Attention Bias and Attentional Control in the Development of Social Anxiety Disorder

Abstract

Although several efficacious treatments exist for social anxiety disorder (SAD), less research has been devoted to identifying specific mechanisms involved in the etiology of SAD using high-risk, longitudinal designs. Given the high prevalence and personal and societal burden associated with a diagnosis of SAD, research is needed to elucidate causal factors at play in the development of SAD to inform innovative prevention programs for at-risk individuals. Theoretical models and empirical research suggest that biased attention toward threat-relevant information is an important factor in the maintenance of SAD. However, relatively little is known about the role of attention bias to threat in the development of SAD, and evidence is inconclusive with regard to whether attention biases lead to increases in anxiety over time. Also, only one study has examined attentional control as a potential factor moderating this relationship despite long-held assertions that "control over cognitive processes" may be an important individual difference factor determining the strength of the relationship between attention bias and development of excessive anxiety. Finally, a few studies have shown that attention bias to threat predicts stress reactivity, but these studies have only been conducted in unselected samples rather than with individuals at risk for developing SAD. Thus, the aims of this study were to examine the moderating effects of risk for SAD and attentional control on the relationships between attention bias to threat and (1) psychological and biological social stress reactivity and (2) development of SAD. The primary aim of the study was to examine the aforementioned relationships using attention bias to threat as assessed using the modified probe detection task (MPDT). In an exploratory analysis, the relationships were examined using an index of attention disengagement bias assessed with the Posner spatial cueing task (PSCT). Attentional control was represented by four indices, analyzed in separate regression analyses given their weak bivariate associations (i.e., Antisaccade task reaction time and accuracy rate, Attention Network Test executive control score, and total score on the Attentional Control Scale). First-year college students at low or high risk for developing SAD completed assessments of attention bias, attentional control, and anxiety during their first month of college. Approximately four months later, they completed a social stressor task and the same self-report measures of social anxiety. At the end of their first year in college, they completed the self-report measures of social anxiety once more, as well as a diagnostic interview for SAD. Correlational analyses indicated that attention bias to threat on the MPDT was associated with concurrent self-reported social anxiety but did not prospectively predict psychological or biological social stress reactivity, self-reported social anxiety, or SAD diagnostic status at the end of the first year in college. Hierarchical regression analyses supported the hypothesized double moderation for concurrent social anxiety, such that high levels of attentional control weakened the association between attention bias toward threat and social anxiety, only among the individuals at high risk for SAD. However, analyses did not support this relationship in predicting prospective outcomes, and several unexpected patterns emerged in which interactions between attention bias and attentional control were observed to predict prospective outcomes, but only among individuals at low risk for developing SAD. Likewise, exploratory analyses using the PSCT index of attention bias revealed unexpected interactions between risk group, attention bias, and attentional control. Considered together, results of the current study highlight the importance of considering individual differences in attention bias and attentional control in the maintenance and development of SAD.