Master thesis: Effects of Noradrenergic Blockade on Reconsolidation and
Extinction.
Student: Wouter Cox
Supervisor: Dr. M. Soeter
Date: 28-07-2015
Abstract
Disruption of the reconsolidation process using ß-adrenergic receptor antagonists is a promising new approach for treating anxiety disorders. Upon reactivation, fear memories enter a labile state during which they are modifiable. However, prolonged reactivation can result in new learning and might constitute a boundary condition for this novel treatment. Furthermore, administering ß-adrenergic receptor antagonists, such as propranolol, could interfere with new learning and by doing so prevent fear-reducing effects of extinction
training. The present research was concerned with whether prolonged reactivation poses a risk to treating anxiety disorders with disruption of the reconsolidation process by means of
propranolol administration. It was expected that return of fear would be greater with
extinction than reactivation after administration of propranolol. Moreover, return of fear was expected to be enhanced by administering propranolol after extinction training. These
hypotheses were tested in a human fear-conditioning paradigm. After acquisition of fear, participants received reactivation with propranolol or extinction training with propranolol or placebo. Afterwards, it was tested whether fear was eliminated or not. Although we found some evidence for extinction being a boundary condition for disruption of the reconsolidation process, the results were largely not in line with the predictions. Possible explanations could be low statistical power and procedural issues, such as unsuccessful extinction training. Suggestions for future research include extending extinction training and studying the role of trait anxiety in disruption of the reconsolidation process. By doing so, improvements in procedures to disrupt reconsolidation of fear memories might be realized.
Introduction
Anxiety disorders are one of the most prevalent and detrimental psychological disorders. The lifetime prevalence is about 29 percent (Kessler et al., 2005) and patients usually experience difficulties in functioning at work and an overall lowered quality of life (Alonso et al., 2004). This pathological fear is traditionally explained by a neutral or
ambiguous stimulus (conditioned stimulus, CS) occurring in close temporal proximity with an aversive outcome (unconditioned stimulus, US). This leads to conditioned fear (Watson & Rayner, 1920) with people believing that a feared stimulus (CS) is followed by an adverse event (US). By doing so, maladaptive fear memory can result in fearful responses to actually harmless stimuli. One of the most successful treatments for anxiety disorders involves exposing patients to the CS without occurrence of the US. These exposure sessions, as embedded in cognitive behavioral therapy (CBT), result in the formation of a new memory pathway (CS-no US association) that inhibits the fear memory (CS-US association) (Bouton, Westbrook, Corcoran & Maren, 2005). As a result, patients experience less fear. However, even though initial effects of CBT are fairly good, relapse rates remain high (Bouton, 2002). A return of fear is thought to be a consequence of the fear memory remaining intact after CBT. Therefore, fear can easily return when the feared stimulus is encountered in a context other than the context used in therapy (renewal), after a single reexposure to the US
(reinstatement) and after merely the passage of time (spontaneous recovery) (Bouton et al., 2005). Thus, due to the high prevalence, adverse effects and the leading treatment suffering from large relapse rates, developing new treatments for anxiety disorders that can overcome the problems of CBT is an important field of study.
Disruption of the reconsolidation process is an intervention that has the potential to permanently erase fear associations and could thereby overcome the problem of high relapse rates. Even though it was previously thought that formed memory cannot be altered, it has been shown that reactivation of a fear memory results in the memory trace entering a labile state (Nader et al., 2000). During this state, protein synthesis is required for the memory to stabilize again. This protein dependent restoring of memory traces is called reconsolidation (Przybyslawski & Sara, 1997). Administration of a ß-adrenergic receptor antagonist, such as propranolol, interferes with the needed protein synthesis so that the memory trace is not reconsolidated in long term memory (Dębiec & Ledoux, 2004). Therefore, it could
permanently erase fear memories and by doing so prevent relapse. Indeed, fear-conditioning studies have repeatedly shown that administration of propranolol before or after retrieving a previously formed fear memory results in an elimination of the eyeblink startle reflex (e.g. Kindt, Soeter & Vervliet, 2009; Sevenster, Beckers & Kindt, 2012), a physiological protective response that is regarded to be a specific measure of fear (Hamm & Weike, 2005).
Successful disruption of the reconsolidation process depends on subtle conditions though. The process can be considered as an adaptive mechanism to incorporate newly
learned experiences trough updating of formed memories. By doing so, memory traces remain relevant in guiding behavior (Dudai, 2009). Importantly, it seems that reconsolidation does not occur when no new information is presented during reactivation, because under this circumstance the memory does not need adjustment. However, when an unforeseen event takes place, the memory trace needs updating in order to better predict future events and reconsolidation is triggered (Lee, 2009). In support of this idea, Sevenster et al. (2012) have shown that, in addition to retrieval of the fear memory, an expectation of threat must be
violated in order for reconsolidation to occur. This means that when in a fear-conditioning procedure a CS (e.g. a picture of a spider) has been repeatedly followed by an US (e.g. a shock), shortly reactivating the memory with an unreinforced CS presentation can induce reconsolidation because the absence of shock was unexpected based on previous learning. However, instead of reconsolidation, violations of CS-US expectations can result in new learning and extinction as well (Rescorla & Wagner, 1972). After all, extinction training is carried out by repeatedly presenting unreinforced CS’s (Barlow, 2008), which leads to the formation of a new memory trace that inhibits the fear memory (Bouton et al., 2005). The difference seems to lie in the length of exposure to the feared stimulus. A short retrieval with a brief violation of threat expectancy can result in reconsolidation, whereas extinction training involves prolonged exposure until expectancy of threat is diminished. In laboratory settings it is relatively straightforward when reconsolidation and extinction is induced, because the number of unreinforced CS presentations can be highly controlled. However, one could argue that this difference is not so clear-cut when attempting to apply these interventions in actual therapy. For instance, what duration of exposure to the feared stimulus resembles short or prolonged exposure is not clear. Therefore, therapists that aim to induce reconsolidation might inadvertently expose patients for too long to the feared stimulus with extinction training as a result. In light of this possibility, it is important to study the consequence of administering agents that block reconsolidation when, in fact, extinction training has occurred.
Based on previous research, administering propranolol after extinction can be suspected to sort unfavourable effects. For instance, Sevenster, Beckers and Kindt (2014) have shown that when a fear memory is reactivated with a few unreinforced CS presentations, administration of propranolol does not result in reduced fear. This finding indicates that multiple unreinforced CS presentations (i.e. extinction training) results in the reconsolidation window to close and new learning to occur with no fear-attenuating effects of propranolol as a consequence. Therefore, extinction seems to be a boundary condition for disruption of the reconsolidation process. Moreover, administering propranolol after extinction training might not only be ineffective, but even counterproductive in reducing fear. Besides reconsolidation, also successful extinction seems to rely on functioning of ß-adrenergic receptors (Mueller & Cahill, 2010). Therefore, the same agents that block reconsolidation might block
consolidation of new inhibitory memory traces after extinction training as well. Hence, administering propranolol after reconsolidation or extinction could have opposing effects on expressed fear. Administration after reconsolidation leads to a removal of the fear memory and less expressed fear, whereas administration after extinction training would lead to provided safe information being neutralized and expressed fear remaining intact.
