Journal Club Week 3: Mechanisms of attentional biases towards threat in anxiety disorders

This week's article was "Mechanisms of attentional biases towards threat in anxiety disorders: An integrative review" (Cisler & Koster, 2010).  The article can be found: Here 

The basics:

• Attentional bias towards threat = differential allocation of attention to threatening vs. non-threatening stimuli
• Compared to non-anxious individuals, people with anxiety disorders may display attentional bias towards threat
• Found in GAD, PTSD, social phobia, specific phobia, panic disorder, and OCD, with some mixed findings for OCD (within some studies, bias found in early trials but not later trials)
• Three possible components of attentional bias towards (or away from) threat:
• Facilitated attention = attention more readily drawn to threatening stimuli
• Difficulty in disengagement = once a threatening stimulus has been noticed, difficulty directing attention away from it
• Attentional avoidance = preferential attention away from threatening stimuli
• Experimental tasks include modified Stroop, dot probe, visual search, and spatial cueing (the latter three can differentiate among the three components)

There are numerous theoretical models of attentional allocation toward threat, but some of them were developed before the bulk of research was conducted, and none fully account for all components of the findings, so I won’t describe them all here, but I will note some interesting features or concepts that appear in one or more models. These are NOT necessarily facts, just ideas! Some of them contradict each other! I wanted to record them because they are interesting hypotheses.

• After perceiving the stimulus, there must be some mechanism for determining whether it is threatening and for determining where to allocate attention (an assumption shared among many of the models—seems pretty self-evident)
• State anxiety may change the sensitivity of either of those mechanisms
• Trait-anxious individuals may have a lower threshold for determining a stimulus is threatening (postulated by several models)
• Conversely, trait-anxious individuals may detect stimuli as equally threatening, but the mechanism for allocating attention may differ (or both)
• There may be multiple stages or types of processing at which the threat-value of a stimulus may be evaluated (e.g., automatic vs. deliberative)
• Some stimuli may be inherently threatening, while for others, threat value may be dependent upon other stored information (e.g., knowledge, goals)
• Trait-anxious individuals may have beliefs or motivations that lead them to perceive stimuli as more threatening
• Anxiety may affect executive function, preventing anxious individuals from pulling attention away from threat stimuli and/or potentiating shifting toward threat stimuli

Empirical findings:

• There is considerable evidence that individuals high in trait anxiety display difficulty disengaging attention from threatening stimuli (it looks like only supraliminal stimuli, but that’s not entirely clear); this may be due to poor attentional control mediated by lower PFC activity (i.e., impoverished top-down control)
• Evidence for facilitated attention toward threatening stimuli is more mixed, and may be moderated by stimulus duration or threat intensity (i.e., appear for briefer and more threatening stimuli); however, the presence of attentional biases and heightened amygdala activation to masked/subliminal stimuli does seem to suggest that there is greater threat reactivity at some early stage of information processing
• Several studies suggest that anxious individuals do deliberately direct their attention away from long-duration threatening stimuli, which may indicate that anxious individuals use distraction as an emotion regulation strategy (interestingly, one theory suggests that they may direct their attention/gaze away while remaining internally preoccupied)

Questions/future directions posed by the authors:

• To what degree are the threat detection and attentional control mechanisms interrelated (at a neural and functional level)?
• Attentional biases seem to reflect both automatic and strategic processing; to what degree are these interrelated?
• Incorporation of cognitive load tasks could help disentangle components of attentional bias by differentially affecting automatic vs. strategic processes
• How are attentional biases related to other anxious symptoms?
• Do attentional bias trainings work, and if so, how?

And here are some questions that came up for me as I was reading the paper. Some of them are more research questions than discussion questions, but I thought I’d include them in case any of you want to provide hypotheses or share any empirical work of which you’re aware that addresses these questions.

• To what degree is difficulty in disengagement responsible for rumination/worry? In other words, is attention preferentially allocated to internal (mental) threats? This would mean that worry-thoughts could arise in anxious and non-anxious people with comparable frequency, but that anxious people would continue to attend to them while others disengage.
• Do these findings leave any role for motivation/beliefs? Many therapies address anxious people’s apparent motivation to worry/attachment to that symptom (i.e., belief that it is adaptive).
• What interventions (other than the computerized attention trainings already being developed) might impact the components of attentional bias?
• How might exposure be expected to affect the components of attentional bias and why?
• How will knowing about attentional biases in anxiety disorders affect your clinical practice?
• Would there be any benefit for a client to know about these attentional biases, and if so, what might be the mechanisms of that benefit? 
• What is the ontogeny of threat bias? 
• Are there circumstances in which characterological bias towards threat could be genuinely adaptive?