Transcript of Maximizing Exposure Therapy for Anxiety Disorders 

Length             : 00:47:02

Speakers         Dr. Jackie Persons, Dr. Michelle Craske

Interviewer:  Welcome.  I'm Jacqueline Persons.  I'm Director of the Cognitive Behavior Therapy and Science Center, which is a small group private practice in Oakland, California.  I'm also a Clinical Professor in the Psychology Department at UC Berkeley.  I'm here today with Michelle Craske who is a Professor of Psychology, Psychiatry and Biobehavioral Sciences, Director of the Anxiety and Depression Research Center, and Associate Director of the Staglin Family Music Center for Behavioral and Brain Health, all at the University of California at Los Angeles.  Thank you so much Dr. Craske for giving us your time here today for this discussion.

Dr. Craske:  You are very welcome.

Interviewer:  I’m meeting with Dr. Craske as part of the series titled Translating Science to Practice that is hosted by and developed by the Society for Science of Clinical Psychology.  The goal of our series is to help clinicians access and use findings from basic science to guide their clinical work.  We’re here today to take up the question, “What do the latest findings in extinction learning and memory mean for clinicians who are doing the exposure-based treatments for patients with anxiety disorders?”  Our discussion focuses on an article published in Behaviour Research and Therapy in 2014.  It’s titled Maximizing Exposure Therapy: An Inhibitory Learning Approach.  Let me acknowledge Dr. Craske’s co-authors, Michael Treanor, Christopher Conway, Tomislav Zbozinek, and Bram Vervliet.  Published in Behaviour Research and Therapy in 2014, did I say that?  This is a fabulous article.  It’s really important for clinicians who are doing exposure-based therapy to know what it says.  At the same time, it’s a very rich article.  I have read it about three times and I learned new things from it every day, so I highly recommend it.  We’ve posted a link to give you access to it on the SSCP webpage.  And we’ll host a listserve discussion following this interview.  And I'm looking forward to having a chance to ask Dr. Craske a bunch of questions about it that were developed by our committee on Science and Practice.  So, thank you again, Dr. Craske.  Could I ask you to start by giving us a summary or a description of the inhibitory learning model account of exposure treatment?

Dr. Craske:  Certainly.  So the inhibitory learning model account, as you indicated, derives from principles of associative learning through fear conditioning and extinction, which is where exposure therapy was first born.  But it takes into account the more recent developments in learning theory, which basically posit that as an individual proceeds through extinction or, in this case, the clinical proxy of extinction, which is exposure, new learning is taking place.  And what is most critical is that that new learning, which we call inhibitory learning, is competing with the original learning, which we call excitatory learning.  So, for example, if an individual had acquired a fear of dogs as a result of being bitten by a dog, the original excitatory learning is that this person perceives dogs as being threatening because they could bite.  As the individual goes through exposure therapy, the new inhibitory learning is that not all dogs bite or this particular dog won’t bite or there is some other meaning that is associated with dogs that’s not as threatening.  Now, here’s the important point.  That new inhibitory learning is somewhat fragile.  And therefore, the goal of exposure therapy is to develop that new inhibitory learning into the most powerful potent kind of learning that we can, so that it will be -- it will resist the tendency to fall apart over time.  And so, in our model of inhibitory learning, we’re targeting ways to enhance the strength of the learning that happens during exposure therapy and ways of trying to retrieve it in the future so that it remains potent and doesn’t fall into what is called the secondary stages and be less competitive with the original excitatory learning.

Interviewer:  So thank you.  That’s very helpful.  So, this might be not such an easy question to answer.  But would -- how is this model different from the models that many of us were trained with-- habituation models or emotional processing models?  What do you see as the key differences in this newer model?

