Associations of General and Multiracial-Specific Discrimination with Psychological Distress and Substance Use in Multiracial College Students
Department of Psychology, Syracuse University, Syracuse, New York, USA
CONTACT Aesoon Park aepark@syr.edu Department of Psychology, Syracuse University, Marley Education Center, 352 Irving Ave, Syracuse, NY 13210, USA.Abstract
Objectives:
Multiracial (alternatively, mixed-race or biracial) college students report higher rates of alcohol/substance use than their monoracial peers, yet factors associated with this disparity remain understudied. Guided by minority stress and tension-reduction frameworks, associations of general and Multiracial discrimination experiences with psychological distress and substance use were investigated.
Methods:
Cross-sectional data obtained from a convenience sample of 193 Multiracial college students (Mage=20.34 [SD = 1.32], 30% Male, 31% fraternity/sorority affiliation) were used. Multiracial college students reported their lifetime discriminatory experiences (general and Multiracial-specific), psychological distress (i.e., depressive and anxiety symptoms), and past-year alcohol/substance (alcohol, cannabis, and tobacco) use frequencies. Fully saturated structural regression models were used to test relations of general discrimination and a latent Multiracial-specific factor with a latent substance use factor with and without psychological distress in the model.
Results:
Without psychological distress in the model, Multiracial-specific (but not general) discrimination experiences were positively associated with substance use (β = 0.29, 95% CI [0.02, 0.57], p = 0.04). With psychological distress in the model, however, neither Multiracial-specific nor general discrimination were directly associated with substance use. Instead, Multiracial (but not general) discrimination predicted psychological distress (β = 0.26, 95% CI [0.06, 0.46], p = 0.01), and psychological distress and substance use were positively associated with each other (β = 0.39 [95% CI [0.05, 0.59], p = 0.003). Ancillary analyses further indicated the mediating role of psychological distress in the discrimination-substance use association.
Conclusions:
The novel findings suggest roles of Multiracial-specific discrimination by peers and family members and psychological distress in heightened substance use among Multiracial college students. Leveraging prospective methodologies is critical to determine temporal order.
Substance use among college students is a significant public health concern in the U.S., and Multiracial college students (defined as individuals with biological parents of two or more racial and ethnic groups; alternatively, mixed-race or biracial; Atkin et al., 2022) may be at heightened risk (Dobani et al., 2024). Alcohol, cannabis, and tobacco are the three most commonly used substances by college students, excluding caffeine (Schulenberg et al., 2021). Although no recent prevalence rates are available specifically for Multiracial college students, national estimates indicate that in the past year, 68% of college-aged individuals (including college students) reported alcohol, 37% reported cannabis, and 26% reported tobacco use (Substance Abuse & Mental Health Services Administration, 2024). Notably, Multiracial college-aged individuals endorse a higher prevalence of substance use than their monoracial peers, with 70% reporting alcohol, 47% reporting cannabis, and 30% reporting tobacco use in the past year (SAMHSA, 2024). These prevalence rates may vary based on college status, as college students report higher alcohol use overall, while their non-college-attending peers report higher rates of past-year cannabis and tobacco use (Schulenberg et al., 2021).
The urgency to better understand and address heightened substance use among Multiracial college students is underscored by their increasing college enrollment, paralleling broader demographic trends classifying the Multiracial sub-population as the fastest-growing racial group in the U.S, with a 250% increase over the past decade (Jones 2020; Hussar & Bailey, 2020). Despite this growth, Multiracial college students remain critically underrepresented in college student literature, limiting an understanding of their health and health-risk behavior, including substance use. Substance use is associated with chronic and acute consequences for college students, including missed classes and lower grades in college, sexual assaults, changes in brain function, lingering cognitive deficits, reduced quality of life, and substance-related mortality (Espinosa et al., 2023; Hingson & White, 2014; Milic et al., 2020). Heightened rates of substance use among Multiracial college students may increase their susceptibility to these adverse consequences, warranting further research on associated factors to clarify potential health disparities.
Discrimination, a recognized sociocultural determinant of health (Pascoe et al., 2022), may be associated with substance use disparities among Multiracial college students. Multiracial individuals commonly report experiencing general racial discrimination due to their minoritized racial status (Oh et al., 2023; Pew Research Center, 2015). In addition to general racial discrimination, they can also experience Multiracial discrimination. In the U.S., the historical reinforcement of race/ethnicity in mutually exclusive categories marginalizes Multiracial people, who may not fully identify with established monoracial categories (e.g., Black, Asian, White; Harris, 2016). This monoracial paradigm can perpetuate monoracism, the belief that people should be categorized into discrete racial/ethnic groups, and consequently, Multiracial discrimination, where individuals of two or more racial/ethnic groups may experience unequal treatment due to their Multiracial status (Atkin & Yoo, 2019; Johnston & Nadal, 2010). Thus, while there is within-group heterogeneity among Multiracial people, Multiracial discrimination may be a shared experience (Harris, 2016). A number of qualitative studies document instances of Multiracial discrimination experienced by Multiracial college students, including invalidation of racial identities, exclusion, challenges to authenticity, pathologizing of Multiracial identities, and familial discrimination (Franco & Carter, 2019; Museus et al., 2016; Nadal et al., 2013).
