The association between smoking behaviour and depressive symptoms in adolescence: the role of biology and of societal influences

Sammanfattning: Tobacco use is common among people who suffer from depressive symptoms. This cooccurrence is associated with several negative health outcomes. The present thesis will contribute to the understanding of if, how, and in which contexts this specific association occurs in early adolescence. We investigated the longitudinal association between tobacco use and depressive symptoms, the potential biological mechanisms behind this association and the extent to which the social context modifies the co-occurrence of these two major health determinants. Data are obtained from the KUPOL study, a population-based cohort of Swedish adolescents. The evaluation of the social context also includes data from the Italian BE-TEEN study. Study I. The possible association between smoking or snus use and depressive symptoms was examined in 3,195 KUPOL study participants. Smoking behaviour was associated with an increased risk of the onset of depressive symptoms among adolescents. Study II. We explored the possible involvement of activation of the hypothalamic-pituitaryadrenocortical axis, measured through salivary cortisol concentration, in the pathway from cigarette smoking to depressive symptoms in 409 KUPOL study participants. There was no evidence for an association between cigarette smoking and the hypothalamic-pituitaryadrenocortical axis levels of activity, and the latter did not predict an increased risk of the onset of depressive symptoms. Study III. Salivary cortisol concentration as a predictor of the onset of tobacco use was investigated in 381 KUPOL study participants. Morning cortisol concentration was associated with an increased risk of smoking and snus initiation as well as duration of use. These findings suggest an association between activation of the hypothalamic-pituitary-adrenocortical axis and tobacco use. Study IV. We evaluated the presence of the association between smoking and depressive symptoms in a restrictive (Sweden) and in a non-restrictive (Italy) tobacco control environment including 3,283 Swedish and 1,947 Italian participants. Cross-sectional associations between smoking and depressive symptoms were found in both the restrictive and non-restrictive tobacco control environments, with the strongest association in the former. Conclusions. This thesis indicates that smoking behaviour is a predictor of the development of depressive symptoms in adolescence. The pathways behind this longitudinal association are not clear. Seemingly, the hypothalamic-pituitary-adrenocortical axis is not implicated in the process from smoking to depressive symptoms. However, findings indicate that the initial activation of the hypothalamic-pituitary-adrenocortical axis may be linked to smoking onset. This opens up for new research avenues on dysregulation of the response to stressors, such as adverse life events, and smoking onset at a young age. Finally, translating this emerging knowledge into prevention programmes may help to develop context-specific interventions and to direct efforts towards specific subgroups of adolescents at high risk of tobacco use and depressive problems.