Abstract
This study examined the relationship between perceived racial discrimination (PRD) and patterns of substance use. Data come from the 2001-2003 National Survey of American Life (N=3,589). PRD was derived from the Major Experiences of Discrimination Scale. Multinomial logistic regression estimated the relationship between PRD and patterns of substance use (i.e., never, single-substance, dual-substance, and polysubstance (3+ substances)) based on six substances; effect modification by ethnicity and sex was assessed by stratification. Study findings indicate that PRD was associated with greater odds of lifetime and current polysubstance use. Results from the effect modification analyses suggested differential associations by ethnicity and sex.
Keywords: Polysubstance use, discrimination, race, ethnicity, minority health
Introduction
Approximately 90% of US Black adults who report experiencing discrimination in their lifetimes attribute race/ethnicity as the primary reason (Kessler, Mickelson, & Williams, 1999). Recently, there has been increasing attention to the ways in which race-salient exposures, such as perceived racial discrimination (PRD), relate to the mental health of minorities in the US (T. T. Clark, 2014; T. T. Clark, Salas-Wright, Vaughn, & Whitfield, 2015; Hunte and Barry, 2012; Kessler, et al., 1999; Soto, Dawson-Andoh, & BeLue, 2011). While a large body of research has documented associations between PRD and related exposures with symptoms of depression (T. T. Clark, 2014; Kessler, et al., 1999) and anxiety (Kessler, et al., 1999; Soto, et al., 2011) among Black Americans, the links between PRD and externalizing behaviors such as substance use and misuse are less clear.
Polysubstance, also known as polydrug use, is defined as the concurrent use of more than one non-prescribed psychoactive substance (Connor, Gullo, White, & Kelly, 2014; Smith, Farrell, Bunting, Houston, & Shevlin, 2011), and is associated with psychiatric comorbidity and other poor health outcomes, including elevated risk of accidental overdose (Connor, et al., 2014; Salom, Betts, Williams, Najman, & Alati, 2016; Schneider, Park, Allen, Weir, & Sherman, 2019; Smith, et al., 2011). Recent investigations have explored the prevalence, predictors, and patterning of polysubstance use in adult (Evans, Grella, Washington, & Upchurch, 2017; Salom, et al., 2016; Schneider, et al., 2019; Smith, et al., 2011) and youth (Chung, Kim, Hipwell, & Stepp, 2013; Moss, Chen, & Yi, 2014; Silveira, Green, Iannaccone, Kimmel, & Conway, 2019; Tomczyk, Isensee, & Hanewinkel, 2016) populations. Existing evidence suggests that PRD is associated with increased risk of substance use and substance use disorders (Assari, Mistry, & Caldwell, 2018; Assari, Mistry, Lee, Caldwell, & Zimmerman, 2019; Borrell et al., 2010; Borrell et al., 2007; Borrell, Kiefe, Diez-Roux, Williams, & Gordon-Larsen, 2013; T. T. Clark, 2014; T. T. Clark, et al., 2015; Hunte and Barry, 2012; Rose et al., 2018), such as the use of alcohol (Borrell, et al., 2010; Borrell, et al., 2007; Borrell, et al., 2013; Rose, et al., 2018), tobacco (Assari, et al., 2018; Borrell, et al., 2010; Borrell, et al., 2007; Borrell, et al., 2013; Rose, et al., 2018), and marijuana (Assari, et al., 2019; Borrell, et al., 2007; Rose, et al., 2018). However, investigation on the relationship between PRD and patterns of substance use, including polysubstance use, among adult samples is limited.
Attempts to understand how exposures like PRD relate to substance use behaviors requires nuanced consideration of how race intersects with sex and other identities. Intersectionality frameworks derived from feminist and critical race theories postulate that variations in social advantages or disadvantages originate from multiple categories of identities, including race, ethnicity, sex, and gender (Cole, 2009; Shields, 2008). That is, when examining the relationship between race and substance use, for example, sex is not simply a variable to adjust for in multivariable analysis; rather, understanding sex differences is an objective of substantive scientific interest. Consistent with this notion, several studies have reported that polysubstance use and subsequent polysubstance-related disorders vary by race/ethnicity and sex/gender (Chung, et al., 2013; Evans, et al., 2017; Moss, et al., 2014; Salom, et al., 2016; Silveira, et al., 2019; Tomczyk, et al., 2016). For example, a recent study using the National Longitudinal Study of Adolescent to Adult Health reported that polysubstance use was associated with substance use disorders in young adulthood, and that this relationship was stronger among adolescent males compared to females (Moss, et al., 2014). Another study explored the influence of ethnicity among US Black women, finding that African American women have higher prevalence of substance use disorders compared to Afro-Caribbean women (Broman, Neighbors, Delva, Torres, & Jackson, 2008).
There are several proposed processes underlying these racial/ethnic and sex differences in substance use disorders, including perceived social and cultural consequences of substance use (Mulia, Ye, Greenfield, & Zemore, 2009; Unger et al., 2002), substance preferences (Wu, Zhu, & Swartz, 2016), and self-regulatory stress coping behaviors (R. Clark, Anderson, Clark, & Williams, 1999; Gerrard et al., 2012). For example, males who experienced discrimination are more likely to externalize behaviors and use substances compared to females (Assari, et al., 2018; Brody, Kogan, & Chen, 2012), while females are more likely to experience other behaviors such as binge eating (Assari, 2018) and exercising more frequently (Brodish et al., 2011) in response to these stressors. Variation in coping behaviors by race/ethnicity also exist. For example, discrimination is associated with alcohol use among Hispanics and tobacco use among non-Hispanic (NH) Whites, while both alcohol and tobacco consumption are associated with discrimination among NH Blacks (Borrell, et al., 2010). Findings indicating that generalized anxiety disorder is associated with discrimination in African Americans but not Afro-Caribbeans and NH Whites suggest possible differential effects of internalizing behaviors by race/ethnicity (Soto, et al., 2011).
To date there has been limited research exploring the relationship between PRD and patterns of substance use, including how ethnicity and sex shape this relationship, within the US Black adult population. Three prior studies on this topic all reported that discrimination was associated with greater substance use outcomes among Black adults (T. T. Clark, 2014; T. T. Clark, et al., 2015; Hunte and Barry, 2012), and one reported that this relationship was primarily mediated by depression for Afro-Caribbean but not African American young adults (T. T. Clark, 2014). These studies helped illustrate the importance of discrimination to substance use outcomes and can be built upon to further understand the patterns in specific ways. For example, none of the previous studies examined the role of attribution (i.e., perception of the primary reason respondents felt they were discriminated against) as an exposure; rather, they evaluated the overall effect of any form of discrimination (i.e., irrespective of attribution). As illustrated by the intersectionality framework, the salience of race, ethnicity, sex, and gender may shape how discriminatory experiences are perceived and evaluated (Bowleg, 2012). However, these studies did not evaluate the relationship between discrimination and patterns of substance use or of polysubstance use specifically as an outcome, instead examining individual substances separately or substance use disorders. In addition, the US Black population is heterogeneous in terms of ethnic identity, birthplace, and other sociocultural norms, and these factors may play a role in the manner people experience, perceive, and attribute racial discrimination, as well as the links between PRD and behaviors like patterns of substance use (Georgiades, Boyle, & Duku, 2007).
