The current study uses longitudinal data from a large, population-based sample, and benefits from the use of standardized instruments for the assessment of depressive symptoms and alcohol use. The latent variable methods applied enable us to examine the association between the effects of mean levels of depressive symptoms, and growth in depressive symptoms, on both the use and harmful use of alcohol. Our findings indicate that these relationships differ across males and females and the alcohol outcome of interest. Despite this variation, increased depressive symptoms during adolescence are generally associated with more alcohol use at 18 years 8 months. The predictive nature of depressive symptoms in both sexes could be helpful for public health and clinical efforts (e.g., education, identification, intervention), as demonstrated by previous findings that a cognitive behavioural intervention approach among teens with depressive symptoms can reduce later substance use [44]. The finding that the nature of the alcohol outcome may be critical in determining whether an association with depressive symptoms exists is an important one, and should be considered when comparing the results of studies that examine
