by Michael D. Anestis, M.S.
I am not particularly well-versed in acceptance and commitment therapy (ACT). I've read outcome studies, seen Steve Hayes give a demonstration of cognitive defusion at the Association for Behavior and Cognitive Therapies (ABCT) conference, written about it a bit on PBB, and discussed the treatment with ACT researchers/practitioners, but some aspects of the theory underlying this approach still elude me. That being said, I am very experienced in the use of mindfulness in therapy and in the basic concept of acceptance, so I don't feel completely lost on this one. When we teach clients about acceptance (e.g., radical acceptance in dialectical behavior therapy), we are asking them to essentially adopt the mantra of the serenity prayer by choosing not to fight the things we can not control (or even some things we can), opting instead to accept them as a part of reality and devoting energy to more positive goals.
Today, I would like to discuss some interesting results published in a paper in press at Behavior Therapy by PBB guest author Brandon Gaudiano, James Herbert, and Steve Hayes. In this paper, they looked at mediators of outcome in ACT for psychosis. You might remember from prior PBB articles that statistical mediators are variables that "explain" the relationship between two other variables. For instance, there is a positive association between height and baldness, meaning that taller people are more likely to be bald; however, this relationship is mediated by sex. In other words, the only reason for that statistical relationship is that men tend to be taller than women. When we account for sex, that relationship disappears. So, in this study, the authors were looking for variables that explain why, in a previous study (Gaudiano & Herbert, 2006), individuals with psychosis randomly assigned to receive ACT in addition to treatment as usual (TAU) reported less distress post-treatment than did individuals randomly assigned to TAU.
As somebody who received a lot of graduate training in traditional cognitive behavioral therapy (CBT), I am accustomed to an approach that emphasizes the direct challenging of thoughts. If somebody has the thought "I am an idiot," I want them to ask what the evidence is for and against that thought and to test that hypothesis and develop more accurate (although not necessarily positive) alternative thoughts that reflect facts (e.g., "I received a 65% on this exam, which lowered my overall grade to a C and means I need to get a much better grade on the final than I was expecting to need"). This approach is called cognitive restructuring and it aims to decrease the degree to which we experience overly negative thoughts capable of amplifying and sustaining negative emotions. ACT, which is a newer "third generation" cognitive behavioral therapy, takes a different approach. According to Gaudiano and colleagues (in press), the ACT skill of cognitive defusion asks the client to change his or her relationship to thoughts rather than trying to change the thoughts themselves. The example they give in the paper (pg 2 of document - page number in journal not yet determined) is experiencing the meta-thought "I am having the thought 'I am going crazy'" instead of seeing the thought as factual and believing "I am going crazy." In other words, thoughts are seen as simply thoughts and no response to thoughts is truly required. Acceptance-based approaches note that we can simply observe our internal experience (e.g., thoughts, emotions) as passing phenomena that do not require us to react. ACT theorists believe that this shifting relationship with our thoughts might explain positive outcomes in treatment studies; however, this assumption is really an empirical question - one that can be tested statistically - so the authors set out to do just that.
Using the data from the Gaudiano and Herbert (2006) randomized clinical trial in which individuals randomly assigned to ACT plus TAU reported lower distress post-treatment than did individuals simply in TAU, the authors looked at participants' reports on the degree to which their hallucinations were believable post-treatment. 14 participants in the ACT condition and 15 participants in the TAU condition had complete data on all the measures and were included in these analyses.
What the authors found was very interesting. As expected, the degree to which participants found their hallucinations to be believable mediated (explained) the relationship between which treatment they were in (ACT or TAU) and their treatment outcome (reported level of distress). Individuals who received ACT tended to see their hallucinations as less believable after treatment and individuals who saw their hallucinations as less believable after treatment tended to report lower levels of post-treatment distress. Importantly, the authors also tested whether hallucination frequency mediated the relationship between treatment type and treatment outcome. After all, it is reasonable to wonder whether distress levels were actually explained by the fact that treatment caused individuals to experience fewer hallucinations. Analyses, however, indicated that this was not the case.
According to these data, ACT does not reduce distress in individuals with psychosis because it makes them hallucinate less often, but rather because it helps them change their relationship with their hallucinations such that they see them as simply thoughts and less believable thoughts at that. The authors acknowledged one significant weakness in their analyses - the mediators were assessed at the same time as the outcomes, which means we can not be certain that believability changed before distress diminished and therefore played a causal role - however this weakness is present in the vast majority of mediational analyses and is far from unique to this work.
ACT is one of those treatments that tests my philosophy that the data matter more than my ability to relate to the treatment (this is also the case with interpersonal psychotherapy for bulimia). It doesn't matter that my instincts say the right move is to challenge the distorted thought and attempt to change it. That certainly can be beneficial, as we've discussed in prior articles; however, it might not be the only approach (or the superior approach) to take. What's great about that is that it means multiple, very distinct approaches exist for successful treatment and further studies examining whether clients' preference for one approach versus the other or other similar variables impact outcome would help us learn whether particular treatments are better candidates for one approach as opposed to the other. Keep in mind that both traditional CBT and ACT are seen as adjunct treatments for psychosis, meaning that they need to be conducted alongside the use of medication rather than utilized as stand alone treatments. Also keep in mind that the research base for ACT is impressive but still new, meaning that caution is still recommended in assessing its value relative to other approaches.
If you would like to learn more about this or other topics discussed on PBB, please refer to our online store for scientifically-based psychological resources.