Motivated by the alarming increases in youth smoking prevalence since 1990, and the failure, to date, of school-based interventions to effect long-term reductions in smoking, this project will investigate the extent to which parent factors, both before and during the child/adolescent smoking acquisition period, can predict the smoking status of the child as he/she exits high school. The study design capitalizes on existing data from a well- maintained, population-based cohort of 4,196 children (now ages 18-21 years), and their parents, from the 20 school districts in the control (no intervention) arm of a long-term randomized trial in school-based smoking prevention. These children constituted the entire third grade enrollment of their school districts in two consecutive school years, making it a population-based cohort. The data available for this project are unique because: (1) the student cohort data spans ages 8-18 (including school dropouts) - virtually the entire period of smoking acquisition, (2) relevant longitudinal data are also available on the students' parents, schools, and communities, (3) the population- based nature of the cohort will yield generalizable results, and (4) findings will be available quickly. Using the rich resources of a large cohort of children will provide a fast, efficient, and low cost method for conducting important research that will yield information about parent characteristics that predict children's smoking. This information will be disseminated immediately to relevant audiences to aid development of future research initiatives and, possibly, development and testing of family-focused smoking interventions, and thus contribute to the smoking and cancer mortality reduction goals of the National Cancer Program.
