Does the way in which we perceive the world make us susceptible to anxiety?

Abstract: One major focus of anxiety research in recent years has been the identification of cognitive factors that promote increased vulnerability to anxiety. Cognitive formulations propose that anxiety is characterised by an increased tendency to attend to negatively valenced emotional information, and that this bias may play a causal role in the development and maintenance of clinical anxiety. Evidence suggests that this anxiety-linked processing bias occurs even in conditions in which the stimuli are masked in order to prevent awareness of the content (i.e., preattentive bias). The present thesis focused principally on the preferential processing of threat-related information that occurs outside awareness. Two different outcome measures were used to index preferential preattentive processing of threat-related information in non-clinical populations: The emotional Stroop task was used to index selective attention to masked presentation of threatening word stimuli, whereas skin conductance responses were used to index selective autonomic responses to masked presentation of threatening pictorial stimuli. The empirical studies in the present thesis showed that elevated levels of trait anxiety promote preferential preattentive processing of negatively valenced information, whereas elevated levels of social desirability (interpreted as defensiveness) generally prevent preferential preattentive processing of negatively valenced information, especially among those at higher levels of trait anxiety, irrespective of outcome measure used. Moreover, previous research has demonstrated that a person’s most common emotional reaction when encountering a stressful event is causally influenced by that person’s habitual tendency to selectively encode the most threatening aspects of all available information. Thus, preattentive bias (as measured with the emotional Stroop task) was used to predict the emotional responses (as seen on self-reported emotional distress and autonomic reactivity) following exposure to a laboratory stressor. This study showed that preattentive bias to negative information had significant effects on both self-reported and physiological measures in response to a laboratory stressor, but in diametrically opposite directions. Specifically, whereas preattentive bias was positively associated with self-reported negative emotional response, it was negatively associated with a physiological indicator of emotional response. The results were discussed in terms of an inability to automatically inhibit the processing of threatening cues, which seems to be a vulnerability marker for anxiety. Whether this bias is ultimately sufficient for the development of clinical anxiety remains to be examined in future research. Additionally, more information is needed before it can be established that the emotional Stroop task can be viewed as a reliable diagnostic tool for determining an individual’s anxiety status.