Age and birth cohort effects on rates of alcohol dependence

Background: Epidemiological studies of traits such as alcohol dependence and depression have often found lifetime rates in younger individuals exceeding those found in older individuals. This suggests additional influences of birth cohort or period effects so that individuals in later-born cohorts have an increased lifetime risk. Methods: Data from the Collaborative Study on the Genetics of Alcoholism were used to investigate secular trends for alcoholism and related conditions and to examine risk predictors while taking the cohort effect into account. We used data on 4099 interviewed parents and siblings of alcohol-dependent subjects and 1054 members of control families. We used survival analysis techniques and the Cox proportional hazards regression model to estimate the relative risk for demographic covariates. We used the relative sample to predict risk in the sibling of the proband and family history information to determine whether there was a bias when deceased individuals were excluded from analysis. Results: In the control sample, we observed a 1.8% lifetime rate of DSM-III-R alcohol dependence in women born before 1940, as contrasted to a 13% rate in women born after 1960, and a 15% lifetime rate in men born before 1940, contrasted with a 28% rate in men born after 1960. As expected, lifetime rates in relatives were increased when compared with controls. Highly significant risk ratios (RR) were observed for gender (RR, 2.3), cohort of birth (RR, 1.5 over a decade), daily smoking (RR, 2.0), heavy smoking (RR, 3.0), and comorbid diagnoses of antisocial personality (RR, 2.2) and depression (RR, 1.6). Analysis of the family history data indicated higher rates of alcohol dependence in relatives who were deceased compared to those who were living. Conclusions: Marked cohort differences were observed and may reflect real changes over time, or artifacts of memory recall, differential mortality, or public awareness. The analysis of all relatives (living or deceased) indicates that associated mortality may, in part, explain the secular trends seen when analyses are restricted to living, personally interviewed individuals.