The present study builds on prior work by examining heterogeneity within the Black population for the relationship between PRD and patterns of lifetime and current substance use, including polysubstance use, in a large, nationally representative sample of US Black adults. In addition, we examined the extent to which the relationship between PRD and patterns of substance use differed by ethnicity, by sex, and by ethnicity and sex, using intersectionality frameworks that call attention to the sources of variation within groups.
Material and Methods
Data
The National Survey of American Life (NSAL) is a nationally representative, cross-sectional study from the National Institute of Mental Health Collaborative Psychiatric Epidemiology Surveys Initiative (Heeringa et al., 2004; Jackson et al., 2004). The NSAL, conducted between February 2001 and June 2003, investigated the mental health status of native (African American) and Caribbean foreign-born (Afro-Caribbean) Blacks in the US (Jackson, Torres, et al., 2004). The study employed a national multi-stage probability sampling strategy and participants were matched on race/ethnicity to interviewers for face-to-face interviews (Jackson, Neighbors, Nesse, Trierweiler, & Torres, 2004). The NSAL was approved by the University of Michigan Institutional Review Board; this analysis used publicly available data from the Inter-university Consortium for Political and Social Research (Alegria, Jackson, Kessler, & Takeuchi, 2016).
Participants
The NSAL sample was comprised of NH White (n=891), Hispanic (n=183) African American (n=3,570) and Afro-Caribbean (US-born and foreign-born, n=1,438) adults (Jackson, Neighbors, et al., 2004). The present study excluded NH White and Hispanic participants (n=1,074) and participants missing data for the Major Experiences of Discrimination Scale (MEDS) (n=641) or substance use, including use of cigarettes, alcohol, marijuana, cocaine, prescription drugs, and other drugs (n=759). Participants missing information on their age at immigration were also excluded (n=19). The final analytic sample consisted of 3,589 participants.
Measures
Perceived Racial Discrimination.
PRD was measured using the 9-item Major Experiences of Discrimination Scale (MEDS) (Williams et al., 2008). The scale assessed lifetime exposures of ways in which participants were treated or how participants perceived being treated by others in nine scenarios (Williams, et al., 2008) (see Supplementary Table 1). If participants responded “yes” to any of the items on the scale, they were prompted to indicate the main reason why they believed the event occurred (i.e., ancestry or national origins, gender, race, age, religion, height, weight, some other aspect of physical appearance, sexual orientation, education or income level). The PRD index aggregated the nine experiences of unfair treatment into a three-category variable, focused on racial attribution. The categories were: none of the nine experiences (i.e., participants who indicated experiencing none of the nine events or participants who did experience at least one event but did not indicate race as one of the attributions), at least one event experienced and attributed to race, and two or more events experienced and attributed to race. These categories were chosen to emphasize differences in not only experiencing unfair treatment and attributing it to race, but to also isolate differences between PRD originating from multiple unique instances, compared to PRD derived from one instance, and discrimination not attributed to race or no experience of discrimination at all. The Cronbach’s alpha internal reliability score of the nine variables used to contextualize scenarios of racial/ethnic discrimination was 0.64.
Patterns of Substance Use.
Participants were asked about lifetime (ever) or current use of the following substances: cigarettes, alcohol, marijuana, cocaine, prescription drugs, and other drugs in the NSAL. This study includes two primary outcomes: lifetime substance use and current substance use.
For lifetime cigarette use, participants answered if they had smoked more than 100 cigarettes in their lifetime, and if they answered “yes”, then they were considered lifetime cigarette users. Because NSAL did not include an “ever” alcohol use variable, participants who ever drank twelve drinks in a year were considered lifetime alcohol users. Participants were asked if they had used marijuana or prescription drugs without recommendation from a health professional, and those who indicated “yes” were considered lifetime users of that respective substance. In addition, participants were asked if they had ever used cocaine or other drugs even once. Questions about participants’ use of prescription drugs, including tranquilizers, stimulants, pain killers, or other drugs, referred to unhealthy use (i.e., use either without recommendation from a health professional or for any reason other than what a health professional advised to use them for). Use of other drugs included heroin, opium, glue, LSD, peyote, or any other drug.
Current use of cigarettes was defined as participants who indicated that they “currently” smoked. Current use of the remaining substances was defined using past-year criteria. For alcohol use, participants who indicated that they drank nearly every day, 3-4 days per week, 1-2 days per week, 1-3 days per month, and less than once a month in the past year were defined as current users. For marijuana, cocaine, prescription drug, and other drug use, current users were defined as participants who responded “yes” to use in the past year.
Two patterns-of-substance use variables were created by categorizing participants into the following mutually exclusive use groups: never used any substance, single-substance users, dual-substance users, and polysubstance users, based on the possible combinations of use among the six substances for both lifetime use and current use, respectively. Dual-substance use indicated use of any combination of two substances, and polysubstance use indicated use of any combination of three or more substances.
Covariates.
Age, immigration status, sex, ethnicity, education, and household income were included in this analysis as potential confounders. Immigration status was categorized into a three-level variable consisting of: US born, aged less than 18 years old at migration, and aged 18 years or older at migration. Participants reported their sex as either male or female, and ethnicity as either African American or Afro-Caribbean. Education was categorized into less than high school, high school graduate, some college, and college or more, whereas household income was operationalized into four groups: less than $30,000, $30,000-$49,999, $50,000-$74,999, and more than $75,000 per year and in US dollars.
Statistical Analysis
Weighted frequencies and percentages of categorical variables and means and standard deviations of continuous variables were compared across the substance use groups. Chi-square and ANOVA tests were computed to evaluate bivariate differences between the four substance use patterns (never, single-substance, dual-substance, polysubstance use). Logistic regression models were computed to obtain odds ratios and 95% confidence intervals for the relationships between the independent variable, PRD (referent: none of the nine experiences), and the dependent variables, lifetime substance use (referent: never) and current substance use (referent: never/ever), of each individual substance and patterns-of use-variable, adjusted for age, sex, ethnicity, education, and household income. Two adjusted multivariable multinomial logistic regression models were used to examine associations between PRD and lifetime patterns of substance use, and PRD and current patterns of substance use, with reporting none of the nine racial experiences as the exposure referent group and never users as the outcome referent group. A two-way interaction for PRD and ethnicity, a two-way interaction for PRD and sex, and a three-way interaction for PRD, ethnicity, and sex were tested in a single model to investigate if associations between PRD and lifetime substance use differed across levels of ethnicity (African American v. Afro-Caribbean), sex (female v. male), and ethnicity and sex (African American female v. African American male v. Afro-Caribbean female v. Afro-Caribbean male), respectively. For analyses involving current patterns of substance use, dual-substance and poly-substance use groups were collapsed. Analyses were stratified by ethnicity, by sex, and by ethnicity and sex for both lifetime and current patterns of substance use to assess effect modification, recognizing that we may have been underpowered to tests three-way interactions, and to explore differences within these subgroups. The Benjamini-Hochberg (false discovery rate) procedure was used to correct for multiple testing by adjusting associated stratum-specific p-values in effect modification analyses (Benjamini & Hochberg, 1995). Sensitivity analyses excluding participants who drank less than once a month in the past year were were robust when using this more conservative measure with respect to alcohol consumption. All associations were examined accounting for the complex survey design by adjusting for primary sampling units, strata, and survey weights. Data were analyzed using Stata, version 15.1 (StataCorp).
