MASTER THESIS
Department of Psychology, University of Amsterdam
A Reconsolidation-based Treatment for Fear of Heights
Name : Tom du Bois Address : Marowijnestraat 3 Zip code and residence : 1058VS, Amsterdam Telephone number : +31612050058
Student ID Card number : 10161945
E-mail address : Thomasjandubois@gmail.com
Words : 7689
Date : 22-07-2017 Supervisor : Jamie Elsey
Table of Contents
1. Abstract ... 3
2. Background ... 4
2.1
Hypotheses ... 7
2.2
Pilot study ... 7
2.3
Adjustment of hypotheses and expectations ... 8
3. Methods and Materials ... 9
3.1
Assessments ... 9
3.2
Behavioural Tests ... 10
3.3
Researchers ... 10
3.4
Participants ... 10
3.5
Procedure ... 10
3.6
Data analyses ... 11
4. Results ... 11
5. Discussion ... 14
6. References ... 18
1. Abstract
Background: In the current study, we aimed to reduce participants’ fear of heights by pharmacologically
disrupting the reconsolidation of fear memory after its reactivation.
Methods: Acrophobic participants (n=16) received a 40 mg dose of propranolol after their fear was reactivated
in a height scenario. In a second session, participants were exposed to the same situation again, and underwent a generalization test, to investigate how their reactions changed.
Results: Giving propranolol after fear memory reactivation reduced tension and anxiety in participants at later
testing. Since there was no control group, we could not guarantee whether this effect was caused specifically by the combination of propranolol and reactivation or other non-specific effects.
Conclusion: The results show the great potential of reconsolidation-based treatments in tackling long-standing
fears. However, the mechanism is not yet fully understood. Triggering a prediction error seems to be a vital part of therapy, and a focus for future research is how to guarantee this happens.
2. Background
Fear has an evolutionary function. It helps to adapt an individual to potential hazards in their surroundings. Anxiety will narrow an individual's behavioural repertoire (fight or flight) to maximise the individual's chance of survival (Fanselow, 1984). Fear memories can be obtained through Pavlovian fear
conditioning; this means that an initially neutral stimulus (the conditioned stimulus or CS+) is associated with an aversive stimulus (the unconditioned stimulus or the US). This is a so-called associative fear memory (Soeter & Kindt, 2015). People have a broad range of associative memories. Those fear memories can become
dysfunctional when they hinder people in their daily functioning, such as when someone becomes intensely afraid of harmless animals or situations that they regularly encounter.
Acrophobia (a specific phobia) is an extreme fear of heights DSM-IV (APA, 1994). Acrophobic people will often avoid height situations like stairs, balconies and bridges. The full range of conditions that can be experienced as dangerous for those people illustrates the fact that acrophobic people can feel impaired by their anxiety. According to research, 1 in 20 adults suffers from acrophobia (Coelho, Waters, Hine, & Wallis, 2009). Anxiety disorders, in general, are highly prevalent in the population (almost one-third of the population suffers from an anxiety disorder sometime in their life: Kessler et al., 2005). The most common way to treat anxiety disorders is cognitive behavioural therapy. CBT is not thought to alter the core fear memory but creates new memories that inhibit the old fear memory (Bouton, 2002). The technique that is used within those therapies is called exposure therapy. For example, people that suffer from panic attacks might have the idea that if they experience an episode of panic, they will get a heart attack. By exposing patients to their fear (by inducing a panic attack), but without the feared consequence taking place (heart attack), patients will learn new information. They can have a panic attack, without them having a heart attack. This process of gathering new information against the negative consequences of the feared stimuli is called inhibitory learning. If you repeat this procedure, anxiety reactions will typically drop. The goal is to make this new association (memory) stronger than the old ‘aversive' one. The questionable thing about inhibitory learning is that the core fear memory is not directly affected. If contextual factors change, the old fear memory could recover spontaneously (Bouton, 2002). A high relapse rate is the consequence of this approach. For this reason, CBT can be very time consuming, since inhibitory learning must occur across a range of situations, in order to prevent spontaneous recovery (Loerinc et al., 2015).
As illustrated above, for a long time, researchers were not trying to change the core fear memory, due to a belief that memories were unchangeable. A different approach to therapy for anxiety disorders came from neuroscientific research. Instead of forming new memories, researchers tried to influence/change the original memory that provoked fear. To explain this mechanism, it is important to know how memory development works.
Memory storage, better known as memory consolidation, is a mechanism in which information goes from a fragile, temporary state (where it is vulnerable to disruption) to a more stable and resistant state in long term memory (figure 1; McGaugh, 2000). Memory consolidation is a process that seems to be regulated through
long-term plasticity in which LTP (long-long-term potentiation) and LTD (long-long-term depression) have an important role in strengthening or weaken neuronal memory traces (Kandel, Dudai, & Hayford, 2014).
Figure 1. Lewis's memory model, in Nader and Hardt (2009).