Despite considerable advances in the Multiracial literature suggesting that Multiracial people experience both general and Multiracial discrimination, the association of these racial stressors with substance use is yet to be fully understood. Limited research has indicated concurrent associations of general discrimination with alcohol and cannabis use among Multiracial adolescents and adults (Choi et al., 2012; Nalven et al., 2021; Wedel et al., 2022). Conversely, findings have also demonstrated that general discrimination was not prospectively associated with alcohol use among Multiracial high school students as compared to monoracial high school students, after controlling for age, sex, and ethnicity (Goodhines et al., 2020). The role of Multiracial discrimination in substance use among Multiracial college students is particularly understudied. Multiracial discrimination may encompass various experiences including, racial identity invalidation (e.g., “I told someone about my racial background(s), but they did not believe me”), familial discrimination (e.g., “A family member said that I am not a “real” member of a racial group(s) with whom I identify”), and perceived racial ambiguity (e.g., “I get asked “What are you?”; Salahuddin & O’Brien, 2011; Franco & O’Brien, 2018; Franco et al., 2021a; Yoo et al., 2016). Such experiences may undermine Multiracial individuals’ sense of belonging within family and peer in-groups, which could be linked to health-risk behaviors, such as substance use (Atkin et al., 2025). Moreover, some forms of Multiracial discrimination may be particularly harmful. Recent research has found that familial discrimination predicts substance use above and beyond general Multiracial discrimination among Multiracial adults (Franco & Carter, 2019).
Minority stress (Meyer, 2003) and tension-reduction models (Kassel et al., 2003; Lazarus & Folkman, 1984; Sher et al., 2007) provide a useful framework for understanding the relationship between discrimination and substance use. These frameworks posit that individuals may use substances to cope with psychological distress (e.g., depressive and anxiety symptoms) stemming from discriminatory experiences. Both general (Jackson et al., 2012; Reid Marks et al., 2020) and Multiracial (Christophe et al., 2024; Coleman & Carter, 2007; Franco & O’Brien, 2018; Salahuddin & O’Brien, 2011; Townsend et al., 2009; Yoo et al., 2016) discriminatory experiences are each predictive of psychological distress when examined separately. However, when general Multiracial and familial discrimination were investigated together, only lack of family acceptance was associated with psychological distress among Multiracial adults, after adjusting for income, education, and gender (Franco et al., 2021b). Overall, these findings underscore the importance of Multiracial discrimination and psychological distress in relation to substance use. However, their roles in substance use among Multiracial college students warrant further exploration, given that Multiracial college students tend to report greater odds of psychological distress (e.g., depressive and anxiety symptoms) than their monoracial peers (Oh et al., 2023) and psychological distress is associated with greater odds of drinking to feel drunk among Multiracial young adults (Atkin et al., 2025).
Three critical gaps exist in the literature on substance use among Multiracial college students. First, although national data indicate elevated rates of substance use among Multiracial college-aged individuals, prevalence estimates specific to Multiracial college students are limited. Additionally, available studies often focus on a single substance (e.g., alcohol) or rely on a composite substance use indicator, limiting insight into substance-specific patterns. Second, Multiracial individuals occupy a uniquely complex sociocultural position that exposes them to multiple forms of marginalization, including Multiracial-specific discrimination, racial discrimination related to a marginalized monoracial identity, and discrimination based on other marginalized identities (e.g., gender). While prior research has linked general and Multiracial-specific discrimination to psychological distress or substance use, to our knowledge, no studies have examined both simultaneously as predictors of psychological distress and substance use. Such investigation can characterize the role of Multiracial discrimination as a minority stressor potentially associated with coping-motivated substance use. Lastly, Multiracial-specific discrimination is multifaceted, encompassing others’ surprise and disbelief, lack of familial acceptance, and overt Multiracial discrimination (Salahuddin & O’Brien, 2011). Limited research has found that familial Multiracial discrimination may predict substance use beyond overt Multiracial discrimination (Franco & Carter, 2019). Yet, other forms of Multiracial discrimination, such as disbelief when Multiracial background is disclosed, remain understudied. Testing a latent factor structure of Multiracial discrimination can clarify which forms of Multiracial discrimination may be behaviorally consequential, further informing research on substance use risk pathways and targeted intervention targets tailored to Multiracial college students.
