Objective: Utilizing a longitudinal prospective design, the purpose of this
study was to: (I) assess the age of onset to begin drinking in relation to
family history of alcoholism using survival analysis, and (2) examine the
importance of selected risk factors in predicting outcome, using a Cox prop
ortional hazards model analysis. Method: Fifty-two children and adolescents
at low risk for developing alcoholism and 73 children and adolescents from
high-risk families were studied. Subjects, spanning the ages of 7 to 18, w
ere evaluated annually (268 total evaluations), providing 2.1 waves of long
itudinal data concerning age of onset along with a number of predictors (po
sitive familial loading of alcoholism, extraversion, and manifest anxiety s
cores). Results: High-risk children showed a significantly earlier age of o
nset to begin drinking. The Cox proportional hazards analysis revealed that
the onset could be predicted by a positive familial loading of alcoholism
and extraversion. Further analyses revealed that extraversion was a mediato
r of the familial density effect. Conclusions: Age of onset to begin drinki
ng has been shown in general population studies to predict the likelihood o
f developing alcohol/abuse/dependence problems. Further follow-up will dete
rmine if these high-risk children develop alcohol abuse/dependence at highe
r rates than control children. Further follow-up would also establish links
between early childhood predictors, such as having an extraverted temperam
ent and development of alcohol problems.