating identity and social contexts (Eid & Parker, 2023), such as rejection or denial of their racial identity (e.g., racial identity invalidation; Green et al., 2024), feelings of not belonging or cultural homelessness (Chaudhari, 2022), multiple forms of discrimination based on their monoracial identities and Multiracial identity experiences from non‐familial and familial individuals (Christophe et al., 2024), and potential internal confusion about their racial identity (Green et al., 2026). These challenges stem from monoracial and essentialist perspectives of race in the United States, which complicate the social perception and acceptance of Multiracial people (Harris, 2016). However, limited work has examined ERI belongingness among Multiracial individuals. Theoretical frameworks Existing theory on Multiracial individuals' ERI has tended to focus on identifying the components involved in ERI and how to measure it (e.g., Gabriel et al., 2023; Rockquemore & Brunsma, 2002) and/or the factors that predict ERI (e.g., Gonzales‐Backen, 2013), and no theories to our knowledge posit how Multiracial youths' ERI is expected to predict mental health or suicide outcomes. In fact, in conceptualizing a theory of ERI across the lifespan, scholars highlight that one particular persistent area of concern is a need for more theory and research regarding Multiracial individuals' ERI development (Rogers et al., 2020; Williams et al., 2020), including conceptualization, factors that predict it, and outcomes associated with it. Although not focused specifically on Multiracial youth, theoretical frameworks that are useful for understanding predictive associations between ERI belongingness, mental health and suicide ideation are the Belongingness Hypothesis (Baumeister & Leary, 1995) and the Interpersonal Theory of Suicide (Van Orden et al., 2010). The Belongingness Hypothesis (Baumeister & Leary, 1995) posits that individuals' need to belong, which authors define as a desire to form and maintain connection with others, is universal among humans and prevents poor effects on well‐being. Relatedly, the Interpersonal Theory of Suicide (Van Orden et al., 2010) posits that lack of belongingness is a proximal predictor of suicide risk. Neither theory focuses specifically on belonging to one's ethnic–racial backgrounds, but both suggest that belongingness predicts better well‐being (e.g., mental health) and reduces suicide risk (e.g., ideation). Accordingly, guided by the Belongingness Hypothesis (Baumeister & Leary, 1995) and the Interpersonal Theory of Suicide (Van Orden et al., 2010), it is possible that Multiracial youths' ERI belongingness to their minoritized, White, and Multiracial identities may predict lower anxiety symptoms, depressive symptoms, and suicide ideation. ERI belongingness predicting anxiety symptoms, depressive symptoms, & suicide ideation ERI belongingness predicting anxiety & depressive symptoms Extensive research has examined the role of ERI in predicting psychological symptoms among monoracial individuals. Studies utilizing both cross‐sectional and longitudinal data have found higher ERI to be associated with mental health promotion and reduced psychological symptoms (Mandara et al., 2009; Walker et al., 2024). For example, Williams et al. (2012) found that higher scores on the Multigroup Ethnic Identity Measure (MEIM), which was a composite of individuals' feelings of affirmation and belonging, ethnic identity achievement, and ethnic behaviors, were associated with reduced symptoms of depression and anxiety among Black adults. Similarly, among Black emerging adults, greater ERI exploration and affirmation were associated with less depressive and anxiety symptoms (Byer‐Tyre et al., 2024). Among Latine adolescents, results from a longitudinal study indicated that higher ERI centrality, private regard, and public regard predicted fewer depressive symptoms (Cross et al., 2018). While the majority of research has focused on monoracial youth, less is known about the predictive role of ERI among Multiracial youth. In fact, a recent scoping review focused on existing work on Multiracial child and adolescent mental health and found that only four studies have been conducted that tested how components of ERI predicted mental health (Williams et al., in press). This limited work provides support to links between ERI and outcomes. For example, among Multiracial adolescents, a composite of ERI exploration and affirmation predicted greater self‐esteem (Fisher et al., 2017). In another study, ERI invalidation was associated with greater depressive symptoms, while Black pride and Biracial pride were associated with fewer depressive symptoms (Green et al., 2024). However, none of these studies assessed ERI belongingness. Across literature with both monoracial and Multiracial individuals, sense of belongingness is an important ERI dimension to consider, yet rarely assessed as a distinct construct (Lee et al., 2022). Though partially captured in the MEIM's affirmation scale, the combination of items related to belongingness and affirmation make it difficult to understand the distinct contributions of each component to psychological outcomes. Specifically, the MEIM defines affirmation as an individual's positive feelings towards their ethnic–racial group, but scholars have called for examining and measuring ERI identity components separately (e.g., Rivas‐Drake et al., 2014). A recent systematic review underscored the importance of assessment and operationalization of ERI belongingness as more than affirmation, drawing from both quantitative and qualitative studies to further clarify ERI belonging as an individuals' sense of fit in their group, acceptance from that group, and sense of self within a broader social context (Lee et al., 2022). In the limited work that has examined belongingness as a distinct, identity‐related construct, findings have been promising. Among Black LGBTQ adults, stronger community belongingness related to race, gender, and sexual orientation were linked to reduced depression and anxiety symptoms (Watts & Thrasher, 2024). Similarly, among Black emerging adults, individuals' sense of belongingness to their racial group was found to be associated with lower anxious arousal (MacNear & Hunter, 2025). Our study builds on this work by focusing on ERI belongingness among Multiracial adolescents, and includes belongingness to three different ethnic–racial identities (i.e., minoritized, Multiracial, and