What Kind of Exposure Do You Want, Goldilocks?

What Kind of Exposure Do You Want, Goldilocks?

Understanding factors that contribute to outcome are crucial as we continue to refine treatments and revise the theories that underlie them. In a study published in the Journal of Anxiety Disorders, Norton and colleagues (2011) examined the role of activation and habituation in exposure therapy. The rationale for study was based on the emotional processing theory, which I’ve written about previously.

By activation, the authors mean how distressed the person becomes during the exposure exercise. Habituation refers to the reduction in distress when someone is confronted with a fear inducing stimuli. I’ve written in greater detail about habituation and some problems in using it as a marker of change. In this study, the focus was on within-session habituation—the degree to which distress reduced during a particular exposure session. In this study, the authors appear favorably disposed towards the emotional processing theory while still acknowledging research that does not support some of its proposed mechanism of action.

The Set-up

The sample consisted of 106 people who were enrolled in studies of a transdiagnostic protocol for a cognitive behavioral treatment for anxiety developed by the first author. The 12-session protocol included cognitive restructuring as well as exposure. According to the article, the protocol can be adapted to a variety of anxiety-related problems. Initial sessions begin with psychoeducation, self-monitoring, and cognitive restructuring, before shifting into 6 sessions of exposure therapy. The remaining sessions shift back to cognitive interventions before ending with relapse prevention. Because of missed session and treatment incompletion, the researchers focused their analyses on the first three exposure sessions.

How Do Activation and Habituation Relate to Dropout and Treatment Outcome?

As the authors examined habitation and activation across three exposure sessions, I’ll step back and focus on the broad findings of the study.

1. Clients were more likely to drop out of treatment if their anxiety was too high or too low during the first exposure session. Dropouts were also higher among those who didn’t show a decrease in anxiety (i.e., habituation) during the first session. (These trends weren’t found in the second and third exposure sessions.)

What does this mean? It’s hard to say for certain, but the findings suggest people with really high anxiety during the first exposure session who don’t experience a reduction in anxiety during the exposure may be at a greater risk for dropping out of treatment. In my mind, it might be worth spending some time preparing clients for this possibility, perhaps even normalizing it.

2. Contrary to what the emotional processing theory predicts, people with lower anxiety overall during exposure tended to have better outcomes at the end of treatment.

This is tricky to interpret, too. The authors suggest that too much anxiety may inhibit treatment, which is not inconsistent with the emotional processing theory.

Should Activation Be “Just Right”?

With so many variables (e.g., three exposure sessions, high/low activation, habituation, dropout, symptom reduction), we should be cautious in interpreting these findings. Should anxiety during exposure like Goldilocks and the Three Bear—not too high, not too low, but just right? Or alternatively, people with high anxiety the first time may be frightened off by treatment, and people with low anxiety may not have been engaged to begin with. Unfortunately, these variables were not manipulated experimentally, so it’s difficult to know what causes what.

Painting in broad strokes here, it does appear that how a client experiences the first exposure session is important. It may be worth processing a client’s experience afterward, responding to any questions or concerns, and renewing commitment to treatment. This study also illustrates how complicated it is to try to understand potential variables when conducting exposure therapy.