The first modern-day researchers that showed interest in editing memory traces after retrieval were Przybyslawski and Sara (1997), who named this phenomenon reconsolidation. The research of Schafe and LeDoux (2000), redirected attention to this topic. They showed with the usage of a protein synthesis inhibitor (anisomycin), which was administered to the amygdala of rats, they could prevent the consolidation of the fear memory after a classic fear conditioning task. Those findings suggest that memory consolidation depends on protein synthesis. Additionally, they discovered that when they gave anisomycin directly after fear conditioning, it prevented the memory from consolidating, whereas the rats that had anisomycin 6 hours after learning retained a fear memory for the conditioned stimulus. This implies that memory consolidation is time dependent. Hence there seems to be a period in which the new memory trace is sensitive to change. Of especial interest, Nader, Schafe and LeDoux (2000) showed that memory that was already consolidated (stored in the long-term memory) was able to go back to this so-called vulnerable state, in which it was receptive to changes under the influence of anisomycin.
The central idea of reconsolidation is that memory can be destabilized after reactivation, producing a labile state in which the memory is vulnerable to disruption. An unstable memory may be updated with new
information, or prevented from restabilizing. We know that a protein synthesis inhibitor like anisomycin can influence the way memory traces get stored in the long-term memory and that it can create amnesia for a
conditioned response (Nader et al., 2000). The theory is that a protein synthesis inhibitor is capable of preventing the memory from going back to a stable state in the long-term memory (line in figure 2.) This way, the memory (the learned response) is weakened (see also figure 2).
This amnesic effect, under the influence of a protein synthesis inhibitor, was found in different kinds of species such as rats and Medaka fish (Eisenberg, Kobilo, Berman, & Dudai 2003). Also, Dudai (2006) argued that this process of reconsolidation is a general mechanism that is used to edit a broad range of memory types. Debiec and LeDoux (2004) discovered that using propranolol (a non-neurotoxic drug), a beta-adrenergic receptor (β-AR) blocker, was also capable of blocking the reconsolidation process of a conditioned fear response. It appears that the beta-adrenergic receptors (β-AR) play a key role in the cascade of intracellular events that lead to protein synthesis. Due to propranolol's receptor blocking properties, the medicine seems to be able to interfere with protein synthesis.
Propranolol was used in human experimental research to investigate if it would have the same reconsolidation blocking properties. Kindt, Soeter and Vervliet (2009) examined this in a Pavlovian fear
conditioning task. They found that propranolol weakened the fear behaviour that was initially triggered by a fear-relevant CS. Their outcome measure for anxiety was the fear-potentiated startle reflex, which is a measure that reflects a negative subjective emotional valence of a stimulus (Brown, Kalish, & Farber, 1951). Soeter and Kindt (2010) demonstrated that propranolol can weaken the emotional content of memory, through disrupting memory reconsolidation, leaving the declarative memory untouched. Also, they discovered that extinction did not cause the effect of the treatment. In a 1-month follow-up study, they found that the propranolol condition did not show any signs of the recovery of the initially learned fear response. However, the placebo-controlled group did show recovery for the fear response. This is explicable through the mechanism of spontaneous recovery. Sevenster, Beckers and Kindt (2013) found that the memory reconsolidation-blocking effect of
propranolol only works when there is a discrepancy between what is expected, and what the actual content of the stimulus is (match-mismatch experience). This so-called prediction error (Rescorla & Wagner, 1972) is vital for the consolidated memory to become destabilised. Only when there was a chance of acquiring new information, the propranolol showed its reconsolidation blocking effect (Sevenster et al., 2012).
As above discussed research suggests, a pharmacological intervention seems to have the potential to change/adapt memories. It is highly relevant to investigate whether propranolol is effective in shutting down profoundly emotional memories, which are often the core problem of anxiety related disorders (like PTSD and social phobia). Soeter and Kindt (2015) showed that they were able to successfully treat a specific phobia, which was not acquired in a laboratory but obtained in the participant's life. Participants showed severe anxiety during an exposure task with a tarantula (participants were selected on spider phobia), after which they received either propranolol or placebo. After treatment, in a follow-up trial, participants in the propranolol condition were able to interact with the spider, even touching the tarantula and letting a dwarf tarantula walk on their hands. Participants in placebo condition still showed severe anxiety. These outcomes illustrate the strong effect that a reconsolidation disrupting intervention can have on anxiety disorders and the clinical significance of it. Participants in the study reported that they could remember being afraid of the spider, but that the fear had disappeared. These findings support the theory earlier mentioned, that propranolol can weaken the emotional valence of memory through preventing memory from reconsolidating (Soeter & Kindt, 2010). In addition, Brunet et al., (2008) discovered that people who had PTSD showed less physiological anxiety responses (in comparison to placebo control group) in follow-up treatment when they got propranolol administered after a task in which participants had to recall a traumatic event the week before.
