Final Report Summary - TESTAAP (The Experimental Study of Threat-Avoidance in Anxiety Patients: Behavioral, Emotional, and Neural Correlates) Disabling forms of anxiety are a major problem for many people world-wide and come with high societal and financial costs. Important steps forward are being made within a global network of translational, trans-disciplinary and trans-diagnostic researchers. These pre-clinical research efforts focus on reducing exaggerated levels of fear, with considerable success. But, clinicians, clinical theorists and clinical researchers argue more and more that avoiding imagined threats is more debilitating than merely fearing those threats. Targeting avoidance may be more important than targeting fear. This research project uses the translational, trans-disciplinary and trans-diagnostic approach to enter the experimental study of (deviant) avoidance behaviors. For that purpose, we developed a new laboratory protocol to study threat-avoidance in human volunteers, and examined subjective, behavioral, physiological and neural correlates of avoidance. This will allow comparing these correlates between healthy individuals and anxiety patients, in order to better understand the mechanisms that push adaptive forms of avoidance into the excessive and uncontrollable forms of avoidance that characterize the anxiety disorders. The question is what drives avoidance and what causes it to derail. Fear is generally seen as the central motivator of avoidance behaviors: we avoid when we are afraid. However, research both in patients and in rodents has shown that levels of fear and avoidance do not always correlate: one can be very fearful and yet not avoid, or vice versa. In this project, we shift the focus towards the relief that follows avoidance, when harm is omitted. Relief is a positive emotion that accompanies the pleasant surprise triggered by unexpected omissions of harm. In line with reward learning theories, we hypothesize that such positive relief experiences can serve as a reward to reinforce foregoing avoidance actions. It is well known that anxiety patients tend to over-expect harm and under-expect omissions of harm. In the case of avoidance, this would yield increased relief experiences that then over-reinforce avoidance behaviors and culminate in pathological levels of avoidance. The core hypothesis of this project is that anxiety patients avoid excessively because they suffer from too much relief. The first objective of the project was to develop and behaviorally validate a novel protocol to study threat-avoidance and the role of relief. Healthy human volunteers first learned that pictures of two lamp colors predicted delivery of an annoying electrical stimulation to their fingers, while a third color did not. Next, they were given the opportunity to press a button when they wanted to avoid the stimulation. Unbeknownst to the volunteers, the button was effective only to one color, not to the other (and it was unnecessary to the third). We observed that volunteers gradually learned to press the button only when it was effective, indicating successful avoidance learning. We further observed that, as volunteers learned when the avoidance was effective, fear indices decreased (subjective ratings of harm expectancy, physiological recording of bodily arousal). In line with our hypothesis, we also observed increased positive relief ratings and bodily arousal during omissions of the aversive stimulation, which subsequently decreased as the omissions became less surprising to the volunteers. This pattern of results effectively validates the new protocol for the experimental study of threat-avoidance and the role of relief as a putative reinforcer. The second objective was to study neural correlates of threat-avoidance and relief, by scanning activity patterns in the brains of healthy volunteers while they engage in the avoidance protocol. We observed that lamp colors associated with the annoying shock triggered activation in fear-related areas in the brain (amygdala, insula), and that fear reductions during effective avoidance was accompanied by activations in a brain area that has been linked to fear regulation (ventromedial prefrontal cortex). Most importantly, omissions of the annoying shock triggered activations in reward-related areas in the brain, in line with the hypothesis that relief serves as a reward to reinforce foregoing avoidance actions. Moreover, activations in a key node in the reward circuitry predicted the persistence of avoidance during follow-up testing 24 hours later. These results establish the new threat-avoidance protocol as an experimental model for translational, trans-disciplinary and trans-diagnostic research into avoidance. The third objective was to investigate the mechanism of maladaptive avoidance by comparing anxiety patients to healthy controls within the newly developed avoidance protocol. For that purpose, we have composed a light version of the avoidance protocol that replaces the annoying electrical shock with a negative picture and that only focuses on the avoidance responses and subjective reports of threat expectancy. Although data collection is still ongoing, the available results already point to increased threat expectancies and increased avoidance frequencies in anxiety patients versus healthy controls. The fourth objective was to investigate the generalization of avoidance within the newly developed avoidance protocol. Overgeneralization of fear has been documented in anxiety patients, but little is known about overgeneralization of avoidance. For that purpose, we added a phase in which a series of morphed colors between the 3 lights (comprising effective, unproductive and unnecessary avoidance). We have now conducted a behavioral validation study in a large cohort of healthy individuals (N = 150) and with assessing various anxiety-related personality traits through questionnaires. Avoidance learning was successful, and generalized towards the unproductive and the unnecessary lights during test. Relief ratings followed the theoretical course of reward prediction error processing. With this project, we developed and characterized a novel avoidance conditioning protocol that will provide a stepping stone towards a better understanding of the mechanistic deficits that trigger excessive and uncontrollable avoidance in patients with pathological anxiety.
