Virtual Reality Training for Social Skills in Schizophrenia - Comparison With Cognitive Training

Social impairments are core features of schizophrenia that lead to poor outcome. Social skills and competence improve quality of life and protect against stress-related exacerbation of symptoms, while supporting resilience, interpersonal interactions, and social affiliation. To improve outcome, it is necessary to remediate social deficits. Existing psychosocial interventions are moderately effective but the effort-intensive nature (high burden), low adherence, and weak transfer of skills to everyday life present significant hurdles toward recovery. Thus, there is a dire need to develop effective, engaging and low-burden social interventions for people with schizophrenia that will result in better compliance rates and functional outcome. In a previous pilot study, the investigators tested the effectiveness of a novel adaptive virtual reality (VR) intervention in improving targeted social cognitive function (social attention, as indexed by eye scanning patterns) in individuals with schizophrenia. 10 sessions of 1-hour VR intervention were sufficient to engage the target mechanism of social attention and improve negative symptoms. Acceptability and compliance were very high among the participants. ' The next phase, supported by a R33 grant will compare the VR social skills training with a control condition. This new protocol includes a control condition for the exposure to computerized training across the 10 sessions and incidental exposure to social interactions (i.e. interactions with experimenters twice a week for 5 weeks) by including a control condition, which involves computerized brain fitness training for 10 sessions.

The effectiveness of the social VR training (n=20) at an optimal dose will be compared with an active control condition (computerized cognitive training) (n=20) in improving social attention (Social Engagement Latency: SEL) and social brain network in a pilot randomized control trial (RCT). Participants with schizophrenia will undergo a baseline assessment of cognition, social functioning, symptoms and social brain network functioning. These behavioral, clinical and functional neuroimaging (fMRI of social imitation task, and resting state connectivity analyses of the social brain network) methods will be used to assess changes. Individuals with schizophrenia will be randomized to either the social skills training or the computerized cognitive training at baseline. They will then participate in 10 sessions of social skills VR training or computerized cognitive training (1 hour per session, twice a week for 5 weeks). In addition, long-term social outcome will be assessed approximately 2 months after the end of training with the Social Functional Scale (Birchwood, 1991), the the Scale for the Assessment of Negative Symptoms (SANS) and the Scale for the Assessment of Positive Symptoms (SAPS). For the long-term outcome assessment, no neuroimaging or behavioral assessments will be conducted because the primary goal of the follow-up is to assess broad levels of social functioning in the real world and because there is not sufficient time within the R33 mechanism. 16 matched control participants will be recruited to obtain behavioral and neuroimaging comparison data for optimal performance levels but these healthy control participants will not undergo social skills training. Lastly, at the end of the R33 project, if this adaptive VR technology shows that it can improve social attention, which in turn improves social outcome in schizophrenia, it will be possible to refine the protocol to make the method more accessible, less burdensome and widely available in the future by moving towards a mobile application.

schizophrenia, social skills training, social brain network, virtual reality, simulation, functional outcome

Two conditions are compared in individuals with schizophrenia: VR social training and active control condition of computerized cognitive training. Healthy controls are run on the baseline tasks to provide the comparison data for the patient groups but the controls do not undergo VR training.

Participants will undergo a virtual reality social skills training program for 10 sessions. Each session takes about an hour. Participants visit the lab twice a week. Therefore, the training duration is 5 weeks.

If there is a significant improvement in social skills for the active treatment condition, the reason might be that the participants were exposed to social environment by coming to the lab and interacting with the research staff twice a week for 5 weeks and/or they used a computerized training tool twice a week for 5 weeks. In order to control for these potential confounds, we included a cognitive training arm. Participants will undergo a commercially available cognitive training program for ten 1-hour sessions (twice a week for 5 weeks).

Social skills game that we developed in the R21 phase will be used across 10 sessions of training in the lab. Each session is about 1 hour long. Participants come to the lab twice a week for 5 weeks.

Commercially available cognitive training program will be used to control for the time spent in the lab and associated social interactions as well as the total exposure to computerized games.

Social engagement latency is defined as the time taken to select an avatar to interact with. SE: is measured in milliseconds. This is the targeted mechanism identified in a previous R21 study.

fMRI of social simulation task will identify the brain networks that are engaged during a social simulation task (see Thakkar, Peterman and Park. American Journal of Psychiatry. 2014)

BLERT (Bell-Lysaker Emotion Recognition Task; Bell et al, 1997). BLERT measures social cognition in people with schizophrenia and it has been reported to have the strongest psychometric properties for use in clinical trials. BLERT presents 21, 10-second video clips to measure the ability to identify seven emotional states: happiness, sadness, fear, disgust, surprise, anger, or no emotion.