The current cross-sectional study aimed to investigate the role of general and Multiracial discrimination in relation to psychological distress and substance use among Multiracial college students. Two models were estimated to test the associations of general and Multiracial discrimination with substance use without (Model 1) and with (Model 2) psychological distress accounted for as a second outcome. Exploratory ancillary analyses evaluated the mediating role of psychological distress in the discrimination-substance use association. Given limited and mixed findings on the association of discrimination, psychological distress, and substance use associations among Multiracial adolescents and adults, a priori hypotheses were not specified.
Method
Participants
Cross-sectional data were drawn from an online health survey study of 193 Multiracial college students (29% Hispanic-White, 19% Asian-White, 8% Black-White, 7% American Indian/Alaska Native-White, 1% Native Hawaiian or Other Pacific Islander-White, 25% dual-racial/ethnic minority, 12% 3+ racial/ethnic categories;Mage=20.34 [SD = 1.32]; 30% Male; and 31% fraternity/sorority affiliation). Multiracial college students were eligible if they were Multiracial, between the ages of 18 and 25, which is the typical age range of college students (Sawyer et al., 2018), enrolled part-or full-time at the university at the time of study participation, English-speaking/reading, and indicated lifetime alcohol use. To determine Multiracial eligibility, participants were given the prompt, “Multiracial is defined as belonging to more than one race and/or ethnicity. Racial and ethnic groups include White, Black, American Indian/Alaska Native, Asian, Hispanic, Latino, of Other Spanish origin, and Native Hawaiian or Other Pacific Islander. Do you consider yourself to be Multiracial?” Participants were provided yes or no response options. Participants were also asked to report the race of their biological mother and father. Participants met criteria for Multiracial status if they self-identified as Multiracial and if their biological parents were from two different racial/ethnic groups or if at least one parent was Multiracial. This approach acknowledges that racial categorization among Multiracial individuals can be based on either identity or ancestry, with these categorizations and classifications shifting across context and time (Atkin et al., 2022).
To assess lifetime alcohol use, participants were asked, “Have you had a drink of beer, wine, or liquor, not just a sip or a taste of someone else’s drink, more than 2 times in your life?” Participants who reported yes were eligible to participate in the study. The lifetime alcohol use eligibility criterion was included to maximize the probability of capturing changes in substance use behaviors, the primary phenomenon of interest. Participants were recruited through the institutional participant pool for psychology research, various Psychology courses, and flyers at a private, northeastern university from January 2021 to December 2021.
Procedure
The University Institutional Review Board approved all study procedures. After completing an electronic consent form, participants completed an anonymous online questionnaire from a location of their choosing and were compensated with either course credit or a $5.00 electronic gift card to Starbucks or Amazon, of their choosing. Approximately 87% of participants were recruited from the participant pool for psychology research or psychology courses and received course credit, whereas 13% were recruited from flyers and received an electronic gift card. Results of a Mann-Whitney U test indicated that there were significant differences by recruitment method in Multiracial discrimination (U = 2823, p<0.001) and psychological distress (U = 2519, p = 0.02), but no differences in alcohol, cannabis and tobacco use frequencies at p<.05. In other words, participants recruited through the psychology participant pool reported higher frequency of Multiracial discriminatory experiences and higher psychological distress, but no differences in alcohol and substance use frequency, compared to participants recruited via flyers. Notably, when the recruitment method was included as a covariate in the main analyses, the pattern of results did not change in terms of significance. As such, recruitment method was ultimately not included in main analyses.
Measures
Past-year substance use frequency
Three items adapted from the National Institute on Alcohol Abuse and Alcoholism (NIAAA; 2011) assessed past-year use frequencies of alcohol (i.e., “During the last year, how often did you usually use alcohol?”), cannabis (“During the last year, how often did you use cannabis?”), and tobacco/e-cigarettes (i.e., “During the last year, how often did you use tobacco and/or e-cigarettes?”). The past-year assessment timeframe was chosen to capture both infrequent (less than yearly, yearly, or less than monthly) as well as more frequent (monthly, weekly, or daily) alcohol/substance use behaviors among Multiracial college students.