In sum, previous human fear-conditioning research suggests that prolonged reactivation of fear memories results in extinction, instead of reconsolidation, and fear memory cannot be erased by administration of propranolol under this circumstance. In addition, administering propranolol might even prevent beneficial effects of extinction training by blocking the consolidation of inhibitory memory traces. The present research aimed to investigate to what extent prolonged reactivation of fear memories poses a risk to treating anxiety disorders with disruption of the reconsolidation process by means of
propranolol administration. More specifically, this research investigated two questions. First, it was studied whether extinction is a boundary condition for disrupting the reconsolidation process. Animal studies have confirmed this to be the case (Lee, Milton & Everitt, 2006). However, to the author’s knowledge, this was the first study that directly compared the effect of administering propranolol after extinction training and reconsolidation in humans. The
second question was whether administration of propranolol blocks fear-inhibiting effects of extinction training. Bos, Beckers and Kindt (2012) already studied the effects of
administering propranolol prior to extinction training on return of fear. Surprisingly, no effect on the fear potentiated startle response was found. Instead, administration of propranolol resulted in a lowered expectation of US occurrence to a previously reinforced stimulus during extinction training and the subsequent day. This finding might indicate that propranolol interferes with consolidation of declarative memory. However, some procedural issues were encountered in the study, such as seemingly unsuccessful extinction training, as measured with the startle response. In addition, since propranolol was administered before and active during extinction training, it is not clear whether it interfered with encoding or consolidation of memory (Bos et al., 2012). Therefore, the present study was conducted with relatively prolonged extinction training and propranolol being administered afterwards in order to gain more insight into its effects on extinction.
A fear-conditioning paradigm was used to test the hypotheses. The experiment was conducted on three consecutive days (see Figure 1). On the first day, fear conditioning took place by presenting two pictures of spiders. One picture (CS1) was followed by an US (i.e. a mild electrical shock), whereas the other picture (CS2) was not. On the next day, participants were randomly assigned to one of three conditions. Participants in the extinction-propranolol condition received propranolol after repeated unreinforced CS1 presentations, whereas participants in the extinction-placebo condition received a placebo. Participants in the reactivation-propranolol condition received propranolol after a single unreinforced CS1 presentation. On day 3, all participants were first exposed to multiple unreinforced CS1 and CS2 presentations. Then three reminder shocks (reinstatement) were delivered, followed by more unreinforced CS1 and CS2 presentations to measure the presence of fear memory.
New learning seems to result in fear remaining intact (Sevenster et al., 2014), whereas reconsolidation leads to an elimination of fear after propranolol has been administered (Kindt et al., 2009). Using startle response as a measure of fear (Kindt et al., 2009), it was, therefore, expected that administration of propranolol after multiple unreinforced CS1 presentations (extinction-propranolol condition) would result in more return of fear on day 3 than after a single unreinforced CS1 presentation (reactivation-propranolol condition). Since ß-adrenergic receptor antagonists could interfere with extinction training by blocking consolidation of inhibitory memory traces (Mueller & Cahill, 2010), it was also expected that administration of propranolol after extinction training (extinction-propranolol condition) would result in more return of fear than when a placebo pill was administered (extinction-placebo condition). Finally, based on the findings of Bos et al. (2012) on declarative memory, it was predicted that expectancies of US occurrence at CS1 presentations would be higher in the extinction-propranolol than in the extinction-placebo condition after extinction training.
Day 1 Acquisition Day 2 Intervention Day 3 Test 5 CS1 (80% followed by US) 5 CS2 1 CS1 + propranololA 10 CS1 5CS1 10 CS2 5CS2 10 CS1 + propranololB 10 CS1 + placeboC 3US’s (reinstatement) Figure 1
Schematic overview of the experimental design. A, B and C correspond with the reactivation-propranolol, extinction-propranolol and extinction-placebo conditions, respectively.
Methods
Participants
Twenty-four undergraduate students (11 men, 13 women) participated in the study. Their age ranged from 19 to 29 years with a mean of 21.96 and a standard deviation of 3.13. Participants with a medical or psychological condition, contraindicative for taking
propranolol, were excluded from the study (n=15). These conditions are cardiovascular, lung, liver or kidney disease, medication use, pregnancy, diabetes, hyperthyroidism, blood pressure lower than 90/60, epilepsy, convulsions and depressive, anxiety or psychotic disorder (Kindt et al., 2009). In addition, participants who did not demonstrate successful fear acquisition were excluded (n=19, see data analyses). Finally, two additional participants were left out of analyses, because one fell asleep and another encountered a technical error during the
experiment. Participants were randomly assigned to the reactivation-propranolol condition (n=7), extinction-propranolol condition (n=9) or extinction-placebo condition (n=8) with the restriction that conditions were matched on State-Trait Anxiety Inventory-Trait (STAI-T) and Fear of Spiders Questionnaire (FSQ) scores as much as possible. They received course credits or a small amount of money (€50,-) for participation. The ethical committee of the University of Amsterdam approved of the study, participants were informed about the nature of the study and consent was obtained before the start of the experiment.
Materials
Questionnaires
State and trait anxiety were measured with the STAI (Spielberger et al., 1970). This questionnaire contains 40 items that are measured on a 4-point scale, ranging from “not at all” to “very much”. The STAI shows good reliability and validity (Metzger, 1976). Fear of spiders was measured using the FSQ. It contains 18 items that are answered on a 7-point scale, ranging from “completely disagree” to “completely agree” (Szymanski & O’Donohue, 1995). This questionnaire also has good reliability and validity (Muris & Merckelbach, 1996). To assess participants’ tendency to respond fearfully to symptoms related to anxiety, the
Anxiety Sensitivity Index (ASI) was used (Peterson and Reiss, 1992). Its 16 items are measured on a 5-point scale, ranging from “very little” to “very much”. This questionnaire shows good psychometric properties (Reiss & Peterson, Gursky & McNally, 1986). Finally, the pleasantness of the US was rated on an 11-point scale, with a minimum score of -5 (unpleasant) and a maximum score of 5 (pleasant).
Stimuli
Conditioned stimuli were pictures of spiders (IAPS numbers 1200 and 1201) (Lang, Bradley and Cuthbert, 2005), displayed on a computer screen. One of the spider pictures (CS1) was in 80 percent of the cases followed by an electrical stimulus (US), whereas another picture was never followed by an US (CS2). Assignment of reinforced (CS1) and
unreinforced (CS2) pictures was counterbalanced across participants.
