Stability and change in attachment security across adolescence

This study examined both continuity and familial, intrapsychic, and environmental predictors of change in adolescent attachment security across a 2-year period from middle to late adolescence. Assessments included the Adult Attachment Interview, observed mother–adolescent interactions, test-based data, and adolescent selfreports obtained from an ethnically and socioeconomically diverse sample of moderately at-risk adolescents interviewed at ages 16 and 18. Substantial stability in security was identified. Beyond this stability, however, relative declines in attachment security were predicted by adolescents’ enmeshed, overpersonalizing behavior with their mothers; depressive symptoms; and poverty status. Results suggest that although security may trend upward for nonstressed adolescents, stressors that overwhelm the capacity for affect regulation and that are not easily assuaged by parents predict relative declines in security over time. Given growing evidence of the importance of attachment security in adolescence and adulthood (Allen, Moore, Kuperminc, & Bell, 1998; Furman & Flanagan, 1997; Kobak, Sudler, & Gamble, 1991; van IJzendoorn & Bakermans-Kranenburg, 1996), one of the central questions in attachment research becomes understanding the sources of continuity and discontinuity in attachment security during this period. Bowlby (1980) described the attachment system as being likely to display significant stability over time but also as open to modification given certain types of environmental input. This study examined both continuity and predictors of change in levels of attachment security from middle to late adolescence. Middle to late adolescence appears as a particularly important and promising time to examine continuities and predictors of change in attachment security. The cognitive and relational transformations of adolescence have the potential to influence significantly adolescents’ developing states of mind with respect to attachment, possibly leading to significant discontinuities over time in these states of mind (Allen & Land, 1999). In adolescence, only two preliminary, small-sample studies (e.g., each with fewer than 40 participants) with low-risk samples have assessed attachment stability in this period,