تصویر از خود فراموش نشدنی؟ نقش منفی تصویر از خود در اختلال اضطراب اجتماعی نوجوانان

Abstract Background and objectives Negative distorted self-images (NSI) allegedly maintain social anxiety in adults suffering from social anxiety disorder (SAD). These NSI are activated in feared social situations and are often linked to past socially traumatic events. However, because empirical evidence on the presence and characteristics of such NSI in adolescents suffering from SAD is limited, the aim of the present study is to examine the nature of NSI in adolescent SAD patients. Methods Using a semi-structured interview, 31 adolescents with a primary diagnosis of SAD and 31 healthy adolescents (HA) who were matched for age and gender, completed a questionnaire set assessing the characteristics of NSI, social anxiety and depression. Results Relative to the HA-group, those suffering from SAD reported experiencing NSI significantly more frequently, more vividly, and with greater distress. No significant differences between the groups emerged regarding a link between the NSI and an autobiographical event. However, NSI were reported as more often having an observer-perspective in the SAD as compared to the HA-group. Hierarchical regression analysis revealed that certain characteristics of the NSI predict social anxiety beyond the influence of depression in adolescents with SAD. Discussion NSI seem to be an important feature of adolescent SAD and phenomenological comparable to NSI in adults suffering from SAD. Conclusions Specific interventions aiming to correct NSI, which have proven to be highly effective in adults, should be developmentally-adapted and evaluated in future studies.

. Introduction According to cognitive models of social anxiety disorder (SAD, e.g., Clark & Wells, 1995), recurrent negative distorted self-images (NSI) that display one's public self, occur in socially threatening situations. For example, a patient with a fear of blushing may imagine herself standing in front of others with a face so deeply red that it glows. Besides representing the visible public self or visible aspects of anxiety, NSI might also comprise the self as odd or weird (Stopa, Brown, & Hirsch, 2012). NSI seem to be a key component in maintaining SAD. They have been shown to stimulate socially phobic individuals to use safety behaviours in order to prevent the feared social mishap (Wild, Hackmann, & Clark, 2008). An example of these safety behaviours would be using make-up to hide blushing. The use of safety strategies is in turn problematic for several reasons. They may cause the patient to appear strange or unfriendly therefore making a bad impression on others (e.g., Clark, 2001). NSI often lead to the inaccurate conclusion that others might see oneself in the same way, thereby leading the person to neglect disconfirmatory environmental evidence (Alfano, Beidel, & Turner, 2008; Hignett & Cartwright-Hatton, 2008). The content of NSI seems to be strongly related to fears such as being criticized or experiencing anxiety symptoms that are obvious to others (Hackmann, Clark, & McManus, 2000; Wild et al., 2008). Accordingly, NSI have a direct negative impact on the feelings and social performance of patients with SAD (Hirsch, Clark, Matthews, & Williams, 2003; Vassilopoulos, 2005). Hackmann et al. (2000) demonstrated that NSI are linked to early socially traumatizing experiences in childhood in socially anxious individuals. However, in a recent study by Moscovitch, Gavric, Merrifield, Bielak, and Moscovitch (2012), such autobiographically related events were reported in high as well as low socially anxious participants, indicating that this link is not in fact a specific feature of NSI in SAD. NSI are often seen from an observer-perspective (seeing oneself from the outside), in contrast to a field-perspective and patients with SAD are flooded by them in social situations (Hackmann, Surawy, & Clark, 1998; Hackmann et al., 2000; Wells, Clark, & Ahmad, 1998; but see Moscovitch et al., 2012). Moreover, NSI seem to be experienced entailing visual components, bodily sensations and sounds (Hackmann et al., 2000). Makkar and Grisham (2011) found that negative self-imagery was associated with higher anxiety levels, higher self-focused attention, more negative self-cognitions and more post-event processing in high socially anxious individuals, compared to visualizing a control image. To date, there are several studies examining the role of NSI in adults suffering from SAD or high social anxiety (e.g., Hirsch, Clark, & Mathews, 2006; Hirsch et al., 2003; Moscovitch et al., 2012; Vassilopoulos, 2005; Wild, Hackmann, & Clark, 2007). In addition, highly effective imagery-based interventions for socially phobic adults have been successfully developed (Hackmann Bennett-Levy, & Holmes 2011; Nilsson, Lundh, & Viborg, 2012; Stopa, 2009). Since the onset of SAD is often apparently in adolescence (Wittchen & Fehm, 2003) a further examination of NSI among adolescents might open up new perspectives for treating SAD. However, only few studies have investigated NSI in adolescent SAD (Alfano et al., 2008; Parr & Cartwright-Hatton, 2009). Alfano et al. (2008) used an experimental design to explore the relationship between NSI and social anxiety in three groups of 12–16 year-old individuals. 