Cognitive bias modification version 2.0 - can immersive virtual reality improve reliability and effectiveness?
Social anxiety disorder is a common problem. Fortunately, novel treatments may increase accessibility, acceptability and effectiveness. An emerging field is Cognitive Bias Modification (CBM). By training a person with social anxiety using the dot-probe paradigm, it has been shown that attentional bias towards threat-relevant stimuli can be changed. Two important 2009 articles demonstrated that increasing attentional flexibility could have impressive results in social anxiety. However, the replication of these effects have sometimes failed, raising the question of its true efficacy. When analyzing the discrepancies, it has become clear that there are two fundamental differences between effective vs. non-effective trials. Whenever there has been a change in bias, a significant improvement in terms of symptoms has always followed. On the contrary, studies not demonstrating a change in bias have all failed in reducing anxiety. In a series of experiments, we would like to test a more effective bias changing paradigm (called "person identity match"). In addition, we want to make use of virtual reality to create a more standardized and immersive training environment. The aim is to create a more reliable bias changing paradigm with research questions identifying the differential effects of 2D vs. 3D, still photos vs. short dynamic video sequences, and the relative effects of lab vs. home-based delivery.
Final report
PROJECT AIM AND DEVELOPMENT
Social anxiety disorder (SAD) is one of the most prevalent mental health problems that negatively impact on social functioning and quality of life. Barriers to traditional face-to-face therapies such as inaccessibility, high cost, long wait time, and stigmatisation can prevent patients from seeking treatment. There is an ongoing effort to develop treatment options that are more accessible, effective, and acceptable. Attentional bias modification (ABM) is a relatively new intervention for SAD that has garnered considerable interest in recent years. ABM operates on the assumption that dysfunctional anxiety is caused by the preferential allocation of attention towards socially threatening information. Attentional training that reduces this underlying bias is thought to also reduce social anxiety.
Attentional training paradigms such as the dot-probe task requires no therapist intervention. This makes ABM well-suited for a self-administered treatment option. Virtual reality (VR) technology removes the restraints of a physical clinic, and provides an immersive and engaging environment within which the therapy can take place. VR offers some unique advantages over traditional therapies, such as extensive control, high flexibility, ease of standardization, and better privacy. These features can address some of the barriers to traditional therapy and make VR-based treatment more appealing to some patients.
The current project aimed to investigate the effectiveness of VR-based ABM in reducing attentional bias and social anxiety. In Study 1, we attempted to replicate previous findings that a dot-probe task can reduce attentional bias and anxiety. We also compared 3D versus 2D stimuli to see if there were any differential effects in training outcomes. In Study 2, we investigated the effectiveness of a novel bias changing paradigm called the person-identity-matching (PIM) task. A further study was planned at the start of the project that aimed to compare stimuli comprised of dynamic video sequences against static images.
SUMMARY OF IMPLEMENTATION
The study design and specification for the first two studies (dot-probe, Study 1 and PIM, Study 2) began in September 2016. VR program development was completed in June 2017. The development of dynamic video sequence stimuli for the third planned study was eventually abandoned due to high cost and technical limitations.
Data collection for the two studies was carried out concurrently between July 2017 and February 2018. In total, 200 participants with elevated social anxiety were recruited from the general population to take part in the two studies. The first article on the dot-probe study was submitted in March 2019 and published in November 2019. The second article on the PIM study was submitted in April 2020 and accepted in June 2020.
MOST IMPORTANT RESULTS AND CONTRIBUTIONS
• The two studies both failed to detect any attentional bias in our participants at baseline. Furthermore, we did not observe any changes in bias after ABM training. We suspect these findings were primarily due to our current method of measuring attentional bias (dot-probe task) being extremely unreliable. We concluded that the most pressing concern of ABM research is to develop tasks that can reliably measure attentional bias, as any attempt at developing an efficacious ABM treatment would be futile without a good estimate of the variable it is designed to change. Perhaps more fundamentally, the concept of attentional bias being a causal factor in anxiety disorders may need to be reevaluated.
• Despite the lack of change in bias, all participants showed reduction in their anxiety scores at post-training and follow-ups. However, in terms of clinically significant change, only a small proportion of the participants (9% in Study 1 and 4% in Study 2) achieved clinically significant change at the 3-month follow-up. Furthermore, the anxiety reduction was not specific to ABM training – in both studies, participants in the placebo groups achieved comparable reduction in anxiety symptoms as those in the active training groups. This suggests that the anxiety reduction was unlikely to be a result of an attentional shift brought about by the ABM. Nevertheless, some components in the training procedures might have some therapeutic effect (e.g. exposure effect).
• Our study was the first to utilize the 3D stimuli in an VR environment. Although the results did not reveal any differential effects between 2D and 3D stimuli in our studies, the use of more immersive and ecological valid stimuli in VR therapies need to be investigated further.
NEW RESEARCH QUESTIONS GENERATED
• The most important research question that arose in light of our findings is how to reliably measure attentional bias, as this is a necessary to move the field forward.
• The mechanism underpinning the general anxiety reduction post-training across all participants remained elusive. Since the current study did not have a wait-list control group, it was difficult to ascertain whether this finding was purely a placebo effect. Identifying therapeutic features not previously hypothesized can contribute to developing better therapies.
• While our study did not systematically measure participant attitudes towards ABM or VR-based treatment, anecdotal experiences from the experimenters reported a wide range of responses ranging from completely skeptical/unconvinced to extremely positive. How these opinions might moderate VR-based treatment outcomes warrants exploration.