White identities). Because of their unique position of being members of multiple ethnic–racial groups and a distinctly Multiracial or ‘Mixed’ identity, Multiracial youth may endorse different levels of belongingness across these three identities, and the impact of belongingness on mental health may vary as well. ERI belongingness predicting suicide ideation Existing research on the relation between ERI and suicidal ideation is inconclusive, but studies in this area are, to our knowledge, focused only on monoracial individuals. Among Black women, stronger racial identity (i.e., higher public regard, private regard, and centrality) was associated with lower levels of suicidal ideation (Mekawi et al., 2020). Additionally, Chang et al. (2017) examined the predictive ability of four ERI constructs‐‐ affirmation and belonging, behaviors and practices, achievement, and other‐group orientation‐‐ on suicide risk (i.e., suicidal ideation and attempts), among Latine emerging adults. Of the four ERI components, only other‐group orientation, defined as an individual's attitudes towards other groups, uniquely predicted decreased suicidal behavior. Research, however, on Multiracial youth and suicide has focused predominantly on ethnic–racial differences in suicidal ideation, leaving a critical gap in understanding the role of ERI in this context (Fix et al., 2022; Nelon et al., 2019; Schlabach, 2013). Empirical support for this notion has been mixed. For example, among adult psychiatric outpatients, general sense of belongingness (i.e., feeling valued and accepted by individuals in one's social environment) was not associated with suicide ideation (Fisher et al., 2015). On the other hand, Ploskonka and Servaty‐Seib (2015) found that college students' stronger sense of belongingness to family predicted lower suicidal ideation. Given that Multiracial youth are at higher risk for suicidal ideation compared with their monoracial peers, research must not generalize findings from monoracial populations (Oh et al., 2023). Instead, a deeper understanding of how sense of ERI belonging relates to suicide ideation specifically among Multiracial youth is necessary. Race‐gender group differences In addition to examining how types of belongingness predict mental health and suicide ideation, it is important to examine nuances in associations. In the current study, we explored differences in these relations by individuals' race‐gender group because Multiracial individuals are not a monolithic group, and there are a multitude of factors that vary across Multiracial subgroups (e.g., Black‐Latine, Asian‐White). Findings from the aforementioned review on Multiracial child and adolescent mental health indicated that only five of the 36 articles specified subgroups of Multiracial individuals (Williams et al., in press). Of the 5 studies, only one study (Scott et al., 2021) examined differences between Latine‐White individuals and Black‐White individuals (among other Multiracial subgroups), but did not examine intersections with other identities beyond race/ethnicity (Williams et al., in press). In the current study, we focused on ethnicity/race and gender specifically (i.e., race‐gender group differences) because these two identities are particularly salient in adolescence, and shape the feedback that youth receive from others (e.g., gendered racism, gendered and racialized stereotypes), which may be particularly impactful for sense of belonging. Therefore, we explored whether there were differences in belongingness and outcomes across Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. Given that no theoretical framework or prior empirical work suggests how these specific relations may vary by each race‐gender group, this aspect of the current study was exploratory and we made no specific hypotheses. The current study The goal of the current study was to test whether three types of ERI belongingness (i.e., belongingness to minoritized identity, belongingness to White identity, and belongingness to Multiracial identity) were associated with anxiety symptoms, depressive symptoms, and suicide ideation among Multiracial adolescents. Additionally, we explored whether there were race‐gender group differences in associations between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. Given that ERI varies by age (Williams et al., 2020) and birthplace (Cheon et al., 2018), adolescents' age and birthplace (i.e., born in the U.S. or outside the U.S.) were included as covariates. METHOD Participants and procedure The current study used data from a larger study of 644 Multiracial youths' cultural experiences, health, and development. All data were collected via Qualtrics Research Panels. Qualtrics Research Panels is a division of the survey company, Qualtrics, that works directly with researchers to help them collect data. Through an immense, pre‐curated, and vetted register of individuals from diverse backgrounds who have indicated willingness to participate in surveys and by working with numerous other organizations around the United States (and beyond) that have created their own research panels, Qualtrics Research Panels facilitates the quick and efficient recruitment of research samples (see Brandon et al., 2013, for more information on Qualtrics Research Panels and similar panels); this service is particularly advantageous when the target demographic is small and/or otherwise hard to reach (e.g., our sample of Black‐White and Latine‐White Multiracial adolescents). Adolescents who were between the ages of 13–17, identified as Latine‐White or Black‐White, and who were living in the United States were invited to participate. Adolescents who met inclusion criteria were directed to an online Qualtrics survey. Youth were asked: “To be eligible to participate in this survey, you must be Multiracial, meaning you have biological parents from two or more of the following groups: White, African American/Black, Asian/Asian American, Pacific Islander, Latino(a), American Indian, Middle Eastern/Arab American,” and they could reply “Yes, I have biological parents of two or more different racial backgrounds.,” or “No, I do not have biological parents of two or more different racial backgrounds.” Youth who replied yes then had to have a parent provide informed consent, and after participants provided assent, they were taken to the survey questions. During the consent process, parents could opt out of having their child presented with