The three past-year substance use frequency variables originally used a 10-point Likert response scale: 0 (I did not use in the past year), 1 (1 or 2 times in the past year), 2 (3 to 11 times in the past year), 3 (Once a month), 4 (2 to 3 times a month), 5 (Once a week), 6 (Twice a week), 7 (3 to 4 times a week), 8 (5 to 6 times a week), and 9 (Every day). All three past-year alcohol (M = 4.42, SD = 2.30, variance = 5.28), cannabis (M = 2.78, SD = 3.16, variance = 9.98), and tobacco (M = 2.65, SD = 3.46, variance = 12.03) frequency variables demonstrated over-dispersion (i.e., variance greater than mean) and Shapiro-Wilk normality tests found the past-year substance use frequency variables to all have abnormally distributed residuals at p<.001. Visual inspection of histograms further indicated excess zero responses for past-year cannabis and tobacco frequency variables and non-zero responses clustering around more frequent (monthly, weekly, and daily) and infrequent (yearly and less than monthly) responses for all past-year substance use frequency variables. Thus, all three substance use frequency variables were transformed into ordinal composite variables, 0 (No yearly substance use), 1 (Infrequent or less than monthly substance use), and 2 (Frequent or monthly use; Wedel et al., 2022). Accordingly, for analyses, past-year alcohol, cannabis, and tobacco use were specified as ordinal variables and estimated as manifest variables comprising a past-year substance use frequency latent factor. Estimating substance use as a latent factor (instead of running three separate models for each of the substance use variables) helps control for potential Type I and measurement errors inherent in alcohol, cannabis, and tobacco use variables by explicitly estimating latent factors of residual/error terms for substance use variables (Tomarken & Waller 2005). Additionally, since the current study aimed to examine whether discriminatory experiences are associated with greater frequencies of substance use, a latent approach captures the underlying commonality or shared variance that alcohol, cannabis, and tobacco use have, resulting in greater interpretability of results (Park et al., 2013; 2014).
Psychological distress
The 4-item Patient Health Questionnaire-4 (Kroenke et al., 2009) assessed past 2-week frequencies of depressive and anxiety symptoms, with two items assessing anxiety symptoms and the other two items assessing depressive symptoms. Responses were rated on a 4-point Likert scale ranging from 0 (Not at all) to 3 (Nearly every day). The Patient Health Questionnaire-4 demonstrates high sensitivity and specificity in screening for psychological distress (Kroenke et al., 2009) and has been shown to be reliable and valid to use among diverse college students, inclusive of Multiracial college students (Khubchandani et al., 2016). For analyses, a sum score (α = 0.91) was used and specified as a continuous outcome variable, with higher scores suggesting greater psychological distress.
Multiracial discrimination
Fifteen items from the 30-item Multiracial Challenges and Resiliencies Scale (Salahuddin & O’Brien, 2011) evaluated Multiracial challenges or specific Multiracial discriminatory experiences. Specifically, out of a total of six subscales, three 5-item subscale scores were used to assess Multiracial discrimination including others’ surprise and disbelief regarding racial heritage (e.g., “When I disclosed my racial background, someone acted surprised”), lack of familial acceptance (e.g., “A family member said something negative about Multiracial/biracial people”), and overt discriminatory experiences attributed to Multiracial status (e.g., “Someone chose not to date me because I am Multiracial”). Participants rated the frequency of occurrence on a 6-point Likert scale ranging from 0 (Never happened to me) to 5 (Happened to me more than 10 times). This response scale is consistent with studies reporting frequency of Multiracial discrimination experiences (Franco et al., 2021b; Franco & Carter, 2019). The Multiracial Challenges and Resiliencies Scale also included a 5-item subscale assessing challenges with racial identity, however, this subscale was excluded from analyses because participants did not rate frequency of occurrence and rather rated how much they disagree or agree with the discriminatory experience (e.g., “I hide parts of myself when interacting with some friends”; 0 [Strongly disagree]–5 [Strongly agree]). Multiracial resiliencies, including the appreciation of human differences subscale, and the Multiracial pride subscale were also excluded from analyses due to not assessing Multiracial specific discriminatory experiences. The three Multiracial challenges subscales assessing frequency of occurrence of discriminatory experiences (i.e., others’ surprise and disbelief regarding racial heritage; lack of familial acceptance; overt Multiracial discrimination) have been validated with Multiracial adults (age range = 18–53; Mage=22, SD = 5.21; 28% Male), with Cronbach’s alpha in the acceptable to good range (αs = 0.79–0.83; Salahuddin & O’Brien, 2011), indicating diverse Multiracial discrimination experiences cluster together. For analyses, a sum score was computed for Others’ Surprise and Disbelief (α = 0.87, possible range = 0–25), Lack of Familial Acceptance (α = 0.82, possible range = 0–25), and Overt Multiracial Discrimination (α = 0.80, possible range = 0–25) subscales, with higher scores indicating a greater frequency of diverse Multiracial discriminatory experiences. For analyses, the three subscales were estimated as manifest variables comprising a Multiracial Discrimination latent factor.