In the current study, we were planning to investigate whether a reconsolidation-based treatment was successful in treating fear of heights in participants within a randomised controlled trial. The use of propranolol appeared to be a promising intervention in the treatment of specific phobias. Our opinion was that it is
important and interesting to investigate whether this intervention was generalisable to other specific phobias as well. Additionally, we were interested in using Virtual Reality for psychological treatment. Virtual Reality is a computer based virtual environment, which uses body tracking sensors to let clients interact with their virtual environment. We were interested in the therapeutic utilities of VR. VR has the potential to expose patients to fearful situations that are somehow difficult to mimic in reallife. It can take a substantial amount of time and effort to find/create a suitable situation for exposure. For instance, it is imaginable that patients have to travel to a site to get proper exposure therapy. In this way, VR could be a good alternative for in vivo exposure, since you could mimic those situations within a digital environment. Furthermore, it has the potential to function like adequate stimuli that could cause fear in participants (which is needed for reactivation in studying
reconsolidation). Emmelkamp, Bruynzeel, Drost and van der Mast (2001) conducted research to investigate whether VR could replace a real life situation as a height stimulus. The study confirmed that exposure therapy in VR was as effective as exposure in vivo. In addition to that research Miloff et al., (2016), reviewed the use of VR within a cognitive behavioural therapy setting. They concluded that VR is useful in treating specific phobias.
The present study was initially designed as a randomised controlled trial, extending Soeter and Kindt's (2015) spider study. We were planning to investigate if propranolol would be successful in reducing anxiety in participants with acrophobia when they were exposed to a height situation. The active treatment group would receive propranolol right after they were exposed to the height situation, and subsequently, after a waiting period of 2 hours they would receive a placebo. The non-active treatment group would receive a placebo right after recall of the fear memory, and would then receive propranolol after 2 hours, such that propranolol reached peak levels outside of the time-window. Theory states that it takes up to 90 minutes for propranolol to peak in bioavailability after administration (Gillman & Goodman, 1996). This way, it could be investigated that memory reconsolidation only got disrupted within 2 hours after reactivation, under the influence of propranolol in the active treatment condition.
2.1 Hypotheses
1. We expected that participants in the active treatment group (propranolol/placebo) report less fear on questionnaires measuring fear related to heights in the post-treatment session, in comparison to the non-active treatment group (placebo/propranolol).
2. We expected that participants in the active treatment group (propranolol/placebo), can perform better, in several behavioural approach tasks (a task developed to assess the severity of anxiety for a particular situation) post-treatment. In comparison to the non-active treatment group (placebo/propranolol).
2.2 Pilot study
Before starting the controlled trials, there was a pilot conducted using a VR setup with 6 participants to check whether manipulations were strong enough in the active treatment group. In the VR environment, participants would take an elevator to the top floor of a building, where they were facing a height of 160 meters.
Participants were then asked to walk up and down a plank that was sticking out of the elevator door. In the test room, there was an original plank on the floor to make the situation more realistic.
The treatment effects that came out of the pilot study were inconsistent. Some pilot participants showed no effect at all. Other participants showed a moderate to strong effect where anxiety dropped substantially in post-treatment in comparison to pre-treatment. Since the effects of the treatment were inconsistent, we were not convinced that testing a large number of participants in a randomized trial was yet warranted. What happens in reactivation can affect treatment, which makes the outcome unpredictable. A better understanding of what happens is important and relevant for the therapeutic usability of reconsolidation focused therapies. Apparently, there were components in the design of the study that prevented the fear from decreasing every time we tested.
A concern that arose was the level of fear that got provoked using the VR plank set-up. The reactivation, in which participants had to balance on a 20-centimeter-wide, unstable plank sticking out of an elevator on the 20th floor, was possibly to strong. It could be argued that the fear that got provoked was not comparable to the fear that people with fear of heights would experience. Participants that would score below subclinical on acrophobia also found the scenario frightening. Additionally, there was another concern with the VR plank set-up. Reactivation in reconsolidation studies should not be much longer than an estimated five minutes because of the lurking effect of inhibitory learning (getting exposed to a fearful situation, without the feared consequence taking place). During the pilot, we noticed that people took a long time to take the first step on the plank, and secondly, to move further on the plank and back. Frequently this exceeded the time that the participant was allowed to be exposed to the stimulus.
The inconsistent effect of the manipulation during the pilot study forced us in making adjustments to the reactivation procedure. Attempts to translate lab work into significant clinical interventions are difficult but crucial for the advancement of new therapeutic interventions. We decided to look for a more ‘natural' stimulus that provoked enough fear for our participants, in a situation that was more likely to occur in the participants daily lives. A suitable location was found at a balcony on the third floor of building L, at REC UvA. Using this set-up for reactivation presented other difficulties since participants varied substantially in their levels of fear. The balcony did not always provoke the desired amount of fear for reactivation. Hence we started looking for a situation in the middle of the first two set-ups regarding levels of provoked fear. The second variation in set-up we used in the experiment was the VR setup with a ledge instead of a plank. This way participants did not have to perform an active task (walking out on the plank like in VR-plank set-up), and at the same time, we were more in control over the time the participants were exposed to the height situation. During the study, based on levels of fear participants reported in the questionnaires and the SCID, the supervisor of the study decided which reactivation the participants would get (Balcony or VR ledge). Furthermore, it was decided to experiment with different waiting periods after reactivation, to see whether those variations affected outcome measures. Participants were randomly assigned to a 1 or 3 hour waiting period after reactivation.