Dr. Craske:  Right.  So, when I think of the inhibitory learning model, there are three basic principles that sort of drive the model.  One is, as I said, exposure is a learning process.  And so, what is critical for this to work is to identify for each person, as they are coming in for treatment, what is it that they need to learn.  So it’s very much focused on a functional analysis of what that person needs to learn.  And how to then design the exposure in a way that will maximize the strength of that learning, as I just described it.  And then third, that -- while the reduction of fear throughout exposure therapy typically does occur, in and of itself, fear reduction may not be the index of whether this new learning is actually happening and how strong that new learning will be in the future.  So that -- those are the principles.  Now, let's see how does this differ from the two basic models of habituation and emotional processing theory.  So clearly, habituation models of exposure therapy focus on fear reduction or reducing the strength of the fear response as the person proceeds through exposure therapy.  And that’s why -- when I was initially trained too, the goal was have the individual stay in the exposure situation until fear declines.

Interviewer:  Yup.

Dr. Craske:  So that was the kind of clinical law, right?

Interviewer:  Yup.

Dr. Craske:  And with the idea being that if it doesn’t decline, then maybe you're doing something terribly wrong.

Interviewer:  Or maybe they're not learning anything.

Dr. Craske:  Or maybe they're not learning anything.  In the inhibitory learning model, fear reduction, in and of itself is not the primary driving force of how to do exposure therapy.  But it may happen and it usually does happen, but it doesn’t have to be the case.  So, in many experimental studies that we’ve conducted, fear does not habituate throughout the given exposure trial, and yet the person is doing well in the long term.  So I think we’ve tended to shift away from the fear reduction, in and of itself, with the basic notion being if -- for a given person -- what they may need to learn is something independent that of -- where the fear reduces.  They may need to learn, in fact, that they can survive fear being elevated for prolonged periods of time, and in that case, fear reduction would not be a good goal because they're not learning what they need to learn.  Emotional processing theory, which, you know, is a magnificent theory, in many ways, and I think it’s been useful in moving our field forward in a providing framework, has -- In the way I see it, emotional processing theory has two components.  It has, first of all, the habituation component, which is that you want fear to habituate within the given trial of exposure. That’s called within-session habituation. Then you want fear to habituate across trials of exposure, that’s between-session habituation, and then that allows corrective learning to take place, which is a shift in the person’s meaning that they give to the stimulus, whether they now see the stimulus as less dangerous or less overwhelming.  That second component, the corrective learning component, has more to do, or I see that has much more overlap with this inhibitory learning model approach.  The initial component of fear habituation has much less to do with the inhibitory learning approach.

Interviewer:  I see.  Thank you.  So now, I -- you’ve already said a lot of the key points here.  So maybe you’ve already answered this question, but it would be fun to hear from you for the clinical audience, what do you see as the, say, three main take home messages that you want clinicians to learn from reading this article and learning the inhibitory learning model?

Dr. Craske:  Okay.  Yes.  I think the three main messages are, as you start to work with an individual and you start to design an exposure practice, instead of thinking, “How do I get rid of the fear?  How do I lessen the person’s fear?” shift your attention over to, “What does this person really need to learn in order to eventually be less afraid or more able to function in their life?”  And so that’s the first question.  What do they really need to learn that’s currently driving them to be so fearful and so avoidant that in the -- that will eventually result in improvement in their distress and functioning, that’s first.

Interviewer:  So that’s -- now, it would be an individualized case conceptualization of what is the key thing this person is not understanding and needs to understand.

Dr. Craske:  Yes

Interviewer:  Okay.

Dr. Craske:  Secondly, how can I design the exposure experience to maximally benefit that learning?  So again, instead of focusing on, “Let's do the exposures in such a way that fear declines,” it’s more about what experience will most -- violate is the word we use, it’s an interesting word, but -- most violate that person’s current expectancies.  So if they're perceiving that they cannot tolerate a certain experience because  -- whether it’d be the length of time they're in a situation or what kind of activity they're engaging in within the situation -- what kind of sensations they're experiencing in the moment of that situation.  Whatever it is that they perceive as being their boundaries that they can get so far that they can't go beyond that boundary, because if they go beyond that boundary, something terrible is going to happen, then we need -- we think we need to design the exposure to go beyond their perceived boundaries in order to violate their expectancy of what they think is going to happen.  So that’s the second main point.  And then the third main point -- so just a summary.  The first one is think about what an individual needs to learn.  Second is design the exposure to maximally violate their expectancy in order to create the most potent learning possible.  And then third, remember that learning is probably taking place, not only during the time that you’re with that client as they're going through exposure, but in the hours and days afterwards as well.  And that period of consolidation, as we call it, is just as critical as in the moment.  Hence, how much their fear has reduced in a moment is not really a good index because it’s not tapping into the underlying mechanisms of learning, which probably are going on for the hours and weeks in between of the exposure sessions.  So we need to think, you know, learning is an implicit process.  It’s happening all of the time, and how do we structure experience around that.

