Abstract
Adolescents residing on the US–Mexico border are at disproportionate risk for substance misuse due to environmental risk factors, including high unemployment, high concentration of liquor licenses, drug-trade–related violence, border militarization, and inadequate access to prevention, treatment, and support services for substance misuse. We used multivariable logistic regression to examine the association between high perceived normalization of drug trafficking and cross-border purchase, and past 30-day substance use (tobacco, alcohol, and marijuana) in a sample of 445 primarily Mexican American adolescents living on the US–Mexico border. Adolescents with higher perceived normalization of drug trafficking scores were more likely to have crossed the border to purchase (OR = 1.17, 95% CI 1.07, 1.28) and use substances (OR = 1.08, 95% CI 1.00, 1.17). The economic and normative environment of border communities must be considered in the formation of interventions targeted at substance use risk prevention and mitigation, and involve binational support from organizations, policy makers, and community members.
Introduction
Research indicates that environmental factors, including access and exposure to substances of abuse, neighborhood disadvantage and disorder, and environmental barriers to treatment, influence adolescent substance use behaviors (Mennis, Stahler, & Mason, 2016). When considering an environmental risk approach (Rhodes et al., 2003), adolescents residing on the US–Mexico border are at disproportionate risk for substance misuse, including high unemployment, poverty, crumbling infrastructure, high concentration of liquor licenses, drug-trade–related violence, and inadequate access to prevention, treatment, and support services for substance misuse (Lee et al., 2013). Additional border stressors include border militarization, perceived racial and ethnic discrimination, and acculturative stress. All have been shown to pose health risks to Hispanic border residents (Borges et al., 2015; Caetano et al., 2008, 2012, 2013; Lange et al., 2002; Sabo et al., 2014; Wallisch & Spence, 2006).
There is a lack of epidemiological data on youth alcohol and other substance use on the US–Mexico border. Older research indicates that Mexican American adolescents living on the US–Mexico border experience higher rates of substance use–related problems compared to non-border Mexican Americans and their non-Hispanic white counterparts (Almodovar, Tomaka, Thompson, Mckinnon, & O’Rourke, 2006; McKinnon, O’Rourke, Thompson, & Berumen, 2004). In a recent study of 445 Arizona border youth, conducted by the present study’s lead author using the same sample, past 30-day alcohol use was higher (33.9%) than average use rates nationally (29.8%) (Kann et al., 2018) and at the state level (20.2%) (Arizona Criminal Justice Commission, 2018). Environmental factors likely contribute to these disparities in adolescent substance use on the US–Mexico border (Mack, 2012; Marshall, 2014; Newcomb & Locke, 2005; Valdez et al., 2021).
Cross-Border Access to Substances
The US–Mexico border stretches across 44 US counties in California, Arizona, New Mexico, and Texas. The majority of individuals living in these counties reside within 100 km of the border (Lee et al., 2013). Community members on both sides of the border regularly cross daily for work, school, or commerce. Poverty mixed with border mobility also facilitate the presence and normalization of drug trafficking and cross-border purchase, which likely contributes to higher rates of substances among US–Mexico border adolescents (Valdez et al., 2019). In Mexico, many border cities have adopted liberal substance use norms and policing in order to attract foreign customers (Lange, Voas, Johnson, 2002; Valdez & Sifaneck, 1997). With lower drinking ages, cheaper alcohol and tobacco, and lenient law enforcement, youth access to substances, including marijuana and illicit drugs, is not uncommon (Lange et al., 2002; Valdez et al., 2019, 2021; Valdez & Sifaneck, 1997). While drinking ages are not lower in Mexican border cities compared to the rest of the country, they are lower compared to the United States; substance use in the US border cities is higher than among Mexican border cities (Lange, Voas, Johnson, 2002; Valdez & Sifaneck, 1997).