suicide‐related items; in these cases, this block of items was skipped. All measures were presented to adolescents in a random order to minimize potential order effects. The survey took approximately 30 min to complete and adolescents received compensation directly from Qualtrics. The larger study was approved by the Institutional Review Board. The current study excluded 43 adolescents who were Black‐White‐Latine Multiracial, given our focus on exploring race‐gender differences (i.e., there were not enough individuals in this group to have a large enough sample for statistical analysis). Additionally, we excluded 18 youth whose gender identity included too few youth to be classified into a race‐gender group for analyses. Specifically, the larger sample included 5 youth who identified as gender non‐conforming, 7 who identified as non‐binary, 4 who identified as “not listed: please specify:” and 2 who did not report their gender identity. These 18 youth were excluded, leaving 312 youth who identified as male and 271 who identified as female. Thus, the current study included 583 Multiracial adolescents ages 13–17 (M = 15.25, SD = 1.38) who identified as Black‐White (n = 274) or Latine‐White (n = 309). Specifically, in the demographic questions asked at the beginning of the demographic survey, youth were asked: “What are your racial backgrounds? Please check all that apply.” Response options were “African American/Black, Asian American/Asian (including Indian), Native Hawaiian or Pacific Islander, White/European, Hispanic/Latinx, Native American/ American Indian, Arab/Middle Eastern/North African American, and Not listed, please specify.” We then created a variable with codes to indicate which youth in our sample selected Latiné and White and Black and White. Regarding social class, youth were asked: Which social class group do you identify with (Hughes et al., 2016), and the majority of adolescents identified as middle class (43%), followed by working class (22%), lower middle class (15%), upper middle class (13%), poor (6%), and affluent (1%). The majority of youth reported that they were born in the U.S (93%). Among the 7% of youth who were born outside the U.S., the majority reported that they have lived in the U.S. longer than 11 years (74%), while 26% reported that they have lived in the U.S. between 5 years and 10 years. Measures Ethnic–Racial Identity Belongingness. We used the home subscale from the iBelong Scale (Lee & Neville, 2024). The original measure asks youth to think about their racial–ethnic‐cultural identity and sense of belonging. In the current study, we provided the 5‐item measure to youth three different times. Youth were asked at the beginning of the measure to select their marginalized label, which was then automatically entered in the items about their marginalized identity. Specifically, they were asked “Which marginalized group are you a part of?” The options were Asian, Latine, or Black. Then youth were instructed: “The following questions are about your racial/ethnic/cultural identity and your sense of belonging. First, answer the following questions thinking about the marginalized part (e.g., Asian, Latino, Black) of your racial/ethnic/cultural identity.” A sample item was “I feel close to other [Latino, Black] people.” For these questions, either “Latino” or “Black” was filled in the blank based on which label applied to youth. For example, Latine‐White youth saw “I feel close to other Latino people,” while Black‐White youth saw “I feel close to other Black people.” The second time, youth were prompted to “Now please answer the questions thinking about the White part of your racial/ethnic/cultural identity” and were given the same 5 items now referring to White people (e.g., “I feel close to other white people).” The third time, youth were prompted to “Now please answer the questions thinking about your Multiracial racial/ethnic/cultural identity.” A sample item was “I feel close to other Multiracial people.” Response options were on a 6‐point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree. Higher scores on each assessment indicated higher minoritized identity belongingness, White identity belongingness, and Multiracial identity belongingness, respectively. Support for validity and reliability has been demonstrated for the home subscale from the iBelong Scale among ethnically/racially diverse emerging adults and adults (Lee & Neville, 2024). Cronbach's alphas were .88 for minoritized identity belongingness, .92 for White identity belongingness, and .91 for Multiracial identity belongingness. Anxiety Symptoms. The Symptoms Checklist‐90 (SCL‐90; Derogatis et al., 1973) was used to assess anxiety symptoms. For the current study, we used a shortened version (Overstreet et al., 2017) that used item response theory to reduce the original anxiety subscale to 4 items. Items assess anxiety symptoms (e.g., “nervousness or shakiness inside”) and ask students to describe “how much discomfort that problem has caused you during the last 30 days, including today.” Response options were on a 5‐point Likert scale ranging from 1 = not at all to 5 = extremely. Higher scores indicated greater anxiety symptoms. Cronbach's alpha was 0.90. Depressive Symptoms. The Symptoms Checklist‐90 (SCL‐90; Derogatis et al., 1973) was used to assess depressive symptoms. For the current study, we used a shortened version (Overstreet et al., 2017) that used item response theory to reduce the original anxiety subscale to 4 items. Items assess depressive symptoms (e.g., “feeling no interest in things”), and ask students to describe “how much discomfort that problem has caused you during the last 30 days, including today.” Response options were on a 5‐point Likert scale ranging from 1 = not at all to 5 = extremely. Higher scores indicated greater depressive symptoms. Cronbach's alpha was .89. Suicide Ideation. A single item from the Virginia Youth Survey (Virginia Department of Health, 2019) was adapted to assess youths' suicide ideation. The item was “During the past 12 months, how often have you had thoughts about attempting suicide?” The response options we used in the current study were coded on a 5‐point Likert scale ranging from 0 = never to 4 = most days or every day, and a higher score indicated more frequent suicide ideation. Covariates and race‐gender grouping variable. Students' age and birthplace were included as covariates. Birthplace was coded as 0 = born outside the U.S. and 