2.3 Adjustment of hypotheses and expectations
We decided not to continue with the randomised controlled trial (we could not compare active with non-active treatment), due to a lack of a consistent manipulation effect in the non-active treatment group during piloting. To make valuable analyses of the acquired data of the active treatment group, however, we adjusted our
hypotheses and expectations. The hypotheses that substantiate expectation 1, 2 and 3 is that propranolol will show its anxiety reducing properties within participants who are afraid of heights.
Expectation 1: As a result of the treatment, we expect that participants score significantly lower on the anxiety questionnaires (AQ and ATHQ) in post-treatment in comparison to pre-treatment.
Expectation 2: As a result of treatment we expect that reported anxiety in the behavioural performance task (fire escape stairs) will be significantly lower in post-treatment in comparison to pre-treatment sessions.
Expectation 3: As a result of treatment, we expect that self-reported distress in the post-treatment session is significantly lower in comparison to the distress numbers within the treatment session.
3. Methods and Materials
3.1 Assessments
Questionnaires
Acrophobia Questionnaire (AQ; Cohen, 1977). The AQ is a self-report 40 item questionnaire which measures fear and avoidance in different imaginary height situations. The AQ has two subscales, one measures anxiety (0-120), this subscale is scored from 0 (not anxious) to 6 (extremely anxious). The other sub-scale measures avoidance (0-60), this scale can be scored from 0 to 2. Those numbers represent: 0 (I would not avoid the situation), 1 (I would try to avoid the situation), and 3 (I would definitely avoid the situation). The Cronbach Alpha for the anxiety subscale is .80, and the Cronbach Alpha for avoidance is .70. The questionnaire is
extensively used and has good psychometric characteristics (Baker, Cohen, & Saunders, 1973).
Attitudes Towards Heights Questionnaire (ATHQ; Abelson & Curtis, 1989). The ATHQ contains 6 questions in which the attitude towards height situation is measured from 0 tot 10. 0 stands for safe/good (positive connotation) or comparable words, were 10 stands for bad/dangerous (negative connotation). The score range of the questionnaire is 0-60, and has a Cronbach Alpha of .81. The ATHQ is a widely used test which seems to be responsive to treatment (Emmelkamp et al., 2001).
The Spiegelberg State-Trait Anxiety Inventory (STAI; Spiegelberg, 1970). The STAI is a 40 item questionnaire which is divided into two subscales. The STAIT is the subscale that measures how anxious participants are in general. The STAIS measures how anxious participants are at the moment they are filling out the questionnaire. The STAI has proven to be a reliable and sensitive measure of anxiety. It has a Cronbach Alpha of .89.
Anxiety Sensitivity Index (ASI; Reiss & McNally, 1986). A 16 item questionnaire developed to measure people’s beliefs about what consequences symptoms of anxiety have on their life (social and somatic). The questionnaire has proven to be a sensitive and valid measure in assessing the impact of anxiety (Peterson & Heilbronner, 1987). The questionnaire has a Cronbach Alpha of .88.
Beck Depression Inventory (BDI; Beck, Steer, & Brown, 1996). The BDI is a 21 item self-score questionnaire, which is used to assess the severity of depression in adults and adolescents. It has proven good reliability and validity. It has a Cronbach Alpha of .92.
3.2 Behavioural Tests
Fire escape stairs. This assessment was developed to investigate how well the treatment effects were generalised to real life height situations. The assessment took place at an outside metal fire escape that goes up 4 stories, with 7 flat levels. Participants were asked to go up as far as they were comfortable. At every flat point of the stairs, participants were asked to rate their anxiety levels from 0 to 100. Participants were rated on how far they went up, counting the flat places on the stairs (0-lowest to 7-highest). The fire escape stairs have been proven a sensitive measure for fear of heights in another study treating acrophobia (Emmelkamp et al., 2001).
Richie’s Plank Experience (Toast, 2016). A virtual reality environment in which participants were exposed to a height experience. In the VR environment, participants could take the elevator to the top floor, where they were facing a height of 160 meters. Participants could walk up and down a plank, while exposed to the height.
Within the current study, we used two different variations of a height situation as reactivation.
1. Ledge performance during Richie’s Plank Experience. In this version of the reactivation, a plank was laid down crosswise, so it formed a ledge (in VR the participants also saw a ledge). The plank measured 200x40cm.
2. Balcony. The second variation on the reactivation was a balcony used on the third floor of building L, at REC UvA. It was a balcony with a safe, but reasonably low railing.
3.3 Researchers
The research was conducted by a PhD candidate and two master students.
3.4 Participants
A total of 16 participants took part in the study. 14 of those individuals were female, and 2 were male. The women ranged in age between 18 and 22 (mean ± SD age, 19.57 ± 1.22). Men ranged in age from 20 and 21 (mean ± SD age, 20.50 ± 0.71). All participants received propranolol. Participants received a small amount of money for taking part in the study or received course credits.
3.5 Procedure
The study was divided over three different sessions.
First session. The first session was intended for baseline measurements. In the first session, all participants underwent a medical screening (blood pressure, medication use, psychiatric history) to check if it was safe to take propranolol. Participants filled in five questionnaires: The Acrophobia Questionnaire (AQ), the Attitude
Towards Heights Questionnaire (ATHQ), the Beck Depression Inventory (BDI), the Anxiety Sensitivity Index (ASI), and the State-Trait Anxiety Inventory (STAI). The AQ and ATHQ were used to get a picture of the participants' acrophobia, and the other three mentioned were used to check for possible confounding factors. Subsequently, participants performed the behavioural test at the fire escape stairs.
