Background Adolescent substance use has a detrimental effect on health capital and health trajectories later in life. Although school socioeconomic context is known to influence academic outcomes and access to tertiary education, less is known on its impact on adolescent behaviours such as substance use. In this study, we used data from Chile, which has a socially segregated schooling system, to examine whether school socioeconomic context relates to recent substance use among Year 8 (Y8) and Year 10 (Y10) students.

Methods Individual-level information on past-month alcohol, tobacco and marijuana use was self-reported by Y8 students in 2017 (N= 91,220; 8,367 school classes; 5,714 schools) and Y10 students in 2018 (N= 164,693; 7,119 school classes; 2,922 schools). Cross-sectional data was collected as part of the Educational Quality Assessment System implemented biannually for Y8 and Y10 students at national level. Parents reported their own education level during this assessment. Schools were classified into five socioeconomic groups (low, lower middle, middle, upper middle, high) according to official Ministry of Education information. Hierarchical data were analysed using multilevel logistic regression. Models were adjusted for parental education and child’s gender. Age information was not available. Analyses were performed using Stata 17.

Results Alcohol was the most frequently consumed substance among Y8 students (14.5%; 95%CI 14.3, 14.7) and Y10 students (34.6%; 95%CI 34.4, 34.8). Of Y8 students, about 7% reported recent smoking (95%CI 7.3, 7.7) and 6% reported recent marijuana use (95%CI 6.0, 6.2). Among Y10 students, these numbers doubled for tobacco (15.7%; 95%CI 15.5, 15.9) and marijuana use (13.7%; 95%CI 13.5, 13.9)

Results from multilevel models showed Y8 students attending high socioeconomic status (SES) schools were less likely to smoke (OR=0.52, 95%CI 0.45–0.62) and use marijuana (OR=0.35, 95%CI 0.28–0.42) compared to those from low SES schools. Among Y10 students, risk of tobacco use was lower in middle SES schools (OR=0.64, 95%CI 0.61–0.74) and risk of marijuana use was about 50% lower in high SES schools than in more socially disadvantaged schools. Among Y10 students there was an inverse social gradient in alcohol use, in which high SES school students were more likely to drink than those attending low SES schools (OR=1.77, 95%CI 1.66–1.90). School SES was unrelated to drinking among Y8 students.

Conclusion The influence of school socioeconomic context on adolescent substance use seems to differ depending on the substance. Students attending socially deprived schools are likely to be more vulnerable to tobacco and marijuana use and their long-term consequences.