1 = born in the U.S. To explore race‐gender group differences, we used responses to various variables that were reported by youth. First, we used individuals' answers to “What are your racial backgrounds. Check all that apply” to create a dataset with only youth who identified as Latine‐White or Black‐White. We then used responses to “With what gender do you identify with?” to create a new variable with 4 codes for youth who were Latina‐White females (indicated they were Latina‐White for race and female for gender; n = 127), Latino‐White males (indicated they were Latino‐White for race and male for gender; n = 182), Black‐White females (indicated they were Black‐White for race and female for gender; n = 144), and Black‐White males (indicated they were Black‐White for race and male for gender; n = 130). This race‐gender variable was included as our grouping variable in analyses. Analytic approach To test hypotheses, a series of nested multigroup path analysis models were specified in Mplus (Muthén & Muthén, 2019) with race‐gender as the grouping variable. In this approach, an unconstrained model in which no paths were constrained by race‐gender group was compared with a fully constrained model in which all paths were constrained to be equal across race‐gender groups, and a chi‐square difference test was conducted (Kline, 1998). A chi‐square difference test that is not significant indicates that there are no significant differences in the model, and the fully constrained model is accepted as the final model, which provides findings for all youth regardless of race‐gender group. A significant chi‐square difference test indicates that there are significant group differences in the model. In this case, the next step is to compare the unconstrained model to a model in which only one path is constrained to be equal across race‐gender groups. If the chi‐square difference test is significant, it indicates that there are significant group differences in the path, and the constraint is removed prior to testing the next path. If the chi‐square difference test is not significant it indicates that there are no significant ethnic–racial group differences in the path, and the constraint is kept, and the next model is compared with this model with the one constrained path. This process is repeated to test for differences in all paths to arrive at a partially constrained final model, in which paths that are significantly different across race‐gender groups are left unconstrained, and paths that are not significantly different across race‐gender groups are constrained to be equal. Three fit indices were used to examine model fit: the comparative fit index (CFI), the root‐mean‐square‐error of approximation (RMSEA), and the standardized root‐mean‐square residual (SRMR). Model fit was considered to be good if the CFI was greater than or equal to .95, the RMSEA was less than or equal to .05, and the SRMR was less than or equal to .05 (Hu & Bentler, 1999). The model chi‐square was used for model comparisons but was not used in evaluating overall fit due to its sensitivity to sample size (Fan et al., 1999). Missing data were handled using full maximum likelihood, which is a recommended approach (Arbuckle, 1996). Age and birthplace were included as covariates. Standardized coefficients are presented below, and unstandardized coefficients are presented in Figure 1. Final partially constrained multigroup model of types of belongingness predicting mental health and suicide ideation. A bolded line indicates that the path was freely estimated among Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males because there was a significant difference by group. For any paths with significant differences (bolded lines), findings are presented in the following order: Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. An unbolded line indicates that the path was constrained to be equal among all adolescents because there was not a significant difference across groups. Coefficients are unstandardized. Age and country of birth were included as covariates predicting outcomes, but are not displayed for ease of illustration. * p < .05. ** p < .01. *** p < .001. RESULTS Descriptive statistics included testing correlations for all variables and covariates among Latina‐White females and Latino‐White males (Table 1) and Black‐White females and Black‐White males (Table 2). We then conducted ANOVA analyses to test mean differences between the four race‐gender groups. These findings indicated that there were significant mean differences across the race‐gender groups for minoritized identity belongingness [F(3, 579) = 3.30, p = .02], White identity belongingness [F(3, 578) = 5.81, p < .001], anxiety symptoms [F(3, 579) = 2.76, p = .04], and depressive symptoms [F(3, 579) = 3.46, p = .02]. Tukey post‐hoc tests (Table 3) indicated that Latino‐White males (M = 5.00, SD = .77) had significantly higher levels of minoritized identity belongingness than Black‐White females (M = 4.73, SD = .87; p = .02). Additionally, Black‐White females (M = 4.43, SD = 1.11) had significantly lower levels of White identity belongingness than Latina‐White females (M = 4.87, SD = .97; p = .00) and Latino‐White males (M = 4.85, SD = .95; p = .00). For anxiety symptoms, there were no significant differences by race‐gender group (p > .05). For depressive symptoms, Latina‐White females (M = 2.46, SD = 1.08) had significantly higher depressive symptoms than Latino‐White males (M = 2.08, SD = 1.03; p = .01). TABLE 1Correlations among study variables and controls for Latina‐White Females (N = 127) and Latino‐White males (n = 182).12345678 Minoritized Identity Belongingness ‐.48***.57***−.28**−.27**−.24**−.06−.07 2White Identity Belongingness .66***‐.39***−.20*−.26**−.24**.07−.15 3Multiracial Identity Belongingness .59***.45***‐−.26**−.24**−.16.08−.10 4Anxiety Symptoms −.20**−.20**−.18*‐.83***.68***−.25**−.06 5Depressive Symptoms −.33***−.29***−.23**.84***‐.60***−.12−.02 6Suicide Ideation −.14−.12−.09.52***.53***‐−.24**−.00 7Birthplace −.02.01−.04.04.10.05‐−.01 8Age .04−.02.04.01.07.05.09‐ Note: Correlations for Latina‐White females are to the top right of the diagonal and correlations for Latino‐White males are to the bottom left of the diagonal. Birthplace was coded 0 = born outside the U.S., 1 = born in the U.S. * p < .05. **p < .01. ***p < .001. Correlations among study variables and controls for Latina‐White Females (N = 127) and Latino‐White males (n = 182). Minoritized