Prenatal Reflective Functioning Related to Childhood Attachment and One’s Own Parenting: Interrelations and Effects on Mother-Infant Interaction Among Drug-Abusing Mothers

Introduction: Reflective functioning (RF) indicates an ability to understand one’s own and other people’s behavior based on underlying mental states, which is fundamental for adaptive parenting. Maternal prenatal drug-abuse (DA) implies severe, cumulative biological and psychosocial risks for child development and parenting (Conners et al., 2004). Difficulties in RF are an integral part of parenting difficulties associated with drug-abuse, and interventions targeting RF are known to improve parenting and child well-being (Suchman et al., 2010). RF can be measured either based on childhood attachment experiences with parents, or related to one’s own parenting, and both types of prenatal RF may predict future parenting. However, their inter-relations and differential effects on parenting are not well-known. This study examines, first, the interrelations between mother’s prenatal attachment-related RF and pre- and postnatal parenting-related RF, among treatment-enrolled DA mothers. Second, we examine, how these two types of RF are associated with mother-infant interaction quality at 4 months and with change in interaction quality from 4 to 12 months.

Methods: The sample comprised 29 mother-infant dyads, receiving psychodynamic parent-infant psychotherapy in outpatient treatment units offering integrative parenting and substance use treatments. Mother’s attachment-related RF was measured with Adult Attachment Interview (AAI-RF) during pregnancy, and parenting-related RF with Pregnancy Interview (PI-RF) during pregnancy and Parent Development Interview (PDI-RF) at child age of 4 months. Mother-infant interaction quality was measured in all dyads with Emotional Availability Scales at 4 and 12 months.

Results: AAI-RF was highly correlated with parenting-RF both pre- and postnatally (r=.59 and r=.51). Hierarchical regressions (covaried with marital status and economical problems) showed that lower PI-RF predicted lower maternal sensitivity (β=.70, p <.01), structuring (β=.80, p <.01), child responsiveness (β=.74, p <.01), and child involvement (β=.75, p <.01), at 4 months, but more positive change from 4 to 12 months in sensitivity (β =-.56. p <.05), structuring (β =-.72. p <.01), child responsiveness (β =-.69. p <.01) and child involvement (β =-.65. p <.05) Higher prenatal AAI-RF predicted more positive change from 4 to 12 months especially related to intrusiveness (β=.53, p <.05).

Conclusion: Our results suggest that AAI-RF and parenting-RF are related, but only partially overlapping constructs. Prenatal parenting-RF may be more predictive of dyadic interaction and easier and especially vital to target in parenting interventions, whereas AAI-RF may indicate more general vulnerabilities in maternal emotion and behavior regulation.