Research project Deleting the Fear: Memory Training to Strengthen Internet-Based CBT for Social Anxiety

Social anxiety is an intense fear of being scrutinised and judged by others. About seven percent of the population has social anxiety at any given time; it often begins in adolescence and tends to persist. The most effective treatment is CBT with exposure, in which people systematically approach the situations they fear and drop their safety behaviours. The treatment works as well when delivered as therapist-supported internet-based CBT. Even so, 30 to 50 percent do not fully recover, and many of those who improve later relapse. One explanation is that exposure leaves the fear memory intact and adds a competing memory that the feared outcome did not happen. The new memory is tied to the context where it was learned, so under stress, in new settings or as time passes, the fear memory can take over again.

The project tests a way to shift that competition. Think of a witness who is asked again and again what the suspect was wearing. Other details from the same event, such as the getaway car, then become harder to remember. It is caused by active suppression in memory, not by the details fading on their own. The effect is called retrieval-induced forgetting (RIF). It has been shown in hundreds of studies. In social anxiety, it could mean that someone who repeatedly recalls that “the audience nodded when I made my point” finds it harder to automatically bring up the memory that “my hands were shaking”. Whether this holds in clinical treatment is what the project sets out to test.

The tool is the RIF audio diary. Despite the name, it is not an ordinary diary but targeted memory training. Right after an exposure exercise, the participant describes aloud on their smartphone exactly what went well in the situation and deliberately leaves the unpleasant parts aside. New recordings follow at increasingly longer intervals. Participants tell the story afresh each time rather than listening to earlier recordings, since the act of recalling is what is expected to suppress the fear memory. The therapist reviews the recordings every week and makes sure they stay with what went well. Building a coherent narrative demands more of memory than the brief cue words used in laboratory research, and what is said aloud is remembered better than what is only read silently. The project therefore tests the hypothesis that the narrative gives a stronger effect. It is the first time RIF has been brought into an internet treatment that can reach many people. The RIF audio diary is built into the treatment platform Iterapi and is designed to be added to existing internet treatments at low cost.

The project comprises four studies. In the first, 128 people with elevated social anxiety give a speech in a standardised stressful situation in the laboratory, and the RIF audio diary is compared with an established laboratory method. The second is a randomised controlled trial in which about 330 participants, aiming for 300 completers, receive nine weeks of internet-based CBT: with a reflection diary, with RIF practice, or with RIF practice where about two thirds of the exposure exercises take place in virtual reality. The results of the first study determine whether RIF practice uses the RIF audio diary or a combined format, and whether a fourth group is needed. The VR exercises are done at home with simple headsets into which participants insert their own phones, to standardise and control fear activation. In the third study, 60 of the trial participants wear a wrist sensor that measures heart rate variability and the skin's electrical conductance while they tag social situations on their phones. This gives a measure of bodily reactions in everyday life that does not rely on self-report. In the fourth study, 20 of them are interviewed about their experience of the treatment, the audio diary and the sensors. Trial participants are followed up one year after treatment.

Despite the project's name, no memories are erased. RIF makes some memories less accessible, but participants can still recall negative experiences when they try. What is expected to change is the automatic, repeated recall of what went wrong that keeps anxiety going, much as in ordinary exposure therapy.

If the hypotheses are supported, more people with social anxiety could recover fully and fewer might relapse, through an add-on that is cheap and easy to use. The principle could also apply to other anxiety disorders in which fear memories compete with safety learning. A null result is also informative, because it can show whether treatment needs to suppress fear memories in order to work. The studies will be preregistered, and the analysis code will be published openly. The project is carried out in collaboration with Linköping University and the University of California, Los Angeles (UCLA).