Identity Belongingness Minoritized Identity Belongingness 2White Identity Belongingness White Identity Belongingness 3Multiracial Identity Belongingness Multiracial Identity Belongingness 4Anxiety Symptoms Anxiety Symptoms 5Depressive Symptoms Depressive Symptoms 6Suicide Ideation Suicide Ideation 7Birthplace Birthplace 8Age Age Note: Correlations for Latina‐White females are to the top right of the diagonal and correlations for Latino‐White males are to the bottom left of the diagonal. Birthplace was coded 0 = born outside the U.S., 1 = born in the U.S. * p < .05. **p < .01. ***p < .001. TABLE 2Correlations among study variables and controls for Black‐White Females (N = 144) and Black‐White males (n = 130).12345678 Minoritized Identity Belongingness ‐‐.41***.43***−.37***−.42***−.26**−.10−.13 2White Identity Belongingness .69***‐‐.22**−.10−.18*−.20*.06−.07 3Multiracial Identity Belongingness .55***.46***‐‐−.07−.17*.02−.07−.14 4Anxiety Symptoms −.22*−.20*−.34***‐‐.83***.42***.11−.12 5Depressive Symptoms −.24**−.20*−.39***.83***‐‐.54***.08.01 6Suicide Ideation −.18*−.19*−.27**.44***.49***‐‐.00.04 7Birthplace −.00.07.01−.24**−.17−.13‐‐.14 8Age .04−.18*−.14.07.07.08.10‐‐ Note: Correlations for Black‐White females are to the top right of the diagonal and correlations for Black‐White males are to the bottom left of the diagonal. Birthplace was coded 0 = born outside the U.S., 1 = born in the U.S. * p < .05. **p < .01. ***p < .001. Correlations among study variables and controls for Black‐White Females (N = 144) and Black‐White males (n = 130). Minoritized Identity Belongingness Minoritized Identity Belongingness 2White Identity Belongingness White Identity Belongingness 3Multiracial Identity Belongingness Multiracial Identity Belongingness 4Anxiety Symptoms Anxiety Symptoms 5Depressive Symptoms Depressive Symptoms 6Suicide Ideation Suicide Ideation 7Birthplace Birthplace 8Age Age Note: Correlations for Black‐White females are to the top right of the diagonal and correlations for Black‐White males are to the bottom left of the diagonal. Birthplace was coded 0 = born outside the U.S., 1 = born in the U.S. * p < .05. **p < .01. ***p < .001. TABLE 3Post‐hoc ANOVA comparisons of mean (SD) differences between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males on main variables.VariableResponse rangeLatina‐White female (n = 127)Latino‐White male (n = 182)Black‐White female (n = 144)Black‐White male (n = 130)Minoritized identity belongingness1–65.00 (.87)5.00 (.77) a 4.73 (.87) a 4.88 (.95)White identity belongingness1–64.87 (.97) a 4.85 (.95) b 4.43 (1.11) ab 4.73 (1.07)Multiracial identity belongingness1–65.10 (.76)5.06 (.86)5.05 (.81)4.99 (.92)Anxiety symptoms1–52.03 (1.07)1.74 (.94)2.00 (1.04)1.93 (.98)Depressive symptoms1–52.46 (1.08) a 2.08 (1.03) a 2.31 (1.10)2.17 (1.09)Suicide ideation0–4.54 (.96).44 (.90).48 (.85).45 (.77) Note: a b denote significant mean differences between groups on a variable. A shared superscript letter indicates that those two groups had means that were significantly different from each other. Means are presented first and standard deviations are presented in parentheses. Post‐hoc ANOVA comparisons of mean (SD) differences between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males on main variables. Note: a b denote significant mean differences between groups on a variable. A shared superscript letter indicates that those two groups had means that were significantly different from each other. Means are presented first and standard deviations are presented in parentheses. We tested our hypothesized research questions using multigroup nested models. Given that our hypothesized model was saturated, we constrained any control paths that were not significant across the four groups to zero to be able to determine model fit. Therefore, age predicting depressive symptoms and age predicting suicide ideation were set to 0 given that neither of these paths was significant for Latina‐White females, Latino‐White males, Black‐White females, or Black‐White males. The results of the nested model comparisons to explore race‐gender group differences using chi‐square difference tests are presented in Table 4. The unconstrained model compared with the fully constrained model was significant, suggesting that the findings were significantly different across race‐gender groups; therefore, we tested which paths were significantly different to arrive at a final partially constrained model, which had good fit: χ2 (df = 32) = 33.44, p = .40; CFI = 1.00; RMSEA = .02 (90% C.I.: .00–.07); SRMR = .04 (see Figure 1). TABLE 4Nested model comparisons and chi‐square difference tests for constraints imposed across Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males to test for significant group differences.Model #Models comparedConstraintsCFIRMSEASRMRχ2(df)Δ χ2 (∆ df)Sig levelConstraint kept0‐Unconstrained1.000.000.012.91 (8) 1 0 and 1 Fully constrained 0.98 0.05 0.05 64.21 (47) 61.30 (39) 0.01 No 2 0 and 2 Minoritized Identity B ➔ Anxiety Symptoms 1.00 0.04 0.03 13.27 (11) 10.36 (3) 0.02 No 30 and 3White Identity B ➔ Anxiety Symptoms1.000.000.024.52 (11)1.61 (3)0.66Yes 4 3 and 4 Multiracial Identity B ➔ Anxiety Symptoms 1.00 0.00 0.03 13.77 (14) 9.25 (3) 0.03 No 5 3 and 5 Birthplace ➔ Anxiety Symptoms 0.99 0.06 0.04 21.97 (14) 17.45 (3) 0.00 No 63 and 6Age ➔ Anxiety Symptoms1.000.000.029.32 (14)4.80 (3)0.19Yes76 and 7Minoritized Identity B ➔ Depressive Symptoms1.000.000.0316.86 (17)7.54 (3)0.06Yes87 and 8White Identity B ➔ Depressive Symptoms1.000.030.0422.91 (20)6.05 (3)0.11Yes98 and 9Multiracial Identity B ➔ Depressive Symptoms1.000.030.0325.24 (23)2.33 (3)0.51Yes 10 9 and 10 Birthplace ➔ Depressive Symptoms 0.99 0.04 0.04 33.38 (26) 8.14 (3) 0.04 No 119 and 11Minoritized Identity B ➔ Suicide Ideation0.990.020.0327.41 (26)2.17 (3)0.54Yes1211 and 12White Identity B ➔ Suicide Ideation1.000.020.0330.57 (29)3.16 (3)0.37Yes1312 and 13Multiracial Identity B ➔ Suicide Ideation1.000.020.0433.44 (32)2.87 (3)0.41Yes 14 13 and 14 Birthplace ➔ Suicide Ideation 0.99 0.04 0.04 41.71 (35) 8.27 (3) 0.04 No Note: Bolded font indicates a model that had significant group differences and unbolded font indicates a model that did not have significant group differences.Abbreviations: B = Belongingness, CFI = comparative fit index, RMSEA = root‐mean‐square‐error of approximation, SRMR = standardized root‐mean‐square residual, Sig = significance. Nested model comparisons and chi‐square difference tests