The Effect of Vicarious- and Direct Extinction Learning after Fear
Memory Activation in Individuals with Lifelong Snake Fear.
Josefien Mooij
10003569
Date: August 2017
Student number: 10003569
Supervisor: Armita Golkar
Abstract
A growing body of research has investigated the possibility of updating fear memory by disrupting the process of reconsolidation through extinction training. In order to disrupt reconsolidation, the fear memory must be successfully (re)activated. Whereas this approach has already been shown effective in experimental fear conditioning studies, the translation to clinically relevant settings is preliminary. Building on recent findings, suggesting that reconsolidation of phobic fear memories can be disrupted by extinction learning shortly after activation of a generic fear memory concept, the current study investigated the possibility of reducing long-term snake fear memories by activating snake fear memory prior to extinction. Moreover, because vicarious- compared to direct extinction has been demonstrated to more effectively attenuate conditioned fear, this study contrasted the efficacy of both extinction techniques in attenuating the expression of long-term fears and examined whether the fear reduction transferred to instrumental approach/avoidance behavior involving the feared stimuli. In this study, snake-fearful participants underwent vicarious- or direct extinction training, 10 min after fear memory activation. One day later, the participants were re-exposed to novel stimuli, after which they performed an approach/avoidance test. The current study did not show a reduction of fear expression from day 1 to day 2. Furthermore, no differences were found in fear expression or instrumental approach/avoidance behavior as a function of vicarious or direct extinction learning. Consequently, this study does not support the idea that extinction training initiated shortly after memory activation can disrupt the reconsolidation of phobic fear memory, and highlights the necessity to better understand the boundary conditions of reconsolidation in order to develop applicable techniques for treating phobic fears.
Keywords: Anxiety disorder, vicarious extinction, observational learning, memory activation,
Table of Content Introduction ... 4 Background ... 4 Current Study ... 7 Methods ... 8 Participants ... 8 Materials ... 8 Stimuli. ... 8 Physiological assessment. ... 9 Subjective assessment. ... 10 Behavioral assessment. ... 10 Procedure ... 10 General procedure. ... 10
Day 1: Activation stage and Extinction stage. ... 11
Day 2: Re-exposure stage and behavioral test. ... 12
Statistical Analyses ... 12 Results... 13 Demographic Characteristics ... 13 Activation Stage ... 14 Extinction Stage ... 15 Re-exposure Stage ... 16
Approach/Avoidance Behavior Test ... 18
Discussion ... 19
References... 25
Introduction
Background
Learning to fear is an adaptive process that exists across species. This process is essential for survival, as it helps an organism avoid harm through learning to predict danger (Öhman & Mineka, 2001). Such learning can be achieved by forming associations between a threatening event and a preceding neutral stimulus. This adaptive process, however, act sub-optimally when the fear persists and generalizes to novel stimuli and contexts, which subsequently are avoided. In this way, fear can become maladaptive and develop into pathological anxiety (Kindt, 2014). Anxiety disorders are the most common class of disorders with a lifetime prevalence of 28.8% (Kessler et al., 2005). Moreover, anxiety disorders are associated with moderate levels of disability, poor levels of vitality (Bijl & Ravelli, 2000) and economic costs (Olesen, Gustavsson, Svensson, Wittchen, & Jönsson, 2012). The leading experimental paradigm to study the mechanisms of associative fear learning and memory is Pavlovian fear conditioning. In this paradigm, subjects learn to associate an innocuous stimulus (conditioned stimulus; CS) with the occurrence of an aversive event (unconditioned stimulus; US), so that the CS subsequently predicts the US and elicits anticipatory fear reactions (conditioned reaction; CR) (LeDoux, 2000).
One of the most widely used interventions for treating anxiety disorders is Cognitive Behavioral Therapy (CBT; Barlow, 2002). A key component of CBT is exposure therapy. In exposure therapy a patient is repeatedly exposed to the feared and avoided situation without experiencing the feared outcome, resulting in fear reduction. Although this is an effective treatment, relapse rates among treated patients range from 19% to 62% (Craske & Mysthowski, 2006). The experimental procedure to investigate the principle underlying exposure therapy is fear extinction learning after fear conditioning. In extinction learning, the fear response (conditioned response, CR) is weakened by repeatedly presenting the CS without the US (Bouton & Swartzentruber, 1991). However, although the acquisition of fear generalizes to stimuli of the same category (Lissek et al., 2008), the effect of extinction is more restricted to the characteristics of a specific cue (Vervliet, Vansteenwegen, Baeyens, Hermans, & Eelen, 2005). Moreover, paralleling clinical findings of high relapse rates, many participants that undergo extinction show an eventual return of fear (Bouton, 2002). Four post-extinction phenomena that could account for the return of fear have been proposed by Bouton (2002); the recovery of the CR after re-exposure to the US (reinstatement); the recovery of the CR when the extinction context is changed (renewal); the recovery of the CR after lapse of time (spontaneous recovery); and the rapid recovery of the CR after the CS is paired with the US again (reacquisition). A possible explanation for these four
phenomena is that extinction does not destroy the original fear association, but that this learning results in a new “safe” memory trace (CS – no US association) that competes with the original CS-US association (Bouton, 2002), keeping the original fear memory intact.
However, promising research in animals has shown that the original fear memory trace can also be changed after reactivation of the fear memory (Nader, Schafe, & LeDoux, 2000; Lee, 2009). According to the reconsolidation hypothesis, conditioned fear memories return to a temporarily labile state after retrieval and require protein synthesis to re-stabilize into long-term memory (Nader, 2003). Subsequently, research in animals and humans has investigated the possibility to change consolidated fear memories during the time window of susceptibility, when the fear memory is malleable. Several pharmacological studies, using the Pavlovian conditioning paradigm in humans, have used the administration of the beta-adrenergic blocker propranolol, before or after reactivation of a conditioned fear memory to disrupt the reconsolidation of the fear memory during this reconsolidation time window (Kindt, 2014). These studies have shown that the administration of propranolol can disrupt the reconsolidation process and prevent the recovery of cue specific fear, but also the fear to stimuli from the same semantic category. This procedure has recently been shown to produce comparable effects on behavioral approach tendencies of spider fearful individuals (Soeter & Kindt, 2015). In addition to this pharmacological approach, alternative behavioral approaches that aim to target the reconsolidation process have been proposed. Schiller et al. (2010) showed that extinction training initiated within the reconsolidation time window (10 min after reactivation), but not outside this window of susceptibility (6 hours after reactivation), prevented the recovery of fear to conditioned, fear-irrelevant stimuli (i.e. colored squares). They conclude that memory activation before extinction enables one to “rewrite” the original memory trace of that cue, and that timing has an important role in this process. A similar effect has been found in “older” (7 day) fear conditioned memories (Steinfurth et al., 2014) and in fear-relevant conditioned stimuli (i.e. snakes/spiders, Thompson & Lipp, 2017). However, other replication attempts have failed to find the same effect, indicating that the results are still equivocal (Soeter & Kindt, 2011; Kindt & Soeter, 2013; Golkar, Bellander, Olsson, & Öhman, 2012).