INTERNATIONAL DIMENSIONS OF THE PROJECT
The stimuli used in the current project were adapted from the BP4D-spontaneous database, provided by Professor Lijun Yin and his research group at State University of New York at Binghamton. The data generated in the current project were made open access and readily available for any collaborative work.
COLLABORATION AND PRESENTATION OF RESULTS
The VR program was developed in collaboration with the VR company Mimerse. The first study was conducted in collaboration with Linköping University. The second study was conducted in collaboration with Linköping University and Uppsala University.
The results of the project were presented at the following conferences
• Swedish Congress on Internet Interventions (SweSRII 2017, Linköping University, Sweden)
• European Society for Cognitive Psychology (ESCOP 2019. Tenerife, Spain)
Social anxiety disorder (SAD) is one of the most prevalent mental health problems that negatively impact on social functioning and quality of life. Barriers to traditional face-to-face therapies such as inaccessibility, high cost, long wait time, and stigmatisation can prevent patients from seeking treatment. There is an ongoing effort to develop treatment options that are more accessible, effective, and acceptable. Attentional bias modification (ABM) is a relatively new intervention for SAD that has garnered considerable interest in recent years. ABM operates on the assumption that dysfunctional anxiety is caused by the preferential allocation of attention towards socially threatening information. Attentional training that reduces this underlying bias is thought to also reduce social anxiety.
Attentional training paradigms such as the dot-probe task requires no therapist intervention. This makes ABM well-suited for a self-administered treatment option. Virtual reality (VR) technology removes the restraints of a physical clinic, and provides an immersive and engaging environment within which the therapy can take place. VR offers some unique advantages over traditional therapies, such as extensive control, high flexibility, ease of standardization, and better privacy. These features can address some of the barriers to traditional therapy and make VR-based treatment more appealing to some patients.
The current project aimed to investigate the effectiveness of VR-based ABM in reducing attentional bias and social anxiety. In Study 1, we attempted to replicate previous findings that a dot-probe task can reduce attentional bias and anxiety. We also compared 3D versus 2D stimuli to see if there were any differential effects in training outcomes. In Study 2, we investigated the effectiveness of a novel bias changing paradigm called the person-identity-matching (PIM) task. A further study was planned at the start of the project that aimed to compare stimuli comprised of dynamic video sequences against static images.
SUMMARY OF IMPLEMENTATION
The study design and specification for the first two studies (dot-probe, Study 1 and PIM, Study 2) began in September 2016. VR program development was completed in June 2017. The development of dynamic video sequence stimuli for the third planned study was eventually abandoned due to high cost and technical limitations.
Data collection for the two studies was carried out concurrently between July 2017 and February 2018. In total, 200 participants with elevated social anxiety were recruited from the general population to take part in the two studies. The first article on the dot-probe study was submitted in March 2019 and published in November 2019. The second article on the PIM study was submitted in April 2020 and accepted in June 2020.
MOST IMPORTANT RESULTS AND CONTRIBUTIONS
• The two studies both failed to detect any attentional bias in our participants at baseline. Furthermore, we did not observe any changes in bias after ABM training. We suspect these findings were primarily due to our current method of measuring attentional bias (dot-probe task) being extremely unreliable. We concluded that the most pressing concern of ABM research is to develop tasks that can reliably measure attentional bias, as any attempt at developing an efficacious ABM treatment would be futile without a good estimate of the variable it is designed to change. Perhaps more fundamentally, the concept of attentional bias being a causal factor in anxiety disorders may need to be reevaluated.
• Despite the lack of change in bias, all participants showed reduction in their anxiety scores at post-training and follow-ups. However, in terms of clinically significant change, only a small proportion of the participants (9% in Study 1 and 4% in Study 2) achieved clinically significant change at the 3-month follow-up. Furthermore, the anxiety reduction was not specific to ABM training – in both studies, participants in the placebo groups achieved comparable reduction in anxiety symptoms as those in the active training groups. This suggests that the anxiety reduction was unlikely to be a result of an attentional shift brought about by the ABM. Nevertheless, some components in the training procedures might have some therapeutic effect (e.g. exposure effect).
• Our study was the first to utilize the 3D stimuli in an VR environment. Although the results did not reveal any differential effects between 2D and 3D stimuli in our studies, the use of more immersive and ecological valid stimuli in VR therapies need to be investigated further.
NEW RESEARCH QUESTIONS GENERATED
• The most important research question that arose in light of our findings is how to reliably measure attentional bias, as this is a necessary to move the field forward.
• The mechanism underpinning the general anxiety reduction post-training across all participants remained elusive. Since the current study did not have a wait-list control group, it was difficult to ascertain whether this finding was purely a placebo effect. Identifying therapeutic features not previously hypothesized can contribute to developing better therapies.
• While our study did not systematically measure participant attitudes towards ABM or VR-based treatment, anecdotal experiences from the experimenters reported a wide range of responses ranging from completely skeptical/unconvinced to extremely positive. How these opinions might moderate VR-based treatment outcomes warrants exploration.
INTERNATIONAL DIMENSIONS OF THE PROJECT
The stimuli used in the current project were adapted from the BP4D-spontaneous database, provided by Professor Lijun Yin and his research group at State University of New York at Binghamton. The data generated in the current project were made open access and readily available for any collaborative work.
COLLABORATION AND PRESENTATION OF RESULTS
The VR program was developed in collaboration with the VR company Mimerse. The first study was conducted in collaboration with Linköping University. The second study was conducted in collaboration with Linköping University and Uppsala University.
The results of the project were presented at the following conferences
• Swedish Congress on Internet Interventions (SweSRII 2017, Linköping University, Sweden)
• European Society for Cognitive Psychology (ESCOP 2019. Tenerife, Spain)