Patterns of drinking and alcohol problems change with age. However, few studies use multiple data points and detailed history spanning early adulthood to middle age. This study reports such data from 373 men in the San Diego Prospective Study.
Data were generated at baseline (T1) at 
age 20, and through face-to-face followup interviews every 5 years in >90% of these eligible Caucasian and relatively higher educated men. Subjects were placed into 4 groups regarding their course: 62.5% with no alcohol use disorder (AUD); 17.2% with AUD onset
On a univariate level, low level of response (LR) to alcohol, family history of AUDs, and higher Novelty Seeking at age 20 predicted AUDs with onset before age 30 (mean age 25), but among these only LR predicted later onset (mean age 38) as well. Additional predictors of AUDs included demography (lower education), and greater involvement with alcohol, drugs, and nicotine prior to T1. Sustained remission from AUDs among alcoholics was predicted by lower T1 and T10 drinking frequencies, and being separated or divorced at T10, along with a trend for higher Reward Dependence.
These data indicate that information available in ages of the late teens to early twenties can help predict the future onset and course of AUDs, and underscore the importance of longitudinal studies in substance use disorders.
Request Reprint E-Mail: mschuckit@ucsd.edu