Results
Participant Characteristics
The distribution of participants by each MEDS item and the PRD index is shown in Supplementary Table 1. Over half of the sample (58.0%) indicated that they either had not experienced any of the nine events of the MEDS, or if they had, they did not attribute any of the events to race. Table 1 describes the weighted distributions of descriptive characteristics overall and by the PRD Index. One-fourth of the sample were lifetime single-substance users, 28% were lifetime dual-substance users, and 31% were lifetime polysubstance users. Nearly 39% were current single-substance users, 19% were current dual-substance users, and 7% were current polysubstance users. Nearly half of the respondents who attributed at least two discriminatory events to race were lifetime polysubstance users; however, only 12.4% of these respondents were current polysubstance users.
Table 1.
Descriptive Characteristics of African American and Afro-Caribbean Adults by Prevalence of PRD: The National Survey of American Life (NSAL), 2001-2003 (N=3,589)
| PRD Indexa | |||||
|---|---|---|---|---|---|
| Descriptive Characteristics | Total (N=3,589) |
No experiences (n=2,184) |
One Experience (n=742) |
At least two experiences (n=663) |
P-valueb |
| Age ± SD | 41.5 ± 0.5 | 41.0 ± 0.6 | 41.2 ± 0.8 | 43.5 ± 0.7 | 0.058 |
| Immigration status, n (%) | 0.091 | ||||
| US born | 2880 (94.3) | 1706 (93.6) | 603 (95.1) | 571 (95.4) | |
| <18 years old at migration | 249 (1.8) | 155 (1.6) | 49 (2.3) | 45 (2.0) | |
| 18+ years old at migration | 460 (3.9) | 323 (4.8) | 90 (2.6) | 47 (2.6) | |
| Sex, n (%) | <0.001 | ||||
| Female | 2220 (53.8) | 657 (37.9) | 336 (50.1) | 376 (65.4) | |
| Male | 1369 (46.2) | 1527 (62.1) | 406 (49.9) | 287 (34.6) | |
| Ethnicity, n (%) | 0.760 | ||||
| African American | 2637 (94.3) | 1602 (94.1) | 538 (94.2) | 497 (94.9) | |
| Afro-Caribbeanc | 952 (5.7) | 582 (5.9) | 204 (5.8) | 166 (5.1) | |
| Education, n (%) | <0.001 | ||||
| Less than high school | 783 (22.6) | 535 (25.6) | 136 (18.4) | 112 (18.2) | |
| High school graduate | 1301 (37.5) | 825 (40.0) | 277 (36.6) | 199 (31.5) | |
| Some college | 891 (24.8) | 512 (22.5) | 187 (26.9) | 192 (29.3) | |
| College or more | 614 (15.1) | 312 (11.9) | 142 (18.1) | 160 (21.0) | |
| Household income, n (%) | <0.001 | ||||
| <$30,000 | 1885 (49.3) | 1233 (53.0) | 360 (46.0) | 292 (42.5) | |
| $30,000-$49,999 | 867 (24.4) | 515 (24.0) | 190 (26.5) | 162 (23.7) | |
| $50,000-$74,999 | 488 (14.7) | 254 (12.6) | 118 (16.7) | 116 (18.6) | |
| $75,000+ | 349 (11.5) | 182 (10.5) | 74 (10.8) | 93 (15.2) | |
| Lifetime substance use, n (%)d | <0.001 | ||||
| Never | 698 (16.7) | 506 (20.4) | 118 (14.1) | 74 (9.1) | |
| Single | 971 (25.0) | 633 (27.9) | 188 (23.1) | 150 (19.1) | |
| Dual | 966 (27.7) | 579 (28.5) | 221 (30.1) | 166 (22.8) | |
| Poly | 954 (30.6) | 466 (23.3) | 215 (32.7) | 273 (49.0) | |
| Current substance use, n (%)d | <0.001 | ||||
| Never/ever | 1326 (34.9) | 880 (38.8) | 260 (33.5) | 186 (25.3) | |
| Single | 1412 (38.9) | 863 (40.1) | 285 (36.5) | 264 (38.1) | |
| Dual | 640 (19.3) | 346 (16.4) | 149 (22.4) | 145 (24.2) | |
| Poly | 211 (6.9) | 95 (4.7) | 48 (7.6) | 68 (12.4) | |
Number of different experiences of discrimination attributed to race
Chi-square (categorical) or ANOVA tests (continuous) p-values comparing descriptive characteristics by lifetime patterns of substance use
Includes US-born and foreign-born participants
Substance use includes cigarettes, alcohol, marijuana, cocaine, abuse of prescription drugs, and use of other drugs; other drugs include heroin, opium, glue, LSD, peyote, or anything else
Perceived Racial Discrimination and Patterns of Substance Use
Table 2 shows results from the adjusted models examining the relationship between PRD and lifetime use of individual substances (separate logistic regression models) and patterns of substance use (multinomial regression model). Two or more PRD experiences was associated with higher odds of lifetime use of each substance. These associations were strongest for illicit drugs such as use of cocaine (Odds Ratio (OR): 2.35; 95% CI: 1.71, 3.21) and other drugs (OR: 3.10; 95% CI: 1.92, 5.01). The weakest association was between PRD and alcohol use (OR: 1.38; 95% CI: 1.11, 1.72). Although some associations between PRD and substance use were not statistically significant, findings were consistent with a dose-response relationship looking across single-, dual-, and polysubstance use. Compared to participants who reported experiencing no PRD, participants who attributed at least two experiences of PRD had nearly four times greater odds of being lifetime polysubstance use.
Table 2.
Associations between the Racial Discrimination Index and Lifetime Patterns of Substance Use (N=3,589)
| Perceived Racial Discrimination Indexa | ||||||
|---|---|---|---|---|---|---|
| One of the nine experiences | At least two of the nine experiences |
|||||
| Substance Use | n | ORb | 95% CI | n | ORb | 95% CI |
| Cigarette | ||||||
| Never | 438 | REF | 317 | REF | ||
| Ever | 304 | 1.48 | (1.22, 1.79) | 346 | 2.22 | (1.63, 3.02) |
| Alcohol | ||||||
| Never | 205 | REF | 159 | REF | ||
| Ever | 537 | 1.28 | (0.95, 1.73) | 504 | 1.38 | (1.11, 1.72) |
| Marijuana | ||||||
| Never | 355 | REF | 279 | REF | ||
| Ever | 387 | 1.59 | (1.32, 1.92) | 384 | 2.22 | (1.66, 2.97) |
| Cocaine | ||||||
| Never | 657 | REF | 540 | REF | ||
| Ever | 85 | 1.47 | (1.07, 2.01) | 123 | 2.35 | (1.71, 3.21) |
| Prescription Drugs | ||||||
| Never | 691 | REF | 606 | REF | ||
| Ever | 51 | 1.39 | (0.84, 2.30) | 57 | 1.76 | (1.17, 2.64) |
| Other Drugsc | ||||||
| Never | 707 | REF | 607 | REF | ||
| Ever | 35 | 1.84 | (1.14, 2.97) | 56 | 3.10 | (1.92, 5.01) |
| Patterns of Substance Use | ||||||
| Never | 118 | REF | 74 | REF | ||
| Single | 188 | 1.17 | (0.80, 1.73) | 150 | 1.37 | (0.84, 2.24) |
| Dual | 221 | 1.42 | (0.95, 2.10) | 166 | 1.37 | (0.86, 2.20) |
| Poly | 215 | 1.97 | (1.43, 2.71) | 273 | 3.97 | (2.42, 6.51) |
Exposure referent group: None of the nine experiences
Number of different experiences of discrimination attributed to race
Estimates are adjusted for age, sex, ethnicity, education, and income
Other drugs include heroin, opium, glue, LSD, peyote, or anything else
PRD was associated with higher odds of current marijuana use, but associations involving the other five substances were null (see Table 3). For marijuana, two or more PRD experiences was associated with nearly 2.5 times higher odds of current marijuana use (OR: 2.46; 95% CI: 1.61, 3.75), compared to those who did not report PRD. The dose-response relationship found in associations of lifetime substance use was also found in associations of current substance use. Particularly, PRD was associated with current dual- and polysubstance use. This association was strongest for those who attributed at least two experiences of PRD, compared to those who did not report PRD, and polysubstance use (OR: 4.87; 95% CI: 3.11, 7.64).