General discrimination
The 9-item Everyday Discrimination Scale (Williams et al., 1997) assessed the frequency of general discriminatory experiences in day-to-day life (e.g., receiving poorer service than others in restaurants or stores; being called names or insulted). Responses were rated on a 6-point Likert scale ranging from 0 (Never) to 5 (Almost every day). The Everyday Discrimination Scale is widely used for assessing general discrimination, which may be attributed to various possible reasons (e.g., gender, race, age, religion, physical appearance, sexual orientation, education, or income level), and has been validated among diverse college students, inclusive of Multiracial college students (Greenfield et al., 2021; Lawrence et al., 2022; Pascoe et al., 2022). For analyses, a sum score (α = 0.82) was used and specified as a predictor variable, with higher scores suggesting a greater frequency of general discriminatory experiences.
Covariates
Age (range = 18–25), sex assigned at birth (0 = female, 1 = male; intersex was an available response option but was not endorsed), and fraternity/sorority (Greek) affiliation (0 = no, 1 = yes) were assessed and included as covariates based on previous findings demonstrating these sociodemographic characteristics’ associations with substance use (McHugh et al., 2018; Park et al., 2008; Poudel & Gautam, 2017), discrimination (Garstka et al., 2004), and psychological distress (Eid et al., 2019) among diverse college students. In all analyses, age, sex, and Greek affiliation were included as covariates.
Data analytic strategies
Descriptive statistics and bivariate correlations were computed using IBM SPSS Version 27 (IBM Corp, 2016). Means, standard deviations for continuous variables (or percentages for dichotomized variables), and bivariate correlations (i.e., Pearson’s correlation coefficients for two continuous variables, Spearman’s coefficients for continuous and dichotomous variables, and Phi coefficients for two dichotomous variables) were calculated. SPSS was used to conduct Shapiro-Wilk normality tests for continuous variables’ distributions. Multivariate outlier analyses based on Mahalanobis distance using a χ2 distribution did not identify any significant multivariate outliers (Mahalanobis D2=2.67–24.40, ps > 0.001). Past-year substance use frequency variables were transformed into an ordinal composite measure to address non-normality (for greater detail, see Measures section) and transformed alcohol (M = 1.47, SD = 0.62, variance = 0.38, skewness= −0.72, kurtosis= −0.45), cannabis (M = 0.88, SD = 0.82, variance = 0.68, skewness = 0.22, kurtosis= −0.49), and tobacco (M = 0.75, SD = 0.85, variance = 0.73, skewness = 0.51, kurtosis= −1.44) use variables and psychological distress (M = 4.49; SD = 3.79; variance = 14.37, skewness = 0.57, kurtosis= −0.77) were within an acceptable kurtosis (< |2|) and skewness (< |2|) range.
For analyses, structural equation modeling was conducted using Mplus Version 8.3 (Muthén & Muthén, 2023). Full information maximum likelihood estimation with robust standard error (MLR) was used to accommodate missing and incomplete data (Graham, 2003), which was largely found in discrimination variables due to participants who skipped the question or responded with “prefer not to answer” (Multiracial discrimination: n = 19; general discrimination: n = 29; see Table 1 for detail). Data, analytic methods, and study materials can be made available to researchers upon request to the corresponding author.
Two fully saturated structural regression models (and thus no model fit indices) were estimated to investigate the relationship of general and Multiracial discrimination (predictors) with substance use (an outcome). The first model (shown in Figure 1(a)) did not include psychological distress, whereas the second model (shown in Figure 1(b)) included psychological distress as another outcome which was correlated with the latent substance use factor. In both models, two continuous latent factors were estimated; Multiracial discrimination was estimated from three manifest/observed variables of others’ surprise and disbelief, lack of familial acceptance, and overt Multiracial discrimination, and substance use was estimated from three manifest/observed ordinal variables of alcohol, cannabis, and tobacco use frequencies. Covariates (i.e., age, sex, and Greek affiliation) were also included in the models as predictors of substance use and psychological distress. As a measure of effect size, standardized regression coefficients and their 95% confidence intervals (CIs) based on 10,000 resamples (Efron, 1987) for each path were used.
For exploratory ancillary analyses, two additional fully saturated path models tested the mediating role of psychological distress in the discrimination-substance use association. The first path model evaluated the mediating role of psychological stress in the association of discrimination with substance use (after accounting for the direct discrimination-substance use relation). The cross-sectional nature of our data precluded temporal ordering of study variables, and thus, a second path model tested an alternative mediating role of substance use in the association of discrimination with psychological distress. Participant age, sex, and Greek affiliation were included as covariates in ancillary models. Bias-corrected bootstrapped 95% confidence intervals (CIs) based on 10,000 resamples (Efron, 1987) were used to determine significance of indirect associations, with CIs that did not include zero, indicating statistical significance.
Results
Descriptive statistics
Means and standard deviations for continuous variables, percentages of dichotomized variables, and bivariate correlations for covariates, discrimination, psychological distress, and substance use variables are reported in Table 1.