Startle probes were presented binaurally over headphones (Sennheiser, HD 25-1 II). The volume and duration were 104 decibel and 40 milliseconds, respectively. The volume of the background noise was 70 decibel.
Electrical shocks (US) were administered for two milliseconds to the wrist of the left hand with two Ag electrodes of 20 by 25 mm with a fixed inter-electrode mid-distance of 45 mm. These electrodes were connected to a constant current stimulator (Digitimer DS7A, Hertfordshire, UK), controlling electrical stimulus delivery. A conductive gel (Signa, Parker) was placed between the electrodes and the skin.
Physiological measures
The eyeblink startle reflex was determined by measuring electromyography (EMG) of the right orbicularis ocili muscle. For this purpose two 7 mm Ag/AgCl electrodes, filled with conductive gel (Signa, Parker), were placed approximately one centimetre beneath the pupil and one centimetre beneath the lateral canthus. A ground electrode was placed on the middle-top of the forehead. The eyeblink EMG electrodes were connected to a front-end amplifier with an input resistance of 10 MOhm and a bandwidth of DC-1500 Hertz. Startle response values were peak level amplitude levels between 20 and 200 milliseconds after startle probe delivery. A notch filter was set at 50 Hertz for removal of interference. The EMG signal was sampled at 1000 Hertz and band-pass filtered subsequently (Butterworth 4th order, 28-500 Hertz).
Electrodermal activity was recorded with two Ag/AgCl electrodes of 20 by 16 mm, placed on the medial phalanx of the middle and fourth finger of the left hand. The skin conductance levels were measured by means of an input device, connected to the electrodes, with a sine shaped excitation voltage (1 V peak-peak) of 50 Hertz.
Blood pressure and heart rate were measured with an electronic sphygmomanometer. Saliva samples were collected with cotton salivettes without chemical stimulants. They were placed in the mouth for three minutes and frozen at -30 degrees Celsius after collection.
Expectancy ratings
Expectancy of US occurrence was measured at every CS presentation on an online 11-point scale, ranging from -5 (“certainly no shock”) to 5 (“certainly a shock”). The expectancy scale was continuously presented at the bottom of the computer screen and participants rated their expectancies by clicking the left button of a computer mouse.
Procedure
The study took place on three consecutive days with a total duration of approximately five hours. The procedure was based on previous fear-conditioning studies on reconsolidation and extinction (e.g. Kindt et al., 2009; Bos et al., 2012). At the start of the first day, a medical screening was carried out to ensure that propranolol could safely be administered. In addition, participants filled in the STAI, FSQ and ASI. Afterwards, the shock intensity level was determined. Starting with an electrical current of 1 milliAmpere (mA), the intensity was gradually built up until participants indicated that they experienced the shock as
uncomfortable, but not painful. Next, the electromyography and skin conductance response electrodes were attached to measure startle reactivity and electrodermal activity, respectively. Participants were seated in front of a computer and received instructions about the
experiment. They were informed that they would look at two pictures of spiders on the computer screen and one picture would “often” and another would “never” be followed by a shock. They were instructed to learn to predict, based on the presented picture, whether a shock would be administered. In addition, participants were asked to use the expectancy scale to indicate their expectation of shock occurrence. It was explained that expectancies should be recorded by pressing the left button of the computer mouse within 5 seconds after picture emergence. Then participants put on the headphones and the acquisition phase was commenced. They heard a background noise throughout all phases of the experiment.
Participants received ten bursts of noise through headphones for habituation to startle probes. Subsequently both pictures of spiders (CS1 and CS2) were presented five times. See Figure 2 for an overview of a single trial. The total duration of a trial was 8 seconds. On the seventh second of presentation, participants heard a burst of noise in order for startle
reactivity to be measured. When the CS1 was presented, participants received a shock half a second after onset of the startle probe on four of the five trials. On the first CS1 and all five CS2 trials participants did not receive a shock. In addition, participants received five bursts of noise (Noise Alone, NA) between trials of CS presentations to measure startle responses to the context of the experiment. The order of CS1, CS2 and NA presentation was randomized within blocks of three trials. Upon completion of the fear acquisition phase, participants filled in the STAI-State and rated the pleasantness of the US. They were asked to remember what they had learned during the session (i.e. which spider picture was and was not followed by a shock). In addition they were given instructions for the sampling of saliva on the next day. They were asked to refrain from caffeine use, alcohol use and exercise within 12 hours and eating, drinking (other than water), chewing gum and smoking within two hours before the next session. Also they were not allowed to brush their teeth within an hour prior to their next appointment.
The following day, ten startle probes were again displayed at the start. Then
participants in the extinction-propranolol and extinction-placebo condition were exposed to ten unreinforced CS1 and NA presentations, whereas participants in the
reactivation-propranolol condition received one unreinforced CS1 and NA presentation (Kindt et al., 2009). Afterwards, 40 milligram propranolol (extinction-propranolol and reactivation-propranolol condition) or placebo (extinction-placebo condition) was administered. Blood pressure was assessed and participants filled in the STAI-State before and 60 minutes after pill administration. At the same time saliva samples were obtained.
On the last day, blood pressure was measured and the STAI-State was filled in upon arrival. Participants were exposed to ten NA, unreinforced CS1 and CS2 presentations after
ten habituation trials. Then they received three unsignaled shocks (reinstatement) and another five NA, unreinforced CS1 and CS2 trials (Bos et al., 2012). At the end of the session they again filled in the STAI-State and rated the pleasantness of the US, before being paid and thanked for participation.
Figure 2
Schematic overview of CS trials. In CS1 trials an US was administered at 7.5 seconds after picture presentation onset, whereas in CS2 trials no US was administered.
Data analyses
The startle response to CS1 and CS2 trials in the acquisition phase were inspected after day 1. Participants who showed a higher startle response to the average of the last two trials of CS2, compared to CS1, were regarded as not demonstrating successful acquisition and excluded from the study. Startle and expectancy rating values of the first and the last two trials of each experimental phase (acquisition, extinction and reinstatement test) were
averaged per CS. These average values were used to analyze patterns within phases (e.g. testing whether expectancy ratings decreased during extinction on day 2) and between phases (e.g. testing whether startle responses increased after reinstatement). Statistical analyses were carried out using the Statistical Package for the Social Sciences (SPSS). For the hypothesis tests mixed analyses of variance were conducted with Condition (reactivation-propranolol, extinction-propranolol and extinction-propranolol) as between-subjects factor, and Stimulus (CS1, CS2 and NA) and Trial (average values of first and last two trials in each phase) as within-subjects factors. The significance level (α) was set at 0.05.