Table 3.
Associations between the Racial Discrimination Index and Current Patterns of Substance Use (N=3,589)
| Perceived Racial Discrimination Indexa | ||||||
|---|---|---|---|---|---|---|
| One of the nine experiences | At least two of the nine experiences |
|||||
| Substance Use | n | ORb | 95% CI | n | ORb | 95% CI |
| Cigarette | ||||||
| Never/ever | 108 | REF | 127 | REF | ||
| Current | 196 | 1.19 | (0.81, 1.73) | 219 | 1.33 | (0.96, 1.84) |
| Alcohol | ||||||
| Never/ever | 131 | REF | 104 | REF | ||
| Current | 414 | 0.87 | (0.63, 1.20) | 407 | 1.13 | (0.84, 1.51) |
| Marijuana | ||||||
| Never/ever | 287 | REF | 272 | REF | ||
| Current | 100 | 1.91 | (1.28, 2.86) | 112 | 2.46 | (1.61, 3.75) |
| Cocaine | ||||||
| Never/ever | 73 | REF | 113 | REF | ||
| Current | 12 | 1.00 | (0.42, 2.39) | 10 | 0.64 | (0.20, 2.05) |
| Prescription Drugs | ||||||
| Never/ever | 41 | REF | 46 | REF | ||
| Current | 10 | 1.19 | (0.36, 3.98) | 11 | 0.87 | (0.22, 3.35) |
| Other Drugsc | ||||||
| Never/ever | 29 | REF | 47 | REF | ||
| Current | 6 | 0.73 | (0.14, 3.64) | 9 | 0.82 | (0.27, 2.49) |
| Patterns of Substance Use | ||||||
| Never | 260 | REF | 186 | REF | ||
| Single | 285 | 0.99 | (0.76, 1.28) | 264 | 1.29 | (0.96, 1.73) |
| Dual | 149 | 1.56 | (1.09, 2.24) | 145 | 2.13 | (1.53, 2.96) |
| Poly | 48 | 2.09 | (1.35, 3.24) | 68 | 4.87 | (3.11, 7.64) |
Exposure referent group: None of the nine experiences
Number of different experiences of discrimination attributed to race
Estimates are adjusted for age, sex, ethnicity, education, and income
Other drugs include heroin, opium, glue, LSD, peyote, or anything else
Sensitivity analyses excluding participants who drank less than once a month in the past year from the definition of current alcohol use showed that the distribution of patterns of substance use by PRD did not statistically significantly change. The associations between current patterns of substance use and PRD were attenuated toward the null for single-, dual-, and polysubstance use, but these shifts in magnitude of associations were minimal.
Effect Modification by Ethnicity and Sex
Tables 4 and 5 display the relationship between PRD and patterns of substance use by ethnicity and sex for lifetime and current use, respectively. Among African Americans, those who reported one PRD experience had nearly two times greater odds of polysubstance use and those who reported at least two PRD experiences had nearly four times greater odds of lifetime polysubstance use (OR: 3.92; 95% CI: 2.32, 6.61), compared to those who did not report PRD. Among Afro-Caribbeans, participants who reported at least two PRD experiences had 3.8 times greater odds of lifetime dual-substance use (95% CI: 1.48, 9.62) and, although only marginally significant, 2.2 times greater odds of lifetime polysubstance use (95% CI: 0.88, 5.29), relative to those who did not report PRD. Associations between PRD and current patterns of substance use were similar to associations involving lifetime patterns of substance use. For example, African Americans who experienced any PRD had greater odds of current dual-/polysubstance use, compared to African Americans who did not report PRD.
Table 4.
Effect Modification by Ethnicity, by Sex, and by Ethnicity and Sex of Associations between the Racial Discrimination Index and Lifetime Substance Use by Ethnicity (N=3,589)
| Perceived Racial Discrimination Indexa | ||||||
|---|---|---|---|---|---|---|
| One of the nine experiences | At least two of the nine experiences |
|||||
| Patterns of Substance Use | n | OR | 95% CI | n | OR | 95% CI |
| By Ethnicity b | ||||||
| African Americans | ||||||
| Never | 80 | REF | 49 | REF | ||
| Single | 123 | 1.13 | (0.74, 1.72) | 105 | 1.31 | (0.78, 2.21) |
| Dual | 160 | 1.38 | (0.90, 2.10) | 113 | 1.27 | (0.78, 2.08) |
| Poly | 175 | 1.95 | (1.39, 2.73) | 230 | 3.92 | (2.32, 6.61) |
| Afro-Caribbeans | ||||||
| Never | 38 | REF | 25 | REF | ||
| Single | 65 | 1.68 | (0.99, 2.85) | 45 | 2.42 e | (1.02, 5.74) |
| Dual | 61 | 1.58 | (0.80, 3.14) | 53 | 3.77 | (1.48, 9.62) |
| Poly | 40 | 1.20 | (0.56, 2.59) | 43 | 2.16 | (0.88, 5.29) |
| By Sex c | ||||||
| Females | ||||||
| Never | 92 | REF | 52 | REF | ||
| Single | 119 | 1.23 | (0.82, 1.83) | 83 | 1.37 | (0.80, 2.34) |
| Dual | 93 | 1.38 | (0.93, 2.03) | 53 | 1.13 | (0.66, 1.93) |
| Poly | 102 | 1.84 | (1.21, 2.78) | 99 | 3.65 | (2.08, 6.40) |
| Males | ||||||
| Never | 26 | REF | 22 | REF | ||
| Single | 69 | 1.12 | (0.45, 2.77) | 67 | 1.46 | (0.62, 3.42) |
| Dual | 128 | 1.44 | (0.57, 3.66) | 113 | 1.64 | (0.68, 3.94) |
| Poly | 113 | 2.08 | (0.90, 4.84) | 174 | 4.56 | (1.94, 10.69) |
| By Ethnicity and Sex d | ||||||
| African American Females | ||||||
| Never | 63 | REF | 35 | REF | ||
| Single | 83 | 1.20 | (0.78, 1.85) | 60 | 1.31 | (0.74, 2.33) |
| Dual | 80 | 1.38 | (0.92, 2.06) | 37 | 1.05 | (0.60, 1.84) |
| Poly | 90 | 1.78 | (1.15, 2.76) | 81 | 3.49 | (1.94, 6.29) |
| African American Males | ||||||
| Never | 17 | REF | 14 | REF | ||
| Single | 40 | 1.07 | (0.40, 2.84) | 45 | 1.43 | (0.58, 3.53) |
| Dual | 80 | 1.40 | (0.51, 3.86) | 76 | 1.55 | (0.61, 3.94) |
| Poly | 85 | 2.15 | (0.86, 5.34) | 149 | 4.73 | (1.92, 11.68) |
| Afro-Caribbean Females | ||||||
| Never | 29 | REF | 17 | REF | ||
| Single | 36 | 1.52 | (0.84, 2.75) | 23 | 2.47 | (0.89, 6.82) |
| Dual | 13 | 0.81 | (0.25, 2.66) | 16 | 3.41 e | (1.03, 11.35) |
| Poly | 12 | 2.42 | (0.80, 7.38) | 18 | 4.22 | (0.92, 19.28) |
| Afro-Caribbean Males | ||||||
| Never | 9 | REF | 8 | REF | ||
| Single | 29 | 1.97 | (0.60, 6.49) | 22 | 2.66 | (0.76, 9.29) |
| Dual | 48 | 1.83 | (0.45, 7.44) | 37 | 3.41 | (0.59, 19.76) |
| Poly | 28 | 0.97 | (0.30, 3.06) | 25 | 1.78 | (0.53, 5.99) |
Number of different experiences of discrimination attributed to race
Estimates are adjusted for age, sex, education, and income
Estimates are adjusted for age, ethnicity, education, and income
Estimates are adjusted for age, education, and income
Association is no longer statistically significant after correction for multiple testing.
