Threat reversal learning and avoidance habits in generalised anxiety disorder

Avoidance and heightened responses to perceived threats are key features of anxiety disorders. These disorders are characterised by inflexibility in dynamically updating behavioural and physiological responses to aversively conditioned cues or environmental contexts which are no longer objectively threatening, often manifesting in perseverative avoidance. However, less is known about how anxiety disorders might differ in adjusting to threat and safety shifts in the environment or how idiosyncratic avoidance responses are learned and persist. Twenty-eight patients with generalised anxiety disorder (GAD), without DSM co-morbidities, and 27 matched healthy controls were administered two previously established paradigms: Pavlovian threat reversal and shock avoidance habits through overtraining (assessed following devaluation with measures of perseverative responding). For both tasks we used subjective report scales and skin conductance responses (SCR). In the Pavlovian threat reversal task, patients with GAD showed a significantly overall higher SCR as well as a reduced differential SCR response compared to controls in the early but not late reversal phase. During the test of habitual avoidance responding, GAD patients did not differ from controls in task performance, habitual active avoidance responses during devaluation, or corresponding SCR during trials, but showed a trend toward more abstract confirmatory subjective justifications for continued avoidance following the task. GAD patients exhibited significantly greater skin conductance responses to signals of threat than controls, but did not exhibit the major deficits in reversal and safety signal learning shown previously by patients with OCD. Moreover, this patient group, again unlike OCD patients, did not show evidence of altered active avoidance learning or enhanced instrumental avoidance habits. Overall, these findings indicate no deficits in instrumental active avoidance or persistent avoidance habits, despite enhanced responses to Pavlovian threat cues in GAD. They suggest that GAD is characterised by passive, and not excessively rigid, avoidance styles.