Results
Questionnaires and shock level
Reactivation-propranolol vs. extinction-propranolol
Using independent t-tests, it was tested whether the reactivation-propranolol and
0 1 2 3 4 5 6 7 8 Time (seconds) Picture presentation Expectancy rating Startle probe Shock
extinction-propranolol conditions differed in shock level, trait anxiety, fear of spiders, anxiety sensitivity and US evaluation (see Table 1 for mean scores and standard deviations per condition). The conditions did not differ in STAI-Trait scores, t (14) = -1.16, p = 0.27; FSQ scores, -1 < t (14) < 1; ASI scores, -1 < t (14) < 1; US evaluation, -1 < t (14) < 1 and shock level, t (14) = 1.29, p = 0.22. Therefore, participants in the two conditions did not score differently on any of the questionnaires or receive unequal shock levels.
Extinction-propranolol vs. extinction-placebo
The extinction-propranolol and extinction-placebo conditions did not differ in STAI-Trait scores, FSQ scores, ASI scores, US evaluation and shock level, (-1 < t (14) < 1, for all measures). Therefore, the two extinction conditions also did not seem to differ in trait anxiety, fear of spiders, anxiety sensitivity, US evaluation and shock level. In addition, it was tested whether administration of propranolol affected state anxiety on day 2 using a mixed ANOVA with Condition (extinction-propranolol and extinction-placebo) as between-subjects factor and Time (pre- and post administration) as within-subjects factor. No Condition x Time effect on STAI-State scores was found, F (1,15) < 1. Hence, propranolol did not seem to affect state anxiety.
Table 1
Mean values (standard deviation) of shock level (mA) and questionnaire scores for all three conditions.
Reactivation-propranolol Extinction-propranolol Extinction-placebo
Shock level 39.43 (37.06) 23.11 (8.1) 28.75 (29.76) STAI-T 33.00 (8.85) 37.89 (8.02) 37.25 (6.34) FSQ 26.29 (14.61) 30.33 (12.47) 26.75 (20.44) ASI 12.29 (9.34) 11.11 (7.44) 12.88 (5.92) US evaluation -3.29 (0.49) -3.33 (1.00) -3.5 (0.93) Blood pressure Reactivation-propranolol vs. extinction-propranolol
As a manipulation check, it was tested whether propranolol exerted a blood pressure decreasing effect. The systolic and diastolic blood pressure levels pre- and post pill
administration are displayed per condition in Table 2. A mixed ANOVA revealed a systolic blood pressure reduction in the reactivation-propranolol and extinction-propranolol condition, Time, F (1,14) = 31.97, p < 0.001. Remarkably, this reduction was larger in the reactivation-propranolol condition than in the extinction-reactivation-propranolol condition, Condition x Time, F (1,14) = 6.19, p = 0.03. With follow-up analyses, a significant drop in systolic blood pressure in both the reactivation-propranolol condition, F (1,6) = 21.94, p < 0.01 and the extinction-propranolol condition was found, F (1,8) = 7.71, p = 0.02. Furthermore, there was a nearly significant diastolic blood pressure reduction, Time, F (1,14) = 4.19, p = .06 that did not differ between the two conditions Condition x Time, F (1,14) = 2.59, p = 0.13. However, additional analyses showed a significant diastolic blood pressure reduction in the
reactivation-propranolol condition, F (1,6) = 7.96, p = .03, but not in the extinction-reactivation-propranolol condition,
F (1,8) < 1. Taken together, administration of propranolol, surprisingly, seemed to result in
more blood pressure reduction in the reactivation-propranolol condition than in the extinction-propranolol condition.
Extinction-propranolol vs. extinction-placebo
A systolic blood pressure reduction was found in the extinction-propranolol and extinction-placebo condition, Time, F (1,15) = 10.44, p < .01. However, no Condition x Time effect was found, F (1,15) < 1, indicating that administration of propranolol after extinction training did not result in reduced systolic blood pressure. In addition no main effect of Time,
F (1,5) < 1 nor a Condition x Time effect was found on diastolic blood pressure, F (1,5) < 1.
This result suggests that diastolic blood pressure did not drop to a larger degree in the extinction-propranolol condition than in the extinction-placebo condition. Therefore,
unexpectedly, administering propranolol after extinction training did not result in a reduction in either systolic or diastolic blood pressure.
Table 2
Mean systolic and diastolic blood pressure levels (standard deviation) pre- and post pill administration per condition.
Reactivation-propranolol Extinction-propranolol Extinction-placebo
Systolic Pre 120.57 (14.95) 113.78 (10.32) 106.88 (12.99)
Post 101.71 (11.70) 106.44 (9.72) 101.75 (12.50)
Diastolic Pre 69.14 (8.11) 67.56 (4.69) 71.38 (6.48)
Post 63.57 (6.05) 66.89 (5.40) 70.13 (7.77)
Fear potentiated startle
Reactivation-propranolol vs. extinction-propranolol
Acquisition. See Figure 3 for the fear potentiated startle response on all days for each condition. With a mixed ANOVA it was tested whether fear acquisition on day 1 was
successful in the reactivation-propranolol and extinction-propranolol condition. A significant Stimulus x Trial effect was found, F (1,14) = 8.67, p = 0.01, indicating that startle responses to CS1 increased to a larger degree than to CS2. In addition, no Stimulus x Trial x Condition effect was observed, F (2,21) < 1. These results suggest that successful fear acquisition occurred and did not differ between the two conditions.
Reactivation. With a repeated measures ANOVA it was found that the startle
response to CS1 decreased from the last two acquisition trials to the reactivation trial, Trial, F (1,6) = 6.25, p = 0.05. Furthermore, the startle response to CS1 and NA did not differ at reactivation, -1 < t (6) < 1. Therefore, it seems that, unexpectedly, the conditioned fear to CS1 was not well consolidated in the reactivation-propranolol condition.
Test. To test whether administration of propranolol after reactivation of the fear memory resulted in more fear reduction than administration after extinction training, the differential startle response from the last two trials of acquisition to the first two trials of extinction on day 3 were compared. There was a significant Stimulus x Trial effect, F (1,14) =
5.38, p = 0.04, indicating a larger startle response decrease to CS1 than CS2 trials in the reactivation-propranolol and extinction-propranolol conditions (see Figure 3). However, no difference in differential startle response reduction was observed between conditions, Stimulus x Trial x Condition, F (1,14) = 1.24, p = 0.28. This finding suggests that the startle response did not differentially decrease to a larger extent in the reactivation-propranolol condition than in the extinction-propranolol condition.