Exposure referent group: None of the nine experiences
Models stratified to include Afro-Caribbeans are also adjusted for immigration status
Table 5.
Effect Modification by Ethnicity, by Sex, and by Ethnicity and Sex of Associations between the Racial Discrimination Index and Current Substance Use by Ethnicity (N=3,589)
| Perceived Racial Discrimination Indexa | ||||||
|---|---|---|---|---|---|---|
| One of the nine experiences | At least two of the nine experiences |
|||||
| Patterns of Substance Use | n | OR | 95% CI | n | OR | 95% CI |
| By Ethnicity b | ||||||
| African Americans | ||||||
| Never/ever | 186 | REF | 136 | REF | ||
| Single | 199 | 0.98 | (0.74, 1.30) | 185 | 1.23 | (0.90, 1.68) |
| Dual/Poly | 153 | 1.72 | (1.25, 2.36) | 176 | 2.63 | (1.92, 3.59) |
| Afro-Caribbeans | ||||||
| Never/ever | 74 | REF | 50 | REF | ||
| Single | 86 | 0.93 | (0.60, 1.45) | 79 | 2.68 | (1.33, 5.39) |
| Dual/Poly | 44 | 0.63 | (0.37, 1.07) | 37 | 2.01 | (0.68, 5.94) |
| By Sex c | ||||||
| Females | ||||||
| Never/ever | 176 | REF | 107 | REF | ||
| Single | 145 | 0.92 | (0.69, 1.23) | 110 | 1.23 | (0.85, 1.77) |
| Dual/Poly | 85 | 1.74 | (1.18, 2.56) | 70 | 2.50 | (1.59, 3.92) |
| Males | ||||||
| Never/ever | 84 | REF | 79 | REF | ||
| Single | 140 | 1.06 | (0.64, 1.75) | 154 | 1.36 | (0.88, 2.10) |
| Dual/Poly | 112 | 1.62 | (0.99, 2.65) | 143 | 2.72 | (1.72, 4.28) |
| By Ethnicity and Sex d | ||||||
| African American Females | ||||||
| Never/ever | 130 | REF | 77 | REF | ||
| Single | 110 | 0.91 | (0.67, 1.23) | 77 | 1.17 | (0.80, 1.72) |
| Dual/Poly | 76 | 1.72 | (1.15, 2.57) | 59 | 2.46 | (1.54, 3.92) |
| African American Males | ||||||
| Never/ever | 56 | REF | 59 | REF | ||
| Single | 89 | 1.09 | (0.63, 1.89) | 108 | 1.32 | (0.83, 2.11) |
| Dual/Poly | 77 | 1.78 e | (1.04, 3.03) | 117 | 2.80 | (1.72, 4.55) |
| Afro-Caribbean Females | ||||||
| Never/ever | 46 | REF | 30 | REF | ||
| Single | 35 | 1.21 | (0.71, 2.06) | 33 | 3.35 e | (1.15, 9.79) |
| Dual/Poly | 9 | 1.53 | (0.49, 4.77) | 11 | 1.58 | (0.48, 5.16) |
| Afro-Caribbean Males | ||||||
| Never/ever | 28 | REF | 20 | REF | ||
| Single | 51 | 0.61 | (0.30, 1.24) | 46 | 1.88 | (0.63, 5.59) |
| Dual/Poly | 35 | 0.40 | (0.15, 1.05) | 26 | 1.80 | (0.37, 8.78) |
Number of different experiences of discrimination attributed to race
Estimates are adjusted for age, sex, education, and income
Estimates are adjusted for age, ethnicity, education, and income
Estimates are adjusted for age, education, and income
Association is no longer statistically significant after correction for multiple testing.
Exposure referent group: None of the nine experiences
Models stratified to include Afro-Caribbeans are also adjusted for immigration status
PRD was positively associated with polysubstance use among both sexes. Among women, those who reported one PRD experience had nearly two times greater odds of lifetime polysubstance use, and those who reported two or more PRD experiences had 3.7 times greater odds (95% CI: 2.08, 6.40) of lifetime polysubstance use, compared to women who did not report PRD. The point estimate for the relationship between the highest level of PRD and lifetime polysubstance use in men was stronger (OR: 4.56; 95% CI: 1.94, 10.69), but less precise, than that for women. Although weaker in magnitude, results stratified by sex for current patterns of substance use were also similar to associations for lifetime patterns of substance use. For example, both females and males who reported at least two PRD experiences had nearly two to three times higher odds of dual-/polysubstance use than females and males who did not report PRD.
Intersection by ethnicity and sex.
African American women who reported one PRD experience had two times greater odds of polysubstance use compared to those who did not report PRD. Among African American men, those who attributed at least two experiences of discrimination to race had nearly 4.7 times greater odds of polysubstance use (95% CI: 1.92, 11.68), compared to those who did not report racial discrimination. For current patterns of substance use, associations were similar with the exception of dual-/polysubstance use in African American males, which was associated with respondents who reported one PRD experience (OR: 1.78; 95% CI: 1.04, 3.03), compared to African American males who did not report PRD.
Among Afro-Caribbean women, those who reported one PRD experience had higher odds of polysubstance use (OR: 2.42; 95% CI: 0.80, 7.38), but this association was not statistically significant. Further, although not statistically significant after false discovery rate correction, Afro-Caribbean women who reported at least two PRD experiences had 3.4 times greater odds of dual-substance use (95% CI: 1.03, 11.35), compared to Afro-Caribbean females who did not report PRD. For current use, Afro-Caribbean women who reported at least two PRD experiences had 3.4 times higher odds of current single-product use (95% CI: 1.15, 9.79), compared to Afro-Caribbean females who did not report PRD, but this association was no longer statistically significant after false discovery rate correction. For Afro-Caribbean females, associations including dual-/polysubstance use were not statistically significant. Furthermore, PRD was not associated with lifetime or current substance use outcomes among Afro-Caribbean men, although most of the point estimates were in the direction of increased risk of use.