Recently, Golkar, Tjaden and Kindt (2017) demonstrated that vicarious extinction learning within the reconsolidation window neutralized the expression of conditioned fear memory. Vicarious extinction learning is a form of safety modeling, during which subjects learn by observing the safety behavior of another individual. This form of modeling has long been exploited as a part of exposure treatment of phobias. In such treatment, the therapist interacts with the phobic stimulus before directly exposing the patient to it(Seligman & Wuyek, 2005). Experimental studies in humans have shown that vicarious extinction learning prevented the recovery of short-term
conditioned fear memory more efficiently than standard, direct extinction (Golkar, Selbing, Flygare, Öhman, & Olsson, 2013) and that the neural processes underlying vicarious extinction are partly distinct from what is observed following direct extinction (Golkar, Haaker, Selbing, & Olsson, 2016).
Although the Pavlovian conditioning paradigm has been useful in investigating extinction learning and reconsolidation, this experimental model has its limitations and does not take into account all factors that are involved in the pathogenesis of anxiety disorders (Beckers, Krypotos, Boddez, Effting, & Kindt, 2013). Most individuals that are subjected to experimental fear conditioning learn the induced fear association, whereas not everyone that experiences a fearful situation develops an anxiety disorder, (Beckers et al., 2013). Moreover, fear memories in individuals with anxiety disorders are not as simple as experimentally acquired fear memories, where only a single association is made between the stimulus, the CS, and the threatening event, the US (Soeter & Kindt, 2015). In addition, although pathological fear is broader than only a single association, previous studies investigating the effect of extinction within the reconsolidation time window have only focused on the cue-specific effects, thus the effects on this single association. It is necessary to evaluate whether these promising behavioral interventions, with fear memory activation preceding the different extinction learning techniques, are effective in not only diminishing experimentally induced fear memory, but also naturally occurring long-term fears.
This was recently addressed by Björkstrand et al. (2016), who compared the effect of extinction training, executed 10 minutes or 6 hours after fear memory activation (within or outside the reconsolidation window), in attenuating the neural expression of long-lasting spider fears. In a 2-day paradigm with activation and extinction performed on day 1 and re-exposure on day 2, spider fearful participants were exposed to four different spider pictures; one picture was presented during the activation stage (activation only), one during the extinction stage (extinction only), one during both the activation- and extinction stage (activation + extinction), and one novel picture was presented only during the critical re-exposure test stage (on day 2), during which fear memory was assessed for all four stimuli. The results showed less amygdala reactivity during the re-exposure test and more approach behavior to all spider stimuli in the 10 minute group, when compared to the 6 hours group, an effect that lasted over 6 months (Björkstrand et al., 2017). Moreover, amygdala responses to the activated and extinguished stimulus decreased from the end of extinction (day 1) to the beginning of re-exposure testing (day 2) in the 10 min group, but not in the 6 hour group. Taken together, these results were interpreted to reflect a disruption of reconsolidation of a generic fear memory concept, rather than the disruption of a cue-specific representation, because the fear attenuating effects during re-exposure in the 10 min group generalized to all four stimuli. These
results are very promising in suggesting a possible clinical utility of adapting the reconsolidation paradigm to modify naturally occurring fears. However, some outstanding questions remain. First, although Björskstrand et al., (2016) assumed that the procedure activated a generic fear memory concept, the decrease in fear expression from end of extinction to the beginning of re-exposure was cue-specific as they only reported this analysis for the activated and extinguished spider picture. Second, as the effect of treatment is usually indicated by a reduction of fear from pre- to post treatment (e.g. Öst, Brandberg, & Alm, 1997), the fear reduction should also be evident from pre- to post extinction training rather than from end extinction to the re-exposure test.
Current Study
The current study investigated the possibility to activate and disrupt long-term fear memory in snake fearful individuals, using a similar paradigm as Björkstrand et al. (2016). To test the assumption that this procedure can activate a generic fear memory concept, the study employed a 2 day paradigm (activation and extinction on day 1 and a re-exposure test on day 2), critically using novel snake stimuli throughout all stages of the experiment, and assessing the reduction of fear from pre- to post extinction training 24 hours later. Furthermore, as previous research has shown that both vicarious - and direct extinction learning initiated within the reconsolidation window can effectively prevent the expression of conditioned fear memories (Schiller et al., 2010; Golkar, Tjaden, & Kindt, 2017), we directly contrasted the efficacy of vicarious and direct extinction after fear memory activation. Fear expression was measured with the fear potentiated startle (FPS) response, which is one of the most reliable measures of fear responding (Lang, Davis, & Öhman, 2000). Moreover, as behavioral avoidance is one of the diagnostic criteria for phobias (American Psychiatric Association, 2013) and contributes to the maintenance of phobias (Barlow, 2002), the proportion of approach-avoidance behavior was measured with a behavioral test, similar to the test reported in Björkstrand et al. (2016).
Assuming that the memory activation procedure successfully activated and disrupted the reconsolidation of a generic fear memory concept, it was hypothesized that extinction training after memory activation would reduce fear expression from pre-extinction to post-extinction training 24 hours later. Furthermore, based on the findings of Golkar, Selbing, Flygare, Öhman and Olsson (2013), showing the superior effect of vicarious extinction over direct extinction in attenuating conditioned short term fear, the second hypothesis was that after fear memory activation, vicarious extinction learning would attenuate the expression of long-term snake fear more effectively than direct extinction learning. Third, it was hypothesized that the superior effect of vicarious extinction learning would translate to increased approach behavior to snake stimuli, compared to direct extinction learning. In addition, it was explored whether the propensity to learn through vicarious
extinction was related to the level of social anxiety, as individuals high in social anxiety show more self-focused attention, which impairs the possibility to learn from another person and enhances the self-awareness of anxious feelings (Clark & Wells, 1995).
Methods
Participants
Participants were recruited through an online advertisement on the University of Amsterdam (UvA) recruitment website. Inclusion criteria were; fear of snakes, assessed by the snake anxiety questionnaire (SNAQ), with a cut-off score of 16 for females and 13 for males (Klorman, Weerts, Hastings, Melamed, & Lang, 1974), no past or present psychiatric conditions as assessed by report, and aged between 18 and 35 years. Additional exclusion criteria were self-reported use of antipsychotic, anti-depressive or anxiolytic medication. A total of 233 undergraduate students filled in the SNAQ online. From these subjects, 119 subjects had an eligible fear of snakes and were contacted to undergo a medical screening and book an appointment. Fifty-two participants dropped out because they were not longer interested to participate or were not reachable (42), did not pass the medical screening (6), or were not in the correct age range (4). One participant was excluded because of absence on the second day. Consequently, the final analyzed sample consisted of 66 participants. The participants were randomly assigned to either the vicarious- or the direct group, controlling for their fear of snakes. The vicarious group consisted of 32 participants (23 female, mean age = 22.56 years, SD = 2.51) and the direct group consisted of 34 participants (25 female, M age = 22.91 years, SD = 3.26). The participants either received course credit or were paid €40. The study was approved by the ethical committee of the UvA and informed consent was obtained from all participants.
Materials Stimuli.
