Better safe than sorry: Avoidance of interoceptive threat following observational fear learning

Learning through observation is often adaptive, allowing individuals to acquire important threat-related information without exposing themselves to danger. However, it can become maladaptive when it leads to excessive fear or avoidance in the absence of actual threat. This study examined whether observational learning of an interoceptive threat promotes avoidance behavior and how conditioned responding changes when avoidance is prevented. Forty healthy participants completed a three-phase observational fear conditioning paradigm. In the observational learning phase, participants watched a demonstrator experience interoceptive threat (US) paired with one conditioned stimulus (CS+), but not with another (CS-). In the following avoidance phase, participants were exposed to the same stimuli and could press a button to avoid the US. In the final, response prevention phase, this avoidance option was removed. Across all phases, participants never directly experienced the interoceptive threat (US). When participants were first directly exposed to the previously observed CSs, results showed differential conditioned responding, with greater SCR, startle eyeblink responses, subjective fear, and threat expectancy to the CS + than CS-. However, avoidance behavior did not reflect this differentiation, as both CS+ and CS- were avoided at comparable rates. During response prevention, conditioned responding to the CS+ was initially elevated compared to the CS-, but physiological and subjective responses attenuated over time. Together, these findings suggest that observational learning of interoceptive threat can result in conditioned responding upon later direct exposure and promote generalized, low-cost avoidance.