Few longitudinal studies exist that evaluate changes in predictors of drug abuse or prevention program effects over multiple risk periods or that systematically address risk factors, program effects, and program effect mediators in the same study. The specific objectives of this longitudinal etiology and prevention study are to evaluate: (1) long-term effects of an earlier community-based prevention program for adolescents on reducing drug abuse and related problem behaviors (drunk driving, social dysfunction, job and school drop-out, unprotected sex) and on increasing participation in prevention and health promotion activities in adulthood (healthy lifestyle choices); (2) changes in mediators of early drug use risk, particularly mediators targeted by the earlier community program; and (3) changes in the strength and pattern of personal, social, and environmental drug abuse predictors as individuals move through four potential risk periods (early adolescence, adolescence, onset and end of early adulthood). Two expanded panels of adolescents will be tracked through early adulthood, half of whom were previously assigned to a control group or a group that received a comprehensive community-based intervention that included school, parent, community organization, health policy change, and mass media program components, beginning in middle school. Three additional waves of survey and archival data on two panels of individuals from Kansas City and Indianapolis (75% white, 23% African-American, 2% Other; 51% female; 63.2% low SES; 23.5% inner city; N=2646) will be collected. Merged with up to nine previously collected waves of data, this proposed study will include subjects spanning the ages 12-29. Panels represent an overall 17% oversample compared to previous waves to enable stratified tests of drug use development and program effects by ethnicity. Based on 1993-94 data, 25.0% reside outside of the original study area, .4% have died, .7% are in prison, 2.5% needed treatment for drug abuse, and 4.4% were in trouble with police over a drug related problem. Hierarchical regression, growth curve, and structural equation analyses will be compared for their ability to predict drug abuse development and program effects with models that include personal, social, and environmental influences. The proposed study is important because it encompasses four potential periods of drug use risk (early adolescence, adolescence, onset of early adulthood, end of early adulthood--ages 12-29); includes multiple waves of data on two diverse populations; is large enough to enable test replication across sites and population sub-groups; incorporates and compares sophisticated analytical techniques to evaluate models of drug use and program effects; and represents the most long-term study of drug abuse prevention program effects in the U.S. Results should inform at least two important questions relevant to drug abuse prevention: whether a drug prevention program aimed at adolescents has the capacity to affect drug abuse, problem behaviors, and healthy lifestyle choices, and whether patterns of risk factors change over different periods, which would suggest that prevention efforts may require a different emphasis depending on age/stage of risk implemented.
