To our knowledge, this is the first study to report national estimates of lifetime substance use and potential etiologic factors among secondary school students in Costa Rica. Alcohol was the most commonly reported substance, followed by e-cigarettes, then tobacco, other illegal drugs, and marijuana. These patterns—excluding e-cigarette use, which was not previously measured—are consistent with findings from Costa Rica’s 2018 National Student Survey, which also identified alcohol as the most frequently used substance (2). However, overall substance use prevalence declined between 2018 and 2021, with the largest reductions observed for alcohol (25%) and marijuana (23%). These decreases should be interpreted in light of the COVID-19 pandemic, as data for the present study were collected in 2021, a period of national restrictions that likely limited adolescents’ opportunities for substance use (32). For example, in the United States, nationally representative data also documented declines in adolescent substance use between 2019 and 2021, with alcohol use dropping by about 22% and marijuana use by 27%. These trends lend support to the idea that pandemic-related disruptions, such as social distancing measures, may have temporarily reduced adolescents’ access to substances (33). These trends support the hypothesis that pandemic-related disruptions, such as social distancing, temporarily reduced adolescents’ access to substances. As such, our findings may reflect substance use patterns specific to 2021 and may underestimate typical prevalence and trends in Costa Rica, limiting their generalizability to subsequent years.

Notably, this study is the first to analyze national data on e-cigarette use among students in Costa Rica. Although e-cigarette use was the second most prevalent substance among students in Costa Rica, its prevalence was lower—potentially due to pandemic-related effects—than reported in other countries in the Americas, such as Colombia (32%) (34), the United States (17%) (35), and Argentina (14.4%) (36). Nevertheless, the findings regarding e-cigarette use point to an emerging public health concern for Costa Rica and the broader Latin American region.

Concerning sociodemographic factors, men were more likely than women to report lifetime use of marijuana and other illegal drugs, whereas women were more likely to report lifetime alcohol use. This pattern aligns with growing evidence of a narrowing gender gap in adolescent substance use, with young women increasingly exhibiting rates comparable to—or exceeding—those of young men. For example, the 2021–2022 international report of the Health Behavior in School-aged Children highlights a significant shift in substance use patterns across 44 countries and regions in Europe, Central Asia, and Canada, challenging historical sex-related differences in consumption that have traditionally shown greater prevalence among men. The report indicates that by age 15, young women often catch up to young men in cigarette smoking and e-cigarette use and surpass them in alcohol consumption (13).

In addition, age was a consistent predictor across substances, with older adolescents generally reporting greater use—except for e-cigarettes and other illegal drugs—consistent with a previous study conducted in Costa Rica using a nationally representative youth sample (22). Finally, living in provinces outside the capital San Jose was associated with a lower likelihood of alcohol, tobacco, e-cigarette, and marijuana use—but not other illegal drug use. Similarly, studying in urban schools was associated with a lower prevalence of e-cigarette use. These findings likely reflect an association between greater population density and increased substance use in urban settings, consistent with prior research documenting higher rates of adolescent substance use in more populated areas (15). Findings may also be influenced by differences in substance availability. For instance, legal substances such as tobacco, might be more easily accessible in urban areas, potentially contributing to the observed patterns, while limited availability in rural settings may shape distinct substance use behaviors. Future studies should explore how ease of access to legal substances vis-à-vis illegal substances might influence patterns of consumption in Costa Rica.

Economic factors also influenced adolescent substance use, with Costa Rican students receiving a greater weekly allowance (i.e., > 20$ per week) showing an increased likelihood of alcohol, tobacco, and marijuana use. This finding is consistent with prior research from various countries demonstrating positive associations between disposable income and substance use behaviors. For example, studies of European school samples have found that greater weekly allowances are associated with increased use of both licit (i.e., alcohol and tobacco) and illicit (i.e., marijuana and ecstasy) substances (14, 37–39).

Further, working while studying was associated specifically with alcohol use, which aligns with prior research showing that part-time employment during adolescence is associated with increased alcohol consumption. For instance, Finnish adolescents working over 10 h per week had a significantly higher odds of reporting heavy drinking (40). Similarly, Canadian high school students with part-time jobs reported greater alcohol use (41). Income from part-time work has also been associated with licit substance use, such as alcohol and tobacco, possibly reflecting the structured nature of employment or personal traits like financial responsibility (38). These findings have implications for school- and family-level prevention. Including financial management education and opportunities for positive involvement—such as those offered in evidence-based parenting programs like Guiding Good Choices (42)—may be valuable components in adolescent prevention curricula in Costa Rica.