In addition, it was tested whether three unsignaled shocks (reinstatement) led to more return of fear in the extinction-propranolol condition than in the reactivation-propranolol condition, by comparing startle responses of two trials before and after reinstatement. No main effect of Trial, F (1,13) = 2.97, p = 0.11, and Stimulus x Trial effect was observed, F (1,13) < 1, indicating that startle responses did not differentially increase after reinstatement. In addition, no increase of differential startle response in the extinction-propranolol condition compared to the reactivation-propranolol condition was found, Stimulus x Trial x Condition,
F (1,13) = 1.72, p = 0.21. Therefore, contrary to the first hypothesis, administration of
propranolol after reactivation did not seem to result in more fear reduction than administration after extinction.
Extinction-propranolol vs. extinction-placebo
Acquisition. A mixed ANOVA revealed that startle responses to CS1 trials increased, compared to CS2 trials during acquisition in the propranolol and
extinction-placebo condition, Stimulus x Trial F (1,15) = 18.12, p = 0.001. This differential increase did not differ between condition, Stimulus x Trial x Condition, F (1,5) < 1. Therefore, acquisition was successful and did not differ between the extinction-propranolol and extinction-placebo conditions.
Extinction. A significant main effect of Trial was found in the extinction-propranolol and extinction-placebo conditions on day 2, F (1,15) = 21.16, p < 0.001. The overall decrease
in startle response (see Figure 3) did not differ between the two conditions, Trial x Condition,
F (1,15) < 1. Furthermore, no Stimulus x Trial effect was observed F (1,15) < 1, suggesting
that participants only showed a general reduction in fear potentiated startle response that did not differ between CS1 and NA trials. Therefore, clear indications of habituation were found, but surprisingly extinction training did not seem to occur.
Test. A significant Stimulus x Trial effect was found from the end of acquisition to the start of extinction on day 3, F (1,15) = 7.02, p = 0.02. However, no Stimulus x Trial x
Condition effect was observed, F (1,15) = 1.52, p = 0.24, indicating that differential startle responses did not return to a larger degree when propranolol, compared to placebo, was administered after extinction training (see Figure 3).
It was also tested whether administration of propranolol resulted in more return of fear from the end of extinction on day 2 to the start of extinction on day 3 than placebo. No main effect of Trial was found, F (1,15) = 2.43, p = 0.14. This finding suggests that the fear potentiated startle responses did not increase from day 2 to day 3 and in general no return of fear took place after extinction training. Also no Stimulus x Trial x Condition effect was found, F (1,15) = 2.53, p = 0.13, indicating that differential startle responses did not increase to a larger degree in the extinction-propranolol condition than in the extinction-placebo condition.
Finally, it was tested whether three unsignaled shocks resulted in an increase of startle response in the extinction-propranolol condition, compared to the
startle responses did not differentially increase after reinstatement. In addition, these startle responses did not differ across conditions, Stimulus x Trial x Condition, F (1,14) < 1. Hence, in contrast to hypothesis two, administration of propranolol after extinction training did not seem to result in more return of fear than administration of placebo.
Figure 3
Fear potentiated startle response during acquisition, extinction and test phases in the
reactivation-propranolol (A), extinction-propranolol (B) and extinction-placebo condition (C).
0 200 400 600 800 1000 1200 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Fear Pot e n tiated S tar tle
Reactivation-propranolol n = 7
CS1 trials CS2 trials NA trials
Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3 0 200 400 600 800 1000 1200 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Fear Pot e n tiated S tar tle
Extinction-propranolol n = 9
CS1 trials CS2 trials NA trials
Acquisition Day 1 Extinction Day 2 Extinction Day 3 Test Day 3 0 200 400 600 800 1000 1200 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Fear Pot e n tiated S tar tle
Extinction-placebo n = 8
CS1 trials CS2 trials NA trials
Acquisition Day 1 Extinction Day 2 Extinction Day 3 Test Day 3
A
B
C
US expectancy ratings
Reactivation-propranolol vs. extinction-propranolol
Acquisition. See Figure 4 for the US expectancy ratings per condition. Differential expectancy ratings were observed on day 1 in the reactivation-propranolol and extinction-propranolol condition, Stimulus x Trial, F (1,14) = 347.62, p < 0.001. In addition, there was no Stimulus x Trial x Condition effect, F (2,21) < 1, suggesting that participants in the two conditions learned to a similar extent that CS1 was followed by an US and CS2 was not.
Reactivation. The CS1 expectancy ratings at the end of acquisition and reactivation did not differ, Trial, F (1,6) = 2.42, p = 0.17. This result indicates that participants in the reactivation-propranolol condition remembered that CS1 was followed by an US after acquisition. Therefore, finding that conditioned fear to CS1 was not well consolidated (see fear potentiated startle results) cannot be explained by cognitive learning of the CS-US contingency.
Test. A differential decrease in expectancy ratings from the end of acquisition to the start of extinction on day 3 was observed, Stimulus x Trial, F (1,14) = 12.07, p < 0.01. The decrease of CS1 expectancy ratings, compared to CS2 expectancy ratings (see Figure 4) did not differ between conditions, Stimulus x Trial x Condition, F (1,14) < 1, suggesting that administration of propranolol after reactivation did not result in different US expectancy ratings than administration after extinction training. Three unsignaled shocks brought about a general increase in CS1 and CS2 expectancy ratings, F (1,13) = 15.78, p < 0.01. However, no differential increase was observed, Stimulus x Trial, F (1,13) = 1.47, p = 0.25. In addition, these expectancy ratings did not differ across conditions, Stimulus x Trial x Condition, F (1,13) = 2.59, p = 0.13. Therefore, administration of propranolol after reactivation and extinction did not seem to result in different expectancies of US occurrence.
Extinction-propranolol vs. extinction-placebo
Acquisition. Expectancy ratings at CS1 trials increased to a larger degree than CS2 trials in the extinction-propranolol and extinction-placebo condition, Stimulus x Trial, F (1,15) = 277.32, p < .001. No Stimulus x Trial x Condition effect was found, F (1,15) < 1, indicating that participants in the two conditions did not differ in successful learning of the CS-US contingencies.
Extinction. On day 2, expectancy ratings at CS1 trials decreased significantly in the two extinction conditions, F (1,15) = 233.08, p < 0.001. No Trial x Condition effect was found, F (1,15) = 2.97, p = 0.11, indicating that the extinction conditions did not differ in learning that CS1 was not followed by an US during extinction training.
