عوامل پیش بینی کننده نتیجه در یک برنامه گروهی شناختی- رفتاری برای کودکان و نوجوانان مبتلا به اختلال اضطراب اجتماعی

Abstract The present study examined predictors of treatment outcome among children and adolescents with social anxiety disorder (SAD). Seventy-five participants (8–13 years) participated in a 12-session cognitive behavioral group treatment (CBT). Potential predictors were the pre-treatment severity of anxious symptoms assessed from both the child's and parent's perspective as well as depressive symptoms (child report only) and general emotional distress in parent (parent self-report). Furthermore, the relationship between treatment outcome and child's self-reported pre–post changes in self-consciousness and maladaptive anxiety regulation was investigated. Pre-treatment level of social anxiety reported by the child was a significant predictor for outcome, i.e. children with higher levels of social anxiety at pretreatment reported a greater reduction in social anxiety at post-treatment. Reduction in self-consciousness and maladaptive anxiety regulation both predicted reduction in social anxiety, although not independently. The results suggest that tailoring intervention to include strategies for emotion regulation of anxiety may improve treatment outcome.

. Introduction Social anxiety disorder (SAD) is among the most common disorders of childhood and adolescence and is associated with impairment in child development (Costello, Egger, & Angold, 2005). Therefore, the development of effective treatments is necessary. Behavioral and cognitive behavioral treatments are among the most effective interventions for the treatment of anxiety disorders, both in children and adolescents. However, current empirical evidence shows that only half to two-thirds of the children benefit in terms of significant symptom reduction or diagnostic recovery at post-treatment (for reviews see In-Albon and Schneider, 2007, Olatunji et al., 2010 and Silverman et al., 2008). Furthermore, there are hints that socially anxious youth do not benefit as much from CBT in comparison to children and adolescents with other anxiety disorders (e.g., Crawley, Beidas, Benjamin, Martin, & Kendall, 2008). To date, studies examining reasons for the differential outcomes are sparse. Previous research on predictors of treatment outcome focused on pretreatment symptomatology in children and parents as well as on change processes (such as emotion regulation), primarily in children with various anxiety disorders. 1.1. Pretreatment symptomatology in children and parents 1.1.1. Severity of anxiety and comorbidity with depressive symptoms in children High symptom severity of anxiety and large impairment has been found to be associated with high comorbidity rates in children and adolescents suffering from anxiety disorders (e.g., Kendall et al., 2010). Several studies have already examined comorbidity as one potential predictor for poor treatment outcome including children with various principal anxiety disorders, such as separation anxiety disorder, generalized anxiety disorder, specific phobia and SAD. The majority of these studies indicate that comorbid disorders generally have no or only small effects on the treatment outcome (for an overview see Olatunji et al., 2010 and Ollendick et al., 2008). However, studies investigating the impact of depressive symptoms instead of disorders as a possible predictor for treatment outcome have produced contradictory findings. Whereas in some studies high symptom levels of depression and/or comorbid mood disorders have been found to predict poorer treatment outcomes, other studies where not able to replicate these findings (e.g., Berman et al., 2000 and Southam-Gerow et al., 2001). Inconsistent results were also found for studies with samples of children with a primary diagnosis of SAD. Crawley et al. (2008) concluded from their study with SAD children that comorbid mood disorders were likely to be responsible for a worse treatment outcome. In contrast, Alfano et al. (2009) found no moderating effect of child reported depressive symptoms on several outcome measures after treating SAD children with Social Effectiveness Therapy for Children (SET-C). Some of the discrepant results may be due to the different assessment methods for depressive symptoms in children (e.g., questionnaire vs. structured interview) and the severity of the symptoms (meeting the threshold for a clinical diagnosis). 1.1.2. Parental psychopathology Models explaining the development of anxiety disorders emphasize the importance of parental influence (e.g., Rapee & Spence, 2004). Therefore, several studies have investigated the relationship between parental psychopathology and treatment outcome in children suffering from heterogeneous anxiety disorders. However, studies on this topic have yielded mixed results. These vary from parental psychopathology having either a positive (e.g. Legerstee et al., 2008), no (e.g. Victor et al., 2007 and Wood et al., 2006) or a negative effect on treatment outcome in children with anxiety disorders. With respect to anxiety disorders in general, studies thus show inconsistent findings concerning the impact of parental psychopathology on treatment efficacy. Moreover, it remains unclear whether disorder-specific differences could explain the above-mentioned findings, due to a lack of studies focusing on one specific anxiety disorder like SAD for example. In sum, studies with inconsistent results investigating pretreatment symptomatology as a possible predictor for treatment outcome focus primarily on children with various anxiety disorders and not specifically on children with SAD, with the exception of two studies (Alfano et al., 2009 and Crawley et al., 2008). 