Two different pictures of snakes were used to activate the generic fear memory concept during the activation stage (S1 and S2) and another snake picture was used to extinguish the fear memory during the extinction stage (S3). The order of these three stimuli was counterbalanced between participants. To assess the expression of snake fear one day after extinction and to assess if fear reduction of the extinction stimulus generalized to similar stimuli, three novel snake pictures (S4, S5, S6) were used during the re-exposure test stage. In addition, two different pictures of mushrooms served as neutral control stimuli during activation. All snake and mushroom stimuli
were obtained from Peira, Golkar, Öhman, Anders and Wiens (2012). The extinction stage consisted of a video-clip with either a calm-looking individual (the learning model), watching repeated presentations of a snake picture (S3) on a computer screen (vicarious group), or a similar video-clip with just the computer screen depicting the same presentation of the snake picture (direct group). The stimuli were always presented for 8 seconds, and a black screen was shown during the inter-trial-intervals (ITIs) for either 15, 20 or 25 seconds, randomized with even probability. During the activation - and re-exposure stage, all stimuli were presented in a pseudo-randomized order with the first picture counterbalanced across participants. The principle of randomization was that a picture from the same category (during activation: snake, mushroom) or of the same sort (during re-exposure: S3, S4, S5) was never presented more than once in a row. In extinction, the same snake stimulus was shown repeatedly, alternated with the NA trials. The pictures, 200 mm high and 270 mm wide, were depicted in the middle of a black background of a 19 inch computer monitor. The screen resolution was 800 × 600 pixels and the refresh rate was set to 60 Hertz. The whole experiment was programmed in Presentation 13.1 (Neurobehavioral Systems, www.neurobs.com).
The FPS response was elicited by a burst of noise (the startle probe) presented 6 seconds after the onset of the visual presentation of all stimuli and during an equal number of ITIs (i.e. Noise Alone; NA trials), that served as a reference blink (Blumenthal et al., 2005). This resulted in 2 NAs during activation, 10 NAs during extinction and 4 NAs during re-exposure. The startle probes had a 40 milliseconds bursts with a volume of 104 decibel, binaurally presented through headphones (Model MD-4600; Compact Disc Digital Audio, Monacor).
Physiological assessment.
The eye-blink startle reflex was measured through electromyography (EMG) recordings of the right orbicularis oculi muscle. For this assessment, two electrodes were placed underneath the right eye; 1 cm from the pupil and 1 cm below the lateral canthus. One ground electrode was placed on the forehead 1 cm below the hairline (Blumenthal et al., 2005). The electrodes were 7 mm Ag/AgCl electrodes, filled with an electrolyte gel (Signa, Parker). The electrodes were connected to a front-end amplifier with an input resistance of 10 MΩ and a bandwidth of DC-1500 Hertz. A notch filter was set at 50 Hertz to remove the unwanted interference. The integration was done by a true-RMS converter (contour follower), using a time constant of 25 milliseconds. The EMG signal was sampled at 1000 Hertz, filtered through a 28-500 Hertz bandpass. FPS response was measured as the peak amplitude levels from 50 to 200 milliseconds after the startle probe presentation. This raw score was normalized using a t-standardization, which gave a normal distribution with an overall mean of 50 and a standard deviation of 10 for each participant.
Subjective assessment.
The severity of snake anxiety was assessed with the Snake Anxiety Questionnaire (SNAQ; Klorman et al., 1974), a 30-item true or false questionnaire assessing feelings and avoidance behavior towards snakes. The SNAQ is a valid questionnaire and is reliable (α=.78; Klorman et al., 1974). Trait anxiety was measured with the Dutch version of the Trait Anxiety Inventory (STAI-T; Spielberger, Gorsuch, & Lusthene, 1970); de Zelf-Beoordelings Vragenlijst (ZBV; van der Ploeg, Defares, & Spielberger, 1981). The second part of this questionnaire was used to assess trait anxiety, consisting of 20 items that are answered on a four-point scale. This inventory has a high validity and a documented reliability of α=.94 (van der Bij, de Weerd, Steegers, & Braspenning, 2003). The social anxiety state was measured with the Liebowitz social anxiety scale (LSAS, Liebowitz, 1987). The scale consists of 24 social interaction- and performance situations for which participants had to rate their fear and avoidance during the past week on a 4-point Likert scale. The LSAS has shown good validity and high reliability (α=.96, Heimberg et al., 1999).
Behavioral assessment.
The approach-avoidance behavior towards the snake stimuli was measured with an approach-avoidance test, similar to the test applied by Björkstrand and colleagues (2016). Participants had to choose between two colored squares on the computer screen, followed by the presentation of either a snake stimulus or a mushroom stimulus, depending on the square they chose. Choosing to watch a snake was most of the trials associated with a monetary reward, whereas choosing to watch a mushroom was never rewarded. The monetary reward was equally divided in five reward levels (levels: € 0, 0.05, 0.10, 0.20 or 0.50 cent), with 4 trials for each reward level. The stimuli were shown on the computer screen for 6 seconds. Between each trial, participants saw a black fixation cross on a white screen for 6 seconds. The test consisted of 20 trials, showing 5 different snake- or mushroom stimuli, depending on what the participants choose to watch (all novel stimuli).
Procedure
General procedure.
The experiment was carried out on two consecutive days, separated with 24 hours (see Figure 1). First, applicants filled in an online version of the Snake Anxiety Questionnaire (SNAQ; Klorman et al., 1974), which they obtained through a Qualtrics link. Consequently, applicants who were eligible for the study were contacted by telephone for a medical screening and to schedule the two sessions. The experiment was always conducted in the afternoon and both sessions were conducted around the same time. The experiment took place in a sound-attenuated room and was conducted on a computer. The computer was placed from a distance of about 1 meter from the
participants. The light was turned off during all stages and the behavioral test. Both days started with a habituation phase consisting of ten NA trials, to reduce initial startle reactivity (Blumenthal et al., 2005). On both days, the participants were instructed to look at the stimuli on the screen during the whole experiment. It was emphasized that they should focus all their attention to the task and to sit as still as possible. Further, they were informed that they could leave the experiment on any moment by pressing the intercom button.
Throughout both days, five cortisol samples were taken from each participant. Three samples on the first day, before and after activation, and after extinction. On the second day, two samples were collected before and after re-exposure. Participants were asked to gently chew on a Salivette for approximately 1 minute per sample. The cortisol data will not be reported herein.
Figure 1. The design of the experiment. This is an exemplar of the stimuli presentations for one
participant. For counterbalancing and randomization, see methods. S1 = snake picture 1, S2 = snake picture 2, S3 = snake picture 3, S4 = snake picture 4, S5 = snake picture 5, S6 = snake picture 6, Mush1 = mushroom picture 1, Mush2 = mushroom picture 2.
Day 1: Activation stage and Extinction stage.
At the beginning of the first day, participants read the information brochure, signed the informed consent and went through the medical screening. Furthermore, the participants filled in the STAI-T (Spielberger et al., 1970), to assess the amount of trait anxiety, and the LSAS (Liebowitz, 1987), to assess the social anxiety state. After filling in the self-report questionnaires, the participants were placed by the computer table, where they were attached to the electrodes.
The participants were informed that they would see stimuli of snakes and mushrooms on the screen and would occasionally hear a loud noise (the startle probe) throughout the whole experiment.
The experiment started with the memory activation stage. Participants were exposed to two snake pictures (S1 and S2), to activate the generic snake fear memory concept, two mushroom pictures that served as neutral controls, and two NA trials. After the activation stage, there was a 10 minute break during which participants could read magazines. After the break the extinction stage started, in which participants that had been assigned to the vicarious group watched video-clips with a calm-looking individual (the learning model), watching 10 presentations of a snake picture (S3) on a computer screen and participants that had been assigned to the direct group watched video clips containing the same 10 presentations of the snake picture only. In the vicarious group, participants were informed that the video clip depicted another person undergoing a similar experiment as they did. The participants were instructed to look at the learning model in the video clip, as well as at the stimuli. The direct group received similar instructions, but without information of the learning model. After the presentation, participants in the vicarious group were asked to rate how anxious they perceived the learning model, on a scale from 1 to 9.
Day 2: Re-exposure stage and behavioral test.