42 non-anxious adolescents were asked to perform two social tasks, with half being instructed to hold a negative image in mind (IMAG-group). The other half was not instructed to use self-imagery (NO-IMAG group). It was hypothesized that the IMAG group would display elevated levels of anxiety and specific negative thoughts, as well as poorer self- and observer-rated performance, compared to the NO-IMAG group. Thus, the resulting pattern in the IMAG group was expected to be similar to the third group consisting of 21 adolescents with SAD, who received no imagery-instruction. In order to assess thoughts before, during, and after the tasks, a mediated-video recall technique was used. Individuals watched the videotape of their tasks and were asked to recall their cognitions at four points in time during the situations. Contrary to the assumptions, no elevated levels of anxiety or differences in observer-rated anxiety or performance occurred in the IMAG-group, compared to the NO-IAMG group. The SAD group was instead found to be significantly more anxious during both tasks and to perform less well in the role-play, compared to both non-anxious groups. Neither in the IMAG-Group nor in the SAD-group, negative on-task thoughts were detected, which was at odds with what had been hypothesized. The authors concluded that NSI might not be as present in adolescent SAD as studies on adults suggest. Unfortunately, the total level of NSI actually occurring in the control and SAD group during the interaction tasks was not assessed, so this conclusion needs to be interpreted with caution. Kley, Heinrichs, and Tuschen-Caffier (2006) manipulated positive and negative self-images in 18 socially-anxious children (Mage = 10.6 years old), using a design similar to that by Hirsch et al. (2003). The children were instructed to listen to two stories and then to retell them once, while holding a NSI in mind and once with a positive self-image in mind, while being videotaped. Two-thirds of the children reported initially having NSI, but none was described from an observer-perspective. Additionally, visualizing a NSI in mind was associated with observer-ratings of poorer performance and more negative on-task thoughts, compared to the positive-image condition. The overall ratings of visibility of anxiety symptoms were low in both conditions and did not differ significantly. Kley et al. (2006) questioned whether images are really seen from an observer-perspective in children. However, data on socially anxious adolescents support the existence of observer-perspective images ( Hignett & Cartwright-Hatton, 2008). In this study, 12–18 year-olds had to complete an anxiety-provoking social task. Next, they were asked to create a mental image of their feelings during the task, as well as report the imagery perspective they adopted. The study revealed that an increase in social anxiety was accompanied with a shift towards an observer-perspective. One common feature of the studies on NSI in children and adolescents is that they used imagery-manipulation techniques, such as creating a mental image, while completing an experimental procedure to assess the influence on behaviour or specific cognitions. However, to the best of our knowledge, there is no study on the nature of NSI in childhood in a clinical treatment-seeking SAD sample. Thus, the natural features of NSI within this age range remain unknown. In general, empirical evidence suggests that cognitive biases found in social anxiety in adults might also be applicable to younger populations (Hodson, McManus, Clark, & Doll, 2008; Kley, Tuschen-Caffier, & Heinrichs, 2011) and that the impact of developmental factors such as age or cognitive development on various anxiety phenomena is marginal (Broeren & Muris, 2009). However, to our knowledge, there is a lack of literature regarding developmental aspects regarding the phenomenon of self-images in adolescents. Thus, the aim of the present study was to explore the nature of NSI in adolescent SAD. Using a cross-sectional design, several characteristics of NSI among adolescents with SAD were examined in n = 31 adolescents with primary diagnosis of SAD, as compared to a group of healthy controls (n = 31). Based on the literature review, we hypothesized that, according to results found in previous adult studies, adolescents suffering from SAD would report NSI significantly more frequently, more vividly, and more distressing than the healthy group. Secondly, we expected NSI to be linked to earlier socially traumatic events and are more likely to be seen from an observer-perspective in adolescents with SAD, compared to healthy adolescents (HA). Thirdly, in exploratory analyses, we examined whether, relative to HA, participants suffering from SAD would a) report emotional states such as anxiety or shame significantly more frequently with a NSI in mind; b) experience the NSI more strongly on different sensual modalities. Furthermore, we explored whether the frequency and several characteristics of NSI would predict social anxiety symptoms over and above the predictive power of depression in the SAD-group.

