Selective attention in social phobia and the moderating effect of a concurrent depressive disorder

Studies using the modified Stroop colour naming task have provided results consistent with the hypothesis that social phobia is associated with an attentional bias towards negative social-evaluative words. However, these results could also have arisen as a consequence of non-attentional processes. For this reason, the present study uses a modified version of MacLeod et al.'s (J. Abnorm. Psychol. 95 (1986) 15) dot-probe task, which provides a more direct measure of attention. Patients with social phobia (n = 28), patients with social phobia and a concurrent depressive disorder (n = 33), and non-patients (n = 40) were presented with word pairs each consisting of a neutral word and a threat word. The results indicated that patients with social phobia show an attentional bias towards social-threat words while non-patients tend to avoid social-threat words. Patients with social phobia and a concurrent depressive disorder behaved like non-patients, indicating that concurrent depression abolishes the attentional bias. Physical threat words were also included in the study. The main analysis indicated that social phobia is also associated with an attentional bias to physical threat. However, a post hoc analysis (which requires replication) suggested that the physical threat bias might have arisen because some social phobia patients also had another anxiety disorder in which physical concerns are likely to have been prominent. Overall, the results emphasise the importance of assessing comorbidity when investigating attentional biases. (C) 2003 Elsevier Science Ltd. All rights reserved.