The present results are potentially relevant for
The present results are potentially relevant for understanding anxiety disorders and their treatment. Both COMT Val158Met (Lonsdorf and Kalisch, 2011, Montag et al., 2012) and neuroticism/anxiety (Clark et al., 1994, Mineka et al., 1998, Weinstock and Whisman, 2006) have been linked to the prevalence of anxiety disorders. Given the present effects exclusive for long-term fear extinction, COMT Val158Met and neuroticism/anxiety might primarily relate to maintenance and treatment resistance (Lonsdorf et al., 2010) which should be investigated in future studies. Moreover, future studies investigating mechanisms of anxiety disorders in subclinical samples may include fearfulness as a potential predictor for the acquisition of dysfunctional fear reactions. The contributions of neuroticism/anxiety and fearfulness vary between different anxiety disorders and fearfulness might be especially relevant for the development of phobias and panic disorder (Kampman et al., 2017, Lang et al., 2000, McNaughton and Corr, 2004), whereas neuroticism/anxiety may be more relevant for maintenance and treatment success. Some limitations should be addressed. First, we used a high number of trials to achieve a sufficient signal-to-noise ratio in ERP measures. This required the use of noise burst USs rather than electric shock US (Sperl et al., 2016), which are more commonly used in behavioral and neuroimaging fear conditioning research. This methodological difference should be kept in mind, when comparing the present results to other studies. Nevertheless, we evoked robust CRs on Day 1 and Day 2. Second, relevant subcortical structures like the amygdala (LeDoux, 2000) cannot be assessed with EEG. Future fMRI studies should therefore address the interplay of COMT Val158Met, neuroticism/anxiety and amygdala in extinction retention. Third, because sex and menstrual on average influence fear learning (Lebron-Milad et al., 2012, Merz et al., 2012) and dopamine (Risbrough et al., 2014) this initial study aimed to control for these factors by only testing males. Future studies will have to investigate interactions between sex and COMT Val158Met in human long-term fear extinction to probe the generalizability of the present findings to females. Fourth, there was an unexpected significant Genotype × Extinction interaction in fear bradycardia during Day 1 acquisition. This effect likely was due to a type I error, given that the factor Extinction had not been manipulated yet. Moreover, interpretation of Day 2 results is not affected as the relevant three-way interaction (Genotype × Contingency × Extinction) was not present on Day 1. Finally, it should be kept in mind that we assessed several independent (COMT Val158Met, fearfulness, neuroticism/anxiety, agentic extraversion) and dependent (LPP, fear bradycardia, SCR, affective self-reports) variables and reported different analytic approaches (averaging across all trials vs. only assessing the first or last ten trials). While this obviously resulted in a high number of tests which increases the likelihood of false positive results, we believe that such an approach is necessary, given that conditioned fear is a dynamic and multilevel phenomenon with potential relevance for various personality domains. It should further be noted that this study and the collected sample were a priori designed to primarily test very specific effects which turned out to be significant, namely the influence of COMT Val158Met on Day 2 cortical and cardiac fear responses. Nevertheless, future replications will be necessary. Here, power analyses suggested that relationships between personality and fear bradycardia as well as between COMT Val158Met and Day 2 fear bradycardia had sufficient statistical power in the present study and could be replicated using comparable or smaller sample sizes. Meanwhile, studies trying to replicate COMT Val158Met effects on Day 2 LPP and Day-1-Day-2 stability in fear bradycardia should have more power.