Individual Differences in Infant-Mother Attachment at Twelve and Eighteen Months: Stability and Change in Families under Stress

Abstract

100 economically disadvantaged mothers and their infants were observed in the Ainsworth and Wittig "strange situation" at 12 and 18 months. Infants were classified as secure, anxiously attached/avoidant, or anxiously attached/resistant. In addition, mothers reported occurrence of stressful events related to the stability of the caretaking environment during the 12--18 month period by completing a 44-item checklist concerning work, finances, family, neighbors, health, etc. 62 infants were assigned to the same attachment classification at both 12 and 18 months (p less than .01). Despite this stability, significantly more infants changes classification than in a recent study of stable middle-class families. With the present sample, anxious attachment was associated with less stable caretaking environments than secure attachment; change from secure to anxious attachment was associated with higher stressful-event scores than stable secure attachment.

Attachment in Chronically Underweight Young Children

“…In contrast to the low proportion of secure attachments found in the group of infants presenting chronic subclinical weight deficits, the distribution of patterns of attachment in the comparison group of adequate weight-for-age children from a similar socioeconomic environment is consistent with the proportion of secure attachments found in other low-income samples in North America (Egeland & Sroufe, 1981;Lyons-Ruth, Gonnell, Grunebaum, Botein, & ZoU, 1984;Vaughn, Egeland, Sroufe, & Waters, 1979). A decrease from the 66% of secure attachments found in the normative Baltimore sample of middle-class children (Ainsworth et al, 1978) to 50% in the present sample is also consistent with the proportion of secure attachments reported in these low-income samples.…”

Change in Atypical Maternal Behavior Predicts Change in Attachment Disorganization From 12 to 24 Months in a High‐Risk Sample

“…The quality of the attachment relationship may be affected not only by developmental shifts but also by the broader context of the relationship. There is evidence that attachment relationships are less stable in higher risk groups, such as socially and economically disadvantaged and maltreating populations, than they are in low-risk, middle-class samples (e.g., Bel-sky, Campbell, Cohn, & Moore, 1996;Lamb, Thompson, Gardner, & Charnov, 1985;Lyons-Ruth, Repacholi, McLeod, & Silva, 1991;Spieker & Booth, 1985;Teti, Sakin, Kucera, Corns, & Eiden, 1996;Thompson, Lamb, & Estes, 1992;Vaughn, Egeland, Sroufe, & Waters, 1979;Vondra, Shaw, Swearingen, Cohen, & Owens, 2001). These populations also feature a higher occurrence of disorganized attachment.…”

Attachment Status in Children Prenatally Exposed to Cocaine and Other Substances

“…One final empirical point is that attachment status is variable over the first years of life. Although some early evidence suggested that attachment classification was stable, at least between 12 and 18 months of age (Waters, 1978), other evidence suggested that instability may be common in samples where significant contextual stresses occur (Vaughn, Egeland, Sroufe, & Waters, 1979). Following this early work, there has been little examination of stability of attachment classifications, with the evidence centering on the conclusions of modest stability in secure versus insecure (National Institute of Child Health and Human Development [NICHD] Early Child Care Research Network, 2001) and disorganized versus not disorganized (van IJzendoorn, Schuengel, & Bakermans-Kranenburg, 1999).…”

“…Attachment theory would predict more substantial stability than observed in these studies with samples that sum to more than 1,600 cases. Both of these studies examined samples where many families were distressed, and stability may be expected to be lower than in more affluent families who have fewer contextual stressors (van IJzendoorn et al, 1999;Vaughn et al, 1979). Still, the observed levels of stability are low in the context of existing theory about attachment (van IJzendoorn et al, 1999).…”