Findings confirmed that social environments significantly shape substance use patterns in Costa Rica. For instance, we found that students with more friends were more likely to have ever consumed alcohol. Prior research shows that larger adolescent social networks contribute to underage drinking (16). Although we did not directly ask about peer substance use, students reporting more friends probably have more close peers who drink, which is a well-established factor influencing youth alcohol use (43). Regarding substance use within the family environment, family binge drinking was associated only with alcohol use, consistent with a study of over 4,700 Finnish adolescents showing that parents’ drinking predicted their children’s alcohol use and intoxication (44). Additionally, family use of illegal drugs was associated with students’ use of all substances except e-cigarettes, a result supported by a 2022 meta-analysis reporting that both maternal and paternal substance use increase the likelihood of child drug use (45).

Surprisingly, students who had smokers in the family reported lower e-cigarette use, a finding that contradicts meta-analytic evidence suggesting a positive association between adolescent e-cigarette use and family smoking (46). Further research is needed to better understand the connection between social environments—particularly the family—and e-cigarette use among Costa Rican students. Finally, consistent with findings from a 12-year longitudinal study conducted in Iceland, which showed that parental monitoring significantly reduced alcohol use and intoxication among adolescents, our results also identified parental monitoring as an important protective factor in the family domain (47). In contrast, rule-setting and expressions of affection did not show a protective effect. Their high prevalence among parents (>80%) may have reduced variability and limited our ability to detect potential associations due to ceiling effects. Additionally, our ability to fully assess these family-level protective factors was constrained by the use of dichotomous indicators (i.e., single-items), which may have reduced the granularity of the information included in the analysis. Future studies would benefit from incorporating validated scales–developed with multiple tested items–to more accurately capture the nuances of these constructs.

Students’ emotional distress emerged as an important individual-level etiologic factor for alcohol use. Specifically, our findings suggest that the three indicators—loneliness, feelings of sadness or hopelessness, and suicidal thoughts—were each associated with more than a 40% increase in underage drinking among students in Costa Rica. Similar findings have been reported in a nationally representative school-based study in South Africa, where feelings of sadness or hopelessness were significantly associated with binge drinking among adolescents (20). Furthermore, our findings mirror other studies reporting that alcohol use was associated with suicidal ideation and attempts among Chinese students (48). Similar associations—particularly between suicidal thoughts and alcohol use—have also been observed in a study conducted in the United States among high school students (21).

When examining the association between emotional distress and tobacco use, results indicate greater tobacco use among students reporting increased loneliness and suicidal ideation. Similar studies using data drawn from the Global School-Based Student Health Survey–led by the World Health Organization–found the same positive associations in the context of loneliness vis-à-vis tobacco use among students in South Asia (49) and the Caribbean (i.e., Dominican Republic, Jamaica, Suriname, Trinidad and Tobago) (50). Furthermore, Dasagi et al. (51), using the nationally representative Youth Risk Behavior Surveillance Survey in the United States, found that heavy cigarette smoking increased suicide risk among adolescents.

In addition, our findings indicate that the use of other illegal drugs was associated with elevated reports of sadness or hopelessness and suicidal thoughts. To our knowledge, no studies have directly examined the association between other illegal drugs use and self-reported single-item measures of feelings such as sadness or hopelessness. However, akin associations have been reported in studies employing validated measures of depressive symptomatology—such as the Short Mood and Feelings Questionnaire (SMFQ) (52), which includes items related to these emotional states. For instance, a longitudinal study conducted in Australia using the SMFQ found that (a) early adolescent depressive symptoms predicted the frequency of mid-adolescent illicit drug use, and (b) mid-adolescent depressive symptoms predicted the frequency of illicit drug use in late adolescence (53). Importantly, a previous study among 2,284 Costa Rican students from 64 middle and high schools also found an association between suicidal thoughts and a measure of substance use that included illicit drugs, consistent with our findings.

Unexpectedly, e-cigarette use was less prevalent among students reporting suicidal thoughts—a finding that contrasts with meta-analytic evidence using studies from the United States, Canada, and South Korea, where adolescent e-cigarette use has been significantly and positively associated with suicidal ideation, planning, and attempts (54). As noted in this meta-analysis, cultural and social factors may shape the association between e-cigarette use and suicidal behaviors. For example, differences in attitudes toward mental health and e-cigarettes—variables not measured in this study and potentially important covariates—could influence both adolescents’ reporting of mental health issues and their engagement with these products, potentially explaining discrepancies across populations (54). Additionally, common sociodemographic predictors of substance use, such as age and sex, were unrelated to e-cigarette use, a pattern similarly observed in recent studies from Mexico (55). Overall, these inconsistencies underscore the need for further longitudinal research on the etiological and contextual determinants of adolescent e-cigarette use in Costa Rica.