اجتناب رفتاری واسطه ارتباط علائم اضطراب و افسردگی در بین بیماران مبتلا به اختلال اضطراب اجتماعی

Abstract This study investigated the relationship between social anxiety, depressive symptoms, and behavioral avoidance among adult patients with Social Anxiety Disorder (SAD). Epidemiological literature shows SAD is the most common comorbid disorder associated with Major Depressive Disorder (MDD), though the relationship between these disorders has not been investigated. In most cases, SAD onset precedes MDD, suggesting symptoms associated with SAD might lead to depression in some people. The present study addressed this question by investigating the mediational role of behavioral avoidance in this clinical phenomenon, using self-report data from treatment-seeking socially anxious adults. Mediational analyses were performed on a baseline sample of 190 individuals and on temporal data from a subset of this group. Results revealed behavioral avoidance mediated this relationship, and supported the importance of addressing such avoidance in the therapeutic setting, via exposure and other methods, as a possible means of preventing depressive symptom onset in socially anxious individuals.

Results 6.1. Descriptive statistics Symptom severity is reported for all participants at baseline and at mid- and post-treatment for those included in the temporal mediational analyses (see Table 2 for pre-treatment, mid-treatment, and post-treatment descriptive statistics for the SPAI, LSAS-Avoidance subscale, and BDI-II). Table 2. Means (S.D.) for temporal participants Baseline Mid-treatment Post-treatment BDI-II 13.29 (10.89) 9.88 (8.24) 6.82 (8.25) LSAS-Avoid 35.47 (12.11) 27.60 (11.82) 23.65 (11.18) SPAI-SP 133.61 (24.37) 112.66 (24.71) 97.21 (30.95) Note: BDI-II, Beck Depression Inventory-II; LSAS-Avoid, Liebowitz Social Anxiety Scale-Avoidance subscale; SPAI-SP, Social Phobia and Anxiety Inventory-Social Phobia subscale. Table options 6.2. Test of mediation at baseline Mediational analyses were conducted according to steps outlined by Baron and Kenny (1986) and Frazier, Tix, and Barron (2004). Steps 1–3 were assessed to establish Paths a, b, c, and c’ as described in Baron and Kenny. First, depressive symptoms were regressed on social anxiety symptoms to establish the “Path c” relationship between the predictor (i.e., social anxiety symptoms) and the outcome (i.e., depressive symptoms), indicating a statistically significant relationship. Next, behavioral avoidance was regressed onto social anxiety to establish Path a (Step 2), indicating a statistically significant relationship. Step 3 was then conducted to test whether behavioral avoidance was significantly related to depressive symptoms. We regressed depressive symptoms simultaneously on social anxiety symptoms and behavioral avoidance symptoms to establish Paths b and c’. The coefficient associated with the relation between behavioral avoidance and depressive symptoms (controlling for social anxiety symptoms) was statistically significant. This significant relationship confirms that the condition for Step 3 was met. Lastly, this third regression equation also provided an estimate of Path c’, the relation between social anxiety symptoms and depressive symptoms, controlling for behavioral avoidance. Path c’ (b = .075, p = .036) was smaller than Path c, indicating partial mediation (see Fig. 1 for summary of path strengths). As path c’ was not equal to zero, behavioral avoidance was not a complete mediator. The drop from .129 to .075 (i.e., from c to c’) was significant, according to the Sobel test ( Sobel, 1982). Lastly, it was calculated that 41.9% of the total effect of social anxiety symptoms on depressive symptoms was mediated by behavioral avoidance, based on the formula recommended by Shrout and Bolger (2002). Results of the hypothesized mediational model with behavioral avoidance as a ... Fig. 1. Results of the hypothesized mediational model with behavioral avoidance as a mediator between social anxiety fear and depressive symptoms (unstandardized regression coefficients presented; *p < .05; Sobel Test: z = 2.17). Figure options 6.3. Test of mediation over the course of treatment Although a formal test of mediation was not conducted because of the lack of a control group, the second part of this investigation examined the mediational effects of behavioral avoidance over the course of treatment to examine why decreases in social fear could lead to later changes in depressive symptoms. Because of the potential shortcomings of utilizing simple raw change scores (e.g., over- and underestimation of change), residualized change scores were calculated using recommendations by Steketee and Chambless (1992) and recently applied in two studies of mediating variables in SAD (Dalrymple & Herbert, 2007; Hofmann, 2004). To examine whether pre-mid changes in behavioral avoidance were associated with later changes in depressive symptoms, we correlated residual gain scores of the LSAS (pre- to mid-treatment) with the residual gain scores of the BDI-II from mid-treatment to post-treatment. The results showed that the Pearson correlation coefficient between the two residual gain scores was moderate in strength and statistically significant: r = −.25 (p < .05).