Learning about threat from friends and strangers is equally effective: an fMRI study on observational fear conditioning

Abstract

Humans often benefit from social cues when learning about the world. For instance, learning about threats from others can save the individual from dangerous first-hand experiences. Familiarity is believed to increase the effectiveness of social learning, but it is not clear whether it plays a role in learning about threats. Using functional magnetic resonance imaging, we undertook a naturalistic approach and investigated whether there was a difference between observational fear learning from friends and strangers. Participants (observers) witnessed either their friends or strangers (demonstrators) receiving aversive (shock) stimuli paired with colored squares (observational learning stage). Subsequently, participants watched the same squares, but without receiving any shocks (direct-expression stage). We observed a similar pattern of brain activity in both groups of observers. Regions related to threat responses (amygdala, anterior insula, anterior cingulate cortex) and social perception (fusiform gyrus, posterior superior temporal sulcus) were activated during the observational phase, reflecting the fear contagion process. The anterior insula and anterior cingulate cortex were also activated during the subsequent stage, indicating the expression of learned threat. Because there were no differences between participants observing friends and strangers, we argue that social threat learning is independent of the level of familiarity with the demonstrator. HighlightsO_LIWe compared observational learning of fear from friends and strangers C_LIO_LIFamiliarity does not enhance social learning of fear in humans C_LIO_LIBayesian statistics confirm absence of differences between friends and strangers C_LIO_LIObservational fear learning activates social and fear networks including amygdala C_LIO_LIAmygdala activations are absent when learned fear is recalled C_LI