Nearly all (93%) Multiracial college students in our sample reported past-year or more frequent alcohol use, with 60% of Multiracial college students reporting past-year or more frequent cannabis use, and 48% of Multiracial college students reporting past-year or more frequent tobacco use, higher than past-year rates reported in a nationally representative sample of Multiracial college-aged individuals (SAMHSA, 2024).
A majority of Multiracial college students in our sample reported experiencing others’ surprise and disbelief (74%), lack of familial acceptance (52%), overt Multiracial discrimination (69%), and general discrimination (71%) at least once in their lives. Multiracial college students reported psychological distress (M = 4.49, SD = 3.79, range = 0–12) indicating mild depressive and anxiety symptoms on average (Kroenke et al., 2009). This average is higher than a largely White college student sample at a large Midwestern university in the U.S. (M = 2.98, SD = 0.08, range = 0–12, 80% White; Khubchandani et al., 2016).
Bivariate correlations indicated that age was positively correlated with past-year alcohol use (r = 0.17), females were more likely to report psychological distress (r=−0.24), and Greek affiliation was positively associated with alcohol use (r = 0.39). Others’ surprise and disbelief, lack of familial acceptance, overt Multiracial discrimination, and general discrimination were positively correlated with one another (rs = 0.31–0.70), and all experiences of discrimination demonstrated moderate positive correlations with psychological distress (rs = 0.26–0.42). While no experiences of discrimination were correlated with past-year alcohol use, all experiences of discrimination were positively correlated with past-year cannabis use (rs = 0.19–0.37), and lack of familial acceptance was positively correlated with past-year tobacco use (r = 0.19). Alcohol, cannabis, and tobacco use were positively correlated with one another (rs = 0.23–0.27; see Table 1).
Structural regression model without psychological distress
The results of the cross-sectional structural regression model evaluating the association of general and Multiracial discrimination with substance use, after adjusting for covariates (age, sex, and Greek affiliation) are presented in Figure 1(a). Paths for covariates are not shown in Figure 1 for simplicity.
Alcohol (β = 0.50, 95% CI [0.25, 1.00], p = 0.02, R2=0.25), cannabis (β = 0.74, 95% CI [0.27, 0.97], p < 0.001, R2=0.55), and tobacco (β = 0.46, 95% CI [0.21, 0.90], p = 0.005, R2=0.21) use significantly and positively loaded onto the latent substance use variable. Others’ surprise and disbelief (β = 0.79, 95% CI [0.67, 0.89], p < 0.001, R2=0.63), lack of family acceptance (β = 0.67, 95% CI [0.51, 0.79], p < 0.001, R2=0.45), and overt Multiracial discrimination (β = 0.89, 95% CI [0.80, 0.98], p < 0.001, R2=0.80) also significantly and positively loaded onto the latent Multiracial discrimination variable. Multiracial discrimination (β = 0.29, 95% CI [0.02, 0.57], p = 0.04), but not general discrimination (β = 0.04, 95% CI [−0.35, 0.28], p = 0.82), was associated with substance use, after adjusting for covariates.
Structural regression model with psychological distress
The results of the cross-sectional structural regression model evaluating the association of general and Multiracial discrimination with substance use and psychological distress, after adjusting for covariates (age, sex, and Greek affiliation) are presented in Figure 1(b). Paths for covariates are not shown in Figure 1 for simplicity.
With psychological distress in the model as an outcome, neither Multiracial discrimination (β = 0.23, 95% CI [−0.06, 0.48], p = 0.11), nor general discrimination (β=−0.07, 95% CI [−0.39, 0.22], p = 0.65) were associated with substance use, after adjusting for covariates. Instead, Multiracial discrimination (β = 0.26, 95% CI [0.06, 0.46], p = 0.01), but not general discrimination (β = 0.15, 95% CI [−0.04, 0.33], p = 0.11), was associated with psychological distress. In turn, psychological distress and substance use were positively associated with each other (β = 0.39, 95% CI [0.05, 0.59], p = 0.003). These results suggest that psychological distress may account for the association of Multiracial discrimination with substance use.
Ancillary analyses
Psychological distress significantly mediated the association of Multiracial discrimination (β = 0.10, 95% CI [0.01, 0.20], p = 0.04), but not general discrimination (β = 0.06, 95% CI [−0.01, 0.15], p = 0.17) with substance use. Similarly, substance use significantly mediated the association of Multiracial discrimination (β = 0.11, 95% CI [0.01, 0.28], p = 0.04) but not general discrimination (β = 0.003, 95% CI [−0.14, 0.11], p = 0.96) with psychological distress.