Discussion
Using a large, nationally representative sample, we found that PRD is a common experience for Black Americans adults and that this experience is associated with patterns of substance use. Nearly four in 10 US Black adults reported at least one PRD experience in their lifetime. Lifetime patterns of substance use were also common, with nearly one-quarter of participants engaging in either dual-substance (26.9%) or polysubstance (26.6%) use. While prevalence of current dual patterns of substance use was similar to lifetime prevalence (24.2% v. 22.8%), current polysubstance use was lower than lifetime prevalence (12.4% v. 49.0%). Participants who experienced PRD had higher odds of both individual substance and lifetime polysubstance use than those who did not experience PRD. PRD was only associated with current marijuana use, but associations between PRD and current patterns of substance use suggested that PRD is related to the use of multiple substances. Overall, these findings illustrate the salience of PRD for substance use outcomes in the US Black population.
Our results are consistent with the handful of prior investigations of PRD and substance use (Assari, et al., 2018; Assari, et al., 2019; Borrell, et al., 2010; Borrell, et al., 2007; Borrell, et al., 2013; T. T. Clark, 2014; Rose, et al., 2018). For example, an analysis using the Coronary Artery Risk Development in Young Adults study operationalized PRD in a similar manner to this study and reported that PRD was related to use of tobacco, alcohol, marijuana, and cocaine (Borrell, et al., 2007). Another study reported that lifetime PRD was associated with higher odds of past-month cigarette use, harmful alcohol use, and past-month marijuana use among young adults (Rose, et al., 2018). Together, these studies indicate a need for further research to investigate how PRD contributes to patterns of substance use such as explicit distinctions among types of substances used within dual-substance and polysubstance use subgroups.
This study employed an intersectionality framework to explore how sex and ethnicity relate to PRD and patterns of substance use. To our knowledge, this is one of the first studies to evaluate the impact of PRD on patterns of substance use, including polysubstance use, with attention to variation by ethnicity and sex within the US Black population. Our analyses showed that the substantive relationship between PRD and lifetime and current substance use differed for African American and Afro-Caribbean adults; for both groups, reporting two or more PRD experiences was associated with higher odds of substance use. For lifetime use, these relationships were more pronounced for dual-substance use among Afro-Caribbeans as compared to polysubstance use among African Americans. In terms of gender, PRD was related to lifetime polysubstance use among both males and females in a dose-response manner, with point estimates that may be stronger among men than women. Associations were weaker in magnitude for current patterns of substance use but suggest that PRD is related to recent use of multiple substances among African Americans, females, and males.
Analyses stratified by ethnicity and sex illustrated that the association between PRD and polysubstance use may be stronger among Afro-Caribbean women relative to Afro-Caribbean men. Although not statistically significant, Afro-Caribbean women who reported at least two PRD experiences had over four times higher odds of being lifetime polysubstance users compared to those who did not report PRD. While PRD was associated with polysubstance use among African American men, findings among Afro-Caribbean men were not statistically significant. The lack of statistically significant findings among Afro-Caribbeans may reflect limitations of sample size and lower statistical power. Overall, our findings contrast with prior studies suggesting that discrimination is a stronger risk factor for substance use among men as compared to women (Assari, et al., 2018; Assari, et al., 2019; Brody, et al., 2012). This was corroborated by the results involving current use where Afro-Caribbean females who reported at least two PRD experiences had over three times higher odds of single-product use, while associations involving Afro-Caribbean males were null. While our results stratified by sex bolster the argument that PRD is related to substance use in both men and women similarly, results by ethnicity and sex highlight that more detailed consideration of intersectional experiences is needed to understand the relationships between PRD and substance use within the US Black population.
Our findings imply that more frequent exposure to racially attributed discrimination is associated with polysubstance use, and these associations vary by ethnicity and sex. While the mechanisms underlying this relationship are still unclear, the stress-process model, which posits that stressors prompt engagement in risky health behaviors, such as substance use, as self-regulatory coping efforts (Ong, Fuller-Rowell, & Burrow, 2009; Pearlin, Lieberman, Menaghan, & Mullan, 1981) may provide a useful framework for understanding our results. Pervasive social stressors, such as PRD, may influence engagement in risky health behaviors as a means to regulate the stress response, a relationship that may change over the life course (R. Clark, et al., 1999). While the mechanisms linking PRD to substance use are still unresolved, there is suggestive evidence that anger (Gibbons et al., 2010), reduced self-control (Gibbons, et al., 2010; Gibbons et al., 2012), and changes in externalizing behaviors, such as onset of problematic alcohol use (Gibbons et al., 2014), may play a role. Our results indicating that PRD and polysubstance use differ by ethnicity and sex supports the underlying linkage explained by intersectionality theory. Future research should continue to employ these frameworks in identifying mechanisms connecting race-salient exposures to health.
Limitations
Findings should be interpreted in light of study limitations. First, the cross-sectional nature of the NSAL precludes our ability to infer temporal relationships between the exposure and outcomes. Substance use generally first emerges in adolescence, and thus PRD may have occurred contemporaneous with, or subsequent to, the patterns of substance use. Second, our goal was to explore the role of racial attribution of discriminatory experiences; as a result, participants who attributed all discrimination to reasons other than race (20.1%) were combined with participants who reported they had never experienced discrimination for any reason (37.9%) in this analysis. Participants who reported never experiencing any form of discrimination and participants who reported experiencing discrimination but who attributed it to reasons other than race may systematically differ in respect to substance use and other covariates. This limitation warrants further research in exploring the various attribution types and their relationships with substance use, especially within ethnically diverse samples. Third, we operationalized PRD as no PRD experiences, one out of nine PRD experiences, and at least two out of nine PRD experiences rather than treating PRD as a count variable. Our motivation was to emphasize the extent to which more PRD was associated with polysubstance use while maintaining adequate sample sizes in ethnicity-by-sex strata. However, we acknowledge that categorizing PRD has the potential to lose important information (Cohen, 1983). For example, participants who experience racial/ethnic discrimination in all nine scenarios might differ in their substance use behaviors from participants who experience it in two scenarios. Finally, NSAL was conducted between 2001 and 2003, which pre-dated recreational marijuana use legalization in several states (Cerdá et al., 2020; Yu, Chen, Chen, Yan, 2020). Thus, in our study, marijuana use was characterized as an illicit substance. Today, legislation and associated de-stigmatization of marijuana use has potentially led to an increase in use among youth and adults (Cerdá et al., 2020; Yu, Chen, Chen, Yan, 2020). Additionally, misuse of prescription drugs is more prevalent today than it was when these data were collected (Scholl, Seth, Kariisa, Wilson, & Baldwin, 2018). Despite these limitations, the NSAL is the only nationally representative sample of Black Americans that contains information on both discrimination and substance use, and thus it provides the richest source of data for examining variation within the Black population.