Test. An increase in CS1 expectancy ratings from the end of extinction on day 2 to the start of extinction on day 3 was observed in the extinction-propranolol and extinction-placebo condition, Trial, F (1,15) = 83.79, p < 0.001. This increase did not differ between conditions, Condition x Trial, F (1,15) < 1, indicating that recovery of CS1 expectancy ratings was not influenced by administration of propranolol. During extinction on day 3, expectancies at CS1 trials decreased to a larger degree than CS2 trials in the two conditions, Stimulus x Trial, F (1,15) = 44.23, p < 0.001. However, no Stimulus x Condition, F (1,15) < 1, nor a Stimulus x Trial x Condition effect, F (1,15) < 1, was found. These results suggest that, in contrast to the findings of Bos et al. (2012), administration of propranolol on day 2 did not influence
expectancy ratings during extinction on the next day. In addition, expectancy ratings
= 17.15, p < 0.01 and increased after three unsignaled shocks on day 3, Stimulus x Trial, F (1,14) = 5.81, p = 0.03. However, the differential decrease from day 1 to day 3 and the reinstatement effect did not differ between conditions, Stimulus x Trial x Condition, F (1,15) < 1; Stimulus x Trial x Condition, F (1,14) < 1, respectively. These results indicate that, surprisingly, administration of propranolol after extinction training did not seem to influence expectancies of US occurrence.
Figure 4
US expectancy ratings during acquisition, extinction and test phases in the reactivation-propranolol (A), extinction-reactivation-propranolol (B) and extinction-placebo condition (C).
-100 -50 0 50 100 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 US e xp e ctan cy r atings
Reactivation-propranolol n = 7
CS1 trials CS2 trials Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3 -100 -50 0 50 100 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 US e xp e ctan cy r atingsExtinction-propranolol n = 9
CS1 trials CS2 trials Acquisition Day 1 Extinction Day 2 Extinction Day 3 Test Day 3 -100 -50 0 50 100 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 US e xp e ctan cy r atingsExtinction-placebo n = 8
CS1 trials CS2 trials Acquisition Day 1 Extinction Day 2 Extinction Day 3 Test Day 3A
B
C
Exploratory analyses
Reinstatement reactivation-propranolol vs. extinction-propranolol
With visual inspection of the fear potentiated startle data, it can be seen that
reinstatement seemed to result in an increase in differential startle responses in the extinction-propranolol condition, whereas startle levels to CS1 and CS2 remained relatively stable in the reactivation-propranolol condition (see Figure 3). Confirmatory reinstatement analyses involved averaging startle levels of the two trials before and after reinstatement. However, there seemed to be quite a large decrease in startle response from the first to the second CS1 trial after reinstatement in the extinction-propranolol condition. This rapid decrease is not surprising, because participants in this condition already received a total of twenty extinction trials (extinction on day 2 and 3). Nonetheless, averaging of the trials could have neutralized a possible reinstatement effect on differential startle responses, even though startle to the first CS1 trial after reinstatement might have been elevated in the extinction-propranolol condition, compared to the reactivation-propranolol condition. For this reason, two additional analyses were conducted. Without averaging the two trials before and after reinstatement, no
significant Stimulus x Trial x Condition effect was found, F (1,13) = 2.46, p = 0.12. However, when comparing the average of the last two trials and the first trial after reinstatement, the Stimulus x Trial x Condition effect was significant, F (1,13) = 7.38, p = 0.02. This finding could indicate that, as expected, administration of propranolol after extinction training resulted in less fear-reducing effects than administration after reactivation. However, this effect was only found with a single exploratory analysis and not with confirmatory analyses. Therefore, the finding needs to be interpreted with caution.
Trait anxiety
High trait anxiety is associated with reduced fear-attenuating effects of disruption of the reconsolidation process (Soeter & Kindt, 2013). Therefore, one could suspect that
participants with high trait anxiety in the reactivation-propranolol condition showed less fear reduction than participants with lower trait anxiety. For this reason, fear potentiated startle data and trait anxiety of individual participants were inspected. One participant had a specifically high score on trait anxiety (46), compared to the group mean (33). Soeter and Kindt (2013) defined high trait anxiety as a score in the highest quartile of the STAI-T (higher than 40). Therefore, this individual participant can be regarded as possessing high trait
anxiety. An additional plot with the average startle response, excluding the highly trait anxious participant, is displayed in Figure 5B. When comparing this pattern with the data of the complete sample in Figure 5A, one can suspect that the startle responses to CS1 and CS2 on day 3 are more similar when only participants with low or average trait anxiety are taken into account, than when also the participant with high trait anxiety is included. With
exploratory analyses it appears that a differential startle response (CS1 vs. CS2) on the first two extinction trials on day 3 is somewhat more noticeable with the complete sample, t (6) = 2.056, p = 0.09 than without the highly trait anxious participant, t (5) = 1.533, p = 0.19. In addition, the pattern of startle responses on day 3 in Figure 5B, compared to Figure 5A, is visually more similar to data of reactivation-propranolol conditions in previous studies (e.g. Sevenster et al., 2012; Kindt et al., 2009). This difference, based on the participant with high trait anxiety is, therefore, in line with high trait anxiety being associated with reduced fear-attenuating effects (Soeter & Kindt, 2013) and could perhaps explain that with the small sample size of the present study, administration of propranolol after reactivation of the fear memory did not seem to be as effective in reducing fear as in previous research.
In addition, the US expectancy ratings of the highly trait anxious participant and the other six participants in the reactivation-propranolol condition with average or low trait anxiety were compared (see Figure 6). Notably, a nearly significant decline in US
expectancies at CS1 trials for the six participants was found from reactivation to the start of extinction on day 3, F (1,5) = 5.66, p = 0.06, whereas the participant scoring high on trait anxiety recorded a maximal expectancy of US occurrence on both trials. Furthermore, with visual inspection of the first two CS1 extinction trials on day 3, it seems that participants with average or low trait anxiety showed a decrease in US expectancy ratings, while the highly trait anxious participant again indicated to be sure (as evidenced by two maximal scores) that an US would be administered. Based on these findings, one could suspect that people with high trait anxiety are relatively resistant to extinction training and need prolonged exposure in order to cognitively learn that a feared stimulus is no longer followed by an US. However, it must be emphasized that only one participant in the reactivation-propranolol condition exhibited high trait anxiety and, therefore, no firm statistical tests could be conducted to thoroughly analyze a role of trait anxiety in both the startle and US expectancy data. For this reason, interpretations based on trait anxiety scores need to be regarded as speculative.
Figure 5
Fear potentiated startle response in the reactivation-propranolol condition with the complete sample (A) and with exclusion of one participant scoring high on trait anxiety (B).
0 200 400 600 800 1000 1200 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Fear Pot e n tiated S tar tle
Reactivation-propranolol n = 7
CS1 trials CS2 trials NA trials
Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3 0 200 400 600 800 1000 1200 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Fear Pot e n tiated S tar tle
Reactivation-propranolol n = 6
CS1 trials CS2 trials NA trials
Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3
A
B
Figure 6
US expectancy ratings of a single participant scoring high on trait anxiety (B) and of the other participants in the reactivation-propranolol condition (A).