SANS is a 25-item clinical symptom rating scale. It consists subscales of alogia, affective blunting, avolition-apathy, anhedonia-asociality, and attentional impairment. (Andreasen, 1983). Each subscale has a single item global score for overall severity of each symptom domain. Each item is rated for severity from least (0) to most severe (5). Andreasen NC: Scale for the Assessment of Negative Symptoms (SANS) . Iowa City, University of Iowa, 1983

Baseline and 5 weeks. In addition, it will also be used to assess long term outcome 2 months after the training has ended.

SAPS has 34-items that evaluate hallucinations, delusions, formal thought disorder and bizarre behavior (Andreasen, 1984). Each subscale has a single global score item to rate the overall severity of each symptom domain. Each item is rated for severity from least (0) to most severe (rating=5) Andreasen NC: Scale for the Assessment of Positive Symptoms (SAPS) . Iowa City, University of Iowa, 1984 (NC Andreasen - The British Journal of Psychiatry, 1989)

Baseline and 5 weeks. In addition, it will also be used to assess long term outcome 2 months after the training has ended.

Social Functioning Scale (SFS; Birchwood, 1990). This is a well validated measure used to assess social functioning. SFS consists of 76 items (7 subscales). Every subscale value is normalized to a Scaled Score (Mean = 100, SD = 15). The SFS full scale score is the mean of the 7 subscales scaled scores. The 7 subscales are: social engagement/withdrawal, interpersonal behavior, pro-social activities, recreation, independence-competence, independence-performance and employment.

Baseline and 5 weeks. In addition, it will also be used to assess long term outcome 2 months after the training has ended.

Inclusion and Exclusion Criteria for Individuals with Schizophrenia: Inclusion criteria: Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5) Axis 1 Diagnosis of schizophrenia Wechsler Abbreviated Scale of Intelligence (WASI) intelligence quotient (IQ) > 90 Currently taking antipsychotic medication No change in current psychotropic medications or housing within the past 30 days. Those patients whose medication or housing situation has changed within a month, we will wait list them until their situation stabilizes. Exclusion criteria: DSM 5 Axis 1 diagnosis other than schizophrenia Diagnosed organic brain disease, brain lesions, history of head traumas, neurological disorders or other conditions that involve the degeneration of the central nervous system (e.g. multiple sclerosis) Substance/alcohol dependence during the past 1 year Tardive dyskinesia Inclusion and Exclusion Criteria for Healthy Control Participants: Inclusion Criteria: WAIS IQ > 90. No DSM-5 Axis 1 diagnosis of psychotic disorders in themselves or their families (e.g. schizophrenia, bipolar disorder). No antipsychotic medications No diagnosed organic brain disease, brain lesions, history of head traumas, neurological disorders or other conditions that involve the degeneration of the central nervous system (e.g. multiple sclerosis) No substance/alcohol dependence during the past 1 year

We are comparing a social skills game in virtual reality with a cognitive training software. After the study is completed, we will make the protocol and the methods available to all those who request this information.

After the study has ended and we have been able to publish the outcome of the study. This is expected to be within 2 years of the completion of the study.

Adery LH, Ichinose M, Torregrossa LJ, Wade J, Nichols H, Bekele E, Bian D, Gizdic A, Granholm E, Sarkar N, Park S. The acceptability and feasibility of a novel virtual reality based social skills training game for schizophrenia: Preliminary findings. Psychiatry Res. 2018 Dec;270:496-502. doi: 10.1016/j.psychres.2018.10.014. Epub 2018 Oct 9.

Torregrossa LJ, Bian D, Wade J, Adery LH, Ichinose M, Nichols H, Bekele E, Sarkar N, Park S. Decoupling of spontaneous facial mimicry from emotion recognition in schizophrenia. Psychiatry Res. 2019 May;275:169-176. doi: 10.1016/j.psychres.2019.03.035. Epub 2019 Mar 20.

Wade J, Nichols HS, Ichinose M, Bian D, Bekele E, Snodgress M, Amat AZ, Granholm E, Park S, Sarkar N. Extraction of Emotional Information via Visual Scanning Patterns: A Feasibility Study of Participants with Schizophrenia and Neurotypical Individuals. ACM Trans Access Comput. 2018 Nov;11(4):23. doi: 10.1145/3282434.