Second session. The treatment session, this session had two variations during research to check if changes would make a difference on reactivation effects. Before treatment the extent in which participants were afraid of heights was examined using the SCID, the participants also had to fill in the STAI-State. After that, participants were exposed to a height situation (VR ledge or Balcony). While participants were facing the height situation,
they were asked a couple of questions. When bending over, and looking down they were asked what they were afraid might happen. Then, participants were asked to rate on a scale from 0 to 100 how sure this event would happen. Subsequently, participants were asked to focus on their body, and how much distress they felt (rating from 0 to 100), finally they were asked where in the body they felt the distress. Right after reactivation,
participants received propranolol. Then, participants had to wait in a room for a fixed amount of time (1 hour or 3 hours) while we monitored their response to the propranolol and checked they did not have an adverse reaction. While in the room, participants were not allowed to do demanding cognitive tasks. They could read a magazine or something comparable. Participants in the 3 hours waiting condition got a placebo pill administered, after 2 hours waiting.
Third session. This was the post-treatment session. Participants were asked to fill in The AQ, ATHQ and the STAI-State. Subsequently, participants were exposed to the same height situation as in session 2. Finally, participants performed the behavioural test at the fire escape stairs again.
3.6 Data analyses
To assess change over time a paired Paired Sample T-test was used, to analyse the results of the AQ-avoidance (AQ-A), AQ-fear (AQ-F), ATHQ, the distress numbers during treatment sessions (Tension) and the reported anxiety during the BAT (Rep. Anxiety BAT). For the anxiety during the BAT, we compared the reported anxiety on the highest level of the fire escape the participants reached in pre-treatment, with the reported levels of fear at the same level on the stairs in post-treatment. The height reached on the stairs was not deemed a useful measure, since the majority of participants could already reach the top of the stairs, or close to the top, before the treatment session.Finally, two Mixed Measures ANOVA’s were utilised for a more
exploratory framework. One was to conduct whether the condition VR-ledge/Balcony made a difference in treatment outcomes (reported tension during treatment sessions). The other Mixed Measures ANOVA was used to assess whether a 1hr vs. 3hr waiting period after administration of propranolol would make a difference in treatment outcomes (reported tension during treatment sessions).
The results of the questionnaires BDI, STAIS and ASI have not been analysed in the current study. Data on those questionnaires were collected for future research, to check whether these outcome measures might predict therapy outcomes.
4. Results
The P-Plot's for the variables (AQ-A, AQ-F, ATHQ, Tension and Rep. Anxiety BAT) did not show normally distributed data, so the data was presumably non-parametric. For this reason, the Paired Sample T-test was exchanged for a non-parametric Wilcoxon-Signed Ranks Test. In Table 1, the results are shown for the different variables. Table 1 shows that the median of the AQ-A post-treatment, Mdn = 12.00, was not significantly lower in comparison to the median of the AQ-A pre-treatment, Mdn = 13.50. This implies that participants did not report a substantial drop in avoidance after treatment. Table 1 does show that the median of the AQ-F post-treatment, Mdn = 44.50, was significantly lower in comparison to the median of the AQ-F pre-treatment, Mdn = 49.00. Participants did report a substantial drop in fear levels after the experiment. It also shows that the median of the ATHQ in post-treatment, Mdn = 37.50, was significantly lower in comparison to
the median of the ATHQ in pre-treatment, Mdn = 41.00. This suggests that participants showed a considerable drop in reported negative associations towards heights. Further, the median of reported tension in the second treatment session, Mdn = 65.00, was significantly lower in comparison to the median of the reported tension in the first treatment session, Mdn = 80.00. This shows that participants reported substantially less distress after treatment. Finally, the median of reported anxiety during the BAT in post-treatment, Mdn = 65.00, was
significantly lower in comparison to the median of reported anxiety in pre-treatment, Mdn = 75.00. Participants reported substantially less anxiety during the BAT after the experiment. The hypotheses that propranolol has anxiety reducing properties within participants who are afraid of heights, is largely supported by this data. Except for the avoidance part of anxiety.
Table 1. Results Wilcoxon-Signed Ranks Test
——————————————————————————————————————————— AQ-A AQ-F ATHQ Tension Rep. Anxiety BAT
———————————————————————————————————————————
Z-Score -1.33 -2.02 -2.60 -3.42 -3.31
Sig. .18 .04 .01 <.01 <.01
———————————————————————————————————————————
Before conducting Mixed Measure ANOVA’s. Assumptions of homogeneity of variance and sphericity were investigated over the dependent variables tension session 2 (TENSION 2) and tension session 3
(TENSION 3). A Levene’s test showed equal variance for the variables: TENSION 2 (F= 1.95, p= 0.19), TENSION 3 (F= 1.02, p= 0.42). Sphericity test is for this data not relevant since the ANOVA is comparing two levels of repeated measures.
