استراتژی های درمان رابطه ای رفتارهای رویکرد اجتماعی در بیماران مبتلا به اختلال اضطراب اجتماعی را افزایش می دهد

Abstract We incorporated strategies based on relational and interpersonal circumplex research within a standard cognitive-behavioral regimen for Generalized Social Anxiety Disorder (GSAD, Generalized Social Phobia) to determine whether these techniques increased the social approach behaviors that facilitate relationship development. Individuals seeking treatment for GSAD were randomly assigned to either the integrated interpersonal cognitive-behavioral group treatment (ICBT) or a wait list condition (WL). Results revealed that the interpersonal techniques were readily implemented by the majority of patients. ICBT produced significant increases in frequency of social approach behaviors and relationship satisfaction, in addition to GSAD symptom reductions comparable to other group CBT regimens. The current research highlights the feasibility and potential benefit of incorporating strategies based on relational and circumplex theories into cognitive-behavioral regimens for GSAD.

8. Results 8.1. Demographic characteristics 8.1.1. Treatment versus wait list To compare demographic and clinical characteristics of ICBT and WL groups, Chi square analyses were conducted on gender, marital status, country of birth (North American versus non-North American), concurrent medication, type of medication, comorbid diagnoses, and past treatment; Analyses of Variance (ANOVAs) were conducted on age and education. The findings revealed no significant between-group differences on any of these variables, all p > .10. See Table 1. The same analyses were conducted to compare treatment completers and non-completers and indicated there were no significant between-group differences on any of the demographic or clinical characteristics, all p > .10. A final set of analyses were conducted to compare completers who participated in follow-up (n = 21) and those lost from follow-up (n = 6). Again, no significant differences emerged on any of the demographic or clinical characteristics, all p > .10. In addition, ANOVAs conducted on the pre- and post-treatment scores indicated there were no significant between-group differences between follow-up participants and follow-up drop-outs on those variables. 8.2. Treatment completion and homework compliance Four ICBT (12.9%) and two WL (8%) participants failed to complete treatment.2 Fisher's exact test revealed no significant difference in proportion of drop-outs between the two conditions, p = .68. By comparison, McEvoy (2007) reported an average 13.2% attrition rate in group regimens and Aderka (2009) a 19.2% rate. For homework compliance, the ICBT participants obtained homework completion ratings of 3 or greater for 95% of the sessions. For 76% of the sessions, completion was rated at 4 (fully completed) or 5 (exceeded assignment). 8.3. Social approach and avoidance To assess whether the IO scales added unique information beyond the symptom measures, a factor analysis was conducted on the three IO scales, the SIAS, SPS, and ADIS-IV Fear/Avoidance composite using Maximum Likelihood extraction with direct oblimin rotation.4 Cattell's (1966) scree test, parallel analysis, and the Chi-square goodness of fit index all indicated that a two-factor solution provided the best fit to the data. The first factor to emerge explained 41.5% of the shared variance and comprised the three IO scales. The second factor explained 24.3% of the shared variance and comprised the three symptom scales. The two factors were negatively correlated (r = −.36) yet were sufficiently independent to be considered two separable dimensions. The factor loadings are presented in Table 2. Table 2. Factor loadings for the interpersonal outcome and SAD symptom measures. Social Approach Social Avoidance SA .41 −.19 RS .65 −.12 SS 1.00 .29 SIAS .02 .79 SPS .09 .65 ADIS-IV-Fear/Avoidance −.13 .63 Note: Interpersonal Outcomes Scales: SA: Social Approach, RS: Relationship Satisfaction; SS: Social Support; SIAS: Social Interaction Anxiety Scale; SPS: Social Phobia Scale; ADIS-IV: Anxiety Disorders Interview Schedule for DSM-IV. Table options 8.4. Treatment outcome The dependent variables were analyzed with 2 (condition: ICBT, WL) by 2 (time: pre, post-assessment) MANOVAs, followed by ANOVAs and simple effects analyses with Sidak corrections for multiple