20390
Friends for Foes: How Friendship May Buffer the Effects of Victimization on Depression in Adolescents with High Functioning Autism

Friday, May 15, 2015: 5:30 PM-7:00 PM
Imperial Ballroom (Grand America Hotel)
M. Jankowski1, L. Sperle2, B. G. Davidson3 and A. R. Neal-Beevers4, (1)Psychology, University of Texas at Austin, Austin, TX, (2)University of Pittsburgh, Pittsburgh, PA, (3)Dept of Psychology, University of Texas at Austin, Austin, TX, (4)Department of Psychology, University of Texas at Austin, Austin, TX
Background: Adolescents with autism spectrum disorders are more likely to be victims of bullying than typically developing individuals (TD; Wainscot et al., 2008).  Research has highlighted the importance of friendships in preventing an escalating cycle of peer victimization in TD children (e.g., Ernest, et al., 1999).  However, these findings may not generalize to children with high-functioning autism (HFA) because of unique deficits in social abilities. 

Objectives: This study sought to examine victimization and depression differences, as well as investigate variations in the presence and quality of friendships in HFA and TD adolescents.  

Methods: Twenty HFA and 36 TD participants 10-21 years of age were matched on chronological and mental age. The ADOS was used to confirm diagnosis. The ADOS Friends and Marriage questions were transcribed and analyzed using thematic analysis (Braun & Clarke, 2008) and focused on three main themes: Poor Quality Friendships, Limited Social Network, and Difficulty Understanding Friendship. A Negative Friendship Total was also calculated.  Participants completed the Social Experiences Questionnaire (Crick & Grotpeter, 1996) yielding Overt Victimization (Overt_Vict), Relational Victimization (Relational_Vict), and Prosocial Support (Prosocial_Sup) scores.  Depression was assessed using the Child Depression Inventory (Kovac, 1992) or the Center for Epidemiological Studies-Depression inventory (CESD; Radloff, 1977) depending on chronological age. Depression scores underwent z-score transformations to ensure that they were on the same scale.    

Results: Separate hierarchical regressions predicting depression revealed: (1) a significant main effect for Relational_Vict and a significant interaction effect for Relational_Vict by Prosocial_Sup  and (2) significant main effect for Overt_Vict but no significant Overt_Vict by Prosocial_Sup interaction.  Main effects for Prosocial_Sup were not observed in either regression model (see Table 1)[1].  To understand the influence of friendship on Prosocial_Sup, additional regression analyses were conducted.  Results suggested that higher Negative Friendship Totals significantly predicted the lower Prosocial_Sup (see Table 2).  Regression models predicting Prosocial_Sup from Limited Social Network, Understanding Friendship, and Poor Quality Friendships were not significant.

Conclusions:  These finding indicate that the amount of prosocial support provided by one’s peers can buffer the negative effects of social exclusion and alienation (relational victimization) on depression symptoms in adolescents.  Furthermore, the presence and quality of friendship significantly predicted the amount of prosocial support from peers.  These findings were similar across groups, however it should be noted that HFA adolescents had poorer quality friendships and more limited social networks than their TD peers.  Thus, it may be that creating and maintaining meaningful friendships can reduce internalizing symptoms in victimized adolescents, regardless of diagnostic group.  However, because HFA adolescents have poorer friendship quality, receive less prosocial support and are more victimized by their peers than TD adolescents, it may be particularly important to understand how friendship buffers the negative effects of victimization to inform research and clinical practice for families affected by autism.


[1] Regression models examining group effects were also conducted, but group main effects and interactions were not significant in any model, and these predictors were removed from the models.