Further research is necessary in order to identify mechanisms linking PRD and substance use, including exploration of the similarities and differences of these processes across social and cultural characteristics. Public health professionals should consider the impact of PRD on patterns of substance use, including polysubstance use. Interventions to address polysubstance use in minority populations may benefit from acknowledging exposures that are disproportionately experienced by these minorities, including PRD.
Supplementary Material
Acknowledgements:
This study used data from the Inter-university Consortium for Political and Social Research’s Collaborative Psychiatric Epidemiology Surveys (CPES). The authors would like to thank the researchers who compiled and harmonized the CPES data for public use.
Funding details:
Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health (NIH) and FDA Center for Tobacco Products (CTP) under Award Number U54CA229974. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration. The NSAL is sponsored by the NIMH under grant number U01-MH57716, with additional support from the Office of Behavioral and Social Sciences Research (OBSSR) at the NIH and the University of Michigan. The funding sponsors had no role in study design, data analysis or interpretation, manuscript preparation, or the decision to publish the results.
Footnotes
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement:
The data that support the findings of this study are openly available in the Inter-university Consortium for Political and Social Research at https://doi.org/10.3886/ICPSR20240.v8.
References
- Alegria M, Jackson JS, Kessler RC, & Takeuchi D (2016). Collaborative Psychiatric Epidemiology Surveys (CPES), 2001–2003 [United States] [Data set]. doi: 10.3886/ICPSR20240.v8. [DOI] [Google Scholar]
- Assari S (2018). Perceived Discrimination and Binge Eating Disorder; Gender Difference in African Americans. J Clin Med, 7(5). doi: 10.3390/jcm7050089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Assari S, Mistry R, & Caldwell CH (2018). Perceived Discrimination and Substance Use among Caribbean Black Youth; Gender Differences. Brain Sci, 8(7). doi: 10.3390/brainsci8070131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Assari S, Mistry R, Lee DB, Caldwell CH, & Zimmerman MA (2019). Perceived Racial Discrimination and Marijuana Use a Decade Later; Gender Differences Among Black Youth. Front Pediatr, 7, p. 78. doi: 10.3389/fped.2019.00078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benjamini Y, & Hochberg Y (1995). Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B (Methodological), 57(1), pp. 289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x. [DOI] [Google Scholar]
- Borrell LN, Diez Roux AV, Jacobs DR Jr., Shea S, Jackson SA, Shrager S, & Blumenthal RS (2010). Perceived racial/ethnic discrimination, smoking and alcohol consumption in the Multi-Ethnic Study of Atherosclerosis (MESA). Prev Med, 51(3–4), pp. 307–312. doi: 10.1016/j.ypmed.2010.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Borrell LN, Jacobs DR Jr., Williams DR, Pletcher MJ, Houston TK, & Kiefe CI (2007). Self-reported racial discrimination and substance use in the Coronary Artery Risk Development in Adults Study. Am J Epidemiol, 166(9), pp. 1068–1079. doi: 10.1093/aje/kwm180. [DOI] [PubMed] [Google Scholar]
- Borrell LN, Kiefe CI, Diez-Roux AV, Williams DR, & Gordon-Larsen P (2013). Racial discrimination, racial/ethnic segregation, and health behaviors in the CARDIA study. Ethn Health, 18(3), pp. 227–243. doi: 10.1080/13557858.2012.713092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bowleg L (2012). The problem with the phrase women and minorities: intersectionality-an important theoretical framework for public health. Am J Public Health, 102(7), pp. 1267–1273. doi: 10.2105/AJPH.2012.300750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brodish AB, Cogburn CD, Fuller-Rowell TE, Peck S, Malanchuk O, & Eccles JS (2011). Perceived Racial Discrimination as a Predictor of Health Behaviors: the Moderating Role of Gender. Race and Social Problems, 3(3), p. 160. doi: 10.1007/s12552-011-9050-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brody GH, Kogan SM, & Chen Y. f. (2012). Perceived Discrimination and Longitudinal Increases in Adolescent Substance Use: Gender Differences and Mediational Pathways. American Journal of Public Health, 102(5), pp. 1006–1011. doi: 10.2105/ajph.2011.300588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Broman CL, Neighbors HW, Delva J, Torres M, & Jackson JS (2008). Prevalence of substance use disorders among African Americans and Caribbean Blacks in the National Survey of American Life. Am J Public Health, 98(6), pp. 1107–1114. doi: 10.2105/AJPH.2006.100727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cerdá M, Mauro C, Hamilton A, Levy NS, Santaella-Tenorio J, Hasin D, … Martins SS (2020). Association Between Recreational Marijuana Legalization in the United States and Changes in Marijuana Use and Cannabis Use Disorder From 2008 to 2016. JAMA Psychiatry, 77(2), p. 165. doi: 10.1001/jamapsychiatry.2019.3254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chung T, Kim KH, Hipwell AE, & Stepp SD (2013). White and black adolescent females differ in profiles and longitudinal patterns of alcohol, cigarette, and marijuana use. Psychol Addict Behav, 27(4), pp. 1110–1121. doi: 10.1037/a0031173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Clark R, Anderson NB, Clark VR, & Williams DR (1999). Racism as a stressor for African Americans. A biopsychosocial model. Am Psychol, 54(10), pp. 805–816. doi: 10.1037//0003-066x.54.10.805. [DOI] [PubMed] [Google Scholar]
- Clark TT (2014). Perceived discrimination, depressive symptoms, and substance use in young adulthood. Addict Behav, 39(6), pp. 1021–1025. doi: 10.1016/j.addbeh.2014.01.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Clark TT, Salas-Wright CP, Vaughn MG, & Whitfield KE (2015). Everyday discrimination and mood and substance use disorders: a latent profile analysis with African Americans and Caribbean Blacks. Addict Behav, 40, pp. 119–125. doi: 10.1016/j.addbeh.2014.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cohen J (1983). The Cost of Dichotomization. Applied Psychological Measurement, 7(3), pp. 249–253. doi: 10.1177/014662168300700301 [DOI] [Google Scholar]
- Cole ER (2009). Intersectionality and research in psychology. American Psychologist, 64(3), pp. 170–180. doi: 10.1037/a0014564. [DOI] [PubMed] [Google Scholar]
- Connor JP, Gullo MJ, White A, & Kelly AB (2014). Polysubstance use: diagnostic challenges, patterns of use and health. Curr Opin Psychiatry, 27(4), pp. 269–275. doi: 10.1097/YCO.0000000000000069. [DOI] [PubMed] [Google Scholar]
- Evans EA, Grella CE, Washington DL, & Upchurch DM (2017). Gender and race/ethnic differences in the persistence of alcohol, drug, and poly-substance use disorders. Drug Alcohol Depend, 174, pp. 128–136. doi: 10.1016/j.drugalcdep.2017.01.021. [DOI] [PubMed] [Google Scholar]
- Georgiades K, Boyle MH, & Duku E (2007). Contextual influences on children’s mental health and school performance: the moderating effects of family immigrant status. Child Dev, 78(5), pp. 1572–1591. doi: 10.1111/j.1467-8624.2007.01084.x. [DOI] [PubMed] [Google Scholar]