Interviewer:  So that last idea suggests that a better index of what the patient learned from the session might be obtained at the beginning of that following session rather than at the end of the session that you are just conducting?  Is it? 

Dr. Craske:  Exactly.  Exactly, that when we’re thinking about how well the person has actually learned and consolidated that learning, we’re thinking about the follow-up, whether that’d be at the next session or, you know, a month after the end of treatment.  It’s always the longevity of the learning that’s most critical to the individual’s, you know, future symptomatic stages.  So, yeah.  The end of the session is not -- the end of an exposure session is just an index of what they performed in a moment.  The learning is tested at later point in time.

Interviewer:  Yes.  So these are very helpful ideas.  Thank you.  So, another question that flows out of what you're telling us is that, as clinicians, we need to think a little more carefully.  And perhaps, you can help us think about how we’re going to orient the patient to the treatment, because usually the patient wants to reduce or eliminate the anxiety or distress.  And our earlier approach to exposure focused very directly on that goal, including within the session.  Now, we’re not so focused on that goal.  And in the article, you used the word toleration.  We’re more focus on learning and we’re less focused on experiencing less anxiety.  So then, we need to think a little more carefully about how we’re going to orient the patient and sell the patient on the treatment.  I'm hoping you can give us your thoughts about that issue.

Dr. Craske:  That’s right.  That’s right.  And I think it’s true for a clinicians as well as patients.

Interviewer:  That’s a very good point.

Dr. Craske:  Everybody wants the fear to go away.

Interviewer:  We do.  We do.  The therapist also wants it to go away.

Dr. Craske:  Yeah, because it feels better.  We don’t like to see people in distress.

Interviewer:  And we don’t want to be in distress ourselves either.

Dr. Craske:  Exactly, exactly.  So it’s very much about that functional analysis that, is the case, conceptualization idea that you brought up before, Jackie. It’s really helping the therapist to gear themselves toward, you know, what is it that the client is needing to learn in order to really have a long-term learning that persists.  And then for the client, it’s helping them to understand, “Okay.  So when you. . . ”  Let's take, for example, an individual who has social anxiety, and what they're most concerned about is being rejected because they perceive if that they are rejected, then they won’t be able to ever establish any relationships in the future because they are unlikeable and unworthy.  If the focus is upon fear reduction, that might mean that they're asked to go into a social situation and stay there until they feel comfortable.  So they’ll go into the social event and have, you know, try to interact with people until they feel comfortable, their fear goes down, and then what have they learned?  Okay, I can be in a situation and my fear declines.  But have they learned anything about the concept of rejection?  And so -- not necessarily.  So what, you know, what may be more effective for that individual is to realize that I can actually experience rejection and go on and interact with another person and not be rejected, or I can experience rejection and follow-up with more social interactions that provide positive reinforcement -- that, in other words, a rejection does not lead to never being in a satisfying relationship.  And of course, I'm speaking, you know, this would take a lot of careful analyses on how to set this up.  But what we would help the client understand is, boy, it sounds like you're fearful that by being rejected, it’s going to mean that you'll never have any relationships.  So wouldn't it be useful?  Wouldn't it be powerful for you to learn that a rejection does not actually relate to never being in a relationship?  How can we design an experience for you to learn that?  How could we set up something so that you could see that a one-time rejection doesn’t mean what you're thinking it means?  So that’s the way we try to present it to the client.

