اثربخشی برنامه نویسی مجدد تصویرسازی برای هراس اجتماعی: مطالعه کنترل شده تصادفی

Abstract Background and objectives There is a need for brief effective treatment of social phobia and Imagery Rescripting (IR) is a potential candidate. The purpose of this study was to examine the efficacy of IR preceded by cognitive restructuring as a stand-alone brief treatment using a randomized controlled design. Methods Twenty-three individuals with social phobia were randomly assigned to an IR group or to a control group. Participants in the IR group were provided with one session of imagery interviewing and two sessions of cognitive restructuring and Imagery Rescripting. Those in the control group had one session of clinical interviewing and two sessions of supportive therapy. Outcome measures including the Korean version of the Social Avoidance and Distress Scale (K-SADS) were administered before and after treatment, and at three-month follow-up. The short version of the Questionnaire upon Mental Imagery and the Traumatic Experience Scale were also administered before treatment. Results Participants in the IR group improved significantly on K-SADS and other outcome measures, compared to the control group. The beneficial effects of IR were maintained at three-month follow-up. It was also found that mental imagery ability and the severity of the traumatic experience did not moderate the outcome of IR. Limitations Further studies are needed to replicate the findings of our study using a large sample. Conclusions The efficacy of IR as a stand-alone brief treatment was demonstrated for social phobia. The findings indicate that IR could be utilized as a cost-effective intervention for social phobia.

Social phobia is a common and enduring anxiety disorder in which images of the self play an important role in maintaining social anxiety. According to the cognitive model developed by Clark and Wells (1995), when individuals with social phobia feel threatened by social situations, they switch their attention from external stimuli to their internal state and construct distorted self-images based on their increased perceptions of bodily sensations. Further, they are convinced that others will see them as they appear in these images, and that this will lead to negative evaluation. Hackmann, Surawy, and Clark (1998) demonstrated that socially phobic individuals were more likely than controls to report experiencing visual images when anxious in social situations. Their images were significantly more negative and more likely to involve seeing themselves from an observer perspective. It was also found that their negative self-images were recurrent and associated with memories of traumatic early social experiences (Hackmann, Clark, & McManus, 2000). Negative self-image has adverse effects on individuals with social phobia in many ways. First, holding a negative image in mind increases anxiety, self-focused attention, and safety behaviors, and thus undermines effective social performance (Hirsch, Clark, Mathews, & Williams, 2003; Vassilopoulos, 2005). Second, individuals with social phobia will judge their performance based on covert information such as bodily sensations and the negative self-imagery that is activated in anxiety-provoking social situations (Clark & Wells, 1995; Rapee & Heimberg, 1997). Third, negative self-images can also play a role in post-event processing. James (2005) found that highly socially anxious individuals had more negatively valenced images during post-event processing than did less socially anxious individuals. Finally, as their negative self-images repeatedly emerge, individuals are likely to accept those images as accurate representations of their present selves (Butler, Fennell, & Hackmann, 2008). In sum, negative self-image, which increases self-focused attention, negative interpretation bias, and safety behaviors, is central to the maintenance cycle of social phobia. One of the key components of the successful treatment of social phobia, therefore, is to help sufferers to view themselves in a more realistic way through replacing their negative self-imagery with a realistic alternative. In cognitive behavior therapy (CBT), guided discovery, the surveying of other people's observations and behavioral experiments have all been used to challenge negative self-image and self-beliefs in socially anxious individuals. Video feedback has proved a particularly powerful way of restructuring distorted self-imagery. Most techniques rely heavily on verbal processing with less emphasis having been placed on imagery. However, emerging evidence has encouraged a renewed interest in the use of imagery as a therapeutic strategy. For example it was demonstrated that, in comparison with verbal processing, imagery had a greater impact on anxiety (Holmes & Mathews, 2005). Pratt, Cooper, and Hackmann (2004) showed that imagery provides a rapid way to access deeper levels of beliefs, including assumptions and core beliefs, when compared to information obtained via verbal cognitions. The use of Imagery Rescripting (IR) with traumatic memories was inspired by the seminal work of Arntz and Weertman (Arntz & Weertman, 1999; Weertman & Arntz, 2007). They devised a therapeutic procedure in which patients revisit adverse early events with an adult perspective. It aims to update the meaning of these events through the use of imagery. First, it involves identifying a childhood memory laden with negative schematic meaning, asking patients to relive the memory from the child's perspective. Then they relive it at their current age, watching what happens to their younger self and may intervene. Finally they relive it from the perspective of the younger self with the adult self present. This time the younger self is often asked what else he/she might need to happen in order to feel better, and receives further