Perceptions of Normalization of Drug Trafficking
Border scholar Howard Campbell (2005) explained that drug trafficking in border towns is a tacitly tolerated activity and a daily phenomenon. While not fully accepted, it is not considered a radically deviant or unusual lifestyle by border residents. Border towns, particularly in rural areas, often experience high rates of unemployment and poverty (Lee et al., 2013) which likely serves as a motivator to participate in trafficking, especially when friends or family are involved (Valdez et al., 2019). Perceived normalization of drug trafficking in the community has its roots in generational dependence on the drug trade for economic reasons, making drug trafficking a predictable adaptation of living in poverty (Campbell, 2004). The desensitization of drug trafficking and the resulting cultural differences, social networks, and the pervasiveness of the border drug trade may exacerbate the risk of adolescent substance use (Campbell, 2005; Fitzgerald and Threadgold, 2004; Molano, 2004).
A previous study by the lead author engaged adolescents in a youth participatory action research project (YPAR) using Photovoice to explore border-bound risk factors for adolescent substance use (Valdez et al., 2019, 2021). These youth researchers defined perceived normalization of drug trafficking as the keen awareness of the common occurrence of drug movement activity within one’s own community including the intergenerational nature of the drug trade and the active presence of drug cartels in the community. Similar to the work by Campbell (2004), youth researchers believed that the normalization of drug trafficking in the community is due to a generational dependence on the drug trade for economic reasons (Valdez et al., 2019). For example, if grandfathers, uncles, cousins, and other family members are involved in the drug trade, youth researchers expressed that this normalizes the drug trade for youth, and that the lack of sufficient jobs coupled with the high-risk/high pay nature of the drug trade can be motivators to becoming involved as well. Notably, youth researchers perceived that the historic and pervasive presence of drug trafficking in the community likely increases access to drugs and may cause desensitization to the drug trade and drug use (Valdez et al., 2019). It is well documented that neighborhood-level drug-related indicators (e.g., neighborhood drug selling or drug use) influence behavioral health and risk taking (e.g., substance misuse) among adolescents (Aneshensel & Sucoff, 1996; Furr-Holden et al., 2008; Milam et al., 2012). Nevertheless, this study is one of very few studies to explore how drug trafficking may influence adolescent substance use in the US–Mexico border region, particularly the ability to cross the border to purchase and use substances. The examination of these factors will provide a more complete understanding of border regions and the experiences of border adolescents’ substance use.
Current Study
Building on this previous work, the present study quantitatively examined associations between perceived normalization of drug trafficking, crossing the border to purchase substances (tobacco, alcohol, and marijuana) in the past 30 days, and past 30-day substance use (tobacco, alcohol, and marijuana) among adolescents living on the US–Mexico border. We hypothesized that participants with high perceived normalization of drug trafficking would be more likely to report cross-border purchase and past 30-day use of substances while controlling for socio-demographic factors like age, gender, socioeconomic status, and family structure.
Methods
Participants and Recruitment
Participants were recruited from a border-region high school and include males and females aged 14–18 years. To ensure balance by age and grade level, participants were recruited from freshman pre-algebra, sophomore geometry, junior algebra I, and senior algebra II classes. Participants must have lived in the city in which the school was located or the sister city on the Mexico side for the last 12 months and be able to write and read in English and/or Spanish. Participants provided informed written assent and obtained informed written consent from a legally authorized representative prior to participating in this study. One week prior to the survey, participants were informed in writing and verbally in class, that because the study may involve participants who are not US citizens or who may have experience with illicit activities, the lead author obtained a Certificate of Confidentiality from the National Institutes of Health, and that the researchers could use this certificate to legally refuse to disclose information that may identify participants in any legal proceedings. Further, due to the sensitive nature of some of the questions, participants were assured that all their information would be de-identified, locked in a safe place, and that only aggregate data would be shared with school officials and the community. Participants included their student identification numbers on their consent forms, which were later linked to their survey. The lead author also informed the student that she was not affiliated with the school. Individuals with missing data (>50% of the survey incomplete) were removed from the analysis.