for constraints imposed across Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males to test for significant group differences. Note: Bolded font indicates a model that had significant group differences and unbolded font indicates a model that did not have significant group differences. Abbreviations: B = Belongingness, CFI = comparative fit index, RMSEA = root‐mean‐square‐error of approximation, SRMR = standardized root‐mean‐square residual, Sig = significance. Results indicated that greater minoritized identity belongingness was associated with less depressive symptoms (β = −.22, p < .001) and suicide ideation (β = −.14, p = .01) for all youth, and less anxiety symptoms for all youth except Latino‐White males (β = −.22, p < .001 for Latina‐White females, β = −.27, p < .001 for Black‐White females, and β = −.21, p = .01 for Black‐White males). Interestingly, greater Multiracial identity belongingness was associated with less anxiety symptoms for only Latino‐White males (β = −.14, p = .03), but was not significant for Latina‐White females, Black‐White females, or Black‐White males (p > .05). Greater Multiracial identity belongingness was also associated with less depressive symptoms for all youth (β = −.10, p = .02), but not associated with suicide ideation for any youth (p > .05). Further, greater White identity belongingness was not associated with anxiety symptoms, depressive symptoms, or suicide ideation for any youth (p > .05). Regarding covariates, birthplace was associated with anxiety symptoms, such that Latina‐White females born in the U.S. (β = −.25, p = .00) and Black‐White males born in the U.S. (β = −.23, p = .00) had less anxiety symptoms than those born outside the U.S. Age was also associated with anxiety symptoms, such that older youth across race‐gender groups had lower anxiety symptoms than younger youth (β = −.05, p = .02). Birthplace was associated with depressive symptoms, such that Black‐White males born in the U.S. had fewer depressive symptoms than those born outside the U.S. (β = −.16, p = .05). Birthplace was associated with suicide ideation, such that Latina‐White females born in the U.S. had less suicide ideation than those born outside the U.S. (β = −.24, p = .00). DISCUSSION The goal of the present study was to test whether three types of ERI belongingness, including adolescents' minoritized identity belongingness, White identity belongingness, and Multiracial identity belongingness, were associated with symptoms of anxiety, symptoms of depression, and suicide ideation. Guided by the Belongingness Hypothesis (Baumeister & Leary, 1995) and the Interpersonal Theory of Suicide (Van Orden et al., 2010), we hypothesized that greater ERI belongingness would inform lower anxiety symptoms, depressive symptoms, and suicide ideation. Some findings were supported, but there were important nuances in these relations based on the type of ERI belonging. In addition, we explored potential race‐gender group differences in the associations between Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males. First, regarding ERI belongingness, results indicated that youth had relatively high levels of belongingness to their minoritized, White, and Multiracial identities (i.e., mean levels above 4 on a 1 to 6 scale). As discussed below in more detail, findings also indicated that belongingness was linked with better mental health, but there were differences across the four race‐gender groups. These findings suggest it is important to not only continue research regarding belongingness and mental health among Multiracial youth, but also to further examine subgroup differences. Below, we discuss the implications of our descriptive findings for belongingness to differing identities, considerations for the linkages between three types of ERI belongingness and associated mental health and suicide outcomes, and close with our study limitations, future directions, and conclusions. Descriptives for belongingness to different identities Given that this was the first study to examine belongingness across Multiracial adolescents' various ethnic–racial identities, and one of few studies to focus on race‐gender group differences in anxiety, depression, and suicide ideation, we examined descriptives for the study. Overall, across the four identified race‐gender groups (Latina‐White females, Latino‐White males, Black‐White females, and Black‐White males), adolescents tended to have relatively high levels of belongingness among the different types of ERI (minoritized, White, and Multiracial). This high endorsement is particularly interesting because belongingness has been one of the least studied components of ERI, sometimes even referred to as the “lost” component of ERI (Lee et al., 2022; Williams et al., 2020). Our findings suggest that among Multiracial adolescents, ERI belongingness is important and warrants future investigation, especially in work that aims to understand the normative development of ERI. We understood that the Multiracial participants in this study may have unique experiences of belongingness based on their specific race and gender identities, but lacked any specific hypothesis given the dearth of research that considers race‐gender differences. The descriptive findings ultimately revealed that Latino‐White males felt a deeper sense of belonging to their minoritized identity than Black‐White females did. Notably, Black‐White females also reported less belonging with their White identity than Latino‐White males and Latina‐White females. The fact that race‐gender differences in belongingness emerged across racial groups (rather than within them) is interesting and illustrates why Multiracial youth should not be studied as a monolith. Moreover, Black‐White females reporting significantly less belonging to their Black and White identities alongside average levels of Multiracial identity belongingness suggests that they may be more psychologically vulnerable to racial or cultural homelessness than Multiracial individuals from other race‐gender groups (Vivero & Jenkins, 1999). What we know about ERI has traditionally been conceptualized and empirically tested with monoracial youth (Rogers et al., 2020; Williams et al., 2020), and to our knowledge, this is the first study to reveal these race‐gender differences in belongingness among Multiracial individuals. Results from this study indicate that not only including Multiracial youth in this work is important, but that intersectional