interventions from the adult self. The powerful effects of the procedure have been reported for personality disorders (Giesen-Bloo et al., 2006; Weertman & Arntz, 2007), post-traumatic stress disorder (Arntz, Tiesema, & Kindt, 2007; Grunert, Weis, Smucker, & Christianson, 2007), ophidiophobia (Hunt & Fenton, 2007), and depression (Wheatley, Brewin, Patel, & Hackmann, 2007). The first attempt to use IR for social phobia was made by Wild, Hackmann, and Clark (2008) and was built on the Arntz and Weertman (1999) three-stage procedure, but differed in that cognitive restructuring was included. The results indicated that IR decreased distress in response to the image and allow individuals to modify the self-beliefs encapsulated in the image. It also produced significant decreases in social anxiety and fears of negative evaluation, thus indicating the potential role of IR as a therapeutic method. Although this study was seminal in demonstrating the substantial benefits of IR for social phobia, it had methodological limitations because it used a within-subject design and a small sample size. Following from this pioneering study, Nilsson, Lundh, and Viborg (2012) tested the efficacy of IR without cognitive restructuring using an experimental between-groups design. Similar to the findings of Wild et al. (2008), IR led to significant improvements in memory and image distress. Decreases in social anxiety and fears of negative evaluation were also demonstrated. These findings suggest that IR by itself can be a powerful intervention tool for social phobia. However, in both studies IR was not regarded as a viable stand-alone treatment option and so its long-term effect was not examined. The current study was designed to examine the efficacy of IR as a stand-alone intervention for social phobia using a randomized controlled trial. Previous research has shown that IR can reduce both the vividness of the early traumatic memory and the distress it causes, and allow changes in beliefs encapsulated in the image. It has also been found to lead to significant decreases in social anxiety and fears of negative evaluation. As was earlier discussed, negative self-imagery is a key maintaining factor in the persistence of social phobia. We propose that IR will be able to bring about changes in negative self-imagery and self core-beliefs, and thus lead to clinical improvement in social phobia. In our preliminary clinical trial, social phobic individuals showed very positive responses to IR including decreases in social anxiety symptoms, fears of negative evaluation, and avoidance behaviors, and did not feel the need for additional treatment after IR. Yoon and Kwon (2010) reported that short-term IR led to dramatic improvement in a Korean socially phobic patient who had not received benefit from group CBT. Recent research examining the efficacy of IR for other disorders has also indicated the therapeutic potential of IR as a stand-alone treatment (Brewin et al., 2009). According to an internet-based survey, only one third of respondents with social phobia had received treatment, despite the severity of their disorder (Erwin, Turk, Heimberg, Fresco, & Hantula, 2004). In the absence of treatment, this disorder is very likely to run a chronic course (Dewit, Ogborne, Offord, & MacDonald, 1999). Therefore, one of the clinical issues which needs to be addressed in the treatment of social phobia is the accessibility and affordability of evidence-based psychological treatments. In Korea, the availability of CBT has thus far been limited by its cost and the lack of trained practitioners, and therefore there is a great need for an effective and brief therapeutic intervention. In the current study we aimed to examine the efficacy of IR as a stand-alone intervention for social phobia. This study followed the procedure used by Wild et al. (2008), which enhanced the usual IR with cognitive restructuring.1 We further attempted to strengthen the efficacy of IR by adding one more session. In our preliminary trial, it was found that patients were better able to consolidate the benefits of the intervention when they had at least two sessions of IR. We predicted that IR would, as compared to the control condition, result in fewer symptoms of social anxiety, reduced fears of negative evaluation, less memory distress, and reduced image vividness and frequency. We also tried to improve some of the methodological limitations of previous studies. First, we used a randomized controlled trial to improve the internal validity of the study. Second, we investigated whether the therapeutic effects were maintained at three months after treatment instead of using a short follow-up period of one week. Third, we also examined the relationship between imagery ability and the outcome of the intervention. There are individual differences in people's ability to use an imagery-based technique, and some individuals may experience difficulty with the task of recalling and creating images. At the moment there is not enough evidence to determine whether imagery ability moderates the efficacy of IR or not. It was found that imagery ability did not affect the efficacy of IR for simple phobia (Hunt & Fenton, 2007). We examined whether imagery ability would moderate the efficacy of IR for social phobia. In addition, we examined whether the severity of traumatic experience would affect the efficacy of IR. When IR was used as a brief stand-alone treatment for depressed patients, it was shown that patients reporting additional memories tended to have more treatment sessions and a larger drop on the BDI (Brewin et al., 2009). Unfortunately, the relationship between the severity of traumatic experience and the outcome was not examined in this study. Therefore our study attempted to explore how the severity of traumatic experience will affect the treatment effect of IR.