Interviewer:  So what that boils down to is spending time laying out the conceptualization to the clients and getting the clients on board.  So they totally understand what it is they need to accomplish in therapy to solve their problems so that, that increases their willingness to do the difficult tasks of therapy.

Dr. Craske:  That’s right.  That’s right.  And of course, that, you know, is very similar, in some ways, to a cognitive therapy or a behavioral testing model of doing exposure therapy.  And there are some overlaps between the inhibitory learning model and that cognitive approach, especially in that initial rationale because we are saying to the client, “What is it that you're expecting to happen?”  Helping the client recognize that the experience that would disconfirm that expectancy would be particularly potent for them.  If they have an experience that is okay but it doesn’t really tap into that expectancy, it’s not going to be as potent.  So we just have them bind into the idea that we want them to have the best learning possible that’s most suited their needs and then actually going ahead, of course, in doing the exposure in such a way that it helps to violate that expectancy.  Now, if they said that my fear is going to be very high, then we’ll come back to say, “How is that going to influence this expectancy?”  So, sometimes the expectancy is tied directly to the fear and the person might say, “If I'm afraid and if I experience fear for a long enough period of time, then I'm going to collapse.”  Okay, then we will need to design an exposure that helps to violate that expectancy.  Let's design an exposure that sustains your fear for a prolonged period of time to test out whether you will collapse or not.  Again, helping the client recognize that if we can violate that belief at that moment, then there will be potent learning taking place.

Interviewer:  Which will ultimately reduce the anxiety they're experiencing on a daily basis.

Dr. Craske:  Exactly.  Yeah.  So the long-term outcome is fear reduction.  We’re not asking people just, you know, tolerate fear.  You know  -- you're always going to be afraid and tolerate it.  No, we’re saying there's a value to tolerating fear in the presence of exposure trials for the long-term reduction of fear.

Interviewer:  Right, and for the learning needed to get the long-term reduction.  Yeah.

Dr. Craske:  Exactly.

Interviewer:  So now, a moment ago, you alluded to. . .  And I would love to talk more about the relationship between doing a behavioral experiment, you know, the cognitive model.  And there's a fabulous book about behavioral experiments.  That edited book by James Bennett-Levy, which lays out in detail, how to test beliefs by doing behavioral experiments.  And there seems quite a bit of overlap between that approach and the inhibitory learning model.  And I would love to hear from you about similarities and differences of those two approaches.

Dr. Craske:  Definitely, yeah.  There is a lot of overlap between the two, but there are also large differences both conceptually and in terms of therapeutic strategy.  So if we start with the conceptual side of things, clearly, the behavioral testing model, which derives from this cognitive theory and cognitive therapy models, is all about disconfirming erroneous beliefs through logic and reasoning and then extending that through behavioral testing.  Whereas, conceptually, the inhibitory learning model is starting from the premise that there is an excitatory or threat-laden and memory that’s being formed and the goal of exposure therapy is to develop an inhibitory memory that will then compete with the excitatory memory.  The excitatory memory, the original fear memory, is very powerful and it generalizes widely.  So we know from a lot of the basic science around fear acquisition and learning and memory that that first fear memory is potent and widely generalizeable, whereas the secondary memory that’s being formed as a result of this new learning experience in exposure therapy is fragile.  It does not generalize.  And that is why we need to do all of these additional things to try to enhance its strength over time.  So, what I'm saying is that at a conceptual level, these are different approaches.  The cognitive model comes from, you know, beliefs that are erroneous and need to be corrected.  The inhibitory learning model is coming from a process model of how learning and memory is formed and how is it retrieve at different points in time.

Interviewer:  Though, could I say something here?  So, is it accurate or would you view it as accurate to say that in the inhibitory learning model, we’re not so much focused on correcting inaccurate beliefs as we are on creating new beliefs that compete with the older beliefs that are causing the anxiety?

Dr. Craske:  That is very -- thank you very much.  That’s a very good way of saying it.  When I think of the inhibitory learning model, I always have in my mind, you know, an original memory on one side of my visual field, and at the other side, a new memory that’s being opened.  And this new memory is always going to be bouncing over and trying to defeat the original memory, but the original memory is powerful.  And so we have to boost up the new memories, so they’re in competition all the time.  And that’s different than the notion of changing …

Interviewer:  Yes.

