The purpose of this study is to examine the prevalence estimates of four substances among adolescents and how they may vary based on Multiracial identity and Hispanic/Latino/a/x ethnicity. To address this, the current study aims to describe the within-group heterogeneity for the Multiracial and Hispanic/Latino/a/x groups for prevalence of any past 30-day use in alcohol, marijuana, combustible tobacco (e.g., cigarettes, cigars), and e-cigarettes. Given variation in the racial composition across states, we opted to focus our investigation on a single state. We used data from the MD YRBS/YTS. Maryland has a population of over 6 million that is 58.5% monoracial White, 31.1% monoracial Black, 10.6% Hispanic/Latino/a/x, and 2.9% Biracial or Multiracial [21]. Findings from the current study will potentially guide new approaches for classifying students by Multiracial identity and Hispanic/Latino/a/x ethnicity in substance use research. Moreover, examining within-group differences is an important step toward identifying structural determinants of risk behaviors among Multiracial and Hispanic/Latino/a/x youth, two marginalized groups.

Data comes from the 2018 Maryland Youth Risk Behavior Survey/Youth Tobacco Survey (MD YRBS/YTS) [22,23], which is conducted biennially in conjunction with the CDC surveillance systems. The Maryland dataset offers the largest YRBS sample of all states participating in the YRBS and is designed to yield prevalence estimates at the county-level for all 24 counties in the state. The 2018 MD YRBS/YTS is a subset of the nationwide YRBS survey combined with the Youth Tobacco Survey (YTS). For each county-level public school system, the MD YRBS/YTS implemented a two-stage cluster design with first stage sampling on the school-level and second-stage sampling of classrooms within schools [22]. YRBS/YTS developed sampling weights for individual students based on student race and ethnicity, sex, and grade to account for school, student nonresponse, and oversampling of Black and Hispanic students [22]. Use of the sampling weights in the MD YRBS/YTS dataset produces estimates for the high school student population in the state of Maryland. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board concluded that the current study is not considered human subjects research and did not require review.

The aim of the current paper was to explore how categorizations of race by Hispanic/Latino/a/x ethnicity impacts prevalence estimates of substance use in adolescents. Our exploration of the Maryland YRBS/YTS sample showed heterogeneity in the racial composition of Hispanic/Latino/a/x, and Biracial and Multiracial youth. Moreover, prevalence estimates of substance use among Hispanic/Latino/a/x and Multiracial youth showed a wider range than estimates using the CDC’s standard classifications of race/ethnicity. These findings demonstrate the importance of estimating substance use with a more precise racial and ethnic classifications.

As previously described, the U.S. population is becoming increasingly diverse with growing Hispanic or Latino/a/x, Biracial and Multiracial populations. Findings from the current study demonstrate that these populations have heterogeneous substance use rates, and the current use of multiple race categories may be under- or over-estimating substance use prevalence. A central implication is the need for future research to take these varying racial and ethnic identities into account. Thus, researchers in the field of substance use epidemiology should incorporate analyses and explorations of substance use among various Biracial and Multiracial groups.

Findings from the current study also highlight the need to address racial identity among Hispanic or Latino/a/x populations when using health behavior surveys. Many Hispanic or Latino/a/x youth in our sample did not identify as any of the five race categories, perhaps because they did not identify with any of the standard OMB racial categories. Examining race with Hispanic/Latino/a/x populations comes with unique and complex challenges due to conflicting conceptualizations between policymakers, researchers, and within the Hispanic/Latino/a/x population itself [25–26]. Given that the Hispanic/Latino/a/x population is often treated as a monolithic racial group in U.S. society and in health research [27], youths’ nonidentification with the standard OMB racial categories may reflect a more racialized panethnic identity [28–29]. The lack of congruence between some Hispanic/Latino/a/x youth’s subjective understanding of their racial identity, in combination with current measurement of race in research methodology, may elicit responses to questions of race that do not speak to their lived identity and have no significant meaning for them [26, 30]. Nevertheless, similar to Biracial/Multiracial youth, the collapsing of all races under a single “Hispanic/Latino/a/x” category ignores a significant amount of heterogeneity and diversity that can impact substance use prevalence estimates. Currently, the YRBS does not include measurement of nationality, further limiting our understanding of the diversity among Hispanic and Latino/a/x youth in the survey.