3.1. The effects of Imagery Rescripting on social anxiety We conducted repeated-measure ANOVAs on the social anxiety scales with the IR and control groups as the between subjects factor (Group) and time (pre-, post-treatment) as the within subjects factor (Time). These analyses revealed no significant main effects of group on the social anxiety scales, but significant main effects of time on the K-SADS, F(1,21) = 50.48, p < 0.001, K-BFNE, F(1,21) = 23.86, p < 0.001, LSAS-Fear, F(1,21) = 11.26, p < 0.01, and LSAS-Avoidance, F(1,21) = 25.81, p < 0.001. More importantly, there were significant interaction effects of group by time on the K-SADS, F(1,21) = 9.00, p < 0.01, K-BFNE, F(1,21) = 11.05, p < 0.01, LSAS-Fear, F(1,21) = 5.27, p < 0.05, and LSAS-Avoidance, F(1,21) = 15.42, p < 0.01. To delineate the significant interaction effects, we used one-tailed paired samples t-tests and calculated the effect sizes (Cohen's d) to measure the magnitude of the differences. In the control group, there were no significant differences between pre- and post-treatment, except for the K-SADS, t(9) = 3.02, p < 0.05. For the IR group, there were significant differences between pre- and post-treatment on the K-SADS, t(12) = 7.19, p < 0.001, K-BFNE, t(12) = 7.62, p < 0.001, LSAS-Fear, t(12) = 4.26, p < 0.01, and LSAS-Avoidance, t(12) = 6.10, p < 0.001. Cohen (1988) proposed a threefold classification of effect sizes: small (0.20–0.40), medium (0.50–0.79), and large (0.80 and above). According to this system, the effect sizes for the control group were mostly small. By contrast, the effect sizes for the IR group were mostly large. Although differences between pre- and post-treatment were significant on the K-SADS for both groups, a notable difference in effect size was observed, with the IR group displaying a large effect size (d = 1.53) and the control group displaying a medium effect size (d = 0.69). 3.2. The effects of Imagery Rescripting on negative self-image, traumatic early memory, and encapsulated belief We conducted repeated-measure ANOVAs on the imagery, memory, and encapsulated belief ratings, using the IR and control groups as the between subjects factor (Group) and the measured times as the within subjects factor (Time). Table 1 shows the mean scores and standard deviations at pre-IR and post-IR for the IR and the control groups, as well as the main effects, the interaction effects, and effect sizes. The analyses revealed significant main effects of group on image distress, F(1,21) = 8.31, p < 0.05, memory distress, F(1,21) = 4.47, p < 0.05, and encapsulated belief, F(1,21) = 5.09, p < 0.05, but no significant main effects of group on image vividness, and memory vividness. There were significant main effects of time on image vividness, F(1,21) = 34.14, p < 0.001, image distress, F(1,21) = 69.02, p < 0.001, memory vividness, F(1,21) = 31.10, p < 0.001, memory distress, F(1,21) = 56.11, p < 0.001, and encapsulated belief, F(1,21) = 75.22, p < 0.001. However, these were modified by significant group by time interactions on image vividness, F(1,21) = 8.05, p < 0.05, image distress, F(1,21) = 18.41, p < 0.001, memory vividness, F(1,21) = 5.51, p < 0.05, memory distress, F(1,21) = 16.76, p < 0.01, and encapsulated belief, F(1,21) = 19.42, p < 0.001. Table 1. Social anxiety, imagery, memory, and