This study determines whether a continuing unresolved disorganized response to attachment-related trauma concurrently relates to dissociative symptoms in adolescents in psychiatric treatment.

Method

Participants

We consecutively recruited 145 adolescents to the study upon their admission to 1 of 5 participating treatment centres in 3 Canadian cities, following such severe family disruptions as unwanted or forced separations, loss of parent by death, and abuse by parent. The sites included inpatient and outpatient programs, longer-term residential settings, and day programs. All participants were adolescents between ages 12 and 19 years. We obtained informed written consent to participate from each adolescent, and for participants under age 18 years, we obtained written consent from their parent or guardian. Exclusion criteria included presence of active psychosis or organic brain or central nervous system disorder. We eliminated 12 of the 145 participants from the study (3 who failed to appear for the attachment interview, 5 who gave incomplete interviews, and 4 whose interviews were not recorded for technical reasons), yielding a 92% participation rate. The complete sample comprised 59 females and 74 males who were predominantly white (86.5%); of the sample, 58.3% were age 15 years or under and 41.7% were older than 15 years. At the time of the study, 53% of the participants were in residential treatment. There were no significant differences between males and females in age (c² = 0.17, df 1, P = 0.67) or residential treatment status (c² = 1.05, df 1, P = 0.31).

Instruments

All participants completed a brief demographic questionnaire, the Youth Self-Report (YSR) and the Adult Attachment Interview (AAI; 9) with minor modifications for adolescent participants.

The AAI is a semistructured interview that focuses on attachment relationships and events in early childhood.

The coding of the AAI is a complex task, requiring several steps based on verbatim transcripts of the audiotaped interviews. Each transcript is rated on fourteen 9-point scales. Each subject is then coded for presence or absence of unresolved and disorganized loss or trauma (U_d rating), and for 1 of 3 primary organized attachment patterns: autonomous-secure (F rating), dismissing (D_s rating), or preoccupied-enmeshed (E rating) (10). These classifications parallel the fourfold infant–caregiver strange situation classifications of insecure-disorganized, secure, insecure-avoidant, and insecure-ambivalent, respectively. Transcripts that do not appear to fit one of the primary F, D_s, or E categories, or show a striking mixture of organizations are coded “cannot classify” (CC). Three trained coders, who did not conduct the interviews, coded the transcripts independently. The coders were blind to all other information about the participants until all transcripts had been coded. One coder scored all transcripts; the other 2 coders scored overlapping subsets of the transcripts. All transcripts were coded by at least 2 coders.

Several studies have established the validity and reliability of the AAI. High concordances between parental AAI and infant strange situation classifications have been reported retrospectively (11,12), as well as prospectively (13,14). Test–retest reliabilities of 78% (kappa 0.63), and 90% (kappa 0.79) have been reported by Bukermans-Kranenburg and van IJzendoorn (15) and Benoit and Parker (13), respectively.

In an analysis of interrater reliability for the AAI in this study, the mean concordance rate across all AAI classifications (U_d, F, D_s, E, and CC) for all pairs of raters was 78.6%. The mean kappa for primary classifications (F, D_s, E, CC) is 0.71

The Youth Self-Report (YSR) is part of a family of instruments, including the CBCL and the Teacher Rating Form(TRF), developed and refined by