Cocaine exposure and developmental outcome from birth to 6 months.

The theoretical framework for many of the early studies of prenatal cocaine exposure has been rooted in the basic concepts of teratology/developmental toxicology. Few have published longitudinal analyses of the complex interplay between the relative effects of prenatal cocaine exposure and perinatal and environmental factors on development. The purpose of this paper was to use structural equation modeling to describe the direct and indirect effects of prenatal drug exposure on developmental outcome from birth to age 6 months. Key variables considered for study include prenatal drug exposure, perinatal medical characteristics, maternal/caregiver/family characteristics, the home environment, and neurobehavioral outcomes. We prospectively enrolled 154 predominantly crack-using women. A priori exclusion criteria included: <18 years old, major illnesses diagnosed prior to pregnancy, chronic use of legal drugs, and any use of illicit drugs other than cocaine and marijuana. From the pool of noncocaine users, 154 subjects were matched to users on pregnancy risk, parity, race, and socioeconomic status. At the end of each trimester, experienced staff conducted private interviews prompting memory of amount and timing of past drug use. Urine specimens were collected at two unanticipated times; positive screens were confirmed by gas chromatography/mass spectroscopy. Measures analyzed include medical (birth) and developmental (birth, 1 month, 6 months) assessments, all performed by blinded evaluators, as well as caregiver characteristics and environmental factors (birth, 1 month). A series of four theoretical models was tested, one for each time point (birth, 1 month, 6 months) and a longitudinal model spanning birth to 6 months. Key findings include direct effects of prenatal cocaine exposure on development at birth in the birth model and on development at birth and 6 months in the longitudinal model. In addition, indirect effects of prenatal cocaine exposure were identified on development at birth, 1 month, and 6 months, mediated through the prenatal use of alcohol and tobacco and the birth head circumference. Implications of these and other findings, including the advantages and limitations of structural equation modeling, are discussed.