encapsulated belief: means, standard deviations, main effects, interaction effects, and effect sizes. Measure Groupa Pre Post Effect sizeb Main effect Interaction effect M (SD) M (SD) d Group F(1,21) Time F(1,21) Group × Time F(1,21) K-SADS IR 106.54 (11.70) 91.77 (7.03) 1.53 0.62 50.48*** 9.00** Control 105.00 (10.04) 99.00 (7.09) 0.69 K-BFNE IR 49.85 (3.53) 43.54 (4.29) 1.61 4.08 23.86*** 11.05** Control 50.10 (4.12) 48.90 (2.89) 0.34 LSAS-Fear IR 39.38 (12.95) 31.92 (13.88) 0.56 0.19 11.26** 5.27* Control 38.50 (11.15) 37.10 (9.34) 0.14 LSAS-Avoidance IR 38.23 (13.40) 22.62 (13.84) 1.15 0.83 25.81*** 15.42** Control 36.00 (12.41) 34.00 (9.94) 0.18 Image vividness IR 84.23 (9.54) 43.23 (21.66) 2.45 2.42 34.14*** 8.05* Control 79.70 (14.41) 65.50 (22.17) 0.76 Image distress IR 85.38 (8.02) 36.77 (18.39) 3.43 8.31* 69.02*** 18.41*** Control 82.00 (7.15) 66.50 (18.86) 1.09 Memory vividness IR 84.08 (10.85) 41.15 (22.00) 2.48 1.19 31.10*** 5.51* Control 77.50 (14.19) 60.00 (24.61) 0.87 Memory distress IR 85.38 (11.08) 34.23 (19.98) 3.17 4.47* 56.11*** 16.76** Control 79.00 (13.08) 64.00 (21.32) 0.85 Encapsulated belief IR 83.69 (12.51) 40.77 (14.98) 3.11 5.09* 75.22*** 19.42*** Control 81.00 (12.65) 67.00 (18.29) 0.89 Note. K-SADS = Korean version of Social Avoidance and Distress Scale; K-BFNE = Korean version of Brief Fear of Negative Evaluation Scale; LSAS = Liebowitz Social Anxiety Scale. *p < 0.05. **p < 0.01. ***p < 0.001. a IR group (n = 13), Control group (n = 10). b Cohen's d was computed using the means and standard deviations of the IR or Control group. Table options To examine the interaction effects more closely, we conducted one-tailed paired samples t-tests and calculated the effect sizes (Cohen's d). For the IR group, the post-IR scores were significantly lower than the pre-IR scores for image vividness, t(12) = 6.14, p < 0.001, image distress, t(12) = 9.16, p < 0.001, memory vividness, t(12) = 6.69, p < 0.001, memory distress, t(12) = 8.42, p < 0.001, and encapsulated belief, t(12) = 9.03, p < 0.001. For the control group, the post-control scores were significantly lower than the pre-control scores for image distress, t(9) = 2.84, p < 0.05, memory distress, t(9) = 2.41, p < 0.05, and encapsulated belief, t(9) = 3.33, p < 0.05, but there were no significant differences between pre- and post-control scores on image vividness, and memory vividness. The effect sizes ranged from medium to large for the control group (average d = 0.89; between 0.76 and 1.09), but the effect sizes for the IR group were all large (average d = 2.93; between 2.45 and 3.43) ( Fig. 3). Full-size image (31 K) Fig. 3. Image vividness, image distress, memory distress, and encapsulated belief scores at the pre- and post-IR stages. Figure options 3.3. Maintenance of Imagery Rescripting effectiveness: three-month follow-up To test whether the effect of the IR was maintained after three months, we performed repeated-measure ANOVA. We analyzed the data of the 12 participants who returned their three-month follow-up questionnaires, from the 13 patients who had participated in the IR group. Table 2 shows the mean scores and standard deviations at each assessment point, as well as the results of repeated contrast. There were significant differences among the scores of three measured