comparisons.5 MANOVAs were selected to control type 1 error rates. Separate analyses were conducted for the three interpersonal outcome scales and the five SAD symptom measures (SIAS, SPS, BFNE, ADIS-IV Fear/Avoidance, ADIS-IV Severity) for both the treatment completer (TC) and intent-to-treat (ITT) samples. In the ITT analyses, the pre-assessment scores were entered for post-assessment scores for the 4 ICBT and 2WL drop-outs as these were the only data available. An examination of both sets of measures indicated that the data met the assumptions of univariate and multivariate normality, linearity, and an absence of significant multicolinearity or singularity.6,7 For efficiency, the multivariate analyses are reported below and the ANOVAs are presented in Table 3 and Table 4. Table 3. Interpersonal outcome measures at each assessment and condition by time ANOVAs. Assessment ICBT Wait list Within Ss ANOVAs, Pre to Post M SD M SD Treatment completers Social Approach Scale Pretreatment 31.78 a 7.13 30.30a 11.34 Time F (1, 48) = 22.07 *** Posttreatment 41.89b 9.76 31.87a 11.72 Condition × Time F (1, 48) = 11.82 *** Follow-up 41.14b 9.32 Relationship Satisfaction Scale Pretreatment 12.65a 4.11 13.17a 5.07 Time F (1,48) = 13.73 ** Posttreatment 16.69b 5.22 13.78a 4.95 Condition × Time F (1,48) = 7.21 ** Follow-up 16.05b 4.96 Social Support Pretreatment 25.04 8.50 23.83 12.60 Time F (1,48) = 7.08 * Posttreatment 29.11 9.94 27.09 14.02 Condition × Time F (1, 48) = 4.12 Follow-up 27.52 11.06 Note: Means with different subscripts differ significantly. * p < .05. ** p < .01. *** p < .001. Table options Table 4. SAD symptom outcome measures at each assessment and condition by time ANOVAs. Assessment ICBT Wait list Within Ss ANOVAs Pre to Post M SD M SD Treatment completers Social Interaction Anxiety Scale Pretreatment 50.93a 11.87 56.70a 8.94 Time F(1, 48) = 40.26 *** Posttreatment 34.93b 12.18 54.43a 10.00 Condition × Time F (1, 48) = 22.78 *** Follow-up 35.29b 12.85 Social Phobia Scale Pretreatment 28.50a 14.99 33.26a 14.43 Time F (1,48) = 12.99 *** Posttreatment 16.83b 10.36 34.04a 16.17 Condition × Time F (1,48) = 17.00 ***. Follow-up 17.38b 10.65 Brief Fear of Negative Evaluation Pretreatment 50.77a 5.29 50.57a 6.62 Time F (1,48) = 25.62 *** Posttreatment 41.56b 8.70 50.57a 7.94 Condition × Time F (1, 48) = 25.62 *** Follow-up 41.43b 8.52 ADIS-IV Fear-Avoidance Composite Pretreatment 52.94a 14.53 58.80a 10.41 Time F(1,48) = 55.18 *** Posttreatment 33.23b 7.40 52.46a 6.60 Condition × Time F (1, 48) = 14.51 *** Follow-up 34.89b 9.75 ADIS-IV Severity Pretreatment 5.3a 0.9 5.4a 0.7 Time F (1,48) = 33.22 *** Posttreatment 3.6b 1.2 5.5a 0.4 Condition × Time F (1, 48) = 40.66 *** Follow-up 3.6b 1.3 Note: Means with different subscripts differ significantly. *** p < .001. Table options 8.4.1. Treatment completers 8.4.1.1. Interpersonal outcomes The TC MANOVA revealed a significant effect for time, F (3, 46) = 7.98 that was modified by a significant condition by time interaction, F (3, 46) = 5.15, p < .01, View the MathML sourceηp2=.25. The ANOVAS and simple effects analyses indicated that ICBT participants displayed significant increases on the SA and RS scales from pre- to post-assessment (p < .01), while the WL group did not (p > .10). The condition by time interaction for SS was not significant (see Table 3). 8.4.1.2. SAD symptom measures The TC MANOVA revealed a significant main effect for time, F (5, 44) = 14.09. p < .001, that was modified by significant a condition by time interaction, F (5, 44) = 8.89, p < .001, View the MathML sourceηp2=.50. The ANOVAs and subsequent simple effects analyses indicated that the ICBT group displayed significant reductions on all five measures from pre- to post-assessment. The WL group displayed significant change (p = .02) only on the ADIS Fear/Anxiety composite and not on the other four measures (see Table 4). 