Study Procedure
The Border Adolescent Substance Use Survey (BASUS) was delivered during math class at the school computer lab. The lead author, [omitted for blind review], administered the Border Adolescent Substance Use Survey (BASUS) to each group (25–30 participants) in two adjacent computer labs for 50-minute intervals during six periods per day for two consecutive days. The participants were only in the lab to complete the survey. Those without consent/assent documentation used the computers while other participants took the survey. Prior to beginning, [lead author name omitted for blind review] described the purpose of the study, the sensitive nature of the questions, the confidential nature of data collection, and invited participants to ask questions. Participants had the option to complete the BASUS in Spanish or English. The participants received a tiny URL to access the BASUS. The BASUS was administered online using the REDCap (Research Electronic Data Capture) platform. Participants took between 20 and 40 minutes to complete the BASUS. As an incentive to participate, BASUS participants received 10 student US$ (to be used in the high school’s student store) for submission of the parental consent and participant assent, and five student US$ for completion of the BASUS. Student identification numbers were used to link participants’ consent and survey documents; this confirmed study completion and allowed incentive distribution. All identifying information was later replaced with a unique code and destroyed. This study received human subject’s approval from the University of Arizona Institutional Review Board.
Measures
All measures were self-reported and taken from the BASUS, a composite survey developed from a previous study (Valdez et al., 2021). The BASUS consists of existing validated instruments and newly created measures to examine the border-bound factors that influence substance use among adolescents (Valdez et al., 2019, 2021). The outcomes of interest were 1) use of substances in the past 30 days and 2) cross-border purchase of substances.
Use of Substances
Use of substances (yes/no) was defined as use of alcohol, marijuana, or tobacco on at least 1 day, within the last 30 days of data collection (past 30-day use).
Cross-Border Purchase of Substances
We defined cross-border purchase of substances (yes/no) as participants crossing the US–Mexican border to purchase substances (alcohol, marijuana, or tobacco) on at least 1 day within the last 30 days of data collection.
Perceived Normalization of Drug Trafficking
We measured the perceived normalization of drug trafficking using a scale that was developed in partnership with border youth from a previously conducted Photovoice study (Valdez et al., 2019), and included in the BASUS. The scale consists of six items that address factors in the community related to drug trafficking (I have been encouraged to move, transport, or carry drugs; I have been encouraged to sell drugs; I have been encouraged to store drugs; In my community, it is normal for high school kids to sell, store of transport drugs; In my community, music that glorifies drug trafficking or the narco lifestyle, influences adolescents to get involved in using, moving, selling or storing drugs; In my community, music that glorifies drug trafficking or the narco lifestyle, influences adolescents to start using substances). Participants answered on a 4-point Likert scale with one being strongly disagree and four being strongly agree. Scores for the perceived normalization of drug trafficking were summed, ranging from 4 to 24, with higher scores representing higher perceived normalization of drug trafficking. We determined the scale’s internal consistency using a Cronbach’s alpha correlation (Cronbach’s α = 0.87).
Covariates
Covariates for this analysis were selected a priori based on our background knowledge. We adjusted for age, gender, ethnicity, social support, country of residence, and socioeconomic status (measured by the level of education of participants’ mother) in our logistic regression models (Barrett & Turner, 2006; Lien, Friestad, Klepp, 2001). Age was self-reported by the participants and measured in years. Participants self-identified as male or female and reported their mother’s education on a 5-point scale with categories ranging from less than a high school education to college education. We considered participants as Hispanic if they self-identified as Mexican American, Mexican, Chicano/a, Tohono O’odham, or Yaqui and classified participants as non-Hispanic if they reported another ethnic group. Participants were considered US residents if they reported living in a community on the US side of the border. Those who reported living in the border sister city were classified as residents of Mexico. Finally, social support was assessed using a modified version from a section of the Border Community and Immigration Stress Scale, which was tested in two border communities with predominantly Hispanic-descent samples (Carvajal et al., 2013). The scale assessed participants’ ability to ask for and receive help from friends and family (I can count on my friends when things go wrong; I can talk about my problems with my family; My family is willing to help me make decisions; I can talk about my problems with my friends). Participants answered on a 5-point Likert scale (1 = strongly agree and 5 = strongly disagree). For consistency, we reverse-coded the scale and created a summed support score ranging from 4–20, with higher scores representing higher social support (Cronbach’s α =0.68).
Statistical Analysis
Descriptive statistics of participants in the BASUS study, stratified by perceived normalization of drug trafficking (n = 445).
