The primary goal of our study was to replicate the finding reported by Dawson and colleagues (2007) of a stronger association between number of past-year stressors and alcohol intake among individuals who initiated drinking at age 14 or younger compared with those who began drinking at age 18 or later. Although the pattern of our findings is consistent with the interaction reported by Dawson and colleagues (2007), the main effect of stressful life events (regardless of dependence) and the interaction of onset age with life events explained relatively little of the variation in past-year drinking. For both sexes, having 1 or more life events accounted for ≤0.5%, and the life event by onset age interaction accounted for ≤1% of the variation in drinking. In contrast, age at drinking onset was a more powerful predictor of variation in past-year drinking in both sexes, accounting for 3 to 6% of the total variance, similar in magnitude to the combined effects of sociodemographic variables.
