As a third finding, the cross-lagged models indicated that student status at age 17 positively predicted any drinking and drinking to intoxication in young adulthood. Thus, adolescents who are more likely to reach a higher educational level increase their alcohol use as young adults more than do those who truncate their education at a lower level. That might be explained by a transient “college effect” where studying young adults drink more than their non-studying peers (O’Malley and Johnston, 2002). Alternatively, it may reflect a more permanent effect in which the association between education and alcohol use reverses in young adulthood from negative to positive (Crosnoe and Riegle-Crumb, 2007). Compatible with the latter explanation, Finnish adults with higher education tend to drink more often and abstain less than those with lower education, although they have a lower rate of alcohol use disorders (Helakorpi et al., 2012; Helasoja et al., 2007; Latvala et al., 2009). Interestingly, student status at age 17 also appeared to predict the frequency of any drinking stronger than the frequency of drinking to intoxication.
