Fear generalization in humans: Impact of prior non-fearful experiences Bram Vervliet a, b, * , Merel Kindt a , Debora Vansteenwegen b , Dirk Hermans b a Department of Clinical Psychology, University of Amsterdam, Roetersstraat 15,1018 WB, Amsterdam, The Netherlands b Department of Psychology, University of Leuven, Belgium article info Article history: Received 15 December 2009 Received in revised form 23 June 2010 Accepted 8 July 2010 Keywords: Fear conditioning Latent inhibition Generalization Prevention Anxiety disorders abstract Fear generalization lies at the heart of many anxiety problems, and is therefore an important target for prevention and/or treatment. Here, we investigated whether fear generalization towards a specific stimulus can be weakened by prior non-fearful experiences with that stimulus. Using the standard human fear conditioning procedure, all participants received paired presentations of a geometric figure and an electric shock. This was followed by a test phase in which a similar but different figure was presented. Electrodermal responding and ratings of shock-expectancy measured the level of fear generalization towards this test stimulus. Crucially, half of the participants had been preexposed to that stimulus (without shock). The results show significantly less generalization in this group, suggesting that prior non-fearful experiences can protect against fear generalization. These results may inspire novel ways to prevent the development of clinical anxiety. Ó 2010 Published by Elsevier Ltd. Introduction Prevention is better than cure, but little is known about effective prevention in the case of clinical anxiety. The conditioning model of anxiety has inspired one well-known technique e “latent inhibi- tion” e that is aimed to reduce the rate of fear conditioning that can occur during a traumatic experience. This technique involves prior non-fearful experiences with the cues from the trauma context before the trauma has occurred. Animal conditioning research has shown that prior nonreinforced exposures to a to-be-conditioned stimulus delays the rate of fear conditioning with that stimulus, when it is later paired with an electric shock (“latent inhibition”, e.g., Lubow, 1973). However, it has proven cumbersome to replicate this finding in human experimental research; the phenomenon seems less robust and depends more on specific experimental parameters (Lubow, 1997; Lubow & Gerwitz, 1995; but see Vaitl & Lipp, 1997). Moreover, even animal research has shown that the latent inhibition effect is fragile: Changes in the perceptual features of the stimulus or the background context at the start of the conditioning phase seriously weaken the latent inhibition effect (Bouton, 1993; Lubow, 1973). Hence, the applicability of latent inhibition procedures is quite limited and probably only efficient in very stable contexts (e.g., for the prevention of dental phobia in children, Davey,1989; Ten berge, Veerkamp, & Hoogstraten, 2002). The latent inhibition technique starts from the basic assumption that excessive fear learning is responsible for the development of clinical anxiety: By weakening the fear learning process, one reduces the chances of developing clinical anxiety. In itself, however, fear learning is a highly adaptive mechanism that can be crucial for survival. The conditioning process underlying the fear mechanism produces fear selectively to those cues that provide unique information about the occurrence of the threatening event (Rescorla & Wagner, 1972). The trauma context is inherently dangerous for the trauma victim, and the experience of fear (and the resulting avoidance) is therefore adaptive in that context. However, fear becomes irrational when it is overgeneralized to everyday contexts/stimuli that are intrinsically safe but share (some) similarity with the trauma context. In the case of post- traumatic stress disorder (PTSD), for example, patients often show an unbridled generalization of fear towards realistically non- dangerous cues (see Feldner, Monson, & Friedman, 2007). This fear generalization is therefore an important target in some psycho- logical treatments of PTSD (e.g., Ehlers & Clark, 2000). Hence, the transfer from normal to abnormal anxiety may be better under- stood by exacerbated fear generalization to related cues or situa- tions (see also Lissek et al., 2005, 2008, 2009). This suggests that prevention strategies may benefit more from targeting fear generalization, rather than the fear learning itself. The question is how fear generalization might be reduced. A number of animal conditioning studies on contextual generalization provide interesting hints in this regard. These studies investigated the * Corresponding author at: Department of Psychology, University of Leuven, Roetersstraat 15, 1018 WB, Amsterdam, The Netherlands. Tel.: þ31 20 525 67 68; fax: þ31 20 639 13 69. E-mail address: b.vervliet@uva.nl (B. Vervliet). Contents lists available at ScienceDirect Behaviour Research and Therapy journal homepage: www.elsevier.com/locate/brat 0005-7967/$ e see front matter Ó 2010 Published by Elsevier Ltd. doi:10.1016/j.brat.2010.07.002 Behaviour Research and Therapy 48 (2010) 1078e1084