Dr. Craske:  Or erasing a former action.  So that’s the guiding principle of the inhibitory learning model, how do we make that competition between the original memory and the new learning memory, such that the newer memory wins and shines.

Interviewer:  So there’s an article by Brewin on this view of cognitive therapy as well, where he’s talking about teaching new beliefs that win the retrieval competition.

Dr. Craske:  Yes, yes.  So it’s very much tied in with cognitive science.  And of course he’s very much a cognitive scientist.  So that’s the, just personally that’s the kind of -- what excites me is where we can pull upon the science of how, you know in learning and memory, the basic fundamental processes, and how do we incorporate that into how we think about during the exposure therapy.

Interviewer:  Exactly, exactly.

Dr. Craske:  And then the other difference, so that’s more the conceptual side of things, in terms of the actual therapeutic strategy, from the inhibitory learning model approach, one principle that’s worth revisiting is that the more that the experience of exposure disconfirms an expectancy that the person has (and that expectancy could be explicit or implicit --the more that the experience disconfirms that expectancy, the greater the learning that takes place.  Now that’s also a fundamental principle of learning, that we do not learn when what we experienced is completely in line with what we thought it was going to be.  Learning is facilitated by, “Oh that’s new, that’s different, that wasn’t quite what I thought.

Interviewer:  Not what I expected.

Dr. Craske:  That’s not what I expected, right.  So, the more we can help clients be willing to engage in an experience that yields an outcome that’s not really what they thought was going to happen, the better.  That means that if we do cognitive restructuring as a therapeutic strategy prior to going into exposure therapy, we are potentially undercutting the value of experience, because we want the experience itself to stand out as a surprise.  So let’s say I have somebody who has panic disorder, and they’re thinking that they’re going to go crazy if they experience, you know bodily sensations of unreality.  If I spend a lot of time before they do exposure therapy, helping them to realize, you know, the chances of you actually going crazy are very slim, you know let’s look at all the data and help you to form different ways of thinking so that you can go into the exposure therapy, realizing you’re probably not going to go crazy, the chances are very low, then theoretically that means the experience of exposure won’t be so surprising for them, and it won’t give them that potent learning.  And so that’s what we now do is, we shift the cognitive restructuring phase, really to after the exposure rather than before. And that’s different, that’s a key therapeutic difference between a behavioral testing model and an inhibitory learning model.

Interviewer:  So now, well I heard you say “theoretically” when you were describing.  So I was wondering if in actual practicality sometimes a certain amount of cognitive restructuring is needed in order to help the person be willing to do the exposure. Because if I literally think that I am going to have a heart attack, then I’m probably not going to be willing to do the exposure.

Dr. Craske:  That’s right, that’s right.  And of course it comes down to what is the person willing to do.  So, methodologically, you know, first of all we establish: what is it that you’re most afraid of happening?  Then we design the optimal exposure to test that out and if the person says I’m willing to go into the optimal exposure which might be, I’m going to travel by a plane to another country on my own without any Zanax or any kind of medication, then that is the optimal exposure for that person. 

Interviewer:  I understand, that would be the optimal learning experience for that person.

Dr. Craske:  If that’s what they needed to do in order to really violate their expectancy that they’re going to die if they’re on their own in an unfamiliar place. Then we would say: What are you willing to test out? And so we might have to back down from the optimal exposure to something that they’re willing to do, and constantly reinforcing the idea that their expectancy of that bad event happening is clearly preventing them from learning something new and how can we design experiences to help them learn something new? And of course by doing this, we as therapists are conveying a sense that this is safe, because we would never be asking our clients to do anything that’s dangerous.  So clients are clearly going to be absorbing that notion that this is okay to do. So I think there’s probably unavoidably some level of thought reappraisal happening. It’s just that we don’t spend time doing time formal cognitive restructuring.

