RESUMO
OBJECTIVES: We compared rates of smoking for 2 groups of youths aged 12 to 14 years: those involved in the child welfare system (CW) and their counterparts in the community population. We then investigated factors associated with smoking for each group. METHODS: We drew data from 2 national-level US sources: the National Survey of Child and Adolescent Well-Being and the National Longitudinal Study of Adolescent Health. We estimated logistic regression models for 3 binary outcome measures of smoking behavior: lifetime, current, and regular smoking. RESULTS: CW-involved youths had significantly higher rates of lifetime smoking (43% vs 32%) and current smoking (23% vs 18%) than did youths in the community population. For CW-involved youths, delinquency and smoking were strongly linked. Among youths in the community population, multiple factors, including youth demographics and emotional and behavioral health, affected smoking behavior. CONCLUSIONS: Smoking prevalence was notably higher among CW-involved youths than among the community population. In light of the persistent public health impact of smoking, more attention should be focused on identification of risk factors for prevention and early intervention efforts among the CW-involved population.