Discussion
The aim of the present study was to investigate two questions. The first question was whether extinction forms a boundary condition for disruption of the reconsolidation process. Second, it was studied whether administration of propranolol blocks fear-inhibiting effects of extinction training. No outcomes of confirmatory analyses were in line with either of the two predictions. Although an exploratory finding showed a larger return of fear after reinstatement when propranolol was administered after extinction compared to reactivation, altogether no clear support was found for the first hypothesis. In addition, no differences in fear reduction were observed between administration of propranolol and placebo after extinction, suggesting that propranolol does not interfere with extinction training. Therefore, based on the present study, prolonged reactivation of fear memories does not seem to pose a risk to treating anxiety disorders with disruption of the reconsolidation process.
However, due to a small sample size the study exhibited low statistical power. When
-100 -50 0 50 100 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 US e xp e ctan cy r atings
Reactivation-propranolol n = 6
CS1 trials CS2 trials Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3 -100 -50 0 50 100 1 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 US e xp e ctan cy r atingsReactivation-propranolol n = 1
CS1 trials CS2 trials Extinction Day 3 Reactivation Day 2 Acquisition Day 1 Test Day 3A
B
assuming medium effect sizes, a total sample size of 24 participants results in power being .39. Therefore, chances of detecting effects with the hypothesis tests were relatively small. In light of the power problem, it seems noteworthy that with exploratory tests administration of propranolol following extinction training resulted in more return of fear after reinstatement than administration following reactivation. This finding can be regarded as support for extinction being a boundary condition for disruption of the reconsolidation process. Nevertheless, since confirmatory analyses were not in line with this prediction, one cannot regard the hypothesis as confirmed. To what extent the surprising difference in blood pressure reduction between the reactivation-propranolol and extinction-propranolol condition has influenced the confirmatory tests is not completely clear. Physiological effects of propranolol have been shown to be unrelated to reducing effects of disruption of the reconsolidation process on fear potentiated startle (Soeter & Kindt, 2013). For this reason, it seems unlikely that a difference in blood pressure reduction has influenced the present findings of startle responses in the two conditions. Therefore, one cannot regard it is a plausible explanation of not finding support for the first hypothesis.
With respect to the second hypothesis, no support was found on any test. Just as in the study of Bos et al. (2012) propranolol did not seem to interfere with fear-reducing effects of extinction. However, both studies share some methodological issues. First, no clear extinction training seemed to occur on day 2 in the two studies, because startle levels to CS1 did not decrease to a larger extent than to NA trials. Therefore, as suggested by Bos et al. (2012), the procedure on the second day might have resulted in mere habituation, instead of extinction training. If indeed no extinction took place, administration of propranolol could clearly not have prevented any fear-reducing effects. In addition, Bos et al. (2012) found a general increase in fear from the end of extinction on day 2 to the start of extinction on day 3. This finding indicates that fear returned in both extinction conditions and, therefore, a ceiling effect could have prevented finding an effect of propranolol on fear recurrence. To overcome this problem, the number of unreinforced CS1 trials during extinction on day 2 was increased in the present study. However, with visual inspection of the startle data of both extinction conditions, one can suspect that also in this study a general fear increase from day 2 to day 3 took place. With a corresponding p-value of .14 (see results, fear potentiated startle,
extinction-propranolol vs. extinction-placebo) and low statistical power, one cannot rule out the possibility that, as in the study of Bos et al. (2012), fear returned in both extinction conditions and a ceiling effect might have occurred. Therefore, in future research on the effects of noradrenergic blockade on extinction it seems advisable to prolong the extinction phase even more. Since extending a single extinction session as in the present study might not be sufficient, one could consider incorporating multiple days of extinction training so that likely a clear fear reduction takes place when a placebo is administered. By doing so, a possible interfering effect of propranolol on extinction could be studied in a better way.
In contrast to the findings of Bos et al. (2012), administration of propranolol did not seem to interfere with expectancies of US occurrence. This discrepancy might be explained by the timing of pill administration, because participants in the present study received propranolol or placebo after extinction training, whereas Bos et al. (2012) administered pills beforehand. Bos et al. (2012) found that propranolol resulted in reduced expectancy ratings at CS1 trials during extinction training, an effect that was maintained one day later. They
suspected that propranolol does not per se interfere with consolidation, but perhaps with encoding of declarative fear memory when propranolol is administered before and active during extinction training. In the present study propranolol was administered after extinction
and no effects were found on US expectancy ratings. Therefore, it seems more apparent that the attenuating effect of propranolol on expectancy ratings in the study of Bos et al. (2012) merely reflects reduced learning, instead of consolidation. This contributes to a broader understanding of the effects of propranolol on declarative memory. It seems that propranolol can only interfere with cognitive fear learning when it is active during extinction training via diminished encoding. Hence, no effects on US expectancies are found when it is administered after extinction training and prior to or following reactivation, as is consistently found in previous fear-conditioning studies on reconsolidation (e.g. Kindt et al., 2009; Sevenster et al., 2012).
Another minor, though interesting indication that corresponds with previous research is that high trait anxiety seems to be related to fear-attenuating effects of disruption of the reconsolidation process. Soeter and Kindt (2013) showed that with higher trait anxiety, administration of propranolol prior to or after reactivation results in less fear reduction. This finding has important implications, because people with high trait anxiety are more prone to develop an anxiety disorder (Mineka & Oehlberg, 2008) and are, therefore, actually the people of interest when developing a new treatment. Soeter and Kindt (2013) suggested that a single unreinforced CS presentation might be insufficient to violate an expectancy of threat in highly trait anxious people and therefore reconsolidation is not triggered maximally under this circumstance. Following this line of thought, one could obviously suspect that prolonged reactivation sessions are better suited to people with high trait anxiety. Yet, as shown by Sevenster et al. (2014), prolonged reactivation sessions seem to result in the reconsolidation window to close and new learning to occur with no fear-attenuating effects of propranolol administration as a consequence. Nevertheless, one could argue that, just as with inducing reconsolidation, also new learning requires prolonged exposure to the feared stimulus in highly trait anxious people. After all, the exploratory analyses of US expectancy ratings suggested that people with high trait anxiety need prolonged extinction to cognitively learn that a feared stimulus is no longer followed by an aversive stimulus. This finding corresponds well with previous research indicating that PTSD and panic disorder patients show delayed extinction on cognitive measures of anxiety, such as US expectancy and valence ratings (Blechert, Michael, Vriends, Margraf & Wilhelm, 2007; Michael, Blechert, Vriends, Margraf & Wilhelm, 2007). In addition, a brain imaging study has shown that people with high trait anxiety display, compared to controls, enhanced activity of the amygdala and anterior
cingulate cortex during extinction training (Barrett & Armony, 2009). Taken together, one can suspect that in highly trait anxious people, the reconsolidation window is shifted compared to people with average or low trait anxiety. More specific, perhaps a longer exposure to the feared stimulus is needed to induce both reconsolidation and new learning (i.e. closing of the reconsolidation window) in these people. This hypothesis could be interesting to test in future research. As a first step, one could consider studying how many reactivation sessions are needed in order for expectancies of US occurrence to drop in highly trait anxious people. Since the single participant with high trait anxiety in the present study showed a maximal expectancy of US occurrence at reactivation and the first two extinction trials (see Figure 6B), three reactivation trials can be regarded as an approximation at this point. By studying larger samples of people with high trait anxiety, it might become apparent exactly how many more reactivations of the fear memory are needed to induce new learning in highly trait anxious people. Based on the result, an indication of the extent to which reactivation sessions can be multiplied before the reconsolidation window closes would be obtained. Then an experiment could be conducted where fear-reducing effects of administering propranolol after
a single and multiple reactivation sessions are compared in people with high trait anxiety. Thereby, procedures to disrupt the reconsolidation process could perhaps be adapted so they are better suited to treat people that are prone to develop an anxiety disorder.