After attachment to the electrodes, the participants received instructions about the second day (similar instructions as day 1). During the re-exposure stage, participants were exposed to three novel snake pictures (S4, S5, S6) presented four times each, to assess the generalization to similar fear-eliciting stimuli. Following the re-exposure test, participants received instructions for the approach-avoidance test and were asked to start when the experimenter had dimmed the light and left the room. As FPS responses were not measured during this test, they could leave their headphones off. In conclusion, the electrodes were removed, after which the participants were informed about the experiment and asked not to speak with other students about the content of the experiment.
Statistical Analyses
The analyses were conducted with SPSS 24.0 for Windows. For the Activation and Extinction stage, two separate mixed analyses of variance (ANOVAs) were carried out with Stimulus (Snake vs. NA) and Block (mean FPS response of two consecutive trials) as within-subject factors, and Group (Vicarious vs. Direct) as a between-subjects factor. In addition, the mean FPS response to the two mushroom stimuli were included as a Stimulus level when analyzing the Activation stage data. For the Re-exposure stage, a mixed ANOVA was used with Stimulus (Snake vs. NA) as within-subject factor, and Group (Vicarious vs. Direct) as between-subjects factor, using the FPS response to the first trial (e.g. Schiller et al., 2010). The reduction in fear expression from
day 1 to day 2 (pre-extinction to post-extinction 24 hours later) was measured with a mixed ANOVA with Stimulus (Snake vs. NA) and Stage (Activation vs. Re-exposure) as within-subject factors and Group (Vicarious vs. Direct) as between-subjects factor, using the FPS response to the first trials of both stages. Thus through all stages, fear expression was analyzed as the mean startle response to the startle probe in the presence of the snake stimuli relative to the mean startle response to the startle probe alone (NA). For the approach avoidance test, an ANOVA was used with Reward level (1-5) as the within-subject factor and Group (Vicarious vs. Direct) as between-subjects factor. The within-subject factor “Reward level” consisted of 5 monetary reward levels, with the mean probability of a participant choosing for the snake stimulus over a mushroom stimulus for each monetary amount separately. Follow-up two tailed t-test were computed for all significant interaction effects and pre-planned comparisons. Finally, an exploratory analysis assessed if social anxiety scores co-varied with fear expression and avoidance behavior. This was explored with a bivariate Pearson’s correlation between the LSAS score and the snake potentiation (the difference between the startle magnitude to the startle probe in the presence of a snake picture and NA) of the first trial during the Re-exposure stage and the mean probability to approach a snake picture in the behavioral test. All FPS responses above or below 3 standard deviations from the mean were excluded from the analysis (Blumenthal et al., 2015). Consequently, all missing data were replaced by the linear trend for that data point, the stimuli and stages separated (Soeter & Kindt, 2015). For all analyses, a significance level (α) of .05 was used and partial eta-squared (ηp2)
were reported as the estimate of effect sizes. When the sphericity assumption (Mauchly’s test) was not met, Greenhouse-Geisser adjustments of degrees of freedom were used.
Results
Demographic Characteristics
An independent t-test was used to test whether there was a difference in demographic characteristics between groups (for means and standard deviations per group, see Table 1). The groups did not significantly differ as regards to gender (t(64) = .15, p = .88), age (t(64) = -.49, p = .63), SNAQ score (t(64) = .12, p = .91), STAI-T score (t(64) = .62, p = .54) and LSAS score (t(64) = -.29, p = .78).
Table 1. Mean Values and Standard Deviations of the Demographic Characteristics for the Vicarious- and Direct Group Vicarious (n = 32) Direct (n = 34) M SD M SD Age 22.56 2.51 22.91 3.26 SNAQ score 20.06 4.45 19.94 3.83 STAI-T score 36.75 9.34 35.47 7.27 LSAS score a 29.44 17.84 30.64 16.07 Fear of model 3.19 0.36 NA NA
Note: NA, not applicable.
a One participant in the direct group did not complete the LSAS questionnaire.
Activation Stage
To compare the mean FPS responses during the activation stage, a 3 (Stimulus) x 2 (Group) repeated measure ANOVA was used. This analysis revealed a significant main effect of Stimulus (F(2, 128) = 8.85, p < .001, ηp2 = .12), and a significant Stimulus x Group interaction (F(2, 128) = 3.62, p = .03, ηp2 = .05). Follow-up t-tests confirmed that the startle potentiation to snake stimuli was significant in both the vicarious and direct group (snake vs. NA; t(31) = 2.47, p = .02 and t(31) = 2.94, p = .01, respectively) (see Figure 2), and there were no between-group differences in response to either the snake or the NA (vicarious vs. direct; t(64) = 1.13, p = .26 and t(64) = 1.47,
p = .15, respectively). The Stimulus x Group interaction was driven by an unpredicted group difference in FPS response to the mushroom stimuli. Thus, a follow-up t-test showed significantly stronger startle responses to mushroom stimuli in the direct group compared to that in the vicarious group (vicarious vs. direct; t(64) = -2.89, p = .01).
To control for this between-group difference in startle response to the mushroom, we additionally ran all following reported analysis including the mean startle response to the mushroom stimuli as a covariate. Including this covariate did not significantly alter any of the reported results (see Appendix A).
Figure 2. Mean startle response during Activation to the snake stimuli (S1 and S2) and the NA as a
function of groups. Error bars represent standard error of the mean (SEM). *p < .05, **p < .01.
Extinction Stage
To assess whether FPS responses towards the snake stimulus decreased from beginning to the end of the extinction stage, a 2 (Stimulus) x 5 (Block) x 2 (Group) repeated measures ANOVA was used (see Figure 3). This revealed a significant main effect of Stimulus, indicating overall higher startle potentiation to the snake stimuli (snake vs. NA: F(1, 64) = 23.49, p < .001, ηp2 = .27) and a significant main effect of Block, indicating an overall decrease in FPS responses during extinction (F(4, 256) = 34.22, p < .001, ηp2 = .35). Neither of these effects differed significantly between groups (Block x Group: F(4, 256) = 1.00, p = .41; Stimulus x Group: F(1, 64) = .33, p = .57). Finally, a 2 (Stimulus) x 2 (Group) repeated measures ANOVA during the last extinction block confirmed that there was no significant startle potentiation left to the snake stimulus at the end of day 1 (Stimulus: F(1, 64) = 2.51, p = .12), and no significant between-group differences (Stimulus x Condition: F(1, 64) = .10, p = .76).
Figure 3. Mean startle responses during Extinction to the snake stimuli (S3) and the NAs. For better
visualization, the groups are plotted separately. Error bars represent standard error of the mean (SEM).
Re-exposure Stage
To assess whether the startle potentiation to snake stimuli differed between the vicarious and direct group one day after extinction learning, a 2 (Stimulus) x 2 (Group) repeated measures ANOVA was used. This analysis resulted in a main effect of Stimulus, showing a significant startle potentiation to the snake stimuli one day after extinction learning (snake vs. NA: F(1, 64) = 6.89,
p =.01, ηp2 = .10). In contrast to the hypothesis, there were no differences between groups (Stimulus x Group; F(1, 64) = .05, p = .82). Analyzing the mean FPS responses across the four blocks of the re-exposure test revealed a similar pattern of results. Thus, a 2 (Stimulus) x 4 (Block) x 2 (Group) repeated measures ANOVA resulted in a significant main effect of Stimulus (F(1, 64) = 14.50, p < .001, ηp2 = .19), indicating a stronger startle potentiation to snake stimuli vs. NA, and
a main effect of Block (F(3, 192) = 20.85, p < .001, ηp2 = .25), showing an overall decrease in startle
potentiation, in the absence of significant between-group differences (Stimulus x Group: F(1, 64) = 1.18, p = .28; Block x Group: F(3, 192) = .48, p = .70).