Interviewer:  And you do the minimum amount of that needed to get the person to have the experiential learning. So the goal is experiential learning.

Dr. Craske:  Exactly, exactly.

Interviewer:  Now the other difference from cognitive therapy is that of course this notion of violating expectancies which overlaps with behavioral testing is only one part of the inhibitory learning model.  We have a number of other strategies that all derived from this concept: how do you strengthen that inhibitory learning that’s going on during exposure, because it is always going to be more fragile than the original learning.  And so we have a number of other strategies such as deepened distinction that comes from the basic science of extinction, it’s how to use multiple cues at the same time so if somebody’s – we’ll take a simple example -- if somebody is fearful of spiders, how do we have not just exposure to a single spider but exposure to a single spider followed by the introduction of the second spider. There’s some value to that because it’s going to help the person lift up their expectancy again, “Oh oh no another spider maybe I’m going to be bitten again.”  And that expectancy now has been lifted up to such a degree that the learning is going to be maximized because, as I said before, when there’s no expectancy of a threatening event, theoretically no learning is going to be taking place. So we always want to raise the expectancy and then violate it by the experience.  And so having multiple cues at one time is one way of doing that.  We do a lot of work on how to attend to the stimulus so keeping a person’s attention focused on where they are and what they’re doing and what they’re facing because the attention is critical to learning.  We would introduce something called variability throughout exposure which means that we might vary up the length of time of that person’s conducting an exposure or the order in which they conduct the exposures or the type of stimuli.  So a very common example for us is when we’re working with socially anxious individuals is to have them approach, go out into our campus around us and approach 30 different people, of all different types and ask them all different questions rather than approaching the same kind of person and asking the same question over and over again, because variability enhances the storage of new learning.  It enhances the capacity of that new learning to be encoded and then retrieved later. And so all these are the strategies, and that’s only some of them that fit in to this sort of approach to exposure, which is a learning based approach, and that make it a different model than a cognitive therapy behavioral testing model. But there definitely are overlaps as well.

Interviewer: I see.  So could you say a little more about attention, because as all of us clinicians know, often patients use distraction to tolerate uncomfortable situations and then as a clinician we need to be clear in understanding, is that a good idea or not? And I think, understanding more about how you’re thinking about the importance of attention and what’s being learned would help guide clinicians here.

Dr. Craske: Right. Definitely from a basic learning model, what we are aiming to do is to help the client understand that this stimulus, and the stimulus might be an internal sensation or external event or activity or situation.  To learn that this stimulus is not a predictor of the bad thing that they were anticipating or expecting to happen, then it’s critical that attention be directed to the stimulus that they have their expectancy based on.  That’s really very, very important.  Now I know because over the years I myself have done some of the studies on distraction.  What happens when you distract?  What happens when you lose attention?  I think sometimes distraction can be part of an exposure therapy practice if it’s going to be helpful to allowing the person to learn what they need to learn.  So let’s say it’s -- I’m going to be in a situation and distract from the fact, not distract, I’m going to shift my attention onto another element in the situation in order to learn that having this physical sensation – no, I just made that confusing.  Distraction can be useful to the degree that it facilitates whatever it is that that person needs to learn, but in that scenario you always want to bring them back to what is the stimulus that they distract me from and I haven’t made that very clear, I am terribly sorry.  But attention to the stimulus facilitates learning and if we distract totally from that stimulus then there wouldn’t be any learning taking place.

Interviewer: Right, if the person has the fear -- if my heart rate goes up really fast, I’ll have a panic attack and I’ll die of a heart attack then during the exposure we would want to focus their attention on how was your heart rate, go -- doing? Is it going up?  Is it like that?

Dr. Craske: Right, right.  Now if they said: I’m afraid that if I shift my attention over to this conversation then there’s definitely a chance that I’m going to have a heart attack because I’m not effectively monitoring my heart rate.

Then in that case you may want to test out, we’ll let see what happens when you distract.

Interviewer: Got it.

Dr. Craske: So that’s the part that I was getting confused on before, on how to use distraction as a tool. But the eventual result would be I’m testing out whether distraction from a physical sensation leads to a negative outcome.