To conclude, based on the present data one cannot infer that extinction is a boundary condition for disruption of the reconsolidation process, nor that propranolol interferes with extinction training. A power problem and procedural issues could have prevented finding support for these hypotheses. Usage of multiple extinction sessions in future research could contribute to better test the effects of propranolol on extinction. Another important direction for future research is the role of trait anxiety in procedures to disrupt the reconsolidation process. With the provided suggestions, a better understanding might be obtained of how this promising new avenue for treating anxiety disorders can be developed for efficacious
application in clinical practice.
References
Alonso, J., Angermeyer, M. C., Bernert, S., Bruffaerts, R., Brugha, T. S., Bryson, H. et al. (2004). Disability and quality of life impact of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatrica Scandinavica, 109(420), 38-46.
Barlow, B.H. (2008). Clinical handbook of psychological disorders: A step-by-step treatment
manual. New York: The Guilford Press.
Barrett, J., & Armony, J. L. (2009). Influence of trait anxiety on brain activity during the acquisition and extinction of aversive conditioning. Psychological Medicine, 39(02), 255-265.
Blechert, J., Michael, T., Vriends, N., Margraf, J., & Wilhelm, F. H. (2007). Fear conditioning in posttraumatic stress disorder: Evidence for delayed extinction of autonomic, experiential, and behavioural responses. Behaviour Research and
Therapy, 45(9), 2019-2033.
Bos, M. G., Beckers, T., & Kindt, M. (2012). The effects of noradrenergic blockade on extinction in humans. Biological Psychology, 89(3), 598-605.
Bouton, M. E. (2002). Context, ambiguity, and unlearning: Sources of relapse after behavioral extinction. Biological Psychiatry, 52(10), 976-986.
Bouton, M. E., Westbrook, R. F., Corcoran, K. A., & Maren, S. (2005). Contextual and temporal modulation of extinction: Behavioral and biological mechanisms. Biological
Psychiatry, 60(4), 352-360.
Dębiec, J., & Ledoux, J. E. (2004). Disruption of reconsolidation but not consolidation of auditory fear conditioning by noradrenergic blockade in the
amygdala. Neuroscience, 129(2), 267-272.
Dudai, Y. (2009). Predicting not to predict too much: How the cellular machinery of memory anticipates the uncertain future. Philosophical Transactions of the Royal Society B:
Biological Sciences, 364(1521), 1255-1262.
Hamm, A. O., & Weike, A. I. (2005). The neuropsychology of fear learning and fear regulation. International Journal of Psychophysiology, 57(1), 5-14.
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
593-602.
Kindt, M., Soeter, M., & Vervliet, B. (2009). Beyond extinction: Erasing human fear responses and preventing the return of fear. Nature Neuroscience, 12(3), 256-258. 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(3), 401-409.
Lang, P.J., Bradley, M.M., Cuthbert, B.M., (2005). International Affective Picture
System (IAPS): Affective ratings of pictures and instruction manual. University of
Florida, Gainesville.
Lee, J. L. (2009). Reconsolidation: maintaining memory relevance. Trends in
Neurosciences, 32(8), 413-420.
Lee, J. L., Milton, A. L., & Everitt, B. J. (2006). Reconsolidation and extinction of
conditioned fear: Inhibition and potentiation. The Journal of Neuroscience, 26(39), 10051-10056.
Metzger, R. L. (1976). A reliability and validity study of the State-Trait Anxiety Inventory. Journal of Clinical Psychology, 32(2),276-278.
Michael, T., Blechert, J., Vriends, N., Margraf, J., & Wilhelm, F. H. (2007). Fear conditioning in panic disorder: Enhanced resistance to extinction. Journal of Abnormal
Psychology, 116(3), 612-617.
Mineka, S., & Oehlberg, K. (2008). The relevance of recent developments in classical
conditioning to understanding the etiology and maintenance of anxiety disorders. Acta
Psychologica, 127(3), 567-580.
Mueller, D., & Cahill, S. P. (2010). Noradrenergic modulation of extinction learning and exposure therapy. Behavioural Brain Research, 208(1), 1-11.
Muris, P., & Merckelbach, H. (1996). A comparison of two spider fear
questionnaires. Journal of Behavior Therapy and Experimental Psychiatry, 27(3), 241-244.
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.
Peterson, R.A., Reiss, S., (1992). Anxiety Sensitivity Index Manual. Worthington, OH: International Diagnostic Systems.
Przybyslawski, J., & Sara, S. J. (1997). Reconsolidation of memory after its reactivation. Behavioural Brain Research, 84(1), 241-246.
Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A.H. Black, W.F. Prokasy (Eds.), Classical conditioning II: Current research and theory (pp. 64-99). New York: Appleton-Century-Crofts.
Reiss, S., Peterson, R. A., Gursky, D. M., & McNally, R. J. (1986). Anxiety sensitivity, anxiety frequency and the prediction of fearfulness. Behaviour Research and
Therapy, 24(1), 1-8.
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.
Sevenster, D., Beckers, T., & Kindt, M. (2014). Prediction error demarcates the transition from retrieval, to reconsolidation, to new learning. Learning & Memory, 21, 580-584. Soeter, M,, & Kindt, M. (2012). Erasing fear for an imagined threat event.
Soeter, M., & Kindt, M. (2013). High trait anxiety: A challenge for disrupting fear memory reconsolidation. PLOS ONE, 8 (11), e75239.
Spielberger, R., Gorsuch, R., & Lushene, R. (1970). Manual for the State-Trait Anxiety
Inventory. Palo Alto, CA: Consulting Psychologists Press.
Szymanski, J., & O'Donohue, W. (1995). Fear of spiders questionnaire. Journal of Behavior
Therapy and Experimental Psychiatry, 26(1), 31-34.
Watson, J. B., & Rayner, R. (1920). Conditioned emotional reactions. Journal of
Experimental Psychology, 3(1), 1-14.
