Predicting clinical outcomes via human fear conditioning: A narrative review

A common assumption in human fear conditioning research is that findings are informative for the etiology and treatment of clinical anxiety. One way to empirically evaluate the external validity of fear conditioning is by prospective studies. We review available prospective research investigating whether individual performance in fear conditioning predicts individual differences in anxiety levels and exposure-based treatment outcome. We focus on fear extinction, generalization, acquisition, and avoidance. Results suggest that reduced extinction and broader generalization predict higher anxiety levels. Results with respect to the predictive value of acquisition for anxiety levels are mixed. With regard to predicting exposure-based treatment outcome, some studies do find an association with extinction whereas others do not. The majority of studies does not find an association with acquisition. Evidence on extinction recall is limited and not consistent. The interpretation of these results requires caution. The number of available studies is limited. It is possible that not all work, in particular studies with only null effects, has found its way to publication. Future research on this topic will benefit from large sample sizes, preregistered hypotheses, full transparency about the conducted analyses and the publication of high-quality studies with null effects.