ABSTRACT
Many individuals with social anxiety disorder (SAD) have depressive symptoms that meet criteria for major depressive disorder (MDD). In our study, we examined the temporal relationship between symptoms of social anxiety and symptoms of depression during the course of an 11-week internet-delivered cognitive behavioral treatment (ICBT) for SAD (n = 170). Specifically, we investigated whether weekly changes in social anxiety mediated changes in depression, changes in depression mediated changes in anxiety, both or neither. In addition, we compared individuals with SAD and MDD (n = 50) and individuals with SAD and no MDD (n = 120) to examine the role of MDD as a moderator of the social anxiety—depression relationship. Lower-level mediational modeling revealed that changes in social anxiety symptoms mediated changes in depression symptoms to a greater extent than vice versa. In addition, mediation among individuals with SAD and MDD was significantly greater compared to individuals with SAD and no MDD. Our findings suggest that ICBT is effective in treating individuals with SAD regardless of comorbid depression, and that focusing ICBT interventions on social anxiety can lead to significant reductions in depression among individuals with SAD.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Ethical approval
This study received ethics approval by the IRB at the University of Haifa (protocol number 288/19).
Notes
1. As the groups (immediate ICBT vs. waitlist/delayed ICBT) did not significantly differ in their changes in social anxiety, depression and their relationship during the active phase of treatment, we analyzed data from all participants together.
2. We examined several models of time including linear time and quadratic time. The linear time model fit the data best and explained the largest proportion of variance. For instance, linear time explained 74.8% of the variance in social anxiety over the course of treatment whereas quadratic time explained 23.5% of the variance in social anxiety over the course of treatment. This reflects a three-fold difference in explained variance. Thus, we used a linear time model in all analyses.