The first 10,000 Adult Attachment Interviews: distributions of adult attachment representations in clinical and non-clinical groups

Abstract

More than 200 adult attachment representation studies, presenting more than 10,500 Adult Attachment Interview (AAI; George, Kaplan, & Main, 1985) classifications, have been conducted in the past 25 years. In a series of analyses on the distributions of the AAI classifications in various cultural and age groups, fathers, and high-risk and clinical samples, we used the distribution of the combined samples of North American non-clinical mothers (23% dismissing, 58% secure, 19% preoccupied attachment representations, and 18% additionally coded for unresolved loss or other trauma) to examine deviations from this normative pattern, through multinomial tests and analyses of correspondence. The analyses were restricted to AAI classifications coded according to the Main, Goldwyn, and Hesse (2003) system. We did not find gender differences in the use of dismissing versus preoccupied attachment strategies, and the AAI distributions were largely independent of language and country of origin. Clinical subjects showed more insecure and unresolved attachment representations than the norm groups. Disorders with an internalizing dimension (e.g., borderline personality disorders) were associated with more preoccupied and unresolved attachments, whereas disorders with an externalizing dimension (e.g., antisocial personality disorders) displayed more dismissing as well as preoccupied attachments. Depressive symptomatology was associated with insecurity but not with unresolved loss or trauma, whereas adults with abuse experiences or PTSD were mostly unresolved. In order to find more reliable associations with clinical symptoms and disorders, future AAI studies may make more fruitful use of continuous AAI scales in addition to the conventionally used categorical classifications.