8.4.2. Intent to treat sample 8.4.2.1. Interpersonal outcomes The condition by time MANOVA revealed a significant effect for time, F (3, 52) = 7.00, that was modified by a significant condition by time interaction, F (3, 52) = 4.42, p < .01, View the MathML sourceηp2=.20. As with the TC sample, the ANOVAS and simple effects analyses revealed that the ICBT group displayed significant increases on the SA, RS, and SS scales from pre- to post-assessment, while the WL group did not. 8.4.2.2. SAD symptom measures The condition by time MANOVA revealed a significant main effect for time F (5,50) = 9.93, p < .001, that was modified by a significant condition by time interaction, F (5,49) = 6.22, p < .001, View the MathML sourceηp2=.38. The ANOVAS and subsequent simple effects analyses revealed that the ICBT group displayed significant reductions (p < .001) from pre- to post-assessment on all five measures, while the WL group displayed change (p < .05) only on ADIS Fear/Anxiety composite and not on the other four symptom measures. 8.4.3. Maintenance of treatment gains Follow-up scores for the ICBT group can be seen in Table 3 and Table 4. A within-subjects MANOVA conducted on the IO scales revealed a significant main effect for time, F (6, 76) = 4.09, p < .001, View the MathML sourceηp2=.24. Simple contrasts with Bonferroni corrections indicated that the ICBT group displayed significant (p < .01) increases on SA and RS from pre to post-assessment. The post- and follow-up scores were not significantly different, which indicates that those changes were maintained at follow-up. For SS, no significant differences emerged between the three time points. A within-subjects MANOVA conducted on the three self-report SAD symptom measures revealed a significant main effect for time, F (6,76) = 6.29, p < .001, View the MathML sourceηp2=.33. 8 The within-subjects MANOVA conducted on the two ADIS-IV ratings revealed a significant main effect for time, F (4,62) = 20.33, p < .001, View the MathML sourceηp2=.56. Simple contrasts indicated that the ICBT group displayed significant (p < .001) reductions on all five SAD symptom measures from pre- to post-assessment. The post- and follow-up scores were not significantly different, which indicates that the initial gains were maintained across the follow-up period. 8.4.4. Magnitude of treatment change Table 5 presents the effect size (ES) comparisons to the benchmark studies of McEvoy (2007) and Aderka (2009). For comparability with these benchmarks, the current ES were based on the pre-treatment to post-treatment (within group) effect sizes for the ICBT group. McEvoy and Aderka selected this approach rather than between-group ES because the studies included in the meta-analyses used different comparison conditions (placebo, wait-list, active treatment), which meant that between-group ES for the different studies were not directly comparable. Following the procedure used by McEvoy (2007), Cohen's d for the SIAS and SPS were computed as follows: d = Mpost−Mpre/SDpre. Following Aderka (2009), we also computed the average reliability-corrected Cohen's d score across the three self-report symptom measures as follows: d = Mpost−Mpre/√rxyxS.D.pre-treatment, where rxy is the test–retest reliability of the outcome measure used. The results indicated that the current ICBT regimen produced similar or greater ES compared with the benchmark figures. Notably, the ES for SPS was lower than that for the SIAS, which was likely due to the symptom profile of the current sample, which obtained high scores on social interaction anxiety and lower scores on the performance fears measured by the SPS. To examine treatment durability, pre- to follow-up Cohen's ds were computed following the same two procedures. As can be seen in Table 5, the ES at follow-up were generally comparable to those at post-assessment. 9 Table 5. Effect Sizes on SAD symptom self-report measures compared to benchmark meta-analyses. Study ES ES Current Study Pre-post Pre-follow-upd McEvoy (2007)a SIAS .74 1.35 1.20 SPS .70 .78 .74 Aderka (2009)b Average .92 1.33c 1.27c Note: ES: Effect Size. a Mean Cohen's d for group format outcome studies, treatment completers. b Weighted mean for reliability-corrected Cohen's d for self-report measures in group format outcome studies, treatment completers. c Mean reliability-corrected d for the SIAS, SPS, and BFNE. d Based on follow-up completers (n = 21). Table options For comparison purposes, we computed ES for the IO scales. Because the scales were developed for this research, no clinical benchmark data were available. Therefore, the ES for the ICBT group were compared to those of the WL group. The within-condition Cohen's d for SA, RS, and SS were 1.41, .98, and .48 for the ICBT group versus 0.13, 0.12, and .26 for the WL group. The mean reliability-corrected d for the three scales = 1.11 for ICBT versus 0.20 for WL.