times for the K-SADS, F(2,22) = 66.94, p < 0.001, K-BFNE, F(2,22) = 34.25, p < 0.001, LSAS-Fear, F(2,22) = 14.10, p < 0.001, LSAS-Avoidance, F(2,22) = 29.76, p < 0.001, image vividness, F(2,22) = 22.71, p < 0.001, image distress, F(2,22) = 57.15, p < 0.001, memory vividness, F(2,22) = 27.56, p < 0.001, memory distress, F(2,22) = 60.41, p < 0.001, and encapsulated belief, F(2,22) = 67.98, p < 0.001. To measure significant differences between pre-IR, post-IR, and at three-month follow-up specifically, Repeated Contrast was analyzed. The IR group's scores on all measures had significantly dropped at post-IR compared to pre-IR, and the reduction on all measures was maintained at three month follow up. At the follow-up assessment, the scores on the K-SADS, and encapsulated belief were further decreased. Among 12 participants, nine participants' scores on the K-SADS and seven participants' scores on the K-BFNE were reduced to be below the cut-off point. Table 2. Means and standard deviations for the IR group at pre-IR, post-IR, and three-month follow-up on social anxiety, imagery, memory, and belief measures (n = 12). Measure Pre-IRa Post-IRb Three-month Follow-upc F(2,22) Repeated contrast M (SD) M (SD) M (SD) K-SADS 107.42 (11.76) 92.08 (7.24) 87.00 (9.35) 66.94*** a > b > c K-BFNE 49.75 (3.67) 43.33 (4.42) 41.42 (5.11) 34.25*** a > b, c LSAS-Fear 40.25 (13.12) 32.75 (14.16) 28.75 (13.07) 14.10*** a > b, c LSAS-Avoidance 39.58 (13.04) 23.67 (13.90) 23.08 (13.76) 29.76*** a > b, c Image vividness 83.75 (9.80) 44.33 (22.24) 36.25 (23.07) 22.71*** a > b, c Image distress 85.83 (8.20) 37.33 (19.09) 27.08 (18.64) 57.15*** a > b, c Memory vividness 84.42 (11.26) 42.92 (22.00) 37.50 (26.42) 27.56*** a > b, c Memory distress 85.83 (11.45) 34.58 (20.83) 27.33 (20.00) 60.41*** a > b, c Encapsulated belief 82.33 (12.02) 40.00 (15.37) 27.33 (18.92) 67.98*** a > b > c Note: K-SADS = Korean version of Social Avoidance and Distress Scale; K-BFNE = Korean version of Brief Fear of Negative Evaluation Scale; LSAS = Liebowitz Social Anxiety Scale. ***p < 0.001. Table options 3.4. The relationship between mental imagery ability level, the severity of traumatic experience, and the effectiveness of the Imagery Rescripting The effectiveness of the IR as a whole on the K-SADS, K-BFNE, LSAS-Fear, and LSAS-Avoidance scores was measured by subtracting the post-IR scores from the pre-IR scores. To examine the relationships between the mental imagery ability level, the severity of traumatic experience, and the effectiveness of the IR, we performed Pearson correlation analysis. The results of this are presented in Table 3. There was no significant relationship found between participants' mental imagery ability level and the IR effect. There was also no significant relationship found between the severity of traumatic experience and the IR effect. Table 3. Correlations of SQMI and traumatic experience scale with effectiveness of Imagery Rescripting (n = 13). ΔK-SADS ΔK-BFNE ΔLSAS-Fear ΔLSAS-Avoidance SQMI 0.05 0.37 0.17 0.23 Traumatic experience scale −0.06 0.11 0.35 −0.04 Note. K-SADS = Korean version of Social Avoidance and Distress Scale; K-BFNE = Korean version of Brief Fear of Negative Evaluation Scale; LSAS = Liebowitz Social Anxiety Scale; SQMI = Short version of the Questionnaire upon Mental Imagery.