Safety behaviors and the persistence of irrational fears

Abstract

Cognitive-behavioral therapy (CBT), in particular, exposure therapy, is the treatment of first choice for anxiety disorders. Patients are repeatedly exposed to the feared, but innocuous stimulus to learn that the catastrophe that they expect does not follow, which typically decreases their fear response (i.e., extinction). Clinical guidelines recommend motivating patients to inhibit all safety behaviors during exposure. Safety behaviors are actions aimed at preventing, escaping, minimizing, or neutralizing a feared outcome. Patients may misattribute the nonoccurrence of the catastrophe to their safety behavior, which hampers exposure effects. However, empirical findings are inconsistent, and show detrimental, neutral, and beneficial effects of using safety behavior during exposure. Our first aim was to test whether the negative effects of safety behaviors on exposure outcomes depend on whether safety behavior precludes the occurrence of threat. A fear conditioning experiment showed that safety behavior that precluded the occurrence of threat prevented extinction, and that safety behavior that minimized the severity of threat allowed extinction for several participants, but prevented extinction for other participants. Additionally, we found that cleaning safety behavior, which logically precludes the occurrence of threat, did not prevent a decrease in threat beliefs about contamination and illness. Together these findings suggest that the negative effects of safety behavior on exposure outcomes do not only depend on whether safety behavior prevents the occurrence of threat. Secondly, we investigated the causal influence of safety behavior on anxiety. Healthy participants clinically representative checking safety behavior on a daily basis for one week. This exacerbated their cognitions about the severity of threat. Hence, checking behavior appears to contribute to the severity of fears in obsessive-compulsive disorder (OCD), which suggests that safety behavior contributes to the exacerbation of pathological anxiety symptoms. Furthermore, in a fear conditioning experiment, we found that using safety behavior after fear extinction can promote a return of fear. This suggests that safety behavior may be involved in relapse after exposure therapy. Third and finally, we explored the potentially beneficial effects of approach-enhancing safety behavior. We found that spider fearful individuals infer safety from approach behavior in spider-relevant scenarios. This suggests that approach itself may signal safety in anxious individuals. However, approach behavior did not add to the beneficial effects of exposure: repeated exposure to a spider by pulling it toward you and by having the experimenter pull it toward you caused similar reductions in spider fear. Safety behavior can maintain irrational fears by preventing extinction learning, contribute to the development of pathological anxiety by directly exacerbating threat beliefs, and may promote a return of fear after exposure therapy. However, safety behavior that allows threat occurrences may not hamper extinction for certain individuals, cleaning safety behavior may not be detrimental to the beneficial effects of exposure in the context of disgust, and safety behavior that motivates approach may directly increase safety perceptions. Pathological anxiety is a severe clinical and societal problem; therefore efforts to translate the current findings to clinical applications are warranted.