Results 3.1. Group differences on explicit measures Table 1 presents demographic data, clinical data and test statistics for adolescents suffering from SAD and for healthy adolescents (HA). Since the two groups were matched for age and gender, they were only compared with regard to educational level, yielding no significant differences between the two groups. Adolescents with SAD were found to exhibit significantly higher scores on all clinical measures in comparison to HA. 3.2. Frequency, vividness and level of distress regarding self-images All adolescents reported having experienced NSI in past social situations. In order to compare the two groups on the frequency, vividness as well as the level of distress of NSI, a MANOVA was employed. A main effect of group was identified (Wilks λ = .09; F(3,58) = 194.93; p < .001; η2 = .91). Bonferroni-corrected post-hoc comparisons revealed that the SAD group reported a higher frequency of the images, in comparison to HA (Mdifference = 1.45, SE = .22; p < .001). Furthermore, adolescents with SAD reported higher levels of distress (Mdifference = 45.97, SE = 7.33; p < .001) as well as greater vividness of negative images (Mdifference = 28.32, SE = 6.66; p < .001) relative to HA. 3.3. Link of images to traumatic experiences in the past and observer-perspective In the SAD group, 64.5% (n = 20) of the adolescents reported that the NSI refers to a traumatic real-life event, whereas 45.2% (n = 14) of the HA reported such a link to a past social situation. A χ2-test showed that this difference was not statistically significant (χ2(1, 61) = 1.97; p = .20). Among adolescents who reported a link to a traumatic event, the mean age when it took place was around 15 for both groups (SAD: Mage = 15.32, SD = 3.2; range: 7–20 years old; HA: Mage = 15.3, SD = 3.2; range: 7–20 years old). Exploratory analysis showed that adolescents with SAD rated the memory of this event to be significantly more distressing (SAD, n = 20; M = 73.9; SD = 28.2; vs. HA, n = 14, M = 23.1, SD = 25.3; t(32) = −5.39, p < .001) as well as more vivid (SAD, n = 20, M = 80.9; SD = 23.7; vs. HA, n = 14, M = 40.7, SD = 32.5; t(32) = −4.18, p < .001) than HA. An independent sample t-test showed that NSI were significantly more likely to be experienced from an observer-perspective in adolescents with SAD, compared to HA (SAD, M = 56.97; SD = 30.74; vs. HA, M = 36.8, SD = 31.74; t(60) = −2.94, p < .01). 3.4. Concomitant emotional states In order to investigate the NSI-related emotional states, a MANOVA was computed. A significant main effect of group (Wilks λ = .55; F(10,49) = 3.99; p < .001; η2 = .45) indicated significant differences between the two groups. Bonferroni adjusted post-hoc comparisons (α = .005) revealed that the SAD-group reported significantly greater feelings of shame (p < .001), anxiety (p < .001), embarrassment (p = .001), feeling foolish (p < .001), insecurity (p < .001), vulnerability (p < .001) and humiliation (p < .001). No differences between the groups emerged for nervousness (p = .009), anger (p = .02) and happiness (p = .03). Means and standard deviations for the concomitant emotional states are depicted in Table 2. 3.5. Sensory modalities in recurrent images A MANOVA with group (SAD, HA) as the independent variable and the five sensual modalities as the dependent variable was computed. The analysis revealed no main effect of group (Wilks λ = .91; F(5,55) = 1.1; p = .37; η2 = .09), indicating that both groups experience NSI in a similar fashion. Most of the images involved visual components followed by physiological and auditory components, while gustatory or olfactory domains rarely seem to be involved. For means and standard errors for each modality among the SAD and HA-group, see Table 2. 3.6. Content of NSI Table 3 depicts the means and standard errors for rating the content of the NSI in social situations. A MANOVA with group as the independent variable (SAD, HA) and the different contents of the images as dependent variables, showed that socially phobic adolescents scored significantly higher for all suggested contents (Wilks λ = .44; F(8,52) = 8.14; p < .001; η2 = .86), in comparison to HA. As indicated in Table 3, the two most frequently reported contents within the SAD group were “feeling self-conscious and uncomfortable in public” and “being criticized for getting something wrong”. 3.7. Prediction of social anxiety by characteristics of NSI for adolescents suffering from SAD To examine whether characteristics of NSI would improve the prediction of social anxiety symptoms beyond prediction by depression, a sequential multiple regression analysis was conducted. Depression entered the regression first. In a second step, the frequency and vividness of NSI, as well as the associated distress and the perspective plus the depression scores were entered. Table 4 displays the standardized regression coefficients (β) and R2 change, after entering all predictors. R differed significantly from zero after each step. When entering the image variables in the second step of regression, depression still contributed to the prediction. R2 was .67 in the second step (F(5,28) = 9.49, p < .001). While the frequency of NSI, vividness and level of distress contributed significantly to the prediction (see Table 3), the perspective was not found to be an independent predictor of social anxiety. An evaluation of the simple correlations between the image-related predictors of the hierarchical regression analysis revealed that the vividness did not correlate significantly with the SPAI-score (r = .12, p = .27). Thus, the vividness was identified to be a suppressor variable, since there is no direct relationship with the dependent variable, although making a significant contribution (β = −.42, p = .02) to the prediction of social anxiety via suppressing error variance in other predictor variables ( Tabachnick & Fidell, 2007). Table 4. Summary of sequential regression analysis for characteristics of images predicting social anxiety in SAD. Predictor Social anxiety (SPAI) ΔR2 β Step 1 .51*** Depression .71*** Step 2 .17* Depression .49** Frequency of images .40* Vividness of images -.42* Level of distress .42* Perspective −.01 Total R2 .67*** n 31 Note. SPAI = Social Phobia and Anxiety Inventory. For step 1 adjusted R2 = .49, for step 2 adjusted R2 = .60. *p < .05. **p < .010. ***p < .001.