Substance use during pregnancy in the state of California, USA.

Most analyses of prenatal substance use focus on individual level correlates and ignore community level variables and the effect of the dependency of respondents within communities. This analysis uses multilevel logistic regression models to more accurately assess the correlates of perinatal substance use in California. Statistical results indicate that a significant portion of substance use can be attributed to neighborhood heterogeneity, and that traditional models of substance use may inaccurately attribute this variation to individual level regression coefficients. Substantive results indicate that levels of neighborhood public assistance had an independent, significant effect on the prevalence of all substances tested for except alcohol. Black women had higher predicted prevalence risks for alcohol and cocaine while White women had higher predicted risks for tobacco, marijuana and amphetamines. Racial contrasts were non-significant for the overall illicit drug category and opiates, after controlling for neighborhood public assistance. Finally, individual level variables, with the exception of age, were not moderated by levels of neighborhood public assistance.