1 - (SYM 4) Factors Associated with the Extinction of Conditioned Threat Responses in Adults with Generalized Anxiety Disorder

Background: Individuals with generalized anxiety disorder (GAD) are known to experience elevated levels of chronic worry, as well as exaggerated threat and stress responses. Threat conditioning and extinction learning paradigms have been widely used in the lab to model how individuals learn to respond to threat and safety. Despite this, studies assessing how these processes unfold in GAD have yielded equivocal findings. Here, we examined how physiological arousal during threat and safety learning differ as a function of self-reported levels of worry and perceived stress in a sample of adults with a primary diagnosis of GAD.

Methods:   Baseline data was analyzed from an ongoing neuroimaging clinical trial of adults with GAD (N=88; Mage=33.66; female=81.6%). Assessments included worry (Penn State Worry Questionnaire) and perceived stress (Perceived Stress Scale). Participants underwent a threat-conditioning paradigm in which two lights (CS+s) were paired with a mild electric shock (US) while a third was never paired with shock (CS-). Participants then underwent extinction training, during which one CS+ and the CS- were presented with no US reinforcement. Mean square-root-transformed SCR responses to CS+, CS- and the slope of conditioned responding across acquisition and extinction were analyzed using separate multivariate regression models, adjusting for age, gender and shock intensity (voltage).

Results: Neither worry nor perceived stress yielded an effect on indices of threat acquisition (all p’s > .05). During extinction learning, we observed no effect of worry or perceived stress on average CS+ or CS- responses. However, the intensity of worry did show a significant effect on the slope of extinction learning (B=-.001; p=.013). Perceived stress yielded no effect on any indices of extinction learning (all p’s > .05).

Discussion: Our finding suggest that self-reported levels of worry and stress observed within a sample of adults with GAD do not affect the initial acquisition of conditioned threat responses, nor conditioned responses to cues during safety learning. Instead, we observed a selective effect of worry severity on the rate at which safety learning is acquired, with greater worry predicting steeper learning. Future work may seek to examine the relation between worry intensity and safety learning rates as they pertain to extinction retention, clinical outcomes and neuroimaging biomarkers.