Critically, to assess if the extinction procedures after memory activation significantly reduced fear expression to the snake stimuli from pre-extinction to post-extinction training, we compared the FPS response during the activation stage (first trial) with the FPS response during
the re-exposure stage (first trial). This 2 (Snake vs. NA) x 2 (Activation vs. Re-exposure) x 2 (Vicarious vs. Direct) repeated measures ANOVA did not show a reduction in startle potentiation to the snake stimuli from Activation to Re-exposure, as shown by the non-significant Stimulus x Stage interaction (F(1, 64) = .23, p = .63). Thus, in the absence of any between-group differences (Stimulus x Group: F(1, 64) = .05, p = .83; Stage x Group: F(1, 64) = 2.07, p = .16), snake potentiation during the activation stage did not significantly differ from snake potentiation in the re-exposure stage in either the vicarious and the direct group (activation vs. re-exposure: t(31) = .31, p = .76, t(33) = .37, p = .71, respectively) (see Figure 4). As previous research (i.e. Björkstrand et al., 2016) has focused on the change in fear expression from end of extinction to beginning of re-exposure, we additionally assessed this change in a 2 (Snake vs. NA) x 2 (last Extinction trial vs. first Re-exposure trial) x 2 (Vicarious vs. Direct) repeated measures ANOVA. This analysis gave a similar result, with a non-significant Stimulus x Stage interaction (F(1, 64) = .78, p = .38).
Figure 4. Snake potentiation, the difference between startle potentiation to the snake stimulus and
startle potentiation to the NA, as a function of group and stage (last trial of Extinction and first trial of Re-exposure). Error bars represent standard error of the mean (SEM).
Approach/Avoidance Behavior Test
The difference in approach behavior to snake stimuli between the two groups was measured with a 5 (Reward level) x 2 (Group) repeated measures ANOVA (see Figure 5). This ANOVA revealed a significant main effect for Reward level, confirming the increase in approach behavior when monetary gain increased (F(2.72, 256) = 116.04, p <.001, ηp2 = .65). However, in contrast to the third hypothesis, there were no differences between groups (Reward level x Group: (F(2.72, 256) = .09, p = .96).
Figure 5. Mean probability to approach a snake stimulus for each monetary reward level (€ 0, 0.05, 0.10, 0.20 or 0.50 cent) during the approach/avoidance behavior test as a function of group. Error bars represent standard error of the mean (SEM).
Social Anxiety
Exploratory Pearson’s correlations were carried out to measure the correlation between the social anxiety score and both the first snake potentiation during re-exposure, and the mean approach behavior. None of these analyses resulted in a significant correlation (see table 2). However, a trend towards a negative correlation between the LSAS score and the first snake potentiation was found during re-exposure in the vicarious group (r = -.32, p = .07), which was not found in the direct group (r = -.17, p = .44). But importantly, the correlation coefficient of the vicarious group did not significantly differ from that of the direct group, as shown by the
non-significant Fisher r-to-z transformation (z = -.95, p = .34), therefore not supporting that this correlation differed between groups.
Table 2. Pearson’s Correlations between LSAS Score, Snake Potentiation and Mean Approach Behavior LSAS score
Vicarious (n=32)
Snake potentiation -.33, p =.07 Mean approach behavior .19, p =.29
Direct (n=33)
Snake potentiation -.09, p =.62 Mean approach behavior .19, p =.28
Notes. Snake potentiation = first presentation on day 2. Mean approach behavior = mean
probability to approach a snake stimulus over 20 trials.
Discussion
In the present study, we directly compared vicarious - and direct extinction learning initiated 10 min after fear memory activation in a sub-clinical population of snake fearful participants. In this two-day memory activation paradigm, no overall reduction of physiological fear expression was shown from day 1 (activation and extinction) to day 2 (re-exposure test), and there were no significant between-group differences in either fear expression or instrumental approach/avoidance behavior as a function of direct or vicarious extinction learning. Collectively, these results do not support the notion that extinction training initiated shortly (10 min) after memory activation can disrupt the reconsolidation of phobic fear memory.
Previous findings have shown that both direct (e.g. Schiller et al., 2010) and vicarious extinction (Golkar, Tjaden, & Kindt, 2017) following reactivation of a fear conditioned memory efficiently prevented the expression of a cue-specific fear memory 24 hours after extinction. These experiments have however been carried out in non-fearful participants, in which fear memories are experimentally induced during a standard fear conditioning session (i.e. fear acquisition), 24 hours before the fear conditioned memory is reactivated and extinguished. More recently, Björkstrand et al. (2016) extended these finding to individuals with long-lasting spider fear, and reported that extinction training initiated shortly (10 min) after memory activation successfully attenuated the neural expression of the spider fear. In particular, they reported that the effects of the 10 min post-activation extinction procedure generalized to novel spider stimuli, suggesting that the procedure
disrupted the reconsolidation of a generic fear memory concept in these spider phobic individuals. As disruption of reconsolidation of experimentally induced fear memories has so far only focused on cue-specific effects, one major aim of the current study was to directly assess the efficacy of different post-activation extinction procedures (direct and vicarious extinction) in disrupting a generic fear memory concept in individuals with a pre-existing, long-term fear of snakes. To assess this, successful fear memory disruption was indexed as a significant reduction in fear expression from pre-extinction levels of fear on day 1 (first trial of memory activation) to post-extinction training levels of fear on day 2 (first trial of the re-exposure test) using novel snake stimuli throughout activation, extinction and re-exposure. However, neither vicarious nor direct extinction 10 min after memory activation resulted in a significant reduction of fear from the first to the second day of testing. This failure to show an overall reduction of fear does not support that this procedure can disrupt a generic fear memory concept of fear in phobic individuals, as suggested by Björkstrand et al. (2016, 2017). There are however several methodological differences between the present study and that of Björkstrand et al., (2016). These differences include the use of fear-potentiated startle and the presentation of a neutral condition during the activation stage in the present study, vs. the use of BOLD-signal and the presentation of a neutral condition during the extinction stage in the study of Björkstrand et al. (2016). It is unclear how and if these differences contribute to the discrepancy in results. Perhaps most relevant is the use of only novel stimuli during all the stages of the present experiment (activation, extinction, re-exposure test). This rationale was based on the assumption derived from Björkstrand et al. (2016); if the memory activation served to activate a generic fear memory concept, then the fear reducing effects of the procedure would generalize to novel stimuli. Notably however, although Björkstrand et al. (2016) did report a cue-specific disruption of fear memory for the activated and extinguished fear cue, they did not report if fear responses were similarly attenuated from day 1 to day 2 in response to novel stimuli only presented on day 2. Because the current paradigm presented only novel stimuli for each stage, it is possible that not re-presenting the memory activation stimulus in this context resulted in insufficient generalization to other stimuli. The utility of the current design could be further investigated by directly comparing groups in which the cue used for memory activation is either present or absent during re-exposure. Finally, whereas the present study focus on the change in fear expression from pre- (activation day 1) to post-extinction (re-exposure day 2), the focus in the study of Björsktrand et al. (2016) is the change in cue specific fear responses from the end of extinction on day 1 to the beginning of re-exposure on day 2. Importantly, a similar analysis of the present data did still not find support for a significant reduction in fear expression.
