Abstract:

Fear over-generalization has been put forward as a potential etiological factor of anxiety disorders. Previous studies have examined the phenomenon of fear generalization among individuals with anxiety disorders and high trait anxiety. However, state anxiety is more common in populations and its impact on fear generalization has not been paid attention to. Thus, in this study, we induced state anxiety in healthy individuals and tested the impact of state anxiety on fear generalization. Thirty-eight healthy participants participated in the experiment and were randomly divided into an experimental group and a control group. The whole experiment consisted of four phases: habituation, acquisition, fear inducement, and generalization. In this experiment, 10 rings of gradually increasing size that were presented on a computer screen served as conditioned stimuli (CS) and generalization stimuli (GS). The rings in the two extreme sizes served as the conditioned danger cue (CS+) and conditioned safety cue (CS−), respectively. The eight intermediately sized rings served as four classes of generalization stimuli (i.e., GS1, GS2, GS3, and GS4), with GS4 being the most similar one to CS+ in size. CS+ was probably paired with an unconditioned stimulus (US), while CS− and GS were unpaired with US. Six fear pictures from the international affective picture system (IAPS) were taken as US. During the experiment, subject online expectancy ratings and skin conductance responses (SCR) were recorded. During the habituation phase, the CS+ and CS− were each presented three times, each without any pictures following the CS+. During the acquisition phase, the CS+ and the CS− were presented six times each, and the pictures were presented 5s before each CS+ offset. During the fear inducement phase, the experimental group passively viewed a 5-min fear video and the control group passively viewed a 5-min video clip of a train traveling. The subsequent generalization phase consisted of six blocks. In each block, eight GSs were presented once without a picture; CS+ and CS− were presented twice each. One CS+ was followed by a picture to avoid the participants forgetting. The results showed that exposure to the fear video significantly increased participants’ state anxiety. The experimental group displayed stronger generalization than control group. Conditioned fear in the experimental group was generalized to rings with up to GS4, GS3, and GS2 in both SCR and online expectancy ratings, whereas generalization in control group was restricted to rings with only GS4 in SCR or GS4 and GS3 in online expectancy ratings. The duration of generalization for the experimental group was longer than the control group in both SCR and online expectancy ratings, indicating that state anxiety slowed generalization extinction. Additionally, state anxiety enhanced the identification of conditioned stimuli in SCR. The results of this study supported both the behavioral inhibition and behavioral activation theories. Regarding online expectancy ratings, the experimental group displayed stronger generalization than the control group, indicating the participants with state anxiety fail to inhibit fear responses in the presence of safety signals (i.e. GS). Regarding SCR, participants with state anxiety showed both the stronger behavioral inhibition to safety signals (i.e. GS) and behavioral activation to danger cues (CS+). Hence, the results indicated that the theories of behavioral inhibition and behavioral activation might occur in the different learning phases: the former one might occur in both explicit and implicit learning and the latter one might occur only in the implicit learning of fear. This study also has clinical implications. For individuals suffering from negative events, to decrease the state anxiety may be an effective method for reducing the fear over-generalization and improving the efficacy of exposure therapy.

Key words: fear generalization, state anxiety, fear video paradigm