Fear learning in unmedicated patients with anxiety disorders: a comparison of delay conditioning, fear reversal, and trace conditioning

Abstract

Anxiety disorders are common and impairing, yet their underlying mechanisms remain incompletely understood. Fear learning provides a critical translational framework for investigating pathological anxiety, bridging laboratory models and clinical phenomena. Prior studies have been limited by important methodological issues, including the inclusion of non-anxiety diagnoses, high comorbidity, and medication use. Here we examined three forms of fear learning— delay conditioning, fear reversal, and trace conditioning—in unmedicated adults with minimally comorbid primary anxiety disorders (Generalized Anxiety Disorder and Social Anxiety Disorder; n = 34) and demographically matched controls (n = 102). Individuals with anxiety disorders showed greater psychophysiological arousal (skin conductance responses) and reduced brain activation (assessed using functional magnetic resonance imaging) in the left dorsolateral prefrontal cortex to the learned safety cue (CS − ) during the early phase of delay conditioning. Differences between individuals with anxiety disorders and controls were not evident for the learned threat versus learned safety (CS+ versus CS − ) contrasts during delay conditioning, fear-reversal, or trace conditioning in psychophysiological arousal, brain activation, or subjective ratings. Taken together, these observations underscore the selectivity of Pavlovian learning deficits among unmedicated individuals with anxiety disorders and highlight differences in learning or using safety-related information to adaptively regulate fear.