Moreover, this study directly compared the effect of vicarious- and direct extinction within this design. Contrary to our hypothesis, vicarious extinction training did not reduce fear expression more than direct extinction training. This hypothesis was based on a previous fear conditioning study reporting that vicarious extinction prevented the recovery of short-term fear more efficiently than did direct extinction (Golkar et al., 2013). However, due to a technical limitation, there was a methodological difference in how the video stimuli was presented in the current study compared to the procedures used in previous vicarious extinction studies (Golkar et al., 2013, 2015, 2016, 2017). In the current study, participants were exposed to brief (8s) film-clips depicting the model viewing the stimuli interrupted by an altering black screen in between stimulus presentations, whereas previous studies have used a continuous film sequence of the model viewing repeated presentations of the stimuli without interruption. Consequently, brief film clips used in this study might have been experienced as less engaging and vivid, and recruited slightly different processes that the continuing films previously used. Follow-up research should directly compare the efficacy of these two vicarious extinction learning techniques.
In the present study, participants expressed comparable responses to the snake pictures throughout activation, extinction and re-exposure, and the responses were higher to the snake pictures than to the NA. However, there was an unexpected strong reaction to the two mushroom pictures during the activation stage that was only present in the direct group. Thus, although the groups did not differ in their reactions to the snake stimuli, the direct group showed significantly stronger startle responses to the mushroom stimuli compared to the vicarious group. This pre-existing group difference in response to the two mushroom pictures was most probably due to noise, because first, this difference in mushroom reaction preceded the experimental group manipulation, and second, participants had been randomly assigned to the experimental groups keeping constant self-reported snake fear and general anxiety levels. Critically, the difference in response to the mushroom stimuli did not appear to influence the main results, as including the response to the mushroom stimuli as a covariate did not alter any of the reported findings (see supplementary results). Further, during the instrumental approach/avoidance task, the groups did not differ in approach behavior towards the snake pictures. As expected, approach behavior towards a snake picture was facilitated by monetary reward, a linear relationship that was present in both groups independent. Exploratory analyses relating self-reported social anxiety and fear expression during the re-exposure test, on the one hand, and the approach behavior on the other hand, showed a trend towards a negative correlation between the level of social anxiety and snake fear during the re-exposure test. This trend was only present for the group that underwent vicarious extinction learning, suggesting that participants with higher levels of social anxiety profited more
from vicarious extinction learning (i.e. expressed lower levels of fear during snake re-exposure). As no previous research has investigated the effect of social anxiety on vicarious extinction learning, the preliminary trend from this exploratory analysis has to be treated with caution. Future research should investigate more closely how social processes such as self vs. other-focused attention involved in social anxiety (Clark and Wells, 1995) influence vicarious extinction learning.
The treatment of choice for phobic patients typically involves exposure therapy, which is the clinical analogue of experimental extinction training. Exposure therapy is a technique in which a patient directlyfaces the feared object or situation (Barlow, 2002). This technique can effectively reduce fears already after 2-3 hours of treatment (Öst, 1989), but relapse rates are high (Craske & Mysthowski, 2006). Thus, unlike reducing conditioned fears acquired in the laboratory, reducing phobic fears generally involves directly facing the feared object, and these fears are typically more resistant to change (Seligman, 1971). Therefore, achieving lasting reduction of lifelong phobic fears merely by exposure to images (i.e. experimental extinction training) of the fear-eliciting object instead of interacting with a real-life object, contradicts previously established routines. Nevertheless, experimental extinction was reported effective by Björkstrand et al. (2016, 2017), who critically attributed the efficacy of the procedure to the activation of the fear memory 10 minutes prior to experimental extinction. The current failure to support the conclusions reported by Björkstrand et al. (2016), could be attributed to a failure to successfully activate the memory prior to extinction, rendering the extinction procedure ineffective. If the current study failed to trigger the activation of a generic fear memory concept, then this raises the question of the processes necessary to achieve this, to better understand how this technique could then effectively be implemented in a clinical setting. This is important, because updating fear memory through reconsolidation is promising in providing long-lasting beneficial effects.
The most important prerequisite for updating memory through reconsolidation, is the (re)activation of the fear memory (Lee, 2009). Research in fear conditioning has shown that the reactivation of specific fear memories needs to meet some conditions. First, the reactivation has to involve a violation of what is expected (i.e. a prediction error) (Sevenster, Beckers, & Kindt, 2012). This, because reconsolidation is an adaptive process in which violation of the expected means that something must be learned and that the fear memory needs to be updated and new information can be incorporated (Lee, 2009). In experimental fear conditioning studies, this is done by first presenting the CS (e.g. a picture of a snake) with the US (e.g. a shock), and subsequently present participants to the CS without the US during reactivation. However, a prediction error is not as straightforward to create when the fear memory is acquired outside the laboratory. Moreover, activation could on the one hand contain too much learning, resulting in extinction learning (Merlo,
Milton, Goozée, Theobald, & Everitt, 2014), or on the other hand contain too little learning, preventing reconsolidation to occur (Sevenster et al., 2012). The other important factor in reconsolidation is the required duration of the reactivation. Both older, as well as stronger, fear memories require a longer duration of reactivation to be able to destabilize the memory (Suzuki, Josselyn, Frankland, Masushige, Silva & Kida, 2004). Although the reactivation of a fear memory has elaborately been investigated in experimental fear conditioning paradigms in both rodents and humans, translating the reconsolidation and the preceding memory activation to phobic fear in humans is still preliminary. The use of reconsolidation in a (sub)clinical sample has been investigated in pharmacological studies, by activating the fear memory followed by propranolol intake, and in behavioral studies, by offering extinction learning after activation of the fear memory. Soeter and Kindt (2015) activated the memory in snake-fearful participants, before giving propranolol, by telling them to touch a tarantula, while they were in the end only briefly exposed to the tarantula, and a case study of Kindt and Emmerik (2016) tended to activate the memory of patients with PTSD, before giving propranolol, by focusing on their hotspots, but terminating this focus when a maximum of distress was reached (to prevent from starting extinction learning). Behaviorally, Telch et al. (2017) activated the fear memory of snake and spider phobics, before offering six 3 min exposure trials, by exposing the patients to the feared object for 10 seconds and asking the patients to memorize an experience with the feared object. The current study, based on the study of Björkstand et al. (2016), tried to achieve memory activation by simply presenting an exemplar picture that resembled the fear memory. Thus, the methods of activating long-lasting fear are still widespread. Nevertheless, it could be questioned whether only showing a picture that represents an exemplar of the feared cue is sufficient to trigger the memory updating process and make fear memory liable to disruption. Although Björkstrand et al. (2016) concluded this because of a superior effect found following extinction learning shortly after activation (i.e. 10 min), compared to later after activation (i.e. 6 hr), the results from the present study do not support a similar conclusion. This difference could be due to a failure to activate the memory in the present study or because any potential effects are weak and subtle. Future studies should further investigate the boundary conditions of the activation of phobic fear memories and its underlying mechanisms, such as the prediction error and the duration of the activation matching the strength and the age of the memory. Consequently, to be able to use this technique in the treatment of phobic fears, an effective, optimized and standardized method for the activation and reconsolidation of phobic fear should still be developed.