Interviewer: So the core thing that the therapist always wants to be anchoring to guide the intervention is what is this person afraid of? And what does this person need to learn in order to get a new idea that inhibits the old idea?

Dr. Craske: Yes exactly.  Okay.

Interviewer:  Okay.  So then I would to talk for a moment about safety behaviors because of course we’ve all been taught that we should all work to be dropping out the safety behaviors. At the same time if I listen to the focus on the conceptualization, I could imagine that a patient might have the idea.  If I go out by myself on an airplane trip I’ll have a panic attack, I’ll have a heart attack and I’ll die and that’s true even if I have a bottle of Zanax in my pocket. In that case, could we agree that doing the exposure with the bottle of Zanax in the pocket would still provide a learning experience for this patient and so therefore might not be contraindicated using that way of thinking about how we design the exposure?

Dr. Craske:  Right. So in that case the primary expectancy that the person’s dealing with is a fear of having a heart attack and dying.

Interviewer:  Yes.

Dr. Craske:  The safe -- the quote safety signal of the Zanax is not having any predictive relationship with that outcome, right?  And because it’s not in any form -- it’s not having an associative relationship, it doesn’t actually matter.

Interviewer:  Yes I see.  Well often patients have what seem to be sort of magically associating,  and they’ll tell you logically it doesn’t make any sense, but I feel more comfortable if I have it.

Dr. Craske:  If that were the case, what makes you feel more comfortable about having that element with you? What is it doing to your expectancy about having a heart attack. If it does come down to, well I think it actually prevents me from panicking and if I don’t panic then I’m not going to have a heart attack then it is serving as a safety signal and that will interfere with effective learning.  And so whenever the object that they perceive as being a safety signal changes the expectancy of the outcome, it will interfere with corrective learning. It will interfere with the development of inhibitory learning.  Does that make sense?

Interviewer:  Yes it does, yes it does.

Dr. Craske:  So that’s what I’m always coming back to the question: what is this object’s function?  How is it functioning for you and what role does the object play for you when you carry it with you?

Interviewer:  Right and what’s the relationship between having that object and your expectancy of about what’s going to happen if you do or do not have it.

Dr. Craske:  Right, exactly and not so much about whether I’ll be safe or not, but what is the expectancy of being fearful?

Interviewer:  The expectancy of danger, okay that’s helpful.  Well it’s helpful to have this kind of conceptual clarity for the therapist. It guides our decision making about what interventions to do and all the details of how to structure it.

Dr. Craske:  Right.

Interviewer:  So now I’m thinking about certain patients who have OCD and generalized anxiety disorder, and what they have is intolerance of uncertainty. And then I’m trying to figure out how to take those concerns and use them in this model -- - and I’m sure you’ve thought about that.

Dr. Craske:  Right, right.  So in a couple of ways, in one way it’s certainly -- with a patient with obsessive compulsive disorder for example then they could have an expectancy that’s so far off in the future that it cannot possibly be tested by exposure therapy, such as if I touch this dirty door handle, then I’ll die from cancer 20 years from now.

Interviewer:  Right.

Dr. Craske:  So we can’t test that out in that an exposure format, but we can test out the notion that are you willing to tolerate that discomfort and function at other tasks despite not knowing whether you will die or not 20 years from now.  So it’s a toleration, it’s an expectancy associated with the uncertainty as you’ve just said.  So, in those cases we design exposures around, let’s test out whether you can tolerate the anxiety or tolerate the uncertainty with the notion that if you perceive the anxiety or uncertainty to be inherently aversive, then it’s a type of expectancy.  So you’re saying the client is predicting that I expect that I cannot tolerate that level of uncertainty or that level of distress, then we would test that out.  Let’s see if we can actually have you engaged in this task induce the uncertainty and the distress and see what happens.  This is very similar to what I think a lot of people already do in fact with those kinds of exposures.

Interviewer:  Yes, yes.  Thank you.  So, now one of the parts of the article that I had trouble with was the idea of that occasional reinforced trials can be therapeutic.

