ATTACHMENT STYLE AS A RISK FACTOR FOR SUICIDE-RELATED BEHAVIORS IN YOUTH
Suicide-related behaviors are of great importance, particularly among adolescents, where their prevalence is notably high. Specifically, suicide is a leading cause of death among adolescents (Xu, Kochanek, Murphy, & Tejada-Vera, 2010) and self-injury affects 13 - 23 % of adolescents in the general adolescent population and 40 - 60% of those in clinical settings (Darche, 1990; DiClemente, Ponton, & Hartley, 1991; Jacobson & Gould, 2007). Though many risk factors of suicide-related behavior have been identified, the most influential theoretical models of suicide attempt [i.e. Joiner’s (2005) interpersonal-psychological theory] and self-injury [i.e. Nock’s (2008) social signaling hypothesis] emphasize the role of an individual’s interpersonal environment and connectedness in suicide-related behaviors. Given the known link between interpersonal functioning and attachment security (see Berlin, Cassidy, & Appleyard, 2008), a focus on examining attachment style as risk factor for suicidal behaviors seems warranted. Few studies, however, have explored this question in adolescents and those that have often suffer from important limitations, leaving this area largely unstudied. Against this background, and addressing several limitations of prior research, the aims of the present study were to (a) determine which attachment styles are associated with suicide-related behaviors and (b) test the mediational role of social cognition in the relation between attachment and suicide-related behaviors. Specifically, 194 adolescents were recruited from an inpatient unit and assigned to one of three attachment styles (dismissing, secure, and preoccupied) and one sub-classification (disorganized), a total of four groups, based on an interview assessing attachment style. First, the link between these attachment classifications and suicide-related behaviors was evaluated with regression analyses, controlling for demographic and psychopathology variables. Specifically, we explored which attachment styles (dismissing, secure, preoccupied, and disorganized) were associated with several categories of suicide-related behavior, including (a) suicidal ideation, (b) a single suicide attempt, (c) multiple suicide attempts, and (d) self-harm. Second, we sought to determine to what extent social cognition mediates the relation between attachment style and suicide-related behaviors and thoughts. While the findings of this study did not identify any significant relations between attachment classification, suicide-related thoughts and behaviors, or social cognitive style, factor analyses were used to explore attachment from a dimensional perspective. One of the three attachment factors identified through factor analyses, the preoccupied attachment factor, was associated with the presence of multiple attempts such that those with multiple attempts displayed less preoccupied anger than those with only one attempt. This finding and the absence of group differences when using a categorical measure of attachment style were discussed.