Abbreviation: SD, standard deviation.
aParticipants with sum of perceived normalization of drug trafficking score less than or equal median of 10.
bParticipants with sum of perceived normalization of drug trafficking score greater than median of 10.
cDefined as persons who used any substances on at least 1 day within the last 30 days of data collection.
dDefined as persons who did not use any substances on at least 1 day within the last 30 days of data collection.
eDefined as persons who crossed the US–Mexican border to purchase substances on at least 1 day within the last 30 days of data collection.
fDefined as persons who did not cross the US–Mexican border to purchase substances on at least 1 day within the last 30 days of data collection.
Primary Analysis
Associations between perceived normalization of drug trafficking and past 30-day substance use among adolescents living in a US–Mexico border city (N = 445).
Abbreviation: OR, odds ratio; SES, socioeconomic status; CI, confidence interval.
aAdjusted for age, gender, socioeconomic status, ethnicity, social support, and cross-border substance purchase.
bOutcome defined as use of substances 0–5 days within the last 30 days before data collection.
cOutcome refined as use of substances 0–5 days within the last 30 days before data collection and adjusted for age, gender, socioeconomic status, ethnicity, social support, and cross-border substance purchase.
dPerceived normalization of drug trafficking used for this analysis was measured on a continuous scale.
eDefined as persons who crossed the US–Mexican border to purchase substances on at least 1 day within the last 30 days of data collection.
fDefined as persons who did not cross the US–Mexican border to purchase substances on at least 1 day within the last 30 days of data collection.
*p-value = .05; **p-value < .05.
Associations between perceived normalization of drug trafficking and past 30-day cross-border substance purchase among adolescents living in a US–Mexico border city (N = 445).
Abbreviations: OR, odds ratio; SES, socioeconomic status; SD, standard deviation; CI, confidence interval.
aAdjusted for age, gender, socioeconomic status, ethnicity, country of residence, social support, and cross-border substance use.
bPerceived normalization of drug trafficking used for this analysis was measured on a continuous scale.
cDefined as persons who used substances on at least 1 day within the last 30 days of data collection.
dDefined as persons who did not use substances on at least 1 day within the last 30 days of data collection.
**p-value < .05.
We assessed the linearity of the log-odds graphically using a locally weighted scatterplot smoothing (lowess). We also evaluated the fit of our models using the Hosmer–Lemeshow test.
Sensitivity Analysis
We conducted a sensitivity analysis to assess the robustness of our primary analysis. For this, we classified substance users as persons who reported using substances more than 5 days in the last 30 days and non-substance users as persons who reported using substances 0–5 days within the last 30 days. We re-ran the models using these new outcome definitions to determine the robustness of our primary analysis.
All analyses were conducted with SAS University Edition (Cary, NC). All statistical tests were based on a significance level of 0.05.
Results
The baseline characteristics are described in Table 1 which is stratified by perceived normalization of drug trafficking, dichotomized using a median split. Participants with low perceived normalization of drug trafficking accounted for 54.2% (n = 241) of all participants. The mean age was 16 years (SD = 1.3) for both groups. Both groups had more females, accounting for 53.1% and 62.3% of persons with low and high perceived normalization of drug trafficking, respectively. Among persons with low perceived normalization of drug trafficking, while 34.6% reported to have used substances, 24.5% purchased substances across the US–Mexico border. For persons with high perceived normalization of drug trafficking, 43.6% were substance users and 35.8% purchased substances across the border. The mean social support score was 7.2 for both groups.
The results of logistic regression analyses for the relationship between substance use and perceived normalization of drug trafficking are shown in Table 2. In the unadjusted analysis, perceived normalization of drug trafficking, as a continuous variable, was associated with past 30-day use (OR = 1.08, 95% CI 1.00–1.17). However, after adjusting for covariates, the relationship was no longer statistically significant (aOR = 1.01, 95% CI 0.91–1.12). We found that while female participants were significantly less likely to use substances (aOR = 0.72, 95% CI 0.44, 1.18), those who resided on the Mexico side of the border were more likely to use substances (aOR = 5.90, 95% CI 2.31, 15.09). The results of sensitivity analysis confirmed the results of the primary analyses. Further exploratory analyses showed that there was no multi-collinearity among the covariates. Table 3 shows the results of logistic regression analyses for the relationship between cross-border substance purchase and perceived normalization of drug trafficking. In the unadjusted and adjusted analysis, we found that participants experiencing higher perceived normalization of drug trafficking were more likely than those who did not to cross and purchase drugs (aOR 1.16, 95% CI 1.04–1.29). We also found that older participants had significantly higher odds of past 30-day cross-border substance purchase (OR 1.26; 95% CI 1.02, 1.54).