Dr. Craske:  Yes this is a complicated one and still needing a lot more research, but let’s see if I can explain this.  So in the experimental sense what this means is that by providing occasional aversive outcomes as the person’s going through extinction or exposure, what’s happening is that the person’s expectancy of having a negative outcome of course is being raised up, because they actually are sometimes having negative outcomes.

Interviewer:  Yes.

Dr. Craske:  Which means that on every subsequent trial, after the negative outcome trial, now the chance for new learning is enhanced because their expectancy has been raised …

Interviewer:  Yes.

Dr. Craske:  … and it’s violated.  Now it’s tricky but I think the best sort of clinical translation is social anxiety.  And the reason for that is that the person’s expectancy is of social rejection, which is a negative outcome. 

Interviewer:  It is aversive for all of us.

Dr. Craske:  It does happen, it’s part of our daily life.  So it’s not unreasonable to deliberately include social rejection in exposure practice.  And the idea being, that if occasional social rejections happen then the person’s expectancy of, “Oh no I’m going to get rejected again.” has been raised.

Interviewer:  Yes.

Dr. Craske:  And then they’re going to the next social situation and they don’t get rejected and they’ll have the experience of, “Oh I just now realized that, yes I expected to get rejected and I wasn’t on this trial.”  So they’re learning: “It’s not always the case that I get rejected.” And that learning is going to be more potent than if they were never rejected, that’s the idea.

Interviewer:  I understand.  So it builds, so it’s really taking a long view and working to build resilience.

Dr. Craske: That’s a good way of thinking about it. In fact, in our experimental work on this topic, that’s exactly how we looked at it, because it seemed that after going through that kind of training with their occasional negative outcomes, the person really seem to be naturally blasé about it.

Interview: “Well, it could happen and now I noticed it did happen and I did not die.”

Dr. Craske: I do not die and it also means that it’s not true for every situation.

Interviewer: Yup. Sometimes it happens, sometimes it doesn’t.

Dr. Craske:  Right.  So, as I say, I think that’s most appropriate for social anxiety and maybe for panic attacks too but of course not appropriate for actual traumas. Certainly we would not want to re-traumatize people.

Interviewer: Yes, very helpful.  Could I ask one final question, which is about affect labeling which of course I’m sure we could talk for a long time about, but -- could I ask you just to say a few minutes about how does affect labeling fit into exposure treatment?

Dr. Craske: Right.  Affect labeling is coming from a slightly different direction of research, but the basic notion is that by using linguistic or verbal processing of emotional states or emotion stimuli, we develop inhibitory pathways at the neuro level, at the brain level, based on the basic experimental work.  In exposure therapy, what it means is we encourage people to simply state their emotional experience or describe the stimulus in terms of how they perceiving it in an emotional way, for example, I feel anxious.  This situation scares me.

Interviewer: The spider.  The spider looks dangerous. 

Dr. Craske: The spider looks dangerous. So it’s clearly not cognitive restructuring, there is no reappraisal going on, it’s just labeling the emotion.  Now when I use this clinically I’m very much using it in a toleration of the situation, a toleration of the experience.  So I’m feeling anxious and I’m here.

Interviewer: Okay.

Dr. Craske: “I’m scared, but I’m going into it anyway” kind of idea. And it also fits quite well with the notion of attention.  So it takes me to what I’m doing and not distracting.  And the person who can really embrace that concept is probably doing it a little bit of the acceptance notion, too.  I’m accepting my emotional states rather than trying to change it.  So it fits with a lot of different conceptual models and clinically I personally think it works very well.

Interviewer: So here there’s a -- maybe we’ll just make a final comment, there’s a fit with Acceptance and Commitment Therapy I assume is part of what you’re referring to?

Dr. Craske: Yes indeed.  Yes.

Interviewer:  I have to tell you how much I learned from this discussion, how much I appreciate your time, Doctor Craske. And we all thank you so much both for your work on these important issues and for giving us your time here today too.

Dr. Craske:  Thank you.