In conclusion, the present findings revealed that both vicarious and direct extinction learning after fear memory activation did not provide any reduction in fear expression and in
addition, no support was found for the superior effect of vicarious extinction learning over direct extinction learning. The findings contrast with the recently reported study by Björkstrand et al. (2016), who showed a fear diminishing effect using a similar design. The discrepancy between the studies could be due to a failure to activate the phobic fear memory in the current study, and future research should investigate the processes that are necessary to reproduce the fear reducing effects of post-activation extinction in phobic participants. This study also stresses the importance of translating the findings found in fear conditioning studies to more ecological settings, in order to develop persistent, non-invasive techniques to treat anxiety disorders. To our knowledge, this was the first study to compare vicarious and direct extinction learning in a sub-clinical sample with older, stronger and more clinical relevant fear memories and this study encourages follow-up studies to detail the potentials and limitations of using experimental techniques in (sub)clinical populations and to develop paradigms that comprehends the complexity and severity of naturally occurring fear and anxiety disorders.
References
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (DSM-
5®). American Psychiatric Pub.
Barlow, D. H. (2004). Anxiety and its disorders: The nature and treatment of anxiety and panic. Guilford press.
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. Social phobia: Diagnosis,
assessment, and treatment, 41(68), 00022-3.
Beckers, T., Krypotos, A. M., Boddez, Y., Effting, M., & Kindt, M. (2013). What's wrong with fear conditioning?. Biological psychology, 92(1), 90-96.
Bijl, R. V., & Ravelli, A. (2000). Current and residual functional disability associated with psychopathology: findings from the Netherlands Mental Health Survey and Incidence Study (NEMESIS). Psychological medicine, 30(3), 657-668.
Björkstrand, J., Agren, T., Åhs, F., Frick, A., Larsson, E. M., Hjorth, O., ... & Fredrikson, M. (2016). Disrupting reconsolidation attenuates long-term fear memory in the human amygdala and facilitates approach behavior. Current Biology, 26(19), 2690-2695.
Björkstrand, J., Agren, T., Åhs, F., Frick, A., Larsson, E. M., Hjorth, O., ... & Fredrikson, M. (2017). Think twice, it’s all right: Long lasting effects of disrupted reconsolidation on brain and behavior in human long-term fear. Behavioural brain research, 324, 125-129.
Blumenthal, T. D., Cuthbert, B. N., Filion, D. L., Hackley, S., Lipp, O. V., & Van Boxtel, A. (2005). Committee report: Guidelines for human startle eyeblink electromyographic studies. Psychophysiology, 42(1), 1-15.
Bouton, M. E. (2002). Context, ambiguity, and unlearning: sources of relapse after behavioral extinction. Biological psychiatry, 52(10), 976-986.
Bouton, M. E., & Swartzentruber, D. (1991). Sources of relapse after extinction in Pavlovian and instrumental learning. Clinical Psychology Review, 11(2), 123-140.
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. Social phobia: Diagnosis,
assessment, and treatment, 41(68), 00022-3.
Craske, M. G., & Mystkowski, J. L. (2006). Exposure Therapy and Extinction: Clinical Studies. Golkar, A., Bellander, M., Olsson, A., & Öhman, A. (2012). Are fear memories erasable?– reconsolidation of learned fear with fear-relevant and fear-irrelevant stimuli. Frontiers in
behavioral neuroscience, 6, 80.
Golkar, A., Castro, V., & Olsson, A. (2015). Social learning of fear and safety is determined by the demonstrator's racial group. Biology letters, 11(1), 20140817.
Golkar, A., Haaker, J., Selbing, I., & Olsson, A. (2016). Neural signals of vicarious extinction learning. Social cognitive and affective neuroscience, 11(10), 1541-1549.
Golkar, A., Selbing, I., Flygare, O., Öhman, A., & Olsson, A. (2013). Other people as means to a safe end: vicarious extinction blocks the return of learned fear. Psychological science, 24(11), 2182-2190.
Golkar, A., Tjaden, C., & Kindt, M. (2017). Vicarious extinction learning during reconsolidation neutralizes fear memory. Behaviour Research and Therapy.
Grillon, C., & Ameli, R. (2001). Conditioned inhibition of fear-potentiated startle and skin conductance in humans. Psychophysiology, 38(5), 807-815.
Heimberg, R. G., Horner, K. J., Juster, H. R., Safren, S. A., Brown, E. J., Schneier, F. R., & Liebowitz, M. R. (1999). Psychometric properties of the Liebowitz social anxiety scale.
Psychological medicine, 29(1), 199-212.
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of general psychiatry, 62(6), 593-602.
Kindt, M. (2014). A behavioural neuroscience perspective on the aetiology and treatment of anxiety disorders. Behaviour research and therapy, 62, 24-36.
Kindt, M., & Soeter, M. (2013). Reconsolidation in a human fear conditioning study: a test of extinction as updating mechanism. Biological psychology, 92(1), 43-50.
Kindt, M., & van Emmerik, A. (2016). New avenues for treating emotional memory disorders: towards a reconsolidation intervention for posttraumatic stress disorder. Therapeutic advances
in psychopharmacology, 6(4), 283-295.
Klorman, R., Weerts, T. C., Hastings, J. E., Melamed, B. G., & Lang, P. J. ~1974!. Psychometric description of some specific-fear questionnaires. Behavior Therapy, 5, 401–409
Lang, P. J., Davis, M., & Öhman, A. (2000). Fear and anxiety: animal models and human cognitive psychophysiology. Journal of affective disorders, 61(3), 137-159.
LeDoux, J. E. (2000). Emotion circuits in the brain. Annual review of neuroscience, 23(1), 155-184. Lee, J. L. (2009). Reconsolidation: maintaining memory relevance. Trends in neurosciences, 32(8), 413-420.
Liebowitz, M. R. (1987). Social phobia. In Anxiety (Vol. 22, pp. 141-173). Karger Publishers. Merlo, E., Milton, A. L., Goozée, Z. Y., Theobald, D. E., & Everitt, B. J. (2014). Reconsolidation and extinction are dissociable and mutually exclusive processes: Behavioral and molecular evidence. Journal of Neuroscience, 34(7), 2422-2431.
Nader, K., Schafe, G. E., & Le Doux, J. E. (2000). Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature, 406(6797), 722-726.
Öhman, A., & Mineka, S. (2001). Fears, phobias, and preparedness: toward an evolved module of fear and fear learning. Psychological review, 108(3), 483.
Olesen, J., Gustavsson, A., Svensson, M., Wittchen, H. U., & Jönsson, B. (2012). The economic cost of brain disorders in Europe. European journal of neurology, 19(1), 155-162.
Öst, L. G. (1989). One-session treatment for specific phobias. Behaviour research and Therapy, 27(1), 1-7.
Peira, N., Golkar, A., Öhman, A., Anders, S., & Wiens, S. (2012). Emotional responses in spider fear are closely related to picture awareness. Cognition & emotion, 26(2), 252-260.
Schiller, D., Monfils, M. H., Raio, C. M., Johnson, D. C., LeDoux, J. E., & Phelps, E. A. (2010). Preventing the return of fear in humans using reconsolidation update mechanisms. Nature, 463(7277), 49-53.
Seligman, M. E. (1971). Phobias and preparedness. Behavior therapy, 2(3), 307-320.
Seligman, L. D., & Wuyek, L. A. (2005). Vicarious extinction. Encyclopedia of behavior modification
and cognitive behavior therapy, 3, 1085-1086.
Sevenster, D., Beckers, T., & Kindt, M. (2012). Retrieval per se is not sufficient to trigger reconsolidation of human fear memory. Neurobiology of learning and memory, 97(3), 338-345. Soeter, M., & Kindt, M. (2011). Disrupting reconsolidation: pharmacological and behavioral manipulations. Learning & Memory, 18(6), 357-366.