Discussion
The current cross-sectional study of 193 Multiracial college students investigated the relation of Multiracial and general discrimination with psychological distress and substance use. Findings demonstrated associations of Multiracial-specific discrimination with substance use, beyond general discrimination. While prospective replication is needed to establish temporality, the exploratory results of this study also suggest that psychological distress may play an intermediary role in the Multiracial discrimination-substance use association. Overall, findings underscore Multiracial discrimination as a racially relevant stressor, potentially contributing to substance use disparities among Multiracial college students.
Substance use was highly prevalent in this sample of Multiracial college students, with past-year cannabis use demonstrating a higher factor loading on the substance use factor compared to past-year alcohol and tobacco use. While substance use prevalence in the study sample exceeds national estimates for Multiracial college-aged individuals 18–25 (NSDUH, 2024), findings are consistent with broader trends demonstrating that cannabis use may be elevated among Multiracial compared to monoracial college-aged individuals. Indeed, national data show that 44% of Multiracial college-aged individuals 18–25 report past-year cannabis use, higher than rates among their Asian (24%), Hispanic/Latino (36%), Black (38%), and White (41%) counterparts. In the absence of prevalence data specific to Multiracial college students, future research with a larger more generalizable sample is warranted. Such research can clarify patterns of substance use among Multiracial college students specifically, and whether observed discrepancies reflect differences in general substance use patterns between college-attending and non-college attending populations (Schulenberg et al., 2021), or whether aspects of the college environment uniquely confer elevated risk for cannabis and other substance use for Multiracial students.
Results of this study indicated that lifetime experiences of Multiracial discrimination were positively associated with past-year substance use frequency among Multiracial college students. Novel to the limited literature, findings demonstrated that non-familial forms of Multiracial discrimination, such as others’ surprise and disbelief (e.g., “When I disclosed my racial background, someone acted surprised”) and overt Multiracial discrimination (e.g., “Someone chose not to date me because I am Multiracial”) had stronger factor loadings to the latent Multiracial discrimination factor than lack of familial acceptance. Results align with qualitative research documenting peer-based Multiracial discrimination on college campuses (Harris, 2016; Museus et al., 2016). Although prior work has shown that familial discrimination predicts substance use above and beyond overt Multiracial discrimination among Multiracial adults (Franco & Carter, 2019), findings of this study suggest that both peer and familial Multiracial discrimination may contribute to substance use risk among Multiracial college students. While replication is needed, results underscore college as a salient developmental context in which both peer and familial Multiracial discrimination may heighten susceptibility to substance use behaviors.
Multiracial-specific discrimination, but not general discrimination, positively predicted substance use frequency among Multiracial college students. Findings extend the limited literature by demonstrating that Multiracial discrimination is a racially relevant stressor for Multiracial college students, distinct from general discrimination, which can be attributed to any marginalized social identity. Consistent with the minority stress framework (Meyer, 2003), the results of this study provide support for Multiracial discrimination as a form of minority stress. As minority stressors are well-established predictors of physical and mental health disparities among individuals from marginalized backgrounds (for a review see, Hoy-Ellis, 2023), future prospective investigation is critical to identify risk pathways specific for Multiracial college students.
One such mechanism may be coping-motivated substance use. Findings indicated that while Multiracial discrimination was positively associated with substance use, when psychological distress was added to the second model, Multiracial discrimination no longer predicted substance use. Instead, Multiracial discrimination was associated with psychological distress, and psychological and substance use were associated with one another, suggesting psychological distress may account for the Multiracial discrimination-substance use association. In support, exploratory ancillary analyses found that psychological distress may play a mediating role in the Multiracial discrimination-substance use association. However, ancillary analyses also supported the reversed mediating role of substance use in the Multiracial discrimination-psychological distress association. Results from the current study are consistent with emerging research showing that perceived racial ambiguity (a form of Multiracial discrimination) increases likelihood for substance use and psychological distress (i.e., depressive symptoms; Atkin et al., 2025) and extends these findings by offering preliminary support for minority stress and tension-reduction models. Multiracial discrimination may predict proximal stress processes (e.g., expectation of rejection; internalized social messages), and in turn, psychological distress, prompting coping-motivated substance use. Prospective research is needed to clarify this potential risk pathway among Multiracial college students. Such insights could inform tailored prevention and intervention efforts. For instance, while discrimination is not a readily modifiable environmental factor, acknowledging Multiracial discrimination as a result of monoracism (Johnston-Guerrero et al., 2020) and fostering adaptive coping strategies to manage psychological distress (Gabriel et al., 2022) may serve as intervention targets for addressing substance use.
