Purpose
Research examining the influences of neighborhood characteristics on adolescent drug use and behavioral problems in developing countries is scarce. We know of only one other study, in addition to the present one, that has examined contextual level influences on adolescent drug use and mental health in the entire Latin American continent. In this presentation we describe the implementation of the “Santiago Longitudinal Study” and discuss findings that highlight the way(s) by which neighborhoods characteristics were associated with adolescent drug use and mental health outcomes.
Methods
Over 1000 youth participated in this NIDA-funded study. Adolescents completed a two-hour interviewer-administered questionnaire with standardized questions assessing substance use behaviors, mental health, peer, parenting, and family relationships. The study also collected neighborhood data that consisted of Chilean Census data, participants' responses to several questions on crime and social capital, and a systematic observation of the neighborhoods. Neighborhood Systematic Observations were conducted by a Chilean social work faculty member and a Chilean MSW student. Together, these observers assessed over 1000 neighborhood blocks using a three-page rating instrument for each of the four corresponding blocks. The instrument included questions assessing various aspects of the environment such as quality of housing exteriors; presence of graffiti, abandoned cars; street dogs; street and sidewalk condition; vacant buildings and construction; quality of parks, plazas, and children's playgrounds, among others. Data were mapped using ArcGIS and analyzed using multilevel analyses.
Results
Data from the adolescent interviews indicate that neighborhoods with higher crime rates are positively associated with substance use and poor mental health outcomes. Multi-level analyses indicated that results depended on the level of aggregation and type of outcome. For example, there was between ‘comuna' (municipality), but not between ‘población' (neighborhood block), variation in Achenbach's YSR anxiety-depression and rule breaking scores. On the other hand, there was between ‘población', but not between ‘comuna', variation in alcohol and marijuana use. Further differences are noted when data from Neighborhood Systematic Observations are included. Alcohol and cigarette use were more common among youth who live in ‘comunas' that have less garbage in the streets. However, anxiety-depression scores were not related to extent of street garbage. At the ‘población' level, the number of local neighborhoods with more homes that have gardens was positively associated with adolescent cigarettes and marijuana use, and inversely with anxiety-depression scores. The number of local neighborhoods with more homes that are better maintained was positively associated with alcohol and marijuana use but not anxiety-depression. Interestingly, different aspects of the built-in environment relate differently to adolescent outcomes and these associations vary by type of outcomes.
Conclusions and Implications
Central themes of this project include a rich description of the implementation of systematic neighborhood assessment in a developing country, coding and spatial analysis of data using GIS, and illustration of results using multilevel analysis. Findings have important implications for understanding the effect of community context on youth outcomes. Neighborhood level factors are associated with individual outcomes, but the associations clearly depend on the spatial units used to define neighborhoods.