Discussion
We examined the association between perception of high perceived normalization of drug trafficking and cross-border purchase and use of substances in a sample of adolescents living on the US–Mexico border. By accounting for personal experience and exposure to trafficking activities (perceptions of normalized trafficking), we hypothesized that a high level of exposure would be associated with increased odds of seeking out and consuming substances across the border. We found that participants who experienced higher perceived normalization of drug trafficking were more likely to cross the border to purchase substances. In univariate analysis, we found that higher perceived normalization of drug trafficking was associated with past 30-day use, but this was not significant in adjusted models.
These results support the findings of earlier studies that, among adolescents, the perception of normalized drug trafficking is associated with substance purchase and/or use (Almodovar et al., 2006; Valdez et al., 2019, 2021; Borges et al., 2018). While Borges (2019) argues that there are misconceptions about the high prevalence of substance use in US–Mexico border communities, studies have shown these communities to actually have high occurrence of substance use and a disproportionately high perceived normalization of drug trafficking compared to non-border communities (Valdez et al., 2019, 2021).
Our findings support our hypothesis that neighborhood-level norms (drug trafficking) and social norms influence adolescent substance use patterns in this border sample. Musick, Seltzer, and Shwartz investigated how specific neighborhood-level norms (adults’ attitudes and behavior with respect to substance misuse) influenced these same individual-level outcomes among youth. According to the authors, norms are communicated through the models neighbors provide of appropriate behavior, as well as through social interaction where residents exchange information about their values and the expected costs of violating rules of conduct (Musick et al., 2008). Their results suggested that norms affect teenagers’ behavior in neighborhoods where residents are willing to enforce rules of conduct. Relatedly, in a study on substance misuse in urban areas and colonias on the Texas–Mexico border, Wallisch and Spence (2006) found that about 82% of border residents agreed or strongly agreed there was a lot of drug trafficking in their area. Interestingly, drug trafficking was seen by some study participants as having possible benefits. About 20% of all respondents agreed or strongly agreed that the drug trade had resulted in some positive economic benefits for this area, and that drug trafficking can be a good way for people to raise themselves out of poverty. In this study, substance use was significantly related to the perception that drugs were easy to get and the perception of high substance misuse in the respondent’s neighborhood (Wallisch and Spence, 2006).
Our analysis explores not just whether adolescents are aware of any source for illicit substance use, but whether the normalization of trafficking is associated with viewing Mexico as a reasonable source. Our findings show that youth who perceive that drug trafficking is normalized are more likely to cross the border to purchase substances. This may also relate to “drug tourism.” The existence of fewer legal restrictions at the borders facilitates more fluid movement across the border by which residents have crossed to purchase substances, particularly in Mexico where they are considered cheaper and more readily available (Cherpitel, 2016). A study of adult residents on the border reported that those who drank in Mexico reported significantly more drinks per week (12.8 vs. 8.7, p < .05), were more likely to have binged (58.3% vs. 35.4%, p < .001), and were more likely to report one or more alcohol problem (35.5% vs. 19.5%, p < .01) than those who did not drink in Mexico (Caetano, Mills, Vaeth, 2013). Among border youth, lower drinking ages, cheaper alcohol and tobacco, lenient identification enforcement, and access to substances, including marijuana and illicit drugs, are not uncommon (Lange et al., 2002; Valdez et al., 2019; Valdez & Sifaneck, 1997).