1.2. Changes in processes during treatment Although research has consequently pointed out the importance of change processes during treatment for outcome measures, these processes are not yet well understood. Only few studies have investigated potential disorder-specific or transdiagnostic processes of change during treatment as potential moderators or mediators for treatment outcome in children suffering from anxiety disorders. In one study, treatment response of children with various anxiety disorders was mediated by changes in anxious self-statements (Kendall & Treadwell, 2007). Alfano et al. (2009) focused on children with SAD only. They found that child-reported changes in loneliness, as one feature of SAD, mediated changes in social anxiety. Among some of the theoretically significant processes for change in SAD symptomatology may be those related to self-consciousness and general maladaptive anxiety regulation. 1.2.1. Self-consciousness and social anxiety Research in adult samples has shown that high levels of self-consciousness or self-focused attention play an important role in the maintenance of social anxiety (Bögels and Mansell, 2004, Clark and Wells, 1995 and George and Stopa, 2008). Furthermore, there is evidence that cognitive behavioral treatment for adults with SAD leads to a significant decrease in self-focused attention and self-focused thoughts, and that these changes are related to lower social anxiety at post-treatment (Hofmann, 2000 and Woody et al., 1997). In contrast to the literature focusing on adults, only few studies have investigated this relation in child or adolescent populations. One study by Higa, Phillips, Chorpita and Daleiden (2008) showed that self-consciousness was significantly correlated with youth self-reported social anxiety and negative affect. Hodson, McManus, Clark and Doll (2008) also found that high socially anxious children reported more self-focused attention than the non-anxious controls. These results provide preliminary evidence that self-consciousness in children may play a similar significant role in SAD as in adults. However, it is currently unknown if self-consciousness changes during treatment and if these potential changes in turn play a significant role in predicting changes in social anxiety. 1.2.2. Maladaptive anxiety regulation strategies A large number of children and adolescents with anxiety disorders are currently not benefiting from available forms of treatment, therefore, several authors suggested that including emotion regulation components could possibly lead to an enhancement of treatment efficacy (e.g., Hannesdottir & Ollendick, 2007). This suggestion is supported by a study in which an emotion regulation treatment component was added to a regular CBT for children with anxiety disorders (Suveg, Kendall, Comer, & Robin, 2006) with generally positive results on the ability to identify emotional states or the understanding of emotion regulation strategies. In a later study, Suveg, Sood, Comer and Kendall (2009) showed that CBT without a supplementary emotion regulation component leads to improvement of worry regulation skills among anxious youths. However, changes in sadness and anger regulation skills were not detected. The change in worry dysregulation predicted change in anxiety scores; change in worry coping, however, did not. In light of these results and the fact that our group treatment generally did not focus on dysfunctional emotion regulation, we investigated only the changes in maladaptive anxiety regulation strategies, as one possible predictor of change in social anxiety after treatment in children with SAD. As described above the limited research on predicting treatment outcome has typically been examined in children and adolescents with heterogeneous anxiety disorders and offers mixed results. Therefore, the purpose of the present study was to specifically examine predictors of outcome in a homogenous sample of children with SAD: (1) the effects of children's pretreatment anxious and depressive symptomatology (2) the effect of parents’ pretreatment psychopathology on changes in children's anxiety and (3) changes in self-consciousness and maladaptive anxiety coping predicting changes in anxiety from pre- to post-treatment. Based on previous findings it was hypothesized that high levels of child or parent reported pretreatment symptom severity of (social) anxiety and child rated depression predicts less change in (social) anxiety from pre- to post-treatment. Furthermore, it was hypothesized that more severe parent psychopathology predicts poorer (social) anxiety change in children. Concerning the change processes during treatment, we hypothesized that potential changes in self-consciousness would predict changes in (social) anxiety in children as it does in adults. In line with the results of Suveg et al. (2009), it was also hypothesized that change in maladaptive anxiety coping would significantly predict changes in (social) anxiety.

5. Conclusions A promising starting point in the investigation of predictors in order to improve treatment outcomes is to better understand underlying change mechanisms during treatment, which may positively or negatively influence treatment outcome. Instead of focusing on parental psychopathology per se, it might be more beneficial to focus on specific parenting processes, which may mediate this association. Corresponding moderating or mediating variables must be taken into account, which itself could be a starting point for supplementary interventions. A better understanding or even a sound theoretical model about what causes change in social anxiety in children may help considerably in identifying predictors of change.