The association of psychiatric symptomatology with patterns of alcohol, tobacco, and marijuana use among Brazilian high school students.


BACKGROUND AND OBJECTIVES: Studies have highlighted psychosocial factors associated with drug use among adolescents. Association of specific psychiatric comorbidity with substance use has not been properly established in Brazil. This study aimed to investigate alcohol, tobacco, and marijuana use by 15-18-year-old high school Brazilian students and to estimate associations with psychiatric symptoms. METHODS: A cross-sectional survey of 4,034 students from 128 public and private schools in Sao Paulo State was carried out using a two-step probability sample. Data were collected through self-report standardized questionnaires including questions on substance use patterns and the Strengths and Difficulties Questionnaire (SDQ). Key outcome variables were past-month use and past-month frequent use of alcohol, tobacco and marijuana. Questionnaires with missing information were excluded, resulting in a final sample of 2,532 adolescents. Weighted data was analyzed through logistic regressions, adjusted by gender and by socio-economic status (SES). RESULTS: Regarding SDQ total score, 43.6% of students had no psychiatric symptoms, 7.9% had subclinical symptoms and 48.5% presented clinically significant symptoms. Respondents with a clinically significant SDQ score were more likely to be past month alcohol (aOR = 1.51; 95%CI 1.22-1.88), tobacco (aOR = 1.82; 95%CI 1.25-2.66), and marijuana (aOR = 1.79; 95%CI 1.21-2.64) users as compared to those with no symptomatology. DISCUSSION AND CONCLUSIONS: Psychopathological symptoms were associated with alcohol, tobacco and marijuana use by 15-18-year-old adolescents. These associations should also be considered when planning public policies of mental health promotion. SCIENTIFIC SIGNIFICANCE: This study discusses the importance of the association between psychopathological symptoms and substance use in a middle-income country, with high level of social inequalities, in a state representative sample. (Am J Addict 2016;25:416-425).