differences among this population must be explored. Links between types of ERI belongingness and outcomes Results from our primary analyses indicated that some paths varied by race‐gender group, and other paths were the same. First, across all adolescents (regardless of their specific race and gender), having a deeper sense of connection with their minoritized identity (i.e., either being Latine or being Black) was associated with less suicide ideation and less depressive symptoms, and a deeper belongingness to their Multiraciality was associated with less depressive symptoms. These findings suggest that individuals who work with Multiracial adolescents who are experiencing mental health challenges should discuss their perceived connection to their minoritized identity and Multiracial identity. For example, clinicians, mentors, and caregivers could help Multiracial youth connect with other Multiracial youth, as well as youth who share their minoritized identities (e.g., other Black youth or other Latine youth) to help them feel a sense of connection and belonging. In a small case study with six Biracial adolescent emerging adults, belongingness was increased via family guidance in being Biracial, extended family impacting their sense of connection to racial groups, and language shaping their sense of inclusion (Eid & Parker, 2023). Thus, family may be particularly important in helping support a sense of belonging, but more research is needed that tests different types of belongingness and the influence of family, as well as individuals outside the family (e.g., peers, teachers, counselors, etc.). Another interesting result was that in the current study, although Multiracial youth had relatively high levels of belongingness to their White identity, this form of identity was the only one that was not associated with any mental health outcomes. Throughout time, per the “one drop rule” Multiracial individuals, especially Black Multiracial individuals, have been classified and perceived by others to be a minority individual regardless of their own opinions, feelings, or labels (Harris, 2016). In a society in which monoracial categories and boundaries have been strongly enforced (Harris, 2016), it is possible that despite Multiracial youth feeling connected with their own Whiteness and with other White individuals, they may receive messages from others that they are not White and do not belong. Because of the messages, bias, and perceptions of others, there may be a more tangled and nuanced relationship with White belongingness, which may prevent the benefits for mental health that belongingness to Multiracial and minoritized identity provides for Multiracial youth. This finding requires further mixed‐methods research to dive deeper into the complexities in youths' experiences related to their belongingness across their identities. Beyond the findings that emerged across adolescents, two specific race‐gender differences were identified, with both of the nuances revolving around anxiety symptoms among Latino‐White males. First, a sense of belongingness to one's minoritized identity (e.g., Black or Latine) was protective against symptoms of anxiety for every race‐gender group except Latino‐White males. The multigroup models further revealed that a sense of belonging to one's Multiracial background was only associated with less anxiety for Latino‐White males. These findings are interesting as Latino‐White males reported comparable levels of belonging to their Latino and Multiracial identities and benefited from both psychologically, albeit in different ways. For these participants, there appears to be a unique relationship between Multiracial belonging and anxiety that is not present for the other race‐gender groups. More research with Latino‐White Multiracial young males is necessary to better understand and formally theorize why these nuances emerged. However, we suspect that racial malleability may be an interesting avenue to explore. Racial malleability (also referred to as racial variability or code switching) is the process through which Multiracial people shift or change which racial identity they identify the most with based on their ecological contexts (Pauker et al., 2018). Racial malleability can be psychologically beneficial as it “enables individuals to access a socially adaptive identity in response to a potentially hostile situation” (Shih et al., 2019). Access to a socially advantageous identity may be protective against anxiety as it is a mental state that is largely amenable to social stimuli, especially during adolescence (Garcia & O'Neil, 2021). In the context of this study, belongingness to one's Multiracial identity may have helped Latino‐White males avoid discrimination and the impacts of it (e.g., anxiety) in a way that belongingness to their Latino identity did not. This protection from discrimination may be more vital for Latino‐White males rather than females as ethnic discrimination has been found to threaten notions of masculinity and heighten feelings of isolation (Cano et al., 2016). This may be particularly true when discrimination comes from Latino peers as discrimination from perceived in‐group members has been linked to lower feelings of commonality among Latine individuals (Sanchez & Espinosa, 2016). However, these hypothesized associations are merely speculative as this study did not assess constructs like malleability, masculinity, or source of discrimination. Future research is warranted. Limitations, future directions, and conclusions There are several limitations to consider. First, we focused on two specific intersections, race and gender, to explore group differences, and our study only included Latine‐White and Black‐White males and females. Interestingly, even with just these two subgroups, the findings revealed significant differences based on race and gender. As Multiracial families continue to increase, it is important to further explore these associations among other subgroups, such as dual minority Multiracial individuals and individuals with other White and minoritized backgrounds (e.g., Black‐Latine, Asian‐White) to determine whether similarities or differences arise. We also need to explore nuances across various intersections, such as race, class, gender, sexuality, ableness, and immigrant identities (Azmitia et al., 2023), to name a few. Overall, this is an area in need of research, and this exploration can be valuable for informing programs