- Gerrard M, Stock ML, Roberts ME, Gibbons FX, O’Hara RE, Weng CY, & Wills TA (2012). Coping with racial discrimination: the role of substance use. Psychol Addict Behav, 26(3), pp. 550–560. doi: 10.1037/a0027711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibbons FX, Etcheverry PE, Stock ML, Gerrard M, Weng CY, Kiviniemi M, & O’Hara RE (2010). Exploring the link between racial discrimination and substance use: what mediates? What buffers? J Pers Soc Psychol, 99(5), pp. 785–801. doi: 10.1037/a0019880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibbons FX, Kingsbury JH, Weng CY, Gerrard M, Cutrona C, Wills TA, & Stock M (2014). Effects of perceived racial discrimination on health status and health behavior: a differential mediation hypothesis. Health Psychol, 33(1), pp. 11–19. doi: 10.1037/a0033857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibbons FX, O’Hara RE, Stock ML, Gerrard M, Weng CY, & Wills TA (2012). The erosive effects of racism: reduced self-control mediates the relation between perceived racial discrimination and substance use in African American adolescents. J Pers Soc Psychol, 102(5), pp. 1089–1104. doi: 10.1037/a0027404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heeringa SG, Wagner J, Torres M, Duan N, Adams T, & Berglund P (2004). Sample designs and sampling methods for the Collaborative Psychiatric Epidemiology Studies (CPES). Int J Methods Psychiatr Res, 13(4), pp. 221–240. doi: 10.2105/AJPH.2012.300780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hunte HE, & Barry AE (2012). Perceived discrimination and DSM-IV-based alcohol and illicit drug use disorders. Am J Public Health, 102(12), pp. e111–117. doi: 10.2105/AJPH.2012.300780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jackson JS, Neighbors HW, Nesse RM, Trierweiler SJ, & Torres M (2004). Methodological innovations in the National Survey of American Life. Int J Methods Psychiatr Res, 13(4), pp. 289–298. doi: 10.1002/mpr.182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jackson JS, Torres M, Caldwell CH, Neighbors HW, Nesse RM, Taylor RJ, … Williams DR (2004). The National Survey of American Life: a study of racial, ethnic and cultural influences on mental disorders and mental health. Int J Methods Psychiatr Res, 13(4), pp. 196–207. doi: 10.1002/mpr.177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kessler RC, Mickelson KD, & Williams DR (1999). The prevalence, distribution, and mental health correlates of perceived discrimination in the United States. J Health Soc Behav, 40(3), pp. 208–230. doi: 10.2307/2676349. [DOI] [PubMed] [Google Scholar]
- Moss HB, Chen CM, & Yi HY (2014). Early adolescent patterns of alcohol, cigarettes, and marijuana polysubstance use and young adult substance use outcomes in a nationally representative sample. Drug Alcohol Depend, 136, pp. 51–62. doi: 10.1016/j.drugalcdep.2013.12.011. [DOI] [PubMed] [Google Scholar]
- Mulia N, Ye Y, Greenfield TK, & Zemore SE (2009). Disparities in alcohol-related problems among white, black, and Hispanic Americans. Alcohol Clin Exp Res, 33(4), pp. 654–662. doi: 10.1111/j.1530-0277.2008.00880.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ong AD, Fuller-Rowell T, & Burrow AL (2009). Racial discrimination and the stress process. J Pers Soc Psychol, 96(6), pp. 1259–1271. doi: 10.1037/a0015335. [DOI] [PubMed] [Google Scholar]
- Pearlin LI, Lieberman MA, Menaghan EG, & Mullan JT (1981). The stress process. J Health Soc Behav, 22(4), pp. 337–356. doi: 10.2307/2136676. [DOI] [PubMed] [Google Scholar]
- Rose SW, Mayo A, Ganz O, Perreras L, D’Silva J, & Cohn A (2018). Perceived racial/ethnic discrimination, marketing, and substance use among young adults. J Ethn Subst Abuse, pp. 1–20. doi: 10.1080/15332640.2018.1425949. [DOI] [PubMed] [Google Scholar]
- Salom CL, Betts KS, Williams GM, Najman JM, & Alati R (2016). Predictors of comorbid polysubstance use and mental health disorders in young adults-a latent class analysis. Addiction, 111(1), pp. 156–164. doi: 10.1111/add.13058. [DOI] [PubMed] [Google Scholar]
- Schneider KE, Park JN, Allen ST, Weir BW, & Sherman SG (2019). Patterns of polysubstance use and overdose among people who inject drugs in Baltimore, Maryland: A latent class analysis. Drug Alcohol Depend, 201, pp. 71–77. doi: 10.1016/j.drugalcdep.2019.03.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Scholl L, Seth P, Kariisa M, Wilson N, & Baldwin G (2018). Drug and Opioid-Involved Overdose Deaths - United States, 2013–2017. MMWR Morb Mortal Wkly Rep, 67(5152), pp. 1419–1427. doi: 10.15585/mmwr.mm675152e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shields SA (2008). Gender: An Intersectionality Perspective. Sex Roles, 59(5), pp. 301–311. doi: 10.1007/s11199-008-9501-8. [DOI] [Google Scholar]
- Silveira ML, Green VR, Iannaccone R, Kimmel HL, & Conway KP (2019). Patterns and correlates of polysubstance use among US youth aged 15–17 years: wave 1 of the Population Assessment of Tobacco and Health (PATH) Study. Addiction. doi: 10.1111/add.14547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith GW, Farrell M, Bunting BP, Houston JE, & Shevlin M (2011). Patterns of polydrug use in Great Britain: findings from a national household population survey. Drug Alcohol Depend, 113(2–3), pp. 222–228. doi: 10.1016/j.drugalcdep.2010.08.010. [DOI] [PubMed] [Google Scholar]
- Soto JA, Dawson-Andoh NA, & BeLue R (2011). The relationship between perceived discrimination and Generalized Anxiety Disorder among African Americans, Afro Caribbeans, and non-Hispanic Whites. J Anxiety Disord, 25(2), pp. 258–265. doi: 10.1016/j.janxdis.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tomczyk S, Isensee B, & Hanewinkel R (2016). Latent classes of polysubstance use among adolescents-a systematic review. Drug Alcohol Depend, 160, pp. 12–29. doi: 10.1016/j.drugalcdep.2015.11.035. [DOI] [PubMed] [Google Scholar]
- Unger JB, Ritt-Olson A, Teran L, Huang T, Hoffman BR, & Palmer P (2002). Cultural values and substance use in a multiethnic sample of California adolescents. Addiction Research & Theory, 10(3), pp. 257–279. doi: 10.1080/16066350211869. [DOI] [Google Scholar]
- Williams DR, Gonzalez HM, Williams S, Mohammed SA, Moomal H, & Stein DJ (2008). Perceived discrimination, race and health in South Africa. Soc Sci Med, 67(3), pp. 441–452. doi: 10.1016/j.socscimed.2008.03.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu LT, Zhu H, & Swartz MS (2016). Trends in cannabis use disorders among racial/ethnic population groups in the United States. Drug Alcohol Depend, 165, pp. 181–190. doi: 10.1016/j.drugalcdep.2016.06.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yu B, Chen X, Chen X, & Yan H (2020). Marijuana legalization and historical trends in marijuana use among US residents aged 12–25: Results from the 1979–2016 National Survey on drug use and health. BMC Public Health, 20(1). doi: 10.1186/s12889-020-8253-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are openly available in the Inter-university Consortium for Political and Social Research at https://doi.org/10.3886/ICPSR20240.v8.