Prenatal cocaine exposure has been associated with alterations in neonatal behavior and more recently a dose–response relationship has been identified. However, few data are available to address the long-term behavioral effects of prenatal exposures in humans. The specific aim of this report is to evaluate the school-age behavior of children prenatally exposed to cocaine.


All black non–human immunodeficiency virus–positive participants in a larger pregnancy outcomes study who delivered singleton live born infants between September 1, 1989 and August 31, 1991 were eligible for study participation. Staff members of the larger study extensively screened study participants during pregnancy for cocaine, alcohol, cigarettes, and other illicit drugs. Prenatal drug exposure was defined by maternal history elicited by structured interviews with maternal and infant drug testing as clinically indicated. Cocaine exposure was considered positive if either history or laboratory results were positive. Six years later, 665 families were contacted; 94% agreed to participate. The child, primary caretaker (parent), and, when available, the biologic mothers were tested in our research facilities. Permission was elicited to obtain blinded teacher assessments of child behavior with the Achenbach Teacher's Report Form (TRF). Drug use since the child's birth was assessed by trained researchers using a structured interview.


Complete laboratory and teacher data were available for 499 parent–child dyads, with a final sample size for all analyses of 471 (201 cocaine-exposed) after the elimination of mentally retarded subjects. A comparison of relative Externalizing (Aggressive, Delinquent) to Internalizing (Anxious/Depressed, Withdrawn, Somatic Complaints) behaviors of the offspring was computed for the TRF by taking the difference between the 2 subscales to create an Externalizing–Internalizing Difference (T. M. Achenbach, personal communication, 1998). Univariate comparisons revealed that boys were significantly more likely to score in the clinically significant range on total TRF, Externalizing–Internalizing, and Aggressive Behaviors than were girls. Children prenatally exposed to cocaine had higher Externalizing–Internalizing Differences compared with controls but did not have significantly higher scores on any of the other TRF variables. Additionally, boys prenatally exposed to cocaine were twice as likely as controls to have clinically significant scores for externalizing (25% vs 13%) and delinquent behavior (22% vs 11%). Gender, prenatal exposures (cocaine and alcohol), and postnatal risk factors (custody changes, current drug use in the home, child's report of violence exposure) were all related to problem behaviors. Even after controlling for gender, other prenatal substance exposures, and home environment variables, cocaine-exposed children had higher Externalizing–Internalizing Difference scores. Prenatal exposure to alcohol was associated with higher total score, increased attention problems, and more delinquent behaviors. Prenatal exposure to cigarettes was not significantly related to the total TRF score or any of the TRF subscales. Postnatal factors associated with problem behaviors included both changes in custody status and current drug use in the home. Change in custody status of the cocaine-exposed children, but not of the controls, was related to higher total scores on the TRF and more externalizing and aggressive behaviors. Current drug use in the home was associated with higher scores on the externalizing and aggressive subscales.


Results of this study suggest gender-specific behavioral effects related to prenatal cocaine exposure. Prenatal alcohol exposure also had a significant impact on the TRF. Postnatal exposures, including current drug use in the home and the child's report of violence exposure, had an independent effect on teacher-assessed child behavioral problems. Furthermore, among the children prenatally exposed to cocaine, change in the child's custody status was a significant predictor of TRF scores. It remains possible that other unmeasured postnatal characteristics of the cocaine-using household may play important roles in teacher-assessed child behavior.

You do not currently have access to this content.