and resources aimed at reducing mental health disparities. Relatedly, there is not an existing theory focused on ERI belongingness and outcomes or regarding race‐gender group differences. We used the Belongingness Hypothesis (Baumeister & Leary, 1995) and the Interpersonal Theory of Suicide (Van Orden et al., 2010) to guide our expectations for belongingness and outcomes, but neither theory focuses on belongingness specifically related to ethnicity/race/culture. Further, we had no theory to pull from for the race‐gender group differences, so this aspect of the study was exploratory. As work in this area moves forward, it will be essential to build theoretical frameworks that outline these relations so that theory‐testing approaches can occur in future research. Furthermore, the current sample lacked sufficient representation of gender non‐conforming youth, limiting our ability to explore racial and gender differences beyond the binary categories of male and female. Existing work has indicated that gender non‐conforming youth, especially those who belong to racially minoritized backgrounds, encounter unique challenges that significantly impact their mental health (Canady, 2022; Lassiter et al., 2020). Specifically, one report found that Multiracial LGBTQ youth are disproportionately at risk for worse mental health indicators (e.g., increased risks of self‐harm compared with cisgender Multiracial peers; Canady, 2022). Research indicates that individuals with dual minority identities, such as being both a racial/ethnic and gender minority, often experience compounded stress due to intersecting forms of discrimination and stigma (Lassiter et al., 2020). While existing studies provide valuable insights, more research is needed to fully understand the complexities and lived experiences of other subgroups of Multiracial youth who are at risk for poorer mental health‐related outcomes. Additionally, while cross‐sectional studies provide valuable insights, they do not capture the possibilities of bidirectional relationships between belongingness and mental health over time. Importantly, existing theory and empirical work documents that ERI and mental health change over time (Rogers‐Sirin & Gupta, 2012; Williams et al., 2020). Longitudinal research is essential to understand how changes in one domain may influence changes in the other. Such research can inform interventions aimed at enhancing well‐being. Furthermore, to reduce participant burden, we included only one item that assessed suicidal ideation and shortened versions of anxiety and depressive symptoms. Future research would greatly benefit from building on this work by implementing more comprehensive assessments that effectively capture these constructs and rigorously test the consistency of the findings. Lastly, although we included multiple types of ERI belongingness in the current study, we did not have multiple types of ERI affirmation (positive feelings) towards their minoritized, White, and Multiracial identities. Early conceptualizations of ERI included affirmation and belongingness items together as affirmation‐belonging. It will be essential to explore different types of affirmation across identities and different types of belonging across identities in the same study to test discriminant validity across the types of ERI components and investigate how each component affects mental health and suicide ideation. For example, do high or low levels of ERI affirmation towards minoritized, White, and Multiracial identities follow similar or different patterns than ERI belongingness in predicting outcomes? This is an important future research direction. Despite these limitations, the current study provides valuable insight into how types of ERI belongingness are associated with anxiety symptoms, depressive symptoms, and suicide ideation among various groups. Specifically, for mental health professionals or community resources geared towards Multiracial youth, it is crucial to consider an individual's intersecting identities. Clinicians should consider the unique identity experiences of Multiracial youth, especially how they navigate belonging to both minoritized and Multiracial identities. Therapeutic practices that affirm and explore complex racial identities may facilitate positive mental health (e.g., less anxiety) for Multiracial youth. Further, individuals who work with Multiracial youth (e.g., educators and counselors) should receive training in identifying patterns of low belongingness, especially in youth whose identities intersect with social boundaries. By enhancing their awareness and skills, they can better support and foster an inclusive environment for Multiracial youth. Overall, initiatives and research need to continue including and understanding the often‐invisible experiences of Multiracial youth, especially those with marginalized gender and racial identities. AUTHOR CONTRIBUTIONS N. Keita Christophe: Conceptualization; investigation; funding acquisition; writing – original draft; writing – review and editing; methodology; project administration; resources; supervision; data curation. Nilee Kerr: Writing – original draft; writing – review and editing. Troy J. Keys: Writing – original draft; writing – review and editing. B. Andi Lee: Writing – original draft; writing – review and editing. Kimberly Lazarus: Writing – original draft; writing – review and editing. McKenzie N. Green: Writing – original draft; writing – review and editing. Arlenis Santana: Writing – review and editing; writing – original draft. Chelsea Derlan Williams: Conceptualization; methodology; investigation; funding acquisition; writing – original draft; writing – review and editing; validation; visualization; formal analysis; project administration; resources; supervision; data curation. Diamond Bravo: Writing – original draft; writing – review and editing. Annabelle L. Atkin: Data curation; supervision; resources; project administration; methodology; writing – review and editing; conceptualization; investigation; funding acquisition; writing – original draft. FUNDING INFORMATION Funding was provided by Virginia Commonwealth University, McGill University, & Purdue University. CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest. ETHICAL APPROVAL STATEMENT This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Institutional Review Board at McGill University. CONSENT STATEMENT Consent was provided by caregivers, and assent was provided by youth.