While it may not be necessary to cross the border to access illicit substances (of which tobacco, alcohol, and marijuana are equally illicit as cocaine and heroin for this age group), but if there is an association (as we found there to be), this is important to note. It has implications for differences in the reach of public health laws, policies, and prevention practices as they encounter legal-political boundaries. Easy access to cheap alcohol may encourage heavy drinking and motivate young people to cross the border to drink (Lange and Voas, 2000; Lange et al., 2002). Binational community-level initiatives, including implementing earlier bar closing policies and stricter identification enforcement, should be considered for reducing access to unsupervised and unrestricted drinking among adolescents in border communities (Voas, Lange, and Johnson, 2002). Additional research on successful community-level strategies for adolescent substance misuse in the border context is necessary.
Public health practitioners, researchers, and policy makers should consider the behavioral health implications of living on the US–Mexico border. Their unique environment will affect the effectiveness of prevention initiatives for substance use and misuse on the border, including the perceived normalization of drug trafficking and cross-border access. Despite media fetishization of narco-culture in the US (e.g., Narcos on Netflix and Border Wars on National Geographic), our findings show that the lived experiences of drug trafficking at the border may increase adolescents’ risk of using. Drug trafficking carries risks such as physical and psychological violence, legal implications, deportations, health implications, and family separation. These are inherently traumatic for adolescents living in this region. The struggling local economy and social disorganization, including disordered neighborhood environments (e.g., graffiti, broken windows, and property damage), negative neighborhood perceptions, and drug-related indicators (e.g., neighborhood substance use) further influence adolescent behavioral health and risk-taking (Mack, 2012; Marshall, 2014; Newcomb & Locke, 2005). For adolescents, these environmental factors may increase hopelessness, further increasing risk for substance use; lack of hope limits how their community is sustained and maintained and how they envision future career opportunities in it (Valdez et al., 2021).
Strengths and Limitations
The findings of this study provide insight into environmental factors associated with substance use among adolescents on the US–Mexico border. However, there are some limitations to our study. We used self-reported measures which could potentially bias our findings. The cross-sectional nature of these data limits our ability to account for trends in use or substance purchase within our analysis. Further, the association between perceived normalization of drug trafficking and crossing the border to purchase is likely bi-directional. Perceptions of normalization of drug trafficking may influence adolescents to cross the border to purchase and use substances, and crossing the border to purchase likely affects their perceptions about drug trafficking. Finally, our sample was drawn from only one rural border community. This minimizes the generalizability of our findings.
Implications for Practice
Public health models are needed to shift focus from prohibition and criminalization to emphasize the health consequences of substance misuse (Babor et al., 2010). Broad action should involve binational support from organizations, policy makers, and community members on both sides of the border. These coalitions can support public health preventive measures, increase access to treatment, improvements in education and vocational opportunities for youth, and reduce the availability of substances for the population (Strang et al., 2012; Valdez et al., 2019). Further, the lack of state and local investment and associated scarcity of resources in the US–Mexico border is especially worrisome. Long-term structural and economic reform is needed to mitigate the root causes of drug trafficking and substance use. As long as the troubled economic state in rural border communities continues, along with the high demand for drugs in the US, jobs in the drug trade and its perceived normalization will continue. Addressing the structural environment (e.g., economic) and social context is imperative to prevent substance misuse among adolescents living on the US–Mexico border.
Footnotes
Author Contributions
Dr Elizabeth Salerno Valdez designed the study and wrote the protocol.Mavis Obeng-Kusi and Dr Benjamin Brady conducted statistical analyses.Drs Brady and Huff-Macpherson conducted literature searches and provided summaries of previous research studies.Dr Valdez wrote the first draft of the manuscript and all authors contributed to and have approved the final manuscript.Drs Bell and Derose provided supervision and contributed to review and final development of the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study described in this manuscript was funded by the Ruth L. Kirschstein National Research Services Award, Individual Predoctoral Fellowship [1F31MD012435-01] by the National Institute of Minority Health and Health Disparities, National Institutes of Health; the Program in Migration and Health–California Endowment, UC Berkeley; and the Center for Border Health Disparities, Arizona Health Sciences, University of Arizona. Funders did not play a role in study design, collection, analysis, or interpretation of data, writing the manuscript, and the decision to submit the manuscript for publication.
