This study has a number of strengths. Firstly, through the use of repeated measures of self-reported depressive symptoms, this study was able to identify of trajectories of these symptoms from early through late adolescence and relate these longitudinal patterns with later alcohol behaviour. This is in contrast to previous studies that have been limited to cross-sectional or a single follow-up [8–11]. Secondly, the availability of data from the AUDIT has permitted us to explore whether the relationship between depressive symptoms and alcohol phenotypes differed depending on the nature of the phenotype (i.e., use vs. harmful use). The presence and strength of the associations between depressive symptoms and alcohol outcomes varied across the sexes; across facets of depressive symptoms (intercept vs. slope); and across alcohol outcomes (use vs. harmful use).