The current study expands upon prior research with substance use among Biracial/Multiracial and Hispanic/Latino/a/x youth by exploring more precise racial categorizations within each of these groups. While there is a growing literature showing high prevalence of substance use among the monolithic group of Multiracial youth [13,15], findings from the current study show that there is a greater variation within Biracial and Multiracial youth. For example, Non-Hispanic Asian-White youth tended to have lower substance use compared to the CDC Multiracial category. Similarly for Hispanic/Latino/a/x youth, whose substance use estimates in previous studies have been moderate compared to other racial and ethnic groups, findings from the current study show some groups with much higher substance use prevalence (e.g., Black Hispanic/Latino/a/x youth) and other with much lower use (e.g., Asian Hispanic/Latino/a/x youth marijuana use). Moreover, the study estimates the use of several substances, including e-cigarettes, among these different racial categorizations. Studies of e-cigarette use are increasingly important given their high prevalence of use among adolescents and increasing evidence that e-cigarette use increases the risk of future combustible tobacco use and negative health effects [31]. Indeed, e-cigarette use among youth has grown so great that recent federal legislation was passed to eliminate the sale of flavored e-cigarette cartridges to reduce their appeal to youth [32]. Results from the current study show that difficulties in youth health and substance use surveillance are partly attributable to imprecise measurements of race among Biracial/Multiracial and Hispanic or Latino/a/x youth.

Although the current study makes a significant contribution to the literature, our findings should be considered in light of their limitations. Though we selected the Maryland YRBS dataset due to its large sample size, there were still subgroups of Biracial and Multiracial youth with small sample sizes, limiting our ability to estimate their substance use with any statistical rigor. Additionally, the use of Maryland data limits the generalizability of results. However, similar analyses conducted with the current sample could be replicated using YRBS data from other states. Moreover, as previously mentioned, the YRBS does not include measurement of nationality to further describe subgroups of Hispanic or Latino/a/x youth (i.e., Mexican, Puerto Rican, Guatemalan, etc.), nor does it include other terms that may be more descriptive of how Hispanic or Latino/a/x youth identify (e.g., “mestizo”, “mulatto”) [33]. Additionally, the YRBS does not measure other contextual factors that might impact stress and substance use among Hispanic or Latino/a/x youth (e.g., acculturation stress, religious affiliation, immigration status). Further, the YRBS does not measure skin tone, which could be a useful measure of Hispanic/Latino/a/x and Multiracial youths’ lived experiences of race and racial discrimination [34]. A further limitation of this research is that some people may report on distant heritage, particularly American Indian heritage. Based on the way race is currently measured, it is a challenge to differentiate those with distant American Indian heritage from those who have a direct cultural affiliation. This difficulty could be addressed by asking about parental race when measuring adolescent race. Additionally, researchers need to conduct qualitative research with adolescents to learn how and why they answer questions on race to aid in the formulation of race measures. Youth-led participatory action research (YPAR) focused on the expertise, lived experiences, and voices of Hispanic/Latino/a/x and Multiracial youth would be key in creating relevant measurement of race and ethnicity that is representative of youths’ racial and ethnic identities [35]. Finally, the current study did not incorporate information on sex, gender identity, or sexual orientation; however, future work is necessary to examine how these factors intersect with race and ethnicity to better understand their contextual influences on adolescent substance use, including experienced homophobia and transphobia as further sources of minority stress.

The current study demonstrates the need for future measurements of race and ethnicity in substance use epidemiology studies to make specific changes. Substance use studies at state and local levels must be mindful about how best to assess race and ethnicity, given the complexity of racial and ethnic identity and population dynamics. Findings from this type of research can help bring to light structural and cultural risk and protective factors for substance use, as well as other health behaviors, that are salient to Hispanic/Latino/a/x and Multiracial adolescents.