Soeter, M., & Kindt, M. (2015). An abrupt transformation of phobic behavior after a post- retrieval amnesic agent. Biological psychiatry, 78(12), 880-886.
Spielberger, C.D., Gorsuch, R.L. & Lusthene, R.E. (1970). Manual for the State-Trait Anxiety
Inventory. Palo Alto: Consulting Psychologists Press.
Steinfurth, E. C., Kanen, J. W., Raio, C. M., Clem, R. L., Huganir, R. L., & Phelps, E. A. (2014). Young and old Pavlovian fear memories can be modified with extinction training during reconsolidation in humans. Learning & Memory, 21(7), 338-341.
Suzuki, A., Josselyn, S. A., Frankland, P. W., Masushige, S., Silva, A. J., & Kida, S. (2004). Memory reconsolidation and extinction have distinct temporal and biochemical signatures. Journal of
Neuroscience, 24(20), 4787-4795.
Telch, M. J., York, J., Lancaster, C. L., & Monfils, M. H. (2017). Use of a brief fear memory reactivation procedure for enhancing exposure therapy. Clinical Psychological Science, 5(2), 367- 378.
Thompson, A., & Lipp, O. V. (2017). Extinction during reconsolidation eliminates recovery of fear conditioned to fear-irrelevant and fear-relevant stimuli. Behaviour Research and
Therapy, 92, 1-10.
van der Ploeg, H. M., Defares, P. B., & Spielberger, C. D. (1981). Handleiding bij de zelf-beoordelings
vragenlijst: een Nederlandstalige bewerking van Spielberger state-trait anxiety inventory STAI-DY.
Addendum 1981. Swets & Zeitlinger.
van der Bij, A. K., de Weerd, S., Cikot, R. J., Steegers, E. A., & Braspenning, J. E. C. (2003). Validation of the Dutch short form of the state scale of the Spielberger State-Trait Anxiety Inventory: considerations for usage in screening outcomes. Public Health
Genomics, 6(2), 84-87.
Vervliet, B., Vansteenwegen, D., Baeyens, F., Hermans, D., & Eelen, P. (2005). Return of fear in a human differential conditioning paradigm caused by a stimulus change after extinction. Behaviour research and therapy, 43(3), 357-371.
Appendix A:Covariance Analyses
All main analyses reported in the results section, with the addition of the covariate “Mushroom stimuli”, which is the mean FPS response to the mushroom stimuli during the activation stage.
Activation Stage
To compare the mean FPS responses during the activation stage, a 2 (Stimulus) x 2 (Group) repeated measure analyses of covariance (ANCOVA) was used, controlling for the Mushroom stimuli. This analysis revealed a significant main effect of Stimulus (F(1, 63) = 14.26, p < .001, ηp2 = .19), and a non-significant Stimulus x Group interaction (F(1, 63) = .05, p = .82, ηp2 = .001).
Extinction Stage
To assess whether FPS responses towards the snake stimulus decreased from beginning to the end of the extinction stage, a 2 (Stimulus) x 5 (Block) x 2 (Group) repeated measures ANCOVA was used, controlling for the Mushroom stimuli. This revealed a significant main effect of Stimulus, indicating overall higher startle potentiation to the snake stimuli (snake vs. NA: F(1, 63) = 21.42, p < .001, ηp2 = .25) and a significant main effect of Block, indicating an overall decrease in FPS responses during extinction (F(4, 252) = 33.57, p < .001, ηp2 = .35). Neither of these effects differed significantly between groups (Block x Group: F(4, 252) = .95, p = .43; Stimulus x Group: F(1, 63) = .002, p = .97). Finally, a 2 (Stimulus) x 2 (Group) repeated measures ANOVA during the last extinction block confirmed that there was no significant startle potentiation to the snake stimuli left at the end of day 1 (Stimulus: F(1,63) = 2.00, p = .16), and no significant between-group differences (Stimulus x Condition: F(1, 63) = .01, p = .93).
Re-exposure Stage
To assess whether the startle potentiation to snake stimuli differed between the vicarious and direct group one day after extinction learning, a 2 (Stimulus) x 2 (Group) repeated measures ANCOVA was used, with Mushroom as a covariate. This analysis resulted in a main effect of Stimulus, showing a significant startle potentiation to the snake stimuli one day after extinction learning (F(1, 63) = 6.32, p = .02, ηp2 = .09). In contrast to the hypothesis, there were no differences between groups (Stimulus x Group; F(1, 63) = .145, p = .71). Analyzing the mean FPS responses across the four blocks of re-exposure revealed a similar pattern of results. Thus, a 2 (Stimulus) x 4 (Block) x 2 (Group) repeated measures ANCOVA, controlling for the Mushroom stimuli, resulted in a significant main effect of Stimulus (F1, 63) = 13.77, p < .001, ηp2 = .18), indicating a stronger startle potentiation to snake stimuli vs. NA, and a main effect of Block (F(3, 189) = 19.58, p < .001,
ηp2 = .24), showing an overall decrease in startle potentiation, in the absence of significant between group differences (Stimulus x Group: F(1, 63) = 1.18, p = .28; Block x Group: F(3, 189) = .22, p = .88).
Critically, to assess if the extinction procedures after memory activation significantly reduced fear expression to the snake stimuli from pre-extinction to post-extinction training, we compared the FPS response during the activation stage (first trial) with the FPS response during the re-exposure stage (first trial), controlling for the Mushroom stimuli. This 2 (Snake vs. NA) x 2 (Activation vs. Re-exposure) x 2 (Vicarious vs. Direct) repeated measures ANCOVA, with Mushroom stimuli as a covariate, did not show a reduction in startle potentiation to the snake stimuli from activation to re-exposure, as shown by the non-significant Stimulus x Stage interaction (F(1, 63) = .42, p = .52). Thus, in the absence of any between-group differences (Stimulus x Group:
F(1, 63) = .01, p = .95; Stage x Group: F(1, 63) = .02, p = .90), snake potentiation during the
activation stage did not significantly differ from snake potentiation in the re-exposure stage in either the vicarious and the direct group. This was measured with a 2 (Snake, NA) x 2 (Vicarious, Direct) repeated measures ANCOVA, controlling for the Mushroom stimuli, resulting in a non-significant main effect of Stimulus (F(1, 63) = .42, p = .52), and a non-significant interaction of Stimulus x Group (F(1, 63) = .21, p = .65). As previous research (i.e. Björkstrand et al., 2016) has focused on the change in fear expression from end of extinction to beginning of re-exposure, we additionally assessed this change in a 2 (Snake vs. NA) x 2 (last Extinction trial vs. first Re-exposure trial) x 2 (Vicarious vs. Direct) repeated measures ANCOVA, controlling for the Mushroom stimuli. This analysis gave a similar result, with a non-significant Stimulus x Stage interaction (F(1, 63) = .64, p = .43).
Approach/Avoidance behavior test
The difference in approach behavior to snake stimuli between the two groups was measured with a 5 (Reward level) x 2 (Group) repeated measures ANCOVA, controlling for the Mushroom stimuli. This ANCOVA revealed a significant main effect for Reward level, confirming the increase in approach behavior when monetary gain increased (F(2.71, 252) = 113.54, p <.001, ηp2 = .64). However, in contrast to our hypothesis, there were no differences between groups (Reward level x Group: (F(2.71, 252) = .05, p = .99).
