You are in: Home » Research at the Unit » Emotion research
Processing styles in anxiety
Past work over the last 20 years has established that vulnerability to anxiety states is associated with characteristic styles of processing emotional events. Specifically, people prone to anxiety attend more to cues having threat-related meanings (e.g. words like disease or failure) than do non-anxious individuals. Similarly, anxiety-prone people tend to perceive the more threatening meanings of ambiguous information (e.g. words like stroke or fit). In current research we are trying to better characterize the nature of these processing styles, the mechanisms that underlie them, and to find ways of controlling them. This latter aim is critical both for establishing the causal relationship between processing styles and emotion, and for exploring the therapeutic potential of modifying them.
Andrew Mathews --a former member of Emotion Group till Sept 2004 - and Bundy Mackintosh in work with Eamon Fulcher, have developed a model in which attention to threat cues is postulated to be the outcome of two opposed processes (see Figure below). One is due to output from a threat evaluation system which increases activation of threat-related representations (e.g. the extended amygdala), and the other due to top-down control processes (e.g. dorso-lateral cortex) that activate competing representations and thus inhibit emotional distracters (Mathews, Mackintosh & Fulcher,1997; Mathews & Mackintosh, 1998; Mackintosh & Mathews, Cognition & Emotion, in press). We are currently working on studies designed to test some of the assumptions underlying this model, using neuroimaging methods.
Attention to threat could arise due to enhanced engagement at its location, or because - once engaged - there is greater difficulty in disengaging from it. With Jenny Yiend -a former member of Emotion Group till Jan 2004 - and Elaine Fox from the University of Essex, we have recently shown that, once attention is drawn to a (picture) cue, anxiety prone individuals are indeed slower to disengage their attention from it in order to find a target elsewhere, but only when the cue has threatening attributes (e.g. Yiend & Mathews, 2000). We are currently exploring whether anxiety-prone individuals also have their attention more readily guided to a location of potential threat. A central cue, such as another person gazing to left or right, tends to direct attention of an observer to the same location. We have found that fearful rather than neutral gaze (see Figure below), enhances this guidance effect, but only in anxiety-prone individuals. Thus, even when no actual danger is present, anxiety is associated with attentional vigilance in locations of potential threat.
The form taken by emotional processing styles that favour threat meanings suggests that they could lead to the accumulation of information about potential dangers, and enhance the perception of personal risk. Despite the apparent plausibility of this causal hypothesis, until recently there has only been indirect correlational evidence to support it. In recent and continuing research (with Bundy Mackintosh, Jenny Yiend and Colin MacLeod) we have established that interpretations of emotionally ambiguous words or events can be biased in either a positive or negative direction by brief training methods. Thus, styles characteristic of either anxiety-prone or non-anxious populations can be induced experimentally. Furthermore, although such training does not necessarily or directly lead to emotional change, congruent changes in anxiety do occur when the induced bias is subsequently used deployed in processing mildly stressful events (see Mathews & MacLeod, 2002 and illustration below). Thus, for the first time, we have shown that emotional processing styles can be induced, and play a causal role in producing emotional changes.
In the next phase of this research, we are investigating how variations in the methods used to induce biases influence their emotional effects, how long they endure, and how much they generalize to new events: all factors of critical importance to possible therapeutic applications.