Limitations and future directions
While these novel cross-sectional findings augment limited findings on the discrimination-substance use relation among Multiracial college students, several limitations call for future research. First, due to the cross-sectional nature of the data, temporal, let alone causal, conclusions could not be drawn. Future prospective investigation is critical to understand risk pathways contributing to Multiracial health disparities. Second, this convenience sample was drawn from a predominantly White college campus. Additionally, significant group differences were found in recruitment methods, such that participants who were recruited using the psychology participant pool (nearly 90% of the sample), reported higher frequency of Multiracial discriminatory experiences and higher psychological distress compared to participants recruited via flyers. Replication of findings is warranted among Multiracial college students in more diverse campus communities to determine generalizability of study findings. Third, sample size restrictions disallowed for investigation of subgroup differences among Multiracial college students. While Multiracial discrimination, as a manifestation of monoracism (i.e., systemic social oppression grounded in the belief that race exists in discrete categories; Johnston & Nadal, 2010, p. 125), may be a shared experience among Multiracial college students, given the large intragroup racial heterogeneity, it is possible that different Multiracial groups have distinct experiences of Multiracial discrimination (Harris, 2016), resulting in subgroup differences in substance use. Indeed, research has found that Multiracial young adults with White ancestry, compared to Multiracial young adults with non-White ancestry were more likely to drink to get drunk, however, subgroup differences were not found when assessing binge drinking or illicit drug use outcomes (Atkin et al., 2025). Additionally, the role of psychological distress may also vary by assessed substance use indicator, potentially through motives for substance use (Atkin et al., 2025). Mixed findings warrant clarification on whether Multiracial subgroups align with their intermediate corresponding monoracial counterparts in relation to alcohol use (e.g., Goings et al., 2020) or reflect variability in discrimination-alcohol use relationships depending on the assessed indicator (i.e., quantity, frequency, alcohol use disorder; Gilbert & Zemore 2016). Thus, well-powered future research is warranted to disentangle the relation between Multiracial discrimination, psychological distress, and varying substance use indicators. Lastly, this study conceptualized discrimination as a particularly salient stressor for Multiracial college students. Consistent with intersectional hypotheses (Moradi et al., 2010), Multiracial individuals may draw on resiliencies to buffer the discrimination-substance use relation (Shih et al., 2019). Future research should consider what constitutes resilience and is protective for Multiracial college students.
Conclusion
Multiracial college students report higher rates of alcohol/substance use than their monoracial peers, yet factors associated with this disparity remain understudied. This cross-sectional study addresses a critical literature gap by examining associations of discrimination, psychological distress, and substance use. The results underscore others’ surprise and disbelief, lack of familial acceptance, and overt Multiracial discrimination as predictive of substance use (beyond general discrimination), with this relation potentially accounted for by psychological distress. Although prospective replication is needed, findings provide preliminary support for minority stress and tension reduction models, identifying a potential risk pathway to elevated substance use among Multiracial college students, a critically understudied but rapidly growing population.
Funding
Preparation of this article was supported by the National Institute on Alcohol Abuse and Alcoholism F31 AA031428 awarded to Fatima Dobani and R01 AA027677 awarded to Aesoon Park. The authors report no conflicts of interest.
| Variable (possible range) | n (%) | M (SD) or % | Correlation Coefficients | 9 | 10 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |||||
| 1. Age (18–25) | 193 (100%) | 20.34 (1.32) | – | |||||||||
| 2. Male sex (1 vs. 0) | 193 (100%) | 30% | .19 | – | ||||||||
| 3. Fraternity/sorority affiliation (1 vs. 0) | 187 (96%) | 31% | −.01 | −.06 | – | |||||||
| 4. Others’ Surprise and Disbelief (0–25) | 174 (90%) | 6.68 (6.96) | −.04 | −.04 | −.02 | – | ||||||
| 5. Lack of Familial Acceptance (0–25) | 174 (90%) | 2.75 (4.58) | −.06 | −.08 | −.09 | .57 | – | |||||
| 6. Overt Multiracial Discrimination (0–25) | 174 (90%) | 4.60 (5.48) | −.03 | −.03 | −.05 | .70 | .60 | – | ||||
| 7. General discrimination (0–45) | 164 (84%) | 15.15 (6.14) | −.04 | .03 | −.21 | .34 | .31 | .53 | – | |||
| 8. Psychological Distress (0–12) | 173 (89%) | 4.49 (3.79) | −.16 | −.24 | −.12 | .27 | .42 | .26 | .29 | – | ||
| 9. Past-year alcohol frequency (0–2) | 184 (95%) | 1.46 (0.61) | .17 | .02 | .39 | .10 | .03 | .02 | −.11 | .09 | – | |
| 10. Past-year cannabis frequency (0–2) | 182 (94%) | 0.88 (0.82) | −.01 | .01 | −.02 | .21 | .37 | .25 | .19 | .29 | .27 | – |
| 11. Past-year tobacco frequency (0–2) | 182 (94%) | 0.74 (0.85) | −.01 | −.02 | −.05 | .09 | .19 | .14 | .05 | .23 | .23 | .25 |