A Mixed Measure's ANOVA with between subject factor scenario (Balcony vs. VRledge) and within-subject factor session (session 2 vs. session 3) was conducted. There was a significant main effect for session (F= 29.35, p <.01). There was no significant main effect for scenario (F= 1.48, p= 0.25). The analyses also showed no significant interaction effect between scenario and session (F= 0.28, p= 0.61). In other words, participants reported a substantial drop in tension when session 2 and session 3 are compared. The drop in reported tension, however, was not dependent on the scenario the participants were tested in. Results are shown in figure 3.
Figure 3. Reported tension during treatment, for different scenarios.
A second Mixed Measure's ANOVA was conducted, with between subject factor waiting period (1hr vs. 3 hr) and within-subject factor session (session 2 vs. session 3). There was a significant main effect for session (F= 35.58, p>0.00). The analyses showed no significant main effect for waiting period (F= 2.37, p= 0.15) and no significant interaction effect between session and waiting period (F= 0.316, p= 0.58). This implies, that participants reported a significant drop in tension when session 2 and session 3 are compared. The drop in reported tension, however, was not dependent on the period participants had to wait after the administration of propranolol. Results are shown in figure 4.
Figure 4. Reported tension during treatment, for different waiting periods.
5. Discussion
The current study aimed to investigate whether a reconsolidation-blocking therapy with a beta-adrenergic receptor (β-AR) blocker, propranolol could reduce fear of heights in people. The study provides some evidence for the fear-reducing properties of propranolol. Partly in line with expectations, participants reported
significantly lower levels of fear on the anxiety questionnaires after treatment with propranolol. Participants reported less fear on the ATHQ. Somehow, the treatment changed the way they perceived heights. In addition to that, participants reported less fear on the AQ sub-scale fear after therapy, but they did not report less avoidance on the sub-scale avoidance. There are two possible explanations for this; one is that participants already reported low avoidance in the baseline measurements, leaving little room for variation. Another possible explanation is that participants do no often face height situations that are mentioned in the avoidance subscale, like a Ferris wheel, or the roof of a high building, such that changes could not be observed. Furthermore, partially conforming to expectations, participants showed a substantial drop in fear while performing the behavioural approach task on the staircase. Likewise, the reported tension during treatment decreased
substantially after treatment. Further analyses, used for a more explorative purpose, suggests that Virtual Reality is a valid instrument for reconsolidation-based treatments. Like the balcony setting, participants that were exposed to a VR environment showed a considerable drop in reported tension after treatment, and this is
congruent with previous research. The variations in hours participants had to wait after the propranolol was administered did not make a clear difference. Likewise, the different scenario's participants were exposed to, did not make a difference in reported levels of physical distress in the post-treatment session. A remark has to be made, concerning this last point, since the supervisor selected the most suitable scenario for every individual participant, based on clinical experience. The main criterion was that the scenario would provoke enough fear to cause retrieval of the fear memory. It could be argued that the balcony scenario, which was in general perceived as less frightening, would not trigger sufficient fear memory reactivation in less fearful participants who had been exposed to the VR instead.
Propranolol seems to be capable of changing the way participants perceived heights in both a
psychological and a physical way. They showed changes in their fear response, where bodily tension dropped as a result of treatment. Besides that, participants reported changes in the way they perceived heights. However, the evidence is not entirely convincing due to several reasons. The reported results in general, are in line with Kindt and Soeter's (2015) research, people tend to interact more easily with the fearful stimuli after treatment with propranolol. The most important difference, is that we did not include a control group to check whether treatment effects were caused by the manipulation. As a result of the pilot study, we decided not to continue with a randomised controlled trial. This decision makes it hard to make any statement over the acquired data and observed effects. For instance, we can not guarantee that the observed reduction in fears, is not being caused by exposure to the feared stimuli. The mechanism which is used in exposure therapy. Nonetheless, Bouton (2002) stated that effective extinction training involves repeated exposure. Since participants were only exposed to the stimulus for less than 5 minutes, the chances are small that extinction caused the observed reduction in fear. It would have been interesting to analyse whether observed reductions in fear of the current study would persist in different situations, and over time to guarantee the treatment effect was not a product of extinction. Fear memory treated with extinction training, have the tendency to re-emerge over time and in a different situation due to the mechanism of spontaneous recovery (Kindt et al., 2009).
Another point of criticism which possibly explains the lack of a consistent manipulation during the pilot study is a motivational aspect. It could be argued that participants that are willing to push their boundaries to overcome their anxiety might have a better chance of good results during treatment. It could be the case that participants that were afraid but did not want to get treated, would show more avoidance behaviour in comparison to very motivated individuals. The vast majority of the participants were first-year psychology students, who scored above subclinical on an acrophobia questionnaire. They did not seek actual help for their anxiety but were asked to take part in the study. During research, it was noticed that the individuals that showed intrinsic motivation, overcoming their fear of heights, also reacted better on treatment. Unfortunately, there is no hard evidence for this hypothesis. This point emphasises the importance of a motivated sample group in which participants are willing to tackle their anxiety.
During the pilot study, we faced some difficulties in translating experimental research into clinical reconsolidation research. The standardised reactivation session did not always cause the results we were hoping to see. There are several factors that can influence the manipulation, and which need to be taken into account when using a standardised reactivation session.
