|| Checking for direct PDF access through Ovid
We examine changes among adolescent girls in substance use during pregnancy and the postpartum period. Three separate latent growth curve analyses assessed the impact of psychosocial, behavioral, and sociodemographic factors on resumption of or change in use of cigarettes, alcohol, and marijuana. The Vulnerable Populations Model for Research and Clinical Practice (Flaskerud & Winslow, 1998) provided the theoretical foundation for this study. This is a secondary analysis of data from a sample of 305 ethnic minority females (245 Latina, 60 African American), aged 13–18 years, who were pregnant at baseline and were participating in an HIV prevention study conducted in inner-city alternative schools in Los Angeles County. Data collected at 4 time points captured changes in substance use from pregnancy through the postpartum period. Baseline predictors included ethnicity/race, partner substance use, childhood abuse history, religiosity, acculturation, depressive symptoms, length of gestation at baseline, and previous substance use. Common predictors of greater resumption and/or greater level of use included greater history of use before pregnancy, partner substance use, childhood abuse, and a longer time since childbirth. African Americans were more likely to be smoking at baseline when they were still pregnant and to use marijuana postpartum; Latinas were more likely to use alcohol over the course of pregnancy and postpartum. Other variables exerted an influence on specific substances. For instance, religiosity impacted cigarette and alcohol use. Findings may assist prenatal care providers to identify and counsel pregnant adolescents at risk for perinatal substance use and to prevent resumption or initiation of substance use after childbirth.