Age-period-cohort effects on substance use in adolescence, 1976-2006

DESCRIPTION (provided by applicant): The identification of age-period-cohort effects in alcohol and drug outcomes provides substantially more information regarding how alcohol and drug use change over time compared to the traditional population rate by year. Further, information on attitudes toward substance use such as disapproval can be analyzed in the age-period-cohort framework to provide information about group-level mechanisms of risk through which lifetime initiation of substance use changes. Unfortunately, age-period-cohort analysis remains an underutilized methodological tool due to the limitations of pre-existing methods to reliably estimate effects. Recently, however, new methodological tools have been developed focusing on aspects of the APC model unaffected by model identification problems, motivating a multi-disciplinary resurgence of interest in APC analysis. The overarching aim of the present research is to apply recent methodological advancements in statistical age-period-cohort effect estimation to quantify the combination of age, period, and cohort effects that give rise to changes in alcohol and drug use as well as disapproval of use in adolescence over time. This research will integrate age-period-cohort effects in substance use with age-period-cohort effects in attitudes toward use to assess heterogeneity in these two related but potentially diverse constructs. Data will be drawn from the Monitoring the Future (MTF) study, a nationally-representative yearly cross-sectional survey of adolescents in the U.S. from 1976 to the present. A unique feature of the MTF data is the standardization of sampling design and measures across 30 years of data collection. This work will focus specifically on estimating age-period-cohort effects in the lifetime initiation of three substances (alcohol, tobacco, and marijuana) and also on estimating age-period-cohort effects in disapproval of alcohol, tobacco, and marijuana. PUBLIC HEALTH RELEVANCE: The present research has implications for public health policy and continued research. Information regarding the similarities and differences between age-period-cohort effects in substance use and disapproval can provide a foundation for productive hypothesis testing. Results of the present study will suggest population- level targets for prevention and intervention with adolescents, as the presence of age, period, and/or cohort effects can provide an initial characterization of effective interventions in populations.