Depression/anxiety disorders and smoking behavior often begin in adolescence as co-occurring phenomena. Epidemiologically, the relationship between them is bidirectional,

To examine both adolescent smoking as a predictor of depression/anxiety disorders, and the effect of these disorders on adolescent smoking.

A representative sample of Catalan adolescents (N = 3008) participating in (PFI) a longitudinal study were interviewed about smoking behavior at baseline (T1) and at two follow-up assessments (T2 and T3). In interviews at baseline (T1) and at T4, the parents were asked whether their daughters/sons had a diagnosis of depression or anxiety. Smoking behavior was defined as smoking at least once a week. Age, sex and household income were included.

Depression/anxiety at T1 increases the risk of smoking behavior at T3 [odds ratio (OR), 1.97, 95% confidence interval (CI) 1.24-3.14] compared with undiagnosed adolescents. After adjusting for age and sex, the risk remains, but after adjusting for sex, age and income, the risk of smoking behavior decreases as household income increases [OR 1.6, 95% CI 0.9-2.85]. Alternatively, smoking behavior at T3 increases the risk of depression/anxiety at T4 [OR 1.7, 95% CI 1.1-2.5] compared with non-smokers. After adjusting for age and sex, the risk remains, but after adjusting for sex, age and income, the risk of depression/anxiety decreases as household income increases [OR 1.48, 95% CI 0.93-2.36].

Our findings provide the first evidence of a two-way relationship between adolescent smoking and depression/anxiety disorders in a community sample in Spain.