Reactivation sessions should trigger a prediction error. The earlier discussed article of Sevenster et al., (2013) states that a prediction error seems to be necessary for the memory to become destabilised. In an experimental setting, it is much easier to generate a prediction error in comparison to clinical cases. Most of the time, the fear memory in experimental settings are laboratory obtained, carefully built using fear conditioning. This way it is relatively easy to adjust some of the information to provoke a prediction error. With clinical fear, obtained outside of the laboratory, it is much harder to control for prediction error. Every fear of patients is unique, so their expectations are also unique, which makes it harder to generate a controlled prediction error. Kindt and van Emmerik (2016) showed that it is possible to produce a prediction error within clinical patients with PTSD, by using imagery rescripting. Briefly, it means that the patient recalled the traumatic event and by rescripting the event at that moment, expectations of negative emotions got violated. This way they were able to generate a prediction error. This emphasises that it may not be desirable to have a standardised reactivation procedure within a clinical application of reconsolidation-blocking therapy. A possible solution for this problem lies in a short interview, like in a SCID questionnaire, where patients tell about their fears. This way, a well-fitted script could be designed, in which it is exactly clear what the patient is afraid of. Subsequently, a clinician could come up with some variations on the script, to contradict the patients’ feared expectations.
Balancing between reconsolidation, limbo and extinction. Sevenster et al., (2013) showed that a prediction error is an important factor in causing memory reconsolidation. However, when participants get exposed for a longer period, an extinction effect could occur in which participants’ anxiety will drop because of habituation effects. In between there is a third state which is called limbo; Merlo et al. (2014) described this as a state in which neither reconsolidation nor extinction will occur. Further; Sevenster, Beckers and Kindt (2014) showed that only when there is one prediction error, reconsolidation is triggered. When there were several prediction errors or no prediction error, the memory was not affected or modulated. These studies point out that generating reconsolidation through prediction error is a delicate matter. Just like the first point of difficulty, this issue also highlights a need for a less standardised reactivation procedure in the treatment of specific phobias in clinical patients.
Ecological validity. Anxiety memory is complex, even with relatively specific fears like spider phobia, or acrophobia, it is still very difficult to translate findings from an experimental setting to the complicated structure of convictions which are part of a patient's anxiety network. It is not clear if there is sufficient correspondence between anxiety disorders and experimental research to specific phobias. An intervention like a pharmacological blockade, may not be effective in treating an anxiety disorder in a clinical setting (just as it has proven to be effective in an experimental one). Furthermore, the way we try to reactivate fear in those experimental settings, may not track the actual memory structure that lies at the foundation of the emotions that patients experience when faced with anxiety. Evidence from recent studies, however, advocate for the use reconsolidation-blocking therapy treating clinical fear. Kindt and Soeter (2015), managed to treat a specific phobia that was not laboratory obtained, and related to those findings, Kindt and Soeter (2016) achieved to damp symptoms of PTSD, using reconsolidation-blocking therapy.
The primary aim of the current study was to see whether strong treatment effects that were observed in Kindt en Soeter's (2015) study could be generalised to other specific phobias. This way, it could be tested whether reconsolidation therapy could be used as a replacement for cognitive behavioural therapy in the future.
Even though the lack of a control group was a shortcoming of the design and a missed chance to gain valuable information, it is noteworthy that with a small sample size of 16, the effect that is observed is significant. Participants experienced less anxiety both physically and psychologically, which is a remarkable result for a relatively short and simple intervention with a sub-clinical sample. The results do not contradict earlier conducted research outcomes. The main difficulty was in obtaining a uniform manipulation for complex and unique memories of individuals. Causing a single prediction error within a limited amount of time seems to be the most important condition that needs to be met. Kindt and Soeter (2016) succeeded in designing an
intervention that theoretically would trigger a prediction error and showed that this intervention seemed to work for individuals with PTSD using imagery rescripting. Due to a standardised procedure, it is very much likely that we did not manage to cause a consistent violation of fear related expectations for all of our participants. As mentioned before, a less standardised approach might be a solution to this problem. Potentially, a protocol could be developed in which the core fear that is causing problems is being examined by a therapist, which they could shape to make it usable for a reconsolidation-based treatment. In an ideal situation,
maladaptive memories could be neutralised using reconsolidation-based treatment, tackling the root problem of the disorder. This new approach for therapy could be used to target both the acute anxiety responses as well as maladaptive cognitions and expectations relating to the fear. In future therapy, those two paradigms could complement each other to reach better results in therapy in a wide range of psychiatric disorders in which dysfunctional memories play a central role.
6. References
Baker, B. L., Cohen, D. C., & Saunders, J. T. (1973). Self-directed desensitisation for acrophobia. Behaviour
Research and Therapy, 11(1), 79–89.
Bouton, M. E. (2002). Context, ambiguity, and unlearning: sources of relapse after behavioural extinction.
Biological Psychiatry, 52(10), 976-986.
Brown, J. S., Kalish, H. I., & Farber, I. E. (1951). Conditioned fear as revealed by magnitude of startle response to an auditory stimulus. Journal of Experimental Psychology, 41(5), 317–328.
