Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 250, 297-301

The Structure of Vulnerabilities for Social Anxiety Disorder

Affiliations

The Structure of Vulnerabilities for Social Anxiety Disorder

Thomas L Rodebaugh et al. Psychiatry Res.

Abstract

Social anxiety disorder symptoms are generally proposed to be related to broad temperamental vulnerabilities (e.g., a low level of approach and high level of avoidance temperament), specific psychological vulnerabilities (e.g., fears of negative and positive evaluation), and additional disorders (e.g., major depressive disorder). However, existing tests of such a model have either not considered depressive symptoms or relied on samples of undergraduates. We examined these and related questions via a latent variable model in a large dataset (N=2253) that combined participants across a variety of studies. The model had adequate fit in the whole sample, and good fit in a subsample in which more participants completed the depression measure. The model indicated that low level of approach and high level of avoidance temperament contributed to fears of evaluation and social anxiety symptoms, and that fears of evaluation additionally contributed independently to social anxiety symptoms. The relationship between social anxiety and depressive symptoms was entirely accounted for by these vulnerabilities: Depressive symptoms were only predicted by avoidance temperament.

Keywords: Depression; Hierarchical models; Social anxiety disorder; Social phobia; Structural equation modeling.

Figures

Figure 1
Figure 1
The hypothesized latent variable model with standardized regression coefficients. The paths from temperamental vulnerabilities to symptoms are such that they are above and beyond prediction by specific vulnerabilities. Paths in dashed lines were hypothesized but not supported by the final model. The following are not shown: individual items, error terms, lower-order factors, and regression coefficients that were statistically nonsignificant (p > 0.05). All paths shown in solid lines are p < 0.009. The depression factor shown is the General depression factor; Somatic and Cognitive subfactors of depression were modeled but not shown here. The only significant effect associated with those subfactors was for the Somatic subfactor correlating strongly with Avoidance (r = 0.65, p < 0.001).

Similar articles

See all similar articles

Cited by 2 PubMed Central articles

Feedback