Attachment behavior of mothers with their preterm infants in the neonatal intensive care unit

Bonnie Jean Behee-Semler, Marquette University

Abstract

Studies of early maternal-infant interaction have been helpful in describing the process of bonding. Characteristics of behavior of mothers has also been examined as predictors of the infant's attachment to his mother It is also recognized that this earliest of all human relationships affects the socioemotional development of the infant. Recently published research examines caregiving as a behavioral system analogous to the attachment behavioral system. Maternal-infant interactions studies provide us with knowledge essential for early identification of problems in the attachment or caregiving systems. To date, subjects for study have been mothers of full term infants and have taken place in the hospital or at home. Mothers of premature infants have also been studied in the home. The findings of these studies are not generalizable to the Neonatal Intensive Care Unit (NICU) as a setting because of the unique characteristics of this environment. Studies of maternal behavior in the NICU have been sparse. In the present study, behavior of mothers in the NICU was described and examined for change over time. The subjects were 21 mothers whose infants were hospitalized in the NICU. Mothers, holding their infants, were videotaped three times behavior. Nurses also rated maternal behavior using a different molar rating scale designed for use by NICU nurses. Results indicate that maternal behavior in the NICU falls in the category of mild to moderate concern for attachment difficulties as measured by reliable raters. Most of the 6 dependent measures were stable over time. The exception was the nurses rating which demonstrated significant difference between Observation 1 and Observation 2. Secondary analysis revealed that younger maternal age and lower SES contributed to this change. The nurses measures were significantly correlated with the reliable rater scores but nurses scores were relatively higher. It is speculated that the higher nurses scores were related the absence of maternal affect items on the nurses scale and heavy weighting of maternal affect on the scale scored by reliable raters.

This paper has been withdrawn.