Brunet, A., Orr, S. P., Tremblay, J., Robertson, K., Nader, K., & Pitman, R. K. (2008). Effect of post-retrieval propranolol on psychophysiologic responding during subsequent script-driven traumatic imagery in post-traumatic stress disorder. Journal of Psychiatric Research, 42(6), 503–506.
Coelho, C. M., Waters, A. M., Hine, T. J., & Wallis, G. (2009). The use of virtual reality in acrophobia research and treatment. Journal of Anxiety Disorders, 23(5), 563-574.
Debiec, J., & Ledoux, J. E. (2004). Disruption of reconsolidation but not consolidation of auditory fear by noradrenergic blockade in the amygdala. Neuroscience, 129(2), 267–272.
Dudai, Y. (2006). Reconsolidation: the advantage of being refocused. Current Opinion in Neurobiology, 16(2), 174– 178.
Eisenberg, M., Kobilo, T., Berman, D. E., & Dudai, Y. (2003). Stability of retrieved memory: inverse correlation with trace dominance. Science (New York, N.Y.), 301(5636), 1102–1104.
Elsey, J., & Kindt, M. (2016). Manipulating human memory through reconsolidation: Ethical implications of a new therapeutic approach. AJOB Neuroscience, 7(4), 225-236.
Emmelkamp, P. M., Bruynzeel, M., Drost, L., & van der Mast, C. A. G. (2001). Virtual reality treatment in acrophobia: a comparison with exposure in vivo. CyberPsychology & Behaviour, 4(3), 335-339.
Fanselow, M. S. (1984). What is conditioned fear? Trends in Neurosciences, 7(12), 460-462.
Goodman, L. S. (1996). Goodman and Gilman's the pharmacological basis of therapeutics (Vol. 1549). New York: McGraw-Hill.
Kandel, E. R., Dudai, Y., & Mayford, M. R. (2014). The molecular and systems biology of memory. Cell, 157(1), 163–186.
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., & 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.
Kindt, M., Soeter, M., & Vervliet, B. (2009). Beyond extinction: erasing human fear responses and preventing the return of fear. Nature Neuroscience, 12(3), 256–8.
Loerinc, A. G., Meuret, A. E., Twohig, M. P., Rosenfield, D., Bluett, E. J., & Craske, M. G. (2015). Response rates for CBT for anxiety disorders: Need for standardised criteria. Clinical Psychology Review, 42, 72–82.
McGaugh, J. L. (2000). Neuroscience - Memory - a century of consolidation. Science, 287(5451), 248–251. Merlo, E., Milton, A. L., Goozée, Z. Y., Theobald, D. E., & Everitt, B. J. (2014). Reconsolidation and extinction
are dissociable and mutually exclusive processes: Behavioural and molecular evidence. Journal of
Neuroscience, 34(7), 2422-2431.
Miloff, A., Lindner, P., Hamilton, W., Reuterskiöld, L., Andersson, G., & Carlbring, P. (2016). Is virtual reality treatment for spider phobia noninferior to traditional one-session treatment?: The results of a
randomized-controlled trial. In 46th European Association of Behavioural and Cognitive Therapies Congress, August
31-September 3, Stockholm, Sweden. (pp. 753-753). The European Association for Behavioural and Cognitive
Therapies.
Nader, K. (2007). A single standard for memory; the case for reconsolidation. Debates in Neuroscience, 1(1), 2-16. Nader, K., & Hardt, O. (2009). A single standard for memory: the case for reconsolidation. Nature Reviews
Neuroscience, 10(3), 224-234.
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.
Otis, J. M., Werner, C. T., & Mueller, D. (2015). Noradrenergic regulation of fear and drug-associated memory reconsolidation. Neuropsychopharmacology, 40(4), 793–803.
Peterson, R. A., & Heilbronner, R. L. (1987). The anxiety sensitivity index: Construct validity and factor analytic structure. Journal of Anxiety Disorders, 1(2), 117-121.
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. Classical Conditioning II: Current Research and Theory, 2, 64-99. Schafe, G. E., & LeDoux, J. E. (2000). Memory consolidation of auditory pavlovian fear conditioning requires
protein synthesis and protein kinase in the amygdala. Journal of Neuroscience, 20(18), RC96.
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. (2013). Prediction error governs pharmacologically induced amnesia for learned fear. Science, 339(6121), 830-833.
Sevenster, D., Beckers, T., & Kindt, M. (2014). Prediction error demarcates the transition from retrieval, to reconsolidation, to new learning. Learning & Memory, 21(11), 580-584.
Soeter, M., & Kindt, M. (2010). Dissociating response systems: erasing fear from memory. Neurobiology of Learning
and Memory, 94(1), 30–41.
Soeter, M., & Kindt, M. (2015). An abrupt transformation of phobic behavior after a post-retrieval amnesic agent. Biological Psychiatry, 78(12), 880–886.
Troncoso, J., & Maldonado, H. (2002). Two related forms of memory in the crab chasmagnathus are
differentially affected by NMDA receptor antagonists. Pharmacology Biochemistry and Behaviour, 72(1–2), 251– 265.
