[Infants' attachment security in a vulnerable French sample]

Attachment is a long lasting emotional link established between infants and their caregivers. The quality of early relationships allows infants to safely explore their environment and contribute to the establishment of a broad range of social skills. Several intervention programs targeting infant attachment have been implemented in different contexts, showing diverse degrees of efficacy.

The present paper describes, for the first time, children's attachment quality distributions in a French multi-risk population, with a preventive intervention, usual or reinforced.

In the CAPEDP study (Parenting and Attachment in Early Childhood: reducing mental health disorder risks and promoting resilience), a sub-sample of 117 women was recruited to assess the effects of this home-visiting program on children's attachment security. With that intent, the Strange Situation Paradigm was used when infants were between 12 and 16 months of age.

In the intervention group, 63% (n=41) of the infants were coded as secure, while 15% (n=10) of them were coded as insecure-avoidant and 22% (n=14) as insecure-ambivalent/resistant. 56% (n=29) of control group infants (usual care) were coded as secure, while 27% (n=14) were coded as insecure-avoidant and 17% (n=9) as insecure-ambivalent/resistant. Even if the percentage of children with a secure attachment in the reinforced intervention group was higher than that of the control group, this difference did not reach the threshold of significance [Chi² (2)=2.40, P=0.30].

Intervention group distributions were closer to normative samples, and these distributions show the clinical impact of our program. In general, preventive interventions focused on attachment quality have moderate effects but, in our case, several factors might have contributed to lower the statistical impact of the program. Firstly, the control group cannot be considered has having received zero intervention for two reasons: (a) the French usual perinatal health system (Maternal and Infant Protection System) is particularly generous and (b) the effect of this usual system might have been increased by the project intensive assessment protocol (6 visits during 28 months). Secondly, it is possible that the full effect of the intervention had not yet been detected because, when a child's attachment was assessed, only two thirds of the intervention visits had been performed (29 of 44 visits). A "sleeper effect" is still possible: we hope that a more clear result will be seen when children are assessed again, at 48 months, in our follow-up study (CAPEDP-A II). By clarifying the mechanisms involved in the development of a secure attachment, our study aims to contribute and refine the development of early preventive intervention strategies in high perinatal and psychosocial vulnerability contexts.