Attachment-Based Representations of Infants in Utero Moderate the Impact of Depression on Child Maltreatment Risk in Expectant Mothers

Purpose: Maternal depression is a significant risk factor for child maltreatment (Hien at al., 2010). However, little research has examined potential factors that moderate this association. Attachment theory is one framework that may guide our identification of potential moderators. Mothers’ attachment-based representations (perceptions) of infants determine parenting behaviors and the quality of the parent child relationship (Benoit, et. al., 1997). Representations include the ways that mothers interpret babies’ behaviors and the primary emotional states that color their perceptions of the infant (Fonagy, 2009). With this framework in mind, the purpose of this study was to examine how maternal representations moderate the influence of depression on mothers’ risk for child maltreatment in expectant mothers.

Methods: Data were collected from 51 low income urban-dwelling expectant mothers in Detroit.  Mothers were interviewed during the third trimester of pregnancy. The Working Model of the Child Interview (Zeanah, et. al., 2005) was used to assess mother’s developing representations of their infant in utero.  The Brief Child Abuse Potential Scale (Ondersma, Chaffin, Mullins, & LeBreqton, 2005) assessed mother’s risk for maltreating her child, and the Edinburgh Perinatal Depression Scale (Cox & Holden, 2003) was used to assess mother’s current level of depression.  Using OLS linear regression and Bootstrapping in PROCESS macro for SPSS, we examined the impact of perinatal depression on expectant mothers’ potential for child maltreatment, and the role of attachment-based representations of the infant in moderating the impact of perinatal depression on expectant mothers’ potential for child maltreatment.

Results:  Perinatal depression was significantly associated (β=.482, p<.01) with mother’s risk for child maltreatment, in that higher depression scores were positively associated with higher scores on the potential for child abuse scale.  Qualities of mothers’ internal attachment-based representations of infants in utero were also significantly associated with child maltreatment potential; greater sensitivity (β=-.339, p<.01), acceptance of the parenting role (β=-.251, p<.05), and reflective functioning (β =-.272, p<.05) were associated with lower risk for child maltreatment, while greater resentment (β=.328, p<.05), parental helplessness (β=.330, p<.01), and perception of infant difficulty (β=.343, p<.01) were associated with greater child abuse potential.  Attachment-based representations moderated the impact of depression on risk for child maltreatment in that positive representations (sensitivity t=-2.41, p<.05; acceptance t=-3.01, p<.01; reflective functioning t=-2.44, p<.05) decreased the impact of maternal depression on child maltreatment potential, while negative representations (helplessness t=5.07, p<.01; infant difficulty t=3.10, p<.01; resentment t=2.98, p<.01) increased the impact of maternal depression on child abuse potential. 

Conclusion:  This study advances our understanding of factors that moderate the relationship between maternal depression and child maltreatment. During assessment and intervention, social workers regularly screen for risk factors such as maternal depression. However, assessing maternal representations and intervening with the mother-infant relationship may be beneficial when working with mothers experiencing depression. Future research should examine the efficacy of attachment-based interventions over time with mothers experiencing depression.