Early initiation of drug use has be shown to be associated with increased risk of later drug abuse and development of drug dependence. The hypothesis tested in this four wave longitudinal study of urban schoolchildren, conducted in Baltimore MD concerned a suspected causal association between neighborhood deterioration and the risk of early initiation of drug use. A total of 925 participants completed four confidential standardized interviews conducted by specially trained interviewers in the spring of four successive years. In addition to drug use assessment, separate standard scales were applied to measure an array of risk factors, including neighborhood deterioration, peer drug use, parental supervision and antisocial behavior. By 1992, a total of 340 youths, then 12-14 years old, reported drug involvement (smoking tobacco, using alcohol without parental permission, use of inhalants, marijuana, or cocaine). An analysis using Kaplan-Meier curves showed that throughout the follow-up interval children reported earlier onset of drug use in areas rated more deteriorated (log rank test with 2 d.f.: 11.59, p-value: 0.003). Under the Cox proportional hazards model, holding constant other factors (peer drug use, parent monitoring, antisocial behavior, etc.), this association still persisted, resulting in a hazard ratio of 1.58 between the lowest and highest tertile of neighborhood deterioration (95% CI: 1.13-2.23). Thus, neighborhood deterioration appeared to be associated with early onset of drug use independent from other suspected factors. We plan to extent this investigation and use geographical software (GIS) to include census tract information, which we now have available.
