Project 2 (Human) - Prenatal Cocaine and Neurobehavioral Factors Underlying Mother-Infant Interaction

Principal Investigators: Karen Grewen, PhD; Kathleen Light, PhD

Project 2 (Human) - Schematic


Cocaine abuse during pregnancy has been linked to problems with maternal-infant bonding and attachment. At its most extreme, maternal cocaine use (which has increased dramatically in women of child-bearing age in the past decade) is related to cases of child abuse or neglect. This vulnerability may derive in part from the physiological effects of the prenatal cocaine exposure on the infant, whose development and behavior may differ from other infants (greater high-pitched and inconsolable crying, excessive startle reactions and irritability, and less frequent or delayed onset of eye contact, smiling, cooing and other social behaviors). Also, the mother’s responses to her infant’s cries and other signals and to holding and interacting with the infant may be altered. Studies have found that women who had used cocaine during pregnancy have more difficulty interacting with their infants and express subtle hostility (insensitivity and irritability) during feeding and play periods. These mothers perceive infant cries as less arousing and requiring less maternal intervention. Furthermore, these mothers tend to pick up and hold their infants less often than other mothers, even though their infants cry more. Also, separation from the mother followed by reunion does not elicit normal emotional behavior in the cocaine-exposed child. These complex alterations in early mother-infant behaviors parallel those seen in animal models, which found that exposure to cocaine in pregnant rats decreases maternal behavior and increases the dams’ inappropriate aggressive behavior.

As in the animal model employed by Johns et al, our initial translational findings suggest that oxytocin (OT, sometimes called the “bonding hormone”) is a partial mediator of cocaine’s effects on mother-infant interaction. We found that mothers with prenatal cocaine exposure (PCE) had lower plasma levels of OT than control mothers with no pre- or postnatal drug history (CTL), and showed decreases rather than increases in OT in response to a speech stressor that immediately followed baby holding. During a normal day at home, the PCE mothers tended to hold their babies less often than the CTL mothers; also, when the CTL mothers held their babies at home, their blood pressure (BP) levels fell, indicating that baby holding was associated with greater relaxation and decreased stress. The PCE mothers, in contrast, had no beneficial reduction in BP associated with baby holding at home, and during videotaped mother-infant interactions, they showed an emotional mismatch (laughing more while their infants cried more).

The proposed studies will extend these findings in several ways. First, it is important to study the effect of prenatal cocaine exposure on the neonatal brain, including gray and white matter and ventricular volumes. Second, it is important to determine whether PCE and CTL infants differ in the quality of their attention-eliciting signals (infant cry characteristics, behavior during mother-infant interactions), and whether these behaviors are associated with differences in brain structure and maturation. Third, we will study whether PCE mothers differ from CTL mothers in their subjective responses to infant cries, face-to-face interactions, and other normal stimuli for maternal behavior, while examining whether OT activity in response to these stimuli is also deficient. We will also assess whether maternal OT response may be correlated with maternal behavioral responses, including frequency of picking up/holding baby at home, maternal ratings of mood and behavioral inclinations after blinded ratings of touch and emotional quality and co-regulation of behavior during videotaped mother-infant interaction. Finally, building on assay methods recently developed by our co-investigator, Sue Carter, a pilot feasibility study is proposed to examine the potential utility of non-invasive methods of assessing OT and vasopressin (VP) activity from urine and saliva samples obtained from both mothers and infants.

Key Personnel:


Principal Investigators:

Grewen, Karen, PhD UNC-Chapel Hill Principal Investigator
Light, Kathleen, PhD University of Utah Principal Investigator




Carter, Sue, PhD University of Illinois-Chicago Co-Investigator
Gerig, Guido, PhD University of Utah Co-Investigator
Gilmore, John, MD UNC-Chapel Hill Co-Investigator
Hindelighter, Alan, MD UNC-Chapel Hill Co-Investigator
Lin, Weili, PhD UNC-Chapel Hill Co-Investigator
Zeskind, Philip (Sandy), PhD UNC-Chapel Hill Co-Investigator




Das Eiden, Rina, PhD SUNY-Buffalo, Research Institute on Addictions Consultant
Garbutt, James (J.C.), MD UNC-Chapel Hill Consultant
Meltzer-Brody, Samantha, MD UNC-Chapel Hill Consultant
Weschberg, Wendee, PhD Research Triangle Institute International Consultant

>> Click here to return to Projects & Cores index.