Period, age, and cohort effects on substance use among American youth, 1976-82.

Period, age, and cohort effects on substance use are differentiated for American youth 18 to 24 years old during the period from 1976 to 1982. The data are provided by the Monitoring the Future project, an ongoing study which employs a cohort-sequential design. Weighted least squares regression is used to find plausible and parsimonious models to account for the observed variation in 12 different classes of drugs, both licit and illicit. The point is made that there are no definitive ways to differentiate among the types of effects; thus, any interpretation is open to debate. Period effects involving increased use occurred for cocaine, amphetamines, and methaqualone, while decreases occurred for barbiturates, tranquilizers, and psychedelics other than LSD. Marijuana showed a curvilinear period effect, first increasing then decreasing. Effects of age were more complex. There were increases in the year after high school for daily cigarette use, but not for monthly use. Monthly and daily alcohol use increased with age. A measure of heavy drinking showed a curvilinear trend, first increasing and then decreasing. Annual use of cocaine showed an increase between the ages of 18 and 21. Annual use of narcotics other than heroin showed a linear age decrease. Clear class (or cohort) effects appeared for cigarette use, with each successive class smoking less.